Journal 1 of 27 Tropical Forest Science 16(3): 271-282 (2004)

CONSERVATION STATU F HOPEAO S SUBALATA ( PENINSULAN )I R MALAYSIA

. Chua*L L. S . Nurulhuda H , w Hamida. Sa . M , G . L h&

Forest Research Institute Malaysia, 52109 Kepong, Kuala Lumpur, Malaysia

Received July 2003______

CHUA, L. S. L., NURULHUDA, H., HAMIDAH, M. & SAW, L. G. 2004. Conservation statu f Hopeaso subalata (Dipterocarpaceae Peninsulan i ) r Malaysia. Hopea subalata (Dipterocarpaceae) is a rare and endemic species of Peninsular Malaysia. It is known occuo t r onl Kanchinn yi g Forest Reserve, Selangor spatiae Th . l distributioe th f no individuals is clumped. Some adult individuals have no regenerating stands surrounding them. While the majority of the individuals lies within the forest reserve boundary, there are at least 10 present outside the boundary. The conservation status of H. subalata possibld an e threat populatios it o st describede nar . Conservation measures for populations existing in the forest reserve and the population outside the reserve are prescribed.

Key words: Spatial distributio nKanchin- g Forest Reserv ethreat- populatioo st n- size class distribution

CHUA, L. S. L., NURULHUDA, H., HAMIDAH, M. & SAW, L. G. 2004. Status pemuliharaan Hopea subalata (Dipterocarpaceae) di Semenanjung Malaysia. Hopea subalata (Dipterocarpaceae) merupakan salah satu spesies endemi i Semenanjunkd g Malaysia. Spesies ini amat langka dan hanya wujud di Hutan Simpan Ranching, Selangor. Populasi spesies memperlihatkan taburan secara berkelompok. Kajian menunjukkan bahawa tidak terdapat pertumbuhan semul i sekelilinad g sesetengah pokok induk. Walaupun kebanyakan pokok tumbuh di dalam kawasan hutan simpan ini, terdapat kira-kir poko0 a1 k yang berad luai ad r sempadannya. Status pemuliharaan ancaman da n kepada populas subalat. iH a dihuraikan. Langkah-langkah pemuliharaan bagi populasi di dalam dan di luar hutan simpan disarankan.

Introduction

Hopea subalata Symington (Dipterocarpaceae endemis i ) Peninsulao ct r Malaysia (Ashton 1982). It is very rare and known to occur only in Kanching Forest Reserve state th Selangorf eo n i face rarit s vieth n t it . I thad f wyo an t Selangor face seversa e conflict between conservatio biodiversitf no socio-economid yan c development,a study was undertaken to locate and enumerate populations of H. subalata and determine its conservation status. Apart from the taxonomic accounts published by Ashton (1982) and Symington (1943), and collections lodged at the Forest Research Institute Malaysia (FRIM)'s Herbarium (KEP), ther s hardlei y yan documentation available to indicate the status of the species in its natural habitat. This paper outlines geographical distribution, current stocking, diameter size

* E-mail: [email protected]. my 2 27 Journal ofTropical Forest Science 16(3): 271-282 (2004)

class distribution and perceived threats to the species. In this paper, the authors provide recommendations thaaimee ar t t improvinda g current conservation measures. Area of population that is located outside the forest reserve was estimated and conservation measures for this population are also suggested.

Material methodd san s

Collections lodged at KEP indicated that the species existed in compartments 1, 2, 5D and 14 of Kanching Forest Reserve. Current forestry records show that compartmentstile ar l 4 extant1 d an . 2 Compartmen , s1 bees ha n D subdividet5 d inte Stat Districd oTh ean compartment. 17 t d Foresan 6 1 t , Offices4 s kindly supplied detailed inventory records suc s logginha g cycle d logginan s g prescriptions of all the compartments. A botanical tea responsibls mwa locatinr efo populatione gth . subalataH f so in this study. Habitat preferences were useguida s da outlino et possible eth e extenspeciee th f o t s distribution. Major topographical features, suc ridgess ha , slopes, riverbanks undulatind an , 4 1 gd lanan compartment5 n di 1 , 16 , 17 , 4 , s2 were inspected. Once the population was located, a careful search was made to identify the outliers; these outliers were used as a guide for the demarcation of the population. This search method enabled the team to cover a larger area but inadvertently some trees could have been misse surveye th dn i . Standard methods of tagging and diameter at breast height (dbh) measurements were used (Manokaran et al 1990). Traditional ecological methodology was modified to examin e conservatioth e n e speciesstatuth f o s . Population boundars ywa demarcated. Impulse 200 lasea , r instrument which measures distances, heights, inclination and azimuth, was used to map population boundary and position e grounth n o d whilinstrumenS GP e t (Garmin Etrex Summit uses o wa dt ) determine coordinates for population boundary. The accuracy for the GPS instrumen . abouGrouns m twa 0 3 t d station points (STNs) were use creato dt e an imaginary boundary around populations. The STNs were marked on the ground. Distances between STNs varied dependin longese th n go t clear linf eo sight available. Tree positio mappes nwa d fro selectee mth d STNs. Both tree-to- tree and array mapping methods were employed during the tree mapping exercise STNse .Th , tree-to-tre arrayd ean data were entered into Roadeng software (Version 3.1), which then generate dtrea e position map. This Roadeng-generated map was converted to .dxf format. Population boundary was then tied to GPS coordinates using Autoca 2000p dMa i software. ascertaio T n whether population compartmentn si wer4 1 ed randomlan s2 y dispersed, i.e. follow a Poisson distribution, the populations were subjected to a Chi-squar e spatiaTh e . ) tesl ( distributiot f tree ncalculateo s swa d usine gth 2 variance-to-mean rati r indeoo f dispersioxo n (ID) (Elliott 1973, Browe. al t e r 1997). e areCalculatio bases th f occupanc ao wa n do D I f no y definee th y b d population boundary. The area of occupancy was subdivided into 5 X 5 m subplots. The index of dispersion calculation is given below. Journal of Tropical Forest Science 16(3): 271-282 (2004)

GPS points, population boundar tred yan e positions were plotted usinc gAr View 3.3. The tree position map generated was then overlaid onto topographic and forest reserve maps and Landsat TM images (30 m). A digital copy of the area's topographi acquires wa p dcma fro Survee mth y Departmen Marcn ti h 2003. Editing and map display were done using Autocad Map 2000i software. The GIS data on the Forest Reserve boundary, including its compartment boundaries, were acquired fro Forese mth t Departmen Peninsulaf o t r Malaysi Novemben ai r 2002. Landsae Th acquires Marcwa 8 t 1 ) imag r m dhfo 0 2001e(3 , courtes Remote th f yo e UnitSensinS GI , d FRIMg an spatia e .Th l dat incorporates awa d usin Viec g3 wAr 3. software.

Plant description

Detailed botanical description givee sar Symington ni n (1943 Ashtod an ) n (1982). Hopea subalata smala s i l tree, withm maximuc 5 3 d an m m heigh 2 1 f o t dbh. The outer bark is smooth, brownish-grey, while its inner bark is a very pale pinkish brown turning yellowish brown upon exposure larged an y , .B tree s found naturally have a poor bole form or a very short clear bole. Hopea subalata in FRIM's dipterocarp arboretum attained 13.06 cm dbh at around 40 years of age (Ng & Tang 1974); there was no girth increment in the subsequent measurement taken year4 4 t sa (Wong pers, M. . . comm.)K , . 274 Journal of Tropical Forest Science 16(3): 271-282 (2004)

Geographical distribution

Although herbarium records indicated that H. subalata occurred in compartments , Kanchin14 d an gD Fores5 , 12 , t Reserve, recent ground survey showed thae th t species was present only in compartments 2 and 14 (Figure 1). No individual was detected in compartment 1 and compartments 4, 16 and 17, i.e. the former compartmen . Tree 5D tcompartmen n si forme2 t populatioda n together with other species outside the forest reserve.

Ecology

Hopea subalata is found on undulating land up to 140 m above sea level. It has a preference for low ridges and gentle hill slopes (Figures 2 and 3) but is also found along stream banks. It grows on granite derived soils. On the hill slopes and ridge tops, it occurs together with the gregarious Dryobalanops aromatica. It is interesting to note that Kanching Forest Reserve has a small quartz ridge. It is one of the many small quartz ridges that border the largest outcrop, i.e. the Klang Gates Ridge. Quartz ridge renownee sar their dfo r high endemism Kanchinn I . g Forest Reserve, this geological aggregatio givy en ma compleris a o et x serie soilf so s that possibly determine the ability of a particular species to colonise specific sites. This perhaps explains why the populations of H. subalata in the reserve are rather discreet.

Figur e1 Distributio Hopeaf no subalata Kanchinn i g Forest Reserve, Selangor, Peninsular Malaysia Journal of Tropical Forest Science 16(3): 271-282 (2004) 275

Figure 2 3-D view of Hopea subalata on a 10 x 10 m grid in compartment 2, Kanching Forest Reserve

vie D Figur griHopeaf wm 3- o compartmen n d0 i 1 e 3 x subalata0 1 a , Kanchinn o 14 t g Forest Reserve

Flowerin fruitind gan g

The H. subalata population at Kanching Forest Reserve fruited in 1921, 1927, 1933 and 1976. The large gap between the observed fruiting years does not indicate non-flowerin fruitingd gan , merel interveninye thath n i t g period collectioo n , n was made tre.A FRIM't ea s dipterocarp arboretum (No. 149) fruite 1955n di , 1958 276 Journal of Tropical Forest Science 16(3): 271-282 (2004)

and 1963. A planted tree near the cafeteria flowered in 1980 while another individual dipterocare inth p arboretum fruite 1985n d i above .Th e informatio acquires nwa d fro herbariue mth m specimen FRIMt sa . During the mast-fruiting year of 2002, observed trees in the Kanching population flowert no d .di However (FR 5 FRIM't , a tree 72 I 4 . 41780s77 no s d dipterocaran ) p arboretum flowered but did not set mature fruits. The non-synchronous flowering and fruiting behaviour between two fairly near populations—FRIM is less than awa1m 0k y from Kanching Forest Reserve—canno explainee tb present da t because its flowering behaviou hardls ri y known. Plantin dipterocare th g n recordi 9 14 tre f p. so arboretueno m indicated that ti 13.7f o wheh m 4c db floweret n i a d ha 1955n di . Although datonls aywa available froe tree speculatme on w , e tha . subalataH t becomes reproductively maturt ea around thisizeh sdb .

Conservation notes

Population compartmentin 2

Figure 4 shows the tree distribution in compartment 2. The corrected X2 value indee th (ddispersiof f xo o ) 28.15s nwa , indicatin gdistinca t clumping pattern (Table 1). The larger trees (> 20 cm dbh), namely nos. 259, 262, 257 and 260 were located at the periphery of the population and were not surrounded by other individuals. In other words, there were no regenerating stands surrounding these reproductively mature individuals indicat. y Thiconcerf o ma s i t i er nthafo t these adult trees (trees with dbh > 13.74 cm) do not produce viable progenies. Alternatively because b , thiy sma e site conditions have influence distributioe dth n of regeneration. It can be seen from Figure 4 that other adult trees (nos. 42 and 173) were surrounded by juveniles of several diameter size classes. During the enumeration, 266 trees of 1 cm dbh were detected. Due to the limitations of the survey method, the population is expected to be higher as some individual havy esma been misse t durindou searche gth distributioe Th . . H f no subalata by dbh size classes is given in Figure 5. Typically, density, based on diameter size class, decreased with increasing size class. As much as 97.8% of the population was dbhbelo largese m c .Th w0 2 t diamete 34. s (tre m 0. 42)rabsencwa c e eno .Th e largef o sizh erdb class indicated tha . subalatatH small-siza s i e dipterocarp tota.A l treeo9 f4 s (18.35 populatione th %f o considere e b n )ca breedins da g individuals, assumptioe baseth n do n thaspeciee th t s becomes reproductively mature when reacheh db s 13.7. 4cm Figure 1 shows that part of the population lies outside the boundary of the forest reserve t leas A .tree0 1 toutliere sar s (nosdbhm e c Th .0 2 1-10) < l al , estimated area of occupancy was 0.17 ha. District Land Office records indicate that this land has been categorised and granted the category 'Lot Kediaman' (residence lot) (Gombak District Land Office records 2/161/8G PT : . File No 7 th (L)t bu ) land has yet to be purchased. Landsat TM images for March 2001 (Figure 6) and recent ground survey indicated that this are foresteds ai . Journal of Tropical Forest Science 16(3): 271-282 (2004) 277

Table 1 test of significant departures of index of dispersion (ID) of Hopea subalata from th2 e Poisson distribution

Compartment x ID

6 97 1.1 0 180 0.3 6 326 3.36 2684.624 28.15

14 177 1019 357 0.32 0.17 1.88 1913.84 16.76

Figur e4 Distributio Hopeaf no subalata compartmenn i , Kanchin2 t g Forest Reserve

Figure 5 Distribution by diameter (dbh) size classes of Hopea subalata trees 1 cm in compartmen , Kanchint2 g Forest Reserve 278 Journal of Tropical Forest Science 16(3): 271-282 (2004)

Population compartmentin 14

Figure 7 shows the tree distribution in compartment 14. The corrected 2 value (d) based on the ID was 16.76, indicating a clumping pattern (Table 1). The larger trees (> 20 cm dbh), namely, no. 7, 8 and 131 were located at the periphery of the population and had regenerating stands.

Figure 6 Landsat TM (2001) for Kanching Forest Reserve, Selangor

Figur e7 Distributio Hopeaf no subalata compartmenn i , Kanchin14 t g Forest Reserve Journal of Tropical Forest Science 16(3): 271-282 (2004) 279

werh db e enumeratem c 1 Atree7 tota f 17 s o f lo thin di s compartment. Their distribution by dbh size classes is given in Figure 8. As much as 98.3% of the dbh populatiolargesm e c .Th 0 2 t< diametes ntree wa 5 tota 2 A 25.. s f slo 0rwa cm (14.12%) were reproductively mature.

IUCN (2001) category

In view of the above results, H. subalata should be given the 2001 IUCN category of CR A4cB2. Followin e IUCgth N (2001 Lisd Re t) categories e speciea , b y sma categorise extincs da t (EX), extinc wile th d n ti (EW) , critically endangered (CR), endangered (EN), vulnerable (VU), near threatened (NT), least concern (LC), data deficient (DD) or not evaluated (NE). Hopea subalata is given an A4 because ther inferren a s ei suspecter do d population size reductio ove% periora 80 f no d of three generations (up to a maximum of 100 years in the future); this period includes both the past and the future. The IUCN CR criteria further state that population reductio understooe b havt t causes no it eno y r ceasey no s ma ma d r do reversiblee b t followinge no th y f o , ma y direc) baser (a an o : n dto observation) (b , inden a abundancf xo e appropriate taxone declina th ) o t (c , aren ei occupancyf ao , extent of occurrence and/or quality of habitat, (d) actual or potential levels of exploitation, and/oeffecte th introduce) f o s(e r d taxa, hybridisation, pathogens, pollutants, competitors or parasites. In the case of H. subalata, its absence from compartments 1, 4, 16 and 17 indicates that there has been a decline in area of occupanc extend yan occurrencef o t ; hence gives i t i , n'c'a additionn .I , because its geographic extenrange e fore th th f m occurrencf o n t ,i o ared occupancyef an a o , is estimate lese b s o dt tha km . 0 gives ni 1 2t B2 i , na

Figure 8 Distribution by diameter (dbh) size classes of Hopea subalata trees 1 cm in compartment 14, Kanching Forest Reserve 0 28 Journal of Tropical Forest Science 16(3): 271-282 (2004)

Discussion

Hopea subalata is confined to certain parts of the forest reserve and has very low adult population densities. These reasons alone necessitate the need for conservation. Studies with local endemic dipterocarps have shown that many species exhibi densitiew tlo s (Baw Ashtoa& n 1991, Bawa 1998). Hopea subalata triploia s i d (2n = 21, also 2n = 20 and 22) (Kaur et al 1978, Somego 1978) and is closely allied phylogenetically to H. rudiformis, H. sphaerocarpa and H. mesuoides of the subsection Sphaerocarpae (Ashton pers, S. . ,.P comm.) insufficieno t e .Du t informatios it n no breeding systems, genetic diversit demographid yan c patterns difficuls i t i , o t t speculate whethe populationss rit t theia , r present sizes sufficientle ar , y viablo et sustai speciee nth s ove lone th r g term noticee W . d that while generaln i , e th , population structur viablees i totae th , l numbe adultf ro smals s i e somd th lan f eo adult trees have no regenerating stands surrounding them. Presumably, some of these tree e loss f individualth fruitee pasto s t th bu ,n i d s from subsequent generation (s) could indicate that these trees cannot produce sufficiently vigorous plant withstano t s d selection. Genetic diversit demographid yan c studiee sar require thiw s ho affectfin dt o t dou s species surviva facn li pressure f eo s from natural selection and man-made threats. Research work on the genetic diversity of two other locally endemic dipterocarps, namely, Shorea lumutensis and H. bilitonensis is nearing completion (Leepers, C. . .,S comm.) date .aTh will provide crucial insights int genetie oth c compositio locaf no l rare populations . subalata,H r .Fo efforte sar currently underwa outlino yt genetis eit c diversit understand yan breedins dit g system. Forest fragmentation, deforestation, even minor change landusn si selectivd ean e harvesting may have long-term deleterious effects on the genetic diversity of H, subalata. 1920se th n I , whe speciee nth discoveres swa Kanchingn di , there were large tracts of primary forests in Selangor. Fragmentation has since taken place; in Kanching Forest Reserve, forestry records indicate that most of the compartments, includin t beegno ncompartmentd loggeha , t wer14 dbu ed destroyean s2 r do damaged durin Japanese gth e occupatio mand nan y part beed sha n rehabilitated wit . aromaticahD (Dipterocarpaceae) (Wan Mamat Saufi, pers. comm.)e Th . rehabilitation must have playe importann da t rol sustaininen i populatione gth n si thes compartmentso etw . The apparent absence of H. subalata from compartments 1, 4, 16 and 17 is of concern. Two scenarios are speculated. Lower collecting intensities in these compartments could indicate that the populations are more dispersed with individuals being more scattere hencd dan e more difficul detecto t t apparene .Th t absence of H. subalata could simply reflect a failure to detect the species in the current survey. The second scenario involves the road and the transmission power line dividethaw tno s Ranching Forest Reserve (Figurpresence Th . e1) roa f eo d transmissiod an n line effectively creates forest fragmentation ove lone rth g term. physicae Th l barriers addition ,i accompanyine th no t g forest clearance activities, either permitted or illegal within the vicinity, and the edge effects that invariably arise from fragmentation are likely to hamper the natural exchange of gene flow and affect its population viability in the long term. Journal of Tropical Forest Science 16(3): 271-282 (2004) 281

Compartment wer4 1 ed originallan s2 y identifie forestre th n di y records sa being productive (Wan Mamat Saufi, pers. comm.). In 1985, however, the entire forest reserve was converted from a productive status (where logging would be permitted) to a protected status and placed under the functional class of recreation forest. Selective harvesting per se, although quoted abov potentiaa s ea l facto genetir rfo c reduction, is not applicable to H. subalata. It is a small size dipterocarp and, by Malaysian forestry standards, has no commercial viability since stocking of poors wa .h db m c tree0 3 s>

Conservation recommendations

overale Parth f to l populatio compartmenn i liet2 s outsid forese th f eo t reserve. From a conservation viewpoint, this forested zone, which separates the forest reserve boundary from the residential/industrial estates, is too narrow. Since a land gran alreads ha t y been issued, this area wil purchasee b l converted dan d into non-forest landus futuree th short-ter a n es i .A m conservation measures i t i , strongly recommended that the State Forest Department and the District Forest Office mak appea n eSelangoe a th o lt r Government Districe th a ,vi t Land Office, preveno t t purchas lane th df eo parcel recommendes i t I . d thagrane th t t File 2/161/8NoG .PT 7 (L), 'Lot Kediaman' Kampung Bat Rawang, u16 , Selangoe rb cancelled, thus renderin t e sitavailablgth eno r conversionefo e longth r - Fo . term woult ,i prudene db relevane th f ti t authorities (District Land Office, Forest Department and the Selangor Government) gazette this relatively small site as an addition to the forest reserve. The annexed land then falls under the jurisdiction, management and legal mechanism of the Permanent Forest Estates, under which it should be placed in the protection forest category. Strict adherence to the regulations will ensure the conservation of the population. Further protection can be decreed by designation of a virgin jungle reserve status or recognising restricted zones in the overall designation of recreation forest status Kanchine foth r g Forest Reserve. additionn I , since both population situatee sar d relatively forese closth o et reserve boundary, it is strongly recommended that the District Forest Office regularly monitor pertinent sections of the boundary to ensure its integrity and prevent future conflicts.

Acknowledgements

We wish to express our gratitude to W. M. Y. Wan Mamat Saufi, Raja Muzaffar Shah and the foresters and rangers of the Hulu Selangor District Forest Office for the permission granted and full field assistance provided. We are grateful to A. Roslan (Selangor State Forest Department), Mokhtar, Mat Shah and Rosli (Gombak District Land Office), the GIS Units of Forestry Department Headquarters and FRIM for providing the necessary information. Special thanks are due to the Forest Engineering Team (Natural Forest Division, FRIM) for their technical assistance. Muhammad an e Le . L Thank. dS Fario t alse e dsar o du (FRIM) , Abd. Latiff Mohmod. 2 28 Journal of Tropical Forest Science 16(3): 271-282 (2004)

Shuko. M d ran (Universiti Kebangsaan Malaysia their )fo r constructive comments Damahuri. S o t . Chand C an . Mustap. ,Y AyauD . , K botanicae d th , aan l teao mwh undertook the field data collection. Support for research came from the Flora Malaysiana Centre Fund, Ministr Primarf yo y Industries, Malaysia under Projec: t6 Conservation Monitoring System for Threatened .

References

ASHTON, P. S. 1982. Dipterocarpaceae. Flora Malesiana. Series 1, 9: 237-552. BAWA, K. S. 1998. Conservation of Genetic Resources in the Dipterocarpaceae. Pp. 45-56 in Appanah S. & Turnbull, J. M. (Eds.) A Review of Dipterocarps—Taxonomy, Ecology and Silviculture. CIFOR and Forest Research Institute Malaysia, Kepong. BAWA, K S. & ASHTON, P. S. 1991. Conservation of rare trees in tropical rain forests: a genetic perspective. . 62-7 HolsingerPp & . 1Falkn i A . (Eds.. E D ,. K , ) Genetics Conservationd an of Rare Plants. Oxford University Press. BROWER, J. E., ZAR, J. H. & VON ENDE, C. N. 1997. Field and Laboratory Methods far General Ecology. 4th edition. WCB/McGraw-Hill, California. ELLIOTT, J. M. 1973. Some Methods for the Statistical Analysis of Samples of Benthic Invertebrates. Scientific Publication No. 25. Freshwater Biological Association, Ambleside, Westmorland. IUCN. 2001. IUCN Red List Categories. Version 3.1. Prepared by the IUCN Species Survival Commission. IUCN, Gland. , JONGO. , SANDS . , CHAN, SOEPADMOK. C , E. , T. . . KAURV , HA H ,ASHTON , & . A. , . E 1978, S . P , . Apomixis ma widespreae yb d among clima e treeth f so x rain forest. Nature (5644)1 27 : 440-442. MANOKARAN , LAFRANKIEN. , , KOCHUMMENV. . ,J , QUAH , KLAHNM. S. . , ASHTON. K , E ,HUBBELLE. & . ,. J S . P , , S. P. 1990. Methodology for the Fifty Hectare Research Plot at Pasoh Forest Reserve. Research Pamphlet No. 104. Forest Research Institute Malaysia, Kuala Lumpur. TANG. 1974& T . . P H . ,. NGS Comparativ . F , e growth rate f Malaysiaso n trees Malaysiane Th . Forester 37(1): 2-23. SOMEGO, M. 1978. Cytogenetical Study of Dipterocarpaceae. The Malaysian Forester 41 (4): 358-366. SYMINGTON . 1943F . C ,. Foresters' Manual of Dipterocarps. Malayan Forest Record. Syonan16 . sNo - Hakubutukan. Kuala Lumpur.