Conservation of biodiversity and of endangered tree species in tropical forests represents a challenge in the era of deforestation and overlogging. Assessments of biodiversity and of conservation species are essential to estimating the conservation value of ecosystems in order to develop appropriate management measures. This paper discusses the floristic composition, the diversity, and the regeneration of the conservation species in a logged forest in southern Vietnam. Twenty five one hectare plots have been set up using circular sub-plots of 8 metres diameter and at a distance of 10m from the centre point of other sub-plots. The sampling is approximately 50% in each plot. Ten years after the last logging, the species richness appears low, when compared with the close protected area, but the diversity is quite high and conservation species still occur. Twelve conservation species have been studied from the point of view of their distribution and regeneration. Some are close to extinction but some others show promise in terms of regeneration. Six floristic groups have been identified and, among them, two are of interest from a conservation point of view. This logged forest still has a high conservation value, and the process that aims at developing protection measures is clearly relevant.
Introduction
Vietnam is one of the world's 10 most biologically diverse countries. Situated at the crossing of three different routes of flora migration, namely the Himalaya-South China route in the North-west, the India-Myanmar route in the West and the Malaysia-Indonesia route in the South, Vietnam has a very high number of plant species [1]. But for many reasons, such as rapid changes in land use, over-harvesting, and long wars, the forests of Vietnam have severely declined, both in quantity and quality [2]. Forests covered 12.3 million ha or 37% of the country's area, in 2005, but 18% was plantation and 70% of the remaining forest was of poor quality (Vietnam Environmental Monitors). Forests have been logged and even overlogged, mostly during the ‘80s. Consequently, forest resources have been highly damaged and many species within Vietnam run the risk of extinction, of which around 150 are critically endangered and vulnerable [3]. These species are called conservation species in this paper. In 1992, the Vietnamese government banned log exports, and then banned logging in many state-managed forests in 1997. A National Biodiversity Action Plan was adopted, in 1995, to guide and focus the nation's conservation effort, and a further plan, covering the period 2006–2015 with a vision for 2020, is being implemented. However, Vietnam has to face two major problems, biodiversity conservation and wood supply. As a result, the terrestrial protected areas system represented only 6.17% of the total country [4] which is insufficient for sustainable conservation. In this context, the question is how logged forests can participate in biodiversity conservation.
Logged forests are generally seen as having a much lower conservation value than protected forests. After logging, the species richness is generally reduced and the composition changed in favor of pioneer trees [5, 6, 7]. However, human-induced disturbances in tropical forests span a wide gradient, depending on land-use type and intensity [8]. For Ashton [9], little information is as yet available on the impact of successive felling cycles in biodiverse tropical rain forests, and such research is now a critical requirement. In spite of the damage caused by logging, secondary forests may act as reservoirs for recolonization and as corridors between remaining primary forest fragments. From a biodiversity conservation perspective, it is important to know how secondary forests compare with primary forests in terms of species richness and species similarity [10]. It is in this context that Tan Phu forest (Fig. 1) has been chosen for study, since it represents one of the last lowland forests in southern Vietnam and has been logged intensively during the last decades.
This paper aims to answer three questions:
Methods
Study area
Tan Phu forest (Fig. 1) is located in southern Vietnam and covers 13 900 ha. Annual rainfall, measured at Talai, is 2,715 mm/yr. Rainfall is mostly spread between mid-May and mid-November. Mean annual temperature is 26°C with an amplitude of 3.7 °C. Soils are sandy and clayic, organic matter is poor (0.2 to 3%), rate of iron (0.1 to 12 mg/100g) and aluminum (0.1 to 2 mg/100g) are high, and pH is acid (5.65 to 4.08). The original forests have been described by Maurand [11]. In the past, forests were characterized by a large number of species belonging to Dipterocarpaceae and Fabaceae, associated in a complex mixture, and by two other families, Lythraceae and Clusiaceae. The main species were Afzelia xylocarpa, Anisoptera costata, Dalbergia bariensis, Dipterocarpus alatus, Dipterocarpus dyeri, Heritiera cochinchinensis, Hopea odorata, Mesua ferrea, Sindora cochinchinensis, Shorea vulgaris, Vatica odorata and Xylia xylocarpa. It is also stated in the Indochina forest department records that logging started in the 1940s. However, there is no detailed data on logging before 1978. For the period 1978–1986, logged trees, over 60 cm in diameter, came from Anisoptera costata, Dipterocarpus alatus, Dipterocarpus dyeri, Heriteria cochinchinensis, Hopea odorata, Lagerstroemia calyculata, Litsea pierrei, Shorea roxburghii, Shorea thorelii, Sindora cochinchinensis, and Xylia xylocarpa. Successive and intense logging continued until 1996, and then logging was banned by the Vietnamese government.
Sampling strategy
We developed a sampling strategy incorporating both the structure and floristic composition of forest stands. First, we conducted a preliminary survey to identify different types of forest and to locate replicate stands for each situation. Based on this survey, forest inventory plots were established across the forest area. The classic method using square sub-plots was tested, but the high density of climbers and shrubs in the undergrowth made setting up sub-plots slow and difficult. We found an alternative method using circular sub-plots which are positioned on the centre point of a 10m square network (Fig. 2). This does not require delimitation of the area. Trees (diameter ≥ 10 cm at dbh) are sampled if the distance from the center of the circle is less than or equal to 4 metres. This was a systematic method of sampling that covered close to 50% of the area and was less time-consuming. Regeneration (diameter < 10 cm and height > 2m) was sampled if the distance from the center of the circle was no greater than 1.2 metre. Tree species were identified using the Flora of Ho [12]. In total, 25 plots, each with 100 circular sub-plots, were established in Tan Phu forest. Plots were mainly located in the eastern part of the forest area as the western part is partly covered by plantation and partly devoted to agriculture (Fig. 1).
Data analysis
Estimation of species richness
The species richness estimators, ICE (Incidence-based Coverage Estimator) and ACE (Abundance-based Coverage Estimator) are modifications of the Chao and Lee [13] estimators discussed by Colwell and Coddington [14]. These have been calculated using EstimateS software [15].
Shannon-Wiener index (H)
A widely used measure of biological diversity is Shannon's index. This index is also referred to as “Shannon's information measure” or “entropy” [16].
Simpson's Index of Diversity 1 - D
The index represents the probability that two individuals randomly selected from a sample will belong to different species. D is the squared sum of the proportion of species relative to the total number of species. The value of the index ranges between 0 and 1, and the greater the value, the greater the sample diversity [17].
Results
1. Description and conservation value of the floristic groups in Tan Phu forest
A graphic representation of the Jaccard similarity index table is presented in Fig. 3. The closest plots are 1 and 22 with a percentage of shared abundance equal to 62%. The most different ones are plots 4 and 12 which share only 15% of their abundance. According to the hierarchic classification, six floristic groups of plots can be identified and their similarity indices are presented in Table 1.
Table 1:
Specific richness (S), Shannon index (H), Simpson index (D) and Jaccard similarity index calculated for the different floristic groups and diversity value for the total of plots
The floristic composition of each group is presented in the following appendices:
1a: This group is mainly characterized by Shorea roxburghii (which represents 16.74% of the total abundance of the species), Syzygium zeylanicum (9.14%), Parinari annamensis (7%), Irvingia malayana (4.39%), and Calophyllum dongnaiensis (3.68%). This group is composed of seven conservation species, all of which are Dipterocarpaceae species which represent 21.85% of the total abundance. Shorea roxburghii is the dominant one.
1b: This is characterized by Diospyros maritima (8.46%), Memecylon sp. (6.63%), Irvingia malayana (5.55%), Vitex tripinnata (5.06%), Vatica odorata (4.31%), Syzygium chanlos (3.56%) and Hopea odorata (3.48%). It has 10 conservation species and among this 8 Dipterocarpaceae which represent 15.25% of the total abundance (N = 85) and 2 Fabaceae species at a very low rate (0.91%).
2a: This is composed of Syzygium chanlos (8.18%), Vatica odorata (6.28%), Dipterocarpus dyeri (4.60%), Diospyros lanceafolia (4.09%) and Grewia paniculata (4.05%). The conservation value is close to those of 1b with 8 Dipterocarpaceae species which represent 13.54% and Sindora cochinchinensis (0.75%).
2b: This is composed of Diospyros lancefolia (5.83%), Acronychia pedunculata (5.58), Litsea pierrei (3.93), Xerospermum noronhianum (3.93), Dipterocarpus dyeri (3.68%), and Syzygium chanlos (3.04%). It contains a total of 6 Dipterocarpaceae species (6.14%) and Sindora cochinchinensis (0.51%)
3a: This is characterized by Lagerstroemia calyculata (16.72%), Xerospermum noronhianum (12.74%), Diospyros sp. (8.60%) and Dipterocarpus alatus (6.79%). In addition, it has four other Dipterocarpaceae species (for a total of 8.44%) and 2 Fabaceae species (0.99%).
3b: This is composed of Cratoxylon formosum (11.29%), Acronychia pedunculata (9.53%), Grewia paniculata (9.47%), Vitex tripinnata (6.85%) and Lagerstroemia calyculata (5.42%). Dipterocarpaceae and Fabaceae species are non-dominant, respectively 4.31% and 2.15%, but at the same time, 11 conservation species are represented.
Globally, the 3 most abundant families are Dipterocarpaceae (11.7%), Myrtaceae (10.5%) and Clusiaceae (7.1%). Lythraceae represents 3.6% and Fabaceae only 1%.
2. Biodiversity value of the Tan Phu forest
The total number of species found for all the 25 plots, equivalent to 12.5 ha, in the Tan Phu forest is 176 and, in terms of diversity, the Shannon index is 6.02 and the Simpson index is 0.98 (Table 1). By using ACE and ICE estimators, the total species richness can be estimated at 207 and 205, respectively. Observation of the species richness accumulation curves (Fig. 4) shows that the species richness is the highest in 3b and 2b with 111 species on 2.5 ha and 89 species on 1.5 ha, respectively. The Shannon index varies from 4.73 in 3a to 5.67 in 2a and the Simpson index changes from 0.93 in 3a to 0.97 in 1b, 2a, and 2b. The similarity is highest between 1a and 1b and lowest between 1a and 3a (Table 1).
In terms of conservation, 12 threatened species are found in the Tan Phu forest; these are Afzelia xylocarpa, Anisoptera costata, Dalbergia bariensis, Dipterocarpus alatus, Dipterocarpus dyeri, Hopea odorata, Shorea guiso, Shorea roxburghii, Shorea thorelii, Sindora cochinchinensis, Vatica odorata, and Xylia xylocarpa.
3. Distribution and regeneration of the conservation species
Distribution of the frequency of trees and regeneration of each conservation species are presented in Fig. 5. Four Dipterocarpaceae species are well represented in Tan Phu. These are, namely, Dipterocarpus alatus (1.4% of the trees), Dipterocapus dyeri and Shorea roxburghii (2.6%), and Vatica odorata (2.97%). These are present in more than half of the total number of plots. The regeneration is numerous for Vatica odorata (3.67%) but limited for the other species and in particular for Dipterocarpus alatus (0.48%), which is found in only four plots. Hopea odorata and Anisoptera costata are less represented, respectively 1% and 0.6% of the trees and 0.4% and 0.48% of the regeneration. Of these two species, trees are present in 18 and 17 plots, respectively, but the regeneration occurs only in 9 and 6 plots. The last two Dipterocarpaceae species, namely Shorea guiso and S. thorelii, are rare (0.27%) and their regeneration is nearly absent (0.02% observed in just one single plot). The Fabaceae species are also barely represented. Sindora cochinchinensis alone is better represented with 0.6% of the adults (in 15 plots) and 0.48 of the regeneration (in 10 plots). Afzelia xylocarpa (0.05% of the total number of trees), Dalbergia bariensis (0.02%), and Xylia xylocarpa (0.29%) are rare. The regeneration is absent for A. xylocarpa and nearly absent for X. xylocarpa (0.03%) but is non-negligible for D. bariensis (0.07%) in comparison with the adult's occurrence.
Discussion
Floristic structure after logging
In Tan Phu, six groups of forest can be identified, and the Jaccard indices range from 40% to 56%. In the neighboring protected area of Cat Tien National Park, similarity indices range from 0 to 37% [19]. These indices are comprised of between 36.13% and 40.25% in a logged forest in Ha Tinh province in central Vietnam [20]. Forest types in Tan Phu share a large number of species; these are more homogeneous than other protected or logged forests in Vietnam. Some species, like Acronychia pedunculata, Buchanania arborescens, Carallia brachiata, Careya arborea, Combretum parvifolium, Diospyros maritima, Elaeocarpus tectorius, Garcinia ferrea, Grewia tomentosum, Irvingia malayana, Parinari annamensis, Syzygium chanlos, S. grandis, Vitex tripinnata, and Xylopia viaelana, occur in 80% of the plots in Tan Phu forest.
Groups 1a, 1b, and 2a could be classified as Dipterocarpaceae forests. The percentage of Dipterocarpaceae trees varied from 21.8% to 13.54%, and these correspond to different levels of species arrangements after logging. Okuda et al. [21] have reported a total density percentage of 9.2% Dipterocarpaceae in a primary forest in Malaysia. Tran et al. [20] have reported a maximum of 15.6% Dipterocarpaceae trees in some forests in central Vietnam. Compared with other forests in Vietnam and even in the region, Tan Phu forest can be considered as an important pool of Dipterocarpaceae. Groups 2b and 3a correspond to a Lythraceae or Myrtaceae forest mixed with Sapindaceae, Ebenaceae, and Dipterocarpaceae species. These are degraded forests in which the dominant species moved from Dipterocarpaceae and Fabaceae to Myrtaceae or Lythraceae families after logging. Group 3b could be classified as Guttifereae-Rutaceae-Tiliaceae forest and corresponds to the extreme degradation of both Dipterocarpaceae and Fabaceae forests (figure 6).
Biodiversity value of Tan Phu forest and in comparison with Cat Tien National Park
The observed species richness ranges from 86 to 89 within 1.5 ha in a Dipterocarpaceae-Myrtaceae forest (1a) and in a Dipterocarpaceae-Ebenaceae forest (2b), respectively. It is equivalent to 85 within 2 ha of a Myrtaceae-Ebenaceae-Dipterocarpaceae forest (1b) and it reaches 107 within 4 ha of a Myrtaceae-Dipterocarpaceae forest (2a). By comparison, only 18 species occur in a 1 ha plot of Dipterocarpaceae forest in Cat Tien National Park. Dipterocarpaceae forests are much richer in Tan Phu than in Cat Tien National Park. Anyway, considering that the number of tree species of 10 cm dbh is within the range of 100–150 species typically recorded in Southeast Asian lowland rainforests [22], the observed species richness is low both in the logged and protected forests. Observed species richness is 73 within a 1 ha Lythraceae-Sapindaceae-Ebenaceae forest (3a) in Tan Phu. By comparison, Blanc et al. [19] have reported 57 to 91 species in 1 ha plots of Lythraceae-Ebenaceae forest in Cat Tien National Park (figure 1); there is no difference between logged and protected forest. In Tan Phu, the highest species richness is observed on 2.5 ha of a Guttifereae-Rutaceae-Tiliaceae forest (3b) where it is equivalent to 111 species. In this latest forest type, the species are mostly pioneer species like Cratoxylon formosum, Acronychia pedunculata, Grewia paniculata, and Vitex tripinnata. In Tan Phu, degraded forests composed of pioneer species and having a low rate of Dipterocarpaceae have higher species richness. In total, the observed species richness reaches 176 on 12.5 ha in Tan Phu and the estimated species richness is comprised of between 205 and 207. In Cat Tien National Park, the same number of species has been recorded in only 5 ha. The total species richness can be considered higher in Cat Tien than in Tan Phu forest. For Chittibabu and Parthasarathy [23] and Zhu et al. [24], although species richness could be higher in some fragments, the species diversity is generally lower in the logged forests than in the protected forest. In Tan Phu, the forest types share a large number of common species, but some important species, like Heritiera cochinchinensis and Mesua ferrea, may have already disappeared.
The diversity ranges from 0.94 to 0.97 and from 5.00 to 5.67, respectively, for the Simpson and Shannon indices in Dipterocarpaceae forests (1a, 1b, 2a). In the neighboring protected forest, the values are 0.51 and 1.98. The diversity of the Dipterocarpaceae forest in Tan Phu is the result of logging. The diversity increases due to both the reduction of native species and to the appearance of pioneer species. The Lythraceae-Sapindaceae-Ebenaceae forest (3a) has the lowest value, equivalent to 0.93 and 4.73, respectively, for the Simpson and Shannon indices. Blanc et al. [19] found values ranging from 0.84 to 0.96 and 4.15 to 5.62, respectively, for the same indices in a similar forest type in Cat Tien National Park. Diversity of Lythraceae forest could be considered equivalent in both logged and protected forest. Finally, in heavily degraded forests (3b), the diversity could be linked to the invasion of some pioneer species and to the reduction of native species. The Simpson and Shannon indices are equivalent to 0.95 and 5.27, respectively. Cumulatively, the diversity values are 0.98 and 6.02 respectively for the Simpson and Shannon indices. The diversity observed in Tan Phu is high and could be linked to the decreasing abundance of native species and to an increasing abundance of pioneer species.
Distribution and regeneration of the conservation species
The Dipterocarpaceae family is the most important one and represents 11.7% of the total number of trees. This family is dominant in two of the six floristic groups and represents 21.85% of the trees. Eight Dipterocarpaceae species, found in Tan Phu, are considered as conservation species. In comparison, Blanc et al. [19] have inventoried only five species of Dipterocarpaceae, which represents less than 10% of the abundance in the Cat Tien National Park. Dipterocarpaceae could also account for 15.6% in some forests in Central Vietnam [20], but it concerns only two species. Tan Phu can be considered as an important area for the conservation of Dipterocarpaceae. However, except for Vatica odorata and to a certain degree D. dyeri and S. roxburghii, the regeneration is limited. S. guiso and S. thorelii are even at risk of extinction in Tan Phu.
Fabaceae species represent only 1% of the abundance in Tan Phu and account for four conservation species. By comparison, Blanc et al. [19] have found 12 species of Fabaceae in Cat Tien National Park. Except for Sindora cochinchinensis which has the highest occurrence and represents 0.6% of the adult trees, the three other species, namely Afzelia xylocarpa, Dalbergia bariensis, and Xylia xylocarpa, are rare and their regeneration is limited. These three species are close to extinction.
Implications for conservation
In the study area, logging has produced a gradient of degradation of the original Dipterocarpaceae and Fabaceae forests. This leads to six actual different floristic compositions in which the abundance of native species has decreased to the benefit of pioneer species. As a result, Dipterocarpaceae species are still dominant in two forest types but are progressively less abundant in the four others, while Fabaceae species have nearly disappeared. Overall, the logged forest of Tan Phu presents a low species richness when compared with the neighboring protected forest of Cat Tien, but the floristic groups observed often have high species richness and diversity values due to the appearance of many pioneer species. The invasion of pioneer species also creates a high percentage of similarity among the groups. From a conservation point of view, the studied forest stands continue to enclose important conservation species, particularly in the Dipterocarpaceae family. Some of these species are abundant and continue to regenerate (for instance, Dipterocarpus dyeri, Shorea roxburghii, and Vatica odorata), but some others, like Anisoptera costata, D. alatus, Hopea odorata, Shorea guiso, and S. thorelii, are scattered and few regenerations are observed. The four Fabaceae species, which are Afzelia xylocarpa, Dalbergia bariensis, Sindora cochinchinensis, and Xylia xylocarpa, are rare and, except for S. cochinchinensis, are close to extinction. In this regard, some species like Heritiera cochinchinensis and Mesua ferrea observed in the past have not been identified in this study and may have disappeared. The logged forest of Tan Phu still has a high conservation value, but urgent action should be undertaken to preserve the conservation species. Two Dipterocarpaceae forests (1b & 2a) identified in the study should be protected in order to preserve most of the Dipterocarpaceae species. The conservation species closest to extinction should be considered under afforestation programs.
Acknowledgments
This study was carried out within the framework of the decentralized cooperation between Dong Nai Province in Vietnam and Rhône Alpes Région in France. The authors are grateful to the authorities of both regions as well as representatives of the Tan Phu forest enterprise for providing them with the opportunity to study. The authors also thank students from the University of Sciences in Ho Chi Minh City and Mr. Dau and Mr. Thanh from Tan Phu for field assistance.
