The Nauclea , Tristaniopsis , and Terminalia forests: Exploring the lesser-known biotopes of Agusan Marsh, Philippines

. The peat swamp forest of Agusan Marsh is characterized by the dominance of a unique and conspicuous assemblage of tree species across different habitat types. Understanding the species richness, diversity, structural patterns, and distribution of these types of forests is crucial for maintaining ecosystem stability and conservation. This study describes three less-explored biotopes dominated by Bangkal ( Nauclea orientalis L.), Dangilas ( Tristaniopsis decorticata Merr.), and Lanipau ( Terminalia copelandii Elmer), using a purposive random sampling method. Across the three sites, 22 tree species, belonging to 17 families and 21 genera, were recorded. Tree density was recorded as the maximum at the Tristaniopsis biotope (650 individuals ha -1 ) and the minimum at the Terminalia biotope (367 individuals ha -1 ). The maximum total basal area (162.93 m2 ha -1 ) was re-recorded in the Tristaniopsis biotope and the minimum area in the Nauclea biotope stand (17.3 m2 ha -1 ). The ANOVA results for stand DBH, tree density, species richness, species diversity, and species evenness varied significantly among biotope types. The Tristaniopsis biotope had the highest species richness and diversity among the three sites. All stands exhibited a high community dependence on forest resources, which may have increased the vulnerability of the forest to human interference, thus necessitating immediate attention to conservation. Site-specific management and conservation measures are proposed to provide opportunities for sustainable forest management.


Introduction
The Philippines is one of the megadiverse countries in the world [1,2] because of its archipelagic nature, consisting of 7,100 islands.The country has many endemic species of plants, birds, mammals, reptiles, amphibians, and invertebrates [3].However, owing to the exponentially growing human population, the Philippines has also been identified as a biodiversity hotspot [4].Verdant rainforests, which harbor a vast diversity of plants and animals, have been cut into timber products [5].Many natural habitats, such as wetlands and Corresponding author: rtsarmiento@carsu.edu.phmangrove areas, have recently been cleared for farming and housing to accommodate the nation's growing population [6,7].This unsound management of natural resources is the primary driver of the Philippines' decline in biodiversity and ecosystem integrity [8].Agusan Marsh, located in the southern part of the Philippines, is a natural wetland that supports an extensive array of life forms [9].Its main habitats include freshwater swamp forests, secondary scrubs, herbaceous swamps, lakes, pools, rivers, and floodplains, which have been converted into rice fields for aquaculture or crops [10].As a key biodiversity area (KBA) in the Philippines, the marsh was declared a protected area under the Philippines National Integrated Protected Area System (NIPAS) [11].1999, it was included in the Ramsar List of the Wetlands of International Importance [12,13].The marsh was recognized as an important wintering ground for migratory birds, as thousands of birds from Japan, China, and Russia escaped the winter winds of the region [14].Over 200 species spend at least part of the year in wetlands, making it one of Asia's most significant transit points for wild birds.[10].
This study assessed the species diversity, composition, and distribution patterns in the three less-explored habitat types of the marsh dominated by Nauclea orientalis L., Tristaniopsis decorticata (Merr.)Peter G.Wilson & J.T.Waterh, and Terminalia copelandii Elmer.Despite being recognized as an important wetland in the Philippines, vegetation surveys undertaken in marshes are limited.This has resulted in a few published reports on marsh biodiversity, including those of Davies [15], Aribal and Fernando [16], Tandang et al. [17], Sarmiento and Garcia [18], and Wetland Link International [19].Because these studies were limited by their geographical scope and timescale, more extensive studies on vegetation community composition and species distribution are required for practical conservation management recommendations.

2
Materials and methods

The study sites
The Agusan Marsh is part of the Agusan River Basin and occupies the middle portion.The Agusan River Basin is the third largest river basin in the Philippines, with a total drainage area of 10,921 km² and an estimated length of 350 km from the headwaters to Butuan Bay.It is located in the eastern part of Mindanao, which draws off most of the water in the Caraga Region and some parts of Compostela Valley Province (Fig. 1).The marsh is geographically located at the heart of Agusan, comprising Bunawan, Loreto, Rosario, San Francisco, Talacogon, and La Paz in Agusan del Sur within coordinates N08.12411° to 08.41346° and E125.81000° to 125.96412°.

The Nauclea Forest
N. orientalis dominates this biotope type.This species, commonly known as Bangkal or Kabak in Manobo, is a medium-to large-sized tree that belongs to the Naucleaceae family.It can grow as tall as 30 m and thrives in secondary forests at low-to-medium elevations.This species is typically found in waterlogged areas and can tolerate prolonged submergences.
The Nauclea Forest biotope is widespread in marshes, especially in Bunawan.It is found along the Agusan River, going to the Bunawan Wharf.This species was also found in the undisturbed areas of Sitio Panlabuhan in Loreto, Agusan del Sur.Some trees show stunted growth and inferior development owing to prolonged inundation.Residents say that the area experiences periodic flooding, but water levels eventually subside during the dry season.However, the forest area becomes flooded year-round, with an influx of upstream sediments [10].

The Tristaniopsis Forest
T. decorticata (Merr.)Peter G.Wilson & J.T.Waterh dominates the medium-pole forest in the peat swamps of the Agusan Marsh.The Dangilas forest, as locals call it, is a unique biotope found only in the pristine areas of Lake Kasawangan, Talacogon, in the Agusan del Sur.The associated species included Calophyllum inophyllum, Arthrophyllum borneense, and Syzygium merrittianum.The underground cover comprised dense Pandanus spp., Exocarpus longifolius, saplings of Cinnamomum mercadoi, and some fern species.The tree can grow straight and rounded boles of up to 30 m in height.

The Terminalia Forest
Recent surveys have indicated a significant decline in tree density and area in the "Lanipau" forest.The biotope type was dominated by Lanipau (T.copelandii) and mainly found in periodically inundated marsh areas.Forest type remains extensive in Rosario and Loreto.However, in La Paz, Talacogon, and Bunawan, few trees exist in small patches owing to extensive land conversion and tree cutting.The local tribal community, Manobo, used this species to build boats and floaters in their houses during the flood season.Approximately six-eight boles were firmly fixed and secured for a steady float in a single house.According to the residents, Lanipau Forest is rich in nutrients and suitable for rice production.In Rosario, vast areas are cleared and burned for agricultural purposes.

Sampling design
Phytosociological analysis was performed from June to December.The study sites were preselected based on the prevailing vegetative cover identified by the local people.We randomly placed three 0.1 ha (18 m radius) circular plots at least 100 m apart to gather quantitative data for each forest type.The DBH of all trees >10 cm was recorded using standard inventory procedures [20,21,22].

Data collection and species identification
Before conducting the field survey, local guides from each municipality were identified through a series of workshops.All tree species were initially identified using local names in the field.The scientific and common names were recorded using literature and taxonomic identification keys, following the latest Angiosperm Phylogeny Group classification version 14 [23] and Rojo [24].Publications referring to Philippine flora include de Guzman et al. [25], Madulid [26], Castro [27], Santos et al. [28], and online identification guides such as Co's Digital Flora of the Philippines [29] and World Flora Online [30], were also utilized to compare plant samples.
For trees not identified in the field, fertile samples were collected in triplicates for further species identification.The collected samples were sent to research collaborators and a reference collection was maintained at the Caraga State University Herbarium for further taxonomic work.Only a limited number of specimens were collected, collated, dried, and pressed using standard field methods [31].

Data Analysis
Species richness, dominance (D), Simpson (1-D), Shannon-Weiner (H'), and evenness (E) were determined using the Paleontological Statistics Software V4.03 [32].In addition, multivariate cluster analysis using Brays-Cluster Analysis was performed to compare the similarities and dissimilarities between habitat types.An ANOVA was also performed to compare the different vegetation parameters of the three vegetation types.Species abundance, dominance, density, and their relative values were computed using a spreadsheet [33], and the relative values were used to obtain the Importance Value Index (IVI) for each species [34].The ratio of abundance to frequency for different species was computed to obtain the distribution patterns of different tree species.This ratio indicates regular (<0.025), random (0.025-0.050), and contagious (>0.050) distributions [35,36].

Site characteristics, species richness, diversity, and evenness
Table 1 lists the stand characteristics for the three biotope types.Species diversity (H') ranged between 0.955 and 1.274, with the maximum recorded in the Dangilas forest.The evenness of the species ranged from 0.275 to 0.866, with maximum evenness recorded in the Bangkal stand.
The Bangkal forest comprises only three tree species and dense ground cover of Pandanus spp.and Saccharum spp.It was the least diverse habitat of the sites, with a Shannon Index value (H') of 0.955 due to the high dominance of N. orientalis, comprising 60% of the biotope.Barringtonia acutangula and Mitragyna rotundifolia accounted for the remaining 40% each.Epiphytes, such as Drynaria spp., Stenochlaena palustris, and lianas, were also abundant in trees, especially in slightly disturbed areas.Saplings of N. orientalis were used as the primary building material for altars during rituals and other cultural gatherings by the local communities.Although there was no recorded cultural preference for this species, it was used perhaps because of its abundance in the area.
In contrast, the Lanipau Forest had a Shannon diversity index of 0.983 and species evenness of 0.334.In this biotope, the tree species T. copelandii had the highest dominance, comprising 75% of the area, with the remaining 25% shared by the other seven species, thus having a low evenness value.Only eight tree species with 110 individuals were recorded during the survey, including Cerbera manghas, B. acutangula, M. rotundifolia, and Pterocarpus indicus.Local people used T. copelandii for boat-making and house floaters but seldom for lumber because of their low economic value.This may be one of the primary reasons why large-diameter trees still exist in the marshlands.

Stand density and basal cover
The average tree density ranged from 367 to 650 individuals ha -1 , with the highest density in the Dangilas forest, followed by the Bangkal forest, and the lowest in the Lanipau forest.This disparity in tree density can be attributed to multiple drivers related to location, access, and resource use [37].For instance, the Lanipau Forest in Rosario is closer to settlements, and trees are felled and utilized to build dugout boats.Residents also accounted for the loss of extensive agricultural land conversion.Historical accounts show that the density of T. copelandii in the past was higher than that reported in a recent survey [38].However, the Dangilas stand is located in the remotest portions of the marsh, and is thus seldom visited and considered pristine.The importance value index was calculated based on the relative values of three major qualitative and quantitative parameters: Density, Dominance, and Frequency.As the IVI is calculated based on the above plant community characteristics, it represents the ecological adaptability of the species and its successful establishment in a particular forest/habitat type [39].The IVi values of the three species were high (Table 2), ranging between 53.86% and 57.89% in their biotopes.For example, T. decorticata had an IVi of 54.75% for all species within the Dangias biotope (Tristaniopsis forest); hence, a high IVi value represents good ecosystem adaptability, with a high rate of successful establishment.The Bangkal stand showed a clear dominance of N. orientalis (53.86%), the Lanipau stand of T. copelandii (57.89%), and the Dangilas stand of T. decorticata (54.75%).Therefore, all stands were identified using the names of the dominant species.Joshi et al. [20] cited that the dominant species is the most populous, found most conspicuously in a particular ecosystem, and generally comprises the maximum density and biomass in an ecosystem [22,40].

Species distribution
Across all sites, different species' abundance by frequency (A/F) ratio varied from 0.03 to 1.37 (Table 2).Most species were classified as having contagious patterns because the A/F ratios were commonly higher than 0.05 [35].However, a few species showed random distribution patterns, including M. rotundifolia, B. acutangula, Garcinia vidalii, Teijsmanniondendron ahernianum, Garcinia rubra, and Ternstroemia philippinensis.
According to Odum [41], as cited by Devi et al. [42], contagious distribution is the most typical pattern, whereas random distribution is found in a uniform environment.Neyman [43] first used the term contagious distribution for a discrete distribution that exhibits clumping, clustering, or contagious effects [44], suggesting that species herd protection occurs [45].Thus, the distribution pattern reflects niche specialization, which may simultaneously cause competition among species or the persistence of species in the ecosystem.

Family composition and size class distribution
Peat and freshwater swamp forests usually show lower global taxonomic diversity than other ecosystems [46].Of the total recorded plant families (17), Clusiaceae was the most dominant family with three species, followed by Fabaceae, Myrtaceae, and Rubiaceae, with two species each.The remaining 13 families were represented by a single species (Figure 2).The diversity of trees in peat and swamp forests was exceptional.The dominant families in the Agusan Marsh are similar to those in other Asian peatlands.Myrtaceae, Sapotaceae, Ebenaceae, Dipterocarpaceae, and Clusiaceae are the most dominant families in Indonesia [47,48].In contrast, Anacardiaceae, Bombacaceae, Burseraceae, Clusiaceae, Dipterocarpaceae, Euphorbiaceae, Leguminosae, Myristicaceae, Myrtaceae, and Sapotaceae are the most dominant families in Malaysia [49].
Fig. 2. Dominant plant families in the selected forest stand on the Agusan Marsh.
Across the three sites, the trees exhibited a higher density in the 0-20 and 21-40 cm diameter classes, showing dominance in young trees.The pattern of the size distribution of the basal area was a sigmoid curve with the highest values of individuals (170 stems ha-1) found in the DBH class of 21-40 cm and reduced as the DBH classes increased consistently until the 100-130 cm dbh class (Fig. 3).Although marshes have been declared wildlife sanctuaries and protected areas, rampant illegal logging activities continue to thrive, even inside the marsh [50,51], resulting in a decline in large trees.
Fig. 3. Girth class structure of trees across all biotope types in the Agusan Marsh.

Comparison of habitat types
Figure (Fig. 4) shows the Bray-Curtis similarity index as a percentage of similarity among the sampling sites.The index compares the relative abundance of a community across locations, where a value of 0 indicates total dissimilarity and 1 indicates total similarity [52,53].Three significant clusters were formed, with a high level of species similarity within a biotope and a significantly high dissimilarity between biotope types.Each cluster combination had low similarity (few species in common), implying that each cluster combination had unique floral characteristics and was distinct from other clusters in terms of species composition.The mean DBH, density, species richness, diversity index, and evenness were analyzed using single-factor ANOVA to determine whether the sites were significantly different from each other (Table 3).The values for all parameters at the three investigated sites varied significantly (p < 0.05) among the biotope types.These findings suggest that the habitat is still diverse, as shown by species richness and evenness of distribution.However, this information shows the need for immediate intervention to prevent a further decline in the diversity of these lesser-known biotope types.The Nauclea and Terminalia forests are among the essential habitat types identified in the Agusan Marsh.However, agricultural development has intruded the area, rendering wetland biodiversity vulnerable [54].Hence, protection and restoration plans are necessary to sustain these biotope types.

Conclusions
The assessment of species diversity, composition, and distribution patterns in the three lessexplored habitat types of the marsh found that species diversity (H') ranged between 0.955 and 1.274, with the maximum in the Dangilas forest.The evenness of the species ranged from 0.275-0.866,with the highest evenness observed in the Bangkal stand.Most species exhibited contagious distribution patterns.However, a few species exhibited random distribution patterns.Of the total recorded plant families, Clusiaceae was the most dominant family, with three species, followed by Fabaceae, Myrtaceae, and Rubiaceae, with two species each.Only one species represented the remaining 13 families.The trees in the marsh exhibited a higher density in the 0-20 and 21-40 cm diameter classes, showing the dominance of young trees, which can be associated with rampant logging activities despite the declaration of the wetland as a protected area.Therefore, the information generated from this study provides a valuable basis for action planning to sustain the biodiversity of wetlands and function as an ecological belt for carbon sequestration, prevention of soil erosion, and absorption of excess nutrients from interconnected agricultural areas to minimize eutrophication.

Fig. 1 .
Fig. 1.Location map of study areas in Agusan Marsh, Agusan del Sur, Philippines

Table 1 .
Characteristics of the different habitat types in Agusan Marsh computed using PAST Statistics.

Table 2 .
Computed relative dominance, relative frequency, relative density, and species importance value index of the three biotopes.

Table 3 .
ANOVA (F and p-values) of vegetation parameters among the study sites.