Thesis Finalise For Minc

FOREST USE AND VERTICAL STRATIFICATION IN FRUITFEEDING BUTTERFLIES OF SIBERUT, INDONESIA

Master thesis by Chung-Lim LUK born 26 January 1983 in Hong Kong

Thesis submitted to the Faculty of Biology, Georg-August-Universität Göttingen, in partial fulfilment of the requirements for the integrated bi-national degree MASTER OF SCIENCE / MASTER OF INTERNATIONAL NATURE CONSERVATION (M.SC. / M.I.N.C.) of Georg-August-Universität Göttingen; Germany and Lincoln University, New Zealand

October 2009

Supervisor/Betreuer: Examiner/Gutachter: Date of Submission/Abgabedatum: German Title/Deutscher Titel:

Dr. Matthias Waltert Prof. Dr. Michael Mühlenberg 30. Oct 2009 Vertikale und horizontale Einnischung bei fruchtsaugenden Schmetterlingen der Insel Siberut, Indonesien

Acknowledgments This study was part of the research supported by the Siberut Conservation Project (SCP). SCP is managed by through a collaboration between the German Primate Centre (DPZ) of Geog-August-Universität Goettingen and Bogor Agricultural University (IPB). Funding of this project was kindly granted by SCP. Academic backstopping of this research was carried out mainly by the Centre for Natural Conservation of Geog-AugustUniversität. First I would like to thank my supervisor Dr. Matthias Waltert, from the Centre for Nature Conservation, for the continuous help, research ideas and academic support. I would like to thank Prof. Keith Hodges and Dr. Thomas Ziegler from the Department of Reproductive Biology at DPZ, giving me the great opportunity to experience the unique Mentawai Forest, not least also for managing all the necessary documents and the financial support. On the Indonesia side I would like to thank Dr. Mudhammed Agil, for issuing all the necessary documents in Indonesia. Thanks Pak Dodo and Aminah for the help of shopping research equipment and for helping with financial issues in Padang. For help in the field I would like to thank my guide Pak Gerson; special thanks also to Dodo, Pak Tasan, Ai and “the Pungut angles” Lia, Ayu, for their “extra help”. Also I would like to thank Dr. Christope Abegg and Pak Johan for managing many issues for me in the project. For identification I would like to thank the staff from The Indonesian Institute of Sciences LIPI, especially Ibu Peggie, and Dr. R.I. Vane-Wright, from Natural History Museum of United Kingdom. I also would like to thank all the staff at the Centre for Nature Conservation, for facilitating my time in Germany Finally I would like to give my biggest thank to my mother, Tak-Mei Wong, for her love, care and support. Without this generous and understanding lady my scientific career would have ended already 3 years ago.

Table of Content 1 Introduction 2 Objectives 3 Study area and methods 3.1 Study area 3.2 Vegetation data 3.3 Butterfly trapping 3.4 Identification 3.5 Statistical analysis

4 Results 4.1 Vegetation structure 4.2 General abundance and diversity 4.3 Differences between forest types 4.3.1 Taxonomical differences between forest sites 4.3.2 Differences between forest sites at species level 4.4 Vertical stratification 4.4.1 Taxonomical differences among vertical height levels 4.4.2 Vertical stratification at species level

5 Discussion 5.1 General 5.2 Furit-feeding butterfly response to disturbance 5.3 Pattern of fruit-feeding buuterfly stratification

1-3 3 4-11 4 7 7 8 9

12-27 12 13 15 18 20 21 24 26

28-37 28 29 32 36

6 Conclusion

38-39

7 References

40-44

Appendices

45-48

Appendix 1 Distribution information and wing size of sampled butterfly species Appendix 2 Timetable of fruit-feeding butterfly sampling Appendix 3 Photographic section

45 46 47

List of Table Table

Content

Page

Table 1 Total number of trees enumerated, tree density and mean basal area, of overstorey and understorey trees, in defined 10m x 10m quadrats in natural and disturbed forest

12

Table 2 Individual numbers of Nymphalidae species trapped in three different heights, and understorey on forest types

14

Table 3 Total number of individuals, species richness, and diversity parameters of fruit-feeding-butterflies trapped in two forest types

15

Table 4 Individual numbers of Nymphalidae trapped regarding to subfamily in natural forest site and disturbed forest site

18

Table 5 Species with abundances significantly different between disturbed and natural forest sites. Statistical test was done by using Mann-Whitney U test

20

Table 6 Numbers of individuals and species richness of fruit-feeding-butterflies trapped at three vertical height levels in Siberut, Indonesia

21

Table 7 Individual numbers of Nymphalidae trapped regarding to subfamily in the three vertical level: 1m, 15m and 30m

24

Table 8 Species considered as indicator regarding to different vertical level

26

List of Figure Figure Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Content Map of the location of the study area in Siberut (extracted from Hadi et al. 2009), and of the 14 trap locations in natural and disturbed forest

Page

5

Cumulative number of Nymphalidae species against individual captured at seven natural forest and seven disturbed forest sites

16

Cumulative number of Nymphalidae species captured against trap days (samples) at seven natural forest and seven forest sites (including two vertical trapping sites in each habitat)

16

Multidimensional scaling of fruit-feeding butterfly communities at understorey trap of five natural forest and five disturbed forest sites

17

Proportional abundance of butterfly according to subfamily in natural forest site and disturbed forest site

19

Cumulative number of Nymphalidae species captured against trap days at the three different vertical heights along the vertical strata of two natural forest and disturbed forest sites

22

Cumulative number of Nymphalidae species captured against trap days at the three different vertical heights along the vertical strata of two natural forest and disturbed forest sites

22

Multidimensional scaling of fruit-feeding butterfly communities at three different trap heights along the vertical strata of two natural forest and two disturbed forest sites

23

Proportional abundance of butterflies according to subfamily at three vertical levels: 1m, 15m and 40m

25

1. Introduction Tropical forests harbouring much of the Earth’s remaining biological diversity, but are experiencing unprecedented rates of deforestation (Laurance 1999; Brooks et al. 2002). The rainforests of South-east Asia are among the most biologically diverse areas in the world (Myers et al. 2000). Among the tropical regions South-east Asia has the highest relative rate of net forest loss and degradation in the humid tropics (Achard et al. 2002), and it is predicted that it could loose up to three-quarters of its original forests and almost half of its species by 2100 (Brooks et al. 2002).

Butterflies (order: Lepidoptera) are among the best-studied insect groups in Southeast Asia in terms of taxonomy and biogeography (D’Abrera 1982, 1985, 1986; Aoki et al. 1982; Tsukada et al. 1985; Tsukada 1991). They are highly sensitive to habitat disturbance and have been used commonly as an indicator taxon for ecological research (Kremen 1994; Koh & Sodhi 2004). However, the responses of insects in general and butterflies in particular to disturbances and deforestation are still relatively poorly known (Koh 2007). For Siberut Island or the Mentawai islands example, no previous ecological studies on the responses of butterflies nor on those of other arthropod communities to disturbances have yet been done.

Siberut and Mentawai are geographically located within Sundaland, which is one of the 34 ‘biodiversity hotspots’ (Mittermeier et al., 2000), regions which not only having high levels of biodiversity and endemic species but are also undergoing immense habitat loss. Siberut is one of the largest islands of the Mentawai Group of Islands, which remained separated from Sumatra for more than 500,000 years (Verstappen 1975). Due to this period of biogeographical separation from the

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

1

mainland, the Mentawai Islands consist of relatively high level of endemic species. In mammals for example, seventeen out of thirty-four species, including all four primate species, found on Siberut are known to be endemic (WWF 1982). More recent ecological studies on the floral and fauna of Siberut and Mentawai were primarily focused on the mega-fauna, especially the four endemic primate species. There has been no previous extensive ecological study on butterflies or insects published. The only publication on the butterfly community of Mentawai is mainly focused on taxonomy and identification (e.g. Aoki et al. 1982, Tsukada et al. 1985 and Tsukada 1991), however the butterfly fauna of the island had been recorded as early as in Hagen (1893).

Many authors had been describing the changes or differences in the butterfly community between disturbed and undisturbed sites at understorey level (e.g., for Asia: Hill et al. 1995; Hamer et al. 1997; Beck and Schulze 2000; Hamer and Hill 2000; Willott et al. 2000; Dumbrell & Hill 2005 for Africa: Kremen 1992, 1994; and for the Neotropics: Lawton et al. 1998; Barlow et al. 2007). However, as butterflies are aerial organisms and research is often ground-based, most of these studies did not achieve to describe the ‘real’ assemblages of butterflies, for which a 3-dimensionally designed research would be more appropriate, especially when examining differences in community structure between habitat types.

The vertical structure of tropical rain forests can be described as having distinctly different vegetation layers (Pomeroy & Service 1986, Schulze et al. 2001; Whitmore 1993). At different vertical heights different abiotic conditions and biotic resources can be observed (e.g. light, food), which provide a gradient of microhabitat within a

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

2

locality. The analysis of vertical distribution patterns and niche breadths of animals along such vertical gradients could provide one of the keys to the understanding of processes underlying species composition in animal communities of multi-layered forest habitats. Studies on vertical stratification of butterflies have also been previously conducted in tropical regions (e.g. Fermon et al. 2005; Schulze et al. 2001; DeVries et al. 1997; Hill et al. 2001).

2 Objectives In South-east Asia, all fruit-feeding butterflies belong to the Nymphalid subfamilies Satyrinae, Morphinae, Nymphalinae and Charaxinae (Corbet & Pendlebury 1992). The aim of this study is to (1) describe the general patterns of vertical stratification in frugivorous butterflies in Siberut and (2) to use them as an indicator group to analyse the impact of human disturbance on the forest. This is done by comparing the vertical distributions, abundance, species richness and diversity of Nymphalidae between natural and disturbed forests; furthermore, the study intends to (3) increase our knowledge about the frugivorous butterfly fauna of Siberut island.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

3

3 Methods 3.1 Study area Siberut is a tropical island, lying at the northern most extremity of the Mentawai group of islands, at a distance of approximately 130 km off the west coast of Sumatra. Due to its separation from Sumatra for more than 500,000 years (Verstappen 1975)., the biodiversity of the Mentawai is quite distinct from the mainland and is characterised by a high degree of endemism, especially when considering its size (Whitten 1982a).

Siberut has a total surface area of 4,030 km² and an estimated population of 25,000 people. The natural vegetation in the island is primarily composed of tropical moist broadleaf forest of the Dipterocarpaceae family (Hadi et al. 2009). However, the environment is now facing the impact of extensive logging and agricultural farming (Fuentes 1996/1997). Apart from priamate study, there have been no extensive studies so far on the ecology of plant or animal communities in Siberut and the Mentawai islands. Til date no ecological study about butterfly or arthropods can be found. The only scientific literature about butterflies are on faunistics and taxonomy (Aoki et al. 1982, Tsukada et al. 1985 and Tsukada 1991)

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

4

Figure 1, Map of the location of the study area in Siberut (extracted from Hadi et al. 2009), and of the 14 trap locations in natural and disturbed forest.

This research was based on work carried out at the field station of the Siberut Conservation Project (SCP). SCP is run by a collaboration between the German Primate Center (DPZ) and Bogor Agricultural University (IPB). The field station is located in the northern part of the island, in the Peleonan forest (1o01’34’S, 98o50’16’E; elevation: 8–180 m above sea level). The Peleonan forest is subject to a long-term international research and conservation program and consists of approximately 5000 hectares of lowland, mostly primary, lowland forest, but also coastal and peat swamp forest and tiny local farms (Quinten 2008). Most of the study area is hilly with elevation measurements ranging from 2-182 m above sea level. The Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

5

forest area is drained by numerous small to medium sized creeks and rivers (Hadi et al. 2009). In total 93 species of trees were recorded in the region, with the tallest tree sampled was 56m, with 73% of the trees were at the height class of 6 -20m. Euphorbiaceae, Myrtaceae, Lauraceae and Moraceae were being the most common tree families found in the site. At species level, species of the genera Mallotus and Knema, as well as Baccaurea sumatrana were most dominant in all dbh classes . (Hadi et al. 2009).

Currently, there is an agreement with local people and the Indonesian officials that the Peleonan forest is being protected for research purpose. At a larger spatial scale, the forest in Siberut is now facing the threat of commercial logging and establishment of oil palm plantations. Large areas of the natural forest were already logged and cleared for wood resources, it is estimated that about 7,000 m³ of timber is exported monthly (Person 2003). As operations expanded, this Figure is likely to increase in the near future (Person 2003).

Our trapping sites were situated along the transact, within 2.0km Northeast of the field station (Figure 1). For the disturbed site chosen, the main cause of anthropological disturbance at the research site was farming by local people, with banana and cocoa being the major crops.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

6

3.2 Vegetation data At each of the fourteen trapping sites, 10 x 10 m vegetation plots were used to assess the vegetation. Plots were centred around each site. In each quadrat, the diameter at breast height (dbh) of all trees with a dbh equal to and greater than 5 cm were recorded. Trees with a dbh of 5-10 cm were defined as understorey trees, and trees >10cm dbh as overstorey trees. Dbh was converted to basal area [m²/ha], calculated separately for understorey and overstorey trees to be used as an index of disturbance.

3.3 Butterfly trapping All trapping activities were conducted in between 6th June and 17th July, 2009. Cylindrical gauze-traps (Rydon 1964) were used, in which a standard portion of mashed, fermenting bananas was placed inside the trap to bait butterflies. All traps were hand-made. Each trap was 30cm in diameter and 60cm long, and comprised a 30x30cm plastic cardboard base. Each trap was stabilised by 2 rings (30cm diameter), with 60 cm long white tissue connecting the rings to produce the cylindrical shape with the bottom open. Nylon strings were used to connected the bottom of the cylinder and the plastic cardboard, to produce a 3cm gap to allow the butterflies to enter. The bait was placed in the middle of the cardboard on a small plate.

Fourteen trapping sites were selected of which seven were in primary forest and seven were situated in disturbed forest. At four of these sites (two primary forest and two disturbed forest sites), traps were established at different vertical heights. At each of these four trapping sites, a tree taller than 30m was chosen in order to suspend traps at different heights. Traps were installed at ground level (1m), in the midstorey (15 m),

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

7

and the canopy (30 m), resulting in a total of 2 x 2 x 3 = 12 traps. Traps were placed for 32 non-rainy days and were checked for butterflies and rebaited every 24hours. A total of 384 trap-days (12 traps x 32 days) were spent to collect fruit-feeding butterflies.

Out of the other 10 trapping sites, five were in natural forest and fiver in disturbed forest. At these 10 sites, no traps were placed in the midstorey or canopy, but only at understorey level (1m):

Traps were left for 16 non-rainy days and were checked, sampled and rebaited every 24hours, with a total of 160 trap-days spent at the 10 understorey sampling sites (5 x 2 x 16).

3.4 Butterfly identification For field identification, most specimens trapped were preserved and were given a “specimen ID”. The preserved specimen was identified later in the laboratory of the Indonesian Institute of Sciences (LIPI). At least three individuals of each species were preserved for later identification. Trapped butterflies of known species identity, were identified and released. The identification was mainly based on the publication of Aoki et al. (1982), Tsukada et al. (1985) and Tsukada (1991)

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

8

3.5 Statistical analysis Differences in capture frequencies between forest sites were analysed using MannWhitney U-statistics. For the vertical stratification of butterfly community, KruskalWallis was used to test for differences among the three levels (1m, 15 m, 30m). The abundance regarding to ‘general’ (the whole nymphalid community), subfamily, and species were tested regarding the factors habitat type and vertical level. All statistical tests were done using the software SPSS 13.0 for Windows

The diversity of the butterfly community in each forest type was analysed using EstimateS 7.52. The species richness estimators, Abundance Based (ACE) and Incidence Based (ICE) were calculated to estimate the size of local species pool. The diversity parameters Fisher’s alpha index, Simpson index, Shannon (or ShannonWeaver) index and evenness were calculated:

The Shannon or Shannon-Weaver index is calculated as:

Where: ni is he number of individuals in species i; the abundance of species i. S is the total number of species. N is the total number of all individuals pi is the relative abundance of each species, calculated as the proportion of individuals of a given species to the total number of individuals in the community

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

9

Evenness is calculated as:

Where: H' is the number derived from the Shannon diversity index and H' max is the maximum value of H'

The Simpson’s index is calculated as:

where S is the number of species N is the total percentage cover or total number of organisms and n is the percentage cover of a species or number of organisms of a species.

Note that D therefore ranges from 0 to 1, with 1 representing infinite diversity and 0 representing no diversity.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

10

Fisher’s alpha diversity index is calculated as:

α = N (1 –x ) x S/N = (1 – x)/ X – ln (x)

Where: N is the total number of all individuals S is the species richness x is estimated by the second solution

Species composition among different vertical strata and between forest types were grouped using a matrix of dissimilarity (1 - Sorensen Index similarity index) and then were analysed using multidimensional scaling.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

11

4 Results 4.1 Vegetation The mean density and basal area of overstorey and understorey trees in natural and disturbed forest are shown in Table 1. No significant difference was found in basal area between the forest types, but the standard deviation was quite high (all > 35%) and basal area of overstorey and understorey trees in natural forest was considerably higher than in disturbed forest (>50% higher in figure). The density of overstory tree in natural forest was significantly higher than disturbed forest (Z=-2.47, p=0.013), however no significant difference was found in the density of understorey trees between forest type (Table 1).

Table 1. Total number of trees enumerated, tree density and mean basal area, of overstorey and understorey trees, in defined 10m x 10m quadrats in natural and disturbed forest. Numbers in brackets represent standard deviation. Statistics from Mann-Whitney U test. Natural (n=7 plots)

Disturbed (n=7 plots)

Z

p value

Overstorey trees enumerated

44

32

-

-

Understorey trees enumerated

25

19

-

-

Overstorey Tree Density [Ind/ha]

628.6 (75.6)

357.1 (190.2)

-2.47

0.013

Understorey Tree density [Ind/ha]

457.1 (207.0)

271.4 (197.6)

-1.55

0.122 (NS)

Total density [Ind/ha]

1085.7 (177.3)

628.6 (111.3)

-3.09

0.002

Overstorey basal area (m²/ha)

49.1 (18.1)

34.5 (20.7)

-1.087

0.277 (NS)

Understorey basal area (m²/ha)

1.82 (0.99)

1.12 (0.64)

-1.22

0.224 (NS)

Total BA [m²/ha]

51.0 (18.5)

35.6 (20.8)

-1.087

0.277 (NS)

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

12

4.2 Overall abundance and diversity A total of 244 fruit-feeding butterfly individuals belonging to 5 subfamilies (Biblidinae, Charaxinae, Morphinae, Nymphalinae and Satyrinae), 14 genera, and 20 species were trapped in both natural and disturbed forests (Table 2).

The Satyrinae showed the highest abundance which contributed to168 individuals, or 69% of total trapped individuals. The Biblidinae were the second most abundant subfamily with 39 individuals (16%), followed by the Morphinae, and the Charaxinae. The Nymphalinae showed the least abundance with only 10 individuals recorded (4%).

Species abundance The most common species found were Neorina lowii (n=62), representing about a quarter of all individuals, followed by Mycalesis maianeas (n=41), Dichorragia nesimachus (n=34) and Mycalesis oresis (n=32).

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

13

Table 2. Individual numbers of Nymphalidae species trapped in three different heights, and understorey on two forest types in Pungut, Siberut, Inodnesia. Trapping was done for 32 rainless days with two replicates for vertical traps, and 16 rainless days, with five replicates for understorey traps. Trapping was being held in between June and July, 2009. The species with an “E” meaning the species is an endemic species to Mentawai Islands. Mycalesis maianeas

16

9

Mycalesis oresis

2 Natural forest 1

12Disturbed forest

Neorina lowii

27 1m

151m

30m

6 1m

154m

30m

Subfamily

Total Species

57

18

11

49

13

9

Biblidinae

Dichorragia nesimachus

3

12

5

2

5

Lexias dirtea

2

Charaxinae

Morphinae

1 5

1

1 Natural

Dist16 urbed

32

13understorey 11

5

1

41

244

5

2

34

2 2

Bassarona teuta

2 3

1

10

2

1

12

2

2

2

4

1

3 1

Lebadea Martha 1

Tanaecia spp.

1

Elymnias nelsoni E

1

2

1

Elymnias panthera

2

1

1

1

1 4

Elymnias spp.

1

2

1

Melanitis phedima

8

1

1

Melanitis zitenius

1

Chung-Lim LUK

1 1

Tanaecia visandra E

Satyrinae

5

1

Zeuxidia amethystus

Bassarona dunya

T62 otal

46

1

Discophora spp

Nymphalinae

41

1

1

2

Amathuxidia amythaon

4

3

Charaxes bernardus Prothoe franck

12

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

2

7

2

10 4

1

11 1

14

4.3 Differences between forest types Overall, numbers of trapped individuals were similar between natural forest (n=132) and disturbed forest (n=112) (Table 3, Chi-square test, p=0.200, χ²=1.639, df=1).

Table 3: Total number of individuals, species richness, and diversity parameters of fruit-feedingbutterflies trapped in two forest types in Siberut, Indonesia, between June and July, 2009. Diversity indices and estimators of total species richness (ACE, ICE) were calculated with 272 samples ((vertical traps: 2 reps x 3 levels x 32 days) + (understorey traps: 5 traps x 16 days)).

Natural forest

Disturbed forest

χ²

p-value

Individuals

132

112

1.639

0.20

Species richness

15

15

Fisher's alpha Index

4.35

4.65

Simpson Index

5.52

7.83

Shannon Index

2.02

2.25

Evenness (E)

0.75

0.81

ACE

16.75

18.22

ICE

16.72

17.68

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

15

16 14

Species

12 10 8

natural

6

disturbed

4 2 0 0

30

60

90

120

Individuals Figure 2. Cumulative number of Nymphalidae species against individual captured at seven natural forest and seven disturbed forest sites ((vertical traps: 2 reps x 3 levels x 32 days) + (understorey traps 5 traps x 16 days)) in Siberut, Indonesia.

16 14

Species

12 10 8 6

natural

4

disturbed

2 0 0

5

10

15 20 Samples

25

30

35

Figure 3. Cumulative number of Nymphalidae species captured against trap days (samples) at seven natural forest and seven forest sites (including two vertical trapping sites in each habitat) in Siberut, Indonesia.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

16

Figure 4. Multidimensional scaling of fruit-feeding butterfly communities at understorey trap of five natural forest (N) and five disturbed forest (D) sites. Scaling is based on Sorensen similarity values. Sites of the same habitat type are connected by lines.

Mainly due to the three common natural forest species Dichorragia nesimachus, Mycalesis maianeas, and Neorina lowii (Table 2), disturbed forest showed slightly higher values in all diversity and species richness indices. However, total fruit-feeding butterfly species richness was similar in natural (15 species) and disturbed forest (16 species). This was also reflected in the Evenness index, indicating the distribution of of individuals to species was more even in disturbed than in natural forest (0.81 vs 0.75). The two species richness estimators, ACE and ICE, indicated species pools of around 18 in natural or 17 in disturbed forest.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

17

4.3.1 Taxonomical differences between forest sites Among the five subfamilies recorded in this study, three, namely Biblidinae, Charaxinae and Satyrinae, were found in very highly significantly different abundances between natural and disturbed forest. Biblidinae (Table 4, n=39, Z=-3.67, p<0.001) and Charaxinae (n=12, Z=-2.966, p<0.001) showed a higher abundance in natural forest than in disturbed forest. Satyrinae (n=168, Z=-4.856, p<0.001) showed opposite results, having a higher abundance in disturbed forest (Table 4). Furthermore, Nymphalidae tended to be more common in natural forest, but this was only nearly significant (Table 4). A MANOVA with the two-dimensional scores resulting from multi-dimensional scaling did not reveal a significant difference in species composition of understorey traps between habitat types (F = 1.788; df = 2, 7; p=0.236, Figure 4).

Table 4. Individual numbers of Nymphalidae trapped regarding to subfamily in natural forest site and disturbed forest site in Pungut, Siberut, Indonesia, between June and July, 2009. Statistical test was done by using Mann-Whitney U test, which the p=0.05 is defined as significance level. No. of spp.

natural forest

disturbed forest

total

Z

p-value

Biblidinae

2

30

9

39

-3.637

<0.001

Charaxinae

2

8

4

12

-2.966

0.003

Morphinae

3

7

8

15

-0.382

0.702 (NS)

Nymphalinae

5

8

2

10

-1.936

0.053(NS)

Satyrinae

8

79

89

168

-4.856

<0.001

Total

20

132

122

244

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

18

100%

80%

60%

Satyrinae Nymphalinae

40%

Morphinae Charaxinae

20%

Biblidinae 0% Natural (n = 132)

Disturbed (n= 112)

Figure 5. Proportional abundance of butterfly according to subfamily in natural forest site and disturbed forest site, in Pungut, Siberut, Inodnesia.

In general, Satyrinae was the dominant subfamily, which contributed 60.0% (n=79) and 67.4%(n=89) to butterflies found in natural and disturbed forest respectively (Figure 5). Bilblidinae were the second most abundant subfamily in both forest types, contributing to 22.3% (n=30) of the natural forest butterflies, but represented only 6.8% (n=9) of the individuals in disturbed forest.The other three (less abundant) subfamilies Charaxinae, Nymphalinae and Morphinae, represented the remaining individuals (total 17.5% in natural forest and 12.5% in disturbed forest) of the Nyphalid count.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

19

4.3.2 Differences between forest sites at species level Of the 20 Nymphalidae species sampled, only four species were found to have a significantly different abundance between forest types (Mann-Whitney U test for species with at least 7 individuals, Table 5). Results wer highly significant (p<0.001) for all four species: Dichorragia nesimachus (Z=-3.037), Mycalesis maianeas (Z=4.231) and Neorina lowii showed significantly higher numbers of individuals in natural forest (n=25, n= 28 & n=41, respectively) than in disturbed forest (n=9, n=13 & n=21, respectively), whereas Mycalesis oresis (Z=-5.259) showed the opposite pattern (Table 5).

Table 5: Species with abundances significantly different between disturbed and natural forest sites. Statistical test was done by using Mann-Whitney U test, with 0.05 as significance level. Only species with total abundance number >7 were tested.

Natural forest

Disturbed forest

Total

p-value

Z

Dichorragia nesimachus

25

9

34

<0.001

-3.037

Mycalesis maianeas

28

13

41

<0.001

-4.231

Mycalesis oresis

4

28

32

<0.001

-3.439

Neorina lowii

41

21

62

<0.001

-5.259

Given that these distribution patterns could be potentially biased due to the combination of possible shifts in vertical stratifiation and the inclusion of understorey trapping sites, an analysis was done excluding the data from the understorey trapping sites: Distributions tested by a Chi-Square test were similar and also significant for Dichorragia nesimachus (χ²=6.259, df=1, p=0.01), Mycalesis oresis (χ²=5.400, df=1, p=0.020) and Neorina lowii (8.526, df=1, p=0.004). For Mycalesis maianeas, the distribution was also similar but not significant at the 5% level (1.960, df=1, p=0.162). Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

20

4.4 Vertical stratification Abundance was generally highest at trap height 1m which amounted to 67.5%, or 106 out of the total of 157 individuals sampled at the four main sites where vertical stratification was studied (Table 6). Regarding to the butterflies trapped at 15m and 30m, 31 (19.7%) and 20(12.7%) individuals were found respectively, with no significant difference between these two levels (Table 6).

Table 6: Numbers of individuals and species richness of fruit-feeding-butterflies trapped at three vertical height levels in Siberut, Indonesia, between June and July, 2009. Statistical test was done by using the Kruskal-Wallis. Different letters indicate significant differences based on Mann-Whitney U statistics. All test with p=0.05 as significance level.

Individuals

Species

1m

15m

30m

p-value

106 a

31 b

20 b

<0.001

Natural Forest

57

18

11

Disturbed Forest

49

13

9

Total

14

8

10

Natural Forest

7

5

6

Disturbed Forest

12

6

5

Total

The distribution of numbers of individuals per height level (Table 6) did not differ between natural and disturbed forest (χ2=0.352, df=2, p=0.839). Likewise, the distribution of the numbers of species did not differ significantly between height levels. Furthermore, there was no difference in vertical height distributions between natural and disturbed forest, although disturbed forest seemingly had an elevated number of species at 1m level compared to natural forest.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

21

16 14

Species

12 10 8 6

1m 15m 30m

4 2 0 0

5

10

15

20

25

30

35

Trap days Figure 6. Cumulative number of Nymphalidae species captured against trap days at the three different vertical heights along the vertical strata of two natural forest and disturbed forest sites in Siberut, Indonesia.

16 14

Species

12 10 8 6

1m 15m 30m

4 2 0 0

30

60

90

120

Individuals Figure 7. Cumulative number of Nymphalidae species against number of individuals captured at the three different vertical heights along the vertical strata of two natural forest and disturbed forest sites in Siberut, Indonesia.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

22

Figure 8. Multidimensional scaling of fruit-feeding butterfly communities at three different trap heights along the vertical strata of two natural forest (N) and two disturbed forest (D) sites. Scaling is based on Sorensen similarity values. Sites of similar trap heights are connected by lines.

Out of the 20 fruit-feeding butterfly species sampled in the study, 70% or 14 species were found at the lowest level. At the midstorey level, only 8 species were found, while 10 species were recorded at the canopy level (Table 6, Figure 6). A species accumulation curve indicated that species sampling at 1m level was fairly complete, reaching a plateau but that at 15m and especially at 30m level species were still added at the end of the study period (Figure 7).

Figure 8 shows results from the.multidimensional scaling of fruit-feeding butterfly communities at three different trap heights along the vertical strata of natural forest and disturbed forest sites. The fruit-feeding butterfly communities of both natural forest and disturbed forest along the stratum did not show a very clear distinction, Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

23

when examining results visually. Also, a MANOVA of the two-dimensional scores from multi-dimensional scaling showed no significant differences in species composition among the vertical stratum (F=2.03; df=4, 10; p=0.165) and between habitat type (F = 1.75; df = 2, 5; p00.265), based on the four vertical trapping sites.

4.4.1 Taxonomical differences among vertical height levels Of the five subfamilies, only Biblidinae and Satyrinae showed significantly different vertical distributions (Table 7; Kruskal-Wallis test, p<0.021 and <0.001, respectively). While most Bibilidinae (2 spp.) were trapped at midstorey levels, Satyrinae were mainly trapped at understorey level (Table 7).

Table 7. Individual numbers of Nymphalidae trapped regarding to subfamily in the three vertical level: 1m, 15m and 30m, in Pungut, Siberut, Inodnesia, between June and July, 2009. Statistical test was done using Kruskal-Wallis Test, which the p=0.05 is defined as significance level.

Biblidinae Charaxinae Morphinae Nymphalinae Satyrinae Total

Chung-Lim LUK

No. of spp. 2b 2 3 5 8a

1m 7 6 6 2 85

15m 17 a 1 3 3 7b

30m 5a 1 1 2 11 b

total 29 8 10 7 103

20

106

31

20

157

p-value 0.021 0.211 0.061 0.051 <0.001

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

24

100%

80%

Satyrinae

60%

Nymphalinae 40%

Morphinae Charaxinae

20%

Biblidinae

0% 1m (n = 106)

15m (n = 31)

30m (n = 20)

Figure 9. Proportional abundance of butterflies according to subfamily at three vertical levels: 1m, 15m and 40m, in Pungut, Siberut, Inodnesia.

None of the subfamilies had highest numbers of individuals at height 30m. For Charaxinae, Mophinae, and Satyrinae, highest number of individuals were found at 1m, which contributed to 75%, 60% and 82.5% respectively (Figure 9). Two of the subfamilies were found to be the most abundant at the mid-storey level: At this vertical level, Biblidinae showed 58% and Nymphalinae showed 42% of the total abundance as compared to the remaining three subfamilies.

At the lowest (understorey) level, Satyrinae were the dominant subfamily, contributing to 80.2% (n=85) of all fruit-feeding butterfly individuals sampled, with the remaining 20% represented by the remaining four sub-families (Figure 4). At the mid-storey level, Biblidinae were the most abundant contributing to 54.8% (n=17) of

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

25

the total 31 individuals sampled. Satyrinae were second in abundance (22.6%, n=7), followed by Morphinae and Nymphalinae (both 9.7%. n = 3). At the highest (canopy) level, Satyrinae and Biblidinae were the two most dominant subfamilies, with 55.0% (n = 11) and 25.0% (n = 5) of individuals (Figure 4).

4.4.2 Vertical stratification at species level For species with a total count higher than 10, Kruskal-Wallis test was done to test if there was any difference in abundance amongst the three vertical levels. Only two species showed significant results (Table 8). Dichorragia nesimachus showed significant higher abundance in 15m (n=17) and 30m (n=5) than in 1m (n=5) (p=0.009). Although at 1m and 30m the species had the same number of individuals, due to the reason of statistics and the butterfly sampled in different locality (i.e. different combination of sample trapped in different replicate, but both height level had five individuals in total), a significant higher result was found at 30m than at 1m.

Table 8 Species considered as indicator regarding to different vertical level. Statistical test was done by using Kruskal-Wallis test, with 0.05 as significant level. Only species with total abundance number > 9 were tested. Abundance and median are based on total butterfly sampled in 32

1m abundance median Dichorragia nesimachus Mycalesis maianeas Neorina lowii

range

15m abundance median

range

abundance

30m median

range

Total

pvalue

5b

1.5

1-2

17 a

2.5

2-10

5a

0

0-5

27

0.009

25 a 33 a

5.5 5

1-13 0-23

0b 5b

/ 1.5

/ 0-2

0b 0

/ /

/ /

25 38

0.005 <0.001

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

26

For Neorina lowii, a significantly higher abundance was found at 1m than at 15m (n=33 & n=5 respectively) (Mann-Whitney U test, p<0.001). At level 30m no individuals of this species were trapped. All twenty-five individuals of Mycalesis maianeas collected were found at level 1m, being significantly different from the other vertical levels (p=0.005, Table 8).

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

27

5 Discussion 5.1 General An overall capture rate of 0.45 fruit-feeding butterflies per day per trap was recorded in this study. This figure is relatively low when compared to other studies in tropical regions. For example Devries and Walla (2001) found an average number of baited nymphalids per day and per trap of 0.8 in floodplain forest in Ecuador. Hill et al. (2001) had nymphalid capture rates of 0.9 individuals per trap day in an unlogged forest plot in Sabah. A capture rate of 2.9 nymphalids was reported by Shahabuddin and Terborgh (1999), from a set of forested islands in Venezuela. A very high capture rate of 8.3 was reported by Fermon (2002), in secondary moist semi-deciduous forest of Côte d’Ivoire.

A total of 244 fruit-feeding butterflies belonging to five sub-families (Biblidinae, Charaxinae, Morphinae, Nymphalinae and Satyrinae) and 20 species were recorded during this study (Table 2). The species richness recorded was relatively low when compared to other vertical stratification studies on fruit feeding butterfly communities in SE Asia region: Sulawesi (Fermon et al. 2005: 43 species) and Borneo (Schulze et al. 2001: 53 species) in particular.

Out of the 20 species of Nymphalindae recorded in this studied, two were not previously found in the Mentawai Islands (Discophora spp & Tanaecia spp.). Further identification will be done to confirm the two unknown species. Tanaecia visandra and Elymnias nelsoni, were the two species sampled in the study area which are endemic to Mentawai. The 10% endemism (2 out of 20 species) recorded for the Nymphalidae family from this study was lower than expected since flora and fauna in

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

28

Siberut have been know to show a high degree of endemism (example: 50% of mammals are endemic, WWF1982). Due to Mentawai Island being originally separated from Sumatra, about 500,000 years ago, all non-endemic Nymphalidae butterfly individuals sampled in the study area are also found in the main island of Sumatra. The status of endemism of butterfly at species and subspecies level can be found in the references of Aoki et al. (1982), Tsukada et al. (1985) and Tsukada (1991).

5.2 Fruit-feeding butterfly response to disturbance Our vegetation analysis showed that tree densities in natural forest were significantly higher than in disturbed forest (1185.7 vs 628.6 Ind/ha). In contrast, basal area was not significantly different between natural and disturbed forest. However the figure of basal area in natural (total: 51.0 m²/ha, dbh ≥5cm ) forest is general higher than in disturbed area (35.6 m²/ha). The reason of non-significant result may due to the high level of variance (Standard derivation >36%) (Table 1). The figures were however consistent with previous vegetation surveys in Siberut (Hadi, in litt: 58.2m²/ha for dbh class >= 10cm, also from Peleonan forest; or the Paitan forest: 47m²/ha, dbh>=15cm, Whitten 1982) Sarawak (Mulu forest: 57m²/ha: Proctor, 1986). Whitten (1982) reports basal areas from peatswamp forests median = 23 m²/ha, dbh ≥15cm). The basal areas found in this study therefore might indicate that study sites are largely representative for conditions of natural and disturbed forests of Siberut.

This study did not reveal major significant differences in nymphalid butterfly abundance or species richness between natural forest and disturbed forest (Table 3).

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

29

Biblidinae and Charaxinae were the two subfamilies, which showed very significant higher abundance in natural forest than in disturbed forest (Table 3) while Satyrinae showed an opposite trend.

Only four species with individuals >7 were found having significant differences between habitat types. Dichorragia nesimachus, Mycalesis maianeas and Neorina lowii found significant higher abundance in natural forest. Mycalesis oresis found opposite trend as significant higher abundance was found in disturbed forest.

These results are coherent to the findings of Fermon (2005), whose study on Sulawesi can be considered as a template for this study, where the two different Mycalesis spp. ((M. horsfieldi & M. itys). showed significantly also showed higher abundances in natural forest. Dichorragia nesimachus was also recorded in Fermon (2005), showing a similar trend. There was no similar species or genera of Neorina lowii found in Sulawesi in previous scientific record.

Concerning diversity and species richness, similar results were found between habitats (Table 3). Findings on the effect of tropical forest disturbance on butterfly species richness and diversity among recent studies show mixed results. Butterflies in SE Asia for example, resulted in no differences (Hill et al 2001, in Borneo), higher (Hill et al 1995, in Buru island, Indoesnesia; Beck & Schulze 2000 in Borneo) and lower ( Hamer et al. 1997, in Sumba island, Indonesia; Willott et al. 2000, in Borneo, Fermon et al. 2005, Sulawesi) species diversity or/and richness in natural forest than in disturbed forest.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

30

The spatial scale maybe one of the factors influencing the result of effect of habitat disturbance on species richness, and this has been widely accepted by most Lepidopterists. The trend suggested that a smaller spatial scale research tended to result in a higher species diversity and richness found at disturbed sites, while at larger spatial scales research showed opposite trend (Koh 2007, Hamer & Hill 2000, Hill & Hamer 2004).

An example can be seen by combing the survey of Hamer & Hill (2000), and Hill & Hamer (2004), about the impact of land use change on the butterfly community sampled. Eleven of 12 studies conducted at spatial scales of < 1ha reported higher species diversity in disturbed sites, while 10 of 15 studies at larger scale of >3.1ha reported higher diversity in undisturbed sites. According to the definition of the Hamer & Hill (2000), this study is categorised as large scale, and did not consistent to the trend suggested.

Species richness and diversity is usually correlated with habitat diversity. We expected that in natural forest, more microhabitat types (e.g. canopy, understorey, gap) could be found, as disturbance produces a more homogeneous vegetation structure compared to undisturbed forest (Hill et al. 1995, Hamer et al. 2003). So a higher species diversity and richness should be found if the spatial sampling scale was large enough to cover the entire habitat. However, disturbance is also known to create opportunities and different niches for additional species, not found in undisturbed forest (Connell 1978), by opening up areas of closed-canopy dense forest as well as creating light gaps. So if spatial scales were small, a higher habitat heterogeneity

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

31

would be found in the sample area resulting in higher species diversity and richness observation in disturbed site.

Evidence showed that climate might also be a factor influencing the result of effect of habitat disturbance on species richness. Hamer et al. (2005) conducted a study of the impact of logging on butterfly community in Bornean forest, where a higher species diversity and richness was found in primary forest during the dry season, while an opposite trend was recorded during the wet season. The cause was attributed to the immature stages of the butterfly being affected by climatic variables. This suggested that a temporal scale might also be an important factor to be considered while conducting research on butterfly. The ecology of different taxonomic groups responding to climate and seasonal abiotic variables would affect the sampling result in ecological research.

5.3 Pattern of fruit-feeding butterfly stratification Only few similar studies on vertical stratification have been done in the past. In general this research was aimed at examining the vertical stratification of nymphalid butterflies by sampling and considering three different levels: ground, midstorey and canopy. The result suggested that in general, the abundance was the highest in the understorey level and significantly higher than at midstorey and canopy levels. In our study, we could not establish different abundances between midstorey and canopy levels. This is consistent to most similar study (Example: in Ecuadorian rainforestDeVries et al 1999, in North Borneo-Dumbrell & Hill 2005, in Sulawesi- Fermon et al. 2005, etc).

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

32

The multivariate analysis and MANOVA showed that a stratification of distinct communities cannot be confirmed (Figure 8). Concerning species diversity, understorey revealed the highest species numbers (14), while only 8 and 10 species were recorded at midstorey and canopy level respectively (Table 6, Figure 6 &7). From examining the species accumulation curve (Figure 6 & 7), it is very likely more species would be observed at the level of midstorey, and in particular overstorey level, if the sample size was increased.

The most widely accepted explanation for the relatively higher abundance of butterflies at understorey level is the availability of food resources like rotting fruits (Schulze et al. 2001). Fallen rotting fruit on the forest ground provide a better amount and diversity of favorable food resource to support a higher species diversity and abundance of fruit-feeding butterfly individuals.

Our results – in line with previous research - showed that midstorey and canopy might share relatively large number of similar fruit-feeding butterfly species. This is because at the canopy and midstorey level, availability of fruit at the vertical position is not stable hence affecting the niche width of the non-understorey fruit-feeder. The niche shared by the fruit-feeder is more likely to be expanded at vertical level. Since the food availability is scare, the fruit-feeder above ground level need to forage in different strata in order to get sufficient food (Schulze et al. 2001).

Fermon et al. (2005) studied vertical structure of fruit-feeing butterfly in Sulawesi, and found that species diversity was highest at midstorey level, suggesting that at this level distinct vertical communities overlap. Schulze et al .(2001) found a quite similar

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

33

result of species diversity and richness which understorey and midstorey had the highest value. Another factor was the difference in canopy height in this research study as compared to previous studies of Fermon et al (2005) and Schulze et al. (2001) were sampling was done at canopy level of 50m, while in this study sampling was done at 30 m. The higher upper canopy level of the two studies compared to this study also result in a larger spatial extent of what is considered “midstorey” and therefore in a bigger niche width, because of the much larger space and probability of rotten food available is higher. Moreover, the higher spatial ratio of midstorey to undstorey size results in a higher chance of more species to be observed.

In the study of Fermon et al. (2005), all six subfamilies recorded showed significant correlations between abundance and vertical height (Spearman rank correlation). The abundance of Charaxinae and Apaturinae was positively correlated with vertical height, while the other four subfamilies: Satyrinae, Morphinae, Biblidinae and Nymphalinae showed a negative relationship with vertical height. In our study, where we used Kruskal-Wallis to test the abundance of the five subfamilies sampled at three different vertical levels. Charaxinae, Morphinae and Nymphalinae showed no significant difference among vertical levels but Satyrinae were found being most abundant at the lowest level, coherent to Fermon (2005) but Biblidinae showed least abundance at the lowest level in contrast to the findings of Fermon (2005). However it should be note that the Biblidinae in our study only consist of two species, out of which one species Dichorragia nesimachus represents 87% of the whole subfamily (details see Table 1), while the trapped species of Biblidinae in Fermon (2005) consists of twelve species. In contrast to the results of Schuze et al. (2001) on the vertical pattern of fruit-feeding butterfly in Bornean forest, the abundance and

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

34

diversity subfamily Satyrinae, Morphinae and Nymphalinae showed a negative relationship with vertical height, while Charaxinae did not show a significant difference in abundance among vertical levels.

Only two species showed significant differences in abundance among three vertical levels. Dichorragia nesimachus was in general found in mid-storey to canopy level, the selection of the vertical strata of the species was being very similar to the study of Fermon (2005). Although no statistical difference in abundance could be found in the Mycalesis maianeas among vertical levels, since it could be only sampled in the understorey level with 25 individuals and statistical test cannot be performed. This species could still be considered as an understorey species as this species showed a similar trend in the studies of Fermon (2005). In previous studies, Charaxes spp. were frequently found at canopy level, and were thus defined as canopy species (e.g. Fermon 2002, Fermon et al. 2005, Schulze et al.2001). In the present study, only two individuals were sampled, one found in each midstorey and canopy level, suggesting Siberut’s Charaxinae having a behaviour similar to previous studies.

Disturbance is known to modify or influence the vertical stratification of butterfly communities in tropical forests (e.g. DeVries 1988; Schulze et al . 2001; Fermon et al . 2005). Previous studies showed that the canopy species of natural forest maybe present in the understorey of disturbed forest. The main suggested reason is that canopy species are known to treat light gaps and forest edges as canopy (DeVeries 1988). The light gaps present in disturbed sites causing the light preferring canopy species will be found in the lower stratum (Dichorragia nesimachus in this study showed this pattern). When considering studies which sampled butterflies only at

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

35

understorey level, species richness or diversity in disturbed forest may have been inflated to give a bias result. True canopy species in natural forest hence cannot be sampled, but might be present in the understorey level of disturbed forest, resulting in (biased) higher species diversity and richness in disturbed forest. In order to study butterfly communities, a 3-dimentional sampling design should be considered to give a better description of the community structure.

Fermon (2002) suggested that any species community should reach its minimum richness and abundance at canopy, which means that more rare species should be found at canopy level. In the study of Schulze (1995), the result suggested a similar trend of rare species of fruit-feeding butterfly with respect to increasing stratum height, but the study also found that nectar-feeding species had an inverse trend with increasing number of rare species at lower stratum. We could come to a tentative conclusion that in contrast to nectar-feeing butterflies, fruit –feeding butterfly species have a lower pressure of resource limitation at lower strata.

5.4 Fruit-feeding butterflies as indicator The result suggested that Dichorragia nesimachus was the best species to be used as indicator species since it was responding most prominently to habitat change in our study area. It was both a common species to be found and was highly sensitive to disturbance. Concerning vertical stratification, the species was more likely to be found in midstorey-canopy level in natural forest, but absent in canopy in disturbed forest suggesting that it might be sensitive to light availability. Mycalesis maianeas, Mycalesis oresis, and Neorina lowii were also good species in indicating disturbance since the abundance of these significantly changed between habitats, and were all

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

36

common species in the region. Mycalesis maianeas and Neorina lowii found higher abundance in natural forest, while Mycalesis oresis showed an opposite trend.

Although some butterfly species show clear response to the change of habitat (e.g. Charaxes spp. in Schulze et al 2001 and in Fermon et al. 2005), the role of butterflies as indicators of habitat change needs much detailed study. Lawton et al. (1998) investigated species richness patterns across a tropical land-use gradient in Cameroon. The habitats investigated ranged from highly modified grasslands to relatively intact closed-canopy forest. The results showed that trends in species richness along the disturbance gradient varied widely for different taxonomic groups, suggesting that conservation monitoring and assessment based on indicator taxa may not be useful.

Simply counting the number of species provides no information on the composition of the species sampled. If species richness alone is used as a measure of biological conservation value, it may be misleading higher abundance or/and species richness might be found in disturbances site, as it may favour wider dispersal of species (Akite 2008). The focus on well-defined assemblages rather than on selected species captures, or species richness, gives more information, and more likely to reflect the process of impacts of fragmentation on ecosystem structure and function (Kitching, 1994)

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

37

6 Conclusion The forest of Siberut is now facing the threat of commercial logging and agricultural farming. Studies on the impact of these anthropological changes on biodiversity are of urgent needed to access conservation concern. In our study, we compared the fruit feeding butterfly assemblages in natural undisturbed forest, and disturbed forest modified by agricultural farming. Overall species abundance and richness showed no significant differences between natural and disturbed forest. Dichorragia nesimachus, Mycalesis maianeas, Mycalesis oresis and Neorina lowii were the species which are sensitive to the change of forest type. Concerning vertical stratification, the understorey showed significantly higher abundance than the midstorey and canopy level. The species richness and diversity is the highest at the understorey level but the trend of the indices at the midstorey level and especially canopy level seem to be still rising. The vertical stratification of species composition in natural and disturbed forest was not clear, however at family level two (Biblidinae & Satyrinae) of the five families showed significant differences in abundance among vertical heights. Satyrinae were dominant at the understorey level, and compose more than 60% of total number of fruit-feeding butterfly sampled and might explain the non-significant result of species composition in natural and disturbed forest.

Butterflies are one of the most ideal taxonomic groups for investigation of habitat change. The lack of consensus among studies makes it impossible to conclude to what extent land use change affects butterfly species richness or diversity. The use of species richness, diversity and species diversity as valuing disturbance impact needs in-depth studies. Future community level studies should be designed carefully, so as

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

38

to be able to examine the true responses of the assemblage of butterfly communities to anthropological disturbance of habitat.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

39

7 References Achard, F., Eva, H.D., Stibig, H.J., Mayaux, P., Gallego, J., Richards, T. & Malingreau, J.P. (2002) Determination of deforestation rates of the world’s humid tropical forests. Science, 297, 999–1002. Aoki, T., Yamaguchi, S. & Uemura, Y. 1982. Satyridae, Libytheidae. In: Tsukada E. (ed) Butterflies of the South East Asian Islands 3. Plapac, Tokyo, Japan, pp. 1–500. Barlow, J., Overal, W.L., Araujo, I.S., Gardner, T.A. & Peres C.A. (2007) The value of primary, secondary and plantation forests for fruit-feeing butterflies in the Brazilian Amazon. Journal of Applied Ecology, 44, 1001-1002 Beck, J. & Schulze, C.H. (2000) Diversity of fruit-feeding butterflies (Nymphalidae) along a gradient of tropical rainforest succession in Borneo with some remarks on the Brook, B.W., Sodhi, N.S. & Ng, P.K.L. (2003) Catastrophic extinctions follow deforestation in Singapore. Nature, 424, 420–423. Brooks, T.M., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Rylands, A.B., Konstant, W.R., Flick, P., Pilgrim, J., Oldfield, S., Magin, G. & Hilton-Taylor, C. (2002) Habitat loss and extinction in the hotspots of biodiversity. Conservation Biology, 16, 909–923. Connell, J.H. (1978) Diversity in tropical rainforests and coral reefs. Science, 199, 1302–1310. Corbet, A.S. & Pendlebury, H.M. (1992) The Butterflies of the Malay Peninsula. Fourth Edition Revised. United Selangor Press Sendirian Berhard, Kuala Lumpur, Malaysia. DeVries, P.J., Walla, T.R. & Greeney, H.F. (1999). Species diversity in spatial and temporal dimensions of fruit-feeding butterflies from two Ecuadorian rainforests. Biological Journal of the Linnean Society 68, 333–353.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

40

DeVries, PJ, et al., (1997) Species diversity in vertical, horizontal, and temporal dimensions of a fruit-feeding butterfly community in an Ecuadorian rainforest. Biological Journal of the Linnean Society. 62, 343-364 DeVries, P.J. & Walla, T.R. (2001) Species diversity and community structure in neotropical fruit-feeding butterflies. Biological Journal of the Linnean Society, 74, 1– 15. Dumbrell, A.J. & Hill, J.K. (2005) Impacts of selective logging on canopy and ground assemblages of tropical forest butterflies: implications for sampling. Biological Conservation, 125, 123–131. DeVries, P.J. (1988) Stratification of fruit-feeding nymphalid butterflies in a Costa Rican rainforest. Journal of Research on the Lepidoptera, 26, 98–108. Fermon H., Waltert, M., Wright, R.I. & Muhlenberg M. (2005) Forest use and vertical stratification in fruit-feeding butterflies of Sulawesi, Indonesia: impacts for conservation . Biodiversity and Conservation. 14, 333–350. Fermon, H. (2002). The butterfly community of a managed West African rainforest: patterns of habitat specificity, diversity, stratification and movement. patterns of habitat specificity, diversity, stratification and movement. Ph.D. Thesis, University of Goettingen, Germany. Fermon, H., Waltert, M., Vane-Wright, R.I. & Mühlenberg, M. (2005) Forest use and vertical stratification in fruitfeeding butterflies of Sulawesi, Indonesia: impacts for conservation. Biodiversity and Conservation, 12, 1–18. Fuentes, A. (1996/7) Current status and future viability for the Mentawai primates. Primate Conservation 17, 111-116 Hadi, S.,Ziegler, T., Walter, M. & Hodges J.K. (2009) Tree diversity and forest structure in northern Siberut, Mentawai islands, Indonesia. Tropical Ecology 50(2), 315-327 Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

41

Hagen, B. (1893) Schmetterlinge von den Mentawej-inseln. Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft 20, 309-340, 2pls Hamer, K.C. & Hill, J.K. (2000) Scale-dependent effects of habitat disturbance on pecies richness in tropical forests. Conservation Biology, 14, 1435–1440. Hamer, K.C., Hill, J.K., Benedick, S., Mustaffa, N., Sherratt, T.N., Maryati, M. & Chey, V.K. (2003) Ecology of butterflies in natural and selectively logged forests of northern Borneo: the importance of habitat heterogeneity. Journal of Applied Ecology, 40, 150–162. Hamer, K.C., Hill, J.K., Lace, L.A. & Langan, A.M. (1997) Ecological and biogeographical effects of forests disturbance on tropical butterflies of Sumba, Indonesia. Journal of Biogeography, 24, 67–75. Hamer, K.C., Hill, J.K., Mustaffa, N., Benedick, S., Sherratt, T.N., Chey, V.K. & Maryati, M. (2005) Temporal variation in abundance and diversity of butterflies in Bornean rain forests: opposite impacts of logging recorded in different seasons. Journal of Tropical Ecology, 21, 417–425. Hill, J.K. & Hamer, K.C. (2004) Determining impacts of habitat modification on diversity of tropical forest fauna: the importance of spatial scale. Journal of Applied Ecology, 41, 744–754. Hill, J.K., Hamer, K.C., Lace, L.A. & Banham, W.M.T. (1995) Effects of selective logging on tropical forest butterflies on Buru, Indonesia. Journal of Applied Ecology, 32, 754–760. Hill, J.K., Hamer, K.C., Tangah, J. & Dawood, M. (2001) Ecology of tropical butterflies in rainforest gaps. Oecologia, 128, 294–302. Kitching, R.L. (1994) Biodiversity political responsibilities and agendas for research and conservation. Pacific Conservation Biology, 1, 279–283. Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

42

Koh L.P. (2007) Impacts of land use change on South-east Asian forest butterflies: a review. Journal of Applied Ecology, 44, 703–713 Kremen C. (1992). Assessing the indicator properties of species assemblages for natural areas monitoring. Ecological Applications, 2, 203–217. Kremen C. (1994). Biological inventory using target taxa: a case study of the butterflies of Madagascar. Ecological Applications, 4, 407–422. Lawton, J.H., Bignell, D.E., Bolton, B., Bloemers, G.F., Eggleton, P., Hammond, P.M., Hodda, M., Holt, R.D., Larsen, T.B., Mawdsley, N.A., Stork, N.E., Srivastava, D.S. & Watt, A.D. (1998) Biodiversity inventories, indicator taxa and effects of habitat modification in tropical forest. Nature, 391, 72–76. Myers N, Mittermeier R.A., Mittermeier, C.G., da Fonseca GAB, Kent J. (2000). Biodiversity hotspots for conservation priorities. Nature, 403, 853–858. Persoon, G.A., 2003: Conflicts over trees and waves on Siberut Island. Geografiska Annaler., 85 B (4), 253–264. Quinten, M. (2008). Survey of the Primate Community of Peat Swamp Forests on Siberut, Mentawai Islands (Indonesia). Master Thesis, University of Goettingen, Germany. Rydon A.H.B. (1964). Notes on the use of butterfly traps in East Africa. Journal of the Lepidopterists’ Society 18, 51–58. Schulze, C. H. (1995) Die Tagfalterzönose eines südostasiatischen Regenwaldes. Muster und Mechanismen der Stratifikation. Diploma thesis, University of Würzburg, Germany.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

43

Schulze, C. H. (1995) Die Tagfalterzönose eines südostasiatischen Regenwaldes. Muster und Mechanismen der Stratifikation. Diploma thesis, University of Würzburg, Germany. Shahabuddin, G., Terborgh, J. W. (1999) Frugivorous butterflies in Venezuelan forest fragments: abundance, diversity and effects of isolation. Journal of Tropical Ecology 15, 703-722. Tsukada E. (1991). Nymphalidae (2). In: Tsukada E. (ed) Butterflies of the South East Asian Islands 5. Plapac, Tokyo, Japan, pp. 1–576. Tsukada E., Nishiyama Y. & Kaneko M. (1985). Nymphalidae (1). In: Tsukada E. (ed) Butterflies of the South East Asian Islands 4. Plapac, Tokyo, Japan, pp. 1–558. Verstappen, H.T. (1975). On palaeoclimates and landform development in Malaysia. pp. 3-36. In: G.J. Bartstra & W. A. Casparie (eds.) Modern Quaternary Research in Southeast Asia. Rotterdam: Balkema. Whitten, T. (1982a). The Gibbons of Siberut. Dent & Sons Ltd., London and Toronto. Whitten, A.J. (1982b). A numerical analysis of tropical rain forest, using floristic and structural data, and its application to an analysis of gibbon ranging behaviour. The Journal of Ecology 70, 249-271. Willott, S.J., Lim, D.C., Compton, S.G. & Sutton, S.L. (2000) Effects of selective logging on the butterflies of a Bornean rainforest. Conservation Biology, 14, 1055– 1065. WWF (1982). Siberut Nature Conservation Area, West Sumatra, Managemant Plan 1983-1988. A WWF report for the directorate of nature conservation, Bogor.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

44

Appendix 1 Distribution information and wing size of sampled butterfly species

Distribution information (based on Aoki et al. 1982, Tsukada et al. 1985 and Tsukada 1991) and wing size of sampled butterfly species.

family Satyridae Nymphalindae Nymphalindae Nymphalindae

distribution 1, 2, 3, 5 1, 2, 3, 4, 5, 6, 1, 2, 3, 4, 5, 6, 7 1, 2, 3, 5, 6, 8 1, 2, 3, 4, 5, 6, 8, 10 / 1 1, 2, 3, 5, 6, 8

remarks on distribution also found in Philipines (Samar, Leyte, Panon, Mindanao & Negros), Malaya to Burma

Amathuxidia amythaon Bassarona dunya also found in Ruteng (Indonesia), Thailand and Myanmar Bassarona teuta also found in W. & S. China, Myanmar, Nepal Charaxes bernardus Dichorragia Nymphalindae also found in S. C. W. China, Japan S. Koera, Taiwan and Myanmar nesimachus Satyridae Discophora sp. Satyridae Elymnias nelsoni Satyridae Elymnias panthera Satyridae Elymnias sp 1 Nymphalindae 1, 2, 3, 4, 5, 6, 8 also found in Indo-China region Lebadea martha Nymphalindae 1, 2, 3, 4, 5, 6 also found in Myanmar, Thailand Lexias dirtea Satyridae 1, 2, 3, 5, 6, 10 also found in from W. Malay to India, Sri Lanka, Formosa, Japan, and S., C. & W. China Melanitis phedima Satyridae 1,2,3,4,5,6,7, 9 also found in S. India, and from W. Malay to S. Burma Melanitis zitenius Satyridae 1, 2, 3, 5 Mycalesis maianeas Satyridae 1, 2, 3, 5 also found from Malaya up to Burma Mycalesis oresis Satyridae 1, 2, 3, 4, 5 Neorina lowii Nymphalindae 1, 2, 3, 4, 5, 6, 8 also found in Ruteng (Indonesia) Indo-China region and Myanmar Prothoe franck Nymphalindae / Tanaecia spp. Nymphalindae 1 Tanaecia visandra Satyridae 1, 2, 3, 5 Zeuxidia amethystus Where 1=Mentawai, 2=Sumatra, 3=Borneo, 4=Palawan, 5=W.Malay, 6=Java, 7=Lombok, 8=Bali, 9=Sumbawa, 10=Sulawesi

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

wing size (mm) 43 29.5 23.5 29 30.5 25 23.5 24.5 26,5 17.5 28 28 31 13 14 39 26 32 27 37

45

Appendix 2 Timetable of fruit-feeding butterfly sampling The date of furit-feeding butterfly sampling of vertical traps (3 traps x 4 locations, 32 days), and undstorey traps (5 traps x 10 locations, 16 days) in research site of Siberut, Indonesia. Each day represents 4 trap location of vertical trap and 5 trap locations of understorey trap sampled. Traps were checked every 24 hours. Month Day

Days per month

Chung-Lim LUK

June July 6 1 7 2 17 3 18 4 19 5 20 6 21 7 22 8 23 9 24 10 25 11 26 12 27 13 29 14 30 15 16 17 15 17

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

46

Appendix 3 Photographic section

Dichorragia nesimachus

Mycalesis maianeas

Chung-Lim LUK

Neorina lowii

Charaxes bernardus

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

47

The local guide with vertical tap

Vertical trap at mid-storey level

Vertical trap at mid-storey level Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

48

Declaration Hiermit versichere ich gemäß § 9 Abs. 5 der Prüfungsordnung für den integrierten binationalen Master-Studiengang Internationaler Naturschutz (engl.: International Nature Conservation) vom 16.08.2006, dass ich die vorliegende Arbeit selbstständig verfasst und keine anderen als die angegebenen Hilfsmittel verwendet habe. Diese Arbeit wurde nicht in der gleichen oder einer ähnlichen Form bereits einem anderen Prüfungsausschuss vorgelegt und wurde bisher noch nicht veröffentlicht. Hereby I affirm – according to § 9 section 5 of the examination regulations for the integrated bi-national Master programme International Nature Conservation (deutsch: Internationaler Natur-schutz) from 16.08.2006 – that I have penned the present thesis autonomously and that I did not use any other resources than those specified above. This work was not submitted previously in same or similar form to another examination committee and was not yet published.

Chung-Lim LUK

Forest use and vertical stratification in fruit-feeding butterflies of Siberut, Indonesia

49