Rend. Fis. Acc. Lincei s. 9, v. 2:79-85 (1991)
Ecologia. - - 3-diversity and phytogeographical patterns in the Ding H u Shan Reserve (Guangdong - South China) forest vegetation. Nota di Evar,a PIGNA-rTI, SANDRO PIGNAT'rI, C*-mNG-Cmv HUaNG, GUANG-QI DING e ZHONG-LL,a'~G HUANG, presentata (*) dal Corrisp. S. PIGNATrL
ABSTRACT. Some statistical features of the forest vegetation in the Ding Hu Shan Biosphere Reserve have been investigated;/~-diversitywith the procedure proposed by Itow and centralisation using an original index based on chorotypes frequence. The primary forest is characterized by high values of diversity and elevate centralisation, whereas secondary vegetation presents lower figures. -
-
KEY WORDS:Tropical forests; China; Diversity; Index of centralisation. RtASSUNTO.- - Diversita di tipo 3 e caratterifitogeografici nella riserva di Ding Hu Shan. Mcuni caratteri statistici della vegetazione forestale della Riserva della Biosfera di Ding Hu Shah nella Cina meridionale sono staff studiati: 3-diversit~l mediante l'applicazione del metodo proposto da Itow e centralizzazione mediante un indice originale. La foresta primaria ~ caratterizzara da valori elevati di diversit~l e di centralizzazione, mentre Ia foresta secondaria presenta valori inferiori.
[[NTRODUCTIO N
The floristic structure of the forest vegetation of the Ding H u Shan Biosphere Reserve has been described by Wang et al. (1982) and, on coenological basis, by Pignatti et al. (1990a) and its phytogeographical characters are discussed in a second Note. In this third (and last) contribution some statistical elaborations of the floristic and phytogeographical information are exposed. The present investigation deals with the problem ofeq-diversity and the treatment of information deriving from the patterns of geographical distribution in the forest vegetation. eq-diversity has been little investigated in subaerial ecosystems and data on tropical vegetation are particularly scarce. Difficulties in recognizing the species occurring in the study areas and high complexity of the vegetation may explain this situation. A further difficulty is given by the fact, that a clear method to evaluate ,8-diversity in subaerial ecosystems is still lacking. T w o different approaches have been proposed: the use of phytosociological material (e.g. Haeupler, 1982) or computations based on measures carried out in the field explicitly for this task. The first one has the advantage to dispose of a large set of available data but results are often erratic and reproducibility is scarce; the second one gives more accurate measures but comparisons are difficult because different Authors used different sampling methods. In the present study the second way was chosen, as an application of the measures carried out by different
(*) Nella seduta del 12 maggio 1990.
80
E, PIGNATTI ET AL.
Authors in tropical and subtropical forests of East Asia and South America (for an overview cf. Itow, 1988). MATERIALS AND METtIODS
Measures of diversity have been carried out in the same places and on the same vegetation stands used for phytosociological relev&: the procedure requires much work and it was possible to investigate only 6 stands of the 10 considered in the phytosociological study of the forest. In each place ca. 100 stems of plant individuals growing along a transect of 5 m width and variable length and belonging to the tree layer have been investigated; only stems at least 5-6 m high and thicker than 5 cm have been taken into cosideration; the number of stems present in the transect was determined for all occurring species. This procedure measures the frequence of arboreal individuals of each species occurring in the forest: each species enters with the number of trees observed along the gradient. The diversity index was obtained applying the Shannon-Weaver formula on these numerical data (program in Basic by Laura Celesti). The index of centralisation is inspired by the concepts proposed by Bertalanffy (1968). As a basis the results of the phytogeographical investigation of the Ding Hu Shan forest (Pignatti et al., 1990b) have been used. The geographical distribution of over 100 species occurring in the forest has been analyzed and the species divided into 4 groups: 1 Endemics; 2 South Eastern-Suboceanic; 3 South Eastern-Subcontinental; 4 Tropical Southeastasian. Species of groups 1 and 2 are restricted to South China, eventually with few localities in the neighboring countries, whereas groups 3 and 4 include species with wider range, covering most of the Asiatic South East. For all relevds the number of species belonging to each of the 4 chorotype groups was calculated. The sum of groups 1 + 2 was divided by the sum of groups 3 + 4. This procedure is original and the result indicates in what measure the flora is centered on the investigated area or is influenced by long range distributions. RESULTS A N D DISCUSSION
Values of the Shannon-Weaver index calculated for 6 stands of the Ding Hu Shan forest are exposed in tab. I (see also fig. 1). These values are regularly increasing from the secondary to the primary forest: the lowest figure is in rel. (1), which is in an extreme pioneer condition, whereas the highest value in rel. (7) corresponds to the climax-like condition of the dense forest in the vicinity of the climatic observation tower. The diversity index is a measure of disorder (high entropy) and low values correspond to a situation of order (negentropic condition). It depends on the number of elements considered and their relative frequency. It is low when elements are few and one of them is strongly dominant. Consequently the high values of the primary
,~-DI\rERSITY AND PHYTOGEOGRAPHICAL P A ~ ' q S
...
81
TABLE I. -- Number of stems observed for each species in the transect. rel.
Acronychia pedunculata Alchornea trewioides Antt~lesma japonicum Aporosa chinensis Aporosa yunnanensis Aquilaria sinensis Averrhoa carambola Calamus rhabdocladus Canarium album Canthium dicoccum Caryota ochlandra Castanopsis cbinenst~ Castanopsis fissa Cinnamomum camphora Craibiodendron kwantungense Cryptocarya concinna Cryptocarya chinensis Cyathea podophylla Diospyros morrisiana Engelhardta roxburghiana Erytbrophleum fordii Eurya groff~i Evodia lepza Ficus fulva Ficus nervosa Ficus sp. A Ficus rp. B Ficus vanolosa Fzksisttkma glaucescens Garcinia multt)qora Garcinta oblongifolia Gironniera subaequalis Glochidium cochincbinensls Ilex pubescens Ithea sinensis var. angustata Lz)zdera chunii Lindera metcalfiana Litsea cubeba Litsea rotundifoka .~lacaranga sampsonii Machilus kwantungensis Machilus chinensis Machilus velutina Melastoma sanguineum Microcos paniculata Microdesmis casearifolta Ormosia fordiana
1
2
5
17
6
7
5
2
10
1 2 5 1
1 2
1 1
8
1
9 1 2
1 1
1
11 7
35
I4
46
i
I 1 8 4
1 2 22
8
6 3
1 2 8 2
2 1 1 1 1 2 2 1 3 1
i
1 1 5 14
3 1 4
2
2 2 2 4 3 1 42
1
E. PIGNATTI ET AL.
82
TABLE I (continued). rel.
1
Photinia prunifolia Phithecellobium clypearia Pinus massoniana Ptero~permum lancaefolium Tetrastigma pIanicaule Randia canthioides Sapium discolor Sarcosperma laurinum Schima superba Schefflera octophylla Syzygium jambos Syzygium levinei Syzygium rehderianum Sterculia lanceolata Symplocos laurina unidentified unidentified unidentified unidentified
2
5
6
7
1
10
1 2 6 1 1 1
1
1 5
64 2
64 6
42 38
15
1
8
1 4
2 6
2
4 1
1 1
A B C C
1 I 2 1
No. of species Stems recortted (tot.)
6 117
10 103
16 104
17 106
24 105
26 102
Shannon-Weaver index
0.96
1.35
1.63
2.29
2.61
2.26
forest indicate a condition with very rich arboreal flora and without distinct dominants. On the contrary the transition to the secondary forest produces a reduction in the quantity of tree species and strong dominance of one or few species (in this case Schima superba together with Castanopsis fissa and Schefflera octophylla). The diversity of the Ding Hu Shan primary forest reaches very. high values, which apparently are typical for the tropical forest. Values obtained with the same procedure in temperate forests are in general inferior to 1.0 (orig. data, unpubl.), as a consequence of the reduced presence of tree species and dominance of one over the others (e.g. oak, beech, spruce). On tropical forests Itow (1984a, I984b, 1986) obtained very elevate values e.g. from Ponape and Kosrae 2.88 to 3.30, but probably using a different method (these measures, repeated with our computing procedure gave values from 1.99 to 2.28). In consequence, the values obtained in the Ding Hu Shan forest can be regarded as very high and indicate a situation of maximal diversity in vegetation. In the present case the high values seem to give a measure of the elevate complexity in the tropical forest. It is generally admitted, following the concepts of Mc Arthur and Odum, that systems with high entropy are stabler than negentropic ones and this appears, at least in the present case, a valid interpretation.
83
~-DIVERSITY AND P! IYTOGEOGRAP! IICAI. PATTERNS ...
Fig. 1. - Values of centralization (columns) and of diversity (line) in the Ding H u Shan vegetation by increasing age of the forest vegetation: secondary forest to the left, primary forest on the right side.
The values of centralisation have a similar distribution: figures are low in the secondary forest and higher in the primary one (see tab. II). This means that the secondary forest is characterized by the presence of dominant species, most of them with wide geographical range; the primary forest on the contrary is richer in local elements or endemics and a better equitability among the arboreal species exists. This T~U3L: II. - No. of species belonging to the 4 chorotypes.
tel. la tel. 1 tel. 2 rel. 3 rd. 4 rel. 5 tel. 6 rel. 7a rel. 7 tel. 8 rel. 9 tel. 10
I. Endemics
2. SE-Suboceanic
3. SE-Subcontinental
4. Tropical SE-Asian
Index of centr,
6 4 5 5 1 7 9 21 12 13 8 12
6 1 2 2 1 2 8 10 9 5 7 5
17 14 18 16 11 32 23 7 9 18 14 10
8 9 9 7 8 8 12 7 8 8 6 4
0.48 0.22 0.26 0.30 0.10 0.22 0.49 2.21 1.23 0.69 0.75 1.21
TaBLe III. -- Percentages of chorotypes m *)~dicatorgroups. Chorotype Indicators of the secondary forest Indicators of the primary forest Indifferent
1.0 8.2 38.9 18.6
2.0 6.7 27.2 2.2,
3.0 57.8 19.4
55.8
4.0 27.3 14.4 23.4
84
E. PIGNATTI ET AL.
result is confirmed if the species are disaggregated into ecological groups: indicators of the primary forest are mostly endemics, whereas among indicators of the secondary forest widespread species are prevailing (table III). The measures of centralization index indicate that the primary, undisturbed forest is richer than the secondary one in species with restricted geographical range. It can be assumed that with human impact widespread species have an advantage and become dominant. In the succession following, a drift from a ubiquitarious dominated flora to a more specialized one should be assumed. This is an example of secondary succession in consequence of heavy disturbance in the forest. In the Alpine belt in Europe primary succdssions show the opposite direction, i.e. from specialized, endemic-rich pioneer associations to climax associations with more widespread species. It seems that pioneer associations show the highest concentration of endemic species and secondary vegetation, as a consequence of impacts, the highest incidence of widespread species; climax vegetation presents the better correspondence with the floristic composition of the phytogeographical zone. BIOGEOGRAPI-tlC AND SYNTAXONOMIC RELATIONS
Little is known about the syntaxonomy of the subtropical forest vegetation of SEAsia: the Chinese literature is not based on phytosociological typology and the data exposed in the present study are insufficient for the definition (even at a tentative level) of higher syntaxa. In any case it appears that this vegetation is completely different from the evergreen vegetation of Japan and Eastern China and belongs to an other still undescribed vegetation class. It was pointed out in a previous investigation (Pignatti et al., 1990a), that the vegetation of the primary, forest at Ding Hu Shah is composed by a very. high number of species endemic in South China and in the surrounding regions. The only investigation based on comparable data deals with the vegetation of Hainan Island (Box et al., 1989): the montane forest at Jianfengling (890 m) has a very rich flora (109 species in a single relevd of 450m 2) mainly differing from the primary forest of Ding Hu Shan and characterized by the presence of Dac~.dium pierrei (Podocarpaceae)and several Fagaceae (Castanopsis, Cyclobalanopsis, Lithocarpus. Quercus). Despite of this, some floristic elements are common to Ding Hu Shan and Jiangfengling and may represent character species of higher syntaxa, e.g.: Schima
superba, Photinia serrulata, Canthium dicoccum, Pithecellobium clypearia, Calamus rhabdocladus, Ardisia quinquegona, Smilax corbularia, Psychotria rubra, Psychotria serpens. The secondary forest is more differentiated and possibly belongs to other vegetation types. CONCLUSION
The primary forest of Ding Hu Shan is characterized by high figures, both in diversity and in centralisation, i.e. tree species are very numerous but with relatively balanced frequence and most of them have a distributional range corresponding to the
~-D1VERSITY AND PHYTOGEOGRAPHICAL PATTERNS ...
85
phytogeographical zone including the forest. In the secondary forest on the contrary tree species are relatively few and mostly widespread; some of them have a marked tendency to become dominant in the tree layer. The forest is one of the very few relicts of a climax like vegetation formerly spread over a wider zone in South China; its flora appears the result of long term evolution under more or less stable conditions. The forest ecosystem develops under tropical climate, i.e. it is relatively protected against heavy environmental constraints such as aridity or frost and in consequence the forest reaches a very high degree of complexity. The vegetation of the primary forest cannot be included in any known vegetation class, and consequently one or more higher syntaxa are to be identified and characterized for the forest vegetation of South China. Research carried out with grants of C.N.R. ,~Gruppo Biologia Evoluzionistica,) and M.P.I. 40%. This paper is dedicated to Professor G. B. Marini Bettolo on the occasion of 75th birthday.
REFERENCES BERTALANF~ L. yon, 1968. General System Theory. Braziller ed., New York. Box E.-O., Fujiw,tRa K., HAO Y.-L., ZHONG YI, Fu Q.-H., Xtao B.-S., 1989. A tropical montane evergreen forest and other vegetation on Hainan Island, southern China. Bull. Inst. Envir. Sc. and Technol. Yokohama Nat. Univ., 16: 75-94. HAE'UVLER H., 1982. Evenness als Ausdruck der Vielfah in der Vegetation. Diss. Bot., 65. Cramer Verl., Vaduz. ITOW S., 1984a. Species diversity of Fagaceae-absent evergreen broadleafforests on three NW-Kyushu satellite islands. Jap. J. Ecol., 34: 225-228. ITOW S., ONO M., SEKI T., 1984b. Species diversity of subtropical evergreen broadleafforests on the Ryukyu and the Bonin Islands. Jap. J. Ecol., 34: 467-472. ITOW S., 1986. Species diversity in equatorial insular forests on Ponape and Kosrae, Micronesta. Ecol. Res., 1: 223-227. ITOW S., 1988. Species diversity of mainland- and island forests in the Pacific area. Vegetatio, 77: 193-200. PIGNAV'rI E., PmNAWI~S., HUANO C.-C., DtNG G.-Q., HU~,XGZ.-L., 1990a. Phytosociologi~alinvest(gations in the Ding Hu Shan forest (Guangdong - South China). Rend. Fis. Acc. Lincei, s. 9, 1:335-356. PIGNAT'rI E., PIGNAVTIS., HU^NG C.-C., DtNG G.-Q, HUANGZ.-L., 1990b. Chorologicalanalys:s of the Ding Hu Shan Flora. Rend. Fis. Acc. Lincei, s. 9, 1:465-472. WaNG Z.-H., HE D.-Q., SONG S.-D., CHEN S.-P., C~EN D.-R., Tu M.-Z., 1982. The vegetation cffDing Hu Shan Biosphere Reserve. Trop. and subtrop. Forest Ecosystem, 1: 77-156. G.-Q. Ding e Z.-L. Huang: Ding Hu Shan Nature Reserve ZHAO QING GUANGDONG (Cina) C.-C. Huang: South China Institute of Botany Academia Sinica - GWNGZHOU (Cina) E. Pignatti: Dipartimento di Biologia Vegetale Universit~l degli Studi di Trieste Piazzale Europa, 1 - 34100 TRmSTE S. Pignatti: Dipartimento di Biologia Vegetale Universita degli Studi di Roma ~*La Sapienza~ Piazzale A. Moro, 5 - 00185 ROMA