732 Agustinus Mardjoko (Recovery of Residual).cdr RECOVERY OF RESIDUAL FOREST ECOSYSTEM AS AN IMPACT OF SELECTIVE LOGGING IN SOUTH PAPUA: AN ECOLOGICAL APPROACH AGUSTINUS MURDJOKO , DJOKO MARSONO , RONGGO SADONO 1* 2 2 and SUWARNO HADISUSANTO3 1 Faculty of Forestry, Papua University, Manokwari 98314, Indonesia 2Faculty of Forestry, Gadjah Mada University, Yogyakarta 55281, Indonesia 3 aFaculty of Biology, Gadjah Mada University, Yogyakarta 55281, Indonesi Received 7 December 2016/Accepted 3 June 2017 ABSTRACT Papua has been experiencing heavy logging activity in its forests for decades. However, only several studies focused on the effect of logging in the forest ecosystem. This research was aimed to analyze recovery processes of the forest ecosystem. The research was conducted in the logged tropical rainforest in South Papua using ecological approach which used tree communities as biotic and soil condition as abiotic indicators. Data were collected in the logging area of PT Tunas Timber Lestari located in the tropical rainforest of South Papua. There were five groups of forests used in this research i.e. unlogged, one year post selectively-logged, five years post selectively-logged, ten years post selectively-logged and fifteen years post selectively-logged forests. Thirty nested plots were laid on each forest group. Canonical Correspondence Analysis (CCA) was applied to analyze the understory and upperstory plant communities. Understory and upperstory plant communities formed different patterns due to logging. Plant communities in the ten and fifteen years post-selectively logged forests were not similar to those in the unlogged forest. Soil organic matter (SOM) content in the selectively logged forests was lower than that in the unlogged forest. These occurrences indicated that the selectively logged forests were still recovering and required more than fifteen years to be fully recovered. Keywords: Canonical correspondence analysis, edaphic factor, logged tropical forest, plant community, soil organic matter INTRODUCTION Tropical rainforests play an important role in ecosystem services, such as logging production (Whitfeld et al. 2014; Putz & Romero 2014). The process of production mechanism in the tropical rainforest has a significant impact on abiotic and biotic elements (Zambrano et al. 2014). Those conditions result in the change in the tropical rainforest as an ecosystem and some circumstances of the secondary successional process take place as a response to ecological alterations. Furthermore, most of the tropical rainforests are experiencing the alterations and the selective logging has a significant impact on ecological factors (Corrià-Ainslie . 2015; et al Flores . 2014). Hence, the logged tropical et al rainforests are counting on the ability of forest recovery itself. Most indicators to analyse forest recovery are based on tree density, basal area (Whitfeld . 2014; Rutten . 2015) and et al et al growth rate of residual trees (Do . 2016; et al Hoang . 2011; West . 2014; Sist . et al et al et al 2014; Susanty . 2015) in the logged forests. et al However, the recovery of disturbed forests should not only be considered based on sustainable timber production, but the ecological elements such as soil conditions and residual trees should also be taken into account as forest recovery indicators. Some areas in lowland tropical forests in South Papua were intended as logging concession for decades (Kuswandi & Murdjoko 2015; Murdjoko *Corresponding author: agustinus.murdjoko.papua@gmail.com BIOTROPIA 4 3 7 230 245 Vol. 2 No. , 201 : - DOI: 10.11598/btb.201 .2 . .7 4 3 732 230 2013; Kuswandi 2014) . Few studies concerning the effects of logging in Papua logged forests were conducted. Some studies focused only on damages, changes in basal area (Gandhi & Mitlöhner 2014), population dynamics of remaining trees (Murdjoko 2013 ; Kuswandi & Murdjoko 2015; Murdjoko 2016b) and et al. biomass stock change (Hendri et al. 2012). Therefore, it is necessary to analyze forest recovery using the ecological approach in South Papua. In this analysis, the primary forest was considered as a stable forest ecosystem (Pennington et al. 2015). Ecological approach took tree communities as biotic factors where many processes such as tree associations, ecological responses of the tree to ecological change as well as successional development can be analyzed based on patterns of tree communities. Besides that, soil condition alters after selective logging (Hattori . 2013) et al mainly the amount of soil properties decrease such as Nitrogen content (Asase . 2014), soil et al organic matter (SOM) (Prasetyo . 2015) and et al other nutrients (Duah-Gyamfi . 2014; Wasrin et al & Putera 1999; Edwards . 2014; Imai . et al et al 2012). Consequently, the edaphic conditions were considered as abiotic indicators to support the explanation of the change in tree communities. This research was aimed to analyze recovery process of selectively logged tropical rainforest ecosystem in South Papua using ecological approach. Our hypotheses were: 1. tree communities in a selectively logged tropical rainforest were considered to be recovered when tree communities in the rainforest were similar to those in the primary forest; 2. the selectively logged tropical rainforest was considered to be recovered when the edaphic indicators in the rainforest were similar to those in the primary forest. MATERIALS AND METHODS Study Area Research was conducted in the logging area of PT Tunas Timber Lestari located in the tropical rainforest of South Papua with geographical position between 140 21` – 140 59` E and 05 50` o o o – 06 42` S (Fig.1). The annual rainfall was between o 231 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. Figure 1 Study area in logging concession of PT Tunas Timber Lestari (Murdjoko et al. 2016c) BIOTROPIA Vol. 24 No. 3, 2017 232 were compared to the unlogged forest to observe the recovery process. The selective logging was carried out by selectively cutting commercial trees having diameter of ≥ 40 cm. Sampling and Data Collection Samples were collected in each forest group using systematic sampling plots. The first plot was placed at 200 m from the main road to avoid edge effect. The plots were rectangular with various sizes i.e. 1. 20 x 20 m for trees (D) having DBH (diameter at breast height) of ≥ 20 cm; 2. 10 x 10 m for poles (C) having DBH of 10 to < 20 cm; 3. 5 x 5 m for saplings (B) having height of > 1.5 m and DBH of < 10 cm; and 2 x 2 m for seedlings (A) having height of < 1.5 m. The four plots were set as nested plot (Fig. 2a). Thirty 3,000 and 4,000 mm with daily moisture range of 75 - 85 %. The edaphic condition was typified as lowland forest with almost flat topography with soil formed by alluvial process (Petocz 1989). The vegetation was dominated by trees belong to Dipterocarpaceae Lauraceae Myrtaceae, and families et al. 2015). (Gandhi & Mitlöhner 2014; Kuswandi Several other plants such as lianas, rattans, ferns, palms, herbs, orchids and pandanus grew and interacted with trees in this forest (Murdjoko et al. 2016a). Five groups of forests were used in this research i.e. unlogged, one year post selectively- logged, five years post selectively-logged, ten years post selectively-logged and fifteen years post selectively-logged forests. The unlogged forest was taken as a primary forest which was a stable forest ecosystem. The selectively logged forests Figure 2 Nested plots to measure individual plant in both unlogged and selectively-logged forests Note: A = plot for seedlings; B = plot for saplings; C = plot for for poles; D = plot for trees; (a) Distance between plots = 100 m; (b) The 30 nested plots were laid on each forest group (unlogged, one year, five years, ten years and fifteen years post selectively-logged forests) 233 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. nested plots were laid in each forest (Fig. 2b) making a total of 150 nested plots for the 5 forest groups (unlogged, one year, five years, ten years and fifteen years post selectively-logged forests). Seedlings and saplings were sampled as understory, while poles and trees were sampled as upperstory in both unlogged and selectively logged forests. Data collected from seedlings, saplings, poles and trees consisted of numbers of individuals, diameter of individuals for those having DBH ≥ 10 cm and species name of individuals. Species identification was carried out by two herbarium technicians. Unidentified samples were set as voucher specimens and sent to the herbarium of "Balai Penelitian dan Pengembangan Lingkungan Hidup dan Kehutanan (BP2LHK) Manokwari" and Herbarium Manokwariense (MAN) Pusat Penelitian Keanekaragaman Hayati Universitas Papua (PPKH-UNIPA), Manokwari. Validation of the species names of the individuals was checked online at http://www.theplantl ist .org/; http://plants.jstor.org and www.ipni.org/ipni/. Soil samples were taken from the center and four corners of the 20 x 20 m plot. The litterfall samples were collected from each plot by making 1 x 1 m rectangular subplots in each plot. The soil and litterfall samples were sent to the laboratory of Balai Pengkajian Teknologi Pertanian Yogyakarta for determining the content of soil organic matter (SOM) for soil samples as well as Carbon (C) content, Nitrogen (N) content and dry weight for litterfall samples. Data and Statistical Analysis Canonical Correspondence Analysis (CCA) was applied to show the relationship among tree species using stem density and environmental factors (SOM, C, N contents and dry weight of litterfall) (ter Braak 1987; ter Braak 1986; Khairil et al. 2014). Plants communities were grouped as: a) understory consisted of small individuals (seedlings and saplings); and b) upperstory consisted of large individuals (poles and trees). Tree communities were formed as a result of interaction among tree species, SOM, C content, N content, dry weight of litterfall and forest groups (unlogged, one year, five years, ten years and fifteen years post selectively-logged). The CCA was computed using R softwarestatistical version 3.3.1. (R Core with VEGAN package Team 2014; Oksanen 2013). The tree et al. communities were grouped using Euclidean distance among tree species The Euclidean . distance among tree communities was calculated as the average and confidence interval of 95%. RESULTS AND DISCUSSION Tree Communities Total tree species in the study area were 163 species and classified as understory (159 species) and upperstory (127 species) (Table 1). Within tree species, there were 106 species consisted of both understory and upperstory. Table 1 Understory (a) and upperstory (b) tree communities formed due to logging activities Note: PF = unlogged forest; X1LF = one year post selectively-logged forest; X5LF = five years post selectively-logged forest; X10LF = ten years post selectively-logged forest; X15LF = fifteen years post selectively-logged forest; ALL = present in all forest groups; NON_AC = not associated 234 BIOTROPIA Vol. 24 No. 3, 2017 Table 1 Continued 235 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. Table 1 Continued 236 BIOTROPIA Vol. 24 No. 3, 2017 Table 1 Continued 237 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. Table 1 Continued 238 BIOTROPIA Vol. 24 No. 3, 2017 Table 1 Continued Those species existed in each forest group (unlogged, one year, five years, ten years and fifteen years post selectively-logged). Patterns of tree communities were formed for each forest group, especially for understory mostly occurred after logging activities. Upperstory were mainly recruited from understory of remnant trees. Several upperstory species were present before logging activities occurred in the forests. Our study presented the results of understory and upperstory communities influenced by logging activities and edaphic conditions. There were three patterns established in our study i.e. 1. tree species formed a tree community in a forest group; 2. tree species present in all forest groups; and 3. tree species did not form a community. Presence of certain tree species as understory in all forest groups was facilitated by ecological alterations, including logging activities. Several tree species existed in all forest groups indicating that those tree species were not influenced by ecological alterations. Distribution of understory tree community was depicted using CCA having 55.34% of the variation for two axes; variation for axis 1 was 30% and variation for axis 2 was 25.34% (Fig. 3; Table 2). ANOVA showed that the model was significant with < 0.05.p 239 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. Table 1 Continued Figure 3 Understory of four tree communities formed due to logging activities symbolized as grey (species grown in PF), green (species grown in X1LF), yellow (species grown in X5LF) and blue (species grown in X10LF-X15LF) Note: PF = unlogged forest; X1LF = one year post selectively-logged forest; X5LF = five years post selectively- logged forest; X10LF = ten years post selectively-logged forest; X15LF = fifteen years post selectively- logged forest; SOM = Soil Organic Matter (%); LF_C = Carbon content in litterfall (%); LF_N = Nitrogen content in litterfall (%); LF_DW = dry weight of litterfall (g) Canonical Correspondence Analysis (CCA) grouped the understory tree species into four tree communities i.e. 28 species in the unlogged forest; 21 species in the one year post selectively-logged forest; 21 species in the five years post selectively- logged forest and 17 species in the ten and fifteen years post selectively-logged forest (Table 1a). Distribution of upperstory tree community was shown of having variation of two axes of 58.26% with 31.24% variation for axis 1 and 27.03% variation for axis 2 (Fig. 4; Table 3). The CCA model was significant at p < 0.05. Edaphic Factors Interactions among SOM, C content, N Table 2 Summary of Canonical Correspondence Analysis (CCA) for understory tree community Importance of components Axes Total Inertia CCA1 CCA2 Eigenvalue 0.2152 0.1818 0.7175 Proportion explained 0.3 0.2534 Cumulative proportion 0.3 0.5534 Figure 4 Upperstory of four tree communities formed due to logging activities symbolized as grey (species grown in PF), green (species grown in X1LF), yellow (species grown in X5LF) and blue (species grown in X10LF-X15LF) Note: PF = unlogged forest; X1LF = one year post selectively-logged forest; X5LF = five years post selectively- logged forest; X10LF = ten years post selectively-logged forest; X15LF = fifteen years post selectively- logged forest; SOM = Soil Organic Matter (%); LF_C = Carbon content in litterfall (%); LF_N = Nitrogen content in litterfall (%); LF_DW = dry weight of litterfall (g) Table 3 Summary of Canonical Correspondence Analysis (CCA) for upperstory tree community Importance of components Axes Total Inertia CCA1 CCA2 Eigenvalue 0.1961 0.1697 0.6277 Proportion explained 0.3124 0.2703 Cumulative proportion 0.3124 0.5826 content, dry weight of litterfall and forest groups (unlogged, one year, five years, ten years and fifteen years post selectively-logged forests) were analyzed using CCA to figure out the fitting edaphic factors as the indicators of logged forest recovery. Results of CCA showed that SOM tended to be higher in the unlogged forest, dry weight of litterfall tended to be higher in the five years post selectively-logged forest and C content of litterfall was higher in the one-year post selectively-logged forest (Fig. 3 & 4; Table 4). Based on this analysis, the ten and fifteen years post selectively-logged forests were still in the recovery process, indicated by lower SOM content in those two logged forests compared to BIOTROPIA Vol. 24 No. 3, 2017 240 the unlogged forest. In contrast, dry weight of litterfall tended to be higher in all logged forests. These results were not in line with research results obtained from logged Bornean rainforest, in which one year post-logged forest produced less litterfall compared to that in the Bornean primary forest. The amount of litterfall in Bornean primary forest was similar to those in the Bornean five years post-logged forest (Prasetyo et al. 2015). This condition suggested that responses of logged forests were depended on ecological circumstances. Furthermore, specific silvicultural treatments should be designed carefully by considering forest condition. Ecological Changes as a Response to Selective Logging Tree communities in the unlogged forest were different from those in the logged forests. The differences were due to ecological changes caused by logging activities resulted in alteration of species composition (Arbainsyah . 2014; et al Verburg & van Eijk-Bos 2003; Lozada . 2012), et al tree density (Decocq . 2014), tree growth rate et al (Murdjoko 2016b) and association patterns et al. among biotic factors, light availability, ambient moisture, temperature, soil properties and litterfall stock as abiotic factors (Murdjoko et al. 2016c). Tree communities were formed as responses of each tree characteristics toward different ecological circumstances in logged forests. Understory and upperstory tree communities had different reactions toward ecological changes (Murdjoko et al. 2016a; Zhu . 2015b). Therefore, there were understory et al and upperstory tree communities consisted of the same species. Tree communities consisted of seedlings and saplings stages that required more light (Karsten . 2014; Flores . 2014). et al et al This is the reason why logged forests had altered tree compositions compared to those in the primary forest. Each logged forest has different species composition of the understory tree community. Species composition of the understory tree community was different among the logged forests. Understory tree community in the one year post selectively-logged forest had very different species composition compared to those in the unlogged forest (Fig. 3). Understory tree community in the five years post selectively- logged forest had very different species composition compared to those in the ten and fifteen years post selectively-logged forests (Fig. 3). These differences in species composition were influenced by changes in environmental conditions (Corrià-Ainslie . 2015; Schnitzer & et al Walter 2013; Duah-Gyamfi . 2014). et al The CCA showed that understory tree community in the one year post selectively-logged forest was mainly influenced by Carbon content of litterfall. Understory tree community in the five years post selectively-logged forest was formed as a response toward dry weight of litterfall. The nitrogen content of litterfall affected the establishment of understory tree community in the ten and fifteen years post selectively-logged forests. Understory tree community in the unlogged forest was influenced by SOM content. Alterations of soil characteristics in the logged forests were caused by the change of microclimate conditions (Asase et al. et al. 2014; Imai 2012). Logging activities were responsible for the widening canopy gap leading to the increase of light availability toward understory tree community (Schwartz 2016). Logging activities were also responsible for the decrease of tree density causing the changes in tree growth rates (Verburg & van Eijk-Bos 2003; Cannon . 1998; Do . 2016). These et al et al Table 4 ANOVA of CCA to analyze interactions among SOM, C content, N content, dry weight of litterfall and forest groups (unlogged, one year, five years, ten years and fifteen years post selectively-logged forests) Edaphic factors Df Sums of square Mean square F.Model R2 P SOM 1 0.746 0.74644 2.438868 0.01442 0.001 * LF_C 1 0.692 0.6916 2.259688 0.01336 0.001 * LF_N 1 0.543 0.54259 1.772822 0.01048 0.005 * LF_DW 1 0.795 0.79469 2.596517 0.01536 0.001 * Residuals 161 49.27566 0.30606 0.94638 Total 165 52.05166 1 Note: *= significant at p < 0.05 Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. 241 conditions triggered space and light competitions among tree species, especially in the seedlings and saplings stages(Laurans . 2014).et al Upperstory tree community had different patterns from the understory tree community. In the unlogged forest, species composition of understory was different from that of upperstory tree community. Conspecific association occurred in the unlogged forest. Not all species grown in the understory tree community grew in the upperstory tree community of unlogged forest (Murdjoko et al. 2016a). Ecological condition occurred in the upperstory tree community was similar to that in the understory tree community. Trees in tropical forest experienced more diameter growth in the upperstory tree community (Zhu et al. 2015a). Upperstory tree community in the unlogged forest had very different species composition compared to those in the five years post selectively-logged forest (Fig. 4). However, similar species composition was observed among upperstory tree communities in the unlogged forest, one year post selectively- logged forest, ten and fifteen years post selectively-logged forests (Fig. 4). Tree species located in the five years post selectively-logged forest was the results of species competition caused by the change of ecological conditions. Thus, the current species were not the same as the previous species because of the duration of the ecological process. Upperstory tree community in the logged forests showed a dynamic establishment of tree community. Each species had different growth rate as a response to logging impact (Murdjoko et al. 2016b). Some species had higher population growth rate than others leading to higher survival rate (Murdjoko 2013; Zuidema et al. 2009). Although recovery process was seen to be happening in the ten and fifteen years post selectively-logged forests, the process still requires more time to reach the fully recovered stage. Implication of Ecological Approach for Sustainable Forest Management This study proposed an ecological approach to determine whether logged forests were recovered in fifteen years. Existing tree communities and edaphic factors, especially SOM, in the unlogged forest were used as a reference of logged forest reaching recovered condition. SOM plays an important role to support nutrient absorption in soil (Mutiso 2013). The soil of South Papua et al. is mainly classified as Ultisols, so the characteristic of soil is infertile (Marshall & Beehler 2012). Selective logging activities did not seem to totally change ecological condition. The logged forest was declared to be fully recovered when its conditions had reached similar condition as those in unlogged (primary) forest, especially in terms of ecological aspects such as the content of SOM, stem density and species composition. Therefore, it is imperative to set permanent sample plots in the unlogged (primary) and logged forests, to conduct intensive and persistent monitoring of ecological conditions and tree growth (Krisnawati & Wahjono 2010; Ruslandi et al. 2017a; Ruslandi et al. 2017b). The monitoring results would be valuable as basic information to further evaluate the silviculture protocol. Useful modifications could be designed by taking ecological perspective into account. CONCLUSIONS Understory and upperstory tree communities formed different patterns due to logging activities. Species composition existed in the tree communities in the ten and fifteen years post selectively-logged forests were not similar to that in the unlogged forest, meaning that the logged forests were still in the recovery process. SOM content in the logged forest was lower compared to that in the unlogged forest, indicating that the logged forests were not fully recovered. These occurrences indicated that it took more than fifteen years for the logged forests to be fully recovered. Long-term studies are necessary to continuously monitor the ecological process in the logged forest in reaching the recovery stage. The recorded influential ecological factors obtained from this study can be used as indicators for logged forest recovery. ACKNOWLEDGEMENTS The research was funded by Beasiswa Pendidikan Pascasarjana Dalam Negeri (BPP- DN) 2014, Manokwari Regency and West Papua Province. The authors are thankful to the BIOTROPIA Vol. 24 No. 3, 2017 242 following people for assisting in fieldwork and species vegetation identification as well as providing useful inputs: Nithanel M. H. Benu (Balai Penelitian dan Pengembangan Lingkungan Hidup dan Kehutanan Manokwari), Dr Purnomo (Faculty of Biology, Gadjah Mada University), Prof Dr Ir Suryo Hardiwinoto, M. Agr. (Faculty of Forestry, Gadjah Mada University), Dr Sena Adi Subrata, M.Sc (Faculty of Forestry, Gadjah Mada University), Dr Ir Soewarno Hasanbahri, M.S. (Faculty of Forestry, Gadjah Mada University), Dr Ir Eko Hanudin, M.Sc (Faculty of Agriculture, Gadjah Mada University), Prof Dr Charlie Danny Heatubun, S.Hut, M.Si, FLS (Faculty of Forestry, Papua University), Dr Ir Relawan Kuswandi, M.Sc, Daud Bano, Dedy Subrata May, Vinoba Chandra, Ari Wibowo, Anton Bongga, Sri Emy Rahayu and all staff of PT Tunas Timber Lestari. We are also grateful to anonymous reviewers for valuable comments . REFERENCES Arbainsyah, de Iongh HH, Kustiawan W, de Snoo GR. 2014. Structure, composition and diversity of plant communities in FSC-certified, selectively logged forests of different ages compared to primary rain forest. Biodivers Conserv 23:2445–72. Asase A, Asiatokor BK, Ofori-Frimpong K. 2014. Effects of selective logging on tree diversity and some soil characteristics in a tropical forest in Southwest Ghana. J For Res 25:171–6. Cannon CH, Peart DR, Leighton M. 1998. Tree species diversity in commercially logged Bornean rainforest. Science 80:1366–8. Core Team R. 2014. R: A language and environment for statistical computing. Vienna (AT): R Foundation for Statistical Computing. Corrià-Ainslie R, Julio-Camarero J, Toledo M. 2015. Environmental heterogeneity and dispersal processes influence post-logging seedling establishment in a Chiquitano dry tropical forest. For Ecol Manage 349:122–33. Decocq G, Beina D, Jamoneau A, Gourlet-Fleury S, Closset- Kopp D. 2014. Don't miss the forest for the trees! Evidence for vertical differences in the response of plant diversity to disturbance in a tropical rain forest. Perspect Plant Ecol Evol Syst 16:279–87. Do TV, Cam NV, Sato T, Binh NT, Kozan O, Thang NT, Mitlöhner R. 2016. Post-logging regeneration and growth of commercially valuable tree species in evergreen broadleaf forest. Vietnam. J Trop For Sci 28:426–35. Duah-Gyamfi A, Swaine EK, Adam KA, Pinard MA, Swaine MD. 2014. Can harvesting for timber in tropical forest enhance timber tree regeneration? For Ecol Manage 314:26–37. Edwards DP, Tobias JA, Sheil D, Meijaard E, Laurance WF. 2014. Maintaining ecosystem function and services in logged tropical forests. Trends Ecol Evol 29:511–20. Flores O, Hérault B, Delcamp M, Garnier É, Gourlet- Freury S. 2014. Functional traits help predict post- disturbance demography of tropical trees PLoS O N E 9 (9 ) : e 105022 . Ava i l a b l e f ro m : https://doi.org/10.1371/journal.pone.010502. Gandhi Y, Mitlöhner R. 2014. Tree species composition, diversity and structure in Tunas logging concession area of Papua-Indonesia. Tree 66:47. Hattori D, Kenzo T, Irino KO, Kendawang JJ, Ninomiya I, Sakurai K. 2013. Effects of soil compaction on the growth and mortality of planted Dipterocarp seedlings in a logged-over tropical rainforest in Sarawak, Malaysia. For Ecol Manage 310:770–6. Hendri, Yamashita T, Kuntoro AA, Lee HS. 2012. Carbon stock measurements of a degraded tropical logged- over secondary forest in Manokwari Regency, West Papua, Indonesia. For Stud China 14:8–19. Hoang VS, Baas P, Keßler PJA, Slik JWF, Steege HT, Raes N. 2011. Human and environmental influences on plant diversity and composition in Ben En National Park, Vietnam. J Trop For Sci 23:328–37. Imai N, Kitayama K, Titin J. 2012. Effects of logging on phosphorus pools in a tropical rainforest of Borneo. J Trop For Sci 24:5–17. Karsten RJ, Meilby H, Larsen JB. 2014. Regeneration and manag emen t o f l e s s e r known t imbe r species in the Peruvian Amazon following disturbance by logging. For Ecol Manage 327:76–85. Khairil M, Wan Juliana WA, Nizam MS. 2014. Edaphic influences on tree species composition and community structure in a tropical watershed forest in Peninsular Malaysia. J Trop For Sci 26:284–94. Krisnawati H, Wahjono D. 2010. Effect of post-logging silvicultural treatment on growth rates of residual stand in a tropical forest. Indones J For Res 2:112–24. Kuswandi R, Murdjoko A. 2015. Population structures of four tree species in logged-over tropical forest in South Papua, Indonesia : an integral projection   model approach. Indones J For Res 2:93–101. Kuswandi R, Sadono R, Supriyatno N, Marsono D. 2015. Keanekaragaman struktur tegakan hutan alam bekas tebangan berdasarkan biogeografi di Papua. J Mns dan Lingkung 22:151–9. Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. 243 Kuswandi R. 2014. The effect of silvicultural treatment on stand growth of logged-over forest in South Papua. Indones J For Res 1:117–26. Laurans M, Hérault B, Vieilledent G, Vincent G. 2014. Vertical stratification reduces competition for light in dense tropical forests. For Ecol Manage 329:79–88. Lozada J, Arends E, Sánchez D, Villarreal A, Soriano P, Costa M. 2012. Vegetation succession of logged forest in the western alluvial plains of vegetation succession of logged forest in the Western Alluvial Plains of Venezuela. J Trop For Sci 24:300–11. Marshall AJ, Beehler BM. 2012. The ecology of Papua: Part one. Hongkong (HK): Periplus Editions. p. 168-9. Murdjoko A. 2013. Recuperation of non-commercial trees in logged forest in Southern Papua, Indonesia. J Man Hut Trop 19:94–102. Murdjoko A, Marsono D, Sadono R, Hadisusanto S. 2016a. Plant species composition and their conspecific association in natural tropical rainforest, South Papua. Biosaintifika J Biol Biol Educ 8:33–46. Murdjoko A, Marsono D, Sadono R, Hadisusanto S. 2016b. Population dynamics of for the period of Pometia post-selective logging in tropical rainforest, Southern Papua, Indonesia. Biosaintifika J Biol Biol Educ 8:321-30. Murdjoko A, Marsono D, Sadono R, Hadisusanto S. 2016c. Tree association with and its structure in Pometia logging concession of South Papua Forest. J Man Hut Trop 22:180-91. Mutiso FM, Hitimana J, Kiyiapi JL, Sang FK, Eboh E. 2013. Recovery of Kakamega Tropical Rainforest from anthropogenic disturbances. J Trop For Sci 25:566–76. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O'Hara RB, ... Oksanen MJ. 2013. Package 'vegan'. Community ecology package, version, 2(9). Pennington RT, Hughes M, Moonlight PW. 2015. The origins of tropical rainforest hyperdiversity. Trends Plant Sci 20:693–5. Petocz RG. 1989. Conservation and development in Irian Jaya: a strategy for rational resource utilization. Leiden (NL): E. J. Brill. Prasetyo E, Hardiwinoto S, Supriyo H. 2015. Litter production of logged-over forest using Indonesia selective cutting system and strip planting (TPTJ) at PT Sari Bumi Kusuma. Procedia Environ Sci 28:676–82. Putz FE, Romero C. 2014. Futures of tropical forests (sensu lato). Biotropica 46:495–505. Ruslandi R, Cropper Jr. WP, Putz FE. 2017a. Effects of silvicultural intensification on timber yields, carbon dynamics, and tree species composition in a Dipterocarp forest in Kalimantan, Indonesia: an individual-tree-based model simulation. For Ecol Manage 390:104–18. Ruslandi R, Cropper Jr. WP, Putz FE. 2017b. Tree diameter increments following silvicultural treatments in a Dipterocarp forest in Kalimantan, Indonesia : a   mixed-effects modelling approach. For Ecol Manage 396:195–206. Rutten G, Ensslin A, Hemp A, Fischer M. 2015. Forest structure and composition of previously selectively logged and non-logged montane forests at Mt. Kilimanjaro. For Ecol Manage 337:61–6. Schnitzer SA, Walter PC. 2013. Treefall gaps and the maintenance of species diversity in a tropical forest. Ecology 82:913–9. Schwartz G. 2016. Profitability of silvicultural treatments in logging gaps in the Brazilian Amazon. J Trop For Sci 28:68–78. Sist P, Mazzei L, Blanc L, Rutishauser E. 2014. Large trees as key elements of carbon storage and dynamics after selective logging in the Eastern Amazon. For Ecol Manage 318:103–9. Susanty FH, Suhendang E, Jaya INS. 2015. Mortality and ingrowth pattern of Dipterocarps in forest recovery in East Kalimantan. BIOTROPIA 22(1):11–23. ter Braak CJF. 1986. Canonical correspondence analysis: a new eigenvector technique for multivariate direct gradient analysis. Ecology 67:1167–79. ter Braak CJF. 1987. The analysis of vegetation- environment relationships by Canonical Correspondence Analysis. Vegetatio 69:69–77. Verburg R, van Eijk-Bos C. 2003. Effects of selective logging on tree diversity, composition and plant functional type patterns in a Bornean rainforest. J Veg Sci 14:99–110. Wasrin UR, Putera AE. 1999. Litterfall in a primary and two logged-over lowland tropical rainforests in Pasirmayang, Jambi. BIOTROPIA 14:36–51. West TAP, Vidal E, Putz FE. 2014. Forest biomass recovery after conventional and reduced-impact logging in Amazonian Brazil. For Ecol Manage 314:59–63. Whitfeld TJS, Lasky JR, Damas K, Sosanika G, Molem K, Montgomery RA. 2014. Species richness, forest structure, and functional diversity during succession in the New Guinea Lowlands. Biotropica 46:538–48. Zambrano J, Coates R, Howe HF. 2014. Effects of forest fragmentation on the recruitment success of the tropical tree at Los Tuxtlas, Poulsenia armata Veracruz, Mexico. J Trop Ecol 30:209–18. Zhu H, Yong C, Zhou S, Wang H, Yan L. 2015a. Vegetation, floristic composition and species diversity in a tropical mountain nature reserve in southern BIOTROPIA Vol. 24 No. 3, 2017 244 Yunnan, SW China, with implications for conservation. Trop Conserv Sci 8:528–46. Zhu Y, Comita LS, Hubbell SP, Ma K. 2015b. Conspecific and phylogenetic density-dependent survival differs across life stages in a tropical forest. J Ecol 103:957–66. Zuidema PA, Brienen RJ, During HJ, Güneralp B. 2009. Do persistently fast-growing juveniles contribute disproportionately to population growth? A new analysis tool for matrix models and its application to rainforest trees. Am Nat 174:709–19. Recovery of residual forest ecosystem: impact of selective logging - Murdjoko et al. 245 Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16