Vegetation Structure and Tree Species Diversity inside and outside a Newly Established Zalon Taung National Park in Northwest Myanmar

,


Introduction
Numerous studies focus on tropical forests since they include some of the most diverse vegetation [1], conserve carbon stocks and timber [2], and provide food, feed, shelter, energy, medicine, and revenue for human progress [3]. Globally, the highest percentage of forests (45%) is in tropical regions [4]. Tropical forests are the biodiversity hotspots of conservation priority [5], as people in poor tropical nations rely on forest resources for daily needs, compared to biodiversity conservation activities [6]. Such the forest resource assessment by the FAO in 2020, the forests cover in Myanmar is 42.19% of the entire country [12]. Currently, 6.43% of Myanmar's total area has been established as protected areas, although the nation has targeted to increase that to 10% by the 2015 Paris Climate Change Conference [13].
Tropical forests are being increasingly afected by population explosion, agriculture expansion, mining, logging, and road building [14]. Although illegal logging activities target the most valuable timber species with promising revenues, logging intensities have been exceptionally high in Southeast Asia, leading to loss and damage to species and ecosystems [15]. Furthermore, collecting frewood from local communities is another distinct driver of the degradation of tropical forests in Southeast Asia [16]. While emphasizing the sustainability, habitat protection, and ecosystem preservation of tropical forests, we need a comprehensive understanding of the diversity and composition of tree species in these forests [17,18]. Understanding tropical forests' diversity and ecological features is vital to sustaining their functions in regulating species diversity, food webs, water and air fltration, microclimate, and soil fertility [19,20]. Considerable research on tropical vegetation focuses primarily on how variations in topography, edaphic conditions, human activities, management types, and land use/land cover impact the diversity, composition, and structure of tree species in the forest reserves [21][22][23][24].
Tropical forests in Myanmar have been studied for their management status and plant community compositions [25,26], soil and environmental characteristics [27,28], tree species diversity distribution along a precipitation gradient [29], and topographic and edaphic variations [30]. Tese studies provided details on Myanmar's tropical forest structure and foristic richness. Most of these studies were conducted in well-maintained protected areas and reserved forests, ideal for conserving biodiversity. However, the unprotected forests of Myanmar have yet to be studied. A substantial portion of the biodiversity in Myanmar and the tropics remains outside of protected areas [31,32]. Recent years have seen an increase in the importance of biodiversity conservation outside protected areas in the global conservation debate [33,34]. Most environmentalists focus on efective land use planning to protect the biodiversity of protected areas outside to achieve conservation goals and livelihood security [35]. A specifc study claimed that while certain species were more prevalent outside the protected area than inside, conservation eforts inside protected areas alone were inadequate to guarantee the long-term survival of endangered species [36]. Te onservation of Southeast Asian biodiversity is geared towards the protected areas [37]; therefore, recording forest tree species outside the protected areas is rare. Supporting the maintenance of unprotected tropical forests near the protected areas might result in a situation where biodiversity conservation is achieved while also boosting the local economy [32]. To combine biodiversity protection with local livelihood security, we must understand human dependence on forest resources and its efects on forest structure and diversity. Disturbances by humans have decreased species diversity and structural characteristics such as stand density, diameter class distributions, and basal area of tropical forests [38,39].
Tis study was conducted inside and outside the Zalon Taung National Park (ZNP) in Banmauk township, Northwest Myanmar. According to the FAO 2020 forest resource assessment, 42.54% of Banmauk township is forested [40]. However, as the population rise and human activities such as illicit logging, gold mining, fuel wood collection, shifting cultivation, and plantation establishment endanger forest resources [41], the extent of forest cover in the Banmauk township is declining at a rate of 0.6% per year between 2010 and 2017 [42]. Te Ministry of Natural Resources and Environmental Conservation of Myanmar issued the ZNP as a protected area in 2022 [43]. However, with the expansion of human populations, their settlements and farmlands contributed to the destruction of natural forests, particularly in the periphery of the ZNP [44]. Te inhabitants use forest resources to generate income, mainly to collect wood, medicinal plants, and frewood [41]. Additionally, the ZNP has a history of selective logging, in which the government extracted teak and other economically signifcant hardwood species for foreign revenue [45]. To implement efective conservation measures under the ZNP designation as a protected area, a quantitative investigation of the foristic composition, richness, and stand structure along this human-dependent natural forest is urgently required.
Tis study aimed to assess tree species diversity, composition, and stand structure inside and outside a newly established Zalon Taung National Park (ZNP) protected area in Myanmar. Our specifc objectives are to (1) identify the tree species diversity and stand structural features; (2) assess the stand density, abundance, and dominance of total tree species and economically important tree species; and (3) evaluate the changes in the tree species composition, diversity, evenness, and stand structure in response to environmental variables. Te results will help to evaluate the vegetation structure, diversity, and composition in and out of the ZNP for long-term conservation planning to balance biodiversity protection and resource use.

Study Area.
We collected vegetation data during April-May 2022 in the Zalon Taung National Park (ZNP) and its adjacent Banmauk unclassifed forest (BUCF), located between 95°44′E and 95°56′E longitude and 24°28′N and 24°42′N latitude in Northwest Myanmar ( Figure 1). As a limitation of this study, we cannot assess the vegetation in the upper part of the ZNP because of ongoing armed conficts between the national military and local armed groups that have rendered it insecure or inaccessible. Te ZNP was designated as a protected area in January 2022 to preserve the cultural value of Zalon Taung Pagoda, the native fora, wildlife, and water resources after it was gazetted in 1994 as Nantkaunghu Reserved Forest, covering 21,616 ha [43]. ZNP is a signifcant protected habitat for hoolock gibbons, pangolins, dhole dogs, Asiatic black bears, and clouded leopards [44]. National parks are included in the category of the protected area system, in which the forest law strictly prohibits harvesting forest resources [46]. Unclassifed forests are included in the category of vacant land or land at government's disposal [47]. Unless the forest law expressly forbids a particular behaviour, such as cutting "reserved tree species" (e.g., teak), villagers can harvest wood and nontimber forest products for subsistence in the unclassifed forest [46]. After Myanmar Timber Enterprise deliberately felled 2,439.54 tons of teak (Tectona grandis) and 5,207.852 tons of kanyin trees (Dipterocarpus alatus) between 2005 and 2010, the government stopped the exploitation in the ZNP [43]. In BUCF, the selective logging of teak and other hardwood trees has continued. Te extraction of nontimber forest products, especially medicinal plants, is noticeable in the ZNP.
Natural vegetation is a diverse range of tropical moist upper mixed deciduous forests, with most species defning the deciduous forest type. Physiographic conditions include steep hills, valleys, fat plains, and fat-topped ridges, with an elevation range of 100-1,600 meters above the sea level. Te typical soil types are Rhodic Ferrasols (red-brown forest soil) and Lithosols (primitive crushed stone). Te region has a tropical monsoon climate with an average of 1,336 mm annual rainfall, with the most rainfall between June and September [48]. Te mean monthly temperature ranges from 6.7°C to 41.9°C; the hottest month is April, and the coldest month is January. Te relative humidity is high throughout the year, averaging 86% [48].
Tere are around 40 villages on the plains surrounding BUCF and the ZNP. Most people who live near and within the territory of BUCF and the ZNP are farmers. Data from the 2014 national census show that the population density in Banmauk township was 33 persons/km 2 , with 83% of the population living in villages and 17% in town areas [49]. Te Shan ethnic group includes 69.51% of the population, followed by the Kadu ethnic group (15.11%) and the Kanan ethnic group (12.55%) [49]. Te local villagers cultivate rice and groundnut, collect frewood, timber, and nontimber forest products, especially medicinal plants, from the forest and work as gold mining labours. Te residents rely on forests for extraction and timber harvesting even though the forests are under the management of the Forest Department of Myanmar.  International Journal of Forestry Research 3 500 m apart ( Figure 1). Te sample locations had various human disturbances, including legal and illicit tree cutting and residents harvesting frewood and medicinal plants. Every tree ≥10 cm diameter at breast height (dbh) was identifed, numbered, and measured for height and diameter using a clinometer and diameter tapes. Te survey team, including local Forest Department staf, local guides, and the botanists of the local nongovernmental organization, assisted in identifying tree species. Te local names of the species were converted into scientifc names following Kress et al. [50]. Seven environmental parameters were determined for each plot, including the elevation, slope, aspect, land class, logging level, and distance to the nearest village and road. We used a GPS (Garmin 62 s) to determine each plot's latitude, longitude, and altitude. Te slope and aspect were derived from the 30 m resolution digital elevation model (DEM) using the coordinates of the sample plots. In each sampling plot, we record the number of cutting and stumps as indicators of logging activities ( Figure 2). We recorded the land class of each sampling plot, such as old-growth forest land (i.e., minimally disturbed land), secondary forest land (i.e., the land disturbed by selective logging and harvesting of forest products), and jungle forest land (i.e., the land heavily disturbed by road and farmland encroachment) ( Figure 3). Additionally, understory vegetation, including bamboo, shrubs, and herbs, was recorded in general, although it was not considered in the woody vegetation analyses.

Data Analysis.
To determine the stand structural characteristics, we calculated the mean dbh (cm), the maximum tree height (m), the total basal area (m 2 ), and tree density in each plot. We used the Shannon-Wiener index [51], Simpson's index [52], and Pielou's evenness [53] to calculate the woody species diversity in each sample plot.
where s is the number of species in the community, and p i is the percentage of individuals found in the i th species. Te Shannon index may range from 1.5 to 3.5, seldom going over 4.5, where high levels suggest a large diversity.
where n is the number of individuals of a single species, and N is the number of individuals in the total population. Te Simpson index has a potential range of 0-1, with high values indicating little diversity and high stand dominance.
S is the number of species, and H′ is the Shannon-Wiener Index. With potential values ranging from 0 to 1.0, strong values indicate good evenness.
Fisher's α diversity [54] was quantifed in each plot because it reduces the sample size and works well on tropical forest vegetation [55]. We used the biodiversity R package [56] in the R software version 4.2.1 [57] to examine the diversity of the species. Te density-diameter distribution of the tree species using dbh classes set at 10 cm intervals characterized the population structures of the two sites (BUCF and the ZNP). We developed rarefaction curves using the iNEXT package [58] in R software to evaluate the direct comparison of species richness between the two sites. Ten, we developed a species rankabundance curve to evaluate the species richness of the most abundant tree species [59]. Additionally, we calculate each species' Importance Value Index (IVI) by summing up each species' relative density, frequency, and dominance [60] (Table 1).
We used the Shapiro-Wilk test to determine the normality of the data before comparing the stand structural features of the economically signifcant and higher IVI species at BUCF and the ZNP [61]. Since the dataset deviated from the normal distribution, we used the Wilcoxon rank sum test to compare dbh, basal area, and height between the sites.
Nonmetric multidimensional scaling (NMDS) ordination in the R vegan package [62] examined tree species composition in 34 plots of three land classes in BUCF and the ZNP. Te sample plots were frst arranged on the graph based on Bray-Curtis distance measure. Ten, using permutation testing, signifcant species and environmental factors were ftted to the ordination graphs. Te Wilcoxon rank sum test compared environmental variables between the two sites. We used the Spearman rank correlation coefcients to quantitatively examine all sampling plots' stand structural characteristics and diversity along with environmental conditions (elevation, slope, distance to settlement, and roads).

Stand Structure and Species Diversity.
Te characteristics of 34 sampling plots were described ( Table 2). Te ZNP had a higher tree density (201 individuals ha −1 ) than BUCF, with a tree density of 109 individuals ha −1 . Te ZNP had a basal area of 20.6 m 2 ·ha −1 whereas BUCF had 15 m 2 ·ha −1 . In the ZNP, Simpson, Shannon, and Fisher diversity indices and Pielou's evenness were higher than in BUCF (Table 3). Te secondary forests in the ZNP had the highest tree density (338 individuals ha −1 ) with the highest basal area (22.4 m 2 ·ha −1 ) ( Figure 4). Te old-growth forest in BUCF had the most signifcant number of trees greater than 30 cm·dbh (128 individuals ha −1 ), followed by the secondary forest in the ZNP (106 individuals ha −1 ) ( Figure 4). Te species richness of all sampling plots and plot number was investigated through rarefaction curves ( Figure 5). Te rarefaction curves for BUCF and the ZNP demonstrated that for the same number of individuals, the species diversity between them was comparable ( Figure 6).
Te population patterns of BUCF and the ZNP were analyzed using the density-diameter distribution ( Figure 7). Most of the trees were between 10 and 15 cm·dbh. Only a handful of the trees in each site were larger than 50 cm·dbh. In contrast to BUCF, the ZNP plots showed more trees larger than 60 cm·dbh. Te tree population structures grew with increasing densities of lower diameter classes and lower densities of higher diameter classes ( Figure 7).  Table 4 lists the families that comprise more than 5% of the composition. In BUCF, the Verbenaceae family dominated but in the ZNP, it was the Euphorbiaceae family. Te IVI values indicate the ecological relevance of the tree species in community structure (Table 5). With an IVI of 26.91%, Protium serratum (Burseraceae) was the most prominent species in BUCF, followed by Dipterocarpus alatus (Dipterocarpaceae), Tectonagrandis (Verbenaceae), Dilleniapentagyna (Dilleniaceae), and Schleichera oleosera (Sapindaceae) with the IVI of 19.83, 19.41, 16.29, and 16.25%, respectively (Table 5). Te most notable species in the ZNP was Dipterocarpus alatus (Dipterocarpaceae), with an IVI of 18.39%, followed by Xerospermum noronhianum (Sapindaceae) (15.77%), Protium serratum (Burseraceae) (13.44%), Baccaurea sapida (Euphorbiaceae) (11.60%), and Dracontomelon dao (Anacardiaceae) (11.16%) ( Table 5). Tese species are considered ecologically signifcant in a given habitat since the IVI value is more than 10 [63]. Te IVI values of all species are provided in the Appendix (Supplementary Material (available here)) Te species rank-abundance curve revealed an abundance of vegetation with variations in BUCF and the ZNP ( Figure 8). A steep gradient of the BUCF stand showed low evenness as high-ranking species have higher abundances than low-ranking species. On the other hand, the modest gradient of the ZNP suggested good evenness since the abundances of various species are more consistent than in the BUCF (Figure 8).  signifcantly difered between the two sites (W �104.5, p < 0.05). Te diameter and basal area of Terminalia crenulata signifcantly difered between the two sites (W � 553.5, p < 0.05). Te height of Protium serratum was signifcantly diferent between the two sites (W � 2510.5, p < 0.01) ( Figure 9).

Diferences in Species Composition and Environmental
Variables. Te fnal stress value for the two-dimensional solution of the NMDS ordination analysis after 999 permutations was 16.61%, which is within an acceptable range of 20% for the statistical reliability of the NMDS ordination [64]. NMDS analysis revealed diferences in the composition of tree species in BUCF and the ZNP with three land classes. NMDS analysis showed that secondary forest plots were signifcantly composed of economically signifcant deciduous tree species such as Tectona grandis, Dipterocarpus alatus, and Shorea siamensis ( Figure 10 and Table 7). Te ZNP secondary forest plots favoured the tropical moist forest habitats such as those occupied by Dracontomelon dao, Malus spp, Syzygium kurzii, Baccaurea sapida, and Note. ZNP � the Zalon Taung National Park, BUCF � Banmauk unclassifed forest, + � counts of cuttings and stumps <3, ++ � counts of cuttings and stumps between 3 and 10, and +++ � counts of cuttings and stumps >10.  (Figure 10 and Table 7). Te elevation, logging level, and access to roads and villages appeared to be the most infuential factors afecting tree species composition in BUCF and the ZNP (Figure 11 and Table 7). In particular, the level of logging and distance to the road correlated with the secondary forest of BUCF (Figure 11 and Table 8). A higher elevation and a greater distance from the villages may favour the substantial composition of tree species, such as Dracontomelon dao, Malus spp, Morus alba, and Swintonia foribunda in the secondary forest of the ZNP according to the NMDS and the Wilcoxon rank sum test ( Figure 11 and Table 9). Te Wilcoxon rank sum test revealed that elevation and distance to the village signifcantly difered, whereas slope, aspect, land classes, logging level, and distance to the roads were not signifcantly different between BUCF and the ZNP (Table 9).

Relationships between Environmental Factors, Forest
Structure, and Biodiversity Indices. Spearman correlation analyses showed that the sample plots' basal area, maximum height, diversity, and evenness were afected by environmental factors such as elevation and distance to the village and road (Table 10). In all the sample plots, elevation had a substantial impact on species diversity and evenness, while the distance to the village and the road signifcantly afected     (Table 10). Te linear regression analyses revealed the relationship between the structural attributes (proportion of the number of trees >30 cm·dbh and the total basal area), diversity, and evenness with the elevation (Figure 12). Te regression was statistically signifcant for the increasing pattern of the percentage of trees greater than 30 cm dbh with increasing elevation (R 2 � 0.183, p < 0.01) (Figure 12(a)). Te total basal area of the trees in the sampling plots increased signifcantly with increasing elevation (R 2 � 0.0976, p < 0.05) (Figure 12(b)). Te higher Simpson's index (R 2 � 0.126, p < 0.05) and Pielou's evenness (R 2 � 0.098, p < 0.05) of the sample plots regarding increased elevation was also significant in the regressions (Figures 12(c) and 12(d)).

Do Tree Density, Dominance, and Diversity Difer between
BUCF and the ZNP? All 34 sample plots were located in the natural forest at altitudes ranging from 168 m to 812 m, dominated mainly by deciduous tree species, which characterizes the study forest as a tropical moist upper mixed deciduous forest [9]. Since all sample plots were in a continuous landscape accessible to nearby villages, local populations' dependency on them would likely be comparable. Tree density (191-201 individuals ha −1 ) in BUCF and the ZNP was lower compared to tree densities reported from the Popa Mountain Park (604-957 individuals ha −1 ) [27], the evergreen montane forest of Ywa Ngan township (312-1372 individuals ha −1 ) [28], and tropical deciduous forests (229-336 individuals ha −1 ) [29] in Myanmar. However, the current research's tree densities were higher than that of a degraded forest (168 individuals ha −1 ) in Oak-twin township [25] and comparable to that of the reserved tropical forest (229 individuals ha −1 ) reported from Seikphyu township [29] in Myanmar. Although the ZNP plots have higher tree density and dominance than the BUCF plots, there was no signifcant diference between them. Te dominance range (15.0-20.6 m 2 ·ha −1 ) in this study was well comparable to the value reported for the human-disturbed forest (17.77-24.47 m 2 ·ha −1 ) in the Popa Mountain Park [27]. However, it was much lower than those recorded in the Alaungdaw Kathapa National Park (60.03 m 2 ·ha −1 ) [25] and the Natmataung National Park (55.63 m 2 ·ha −1 ) [65], which have been protected since 1984 and 1994 in upper Myanmar.
Plant diversity indices provide more accurate measurements of species diversity and abundance in forests than species counts alone [66,67]. Te ZNP had higher diversity indices and Pielou's evenness than BUCF, although not so signifcantly. Te steeper gradient rank-abundance curve of the BUCF revealed unequal resource sharing (Figure 8), with Protium serratum and Dipterocarpus alatus having much greater relative dominance than neighbouring species (Table 5). Te growing human demand for wood for housing and fuel causes the repeated exploitation of preferable tree species, which can alter the species composition in the BUCF, where the forest law does not strictly prohibit the exploitation of forest products.   Te secondary forest plots had the highest species diversity, while the jungle forest plots had the lowest species diversity in rarefaction-based species richness ( Figure 5). Te limited diversity of woody species in the jungle forest today is due to the dominance of bamboo and shrubs, notably Pseudostachyum polymorphum, Dendrocalamus strictus, Ziziphus oenoplia, Caryota mitis, and Arenga nana after it was previously disturbed by the heavy extraction of forest products by road and farmland invasion. Te people's substantial dependence on unlawful access to forest resources may cause similar patterns of rarefaction-based species diversity in BUCF and the ZNP (Figure 6). We    anticipated that the similar level of species richness in the ZNP to the nearby unclassifed forest (BUCF) was caused by previous heavy logging, the exploitation of forest products, and human intrusion. Te ZNP became a protected area (PA) in 2022. We predict that the ZNP plots will evolve into excellent forest conditions if they are sufciently protected under the PA status. Te large-diameter trees (dbh > 60 cm) considerably contribute to biomass; thus, the increasing frequency of these giant trees in the ZNP (Figure 7) may increase structural variability and the forest's capacity to store carbon [68]. According to the distribution of the families, Verbenaceae (12.9%), Burseraceae (9.9%), and Dipterocarpaceae (9.9%) are the families with the highest representation in the BUCF. Te family with the highest representation in the ZNP plots is Euphorbiaceae (7.2%), followed by Dipterocarpaceae (6.7%) and Combretaceae (6.5%). Te highest abundance of Euphorbiaceae in the ZNP mirrored a feature of tropical forests reported by the study of the tropical forest in the Congo [69]. However, Verbenaceae, Burseraceae, Dipterocarpaceae, and Combretaceae are distinctive families to the tropical mixed deciduous forests frequently disturbed by logging activities [29]. Te abundance of Fagaceae and Fabaceae in some plots proves the old age or maturity of the    forest [28,69]. Each forest lacked a few families. For example, Apocynaceae, Caesalpiniaceae, Convolvulaceae, Lecythidaceae, and Ulmaceae were found in BUCF but not the ZNP. Oleaceae, Capparaceae, Juglandaceae, Lythraceae, and Tetramelaceae did not appear in BUCF while present in the ZNP. Elevation patterns, logging levels, and access to roads and settlements throughout the sample plots help explain the compositional variances of diferent families (Figures 10 and 11). Te IVI gives a total picture of the social structure of species in a community and can be used to form an association of dominant species [70]. Our results indicated that Protium serratum, Dipterocarpus alatus, Tectona grandis, Dillenia pentagyna, Schleichera oleosa, and Xerospermum noronhianum had the highest IVI (>15%), refecting their relatively high ecological importance in the study area. However, compared to the ZNP, BUCF revealed lower IVI and proportionately fewer trees per unit area for Baccaurea sapida, Dracontomelon dao, Terminalia crenulate, and Xerospermum noronhianum. Dalbergia oliveri, Dillenia pentagyna, Schleichera oleosa, and Tectona grandis had lower IVI and densities in the ZNP than in BUCF plots. Importantly, Tectona grandis (teak) had the best relative density in BUCF but severely decreased numbers in the ZNP. Te lower tree density of teak in the ZNP may be due to the efects of previous extensive logging by the government before its declaration as a national park. Due to its better wood quality and higher demand than other species, the overuse and overharvesting of teak causes the deterioration of forests and biodiversity [68,71]. Our results highlighted that Tectona grandis, Dalbergia oliveri, and Protium serratum had been overharvested in the ZNP plots resulting in a lower relative density and IVI. Tis fnding suggested that the last logging activities signifcantly afected the abundance and dominance of certain commercially important tree species in the ZNP. In the ZNP, 23 species (27.38%) of all species exhibited the lowest IVI (<1%), while 32 species (35.16%) of those in BUCF did the same. Te species with the lowest IVI ratings, including Melastoma spp. (Melastomataceae), Atalantia monophyla (Rutaceae), Zanthoxylum spp. (Rutaceae), and Dalbergia spp. (Fabaceae), are all tolerant to considerable shade. Due to their higher mortality, slower development, and reduced dispersion capacity, shadetolerant species are more vulnerable to disturbance and Te correlation is considered signifcant at p < 0.05 ( * ), p < 0.01 ( * * ), and p < 0.001 ( * * * ). fragmentation [72] than light-demanding species such as Tectona grandis, Terminalia crenulata, and Dillenia pentagyna, which showed greater IVI. Controlling fragmentationinduced human activities and using specifc silvicultural techniques, such as enrichment planting, are necessary if these tree species are to be conserved and managed to enhance their ecological value.

Proportion
Variations in structural attributes (tree density, dbh, basal area, and height) were afected by many factors and varied between species [73]. Te structural characteristics of woody vegetation were associated with forest types, land uses, human activities, and environmental factors such as elevation, slope, aspect, soil, and light conditions [74,75]. Te peculiarities of the habitat preferences and favoured light conditions afect Tectona grandis's much higher tree density in BUCF than in the ZNP [66]. Human activities like logging and exploiting forest products leave gaps in the forests, which may help certain species respond positively to light conditions [76]. In the ZNP, Dipterocarpus alatus performed slightly better than BUCF in tree density, dbh, basal area, and height. Te dominance of Dipterocarpus alatus (IUCN endangered species) demonstrated its habitat preferences and the signifcance of its protection. Dalbergia oliveri (IUCN endangered species) showed lower tree density, dbh, basal area, and height compared to Dipterocarpus alatus, demonstrating its lower dominance in the tree community. Te fact that the villager favour Tectona grandis, Dalbergia oliveri, and Terminalia crenulata for housing, frewood, and income may help to explain the frequency and dominance of Dipterocarpus alatus in the study area (source: the authors' unpublished interview data in April 2022).

Do Variations in Species Composition Relate to Environmental Factors?
Te logging intensity appears to be highly infuential on tree species present in the jungle and secondary forest of BUCF ( Figure 11). Logging operations favour light-demanding species such as Tectona grandis, Aporusa roxburghii, Schleichera oleosa, and Dipterocarpus alatus in the BUCF jungle and secondary forests since light infuences the abundance and dominance of pioneer and light-demanding species [77,78]. Te dominance of Dracontomelon dao in ZNP's secondary forest seemed to be infuenced by elevation (Figures 10 and 11). Tis endorsed the species' predilection for elevation up to 500 m above the sea level in the ZNP [79]. Te prevalence of Dillenia pentagyna and Terminalia chebula in BUCF plots near roads may be owing to their less-desirable timber (Figures 10 and  11). Local populations use these two species for food and medicine, not for wood. Te greater distance to the village afects the considerable composition of Swintonia foribunda (used as frewood) in the secondary forest of the ZNP, presumably due to the species' lessened stress from being far away from populated areas. NMDS ordination indicated that BUCF plots clustered closely while ZNP plots were more dispersed (Figures 10 and 11). Te presence of diferent species in each plot may help explain some of the dispersion seen in the ZNP. In the current study, we exclusively evaluated parts of the ZNP that are accessible and are under signifcant human danger, restricting the species composition and richness of the species. We suggested that additional edaphic factors, such as soil properties and moisture, along with canopy openness, may afect changes in tree composition.

Do the Stand Structure and Floristic Diversity Relate to
Environmental Factors? In the current research, higher altitudes, which are not easily accessible to gather forest resources, tended to have a signifcantly higher number of trees, more giant than 30 cm·dbh, a larger basal area, and more diverse, and evenly distributed tree species (Figure 12). Evidence suggests that human activities such as harvesting wood and forest products at lower elevations diminish the diversity of tropical forest tree species [80,81]. Te steeper slope has a favourable relationship with species diversity and evenness owing to difcult logging access but a negative link with the basal area and height may be due to a natural process of loosening soil structure [82], which may infuence the growth of trees (Table 10). Tere was a favourable relationship between tree species' dominance, diversity, and evenness and the distance of sample plots from the villages and roads (Table 10). Due to easy accessibility from the villages and road network, unlawful logging reduces tree species' dominance (basal area), diversity, and evenness. According to several studies, tree species' abundance, dominance, and diversity are all impacted by logging and the exploitation of forest products [83][84][85][86]. However, some studies claimed that logging creates gaps in the forest that allow shade-intolerant species to regrow and thrive, increasing their range and dominance [78,87,88].

Consequences for Conservation inside and outside of the Protected Area.
Two of the 116 total species, Dalbergia oliveri and Dipterocarpus alatus (IUCN endangered species), were discovered in both BUCF and the ZNP, demonstrating their moderate dominance and tree density per hectare. Te remnant forests of BUCF (outside PA) had a higher frequency and dominance of these two endangered species, suggesting its potential to boost biodiversity conservation. However, several studies have shown that protected areas promote greater abundance and dominance of endangered species, which contradicts this fnding [35,[89][90][91]. Te studied forests provide ecological services and assist local economies by giving timber, frewood, medicinal plants, and food supplies. Te profusion of high-quality timber species like Tectona grandis, Dalbergia oliveri, Dalbergia stipulaceae, Chukrasia velutina, Gmelina arborea, Quercus glauca, and Shorea spp. showed that forests could be crucial seed sources for future regrowth. However, an efcient conservation approach is needed to reduce overexploitation, which perils these species' survival, particularly in BUCF. Te three prominent species in both sites, Baccaurea sapida, Dracontomelon dao, and Dillenia pentagyna provide the bulk of edible food supplies for surrounding local populations. Cinamomum obtusifolia, Millettia cinerea, Garcinia paniculata, and Millingtonia hortensis were known to be medicinal and economically important species for local communities via the sale of these medicinal plant parts; yet, they prevailed in the ZNP.
Our study showed that the ZNP requires efective conservation compared to species richness in other protected areas in Myanmar [25,26,65]. However, the proximity of the sample plots to human-dominated areas may explain the restricted diversity of the species. In addition, selective logging was widespread in the ZNP before 2011, and local people's continuing harvest of forest products might afect the dominance and diversity of economically important species. We suggested that the ZNP would recover richer vegetation by the year after protection. Te whole part of the ZNP is said to have various types of forest, including evergreen, deciduous, dry dipterocarp, low indaing (a seasonally dry tropical forest), and native fora and wildlife [43], although ofcial statistics are not available. Te whole part of the ZNP could contain more species diversity than the current research. For instance, a recently discovered endemic plant species from Myanmar, Sapria myanmarensis (Rafesiaceae), was only found in Kachin State and the Sagaing Region, encompassing the ZNP [92].
Many tree species recorded in the BUCF with a high dominance were commercially valuable timber species. It was mistakenly believed that the BUCF region has minimal conservation importance, as it is an open forest environment and is not ofcially protected as a reserved forest. Our research indicated that while BUCF jungle forest plots have low basal area, old-growth and secondary forest plots are rich in economically relevant species. However, the BUCF is imperilled by road development, human habitation, agriculture encroachment, and illicit logging. Te continued fragmentation of the habitat of the BUCF will probably increase the disturbance to the nearby ZNP. Many studies reported that increased human disturbances quickly become problematic in conserving protected areas in Myanmar and tropical Asia [93][94][95][96]. For this reason, we urge the remaining forests of the BUCF to be acknowledged as an ecosystem of a high conservation value to protect its biodiversity and economic viability. Community forest management ensures the participation of the local community in forest management by granting 30-year land use rights and removing forest products [97]. Community forests should be established in the BUCF's degraded forest land to enable local engagement in forest management and potentially lessen stress inside the ZNP.

Conclusion and Recommendation
Trough our study, we determined the vegetation structure and diversity in and around the ZNP and how the composition of tree species had varied in response to various environmental factors. Despite the human impact on forests, the high richness and basal area of valuable timber species suggest that these forests remain essential for human well-being. Numerous medicinal plant species, including Lannea coromandelica, Cinamomum obtusifolia, Millettia cinerea, Garcinia paniculata, and Millingtonia hortensis, are present in the ZNP, demonstrating a high congruence of conservation values and meriting protection for both biodiversity and human security. Te dominance of two endangered tree species, Dipterocarpus alatus and Dalbergia oliveri, in BUCF presents a high conservation value and demands special attention in conservation planning. Te research found that even if the densities of commercially relevant tree species are declining in the ZNP and maintaining their big-diameter trees is favourable. Establishing permanent sample plots would be a helpful alternative for future evaluation to track the efectiveness of the protected area and assess the dynamics and natural regeneration of certain tree species in the ZNP. Moreover, the assisted natural regeneration and enrichment planting of valuable tree species should be promoted in degraded areas of the ZNP.
As the primary goals for establishing protected areas are to conserve forests and biodiversity, park managers focus mainly on conservation inside the park. We can understand the importance of protecting the ZNP as a PA. However, to efectively conserve the biodiversity of the ZNP, it is necessary to preserve it in the adjacent BUCF. Due to easy accessibility, unclassifed forests are more prone to species compositional fuctuation and loss. Tus, forest managers should take action to regularize the local exploitation of specifc tree species in BUCF. Controlling frewood collection and illicit logging will be crucial in unclassifed forest ecosystems. Encouraging fastgrowing tree species for frewood plantations in nearby villages might satisfy the people's fundamental demands for frewood and poles. Te local populations who depend on and live close to forests know the potential of tree species diversity and the value of ecological sustainability [98]. We expect nearby communities to actively participate in forest management and conservation eforts to ensure long-term sustainability. Myanmar's biodiversity and protected areas legislation allows the sustainable use of the bufer zone [99]. Terefore, creating bufer zones within and outside the ZNP may balance the competing interests of biodiversity protection and socioeconomic development. Te favourable stand density and species richness of BUCF provide a future avenue for conservation. In the jungle areas of the BUCF, community forestry development might encourage local participation in forest management, therefore decreasing poverty and reliance on the forest in the park's core part. Tis study does not indicate that the protected areas are unnecessary. Instead, the fndings of this study suggest that if the remaining natural habitats in the BUCF, where humans collect forest resources, are efectively managed to encourage native species to persist, we may gain enormous biodiversity both inside and outside the protected area.

Data Availability
Some of the data used to support the fndings of this study are included within the article and the supplementary materials, and the remaining data (fgures and tables in R and Excel) are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Acknowledgments
Te PhD program of the corresponding author is being funded by the Japanese Government MEXT scholarship program. Te authors would like to thank MEXT as a fnancial supporter to conduct this research. Open Access funding was enabled and organized by JUSTICE Group 1 2023.

Supplementary Materials
Appendix: the IVI lists of tree species in the ZNP and BUCF. (Supplementary Materials)