Conservation management of rare and predominantly selfing tropical trees: an example using Hopea bilitonensis (Dipterocarpaceae) Soon Leong Lee, Lillian S. L. Chua, Kevin K. S. Ng, Mamat Hamidah, Chai Ting Lee, Chin Hong Ng, Lee Hong Tnah & Lay Thong Hong Biodiversity and Conservation ISSN 0960-3115 Biodivers Conserv DOI 10.1007/s10531-013-0566-5 1 23 Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media Dordrecht. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”. 1 23 Author's personal copy Biodivers Conserv DOI 10.1007/s10531-013-0566-5 ORIGINAL PAPER Conservation management of rare and predominantly selfing tropical trees: an example using Hopea bilitonensis (Dipterocarpaceae) Soon Leong Lee • Lillian S. L. Chua • Kevin K. S. Ng Mamat Hamidah • Chai Ting Lee • Chin Hong Ng • Lee Hong Tnah • Lay Thong Hong • Received: 29 July 2013 / Accepted: 19 September 2013 Ó Springer Science+Business Media Dordrecht 2013 Abstract Hopea bilitonensis is an extremely rare and predominantly selfing dipterocarp in Peninsular Malaysia. A comprehensive research was initiated to assess the ecological genetics of H. bilitonensis to elucidate specific ecological and genetic requirements and subsequently to recommend conservation strategies. The objective for conservation of a rare plant such as H. bilitonensis differs from that of a common plant. For common plants, the conservation strategies are to prevent the species from becoming endangered. In contrast, for rare plants, the final race against extinction is being fought. Tropical forests are rich in plant species diversity and obtaining adequate knowledge to set conservation strategies for the majority of these species might be difficult. Thus, it is suggested that for the conservation of tree species, the species can be grouped according to their life history traits. The information generated for a species can then be adapted to species that have S. L. Lee (&)  L. S. L. Chua  K. K. S. Ng  M. Hamidah  C. T. Lee  C. H. Ng  L. H. Tnah Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia e-mail: leesl@frim.gov.my L. S. L. Chua e-mail: lilian@frim.gov.my K. K. S. Ng e-mail: kevin@frim.gov.my M. Hamidah e-mail: hamidah@frim.gov.my C. T. Lee e-mail: leechait@frim.gov.my C. H. Ng e-mail: chinhong@frim.gov.my L. H. Tnah e-mail: leehong@frim.gov.my L. T. Hong Regional Office for Asia, the Pacific and Oceania, Bioversity International, P.O. Box 236, UPM Post Office, 43400 Serdang, Selangor, Malaysia e-mail: l.hong@cgiar.org 123 Author's personal copy Biodivers Conserv similar types of life history traits. We have recently generated the ecological genetics information for a rare and predominantly outcrossed dipterocarp (Shorea lumutensis). This study on H. bilitonensis will provide ecological genetics information for the conservation of rare and predominantly selfing dipterocarps. Keywords Genetic diversity  Demography  Microsatellites  Life history traits  Mating system  Critically endangered plant Introduction The fate of many rare plant species is increasingly uncertain due to our overall ignorance of the biology of the species. To conserve a rare plant species, conservation programs must be guided by the biology of the species. We cannot conserve effectively what we do not understand. The augmented extinction of rare and endangered plant species has led to concerns for their viability and requires more comprehensive conservation strategies and efforts to prevent further loss of biodiversity, especially in the tropics. Ecological interactions between plants and their environment can influence population growth rates via their effects on fecundity, growth, or survivorship of individuals (Blundell and Peart 2001; Peters 2003). Hence, characterizing the habitat requirement of a species is critical to sound conservation practices (Simberloff 1988; Brussard 1991). In addition, genetic aspects of rarity should also be given attention because the long-term survival of a species is eventually associated with the genetic diversity available to a particular species. Levels of genetic diversity determine the evolutionary potential, adaptability and fitness of a species in response to environmental changes (Lande and Barrowclough 1987; Huenneke 1991). Loss of genetic diversity is likely to decrease the ability of a species to respond to environmental changes and will potentially discard biological information useful to human. Mating patterns in most flowering plants are governed by complex interactions between reproductive traits and the ecology of populations (Holsinger 1996). The mating parameter with the largest influence on genetic structure is the selfing rate, the proportion of mating that result from self-fertilization (Barrett and Kohn 1991). Selfing has direct genetic consequences, including its effect on the intensity of inbreeding depression (Charlesworth and Charlesworth 1987) and the partitioning of genetic diversity within and among populations (Hamrick and Godt 1989). Between-population variation in genetic diversity tends to be much higher in selfing plants compared to outcrossing plants and the total genetic diversity across an entire species tends to be higher in perennial outcrossers compared to annual selfers (Hamrick and Godt 1989). In terms of conservation, for highly inbred species where population differentiation is great and within-stand genetic diversity is low, conserving many reserves would be better than a few. On the other hand, for species that are typically highly outcrossed and where within-stand genetic diversity is high, large reserves would be better to maintain the important component of within-stand variability. The objective for conservation of a rare species differs from that of a common species. For a common plant, the conservation strategies are to prevent the species from becoming endangered. In contrast, for a rare plant, the final race against extinction is being fought. Tropical forests are rich in plant species diversity. For example, an area of 50 ha in the Pasoh Forest Reserve, Peninsular Malaysia, was reported to consist of 814 different tree species (Kochummen 1997). For the majority of these species, obtaining adequate knowledge for conservation strategy recommendation might be challenging. One of the 123 Author's personal copy Biodivers Conserv approaches suggested for tree species conservation is to group species according to their life history traits (Lee et al. 2000). The information generated for a species then can be adapted to species that have similar types of life history traits. Accordingly, we have recently used ecological genetics information for a rare and predominantly outcrossed dipterocarp (Shorea lumutensis; Lee et al. 2006) to set conservation strategies to protect rare species against extinction. As an extension, in this paper, we present the ecological genetics information for Hopea bilitonensis Ashton, a rare but predominantly selfing dipterocarp in Malaysia. Hopea bilitonensis (Dipterocarpaceae) is a small smooth-barked tree species with stilt roots. It is locally common on the sandy islands of Banka and Billiton in East Sumatra but was only once recorded by Ogata (KEP 110201; 15 February 1968; Gunung Gajah, Kampar, Perak) on a limestone forest in the central northwest of Peninsular Malaysia (Ashton 1982). Botanical descriptions and geographical distribution of H. bilitonensis, presented by collections up to 1968 and given in Ashton (1982), provide scarce information on the ecology, genetic composition and breeding system of this species. The species is categorized as critically endangered according to the IUCN Red List of Threatened Species (CR A1c?2c, B1?2c; version 2.3, IUCN 1994) and Malaysia Plant Red List (CR A4c, B1ab[iii], Chua et al. 2010). The specific objectives of this study are: (1) to outline spatial distribution, demographic structure, seed germination behavior and threats to the species in Peninsular Malaysia; (2) to assess the levels of genetic diversity, spatial genetic structure and population genetic structure of the species in Peninsular Malaysia; (3) to characterize the mating system of the species; and (4) to provide an integrated conservation strategy for a rare and predominantly selfing species using ecological genetics approach. Materials and methods Population survey Population surveys were conducted at the Tempurung limestone massif in the district of Kinta in the state of Perak, Peninsular Malaysia (Fig. 1). Large parts of the massif are forested but the entire outcrop lies in the State Land. Forested areas under State Land are managed by the District Land and Forest Offices and are under the jurisdiction of the State Government. Two populations were spotted (Fig. 1). The first population is sited at Gunung Gajah (Latitude: N 4°25.7580 , Longitude: E 101°11.5380 , altitude: *250 m at sea level), at the southwestern tip of the massif while the second population is on a ridge leading to the summit of Gua Tempurung (Latitude: N 4°25.7560 , Longitude: E 101°11.5600 , altitude: *280 m at sea level). The occurrence of H. bilitonensis at Gua Tempurung is a new locality record for Peninsular Malaysia. Population mapping Ground station points (STN), marked by polyvinyl carbonate pipes were placed on the slopes. Distances between STNs varied depending on the longest clear line of sight available. Tree positions were mapped from selected STNs, using tree-to-tree and array mapping methods. Impulse 200 (Laser Technology Inc.), a laser instrument which measures distance, height, inclination and azimuth, was used to map population boundary and tree position on the ground while a global positioning system (GPS) instrument (Garmin Etrex Summit) was used to determine coordinates for population boundary. The STNs, 123 Author's personal copy Biodivers Conserv Fig. 1 Locations of Gunung Gajah and Gua Tempurung within the Tempurung massif tree-to-tree and array data were entered into Roadeng software (Ver. 3.1), which then generated a tree position map. This Roadeng-generated map was converted to .dxf format. The population boundary was then tied to the GPS coordinates using Autocad Map 2000i software. The GPS points, population boundary and tree position were also plotted using Arc View 3.3 software to determine the coordinates of individual trees for spatial studies. Demographic structure and spatial distribution The populations were enumerated using standard method of tagging and diameter at breast height (dbh) was measured at 1.4 m height. The demographic structure of the species was examined by assigning individuals to one of five size classes (dbh): 1.0–4.9, 5.0–9.9, 10.0–14.9, 15.0–19.9, and [20 cm, and fitted to inverse J-shaped curve, the shape distribution of natural tree populations with abundant regeneration (Condit et al. 1998). As the terrain of Gua Tempurung is extremely demanding, detailed mapping was done only for the Gunung Gajah population. Hence, the spatial distribution was only tested on Gunung Gajah population at two size classes (1–10 and [10 cm), for clumping and over-dispersal using univariate second-order spatial pattern analysis, based on Ripley’s (1976) K-function. Four continuous distance classes, each of 5 m, were considered, from 0–5 to 15–20 m. Confidence limits were estimated using the bootstrap method as explained in the program Spatial Point Pattern Analysis (Haase 1995); the location of individuals was randomized in 19 Monte-Carlo trials to determine a 95 % confidence interval within each 5-m distance. Phenological observation and germination study Phenological observations were carried out using binocular from January 2003 to December 2005. For germination study, six seed batches were collected from Gunung 123 Author's personal copy Biodivers Conserv Gajah and Gua Tempurung during the fruiting event in October 2005. The seeds were sown in a medium consisting of forest topsoil and sand in the ratio of 3:1 in black perforated polythene bags. Seed were assumed to have germinated when the radicle was [1 cm in length. Sample collections and DNA extraction For genetic diversity studies, a total of 77 and 100 individuals of H. bilitonensis were collected from Gunung Gajah and Gua Tempurung, respectively. For mating system study, seeds were directly collected from eight randomly selected trees evenly distributed within Gunung Gajah. The genomic DNA was extracted from leaf tissues and seeds using the procedure of Murray and Thompson (1980) with modification, and further purified using High Pure PCR Template Preparation Kit (Roche Diagnostics). Microsatellite analysis The samples were genotyped for 13 microsatellite loci, developed for H. bilitonensis, i.e. Hbi016, Hbi019, Hbi022, Hbi055, Hbi116, Hbi160, Hbi161, Hbi221, Hbi247, Hbi303a, Hbi316, Hbi325a and Hbi329 (Lee et al. 2004). Microsatellite amplifications were performed in a 10-ll reaction volume, containing 10 ng DNA, 50 mM KCl, 20 mM Tris–HCl (pH 8.0), 1.5 mM MgCl2, 0.2 lM of each primer, 0.2 mM of dNTP mix (Promega), and 1 U of Taq DNA polymerase (Promega). The reaction was subjected to amplification on a GeneAmp 9700 thermal cycler (Applied Biosystems), for an initial denaturing step at 94 °C for 4 min, followed by 40 cycles each at 94 °C for 1 min, 46–58 °C annealing temperature for 30 s, and 72 °C for 45 s. A final extension step at 72 °C for 30 min was performed after the 40 cycles. Genotyping was done on 5 % denaturing (6 M urea) polyacrylamide gels. Electrophoresis was carried out with 19 Tris–borate–EDTA (TBE) buffer on an ABI Prism 377 automated DNA sequencer (Applied Biosystems). Allele sizes were scored against the internal size standard and the individuals were genotyped using GeneScan Analysis 3.1 and Genotyper 2.1 software (Applied Biosystems). Allelic frequencies were determined for each locus in each population. Based on these data, the following levels of genetic diversity were estimated: average number of alleles per locus (Aa), allelic richness (Rs; Petit et al. 1998), gene diversity (He; Nei 1987) and fixation index (Fis; Nei 1987). The significant positive or negative of Fis was tested using 520 randomization (default parameter in FSTAT; Goudet 2002) for each locus and across loci for each population. Genetic structure within population was analysed using the Moran’s I coefficient (Sokal and Oden 1978). The spatial distribution of alleles was tested at two size classes (1–10 and [10 cm) for five distance intervals, each of 5 m, from 0 to 30 m. Significant deviation from random spatial distribution at 95 % confidence interval was tested using Monte Carlo simulations (1,000 permutations) as explained in the program spatial genetic structure (Degen et al. 2001). Genetic structure between populations was quantified using R-statistics (Rst; Slatkin 1995), an analogue of Fst developed for microsatellite loci under the assumption of a stepwise mutation model, which is likely at many microsatellite loci (Jarne and Lagoda 1996). Permutation tests (10,000 permutations, randomizing alleles) were used to test whether the estimates of Rst were significantly greater than zero. The significance value of Rst was corrected for multiple tests using the sequential Bonferroni test (Rice 1989). 123 Author's personal copy Biodivers Conserv The genetic relatedness among individuals was quantified using DSA shared allele distance (Chakraborty and Jin 1993) and cluster analysis on shared allele distances via the neighbour-joining method (Saitou and Nei 1987). Relative strength of the nodes was determined using bootstrapping analysis (1,000 replicates) applying the program PowerMarker (Liu and Muse 2005). The minimum population size to maintain current levels of genetic diversity was estimated according to Lee et al. (2002). The genotype data of the collected samples from Gunung Gajah and Gua Tempurung populations were pooled (total number of samples was 177) for simulation analysis. To determine the minimum population size required for maintaining the total number of alleles (At), 170 of the 177 samples were sampled without replacement for 1,000 times using a computerized algorithm. The At was calculated. The At was also estimated for sample sizes of 170–10, with reduction of ten samples for each interval. The percentage mean At with the confidence envelops were plotted against the sample sizes to reveal trends. The multilocus population outcrossing rate (tm) and the average single-locus population outcrossing rates (ts) were estimated from seed parent genotypes based on six microsatellite loci using MLTR program (Ritland 2002). The standard errors of tm, ts, and (tm - ts) were estimated using 250 bootstraps with the maternal family as unit of resampling. Pollen and ovule gene frequencies were estimated jointly with the outcrossing rate using the Newton–Raphson numerical iteration method. If mating occurs between relatives (biparental inbreeding), some outcrossing events would be confounded with selfing events. The difference (tm - ts) is an estimate of minimal fraction of apparent selfing events due to biparental inbreeding. Results Demographic structure and spatial distribution Generally, the forest structure at Tempurung massif had two strata, i.e. ground and understorey, with a dwarf main canopy overhead. The vegetation became increasingly sparse as the elevation increased. At the foothills and lower slopes, small-sized trees of up to c.40 cm dbh of Shorea glauca (Dipterocarpaceae), Vatica harmandiana (Dipterocarpaceae), Maranthes corymbosa (Chrysobalanaceae), Paranephelium macrophyllum (Sapindaceae), Mallotus brevipetiolatus (Euphorbiaceae) and Diospyros retrofacta (Ebenaceae) were common while Streblus ilicifolia (Moraceae) occupied the understorey. At the upper slopes, these were replaced with Isonandra perakensis var. perakensis (Sapotaceae), Pouteria obovata (Sapotaceae), Vitex siamica (Verbenaceae), Pistacia malayana (Anacardiaceae), Homalium kunstleri (Flacourtiaceae) and Cleistanthus gracilis (Euphorbiaceae). Paraboea verticillata (Gesneriaceae), a limestone herbaceous endemic, was restricted to exposed slopes at c.350 m altitude. Vegetation along the upper slopes leading to the ridge was comparatively sparser and dominated by Pandanus sp. (Pandanaceae). Throughout the trail, S. glauca and V. harmandiana were frequently encountered. At Gunung Gajah, 77 trees of 1-cm dbh and greater were found on the slopes of the outcrop, occupying an approximate area of about 0.28 ha. The population structure showed a typical inverted-J shape curve (Fig. 2a), indicating abundance of regenerations. A total of 54.5 and 72.7 % of the population had a diameter of less than 5 and 10 cm dbh, respectively. The largest diameter was 23.4 cm (tree no. 30). Figure 2a also indicates that the population has a rather even stand structure with a very small proportion of trees (6.5 %) in the largest diameter class. 123 Author's personal copy Biodivers Conserv Proportion (%) A 60 50 40 30 20 10 0 1.0-4.9 5.0-9.9 10.0-14.9 15.0-19.9 20.0-24.9 dbh classes (cm) Proportion (%) B 60 50 40 30 20 10 0 1.0-4.9 5.0-9.9 10.0-14.9 15.0-19.9 20.0-24.9 dbh classes Fig. 2 Distributions by diameter (dbh) size classes of H. bilitonensis trees 1 cm and above in a Gunung Gajah and b Gua Tempurung A total of 243 trees of 1-cm dbh and greater were found on the slopes of Gua Tempurung, occupying an approximate area of about 0.90 ha. Unlike Gunung Gajah, the population structure here had a skewed normal curve, indicating that mortality rates for trees in the two smallest diameter classes are disproportionately different (Fig. 2b). Coppicing is a common occurrence in the population; at least 19.3 % of trees coppiced and such trees were found in all diameter classes except the largest class; 53 % of trees in the diameter class 5.0–9.9 cm coppiced followed by 32 % in the diameter class 10.0–14.9 cm. The spatial distribution pattern analysis of trees in Gunung Gajah at a range of 0–20 m and 95 % probability indicates a highly significant spatial clustering of trees at all distances (Fig. 3). The observed Ĺ(t) - t lies well above the expected value of 0 and above the upper 95 % quantile for all distances (Fig. 3a). The magnitude of this deviation fluctuates but remains large for all distances. High levels of significance and observed Ĺ(t) - t were also recorded for trees grouped into various size classes (Fig. 3b, c). In all cases, trees were confined to within 20 m from the baseline. Phenological observation and germination Phenological observations from January 2003 to December 2005 noticed fruiting events in 2003 and 2005. Dried fruits were available on the ground in September 2003 (see Chan FRI 46675) and mature fruits in February 2005 (Chan FRI 36169). The shortest interval 123 Author's personal copy Biodivers Conserv A 4 3 (t)-t 2 1 0 0 5 10 0 5 10 0 5 15 20 -1 -2 B 4 3 (t)-t 2 1 0 -1 15 20 -2 C 6 4 (t)-t 2 0 10 15 20 -2 -4 -6 t (m) Fig. 3 Univariate second-order spatial pattern analysis of trees in Gunung Gajah using Ripley’s K function for a all the H. bilitonensis trees; b trees below 10 cm dbh; c trees above 10 cm dbh. Continuous lines represent the sample statistics Ĺ(t) - t and the dotted lines the 95 % confidence envelope for t = 0–20 m period between fruiting events is likely to be from one and the half to two years. During the fruiting event in 2005, 12 trees in Gunung Gajah and five trees in Gua Tempurung flowered. Of these 17 trees, five had a dbh less than 10 cm and the smallest diameter tree that flowered was at 4.8 cm dbh. Based on this fruiting episode, we can assume that trees above 4.8 cm dbh are reproductively mature. The mean germination percentage for six seed batches was 79.4 ± 9.73 with 88 % being the highest (Fig. 4). The number of days required to reach peak germination was nine. For some seed batches, germination had occurred prior to sowing; dipterocarp seeds are recalcitrant and have been observed to germinate in collection bags. Only one batch, i.e. 2005–0356 (Pk 137) had delayed germination. The germination results are within the range of estimates reported for other dipterocarp species (Ng and Mat Asri 1991). 123 Author's personal copy Biodivers Conserv 2005-0351 (Pk 2)* 2005-0353 (Pk 137A) 2005-0355 (Pk 4)* 2005-0352 (Pk 3)* 2005-0354 (Pk 148) 2005-0356 (Pk 137) 100 % germination 80 60 40 20 0 0* 2 4 6 8 10 12 14 days Fig. 4 Germination of H. bilitonensis seed batches from Gunung Gajah* and Gua Tempurung Genetic diversity The number of alleles observed at each locus ranged from two (Hbi016 and Hbi022) to 11 (Hbi316) for Gunung Gajah and three (Hbi160 and Hbi221) to eight (Hbi316) for Gua Tempurung (data not shown). The gene diversity (He) was found to be higher in the Gunung Gajah (0.65) compared to Gua Tempurung (0.57) (Table 1). However, the mean number of alleles per locus (Aa) and the allelic richness (Rs) showed rather similar values in Gunung Gajah (5.0 and 4.90, respectively) and Gua Tempurung (4.9 and 4.68, respectively). The fixation indices (Fis), calculated for all loci in each population showed significantly positive or negative in nine loci (Hbi055, Hbi116, Hbi161, Hbi221, Hbi247, Hbi303a, Hbi316, Hbi325a and Hbi329) at Gunung Gajah and 10 loci (Hbi019, Hbi055, Hbi116, Hbi161, Hbi221, Hbi247, Hbi303a, Hbi316, Hbi325a and Hbi329) at Gua Tempurung (data not shown). Across loci, significant positive value of Fis (0.271; p \ 0.05) was observed in Gua Tempurung, an indication of excess of homozygotes (Table 1). Although negative Fis was detected in Gunung Gajah, this was not significantly different from zero. The spatial distribution of alleles study showed significant spatial genetic structure at the distance of 0–10 m (Fig. 5a). However, when the trees were grouped into various size Table 1 Genetic diversity parameters (Aa, Rs and He) and fixation index (Fis) values of two natural populations of H. bilitonensis (Gunung Gajah and Gua Tempurung) based on 13 microsatellite loci Population Gunung Gajah Gua Tempurung Mean Sample size Aa Rs He Fis -0.025 77 5.0 (2.6) 4.90 (2.47) 0.65 (0.04) 100 4.9 (1.6) 4.68 (1.57) 0.57 (0.02) 0.271* 89 5.0 (0.1) 4.79 (0.16) 0.61 (0.06) 0.123 Values in parentheses are standard deviations and * indicates value significantly greater than zero (p \ 0.05) 123 Author's personal copy Biodivers Conserv classes, significant spatial genetic structure was only observed for trees below 10 cm dbh at the distance of 0–5 m (Fig. 5b) but not on trees above 10 cm dbh (Fig. 5c). The population genetic structure study revealed high levels of differentiation between the two populations. The Rst values ranged from 0.001 (Hbi316) to 0.441 (Hbi303a), with a mean A 0.1 0.08 Moran's I 0.06 0.04 0.02 0 -0.02 -0.04 -0.06 -0.08 5 10 15 20 25 30 25 30 25 30 Distance class Moran's I B 0.12 0.1 0.08 0.06 0.04 0.02 0 -0.02 -0.04 -0.06 -0.08 5 10 15 20 Distance class C 0.3 Moran's I 0.2 0.1 0 -0.1 -0.2 -0.3 5 10 15 20 Distance class Fig. 5 Correlograms of average Moran’s I coefficients of trees in Gunung Gajah for a all the H. bilitonensis trees; b trees below 10 cm dbh; c trees above 10 cm dbh. Distance classes were defined at five intervals, each of 5 m, from 0 to 30 m. Continuous lines represent the sample statistics and dotted lines represent 95 % envelopes of average I distribution after 1,000 permutations of individual multi-genotypes within each diameter class 123 Author's personal copy Biodivers Conserv of 0.116. This indicated that 11.6 % of the total genetic diversity was distributed between the two populations. The cluster analysis among individuals formed two common genetic clusters which clearly divided the individuals according to population (except GT069, which belongs to Gua Tempurung was grouped with Gunung Gajah cluster; Fig. 6). Within each population, the individuals were further divided into various small clusters. The minimum population size to maintain current levels of genetic diversity (number of alleles) is shown in Fig. 7. The basic relationship between At with sample size was logarithmic. To maintain 95 % of alleles, 107 individuals are required. Mating system The mating system study showed that H. bilitonensis from Gua Gajah reproduced mainly through selfing, with *96 % of the seeds produced from selfing (Table 2). The mean G G T0 G T0 39 G T02 59 GT T06 2 GT 09 3 GT 070 3 GT 067 GT 045 GT 068 GT 079 0 GT 76 0 G 0 57 T GT0 31 6 GT0 1 60 G T 0 0 GT07 6 4 GT07 5 GT082 GT081 GT044 GT041 GT042 GT077 GT038 GT088 GT087 GT058 GT086 9 GT09 0 GT08 71 GT0 9 0 GT0 7 0 GT0 17 GT0 23 0 GT 14 0 GT 021 GT 019 GT 018 GT 34 0 GT 036 1 GT 00 1 GT 01 0 GT T02 2 G T01 05 G T0 G GT008 GT003 GT035 GT010 GT004 * * * * * * * * * * GT046 GT043 GT083 GT029 GT028 GT032 5 GT02 0 GT03 4 2 GT0 8 4 GT0 47 GT0 54 0 GT 53 0 GT 050 GT 052 GT 051 GT 064 GT 055 GT 049 6 GT T05 2 6 0 G 0 GT T04 97 G T0 G G G T0 GT T09 02 GT 09 6 GT 02 5 GT 100 6 GT 098 GT 065 GT 066 * * 0 GT 72 GT 073 0 GT 85 0 GT0 84 * GT0 90 94 GT 0 7 GT0 8 92 GT09 1 GT089 * GT037 GT016 GT015 * * GT013 GT033 * GT027 * * * * * * * * * * * * GT GG 069 GG 008 GG 002 GG 003 * * * GG 017 GG 018 G 02 GGG02 7 G 03 1 G G0 0 GG G04 56 G 0 6 G G 0 20 G 68 06 9 * * * * * * * * 0 07 G 64 G G 0 59 G 0 8 GG G05 7 G G07 6 G G07 5 G 07 GG 054 GG 074 GG 071 GG 072 GG 073 GG 67 0 GG 48 GG0 61 GG0 0 5 GG0 9 GG04 5 00 G 45 G G 0 25 G G 0 29 G G0 44 G 0 6 GGG01 1 G G01 6 G 02 GG 051 GG 060 GG 052 GG 053 G G 05 5 GG 57 GG0 62 GG0 63 GG 0 5 GG06 6 GG06 GG036 GG035 GG040 GG032 GG034 GG031 GG042 GG039 GG006 GG001 G G 02 8 GG02 4 GG0 3 GG0 7 GG0 43 GG 19 GG 041 GG 038 GG 033 GG 012 GG 047 GG 004 G 02 G G01 3 G G0 0 G G 0 09 G G 0 22 G G 0 15 G G 0 14 G 1 00 3 7 * * * * * * * * ** * * * * Fig. 6 Genetic relatedness among individuals of H. bilitonensis from Gunung Gajah (GG) and Gua Tempurung (GT) using shared allele distance (Chakraborty and Jin 1993) and cluster analysis via the neighbour-joining method (Saitou and Nei 1987). Relative strength of the nodes was determined using bootstrapping analysis (1,000 replications) and the * indicates strength of node more than 50 % 123 Author's personal copy Biodivers Conserv 100 95 Total number of alleles (%) 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 Number of individuals Fig. 7 Changes in percentage of allele maintained with changes in the number of individuals of H. bilitonensis removed. All values were based on a mean of 1,000 resamplings. Dotted lines represent the confidence envelops Table 2 Outcrossing rates of H. bilitonensis from Gunung Gajah No. of seeds tm ts tm - ts Individual GG001 24 0.00 (0.00) – – GG003 24 0.04 (0.01) – – GG004 24 0.00 (0.00) – – GG019 24 0.05 (0.01) – – GG022 24 0.00 (0.00) – – GG023 24 0.05 (0.01) – – GG026 24 0.19 (0.02) – – GG057 24 0.00 (0.00) – – Population 192 0.037 (0.002) 0.018 (0.001) 0.019 (0.001) Values in parentheses are standard errors. The tm and ts are multilocus and single locus outcrossing rates, respectively single locus outcrossing rate (ts) was 0.018, while the multilocus outcrossing rate (tm) was 0.037. Genetic substructuring, in terms of biparental inbreeding (detectable by the difference between multilocus and single locus outcrossing rates), was not significant (tm ts = 0.019). At the individual level, multilocus outcrossing rates ranged from 0.00 (GG001, GG004, GG022 and GG057) to 0.19 (GG026). Variability in outcrossing rates 123 Author's personal copy Biodivers Conserv among individuals may reflect heterogeneity in the pollen pool, differences in the mating neighborhood of individuals, population substructure, or differences in self-compatibility. Discussion The study showed that H. bilitonensis is rare and has a restricted habitat niche in the Tempurung massif. Although the entire massif is limestone and most ridges leading to the summit are thought to provide suitable sites, randomly conducted ground checks and its absence from limestone flora accounts indicate that the species is not widespread in the massif. It is locally abundant where it occurs and it is simply not possible to overlook its presence upon encounter. Ashton (1982) reported that H. bilitonensis is locally common on the sandy islands of Banka and Billiton in East Sumatra and this disjunct distribution could possibly date back to the Pleistocene. During this period, sea levels were low and the levels reached the lowest minimum during middle Pleistocene, thereby exposing the Sunda Platform and the Sahul Shelf as an extensive land mass (Morley and Flenley 1987). Vegetation on the limestone hills of Tempurung has some degree of affinity with the coastal lowland and hill forests. Several species associated with but not necessarily restricted to the coastal forests such as S. glauca, M. corymbosa, P. obovata and S. ilicifolia were common in the area. According to Crowther (1978), several features of low-lying karst areas, many of these present at Gua Tempurung, were developed during the late Pliocence and Pleistocene periods. These karst areas could have been more widespread when the sea level was much lower. When sea level rose during the Holocene period (Tjia et al. 1984), certain areas of the massif might have been undercut by sea (Walker 1956). As a result, species associated with coastal environment remained present here. The spatial distribution pattern analyses showed significant aggregation for trees in all size classes. The spatial distribution pattern in plant populations can be influenced by the pattern and distance of seed dispersal (Plotkin et al. 2000). Many tropical tree species show spatial aggregation at varying scales, generally from higher to looser aggregation with age increase (Condit et al. 2000; Plotkin et al. 2000; Ng et al. 2004, 2006). This change with age is also apparent in H. bilitonensis which might indicate that seed dispersal is limited. Chan (1977) reported that terrestrial animals, with the exception of wild pigs, do not favour dipterocarp fruits, and fruits are dispersed by wind or gravity. Similarly, several studies have shown that dipterocarp fruits are never dispersed far from the parent trees (Turner 1990). Spatial genetic structure of plants within a natural population is primarily influenced by the pattern and distance of pollen and seed dispersals (Ennos 1994). Spatial genetic structure shall not be detected if both pollen and seed dispersals are random within a population (Kalisz et al. 2001). However, when both pollen and seed dispersals are restricted, intense spatial genetic structuring will result within population and genetic substructuring of population will evolve over time as described in the isolation by distance model (Sokal and Wartenberg 1983). Spatial genetic structure was detected for H. bilitonensis below 10 cm dbh. A positive relationship between spatial genetic structure and spatial proximity for several dipterocarp species has also been reported (Konuma et al. 2000; Takeuchi et al. 2004; Harata et al. 2012). Limited gene flow might be responsible for the observed structure in H. bilitonensis. Furthermore, given that the two populations are located within the Tempurung massif, the high population differentiation observed in this 123 Author's personal copy Biodivers Conserv study might indicate that limited gene flow due to the combined effects of selfing, occasional outcrossing and localized seed dispersal. Hopea bilitonensis flowers more or less biennially with each flowering episode involving only a small proportion of the population. The biennially flowering and the ability of a fruit crop to germinate within the expected range during an out-of-phase sporadic fruiting imply that there is ample opportunity for continuous recruitment. However, several studies showed that seed and seedling mortality due to abortion and insect predation was extremely high during out-of-phase sporadic fruiting for Shorea (Chan 1980; Appanah and Mohd Rasol 1994) and in some cases, recruitment was equally poor after mast fruiting (Blundell and Peart 2004). At Gua Tempurung, recruitment rates are disproportionately different between each seedling recruitment phase and the next. This is most likely due to the varying degree of flowering and fruiting intensity in the population. Several closely-spaced general flowering incidences could have occurred in the population but such incidences were not detected in the Gunung Gajah population. The period from germination to establishment is one of the most critical phases for plant populations (Silvertown 1987). In his work with Dryobalanops aromatica and D. lanceolata, Itoh (1995) showed that at the primary phase of establishment, both species had lower survivorship on the ridge than in the valley. Effects from desiccation and root predation were suggested as reasons. Root elongation on the ridge was delayed during the first 20-day period, caused by drying litter. Requiring a longer development period increases the opportunity for root predation. Ashton et al. (1995) also found that dipterocarp seedling growth in a Sri Lankan rain forest was slowest on ridges compared to slopes and valleys and for some species, survival was the least on ridges. Although various studies have suggested the influence of seasonal variation in soil water on the survival and growth of dipterocarps (Brown 1993; Palmiotto et al. 2004), the effects are still unclear (Reich and Borchert 1984; Bebber et al. 2004). Crowther (1978) described the slopes of the Gunung Gajah-Gua Tempurung landmass as mostly precipitous. Soil cover was almost continuous at the lower part, becomingly increasingly intermittent and lighter in bulk density at higher elevations. At the ridge, there was almost no mineral soil and peat was restricted to rock crevices. The landmass received little direct rainfall and water that reached crevices and cracks came either as condensation, storm runoff or seepage from joint planes. Clearly, such an environment, enhanced by high light intensity put pressure onto germination and early phase establishment of H. bilitonensis. It appears that confinement to a restricted niche and small population size have not excessively depleted the genetic diversity within H. bilitonensis. Despite being predominantly selfing, the levels of genetic diversity of H. bilitonensis were comparable with those of Shorea leprosula, S. ovalis, S. curtisii and S. macroptera (Ng et al. 2004, 2006), S. lumutensis (Lee et al. 2006) and S. xanthophylla, Parashorea tomentella and Dipterocarpus grandiflorus (Kettle et al. 2011). Variously viewed as either a cause or a consequence of rarity, limited genetic diversity has been reported for many rare and endangered plant species (reviewed by Hamrick and Godt 1989). This study supports the overgeneralization view that rare species have less variability than more widespread ones (Gitzendanner and Soltis 2000). Other studies have shown that rare plant species and species on narrow ranges can have high levels of genetic diversity (e.g. Lewis and Crawford 1995; Neel and Ellstrand 2001; Cao et al. 2009). In many aspects, the biology of rare plants that are locally common is similar to that of widespread congeners. The ability to occupy suitable sites is limited only by its seed dispersal. Re-sprouting or coppicing has been suggested to promote long-term persistence of genotypes which contributes to the maintenance of diversity and evolutionary potential within a refugial population (Rossetto and 123 Author's personal copy Biodivers Conserv Kooyman 2005). The ability to coppice must contribute some form of evolutionary advantage for a predominantly selfed taxa to maintain levels of genetic diversity comparable with its predominantly outcrossed relatives. Conservation recommendations The results from current ecological and genetic studies showed that the species is locally common in its niche and there is no indication of major biological bottlenecks. Variability in outcrossing rates among individuals with a tendency for selfing and coppicing are clearly evolutionary advantages to ensure long-term persistence but ultimately the species is dependent on its natural habitat for its existence. This study showed that H. bilitonensis in Malaysia comprises only two populations within the Tempurung massif. The Tempurung massif is a State Land and jurisdiction over its land use lies with the State Government of Perak. The Land Office records (unpublished) showed that parts of the Tempurung massif have been licensed out for quarrying activities. Quarrying activities in certain areas of the foothill have begun and are slowly progressing uphill. In this respect, both populations are seriously threatened from an ecological and conservation point of view, and the elimination of the species from Malaysia is likely if nothing is done to alert the authorities. Land use in State Land varies disproportionately, reflecting past values placed by man on a particular site and affecting future values associated with its use. Gua Tempurung is an ecotourism destination, offering a variety of cave systems to explore. This explains why Gua Tempurung has remained forested while its neighbouring outcrop Gunung Gajah was someway disturbed. This short-term assurance may seriously lapse; we therefore recommend, from a long-term perspective, that the entire Tempurung massif be gazetted as Permanent Reserved Forest under the functional class of amenity forest. Hamrick (1993) suggested that for tropical tree species, if 80 % of the total genetic diversity resides within a population, five strategically placed populations should capture 99 % of their total genetic diversity. The present study showed that the species has [80 % of its total genetic diversity residing within the population. However, as the species comprises only two populations, establishment of in situ conservation areas is limited to two. As the species exhibits high selfing rate, the minimum population size required to maintain 95 % of its genetic diversity is 107 individuals, which is much lower if compared with predominantly outcrossing species, such as S. lumutensis (Lee et al. 2006) and Intsia palembanica (Lee et al. 2002). If the number of 107 is to be considered and limited to two in situ conservation areas, the total number of individuals to be conserved is only 214. When planning a conservation area, however, a minimal population size should be regarded only as a last resort and an extreme compromise. Units of in situ conservation should constitute a much larger population or area to prevent major bottlenecks and rapid population crashes. In addition, the natural state of the habitat helps mitigate stochastic pressures. Conserving H. bilitonensis in its native habitat is clearly a first step but ex situ conservation is also necessary to provide insurance against catastrophic events and to facilitate the possibility of reintroduction in the future. In view of the approaching loss at Gunung Gajah, we recommend that a concerted rescue operation be conducted to salvage the species’ germplasm. Since the species exhibits high selfing rate, in order to capture the maximum levels of genetic diversity, at least 50 unrelated mother trees should be considered for germplasm collections. Selections of genetically unrelated mother trees can be 123 Author's personal copy Biodivers Conserv guided using Fig. 6. This germplasm should then be distributed and established at the respective State Forestry Departments and the network of Botanic Gardens in Malaysia. Acknowledgments We thank the Forest Department of Perak for granting us permission to access the forest reserves. The foresters and rangers of the Kinta/Manjung District Forest Office provided assistance during the field work and the officers from the District Land Office provided the information on the land status. We are grateful to the late Chan Yee Chong, Ghazali Jaafar, Yahya Mahani, Ramli Ponyoh, Mariam Din, Sharifah Talib, Damahuri Sabari, Mustapa Data and Ayau Kanir for their excellent assistance in the laboratory and field. We also extend our thanks to Chen King Min, Quarry Manager, Superior Lime Sdn. Bhd. and the GIS Units of Forestry Department Headquarters and FRIM for useful information given. Special thanks go to the Forest Engineering Team (Natural Forest Division, FRIM) for their technical assistance. This study was supported in part by the IRPA research grant (09-04-01-0013-EA001), the Timber Export Levy Fund (Flora Malaysiana Centre Programme, Project 6), and the Bioversity International Agreement No. APO 06/025. References Appanah S, Mohd Rasol AM (1994) Fruiting and seedling survival of diptercarps in a logged forest. J Trop For Sci 6:215–222 Ashton PS (1982) Dipterocarpaceae. Flora Malesiana Ser 1(9):237–552 Ashton PMS, Gunatilleke CVS, Gunatilleke IAUN (1995) Seedling survival and growth of four Shorea species in a Sri Lankan rainforest. J Trop Ecol 11:263–279 Barrett SCH, Kohn JR (1991) Genetic and evolutionary consequences of a small population size in plants: implications for conservation. In: Falk DA, Holsinger KE (eds) Genetics and conservation of rare plants. Oxford University Press, New York, pp 3–30 Bebber DP, Brown ND, Speight MR (2004) Dipterocarp seedling population dynamics in Bornean primary lowland forest during the 1997–8 El-Nino Southern Oscillation. J Trop Ecol 20:11–19 Blundell AG, Peart DR (2001) Growth strategies of a shade-tolerant tropical tree: the interactive effects of canopy gaps and simulated herbivory. J Ecol 89:608–615 Blundell AG, Peart DR (2004) Seedling recruitment failure following dipterocarp mast fruiting. J Trop Ecol 20:229–231 Brown ND (1993) The implications of climate and gap microclimate for seedling growth conditions in a Bornean lowland rain forest. J Trop Ecol 9:153–168 Brussard PF (1991) The role of ecology in biological conservation. Ecol Appl 1:6–12 Cao CP, Gailing O, Siregar IZ, Siregar UJ, Finkeldey R (2009) Genetic variation in nine Shorea species (Dipterocarpaceae) in Indonesia revealed by AFLPs. Tree Genet Genomes 5:407–420 Chakraborty R, Jin L (1993) Determination of relatedness between individuals using DNA fingerprinting. Hum Biol 65:875–895 Chan HT (1977) Reproductive biology of some Malaysian dipterocarps. University of Aberdeen, Dissertation Chan HT (1980) Reproductive biology of some Malaysian Dipterocarps. II. Fruiting biology and seedling studies. Malays For 43:438–451 Charlesworth D, Charlesworth B (1987) Inbreeding depression and its evolutionary consequences. Annu Rev Ecol Syst 18:237–268 Chua LSL, Suhaida M, Hamidah M, Saw LG (2010) Malaysia plant red list: Peninsular Malaysia Dipterocarpaceae. Research Pamphlet No. 129. Forest Research Institute Malaysia, Kuala Lumpur, p 57 Condit R, Sukumar R, Hubbell SP, Foster RB (1998) Predicting population trends from size distributions: a direct test in tropical tree communities. Am Natur 152:495–509 Condit R, Ashton PS, Baker P, Bunyavejchewin S, Gunatilleke S, Gunatilleke N, Hubbell SP, Losos E, Manokaran N, Sukumar R, Yamakura T (2000) Spatial patterns in the distribution of tropical tree species. Science 288:1414–1418 Crowther J (1978) The Gunong Gajah-Tempurong massif, Perak, and its associated cave system, Gua Tempurung. Mal Nat J 32:1–18 Degen B, Petit R, Kremer A (2001) SGS-Spatial Genetic Software: a computer program for analysis of spatial genetic and phenotypic structures of individuals and populations. J Hered 92:447–449 Ennos RA (1994) Estimating the relative rates of pollen and seed migration among plant populations. Heredity 72:250–259 123 Author's personal copy Biodivers Conserv Gitzendanner M, Soltis PS (2000) Patterns of genetic variation in rare and widespread plant congeners. Am J Bot 87:783–792 Goudet J (2002) FSTAT version 2.9.3.2: a computer package for PCs which estimates and tests gene diversities and differentiation statistics from codominant genetic markers. http://www.unil.ch/izea/ software/fstat.html Haase P (1995) Spatial pattern analysis in ecology based on Ripley’s K-function: introduction and methods of edge correction. J Veg Sci 6:575–582 Hamrick JL (1993) Genetic diversity and conservation in tropical forest. In: Drysdale RM, John SET, Yapa AC (eds) Proceedings of the ASEAN–Canada Symposium on Genetic Conservation and Production of Tropical Tree Seed, ASEAN–Canada Forest Tree Seed Center, Saraburi, pp 1–9 Hamrick JL, Godt MJW (1989) Allozyme diversity in plant species. In: Brown AHD, Clegg MJ, Kahler AL, Weir BS (eds) Plant population genetics, breeding and genetic resources. Sinauer Associate, Sunderland, pp 43–63 Harata T, Nanami S, Yamakura T, Matsuyama S, Chong L, Diway BM, Tan S, Itoh A (2012) Fine-scale spatial genetic structure of ten dipterocarp tree species in a Bornean rain forest. Biotropica 44:586–594 Holsinger KE (1996) The scope and limits of conservation genetics. Evolution 50:2558–2561 Huenneke LF (1991) Ecological implications of genetic variation in plant populations. In: Falk DA, Holsinger KE (eds) Genetics and conservation of rare plants. Oxford University Press, New York, pp 31–44 Itoh A (1995) Effects of forest floor environment on germination and seedling establishment of two Bornean rainforest emergent species. J Trop Ecol 11:517–527 IUCN (1994) IUCN Red List Categories and Criteria version 2.3. IUCN, Gland Jarne P, Lagoda PJ (1996) Microsatellite, from molecules to populations and back. Trends Ecol Evol 11:424–429 Kalisz S, Nason JD, Hanzawa FM, Tonsor SJ (2001) Spatial population genetic structure in Trillium grandiflorum: the roles of dispersal, mating history and selection. Evolution 55:1560–1568 Kettle CJ, Hollingsworth PM, Burslem DFRP, Maycock CR, Khoo E, Ghazoul J (2011) Determinants of fine-scale spatial genetic structure in three co-occurring rain forest canopy trees in Borneo. Perspect Plant Ecol Evol Syst 13:45–54 Kochummen KM (1997) Tree Flora of Pasoh Forest. Malayan Forest Records No. 44, Forest Research Institute Malaysia, Kepong, Kuala Lumpur Konuma A, Tsumura Y, Lee CT, Lee SL, Okuda T (2000) Estimation of gene flow in the tropical rain forest tree Neobalanocarpus heimii (Dipterocarpaceae), inferred from paternity analysis. Mol Ecol 9:1843–1852 Lande RC, Barrowclough GF (1987) Effective population size, genetic variation, and their use in population management. In: Soulé ME (ed) Viable populations for conservation. Cambridge University Press, Cambridge, pp 87–124 Lee SL, Wickneswari R, Mahani MC, Zakri AH (2000) Genetic diversity of Shorea leprosula Miq. (Dipterocarpaceae) in Malaysia: implications for conservation of genetic resources and tree improvement. Biotropica 32:213–224 Lee SL, Ng KKS, Saw LG, Norwati A, Siti Salwana MH, Lee CT, Norwati M (2002) Population genetics of Intsia palembanica (Leguminosae) and genetic conservation of Virgin Jungle Reserves (VJRs) in Peninsular Malaysia. Am J Bot 89:447–459 Lee SL, Tani N, Ng KKS, Tsumura Y (2004) Characterization of 15 polymorphic microsatellite loci in an endangered tropical tree Hopea bilitonensis (Dipterocarpaceae) in Peninsular Malaysia. Mol Ecol Notes 4:147–149 Lee SL, Ng KKS, Saw LG, Lee CT, Norwati M, Tani N, Tsumura Y, Koskela J (2006) Linking the gaps between conservation research and conservation management of rare dipterocarps: a case study of Shorea lumutensis. Biol Conserv 131:72–92 Lewis PO, Crawford DJ (1995) Pleistocene refugium endemics exhibit greater allozyme diversity than widespread congeners in the genus Polygonella (Polygonaceae). Am J Bot 82:141–149 Liu K, Muse SV (2005) PowerMarker: integrated analysis environment for genetic marker data. Bioinformatics 21:2128–2129 Morley RJ, Flenley JR (1987) Late Cainozoic vegetational and environmental changes in the Malay Archipelago. In: Whitmore TC (ed) Biogeographical evolution of the Malay Archipelago. Clarendon Press, Oxford, pp 50–59 Murray M, Thompson WF (1980) Rapid isolation of high molecular weight plant DNA. Nucleic Acid Res 8:4321–4325 Neel MC, Ellstrand C (2001) Pattern of allozyme diversity in the threatened plant Erigeron parishii (Asteraceae). Am J Bot 88:810–818 123 Author's personal copy Biodivers Conserv Nei M (1987) Molecular evolutionary genetics. Columbia University Press, New York Ng FSP, Mat Asri NS (1991) Germination and seedling records. FRIM Research Pamphlet No. 108, Forest Research Institute Malaysia, Kuala Lumpur, pp 36–55 Ng KKS, Lee SL, Koh CL (2004) Spatial structure and genetic diversity of two tropical tree species with contrasting breeding systems and different ploidy levels. Mol Ecol 13:657–669 Ng KKS, Lee SL, Saw LG, Plotkin JB, Koh CL (2006) Spatial structure and genetic diversity of three tropical tree species with different habitat preferences within a natural forest. Tree Genet Genomes 2:121–131 Palmiotto PA, Davies SJ, Vogt KA, Ashton MS, Vogt DJ, Ashton PS (2004) Soil-related habitat specialization in dipterocarp rain forest tree species in Borneo. J Ecol 92:609–623 Peters HA (2003) Neighbor-regulated mortality: the influence of positive and negative density dependence on tree populations in species-rich tropical forests. Ecol Lett 6:757–765 Petit RJ, El Mousadik A, Pons O (1998) Identifying population for conservation on the basis of genetic markers. Conserv Biol 12:844–855 Plotkin JB, Potts M, Leslie N, Manokaran N, LaFrankie J, Ashton PS (2000) Species-area curves, spatial aggregation, and habitat specialization in tropical forests. J Theor Biol 207:81–99 Reich PB, Borchert R (1984) Water stress and tree phenology in a tropical dry forest in the lowlands of Costa Rica. J Ecol 72:61–74 Rice WR (1989) Analysing tables of statistical tests. Evolution 43:223–225 Ripley BD (1976) The second-order analysis of stationary processes. J Appl Probab 13:255–266 Ritland K (2002) Extensions of models for the estimation of mating systems using n independent loci. Heredity 88:221–228 Rossetto M, Kooyman RM (2005) The tension between dispersal and persistence regulates the current distribution of rare paleo-edemic rain forest flora: a case study. J Ecol 93:906–917 Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 11:553–570 Silvertown JW (1987) Introduction to plant population ecology, 2nd edn. Longman, London Simberloff D (1988) The contribution of population and community biology to conservation science. Annu Rev Ecol Syst 19:473–511 Slatkin M (1995) A measure of population subdivision based on microsatellite allele frequency. Genetics 139:457–462 Sokal RR, Oden NL (1978) Spatial autocorrelation in biology. I. Methodology. Biol J Linn Soc 10:199–228 Sokal RR, Wartenberg DE (1983) A test of spatial autocorrelation analysis using an isolation-by-distance model. Genetics 105:219–247 Takeuchi Y, Ichikawa S, Konuma A, Tomaru N, Niiyama K, Lee SL, Norwati M, Tsumura Y (2004) Comparison of the fine-scale genetic structure of three dipterocarp species. Heredity 92:323–328 Tjia HD, Sujitno S, Suklija Y, Harsono RAF, Rachmat A, Hainim J, Djunaedi (1984) Holocene shorelines in the Indonesian Tin Islands. Mod Quat Res Southeast Asia 8:103–117 Turner IM (1990) The seedling survivorship and growth of three Shorea species in a Malaysian tropical rain forest. J Trop Ecol 6:469–478 Walker D (1956) Studies in the quarternary of the Malay Peninsula. I. Alluvial deposits of Perak and changes in the relative levels of land and sea. Fed Mus J 2:19–34 123 View publication stats