Smallholder rubber plantation expansion and its impact on local livelihoods, land use and agrobiodiversity, a case study from Daka, Xishuangbanna, southwestern China

Abstract

Crop inventory, botanical surveys, and questionnaire investigations of 60% of households in Daka from 1998 to 2004 (a Hani nationality village in the mountainous region of Xishuangbanna) were undertaken. We focused on virtual extinction of agrobiodiversity in smallholder rubber plantations, which have improved smallholder livelihoods but have affected land-use strategies. Income per capita has increased from US$128.3 in 1998 to US$561.7 in 2004 because of an increase of income from rubber from US$75.8 in 1998 to US$451.4 in 2004. The number of cultivated upland rice varieties decreased from seven in 2001 to one in 2004 because of a sharp increase in rubber price after 2002. Rubber plantations increased from 17.7 ha in 1998 to 82.2 ha in 2004, while swidden fields decreased from 20.4 ha in 1998 to 12.7 ha in 2004. It appears that traditional upland rice production and number of varieties is being seriously eroded by encroachment from rubber plantation. Stabilization of agrobiodiversity loss is necessity while still improving the rural economy.

Keywords:

• agrobiodiversity
• shifting cultivation
• land-use strategy
• smallholder livelihoods
• smallholder rubber plantation
• rubber price fluctuations
• Xishuangbanna

Introduction

Agrobiodiversity is defined as ‘management and direct use of biological species, including all crops, semi-domesticated and wild species’ (Guo et al. 1996). Biodiversity conservation through agrobiodiversity is focused, because most biodiversity exists in human-managed agricultural/forestry ecosystems (Pimented et al. 1992; Brookfield 2001; Bisht et al. 2006). Socioeconomic analyses have increased understanding of incentives necessary to maintain on-farm conservation of agrobiodiversity (Huang et al. 2005; Mai et al. 2006; Sekhar 2007). Internationally, the relationship between agrobiodiversity and socioeconomic development has been extensively explored, e.g., the impact of land conversion on agrobiodiversity (Zimmerer 2004; Xu et al. 2005), as land-use transformation has taken place worldwide (Padoch et al. 1998; Miyamoto 2006). Xishuangbanna is home to unique agricultural systems, crop species, and livestock, which have sustained the indigenous farming communities over time. There can be little doubt that the agrobiodiversity of this region had great potential to provide a fundamental and resilient basis for the regeneration necessary to achieve sustainability of local agriculture and standards of living of the indigenous people. For example, The Hani used to cultivate over 100 upland rice varieties in Mengsong, Xishuangbanna, of which over 30 varieties have recently been conserved (Xu et al. 1997); and 55 species and varieties of cultivated plants on Jinuo swidden lands in Xishuangbanna have been reported (Long et al. 1997). Within a single village, there are still several upland rice varieties grown that are adapted to different soils, traditional cultures, and wealth status (Fu and Chen 2002; Gong et al. 2004). Recently, however, as a result of rapid changes in land use, the agrobiodiversity of Xishuangbanna has been changed through population growth, socioeconomic development, and especially commercial plantation (Tisdell and Zhu 1996; Guo, Padoch, Coffey, et al. 2002; Liu et al. 2006). What local, provincial, national, cross-border, and global forces determine these aspects of agrobiodiversity? How is agrobiodiversity in Xishuangbanna enhanced or adversely affected by indigenous culture and local basic needs, traditional and modern technological innovations, and local, national and international development and policies, e.g. tourism, market access and communication, elimination of shifting cultivation? It is urgent to combine biodiversity inventories with household surveys of resource use technologies and conservation practices of people, together with socioeconomic profiling to analyze conservation incentives for agrobiodiversity by local households. We have studied the threats to agrobiodiversity posed by agroecosystem changes in tropical mountains of Xishuangbanna for some time (Fu et al. 2005).

Rubber plantation has become the most important industry in the economy of Xishuangbanna. Demand for rubber is leading to the loss of the high biodiversity of rain forests in SW China (Li et al. 2007) and has replaced traditional swidden–fallow agroecosystems (Zhang et al. 2001; He 2004; Fu et al. 2005; Xu 2006), undermining their key role in conserving traditional agroecosystems and crop variety diversity (Perreault 2005). This paper aims to determine the status of smallholder rubber plantations and their impact on livelihoods and land-use strategies of smallholders because peasant agriculture is important for conservation of both crops and wild plant resources. We focus on the impact of smallholder rubber plantations on agrobiodiversity in Xishuangbanna through on-farm investigation of Daka, a village of the Hani/Akha nationality, within the larger administrative area of Daka in a mountainous region of Xishuangbanna.

Methods

Study area

Daka is in Menglun township, Mengla County, Xishuangbanna Dai Autonomous Prefecture (Figure 1). The village (N21°41′, E101°25′) is about 10 km from Menglun town. The climate is seasonally tropical, with an average annual temperature of 21.5°C and rainfall average of 1563 mm/year. Daka covers an area of 727 ha; the original vegetation was tropical seasonal rain forest, but land now consists of community forest, rubber plantations, and swidden–fallow. This small village is a rural subsistence community; the most recent census recorded 337 people in 69 families (167 males and 170 females) in 2005.

Figure 1. The study area, Daka in southern Yunnan Province, China.

Field methods

We performed a socioeconomic investigation involving 60% of village households (30 sample households) using an annual questionnaire from 1998 to 2005, with questions on demography, culture, gender, land-use history and scale, distribution patterns and crop yields, produce and labor input, income, and limiting factors, using household-based agrobiodiversity assessment methods (HH-ABA) (Guo, Padoch, Fu, et al. 2002). This survey aimed to monitor the land use and level of agrobiodiversity within households and on their lands. Furthermore, we did complete inventories of variety diversity of upland rice and maize in Daka.

The botanical survey examined 20% of random households in Daka in a plot survey, with the assistance of local households. A total of 27 quadrats (20 × 20 m) in 3-year fallow fields, upland rice fields, and rubber plantations, belonging to nine households, were studied. We undertook on-plot investigation of utilized plants of different agroecosystems as well as rapid inventory of upland crops and varieties.

Data analysis used Simple Linear Correlation (Pearson r) and t-tests of significance of rubber price and upland rice production through SPSS software.

Results

Status of smallholder rubber plantations in Xishuangbanna

Rubber has been grown in southern Yunnan since the 1940s, and production increased in the 1950s on state farms established under the post-liberation regime. Rapid development of rural rubber plantations has provided more stable income for rural households since 1980. By the end of 2004, the rubber plantation area of state farms (51.5%) was 85,143 ha, and rural plantations (35.4%) produced 55,245 t. However, new rubber plantations in Menglun, established between 1988 and 2003, were derived from forested areas (42%) and fallow fields (23%) (Liu et al. 2006). Households in Daka began to cultivate rubber in 1975, and plantations increased from 0.7 ha in 1985 to 160.3 ha in 2005, while swidden–fallow fields declined from 61.3 ha in 1985 to 32.7 ha in 2005.

Smallholder rubber expansion has relied on changes in land policy. Following the national reforms, after 1980, Xishuangbanna abolished communal farming and implemented the household responsibility system. Peasant food production was encouraged and rose to unprecedented levels. At the same time, households could earn cash income through sale of surplus food crops and cash crops, which encouraged diversification of cash crops and varieties. Furthermore, reform in the forestry sector began in 1981, when the State issued its Linyesanding (a regulation governing forest tenure and production to stabilize tenure of hill areas and forests, delimit areas of individually-held hills and forests, and set out responsibility for forest production), immediately followed by the more detailed policy Liangshangyidi (freehold and contracted forestlands and swidden fields) in 1982. Under these reforms, both freehold plots and collectively-held forests with fixed identifiable boundaries were exclusively leased or contracted to individual households (Guo and Padoch 1995). For example, the Linyesanding, demarcated the area of state forestlands and community forestlands as 381,306 ha and 221,462 ha, which were 55.5% and 32.2% of total forestlands, respectively, in Mengla County, where the latter had a standard area of 13.16 ha per family. Furthermore, during the following Liangshanyidi, 69,796 ha of contracted forestlands were demarcated at 4.15 ha per family and 0.73 ha per capita. At the same time, 5988 ha of freehold forestlands and 99,462 ha of swidden fields were demarcated as 0.06 ha per capita of freehold forestland and 1.05 ha per capita of swidden fields. By the end of 1983, there were 137,000 ha, 123,000 ha, and 183,000 ha of contracted forestlands, freehold forestland, and swidden fields demarcated in the whole prefecture. Small rubber plantations were established from swidden and family mountain land that used to be cultivated for fuelwood or fruit. Some rubber was planted on what had been community forest. Part of the collective forest maintained to conserve the catchment for drinking and irrigation water in Daka was cleared for rubber plantation in 1988 (Fu et al. 2005).

Smallholder rubber plantation is usually practiced in agroforestry systems. In Hainan, rubber–tea intercropping is popular and generates higher value than either rubber or tea monoculture under current socioeconomic conditions (Guo et al. 2006). In Xishuangbanna, this system also offers some measure of protection from cold weather damage to the base of the rubber trees. Intercropping provides an important means of raising not only productivity and land-use efficiency of smallholder rubber lands, but also income generation during the unproductive immature phase of the rubber tree (Rodrigo et al. 2001). Agroforestry systems practiced in tropical areas are subject to changes in land management from both external and internal causes; due to increasing population, decreasing land availability, changes in administrative guidance and economic support by government, and market variations (Zeng et al. 2001).

Rubber-related agroforestry systems can be classified into three models: short-term interplanted, middle-term interplanted, and long-term interplanted. The dominant short-term model is for upland rice, maize, and other cereal crops to be interplanted with immature rubber, with random planting of Capsicum frutescens. Pineapple is interplanted with rubber in the middle-term model. In the dominant long-term model, tea and black piper (Piper nigrum) are interplanted with mature rubber, while coffee, Chinese cardamom (Amomum villosum), Gynostemma pentaphylla, and inferior cardamom (Amomum kravanh) are randomly interplanted (Chen et al. 1995). Intercropping species develop and change with the state of the market. Type of intercrop was found to be a significant factor influencing adoption of intercropping of smallholder rubber in Kerala, India (Rajasekharan and Veeraputhran 2002). There are several rubber-related agroforestry systems practiced by local smallholders in Daka. For example, upland rice and taro are interplanted with immature rubber for short-term production, while pineapple and passion fruit are interplanted with immature rubber for middle-term production (Figure 2).

Figure 2. Representative rubber-related agroforestry systems in Daka. Photo by Yongneng Fu.

Effect of smallholder rubber plantation on smallholder livelihoods

Income varies greatly among households of Daka, primarily due to variations in income from rubber. Mean income from rubber per family was US$124 in 1999 (maximum US$313), and US$1028 in 2003, ranging from US$545 to US$1937. In recent years, the amount of rubber plantation has been regarded as a principal indicator of wealth status among households. However, the land distribution system quickly leads to variations in income among households. Land was distributed to each household on the basis of actual household size in 1983, with no changes following death, marriage, or birth of children; so some households now own less land while other households own more (Wei 1997). Households with more land prefer to transfer swidden–fallow into cash crop plantations, especially rubber plantations. However, households with less land still retain swidden–fallow for survival, even though they want cash crop plantations.

Rubber plantations have markedly improved smallholder income in Xishuangbanna, similar to farmers in other regions who have experienced transformation from traditional subsistence agriculture to commercial agriculture based on high-value cash crops. Income per capita increased from US$128 in 1998 to US$561.7 in 2004, while rubber income increased from US$76 in 1998 to US$451 in 2004. Daka's income from rubber increased from 59% in 1998 to 80% in 2004 (Table 1). Some households now wish to give up cereal crop production and buy rice from the market using income from rubber.

Table 1. Income per capita and income proportion of rubber in sampled households in Daka from 1998 to 2004

	Income from rubber (US$)	Income per capita (US$)	Proportion of rubber (%)
1998	76	128	59.2
1999	26	79	32.7
2000	107	167	64.1
2001	98	155	63.5
2002	141	234	60.3
2003	199	298	66.8
2004	451	562	80.2

Effect of smallholder rubber plantations on smallholder land-use strategies

Local land-use strategies, such as subsistence farming and cash crop interaction responses to external and internal pressures, have been affected by global change (Mertz et al. 2005). The increasing scale and speed of cash crop plantation, dominated by rubber, is the most prominent change in land use in Daka, where 80% of sampled households had newly-planted rubber in 1999. Rubber plantations have encroached deeply into the neighboring forest, perennial and annual crop fields, and even into fuelwood plantations of Cassia siamea. Rubber now covers the largest area among all land uses (Table 2). Although most households still have small areas of Chinese cardamom in small forest patches, all smallholders are progressively altering their traditional land-use patterns and labor allocation towards a greater emphasis on cash crops, overwhelmingly rubber plantations.

Table 2. Land-use strategies of 30 sampled smallholders in Daka in 1998 and 2004

		Fallow	Swidden field	Rubber plantation	Paddy	Tea	Fuel	Fruit
1998	Area (ha)	6.2	20.4	17.7	4.2	0.6	0.1	8.0
	Smallholders	6	29	25	27	6	2	19
2004	Area (ha)	6.7	12.7	82.2	7.4	2.9	0.7	1.6
	Smallholders	10	15	29	30	26	8	5

Effect of smallholder rubber plantations on agrobiodiversity

The decline of upland rice varieties in Daka

Mountain-dwelling people need to cultivate several crops in their swidden fields for family consumption as they lack sufficient supply from outside. Elsewhere in Southeast Asia, farmers diversify agricultural production to hedge against the risk of food shortages in subsistence-oriented farming systems (Frei and Becker 2004). Simultaneously, households need rich agrobiodiversity to alleviate risks associated with climate and markets for crops. The different maturation periods of different upland rice varieties is of benefit to avoid the effects of climate disasters on crop yields. Diversity of cash crops and varieties is also of benefit to avoid loss of cash income from market price variations. For example, a sharp price decline for passion fruit since 1999, as well as chilling damage due to exceptionally cold weather in December 1999, had a significant effect on all cash crop plantations of households in Xishuangbanna.

On-farm conservation of upland rice varieties is threatened by the loss of swidden–fallow fields and their replacement by rubber plantations. In a nearby Dai village, Mangdaju, 13 traditional upland rice varieties were once cultivated, but since traditional shifting cultivation has virtually ceased in favor of planting rubber, many households in Mangdajiu, since about 1998, have given up upland rice cultivation (Guo and Fan 2004). Similarly, in another Dai village, Manlun, most upland and even some wet rice varieties have been abandoned (Long and Li 2006). At Daka, four traditional upland rice varieties, Shangaoduo, Jiudifang, Maoxianduo, and Agguoganli, have disappeared. As inventoried, there are still seven upland rice varieties cultivated (Apineiluoqie, Heijieba, Mowangu, Kaopie, Huoniola, Huoniosi, and Huoniole). The remaining upland rice varieties (Heijieba and Apineiluoqie) are usually cultivated in poor soil. However, households plan to cultivate rubber in poor fields instead of upland rice and only leave one plot of fertile soil for upland rice varieties that need fertile soil – Mowanggu. Among the 30 sampled households, there were 24, 7, and 1 who cultivated upland rice varieties in 2001, 2002 and 2003, respectively (Figure 3). There was a negative correlation between upland rice varieties cultivated by local smallholders and global rubber price. For example, rubber price declined from 1997 to 2001, then recovered sharply from 2002 to 2006. Correspondingly, the number of cultivated upland rice varieties increased from two to seven from 1998 to 2001, then decreased from three to one from 2002 to 2004 (Figure 3). The correlation of rubber price, upland rice varieties, and rice sowing area was tested with Simple Linear Correlation and t-tests, showing that rubber price had a significant negative correlation with the number of cultivated traditional upland rice varieties. In addition, the correlation coefficients between rubber price and upland rice varieties was significant (p = 0.01, Table 3).

Table 3. Simple Linear Correlation (Pearson r) and t-tests of significance of rubber price and upland rice production in Daka from 1998 to 2004

	Rubber price	Variety	Rice sowing area
Rubber price	1
Variety	−0.76∗∗	1
Rice sowing area	−0.94∗	0.69∗	1
∗P < 0.05.
∗∗P < 0.01.

Figure 3. Change in rubber price and number of households cultivating upland rice varieties in Daka.

The decline in other upland crops in Daka

Of the 24 other traditional upland crops, many are threatened as a consequence of traditional agroecosystem loss (Fu et al. 2005). Traditional cultivated crops, such as common perilla, gingeli, Coix lachryma-jobi, peanut, and cotton have gradually been abandoned (Fu et al. 2008). In Daka, plantation area of cotton has declined from 3.3 ha in 1985 to 0.5 ha in 2005, and maize, like upland rice, is cultivated within a smaller area. Among field surveys of nine households, there were 12 crops cultivated in upland rice field, and only one household cultivated towel gourd, cotton, sweet potato, maize, grain sorghum, and Coix lachryma-jobi (Table 4). Furthermore, there were six maize varieties in Daka: glutinous types A du du bia with purple kernels, A du du xi with yellow kernels, and A du du pu with white kernels; and non-glutinous types A du du pu with white kernels, A du du xi with yellow kernels, and A du du le with red kernels (Figure 4). In addition, there were two varieties of white gourd (Tao huo and Tao huo san die). Rubber expansion has led to serious erosion of maize, other traditional crops, and upland rice production.

Table 4. Cultivated species in upland rice fields of nine householders in Daka

Scientific name	Local name	Use	Used part	Common name	Household number
Luffa cylindrica	Bie ba	vegetable	fruit	towel gourd	1
Gossypium barbadense	Ca la	fiber	seed	cotton	1
Ipomoea batatas	A na	starch	tuber	sweet potato	1
Coix lachryma-jobi	A qie	ethnomedicine	seed		1
Sorghum vulgare	Xiang lao	cereal	seed	grain sorghum	1
Zea mays	A du	cereal	seed	maize	1
Carya papaya	De ma a ao	fruit	fruit	papaya	2
Arachis hypogaea	Ni piao	vegetable	seed	peanut	2
Colocasia esculenta	Nu ma	starch	tuber	taro	2
Oryza sativa	Qie	cereal	seed	upland rice	6
Hevea brasiliensis	Jia ge a bo	latex	gum	rubber	8
Passiflora altebilobata	Xifanlian	beverage	fruit	passion fruit	9

Figure 4. Traditional cultivated upland crops of maize, white gourd, and peanut in Daka. Photo by Yongneng Fu.

The decline in plant species in fallow fields by rubber plantation

There were 86, 125, and 138 plant species among nine households with rubber plantations, upland rice fields, and 3-year fallow fields, respectively (Table 5), indicating that rubber plantations contain less species than other field types. Among these nine households, there were 49, 51, and 51 utilized species in rubber plantations, upland rice fields, and 3-year fallow fields, respectively. It is reported that rubber agroforestry systems conserve more species than monoculture rubber plantations (Beukema et al. 2007). For example, the species richness indices of rubber interplanted with tea, with pineapple, or in monoculture were 6.5, 6.4, and 5.5 respectively in Daka (Fu et al. 1998).

Table 5. Plant species and crops of different agroecosystems in nine households in Daka

Agroecosystem	Household	Size (m^2)	Number of species	Number of utilized species	Cultivated species/crops
3-year fallow field	Mean	400	33	15	3.4
	Sum	3600	138	51	5
Rubber plantation	Mean	400	23	13	4.1
	Sum	3600	86	49	16
Upland rice field	Mean	400	42	22	3.9
	Sum	3600	125	51	11

Discussion

While there is no doubt that cash cropping is now permanently entrenched in Xishuangbanna, the history of the last 30 years reveals a farming population lacking in confidence in the future of their livelihoods. Price fluctuations have triggered rapid shifts in crop choice, just as they have also triggered abandonment of many old varieties and practices. In Daka in 1998, the cash crop households preferred to plant passion fruit, pomelo (Citrus grandis), common teak, litchi (Litchi chinensis), and Chinese cardamom; however, the price of rubber was declining at that time. Passion fruit cultivation was virtually abandoned by most households since a price decline in 1999. Some smallholders developed passion fruit plantations on poor soils on sloping land. In recent years, the price of rubber has been strong and rising, and China's large expansion in motor vehicle production suggests it may remain good for the foreseeable future. But Xishuangbanna is climatically marginal for rubber, as the 1999 cold snap sharply demonstrated, with a decrease in Daka of 25%. China is now part of the WTO and, at the same time, is negotiating free-trade arrangements with its tropical neighbors in Southeast Asia, who include the world's principal exporters of natural rubber. The new Kunming to Bangkok multi-lane highway, of which the Chinese section is now largely complete, passes through Xishuangbanna, and its presence is a reminder both of latent threats and new opportunities. If the price of rubber weakens, how do households cope with their livelihood activities? Some farmers are already taking precautions. Tea, an ancient crop in Yunnan, has been interplanted with rubber by some households in Daka, not because of agronomic advice from the botanical garden scientists, but simply as a method of risk alleviation (Figure 2).

Daka is within reach of the fast-growing small town of Menglun, and to the growing road network of southern Yunnan. The regional market offers scope for marketing subtropical crops to upland Yunnan, but into an increasingly competitive market. Loss of both natural biodiversity through agricultural expansion, and agrobiodiversity through rapid changes in land use are very substantial and serious in Xishuangbanna. If there is to be a halt to this process, agroforestry must be rebuilt in an economically rewarding manner. But whether it is still possible to maintain traditional swidden–fallow ecosystems and upland rice varieties, as well as other traditional upland crops is unknown. Agroforestry solutions, based on rubber, seem to offer the only feasible short-term solution. It seems clear that stabilization of biodiversity loss must of necessity go hand in hand with stabilization and improvement of the rural economy.

Acknowledgements

The support of United Nations University Project on People, Land Management and Environmental Change is gratefully acknowledged. We are deeply indebted to the many farmers in Daka who shared their time and knowledge with us.