Vietnam lies within the Indo-Burma Region, recognised as one of the 36 global biodiversity hotspots. The country's significant habitat loss and the presence of numerous rare species make it a critical area that demands high priority in biodiversity conservation efforts (Duke et al. 2010,Habel et al. 2019). Bruguiera hainesii C.G.Rogers 1919, has a very limited distribution in coastal areas of Southeast Asia and belongs to the Rhizophoraceae family (Tran et al. 2023). Despite the species being situated within a protected forest area, where it faces less impact from human activities like agricultural development and shrimp farming, climate change is accelerating the loss of mangrove habitats. Elevated temperatures and rising sea levels are worsening coastal erosion. For many years, B. hainesii has not been studied for its ecological characteristics due to its narrow distribution on Con Dao Island, which is located more than 100 km from the mainland (Tran et al. 2023). Although B. hainesii is thought to be a narrow-range species, efforts must continue to balance the development and preservation of this species' ecosystem, as biodiversity in mangrove forests plays a crucial role in maintaining the health of coastal ecosystems. This can be particularly achieved through scientific research to better understand its growth, facilitating the propagation of B. hainesii and other threatened species (Hoang 2013)

Research indicates that, under high-emission scenarios, the presence and relative density of mangrove forests are likely to increase significantly in the northern Gulf of Mexico and along the south-eastern coast of the United States (Bardou et al. 2024). Research suggests that warming temperatures and changing precipitation patterns could create favourable conditions for mangroves to expand their range and density in these areas. This expansion could lead to shifts in local ecosystems, influencing coastal biodiversity, carbon storage and shoreline protection in regions that historically have not had extensive mangrove forests (Smee et al. 2017). Understanding these changes is essential for anticipating the broader impacts on coastal communities, infrastructure and fisheries, making this research critical for future environmental and policy planning.

Another study hypothesises that black mangroves will expand their range to higher latitudes in the 21^st century due to global climate change (Osland et al. 2019). The main factor driving this change is the decrease in both the frequency and intensity of coastal cold spells, which historically constrained the growth of mangrove stands and the size of individual trees. This reduction in cold spells has allowed mangroves to expand their range. One potential area for such expansion is the Gulf of Mexico, situated at the northern edge of the black mangrove’s habitat. The Gulf’s relatively milder temperatures could facilitate the northward movement and growth of black mangrove populations, which were previously limited by colder conditions (Comeaux et al. 2012). Over approximately 35 years, mangrove coverage increased by 4.3% from 1980 to 2015, with the most significant expansion in Texas and southern Florida, owing to mild winters without subzero temperatures (Giri and Long 2016).

An important factor that conservationists might find crucial for assessing a species' adaptation to climate change is the ability to identify specific indicators (Pham et al. 2024). These indicators can provide valuable insights into how a species responds to shifting environmental conditions, such as changes in temperature, precipitation or habitat availability (Pham et al. 2024). By monitoring these indicators, conservationists can better understand the species' resilience and vulnerability to climate-related impacts. This information is vital for developing effective conservation strategies and management plans aimed at mitigating the effects of climate change on biodiversity and ensuring the long-term survival of the species in question. In the past decade, the rapid development of machine-learning has provided significant benefits, serving as an effective tool for creating potential species distribution maps, which are essential in biodiversity conservation programmes (Beery et al. 2021). These maps help identify areas where species may survive and thrive in the future, thereby supporting more effective ecosystem management and protection. In Vietnam and some neighbouring countries, ecological modelling studies have been successfully conducted, offering clearer insights into the potential habitats of wildlife species (Pham et al. 2024). These models use machine-learning algorithms to analyse environmental and biological data, predicting species distributions (ENM) under changing environmental conditions. This is particularly important in the context of climate change and altering natural landscapes, where species need to adapt to new conditions to survive (Crego et al. 2022).

This study aims to present new findings on the total number of B. hainesii individuals in Vietnam. Using a machine-learning algorithm (Maxent), the research analyses and selects an optimal modelling approach to identify suitable areas for the species, based on 19 bioclimatic factors. The analysis results include both current and future predictions for the period 2080–2100, based on the ACCESS-CM2 climate change scenario. The research process helps identify parameters affecting species distribution, providing crucial information for effective conservation measures in the context of climate change.

Occurrence and environmental data

Field surveys were carried out at Con Dao National Park, located in the Con Dao Islands of Ba Ria-Vung Tau Province, Vietnam. The occurrence of B. hainesii was individually measured for several parameters, including height, trunk diameter and crown diameter, to obtain a comprehensive dataset on the population structure of species in this region; the geographic coordinates and elevation data were gathered using a Garmin GPSMAP 64s receiver, with all measurements recorded according to the WGS 84 coordinate system. After examining the morphology and photographing the habitat, the specimens were numbered according to the individual trees. The specimen photos have been stored in the photo archive of the Institute of Tropical Ecology, Vietnam-Russia Tropical Center. Specimen identification was based on the documents by Tran et al. (2023).

Additionally, we used species distribution data from the GBIF biodiversity database. In total, data from 13 population distribution areas were collected from all countries where they are found worldwide, including Vietnam, Thailand, Indonesia, Malaysia, Singapore, Australia and Papua New Guinea (Fig. 1).

Figure 1.

Distribution location of B. hainesii.

Nineteen bioclimatic variables were collected from Worldclim (version 2.1) for modelling and are considered to influence the distribution of species in natural environments (Phillips et al. 2006, Fick and Hijmans 2017). Correlations between environmental variables greater than 0.8 were removed to avoid overfitting the model. Finally, the remaining environmental parameters were used in the model, including seven variables (bio1: annual mean temperature; bio2: mean diurnal range (mean of monthly (max temp - min temp)); bio8: mean temperature of the wettest quarter; bio9: mean temperature of the driest quarter; bio10: mean temperature of the warmest quarter; bio12: annual precipitation; bio15: precipitation seasonality, coefficient of variation).

Ecological Niche Model (ENM)

Predicted future distribution of suitable habitats of B. hainesii

Validation of a model

Reliability of the analysis results

Mapping suitable distribution zones

Based on the results of the ecological survey conducted in November 2013, we recorded seven individuals of B. hainesii on the Con Dao Islands. For the first time, a study on the biological and ecological characteristics of B. hainesii has been conducted, with the following results:

Description

Ecological notes

In Vietnam, this species primarily grows in mangrove forests on a small island (Hon Ba Island, part of the Con Dao Archipelago). The habitat of B. hainesii here is characterised by saline water, partially submerged by tides, with seasonal variations in seawater levels. The mangrove forests where B. hainesii thrives are noted for their soft, thick mud layer and frequent inundation. The trees often grow in shallow water areas where there is an exchange of water between the sea and the land. Analysis of 24 soil samples from areas where this species is distributed showed that the soil composition is predominantly sandy (> 89%), with minimal clay and silt (Fig. 2). The soil pH_KCl ranges from slightly acidic to neutral (4.62-6.77). Organic matter content is moderate (OM = 3.27%). The total nitrogen (N) content is moderate (0.12%), phosphorus (P) is low (0.04%) and potassium (K) ranges from moderate to high (0.38%) (Fig. 3).

Figure 2.

Soil properties (min/max, standard deviation) around the B. hainesii distribution site, Vietnam.

Figure 3.

NPK content (min/max, standard deviation) around the B. hainesii distribution site in the soil layer at a 10 - 30 cm depth, Vietnam.

Conservation status

Current and future appropriate bioclimate zone of B. hainesii

Based on the Maxent algorithm for B. hainesii, the model generated using Maxent software achieved an AUC of 0.862, which outperforms the model created with GEE (AUC: 0.66). The high AUC value (Fig. 4) from Maxent indicates that the bioclimatic variables used in the model effectively explain the distribution range of B. hainesii, highlighting the critical role of climatic factors in shaping and maintaining its habitat. In contrast, the model generated using GEE, which incorporated both presence and pseudo-absence data, exhibited a lower AUC, potentially due to inaccuracies or the random distribution of pseudo-absence data. This finding suggests that the use of pseudo-absence data should be approached with caution, particularly for species with limited presence data or those inhabiting specialised environments such as mangrove forests.

Figure 4.

AUC curve.

Consequently, the Maxent software model was selected for analysis and evaluation of both current and future scenarios. The predictions suggest that the most suitable habitat for B. hainesii is primarily located in Southeast Asia. This region includes countries such as Vietnam, Thailand, Indonesia, Malaysia, the Philippines, Singapore and Papua New Guinea. Additionally, parts of northern Australia are also predicted to offer suitable conditions for this species (Fig. 5). This information is valuable for conservation efforts, as it helps identify and prioritise regions where habitat protection and restoration activities are most needed to support the survival and growth of B. hainesii.

Figure 5.

Habitat Suitability Index of B. hainesii at present.

In the model used to assess the habitat suitability for B. hainesii, three specific climatic variables played a crucial role in determining its performance. These factors include the mean diurnal range (bio2), which represents the difference between the maximum and minimum temperatures within a day; the mean temperature of the driest quarter (bio9), which indicates the average temperature during the least rainy season of the year; and the annual mean temperature (bio1), which is the average temperature throughout the entire year. Together, these factors contributed to the model's effectiveness in predicting habitat suitability (Table 1).

Table 1.

Download as

CSV

XLSX

Percent contribution and permutation importance of the variables. (bio1: annual mean temperature, bio2: mean diurnal range (mean of monthly (max temp - min temp)), bio8: mean temperature of wettest quarter, bio9: mean temperature of driest quarter, bio10: mean temperature of warmest quarter, bio12: annual precipitation, bio15: precipitation seasonality, coefficient of variation).

Variables	Percent contribution	Permutation importance
bio2	76	62.5
bio9	10.3	5.6
bio1	10.1	0
bio8	2.8	30.8
bio15	0.7	0
bio12	0	1.1
bio10	0	0

The study's results indicate that the potentially suitable distribution area for B. hainesii under future scenarios (ACCESS scenario SSP2-4.5 for period: 2080-2100) is predicted to expand, with a significant increase in the total area of high habitat suitability index (HSI) across the entire region. The results of the Jackknife test on the importance of environmental variables show that the variable that achieves the highest gain when analysed independently is bio2, suggesting that this variable provides the most useful information when considered independently. The environmental variable that caused the greatest reduction in model gain when excluded was also bio2, this suggesting that it holds the most distinct information not found in other variables. These values represent averages over multiple models run (Figs 6, 7).

Figure 6.

Habitat Suitability Index of B. hainesii in the future (2080-2100) using the ACCESS scenario SSP2-4.5.

Figure 7.

Jackknife test of the model.

Mangroves are capable of adapting to climate change due to their ecological characteristics and crucial role in coastal ecosystems. These ecosystems are resilient to high salinity conditions, fluctuating water levels and tidal flows. Mangrove species have generally developed complex adaptive mechanisms in morphology, anatomy, physiology and molecular biology that enable them to thrive in stressful environments (Srikanth et al. 2016).

There has been no study on the adaptive capacity of B. hainesii under significant climate change impacts. However, recent local assessments suggest that B. hainesii might have expanded to Vietnam due to ocean currents transporting it to the Con Dao Islands. Another reason that the distribution of B. hainesii and mangroves, in general, might adapt well to climate change is their mobility. Mangroves can naturally expand and colonise new areas as conditions become more favourable. In western Jamaica, mangrove ecosystems were able to persist during the mid-Holocene as their sedimentation rates outpaced the rising sea levels (Hendry and Digerfeldt 1989). Suppose migration or inland expansion does not occur rapidly enough to counteract rising sea levels. In that case, mangroves may become increasingly diminished and potentially face extinction. The suitable climatic zones for B. hainesii are likely to expand significantly, which also shows that the regions such as Vietnam, Indonesia, Singapore and Malaysia, where high temperatures and precipitation, support their habitats. However, mangrove expansion is not a straightforward process. Despite potentially favourable climatic conditions, other factors such as land loss due to urbanisation, resource extraction and marine ecosystem changes could create barriers to this expansion (McLeod and Salm 2006).

Different mangrove species have distinct tolerances for the duration, frequency and depth of flooding (Semeniuk 1994). B. hainesii, as a tree species found in high tidal zones, has stable adaptations to coastal shape, soil and salinity and changes in these factors could alter the species' tolerance (Hoang 2013). Given the very small population of this species and the lack of research on its potential for relocation (replanting) to new areas, further studies are necessary. To evaluate the impact of rising sea levels on the B. hainesii habitat, it is important to consider factors affecting ecological balance, including coastal processes, local geomorphology and salinity of soil.

Climatic changes can exacerbate impacts on mangroves before sea-level rise, sediment and nutrient supply, as well as salinity regimes. In examining the effects of sea-level rise on river estuaries, Kennish (2002) highlighted the crucial influence of local environmental conditions. He stressed that factors such as tidal range and sediment supply play a fundamental role in shaping how mangrove ecosystems respond to rising sea levels. According to his analysis, mangrove communities situated in regions with high tidal ranges and abundant sediment, such as northern Australia, exhibit greater resilience to sea-level rise compared to those in areas with lower sediment availability and smaller tidal ranges, such as the Caribbean islands. This contrast is evident from studies like those of Woodroffe (1995), who observed the enhanced stability of mangrove systems in sediment-rich environments and Parkinson et al. (1994) who documented the vulnerability of mangroves in sediment-poor areas with limited tidal fluctuations. Understanding these local conditions is essential for predicting the potential impacts of sea-level rise on mangrove habitats and for developing targeted conservation strategies to protect these vital coastal ecosystems (Ward et al. 2016).

Although the global distribution of B. hainesii may increase in the future, mangroves often experience stress beyond their survival thresholds due to various factors (e.g. hydrological changes, artificial sedimentation, subsidence, climate change). We recommend that rising sea levels are likely to reduce the geographical distribution of this species in Con Dao NP due to the island's small size and limited sediment conditions, making it difficult to access external sediment sources. Compared to riverine mangroves, mangroves on small islands are less likely to withstand sea-level rise effectively, compared to those with external sediment sources (Pernetta 1993).

B. hainesii demonstrates significant potential to adapt to climate change, with projections indicating that its climatic range could expand due to increases in temperature and precipitation between 2080 and 2100. This suggests that the species may be resilient to some aspects of climate change, particularly those affecting temperature and rainfall. As such, B. hainesii might be able to colonise new areas that become more suitable under future climatic conditions, which could potentially enhance its distribution and ecological role. These findings highlight the species' ability to adjust to changing environmental conditions, which is critical for its long-term survival in the face of global climate challenges. However, the results of this study also suggest that climatic factors alone are not responsible for the global rarity of B. hainesii. Despite the potential for adaptation, the species' low regeneration capacity remains a significant limiting factor. This limitation could be linked to physiological constraints, such as poor seed viability, limited pollination or challenges in seedling establishment, which could hinder its ability to recover and expand. These issues are more critical than climate-related factors in determining the species' population dynamics. Therefore, addressing these physiological barriers is essential for improving the regeneration and overall survival of B. hainesii. In light of these findings, conservation efforts for B. hainesii should prioritise habitat protection, particularly in regions where the species already exists. This can provide a stable environment for the remaining populations, ensuring their survival in the short term. Simultaneously, research into overcoming the physiological barriers to regeneration should be a central focus. Strategies such as enhancing seedling establishment, improving pollination rates or exploring genetic approaches could help increase the species' regeneration capacity. By addressing both habitat preservation and physiological limitations, we can ensure the long-term sustainability of B. hainesii and the critical coastal ecosystems it supports.

This research was funded by the Basis project of Joint Vietnam-Russia Tropical Science and Technology Research Center as part of a project dedicated to conserving genetic resources for Bruguiera hainesii in Vietnam, from 2023 to 2025 (Vietnamese: Nghiên cứu bảo tồn và phát triển nguồn gen loài Vẹt hainesii (Bruguiera hainesii C.G. Rogers) cực kỳ nguy cấp ở Vườn quốc gia Côn Đảo, Bà Rịa - Vũng Tàu). We are grateful to the directorates of the Con Dao NP for their support of our fieldwork and for issuing relevant permits. We thank Le Hong Son (Department of Science and International Cooperation, Con Dao NP), Dau Nhu Kien, My Duy Hai and Thai Duc Tho (Hon Ba Forest Ranger, Con Dao NP) for their assistance in the field.