Plant Species Composition and Their Conspecific Association in Natural Tropical Rainforest , South Papua

Papua has so many wide lowland areas that cover high diversity in plant from lifeforms. In contrast, there is a lack of information concerning species diversity. The objectives of the research were to describe species of plant life-forms and to analyze conspecific associations between small individuals and large individuals in natural tropical rainforest, south Papua. Then, 46 nested plots were placed systematically in natural tropical rainforest of Boven Digoel Regency, Papua. Density, frequency, dominance and importance value index (IVI) were analyzed to describe plant diversity, while detrended correspondence analysis (DCA) was to describe conspecific association between small individuals and large individuals. Results found 2040 individuals in this forest consisting 194 plant species. Those are from pteridophytes, angiosperms and gymnosperms. The plant life-forms are herbs, orchids, palms, pandans, shrubs, rattans, vines and trees in which of the plant life-forms, trees are the most abundant according to number of species. In the dominant plant species, two types of conspecific association occurred between small individuals and large individuals. Analysis of DCA revealed that there are large plant species grow along with their small individuals. Another type is that small individuals grow far from their large individuals.

researches have been conducted to reveal diversity based on plant species (Kuswandi, Sadono, Supriyatno, & Marsono, 2015).Some plants species have not completely identified.Therefore, some plant species have recently classified taxonomically e.g.(Heatubun, Zona, & Baker, 2014;Hughes, Barber, Heatubun, & Gagul, 2015;Maturbongs, Dransfield, & Mogea, 2015).However, there are still lack of information regarding abundance, distribution and association of the plant species that grow in natural forest.Nowadays, there are many threats, one of which is conversion of forest area to other land uses resulting in habitat loss.Hence, it is necessary now to find out related to the species diversity in natural tropical rainforest.
We formulated research questions as follows: 1.What kind of species composition based on plant life-form?and what species were dominant among them?; 2. Whether there are any association between conspecific small and large individuals of dominant species?.Thus, the aims of this research were to identify species of each plant life-forms and to describe conspecific associations between small and large individuals in natural tropical rainforest, south Papua

Study area
Location took place in natural tropical rainforest of Boven Digoel Regency, Papua.The location is situated in 6 o 26'0''S -6 o 30'0''S and 140 o 52'30''E-140 o 57'0'' Ebetween Muyu and Uwim Merah River in the western part and Fly River in eastern part near border of Indonesia and Papua New Guinea, while in the northern part is mountainous area of Papua and in the southern part is ex-timber concession (Figure 1).Annual mean of rainfall range from 3000 to 4000 mm and monthly rainfall is characterized as high which range from 221 to 426.This location
Papua is the western part of New Guinea Island that comprises high diversity of plants.In general, the plant can be grouped as forest, woodland, scrub, savanna, grassland, mixed herbaceous vegetation, pioneer vegetation, mangrove vegetation and garden (Petocz, 1989).Lowland in Papua is a wide area that cover high diversity in plant from life-form (Paijmans, 1970).Many

Sampling
Total sampling plots in this research were 46 plots which were sampled using systematic plots.Plots were placed along transect and distance between plots was 100 m.First, plots with size set 20 m x 20 m (0.04 ha) were set for large trees and large individuals that are typified to have diameter at breast height (dbh) greater than 20 cm.Plots with size 10 m x 10 m (0.01 ha) were set for poles in which individuals are characterized to have diameter between 10 and 20 cm.Plots with size 5 m x 5 m (0.0025 ha) were set for saplings where individuals are characterized as height greater than 1.5 m and diameter less than 10 cm.Plots with size 2 m x 2 m (0.0004 ha) for seedlings where individuals are typified to have height less than 1.5 m.Then, the plots were designed and laid as nested samples (Forestry Department, 1989). .

Data collection
Plant species of pteridophytes, angiosperms and gymnosperms were taken in plots and every individual was identified according to the scientific name.Species identification was performed by two vegetation identifiers from herbarium technician.Unidentified samples were set as voucher specimens and sent to Herbarium of "Balai Penelitian Kehutanan Manokwari" and Herbarium Manokwariense (MAN) Pusat Penelitian Keanekaragaman Hayati Universitas Papua (PPKH-UNIPA) to be identified.Then, the vali-dation of international scientific name of plant species and family was checked through online at http://www.theplantlist.org/;http://plants.jstor.organd www.ipni.org/ipni/.Numbers of each species in plot were documented, whereas the diameter at breast height (1.3 m) or 20 cm above buttress were tallied for poles and trees except for seedling and saplings phase (Bertault & Sist, 1997).

Data analysis
Diversity index.Shannon-Weiner diversity index was singled out as parameter and to describe distribution of each species in term of number of individuals by computing evenness (E) (Smith & Wilson, 1996;Spellerberg & Fedor, 2003).The diversity index is calculated as , where is Shannon-Weiner diversity index, is num-is Shannon-Weiner diversity index, is num-is Shannon-Weiner diversity index, is number of sample in which species i is present.Evenness was measured using in which S is number of species.Basal area (BA).Basal area was only performed for tree.Basal area (BA) will be computed by taking into account diameter of tree species.The computation of BA is as , where BA i is basal area (m 2 ) of tree species i, D i is diameter (m) of tree species i, 0.7854 is π divided by 4.Then, BA per hectare where BA of tree species is divided by area of plots (m 2 ha - 1 ) as density.The BA for each tree species is to describe how large the tree species dominate in location (Folega et al., 2014).Density.Density is then used to describe number of each species per hectare (ind ha -1 ).Density was carried out for all plant life-forms.Frequency.All species of plant life-forms was described using frequency.Furthermore, number of plots where tree species i is present are divided by total number of sample plots.Hence, the frequency is calculated as where Fr i is frequency of species i, n i is number of plots in which species i is found, N is total number of sample plots.Importance value index (IVI).The important values index was performed only for trees and was calculated to figure out distribution of each tree species in term of dominance (Cottam & Curtis, 1956).The index is determined by adding relative frequency, relative density and relative dominance as where IV i is important value index of tree species i, RFr i is Relative Frequency of tree species i, RDe i is Relative Density of tree species i, RDo i is Relative Dominance of tree species i. Detrended correspondence analysis (DCA).DCA is a multivariate statistical technique DCA (Hill & Gauch, 1980) and was applied to describe conspecific association between small individuals and large individuals.

Taxonomic diversity and abundance of species
The results found 2040 individuals in this forest consisting 194 plant species.Then, those species were grouped plant life-form as ferns (7 species), herbs (11 species), orchids (3 species), palms (6 species), pandan (1 species), rattans (3 species), shrubs (26 species), trees (126 species) and vines (11 species).Of number of individuals in each plant life-form, tree is the most abundant in concession forest (65.75 %), while the least scarce plant life-form is pandan (0.98 %).Other plant life-form such as fern, herb, orchid, palm, rattan, shrub and vine are between about 2 % and 8 % distributed in this area as can be seen in Table 1.

Rattans
Rattans from family Arecaceae covers 3 three species which are mentioned based on descending order in density as Calamus sp. with density 3030 ind ha -1 , Calamus keyensis Becc.with density 172 ind ha -1 and Korthalsia sp. with density 54 ind ha -1 (Table 7).

Seedlings
There are 78 species of tree categorized as seedlings and then the 10-dominant species of trees are displayed based on Importance Value Index (IVI).The density of the dominant species is above 800 ind ha -1 (Table 9).Of the species of seedlings, Hopea celtidifolia Kosterm. is the densest species growing in this forest with density almost 5000 ind ha -1 , where Vatica rassak Blume and Hopea iriana Slooten showed density between 2000 ind ha -1 and 3000 ind ha -1 .Ten species of dominant species were singled out and displayed in Table 9.

Saplings
Saplings contain 86 species in total and the 10-most dominant sapling of tree have been singled out based on IVI (Table 10).Density of tree saplings are greater than 100 ind ha -1 except for Timonius rufescens (Miq.)Boerl.where its density is only 78 ind ha -1 .The highest density of saplings is Syzygium anomalum Lauterb. in which number of individuals per hectare is almost 300 (Table 10).

Poles
There are 60 species of treeclassified as pole.Density of 10-most dominant poles range from 11 ind ha -1 to 37 ind ha -1 with dominance basal area of its pole per hectare is between 0.18 m 2 ha -1 and 0.67 m 2 ha -1 .Gymnacranthera farquhariana (Hook & Thomson) Warb.and Hopea celtidifolia Kosterm.are the two-densest species of poles in which number of individuals and basal area per hectare are above 30 and 0.43 respectively (Table 11).

Large Trees
Species recorded in large trees are 76 species in which of the 10-most dominant species of large trees (Table 12), their density cover from 5 ind ha -1 to 23 ind ha -1 along with a range of dominance is from 0.54 m 2 ha -1 to 2.33 m 2 ha -1 .The high density of large tree is Vatica rassak Blume with about 20 ind ha -1 and its basal area is 2.33 m 2 ha -1 , whereas Syzygium anomalum Lauterb., Planchonella anteridifera (White & Francis ex Lane-Poole) H. J. Lam, Gironniera subaequalis Planch.
and Hopea celtidifolia Kosterm.showed their number of individuals per hectare at least above 10 (Table 12).

Distribution and dominance of species of lifeform in natural forest
The distribution of vegetation species, regardless of life-forms, is seemingly uniform since the average evenness (E) is 0.9 and its standard deviation is 0.09 for 46 plots.It suggests that most of 194 plant species, there are not species that is mainly dominant in natural forest.Another indication is that natural forest is species-mixed forest (Katovai, Katovai, Mesquita et al., 2015) and species-rich area.Even though we did not take into account the Bryophytes (Mosses and liverworts), there are ferns, herbs, orchids, palms, pandans, rattans, shrubs, trees and vines as categories of life-form of plants that showed high number of species.The plant diversity can be described from understory to upper story in which they have grown symbiotically as natural forest (Givnish, 1999).It can be said that forest condition is in successional climax (Swaine et al., 1988).
In understory, the most dominant fern is Selaginella sp.(Selaginellaceae) which were found in about 25 % of natural forest (frequency is 0.24), while other ferns were only found in less than one per cent of this forest.This species is small ferns grow mostly in group (Kessler, Salazar, Homeier, & Kluge, 2014;Sharpe & Shiels, 2014;Tryon, 1971;Watkins & Cardelús, 2009).In herbs, the most dominant species is Mapania sumatrana (Cyperaceae) in which they are distributed in 22 % of this forest (frequency is 0.22).The species also grow more than 1000 ind ha -1 (Table 3).This herb is morphologically like pandan.The second-most dominant species is Alpinia sp.(Zingiberaceae) that are present in 13 % of this forest (frequency is 0.13).The species is typical of understory plants in tropical forest and grow as perennial herbs (Cicuzza et al., 2013;Folega et al., 2014).Other species herbs were only found lower than one per cent of this forest.Orchids found in natural forest were terrestrial plants in which Corymborkis veratrifolia is the most dominant species with distribution 15 % of this forest (frequency is 0.15), whereas the two-other species showed the distribution less than one per cent in this forest.Those ferns, herbs and orchids have short time to reach reproductive stage, resulting in producing new individuals.Hence, those life-forms can grow abundantly in understory of natural forest compared to other species of their life-form.Species of palms recorded in this forest vary in size that is from small plants to large plants.Espe-cially, Hydriastele beccarianais the most dominant species of palm in small plant.This species is distributed about 30 % of this forest (frequency is 0.35).This species grow either as solitary palm or assembly palm (Baker & Loo, 2004).Regeneration of Hydriastele beccariana can be from germination of its seed (Wang & Augspurger, 2006) produced by mature individual and can also reproduce new individuals through suckering.The suckers can grow from small individuals (Frangi & Lugo, 1998).This is as result of how this palm can grow more abundant than some palms.There is only one species of pandan found in this forest, namely Pandanus sp. that is distributed only more or less one per cent of this forest (frequency is 0.11).Of the tree rattan species, Calamus sp. is the most dominant rattan.In contrast, this species was only found rarely in only maximum about one per cent of this forest.This species is not distributed widely, but this species tend to grow in assemblage as the density about 3000 ind ha -1 .The   stem of rattans can grow in any direction with hundreds of meters long as long as the stem can reach sunlight by using leaf-climbers as morphological support (Tomlinson et al., 2001) There are many types of vines, one of which is Derris ellipticawhere this species is the most dominant vine categorized as small plant (H.Zhu, 2008).This species was found in 30 % of this forest (frequency is 0.3).The density is more or less 1000 ind ha -1 and mostly this species is as creeping plant rather than climbing plant.Two other species, namely Freycinetia sp. and Flagellaria indica were found only in more or less one per cent of this forest.The rest species of vine were only found less than one per cent of this forest.In trees, the most dominant species of seedling is Hopea celtidifolia (IVI=12.23 %).

Conspecific association between small individuals and large individuals among dominant plant species
There are 19 species selected as dominant species and each species consists of small individuals as juveniles and large individuals as mature individuals.Then, those were analysed using of DCA are presented in Figure 2 and Table 13.Total inertia of DCA is 4.5 where eigen values for first axes and second axes are 0.71 and 0.37 respectively.Percentage variance of species data are 15.8 % (axes 1) and 8.28 % (axes 2), then cumulated percentage of the-two axes is 24.07 % to explain variation of the species distributed in graph of DCA.Based on analysis (Figure 1), association between small and large plant for each dominant species can be grouped as three groups.There are two kinds of pattern of association between small and large plants in tropical rainforest.The association describe an effect of neighbourhood whether same species or different species (Giles et al., 2004).First, small and large plants grow close together in an area where seedlings and saplings of plants are commonly found near their large plants.It indicates that seeds are produced and are dropped by large trees (Mesquita et al., 2015;Zambrano, Coates, & Howe, 2014).Those seeds mainly germinate around large trees as their parent trees.In this case, we found that Vatica rassak, Hydriastele bec-cariana, Syzygium anomalum, Canarium asperum, Gymnacranthera farquhariana and Prunus arborea as association of first group and another second group are Melicope elleryana, Hopea celtidifolia and Gironniera subaequalis.Seeds of those species are observably not as feed of animals.Therefore, the seeds are not dispersed by animals.Besides that, the seed size like Vatica rassak is big, resulting in the seeds are fallen directly below parent trees (Y.Zhu et al., 2015).During research, we also observed that the seeds of those species were near their large trees.
Secondly, there are also group of plants in which small plants are not associated with large plants.The small plants tend to grow far from their large plants such as found in first group, second group and third group.Species such as-Hopea iriana, Rhodomyrtus sp., Blumeodendron tokbrai, Calophyllum caudatum, Orania disticha, Parastemon versteeghii, Pimelodendron amboinicum, Planchonella anteridifera, Pometiaacuminate, Virola surinamensisand Gironniera subaequalis showed the distance between small and large plants, showing a negative dependence to parent trees (Pérez-Méndez et al., 2015).Those small individuals are not beneath their large trees.For that reason, the small plants were distributed separately with large plants as shown in DCA result (Figure 1).In the field of tropical forest, the small plants rarely grow below their large plants as parent trees, or in some cases the small individuals are absent (Baldeck et al., 2013).We also noted that it is also very infrequent to see fruits or seeds of those species near their large plants.That is probably because of that the presence of animals such as birds, bats, and rodents feeding the fruits.Other study showed that the presence of rodents in tropical forest will contribute in recruitments of plants and its distribution (Velho, Isvaran, & Datta, 2012).Then, the animals play a crucial role in seed dispersals of those species (Chapman & Chapman, 1995).Besides that, the seed size is small, bringing about the seeds are easier to be dispersed by the animals (Seidler & Plotkin, 2006;Theimer, Gehring, Green, Connell, & Theimer, 2016).Probably, there are other reasons to explain the phenomena like allelopathic potential in inhibiting growth of other plants including their young individuals (Cummings, Parker, & Gilbert, 2012;Ladwig, Meiners, Pisula, & Lang, 2013;Silva Matos & Belinato, 2010;Zhang & Fu, 2009), but we did not observe during research.Hence, more details of discussion are not presented in this paper.
Finally, results of this study give a contribution of plant diversity in natural tropical rain forest which from pteridophytes, angiosperms and gymnosperms.The densities of each species can be used as population condition in natural rainforest.Those information of plant diversity can be specifically used as future researches on how local people utilize them such as medicinal plants e.g.(Abdullah & Mustikaningtyas, 2010; Irawan, Fitmawati, & Herman, Therefore, conservation and management programs can be supported by using this information.Moreover, there are number of species have not been identified where in this research species were taxonomically named as "sp.".This condition reveals that future researches especially regarding taxonomic topics are needed to uncover plant biodiversity taxonomically in natural tropical forest.
The conspecific association has discovered that in forest the individuals are not consistently present and grow along with the large individuals.In some cases, small individuals as juvenile phase are missing even though the mature individuals are present.

CONCLUSIONS
There are species found from pteri-pteridophytes, angiosperms and gymnosperms that were grouped as herbs, orchids, palms, pandans, shrubs, rattans, vines and trees.Trees are the most abundant according to number of species.In the, dominant plant species, two types of conspecific association occurred between small individuals and large individuals.Analysis of DCA revealed that there are large plant species grow along with their small individuals like Vatica rassak, Hydriastele beccariana, Syzygium anomalum, Canarium asperum, Gymnacranthera farquhariana and Prunus arborea.Another type is that small individuals grow far from their large individuals such as Blumeodendron tokbrai, Calophyllum caudatum, Orania disticha, Parastemon versteeghii, Pimelodendron amboinicum, Planchonella anteridifera, Pometia acuminata and Virola surinamensis.

ACKNOWLEDGMENTS
This study was financially supported by Beasiswa Pendidikan Pascasarjana Dalam Negeri (BPP-DN) 2014 and Manokwari Regency scholarships.We thank Daud Bano, Dedy Subrata May for accommodation support; Vinoba Chandra, Ari Wibowo, Anton Bongga and Sri Emy Rahayu and all staff in PT Tunas Timber Lestari for data and field assistance; and Nithanel M. H. Benu (BPK Manokwari), Dr. Purnomo (Faculty of Biology, Gadjah Mada University) for checking scientific name of plants.We would be grateful for anonymous reviewers for valuable comments.

Figure 1 .
Figure 1.Location of research

Figure 2 .
Figure 2. Detrended correspondence analysis (DCA) to analyze association of small plants and large plants of each species.Letter before dash (-) S stands for small plant and L stands for large plant.Distribution of individuals in the graph show association where if they are closer, meaning that the associations are strong and vice versa.

Table 1 .
Density of individual based on plant lifeform

Table 2 .
Families and species density of fern

Table 3 .
Families and species density of herbs

Table 4 .
Families and species density of orchids

Table 5 .
Families and species density of palms

Table 7 .
Families and species density of rattans

Table 8 .
Families and species density of vines

Table 9 .
The 10-dominant species of seedlings

Table 10 .
The 10-dominant species of saplings

Table 11 .
The 10-dominant species of poles

Table 12 .
The 10-dominant species of trees

Table 13 .
Summary of detrended correspondence analysis (DCA) First group consists of species with small and large plants such as Vatica rassak, Hydriastele beccariana, Syzygium anomalum, Canarium asperum, Gymnacranthera farquhariana and Prunus arborea; species with only small plants such as Hopea iriana and Rhodomyrtus sp.;