A plant community is a collection or association of plant species within a specific location, which forms a relatively uniform patch, distinguishable from neighbouring patches of different forest types (Singh
Although the tropical dry forests (DF) of Vietnam have been disturbed by humans for years, they still cover approximately 650,000 ha of the country's land surface (Huy
The study area is situated in the Central Highlands, which is one of the eight agroecological regions of Vietnam. It is located between the north latitude 12°45′N–13°20′N and east longitude 107°30′E–108°30′E. This region has an average altitude of 500–600 m.a.s.l., with a hot and humid equatorial climate during the entire year. The area has a unimodal rainfall regime of up to 1600 mm of mean annual rainfall, considering the period of interest from 2004 to 2020. The longer dry season stretches from November to April. The mean annual temperature is about 25.5°C, while 12.4°C and 28.3°C are the minimum and maximum temperatures, respectively (Huy
The classification of the Forest Inventory and Planning Institute of Vietnam for tropical dry forests is based on their different standing volume levels. For instance, rich dry forests (DFR) comprise standing woody volume levels of > 200 m3 ha−1, the medium dry forest (DFM) of 101–200 m3 ha−1, the poor dry forest (DFP) of 51–100 m3 ha−1 and the very poor dry forest (DFV) of 10–50 m3 ha−1, respectively. This is how the four different explored DF types were classified in this study.
For the data collection of tree species, a total of 103 quadrats of 0.1 ha each (31.63 m × 31.63 m) were laid randomly as follows: 10 plots in DFR, 19 in DFM, 25 in EBP and 45 in DFV, respectively (Figure 1). Within each sample plot several variables were measured for all trees ≥ 5 cm in diameter, including species name, diameter at breast height (
Figure 1
Geographic location of the study area. The upper left panel shows the location of Vietnam's Central Highlands Region (pink colour); the lower left panel shows the location of the study area (light green colour) and the right panel shows the vegetation cover of tropical dry dipterocarp forests (fluorescent green colour) and the distribution of sampled plots for the four different dry forest (DF) types.

For the seedling and sapling survey (i.e., individuals with height > 30 cm and DBH < 5 cm), five subsample plots of 4 m × 4 m, i.e., each of 16 m2, were laid out in four corners and in the centre of the sampled plot (31.63 m × 31.63 m). Therefore, a total of 515 plots covering 0.824 ha were used for the seedling and sapling survey.
The data were analysed to quantify tree species diversity, composition, density, and forest structure across the four DF types. The importance value index (IVI) was used to assess the distribution of species abundance. It was computed as a sum of relative density, relative frequency, and relative basal area. The IVIi for species and
Rdi (relative density of species i) was calculated as:
RBAi (relative basal area of species i) was calculated as:
The quantitative tree species diversity indicators that were used in this study include:
Margalef's Index (Margalef, 1958):
Species rank distribution curves were also used to determine the relative importance of tree species in terms of diversity (Matthews & Whittaker, 2015) in the four forest types. Further, species similarity among different regions was computed using Jaccard's Coefficient of Similarity (Jaccard, 1912). The value of the coefficient varies between 0 and 100%, where 100 means the two sites have the same composition (they share the same number of species), and 0 means the two sites do not share any species.
Tree population structure was analysed using different DBH size class distributions across the four types of forest. To improve the visualization and comparison of tree diameter distribution among the four forest types, the individual stems recorded in each forest type were classified into fourteen different DBH classes with 5 cm interval between one another.
Due to lack of normality and homogeneity of variance, Kruskal–Wallis tests were used to determine significant differences between the forest types. Post hoc pairwise multiple comparisons were performed using Wilcoxon rank sum test. Finally, we performed Chi square tests to determine significant differences between tree diameter distributions for the four forest types. All significant differences reported here refer to
Figure 2 shows a ‘collector's curve’ of woody species densities plotted as a function of the number of sites enumerated. Cumulative densities of species substantially increased with each additional plot enumerated.
Figure 2
Tree species-area curve between species richness and the 103 main plots that were sampled within a total of 10.3 ha of natural forest in Vietnam's Central Highlands Region.

However, the rate of increase diminished with increasing numbers of plots after the seventh site. The ‘collector's curve’ flattened out as more woody species specimens were enumerated.
A total of 4914 trees belonging to 87 species from 37 families were recorded within a sampled area of 10.3 ha (103 plots). Of the total recorded individuals, 32 species (19 families) were found in DF rich forest, 54 species (30 families) were found in DF medium forest, 52 species (28 families) were found in DF poor forest, and 66 species (30 families) were found in DFV.
Across the four forest types studied, we found 15 rare and threatened species listed on the International Union for Conservation of Nature (IUCN) Red List (Supplementary Table S1), ranging from Near Threatened (NT) to Endangered (EN). The communities with the highest number of rare and threatened species were DF medium forest and DF very poor forest.
The five most abundant families were Dipterocarpaceae (n = 2953), Combretaceae (414), Fabaceae (254), Phyllanthaceae (230) and Leguminosae (158) and their most dominant species in each family were
Basal area ranged from 5.74 m2 ha−1 (DFV) to 23.7 m2 ha−1 (DFR) while stem density ranged from 431.05 stems ha−1 (DFM) to 562.40 stems ha−1 (DFP) (Figure 2). DFR had significantly higher species richness, family richness, basal area, mean diameter and mean height than DFP and DFV. There were no significant differences between DFR and DFM, except for species richness and basal area which were significantly higher in DFR than in DFM. Similarly, no significant differences in stem density were observed among forest types (Figure 3).
Figure 3
Species richness (number of species per plot); stem density (number of individual stems per plot); family richness (number of families per plot); basal area (sum of basal area per plot); mean diameter (mean diameter per plot); mean height (mean height per plot) for the four dry forest (DF) types studied.

Species rank abundance distribution curves show a similar pattern with inverted J-shaped species distributions with relatively flatter curves with high species richness and evenness for the four forest types. All of them had 5–10 species with high richness (larger distance between two adjacent species).
Figure 4
Relationship between species relative importance and species rank in abundance for the top 5 tree species recorded in the four dry forest (DF) types studied.

Analysis of the Jaccard's Coefficient of Similarity showed that all the DF types in the Central Highlands of Vietnam have an average similarity of more than 49% in terms of tree species presence. DFM and DFP shared the maximum similarity of 59.09% followed by DFP and DFV (58.11%) respectively. However, the lowest similarity (36.11%) in tree species occurrence was observed between DFR and DFV (Table 1).
Similarity Index (J’) matrix between four dry forest (DF) types in Vietnam's Central Highlands Region.
Forest types | DFR | DFM | DFP | DFV |
---|---|---|---|---|
DFR | 45.76% | 43.10% | 36.11% | |
DFM | 59.09% | 53.85% | ||
DFP | 58.11% | |||
DFV |
As in other tropical dry dipterocarp forests of Vietnam,
Figure 5
Top ten tree species and their contribution based on their importance value index (IVI), in terms of relative basal area (RBA), relative density (RD) and relative frequency (RF) for the four dry forest (DF) types studied in Vietnam's Central Highlands Region.

Generally, the diversity of the four forest types is at a moderate level with the index of 1 < H’ < 3 (Table 2). DFR had a higher species richness (R), species diversity (H’) and dominance (D) than other vegetation types (
Consolidated details of tree diversity in four types of tropical dry dipterocarp forests in Vietnam's Central Highland Region.
Forest types | Margalef's Index (R) | Shannon-Wiener diversity (H’) | Simpson Index (D) | Evenness Index (E) |
---|---|---|---|---|
DFR | 8.43±1.03a | 2.29±0.16a | 0.87±0.02a | 0.32±0.02a |
DFM | 5.90 ±1.62b | 1.85±0.38b | 0.78±0.12b | 0.34±0.04a |
DFP | 5.44±1.80b | 1.80±0.32b | 0.77±0.08b | 0.35±0.07a |
DFV | 5.88±2.00b | 1.81±0.37b | 0.76±0.11b | 0.34±0.06a |
Kruskal-Wallis chi-squared | 10.553 | 10.572 | 9.9064 | 1.4935 |
p-value | 0.0144 | 0.01428 | 0.01938 | 0.6838 |
Note: Different letters (a, b) within the same column denote significant differences among indices (
In the tropical dry dipterocarp forests of the region, the density of individual tree species decreased with increasing tree size. The average number of individual trees in 0.1 ha plots was 47.71 ± 24.92 (ranging from 15 to 140 individuals). The maximum density of trees per unit area was 410.38 stems ha−1 and it was observed in the DBH class of 5–25 cm, which contributed to 86.02% of the total tree population (Figure 6). About 15.7% of the trees in the sampled plots were included in the 5–10 cm diameter class, while the 10–15 cm class accounted for 34.3%. The highest tree stand density and species richness were found in these two diameter ranges, which accounted for 49.9% of enumerated trees. The above 15 cm of DBH classes contributed 50.1% to the total number of sampled trees. In other words, there was a normal population structure or ‘inverted J-shaped’ distribution of diameters in four forest types. This type of tree population structure implies good reproduction but hinders recruitment capacity (Angessa
Figure 6
Frequency (%) of trees according to different DBH classes for the four dry forest (DF) types studied. Notes: DFR: rich dry forest; DFM: medium dry forest; DFP: poor dry forest; DFV: very poor dry forest.

Among the four types of tropical dry dipterocarp forests, the best tree regeneration was observed in DFV with 21,734 seedlings and saplings per hectare. The number of seedlings and saplings was the lowest (6,825 individuals) in the sampled areas of DFR. Adequate regeneration of trees was noticed in DFM (13,829 individuals) and DFP (13,095 individuals) as compared to the other forest types under study. The regeneration status of species like
Figure 7
Regeneration status for the top 10 tree species recorded in the four dry forest (DF) types studied. Notes: DFR: rich dry forest; DFM: medium dry forest; DFP: poor dry forest; DFV: very poor dry forest.

There are large variations in species richness, basal area, and stem density in tropical dry forest ecosystems due to different climatic conditions, topography, and elevation gradients (Supplementary Table S3). Our results showed that DFR had significantly higher species richness, family richness, basal area, tree size and diversity than DFP and DFV, respectively. This variation may be due to the difference in management activities and past disturbance between both categories of forests (Sapkota
Similarly to other tropical dry forests of Vietnam, the five most abundant families in the study area were Dipterocarpaceae, Combretaceae, Fabaceae, Phyllanthaceae and Leguminosae (Nguyen & Baker, 2016; Do
In this study, the total basal area ranged from 5.74 m2 ha−1 in DFV to 23.7 m2 ha−1 in DFR. The basal area of the present study was very similar to that found in the tropical dry forests of Nepal (Bhatta & Devkota, 2020), India (Chaturvedi & Raghubanshi, 2014), Brazil (Coelho
Tree diameter distribution reflects the disturbance effect within the forests and it is helpful in detecting trends of regeneration patterns (Poorter
The forest wealth depends on the potential regenerative status of species composing the forest stand, in space and time (Jones
Floristic inventories and vegetation analysis for the tropical dry forest of the Central Highlands Region in Vietnam showed that tree stem density, basal area and Shannon Index were relatively low in the four forest types studied, particularly due to past disturbance such as anthropogenic, selective logging, deforestation for agricultural purposes and natural fires. The large population and regeneration of sapling trees in DF in the Central Highlands Region, suggest that this is a second-growth forest with high regeneration potential. With agricultural communities located in the buffer zone for the protection of the forest with a fast-growing economy, the pressure on forest resources is increasing. Therefore, a systematic management plan is required for their conservation and sustainable use.
Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Similarity Index (J’) matrix between four dry forest (DF) types in Vietnam's Central Highlands Region.
Forest types | DFR | DFM | DFP | DFV |
---|---|---|---|---|
DFR | 45.76% | 43.10% | 36.11% | |
DFM | 59.09% | 53.85% | ||
DFP | 58.11% | |||
DFV |
Structure, composition, and diversity of tree species in tropical dry forests of the present and previous studies.
ID | Study area | Sampled area (ha) | Minimal DBH (cm) | Number of species | Number of families | Stand density | Stand basal area | Shannon-Weiner | Simpson | Evenness | Margalef's | Seedling (individuals / ha) | Sapling (individuals / ha) |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Dadeldhura district, Western Nepal ( |
1.05 | >5 | 42 | 20 | 3162.9±194.3 | 13.77±0.88 | 15905±927.3 | 1876.2±143.4 | ||||
2 | 3 provinces from Central Highlands of Vietnam ( |
6.00 | >10 | 157 | 622 | 2.06 | 0.80 | ||||||
3 | Huai Kha Khaeng Wildlife Sanctuary, Thailand ( |
0.75 | >4.5 | 54 | 667–1558 | 12.9–23.6 | 13000–40000 | 573–4570 | |||||
4 | Fasiakhali Wildlife Sanctuary, Bangladesh ( |
3.2 | >11 | 32 | 19 | 249 | 46.17 | 2.05 | 0.75 | ||||
5 | Eastern Ghats, Southern India ( |
7.9 | >10 | 128 | 45 | 994 | 5.1±3.6 | 1.1–3.5 | 0.4–1.0 | 0.5–0.9 | |||
6 | The North-Western Coast of Costa Rica ( |
0.9 | >5 | 96 | 41 | ||||||||
7 | YokDon National Park, Vietnam ( |
2.8 | >1 | 38 | 1229 ±523 | 19.12 ±5.89 | 13383 | ||||||
8 | The Minas Gerais, Brazil ( |
1.8 | >5 | 59 | 19 | 510–1091 | 3.13–25.34 | 0.9–1.3 | |||||
9 | Central America ( |
0.7 | >2.5 | 204 | 58 | 1350–2650 | 17.7–25 | ||||||
10 | Guanacaste Province, Costa Rica ( |
13.44 | >2 | 135 | |||||||||
11 | Sri Lanka ( |
0.12 | >5 | 446–1500 | 11.4–44.8 | 0.6–0.9 | 694–3400 | 773–4475 | |||||
12 | The North West Region of Cameroon ( |
12.3 | >10 | 178 | 42 | 157–404 | 6.8–32.4 | 2.7–3.5 | |||||
13 | Michoacán, Mexico ( |
0.36 | >1 | 78 | 24 | 700–5600 | 9–44 | 0.41–3.5 | 0.04–0.71 | ||||
14 | Bandipur, South India ( |
>1 | 66 | 2.46 | 0.81 | ||||||||
15 | The Accra Plains in South-East Ghana ( |
0.49 | >10 | 16 | 1332 | ||||||||
16 | Northern Sri Lanka ( |
0.12 | >5 | 32 | 18 | 446 | 1.54–2.29 | 0.86–0.94 | |||||
17 | Vindhyan Highlands, India ( |
20 | >10 | 52 | 126–490 | 3.1–18 | 1.5–2.5 | 1.7–3.7 | 5040–14167 | 540–1383 | |||
18 | Minas Gerais, Southeastern Brazil ( |
0.3 | >5 | 50 | 21 | 1076–1226 | 17.8–29.3 | 1.52–2.58 | 0.48–0.64 |
Nationally rare and IUCN red list of threatened species recorded in four different dry forest (DF) types of Vietnam's Central Highlands.
Species name | IUCN | DFR | DFM | DFP | DFV |
---|---|---|---|---|---|
EN | - | - | - | 1 | |
EN | - | 1 | - | - | |
EN | - | - | - | 1 | |
EN | - | 1 | 3 | 12 | |
EN | 5 | 2 | 21 | 41 | |
LR | - | 1 | - | - | |
NT | 8 | 157 | 250 | 376 | |
NT | 7 | 158 | 242 | 323 | |
NT | - | 15 | 15 | 37 | |
VU | - | 2 | 1 | 33 | |
VU | - | - | - | 1 | |
VU | - | 1 | - | - | |
VU | - | 24 | 1 | 5 | |
VU | - | - | - | 2 | |
VU | 26 | 85 | 309 | 448 |
List of species richness found in four different dry forest (DF) types of Vietnam's Central Highlands. Tree species are ranked from the highest to the lowest based on their importance value index (IVI).
DFR | DFM | DFP | DFV | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
No | Species | IVI | No | Species | IVI | No | Species | IVI | No | Species | IVI |
1 | 52.1 | 1 | 37.8 | 1 | 52.2 | 1 | 40.8 | ||||
2 | 42.8 | 2 | 37.4 | 2 | 41.6 | 2 | 37.8 | ||||
3 | 35.6 | 3 | 36.3 | 3 | 40.9 | 3 | 32.7 | ||||
4 | 32.6 | 4 | 19.1 | 4 | 22.6 | 4 | 23.1 | ||||
5 | 15.2 | 5 | 18.1 | 5 | 17.8 | 5 | 14.8 | ||||
6 | 13.1 | 6 | 15.6 | 6 | 14.1 | 6 | 13.8 | ||||
7 | 11.6 | 7 | 11.2 | 7 | 12.1 | 7 | 11.5 | ||||
8 | 8.6 | 8 | 10.2 | 8 | 10.1 | 8 | 10.2 | ||||
9 | 7.3 | 9 | 8.8 | 9 | 6.8 | 9 | 9.9 | ||||
10 | 6.3 | 10 | 7.4 | 10 | 6.1 | 10 | 6.8 | ||||
11 | 6.3 | 11 | 7.1 | 11 | 5.8 | 11 | 6.3 | ||||
12 | 5.3 | 12 | 6.6 | 12 | 5.1 | 12 | 5.9 | ||||
13 | 4.9 | 13 | 6.5 | 13 | 4.3 | 13 | 5.6 | ||||
14 | 4.8 | 14 | 5.6 | 14 | 4.3 | 14 | 4.6 | ||||
15 | 4.6 | 15 | 5.6 | 15 | 3.5 | 15 | 4.5 | ||||
16 | 4.5 | 16 | 5.0 | 16 | 3.4 | 16 | 4.2 | ||||
17 | 4.1 | 17 | 4.2 | 17 | 3.3 | 17 | 4.1 | ||||
18 | 3.5 | 18 | 4.1 | 18 | 2.9 | 18 | 4.0 | ||||
19 | 3.3 | 19 | 3.9 | 19 | 2.9 | 19 | 3.9 | ||||
20 | 3.3 | 20 | 3.9 | 20 | 2.8 | 20 | 3.2 | ||||
21 | 3.2 | 21 | 3.4 | 21 | 2.3 | 21 | 3.1 | ||||
22 | 3.1 | 22 | 3.4 | 22 | 1.9 | 22 | 3.1 | ||||
23 | 3.0 | 23 | 2.9 | 23 | 1.9 | 23 | 2.5 | ||||
24 | 2.6 | 24 | 2.7 | 24 | 1.8 | 24 | 2.4 | ||||
25 | 2.5 | 25 | 2.6 | 25 | 1.8 | 25 | 2.3 | ||||
26 | 2.5 | 26 | 2.5 | 26 | 1.8 | 26 | 2.2 | ||||
27 | 2.4 | 27 | 2.1 | 27 | 1.8 | 27 | 2.1 | ||||
28 | 2.3 | 28 | 2.0 | 28 | 1.7 | 28 | 1.9 | ||||
29 | 2.2 | 29 | 1.8 | 29 | 1.6 | 29 | 1.9 | ||||
30 | 2.1 | 30 | 1.5 | 30 | 1.6 | 30 | 1.7 | ||||
31 | 2.1 | 31 | 1.5 | 31 | 1.6 | 31 | 1.6 | ||||
32 | 2.0 | 32 | 1.4 | 32 | 1.4 | 32 | 1.5 | ||||
33 | 1.4 | 33 | 1.4 | 33 | 1.5 | ||||||
34 | 1.2 | 34 | 1.3 | 34 | 1.4 | ||||||
35 | 1.1 | 35 | 1.1 | 35 | 1.4 | ||||||
36 | 0.9 | 36 | 1.1 | 36 | 1.3 | ||||||
37 | 0.9 | 37 | 1.0 | 37 | 1.3 | ||||||
38 | 0.9 | 38 | 1.0 | 38 | 1.3 | ||||||
39 | 0.9 | 39 | 1.0 | 39 | 1.2 | ||||||
40 | 0.8 | 40 | 1.0 | 40 | 1.1 | ||||||
41 | 0.8 | 41 | 0.8 | 41 | 1.0 | ||||||
42 | 0.8 | 42 | 0.7 | 42 | 0.9 | ||||||
43 | 0.8 | 43 | 0.6 | 43 | 0.9 | ||||||
44 | 0.7 | 44 | 0.6 | 44 | 0.9 | ||||||
45 | 0.7 | 45 | 0.6 | 45 | 0.9 | ||||||
46 | 0.7 | 46 | 0.6 | 46 | 0.8 | ||||||
47 | 0.7 | 47 | 0.6 | 47 | 0.7 | ||||||
48 | 0.7 | 48 | 0.6 | 48 | 0.7 | ||||||
49 | 0.7 | 49 | 0.5 | 49 | 0.7 | ||||||
50 | 0.6 | 50 | 0.5 | 50 | 0.6 | ||||||
51 | 0.6 | 51 | 0.5 | 51 | 0.6 | ||||||
52 | 0.6 | 52 | 0.5 | 52 | 0.6 | ||||||
53 | 0.6 | 53 | 0.6 | ||||||||
54 | 0.6 | 54 | 0.5 | ||||||||
55 | 0.5 | ||||||||||
56 | 0.5 | ||||||||||
57 | 0.5 | ||||||||||
58 | 0.5 | ||||||||||
59 | 0.4 | ||||||||||
60 | 0.4 | ||||||||||
61 | 0.3 | ||||||||||
62 | 0.3 | ||||||||||
63 | 0.3 | ||||||||||
64 | 0.3 | ||||||||||
65 | 0.3 | ||||||||||
66 | 0.3 |
Consolidated details of tree diversity in four types of tropical dry dipterocarp forests in Vietnam's Central Highland Region.
Forest types | Margalef's Index (R) | Shannon-Wiener diversity (H’) | Simpson Index (D) | Evenness Index (E) |
---|---|---|---|---|
DFR | 8.43±1.03 |
2.29±0.16 |
0.87±0.02 |
0.32±0.02 |
DFM | 5.90 ±1.62 |
1.85±0.38 |
0.78±0.12 |
0.34±0.04 |
DFP | 5.44±1.80 |
1.80±0.32 |
0.77±0.08 |
0.35±0.07 |
DFV | 5.88±2.00 |
1.81±0.37 |
0.76±0.11 |
0.34±0.06 |
Kruskal-Wallis chi-squared | 10.553 | 10.572 | 9.9064 | 1.4935 |
p-value | 0.0144 | 0.01428 | 0.01938 | 0.6838 |
Effect of interaction between Fusarium solani andRhizoctonia solani on damping-off and root rot disease ofTetraclinis articulata seedlingsEffect of growth conditions on wood properties of Scots pine ( Pinus sylvestris L.)Suitability of the boreal ecosystem simulator (BEPS) model for estimating gross primary productivity in hemi-boreal upland pine forest Composition of live, dead and downed trees in Järvselja old-growth forest Changes in the forest structure of the Bashkortostan Republic over two decades The effect of implementation on successful forest management policy, moderated by actor-network and stakeholder collaboration Facilitating long-term 3D sonic anemometer measurements in hemiboreal forest ecosystems Optimizing the pine wood drying process using a critical diffusion coefficient and a timed moistening impulse Comparative study of Central and Eastern European alliances of thermophilous oak forests ( Quercion petraea ,Betonico-Quercion , andLathyro-Quercion ) within the temperate zone of EuropePuhtpuistute hinna- ja kasumiküpsuse empiiriline analüüs Eesti majandusmetsades ning selle võrdlus uuendusraiet lubava vanusega Variations in habitat metrics along plantation chronosequences: contrasting tree plantations in subtropical forest and grassland The changes of extractive contents of young Swietenia mahagoni (L.) Jacq trees during heartwood formationTree diversity and species composition of tropical dry forests in Vietnam's Central Highlands Region Seed viability and optimal germination conditions of Cedrus atlantica (Manetti ex Endl.) Carrière