Succession on tin-mined land in Bangka Island

A quantitative study of floristic composition and vegetation structure was conducted at Bangka Island, Indonesia. Six different vegetation types were chosen, riparian forest, abandoned farmland, and natural regeneration of tin-mined lands of different ages: 0 and barren, 7, 11 and 38 years’ old tin-mined land. The seedling composition of the oldest tin-mined land was less than 2 % similar to that of a riparian forest. Natural regeneration on 7-year old tin-mined land began with herb species belonging to Cyperaceae, Poaceae, and Melastomaceae; followed by herb species belonging to Asteraceae and Poaceae on 11-year old; then by Poaceae and shrub species of Myrtaceae on 38-year old tin-mined land. Older tin-mined land tended to have less sand, higher nutrients and cation-exchange capacity. The phosphate solubilizing bacteria showed a gradual increase along with the more newly abandoned tin-mined land but decreased in barren tin-mined land, while the number of arbuscular mycorrhizal fungal spores showed the opposite.


INTRODUCTION
Tin deposits in Bangka Island have been exploited for about a hundred years with increased intensity during the Dutch colonial period.Bangka is the largest tin producing island in Indonesia, contributing 40 % of world demand of tin (ASTIRA 2005).Tin mining leaves disturbed land and ex-tin mining ponds ('kolong'), damages natural drainage and habitats, and causes pollution.Reliance on natural succession to restore sand tin tailings without any human intervention requires a long time, during which the tin tailings remain economically barren (Mitchell 1959, Ang 1994, Elfis 1998).
To obtain a more detailed picture of succession on tin-mined land, especially sand tin tailings in Bangka Island, floristic and species composition, soil properties, and soil flora and fauna of different ages of natural regeneration of tin-mined sites was needed.This study will be useful for the re-establishment of diverse native forests.Dominant among the species planted in Bangka since 1993 are a number of exotic species (Nurtjahya 2001), but ecological caution suggests it is unwise to continue to rely on such a limited species mix for all future rehabilitation efforts (Lamb & Tomlinson 1994).While there is no list of local tree species as candidates for revegetating tin-mined land in Bangka Island, some scholars, however, have suggested some local species (Sambas & Suhardjono 1995, Van Steenis in Whitten et al. 2000).Other suggested sources for species that are suitable for revegetating sand tin tailings are Padang-vegetation (Van Steenis 1932), regarded as degraded heath forest (Whitten et al. 2000); coastal mixed forest with Barringtonia formation (Whitten et al. 2000), and heath forest (MacKinnon et al. 1996).
The aim of this study was to understand the succession on tin-mined land in Bangka Island, in order to develop better criteria for the selection and identification of local species with the potential to revegetate these lands.

Study sites
A quantitative ecological study of tin tailings in the Bangka Regency, Province of Bangka Belitung, Indonesia was carried out from May 2004 to November 2005 on six different types of vegetation (Fig. 1, 2).The study involved field surveys, interviewing locals, and consulting the tin mining map ( PT Tambang Timah 2004), and land use and land system maps (Bakosurtanal 1986, Widagdo et al. 1990).Study plots were a 13 ha riparian forest at Sempan (01°53'38.5"S105°58'14.5"E),a 4-year old 1.6 ha abandoned farmland ('ladang' at Sempan), and natural regeneration of tin-mined lands of different ages: 0-year old 2 ha barren tin-mined land at Riding Panjang (01°59'53.4"S106°06'45.3"E),7-year old 0.5 ha land at Sempan, 11-year old 0.6 ha at Gunung Muda (01°37'0.01"S105°54'47.9"E),and Abstract A quantitative study of floristic composition and vegetation structure was conducted at Bangka Island, Indonesia.Six different vegetation types were chosen, riparian forest, abandoned farmland, and natural regeneration of tin-mined lands of different ages: 0 and barren, 7, 11 and 38 years' old tin-mined land.The seedling composition of the oldest tin-mined land was less than 2 % similar to that of a riparian forest.Natural regeneration on 7-year old tin-mined land began with herb species belonging to Cyperaceae, Poaceae, and Melastomaceae; followed by herb species belonging to Asteraceae and Poaceae on 11-year old; then by Poaceae and shrub species of Myrtaceae on 38-year old tin-mined land.Older tin-mined land tended to have less sand, higher nutrients and cation-exchange capacity.The phosphate solubilizing bacteria showed a gradual increase along with the more newly abandoned tin-mined land but decreased in barren tin-mined land, while the number of arbuscular mycorrhizal fungal spores showed the opposite.38-year old 2 ha at Riau (01°44'33.8"S105°51'66.4"E).All plots located are below 50 m asl.

Data collection and analysis Soil properties, arbuscular mycorrhizal fungi, and phosphate solubilizing bacteria
Soil was sampled at 0 -20 cm and 20 -40 cm depth using an 8 cm auger.Diagonal composite (Setyorini et al. 2003) sample of nine subsamples and single replicate were analyzed.
A composite of eight rhizosphere points of 500 g samples for phosphate solubilizing bacteria (PSB) and arbuscular mycorrhizal fungi (AMF) spores, which were taken at 0 -10 cm and 10-20 cm depth and three replicates, were collected using a modified CSM-BGBD Project protocol ( 2004) under three highest important value index (IVI) domination at each site.Fungal spores were recovered by wet sieving and decanting (Gadermann & Nicolson 1963), and genera were identified (Schenck & Perez 1988;INVAM).Soil solution was seeded in the surface of Picovskaya agar and colonies showing transparent halos around them were counted, and re-isolated and maintained in nutrient agar.

Vegetation analysis
A minimum study plot size of 0.2 ha per study site was determined on basis of the species-area curve (Setiadi & Muhadiono 2001).The study was conducted on 20 contiguous plots of 10 by 10 m at each of the study sites using the modified quadrat sampling technique of Oosting 1956 (Soerianegara & Indrawan  1995).Quadrats of 10 by 10 m were used for measuring trees > 20 cm dbh, and for poles with diameter 10 -20 cm; 5 by 5 m for saplings with height taller than 1.5 m and diameter less than 10 cm; and 1 by 1 m quadrats for seedlings with height less than 1.5 m.For trees and poles the number of individual plants for each species and diameter class was recorded, for saplings and seedlings only the species.Herbarium specimens were collected for identification at Herbarium Bogoriense, Bogor.

Soil properties, arbuscular mycorrhizal fungi, and phosphate solubilizing bacteria
A comparison of soil properties (Table 1) shows that sand fraction in forest at 0-20 cm and 20-40 cm depth (78 and 66 %) and abandoned farmland (47 and 48 %) was less than in the four tin-mined lands (80-97 %).Except for the 11-year old tin-mined land, C/N ratio of tin-mined lands was higher than forest and abandoned farmland.The concentration of P 2 O 5 , K 2 O, K, and Na of undisturbed land generally was higher than of disturbed land, and gradually decreases along with the newly abandoned tin-mined land.The cation-exchange capacity (CEC) of all tin-mined lands was very low (0.4-3.9).
The concentrations of other soil properties showed different relations to each of tin-mined lands.

Study sites
Texture pH Soil organic matter HCl 25% Cation-exchange Barren tin-mined -----------0-20 6.0 2.0 1 land 0 yrs Glomus 100%  Table 2 shows the development of some microbiological parameters.The average number of arbuscular mycorrhizal fungi (AMF) spores per 50 g soil per dominant plant species in 0-20 cm in forest soil (15.0) was less than that of abandoned farmland (45.3) and the number of spores in 38-year old tin-mined land was the highest (261.7).The number of AMF spores increases strongly with the period of abandonment of tin-mined land, with lowest values in 0-year old barren tin-mined land.The number of genera was between three and five, and Glomus Tul.& C.Tul.
(Glomaceae) was dominant (44 -100 %).In contrast to AMF, the average number of phosphate solubilizing bacteria (PSB) in tin-mined lands showed a slight increase at 7 years, and a gradual decrease after that.The average number of PSB per g soil in 0 -20 cm in forest soil (4.4 10 5 ) was less than that of abandoned farmland (12.7 10 5 ) and the smallest number of spores was found in 38-year old tin-mined land (1.3 10 5 ).

Floristic composition
The number of individuals, species, and families was highest in riparian forest, and gradually decreased in abandoned farmland, 38-year old tin-mined land, 11-year old tin-mined land, 7-year old tin-mined land (Table 3).Trees (11 species) and poles were only found in forest (24 species) and abandoned farmland (4 species).Saplings were found at all study sites except on the 7-year old tin-mined land.Similarities between study sites with different regeneration phases could be assessed only on basis of seedlings (Table 4), and show that similarity is generally low (less than 28 %), and that the similarity index (seedlings) of the oldest tin-mined land and that of riparian forest is still only 1.5 %.
A comparison of the development of structural properties and diversity of the vegetation is presented in Table 5.

Species composition (see Appendix)
Species

DISCUSSION
It appears that the combination of high sand component, and low macronutrients, CEC, and soil organic matter in the tinmined lands determine the floristic composition and species composition.Older tin-mined lands were shown to be more fertile than younger ones, and to be richer floristically and in species composition.From the PSB and AMF point of view, it appears that the presence of PSB in tin-mined lands reflects phosphate concentration availability in soil.In contrast to PSB, the number of AMF spores in tin-mined lands was likely supported by the soil fertility and its vegetation development above ground.As the availability of phosphate in the soil is low, root hairs and spore production seemed to increase.In general, with increasing soil fertility, PSB population increased while the number of AMF spores tended to decrease.The severe conditions in barren tin-mined land supported only one Glomus species.
Concentration and distribution of species in disturbed study sites was higher than for undisturbed sites because the number of species and the number of plants at disturbed sites were less.Therefore, the species richness of disturbed sites was poorer than for undisturbed sites.The floristic composition of older abandoned tin-mined lands is gradually more similar to the site with forest, but, similarity of forest to 38-year old tin-mined land is still low.The high dissimilarity among study sites confirms that succession takes a long time and supports the projection in Singkep Island (Elfis 1998).
Succession is initiated by herb species, followed by shrub species, but the time required for poles and finally trees to appear apparently exceeds the 38 years covered in this study.The slow succession reflects the long period of time needed to increase soil fertility, accumulate organic material and so improve the quality of the soils' microclimate.During this time, apparently also minerals are added, reducing the sand and increasing the silt and clay component.Much older tin-mined lands are needed to present a more comprehensive conclusion.Unfortunately, illegal mining and illegal re-mining in revegetated tinmined lands (Bangka Pos 2004) make these difficult to find.

Table 4
Number of individuals, species, and families in study sites, per growth phase.Similarity indices between different study site, per growth phase.

Table 1
Soil properties of study sites.

Table 2
Status of phosphate solubilizing bacteria (PSB) and arbuscular mycorrhizal fungi (AMF) in study sites.

Table 5
Structure indices for study sites, per growth phase.
To accelerate succession, potential species candidates can be identified, starting with herb species belonging to Cyperaceae and Poaceae, or shrub species belonging to Melastomataceae and Myrtaceae.Floristic composition and Importance value index of 0-year, 7-year, 11-year, 38-year, abandoned farmed land, and forest.Tristaniopsis whiteana (Griff.)Peter G.Wilson & J.T.Waterh.