2.1 Assessing the Begonia diversity occurring in Thailand

In support of the aims of this study, it is important to note that the taxonomy of Thailand's Begonia species is well-established (Peng et al., 2017; Phutthai & Hughes, 2016; Phutthai & Hughes, 2017a; 2017b; Phutthai et al., 2012, 2014, 2019, 2021; Phutthai & Sridith, 2010; Radbouchoom et al., 2023). Thus, the study does not need to address taxonomic ambiguities that could otherwise limit the reliability of results regarding the spatial distribution of species diversity and the assessment of extinction threats.

In total, 2484 occurrence records of Begonia species in Thailand were obtained via the integration of several data collection strategies. First, a series of field surveys were conducted across Thailand with the aim of expanding records of these plants' distribution and ecological preferences, with a particular focus on occurrences outside protected areas. These field surveys, conducted between 2020 and 2021, covered multiple provinces, including Chiang Mai, Chumporn, Kanchanaburi, Lampang, Ratchaburi, Ranong, Tak, Trang and Krabi. Second, occurrence data were assembled by studying voucher specimens deposited in two key herbaria in Thailand: the Forest Herbarium in Bangkok (BKF) and the Queen Sirikit Botanic Garden Herbarium in Chiang Mai (QBG). In addition, information was obtained by studying voucher specimens outside Thailand, specifically AAU, ABD, BM, BK, E, K, HAST, LUN, L, P, PE, QUB, and SING. The abbreviations used follow the Herbarium Codes as given in the Index Herbariorum (Thiers, 2024). Specimen searches utilised databases accessible via institutional websites, the Begonia Resource Center (https://padme.rbge.org.uk/Begonia/home; Hughes et al., 2015), and the Global Biodiversity Information Facility (GBIF; https://www.gbif.org). Finally, publications on the flora of Thailand were studied to extract further distribution data (Phutthai et al., 2012, 2014, 2019, 2021; Phutthai & Sridith, 2010). The species taxonomy followed recent treatments (Moonlight et al., 2018). The taxonomic accuracy of all accessions was carefully verified by the two taxon experts on the author team (TP and SR) accessions to ensure the data's integrity. The final data set was cleaned by removing occurrences based on accessions with ambiguous species identification, incorrect coordinates or erroneous placement within Thailand. Wherever possible, spatial coordinates were assigned to accessions that had not been previously geo-referenced. Voucher collections lacking sufficient information about the collection site were excluded. All maps used in this study were handled using QGIS.org (2021).

2.2 Evaluation of the spatial distribution and identification of hotspots

The assembled occurrence records for all 60 species of Begonia known to occur in Thailand were used to reconstruct the spatial distribution of this diversity across the country. Specific attention was given to the seven floristic regions: Northern, North-Eastern, Eastern, Central, South-Eastern and Peninsular (Phutthai et al., 2009; Trisurat et al., 2011). The software Biodiverse 4.0 was used to estimate three widely applied biodiversity indices: Species Richness (SR), Weighted Endemism (WE), and Corrected Weighted Endemism (CWE) (Laffan et al., 2010). These values were estimated utilising a 0.25-degree grid cell map of Thailand. By considering both the spatial distribution of species richness and weighted endemism across all grids, hotspots of Begonia diversity in Thailand were localised.

2.3 Assessment of the extinction threats experienced by Thailand's Begonia species based on their restricted geographic range (=criterion B)

The vulnerability of each Begonia species to extinction threats was evaluated by applying the IUCN Criteria B1 & B2, implemented through the ‘ConR’ package in R version 4.3.0 (Dauby, 2017; R Core Team, 2023). Given that for most species, only distributional data is available, the application of the restricted area criterion (see IUCN, 2017–2020) was deemed most appropriate. Future studies will hopefully also provide estimates of population size for each species, which would allow for the consideration of Criteria C and D as well. In this study, the available data was restricted to recorded occurrences, allowing the application of Criterion B1, which focuses on the extent of occurrence (EOO), and Criterion B2, which explores the area of occupancy (AOO) based on currently occupied suitable habitats. The analysis was conducted using a 0.25-degree grid cell resolution. Species with fewer than three occurrence records were classified as Data Deficient (DD). For species with more than three occurrences, their threat status was determined by comparing the estimated EOO value and AOO values: Critically Endangered (CR) taxa had an EOO of less than 100 km² and AOO of less than 10 km², Endangered (EN) taxa had an EOO of less than 5000 km² and AOO of less than 500 km², Vulnerable (VU) taxa had an EOO of less than 20,000 km² and AOO of less than 2000 km², and Near Threatened (NT) or Least Concern (LC) taxa had an EOO of more than 20,000 km² and AOO of more than 2000 km² (see IUCN, 2017–2020). In cases where EOO values were lower than AOO values, both values were treated equally. The analyses generated maps visualised in 0.25-degree grid cells, representing the number of recorded species and the proportion of threatened species within each cell.

2.4 Inferring conservation capacities

Based on the location within or outside protected areas, species occurrences were categorised as either in situ protected or not. This categorisation was achieved by overlaying species distribution maps with the map of protected areas in Thailand. The protected areas included in the analysis were those fitting the IUCN Category 1a (Strict Nature Reserve) and IUCN Category II (National Park), as provided by the Department of Natural Resources and Wildlife of Thailand. This assembled map specified the distribution of 268 protected areas suitable for in situ conservation of Begonia, namely national parks, wildlife sanctuaries, and nonhunting areas. Specific attention was given to species whose entire range occurs outside protected areas, as these species are likely more vulnerable and in need of conservation efforts. The assessment also aimed to identify gaps in conservation coverage by identifying hotspots that fall outside of protected areas. To evaluate ex situ conservation capacities, we identified the number of Begonia species currently cultivated in botanical gardens, with special attention paid to the Queen Sirikit Botanic Garden. Furthermore, we searched for evidence suggesting the existence of specific ex situ programmes, such as targeted reproduction of Begonia species in botanical gardens followed by replanting in the wild or the establishment of resources aiming to preserve the genetic diversity of these plants, such as seed banks.

3.1 Spatial distribution of the 60 species of Begonia in Thailand

Although Begonia species are reported from various parts of Thailand, the majority of occurrence records are concentrated in three floristic regions (Figure 1; Supporting Information: File S1). The Northern floristic region is particularly notable, contributing 46% of the total occurrences in Thailand. The Peninsula floristic region follows, accounting for 26%, while the Southwestern floristic region contributes 12%. This uneven distribution is reflected in the diversity indices applied, which reveal disparities in species richness, weighted endemism and corrected weighted endemism (Figures 2 and 3). Consequently, several hotspots were detected: five in the Northern floristic region, two in the Peninsula floristic region, and two in the Southwestern floristic region (Figure 2d). Of these nine hotspots, only one is located outside a protected area, specifically in Chiang Rai province (Figures 2d and 3d). This gap in the protection of diversity hotspots was found to be located in the most northern parts of the Northern floristic region.

3.2 Inference of the vulnerability to extinction threats considering the restricted geographic range

The vast majority of the 60 species are known from only one or a few locations, indicating their vulnerability due to restricted geographic range. The most notable exceptions were widespread species such as B. integrifolia Dalzell (71 locations), B. palmata D.Don (30 locations), and B. sinuata Wall. ex Meisn. (30 locations). Among the 26 taxa endemic to Thailand, seven taxa were recorded from a single location, five taxa from two locations, two taxa from three locations, and three taxa from four locations. The remaining endemic species were distributed across five to six locations. The extinction threat assessment covered all 60 species, focusing solely on their occurrences within Thailand (Figure 4; Supporting Information: File S2). Using the ‘ConR’ package and applying the IUCN Category B2a, 31 species were classified under this criterion, while 28 species were classified under B1a+B2a, and one species under B1a. According to the AOO criterion, 13 species were evaluated as Least Concern or Nearly Threatened. The remaining species were inferred to be threatened, with 10 species as Critically Endangered, 24 species as Endangered, and 13 species as Vulnerable. The EOO criterion could not be applied to 10 species. For the remaining species, 27 were evaluated as non-threatened, while the rest were categorised as threatened, with three species as Critically Endangered, 10 species as Endangered, and 10 species as Vulnerable (Figure 4).

3.3 Capacity to protect Begonia diversity in Thailand

Approximately 63% of all recorded Begonia occurrences were located in protected areas, although these areas only cover 23% of Thailand's land area. 53 species (88.3%) had at least some occurrence within protected areas, whereas seven species (11.7%) were exclusively found outside these areas (Figure 4; Supporting Information: File S2). The effectiveness of in situ conservation provided by protected areas varied among species. Some widespread species had a relatively small proportion of their known locations within protected areas, whereas certain rare species were entirely confined to these protected zones. Despite the high ornamental value of Begonia species, only eight species were recorded as being cultivated in botanical gardens, with most located at the Queen Sirikit Botanic Garden (QBG) (Figure 4d). The species found at QBG showcased promising ornamental properties, making them suitable for exhibition purposes. Additionally, QBG also harboured several rare and endangered species, which were observed to thrive under controlled conditions (BGO, 2020). In total, 11 species were known from only a single location, of which seven species were found to be endemic to Thailand (B. exposita Phutthai & M. Hughes, B. festiva Craib, B. khaophanomensis Phutthai & M. Hughes, B. pengchingii Phutthai & M. Hughes, B. sirindhorniana Phutthai, Thananth., Srisom & Suddee, B. smithiae E.T. Geddes, B. subviridis Craib). The remaining four species were also found in neighbouring countries to the south (B. barbellata Ridl, B. kingiana Irmsch.) or northern neighbour nations (B. hymenophylla Gagnep., B. surculigera Kurz). All eleven of these species were evaluated as Critically Endangered or Endangered based on the AOO category. Four of these species—B. exposita, B. pengchingii, B. sirindhorniana, and B. surculigera Kurz—were among the seven species with no known locations within protected areas. The other three species—B. bella Phutthai, B. fulgurata C.I Peng, C.W. Lin & Phutthai, and B. silletensis (A. DC.) C.B. Clarke—were known from two or three locations, with the first two being endemic to Thailand. Besides the seven species known from a single location occurring in protected areas, three species—B. incondita Craib, B. sandalifolia C.B. Clarke and B. tenasserimensis Phutthai & M. Hughes—known from two locations, both in protected areas. Conversely, some species with more than 10 known locations had only a small proportion of these within protected areas. For example, B. integrifolia Dalzell had only 4.2% of its known locations in protected areas. Similarly, some endemic species with more than five known locations had 50% or fewer occurrences within protected areas, such as B. saxifragifolia Craib, with only 18.2% of its locations in protected areas (Supporting Information: File S2).

4.1 Taxonomy challenge

Determining the number of species that need protection against extinction threats is significantly challenged by the lack of adequate taxonomic research, which is key for effective conservation (Sandall et al., 2023; Thomson et al., 2018). Unfortunately, taxonomic expertise is often poorly established and maintained for many plant lineages (Löbl et al., 2023). This is regrettable because advancements in taxonomic research are essential to enable progress towards the zero extinction target. These advancements are required not only to discover and describe the many species that remain unidentified but also to correct errors in existing taxonomic classifications as new biological insights are gained. Such research may lead to the redefinition of previously described species, either by recognising over-looked species or by synonymising unnatural taxonomic units. To support access to taxonomic expertise, new digital infrastructures are being developed to facilitate rapid access to and exchange of information on all plant species (Sandall et al., 2023). Notable examples include the Plants of the World Online (POWO, 2024) by the Royal Botanic Gardens at Kew (https://powo.science.kew.org) and World Flora Online (https://www.worldfloraonline.org). This study benefited from ongoing efforts to clarify the taxonomy of Begonia both in Thailand and globally. These efforts have provided a robust framework for species recognition and access to valuable metadata of the taxa through resources like the Begonia Resource Center (https://padme.rbge.org.uk/begonia/) and the World Flora Online (WFO) Begonia treatment (http://www.worldfloraonline.org/taxon/wfo-4000004308). Unfortunately, similar resources are lacking for many other plant lineages.

4.2 Occurrence data quality

The research approach in this study relies heavily on the quality of the occurrence data obtained through field surveys and historical records. The study took advantage of substantial efforts that were taken to document Begonia species across Thailand. However, there were some limitations due to the lack of detailed geographical information in certain historical records. The resulting inaccuracies in occurrence data need to be addressed through targeted field surveys aimed at improving the precision of geographic locations. Furthermore, some regions of Thailand require further field surveys to overcome the difficulties in data collection, particularly for species with specific ecological preferences, such as those that grow on poorly accessible rock outcrops. Another problem is the absence of continuous monitoring to confirm the ongoing existence of these species at previously recorded locations, which is especially important for older records. Finally, the utility of occurrence records could be greatly enhanced by linking informative metadata, such as details about the ecosystems where certain species are found and the number of mature and/or young individuals observed. For some occurrences, this highly valuable data is available, but it is not the norm. We suggest that future research efforts should establish protocols to consistently gather and incorporate this metadata to improve the overall quality and applicability of the occurrence data.

4.3 Assessing vulnerability to extinction

The study set out to tackle the challenge of utilising occurrence data to assess the vulnerability of Thailand's Begonia species to extinction threats. The assessment employed corresponds to Criterion B of the IUCN threat status assessment protocol. However, the full application of the IUCN protocol is often constrained by the limited evidence available for many species. Despite this limitation, the results of these simplified assessments were largely consistent with previously published evaluations. Our data reaffirmed the trend that the majority of Begonia species in Thailand are vulnerable to extinction threats, although not all species fall into this category—B. hatacoa Buch.-Ham. is arguably the most notable exception. While the Begonia treatment presented in World Flora Online lists this species as Least Concern, our assessments using the AOO (Endangered) and EOO (Vulnerable) criteria categorised it as threatened in Thailand. However, this species is not endemic to Thailand, which may explain the discrepancy between our assessment and the global status. This difference highlights the influence of geographic scope on threat assessments. Given the lack of demographic data for many species, further refinement of these vulnerability assessments remains challenging.

4.4 Assessing the effectiveness of current conservation measurements

Thailand's protected areas provide substantial resources for in situ conservation of Begonia species. Only seven species had no occurrences within protected areas. In contrast, 21 species had at least 50% of their occurrences in protected areas. Focusing specifically on very rare species—those with only one to three known locations in Thailand—the study found that seven species were without a recorded location in protected areas. This includes four of the 11 species known from a single location in Thailand, two of the 11 species with two locations in Thailand, and one of four species with three locations in Thailand. However, it is important to note that seven species with a single known location are indeed found within protected areas, as are three out of the 11 species with two known locations. Among the 24 species endemic to Thailand, five lacked protection from the country's protected areas, whereas four species had all their known locations within these areas.

Regarding biodiversity hotspots, the study identified only one out of nine hotspots located outside protected areas, further emphasising the critical role these areas play in in situ conservation. Thailand arguably already has substantial resources to protect the Begonia diversity via in situ interventions. In contrast, ex situ capacities are still underdeveloped despite the existence of several well-managed botanical gardens. In spite of their high ornamental value and the possibility of reproducing many species vegetatively, no evidence was found of any targeted cultivation program focusing on reproducing individuals for conservation purposes. Furthermore, there is a notable absence of initiatives to protect the genetic diversity of these species through seed banks or tissue banks.

4.5 From priority species lists towards conservation actions

By employing relatively straightforward criteria—such as the range criterion categories AOO and EOO, the number of known locations, endemism, and percentage of locations within protected areas—the study established a priority species list for conservation interventions (Supporting Information: Files S2 and S3). Four species were ranked on the top due to particularly high conservation concerns: (1) B. bella Phutthai, B. fulgurata, B. pengchingii, and B. sirindhorniana. These species are all endemic to Thailand, known from only one or two locations and fall under the Critically Endangered or Endangered categories based on the range criteria. Moreover, all known occurrences of these species are outside protected areas. Other endemic taxa appear to be under less immediate threat, primarily due to some level of conservation intervention through their presence within protected areas. For instance, B. saxifragifolia is endemic to Thailand and is classified as Least Concern because it is known from a total of 11 locations. However, the fact that only two of these locations are within protected areas raises concerns about the ability to protect this species. Two species with all occurrences outside protected areas were not endemic to Thailand. For such species—non-endemic with all occurrences outside protected areas—conservation actions might require collaboration with neighbouring countries. Such collaborations are particularly pertinent for species occurring in Thailand's Peninsula floristic region and the Malay Peninsula, involving 11 shared species. Likewise, collaboration is necessary for species shared with Myanmar (27 species) and Lao (12 species).

4.6 Steps forward

Arguably, the main shortcoming of the current inventories of Begonia species in Thailand is the lack of demographic data. Accurate estimates of the number of individuals and regeneration capacity of populations via sexual reproduction are especially important for evaluating the extinction threats for species highlighted as a priority. Targeted field surveys have to be carried out to gather this data, which will improve threat assessments and guide more effective conservation interventions. Another significant gap is the absence of information on the genetic diversity of Begonia species in Thailand. An increasing number of studies have explored the genetic diversity of Begonia species (Chan et al., 2018; Li et al., 2018; Nakamura et al., 2014; Tseng et al., 2019; Twyford et al., 2014, 2015; Xiao et al., 2023), but none have focused on species occurring in Thailand. These studies support the hypothesis that many species of Begonia have limited dispersal capacities, resulting in population structures shaped by limited gene flow and random genetic drift. These genetic insights need to be taken into account when designing conservation interventions that protect the genetic integrity and genomic health of the conserved species (Allendorf et al., 2010; Theissinger et al., 2023).

To address these conservation gaps, a multi-component strategy is required. First, the existing in situ conservation interventions must be expanded. The required capacity building may be achieved by expanding the existing protected areas to cover currently unprotected biodiversity hotspots in Northern Thailand and implementing the concept of micro-reserves (Fos et al., 2017; Laguna et al., 2013). Given the narrow ranges and highly specialised ecosystem preferences of priority species, micro-reserves focused on unique habitats such as karst outcrops—home to species such as B. exposita, B. pengchingii, and B. surculigera—are likely to be effective. The focus on the habitat of a single species resolves the conflict between providing for threatened species and the provision of natural resources, for example, land, to the local communities. However, translocation from highly threatened locations to sites in protected areas may have to be considered (Diallo et al., 2023). In addition to in situ measures, there is an urgent need to establish resources for ex situ conservation. This should include cultivation programmes in botanical gardens and the creation of germplasm banks for long-term protection, for example, seed banks, tissue culture banks and DNA banks (Elliott & Greuk, 2022). Successful reproduction in ex situ programmes may also facilitate the translocation of species into restored habitats, such as karst landscapes rehabilitated after mining activities end. By addressing these gaps, Thailand can strengthen its conservation capacities and move closer to achieving the goal of zero species extinction for its unique Begonia diversity.