The Southeast Asian Journal of Tropical Biology Vol. 32 No. 3, 2025: 373 - 384 DOI: 10.11598/btb.2025.32.3.2514 ISSN: 0215-6334 | e-ISSN: 1907-770X 373 REMNANT OF THE DRYLAND COASTAL VEGETATION OF THE SONGKHLA LAKE BASIN, SONGKHLA, THAILAND Anusit Cheechang* and Kitichate Sridith Department of Biology, Faculty of Science, Prince of Songkhla University, Songkhla 90230, Thailand ARTICLE HIGLIGHTS This article identifies and characterizes three native plant communities in the Songkhla Lake Basin, highlighting their distinct species composition, coastal zonation patterns, and ecological value. It reveals high plant diversity across remnant dryland coastal habitats and documents key native species that define each community type. The study emphasizes the negative impacts of human activity and land policies on native vegetation and offers baseline floristic data that support conservation, ecological restoration, and sustainable land-use planning. Article Information Received : 9 April 2025 Revised : 18 May 2025 Accepted : 28 May 2025 *Corresponding author, e-mail: c.anusit@gmail.com Research Paper ABSTRACT The remnant dryland coastal vegetation of the Songkhla Lake Basin, Songkhla, Thailand, was investigated in this study. Twenty vegetation plots were established to represent each vegetation subtype. Using Braun-Blanquet cluster analysis and the Jaccard similarity index, the vegetation was classified into three distinct communities: (1) the Coastal Woodland Community, located farthest inland and dominated by Dipterocarpus alatus Symington; (2) the Coastal Scrub-Shrubby Tree Community, situated inland from the coastal zone, with Dipterocarpus chartaceus Symington, Neolitsea zeylanica (Nees & T. Nees) Merr., and Psychotria asiatica L. identified as dominant species; and (3) the Coastal Scrub Community, occurring near the shoreline and characterized by densely clustered vegetation dominated by Mischocarpus sundaicus Blume, Planchonella obovata (R.Br.) Pierre, and Vitex pinnata L. Vegetation profiles for each community were delineated. Significant degradation and decline of native plant communities in the Songkhla Lake Basin have been caused by conservation activities involving the introduction of fast- growing non-native species and by governmental land tax policies. Keywords: dryland coastal vegetation, plant community structure, Songkhla Lake Basin INTRODUCTION Dryland coastal vegetation is recognized as a unique but often underappreciated component of tropical coastal ecosystems (Swiderska et al. 2018). These communities are typically found on sandy floodplains and elevated sandbars, where harsh environmental conditions are endured, including fluctuating salinity, limited freshwater input, and intense wind exposure. Vegetation structure and species composition are influenced by these stressors, making such areas valuable for investigations of plant resilience and adaptation in marginal environments (Freeman et al. 2023). Additionally, drylands are shaped by aeolian processes, in which sparse vegetation and erodible surfaces are subjected to strong winds. While reductions in wind speed and sediment transport can be facilitated by vegetation through drag, the overall effects are considered complex due to turbulence, porosity, and plant flexibility. Consequently, uncertainty remains regarding the precise role of vegetation in wind-driven erosion and deposition (Mayaud & Webb 2017). In Southeast Asia, ecological research in coastal regions has primarily been focused on mangrove forests and freshwater wetlands (Tatu & Anderson 2017; Daga & Salmo 2022; Liu et al. 2022; Veettil et al. 2024), whereas dryland coastal vegetation has been given disproportionately little attention (Marod et al. 2020). In Thailand, this knowledge gap is particularly evident, as much of the remaining dryland vegetation has been reduced to fragmented patches due to land-use changes, coastal development, and environmental degradation. These remnant patches are considered important since native species, ecological functions, and biogeographical signals may still be preserved, providing critical insights for understanding coastal ecosystem transitions and supporting conservation planning (Tulloch et al. 2015). Copyright (c) 2025@author(s). https://doi.org/10.11598/btb.2025.32.3.2514 https://creativecommons.org/licenses/by-nc-nd/4.0/ BIOTROPIA Vol. 32 No. 3, 2025 374 Previous studies have provided baseline floristic data and community classifications for portions of the southern coastline of Thailand (Laongpol 2003, 2010; Sridith 2014; Taing 2015). While these contributions are considered valuable, the focus was largely placed on species inventories and landscape profiles without explicit consideration of vegetation dynamics, environmental drivers, or conservation implications. Furthermore, the extent to which freshwater inflows, seasonal salinity, and geomorphological processes influence dryland coastal plant communities in Thailand has remained poorly documented. This study was conducted to address these knowledge gaps by investigating the remnant dryland coastal vegetation of the Songkhla Lake Basin, southern Thailand, through a phytosociological approach. Specifically, the objectives were defined as follows: 1) the structure and composition of the remaining dryland coastal plant communities were characterized, and 2) their distribution patterns were analyzed in relation to environmental conditions. By documenting and analyzing these vegetation remnants, critical knowledge was generated to support the conservation and management of coastal ecosystems, particularly in identifying potential areas for habitat protection, ecological restoration, and adaptive land-use planning within one of Thailand’s most ecologically complex coastal regions. MATERIALS AND METHODS Study Area Songkhla Lake, located in southern Thailand, is recognized as the largest lagoon system in Southeast Asia, covering approximately 1,040 km². It is situated across the provinces of Songkhla, Phatthalung, and parts of Nakhon Si Thammarat, between latitudes 6°40′ - 7°50′ N and longitudes 100°00′ - 100°40′ E. The lake is divided into four main sections: Thale Noi, a freshwater lake in the north; Upper Songkhla Lake (Thale Luang); Middle Songkhla Lake, which constitutes the brackish zone; and Lower Songkhla Lake, which directly connects to the Gulf of Thailand. The hydrology of the system is influenced by freshwater inflows from numerous small rivers and by seasonal seawater intrusion. Freshwater conditions dominate during the flood season (November - December), whereas brackish conditions prevail in the dry season due to reduced river inflows. Geomorphologically, the lagoon was formed by a series of coastal sandbars that created a semi-enclosed basin comprising diverse habitats, including peat swamps, estuarine areas, floodplains, and coastal dunes. These varied environments are considered ecologically significant due to the high biodiversity they support (Pongsaputra 1991; Tookvinas & Sirimontaporn 1988; Yokokawa 1984) (Fig. 1B). Figure 1 Study area and sampling plot Notes: A = Map of Thailand; B = Part of Songkhla Lake; C = Sampling plots (black dots), small relic patches (red dots). Remnant of dryland coastal vegetation in Songkhla Lake Basin - Cheechang and Sridith 375 Data Collection Field Survey Rapid land-use changes and human activities have resulted in a significant reduction of natural forest cover in the Songkhla Lake Basin, leaving only small and scattered forest remnants. Within these patches, notable canopy structures and remnants of native plant species are still preserved, and they were selected as the primary focus of this study. Site selection was carried out based on three main criteria: 1) the presence of natural vegetation as determined from satellite imagery and preliminary field visits; 2) accessibility to the site; and 3) the absence of invasive alien plant species. Field surveys were conducted from January to December 2023 to verify the presence of vegetation and to assess the suitability of the sites for phytosociological sampling. Vegetation Data Collecting A total of 20 study sites were selected across the Songkhla Lake Basin in Songkhla Province. Each selected site (indicated by black dots in Fig. 1C) encompassed an area of approximately 6,000 to 8,000 m² and was used for the collection of vegetation data. The coordinates of all study sites were recorded using GPS and subsequently mapped onto satellite imagery. During the survey, additional small forest patches (< 2,500 m²) were also identified (indicated by red dots in Fig. 1C). These patches were not selected for detailed sampling due to their limited species diversity and evidence of disturbance; however, they were noted as areas of ecological interest to reflect the extent and fragmentation of the remaining dryland vegetation in the basin. Vegetation data were collected using standard phytosociological methods. Sampling plots were established as 20 × 20 m for the tree layer, 5 × 5 m for shrubs, and 1 × 1 m for herbaceous plants. Within each plot, species composition, abundance (using the Braun-Blanquet scale), plant height, and vegetation structure were recorded. Vegetation Analysis The vegetation was analyzed following the conceptual framework and methodologies of the Zürich-Montpellier School (Braun-Blanquet 1964; Bridson & Forman 1998). Study plots were systematically established at the selected sites to ensure that each stand encompassed a minimum area with homogeneous physiognomic features, thereby maintaining uniformity. For each plot, a comprehensive species inventory was compiled by stratifying the vegetation into vertical layers. Multi-structured plant communities were classified into two or three layers, i.e., tree, shrub, and herb, based on their structural characteristics. The Braun-Blanquet method was employed to estimate species cover and sociability within each vegetation layer. Discrete plant communities were subsequently classified through the synthesis of the collected vegetation data. Profile diagrams and photographic records were constructed to illustrate stand structure at representative sites. To further analyze plant community associations, classical clustering analysis was performed using the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) based on the Bray- Curtis similarity index in PAST version 4.14. In this method, species were grouped into clusters according to similarity, beginning with the most similar pair and progressively merging less similar groups (Miyawaki & Suzuki 1980). The resulting dendrogram was used to elucidate ecological relationships among vegetation types across the study plots. In addition, the Jaccard Similarity Index (JSI) was calculated to assess the similarity in species composition between different plots. The index is expressed as: J (A, B) = A The value of the Jaccard Similarity Index is expressed on a scale ranging from 0 to 1, where a value of 0 is interpreted as indicating no shared species (representing entirely dissimilar communities), whereas a value of 1 is interpreted as indicating identical species composition (representing completely similar communities) (Bisandu et al. 2019). These analyses were used to provide insights into floristic relationships, habitat fragmentation, and ecological patterns among the remnant forest patches within the Songkhla Lake Basin. ⋂ B A ⋃ B BIOTROPIA Vol. 32 No. 3, 2025 376 RESULTS AND DISCUSSION Species Richness, Composition, and Plant Community A comprehensive survey of vascular plants within the study plots documented 64 species, which were distributed across 59 genera and 42 families. Among these, three species were identified as ferns: Drynaria quercifolia (L.) J.Sm., Pyrrosia longifolia (Burm.f.) C.V.Morton, and Pyrrosia piloselloides (L.) M.G.Price. The family Fabaceae exhibited the highest species richness, comprising four species, followed by Apocynaceae, Dipterocarpaceae, and Myrtaceae, each represented by three species. Several families, including Annonaceae, Combretaceae, Dilleniaceae, Lamiaceae, Malvaceae, Melastomataceae, Orchidaceae, Phyllanthaceae, Primulaceae, Rubiaceae, Rutaceae, and Salicaceae, were represented by two species each, while 29 families were represented by a single species (Table 1). This composition reflects the diverse floristic structure of the study plots, with Fabaceae contributing the greatest species richness among the recorded families. Table 1 Plant species, their taxonomic families, and associated plant communities No Species name Family Com. Ab. 1 Uvaria ferruginea var. cherrevensis (Pierre ex Finet & Gagnep.) Meade & J.Parn. Annonaceae C + 2 Uvaria siamensis (Scheff.) L.L.Zhou, Y.C.F.Su & R.M.K.Saunders C + 3 Amphineurion marginatum (Roxb.) D.J.Middleton Apocynaceae B + 4 Dischidia major (Vahl) Merr. B + 5 Hoya verticillata (Vahl) G.Don var. verticillata B + 6 Alocasia sanderiana W.Bull Araceae A + 7 Caryota mitis Lour. Arecaceae A + 8 Dianella ensifolia (L.) Redouté Asphodelaceae B + 9 Garcinia celebica L. Clusiaceae C + 10 Combretum trifoliatum Vent. Combretaceae C + 11 Terminalia catappa L. C + 12 Cyanotis cristata (L.) D. Don Commelinaceae B + 13 Dillenia hookeri Pierre Dilleniaceae C + 14 Tetracera indica (Christm. & Panz.) Merr. C + 15 Dipterocarpus alatus Roxb. ex G.Don Dipterocarpaceae A ++++ 16 Dipterocarpus chartaceus Symington B,C ++++ 17 Anthoshorea roxburghii (G.Don) P.S.Ashton & J.Heck. C + 18 Elaeocarpus robustus Roxb. Elaeocarpaceae C + 19 Vaccinium bracteatum Thunb. Ericaceae B + 20 Derris trifoliata Lour. Fabaceae C + 21 Guilandina bonduc L. C + 22 Biancaea sappan (L.) Tod. C + 23 Abrus precatorius L. B,C ++ 24 Vitex pinnata L. Labiatae C +++ 25 Clerodendrum indicum (L.) Kuntze Lamiaceae C + 26 Vitex glabrata R.Br. C + 27 Neolitsea zeylanica (Nees & T. Nees) Merr. Lauraceae B,C ++ 28 Sterculia villosa Roxb. ex Sm. Malvaceae C + 29 Microcos tomentosa Sm. C+ 30 Memecylon edule Roxb. Melastomataceae B,C + 31 Memecylon ovatum Sm. C + 32 Sandoricum koetjape (Burm.f.) Merr. Meliaceae A + Remnant of dryland coastal vegetation in Songkhla Lake Basin - Cheechang and Sridith 377 No Species name Family Com. Ab. 33 Tinospora crispa (L.) Hook.f. & Thomson Menispermaceae A + 34 Streblus asper Lour. Moraceae A,B,C ++ 35 Syzygium antisepticum (Blume) Merr. & L.M.Perry Myrtaceae C + 36 Syzygium grande (Wight) Walp. A,B + 37 Rhodomyrtus tomentosa (Aiton) Hassk. B,C + 38 Ochna integerrima (Lour.) Merr. Ochnaceae C + 39 Olax scandens Roxb Olacaceae C + 40 Tetrapilus brachiatus Lour. Oleaceae C + 41 Cymbidium findlaysonianum Lindl. Orchidaceae B + 42 Dendrobium crumenatum Sw. B + 43 Pandanus tectorius Parkinson Pandanaceae C + 44 Phyllanthus vitis-idaea (Burm.f.) J.Koenig ex Roxb. Phyllanthaceae C + 45 Phyllanthus subscandens var. subscandens C + 46 Bambusa bambos (L.) Voss Poaceae A + 47 Drynaria quercifolia (L.) J.Sm. Polypodiaceae B + 48 Pyrrosia longifolia C.V.Morton B + 49 Pyrrosia piloselloides (L.) M.G.Price B + 50 Ardisia crenata Sims Primulaceae B + 51 Ardisia elliptica Thunb. B + 52 Ziziphus oenopolia (L.) Mill. var. brunoniana (C.B.Clarke ex A.W.Hill) Tardieu Rhamnaceae C + 53 Carallia brachiata (Lour.) Merr. Rhizophoraceae A,C + 54 Catunaregam tomentosa (Blume ex DC.) Tirveng. Rubiaceae B,C + 55 Psychotria asiatica L. B ++ 56 Atalantia monophylla (L.) DC. Rutaceae C + 57 Glycosmis pentaphylla (Retz.) DC. B + 58 Flacourtia indica (Burm. f.) Merr. Salicaceae C + 59 Casearia grewiifolia Vent. A + 60 Dendrotrophe umbellata (Blume) Miq. Santalaceae B + 61 Lepisanthes rubiginosa (Roxb.) Leenh. Sapindaceae A,C + 62 Mischocarpus sundaicus Blume C +++ 63 Planchonella obovata (R.Br.) Pierre Sapotaceae C +++ 64 Alpinia malaccensis (Burm.f.) Roscoe Zingiberaceae A + Notes: Com. = community; Ab. = abundance; A = Coastal woodland community; B = Coastal scrub-shrubby tree shrub community; C = Coastal scrub community. The Songkhla Lake Basin is situated on the Sathing Phra Peninsula, a narrow and elongated landform extending along the longitudinal axis of Songkhla Lake from its northern to southern reaches. The plant communities within this region have been subjected to prolonged anthropogenic disturbances over time (Prapruit et al. 2023). Nevertheless, certain areas still retain remnants of the original vegetation, which can be used as important indicators for reconstructing the potential native plant composition that once characterized the Songkhla Lake Basin ecosystem. The cluster analysis and Jaccard similarity matrix based on species composition distinguished three types of plant communities (Fig. 5; Table 2). The Woodland community is represented by the plant community situated farthest inland. The Scrub community is represented by the plant community occurring closest to the coastline. The Scrub- Woodland mixed community is represented by the transitional zone between these two communities, reflecting a gradient of environmental conditions and species composition. The proposed plant communities and their species richness are presented in Figure 2. BIOTROPIA Vol. 32 No. 3, 2025 378 A. C. B. Figure 2 Coastal vegetation Notes: A. Coastal Woodland Community (Plot No.20); B. Coastal Scrub-Shrubby Tree Community (Plot No.10); and C. Coastal Scrub Community (Plot No.3). (A) Coastal Woodland Community The plant community located farthest inland was represented by four sampling plots (Plots 17 - 20), in which 12 vascular plant species were recorded. Dipterocarpus alatus Symington, a characteristic emergent tree species, dominated the canopy layer. The understory layer was composed of Bambusa bambos (L.) Voss, Carallia brachiata (Lour.) Merr., Casearia grewiifolia Vent., Lepisanthes rubiginosa (Roxb.) Leenh., Sandoricum koetjape (Burm.f.) Merr., and Syzygium grande (Wight) Walp., which were characteristic components of the coastal sand ridge plant community. Their presence had also been documented in previous studies (Congdon 1982; Sridith 2002; Laongpol 2010), contributing to the structural complexity of the community. The ground layer was represented by two herbaceous species, Alocasia sanderiana W. Bull and Alpinia malaccensis (Burm.f.) Roscoe, and one climbing species, Tinospora crispa (L.) Hook.f. & Thomson, was also recorded. Additionally, seedlings of Ardisia elliptica Thunb., Caryota mitis Lour., and Streblus asper Lour. were observed (Figs. 3 & 2A). According to Laongpol (2010), plant communities containing Dipterocarpus alatus Symington were classified as dune woodland communities. This classification is consistent with the findings of the present study; however, the observed species diversity was lower than that reported in previous research, likely due to human habitation and various land-use activities in the surrounding area, which may have impacted the vegetation at the study site. Results from cluster and similarity analyses indicated that this plant community, comprising Plots 17, 18, 19, and 20 (Fig. 6), exhibited the highest level of similarity within the group. The short branch lengths in the dendrogram, coupled with moderate similarity indices in the Jaccard matrix (e.g., Plot 17 vs. Plot 18: 0.25; Plot 19 vs. Plot 20: 0.13; Table 2), suggest that these plots share a common set of species. This high degree of resemblance is attributed to their proximity to similar environmental conditions. (B) Coastal Scrub-Shrubby Tree Community The area inland from the coastal zone was surveyed using 12 sampling plots (Plots 7 - 14), in which 24 plant species were documented. The canopy layer was predominantly characterized by Dipterocarpus chartaceus Symington, which emerged as the dominant species and was consistently recorded across all surveyed plots. Neolitsea zeylanica (Nees & T. Nees) Merr. was present within the sub-canopy layer, while Psychotria asiatica L. was ubiquitously recorded, highlighting its ecological significance in this stratum. The understory layer was primarily composed of Dianella ensifolia (L.) Redouté, representing the most abundant species at this level. A notable presence of the climbing species Hoya verticillata (Vahl) G.Don var. verticillata was observed Remnant of dryland coastal vegetation in Songkhla Lake Basin - Cheechang and Sridith 379 throughout the plots. Furthermore, two orchid species, Cymbidium findlaysonianum Lindl. and Dendrobium crumenatum Sw., were recorded, indicating a diversity of epiphytic flora. Fern species, including Drynaria quercifolia (L.) J.Sm., Pyrrosia longifolia (Burm.f.) C.V.Morton, and Pyrrosia piloselloides (L.) M.G.Price, were also documented, contributing to the compositional and structural complexity of the community (Figs. 4 & 2B). Compared to previous studies by Laongpol (2010), several plant species recorded in the present study are classified within the dune scrub community, including Dianella ensifolia (L.) Redouté, Hoya verticillata (Vahl) G.Don var. verticillata, and Dischidia major (Vahl) Merr. Figure 3 Community A (Plot No. 17) Notes: 1 = Alocasia sanderiana W.Bull; 2 = Caryota mitis Lour.; 3 = Bambusa bambos (L.) Voss; 4 =Streblus asper Lour.=; 5 =Carallia brachiata (Lour.) Merr.; 6 = Dipterocarpus alatus Roxb. ex G.Don; 7 =Lepisanthes rubiginosa (Roxb.) Leenh.; 8 =Sandoricum koetjape (Burm.f.) Merr.; 9. Alpinia malaccensis (Burm.f.) Roscoe. Figure 4 Figure 4 Community B (Plot No. 8) Notes: 1 =Cyanotis cristata (L.) D.Don; 2 = Neolitsea zeylanica (Nees & T.Nees) Merr.; 3 = Psychotria asiatica L.; 4 = Hoya verticillata (Vahl) G.Don var. verticillata; 5 = Caryota mitis Lour.; 6 = Drynaria quercifolia (L.) J.Sm.; 7 = Abrus precatorius L.; 8 = Dipterocarpus chartaceus Symington. BIOTROPIA Vol. 32 No. 3, 2025 380 (C) Coastal Scrub Community This plant community, located near the shoreline, is characterized by densely clustered vegetation and comprises eight plots (Plots 1 - 5, 15, and 16). It exhibits the highest species diversity among all communities, with 39 species recorded. The dominant species within this community include Mischocarpus sundaicus Blume, Planchonella obovata (R.Br.) Pierre, and Vitex pinnata L. Dipterocarpus chartaceus Symington was the most frequently encountered tree species. Several tree species in this community exhibited a shrubby crown structure, despite typically developing regular, well-formed crowns when growing further inland. Notably, 29 species recorded in this community were absent from the previously described plant communities, highlighting its unique floristic composition (Figs. 5 & 2C). The natural forest patches studied exhibited clear zonation. Salt-tolerant plant species were observed in the frontal zone closest to the saltwater. Behind this zone, species such as Pandanus tectorius (Merlin 1999) and Planchonella obovata (R.Br.) functioned as a buffer against salinity intrusion. Further inland, dipterocarp species (Dipterocarpus alatus Roxb. ex G.Don, Dipterocarpus chartaceus Symington) were present; these species are less salt-tolerant but were protected by the vegetation in the frontal zones. When the frontal zone vegetation was removed, the inner zones were also negatively affected due to increased salinity. In certain cases, human-made structures or cultivated plants were employed as barriers to limit salt intrusion (Su et al. 2025). The gradual disappearance of dipterocarp species was primarily attributed to human harvesting rather than natural mortality, as natural plant death was rarely observed; anthropogenic activities were the predominant cause of vegetation loss. The analysis results included plots 4, 7, 8, 9, 10, 11, 12, and 13 (Fig. 6). Notably, several plots within this community, particularly plots 9, 10, 11, and 12, exhibited high similarity values (up to 0.67, Table 2), as reflected by the shorter branches in the dendrogram. This finding indicates a strong floristic cohesion, which may be attributed to shared habitat characteristics, such as moderate exposure to environmental stress or similar successional stages. The relatively higher species turnover observed between plots in this community, compared to Community A, suggests the presence of microhabitat heterogeneity or slight variations in disturbance regimes. The findings of the present study are largely consistent with those reported by Laongpol (2010) regarding the composition of this plant community. Analysis of Community C revealed the greatest heterogeneity, with plots 1, 2, 3, 5, 6, 14, 15, and 16 exhibiting longer branch lengths and lower similarity values across the matrix. The isolation of plots such as 1, 2, and 5 reflects their distinct species assemblages. These patterns suggest that Community C may encompass a broader range of environmental conditions or historical land-use impacts, resulting in increased species variability. Additionally, the presence of unique species not shared with other communities indicates reduced influence from coastal stressors and potentially more stable or mature vegetation. An interesting ecological observation is that tree species within certain plots exhibit a shrubby crown form near the coastline but develop a normal canopy structure further inland. This morphological adaptation is likely a response to abiotic stressors such as strong winds, high salinity, or nutrient-poor soils in the coastal zone, reinforcing the clustering patterns observed in Community A (Jiang et al. 2025). Overall, the strong correspondence between the dendrogram and the similarity matrix supports the robustness of the clustering results. The differentiation among the three communities appears to reflect gradients of environmental stress, disturbance history, and microhabitat variation. Future analyses incorporating specific environmental parameters, such as soil salinity, elevation, and wind exposure, are recommended to further elucidate the ecological drivers shaping these plant communities. Remnant of dryland coastal vegetation in Songkhla Lake Basin - Cheechang and Sridith 381 Figure 5 Community C (Plot No. 4) Notes: 1 = Neolitsea zeylanica (Nees & T. Nees) Merr.; 2 = Vitex pinnata L.; 3 = Olax scandens Roxb; 4 =Derris trifoliata Lour.; 5 = Mischocarpus sundaicus Blume; 6 = Ziziphus oenoplia (L.) Mill.; 7 = Abrus precatorius L.; 8 = Planchonella obovata (R.Br.) Pierre; 9 = Syzygium antisepticum (Blume) Merr. & L.M.Perry; 10 = Garcinia celebica L; 11 = Flacourtia indica (Burm. f.) Merr. Figure 6 The cluster analysis dendrogram presents the results of a hierarchical cluster analysis (HCA) based on the similarity of 20 vegetation plots Notes: The plots are grouped into three distinct plant communities, labeled as: Community A comprises plots 17, 18, 19, and 20. These plots are highly similar, as indicated by the short branch lengths within this cluster, suggesting strong compositional resemblance among these plots. Community B consists of plots 4, 7, 8, 9, 10, 11, 12, and 13. Although this community exhibits more internal variation than Community A, the plots still share significant floristic similarities. Community C includes plots 1, 2, 3, 5, 6, 14, 15, and 16. Plots within this group demonstrate distinct characteristics, especially plots 1, 2, and 5, which show longer branch lengths, indicating lower similarity to other plots. BIOTROPIA Vol. 32 No. 3, 2025 382 Table 2 Jaccard similarity matrix between the plots Conservation Measures The dryland coastal vegetation of the Songkhla Lake Basin, Songkhla, Thailand, is considered significant for regulating water dynamics in the coastal dunes. It contributes to the maintenance of water levels accumulated in the sandbars, originating both from mountain runoff along the descending slopes of the basin, driven by gravity, and from rainfall that is collected within the sandbars themselves (Finkl & Makowski 2017). The composition, structure, and diversity of the plant community have been significantly influenced by reforestation efforts and government policies. In past reforestation programs, certain tree species may have been introduced, natural regeneration processes altered, and land-use patterns modified, all of which affected native vegetation dynamics (Willium et al. 2024). Furthermore, government policies concerning land management, conservation, and development have shaped the establishment and persistence of these plant communities over time (Talerngsri 2020). Conservation activities have included reforestation projects that introduced fast-growing invasive species, such as Casuarina equisetifolia L. and Acacia mangium Willd., which inhibit the growth of native plants and destroy existing seed banks and seedling reserves of indigenous species. Coastal zones are characterized by high biodiversity and complex interconnections among terrestrial and marine ecosystems, as well as social and governance systems. Despite ongoing conservation efforts, unsustainable resource use continues to persist, presenting significant governance challenges due to the complexity and cross-sectoral nature of these issues. Effective management requires an interactive governance approach that accounts for the relationships among ecological, social, and governing systems at multiple levels: understanding the complexity of coastal zones (first order), improving legal and institutional frameworks (second order), and establishing principles and values to guide challenging decisions (meta order) (Satumantpan & Chuenpagdee 2022). In the Songkhla Lake Basin, the destruction of dryland coastal vegetation has been notably influenced by government policies (Wear & Wibbenmeyer 2023). For instance, new land tax laws have compelled landowners to remove remaining natural plant communities classified as forest patches to avoid high taxes. To comply, crops specified under the law have been cultivated in place of native vegetation, leading to the near- elimination of indigenous plant communities on the dry coastal dunes of the basin. However, the environmental impacts of these government policies have been insufficiently studied. This lack of comprehensive research has allowed the continued destruction of valuable coastal vegetation, resulting in the rapid loss of ecologically important areas. Remnant of dryland coastal vegetation in Songkhla Lake Basin - Cheechang and Sridith 383 CONCLUSION The remnant dryland coastal vegetation of the Songkhla Lake Basin represents a highly diverse yet fragile ecosystem shaped by both natural gradients and intense anthropogenic pressures. Three distinct plant communities, i.e., coastal woodland, coastal scrub-scrubby tree, and coastal scrub communities, reflect ecological adaptations to inland-to-shoreline conditions but remain increasingly threatened by land-use change, harvesting, invasive species, and misguided policies. The primary drivers of degradation are not natural processes but human interventions, particularly land tax laws and inappropriate reforestation practices. Effective conservation requires an integrated governance approach that recognizes ecological functions, reforms counterproductive policies, and prioritizes the intrinsic value of native ecosystems. Without timely and coordinated action, these unique and irreplaceable plant communities face imminent and irreversible loss. ACKNOWLEDGMENTS This research was generously supported by the Science Achievement Scholarship of Thailand (SAST) and the Division of Biological Science, Faculty of Science, Prince of Songkhla University, Hat Yai, Songkhla Province, Thailand, which provided all research facilities for the present study. REFERENCES Bisandu DB, Prasad R, Liman MM. 2019. Data clustering using efficient similarity measures. JSMS 22(5):23. DOI: 10.1080/09720510.2019.1565443 Braun-Blanquet J. 1964. Pflanzensoziologie: Grundzüge der Vegetationskunde [Plant sociology: fundamentals of vegetation science]. Third Edition. Vienna (AT): Springer- Vienna. Bridson D, Forman L. 1998. The herbarium handbook. Third Edition. London (UK): Whitstable Litho Printers Ltd. 256p. Congdon G. 1982. The vegetation of Tarutao Natural Park. NHBSS 30:135-98. Freeman K, Subedi SC, Ross MS. 2023. Coastal dry tropical forests in Florida and the Caribbean in peril: A review. Biotropica 1(3):1-13. DOI: 10.1111/btp.13285 Gerona-Daga MEB, Salmo SG III. 2022. A systematic review of mangrove restoration studies in Southeast Asia: Challenges and opportunities for the United Nation’s decade on ecosystem restoration. Front Mar Sci 9:987737. DOI: 10.3389/fmars.2022.987737 Jiang Z, van Zanten M, Sasidharan R. 2025. Mechanisms of plant acclimation to multiple abiotic stresses. Commun Biol 8:655. DOI: 10.1038/s42003-025-08077-w Laongpol C. 2003. Flora and vegetation along the coast in Narathiwat [Thesis]. Songkhla (TH): Prince of Songkhla University. [Internet]. [cited 2024 Oct 9]. Available from: http://kb.psu.ac.th/psukb/handle/2553/1369 Laongpol C. 2010. Phytosociology of the terrestrial vegetation along the coasts of the Peninsular of Thailand. [Dissertation]. Songkhla (TH): Prince of Songkhla University. [cited 2024 Oct 9]; Available from: http:// kb.psu.ac.th/psukb/handle/2016/10359 Liu Y, Zhang H, Cui Z, Zuo Y, Lei K, Zhang J, …, Ji P. 2022. Precise wetland mapping in Southeast Asia for the Ramsar Strategic Plan 2016-24. Remote Sens 14(22):5730. DOI: 10.3390/rs14225730 Marod D, Sungkaew S, Mizunaga H, Thinkampheang S, Thongsawi J. 2020. Woody plant community and distribution in a tropical coastal sand dune in southern Thailand. J Coast Conserv 24(4). DOI: 10.1007/s11852- 020-00761-9 Mayaud JR, Webb NP. 2017. Vegetation in drylands: Effects on wind flow and aeolian sediment transport. Land 6(3):64. DOI: 10.3390/land6030064 Merlin M. 1999. Hawaiian forest plants: An illustrated field guide. Honolulu (US): Pacific Guide Books. 81 pp. Miyawaki A, Suzuki K. 1980. Process of phytosociological studies and vegetation mapping. IEST 6(1):65-76. Pongsaputra P. 1991. Illustrated landforms of Thailand. Bangkok (TH): Darnsutha Press. 300p. Prapruit P, Chuaduangpui P, Choosuk C. 2023. Adaptation and coping ability of rice farmers regarding the saltwater intrusion along the upper area of Sathing Phra Peninsula of Thailand. Bulg J Agric Sci 29(1):70-9. Satumantpan, S., Chuenpagdee, R. 2022. Interactive governance for the sustainability of marine and coastal resources in Thailand. Environ Nat Resour 20(6):543-52. Sridith K. 2014. The remnant of vegetation on a coastal sandbar in Songkhla Province, Peninsular Thailand. TFB (Botany) 30:49-58. Su Q, Kambale RD, Tzeng J-H, Amy GL, Ladner DA, Karthikeyan R. 2025. The growing trend of saltwater intrusion and its impact on coastal agriculture: Challenges and opportunities. Sci Total Environ 966:178701. DOI: 10.1016/j.scitotenv.2025.178701 Swiderska K, King-Okumu C, Islam M. 2018. Ecosystem- based adaptation: A handbook for EbA in mountain, dryland and coastal ecosystems. London (UK): International Institute for Environment and Development (IIED). Taing L. 2015. The woody plants architectural models in the Songkhla Lake Basin, Peninsular Thailand [Thesis]. Songkhla (TH): Prince of Songkhla University. [cited 2024 Nov 13] Available from: http://kb.psu.ac.th/psukb/ handle/2010/10204. BIOTROPIA Vol. 32 No. 3, 2025 384 Talerngsri K. 2020. Forests, peoples, and governments: Persistent land-use conflict in northern Thailand. Econ Peace Secur J 15(2). DOI: 10.15355/epjs15.2.19 Tatu K, Anderson J. 2017. An introduction to wetland science and South Asian wetlands. In: Kumar Prusty BA, Chandra R, Azeez PA (Editors), Wetland Science: Perpectives from South Asia. p.3-30. New Delhi (IN): Springer New Delhi. DOI: 10.1007/978-81-322-3715-0_1 Toonkvinas S, Sirimontaporn P. 1998. Ecological properties review of Songkhla Lake in Report of Thailand and Japan Joint Coastal Aquaculture Research Project. Tulloch AIT, Barnes MD, Ringma J, Fuller RA, Watson JEM. 2016. Understanding the importance of small patches of habitat for conservation. J Appl Ecol 53(2):418-29. DOI: 10.1111/1365-2664.12547 Veettil BK, Tran X-L, Lopes MS. 2024. Mangrove vegetation changes and shoreline erosion along the central coastline of Vietnam: A study from Quang Ngai Province. Environ Dev 52:101065. DOI: 10.1016/j.envdev.2024.101065 Wear D, Wibbenmeyer M. 2023. Land-use change, no-net- loss policies, and effects on carbon dioxide removals. DOI: 10.21203/rs.3.rs-3739874/v1 Williams BA, Beyer HL, Fagan ME, Chazdon RL, Schmoeller M, Sprenkle-Hyppolite S, …, Crouzeilles R. 2024. Global potential for natural regeneration in deforested tropical regions. Nature 636:131-37. DOI: 10.1038/s41586-024- 08106-4 Yokokawa T. 1984. Report on aquaculture ground survey of Songkhla Lake. Songkhla (TH): National Institute of Coastal Aquaculture.