Kevin Kit Siong Ng

Spatial structure and genetic diversity of two tropical tree species with contrasting breeding systems and different ploidy levels.

Analyses of the spatial distribution pattern, spatial genetic structure and of genetic diversity were carried out in two tropical tree species with contrasting breeding systems and different ploidy levels using a 50‐ha demographic plot in a lowland dipterocarp forest in Peninsular Malaysia. Shorea leprosula is a diploid and predominantly outcrossed species, whereas S. ovalis ssp. sericea is an autotetraploid species with apomictic mode of reproduction. Genetic diversity parameters estimated for S. leprosula using microsatellite were consistently higher than using allozyme. In comparisons with S. leprosula and other tropical tree species, S. ovalis ssp. sericea also displayed relatively high levels of genetic diversity. This might be explained by the lower pressure of genetic drift due to tetrasomic inheritance, and for autotetraploids each locus can accommodate up to four different alleles and this allows maintenance of more alleles at individual loci. The observed high levels of genetic diversity in S. ovalis ssp. sericea can also be due to a random retention of more heterogeneous individuals in the past, and the apomictic mode of reproduction might be an evolutionary strategy, which allows the species to maintain high levels of genetic diversity. The spatial distribution pattern analyses of both species showed significant levels of aggregation at small and medium but random distribution at the big diameter‐class. The decrease in magnitude of spatial aggregation from small‐ to large‐diameter classes might be due to compensatory mortality during recruitment and survival under competitive thinning process. Spatial genetic structure analyses for both species revealed significant spatial genetic structure for short distances in all the three diameter‐classes. The magnitude of spatial genetic structure in both species was observed to be decreasing from smaller‐ to larger‐diameter classes. The high spatial genetic structuring observed in S. ovalis ssp. sericea at the small‐diameter class is due primarily to limited seed dispersal and apomictic mode of reproduction. The similar observation in S. leprosula, however, can be explained by limited seed and pollen dispersal, which supports further the fact that the species is pollinated by weak fliers, mainly of Thrips and Megalurothrips in the lowland dipterocarp forest.

Population genetics of Intsia palembanica (Leguminosae) and genetic conservation of virgin jungle reserves in Peninsular Malaysia.

A field survey of Virgin Jungle Reserve (VJR) compartments in Peninsular Malaysia allowed us to identify six populations of Intsia palembanica for this study. These were Pasoh Forest Reserve (FR) (Pasoh), Sungai Lalang FR (Lalang), Bukit Lagong FR (Lagong), Bubu FR (Bubu), Bukit Kinta FR (Kinta), and Bukit Perangin FR (Perangin). About 40 adult individuals were sampled in each population. In addition, progeny arrays were collected from nine mother plants at Lagong for a mating system study. A total of nine allozymes, encoded by 14 putative gene loci, were consistently resolved in I. palembanica. The mating system study showed that the species exhibited a mixed-mating system, with multilocus outcrossing rate of 0.766. The levels of diversity were comparably high (mean number of alleles per polymorphic locus = 2.4, mean effective number of alleles per polymorphic locus = 1.64, and mean expected heterozygosity (H(e)) = 0.242), and the majority of the diversity was partitioned within population (G(ST) = 0.040 and F(ST) = 0.048). Significant levels of inbreeding were detected in Bubu and Perangin. Probability tests of recent effective population size reduction using the Infinite Allele Model showed the occurrence of genetic bottlenecks on Lalang and Kinta. Two genetically unique populations (Pasoh and Perangin) were inferred using jackknife analysis. By using the neutral mutation rates, effective population size (N(e)) to maintain the H(e) was 80-800 000 individuals. A simulation study based on pooled samples, however, circumscribed the N(e) to 200 and 210 individuals. Implications of the study for managing the species and the VJRs are discussed.

Paternity analysis-based inference of pollen dispersal patterns, male fecundity variation, and influence of flowering tree density and general flowering magnitude in two dipterocarp species

BACKGROUND AND AIMS Knowledge of pollen dispersal patterns and variation of fecundity is essential to understanding plant evolutionary processes and to formulating strategies to conserve forest genetic resources. Nevertheless, the pollen dispersal pattern of dipterocarp, main canopy tree species in palaeo-tropical forest remains unclear, and flowering intensity variation in the field suggests heterogeneity of fecundity. METHODS Pollen dispersal patterns and male fecundity variation of Shorea leprosula and Shorea parvifolia ssp. parvifolia on Peninsular Malaysian were investigated during two general flowering seasons (2001 and 2002), using a neighbourhood model modified by including terms accounting for variation in male fecundity among individual trees to express heterogeneity in flowering. KEY RESULTS The pollen dispersal patterns of the two dipterocarp species were affected by differences in conspecific tree flowering density, and reductions in conspecific tree flowering density led to an increased selfing rate. Active pollen dispersal and a larger number of effective paternal parents were observed for both species in the season of greater magnitude of general flowering (2002). CONCLUSIONS The magnitude of general flowering, male fecundity variation, and distance between pollen donors and mother trees should be taken into account when attempting to predict the effects of management practices on the self-fertilization and genetic structure of key tree species in tropical forest, and also the sustainability of possible management strategies, especially selective logging regimes.

Spatial structure and genetic diversity of three tropical tree species with different habitat preferences within a natural forest

Analyses of the spatial distribution pattern, spatial genetic structure and genetic diversity were carried out using a 33-ha plot in a hill dipterocarp forest for three dipterocarps with different habitat preferences, i.e. Shorea curtisii on the ridges, Shorea leprosula in the valleys and Shorea macroptera both on the ridges and in the valleys. The significant spatial aggregation in small-diameter trees of all the three species was explained by limited seed dispersal. At the large-diameter trees, only S. macroptera showed random distribution and this might further prove that S. macroptera is habitat generalist, whilst S. curtisii and S. leprosula are habitat specific. The levels of genetic diversity estimated based on five microsatellite loci were high and comparable in all the three studied species. As the three studied species reproduced mainly through outcrossing, the observed high levels of genetic diversity might support the fact that the plant mating system can be used as guideline to infer the levels of genetic diversity, regardless of whether the species is habitat specific or habitat generalist. The lack of spatial genetic structure but significant aggregation in the small-diameter trees of all the three species might indicate limited seed dispersal but extensive pollen flow. Hence, if seed dispersal is restricted but pollen flow is extensive, significant spatial aggregation but no spatial genetic structure will be observed at the small-diameter trees, regardless of whether the species is habitat specific or habitat generalist. The inferred extensive pollen flow might indicate that energetic pollinators are involved in the pollination of Shorea species in the hill dipterocarp forests.

Molecular database for classifying Shorea species (Dipterocarpaceae) and techniques for checking the legitimacy of timber and wood products.

The extent of tropical forest has been declining, due to over-exploitation and illegal logging activities. Large quantities of unlawfully extracted timber and other wood products have been exported, mainly to developed countries. As part of the export monitoring effort, we have developed methods for extracting and analyzing DNA from wood products, such as veneers and sawn timbers made from dipterocarps, in order to identify the species from which they originated. We have also developed a chloroplast DNA database for classifying Shorea species, which are both ecologically and commercially important canopy tree species in the forests of Southeast Asia. We are able to determine the candidate species of wood samples, based on DNA sequences and anatomical data. The methods for analyzing DNA from dipterocarp wood products may have strong deterrent effects on international trade of illegitimate dipterocarp products. However, the method for analyzing DNA from wood is not perfect for all wood products and need for more improvement, especially for plywood sample. Consequently, there may be benefits for the conservation of tropical forests in Southeast Asia.

Nuclear and chloroplast DNA phylogeography reveals Pleistocene divergence and subsequent secondary contact of two genetic lineages of the tropical rainforest tree species Shorea leprosula (Dipterocarpaceae) in South‐East Asia

Tropical rainforests in South‐East Asia have been affected by climatic fluctuations during past glacial eras. To examine how the accompanying changes in land areas and temperature have affected the genetic properties of rainforest trees in the region, we investigated the phylogeographic patterns of a widespread dipterocarp species, Shorea leprosula. Two types of DNA markers were used: expressed sequence tag‐based simple sequence repeats and chloroplast DNA (cpDNA) sequence variations. Both sets of markers revealed clear genetic differentiation between populations in Borneo and those in the Malay Peninsula and Sumatra (Malay/Sumatra). However, in the south‐western part of Borneo, genetic admixture of the lineages was observed in the two marker types. Coalescent simulation based on cpDNA sequence variation suggested that the two lineages arose 0.28–0.09 million years before present and that following their divergence migration from Malay/Sumatra to Borneo strongly exceeded migration in the opposite direction. We conclude that the genetic structure of S. leprosula was largely formed during the middle Pleistocene and was subsequently modified by eastward migration across the subaerially exposed Sunda Shelf.

Evidence for a higher number of species of Odontotermes (Isoptera) than currently known from Peninsular Malaysia from mitochondrial DNA phylogenies.

Termites of the genus Odontotermes are important decomposers in the Old World tropics and are sometimes important pests of crops, timber and trees. The species within the genus often have overlapping size ranges and are difficult to differentiate based on morphology. As a result, the taxonomy of Odontotermes in Peninsular Malaysia has not been adequately worked out. In this study, we examined the phylogeny of 40 samples of Odontotermes from Peninsular Malaysia using two mitochondrial DNA regions, that is, the 16S ribosomal RNA and cytochrome oxidase subunit I genes, to aid in elucidating the number of species in the peninsula. Phylogenies were reconstructed from the individual gene and combined gene data sets using parsimony and likelihood criteria. The phylogenies supported the presence of up to eleven species in Peninsular Malaysia, which were identified as O. escherichi, O. hainanensis, O. javanicus, O. longignathus, O. malaccensis, O. oblongatus, O. paraoblongatus, O. sarawakensis, and three possibly new species. Additionally, some of our taxa are thought to comprise a complex of two or more species. The number of species found in this study using DNA methods was more than the initial nine species thought to occur in Peninsular Malaysia. The support values for the clades and morphology of the soldiers provided further evidence for the existence of eleven or more species. Higher resolution genetic markers such as microsatellites would be required to confirm the presence of cryptic species in some taxa.

Isolation and characterization of microsatellite markers for an important tropical tree, Aquilaria malaccensis (Thymelaeaceae)

PREMISE OF THE STUDY Aggressive collections and trade activities in recent decades have resulted in heavy pressure on the natural stands of Aquilaria malaccensis and concerns over its long-term survival potential. To aid DNA profiling and assessment of its genetic diversity, microsatellite markers were developed for the species. METHODS AND RESULTS Seventeen polymorphic microsatellite markers were developed for A. malaccensis using an enrichment protocol. The markers were screened on 24 samples from a natural population. The number of alleles ranged from two to 11, and the observed heterozygosity ranged from 0.042 to 0.957. No significant deviation from Hardy-Weinberg equilibrium was detected after conservative Bonferroni correction. CONCLUSIONS This is the first report on the development of microsatellite markers in A. malaccensis. The markers will be used to establish a DNA profiling database and to estimate the genetic diversity and population genetic structure of the species.

Expressed sequence tag–simple sequence repeats isolated from Shorea leprosula and their transferability to 36 species within the Dipterocarpaceae

Simple sequence repeats (SSRs) derived from expressed sequence tags (ESTs) are valuable markers because they represent transcribed regions and often transferable to related taxa. Here, we report the development and characterization of EST–SSRs from Shorea leprosula. Fifty‐four sequences containing SSRs were identified in 2003 unigenes assembled from 3159 ESTs. Twenty‐four EST–SSRs were developed, of which four gave multiple amplifications, five were found to be monomorphic and 15 showed polymorphism, with allele numbers ranging from two to 17 in a single Pasoh Forest Reserve population of 24 individuals. The observed and expected heterozygosities ranged from 0.05 to 0.91 and from 0.16 to 0.93, respectively. Cross‐species transferability of the 15 loci to 36 species within Dipterocarpaceae revealed between four and 14 loci that gave positive amplification and 10 loci were found to be transferable to more than 15 species.

Forensic timber identification: a case study of a CITES listed species, Gonystylus bancanus (Thymelaeaceae)

Illegal logging and smuggling of Gonystylus bancanus (Thymelaeaceae) poses a serious threat to this fragile valuable peat swamp timber species. Using G. bancanus as a case study, DNA markers were used to develop identification databases at the species, population and individual level. The species level database for Gonystylus comprised of an rDNA (ITS2) and two cpDNA (trnH-psbA and trnL) markers based on a 20 Gonystylus species database. When concatenated, taxonomic species recognition was achieved with a resolution of 90% (18 out of the 20 species). In addition, based on 17 natural populations of G. bancanus throughout West (Peninsular Malaysia) and East (Sabah and Sarawak) Malaysia, population and individual identification databases were developed using cpDNA and STR markers respectively. A haplotype distribution map for Malaysia was generated using six cpDNA markers, resulting in 12 unique multilocus haplotypes, from 24 informative intraspecific variable sites. These unique haplotypes suggest a clear genetic structuring of West and East regions. A simulation procedure based on the composition of the samples was used to test whether a suspected sample conformed to a given regional origin. Overall, the observed type I and II errors of the databases showed good concordance with the predicted 5% threshold which indicates that the databases were useful in revealing provenance and establishing conformity of samples from West and East Malaysia. Sixteen STRs were used to develop the DNA profiling databases for individual identification. Bayesian clustering analyses divided the 17 populations into two main genetic clusters, corresponding to the regions of West and East Malaysia. Population substructuring (K=2) was observed within each region. After removal of bias resulting from sampling effects and population subdivision, conservativeness tests showed that the West and East Malaysia databases were conservative. This suggests that both databases can be used independently for random match probability estimation within respective regions. The reliability of the databases was further determined by independent self-assignment tests based on the likelihood of each individuals multilocus genotype occurring in each identified population, genetic cluster and region with an average percentage of correctly assigned individuals of 54.80%, 99.60% and 100% respectively. Thus, after appropriate validation, the genetic identification databases developed for G. bancanus in this study could support forensic applications and help safeguard this valuable species into the future.