Local Uses of Tree Species and Contribution of Mixed Tree Gardens to Livelihoods of Saleman Village Near the Manusela National Park, Seram Island, Maluku (Indonesia)

Local uses of tree species and contribution of mixed tree gardens to livelihoods of Saleman village near the Manusela National Park, Seram Island, Maluku (Indonesia)

Master’s thesis Ariane Cosiaux Master “Biodiversité Végétale Tropicale” 2012

Ariane Cosiaux 2012 Master’s internship from 7 March to 17 August 2012 at : Center for International Forestry Research (CIFOR) Jalan CIFOR, Situ Gede Bogor Barat 16115 Indonesia

Supervisor : Yves Laumonier CIRAD scientist in the research unit « Tropical Forest Goods and Ecosystem Services »

Aknowledgments

I first want to thank Yves Laumonier for allowing me to do this internship, for his availability and advices. And I want to thank him for allowing me to discover the amazing Island of Seram.

I would like to thank also Pak Purwanto from LIPI for his kindness and his invaluable advices for the ethnobotanical part of my study. I am grateful to him for his help to the realization of my questionnaire in bahasa Indonesia and his encouragements during the first two weeks of my field work.

I am so grateful for the hospitality of the Saleman village. First I want to thank Nenek In and Hapipa for hosting me during my field work. Then I want to thank my field assistant Jul who help me and take care of me. A special thank goes to Bapak Hadir who taught me the name of plants in bahasa Saleman, for his kindness and availability. I also would like to thank Bapak Mustafa and Mama Ratna who also take care of me and help me for collecting voucher specimens. Finally, I like to thank Andre, Yaren and all the member of Nenek In family.

At CIFOR, I want to thank the other member of the Colupsia project, Pak Wyono and Pak Danan for their help and thanks to Popi Astriani who took care of all the administrative documents.

I am also grateful to Pak Yan for drying my voucher specimens and to Pak Ismael, Bogor Herbarium, for all species identification.

A final thank to Edith, Anis, Caroline, Marine, Anne Sophie, Thibaut, Nicolas, Jeffrey, Suzanne (my friends in Bogor) and my family for their support.

Contents

Introduction ...... 1 A. General context ...... 1 1) Deforestation and forest degradation in Indonesia ...... 1 2) Importance of human-modified landscapes ...... 1 B. The Colupsia project ...... 2 1) Partners ...... 2 2) A study area in Maluku ...... 2 3) Challenges ...... 3 C. My study within the Colupsia project ...... 3 1) Complex agroforestry systems in Indonesia ...... 3 2) Aim of the study ...... 6 Site and Methods ...... 6 A. Study area ...... 6 1) Island of Seram and the regency of Central Moluccas ...... 6 2) The village of Saleman ...... 7 B. Participatory mapping and identification of local practices ...... 9 1) Participatory mapping ...... 9 2) Management of mixed tree gardens: identify local practices ...... 10 C. Typology of mixed tree gardens ...... 10 1) Sampling design ...... 10 2) Recorded variables ...... 10 3) Data analysis ...... 11 D. Ethnobotanical knowledge ...... 12 1) Interviews on local uses of tree species in mixed tree gardens ...... 12 2) Free listing ...... 13 3) Plant voucher specimens ...... 15 Results ...... 15 A. Customary landscape, the petuanan of Saleman ...... 15 1) Customary governance of landscape management ...... 15 2) Participatory mapping and folk classification of landscapes ...... 17 3) Mixed tree gardens: general characteristics and local management of the Kebun besar ...... 19 B. Structure and floristic composition of mixed tree gardens...... 19 1) Structure ...... 19 2) Floristic composition ...... 24 C. Local uses of tree and other plant species ...... 26 1) Local uses of mixed tree garden species ...... 26 2) The other plants used by communities ...... 27 3) Important species for each use ...... 28 Discussion ...... 30 A. Mixed tree gardens: an integrated land use ...... 30 B. Mixed tree gardens: complex agroforestry systems similar to natural forest ...... 32 C. Importance of mixed tree gardens for local population ...... 32 D. The mixed tree gardens inside the Manusela National park ...... 33 Conclusion ...... 34

Introduction

A. General context 1) Deforestation and forest degradation in Indonesia Currently, there is little doubt that tropical ecosystems are exceptionally rich and contain much of the terrestrial biodiversity. However, the rapid and extensive forest degradation, involving modification of ecosystems and fragmentation of habitats lead to an alarming loss of biodiversity (Laurence 1999). Most of the 25 “biodiversity hotspots” defined by Myers et al. (2000) are located in the tropics and characterized by high level of endemism and habitats loss. Indonesia overlaps with two of these: the Sundaland (Western Indonesia) and the Wallacea (Eastern Indonesia). As part of the South-Est Asia, the environmental degradation was severe during the recent decades in Indonesia (Sodhi et al. 2004). From 1990 to 2005, Indonesia has lost 21,32 Mha of forest (17,56% of the forest cover), with a mean rate of deforestation on the period 1990-2000 (1,78 Mha/year) three times bigger than on the 2000-2005 period (0,58 Mha/year) (Hansen et al. 2009). Even if the deforestation rates have decreased, the loss of forest still high with more than 500000ha/year lost on the period 2005- 2010 (FAO, 2010). These high rates of deforestation are due to conversion of forest in agricultural lands, commercial logging, fire and mining (Sodhi et al. 2004).

2) Importance of human modified landscapes Facing this severe forest loss, one challenge for the conservationists is the creation of new protected areas (Sodhi et al. 2004). However, the Indonesian protected areas (24Mha) varies widely in their effectiveness (Curran et al. 2004), and may be the source of dispute with local populations (Aumeeruddy 2003, Hariyadi et al. 2012) which are seen as responsible of the forest degradation by governments in tropical countries. In many cases, when protected areas are created, local populations are not allowed to carry on natural resource extraction or agricultural activities leading to conflicts with the local authorities. Moreover, in the tropical regions, 70% of the land is pastures, agriculture or a mixture of managed landscapes (Perfecto and Vandermeer 2008) and the remnant forests cannot all be put under protection status. Indeed, the challenge is finding approaches which can reduce the decrease of forest cover, assure livelihoods for rural communities and conserve biodiversity outside protected areas (Bhagwat et al. 2008). Complex agroforestry systems is a type of land use which can provides solutions to forest degradation, rural livelihoods and to the loss of biodiversity in the tropics (Michon and De Foresta 1995, Swallow et al. 2006, Perfecto and

Vandermeer 2008, Harvey et al. 2008). This kind of land use is widespread in the Indonesian rural landscape.

B. The Colupsia project 1) Partners In this context of forest degradation an conversion in other land uses, the Center for International Research on Agronomy and Development (CIRAD), together with its partner the Center for International Forestry Research (CIFOR) are conducting a four year project called “Collaborative land use planning and sustainable institutional arrangements for strengthening land tenure, forest and community rights in Indonesia” (Colupsia). The project is financing by the European Commission and it is leading in partnership with several Non Governmental Organizations and local universities in Indonesia. The project works in two regions of Indonesia: Kalimantan and Maluku.

2) A study area in Maluku In Maluku, the pilot site is on the island of Seram. The island is part of the Wallacea biogeographic region and harbors a large diversity of vegetation types, from coastal mangrove to tropical alpine vegetation (Edwards et al. 1993, Monk et al. 1997). In the central part of the island lies the Manusela National Park, a conservation area created in 1997, covering an area of 189 000ha which represents about 10% of the total island area. The island still largely forested even if there was commercial logging activities in the 80’s and some estate plantations. Historically, Seramese people were practicing subsistence activities like hunting- gathering and swidden agriculture or more intensive forms of permanent agriculture on the coast (Ellen 1997). Like in Papua and all the Moluccas, a common characteristic of Seramese people is their dependency to the sago palms (Metroxylon sagu) that constitutes staple food. Sago starch is extracted both from swamp forest area and from gardens closer to settlement where it is planted. This habit reduces considerably the need to cut forests since highland rice and subsequent forest opening is not needed (Ellen 1993, 1999, Sasaoka and Laumonier 2011). In addition, the planting of “spice” trees like nutmeg (Myristica fragrans) and clove (Syzygium aromaticum) occurred very early in the Moluccan history because of the European demand for spices since the sixteenth century (Ellen 1997). Several agricultural practices still occur today in Seram: shifting1 and permanent cultivation. The swidden cultivation consists in

1 Shifting cultivation encompasses swidden cultivation and slash and burn cultivation. The both starts by slashing and burning a patch of forest. 2 opening dry fields for a short period before leaving it as fallow for a long period, integrated into a rotational system. People usually plant yams (Dioscorea), taro (Colocasia, Alocasia) or introduced species like cassava (Manihot esculenta) and sweet potato (Ipomea batatas). In contrast, slash and burn cultivation is practiced by pioneer farmers, who open new fields every year without rotational system, cutting primary forest for plant cash crops. The permanent cultivation of Seram Island is described as complex agroforestry systems including home gardens and mixed tree gardens (Monk et al. 1997).

3) Challenges For the collaborative land use planning process, it is important to identify and characterize the different land uses and understand their importance for the local communities. Indeed, it’s necessary to take into account the perception of the local communities toward forested landscape and natural resources which enables a better understanding of local priorities, local challenges according to rural livelihoods, management of the forested landscape and biodiversity conservation.

C. My study within the Colupsia project 1) Complex agroforestry systems in Indonesia Definitions and examples Complex agroforestry systems are tree-crop based systems with a forest-like structure (Michon and De Foresta 1995, Torquebiau 2007). Such systems account a high number of component (trees, lianas, herbaceous) and present a multistoried structure close to those observed in primary or secondary forests (Michon and De Foresta, 1995, 1995). In Indonesia, one type of complex agroforestry system is the homegarden called Pekarangan in Java (Wiersum 1982, Christanty et al. 1986). Homegardens are fenced-in gardens surrounding individual houses, in which several fruits and others trees are planted together with vegetable herbs and annual crops (Kumar and Nair 2004). They are generally intensively tended. However most of complex agroforestry systems are smallholder mixed tree plantations or “agroforests” on lands outside the village and less intensively managed than homegardens. These systems are complex mixed tree plantations with high species diversity, tree density and a complex vertical structure. In the literature they are also designed as “mixed tree gardens” (Michon et al. 1986) or sometimes “forest-gardens” (Wiersum 1982, Marjokorpi and Ruokolainen 2003, Wiersum 2004, Belcher et al. 2005). When the term agroforest is employed, it implies the existence of a large block of “forest” composed by a mosaic of

3 similar agroforestry plots. Then, forest gardens is employed as synonym of agroforest but it often corresponds to tree based systems with a higher proportion of wild trees (Wiersum 1997a, 1997b, 2004). As example we can cite the multistoried agroforestry gardens in West Sumatra known as parak characterized by a complex association of fruit trees like durian (Durio zibethinus), species for commercial purposes (Myristica fragrans, Cinnamomum burmani) and others useful trees (Michon et al. 1986). In the southeast Sumatra, the “jungle rubber” constitutes also complex agroforestry systems with high species diversity (268 plant species in a 1000 m² plot) and a structure similar to secondary forest (Gouyon et al. 1993). In West Java, bamboo- tree gardens with high species diversity have been studied by Okubo et al. (2010). According to Michon and De Foresta (1995), the villages in the island of Ambon are surrounded by agroforests which combine nutmeg and clove trees with fruit trees and forest nut trees. On the Saparua Island, there are examples of complex agroforestry systems referred as dusun by Kaya et al. (2002). The major components of these agroforestry systems are clove, nutmeg and coconut palms (Cocos nucifera) with a cortege of other tree species. In the North of Seram (Masihulan village), complex agroforestry systems such as durian, coffee or cocoa (Theobroma cacao) gardens have been studied by Vallet (2011).

Local management These tree based systems are “reconstructed forests” by human. The establishment of a new agroforestry plot starts by a total removal of patch of primary or secondary forest (Gouyon et al. 1993, Foresta and Michon 1996, Michon and De Foresta 1997). Sometimes wild trees are maintained on the plot to make shade or supply wood. These mixed tree gardens are generally established after a phase of slash and burn agriculture based on food crops and thus, closely associated with sifting cultivation (De Foresta and Michon 1996, Michon and De Foresta 1997, Kaya et al. 2002). The management of the agroforestry systems by farmers presents an intermediate level of input (clearing, cutting, and planting) and is complemented by the utilization of natural processes to assure soil fertility, tree regeneration or control weeds (Michon et al. 1986, Belcher et al. 2005).

Contribution to rural livelihoods The first purpose of most of mixed tree gardens is supplying cash incomes to rural households. In all of them, a commercially valuable main tree crop is planted to assure the main or only source of cash income to farmers (De Foresta and Michon 1996, Belcher et al.

2005). In the mixed tree gardens of West Sumatra, cinnamon (Cinnamomum burmani), nutmeg and coffee are planted for commercial purpose (Michon et al. 1986). In the South-Est of Sumatra, “jungle rubber” assures the major source of income of producers, and 70% of the rubber latex exported from Indonesia is produced by smallholders in rubber gardens (Gouyon et al. 1993, Belcher et al. 2005). In bamboo-tree gardens of Java, clove and coffee assure the same economic function. In Maluku, nutmeg and clove also assure cash income for local communities (Kaya et al. 2002) like in forest gardens of Sulawesi where cocoa, vanilla (Vanilla planifolia) and nutmeg are the main cash crops. Moreover, supplementary incomes are generated when farmers sell fruits or other commercially valuable products (Wiersum 1982, Gouyon et al. 1993, Murniati et al. 2001) in the local market. Finally, the tree gardens systems provide lots of subsistence products like firewood, house construction wood, fruits, vegetables or medicines. In bamboo tree gardens of Sumatra, 8 categories of uses and 42 tree species have been identified which assure the daily needs of farmers (Okubo et al. 2010). In Sumatra, durian gardens provide construction wood, fruits and vegetable (Michon et al. 1986). In traditional gardens of Central Sulawesi, Brodbeck et al. (2003) have inventoried a total of 43 to 95 species with subsistence uses. In the traditional tree farming systems dudukuhan in West Java, tree species assure the household subsistence needs in supplying food, medicines, construction wood, and firewood (Manurang et al. 2004).

Contribution in biodiversity conservation As multistoried tree base systems which mimic natural forests with a high species diversity, agroforestry systems can contribute to biodiversity conservation. First, tree gardens can conserve the “planned biodiversity”2 (Swallow et al. 2006, Perfecto and Vandermeer 2008) like the indigenous fruit tree species. Secondly, tree gardens may conserve the “unplanned associated biodiversity”3 which is found in complex agroforestry systems. Indeed, many wild spontaneous tree species are found in tree gardens and are maintained and managed by farmers (Michon et al. 1986, Gouyon et al. 1993, Kaya et al. 2002). Moreover, lots of shrub, lianas or epiphytes and herbs are found in complex tree based systems like in the rubber jungle of Sumatra where 268 plants species in which 91 tree, 27 shrubs, 97 lianas and 28 epiphyte species have been inventoried (Gouyon et al. 1993, Michon and De Foresta 1995). Then, many of these systems are viewed to be important for the conservation of wild

2 The planned biodiversity refers to annual and perennial crops purposefully introduced and maintained in an agro-ecosystem, by the farmer. 3 Unplanned associated biodiversity includes all flora and fauna species that exist in, or colonize the agroecosystem (FAO). 5 fauna in supplying a diversity of habitat (Swallow et al. 2006, Bhagwat et al. 2008) and facilitating the movement of mammals, birds of butterflies between patches of natural forest. Finally, complex agroforestry systems is an opportunity to reduce the pressure on the protected areas in forming forested transition zone between protected forests and human dominated areas (Murniati et al. 2001).

2) Aim of the study In this study we investigate the importance of mixed tree gardens for the Saleman community, a village near the Manusela National Park, on Seram Island, Maluku. The study is divided in three key points. The first one is the understanding of the place occupied by the mixed tree gardens in the village territory. The second one consist of describing the structure and tree species composition of the mixed tree gardens and the third one is assess the importance of mixed tree gardens as regards to livelihoods through the inventory of useful plant species and products supplied by these agroforestry systems.

Site and Methods A. Study area 1) Island of Seram and the regency of Central Moluccas Seram, the second largest island in the in eastern Indonesia (Figure 1), covers 17 429 km², being about 340 km long and 60 km wide. The island of Seram is part of the Australasian continent, but since its emergence about 3 to 5 Ma, it has never been connected to New Guinea or Australia by a land bridge (Audley-Charles 1993). It is speculated that it has been colonized by plants and animals by “island hopping” through different roads. Its floristic and faunistic composition includes species that have arrived from both Asia and New Guinea-Australia (Edwards 1993) and some endemics. The island is very mountainous with the highest peak culminating at 3027 m (Gunung Binaya). The island is non-volcanic, mainly composed of sedimentary rocks including limestone massifs and metamorphic rocks (Payton 1993). The climate of Seram is tropical ever wet (only 4 months with precipitations ≤ 100mm), with temperatures at sea level varying between 25° to 30°C. The climate is less humid in the North coast, with a mean annual rainfall about 2 000 mm with a drier season from April to September (Lembaga Meteorologi dan Geofisika, 1969), whereas the south coast receives more than 3 000 mm, with a precipitation peak in July.

The Central Maluku regency covers 11 595 km² (275 097 km² when including the sea territory) and encompasses other islands such as Ambon, Haruku, Saparua and Banda (PBS 2009) (Figure 1). It is divided into districts. In 2010, the regency had a total population of 361 698 people, with an average density of 31 persons/km² (PBS, 2012), and livelihoods is mainly based on agriculture and fisheries.

Central Maluku Regency

Figure 1: Map of Seram Island, Maluku, Eastern Indonesia (Source: Colupsia project)

2) The village of Saleman The study was conducted in the village of Saleman Negeri Saleman (-2°57’S, 129°6’E) and its surrounding agroforests and forested landscape during a period of three months. The village is located on the north coast near the north-west edge of the Manusela National Park (Figure 2). It is accessible by a damaged road. The total population reaches 1666 inhabitants (BPS, 2010). At the end of the 80’, the village population was only about 300 inhabitants (Edward et al. 1993). The villagers are Muslim, and the two most common occupations for livelihood are agriculture and fisheries. The level of education is low; few people have college diplomas or bachelor degrees, but most of the population is able to read and write. Historically, Saleman people belong to the Seramese Nuaulu indigenous group (Ellen 1993). They speak Indonesian language, Ambonese and a local Nuaulu language bahasa Saleman that is spoken in only 4 villages of the north coast (Sawai, Wahai, Hatue and Saleman). This Nuaulu language belongs to the Austronesian phylum (Ellen 2006).

Figure 2: Draft vegetation map of the study site, North coast of Seram Island Maluku, Indonesia (Source: Colupsia project)

The village is on the beach, and is surrounded by hilly landscape (Figures 3a, 3b). The most common vegetation types around the village are lowland and hill limestone forest, logged-over (depleted) lowland forest and mixed gardens (Figure 2). Mixed gardens are a source of conflict with the neighboring national Manusela park authorities because 600 ha gardens are inside the Park (Pers. com., Park authorities, 2012). In addition in 2006 a conflict arose between Saleman village and the neighboring village Horale: 10 000 ha of lands are mutually claimed between Saleman village and Horale village. In 2008, Horale village has even been attacked by Saleman people, many houses have been burnt.

a b

Figure 3: The village of Saleman, Seram Island, Maluku, Eastern Indonesia. The village is on the North Coast of Seram. It is on the beach (a) and is surrounding by a hilly forested landscape (b). (Photos: A. Cosiaux)

B. Participatory mapping and identification of local practices 1) Participatory mapping At the beginning of the study a mapping exercise has been conducted following general principles of IFAD 2009. The mapping has been conducted with customary leaders adat to understand their perception of their territory and the place occupied by the gardens in the landscape. A base map has been previously drawn on a transparency paper, representing neighboring villages, roads, principal rivers and mountain peaks around the village (Bujang 2004). Firstly, informants were asked to give the name of the main rivers, mountains, and locations around the village and draw the boundaries of the village lands, the borders of the Manusela National Park and the eventual conflict areas. Secondly, I asked them to enumerate and describe the different land uses found inside the village territory and sketch them on the map (Martin 1995, Vogl et al. 2004). Once the generic categories of land use type’s perception are identified, I asked them to describe each land use in order to identify the eventual local subcategories. For example, the generic category “garden” may include

9 subcategories such as “vegetable garden”, “fruit garden” (Vogl et al. 2004). Finally, they have been asked about activities inside each land use type. This exercise was crucial to identify the different land use around the village, especially the mixed tree gardens for which I will conduct study in detail.

2) Management of mixed tree gardens: identify local practices To identify the local practices of management of mixed tree gardens, 15 structured interviews of farmers have been conducted. A set of questions have been previously prepared, written that guides interviews. Questions have been asked about the attributes of the garden and then farmers have been asked about the management practices, from the creation of a new garden to the usually management practices in an old garden. Other questions address harvest, tree diseases or problems met by farmers (Appendix 1). All the interviews have been conducted in Indonesian without translator.

C. Typology of mixed tree gardens 1) Sampling design The selection of the sample sites have been based on the participatory map and discussions with local informants. A total of 22 gardens (approximately 28ha) have been visited. All the gardens were located no more than 4 km from the village. In order to study the vegetation characteristics, small plots of 20 x 20 m has been established in each garden. Plots have been selected as much as possible at similar elevation, when possible on similar slopes and in area relatively homogeneous according to environmental conditions. The mean slope within type of gardens varies between 4° (±6) and 16° (±9). All the studied gardens are old gardens established more than 30 years ago. A total of 5 plots (0.2 ha) in each type of mixed tree garden have been sampled. Plot size has been decided according to the methods and empirical knowledge developed previously by the Colupsia project while studying other agroforestry systems in other locations (Comptour 2011, Vallet 2011). A 400m² plot sizes enables to visit several gardens encompassing all garden diversity, and assure a relative homogeneity inside each plot (slope, soil etc) and involve a compromise between covering a wide sample of gardens and limit the work time attributed for each plot.

2) Recorded variables In each plot all trees with a diameter at breast height (dbh) ≥ 5cm have been inventoried, height estimated, vernacular names recorded and voucher specimens collected.

For each type of mixed tree garden, a sketch “profile diagram” of one 400m² plot has been drawn. The position of the tree inside the plot, the height and the crown span has been respected. Moreover, in each garden visited, other tree species found outside the plot have also been inventoried (local name and voucher specimens).

3) Data analysis Structure The overall tree density (all individuals with dhb ≥5cm), the density, the relative density of regenerating trees (5cm ≤ dbh <10cm) and only trees with a dbh ≥10cm have been calculated. To analyze diameter distribution, all trees with a dbh≥5cm were allocated to diameter classes (5-9,9cm, 10-19,9cm, 20-29,9cm etc). In order to analyze height distribution in each type of garden, all trees were allocated to 1 of 4 height classes (2-7m, 7,1-15m, 15,1- 25m, >25m). Then for each type of garden the number of continuous layers have been determined based on the observation on the field and the previously heights distribution. The mean heights of the continuous layers have been calculated. The total basal area G has also been calculated. Density, relative density and the total basal area have been systemically determined for all mixed tree gardens (25 plots) and for each type of mixed tree garden (5 plots).

Floristic composition The number of family and the species richness (S) have been determined at garden scale (based on all tree species inventoried in one gardens) and at plots scale (based on the tree species inventoried inside the plots only). The index of Margalef (Dmg) was calculated (Magurran 2004) at plots scale. The importance value index (IVI) (Cottam and Curtis 1956) of each species was calculated as the sum of relative density, relative frequency and relative dominance. For relative dominance, we use the basal area. Based on the IVI the most important species have been identified for each type of garden and for all gardens. With regards to vertical structure for each type of garden, the most abundant species and the dominant species of each layer have been identified respectively based on their relative density and relative dominance. Then, the Simpson diversity index (1 – D) (Marcon 2011) was calculated to examine the α-diversity of all gardens and of each garden type. We also used the index of Sorensen (Magurran 1988) of plant similarity to determine the β-diversity among the type of gardens. All the formulas to calculate the indexes and variables are summarized in the (Table 1).

Table 1: Biodiversity and species richness indexes and others variables calculated for the data analysis

Parameters and indexes Formulas Explanations Variables and indexes used to analyze the structure and floristic composition of gardens Density N Number of tree per unit area Relative density (Ni/N)*100 With Ni the number of individuals of one species Individual basal area g π*(D/2)² With D diameter of the tree

Total basal area G G = N the number of individuals Relative frequency (number of plots containing a taxon / total frequencies of all taxa)*100 Relative dominance (basal area of a taxon/total basal area of taxa)*100 Importance Value Index (relative density + relative frequency + relative dominance) (IVI) Species richness S Number of species The index of Margalef This species richness index take into account the total (S-1) / ln N Dmg number of inventoried individuals Simpson diversity index D = Σ Ni(Ni- Probability that two randomly selected individuals will (1-D) 1)/N(N-1) not belong to the same species. It ranges from 0 to 1. Where j the number of species common to both sites, a The β-diversity Index of 2j / (a + b) the number of species in the site A and b the number of Sorensen Cs species in the site B Index use to compare free lists content With ns the number of shared species by men and SI = ns / (nm + nw – Similarity index women, nm the number of species cited by men and nw ns) * 100 the number of species mentioned by women.

D. Ethnobotanical knowledge In order to determine the role of mixed tree gardens in household livelihoods, ethnobotanical data have been collected on the local uses of plants by the Saleman people. 1) Interviews on local uses of tree species in mixed tree gardens To identify the local uses of tree, 15 structured interviews (the same farmers interviewed on garden management) have been conducted. During these interviews, I walk with the farmer through his garden, and ask the vernacular names in Indonesian and in local language (Bahasa Saleman) of each tree species. When the farmers don’t know the vernacular name in local language, the name has been asked to a key informant (which fluent in local language and know all the vernacular names of plants). After that, for each inventoried species, questions have been asked about the local uses following procedure recommended by Martin (1995) and Vogl et al. (2004) (Appendix 1). When a species was commercial, the selling prices and locations have been recorded. Theses interviews have led to identify one part of the local uses of each tree species, and one part of the use categories supplied by mixed tree gardens.

2) Free listing Data collection Free listing is a structured interviewing method used in ethnographic research for eliciting systematic data on a “cultural domain”4 (Gravlee 1998, Quinlan 2005). The free listing consists of asking the respondents to list as many names of X as he or she can, where X is the cover term for a cultural domain (e.g. kind of wood for building). Free listing generates two type of information about a cultural domain: which items belong to it and how items are structured inside the list (Gravlee 1998). A total of 27 household heads have been asked to list all the plants they know for a set of 16 uses. For most coherent domains, 20 to 30 informants are usually sufficient (Weller and Romney 1988). The 27 respondents have been randomly selected to ensure including inhabitants from all parts of the village. Finally, 15 men and 12 women have been interviewed. The uses have been previously determined according to uses identified during farmers’ interviews. The first purpose of this free listing exercise is to identify the plants known by villagers for each use and the second purpose is to determine the most culturally important plants for each use. All growth forms (not only tree species) and all origins (gardens and forest) have been included in the free listing exercise, in order to encompass all plant diversity used by communities. This method allows collecting supplementary data about local uses of mixed tree gardens species and identify other important useful species. Indeed the free listing associated with structured interviews enables to increase substantially the quality and quantity of data (Brewer 2002).

Data analysis The two most important results from free listing is the frequency and order with which each item is cited across all respondents (Gravlee 1998, Quinlan 2005). According to cognitive anthropology5, items frequently mentioned and in the top of the lists are culturally important. Moreover, lists are mostly featured by a core-periphery structure: a small set of core items are cited by most respondents, while a much larger set of peripheral items are mentioned a single respondent (Gravlee 1998). Another result is the differences in list length and content which light the intra-cultural variation of respondents’ knowledge (Quinlan,

4 A cultural domain is an organized set of words, concepts or sentences which refers to a single mental category or semantic domain (Weller and Romney 1988). 5 Cognitive anthropology consists of studying the content of thought in communities of individuals. As example cognitive anthropologists seek to discover the pattern of shared knowledge or how collective understandings of the world emerge in social groups.

2005). First of all, to determine if there are quantitative differences between men and women mentioned species, mean list length of men and women for each use have been compared using Student’s t-test. Next, for each use, regarding to frequency of each species mentioned, the core species have been identified. The following analysis focuses only on the core species. To quantify the qualitative differences between men and women species a similarity index (%) has been calculated for each use (Table 1). Then, the composite salience value of each species for each use has been determined for men and women separately. Salience is a statistic accounting for rank and frequency of each item cited in the list (Smith 1993). First, for each use, the salience of listed species (S) has been calculated for each freelister (Table 2).

Table 2: An example of how determinate the salience S of each item mentioned by a freelister (Quilan 2005) Items mentioned by the freelister are illnesses. “Vomiting” is the first mentioned item of the list and “sprains” the last one. To determined the Salience S of each mentioned illness, first you rank items inversely (the final item listed equal one), then you divide the rank by the number of mentioned items.

Then, for each use, a composite salience value of each species has been calculated (Table 3). For each use, and for men and women, bar graph has been realized, representing the species classified by descendent order regarding to composite salience value enabling to determined highly salient, salient, less salient plant species based on the observation of salience breaks (or salience threshold) (Quinlan 2005). The free lists have been analyzed using the software ANTHROPAC 4.0 which calculates index of cultural importance (salience value) Analysis has been conducted based on the vernacular names of species.

Table 3: An example of how calculate the Composite Salient Value of each illness cited by 3 free listers (Quilan 2005). The composite salience value of one item is the sum of all Salience score of that item, and then divided by the number of free lister.

3) Plant voucher specimens Plant voucher specimens of each tree species inventoried in mixed tree gardens and each species mentioned in the free listing have been collected (leaves, fruit and flowers when possible) in 5 duplicates. The specimens have been conserved in newspapers soaked with denatured alcohol (spiritus) before drying, and brought to Bogor Herbarium for identification. It was not possible to collect voucher specimens of 30 species known only through their vernacular names (Appendix 2).

Results A. Customary landscape, the petuanan of Saleman 1) Customary governance of landscape management The village of Saleman is an old village established since more than a century which still applies customary rules as regards to local governance, land management and forest products extraction. Seven clans soa are represented in the village: Makuituin, Aloatuan, Rumaolat, Ialuhun, Aloahiit, Upuolat and Makatita cited in their order of arrival during the establishment of the village. Makuituin is the founder clan of the village. A local government pemerintah desa assures the management of the everyday life in the village. In addition, a customary government pemerintah adat is in charge of customary rules and rituals, allocation and management of the village land. Makuituin clan is heading the customary government as Raja tanah while Makatita clan is heading the local government as Raja.The Rumaolat clan assures the function of war leader Kapitan. The Makatita clan, which had last arrived, is not a member of the customary government and has no responsibilities or decision regarding the land. The petuanan desa (village land territory) is further divided into lahan soa (land of clans) (Figure 4) but Makatita has no lands. Since the first settlement, all the lands belonged to Makuituin the founder clan, which later distributed land for each other clan except to Rumaolat which shares the same lands as Makuituin. Although each clan has a specific land, a member of a clan can use the land of another clan (e.g. to make a new garden) provided that he requests the authorization.

Cocoa gardens area Durian Clove gardens gardens Nutmeg area area gardens area New National Park boundary requested by customary leaders

Coconut Conflict area with the gardens area village of Horale

Figure 4: Participatory map of the Petuanan desa (village land territory) of the village of Saleman realized by the customary leaders (Seram Island, Moloccas, Indonesia) 16 The tanah marga refer to lands of each clan.

2) Participatory mapping and folk classification of landscapes Saleman villagers distinguish 4 categories of landscapes: ewang, aung, aka and dusun. Ewang is the natural forest not-disturbed by human activities or natural hazards. This forest is far from the village (≥ 5km) (Figure 4). Aung corresponds to a forest already disturbed by human activities (e.g. timber extraction) or natural hazards (e.g. fire, landslip). It includes two subcategories: the first one is hutan sudah ditebang which corresponds to disturbed forests where people have just taken timbers; the second one is bekas kebun or bekas aka. Bekas kebun or bekas aka correspond to secondary regrowth, abandoned gardens (fallow) regenerating into secondary forest. Saleman farmers often leave their tree gardens if they find better job outside the village. Kebun in Indonesian or aka in Saleman language describe the gardens where people plant perennial, annual crops or both. This category includes three subcategories: Kebun kecil or aka kiiti, kebun besar or aka maina and kebun khusus or aka khusus. Kebun kecil are small gardens (fenced) surrounding individual houses where people plant a few trees and mostly annual crops (Figure 5a). Small gardens are not developed around all houses. Kebun besar, are not-fenced gardens outside the village in which several tree species are planted (and sometimes a few non tree species). These mixed tree gardens are located around the village (at distance ≤ 5km) (Figure 4). Saleman villagers distinguish 6 types of mixed tree gardens: kebun pala (nutmeg garden), kebun cengkeh (clove garden), kebun kelapa (coconut garden), kebun coklat (cocoa garden) (Figure 6), kebun durian (durian garden) and kebun langsat (Lansium domesticum garden). Kebun khusus corresponds to special banana gardens called kebun pisang or aka uri located on the edge of roads. Finally, dusun corresponds to a land use form inside the forest which consists in extracting and managing a forest product. A dusun covers several hectares and is managed collectively by villagers. The term dusun refers to two subcategories: dusun damar and dusun sagu (Figure 7) where damar resin (Agathis dammara) and sago starch (Metroxylon sagu) are respectively collected. The dusun damar is relatively far from the village, while the dusun sagu are located near a river approximately 5 km from the village (Figure 4).

Figure 5a: A small garden surrounding a house ("Kebun kecil" or "Aka kiiti") Photo: A.Cosiaux 2012

Figure 5b: Coconut seedlings in the small garden Photo : A.Cosiaux 2012

Figure 6: A mixed tree garden ("Kebun coklat" or "Aka coklat") Photo : A.Cosiaux 2012

Figure 7: A managed swamp forest: the "dusun sagu" Photo : A.Cosiaux 2012

3) Mixed tree gardens: general characteristics and local management of the Kebun besar As one can see on the figure 4, and providing that the official park boundary is really at that location (no markers in the field), a lot of mixed tree gardens are located inside the Manusela National Park. When the park border has been created, gardens were already there. Since the border creation, six villagers have been arrested by the national park guards and put in jail for a few months because they had opened new gardens inside the park. Currently, the customary leaders request moving the border of the park. Saleman is also at the same time in conflict for land dispute with the neighboring village Horale: people have abandoned their coconut and langsat gardens near Horace. Langsat gardens have not been studied because they are in the conflict area (Figure 4). The size of the mixed tree gardens is usually between 0,5 and 1,5 ha. The farmers plant trees like durian or maintain big wild trees to set the limits. A mixed tree garden is a private land, managed by a family and transmitted by heritage. When a new garden is created, it belongs to the person who opened the forest and planted trees. Like in many places in traditional rural Indonesia, planting a tree is considered as a way to get property rights on the land. Creating a new mixed tree garden starts by cutting the vegetation, then burning it (but not always). The small gardens surrounding houses are used as nursery in preparing seedlings (Figure 5b) to be planted inside the new plot. In the past, the establishment of new garden started by a period of a few months, during which taro (Colocasia esculenta) or cassava (Manihot esculenta) were planted, and the garden was systematically fenced. Currently, farmers directly plant tree seedling and abandoned this practice (with the exception of few people). The period from opening the forest to plant the seedlings is of approximately 3 months. The first fruiting period varies between species (Appendix 3). Farmers used to work in their gardens 3 days per week and usually have several gardens (1 to 5).

B. Structure and floristic composition of mixed tree gardens 1) Structure Density, diameter and basal area A total of 609 trees has been recorded in the 1ha of mixed tree garden (25 plots). The highest tree density is found in the cocoa garden (170 trees/0,2ha), and the lowest in nutmeg garden (79 trees/0,2ha). The relative density of regenerating trees (5≤dbh<10cm) is low and varies from 19% for nutmeg gardens to 34% for coconut gardens (Table 4). Durian gardens have the highest total basal area (14m²/0,2ha) twice bigger than coconut and nutmeg gardens

which both have a total basal area of 6m²/0,2ha. Cocoa and clove gardens both have the lower total basal areas (4m²/0,2ha) more than 3 times lower than durian gardens.

Table 4 : Density, relative density, mean dbh and total basal area of the 1ha of sampled gardens and the 0,2ha sampled in each type of mixed tree gardens. mean relative relative dbh G trees dbh Type of gardens Density all density Density density mean dbh G all trees density trees trees ≥ ≥10 cm trees trees <10cm trees ≥10cm trees all trees (cm) (m²) ≥10cm 10 cm (m²) <10cm (cm) all (1ha) 609 181 428 30 70 20 ± 19 25 ± 20 35 34 cocoa (0,2ha) 170 52 118 31 69 15±10 25 ± 20 4 4 coconut (0,2ha) 130 44 86 34 66 20± 15 26 ± 15 6 6 clove (0,2ha) 114 34 80 30 70 18 ±13 22 ±13 4 4 durian (0,2ha) 116 36 80 31 69 24 ± 31 32 ± 34 14 14 nutmeg (0,2ha) 79 15 64 19 81 25 ± 20 29 ± 20 6 6

The mean diameters of trunks (dbh) vary from 15cm (±10) for cocoa gardens to 25 (±20) for nutmeg gardens (Table 4). In cocoa, clove, coconut and durian gardens, most of individuals (From more than 50 to 80%) were in the small diameter size classes (between 5 and 19,9cm) (Figure 8). In nutmeg gardens, more than 70% are distributed in the 3 first diameter classes (5 to 29,9cm). Durian gardens have more individuals in the bigger size class than other type of gardens (>10%), they represent indeed a considerable phytomass. Coconut gardens presents also a high proportion (>20%) of individuals in the diameter class 30- 39,9cm.

30 percent 20

0 cocoa clove garden coconut nutmeg durian garden garden garden garden

5-9,9cm 10-19,9cm 20-29,9cm 30-39,9cm 40-59,9cm >60cm

Figure 8: Relative distribution of trees (dbh ≥ 5cm) among diameter classes for 0,2ha of each type of gardens 20

Vertical structure The relative distribution oh height varies between garden types (Figure 9). In cocoa, clove, coconut and durian gardens, most of individuals (>70%) are distributed among the height classes 2-7m and 7,1-15m. In nutmeg gardens, most of trees are distributed in the height classes 7,1-15m and 15,1-25m (>70%).

100

percent 40

0 cocoa clove coconut nutmeg durian garden garden garden garden garden

2-7m 7,1-15m 15,1-25m >25m

Figure 9: Relative distribution of trees (dbh≥ 5cm) among height classes for 0,2ha of each type of gardens

In cocoa gardens there is only one continuous layer (2-7m) dominated by a discontinuous layer between 7 and 15m (Figure 10). In the continuous layer, Theobroma Cacao is the most abundant (83%) and the dominant species (73% of the total basal area). In clove gardens, 2 continuous layers (2-7m and 7,1-15m) surrounded by only a few trees >15m can be identified (Figure 11). Theobroma cacao mixed with Syzygium aromaticum are the most abundant (66%) and the dominant species (64%) of the lower layer. The second layer is also dominated by Syzygium aromaticum (44%). In coconut gardens, there are two lower layers more or less continuous (2-7m and 7,1- 15m) with no clearly dominant species. The most abundant species of the second one is Cerbera floribunda (20%). The upper discontinuous layer (15,1-25m) is dominated by Cocos nucifera (Figure 12). Then, nutmeg gardens are characterized by two more or less distinct medium continuous layers (Figure 13) in which Myristica fragrans are clearly dominant.

Figure 10: Sketch profile of a 20x20m plot in cocoa garden, Figure 11: Sketch profile of a 20x20m plot in clove garden, Seram Island, Moluccas, Indonesia. All trees with a diameter ≥ Seram Island, Moluccas, Indonesia. All trees with a diameter ≥ 5cm have been drawn. 5cm have been drawn.

Species composition Species composition Lower layer Lower layer Theobroma cacao coklat: 2, 4, 5, 7, 8, 9, 11, 12, 13, 15, 16, 17, Theobroma cacao coklat: 1, 2, 3, 4, 6, 9, 10, 11, 13, 14, 15, 16, 19, 21, 23, 24, 25, 27. 18, 19, 21, 22, 23, 24, 25. Coffea canephora kopi: 10, 14. Gliricidia sepium pohon pelindung: 28. Cinnamomum verum kayu manis: 1. Upper layer

Upper layer Syzygium aqueum jembu air: 5. Syzygium aromaticum cengkeh: 3. Syzygium aromaticum cengkeh: 7, 8, 12, 22, 26. Spondias cytherea kedongdong: 6. Cordia myxa kayu semang: 17.

Syzygium malaccense Jembu Makui: 20 Mallotus ricinoides kayu kapur: 18. Lansium domesticum langsat: 20.

Species composition

Lower layers 2-15m Cynometre cauliflora namu-namu: 2, 8, 22. Citrus aurantifolia lemon cina: 5, 18. Cynometre cauliflora namu-namu: 2 Mangifera indica mangga java: 6, 24. Syzygium malaccense jembu makui: 7, 23. Areca catechu pinang: 9, 10. Flacourtia rukam tomi-tomi: 11. Annona muricata Nangka belanda: 12, 21. Neuburgia moluccana jati hutan: 13, 14, 15. Buchanania arborescens jembu meteng hutan : 25.

Upper Layer Artocarpus altilis sukun : 1. Cocos nucifera kelapa : 3, 4, 16, 19, 20.

Figure 12: Sketch profile of a 20x20m plot in coconut garden, 22 Seram Island, Moluccas, Indonesia. All trees with a diameter ≥ 5cm have been drawn.

Figure 13: Sketch profile of a 20x20m plot in nutmeg garden, Seram Island, Moluccas, Indonesia. All trees with a diameter ≥ 5cm have been drawn.

Species composition Myristica fragrans pala : 1, 4, 5, 6, 8, 9, 12, 13, 14, 15. Durio zibethinus durian : 2, 10. Lansium domesticum langsat : 3, 11. Areca catechu pinang : 7.

Figure 14: Sketch profile of a 20x20m plot in durian garden, Seram Island, Moluccas, Indonesia. All trees with a diameter ≥ 5cm have been drawn.

Species composition Lower layer 2-15m Theobroma cacao coklat: 1, 2, 10, 16, 23, 25. Syzygium aromaticum cengkeh: 5, 9. Mangifera odorata mangga kuini: 8.

Hibiscus tiliaceus kayu baru: 13. Mangifera indica Mangga java: 15. Lansium domesticum langsat : 22.

Medium layer 15,1-25m Table 5: Mean heights (m) of layers in each type of garden Areca catechu pinang: 4, 6, 7, 17. Durio zibethinus durian: 11, 12, 14, 18, 20, 21.

cocoa coconut clove durian nutmeg

gardens gardens gardens gardens gardens Upper layer Durio zibethinus durian: 3, 19, 24. lower layer 2-7m 5 ± 1 5 ± 1 5 ± 1 5 ± 1 - low layer 7,1-15m 10 ± 2 11 ± 3 12 ± 2 11 ± 3 11 ± 2 medium layer 15,1-25m - 23 ± 3 - 18 ± 3 21 ± 3 upper layer > 25m - - - 37 ± 7 -

Finally, durian gardens are the most complex with 4 layers more or less continuous. Durio zibethinus exclusively dominates the canopy layer characterized by a mean height of 37m (±7m) (Figure 14). The species is also dominant (76,95%) in the medium layer mixed with S. aromaticum (16,07%). Theobroma cacao is dominant in the lower layer (82,39%) and the two species Syzygium aromaticum (47,07%) and Areca catechu. (16,95%) are dominant between 7 and 15m. The mean height of each layer is summarized in Table 5. The dominant and most abundant species of each layer are summarized in (Appendix 4).

2) Floristic composition Within the 22 mixed tree gardens visited (28ha), 110 tree species belonged to 41 families have been inventoried with respectively 76 species in cocoa gardens (8ha), 41 species in coconut gardens (5ha), 40 species nutmeg gardens (5ha), 37 species in clove gardens (5ha) and 32 species in durian garden (5ha) (Appendix 5). With regard to the number of species found within families, Moraceae (8 species) and Fabaceae (8 species) are the most diverse tree families. They are followed by Anacardiaceae (7 species), Euphorbiaceae (6 species), Myrtaceae (6 species) and Verbenaceae (6 species). Then, within the 25 sampled plots (1ha), a total of 53 species belonged to 27 families with 32 species in coconut gardens, 27 in cocoa gardens, 17 in clove gardens, 14 in durian gardens and 10 in nutmeg gardens have been inventoried (Table 6). Based on the qualitative index of Sorensen, the highest floristic similarity was found between durian and clove gardens (0,52) as well as durian and nutmeg gardens (0,50) (Table 7). At plots scale level when all garden plots are analyzed together, the Margalef index is 8,11. Among each garden types, the richness index of Margalef varies from 2,06 (nutmeg garden) to 6,37 (coconut gardens) (Table 6). For the 25 plots, the Simpson index is 0,87 and it varies within garden types from 0,91 (coconut gardens) followed by durian gardens (0,84), clove garden (0,82) nutmeg garden (0, 59) and cocoa gardens (0,54)(Table 6). Indeed, in taking in account the relative abundance of each species, coconut gardens stills the most diversified whereas cocoa gardens are the less one. Nutmeg gardens with the lower species richness have also a low Simpson index value. Among the 53 tree species inventoried in the 25 plots, 6 species have a major role according to their importance value index. Not surprisingly Durio zibethinus, Theobroma cacao, Myristica fragrans, Cocos nucifera, Syzygium aromaticum and Lansium domesticum account for 62,94% of the importance value (Figure 15). The three most important species of each type of gardens are summarized in the Table 8. They encompass 8 species. In each garden type the species which give its name to the garden have the highest IVI.

Table 6: Families, Species richess and diversity indexes for each type of garden at plots scale and at gardens scale

At plots scale At gardens scale Simpson Number of Margalef S (trees dhb ≥ Diversity Type of Number of Plots families (all S (all trees ) Index (all S 10 cm) index (all garden families trees) trees) trees) all (1ha) 27 53 45 8,11 0,87 All (28ha) 41 110 cocoa 17 27 22 5,06 0,54 cocoa (8ha) 37 81 garden(0,2ha) coconut 20 32 23 6,37 0,91 coconut (5ha) 24 42 garden (0,2ha) clove garden 13 17 16 3,38 0,82 clove (5ha) 23 39 (0,2ha) durian garden 11 14 9 2,73 0,84 durian (5ha) 23 34 (0,2ha)

nutmeg garden 7 10 9 2,06 0,59 nutmeg (5ha) 21 41 (0,2ha)

Table 7: β-diversity between the 5 types of garden based on the qualitative Sorensen index type of garden cocoa durian nutmeg coconut clove

cocoa - 0,44 0,32 0,41 0,36 durian - - 0,50 0,39 0,52 nutmeg - - - 0,24 0,37 coconut - - - - 0,37 clove - - - - - Distribution of IVI (%) Durio zibethinus Murr.

19,6 Theobroma cacao

Myristica fragrans

15,9 Cocos nucifera

Syzygium aromaticum 6,1 8,0 6,6 6,7 Lansium domesticum Corr.

Figure 15: The distribution of the Importance Value Index among the 53 species. The six most important species are listed on the right.

Table 8: The three most important species of each type of gardens according to their Importance Value (%)

Cocoa gardens Coconut gardens Clove gardens Durian gardens Nutmeg gardens

IVI IVI IVI IVI Species Species Species Species IVI (%) Species (%) (%) (%) (%) Syzygium Theobroma cacao 38,5 Cocos nucifera 25,1 19,6 Durio zibethinus 42,3 Myristica fragrans 42,9 aromaticum Cerbera Lansium Durio zibethinus 7,2 6,6 18,0 Theobroma cacao 13,9 Durio zibethinus 16,6 floribunda domesticum Artocarpus Kleinhovia Syzygium 6,1 6,4 Theobroma cacao 16,8 10,8 Lansium domesticum 11,5 heterophylla hospita aromaticum

sum 51,8 38,1 54,4 66,9 71,0

C. Local uses of tree and other plant species 1) Local uses of mixed tree garden species Among the 110 tree species inventoried in the mixed tree gardens (Appendix 5), 102 are useful, 19 species have 1 use only, 67 species between 2 and 5 uses, 15 species between 6 and 10 uses and just 1 species with more than 10 uses. Only 8 species are not used by the local population. The 8 species with no use are spontaneous and among the 102 useful species, 43 are planted and 59 are spontaneous. The number of uses per species varies between 1 and 18 (Appendix 5). During the interviews of farmers, 24 uses have been identified (Table 9).

Table 9: Number of tree species per use and proportion of planted and spontaneous species for each use Category of Number of Proportion of planted Proportion of Uses uses species species (%) spontaneous species (%) Source of Cash crops 4 100 0 income Supplementary cash crops 17 94 6 House construction 58 26 74 Construction Canoe contruction 32 16 84 House furniture 27 26 74 Firewood 81 38 62 Fishing tools 8 38 63 Tools Rope 7 43 57 Staple food 5 100 0 Fruit for food 38 84 16 Vegetable 8 63 38 Food Condiment 15 67 33 Drink 10 70 30 Sugar 2 50 50 Oil 1 100 0 Medecine 27 59 41 Cosmetic 13 62 38 Ornemental 10 90 10 Fence 4 50 50 Garden Fertiliser 2 50 50 management Protection tree 1 100 0 Basketery 1 100 0 Ritual 4 75 25 Other 1 100 0

The two uses “cash crops” and “supplementary cash crops” concern tree species which supply cash income. The 4 species Myristica fragrans, Syzygium aromaticum, Theobroma cacao and Coffea canephora are the main source of cash income of farmers in Saleman (See the selling prices and estimation of one year production in Appendix 6). The productions are sold in a cooperative in Masohi (the closest city). Coffee is only planted by a few people and plays a minor role in households’ economy. Another 17 species, mainly fruit trees like Durio zibethinus, Cocos nucifera or Mangifera indica, may supply supplementary incomes. Farmers sell fruits in the village or sometimes in the city when the production is important. The other uses are linked to households’ subsistence. According to construction materials, 3 uses have been identified: “house construction”, “canoe construction” and “house furniture” with 58, 32 and 27 species. For example, Intsia bijuga is used for house construction, Canarium oleosum for the construction of canoe or Polyalthia lateriflora used to make furniture. Next, 81 species is used as firewood like clove tree or Cryptocarya densiflora. Regarding species used to make tools, people use 8 species for making fishing tools (e.g. sago palm) and 7 species for making rope (e.g. Kleinhovia hospita). Then, 7 uses cover food supply: 5 staple food species, 38 fruit tree species, 8 vegetable species, 15 condiment species, 10 species used to make drink, 2 species for making sugar (Cocos nucifera and Arenga pinnata) and 1 species for oil production have been identified. Next, 27 medicinal trees, 13 species used as cosmetic and 10 ornamental trees have been recorded. Cocos nucifera is used for basketry, and Musa sp., Arenga pinnata and Areca catechu are used during customary rituals. Finally, Saleman villagers use 3 species to make fence, 2 species as fertilizer and 1 species as shade tree. Among all the mixed tree gardens species, the five one which have the more uses are coconut palm, sago palm, banana trees, Canarium oleosum and Areca catechu which count respectively 18, 8, 8, 8 and 7 uses. Generally speaking, we note that species used for firewood, construction and tools are mainly wild species (more than 50%), whereas species assuring other uses are mainly planted species (Table 9).

2) The other plants used by communities During the free listing a total of 125 other species belonged to 58 families have been identified (Appendix 5). Among these 125 species, 43 are trees and 82 are herbaceous and lianas species. Regarding to tree species, 17 have been sampled in small gardens surrounding houses and 26 in forest (secondary forest and dusun). Concerning non tree species, 24 have been sampled in mixed tree gardens, 54 in small gardens surrounding houses and 14 in forest. Indeed we can see that mixed tree gardens can also contain herbaceous or lianas useful

27 species and small gardens surrounding houses supply more useful non tree species than other land use type. To summarize, a total of 235 species have been inventoried including 227 useful species.

1) Important species for each use category Intracultural variations For the 16 categories of use (Appendix 7) studied during the free listing exercise, the number of species mentioned (using vernacular names) varies between the respondents. In comparing the mean list length of men and women for each use, significant difference has been identified for 6 categories of uses (Appendix 8). The number of species mentioned by men for the use categories “canoe construction”, “fishing tools” and “sugar” is higher than number of species cited by women whereas women have mentioned more species than men for the 3 uses “vegetable”, “medicine” and “cosmetic”. Next, qualitative differences have also been identified between species which men and women have mentioned. The number of shared items is low for 10 uses for which the similarity index is ≤ 50% like “house construction”, “staple food” or house furniture” (Appendix 9). The similarity index is higher than 60% for 4 use categories (“fruit for food”, “vegetable”, “condiment”, “oil” and “sugar”). Items cited by men and women for “oil” and “sugar” are the same (SI=100%). Indeed, we can see, there are differences in list length and content which highlight intra-cultural variations in local knowledge.

From highly salient to less salient species Highly salient, salient and less salient items have been identified for men and women for each use (Appendix 10). In order to illustrate the results 3 uses have been chosen. Firstly, for the construction of house, the highest salient item is kayu besi according to men and women (Figure 16a, b) with a high composite salience value. The local name kayu besi encompass 2 species: Intsia bijuga and Pongamia pinnata. Then, for women kayu gufasa (Vitew cofassus) and kayu lenggua (Pterocarpus indicus) can be identified as salient species but less than kayu besi (Figure 16a). Indeed a threshold in the composite salience value is observed between the first item and the second one. Next, another salience threshold is distinguished between kayu lenggua and the following item which enables to identify the less salient items kayu palaka (Octomeles moluccana) and durian (Durio zibethinus). These species are less salient, but they have been mentioned by about a half of the respondents. For men, kayu gufasa can also be identified as salient species following the highly salient species

SALAWAKU MABUK MERAH TITI ASOER BELIMBING HUTAN ALALI MABUK TITI_HUTAN TITI BELO_HITAM GIAWAS KENARI HUTAN BURUNG BELO_HITAM KENANGAN MERANTING MAKILA SIKI/SIRIN KELAPA MERANTING SIKI SAMAMA PAPAYA PULI HUTAN PULI TAWANG DURIAN DURIAN KENARI PALAKA LENGGUA LENGGUA PALAKA GUFASA SIKI GUFASA BESI BESI 0 0,2 0,4 0,6 0,8 1 0 0,2 0,4 0,6 0,8 1 Composite salience value Composite salience value

Figure 16a: Composite Salience Value of construction wood Figure 16b: Composite Salience Value of construction wood (Men) (Women)

LABU DURIAN KELAPA SUKUN LABU JAGUNG JAGUNG PATATAS PISANG KELADI KUMBILI KELAPA PISANG PATATAS SINGKONG KELADI SAGU SINGKONG SAGU 0 0,2 0,4 0,6 0,8 1 0 0,2 0,4 0,6 0,8 1 Composite salience value Composite salience value

Figure 17a: Composite Salience Value of staple food species Figure 17b: Composite Salience Value of staple food species (Men) (Women)

BUAH_BORI TALI_BORI ROTAN BAMBU BAMBU 0 0,2 0,4 0,6 0,8 1 0 0,2 0,4 0,6 0,8 1 Composite salience value Composite salience value

Figure 18a: Composite Salience Value of species used as fishing tool Figure 18b: Composite Salience Value of species used as fishing (Women) tool (Men) 29

kayu besi (Figure 16b). Finally kayu siki and kayu palaka can be identified as less salient species. Kayu siki also encompass 2 species: Palaquium obovatum and Palaquium obtusifolium. Secondly, for the category of use “staple food”, the most salient item for men and women is sagu (Metroxylon sagu) followed by singkong (Manihot esculenta) also salient (Figure 17a, b). As less salient “items”, for men we distinguish keladi (Colocasia) and patatas (Ipomea batatas) (Figuge 16b) and for women pisang (Musa sp.) and keladi (Figure 17a). Thirdly, the only important species for fishing is bambu with a very high composite salience value (Figure 18a, b). The local name bambu refers several species belonged to the sub-family of Bambusoideae.

The contribution of mixed tree gardens and other land uses in livelihood For each use, the species identified as highly salient, salient and less salient have different characteristics according to growth form and origin (mixed tree gardens, small gardens or forest) (Appendix 11). Indeed species used as construction wood, fishing tools, cosmetic, ornamentation and for making drink are tree species found in mix tree gardens apart from kunyit (Curcuma domestica). Then, species used as firewood are mainly found in mixed tree gardens, just two of them kayu pakat (Casearia grewiaefolia) and kayu sihata (Mallotus peltatus) have been sampled in forest. For cooking oil production, the only used species, coconut tree, is only found in mix tree gardens. On the contrary, salient species used as vegetable and condiment are from small gardens surrounding house. Most of them are herbaceous like sare (Cymbopogon nardus) or shrub like cili (Capsicum annuum, Capsicum frutescens). For staple food, Metroxylon sagu is mainly from managed swamp forest whereas the others important species are found in small garden near houses. All fruit species are trees from mixed tree gardens apart from papaya (Papaya carica) exclusively planted in small gardens. Finally, the most salient medicinal plants are both from mixed tree gardens and small gardens and are herbaceous or tree species.

Discussion A. Mixed tree gardens: an integrated land use The village lands are composed by a mosaic of tree garden systems, secondary forest and not disturbed forests as in many rural areas in Indonesia (Michon et al. 1986, Hariyadi and Ticktin 2012). In the past, there were some upland padi fields (ladang) around the village,

30 like in other part of west Seram (Ellen 1993), but Saleman farmers abandoned this practice. The land use system is still controlled by customary law (adat). The village land is divided among clans and its management is regulated by customary practices which still widespread in Maluku (Ellen 1985). The two categories of gardens in Saleman are agroforestry systems. The small garden kebun kecil or aka kiiti surrounding houses including fruit trees and a majority of herbaceous species used for self consumption by households may be compared to home-gardens (Kumar and Nair 2004, Kabir and Webb 2009) which are common in Java or Sulawesi (Christanty et al. 1986, Kelhenbeck 2004). However, not all households have small gardens. Indeed this kind of land use seems to be a minor component of the land use system. One reason may be the lack of available area between houses or near the village (since the expansion of the village in the past 20 years) or the soil fertility because the village is on the sandy beach front. The second type of garden is the mixed tree gardens called kebun besar or aka maina. Their establishment consists in cutting a patch of forest and planting tree seedlings. The slashed plot is not systematically burnt before cropping and farmers don’t use fertilizer. Indeed fertility is assured by natural nutrient cycling processes which constitute one of the characteristic of lots of complex agroforestry systems (Michon et al. 1986, Wiersum 2004, Belcher et al. 2005). In the past the practices were different: in recent decades, farmers abandoned planting food crops in the first phase of establishment of a new tree garden. This change is associated to the abandon of rice fields. One reason may be the increasing need of cash income which leads farmers to plant directly trees and focus their input on commercial crops. Another factor may be the easier access to market to sell commercially valuable products which starts in the 80’s (Ellen 1997, 1999) accompanied by the introduction of new cash crops like cocoa (Clough et al. 2009). These smallholders tree plantations integrated in the forested landscape around the village are private plots owned by farmers as regard to customary law. In several other regions in Indonesia, like in Sumatra near the Kerinci Seblat National Park, where lands are historically under customary laws and owned by clans, land tenure gradually shift from common property to more private property when cash crops are introduced in the tree garden systems (Belcher et al. 2005; Hariyadi and Tickin 2012). Then, the mixed tree gardens are a source of conflict with the Manusela National Park authorities. Similar conflicts between local populations and park authorities on the periphery of conservation areas have already been recorded in many areas in Indonesia (Aumeeruddy and Sansonnens 1994, Murniati et al. 2001, Hariyadi and Tickin 2012). Before the creation of the park boundaries in 1997, 600 ha

31 of mixed tree gardens were already established there. Since then, the farmers continue to make new gardens especially because the boundaries are not clear (old original boundaries with concrete post, new boundaries with wood stick only). In this context of conflict, the expansion of gardens may actually ensure “property rights” when planting trees is a mean to get long term use rights on a land (Michon 2003).

B. Mixed tree gardens: complex agroforestry systems similar to natural forest The mixed tree gardens are multistoried with 2 to 4 layers. Like in others part of Indonesia, durian gardens have the most complex structure (Michon et al. 1986). The less complex cocoa gardens are not-shaded. In Sulawesi, lots of smallholder cocoa plantations are also not-shaded and present a simple structure because farmers remove tree cover to increase production yields (Clough et al. 2009) The dominant species of the mixed tree gardens of Saleman, are also dominant in the “dusun system” of Saparua Island studied by Kaya et al. (2002). Indeed, similarities exist within agroforestry systems in Maluku. In comparing the 1ha sampled mixed tree gardens with 1ha lowland limestone forest in Seram (Ranlund 2011), we note that the density of trees (dbh≥10cm) in mixed tree gardens (428) is close to the density in primary forest (438). However, the total basal area is clearly higher in primary forest. As regard to species richness, 95 tree species have been found in primary forest and 45 species in gardens (25 plots). In other study of lowland forest of Seram, 17 to 54 species have been inventoried in 0,25ha plots (Edwards 1993). Indeed, the phytomass and the species richness are lower in mixed tree gardens. However, at gardens scale a total of 110 species belonged to 42 families have been identified in the 22 gardens. These results highlight the effect of sampling effort on the assessment of species richness and also the possible impact of farmer choice and practices on the spatial distribution of species. Some species are always planted or grow spontaneously in border of the garden. A plot settled in the middle of the garden is not able to assure the sample of all species. Finally, we can suggest that the highest floristic similarity observed between durian and clove garden is due to the local agroforestry practices and farmer choices.

C. Importance of mixed tree gardens for local population The mixed tree gardens, with 110 trees and 24 lianas and herbaceous useful species, can supply fruits, medicine, construction wood, vegetable or cash income in selling commercial products like lots of tree garden system in Sumatra, Java or Sulawesi (Michon et

32 al. 1986, Brodbeck et al. 2003, Manurang et al. 2004, Okubo et al. 2010). The 3 cash species Theobroma cacao, Syzygium aromaticum and Myristica fragrans are the major source of income for Saleman farmers like Hevea brasiliensis (Gouyon et al. 1993) or Shorea javanica for farmers of Sumatra. Then, mixed tree gardens play a major role in the auto-subsistence of farmers’ households in supplying firewood (81 species), house construction wood (58 species), medicines (27 species) or fruits (38 species). These results confirm the key role of complex agroforestry systems in the auto-subsistence of rural households already identified in other tree based systems of Indonesia (Christanty et al. 1986, Michon et al. 1986, Gouyon et al. 1993, Brodbeck et al. 2003, Okubo et al. 2010). However, the mixed tree gardens can’t assure all livelihoods. Other land use type supply complementary products or incomes. The small gardens surrounding houses are the major source of vegetable, condiment and staple food. They also supply medicinal plants and fruits for the daily auto-consumption. The natural forest and the managed forest like dusun sagu also supply useful species and contribute to the rural livelihoods (Sunderlin et al. 2004). The dusun dammar, another example of managed forest habitat, takes part in cash income supply although the market has been up and down. Indeed I confirm here that the mixed tree gardens are integrated in an overall land use system, complementing other land uses and complemented by other land uses (Belcher et al. 2005, Wiersum 2004). However, some products, especially the rice, which people use every day, are bought. In this case, cash incomes from mixed tree gardens allow Saleman villagers to buy staple food.

D. The mixed tree gardens inside the Manusela National park Thus, the local challenge is the reconciliation between the management of the conservation area with the social and economic needs of the local population. The mixed tree gardens currently inside the boundaries of the Manusela National Park are characterized by a forest like structure more or less complex and account an appreciable amount of wild tree species. Indeed, these gardens may be an opportunity to conserve wild tree species outside the natural forest and supply a diversity of habitats for the wild fauna (birds, mammals or insects) and other plant species (lianas, epiphytes, herbaceous). Several studies have shown the potential of agroforests to conserve biodiversity (Michon and De Foresta 1995, Bos et al. 2007, Beukema et al. 2007,). A solution to present conflicts could be the promotion of this agroforestry system as a buffer zone limiting further access to the park (Aumeeruddy and Sansonnens 1994, Murniati et al. 2001), a transition zone between the park and the surrounding villages.

Conclusion The results of this study have shown that tree gardening systems of Seram Island are another example of reconstructed-forest by tropical farmers. These tree gardens, characterized by a multistoried structure and high species richness play a key role in the economy of Saleman farmers. The two mains socio-economic function of mixed tree gardens consist of assuring cash incomes and supplying forest products essential for poor-rural communities. Moreover, this land use is not isolated, but integrated in a complex land use system where each land use assure different socio-economic functions and complement each others. Then, in providing diverse habitats for wild species (plant and animals), the promotion of mixed tree gardens as a buffer zone between the park and the surrounding villages may be a solution to the conflict with the Manusela National Park authorities. Moreover, it would be interesting to study the potential of conservation of these gardens as regards to wild fauna, or other taxa, and also the role of these systems according to ecosystems services. In the framework of Colupsia project, the present results associated with the previously studies in others villages around the Manusela National Park, may offer interesting propositions concerning the land use planning which take in account the local population priorities.

References Aumeeruddy, Y., Sansonnens, B., 1994. Shifting from simple to complex agroforestry systems: an example for buffer zone management from Kerinci (Sumatra, Indonesia). Agroforestry systems 28, 113–141. Aumeeruddy-Thomas, Y., 2003. Conflits de pouvoir et de représentations à l’interface des sociétés et des aires protégées: exemple du Parc National Kerinci Seblat en Indonésie. Bois et forêts des tropiques 4, 77–92. Belcher, B., Michon, G., Angelsen, A., Pérez, M.R., Asbjornsen, H., 2005. The socioeconomic conditions determining the development, persistence, and decline of forest garden systems. Economic Botany 59, 245–253. Beukema, H., Danielsen, F., Vincent, G., Hardiwinoto, S., Andel, J., 2007. Plant and bird diversity in rubber agroforests in the lowlands of Sumatra, Indonesia. Agroforestry Systems 70, 217–242. Bhagwat, S.A., Willis, K.J., Birks, H.J.B., Whittaker, R.J., 2008. Agroforestry: a refuge for

tropical biodiversity? Trends in Ecology & Evolution 23, 261–267. Bos, M.M., Steffan-Dewenter, I., Tscharntke, T., 2007. The contribution of cacao agroforests to the conservation of lower canopy ant and beetle diversity in Indonesia. Biodiversity and Conservation 16, 2429–2444. BPS, 2009. Maluku dalam angka. BPS, 2010. Population of Indonesia by Village. BPS, 2012. Maluku dalam angka. Brodbeck, F., Hapla, F., Mitlöhner, R., 2003. Traditional forest gardens as “safety net” for rural households in Central Sulawesi, Indonesia, in: International Conference on Rural Livelihoods, Forests and Biodiversity, Bonn, Germany. pp. 19–23. Brewer, D. 2002. Supplementary interviewing techniques to maximize output in free listing task. Field methods 14, 108-118. Bujang, M., 2004. A Community Initiative: Mapping Dayak’s Customary Lands in Sarawak. Christanty, L., Abdoellah, O.L., Marten, G.G., Iskandar, J., 1986. Traditional agroforestry in West Java: the Pekaranagan (homegarden) and Kebun-Talun (annual-perennial rotation) cropping systems, in: Traditional Agriculture in South Est Asia. pp. 132–158. Clough, Y., Faust, H., Tscharntke, T., 2009. Cacao boom and bust: sustainability of agroforests and opportunities for biodiversity conservation. Conservation Letters 2, 197–205. Comptour, M., 2011. Exploring trees diversity and local perceptions in primary forest and other human land uses in West Kalimantan, Indonesia (Master thesis). Cirad-Cifor. Cottam, G., Curtis, J.T., 1956. The use of distance measures in phytosociological sampling. Ecology 37, 451–460. Curran, L.M., Trigg, S.N., McDonald, A.K., Astiani, D., Hardiono, Y.M., Siregar, P., Caniago, I., Kasischke, E., 2004. Lowland forest loss in protected areas of Indonesian Borneo. Science 303, 1000–1003. Edwards, I.D., Macdonald, A.A., Proctor, J., Operation Raleigh (Organization), Raleigh International, 1993. Natural history of Seram : Maluku, Indonesia. Intercept Ltd., Andover, Hampshire, England. Edwards, I.D., Proctor, J., Riswan, S., 1993. Rain Forest Types in the Manusela National Park, in: Natural History of Seram: Maluku, Indonesia. Edwards, I.D., Macdonald, A.A., Proctor, J., pp. 63–74. Ellen, R., 1985. Patterns of indigenous timber extraction from Moloccan rain forest fringes. Journal of biogeography 559–587.

Ellen, R., 1993. Human impact on the Environment of Seram: Maluku, Indonesia, in: Natural History of Seram. Edwards, I.D., Macdonald, A.A., Proctor, J., pp. 191–205. Ellen, R., 1997. The human consequences of deforestation in the Moluccas. Civilisations. Revue internationale d’anthropologie et de sciences humaines 176–193. Ellen, R., 1999. Forest knowledge, forest transformation: political contingency, historical ecology and the renegotiation of nature in central Seram. Studies in Environmental Anthropology no. 4. Ellen, R., 2006. The Cultural Relations of Classification: An Analysis of Nuaulu Animal Categories from Central Seram. Cambridge University Press. Ellen, R.F., 1998. Indigenous knowledge of the rainforest: perception, extraction and conservation. GEOJOURNAL LIBRARY 44, 87–100. FAO, 2010. Key Finding, in: Global Forest Resource Assessment. Foresta, H. de, Michon, G., 1996. Etablissement et gestion des agroforêts paysannes en Indonésie : quelques enseignements pour l’Afrique forestière, in: Hladik, C.M., Hladik, A., Pagezy, H., Linares, O.F., Koppert, G.J.A., Froment, A. (Eds.), L’alimentation en forêt tropicale : interactions bioculturelles et perspectives de développement : 2. Bases culturelles des choix alimentaires et stratégies de développement, L’Homme et la Biosphère. UNESCO, Paris, pp. 1081–1101. Gouyon, A., Foresta, H., Levang, P., 1993. Does “jungle rubber”deserve its name? An analysis of rubber agroforestry systems in southeast Sumatra. Agroforestry systems 22, 181–206. Gravlee, L., 1998. The uses and Limitations of Free Listing in Ethnographic Research [WWW Document]. Research methods in Cognitive Anthropology. URL http://gravlee.org/ang6930/freelists.htm Hansen, M.C., Stehman, S.V., Potapov, P.V., Arunarwati, B., Stolle, F., Pittman, K., 2009. Quantifying changes in the rates of forest clearing in Indonesia from 1990 to 2005 using remotely sensed data sets. Environmental Research Letters 4, 034001. Hariyadi, B., Ticktin, T., 2012. From Shifting Cultivation to Cinnamon Agroforestry: Changing Agricultural Practices Among the Serampas in the Kerinci Seblat National Park, Indonesia. Human Ecology 40, 315–325. Harvey, C.A., Komar, O., Chazdon, R., Ferguson, B.G., Finegan, B., Griffith, D.M., MartíNez-Ramos, M., Morales, H., Nigh, R., Soto-Pinto, L., Van Breugel, M., Wishnie, M., 2008. Integrating Agricultural Landscapes with Biodiversity Conservation in the Mesoamerican Hotspot. Conservation Biology 22, 8–15.

IFAD, 2009. Good practices in participatory mapping. Kabir, M.E., Webb, E.L., 2008. Household and homegarden characteristics in southwestern Bangladesh. Agroforestry Systems 75, 129–145. Kaya, M., Kammesheidt, L., Weidelt, H.J., 2002. The forest garden system of Saparua island Central Maluku, Indonesia, and its role in maintaining tree species diversity. Agroforestry systems 54, 225–234. Kehlenbeck, K., Maass, B.L., 2004. Crop diversity and classification of homegardens in Central Sulawesi, Indonesia. Agroforestry Systems 63, 53–62. Kumar, B.M., Nair, P.K.R., 2004. The enigma of tropical homegardens. Agroforestry Systems 61, 135–152. Laurance, W.F., 1999. Reflections on the tropical deforestation crisis. Biological Conservation 91, 109–117. LMG, 1969. Mean rainfall on the islands outside Java and Madura (period 1931-1960). Magurran, A.E., 1988. Ecological diversity and its measurement. Princeton University Press, Princeton, N.J. Magurran, A.E., 2004. Measuring biological diversity. Blackwell Pub., Malden, Ma. Manurung, G.E.S., Roshetko, J.M., Budidarsono, S., 2004. Traditional tree farming systems in West Java and their importance to local people. World Agroforestry Centre (ICRAF)–Bogor, Indonesia. Marcon, E., 2011. Mesurer la biodiversité. Marjokorpi, A., Ruokolainen, K., 2003. The role of traditional forest gardens in the conservation of tree species in West Kalimantan, Indonesia. Biodiversity and Conservation 12, 799–822. Martin, G.J., 2004. Ethnobotany : a methods manual. Earthscan, London; Sterling, VA. Michon, 1996. Agroforests: An Original Agroforestry Model from Smallholder Farmers for Environmental Conservation and Sustainable Development. [WWW Document]. Michon, G., 2003. Déforestation et reconstructions forestières en Indonésie : de la transformation des paysages aux recompositions sociales et politiques. Bois et forêts des tropiques 4, 65–75. Michon, G., de Foresta, H., 1995a. Agroforest: an original agro-forestry model from smallholder farmers for environmental conservation and sustainable development, in Traditional Technology for Environmental Conservation and Sustainable Development in the Asian-Pacific Region. Ishizuka, K., S. Hisajima, and DRJ Macer 52–58.

Michon, G., de Foresta, H., 1995c. The Indonesian agroforest model forest resource management and biodiversity conservation, in: Conserving Biodiversity Outside Protected Areas: The Role of Traditional Agro-ecosystems. Patricia Halladay and D A Gilmour, pp. 90–106. Michon, G., de Foresta, H., 1997. Agroforests: pre-domestication of forest trees or true domestication of forest ecosystems? Netherlands journal of agriculture science 461– 462. Michon, G., Mary, F., Bompard, J., 1986. Multistoried agroforestry garden system in West Sumatra, Indonesia. Agroforestry Systems 4, 315–338. Monk, K.A., Fretes, Y.D., Reksodiharjo-Lilley, G., 1997. The ecology of Nusa Tenggara and Maluku. Periplus Editions. Murniati, Garrity, D.P., Gintings, A.N., 2001. The contribution of agroforestry systems to reducing farmers’ dependence on the resources of adjacent national parks: a case study from Sumatra, Indonesia. Agroforestry Systems 52, 171–184. Myers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A.B., Kent, J., others, 2000. Biodiversity hotspots for conservation priorities. Nature 403, 853–858. Okubo, S., Parikesit, Harashina, K., Muhamad, D., Abdoellah, O.S., Takeuchi, K., 2010. Traditional perennial crop-based agroforestry in West Java: the tradeoff between on- farm biodiversity and income. Agroforestry Systems 80, 17–31. Payton, R.W., 1993. Soils of the Manusela National Park, in: Natural History of Seram: Maluku, Indonesia. Edwards, I.D., Macdonald, A.A., Proctor, J., pp. 19–61. Perfecto, I., Vandermeer, J., 2008. Biodiversity Conservation in Tropical Agroecosystems. Annals of the New York Academy of Sciences 1134, 173–200. Quinlan, M., 2005. Considerations for Collecting Freelists in the Field: Examples from Ethobotany. Field Methods 17, 219–234. Ranlund, A., 2011. Structure and tree diversity of lowland limestone forest on Seram Island, Indonesia (Master thesis). Cirad-Cifor. Ruf, F., Schroth, G., 2004. Chocolate forests and monocultures : a historical review of cocoa growing and its conflicting role in tropical deforestation and forest conservation, in: Agroforestry and Biodiversity Conservation in Tropical Landscapes / Schroth Gotz (ed.), Da Fonseca Gustavo A.B. (ed.), Harvey Celia A. (ed.), Gascon Claude (ed.), Vasconcelos Heraldo L. (ed.), Izac Anne-Marie N. (ed.). Island Press, Washington, pp. 107–134. Sasaoka, M., Laumonier, Y., 2011. The influence of “sago-based vegeculture” on forest

landscapes in c... Smith, J.J., 1993. Using ANTHROPAC 3.5 and a spreadsheet to compute a free list salience index. Cultural Anthropology Methods 5, 1–3. Sodhi, N.S., Koh, L.P., Brook, B.W., Ng, P.K.L., 2004. Southeast Asian biodiversity: an impending disaster. Trends in Ecology & Evolution 19, 654–660. Sunderlin, W.D., Angelsen, A., Belcher, B., Burgers, P., Nasi, R., Santoso, L., Wunder, S., 2005. Livelihoods, forests, and conservation in developing countries: An Overview. World Development 33, 1383–1402. Swallow, B., Boffa, J.M., Scherr, S.J., 2006. The potential for agroforestry to contribute to the conservation and enhancement of landscape biodiversity. World Agroforestry into the future. Nairobi: World Agroforestry Centre (ICRAF) p 95–101. Torquebiau, E., CIRAD, 2007. L’agroforesterie : des arbres et des champs. Harmattan ; CIRAD, Paris; Montpellier. Vallet, A., 2011. REDD+, communities and carbon accounting near the border of a National Park in Seram, Moluccas, Indonesia. Cirad-Cifor. Vogl, C.R., Vogl-Lukasser, B., Puri, R.K., 2004. Tools and Methods for Data Collection in Ethnobotanical Studies of Homegardens. Field Methods 16, 285–306. Weller, S.C., Romney, A.K., 1988. Systematic Data Collection. SAGE. Wiersum, K.F., 1982. Tree gardening and taungya on Java: examples of agroforestry technics in the humid tropics. Agroforestry Systems 53–70. Wiersum, K.F., 1997. Indigenous exploitation and management of tropical forest resources: an evolutionary continuum in forest-people interactions. Agriculture, ecosystems & environment 63, 1–16. Wiersum, K.F., 1997b. From natural forests to tree crops, co-domestication of forests and tree species, an overview. Netherlands journal of agriculture science 425–438. Wiersum, K.F., 2004. Forest gardens as an “intermediate”land-use system in the nature– culture continuum: characteristics and future potential. Agroforestry Systems 61, 123– 134.

Appendixes

Appendix 1: Questionnaire for farmers interviews used on the field

Respondant: Date: Age: Sex:

Attributes of the mixed tree garden Type of garden: *Cocoa *Coconut *Clove *Durian *Nutmeg Location name: Clan: Size: Age: Type of soil: Boundary: *Fence *Planted trees *Wild trees *Others Way to get the garden: *Heritage *Open the forest *Buy *Rent Do you have other gardens? *Yes *No If yes which type of garden? *Cocoa *Coconut *Clove *Durian *Nutmeg

Establishment and management of mixed tree gardens A. Establishment of a new garden Explain step by step, how to make a new garden from opening the forest to plant seeds/seedlings. How long from opening the forest to plant seeds/seedlings? Origin of seeds/seedlings: First fruiting period:

B. Farming practices Usual activities in gardens: *Planting *Clearing *Cutting *Remove weeds *Harvest *Others Utilization of fertilizer: *Yes *No Pests/Diseases: *Yes *No How to remove pests and diseases?

C. Main cash crops production (for one garden) Number of period of harvest per year: Production per period of harvest: Locate of production sale: Selling prices:

D. Local uses of tree species For each tree species present in the gardens, the following questions have been asked: Life form: Local names in bahasa Indonesia: Local names in bahasa Saleman: Local uses: Preparation:

Notes:

Appendix 2: The 29 species known only through their vernacular names not possible to sample.

Vernacular names in Vernacular names in Vernacular names in Vernacular names in Notes Notes bahasa indonesia bahasa Saleman bahasa indonesia bahasa Saleman

Benteng - Tree Kayu Kuning/Mengkudu Ai kuni Tree Buah botol - - Kayu Makila Ai Awarela Tree Bunga Kamboja - - Kayu Manggi manggi Ai Aati Tree Bunga melati - - Kayu Manggi manggi Ai Palun Tree Daun biana Lamiaceae Keladi hutan - Araceae Daun Bobo - Palm Kersen - Tree Daun dilan - - Kisasi - Generic name for epiphytes ferns Daun gaga - - Lemon suanggi Usi hirin Tree Daun Galiji - - Obat gunung - Herbaceous Daun Santang - - Paha - Pandanaceae Daun suplir - - Pohon Galoba hutan - - Daun tikar Buri buri Pandanaceae Rumput Teki - - Jembu kenop - Tree Tali cincin - Liana Kayu Asoer - Tree Tali Rumah Ayaa luman Liana Kayu Kira kira - Tree Tali ruri - Liana

Appendix 3: Local practices of management of the 5 main tree garden species.

Practices to First fruiting period (in Species origin of seeds and/or seedling desease/pests eradicate Fertilizer year) desease/pests seeds harvested on old cocoa trees / seedling Theobroma cacao 2 to 3 yes fire no harvested in gardens

Cocos nucifera seedling from Pulau tujuh (Island near Saleman) 5 to 7 yes fire no

seedling from Est Seram or harvested in gardens / Myristica fragrans 5 to 7 no no seeds harvested on old nutmeg trees -

Syzygium aromaticum Seed harvested on old clove trees 6 to 7 yes fire no

Durio zibethinus seeds harvested on durian trees 15 no - no

Appendix 4: Dominant and most abundant species of each layer. Rdo is the relative dominance and Rde the relative density.

cocoa gardens coconut gardens clove gardens

Most Dominant Most abundant Dominant Most abundant Dominant Rdo Rde Rdo Rde Rdo abundant Rde species species species species species species Theobroma Theobroma Theobroma Theobroma cacao cacao lower layer 2-7m 73% 83% - - - - 65% 66% cacao cacao Syzygium Syzygium aromaticum aromaticum Cerbera Syzygium Syzygium low layer 7,1-15m ------20% 44% 37% floribunda aromaticum aromaticum Cocos medium layer 15,1-25m - - - - 73% Cocos nucifera 82% - nucifera upper layer > 25m ------durian gardens nutmeg gardens

Dominant Most abundant Dominant Most abundant Rdo Rde Rdo Rde species species species species Theobroma Theobroma lower layer 2-7m 82% 56% - - cacao cacao Syzygium Syzygium aromaticum Myristica Myristica low layer 7,1-15m aromaticum 64% Areca catechu 57% 61% 58% fragrans fragrans Areca catechu Lansium domesticum Durio Durio zibethinus zibethinus Myristica Myristica medium layer 15,1-25m 91% Syzygium 70% 49% 66% Syzygium fragrans fragrans aromaticum aromaticum Durio upper layer > 25m 100% Durio zibethinus 100% - - zibethinus

Appendix 5: List of all the inventoried plant species Legend GF: growth form; T: tree; H: herb; L: liana; S:shrub; NU: number of uses Uses: CC: cash crops, SCC: supplementary cash crops, HS: house construction, CaC: canoe contruction, FW: firewood, HF: house furniture, FT: fishing tool, Ro: Rope, SF: staple food, FF: fruit for food, Ve: vegetable, Co: condiment, Dr: drink, Su: sugar, Oi: oil, Me: medicine, Cos: cosmetic, Or: ornamental, Fe: fence, Fer: fertilizer, Pt: protection tree, Ba: basketry, Ri:ritual, Oth: Other.

Vernacular name in "bahasa Vernacular name in Genus species Family GF CU NU indonesia" "bahasa Saleman"

Tree species in mixed tree gardens

Aglaia sapindina Harms Melliaceae Langsa hutan Ai Mahu wapa T FW 1

Alangium javanicum (Blume) Wang. Alangiaceae Kayu gufasa talaga / gufasa Ternate Ai Laharu T HC, FW, HF 3

HC, CaC, FW, HF, FT, Alstonia scholaris (L.) R.Br. Apocynaceae Kayu puli / pule Ai Itun T 6 Me Annona muricata L. Annonaceae Nangka Belanda /Sirsak Duria Arata T SCC, FW, FF, Dr, Me 5 Anthocephalus macrophyllus Havil. Rubiaceae Kayu samama Ai Kimama T HC, CaC, FW 3 Aphanamixis polystachya (Wall.) R.N. Melliaceae Kayu mabuk Ai museni T HC, CaC, FW, HF, Me 5 Parker Aquilaria versteegii H. Hallier Thymelaeaceae Kayu gaharu - T Oth 1 SCC, HC, HF, FF, Me, Areca catechu L. Areacaceae Pinang Hua T 7 Or, Ri HC, HF, Ro, FF, Dr, Arenga pinnata (Wurmb) Merr. Arecaceae Pohon Aren Pohon Gamutu/Ai La T 7 Su, Ri

Sukun Ai Suun HC, CaC, FW, FT, SF, Artocarpua altilis (Parkinson) Fosberg Moraceae T 8 Sukun hutan Ai Ulun FF, Veg, Me Nangka bubur - Artocarpus heterophylla Lam. Moraceae T SCC, FW, FF, Ve 4 Nangka salak - Artocarpus integra Merr. Moraceae Campedak / Campadak Laka Laka T FW, FF 2 Averrhoa carambola L. Oxalidaceae Belimbing manis - A FF 1 Averroha belimbi L. Oxalidaceae Belimbing asam - T FW, FF, Co 3 Barringtonia sp. Lecythidaceae - Ai Hutun T HC, FW, Co, Dr 4

Buchanania arborescens (Blume) Blume Anacardiaceae Jembu Meteng hutan - T FW 1

Buchanania macrocarpa Lautrb. Anacardiaceae Ketapang hutan Ai Kaea T HC, CaC, FW 3 Calophyllum soulatri Burm.f. Clusiaceae Bintanggur/ Bintanggur hitam Ai taun T CaC, FW 2 Cananga odorata Hook.f. & Thoms. Annonaceae Kayu kenangan Ai Sipalin T HC, CaC, HF, FT, Cos 5 SCC, HC, CaC, FW, Canarium oleosum (Lamk) Engl. Burseraceae Kayu kenari / kanari Ai Iyan T 8 HF, FF, Co, Me Caryota rumphiana Mart. Arecaceae Pohon Nibung Koran T HC, HF 2 Cerbera floribunda K. Schum Apocynaceae Kayu mata buta Leka leka T FW 1 Cinnamomum verum J.S.Presl. Lauraceae Kayu manis Ai Mulele T Co 1 Lemon cina Usi sina / Usi timun Citrus aurantifolia Swingle Rutaceae T SCC, FW, FF, Co, Me 6 Lemon nipis/ Jeruk nipis Usi timun Usi Alata / Usi Mulele Lemon manis buah besar Citrus maxima Merr. Rutaceae maina T FF, Co 2 Lemon papeda Usi Kamasi Usi mulele biasa / Usi Citrus nobilis Lour. Rutaceae Lemon manis buah kecil T SCC, FW, FF, Co, Dr 5 Mulele kiiti

SCC, HC, FW, HF, FT, Ro, SP, FF, Ve, Co, Dr, Cocos nucifera L. Arecaceae Pohon kelapa Luin T 18 Su, Oi, Me, Cos, Or, Ba, Ri

Codiaeum variegatum Blume Euphorbiaceae Gadihu - T Or 1 Coffea canephora Pierre var Rubiaceae Kopi - T CC, FW, FF, Dr 4 RobustaChevai Colona scabra Burret Tiliaceae Kayu marong Ai Huluti T HC, CaC, FW, Ro 4 Cordia myxa Forsk. Borraginaceae Kayu semang Ai Alupun T HC, CaC, FW, FT 4 Cryptocarya densiflora Blume Lauraceae Kayu samar putih / kayu samarang Ai rola masahun putih T HC, FW 2 Cynometra cauliflora L. Fabaceae Namu Namu Lambuti T FW, FF 2 Dalbergia sp. Fabaceae Kayu Salawaku - T HC, CaC, FW, Fer 4 Decasperbum bracteatum (Roxb.) A.J. Myrtaceae Kayu merah daun halus Ai musina T FW 1 Scott. Dendrophthoe falcata (L.f) Danser Loranthaceae Benalu Ai Manunpang T Me 1 Diospyros korthalsiana Hiern. Ebenaceae Kayu belo hitam Alametin T HC, FW, HF 3 Diospyros pilosanthera Blanco Ebenaceae SCC, HC, CaC, FW, Durio zibethinus Murr. Bombacaceae Durian - T 7 HF, SF, FF Dysoxylum densiflorum (Blume) Miq. Meliaceae Kayu mabuk Ai museni T HC, CaC, FW, HF, Me 5 Endospermum moluccanum Becc. Euphorbiaceae Kayu Raja/ Kayu Semut Ai Kineri T 0

Erythrina variegata L. Fabaceae Galala/Kayu dadak T HC, Co 2

Ficus capiosa Steud. Moraceae Gohi Kayohiti T Ve 1 Ficus congesta Roxb. Moraceae Pohon Gondal buah kecil Ai Mosu T 0

Ficus lepicarpa Blume Moraceae Sirih Popa Asapulan T HC, FW 2 Ficus septica Reinw.ex Blume Moraceae Asapulan T 0

Ficus subulata Blume Moraceae kayu beringin Waringin/Ai Lulu T FW, Ro, Co 3 Ficus variegata Blume Moraceae Pohon gondal buah besar Ai Mosu T HC, CaC, FW, Me 4 Flacourtia rukam Zoll.& Mor. Flacourtiaceae Tomi Tomi Tom Tom T FW, FF, Co, Or 4 Garcinia mangostana L. Clusiaceae Manggis/Mangustan - T SCC, FW, FF, Me, Cos 5 Garcinia tetrandra Pierre Clusiaceae Manggis hutan/Mangustan hutan Tainuhun T HC, FW, Co 3 Geunsia pentandra (Roxb.) Merr. Verbenaceae Kayu tembakau hutan Ai tabaku T Fw 1 Gliricidia sepium H.B.K. Fabaceae Pohon pelindung /Kayu hidup - T FW, Fe, Fer, Pt 4

Gmelina moluccana Backer ex K.Heyne Verbenaceae Kayu titi Pantai Ai Pain T HC, CaC, FW, HF 4

Gnetum gnemon L. Gnetaceae Ganemon Ai Kirama T FW, Ve 2 Gulubia costata Becc. Arecaceae Pinang hutan Mapua T HF, 1 Hibiscus tiliaceus Linn. Malvaceae Kayu Baru Ai Halun T HC, CaC, FW, Me 4 Homalium foetidum Benth. Flacourtiaceae Kayu samar / Kayu Samarang Ai rola masahun T HC, CaC, FW, HF 4

Horsfieldia bivalvis Merr. Myristicaceae Lobi lobi buah kecil/lobi lobi makan Topi Topi T HC, FW, FF 3

Horsfieldia sylvestris Warb. Myristicaceae Lobi lobi buah besar/Lobi lobi hutan Topi Topi kamuka T HC, CaC, FW 3

Hydnocarpus sp. Flacourtiaceae Gayang hutan Ai Ka T 0

Inocarpus fagiferus (Parkinson) Fosberg. Fabaceae Gayang pantai Ai Ka T HC, FW, FF 3

Intsia bijuga Kuntze Fabaceae Kayu besi Ai Tolati T HC, CaC, FW, HF 4 Jatropha curcas Euphorbiaceae Jarak pagar Ai Latu T FW, Me 2 HC, FW, Ro, Me, Or, Kleinhovia hospita Linn. Sterculiaceae Kayu kinar Ai Lohan T 6 Fe SCC, FW, FF, Me, Lansium domesticum Corr. Meliaceae Langsa Ai Mahu T 6 Cos, Or

Daun Gatal Sinan Laportea decumana Wedd. Urticaceae T Me 1 Kayu daun gatal Ai sinatan Leea indica Merr. Leeaceae - Alapakan berduri T 0

Litsea diversifolia Blume Lauraceae Alpukat hutan - T FF 1 Macaranga hispida Muell. Arg. Euphorbiaceae Hanua Ai Hulua T HC, FW 2 Mallotus ricinoides Muell. Arg. Euphorbiaceae Kayu kapur Ai Losa T Me 1 Mangga Biasa/Pantai/Buah SCC, HC, FW, FF, Dr, Mangifera indica L. Anacardiaceae Ai papalam T 7 Kecil/Isap/Java/ Mangga golek Cos, Or Mangifera laurina Blume Anacardiaceae Mangga hutan/asam Aun T HC, CaC, FW, FF 4 Mangga Kuini Kuin Mangifera odorata Griff. Anacardiaceae T SCC, HC, FF, FW Cos 5 Bacang/Becang Pati Manihot esculenta Crantz. Euphorbiaceae Singkong Kasbi T FW, SF, FF, Ve 4 HC, HF, FT, Ro, SF, Metroxylon sagu Rottb. Arecaceae Pohon sagu Hatan T 8 FF, Or, Fe FT, Ro, SF, FF, Ve, Musa sp. Musaceae Pisang Uri T 8 Me, Or, Ri Mussaenda reinwardtiana Miq. Rubiaceae Kayu Ge/Bunga Kupu kupu - T Oth 1 Myristica fatua Houtt. Myristicaceae Pala hutan buah besar/Pala hutan - T 0

CC, FW, FF, Co, Me, Myristica fragrans Houtt. Myristicaceae Pala - T 6 Or Neonauclea moluccana Merr. Rubiaceae Kayu kasa Kimasi mula T HC, FW, Cos 3 Nephelium lappaceum L. Sapindaceae Rambutan - T SCC, FW, FF 3 Neuburgia moluccana (Scheff.ex Boerl.) Loganiaceae Jati hutan - T HC, FW 2 Leenh. Octomeles moluccana Teijsm. & Binn.ex Datiscaceae Kayu palaka / pulaka Ai Puraran T HC, CaC, FW, HF 4 Hassk Palaquium obovatum Engl. Sapotaceae Kayu siki Ai Sirin T HC, CaC, FW, HF 4 Persea americana Mill. Lauraceae Alpukat T FF, Dr, Me, Cos 4

Pertusadina multifolia (Havil.) C.E. Rubiaceae Kayu nisan Ai Losuti/Ai Mahalu T HC, CaC, FW, FT 4 Ridsdale Piper aduncum L. Piperaceae Sirih Hutan Ai Kamu T 0

Podocarpus neriifolia D.Don Podocarpaceae Kayu Cina - T HC, FW, HF 3 Polyalthia lateriflora King Annonaceae Kayu meranting Ai loren T HC, CaC, FW, HF, Fe 5 Pongamia pinnata Merr. Fabaceae Kayu besi pantai Ai Awasan T HC, FW 2 Premna sp. Verbenaceae Kayu titi hutan Ai Lusan T HC, CaC, FW, HF 4 Psidium guajava Linn. Myrtaceae Giawas/Jembu batu/Jembu biji - T SCC, FW, FF, Dr, Me 5

Pterocarpus indicus Willd. Fabaceae Kayu lenggua Ai Lalan T HC, CaC, FW, HF, Cos 5

Pterocymbium tinctorium (Blanco) Merr. Sterculiaceae Kayu kalabasa hutan Ai Kalabasa apa T HC, FW 2

Salacca zalacca (Gaertn.) Voss Areacaceae Salak - T SCC, FF 2 Semecarpus forstenii Blume Anacardiaceae Ninat Ai Ninati T FW 1 Spondias cytherea Sonnerat Anacardiaceae Kedongdong - T HC, CaC, FW, FF, Me 5 Syzygium aqueum Alston Myrtaceae Jembu Air Poputi T SCC, HC, FW, FF, Me 5

Syzygium aromaticum(L.) Merr. & Perry Myrtaceae Cengkeh - T CC, FW, FF, Co, Me 5

Jembu Makui / Makui talukun

Syzygium malaccense (L.) Merr.& Perry Myrtaceae T SCC, HC, FW, FF, Cos 5 Makui talukun -

Syzygium sp. Myrtaceae Kayu meran Ai musina T HC, CaC, FW, HF 4 Tectona grandis L.f. Verbenaceae Jati mas - T HC, CaC, FW, HF 4

HC, CaC, FW, FF, Me, Terminalia catappa L. Combretaceae Ketapang pantai Ai Tanisa T 6 Cos HC, CaC, FW, Ve, Co, Terminalia microcarpa Decne Combretaceae Kayu alali Ai Alali T 6 Dr Theobroma cacaoL. Sterculiaceae Coklat - T CC, FW, FF 3 Trevesia sundaica Miq. Araliaceae Kayu gurita Ai Kunita T Me 1 Rhus taitensis Guill. Anacardiaceae Belimbing hutan Leusaki T FW 1 Uvaria littoralis Blume Annonaceae kayu kasa Kimasi mula T HC, FW, Cos 3 Vitex cofassus Reinw.ex Blume Verbenaceae Kayu gufasa / gufasa biasa Ai Pasan T HC, CaC, FW, HF, FT 5

Other plant species used by local population

Tree species Aleurites moluccana Willd. Euphorbiaceae Kayu kamiri Ai Hau T HC, FF, Co 3 Annona reticulata L. Annonaceae Buah nona - T FF 1 Annona squamosa L. Annonaceae Sirih Kaya - T FF 1 Artocarpus teysmanni Miq. Moraceae Nangka hutan - T PC 1 Bougainvillea sp. Nyctaginaceae Bugen fil - T Or 1 Carica papaya L. Caricaceae Papaya Uri mahu T FF, Ve, Dr, Me, Cos 5 Casearia grewiaefolia Vent. Flacourtiaceae - Ai Pakat T FW 1 Casuarina equisetifolia Blanco Casuarinaceae Kasawari - T Or 1 Ceiba pentandra (L.) Gaertn. Bombacaceae Kapok hutan Ai Kapusi yapa T Me 1 Chrysophyllum roxburghii G. Don Sapotaceae - Ai lolan T HC, FW 2 Citrus hystrlx OC. Rutaceae Lemon purut Usi matatahi T Co 1 Citrus medica L. Rutaceae Lemon papaya - T Co 1 Clerodendron floribundum R.Br. Verbenaceae Pohon Matel - T Ve 1 Duabanga moluccana Blume Sonneratiaceae Giawas hutan Ai Inotu T HC, CaC 2 Eugenia polyantha Wight Myrtaceae Daun Salam - T Co 1 Fagraea racemosa Jack Loganiaceae Kopi hutan - T FW 1 Ficus sp.1 Moraceae Kayu Ulit Ai Uliti T FW, FF 2 Gastonia serratifolia (Miq.) Araliaceae Kayu patatulan Ai patatulan T FW 1 W.R.Philoson Graptophyllum pictum Griff. Acanthaceae Alipuru - T Or 1 Hydnophytum formicarum Jack Combretaceae Posi posi Ai Uloati T HC 1 Koordersiodendron pinnatum Merrill Anacardiaceae Belimbing hutan Ai Selang T HC, FW 2 Lawsonia inermis L. Lythaceae Pohon pacar Ai Naka T Cos 1 Leea indica Merr. Leeaceae - Alapakan tidak berduri T 0

Macaranga involucrata (Roxb.) Baill. Euphorbiaceae Kayu hurori Ai Hurori T FW 1 Mallotus peltatus Muell. Arg. Euphorbiaceae Kayu sihata Ai Sihata T FW 1 Maranthes corymbosa Blume Chrysobalanaceae Kayu mas - T FW 1 Moringa pterygosperma Gaertn. Moringaceae Kelor - T Ve, Me 2 Ai Teha Palaquium obtusifolium Burck. Sapotaceae Kayu siki T HC 1 Ai tolu Phaleria capitata Jack Thymelaeaceae Pohon Jiba Gaharu buaya T Ro 1 Phyllanthus microcarpus Muell. Arg, Euphorbiaceae Kartok - T Ve 1 Planchonella obovata H.J.Lam Sapotaceae Kayu Hitam Ai Eta T FW 1 Polyalthia rumphii (Blume) Merr. Annonaceae Kayu pelasun - T HC, FW 2 Pometia pinnata J.R.& G.Forst. Sapindaceae Tawang Ai Taan T HC, FW, HF 3 Sauropus androgynus Merrill Euphorbiaceae Kartok - T Ve 1

Scaevola sericea Forst. f. Goodeniaceae Papaceda - T FW, Me 2 Sesbania grandiflora (L.) Poir. Fabaceae Kayu turi Ai Turi T Me 1 Sindora galedupa Prain Fabaceae Kayu minyak - T HC, CaC, FW, HF, Me 5 Sterculia macrophylla Vent. Sterculiaceae Papaya hutan - T HC, HF 2 Syzygium samarangense (Blume) Merr. Myrtaceae Jembu mawar - T FW, FF, Me 3 & Perry HC, FW, FF, Co, Me, Tamarindus indica L. Fabaceae Asam Java - T 6 Cos Syzygium sp.3 Myrtaceae Jembu hutan T FW, FF 2

Agathis dammara(Lamb.) Rich. Araucariaceae Agathis/ pohon damar - T HC, FW, Ve 3 Non tree species Acalypha indica Linn. Euphorbiaceae Daun pus - H Me 1 Adenia heterophylla (Blume) Koord. Passifloraceae Matel maraya - L Ve 1 Adenia sp. Passifloraceae Patola ular - L Ve 1 Aglaonema sp. Araceae Sri Rejeki - H Or 1 Alpinia galanga (L.) Willd. Zingiberaceae Lengkuas merah - H Co, Me 2 Alpinia sp. Zingiberaceae Lengkuas putih - H Me 1 Amaranthus hybridus Linn. Amaranthaceae Bayam - H Ve 1 Anamirta cocculus (L.) Wight & Arn. Menispermaceae Buah bori Simuliti huan L FT, Cos 2 Ananas comosus (L.) Merr. Bromeliaeae Nanas - H FF, Dr, Or 3 Andrographis paniculata Nees. Acanthaceae Sambiloto/Daun Holand - H Me 1 Arachis hypogaea L. Fabaceae Kacang tanah - H Ve 1 Arcangelisia flava Merr. Menispermaceae Tali kuning Ayaa Kuni L Ro, Me 2 Brassica oleracea L. var. capitata L. Brassicaceae Kol - H Ve 1 Brassica rugosa (Roxb) L. Brassicaceae Sawi Sasawi H Ve 1 Cili panjang manisa Capsicum annuum Linn. Solanaceae Cili besar manisa S Co 1 Capsicum frutescens Linn. Solanaceae Cili kecil manisa Casssytha filiformis Linn. Lauraceae Tali putri Ayaan putri L FF, Me 2 Chromolaena odorata (L.) R.M.King & Asteraceae Sunga sunga - H Me 1 H.Robinson Clerodendron fragrans (Vent.) Willd. Verbenaceae Daun pata - H Me 1 Clerodendron villosum Blume Verbenaceae Picahpiring - S Me 1 Colocasia esculenta (L.) Schott Araceae Talas Keladi H SF, FF, Or 3 Commelina bengalensis Forst. Commelinaceae Sambung tulang - H Me 1 Costus speciosus (Koenig) Smith Zingiberaceae Lampuyang - H Co, Cos 2 Crinum asaticum Blanco Amaryllidaceae Bawang laut - H Or 1 Cucumis sativus L. Cucurbitaceae Ketimun/Katimun Papinyu H FF, Ve, Dr 3

Cucurbita moschata Duchesne ex Poir. Cucurbitaceae Labu - H SF, Ve 2 Kunyit hutan/ Kuning hutan/Kuning Kapali Curcuma domestica Valeton Zingiberaceae pawar H Co, Me, Cos 3 Kunyit/Kuning (ditanam) Kuni Cymbopogon citratus Stapf. Poaceae Sare banda - H Cos 1 Cymbopogon nardus Rendle Poaceae Sare - H Co, Cos 2 Cyperus kyllingia Endl. Cyperaceae - Ai Hua gonggong H Cos 1 Derris elliptica Benth. Fabaceae Tali Bori Ayaan minu L FT, Ro, Me 3 Desmodium umbellatum DC. Fabaceae Daun bedak - H Cos 1

Diplazium esculentum (Retzius) Swartz Athyriaceae Paku Paku/Pakis - H Ve, Or 2

Etlingera elatior (Jack) R.M. Smith. Zingiberaceae Galoba Manipa H Co 1 Euphorbia prunifolia Euphorbiaceae Daun Kastarole/Salamake - H Me 1 Hiptage sp. Malpighiaceae Tali Papua - L Ro 1 Imperata cylidrica (L.) Beauv. Poaceae Alang alang / Kusu Kusu Enin tamun H Me 1 Ipomoea aquatica Forsk. Convolvulaceae Kangkung maraya - H Ve 1 Ipomoea batatas Poir. Convolvulaceae Patatas/Ubi Jalar Uhin H SF, FF, Ve 3 Ipomoea pes-caprae (L.) R.Br. Convolvulaceae Katang Katang - H Me 1 Kaempferia galanga Linn. Zingiberaceae Cangkor Kasu Ku H Co, Me 2 Luffa acutangula (L.) Roxb. Cucurbitaceae Patola/Gambas - L Ve 1 Lycopersicon esculentum Solanaceae Tomat Tamati H FF, Ve, Co, Dr, Cos 5 Merremia peltata Merr. Convolvulaceae Tali Hulali/Hulale Hulanin aan L Ro 1 Momordica charantia L. Cucurbitaceae Papari/Pare Kaparia H/L Ve, Me 2 Ai toto kapi kapi/Ai totun Nasturtium indicum DC. Cruciferaceae H Cos 1 makapi Nomaphila stricta (Vahl.) Nees. Acanthaceae Akar burung - H Me 1 Nothopanax scutellarium Merr. Araliaceae Mangkok - H Cos 1 Ocimum basslicum Linn. Lamiaceae Kumangi - H Co 1 Orthosiphon aristatus Miq. Lamiaceae Kumis kucing - H Me 1 Paederia verticillata Blume Rubiaceae Tali konto konto Ai toto otoki L Ro 1 Pandanus polycephalus Lam. Pandanaceae Daun tikar Kihutotun HF 1

Pandanus sp. Pandanaceae Daun Pandan - H Co, Me 2 Peperomia pellucida H.B.& K. Piperaceae daun dingin - H Cos 1 Phaseolus vulgaris L. Fabaceae Bonces/Bunces - H Ve 1 Phyllanthus niruri Linn. Euphorbiaceae Daun cinta cinta Muhu Nalan H Me 1 Physalis minima Linn. Solanaceae Tiplukan - S Me 1 Piper betteL. Piperaceae Sirih Buah Kamu matan L FF, Me, Ri 3 Piper nigrum L. Piperaceae Lada - Co 1

Piper sp. Piperaceae Sirih Lele Kamu laka L Me 1

Pleetranthus rotundifolius (Poir.) Spreng Lamiaceae Kumbili/Ubi Kumbili Esi Ahu H SF, FF 2

Pluchea Indica (L.) Less. Asteraceae Beluntas - H Me 1 Calamoideae Areacaceae Rotan/Tali Rotan Ulesa/Ayaan Ulesa L HF, FT, Ro, Or, Ba 5 Saccharum officinarum L. Poaceae Tebu Tohun H Su 1 Sechium edule (Jacq.) Sw. Cucurbitaceae Labu siam/siang - H Ve 1 Selaginella wildenowii (Desv.) Bakh. Selaginellaceae Soka Soka - H Me 1 Senna alata Roxb. Fabaceae Kupang Kupang - H Me 1 Sida acuta Burm.f. Malvaceae Sapu Sapu Ternate - H Me 1 Solanum melongena Solanaceae Terong/ Aubergine Paloli H VE 1 Stachytarpheta jamaicensis Gard.. Verbenaceae Kakurang - H Me 1 Tecomanthe ternatensis v. Steenis Bignoniaceae Tali bela empat Ayaan Rina rina L Ro 1 Tetracera nordtiana F.Muell. Dilleniaceae Tali api - L Ro 1

Tetrastigma papilosum (Blume) Planch. Vitaceae Tali Saku Ayaa Solun L Ro 1

Tinospora crispa Diels. Menispermaceae Antawali Tali gurita L Me 1 Vanilla mexicana Mill. Orchidaceae Vanili - L CC 1 Vigna radiata (L.) Wilezek Fabaceae Kacang hijau/Kacang tumbuh - H Ve 1

Vigna unguiculata (L.) Walp Fabaceae Kacang panjang H Ve 1

Zea maysL. Poaceae Jagung Alamahu H SF, FF, Ve, Cos 4 Zigiber officinaleRoscoe Zingiberaceae Halia/Jahe Sohi H Co, Me 2 Zingiber sp. Zingiberaceae Tumbuh lawan - H Cos 1

Anredera cordifolia (Ten.) van Steenis Basellaceae Penahong L Me 1

Bambusoidaeae Poaceae Bambu Tipau/Luleba H HF, FT, Ro, Ve, Or, Ba 6

Appendix 6: Selling prices and estimation of one year production of the main cash crops (production/year/garden) Legend: IDR: Indonesia Rupiah

number of production per production per production production cash incomes cash incomes cash incomes cash incomes selling price Production harvest per harvest (kg) harvest (kg) per year (kg) per year (kg) per year (min) per year per year per year per kg year (min) (max) (min) (max) IDR (max) IDR (max) euros (max) euros

Clove 1 120 300 120 300 150000 18000000 45000000 1500 3750 Nutmeg 24 70 48 140 75000 3600000 10500000 300 875 Nutmeg 2 7 20 14 40 140000 "flower" (mace) 1960000 5600000 163 467 Cocoa 4 45 75 180 300 16000 2880000 4800000 240 400

Uses house construction firewood canoe contruction house furniture rope fishing tool staple food fruit vegetable condiment sugar drink

oil

medicine

ornemental

cosmetic

Appendix 7: The 16 uses of plant species by Saleman people, studied through the free listing. 27 individuals have been interviewed (15 men and 12 women) to list all the species they know for the 16 uses. 49

Appendix 8: Mean number of items mentioned by men and women for each use. In comparing the mean list length of men and women for each use, significant difference has been identified for 6 categories of uses. The number of items mentioned by men for the use categories “canoe construction”, “fishing tools” and “sugar” is higher than number of items cited by women whereas women have mentioned more species than men for the 3 uses “vegetable”, “medicine” and “cosmetic”.

house canoe Mean list lenght firewood house furniture rope fishing tools staple food fruit construction contruction Women 7,3 8,1 3,3 5,0 3,8 1,0 4,6 17,1 sd 3,5 5,0 1,7 3,0 1,7 0,9 1,5 6,9 Men 10,5 9,1 5,7 5,7 5,1 2,3 5,1 13,6 sd 6,5 4,0 2,5 3,0 3,2 1,4 1,7 5,4 Student t-test p value 0,210 0,409 0,007 0,534 0,196 0,007 0,512 0,118 ** ** vegetable condiment sugar drink medicine ornement cosmetic Oil mean list lengt women 13,6 8,8 0,9 2,3 10,3 3,8 3,8 1 sd 4,3 4,1 0,8 1,6 4,8 2,0 2,7 0 mean list lenght men 8,9 7,5 1,6 2,0 6,5 3,6 1,1 1 sd 2,6 2,8 0,5 1,6 2,7 1,9 1,3 0 p value 0,004 0,411 0,022 0,640 0,027 0,738 0,006 - ** * * **

Appendix 9: Similarity (%) between species mentioned by men and women for each use during the free listing

100

% 40

Appendix 10: Highly salient, salient and less salient species for each use. Legend HS: highly salient, S: salient, LS: less salient, MTG: mixed tree gardens, SG: small gardens surrounding houses, F: forest.

Important Species for each use category Use Salience Men Women Origin Growth form HS kayu besi kayu gufasa S kayu gufasa House construction kayu lenggua MTG Tree kayu siki kayu palaka LS kayu palaka durian kayu alali HS kayu titi kayu titi pantai Canoe contruction S kayu titi durian MTG Tree kayu titi hutan LS - kayu palaka kayu lenggua kayu lenggua HS kayu besi kayu besi kayu gufasa kayu palaka S kayu gufasa House furniture kayu siki MTG Tree kayu siki kayu titi LS - kayu palaka durian

HS kayu samar kayu besi

kayu besi kayu cengkeh S kayu belo hitam kayu pakat Firewood kayu gufasa kayu sihata MTG (F) Tree

kayu cengkeh Kayu langsat LS durian kayu samar kayu pakat Fishing tools HS Bambu MTG Herbaceous Rotan HS Rotan kayu beringin Liana Rope F (MTG) S Tali hulali Bambu Herbaceous LS Tali bela empat Tali hulali HS Pohon Sagu Herbaceous Staple food S Singkong F/SG (MTG) Tree Keladi Pisang LS Patatas Keladi HS Durian

Langsat Papaya Langsat S Kelapa Fruit for food Pisang MTG (SG) Tree Lemon manis Pisang

Mangga golek LS - Salak Vegetable HS Kangkung SG Herbaceous

Papaya (Tree) S Bayam Singkong Bayam LS Kacang panjang HS Cili Sare Lengkuas S Sare Cili Herbaceous Condiment SG (MTG) Kunyit Cengkeh (Tree) Pala Kunyit LS Cengkeh Kayu manis Lada Halia Lengkuas Daun salam

HS Gamutu Kelapa Drink S Kelapa Gamutu MTG/F Tree LS - HS Gamutu Kelapa Sugar S Kelapa Gamutu MTG/F Tree LS - Oil HS Kelapa Kelapa MTG Tree HS Papaya Kumis kucing

Kumis kucing Daun holand Herbaceous Medicine S MTG/SG kunyit Cengkeh (Tree Liana) Papaya LS Tali kuning Daun cinta cinta HS Mangga biasa Langsat Tree Cosmetic Kunyit MTG S - (Herbaceous) Mangga biasa

HS Kelapa Ornemental Pinang Gadihu MTG Tree S Gadihu Pinang