a quantitative analysis of plant community structure in an...

A quantitative analysis of plant community structure

in an abandoned rubber plantations on Kho-Hong Hill,

southern Thailand

Sara Bumrungsri1, Ekapong Sripao-raya2 and Charan Leelatiwong3

AbstractBumrungsri, S., Sripao-raya, E. and Leelatiwong, C.

A quantitative analysis of plant community structure in an abandoned rubber

plantations on Kho-Hong Hill, southern ThailandSongklanakarin J. Sci. Technol., 2006, 28(3) : 479-491

The present study aimed to characterize plant community structure of rubber plantations abandoned

26 years previously and now a protected area of Prince of Songkla University on the west slope of Kho-Hong

Hill. Trees whose girth at breast high were at least 30 cm were recorded from thirty-one plots (10*10 m) which

were laid out systematically at every 100 m along three transects. Among native trees, this plant community

is dominated by Schima wallichii Choisy, Castanopsis schefferiana Hance, Memecylon edule Roxb., Diospyrosfrutescens Blume, and Diplospora malaccensis Hook.f. Trees of the families Myrtaceae, Theaceae, Clusiaceae,

Fagaceae, and Rubiaceae, and saplings of Clusiaceae, Myrtaceae, Theaceae, Rubiaceae and Euphorbiaceae

were the most common. This plant community was characterized as a late seral stage post-cultivation

succession. The basal area of rubber trees was positively significantly related to the species richness of native

trees but negatively related to the density of native trees in each plot. Although abandoned rubber plantations

create environmental conditions which effectively catalyze forest succession, dense rubber trees could slow

succession of native trees by competition for resources. Further ecological, educational and recreational

studies are discussed. Zoning this area to be a strict nature reserve and a conservation area is recommended.

Key words : rubber plantation, forest succession, secondary forest, Kho-Hong Hill

1Ph.D.(Conservation Biology), 2M.Sc.(Zoology), 3M.Sc.(Botany), Department of Biology, Faculty of Science,

Prince of Songkla University, Hat Yai, Songkhla, 90112 Thailand.

Corresponding e-mail: [email protected]

Received, 26 July 2005 Accepted, 17 November 2005

ORIGINAL ARTICLE

Songklanakarin J. Sci. Technol.

Vol.28 No.3 May - Jun. 2006 480

Plant community in abandoned rubber plantation

Bumrungsri, S., et al.

∫∑§—¥¬àÕ

“√– ∫”√ÿß»√’ ‡Õ°æß»å »√’‡ª“√¬– ·≈– ®√—≠ ≈’≈µ‘«ß»å

‚§√ß √â“ß —ß§¡æ◊™„πæ◊Èπ∑’Ë «π¬“ßæ“√“∑‘Èß√â“ß∫π‡¢“§ÕÀß å ¿“§„µâ¢Õßª√–‡∑»‰∑¬

«. ß¢≈“π§√‘π∑√å «∑∑. 2549 28(3) : 479-491

°“√»÷°…“§√—Èßπ’È¡’«—µ∂ÿª√– ß§å‡æ◊ËÕ«‘‡§√“–Àå‚§√ß √â“ß —ß§¡æ◊™„πæ◊Èπ∑’Ë «π¬“ß∑’Ë∑‘Èß√â“ß¡“ 26 ªï∫√‘‡«≥‡™‘ß

‡¢“§ÕÀß å ´÷Ëß‡ªìπ‡¢µÕπÿ√—°…å∏√√¡™“µ‘¢Õß¡À“«‘∑¬“≈—¬ ß¢≈“π§√‘π∑√å ∑”°“√«“ß·ª≈ß¢π“¥ 10x10 ‡¡µ√ ∑ÿ°Ê

√–¬– 100 ‡¡µ√µ“¡‡ âπ·π« 3 ‡ âπ «—¥¢π“¥·≈–®”·π°™π‘¥µâπ‰¡â∑’Ë¡’¢π“¥‡ âπ√Õ∫«ß∑’Ë√–¥—∫Õ°µ—Èß·µà 30 ´¡.

¢÷Èπ‰ª æ∫«à“‰¡âæ◊Èπ‡¡◊Õß∑’Ë¡’§à“¥—™π’§«“¡ ”§—≠ Ÿß§◊Õ ¡—ßµ“π (Schima wallichii Choisy) °àÕ‡¢’È¬«À¡Ÿ (Castanopsis

schefferiana Hance) æ≈Õß‡À¡◊Õ¥ (Memecylon edule Roxb.) æ≈—∫°≈â«¬ (Diospyros frutescens Blume) ‡¡◊ËÕ

æ‘®“√≥“„π√–¥—∫«ß»åæ∫«à“ «ß»å∑’Ë‡¥àπ§◊Õ «ß»å Myrtaceae Theaceae Clusiaceae Fagaceae ·≈– Rubiaceae ¢≥–∑’Ë

≈Ÿ°‰¡â„π«ß»å∑’Ëæ∫¡“°§◊Õ«ß»å Clusiaceae Myrtaceae Theaceae Rubiaceae ·≈– Euphorbiaceae ªÉ“∑¥·∑ππ’Èπà“®–

Õ¬Ÿà„π√–¬–°“√∑¥·∑π¢—Èπ∑â“¬ ¢π“¥¢Õßæ◊Èπ∑’ËÀπâ“µ—¥√«¡¢Õß‰¡â¬“ßæ“√“¡’§«“¡ —¡æ—π∏å∑“ß∫«°°—∫§«“¡À≈“°™π‘¥

¢Õß‰¡âª√–®”∂‘Ëπ·µà¡’§«“¡ —¡æ—π∏å∑“ß≈∫°—∫§«“¡Àπ“·πàπ¢Õß‰¡âª√–®”∂‘Ëπ„π·ª≈ß»÷°…“ ∂÷ß·¡â«à“ «π¬“ß∑’Ë∑‘Èß

√â“ß®–¡’ ¿“æ·«¥≈âÕ¡∑’Ë™à«¬‡√àß°“√øóôπµ—«¢ÕßªÉ“µ“¡∏√√¡™“µ‘ µâπ¬“ßæ“√“∑’ËÀπ“·πàπÕ“®®–≈¥§«“¡‡√Á«¢Õß°“√

øóôπµ—«¢Õß‰¡âªÉ“‚¥¬°“√·°àß·¬àß∑√—æ¬“°√ ‡π◊ËÕß®“°¡’°“√„™âæ◊Èπ∑’Ëπ’È‡æ◊ËÕ°“√»÷°…“∑“ßπ‘‡«»«‘∑¬“ ‡ªìπ·À≈àß‡√’¬π√Ÿâ

·≈–‡æ◊ËÕ —π∑π“°“√ ®÷ß§«√·∫àß°“√„™âæ◊Èπ∑’Ë‡ªìπ‡¢µ ß«π·≈–‡¢µÕπÿ√—°…å

¿“§«‘™“™’««‘∑¬“ §≥–«‘∑¬“»“ µ√å ¡À“«‘∑¬“≈—¬ ß¢≈“π§√‘π∑√å Õ”‡¿ÕÀ“¥„À≠à ®—ßÀ«—¥ ß¢≈“ 90112

Throughout the tropics, enormous areas ofprimary forest have been converted to agriculturalland and this phenomenon is a major contributorfor biodiversity loss (Bryant et al., 1997). In someinstances, cultivated land is abandoned and replacedby secondary forest. Although it differs fromoriginal forest, this post-cultivation secondaryforest supports a number of forest taxa, even whenisolated from intact forest. Like other tropicalcountries, Thailand has lost a large portion of itsforest in the last few decades. Fifty three percentof the country's area was covered with forest in1961, but this had declined to 33.15% in 2000(Department of National Park, Wildlife and PlantConservation, 2003). Rubber or 'yang-pa-ra'(Heavea brasiliensis Willd. ex A. Juss.) plantations,the major cultivation practice in southern Thailand,replace large tracts of tropical forest. Rubberplantation increased from 1.58 million ha in 1978to 2.0 million ha in 1992. In that year, 25% of thearea of southern Thailand was covered with rubberplantations whereas only 18% was forest (RoyalForest Department, 1993).

Kho-Hong Hill, the largest isolated forestremnant close to the city of Hat Yai, was oncecovered with tropical forest, however, the majorityof the hill area (>50%) is currently covered byrubber plantations. During 1976-1980, Prince ofSongkla University bought some rubber plantations,and sixty hectares of them had no further disturb-ance or cultivation, so that secondary forest becameestablished. This secondary forest is now protectedas the university's nature reserve. This protectedforest is a catchment for water supply to theuniversity and has become increasingly used forresearch, education, recreation, and plant conserv-ation. An inventory carried out by Maxwell (1986)indicated relatively high diversity of at least 596species of vascular plants on this hill. Severalundergraduate research projects have been done inthe area, mainly on animals such as birds, termites,ants and canopy beetles (Charoensap, 1996;Ratcharote, 1999; Sritangnan, 1999; Wattanasit etal., 2005). Further research on forest regeneration,other ecological studies and nature interpretationare planned (Prince of Songkla University, 2002).


Vol.28 No.3 May - Jun. 2006


Bumrungsri, S., et al.481

However, the area is still lacking a quantitativedescription of the forest community. It is alsoimportant to describe the community structure ofabandoned rubber plantations of known age, sinceit may provide insights into the process of forestrecovery. In addition, it is interesting to establishthe extent to which a high density of planted rubbertrees affects the establishment of native plants. Theobjectives of the present study are to describe thecommunity structure of the secondary forest in anabandoned rubber plantation, and to determine therelationship between the density of rubber trees andthe density and diversity of native plants.

Study area

The present study was carried out in aprotected forest of Prince of Songkla University onKho-Hong Hill (07º 00.4' N, 100º 30.7' E.). Thisisolated hill lies in a North-South direction and issurrounded by settlements and agricultural land.Most of the study area (60 ha) was an abandonedrubber plantation mixed with a disturbed forest.These plantations had been abandoned since theywere bought by the university during 1976-80. Thestudy area is characterized with slight undulatingterrain ranging in altitude between 30 and 100 masl, situated on the western side of the hill. Thesummit rises to 371 m. Former rubber plantationscovered most of the lower elevations and flat areas.The 10 years climate was hot (mean minimum-mean maximum temperatures = 24.1-32.5ºC,average 28.3ºC) and relatively humid (averagerelative humidity of 72%) with 2,118 mm meanannual rainfall. Generally, there was a three monthdry season (February to April) with rainfall lessthan 100 mm, whereas rain was heavy in October-December (Kho-Hong Meteorically Station, 2004).This tropical climate is classified as tropical wetseasonal according to Walsh (1996). The soilseries belonged to the Ranong-Phato Association,subgroup typic paleudults. They are dark brownto brown over yellowish brown, very shallow/shallow to moderately deep, well drained andhave moderate permeability. Soils are poor withmoderately low organic matter (Soil Survey

Division, 1973). Bedrocks are heavily crushed andweathered carboniferous mudstones (Maxwell,1986).

Methods

Plots of 0.01 ha (10*10 m) were laid outsystematically at every 100 m along three transects(800-1200 m long) in a North-South direction(Figure 1). Transects were 150 m apart. When plotswere placed on slopes, their size was adjusted byx = 10tan φ where x is the adjusted length, and φis the degree of slope. Trees whose girth at breasthigh (1.3 m) (GBH) was at least 30 cm wereidentified and recorded for their GBH. Saplingsand shrub (GBH less than 30 cm but greater orequal to 5 cm) were identified and counted in a4*4 m subplot set at the south-west corner in eachplot. Voucher specimens of unknown species werecollected for later identification by comparingwith specimen in the Prince of Songkla Herbarium.Aspect and location of each plot were also noted.

A number of quantitative attributes werecalculated. These were density (D) (total numberof tree of species x/total sampling area), frequency(F) (number of plots with species x/ total numberof plots), dominance (basal area of species x). Thecontribution of each species to the forest communityin relation to other species was determined asrelative density (R

den) (density of species x/ density

of all species), relative frequency (Rf) (frequency

of species x/ sum frequency of all species), relativedominance (R

do) (basal area of species x/ basal area

of all species). The performance of each specieswas determined from the sum of relative contri-butions referred to as the Importance Value Index(IVI) when IVI = R

den + R

f + R

do.

A profile diagram of trees in a selected10x30 m plot was drawn and the composition andthe vertical forest structure were described.

To investigate the possible effect of rubbertrees on native trees and saplings, total rubber treebasal area and the number of rubber trees werecorrelated with the number of native saplingspecies, the number of native sapling individuals,the number of native tree species and number of


Vol.28 No.3 May - Jun. 2006 482



Figure 1. Distribution of sampling plots along thre transects in N-S direction over

the protected nature reserve of Prince of Songkla University.

native individual trees in each plot. All statisticalanalyses were carried out using SPSS (11.0).

Results

Quantitative analysis

Two hundred and thirty seven individualsof 49 tree species with GBH equal or larger than30 cm were measured in thirty-one 10*10m plots(0.31 ha). In addition, 249 individuals of 52 speciesof sapling and shrub were found in 31 plots of4*4 m (0.05 ha). From a total native tree speciesdensity of 529 trees/ha, Schima wallichii Choisy,Castanopsis schefferiana Hance, Memecylon eduleRoxb. and Diospyros frutescens Blume showedthe highest density from 35.5-54.8 trees/ha whilethe density of rubber tree was 235.5 trees/ha.Seventy percent of native species were limited in

distribution to only 1 or 2 plots. The four specieswith highest density and Diplospora malaccensisHook.f. also had the widest distribution amongnative species and contributed 19-29 % of all plots.The estimated total basal area of all native treeswas 11.14 m

2/ha while the basal area of rubber

trees is 7.87 m2/ha. The majority of trees (48%)

had GBH between 30 and 39 cm. The GBH of sixtrees from five species exceeded 100 cm. Thesewere Albizea splendens Miq. (200 cm), Syzygiumglaucum (King) P. Chantaranothai & J.Parn. (160cm), Aquilaria malaccensis Lamk. (157 cm),Castanopsis schefferiana Hance (143.5 cm), andrubber tree (128 and 105 cm). Outside the sampledplots, several trees were as large as 300 cm GBHsuch as Instia palembanica Miq., Sindoraechinocalyx Prain and Horsfieldia macrocomaWarb. In terms of summed basal area, Castanopsis





schefferiana Hance, Schima wallichii Choisy andAlbizea splendens Miq. were dominant. Inconclusion, Schima wallichii Choisy, Castanopsisschefferiana Hance, Diospyros frutescens Blume,Memecylon edule Roxb. and Diplosporamalaccensis Hook.f. had the highest IVI (30.5,28.67, 17.57, 17.42 and 14.26, respectively, Table1).

At the family level of native trees, Myrtaceae(12.1%), Theaceae (11%), Clusiaceae (9.7%),Fagaceae (9.4%), Ebenaceae, Leguminosae andRubiaceae (5%) had the highest IVI among the 25families recorded. These families contributed morethan 50% of all the IVI among native trees.

From a total sapling and shrub density of5,020 individuals/ha, Mesua kunstleri (King)Kosterm. (806 ind./ha), Syzygium zeylanicum (L.)DC. (362 ind./ha), Syzygium lineatum (DC.) Merr.& L.M.Perry, Syzygium cerasiforme (Blume) Merr.& L.M.Perry, Schima wallichii Choisy (282.3 ind./ha) had the highest density. Saplings of rubber treeswere also well established (282.3 ind./ha) (Table2). At the family level, sapling of Clusiaceae (12%)Theaceae (11%), Myrtaceae (9%), Ebenaceaeand Fagaceae (9%) and Memecylaceae (7%)dominated the studied area.

Description of the forest community structure

based on the profile diagram

Vertical stratification (Figure 2) of the forestwas not clearly defined. Three strata probablyoccured; top layer, middle layer and shrub layer.The top layer was 18-25 m high and was domin-ated by Castanopsis schefferiana Hance, Schimawallichii Choisy, Syzygium glaucum (King) P.Chantaranothai & J. Parn., Syzygium claviflorum(Roxb.) Wall. ex A.M. Cowan & Cowan, Syzygiumlineatum (DC.) Merr.& L.M.Perry,Vitex pinnataL., Sindora echinocalyx Prain, Cynometraramiflora L., Elaeocarpus floribundus Blume,Swintonia floribunda Griff. and Albizea spendensMiq.. When present, rubber trees occupied the toplayer. The middle layer was between 7-15 m highwhich was dominated by Garcinia nigrolineataPlanch. ex. T. Anderson, Garcinia merguensisWight, Diplospora malaccensis Hook.f., Diospyros

frutescens Blume, Crypteronia paniculata Blume,Memecylon edule Roxb., Calophyllum calaba L.,Psydrax nitida (Craib) K.M. Wong. A number ofshrubs including Mesua kunstleri (King) Kosterm.,Mallotus oblongifolius M˙ull. Arg., Macarangalowii King ex Hook.f., Ixora javanica DC.,Greenea corymbosa (Jack.) K. Schum., Crotonargyratus Blume, Memecylon edule Roxb.,Streblus taxoides (Heynes) Kurz and Prunus grisea(C. Muell.) Kalkman dominated the shrub layer(1-3 m heigh). Seedlings of canopy and subcanopytrees such as Hopea ferrea Laness., Aquilariamalaccensis Lamk., Garcinia spp., Diospyrosspp., and Canthium spp. were common. Woodyclimbers such as Spatholobus harmandii Gagnep.,Linostoma pauciflorum Griff., Ancistrocladustectorius (Lour.) Merr., and Artabotrys sp. werepresent. Rattan (Calamus spp.), Amomumuliginosum K.D. Koenig, Elettariopsis curtisiiBaker, Etlingera littoralis (A. Koenig) Giseke,Phrynium pubinerve Blume, Pandanus humilisLour., Molineria latifolia Herb. ex Kurz, Licualaspinosa Thunb., Homalomena sagittifolia Jungh.ex Schott and Alocasia lowii Hook. were also oftenfound in moist areas along temporary streams.Remnant large trees including Cynometra malac-censis L., Sindora echinocalyx Prain, Eleocarpusfloribunda Blume, Instia palembanica Miq. werepresent along the main stream. Creeping bamboo(Dinochloa scandens (Blume) Kuntze) was alsocommon, especially along hill-ridges. In tree gaps,seedlings of Macaranga spp. were common whilea number of early pioneer species such as Alstoniamacrophylla Wall. ex G. Don, Mallotus paniculatusM˙uell. Arg., Microcos tomentosa Sm., Erythro-xylum cuneatum (Miq.) Kurz, Peltophorumdasyrachis (Miq.) Kurz and Macaranga triloba(Reinw. ex Blume) M˙uell.Arg. were present alongdistinct forest edges. Snags were hardly found.

Although the forest covers a relatively smallarea (ca. 60 ha), a reconnaissance survey indicatedsome variations in dominant species in relation tophysical microhabitats. On hill-ridges which werecharacterized by relatively shallow soil overmudrock fragments, Castanopsis schefferianaHance, Crypteronia paniculata Blume, Syzygium


Vol.28 No.3 May - Jun. 2006 484


Bumrungsri, S., et al.T

ab

le 1

.Im

port

an

ce v

alu

e in

dex

of

nati

ve

tree

s w

hose

gir

th a

t b

reast

hig

h a

t le

ast

30 c

m fr

om

31 p

lots

(10*10 m

.) i

n a

ban

don

ed r

ub

ber

pla

nta

tion

s of

Pri

nce

of

Son

gk

la U

niv

ersi

ty.

S

cien

tifi

c n

am

esV

ern

acu

lar

Den

sity

Fre

qu

ency

Basa

l are

aR

elati

ve

Rel

ati

ve

Rel

ati

ve

Imp

ort

an

ce

nam

es(t

ree/

ha)

(m2/h

a)

Den

sity

Fre

qu

ency

Dom

inan

ceV

alu

e In

dex

Schi

ma

wal

lichi

i Cho

isy

Man

gtan

54.8

429

.03

1.24

10.3

79.

0011

.13

30.5

0C

asta

nops

is s

chef

feri

ana

Han

ceK

ow k

euw

mu

45.1

619

.35

1.57

8.54

6.00

14.1

328

.67

Mem

ecyl

on e

dule

Rox

b.Pl

ong

mue

at35

.48

19.3

50.

546.

716.

004.

8617

.57

Dio

spyr

os fr

utes

cens

Blu

me

Phla

p kl

uai

35.4

822

.58

0.41

6.71

7.00

3.71

17.4

2D

iplo

spor

a m

alac

cens

is H

ook.

f.22

.58

19.3

50.

444.

276.

004.

0014

.26

Syzy

gium

gla

ucum

(K

ing)

P.C

hant

aran

otha

i & J

.Par

n.D

aeng

12.9

06.

450.

952.

442.

008.

5212

.96

Gar

cini

a m

ergu

ensi

s W

ight

Nua

n25

.81

9.68

0.35

4.88

3.00

3.18

11.0

6M

icro

cos

tom

ento

sa S

m.

Phla

pphl

a3.

233.

231.

030.

611.

009.

2210

.83

Alb

izea

spl

ende

ns M

iq.

22.5

812

.90

0.28

4.27

4.00

2.48

10.7

5C

rypt

eron

ia p

anic

ulat

um B

lum

eK

a aa

m16

.13

12.9

00.

233.

054.

002.

099.

14A

cron

ychi

a pe

dunc

ulat

a (L

.) M

iq.

Ka

uam

19.3

59.

680.

273.

663.

002.

449.

10Si

ndor

a ec

hino

caly

x Pr

ain

Ma

kha

12.9

06.

450.

372.

442.

003.

377.

80Vi

tex

pinn

ata

L.

Tin

nok

3.23

3.23

0.63

0.61

1.00

5.68

7.29

Aqu

ilari

a m

alac

cens

is L

amk.

Mai

hom

12.9

09.

680.

162.

443.

001.

426.

86G

arci

nia

nigr

olin

eata

Pla

nch.

ex.

T. A

nder

son

Cha

mua

ng12

.90

6.45

0.20

2.44

2.00

1.80

6.24

Syzy

gium

gra

nde

(Wig

ht)

Wal

p.M

ao16

.13

3.23

0.17

3.05

1.00

1.55

5.60

Syzy

gium

cer

asifo

rme

(Blu

me)

Mer

r. &

L.M

.Per

ry9.

686.

450.

131.

832.

001.

195.

01C

alop

hyllu

m c

alab

a L

.9.

686.

450.

121.

832.

001.

114.

93P

runu

s gr

isea

(C

. Mue

ll.)

Kal

kman

Nuu

t ton

9.68

6.45

0.08

1.83

2.00

0.69

4.52

Syzy

gium

line

atum

(D

C.)

Mer

r. &

L.M

.Per

ryK

hwaa

t6.

456.

450.

121.

222.

001.

074.

29Sy

zygi

um z

elan

icum

(L

.) D

C.

Sam

et d

aeng

9.68

3.23

0.13

1.83

1.00

1.14

3.97

Syzy

gium

syz

ygio

ides

(M

iq.)

Mer

r. &

L.M

.Per

ry6.

456.

450.

081.

222.

000.

703.

92M

esua

kun

stle

ri (

Kin

g) K

oste

rm.

Bun

nnak

paa

9.68

3.23

0.12

1.83

1.00

1.04

3.87

Hop

ea fe

rrea

Pie

rre

Ta k

hian

hin

9.68

3.23

0.11

1.83

1.00

0.98

3.81

Ela

eoca

rpus

flor

ibun

dus

Blu

me

Kaa

lon

6.45

3.23

0.16

1.22

1.00

1.47

3.69

Unk

now

n 1

(Thy

mel

iace

ae)

6.45

6.45

0.05

1.22

2.00

0.46

3.68

Unk

now

n 2

6.45

6.45

0.05

1.22

2.00

0.44

3.66

Cyn

omet

ra m

alac

cens

is L

.M

ang

khaa

k9.

683.

230.

071.

831.

000.

643.

47C

haet

ocar

pus

cast

anoc

arpu

s T

hwai

tes

Khe

e no

n3.

233.

230.

190.

611.

001.

723.

33E

laeo

carp

us p

etio

latu

s W

all.

ex C

. Mue

ll.D

ee n

guu

6.45

3.23

0.10

1.22

1.00

0.93

3.15





Ta

ble

1.

(co

nti

nu

ed)

S

cien

tifi

c n

am

esV

ern

acu

lar

Den

sity

Fre

qu

ency

Basa

l are

aR

elati

ve

Rel

ati

ve

Rel

ati

ve

Imp

ort

an

ce

nam

es(t

ree/

ha)

(m2/h

a)

Den

sity

Fre

qu

ency

Dom

inan

ceV

alu

e In

dex

Dio

spyr

os s

umat

rana

Miq

.L

ak k

hoei

lak

klue

a6.

453.

230.

071.

221.

000.

592.

81C

rato

xylu

m c

ochi

nchi

nens

e B

lum

eT

iu k

liang

6.45

3.23

0.05

1.22

1.00

0.46

2.68

Kne

ma

glob

ular

ia W

arb.

Lue

at r

aet

3.23

3.23

0.10

0.61

1.00

0.89

2.50

Art

ocar

pus

elas

ticus

Rei

nw. e

x B

lum

eK

a ok

3.23

3.23

0.05

0.61

1.00

0.47

2.08

Mal

lotu

s pa

nicu

latu

s M

˙ ull.

Arg

.So

i daa

o3.

233.

230.

050.

611.

000.

472.

08E

ryth

roxy

lum

cun

eatu

m (

Miq

.) K

urz

Kra

i tho

ng3.

233.

230.

050.

611.

000.

412.

02P

ittos

poru

m fe

rrug

ineu

m A

it.3.

233.

230.

040.

611.

000.

371.

98U

nkno

wn

33.

233.

230.

040.

611.

000.

351.

96Sw

into

nia

schw

enck

ii Te

ijsm

. & B

inn.

Kaa

n th

ong

3.23

3.23

0.04

0.61

1.00

0.35

1.96

Dio

spyr

os s

p. 1

3.23

3.23

0.04

0.61

1.00

0.33

1.94

Mor

inda

elli

ptic

a R

idl.

Yo

thue

an3.

233.

230.

040.

611.

000.

321.

93M

orin

da s

pp.

Yo

3.23

3.23

0.02

0.61

1.00

0.21

1.82

Met

aden

ia tr

icho

tom

a B

akh.

f.K

ham

in to

n3.

233.

230.

030.

611.

000.

301.

91C

alle

rya

atro

purp

urea

(W

all.)

Sch

ot3.

233.

230.

030.

611.

000.

301.

91R

inor

ea a

ngui

fera

Kun

tze

Ngo

h pa

a3.

233.

230.

020.

611.

000.

221.

83C

leis

tant

hus

sum

atra

nus

M˙ ue

ll. A

rg.

Fin

daen

g3.

233.

230.

020.

611.

000.

211.

82G

arci

nia

hom

bron

iana

Pie

rre

Waa

3.23

3.23

0.02

0.61

1.00

0.21

1.82

Adi

nand

ra in

tege

rrim

a T.

And

ers.

ex

Dye

r3.

233.

230.

020.

611.

000.

211.

82U

nkno

wn

43.

233.

230.

020.

611.

000.

211.

82

Tota

l52

9.03

322.

5811

.14

100.

0010

0.00

100.

0030

0.00


Vol.28 No.3 May - Jun. 2006 486



Table 2. Sapling density.

Vernacular names Scientific names Density (tree/ha)

∫ÿππ“§ªÉ“ Mesua kunstleri 806.45‡ ¡Á¥·¥ß Syzygium zelanicum 362.90¬“ßæ“√“ Hevea brasiliensis 282.26¢«“¥ Syzygium lineatum 282.26≈“ßπ“ß Syzygium cerasiforme 282.26¡—ßµ“π Schima wallichii 282.26®‘°π¡ Barringtonia macrostachya 221.77æ≈Õß‡À¡◊Õ¥ Memecylon edule 201.61µ—ßÀπ Calophyllum calaba 181.45¢’ÈÀπÕπ Chaetocarpus castanocarpus 161.29‚∑– Rhodomyrtus tomentosum 141.13°–Õ«¡ Acronychia pedunculata 120.97‡¢Á¡∑Õß Ixora javanica 120.97π«≈ Garcinia merguensis 100.81æ≈—∫°≈â«¬ Diospyros frutescens 80.65Tarenna sp. Tarenna sp. 80.65Memecylon sp. 1 80.65µ–‡§’¬πÀ‘π Hopea ferrea 80.65™–¡«ß Garcinia nigrolineata 60.48æ≈—∫æ≈“ Microcos tomentosa 60.48°â“π∑Õß Swingtonia schwenckii 60.48Unknown 1 Thymeliaceae 60.48°“≈π Elaeocarpus floribundus 60.48µ–‰À≈ Prismatomeris tetrandra 60.48¢àÕ¬Àπ“¡ Streblus ilicifolius 60.48æ‘°ÿ≈ªÉ“ Adinandra integerrima 60.48

Mallotus lowii 40.32¡–‰ø°“ Baccaurea parviflora 40.32°àÕ‡¢’È¬«À¡Ÿ Castanopsis schefferiana 40.32Leguminosae 40.32¢¡‘Èπµâπ Metadina trichotoma 40.32¡«°°Õ Olea salicifolia 40.32Ωîòπ·¥ß Cleistanthus sumatranus 20.16Diplospora sp 20.16¡Õ°¡—π Glocidion superbum 20.16°“·´– Callerya atropurpurea 20.16 â“π„À≠à Dillenia obovata 20.16°–Õ“¡ Crypteronia paniculatum 20.16°–ÕÕ° Artocarpus elasticus 20.16°–∑—ß„∫„À≠à Litsia grandis 20.16Unknown 2 20.16Unknown 3 20.16Unknown 6 20.16°√–¥Ÿ°§à“ß Aporosa aurea 20.16





Figure 2. Profile diagram and vertical stratification of trees in a 10*30 m plot.

Table 2. (Continued)

Vernacular names Scientific names Density (tree/ha)

À“¥√ÿ¡ Artocarpus dadah 20.16µ’ππ° Vitex pinnata 20.16°“·√âßÀ‘π Koilodepas longifolium 20.16¡–®È”°âÕß Ardisia colorata 20.16À«â“À‘π Syzygium claviflorum 20.16Ficus sp 20.16À“¥ Artocarpus lacucha 20.16πŸ¥µâπ Prunus grisea 20.16Tarenna sp. 20.16

total 5020.16


Vol.28 No.3 May - Jun. 2006 488



lineatum (DC.) Merr. & L.M. Perry, Syzygiumglaucum (King) P.Chantaranothai & J.Parn., Ixorajavanica DC. dominated. This microhabitat alsoseems to be more pristine than the others. On slightslopes and relatively flat areas, where moreorganic matter and sediment accumulated, Garciniaspp., Diospyros spp., Memecylon edule Roxb. andArtocarpus elasticus Reinw. ex Blume werecommon. This community covered most of thestudy area. In a small area of very shallow soilover rock outcrops (ca. 1 ha), a few plant speciessuch as Syzygium zeylanicum (L.) DC., Schimawallichii Choisy, Rhodomyrtus tomentosa (Aiton)Hassk., Calophyllum calaba L. and a parasiticplant, Dendrotrophe buxifolia (Blume) Miq.,sparsely covered the area, and canopy trees wererelatively low (< 8 m.).

Correlation of rubber tree density and commu-

nity characters of native plants.

Although the density of rubber trees in eachplot was not significantly correlated to the numberof sapling species (r = 0.02, p = 0.93), the densityof saplings (r = -0.16, p = 0.47), the number of treespecies (r = 0.03, p = 0.91) or the density of nativetrees (-0.24, p = 0.27) in plots, the basal area ofrubber trees was positively significantly correlatedto the number of native species (r = 0.57, p = 0.002,1 tailed), and partially negatively significantlyrelated to the density of native trees in each plot(r = -0.75, p<0.001, 1 tailed, controling for numberof all trees).

Discussion

In post-cultivation forest succession, floristiccomposition, structural characters and rate ofsuccession reflect land-use history, soil character-istics and proximity to old-growth forest (Tan-thana, 1988; Grau et al., 1997; DeWalt et al., 2003).Tanthana (1998) compared successional plantcommunities in different ages (up to 10 years-old)of abandoned rubber plantations at Ton Nga ChangWildlife Sanctuary, and revealed that native treedensity increases with time, although at a longertime scale (50 years) the opposite is true (Grau et

al., 1997). DeWalt et al. (2003) also pointed outthat the similarity between secondary forest and oldgrowth forest in tree species composition increaseswith forest age. Trees of Euphorbiaceae, Moraceae,Lauraceae, Sterculiaceae were dominant, andseedling, sapling and sprouting of Rubiaceaeand Euphorbiaceae were the most common inabandoned rubber plantations in a wildlife sanc-tuary which was 26 km from the study area. Thevariation in major families between the presentstudy and that of Tanthana (1998) may reflectecological variation especially proximity topropagule resources and soil conditions. In primaryforest of Ton Nga Chang Wildlife Sanctuary,Sawangchote (1998) studied ten 0.1 ha standsand found that trees of Dipterocarpaceae, Euphor-biaceae, Annonaceae, Sapindaceae were the mostdominant, respectively. Although soil at Ton NgaChang and Koh-Hong Hill are both in Ranong-Plato series, soil in the latter may be drier, as it ismuch shallower, with more weathered mudrockparticles and likely more drainage (S. Bumrungsri,pers. obs.). Dominance of Schima wallichii Choisy,Castanopsis schefferiana Hance, Syzygium spp.and Mesua kunstleri (King) Kosterm. which areoften found in dry habitats, suggests that Kho HongHill is drier. Syzygium spp. is also well dispersedin coastal sandbars communities in SongkhlaProvince which normally experience droughtconditions during the dry season (Sridith, 2002).Similarly, Maxwell (1986) also pointed out thatthis forest is drier, and postulated that this wascaused by continuous disturbance.

It is suggested that in a tropical moist climate,secondary succession proceeds rapidly with someecosystem characteristics such as leaf litter mass,soil nutrient processes differing little from valuesof old growth forest over 40-50 years (Grau et al.,1997; DeWalt et al., 2003). It would be interestingto compare between the structure of this 26-year-old secondary forest with that of patches of primaryforest that are present in military bases. DeWaltet al. (2003) found 21% similarity in foreststructure between 20 year-old secondary forestand old growth stands in moist neotropical foreston Barro Colorado Island. They also found that





it takes at least 70 years for secondary forest toresemble old growth stand structure. However, inthe less fertile soils of the Rio Negro Basin ofVenezuela, it was suggested to take up to 190 years(Saldarriaga et al., 1988).

Grau et al. (1997) suggested that pre-abandonment vegetation influences forest com-position for several decades. Specifically, forestrecovery in former rubber plantations differs fromother kinds of cultivation, as farmers normallyweed without burning in rubber plantations. Thus,when they were abandoned, rubber plantations stillhouse some plant propagules. Furthermore, severalstudies have indicated that tree plantations arean effective tool for catalyzing forest successionby overcoming barriers to natural regeneration(review by Parrotta, 2000). This is due to theirinfluence on understorey microclimate, develop-ment of litter, humus accumulation and structuralcomplexity of the vegetation. Apart from litterquantity, soil surface microclimate and shading inplantations is also responsible for biomass of thesoil fauna which plays a vital role in tropical forestas soil modifiers and nutrient cyclers. Thus, furtherinvestigation on how much litter fall and how fastthe decomposition rate there is in this 26-year-oldabandoned plantation is recommended.

Richards (1996) recognized four stages ofpost-cultivation succession; invasion, early seral,late seral, and mature. From his description, thepresent community could be assigned to the lateseral stage. In this stage, which is comparable withthe building phase of a primary forest gap, theforest becomes mixed with late seral species (long-lived and slow growing pioneers) and primaryforest species. Cryptononia paniculata (Blume),Vitex pinnata L., Acronychia pedunculata (L.)Miq., Artocarpus elasticus Rienw. Ex Blume areexamples of late seral species in this area, whereasspecies in the genera Memecylon, Diospyros,Syzygium, Garcinia, and Hopea are often foundin primary forest.

Correlation of rubber tree density and commu-

nity characters of native plants.

The negative correlation between rubber

tree biomass and native tree density may be due tothe fact that although abandoned rubber plantationcreate suitable environmental conditions for nativetree establishment, rubber trees are likely to becompetitors for light and soil water. Althoughrubber trees probably do not have allelopatric effecton nearby plants (Tanthana, 1998), this negativecorrelation suggests that, at a certain stage ofdevelopment, these exotic tropical trees whichperform well in the wet tropics can slow success-ion of native species probably by resource com-petition. A number of studies indicated negativeeffects of invasions of exotics on native commu-nities (review by Levine et al., 2003), and com-petition is by far the most common explanation.An empirical study on whether rubber trees havea negative impact on plant community structureand the underlying mechanisms of such impact isneeded. At present, rubber trees are well distributedin the study area (70% of plots), and they con-tributed 30-40% of density and basal area of treeswith a diameter larger than 10 cm. Furthermore,rubber saplings are among those which have thehighest density, although they accounted for only5.6 percent of total saplings. Alternatively, thepositive correlation between the basal area ofrubber trees and the species diversity of nativetree was significant. It is possible that in siteswhere the basal area of rubber trees is larger, theenvironmental conditions (e.g. microclimate, litterquantity, soil fertility) are also more suitable forthe establishment of a number of primary forestspecies as well.

Further studies and management recommend-

ation

Considering the proximity of the area to auniversity, this forest provides excellent opport-unities for ecological, educational and recreationalactivities. A GIS map of its boundary and thedifferent habitats on Kho Hong Hill based onupdated aerial photographs, together with envi-ronmental factors such as soil type and rock parentmaterial should be made. For botanical studies,the flora of Kho Hong Hill should be updated,since Maxwell (1986) carried out his intensive


Vol.28 No.3 May - Jun. 2006 490



survey of the flora twenty years ago. For furtherecological studies, a database and a field guidebased on morphological characters of each speciesare strongly recommended. This is an essentialfirst step and further studies on plant phenology,plant ecology, forest succession, and plant-animalinteraction could be undertaken. Permanent plotsshould be established in representative areas ofthis forest and intensive studies on forest success-ion should be carried out. These plots could bespecifically research areas for undergraduate andpostgraduate studies under the supervision ofacademic committees.

It is suggested that secondary forest canprovide vital resources for nesting, foraging andprotection for a number of animal taxa. DeWaltet al. (2003) indicated that understory fleshy fruitavailability was highest in young secondary forest(20 years old) compared to older secondary andold growth forest. Thus, secondary forest at thisage may provide good food resources for wildanimals especially generalist frugivore/insectivorebirds and mammals. Several species of fruigovorusbirds e.g. bulbul (Pycnonotus spp.) are common inforest habitats (Charoensap, 1996; S. Bumrungsri,pers. obs.). Several studies have indicated thatavian diversity in secondary forest is as high as itis in primary forest, although endemic species arefewer (Estrada et al., 1997; Waltert et al., 2004).As fauna is also important for forest recovery, aninvestigation of faunal diversity, population andhabitat use is needed, especially for those animalsthat are pollinators, seed dispersers, and de-composers. This forest patch is actually isolatedfrom nearby patches by current rubber plantationsand may suffer from small population effects.Conservation biology-based practices such asrestocking of plants and animals, natural corridorestablishment and monitoring could be imple-mented. In addition, environmental education andecotourism could be potentially applied in thisforest, especially for those school students fromHat Yai city. Forest nature trails with informationprovided could be constructed in different micro-habitats including a temporary waterfall (8 mhigh). However, without any regulations, these

activities could possibly upset the area and coulddisturb ecological studies. Zoning of this protectedarea to 1.) strict nature reserve (for research only)and 2.) conservation area (education and recreat-ion uses) is recommended. The particularly richflora along the main stream (Maxwell, 1986) isparticularly in need of protection.

Acknowledgements

The authors thank N. Noon-anan, S. In-tamusik, S. La-ma, P. Chaochiangtung and severalstudents of Department of Biology for their helpduring field work as part of the Tropical ForestEcology course. Kho Hong Meteorical Stationkindly provided climate information. Thanks arealso due to the Landscape and Building section ofPrince of Songkla University for providing mapsand information about site history, and to AssistantProfessor Dr. Kittichate Sridith, Professor PaulRacey and Dr. Prakart Sawangchote for theirdiscussion. Three anonymous reviewers improvedthe original manuscript.

References

Bryant D., Nielsen, D. and Tampley, L. 1997. The LastFrontier Forest: Ecosystems and Economics onthe Edge. World Resources Institute, Washington,D.C.

Charoensap, P. 1997. Species Diversity and HabitatDistribution of Birds in Prince of Songkla Uni-versity, Hat Yai Campus. BSc Thesis. Depart-ment of Biology, Prince of Songkla University.(in Thai)

Department of National Parks, Wildlife and PlantConservation. 2003. Forestry Statistical Data.http://www.dnp.go.th/stat/stat46/stat2546.htm(4/01/05).

DeWalt, S.J., Maliakal, S.K. and Denslow, J.S. 2003.Changes in vegetation structure and compositionalong a tropical forest chronosequence: implic-ations for wildlife. Forest Ecol. Manag., 182:139-151.

Estrada, A., Coates-Estrada, R. and Meritt, D. Jr. 1997.Anthropogenic landscape changes and aviandiversity at Los Tuxtlas, Mexico. Biodivers.





Conserv., 6:19-43.Grau, H.R., Arturi, M.F., Brown, A.D. and Acenolaza,

P. G. 1997. Floristic and structural patterns alonga chronosequence of secondary forest successionin Argentinean subtropical montane forests.Forest Ecol. Manag., 95: 161-171.

Kho-Hong Meteorically Station. 2004. Climate recordsat Hat Yai. Unpublished report.

Levine, J.M., Vila, M., D'Antonio, C.M., Dukes, J.S.,Grigulis, K. and Lavorel, S. 2003. Mechanismsunderlying the impacts of exotic plant invasions.Proc. R. Soc. Lond. B, 270: 775-781.

Maxwell, J.F. 1986. The Flora of Kor-hong Hill,Hat Yai District, Songkhla Province, Thailand.Department of Biology, Prince of SongklaUniversity. Songkhla.

Parrotta, J.A. 2000. Catalysing natural forest restorationon degraded tropical landscapes. Pp. 45-56. inEilliott, S.J., Kerby, J., Blakesley, D., Hardwick,K., Woods, K. and Anusarnsunthorn, V. (eds.).Forest Restoration for Wldlife Conservation.International Tropical Timber Organisation andthe Forest restoration Research Unit, Chiang MaiUniversity, Thailand.

Prince of Songkla University. 2002. Master Plan forPlant Genetic Conservation Project under HerPrincess Siridhorn. Unpublished report. (in Thai)

Ratcharote, S. 1999. Survey of Mound Building Termitesat Prince of Songkla University, Hat Yai Campus.BSc Thesis. Department of Biology, Prince ofSongkla University. (in Thai)

Richards, P.W. 1996. The Tropical Rain Forest: anEcological Study. Second Edition with contri-butions by R.P.D. Walsh, I.C. Baillie, and P.Greig- Smith. Cambridge University Press,Cambridge.

Royal Forest Department. 1993. 1993 Annual Report.Royal Forest Department. Bangkok.

Saldarriaga, J.G., West, D.C., Tharp, M.L. and Uhl, C.1988. Long-term chronosequence of forestsuccession in the upper Rio Negro of Columbiaand Venezuela. J. Ecol., 76: 938-958.

Sawangchote, P. 1998. Phytosociological Structure ofLower Tropical Rain Forest at Ton Nga ChangWildlife Sanctuary, Songkhla Province. M.Sc.Thesis, Prince of Songkla University. 171 pp.(in Thai)

Soil Survey Division. 1973. Detailed ReconnaissanceSoil Map of Songkhla Province. Department ofland development. Ministry of Agriculture andCooperation.

Sridith, K. 2002. The remnant of vegetation on thecoastal sandbar in Songkla Province, PeninsularThailand. Thai. For. Bull., 30: 49-58.

Sritangnan, Y. 1999. Plant Community Structures in anAbandoned Rubber Plantation, Rubber CuttingArea and Secondary Natural Forest. B.Sc. Thesis.Department of Biology, Prince of SongklaUniversity. (in Thai)

Tanthana, N. 1998. Natural Reforestation of AbandonedRubber Plantations. M.Sc. Thesis, Prince ofSongkla University. (in Thai)

Walsh, R.P.D. 1996. Climate. Pp. 159-205 in Richards,P.W. 1996. The Tropical Rain Forest: an Eco-logical Study. Second Edition with contributionsby R.P.D. Walsh, I.C. Baillie, and P. Greig-Smith. Cambridge University Press, Cambridge.

Waltert, M., Mardiastuti, A. and Muhlenberg, M. 2004.Effect of land use on bird species richness inSulawesi, Indonesia. Conserv. Biol., 18: 1339-1346.

Watanasit, S., Noon-anant, N. and Binima, N. 2005.Preliminary survey of ants at a reserve area ofPrince of Songkla University. SongklanakarindJ. Sci. Technol., 27: 39-46.

a quantitative analysis of plant community structure in an...

Documents