Population, morphological, and genetic characteristics of pelawan trees on Bangka Island, Indonesia: implications for conservation

Abstract

Tristaniopsis spp. (pelawan) trees hold great significance in ensuring biodiversity sustainability and supporting the local species within Kepulauan Bangka Belitung Province. Moreover, it plays a crucial role in fostering the social and cultural aspects of the local community. However, the pelawan population has faced great threats due to land conversion, unmanaged forests, and mining. To address these challenges, the local government has shown a commitment to expanding the in-situ conservation areas beyond Namang Biodiversity Park. In order to enhance the conservation status of this species, it is essential to conduct a comprehensive study of the tree’s distribution, morphological characteristics, species delimitation, and genetic diversity. Therefore, this study aimed to obtain information on pelawan distribution, morphological characteristics, species confirmation, and genetic diversity. It was conducted at five natural populations, namely KK (Kota Kapur), BB (Berbura), LAP (Labuh Air Pandan), PN (Pangkal Niur), and LB (Limbung), and the genetic character was observed using ITS and RAPD markers. The first instance of establishing a nucleotide sequence database for pelawan on Bangka Island was also represented, showing the presence of two distinct taxa groups based on ITS sequence nucleotides. Based on morphological characters (bark color and leaves), the two groups were determined to be pelawan merah/orange (Tristaniopsis sp. 1) and pelawan air (Tristaniopsis sp. 2). Significantly, the genetic diversity analysis yielded a range of H_E values, from 0.283 (KK) to 0.353 (PN). The PCoA and dendrogram analyses exhibited the formation of a cluster comprising populations from Bangka (KK, BB, LAP, and PN), while LB formed a separate cluster. The result showed significant genetic differentiation between regions and populations, without mixing seeds or seedlings. It was suggested that the conservation strategy for pelawan trees on Bangka Island should be developed with several considerations. In-situ conservation sites should be designated based on the presence of pelawan merah/orange and pelawan air taxa. Additionally, ex-situ conservation plots needed to be established, with at least one location dedicated to each taxonomic group. To preserve the remaining natural genetic resources, KK and LB populations, which had the lowest H_E value and different genetic structures should be prioritized. Strategies for ex-situ conservation included KK, LB, and PN. The prioritization of PN for conservation was necessary, and further genetic study of the pelawan should involve larger populations representing the natural distribution of each species.

Keywords:

• Pelawan
• distribution
• specimen
• ITS
• RAPD

Introduction

Tristaniopsis Brongn. & Gris is a significant genus within the Myrtaceae family. This genus comprises a total of 50 species, naturally found in regions ranging from Burma and Thailand in the north to Malesia, and extending to eastern Australia and New Caledonia (Ahmad and Wilson 2015). Furthermore, 25 species of this genus can be found in the Malesian region, including Indonesia, specifically Borneo, Sumatera, Java, and Papua. Within Indonesia, Tristaniopsis spp., commonly referred to as the pelawan tree, possesses considerable economic significance due to its multiple applications in food production, medicine, energy generation, adhesion, boat construction, and building materials. Additionally, it holds ecological value as an essential key species for biodiversity sustainability, with populations featuring more than 10 individual trees each (Falk and Holsinger 1991). Its presence holds significant importance in the social and cultural fabric of the surrounding community, owing to its considerable economic advantages. This species serves as a natural habitat for Heimioporus sp. mushrooms and as a nectar source for Apis dorsata bees (Akbarini 2016). Moreover, it possesses remarkable cultural value, particularly during traditional events, acting as a key species of recognition (Akbarini et al. 2017). Mushrooms and honey obtained from these trees have emerged as highly valuable non-timber forest products within the province. The usage of pelawan leaf tea as a medicinal herbal beverage has gained popularity in treating various ailments, such as kidney stones, diabetes, as an antioxidant, and in stroke prevention (Enggiwanto et al. 2018; Sudirman et al. 2019; Keim et al. 2021; Rosianty et al. 2022).

The pelawan population has encountered significant challenges due to various factors that have posed threats to its existence. Forest conversion, driven by copper mining, infrastructure development, the establishment of plantations (Turjaman et al. 2019), and charcoal production (Yarli 2011; Siahaan et al. 2020), has led to the degradation of pelawan habitats. This degradation is particularly detrimental to T. merguensis, as its slow growth and the limited availability of cultivation information have hindered optimal conservation efforts (Triadiati et al. 2020).

Tristaniopsis merguensis (Siahaan et al. 2020), T. obovata (Turjaman et al. 2019), and T. whiteana (Yulisma et al. 2018; Juliana et al. 2021) species of Tristaniopsis were also reported in Bangka Belitung. The local community categorizes pelawan trees into three groups, namely pelawan merah, pelawan air, and pelawan tudal (Karma 2019, personal communication). Pelawan merah displays a range of reddish bark, varying from light to dark red, while pelawan air is characterized by its light gray bark, typically found near river banks, and pelawan tudal primarily grows in highland areas. Without genetic information or specific morphological descriptions, it is challenging to provide precise confirmation of their differences.

In the late twentieth century, conservationists and agricultural scientists shared a common focus on the conservation, management, and utilization of genetic resources. To achieve this, the development of marker systems that are repeatable, efficient, and applicable across various organisms with differing degrees of relatedness has become essential. Molecular markers serve as valuable tools in the management of plant genetic resources. Furthermore, the utilization of RAPD (Random Amplified Polymorphic DNA) markers, consisting of short primers with arbitrary nucleotide sequences, enables the amplification of genomic DNA segments from a wide range of species (Williams et al. 1990). As a dominant marker, RAPD solely identifies homozygote dominant or recessive alleles and cannot distinguish heterozygotes. These markers have been employed in studies focusing on population genetics (Tsuda et al. 2004) as well as plant tree improvement (Rajora and Rahman 2003).

The conservation program for pelawan in Bangka was initiated through the establishment of a biodiversity park and in-situ conservation efforts by the local government in Namang Village, Central Bangka (Akbarini et al. 2017). Even though this program represents a step toward ensuring the sustainability of pelawan and addressing potential threats, its current scope is insufficient. The expansion of ex-situ and in-situ conservation areas is deemed necessary. To establish a conservation baseline, gathering information on genetic diversity, distribution patterns, and species confirmation within their natural populations is important. Therefore, this study acquires comprehensive data concerning the distribution of pelawan, confirms the presence of Tristaniopsis species across five populations on Bangka Island, assesses the genetic diversity, and conducts morphological observations on leaves and barks.

Materials and methods

Distribution of populations of pelawan sample on Bangka Island

Five natural populations of Tristaniopsis spp. were sampled, as shown in Figure 1 and Table 1. The local authority initially promoted five different sites as candidates for pelawan conservation areas on Bangka Island. These five populations were spread in two different geographic areas, namely four in Bangka, including Kota Kapur (KK), Labuh Air Pandan (LAP), Berbura (BB), and Pangkal Niur (PN), as well as one in West Bangka, namely Limbung (LB), as shown in Table 2. Each population was assessed to determine its condition, threats, potential characteristics, and genetic data supporting both ex-situ and in-situ conservation strategies. There was no specific population-pertained putative taxon from the five populations on Bangka Island, but different species of Tristaniopsis spp. were retrieved at the riparian area of PN.

Figure 1. Site distribution of pelawan in Bangka Island, five locations were observed.

Table 1. Population and geographical location of five populations of pelawan.

Population name	Population code	Location	Region	GPS coordinate
				South	East
Kota Kapur	KK	Mendo Barat	Bangka	105° 51′ 46.90″	2°13′ 30.50″
Labuh Air Pandan	LAP	Mendo Barat	Bangka	105° 49′ 23.30″	2°8′ 47.50″
Berbura	BB	Riau Silip	Bangka	105° 52′ 3.58″	1°49′ 22.99″
Pangkal Niur	PN	Riau Silip	Bangka	105° 45′ 21.87″	1°49′ 58.70″
Limbung	LB	Jebus	West Bangka	105° 33′ 53.08″	1°44′ 45.88″

Table 2. Various information relating to potential areas for pelawan conservation.

No	Description	KK	LAP	BB	PN	LB
1	Road access	Motorbike	Car	Motorbike	Car	Motorbike
2	Total area for in-situ conservation	5 ha	3–5 ha	≥10 ha	≥10 ha	≥10 ha
3	Flowering	None	None	None	Few	None
4	Natural regeneration	Few	Many	Few	Many	Many
5	Safety Issue	Safely managed by local forest farmer community	Safely managed by local community for Bee keeping area.	Bushfire history, land use change for pepper cultivation, State-owned Forest production area.	Protected forest, partial rejection by local community due to uninformed programs.	Land use change for pepper cultivation
6	Management authority	Production Forest Management Unit Sigambir Kotawaringin	Production Forest Management Unit Sigambir Kotawaringin	Production Forest Management Unit Bubus Panca	Local district (Village Forest Protection Area)	Production Forest Management Unit Bemban Antan

Referring to Koppen, Bangka Island had a climate type A (tropical forest) with an average rainfall of 2400 mm/year (Badan Pusat Statistik 2023). The soil type on this island is dominated by yellowish-brown podzolic soil with sandstone and quartzite bedrock. Meanwhile, the average soil acidity (pH) was 5, with the mineral content of tin ore, quartz sand, and kaolin (kaolinite) (DESDM 2023).

As indicated in Table 2, accessibility to these conservation sites varied due to their distinctive landscape characteristics. Additionally, a thorough evaluation of the land’s historical background, ownership, and socio-economic response to the proposed conservation plan was conducted within the designated areas. These assessments are crucial in anticipating the variables impacting the successful management of pelawan conservation area (Maxted et al. 2000; Gradl et al. 2022).

For the DNA studies, genetic materials were collected from five natural populations, namely KK, LAP, BB, PN, and LB, as shown in Table 1. The GPS coordinates were recorded to document the precise location of each sampled pelawan tree. The assessment of each population was conducted to gather information regarding the status, potential threats, and factors supporting ex-situ and in-situ conservation programs, while also investigating the genetic characteristics.

Plant sampling for genetic characters

For DNA barcoding, 12 leaf samples were collected from the five populations, including two pelawan air (the only two trees in the PN forest), and 10 pelawan merah samples (Table 3). While for population genetic studies, only leaves from pelawan merah trees were collected from the five populations, with the total sample being 96, as shown in Table 4. All leaf samples collected were kept dry using silica gel and stored at room temperature (25 °C) until DNA extraction.

Table 3. Samples collection for DNA barcoding study from five populations pelawan in Bangka Island.

DNA ID	Local name	Population	GPS coordinate		Bark color
			South	East
PN1	Pelawan air	PN	105° 45′ 22.10″	1° 4′ 58.86″	Pale red
PN2	Pelawan air	PN	105° 45′ 24.50″	1° 50′ 2.38″	Pale red
PN3	Pelawan merah	PN	105° 45′ 21.87″	1° 49′ 58.70″	Orangish-red
PN4	Pelawan merah	PN	105° 45′ 22.10″	1° 49′ 58.64″	Orangish-red
LB5	Pelawan merah	LB	105° 33′ 53.08″	1° 44′ 45.88″	Orangish-red
LB6	Pelawan merah	LB	105° 33′ 52.05″	1° 44′ 47.08″	Orangish-red
BB7	Pelawan merah	BB	105° 52′ 4.22″	1° 49′ 22.89″	Red
BB8	Pelawan merah	BB	105° 52′ 4.64″	1° 49′ 22.79″	Red
LAP9	Pelawan merah	LAP	105° 49′ 25.90″	2° 8′ 45.90″	Red
LAP10	Pelawan merah	LAP	105° 49′ 19.20″	2° 8′ 49.00″	Red
KK11	Pelawan merah	KK	105° 51′ 43.60″	2° 13′ 33.10″	Red
KK12	Pelawan merah	KK	105° 51′ 49.90″	2° 13′ 29.50″	Red

Table 4. Samples collection for population genetic study from five populations of pelawan in Bangka Island.

Population code	Location	Region	Number of samples
KK	Mendo Barat	Bangka	16
LAP	Mendo Barat	Bangka	20
BB	Riau Silip	Bangka	20
PN	Riau Silip	Bangka	20
LB	Jebus	West Bangka	20

Morphological observation on leaves and barks

During field surveys, leaf specimen of pelawan merah and pelawan air were collected for morphological characterization. Only one specimen for each species was used phenotypic study. The morphological characters were examined following the Kew plant glossary (Beentje 2020), and qualitative data on leaf arrangement, shape, base, apex, margin, color, leaf surface, venation, and tree bark type were collected. Furthermore, SEM micrographs also depicted the dorsal surfaces of the leaves. Determination and identification were also performed based on the Kew Tropical Plant Families Identification Handbook (The Royal Botanic Garden 2020), the National Herbarium of The Netherlands (https://bioportal.naturalis.nl/), specimen images on JSTOR Global Plants, USA (https://plants.jstor.org/), and picture guides of forest trees published by the Center for Asian Conservation Ecology, Kyushu University (https://sites.google.com/site/pictureguides/).

DNA extraction for DNA barcoding and RAPD analysis

The total genomic DNA was extracted from dried leaf specimens of pelawan using a modified Cetyl Trimethyl Ammonium Bromide (CTAB) method (Shiraishi and Watanabe 1995). Subsequently, the dried leaf was ground inside a 2 ml microtube using two 5 mm stainless steel beads (Qiagen) on a mini-bead beater-8 machine (Biospec) for 5 min or until fine powder formed. A 1.5 ml CTAB buffer was added to the leaf powder and shaken repeatedly for 2 min or until the liquid was homogenized. It was incubated at 65 °C for 1 h, then the supernatant was removed into a new 2 ml microtube, and 800 μl chloroform was added and mixed using a rotator RT-50 machine (Taitec) for 20 min. Furthermore, the mixture was separated by centrifugation at 12,000 rpm for 10 min using a Centrifuge 5417 R machine (Eppendorf), and the procedure was repeated by mixing 700 μl supernatant with 700 μl chloroform in a new 1.5 ml microtube. A volume of 600 μl supernatant from the second procedure was removed into a new 1.5 ml microtube, and DNA was precipitated by adding 20 μl NaOAc and 650 μl isopropanol. DNA pellets were rinsed using 1 ml of 70% cold ethanol, then re-rinsed using 1 ml of 100% cold ethanol. The pellets were dried using a vacuum system for ∼45 min, eluted in 200 μl of sterile water, and stored at 5 °C before conducting amplification using RAPD markers and a DNA barcoding marker.

Amplification of DNA barcoding, sequencing, and phylogeny construction

The total DNA was diluted 10 times using sterile water and used as a template for the PCR to amplify the ITS region. Amplification of the internal transcribed spacer (ITS) region was conducted using a combination of an ITS5p primer (5′ CCT TAT CAY TTA GAG GAA GGA G 3′) (Cheng et al. 2016) and an ITS4 primer (5′ TCC GCT TAT TGA TAT GC 3′) (White et al. 1990) to target the sequence of ITS1, 5.8s, and ITS2. Furthermore, the amplification was performed in 50 µl PCR mix containing 10 µl genomic DNA, 25 µl 2× My Taq HS Red Mix (Bioline), 0.5 µM of each primer, 0.2 µg bovine serum albumin (BSA), and sterile water. DNA amplification was also performed in a ProFlex PCR system (Applied Biosystem) with a program PCR consisting of denaturation at 94 °C for 5 min, 35 cycles of 94 °C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 60 s, and final extension at 72 °C for 7 min. This was similar to a previous study conducted on Pericopsis mooniana (Prihatini et al. 2020).

The electrophoresis to visualize the ITS DNA amplicon was conducted on agarose gel In a system using 1× TBE buffer at 110 voltages for ∼45 min. All the ITS DNA amplicons were sent to 1st Base (https://order.base-asia.com/) for standard sequencing using the Sanger sequencing methods. Meanwhile, DNA chromatograms were viewed and edited manually to remove poor-quality sequences at each end using Chromas software (http://technelysium.com.au/wp/chromas/). The DNA sequence showing good-quality chromatograms was then deposited at GenBank and used for barcoding and phylogeny analysis. Each sequence was blasted using the BLAST option against the nucleotide collection (Altschul et al. 1990), available on the NCBI website, to obtain the closest match and sequence DNA references (Altschul et al. 1990). The nucleotide sequences of the first- and second-best matches of species were retrieved from Genbank and included in further analysis to calculate sequence divergences and phylogeny reconstruction.

Edited DNA sequences were aligned using Clustal W (Thompson et al. 1994) on Bioedit software version 7.0.5.3 (Hall 1999). The alignment data were imported into MEGA11: Molecular Evolutionary Genetics Analysis version 11 software (Tamura et al. 2021) to construct a phylogeny tree. The construction aimed to verify the species identity of the specimens and assess the resolution of ITS barcoding in the phylogeny construction of Tristaniopsis. DNA sequences retrieved from GenBank nucleotide databases, which had the highest similarity shown by the BLAST online software, six nucleotide sequences of Tristaniopsis spp. From Sabah (Ahmad 2011), and an outgroup sequence from distantly related taxa were included in the analysis (Table 5). Furthermore, the phylogenetic tree was constructed using the Maximum Likelihood and the Tamura and Nei (1993) substitution model with a 500-bootstrapping test on MEGA11. This involved nucleotide references to related taxa retrieved from Genbank and the sequence of seven Tristaniopsis species from a previous study (Ahmad 2011).

Table 5. DNA sequences retrieved from GenBank nucleotide databases and six nucleotide sequences of Tristaniopsis from Sabah (Ahmad 2011) were included in the phylogeny reconstruction of pelawan.

Accession number	Species	Specimen ID	Origin	Sequence length (bp)
2EF041514	Tristaniopsis laurina	2080Tris	Australia	697
KM064886	T. laurina	M55 NZFRI 29049	New Zealand	683
n/a	T. elliptica	–	Borneo	637
n/a	T. grandifolia	–	Borneo	637
n/a	T. kinabaluensis	–	Borneo	637
n/a	T. rubiginosa	–	Borneo	624
n/a	T. whiteana	–	Borneo	709
n/a	T. whiteana Sentosa	–	Sentosa Island, Singapore	711
n/a.	T. sp nov 5	–	Borneo	608
KM064877	Metrosideros albiflora	M134 AK280103	New Zealand	677
KM064847	M. carminea	M335 NZFRI 29309	New Zealand	679
KM064856	M. erforate	M111 NZFRI 29090	New Zealand	676
KM064983	M. parkinsonii	M68 NZFRI 29065	New Zealand	674
EF026608	Tristania neriifolia	n/a	Australia	608
MK313872	Psidium cattleyanum	PSI 15	Brazil	786

DNA amplification and RAPD analysis

The genetic analysis of pelawan was carried out using RAPD markers. According to Table 6, a total of 85 RAPD markers were used in the screening process, and only primers that produced polymorphic and stable alleles were selected for the analysis. To test the reproducibility and polymorphisms of the RAPD fragments, 85 primers developed from an Operon 10-mer kit (Qiagen Co. Ltd. Tokyo) were evaluated in tiny samples and employed in three runs. Stable markers in amplification after repeated runs of three were then employed to increase the accuracy of the genetic diversity assessment using RAPD markers. In addition, the DNA amplification using RAPD markers was conducted following the previous study on Callophylum inophyllum (Nurtjahjaningsih et al. 2015). The 10 µl PCR reaction was consisted of 10× Stoffel buffer, 3 mM MgCl₂, 0.2 mM dNTPs, 0.05 unit of AmpliTaq Stoffel polymerase, 10 µM primer, and 10 ng/µl DNA. The reaction mixture was started with initial denaturalization for 5 min at 94 °C, continued with 45 cycles of denaturation for 90 s at 94 °C, annealing for 30 s at 37 °C and extension of 1.50 min at 70 °C, and finished with final extension for 5 min at 70 °C. The PCR process was conducted using the thermal cycler GeneAmp PCR system 9700 (Applied Biosystem).

Table 6. Primer name and number of RAPD markers for screening, amplified, polymorphic, and stable on pelawan.

	RAPD primers for screening	Amplified RAPD primers	Polymorphic and stable primers
	OPA1,OPA2,OPA3,OPA4,OPA5, OPA6,OPA7,OPA8,OPA9,OPA10, OPA11,OPA12,OPA13,OPA14,OPA15,OPA16,OPA17,OPA18, OPA19,OPA20,OPB4,OPB6, OPB7, OPB9, OPB13, OPB17, OPC1, OPC4, OPC6, OPC10, OPD1, OPD4, OPD5, OPD16, OPQ1, OPQ2, OPQ3, OPQ4, OPQ5, OPQ6, OPQ7, OPQ8, OPQ9, OPQ10, OPQ11, OPQ12, OPQ13, OPQ14, OPQ15, OPQ16, OPQ17, OPQ18, OPQ19, OPQ20, OPR1, OPR2, OPR3, OPR4, OPR5, OPR6, OPR7, OPR8, OPR9, OPR10, OPR11, OPR12, OPW1, OPW2, OPW3, OPW4, OPW5, OPW6, OPW7, OPW8, OPW9, OPW10, OPW11, OPW12, OPY1, OPY2, OPY3, OPY4, OPY8, OPY14	OPA2, OPA3, OPA13, OPB7, OPB17, OPC6, OPD2, OPD5, OPQ1, OPQ4, OPQ6, OPQ9, OPQ12, OPQ13, OPQ16, OPQ17, OPQ18, OPR1, OPR3, OPR4, OPR8, OPW1, OPW3, OPW4, OPW5, OPW9	OPA9, OPA10, OPA19, OPD2, OPD5, OPR3, OPR8, OPW3, OPW5, OPW9
Total	85	26	10

Data analysis for population genetic studies

Genetic diversity within the population was estimated using the H_E value, while the structure was assessed through AMOVA, and the GenAlex computer program (Peakall and Smouse 2012) was employed for both analyses. In the AMOVA analysis, a grouping of 96 individuals into two regions, namely Bangka and West Bangka, was conducted. The Bangka region consisted of four populations (PN, BB, LAP, and KK), while West Bangka included only LB.

To observe the correlation between populations, a dendrogram was constructed using the Poptrew computer program (Takezaki et al. 2014). Principal coordinate analysis (PCO) was performed using the Euclidean metric (Gower 1966), and the GENALEx program was used to generate graphical representations illustrating the relationships between different populations.

Results

Morphological observation on leaves and barks

Tristaniopsis spp., commonly known as pelawan, belonged to the family Myrtaceae and exhibited several shared characteristics. These included peeling barks, colored flowers, the absence of stipules, simple leaves with entire margins, and intra-marginal veins. Furthermore, the pelawan tree released a fresh scent when squeezed due to the presence of pellucid gland dots. The flowering was characterized by numerous stamens and the absence of latex. Other plant families that shared similarities with Myrtaceae included Rutaceae, Rubiaceae, and Malpighiaceae (The Royal Botanic Garden 2020).

During field collection, it was not possible to obtain flower and fruit organs, resulting in many specimens being sterile. Table 7 and Figure 2 provide a qualitative comparison between two Tristaniopsis species.

Figure 2. Herbarium of three Tristaniopsis sp. retrieved from Royal Botanic Gardens Kew.

Table 7. The comparison of morphological characteristic of two Tristianopsis species (Myrtaceae) from Bangka Island.

No	Organ	Pelawan Merah/Orange (Tristianopsis sp. 1)	Pelawan Air (Tristianopsis sp. 2)
1	Tree bark type	Peeling	Peeling
2	Bark color	Orangish-red or red	Pale red
3	Stipule	Absent	Absent
4	Leaves type	Simple	Simple
5	Leaves arrangement	Alternate and spirally arranged: dextrorse (right)	Alternate and spirally arranged: dextrorse (right)
6	Leaves margin	Entire, thickened, revolute	Entire, thickened, revolute
7	Leaves shape	Obovate, lanceolate	Oblanceolate
8	Leaves base	Cuneate	Cuneate
9	Leaves apex	Attenuate, acuminate	Acute, attenuate, obtuse, rounded
10	Adaxial surface of lamina	Punctate, green color	Punctate, green color
11	Abaxial surface of lamina	Villous, cream color	Villous, cream color
12	Venation	Pinnately veined with intra-margin, secondary vein with very closely parallel venation, running from midrib to the intra-margin. Tertiary veins are not visible.	Pinnately veined with intra-margin, secondary veins sparse, running from midrib to the intra-margin, and linked by clearly visible mesh-shaped tertiary veins.
13	Pellucid gland dots	Present with dense gland dots	Present with dense gland dots
14	Scanning electron microscope of the abaxial and adaxial leaf surfaces	Midrib has canaliculate in adaxial leaf surfaces. Abaxial leaf surface has trichomes and dense stomata.	Midrib has canaliculate in adaxial leaf surfaces. Abaxial leaf surface has trichomes and dense stomata.

rDNA ITS amplification and characterization of pelawan in Bangka Island

The amplification using a pair of ITS5p/ITS4 primers in the PCR produced rDNA ITS amplicon for all 12 pelawan specimens. The amplicon length from these specimens was ∼750 bp, as shown in Figure 3. The success rate of amplification using ITS5p/ITS4 primers was 100%, with all the specimens producing a strong and clear single band, except for one specimen (BB7), which produced faint double bands. A high success rate was also obtained for DNA sequencing since the specimens (12) produced clear DNA sequence chromatograms. The DNA sequence length was varied between 640 and 644 bp and submitted to GenBank, with accession numbers OL681855–OL681866 (Table 8). Furthermore, the BLAST analysis of the DNA ITS sequence of pelawan air (PN1 and PN2) specimens had the highest match to a range of T. laurina specimens from New Zealand (Buys et al. 2016) and Australia (Chong 2008) with ∼94% (609/646) sequence similarity (Table 8). According to the results, the closest matches to the pelawan air specimens were identified, where Metrosideros carminea, M. albiflora, M. parkinsonii, and Psidium cattleyanum exhibited a sequence similarity of 91% (588 out of 646 base pairs), 91% (587 out of 644 base pairs), 91% (586 out of 644 base pairs), and 91% (586 out of 645 base pairs), respectively. These results further highlighted the limited availability of Tristaniopsis DNA sequences in public databases, such as GenBank.

Figure 3. Amplification of twelve specimens of pelawan air (PN1, PN2), pelawan orange, and pelawan merah (PN3-KK12) using ITS5p/ITS4 primers; DNA fragment lengths were compared to a 100 basepair (bp) DNA ladder.

Table 8. DNA ITS amplification and sequencing product of 12 pelawan specimens collected from Bangka Island and their accession number in the Genbank database.

Specimen ID	Sequence length (bp)	Blast matching results	Genbank accession No
PN1	644	94% (609/646) similar to specimens of Tristaniopsis laurina, including KM064886	OL681855
PN2	644	Similar to PN1	OL681856
PN3	640	94% (605/646) similar to specimens of T. laurina, including KM064886	OL681857
PN4	640	Similar to PN3	OL681858
LB5	640	Similar to PN3	OL681859
LB6	640	Similar to PN3	OL681860
BB7	640	Similar to PN3	OL681861
BB8	640	Similar to PN3	OL681862
LAP9	640	Similar to PN3	OL681863
LAP10	640	93% (601/646) similar to specimens of T. laurina	OL681864
KK11	640	Similar to PN3	OL681865
KK12	640	Similar to PN3	OL681866

The best match species to pelawan merah (PN3-4, LB5-6, BB7-8, LAP9, KK11, and KK12) was T. laurina with 94% (605/646) sequence similarity, followed by Psidium longipetiolatum, M. albiflora and M. perforata, and M. carminea with 91% (586/645), 91% (584/644) and 91% (585/646) similarities. The exception of pelawan merah/orange was LAP10, which showed less nucleotide similarity (93%) to T. laurina. The sequences that displayed high similarity to all the examined specimens through BLAST analysis (T. laurina, M. albiflora, M. perforata, M. carminea, and P. cattleyanum) were obtained from the GenBank nucleotide databases. These sequences were included in the analysis with seven nucleotide sequences of pelawan obtained from a previous study (Ahmad 2011) and an outgroup sequence from a distantly related taxon (P. cattleyanum). The nucleotide of Tristaniopsis from Sabah (Ahmad 2011) was retrieved from the thesis manuscript as the sequences were not submitted to GenBank.

Phylogeny of pelawan based on ITS DNA

The specimens of pelawan, as shown in Table 3 produced a clear DNA sequence chromatogram, and the nucleotide sequences obtained were used for phylogeny reconstruction. The phylogeny tree involving nucleotide references of related taxa from Genbank and the sequence of seven Tristaniopsis species retrieved from a previous study (Ahmad 2011) (Figure 4), separated pelawan air from other pelawan trees. Furthermore, the phylogeny tree also separated T. laurina, the closest match (94%) to pelawan merah. The ITS barcoding marker provided high resolution, which resolved pelawan air (Tristaniopsis sp. 2) from pelawan merah (Tristaniopsis sp. 1). The ITS barcoding marker resolved two clades of Tristaniopsis, namely an Asia (Sabah and Bangka) and a Pacific clade (New Zealand and Australia). This marker also resolved the Tristaniopsis genus from other related genera (Metrosideros and Psidium) based on the nucleotide data on Genbank.

Figure 4. Molecular phylogenetic analysis of pelawan from Bangka Island by the Maximum Likelihood method and Tamura and Nei (1993) substitution model. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test 500 replicates are shown below the branches. Branches corresponding to partitions reproduced in <50% of bootstrap replicates are collapsed. This analysis involved 26 nucleotide sequences. There was a total of 650 positions in the final dataset, and Psidium longipetiolum was used as an outgroup. Samples generated in this study were highlighted.

Characteristics of the screened RAPD primers

A total of 10 out of 85 RAPD primers screened produced DNA bands, unambiguous, and reproducible, and were selected to study population genetics of Tristaniopsis spp. The number of alleles produced varied from 8 to 14, with a total of 106 alleles. The size ranged between 300 and 1400 bp, while polymorphic loci ranged from 92.31 to 100%, as shown in Table 9. These markers were used for further analysis to observe the genetic diversity of pelawan. Figure 5 shows a sample of the electropherogram of the DNA amplification band using OPD-02 primer.

Figure 5. Electropherogram of DNA amplification band using OPD-02 primer. Note: M: marker; bp: base pair; 1–24: Tristaniopsis sample.

Table 9. Characteristic of 10 RAPD primers on pelawan.

RAPD primers	Sequence 5′-3′	Annealing temperature (°C)	Number polymorphic of alleles	Polymorphic allele size (bp)	Polymorphism (%)
OPA-09	GGGTAACGCC	37	13	450, 500, 550, 600, 700, 750, 800, 900, 950, 1000, 1100, 1200, 1300	92.31
OPA-10	GTGATCGCAG	37	9	500, 550, 600, 650, 700, 850, 900, 1000, 1100	100
OPA-19	CAAACGTCGG	37	12	550, 600, 690, 700, 750, 800, 850, 950, 1100, 1200, 1300, 1400	100
OPD-02	GGACCCAACC	37	14	400, 500, 550, 600, 750, 800, 890, 900, 950, 1000, 1050, 1100, 1200	100
OPD-05	TGAGCGGACA	37	10	300, 400, 450, 530, 570, 650, 750, 800, 1000, 1200	100
OPR-03	ACACAGAGGG	37	11	300, 480, 490, 500, 550, 600, 700, 750, 800, 900, 1000	100
OPR-08	CCCGTTGCCT	37	12	300, 350, 400, 600, 630, 700, 730, 800, 850, 1050, 1100, 1300	100
OPW-03	GTCCGGAGTG	37	8	550, 600, 700, 750, 800, 850, 1000, 1100	100
OPW-05	GGCGGATAAG	37	9	400, 450, 530, 550, 650, 800, 850, 990, 1050	100
OPW-09	GTGACCGAGT	37	8	400, 500, 550, 650, 800, 950, 1000, 1200	100
	Total		106

Genetic diversity of pelawan

Table 10 shows the number of polymorphic loci and unbiased expected heterozygosity (uH_E) for each population. The average number of polymorphic loci was 103.2, ranging from 101 (KK) to 106 (LAP). The average uH_E was 0.317, with values ranging from 0.283 (KK) to 0.353 (PN).

Table 10. Population name, number of samples, number of polymorphic loci, unbiased expected heterozygosity (uH_E) values, and the standard of error (SE) within five populations.

Population code	Number of samples	Number of polymorphic loci	Mean uHE ± SE
KK	16	101	0.283 ± 0.016
BB	20	103	0.319 ± 0.015
LAP	20	106	0.327 ± 0.014
PN	20	104	0.353 ± 0.014
LB	20	102	0.302 ± 0.016
Total/Mean	96	103.2	0.317 ± 0.007

The genetic diversity within the population was estimated by heterozygosity (H_E) values, as shown in Table 11. The mean value of genetic diversity in the five contrarian populations was moderate (mean H_E = 0.317). Meanwhile, the lowest and highest H_E genetic diversity value was found in the KK and PN populations at H_E = 0.283 and H_E = 0.353.

Table 11. Population name, number of samples (N), expected heterozygosity (H_E) values, and the standard of error (SE) of pelawan in each population.

Population name	N	HE	SE
KK	16	0.283	0.016
BB	20	0.319	0.015
LAP	20	0.327	0.014
PN	20	0.353	0.014
LB	20	0.302	0.016
Mean		0.317	0.007

Genetic structure among populations

For the sampled members of the four populations, the top three principal coordinates from the PCO based on the Euclidean metric described 42.02, 26.76, and 17.72% of the overall variance. Figure 6 shows the plots of the first two coordinates, depicting the distribution of the four populations in Bangka as a single group, while the LB in West Bangka is positioned in a separate group.

Figure 6. A visualization of the first two primary coordinates from the PCO analysis based on Nei DA distances (Nei 1983) for the five populations of Tristaniopsis spp. The populations of Bangka grouped into one cluster, while those of West Bangka formed another cluster.

A dendrogram analysis shows a genetic relationship among five populations, and this classifies genetic distance according to geographic position. The population of PN, LAP, and BB forms a cluster with KK, while LP forms another cluster. The results of the PCO and the dendrogram indicate a similar pattern, where five populations in Bangka (KK, LAP, BB, and PN) and West Bangka (LB) are genetically distinct. As shown in Figures 6 and 7.

Figure 7. Dendrogram of five populations of pelawan in Bangka Island.

AMOVA analysis determined the hierarchy of 96 trees grouped in five populations (KK, LAP, BB, PN, and LB) from Bangka and West Bangka, as shown in Table 12. The amount of the genetic variance partitioned between, among, and within the population was 3.8, 8.6, and 87.6%, respectively. Each value was significant (p < 0.0001), and the AMOVA results supported the dendrogram analysis. Furthermore, the LB population was significantly different from Bangka, while PN was significantly different from the others.

Table 12. Analysis of molecular variance (AMOVA) of Tristaniopsis spp. used in this study for 96 individuals in five populations from two regions.

Variation sources	df	SS	MS	Variance (%)
Among regions	1	81,414	81,414	3.8***
Among populations	3	161,074	53,691	8.6***
Among individuals	91	1,708,700	18,777	87.6***
Total	95	1951.188		100

df: degree of freedom; SS: sum of square; MS: mean of square.

*** Significant p < 0.001.

Discussion

Tristaniopsis species (Myrtaceae) from Bangka Island are morphologically diverse. Morphological characteristics, namely stem color and leaf shape, can be used to differentiate species. Firstly, the bark color serves as a distinguishing factor, where Tristaniopsis sp. 1 exhibits bark that peels with an orangish-red hue, while Tristaniopsis sp. 2 showcases a pale-red bark. Secondly, the leaf morphology provides further differentiation. Tristaniopsis sp. 1 features obovate-lanceolate leaves with an attenuate apex, acuminate tip, and closely spaced secondary veins but the tertiary veins are not readily observable in this species. Meanwhile, Tristaniopsis sp. 2 has oblanceolate-shaped leaves, apex acute, attenuate, obtuse, rounded, and secondary veins sparse, and linked by visible mesh-shaped tertiary veins. The two species also differ in secondary and tertiary veins appearance. Tertiary venations are also widely used in research related to leaf architecture (Hernandez et al. 2020; Meinata et al. 2021).

The utilization of molecular characters offers a valuable method to augment rapid and precise species identification. DNA barcoding presents an accurate and alternative approach for identifying unknown species. However, the study of ITS characteristics of Tristaniopsis is still limited. This study has successfully amplified and sequenced the ITS region of Tristaniopsis on Bangka Island, Indonesia, with a higher success rate (100%) than other barcoding markers in Tristaniopsis (Ahmad 2011; Bolson et al. 2015; Buys et al. 2016). This study is the first report of Tristaniopsis barcoding distributed on Bangka Island. It provides the nucleotide sequences database of pelawan on Bangka Island, and showed that there were two groups, based on ITS sequence nucleotide, namely pelawan merah (Tristaniopsis sp. 1) and pelawan air (Tristaniopsis sp. 2). The application of DNA barcoding found that the variation of bark color on pelawan merah (from orangish-red to red) is not associated with the genetic structured based on ITS sequences. In the BLAST analysis using sequences of Bangka’s Tristaniopsis, the success rate of species identification was low, with the highest match to NCBI sequence reference at 94%. The low match score was related to a low number of Tristaniopsis DNA sequences submitted to public DNA databases, such as NCBI with a limited number of species. The Genbank database has 25 entries of Tristaniopsis species, including T. laurina, T. burmanica, T. collina, T. exiliflora, and Tristaniopsis sp. The BOLD database has nine sequence entries of Tristaniopsis (T. laurina, T. collina, T. exiliflora, and Tristaniopsis sp.). This study generated 12 DNA ITS sequences, and suggested that further study involving more pelawan should provide more nucleotide sequence databases for the identification of Tristaniopsis. However, based on ITS sequences, the species are not closely related to T. laurina or T. whiteana.

This is the first study on the genetic diversity of Tristaniopsis spp using RAPD markers. The ten markers showed a high percentage of polymorphism from 92 to 100% compared to other species at 62 to 75% (Normala et al. 2021). These primers have a high polymorphism and consist of many alleles used for population genetic analysis.

There is no information on the genetic diversity of Tristaniopsis spp. and cannot be compared to other species. Hamrick and Godt (1990) analyzed 449 species and estimated H_E of 0.157 in species. The genetic diversity of the pelawan populations is also higher than other species in Myrtaceae such Eucalyptus (H_E = 0.200) (Li 2004). Therefore, the mean H_E of the remnant population of Tristaniopsis spp in Bangka was considered moderate level (H_E = 0.317) compared with other natural tree species.

Despite a decline in the number of populations and individual trees on Bangka Island, the genetic diversity of the five populations remained at a moderate level. The high genetic diversity observed in these populations can be attributed to two key factors. Firstly, the number of remaining individual trees in each population is still considerable, due to effective management by the local communities. Even though large trees have been cut down, the pelawan trees have the ability to sprout easily, preventing their extinction (Sudirman et al. 2019).

The second factor contributing to the high genetic diversity is the random and crossing mating system prevalent in these populations. This ensures that the species exhibit high levels of heterozygosity to maintain genetic diversity. The relationship between mating systems and genetic diversity has been previously reported by Han et al. (2009) and Imai et al. (2016). Even though the information specific to the mating system in pelawan is lacking, the abundant flowering and insect-pollinated nature suggest a random mating system (Nurtjahjaningsih et al. 2019).

The range of genetic diversity of the five pelawan populations was small (H_E = 0.283–0.353), and the highest estimates were recorded from PN. The highest level of genetic diversity in this population may be due to the relatively large population as shown in Table 11. The PN population is a customary forest, where the local peoples are actively maintaining the integrity of the population. Therefore, the genetic diversity is expected to be higher than the other populations. The population also has more natural regeneration compared to the others (Table 4). This showed that reproduction success among mother trees in this population might be high. In this study, the KK and LB populations exhibited genetic diversity at the lower level of the observed range. LB is managed by a production forest management unit and one of the largest forests of Tristaniopsis spp. in Bangka Belitung Province. However, KK showed a low number of trees relative to the other populations, with the absence of flowers and poor natural regeneration. This population is characterized by failure in the reproduction system (Nurtjahjaningsih et al. 2019). Butcher et al. (2005) also reported that a limited number of low-density flowering trees might affect the abundance and behavior of pollinators. In contrast, LB is managed by a production forest management unit and one of pelawan in Bangka Belitung Province. The conservation in this area is competing with pepper farmers and fuelwood conducted by local people. The farmers or local people were cutting down young stands for poles of the pepper trees or household fuel.

The level of genetic distance among the five populations was moderate level (D_A = 0.04). This is similar to other species in Myrtaceae, such Eucalyptus urophylla (the fixation index (F_ST), which is an analog value with D_A = 0.031, Payn et al. 2008), lower than a vulnerable-extinct species, E. benthamii (F_ST = 0.105, Butcher et al. 2005).

The PCoA and a dendrogram analysis showed that LB (West Bangka) was genetically separated from the other populations of Bangka. Furthermore, AMOVA indicated a significant difference between West Bangka and Bangka (% variance = 3.8%). The local people may cultivate using seeds from surrounding places without admixture with other seed sources. Even though there were no private alleles in any populations, maintaining the genetic differentiation between two regions was an important thing useful for broadening genetic bases. Declining population size also contributes to a high level of genetic differentiation among populations (Payn et al. 2008). The observed significant genetic variance between regions and among populations can be attributed to the substantial distances exceeding the typical range over which insect vectors transport pollen (Nurtjahjaningsih et al. 2019). Furthermore, regional variations in the dynamics of widely dispersed species can be influenced by spatial variation in landscape barriers and climate gradients (Fahey et al. 2019). In the absence of long-distance dispersal mechanisms, tree species remain susceptible to rapid range shifts driven by climate changes.

In contrast to the idea of rapid range expansion, the presence of genetic diversity and a stronger population structure suggests a closer association with spatial conservatism and recurrent isolation (Hewitt 1996). The Pleistocene era also referred to as the “Ice Age,” occurred from ∼1,808,000 to 11,600 years ago. During this period, global temperatures were significantly lower, around 15 °C below current levels known as the age of glaciation. During this time, Kalimantan, Sumatra, Java, Bangka Belitung Islands, and Riau Islands were part of the mainland of Asia, forming a contiguous landmass (Bird et al. 2005; Wurster et al. 2010; Raes et al. 2014). The finding of elephant teeth from Elephas sumatranus in 1804 by F. Martin in a layer of tin deposits supports the idea that Bangka was connected to West Kalimantan, Sumatra Islands, and the mainland of Asia during the Pleistocene. This landmass disintegrated into small islands and confined shallow straits.

Certain sites have limited access and space to obtain more complete data and information. These issues may put certain sites in less favorable options as the genetic reserve area. Accessibility and the vast area provided should be considered to support monitoring, material mobility, and site development (O’Neill et al. 2001). Furthermore, site security and landholder should be considered in developing a management plan for the site to assure its sustainability (Grodzińska-Jurczak and Cent 2011).

In line with the arguments, LAP and PN are the proper sites to be developed as genetic reserve areas of pelawan. They have competitive advantages in terms of accessibility and security supporting the sustainability of the conservation program (O’Neill et al. 2001). Even though during the survey, the local community in PN showed resistance to the objective of using their traditional forest area as a conservation area, the rejection was based on the limited information distributed. The resistance observed can be seen as a positive reflection of the local community’s awareness and active management of their forests. This challenge can present a valuable opportunity by harnessing the awareness of the community and integrating the concept into participatory forest management, aligning with the conservation objectives (Tolo 2014; Tabb 2023). It is important to disseminate information regarding the benefits of conserving and managing pelawan stands among the community. This approach can also contribute to the long-term sustainability and security of the site.

Implications for conservation

The number of individual trees and the population of pelawan decrease yearly due to overexploitation and forest conversion (Nurtjahya et al. 2017; Siahaan et al. 2020). Therefore, a conservation strategy is needed to protect these species. The strategy for pelawan trees on Bangka Island should be developed with several considerations, such as (1) determinate pelawan trees as two taxa (pelawan merah/orange and pelawan air), (2) the populations that meet the conditions for in-situ conservation sites should be designated based on two taxa, and (3) ex-situ conservation plots should be established with at least one location for each taxon. Additionally, a comprehensive genetic investigation of the Tristaniopsis species necessitates the inclusion of more extensive populations that accurately reflect the natural distribution of every individual species.

Based on this genetic study, pelawan on Bangka Island was divided into two taxa groups, pelawan merah and pelawan orange in the same taxon (Tristaniopsis sp.1), and pelawan air in different taxon (Tristaniopsis sp.2). Pelawan merah and orange were naturally distributed in the dry land, while pelawan air is found close to river banks.

The genetic diversity in pelawan is closely linked to its geographical position since populations in close proximity exhibit genetic similarities. Substantial genetic variations are also evident between regions and among populations. According to Tian et al. (2012), the high genetic diversity observed in the remaining populations indicates that species conservation efforts carried out in-situ can effectively preserve their evolutionary potential. Considering the genetic diversity and the number of individual trees, the population of PN should be given the highest priority in the development of conservation strategies. This particular population of pelawan in PN shows the highest genetic diversity and maintains a sufficient number of mature trees. On the other hand, LB is recommended to enhance genetic diversity through an infusion strategy.

The current condition of pelawan genetic diversity and population structure develops appropriate sampling strategies for optimization and implementation of ex-situ conservation. The preservation of KK and LB should be a high priority to protect the remaining natural genetic resources. Ex-situ conservation strategies involve KK, LB, and PN, with the highest HE value. Therefore, an ex-situ conservation plot should be established by planting several samples, and each population must be represented by a suitable number of individual trees according to their genetic diversity.

Conclusions

In conclusion, a conservation strategy is important to preserve the existing species. This study provided genetic information, namely species identification and diversity of 5 populations, with baseline data for the conservation activities of pelawan on Bangka Island. Tree data and the distribution of existing trees from each population are also important information for the construction of in-situ and ex-situ conservation plots. The selection of PN as an in-situ conservation area is an important step toward improving pelawan on Bangka Island. The establishment of ex-situ conservation plots necessitates careful consideration of population distribution and genetic diversity but conducting an environmental analysis is imperative. For the development of the pelawan strategy, the genetic study of Tristaniopsis needs to be continued by increasing the genetic materials distributed in Indonesia.

Institutional Review Board statement

Not applicable.

Informed consent statement

Not applicable.

Author contributions

Y.W., A.Y.P.B.C.W., I.L.G.N., I.P., L.H., Y.H., A.I. P., A.S., T.H., S., Y., and M.H.S.: contributed in conceptualization, methodology, software, validation, formal analysis, investigation, resources, data curation, writing—original draft preparation, writing—review and editing, visualization, supervision, and project administration. All authors had an equal role as main contributors in discussing the conceptual ideas and the outline, providing critical feedback for each section, and writing the manuscript. All authors also have read and agreed to the published version of the manuscript.

Acknowledgments

The authors are grateful to the Forestry Agency of Bangka Belitung Province for fully supporting the genetic materials.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Not applicable.

Additional information

Funding

Part of this study was funded by the Local Government of Bangka Belitung.