Effects of landuse change and forest fragmentation on the biodiversity and ecosystem functioning in the tropical lowlands of sri lanka

111 CHAPTER 7 HABITAT CHANGE IMPAIRS ECOLOGICAL SERVICES PROVIDED BY SCARABAEINAE DUNG BEETLES IN THE TROPICAL LOWLANDS OF SRI LANKA .... Throughout this thesis, selected modified land

EFFECTS OF LANDUSE CHANGE AND FOREST FRAGMENTATION ON THE BIODIVERSITY AND ECOSYSTEM FUNCTIONING IN THE TROPICAL

LOWLANDS OF SRI LANKA

ENOKA PRIYADARSHANI KUDAVIDANAGE B.Sc (Zoology), M.Sc (Environmental Sciences)

A THESIS SUBMITTED FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY DEPARTMENT OF BIOLOGICAL SCIENCES NATIONAL UNIVERSITY OF SINGAPORE

2011

“The forest is a peculiar organism of unlimited kindness and benevolence that makes no demands for its sustenance and extends generously the products of its life activity; it provides

protection to all beings, offering shade even to the man who destroys it.”

— Gautama Buddha

Dedicated to my parents

ACKNOWLEDGEMENTS

I am grateful to late Prof Navjot Sodhi, my ex-supervisor whose unique combination of

intelligence, determination, intimidation and social skills defined my path into the field of

conservation biology I am indebted, to my supervisor Prof Edward Webb for persuasion and guidance that enabled me to complete the PhD thesis, and to the constant encouragement

provided by Prof Richard Corlett I extend my gratitude to Prof Sarath Kotagama, my supervisor

in Sri Lanka, and mentor for the last 15 years, for his guidance provided throughout my

academic life I thank all examiners for their most valuable and helpful comments that were adopted in the final version of my thesis

Many thanks to all the past and present staff member and colleagues of the Conservation Ecology, Applied Plant Ecology laboratories and the terrestrial ecology group including Daniel Friess; David Bickford; Mary Rose Posa, Nanthinee Jeevanadam, Reuben Clement, Arvin Diesmos and Lee Wei Kit; the technical and administrative staff including Tommy Tan, Reena Devi Samayanadan, Priscilla Lee, members of the Biodiversity group and the Raffles Museum for Biodiversity research of the Department of Biological Sciences for all the encouragement assistance and inspiration extended in one-way or another Special thanks towards Janice Lee and Qie Lan, my fellow dung girls for the comradeship of dung beetles Prof David J Lohman (Lohman Lab, City College of New York) is kindly acknowledged for extending a generous helping hand in many challenges faced during my PhD study including correcting manuscripts, assisting analytical problems, helpful discussions and providing resources

I’m grateful to Darren J Mann, Eleanor Slade and the staff of the Oxford University Museum of Natural History for patiently guiding me through the intense process of species identification and for the opportunity to use the scarab beetle collection at the American Museum

of Natural History Thanks to the ScarabNet research team including Finbarr Horgan for sharing with me their knowledge about dung beetles I acknowledge all my collaborators; Thomas

Wanger (Center for Conservation Biology , Stanford University), Neil Collier (Lohman Lab, City College of New York) for analytical work; Priyadarshan Dharmarajan (Ashoka Trust, India), Manori Gunetilleke (National Museums, Sri Lanka) and Deepchandi Lekamge

(Sabaragamuwa University) for the taxonomic study

I acknowledge the Department of Wildlife Conservation and the Forest Department, Sri Lanka for granting me access to work in the protected area and permission to export specimen for taxonomic work I thank the Vice- Chancellor of the Sabaragamuwa University of Sri Lanka and the Department of Natural Resources for kindly facilitating my graduate studies Gratitude extended to Professors Nimal Gunatilleke, Savithri Gunatilleke and Drs Eben and Uromi

Goodale for expertise knowledge, guidance and helpful discussions Ravi Amarasinghe,

Chamitha de Alwis, Vimukthi Herath, Krishan and Udeni Ariyasiri, Vishan, Amila Perera, Nuwan Hegodaaarachchi, the Wildlife Conservation Society of Galle and the local assistants who supported the field work are acknowledged

The Field research was funded by NAGAO Natural Environmental Foundation, Japan and Joe Grove memorial award, UK, while the taxonomic study was funded by the Oxford

University Museum of Natural History, London The World Bank sponsored IRQUE project and

the National University of Singapore Research Scholarship supported the PhD studies

Finally, I thank everybody including my family who stood by me extending their constant support, patience and understanding through the graduate studies I owe all that I am to my parents, who made it their lifetimes’ goal to help me fulfill my dreams I remember with heartfelt gratitude and warmth, my father who passed away one month before I entered the graduate

school of his choice and my mother who stood by me all my life before she passed away seven months ago She shared my love for nature and all living things, supported my chosen career and enjoyed every one of my field expeditions, participating with great enthusiasm This thesis is a product of her faith in me

ACKNOWLEDGEMENTS iii

THESIS SUMMARY 1

LIST OF FIGURES 4

LIST OF TABLES 5

APPENDICES 6

CHAPTER 1: GENERAL INTRODUCTION 7

CHAPTER 2: STUDY SITE 13

2.1 Sri Lanka and the Lowland Wet Zone: a biodiversity hotspot 13

2.2 From game protection to biodiversity conservation: the history and status of biodiversity conservation in the wet zone of Sri Lanka 15

2.3 Study sites: Sinharaja, Kanneliya and Kottawa 20

2.3.1 Sinharaja Man and Biosphere Reserve 20

2.3.2 Kanneliya conservation forest 23

2.3.3 Kottawa Kombala conservation forest 24

CHAPTER 3: AMPHIBIAN AND BUTTERFLY DIVERSITY ACROSS A LOWLAND TROPICAL LAND-USE GRADIENT IN SRI LANKA 28

3.1 Introduction 28

3.2 Materials and methods 29

3.2.1 Study area 29

3.2.2 Sampling diversity and abundance 30

3.2.3 Environmental variables 31

3.2.4 Analysis of species richness and abundance data 32

3.2.5 Analysis of environmental variables 33

3.3 Results 34

3.3.1 Butterflies 34

3.3.2 Amphibians 35

3.3.3 Environmental determinants 37

3.4 Discussion 38

3.4.1 Butterfly response to land-use change 38

3.4.2 Amphibian response to land-use change 40

3.4.3 Environmental predictors of butterfly and amphibian diversity patterns 42

3.5 Conclusions 43

4.1 Introduction 51

4.2 Dung beetles (Insecta: Coleoptera: Polyphaga: Scarabaeoidea: Scarabaeidae: Scarabaeinae and Aphodiinae) 51

4.2 Materials and methods 53

4.2.1 Study site 53

4.2.2 Beetle sampling and identification 54

4.3 Results and discussion 55

4.3.1 Species distribution overview 55

4.3.2 Species distribution in contrast to previous records 56

4.3.4 Possible natural and anthropogenic environmental determinants of observed distribution

patterns of dung beetle species 58

4.3.5 Limitations and future work 62

CHAPTER 5: EFFECTS OF LAND-USE CHANGE ON THE COMMUNITIES OF DUNG BEETLES IN LOWLAND TROPICS OF SRI LANKA 67

5.1: Introduction 67

5.4 Materials and methods 70

5.4.3 Beetle sampling 71

5.5 Data analysis 72

5.8 Conclusion 84

CHAPTER 6 DUNG BEETLE COMMUNITIES IN LOWLAND FOREST FRAGMENTS 90

6.1 Introduction 90

6.2 Study site 94

6.3 Materials and methods 94

6.3.1 Dung beetle sampling 94

6.3.2 Environmental variables and fragment characteristics 95

6.3.3 Fragment characteristics 95

6.4 Data analysis 96

6.4.1 Summary statistics and community parameters 96

6.4.2 Comparison of species richness, abundance across the fragments 96

6.5 Results 99

6.5.1 Community comparison 100

6.5.2 Relating species richness, abundance, fragment characteristics and environmental variables 101

6.6 Discussion 104

6.7 Conclusion 111

CHAPTER 7 HABITAT CHANGE IMPAIRS ECOLOGICAL SERVICES PROVIDED BY SCARABAEINAE DUNG BEETLES IN THE TROPICAL LOWLANDS OF SRI LANKA 122

7.1 Introduction 122

7.2 Materials and methods 123

7.2.1 Study area 123

7.2.2 Sampling design for dung removal and dung beetle trapping 123

7.2.3 Dung removal experiments 124

7.2.4 Beetle sampling 125

7.2.5 Statistical analysis 125

7.3 Results 126

7.4 Discussion 127

7.4.1 Functional responses of dung beetles 127

7.4.2 Economic implications for conservation 129

7.5 Conclusions 131

CHAPTER 8: GENERAL DISCUSSION – RESEARCH FINDINGS AND THEIR APPLICABILITY 137

8.1 Ecological trends 137

8.2 Conservation implications 143

BIBLIOGRAPHY 150

APPENDICES 184

functioning in Sri Lanka using community data This study was conducted in the lowland wet zone, the most species-rich zone of Sri Lanka, but which has been severely affected due to logging and conversion to agriculture and home gardens This is the first ecological study to

be conducted in the area that encompasses several modified land use types Throughout this thesis, selected modified land use types; old selectively logged forest, monoculture

plantations, home gardens and forest fragments, are compared with primary forests to

evaluate the effects of habitat disturbance Amphibians, butterflies and Scarabaeinae dung beetles were chosen as focal taxa in this study because they are known to be among the best indicators of habitat disturbance due to their sensitivity to habitat changes and cost efficiency

of sampling In the second chapter, I provide background information to Sri Lanka’s

conservation history at the study sites Chapter three examines effects of anthropogenic land use and selective logging by surveying the diversity and community composition of

amphibians and butterflies in primary forest and across the chosen land use types I found that amphibians, specially endemics and direct developing species were more susceptible to habitat modification than butterflies in the lowland wet zone landscape The environmental determinants of the communities indicated that structural variables of the habitats were more important for amphibians, while butterflies communities were more responsive to climatic variables In chapter 4, I include a taxonomic update, distribution maps, and a photographic

I report the diversity and abundance of Scarabaeinae beetles in multiple land use areas over a wide geographic range and twenty forest fragments in the lowland wet zone I found that diversity and abundance negatively responded to anthropogenic land use in tea plantations and home gardens, primarily through altered abundance and community composition; total species richness was less affected Communities in more than 70% of forest fragments were significantly different from the primary forest and those differences were best explained by

fragment area, area to edge ratio and some abiotic and structural environmental variables (i.e

soil temperature, soil pH, maturity of the forest indicated by DBH profile) Finally, I relate dung beetle species richness and abundance to ecosystem functioning by studying dung removal across the same gradient of land-use change, and then discuss how disturbances can affect dung removal service and nutrient recycling Dung removal was negatively affected by land use change, primarily through altered abundance and functional group diversity Further,

I discuss the importance of restoring nutrient enrichment of soil through dung removal, and the potential economic benefits for agriculture This dissertation provides the most coherent picture to date of how amphibians, butterflies, and mostly Scarabaeinae dung beetles are affected by land use change and forest fragmentation in Sri Lanka, and how ecosystem

functioning of dung beetles is influenced by habitat modification In addition, the study surmounts some of the hurdles to tropical conservation research by supplementing the limited knowledge on ecological effects of habitat disturbance in South Asia specifically by

highlighting an ecologically little known country in the region The research findings can be

1

2

LIST OF FIGURES

Figure 2.1 Major climatic zones of Sri Lanka

Figure 2.2 Map of Sri Lanka and the study sites

Figure 2.3 Categories of peer reviewed publications from the Sinharaja forest

Figure 3.1 Map of the Sinharaja forest-lowland sector

Figure 3.2 Nonmetric multidimensional scaling ordination of Bray-Curtis similarity

indices for butterflies and amphibians Figure 4.1 Major bio-climatic zones of Sri Lanka and sampling locations of the

island wide dung beetle survey Figure 4.2 Dung beetle trap designs used during the study

Figure 4.3 Distribution of dung beetle species according to body size, across the

bio-climatic zones Figure 5.1 Nonmetric multidimensional scaling ordination of Bray-Curtis similarity

indices for dung beetles Figure 5.2 Relating the land use and environmental variables to dung beetle richness

abundance data using Constrained Correspondence Analysis Figure 5.3 Recursive partitioning analysis of the average species richness in the

landscape

Figure 6.1 Locations of the sampled forest fragments

Figure 6.2 Recursive partitioning analysis of the species richness and abundance data

vs fragment characteristics

Figure 6.3 Constrained Correspondence analysis: species richness and abundance

data vs fragment characteristics and environmental variables Figure 7.1 Average percentage of cow dung removal by dung beetles in four habitat

types and two seasons Figure 7.2 Correlation of fertilizer use and tea production in Sri Lanka ( 1961 –

2002)

1

2

LIST OF TABLES

Table 3.1 Summary statistics of butterfly and amphibian sampling

Table 3.2 Environmental determinants of butterfly species richness and abundance

Table 3.3 Environmental determinants of amphibian species richness and

abundance Table 5.1 Summary statistics of dung beetle species richness, abundance and

biomass of habitat types Table5 2 Summary of results from the generalized mixed-effects models for the

effect of land-use change on species richness Table 5.3 Community similarity among land use types quantified by Morrisita

Horn index Table 6.1 Summary statistics of dung beetle species richness and abundance of

individual forest fragments

T able 6.2a Summary of community parameters calculated for 20 forest fragments

Table 6.2b Community similarity among fragments compared using two measures

Table 6.3 Generalized linear models: fragment characteristics as determinants of

dung beetle species richness

Table 7.1 Dung removal as a function of temporal and spatial parameters in the

lowland wet zone of Sri Lanka

Table 7.2 Dung removal as a function of species richness and abundance of

different dung beetle guilds

1

2

APPENDICES

Appendix 3.1 Distribution and endemism of butterfly species in a lowland wet zone

landscape, Sri Lanka Appendix 3.2 Species accumulation, population abundance and rarefaction curves for

butterflies in sampled habitat types Appendix 3.3 Summary statistic of the multi-response permutation procedure results

for community comparison of amphibians and butterflies Appendix 3.4 Distribution, endemism and conservation status of amphibian species in

a lowland wet zone landscape, Sri Lanka Appendix 3.5 Species accumulation, population abundance and rarefaction curves for

amphibians in sampled habitat types

Appendix 3.6 (a) Environmental variables that best describe the species richness and

abundance of butterflies and amphibians in all sampled habitats and (b) the correlation matrix

Appendix 4.1 Species checklist of the Scarabaeinae beetles recorded in Sri Lanka

Appendix 4.2 Dung beetle fauna of Sri Lanka (Plates 1-14)

Appendix 4.3 A representative set of the distribution maps of Scarabaeinae beetles in

Sri Lanka Appendix 5.1 Species accumulation and population abundance curves for dung beetles

in sampled habitat types

Appendix 5.2 Dung beetle abundance and seasonal variation across land use types

Appendix 5.3 Proportionate abundance of functional groups of dung beetles across

land use types

Appendix 5.4 Environmental and structural variables that best describe the species

richness and abundance of dung beetles in all sampled habitats Appendix 5.5 Attributes of Scarabaeinae and Aphodinae dung beetle species recorded

from the study sites Appendix 6.1a Species accumulation curves for dung beetles in all sampled fragments

Appendix 6.1b Population abundance curves for dung beetles in all sampled fragments

Appendix 6.2 Influence of fragment characteristics on standardized dung beetle

community parameters

Appendix 6.3 Fragment characteristics used for the analysis

Appendix 6.4 Environmental variables used for the analysis

Appendix 8.1 Comparison amphibian, butterfly and dung beetles across forests and

home gardens

CHAPTER 1: GENERAL INTRODUCTION

A century ago, environmental protection was primarily a means to sustain wildlife for

scientific and recreational interest This theme has shifted over time, driven by the

exponential growth of human populations and rapid economic development (Sodhi et al.,

2007) There is enormous pressure on natural resources in the tropics, particularly fauna and flora (Jha and Bawa, 2006) Wildlife is now encompassed by the term “biodiversity” and is conserved under general legislation instead of specific conservation laws Sustainable

utilization has replaced harvesting, and ecosystem approaches are preferred over species approaches Conservation has become a participatory process Adaptive management,

monitoring and evaluation have become necessary in regulations As the severity of global

biodiversity loss continues to increase (Reid and Souza, 2005; Sodhi et al 2007),

conservation is no longer a matter of interest but a dynamic, multidisciplinary field and a battle for survival

The current global biodiversity crisis is driven primarily by the clearance and

conversion of tropical rainforests, which contain most of the planet’s species (Turner and Corlett, 1996) Unprecedented loss of tropical forest cover is due mainly to the rapid

expansion of agriculture and other human-modified habitats (Jang et al, 1996; Whitmore,

1997; Laurance, 1999; Sodhi and Brook, 2006) Tropical forests are either completely cleared and converted to anthropogenic uses or partially cleared, leaving forest patches embedded in

a matrix of plantations, home gardens and other anthropogenically modified habitats (Turner, 1996) These patches are also interchangeably called as forest fragments or remnants

(Lindenmayer and Fischer, 2007) Both types of habitat disturbance severely impact native forest biota Resulting changes in the environment include altered biotic community

composition, behavioural changes in animals, and extinction due to the inability to adjust to

habitat changes Disruption of original community structures affects species that are

functionally important, thus altering ecosystem functions (Terborgh et al 2001) The

apparent link between biodiversity and ecosystem functioning has accelerated efforts to gain

scientific knowledge and implement conservation measures (Loreau et al., 2001) Frequent

changes in wildlife and forestry regulations, designation of protected areas, and joining international conventions on conservation are among the political measures adopted by many tropical countries to help curb biodiversity loss The scientific community is focused on understanding the response of biotic communities to habitat change and on providing

empirical frameworks to mitigate the loss (Tilman, 1999; Fahrig, 2003; Balmford and Bond, 2005)

Scientific evidence necessary to document the vulnerability of species and to identify taxonomic groups most susceptible to disturbance is still scarce in some parts of the tropics (Dunn, 2004b) Biodiversity science faces a number of problems in many South Asian

countries (Bawa, 2006) where logistical hurdles delay implementation of management

practices suggested by research Quantification of the effects of anthropogenic modifications

on habitat quality and the effectiveness of conservation measures are still inadequate “Cut and paste” resource management where decisions based on research findings from one

country or region are applied elsewhere without suitability trials may be more harmful than inaction I have observed this in Sri Lanka and it is known to be practiced in other Asian countries Adopting a “one-size-fits-all” approach is another issue partially arising from lack

of data where forest areas are protected without monitoring degradation caused by human impacts or considering the ecological value of habitats is detrimental to the remaining

biological communities and a waste of limited resources

Research on tropical forest fragmentation and modification has come mainly from

studies in the Neotropics (Didham et al, 1998; Terborgh et al, 2001; Fahrig, 2003;

Cushman, 2006); well-replicated studies in South and Southeast Asia are limited This dearth

of research in Asia is disturbing because the Asian tropics are the most imperiled on Earth

(Sodhi et al., 2004) Because of limited resources, conservation efforts are prioritized based

on the magnitude of threats (Spector and Forsyth, 1998; Spector, 2002) With the exceptions

of Australia and tiny Singapore, every country with tropical rain forest is a developing

economy and thus poorly endowed to deal with conservation threats Further, the

undervaluation of biodiversity resources by the global market economy provides developing countries with less incentive for conservation (Lapham and Livermore, 2003) Their inability

to identify threats to biodiversity often handicaps their ability to act on conservation

emergencies within their own borders

Large, connected primary habitats and regions of rapid species turnover are

prioritized in planning efforts for their conservation value (Wikramanayake et al., 1998; Foreman et al., 2000; Spector, 2006) In light of this, understanding the habitat value of

anthropogenic land-use areas is often neglected Species richness, abundance and community composition in anthropogenically modified areas varies according to habitat type and quality Nevertheless, many human-dominated landscapes may prove to be valuable habitats and could be enhanced for their habitat value with suitable management

The present study examines the effects of land use change and forest fragmentation

on biodiversity and ecosystem functioning in the tropical lowlands of Sri Lanka More

specifically, the study provides an understanding of how amphibians, butterflies, and

Scarabaeinae dung beetles are affected by habitat disturbance in Sri Lanka and how

ecosystem functioning of dung beetles is influenced by changing habitats Sri Lanka is an

poorly understood This paucity of scientific data hampers the evaluation of existing

conservation practices and informed management of forest and modified habitats

Amphibians, butterflies and Scarabaeidae dung beetles are globally known to be

among the best indicators of habitat disturbance (Gardner et al, 2008) due to their sensitivity

to habitat changes and cost-efficient sampling (Gardner et al, 2008), and were therefore

chosen as focal taxa in this study Effective indicator responses are measurable within

limitations of sampling protocols (Gardner, 2010) Variation in the diversity of indicator taxa reflects the link between ecological conditions and management or disturbance In addition to understanding taxon specific responses to disturbance, studying a known indicator in novel terrain allows assessment of its response variability and utility as an indicator or a

“calibration” of its indicator value in a different environment (Gardner, 2010)

Few studies have explored the responses of multiple taxa under different management

regimes or interventions in the same landscape (Lawton et al., 1998; Schulze et al., 2004; Barlow et al., 2007; Smith et al., 2008) The use of multiple taxa provides a better

understanding of ecological consequences of a habitat change as responses by each group may differ (Gardner, 2010)

In this study, the choice of habitat types and the bioclimatic zone were based on the outcomes of forest cover loss in Sri Lanka during the past 150 years The lowland wet zone was the most severely affected due to logging, but it is also the most species-rich (Chapter 2) Throughout this thesis, selected modified land use types, old selectively logged forest,

monoculture plantations, and home gardens are compared with primary forests as a reference point to evaluate the effects

in Sri Lanka In my fourth chapter, I have included a taxonomic update of Sri Lankan

Scarabaeinae beetles Much of Sri Lanka’s invertebrate fauna remains unknown except for work done in early 20th century and published in Fauna of British India Sri Lanka’s dung

beetles were first included in Arrow, (1931), followed by a few subsequent addition of

species in Balthasar (1963) I collected dung beetles across the country and prepared a

reference collection to identify dung beetles gathered as data I found that the distribution of dung beetle species across the island has changed in comparison to previous records In the fifth and sixth chapters, I explore the diversity and abundance of Scarabaeinae and

Aphodiinae dung beetles (referred to as dung beetles here after) in multiple land use areas and forest fragments in the lowland wet zone Dung beetles were selected because of their well-known indicator potential (described in chapter 4), and the feasibility of sampling over such a large landscape, and their link to ecosystem functioning In my seventh chapter, I relate diversity to ecosystem functioning by studying dung removal by dung beetles across a gradient of land-use change, then discuss how disturbances can affect useful functions

through altered community structure Further, I discuss the importance of restoring

ecosystem functions and potential economic benefits

The specific objectives of the study were to assess the responses of three selected taxa

to habitat disturbance and to relate altered diversity to function The general objectives of the

conservation of pristine forests and management of anthropogenic land use areas to increase their conservation value

CHAPTER 2: STUDY SITE

2.1 Sri Lanka and the Lowland Wet Zone: a biodiversity hotspot

Sri Lanka is a small but biologically rich island of 6,570,134 ha located at the southern point

of the Indian sub-continent between 5˚ 54' and 9˚ 52' A variety of biomes are found in Sri Lanka, and this is in large part responsible for the country’s high species richness

Topographically, a south central massif 2500 m above sea level (asl) is surrounded by broad, lowland plains that reach 75 m asl Climate varies across the island due to a combination of varying rainfall patterns driven by elevational gradients and monsoons Three, broad climatic regions are recognized: wet zone, dry zone, and intermediate zone (IUCN, 1993; Fig 2.1a)

Whereas the entire dry zone is lowland, the other two zones are further subdivided on the basis of altitude (MENR, 2002) In the wet zone, forests have been categorized as

tropical, wet lowland evergreen forests (0-1000 m); wet sub-montane forests (1000-1500 m); and wet montane forests (1500-2500 m) Tropical dry, mixed evergreen forests, riverine vegetation, tropical moist evergreen forest, thorny scrub and mangrove swamps are other types of vegetation found in the island (NBSAP, 1999)

Sri Lanka was connected to India during the Pleistocene ages and the fauna is

evolutionarily related to the Indian mainland However, after the sea level rise, isolation and limited migrations allowed a large number of endemic species to evolve over a period of 25

million years in the mid-Paleocene to late Eocene Epochs (Bossuyt et al., 2004) Southwest

of Sri Lanka is known as the Highland Complex (IUCN and MENR, 2007) is of considerable biogeographic significance as it holds an exceptional diversity of a relict biota distinct from that of the Western Ghats of India (Ashton & Gunatilleke, 1987) The biota of Sri Lanka's south-western rainforests and the montane forests also show evidence of significant and prolonged isolation from both the island's dry zone and from peninsular India

a result of been frequent migrations (Ashton and Gunatilleke, 1987)

The southwest section of the island where this study was conducted falls within the wet zone, covering an area of 20,000 km2 Lowland, wet evergreen forest dominant

vegetation in the southwest lowlands; (mean annual temperature 28.0˚ C, rainfall 2300 –

5000 mm, humidity 75-85%; Pemadasa and Gunatilleke, 1981) Many of the endemic species are concentrated in these scattered fragments of rainforest, which are surrounded by agro-ecosystems and human settlements (IUCN, 1997) The forest is stratified and characterized

by a tall, dense canopy of about 30 m with emergents reaching 45 m or more Vegetation is dominated by tree species in the families; Dipterocarpaceae, Clusiaceae, Myrtaceae and

Sapotaceae (Gunatilleke et al., 1996) A broad sub-canopy of 15-30 m and a shrub layer

mainly consisting of tree saplings and a sparse herbaceous ground flora complete the forest profile (IUCN and WCMC, 1997)

The high degree of endemism has led to southwest Sri Lanka being designated a

biodiversity hotspot (Myers et al., 2000) About 75% of all endemic species, including 88%

of the flowering plants (Dassanayake et al., 1980–2004), 80% of the fresh water crabs (Bahir

et al., 2005), and a major proportion of amphibians (Meegaskumbura et al., 2002;

Manamendra-Arachchi and Pethiyagoda, 2005) and land snails (Naggs and Raheem, 2000;

Naggs et al., 2005) are recorded in this area Nearly 95% of the wet zone forests were

converted to agriculture or other land uses in the last century (Myers et al., 2000; Mittermeier

et al., 2004; Bahir et al., 2005) There is a great demand for land for development as the wet

zone contain 67% of the island’s 19 million human inhabitants (Anon., 2003)

conservation This transition was driven primarily by public protest for the massive

deforestation, intervention by the scientific community, shifting public opinion towards forest conservation, and obligations to international conventions The concepts of biodiversity conservation and environmental management have been evolving worldwide as natural environments deteriorate and biodiversity declines; policies and regulations have kept pace in Sri Lanka as elsewhere (Kotagama pers com)

Sri Lanka’s protected areas are among the most extensive in the world as a proportion

of total land area Around 15% of the island is protected and only about 18% of protected areas fall within the biologically richest wet zone (Fig 2.1; IUCN and WCMC, 1997) The national government holds the primary jurisdiction for conservation of the country’s

biodiversity, managed by two governmental departments: the Forest Department and the Department of Wildlife Conservation The Department of Forest Conservation administers forest reserves and wilderness areas confined mostly to the wet zone The Department of Wildlife Conservation (DWLC) is responsible for maintaining national parks, nature

reserves, and wildlife in other wilderness areas that occur mainly in the dry zone Until

recently, both of these were administered by the Ministry of Environment and Natural

Resources, but recent political changes have prompted the DWLC to now be governed by the

(CITES), the Convention on Wetlands of International Importance (RAMSAR Convention) and, most significantly, the Convention on Biological Diversity (IUCN, 1994)

Sri Lankan culture has safeguarded wild animal and plant populations for over 2,200 years guided by the conservation ethics and beliefs of Buddhism Mihintale, Anuradhapura, was the world’s first wildlife sanctuary created between 307-266 B.C (Abeywickrama, 1987) Forest protection in Sri Lanka during the modern era was initiated to slow timber felling for export, security, and agriculture The advent of coffee and tea agriculture in 1835, lead to massive deforestation and hunting, especially of large mammals (Gunawardena, 2005;

Jayasuriya et al., 2006) The Wastelands Ordinance of 1840 designated all unoccupied and

uncultivated lands (including forests) for the establishment of tea, rubber, coconut and

cinnamon plantations—even land previously held by local communities and individuals The over-exploitation of forests through agreements between contractors and government

officials severely degraded the silvicultural value of dry zone forests during this era (NBSAP, 1999)

During 1882-1889 attempts were made address deforestation and, to promote

sustainable management These include the appointing of a conservator of forest and the establishment of the Forest Department in 1889 While directive of the department was primarily to conserve and manage state forests to maintain a sustainable yield, the actions were clearly utilization-oriented ((NBSAP, 1999; Wijesinghe, 2003) During the same period, concerns over poaching carried by animal traders prompted immediate legislation to forbid the practice This ordinance readjusted customs duties on firearms, restricted the market of

The first three decades of the 20th century saw more regulations consolidating existing laws relating to mammal, fish, and wild bird protection (Wijesinghe, 2003) These changes also halted all forms of commercial exploitation of wild animals, imposed strict regulations

on hunting and increased the protection status of demarcated wilderness areas The first forest policy was ratified by the State Council in December 1935 to serve several functions: 1) coordinate conservation activities for indigenous fauna and flora; 2) conserve water supplies and prevent erosion; and 3) make the island self-supporting in timber and other essential forest products (MENR, 2002) By this time, the administration of all forests was placed under the Ministry of Agriculture and Lands, and the role of the Forest Department had completely shifted to timber harvesting 1937 saw the most significant legislation in the conservation history; the Fauna and Flora Protection Ordinance (FFPO) -No.2 of 1937

(Chapter 325 of the Legislative Enactments), which was declared to recognize the

preservation of indigenous fauna and flora (MENR, 2002) Many national parks, strict nature reserves and sanctuaries were created under the ordinance Some were challenged because areas designated for animal conservation by the act were earmarked for timber extraction by the Forest Department Most areas designated as strict nature reserves for animal

conservation were of poor timber value in both dry and wet zones The Department of

Wildlife Conservation was established in 1949 and management responsibilities for all

The beginning of the Forest Department’s transition from timber supplier to

environmental stewardship started with the redefining of the forest policy in 1951(NBSAP, 1999) However, logging continued amidst the rapidly developing conservation and scientific interest by local and international parties A new forest policy was drafted in 1980 to deal with major concerns in environmental conservation for scientific and socio-economic reasons (NBSAP, 1999) It also ensured a supply of wood for agriculture, fuel, and domestic timber uses Local community participation was encouraged through social forestry (NBSAP, 1999)

In 1987, the World Bank launched a controversial project to prepare a forestry sector master plan The master plan, although supported by the government and intended to support

conservation, was primarily concerned with maximizing timber extraction Out of 278,000 ha

of wet zone and montane forests, 159,000 ha were to be protected, while 119,000 ha were to

be harvested However, protection of montane forests under this plan was redundant and unnecessary because these areas were already protected as important watersheds Forests destined for logging were all lowland rainforest of extreme biodiversity value (Wijesinghe, 2003) The Director of the Wildlife Department initiated formulating a National Policy for Wildlife Conservation and in 1990, with the commencement of the National Conservation Review (NCR), a moratorium on logging wet zone forests was enacted and later extended to include to the dry zone (Green and Gunewardene, 1997) The NCR thus completely shifted the role of the Forest Department to biodiversity conservation When the NCR ended in 2000, the concept of Conservation Forests was established As legislative changes were forcing the Forest Department to be conservation-oriented, the role of timber harvesting passed to the newly established Sri Lanka timber corporation

resources Plantation forests were established to supply timber and wood for paper

production To reforest the logged areas, pine trees (Pinus caribea.) were planted, and as a

result, Sinharaja rain forest of the wet zone (one of the three main study sites described later)

has a belt of Pinus around its periphery separating it from nearby villages Use of Pinus was challenged eventually, and today, broad-leafed Cassia is recommended instead (Ashton et al., 1997) Since the early 1990s, there have been many national conservation plans such as

the National Environmental Action Plan; 1992-1996, and other programs launched by the government and backed by international groups (i.e The final transformation of the Forest Department into a conservation entity came in 2010 when its name was changed to the

Department of Forest Conservation

Sri Lankan culture is predisposed to nature conservation, as indicated in the second Republican Constitution (1978), which states that, “The State shall protect, preserve and improve the environment for the benefit of the community” (Article 27.14) and that “it is the duty of every person in Sri Lanka to protect nature and conserve its riches" (Article 28F;

Jayasuriya et al., 2006) When the 30-year civil war ended in 2009, the country opened to

rapid development and commercial exploitation Exploitation of fauna and flora for scientific purposes and for export sale is strictly monitored by the government departments In contrast, the booming interest in boosting the country’s economy through intensified tourism activities has brought a new set of opportunities, threats, and challenges to biodiversity conservation Publicity campaigns are attracting global attention to Sri Lankan wildlife, especially the charismatic species, and much infrastructure has been developed in the periphery of protected areas and adjoining landscapes National parks and other protected areas, which serve as the

organizations undertake conservation activities Some of them are small-scale, localized, and unmonitored, while there are a few conducting effective, long-term programs with lasting effects, particularly to mitigate human-elephant conflict and awareness of this issue Sri Lankan citizens are, in general, highly informed about conservation, and prevent exploitation, most recently with the aid of media and social networking

2.3 Study sites: Sinharaja, Kanneliya and Kottawa

The study was conducted in the lowland wet zone landscape around 20 forest fragments of 10-200 ha and three large forests: Sinharaja, Kanneliya and Kottawa (Fig 2.2) Only 4.6% of the wet zone (800 km2) now contains natural forest comprising about 140 fragments

(Jayasuriya et al., 2006)

2.3.1 Sinharaja Man and Biosphere Reserve

Sinharaja forest is a fragile and least resilient forest with high species richness, high

endemicity, and a unique climax vegetation (Gunetilleke et al., 2004) Sinharaja forest

(11,331 ha: elevation 200-1150 m asl) is located in the wet zone in southwest Sri Lanka (6˚ 21–26’ N, 80˚ 21–34’ E) The mean monthly temperature ranges from 18–27˚ C The rainfall ranges from 3750–5000 mm, and falls primarily during the southwest monsoon (May–July)

intense public protest The Decision was mostly driven by the intervention of the scientific community and the public, although there was no scientific evidence to support legislative decision-making at the time (Wijesinghe, 2003) The same year, 8,500 ha of Sinharaja’s lowland forest, comprising 65% primary forest and 34% fern lands/secondary forest, were declared an International Biosphere Reserve In the 1980s, an additional 2,687 ha of lower montane forests on the eastern end was included in the Sinharaja reserve, raising the total land area to 11,187 ha In 1985, the Forest Department established a live forest boundary by

planting Pinus caribaea Subsequently, the Sinharaja National Heritage Wilderness Area was

declared, and in 1990, UNESCO recognized it as a Natural World Heritage site (MENR, 2002)

Early research in Sinharaja was limited to gathering herbarium specimens—even as far back as 1855 (Dassanayke and Clayton, 1996) Today, many local and internationally peer reviewed scientific publications have resulted from research at Sinharaja Fig 2.3

categorizes these peer reviewed international publications from 1985 to date according to the taxa studied Most of these works, including this study, have been conducted in the western sector of the reserve, accessed via Kudawa A major proportion of the studies in Sinharaja are

on long-term forest dynamics, plant phenology (i.e., Shorea species), studies of non-timber

(Gunetilleke et al., 2004) A 50-hectare forest dynamics plot affiliated with the CTFS

network (Center for Tropical Forest Science of the Smithsonian Tropical Research Institute)

is located within Sinharaja The National Conservation Review sampled the western, eastern, northern, and southern sectors and recorded 337 species of woody plants in 71 families and

191 genera Of these, 192 species (57%) were endemic (IUCN and WCMC, 1997), and these numbers have increased over the time due to taxonomic changes One hundred and seventy nine plant species, of which 33% are endemic, are used as non-timber forest products by

local people (Gunetilleke et al., 1993)

An initial biodiversity survey revealed that Sinharaja harbours 19% of the fish, 53%

of the amphibians, 57% of the reptiles, 90% of the birds and 67% of the mammals endemic to Sri Lanka (Kotagama, 1989) While species checklists of numerous taxa exist, few have been studied in detail A long-term research program on avian flocks by a single team lead by Sarath Kotagama and Eben Goodale (1982-2010) constitutes a major proportion of the peer-reviewed publications of fauna Sarath Kotagama (1985-1986), and Mayuri Wijesinghe (2004-2006) have documented small mammals of Sinharaja Amphibian studies mainly focused on revising taxonomy and describing new species Newly described species include living species and undescribed museum specimens that can no longer be located in the wild

and are presumed extinct (Bahir et al., 2005) Invertebrate studies in Sinharaja have focused

on mosquitoes (Diptera: Culicidae), tiger beetles (Coleoptera: Cicadellidae), snails

(Mollusca: Gastropoda) and most recently on ants (Hymeoptera: Formicidae)

The ecological effects of selective logging and land use change in Sinharaja are less

well known Kotagama et al., (1986) explored the effects of selective logging on small

mammal diversity Wijesinghe and Brooke (2004, 2005) followed by comparing small

mammal distribution patterns and vulnerability of endemics across a land-use gradient from

primary forests and logged forest to Pinus plantations and abandoned human land use areas surrounding Sinharaja Gunewardana et al., (2010) compared ant diversity in primary forest

and selectively logged forest

Research on the abiotic environment at Sinharaja include: records of

physico-chemical and biological properties of soil across primary and disturbed areas (Maheswaran

and Gunetilleke, 1988, Hafeel and Gunatilleke, 1989); endomycorrhiza in forests and Pinus

plantations; leaf litter decomposition rates and nitrogen fixation in the forest and fern-land

(Maheswaran and Gunetilleke, 1988, 1990); seed bank germination (Uduporuwa et al.,

1997) As summarized in Gunetilleke et al., 2004, many of the socio-economic studies are

anthropological surveys that focus on the utilization of the forest as a resource by local

people, ecological valuation of forest resources, community-based participatory management, and overexploitation through hunting and forest product harvesting and negative

environmental consequences of utilization

2.3.2 Kanneliya conservation forest

The forest of 6,114.4 ha (6˚ 09-18’ N and 80˚ 19-27’ E) is located in the southwestern Sri

Lanka in the southern part of the wet zone It is part of the

Kanneliya-Dediyagala-Nakiyadeniya (KDN) complex The average annual rainfall is about 4,445 mm and falls mainly during the Southwest monsoon season from mid May to the end of September, and the mean monthly temperature is around 27˚ C (IUCN and WCMC, 1997) Selective logging

in Kanneliya was halted by the moratorium on all logging within the wet zone, and the forest was declared an International Biosphere Reserve in 2004 (Jayasuriya and Abayawardana, 2008)

(e.g., Nannophrys guentheri, Ramanella palmata and Icthyophis glutinosus), reptiles, and fish

(NBSAP 1999) Kanneliya was included in an internationally funded scientific project

conducted by the Ministry of Environment and Natural Resources from 2000-2006, to

promote protection of the lowland forest ecosystems through community participation

There are about 78 villages surrounding the entire KDN complex Their primary source of income is from tea and paddy cultivation More than half the community is of the government identified poor income level Most home gardens are partially cultivated with food or commercial crops The locals are still dependent on the non timber forest products from the forest complex although utilization is much low due to the protection status

(Bandaratillake, 2003) The KDN management plan of 1995 has outlined the involvement of local people in conservation through community participation

2.3.3 Kottawa Kombala conservation forest

Kottawa Kombala conservation forest is protected under the Forest Department and is located

in Galle District (6˚ 09-18’ N and 80˚ 19-27’ E; ~2000 ha) The forest reserve includes

Kottawa Rainforest and Arboretum margined by a main road and Hiyare selectively logged

Gunatilleke (1978); Pemadasa and Gunatilleke (1981); Gunatilleke and Gunatilleke, (1985); Gunatilleke and Ashton (1987), and participatory forest management studies by Yamamoto (2000) Field observations in Kottawa show the return of some of the canopy and sub-canopy endemic plant species following selective logging However, it is poorly known how endemic

herbs and shrubs recovered after disturbance (Gunatilleke et al., 2004)

The earliest known biodiversity research work on Kanneliya and Kottawa was

conducted in the early 1980 through the Nation Conservation Review Programme (NCR Data) by the Forest Department with the collaboration of International Union for

Conservation of Nature (IUCN) The more recent 5-year survey of the biodiversity of Sri Lankan natural forests established a faunal checklist for Kanneliya and Kottawa These studies highlighted the importance of the remaining rainforests of the wet zone for the

conservation of woody plant, animal diversity and for the protection of watersheds (Green and Gunawardena, 1997) Some of the peer-reviewed publications also include ecological

studies examining the effects of forest fragmentation on snails (Raheem et al., 2008, 2009), the distribution of Slender Loris (Nekaris et al., 2005).and species descriptions of fish,

amphibians, lizards, mollusks and insects

*The above section was written from personnel communication with Prof S.W Kotagama, Professor of Environmental Sciences and former Director of the Department of Wildlife Conservation Except for the cited references, facts in this section are extracted from Prof Kotagama’s unpublished notes with permission

Number of peer reviewed publications

Figure 2.3: Peer reviewed publications, categorized by subject, resulting from research conducted in Sinharaja rainforest (1985-2010) Data were obtained from Web of Science, Google, and Google Scholar searches using the following search terms: Sinharaja, wet zone forest, lowland wet zone, Sri

Megadiverse lowland rainforests are being felled throughout tropical Asia for agricultural

expansion, bio-fuel production and urban development (Sodhi et al., 2010) While research

findings on the response of biotic communities to land-use change have helped to mitigate tropical biodiversity loss in other parts of the world (Balmford and Bond, 2005), such studies are urgently needed in tropical Asia, which suffer the highest tropical deforestation rates

globally (Sodhi and Brook, 2006; Sodhi et al., 2009)

Sinharaja Man and Biosphere Reserve (hereafter, Sinharaja), is the largest and one of the most species-rich lowland rainforests remaining in Sri Lanka Although part of it has been selectively logged (Gunatilleke and Gunatilleke, 1980), the forest has regenerated over several decades The forest is now protected and surrounded by home gardens, tea

plantations, and non-native Pinus caribea (pine) trees acting as a buffer between the forest and the adjoining villages (Ashton et al., 1997)

Studying distribution across a land-use gradient allows to understand species

responses to disturbance (i.e Wijesinghe and Brooke, 2005; Koh, 2007, Sodhi et al., 2007;

Gardner, 2010), and ecological characteristics that make them adaptable or vulnerable

(Connell and Orians 1964) Most faunistic studies on the effects of habitat disturbance focus

on a single taxon and disturbance type (Dunn, 2004b) A multi-taxon approach will

ultimately lead to a more representative picture of land-use change effects because responses

to disturbance may differ between taxa (Gardner, 2010) Further, comparing the effects of multiple types of land-use will allow for more accurate predictions of diversity in

increasingly disturbed and fragmented habitat matrices Examining taxonomically and

Species richness, abundance, and community composition of butterflies and

amphibians in primary forests, are compared with other land use types (selectively-logged

forests, home gardens, and Pinus caribea plantations) in and around Sinharaja forest

Through this comparison, I aimed to assess the impacts of habitat change, resulting from the conversion of forest to human modified habitats As there were no temporal (pre-habitat conversion) data, the associations between species richness/abundance/community

composition and different land use types were used as responses by each taxon Best

environmental predictors of amphibian and butterfly diversity were determined using

Bayesian regression modeling, and compared our results with similar studies from other regions The research findings are used in Chapter eight, to make scientifically informed recommendations for the conservation of forest habitats and management of anthropogenic land use areas in Sri Lanka

3.2 Materials and methods

3.2.1 Study area

The present study was conducted in the lowland section of the Sinharaja Man and Biosphere Reserve (Figure 3.1; see Chapter two for details of the study site) Butterflies and amphibians were sampled in primary forests (PF), selectively-logged forests (SLF), home gardens (HG),

and Pinus caribea plantations (PP) The two forest habitats were collectively termed “forest”

and the other two habitats of human origin were termed “anthropogenic habitats” in some parts of the discussion for the interpretation of data The land use gradient is PF< SLF< PP<HG but the boundaries are not distinct or clear-cut The selectively logged forest is