Species Composition and Environmental Factors, Including Human Impacts, on Coastal Sand-Dunes and Maritime Strand- Forests in Southern Thailand

TROPICS Vol. 14 (3) Issued March 15, 2005

Species composition and environmental factors, including human impacts, on coastal sand-dunes and maritime strand- forests in Southern Thailand

1) 2) Daisuke HAYASAKA and Kazue FUJIWARA

1) Graduate School of Environment and Information Sciences, Yokohama National University, Hodogaya-ku, Tokiwadai 79-7, Yokohama, 240-8501 JAPAN E-mail: [email protected] 2) Faculty of Environment and Information Sciences, Yokohama National University, Hodogaya-ku, Tokiwadai 79-7, Yokohama, 240-8501 JAPAN

ABSTRACT The relationship between species composition and environmental factors, including human activities, such as beach cleaning, mowing and trampling, in Southern Thailand’s maritime strand forests and sand dunes, -, was studied in order to assist in the development of a plan for the conservation of coastal plant species. Species composition at the study sites was studied using the phytosociological method (full-floristic inventory and total cover estimate of each species in quadrats). Quadrats are homogeneous representative samples of vegetation, varying from 9 to 100m2. The following environmental data were collected at the site: width of sand dune, beach type, beach management type (beach cleaning), seawater salinity, slope, aspect, micro-topography and the amount of discarded garbage. Based on these data, two results were found: 1. The species composition of the sites was shown to be similar by the Bray-Curtis method, based on phytosociological relevé samplings; and 2. The main factors determining species composition were wind speed, rainfall, soil texture, salinity and especially beach management practices, based on the CCA method. Most of the general coastal species use thalassochory (seed dispersal by sea currents) and are distributed from tropical to subtropical regions, including the Okinawa Islands of Japan. The diversity of coastal plants was sustained by beach management practices on abandoned beaches, such as garbage cleaning and mowing. There were many shrubs (microphanerophytes) and sub-canopy trees (mesophytes) as a result of burning and cutting. Regeneration of maritime strand-forest species could be seen in open areas that resulted from burning and cutting. The spectra of life forms (dormancy types) varied from site to site, due to varying environmental conditions and human activities.

Key words: beach management, Bray-Curtis method, canonical correspondence analysis (CCA), coastal sand-dune, maritime strand-forest, similarity, species composition, southern Thailand

INTRODUCTION Sites with water environmental conditions such as seacoasts, sand dunes, mangroves, swamps, marshes and rivers, and the organisms living in these habitats, are sensitive to environmental change (Omura et al., 1999). Coastal areas of Thailand are intensively exploited to sustain the current high level of economic activity of this country. Many researchers have investigated the distribution and classification of coastal sand-dune vegetation in temperate and boreal regions and in South America (Tatewaki & Ro, 1960; Chapman, 1964; Richmond & Muller-Dombois, 1972; Ohba et al., 1973; Morrison & Yarranton, 1974; Barbour et al., 1976; Moreno-Casasola & Espejel, 1986; Allen & Allen, 1988; Olff et al., 1993). The relationship between vegetation and micro-topography of sand dunes has been studied mainly by Nakanishi (1982), Nakanishi & Fukumoto (1987; 1993; 1994), Seelinger (1992), Olff et al. (1993), van der Veen et al. (1997) and Moreno- Casasola & Vazquez (1999). Although there were many conservation studies for river and marsh species (National Research Council, 1992; van der Valk et al., 1992; Koike et al., 2003), there are few studies for coastal sand-dune vegetation, especially in Southeast Asia. Although there are more species with a high extinction risk in marshes and the rivers than at the coast, coastal plant species are also sensitive to environmental changes. Hewett (1985) and Sasaki et al. (2002) reported that species diversity in sand dunes was influenced by grazing and soil compaction by mobile vehicles. In this paper we describe the relationship between species composition and environmental factors in Southern Thailand, along with human activities, such as beach cleaning, mowing and trampling. The goal of the study is to assist in the development of a plan to conserve coastal plant species. As the distance from the shoreline increases, the maritime influence lessens and the coastal sand-dune vegetation changes, depending on wind, salt spray and sand-dune creation. The important environmental factors of sand-dune vegetation are salt spray, depth of the water table, wave action and sand movement by wind (Randall, 1970; Nakanishi, 1982; Middleton & Schimpf, 1986; Moreno-Casasola, 1986; Nakanishi & Fukumoto, 1987; Barker et al., 1989; Martinez et al., 1997; Moreno-Casasola & Vazquez, 1999). These factors may act alone, or in combination with others, and impacts of each will vary according to local environmental conditions. Thus, differences in the community 246 Daisuke HAYASAKA and Kazue FUJIWARA and species composition of the sites relate to local variations in abiotic conditions, such as salt spray, sand movement, as well as human impacts.

MATERIALS AND METHODS The study was carried out in the Phuket Islands and Songkhla Province (Figure 1). The Phuket Islands are located between 8°07’ and 7°53’ N latitude and between 98°19’ and 98°24’E longitude; they face the Andaman Sea. In Phuket, there are two national-park beaches (Nai Yan and Mai Khao Beach), and two resort beaches (Kamala and Karon Beach). Songkhla is located at 7°12’N latitude and 100°03’E longitude, and faces the South China Sea. In Songkhla, there are a resort beach (Sai Kaeo Beach) and two recreation beaches for Thai citizens only (Chalatat and Samila Beach). Climatic data (temperature, relative humidity, sunshine duration, wind speed, rainfall) and Kira’s (1977) Warmth Index are shown in Table 1. The two regions have similar climate conditions, except for rainfall, which is between 2317 to 2504mm in Phuket, but only 2035mm in Songkhla. The dry season is from October to April, and the wet season is from May to September.

Japan China

Thailand Myanmar Cambodia

Phuket Islands Gulf of Thailand Nai Yan Beach Mai Khao Beach Karon Beach Songkhla Province Kamala Beach Chalatat Beach Andaman sea Samila Beach Sai Kaeo Beach Malaysia Fig. 1. Map of Thailand, South-east Asia, showing the study sites. Nai Yan and Mai Khao are in the northern Phuket Islands, Kamala is in central Phuket, and Karon is in southern Phuket. Samila, Chalatat and Sai Kaeo are in Songkhla Province.

Table 1. Climatic characteristics of the study sites (data for the period 1961-1990). Phuket Phuket Airport Songkhla Latitude 7 53N 8 07N 7 12N Longitude 98 24E 98 19E 100 03E

Temperature (°C) 28.1 27.4 27.9 Relative humidity (%) 75 80 77 Rainfall (mm) 2316.8 2503.6 2035.1 Warmth Index (WI) 277.6 268.7 274.9

※ For Nai Yan and Mai Khao Beach, data are from Phuket Airport. For Kamala and Karon Beach, data are from Phuket. For Chalatat, Samila and Sai Kaeo Beach, data are from Songkhla.

Species composition at the study sites was studied using the phytosociological method (Braun-Blanquet, 1964; Fujiwara, 1987), based on investigations between September 2003 and March 2004. The number of field relevés at the seven study sites was 338, ranging in size from 1 to 100m2. At the sites, the following data were collected: width of sand dune, distance from the shoreline, elevation, beach type, beach management type (e.g. beach cleaning), seawater salinity, slope, aspect, micro-topography and the amount of Species composition and environmental factors, including human impacts, on coastal sand-dunes and maritime strand-forests in Southern Thailand 247 discarded garbage (Table 1). Beach use types were established as follow: (1) natural-park beach, (2) recreation beach for natives and (3) resort beach for visitors. Rank variables for beach management activity, including cleaning seaweed and discarding garbage, and mowing, were as follows: (1) very little cleaning or no cleaning, (2) cleaning by manpower only, (3) cleaning by manpower and mowing, (4) beach cleaning by manpower, mowing and burning. A handheld refractometer (Bellingham＋Stanley Inc) was used to measure seawater salinity. Soil texture was classified as follows based on grain diameter: fine gravel, above 2.0mm in diameter; coarse sand from 2.0 to 0.2mm; fine sand, from 0.2 to 0.02mm; and silt, under 0.02mm. The amount of discarded garbage in each quadrat was classified as follows: (＋) under 1%, (1) under 5%, (2) under 25%, (3) under 50%, (4) under 75% and (5) above 75%. Species’ life forms were categorized into dormancy forms based on Raunkiaer (1934) and Numata (1947). Plants were classified as annuals (therophyte: Th), perennials (chamaephyte: Ch, hemicryptophyte: H, geophyte: G), dwarf shrubs (nanophanerophyte: N), shrubs (microphanerophyte: M), or sub-canopy trees (mesophyte: MM). The sites were shown to have similar species composition, based on the Bray-Curtis method (1957). To create a diagram we calculated the percentage of similarity of Sørensen (PS) and the percentage of distance (PD) and applied the equations to the Bray-Curtis method, where 2Σmin( xi, yi) PS = i Σ( xi+yi) i xi and yi represent the appearance ratio of each species (i) to the number of research quadrats on x and y study beaches; and PD = 100－PS The relationships between species composition and environmental factors, including human activities, were clarified using Canonical Correspondence Analysis (CCA) (ter Braak, 1986, 1994, 1995). Before the CCA was carried out, we excluded the correlated factors in order to reduce co-linearity. The independent variable was the total cover estimate of each species on each quadrat, and dependent variables were site conditions and environmental factors, including human impacts.

RESULTS Beach characteristics There are very few visitors in Nai Yan and especially in Mai Khao Beach, because of their inaccessibility so there is little beach management. On the other hand, many visitors come to Kamala Beach, and there are many structures, such as beach houses as well as beach chairs. Beach management at Karon Beach has been discontinued. Few people visit Chalatat and Samila Beaches, because they are recreation parks for Thai citizens. Beach cleaning, mowing and bonfire burning are performed at Chalatat and Samila Beaches. There is burning and cutting at Sai Kaeo Beach. Coastal erosion is extreme at Mai Khao and Sai Kaeo Beach. The sand dunes are narrow at Karon, Chalatat and Sai Kaeo Beaches (Table 2).

Similarity among sites, based on species composition and dormancy forms The similarity in site species composition (Bray-Curtis method) and the occurrence frequency ratio of species as a percentage of all quadrats among sites are shown in Figure 2 and Table 3. Chalatat and Samila Beach, and Nai Yan, Mai Khao and Karon Beach were similar to each other. Sai Kaeo and Kamala Beaches were quite different from the other beaches. The general coastal species on sand dunes in Southern Thailand were Casuarina equisetifolia, Terminalia catappa, Hibiscus tiliaceus, Premna obtusifolia, Ipomoea pes-caprae, Vigna marina, Thuarea involuta, Canavaria lineata, Lepturus repens and Cassytha filiformis. Common weed species were Panicum repens, Tridax procumbens, Achyranthes bidentata, Mimosa pudica, Dactyloctenium aegyptium, Borreria laevis, Eragrostis multicaulis, Cenchrus brownii and Vernonia cinerea. Regional coastal species in Songkhla were Vitex rotundifolia, Ipomoea storonifera, Spinifex littoreus, Cyperus polystachyos and Fimblistylis sericea. Pandanus odoratissimus, Scaevola sericea, Clerodendron inerme, Remirea maritima and Alysicarpus vaginalis were observed mainly in Phuket. No clear difference in weed species could be found between Phuket and Songkhla. Site-specific species were as follows: Tribulus terrestrial, Ruellia tuberose, Imperata cylindrica and Cynodon dactylon occurred only at Chalatat Beach. Many coastal or weed species were also found only at Samila Beach, including Hernandia nymphaeifolia, Caesalpinia bonduc, Excoecaria agallocha, Sesuvium portulacasrrum, Breynia rhamnoides, Sida acuta, Turnera ulmifolia and Cyperus flavidus. Only Sai Kaeo beach had Helicteres hirsute, Morinda citrifolia, Jasminum bifarium, Cassia glauca, Lippia nodiflora, Fimblistylis cymosa, Oxystelma esculentum, Ixeris sp., Leonurus sp. and Commelina paludosa. Erythrina indica, 248 Daisuke HAYASAKA and Kazue FUJIWARA

Calophyllum inophyllum, Vitex trifolia, Gloriosa superba, Euphorbia cyathopora, Melinis repens, Anisolemes indica,

－ Vinca rosea and an Acchimenes sp. occurred only at Kamala Beach. Crinum asiaticum var. japonicum, Paspalum dilatatum and Emilia sonchifolia appearred only at Nai Yan Beach. Mai Khao Beach had Solanum sp., Paederia scandens and Paspalum conjugatum. The dormancy-form patterns among the sites are shown in Figure 3. The ratios of therophytes, chamaephytes, hemicryptophytes, microphanerophytes and mesophytes were not significantly different among the sites. Mai Khao beach was, however, characterized by a low proportion of therophytes and a high proportion of chamaephytes, compared to the other sites. The proportion of beach nanophanerophytes was lower at 2527.0 2527.0 beaches with little or no management (Nai Yan, Mai Khao and Karon Beach) than at other beaches. At Phuket there was a higher proportion of geophytes than at Songkhla.

Relationship between species composition and

－ environmental factors mowing

＋ The effects of environmental factors and beach management practices on species composition, as determined by CCA, are shown in Figure 4. The following highly correlated factors were excluded from

mowing manpowerthe little management results: little management abandoned temperature, relative humidity, sunshine ＋

firing duration, warmth index (WI), distance from the shoreline,

＋ elevation, fine gravel rate, coarse sand rate and fine sand rate. Three axes were extracted by CCA analysis. The totals of the axes 1, 2 and 3 were 36.6%, 23.7% and 19.5%. Eigen values for axes 1, 2 and 3 were 0.555, 0.361 and 0.296, respectively. Correlations between axes and mowing

＋ six factors are shown in Table 4. With axis 1, seawater salinity and wind speed had a positive relationship but silt rate and rainfall had a negative relationship. Beach management type (pressure) showed a negative relationship with axis 2. There were no significant factors 0 0.009 0 0 0 0.02 0 53 68 51 59 46 43 29 3.5 2.9 3.5 3.2 3.2 3.2 3.1 3.43 3.4337.7 3.43 50.00 1.56 50.98 2.59 38.98 2.59 45.65 1.56 55.81 68.97 0.7770.2090.014 0.525 0.46 0.006 0.832 0.16 0.008 0.227 0.69 0.083 0.454 0.494 0.052 0.972 0.006 0.002 0.376 0.573 0.051 mowing 53-66 115-177 62-94 64-120 54-88 135-158 52-99 2626.1 2626.1 2626.1 for axis 3. ＋ Figure 4 indicates that the species composition at Chalatat Beach Samila Beach Sai Kaeo Beach Kamala Beach Nai Yan Beach Mai Khao Beach Karon Beach Sai Kaeo Beach was influenced by high seawater salinity and intensive beach management. Kamala Beach had high rainfall, high silt rate and weak wind. Chalatat and Samila Beaches had strong wind, low rainfall and low silt rate. Nai Yan, Mai Khao and Karon Beaches had little beach management and low seawater salinity.

DISCUSSION This is the first report on the relationship between Table 2. Beach characters and environmental factors including human impacts such as beach cleaning mowing at the sites. Wind speed (m/s) Salinity (%) Fine gravel rate (%) Coarse sand rate (%) Fine sand rate (%) Silt rate (%) Number of species Proportion of coastal species (%) Beach type Beach management type manpower recreation beach manpower recreation beach manpower resort resort national park Beach national park Beach resort Beach environmental conditions Width of sand dunes (m) Sunshine duration (hr.) Species composition and environmental factors, including human impacts, on coastal sand-dunes and maritime strand-forests in Southern Thailand 249

Table 3. The occurrence frequency ratio of coastal species as percentage of all quadrats in sites. Local Site Number of relevés 42 48 36 42 53 63 54 Nomenclature Dormancy form Chalatat Samila Sai Kaeo Kamala Nai Yan Mai Khao Karon Casuarina Casuarina equisetifolia equisetifolia MM 10.64 19.2 46.15 19.23 11.11 24.00 ･ Terminalia Terminalia catappa catappa MM ･ 4.12 8.33 7.14 30.19 7.94 7.40 Pongamia Pongamia pinnata pinnata MM 2.38 2.08 8.33 23.81 ･ 6.35 ･ Hibiscus Hibiscus tiliaceus tiliaceus M 4.76 8.33 25.00 ･ 11.32 ･ 3.70 Premna Premna obtusifolia obtusifolia M 11.90 ･ 23.08 30.95 ･ 19.05 ･ Ipomoea Ipomoea pes-caprae pes-caprae G 61.90 68.75 ･ 80.95 90.57 39.68 96.30 Thuarea Thuarea involuta involuta H 11.90 20.83 47.22 ･ 13.21 33.33 18.52 Lepturus Lepturus repens repens H ･ 2.08 8.33 4.76 52.83 20.63 1.85 Canavalia Canavalia lineata lineata Ch 30.95 50.00 ･ 7.14 79.25 79.37 50.00 Cassytha Cassytha filiformis filiformis Ch ･ 6.25 11.11 11.90 ･ 15.87 22.22 Vitex Vitex rotundifolia rotundifolia N 33.33 8.33 69.44 ････ Ipomoea Ipomoea stolonifera stolonifera G 19.04 18.75 47.22 ･･7.94 ･ Sporobolus Sporobolus virginicus virginicus H 21.43 20.83 8.33 ･ 3.77 ･･ Spinifex Spinifex littoreus littoreus H 40.48 37.50 77.78 ････ Fimbristylis Fimbristylis sericea sericea H 61.90 58.33 63.89 2.38 ･･1.85 Cyperus Cyperus polystachyos polystachyos Th 21.42 20.83 8.33 2.38 ･･･ Hedyotis Hedyotis sp. sp.1 1 Th 73.81 37.50 94.44 ･･･1.85 Sesuvium Sesuvium portulacasrrum portulacasrrum Ch 11.90 14.58 ･････ Derris Derris trifoliata trifoliata N 2.38 2.08 ･････ Chenopodium Chenopodium ficifolium ficifolium Th 2.38 2.08 ････ Zoysia Zoysia matrella matrella H 61.90 ･･11.90 ･･･ Tribulus Tribulus terrestris terrestris Th 23.81 4.12 ･････ Hernandia Hernandia nymphaeifolia nymphaeifolia MM ･ 4.12 ･････ Caesalpinia Caesalpinia bonduc bonduc MM ･ 4.12 ･････ Excoecaria Excoecaria agallocha agallocha M ･ 2.08 ･････ Croton Croton cascarilloides cascarilloides N ･ 2.08 ･････ Canavalia Canavalia cathartica cathartica H ･ 8.33 ･････ Helicteres Helicteres hirsuta hirsuta M ･･25.00 ･ 1.89 ･･ Morinda Morinda citrifolia citrifolia M ･･8.33 ････ Jasminum Jasminum bifarium bifarium N ･･16.67 ････ Cassia Cassia glauca glauca N ･ 2.08 25.00 ････ Lippia Lippia nodiflora nodiflora G ･ 4.12 47.22 ････ Fimbristylis Fimbristylis cymosa cymosa H ･ 11.11 ････ Pandanus Pandanus odoratissimus odoratissimus M ･･11.90 26.42 14.70 3.70 Scaevola Scaevola sericea sericea M ･･･16.67 32.08 28.57 1.85 Ischaemum Ischaemum muticum muticum M ･･･71.42 18.87 53.97 46.30 Clerodendron Clerodendron inerme inerme M ･･･4.76 ･ 23.81 1.85 Vigna Vigna marina marina N ･･･4.76 1.89 6.35 12.96 Cyperus Cyperus stoloniferus stoloniferus G ･ 2.08 ･ 30.95 22.64 ･･ Hedyotis Hedyotis sp. sp.2 2 G ････16.98 49.21 37.04 Ipomoea Ipomoea gracilis gracilis G ･･･23.81 18.87 23.81 ･ Remirea Remirea maritima maritima H ･ 6.25 ･･28.30 36.51 7.40 Alysicarpus Alysicarpus vaginalis vaginalis H 2.38 ･･35.71 13.21 ･ 12.96 Cuscuta Cuscuta chinensis chinensis Ch ･･2.08 4.76 1.89 23.81 ･ Erythrina Erythrina indica indica MM ･･･4.76 ･･･ Calophyllum Calophyllum inophyllum inophyllum MM ･･･4.76 ･･･ Vitex Vitex trifolia trifolia M ･･･23.81 ･･･ Gloriosa Gloriosa superba superba H ･･･4.76 ･･･ Wedelia Wedelia biflora biflora H 4.76 ･･2.38 49.06 53.97 ･ Euphorbia Euphorbia atoto atoto Ch ････15.09 57.14 ･ Crinum asiaticum var.var. japonicumjaponicum H ････11.32 ･･ etc.etc. 250 Daisuke HAYASAKA and Kazue FUJIWARA

Table 3. Continued The occurrence frequency ratio of weed, planted or ornamental species as percentage of all quadrats in sites. Local Site Number of relevés 42 48 36 42 53 63 54 Nomenclature Dormancy form Chalatat Samila Sai Kaeo Kamala Nai Yan Mai Khao Karon Lucaena glauca M 19.05 16.67 8.33 23.81 11.32 ・ 3.70 Panicum repens H 9.52 6.25 ･ 38.10 7.55 12.70 7.41 Tridax procumbens H 9.52 12.50 25.00 ･ 3.77 15.87 ･ Achyranthes bidentata bidentata Ch 7.14 4.17 25.00 30.95 5.66 7.94 ･ Mimosa pudica Ch 19.05 8.33 69.44 35.71 16.98 6.35 14.81 Dactyloctenium aegyptium aegyptium Th 9.52 2.08 47.22 7.14 3.77 6.35 ･ Borreria laevis Th 14.29 2.08 30.55 50.00 7.55 ･･ Phyllanthus simplexsimplex Th 4.78 6.25 8.33 16.67 1.89 3.70 Eragrostis multicaulismulticaulis Th 33.33 52.08 8.33 7.41 1.89 ･ 18.52 Cenchrus browniibrownii Th 9.52 4.17 ･ 26.19 ･ 7.94 3.70 Vernonia cinerea Th 11.90 4.17 8.33 23.81 ･･･ Calotropis giganteagigantea M 4.78 6.25 ･ 26.19 ･･･ Lantana camaracamara ver. ver. aculeata aculeata N 2.38 2.08 ･･3.77 ･･ Sonchus arvensisarvensis H 4.78 4.17 47.22 ･ 3.77 ･･ Eupatorium odoratum H 4.78 2.08 38.89 ･ 1.89 ･･ Momordica charanria Th 4.78 2.08 ･ 4.78 3.77 ･･ Digitaria adscendensadscendens Th 2.38 ･ 25.00 4.78 1.89 ･･ Stachytarpheta indica indica H 2.38 2.08 ･････ Chrysopogon aciculatus aciculatus H 4.78 6.25 ･････ Chloris barbata Th 4.78 4.17 ･････ Mukia maderaspatana Th 2.38 ･ 69.44 30.95 ･･3.70 Eleusine indicaindica Th ･ 2.08 8.33 4.78 ･･･ Lonicera japonica N ･･38.89 7.14 ･･･ Desmodium triflorumtriflorum H ･･38.89 11.90 ･･･ Oplismenus composituscompositus Ch 4.78 ･･････ Ruellia tuberosa G 2.38 ･･････ Imperata cylindricacylindrica G 4.78 ･･････ Cynodon dactylondactylon H 9.52 ･･････ Breynia rhamnoides N ･ 2.08 ･････ Sida acutaacuta N ･ 2.08 ･････ Sida sp.sp. N ･ 4.17 ･････ Turnera ulmifolia G ･ 2.08 ･････ Houttuynia cordata G ･ 4.17 ･････ Cyperus flavidus flavidus Th ･･25.00 ･ 3.77 ･･ Oxystelma esculentum esculentum G ･ 2.08 16.67 ････ Ixeris debilisdebilis H 2.38 ･ 47.22 ････ Fimbristylis complanatacomplanata H ･･52.78 ････ Commelina paludosa Th ･･52.78 ････ Euphorbia cyathopora N ･･･2.38 ･･･ Euphorbia geniculata N ･･･4.78 ･･･ Alocasia macrorrhiza G ･･･4.78 ･･･ Hewittia sublobatasublobata G ･･･4.78 ･･･ Ludwigia adscendens H ･･･23.81 ･･･ Melinis repens Th ･･･19.04 ･･･ Anisomeles indica indica Th ･･･11.90 1.89 ･･ Vinca rosea Th ･･･7.14 ･･･ Euphorbia hirta Th ･･･30.95 3.77 ･･ Cassia garrettiana －････16.98 ･･ Paspalum dilatatum H ････3.77 ･･ Emilia sonchifoliasonchifolia Th ･････6.35 3.70 Coccinea indicaindica G ･････7.94 ･ Paederia scandens Ch ･････9.52 ･ etc. Species composition and environmental factors, including human impacts, on coastal sand-dunes and maritime strand-forests in Southern Thailand 251

80 ▲ Chalatat － Samila ◆ Sai Kaeo ＊ Kamala 70 □ Nai Yan ● Mai Khao 60 ○ Karon 50 40 30 20

Score of the study sites based on Score species composition. 10 0 0 102030405060708090100 Score of the study sites based on species composition. Fig. 2. Ordination diagram for the similarity of the species composition of the study sites, based on PD and the Bray-Curtis method.

Table 4. Correlations between ordination axes and six environmental factors, including human impacts. Correlations (p<0.01) Axis1 Axis2 Axis3 Selected variances Width of sand dune (m) 0.421 -0.085 -0.135 Sea salinity (%) 0.734* -0.326 -0.222 Silt rate (%) -0.84* -0.51 -0.165 Wind speed (m/s) 0.817* 0.428 0.174 Rainfall (mm) -0.904* -0.207 0.363 Beach management pressure 0.566 -0.707* -0.383

100% Mesophyte

90% Microphanerophyte Nanophanerophyte 80% Geophyte 70% Hemicryptophyte 60% Chamaphyte Therophyte 50%

40% 1 Chalatat Beach 2 Samila Beach 30% 3 Sai Kaeo Beach 4 Kamala Beach 20% 5 Nai Yan Beach 6 Mai Khao Beach 10% 7 Karon Beach

Proportion of dormancyProportion forms in the study sites 0% 1234567 The number of study sites Fig. 3. Proportions of dormancy-forms at the sites. (see numbers on horizontal axis) species composition and environmental factors, including human activities, in Southern Thailand. Generally, coastal species, except for Casuarina equisetifolia and Lepturus repens, showed thalassochory (Table 3). Most of these species exist from tropical to subtropical regions, including the Okinawa Islands of south-western Japan. Ridley 252 Daisuke HAYASAKA and Kazue FUJIWARA

2 □ Chalatat ◆ Samila ▲ Sai Kaeo ○ Kamala 1.5 × Nai Yan ＊ Mai Khao ＋ Karon 1

0.5 wind speed

axis 2 0 rainfall seawater salinity silt rate -0.5 beach management pressure

-1

-1.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 axis 1 Fig. 4. CCA ordination diagram for the study sites and environmental variables including human activities (bold characters), based on total cover estimates of the plant species.

(1930) suggested that the origin of plants with thalassochory is not the continental coasts of South America or South Africa, but rather Asian regions where many islands exist. Ipomoea pes-caprae and Canavalia lineata reach Yamagata Prefecture in northern Japan, at 38°15’N latitude, but the population available to propagate is limited only to as far north as Kagoshima, southern Kyushu (Nakanishi, 1987). Crinum asiaticum var. japonicum reaches Chiba Prefecture in central Japan, at 35°44’N latitude, its northern limit (Koshimizu, 1938). On the other hand, although most regional species of the two regions (Phuket and Songkhla) also showed thalassochory, a significant distinction was found based on species composition. We concluded that some species were confined to either Phuket or Songkhla, because they face different seas, the Indian Ocean or Gulf of Thailand, respectively. Many maritime strand-forest species, which are sensitive to sand, damaged by strong wind and have high water requirments, occurred at Phuket, but could not settle at Songkhla (Table 1, 3). There were many weed species at Songkhla, but there were no significant distinction among the weeds, planted or ornamental species of the two regions. Hewett (1985) and Jungerius et al. (1995) stated that grazing and mowing were the most important factors changing floral species diversity in North Wales and the Netherlands. It was found that coastal species were significantly reduced by soil compaction due to trampling and off-road vehicles (Godfrey & Godfrey, 1980), and in Hokkaido, northern Japan, species diversity declined with decreasing plant cover (Matsushima et al., 2000). In Thailand, abandoned beaches with no beach cleaning or mowing, such as Nai Yan, Mai Khao and Karon, showed high proportions of coastal species. Many weed species, however, appeared on beaches with active beach management, i.e. at Chalatat and Kamala Beach (Table 1). Thus, it was found in this study that the diversity of coastal plants is enhanced by discontinuing beach cleaning and mowing. Table 3 shows species at beaches with very little or no management and much accumulated garbage (Nai Yan, Mai Khao and Karon Beach): Wedelia biflora, Euphorbia atoto, Paspalum dilatatum, Paspalum conjugatum and Solanum sp. This research suggests that Wedelia biflora and Euphorbia atoto are indicators of beach naturalness; the others grew on high-nutrient sites, because seaweeds and garbage acted as fertilizers. Some specific species in Kamala Beach such as Erythrina indica, Calophyllum inophyllum, Vitex trifolia and Gloriosa superba might be secondary plants due to beach cleaning, and they were sensitive to sand movement by strong winds. Other specific species of Chalatat and Samila Beaches such as Zoysia matrella, Tribulus terrestris, Hernandia nymphaeifolia, Caesalpinia bonduc, Excoecaria agallocha, Croton cascarilloides and Canavalia cathartica might be indicators for wind-beaten, dry conditions. There were many shrubs (microphanerophytes) and sub-canopy trees (mesophytes) at Sai Kaeo Beach, due to burning and cutting. It was suggested that regeneration of maritime strand-forest species could be seen in open areas created by burning and cutting. Therefore, it was concluded that most coastal plants were replaced by secondary vegetation due to the impact of human activities, such as beach cleaning and mowing, as was reported in other publications (Godfrey & Godfrey, 1980; Hewett, 1985; Jungerius et al., 1995). Figure 3 shows the wide life-form spectra (dormancy forms) among the sites. Regional patterns appeared for both dormancy type and species composition. We concluded that: 1. The nanophanerophyte proportion at Nai Yan, Mai Khao and Karon Beaches in Phuket was lower than at Samila, Chalatat and Sai Kaeo Beaches in Songkhla and at Kamala Beach in Species composition and environmental factors, including human impacts, on coastal sand-dunes and maritime strand-forests in Southern Thailand 253

Phuket, because the site was stable (anthropogenic influences were stopped). Thus nanophanerophytes will be replaced by microphanerophytes and mesophytes; and 2. The geophyte proportion at Phuket increased to a level comparable with Songkhla, because, although Phuket has lower rainfall than Songkhla (Table 1), geophyte species with dormant buds deep in the ground are able to access groundwater. The proportion of chamaephyte was highest at Mai Khao Beach, because human had impacted this area the least. Chamaephyte species with resistant buds above ground increase.

ACKNOWLEDGEMENTS We are grateful to Thawatchai Santisuk and Thawatchai Wongprasert for identifying plants in Thailand. We are indebted to Kunio Suzuki for study-site selection. We thank Takao Kikuchi for technical advice. We wish to thank Elgene O. Box and Akiko Sakai for their critical reviews of drafts of this manuscript.

REFERENCES Allen, E. B. & Allen, M. F. 1988. Facilitation of succession by the nonmycotrophic colonizer Salsola kali (Chenopodiaceae) on a harsh site: Effects of mycorrhizal fungi. American Journal of Botany 75: 257-266. Barbour, M. G., de Jong, T. M. & Johnson, A. F. 1976. Synecology of beach vegetation along the Pacific coast of the United States of America: a first approximation. Journal of Biogeography 3: 55-69. Barker, J. R., Herlocker, D. J. & Young, S. A. 1989. Vegetal dynamics in response to sand dune encroachment within the coastal grasslands of central Somalia. African Journal of Ecology 27: 277-282. Braun-Blsnquet, J. 1964. Pflanzensoziologie. 3 Aufl. Springer-Verlag, Wien. 865 pp. Bray, J. R. & Curtis, J. T. 1957. An ordination of the upland forest communities of southern Wisconsin. Ecological Monographs 27: 325-349. Chapman, V. J. 1964. Coastal vegetation. Pergamon Press, Oxford. 245pp. Fujiwara, K. 1987. Aims and methods of phytosociology or “vegetation science”. Plant ecology and taxonomy to the memory of Dr. Satoshi Nakanishi. Pp. 607-628. Kobe Geobotanical Society, Kobe Godfrey, P. J. & Godfrey, M. M. 1980. Ecological effects of off-road vehicles on Cape Cod. Oceanus 23: 56-66. Hewett, D. G. 1985. Grazing and mowing as management tools on dunes. Vegetatio 62: 441-447. Jungerius, P. D., Koehler, H., Kooijman, A. M., Mücher, H. J. & Graefe, U. 1995. Response of vegetation and soil ecosystem to mowing and sod removal in the coastal dunes ‘Zwanenwater’, The Netherlands. Journal of coastal conservation 1: 3- 16. Kira, T. 1977. A climatological interpretation of Japanese vegetation zones. Pp. 21-30 in Miyawaki, A. & Tüxen, Ｒ. (ed.). Vegetation science and environmental protection. Maruzen, Tokyo. 577 pp. Koike, F., Enomoto, T. & Shimada, N. 2002. Metapopulations of Utricularia bifida and Utricularia racemosa in spring-fed hillside marshes in the southern Tohoku. Japanese Journal of Conservation Ecology 8: 43-49 (in Japanese with English summary). Koshimizu, T. 1938. On the “Crinum Line” in the Flora of Japan. Botanical Magazine 52: 135-139. Martinez, M. L., Vazquez, G. & Sanchez, C. S. 2001. Spatial and temporal variability during primary succession on tropical coastal sand dunes. Journal of Vegetation Science 12: 361-372. Matsushima, H., Aikoh, T., Kondo, T. & Asakawa, S. 2000. Study on the change of beach plants cover area on Ishikari coast. Environmental Information Science 14: 295-300 (in Japanese with English summary). Middleton, B. A. & Schimph, D. J. 1986. Sand movement and vegetation in the Apostle Islands, Lake Superior. Canadian Journal of Botany 64: 1671-1674. Moreno-Casasola, P. 1986. Sand movement as a factor in the distribution of plant communities in a coastal dune system. Vegetatio 65: 67-76. Moreno-Casasola, P. & Espejel, I. 1986. Classification and ordination of coastal sand dune vegetation along the Gulf and Caribbean sea of Mexico. Vegetatio 66: 147-182. Moreno-Casasola, P. & Vazquez, G. 1999. The relationship between vegetation dynamics and water table in tropical dune slacks. Journal of Vegetation Science 10: 515-524. Morrison, R. G. & Yarranton, G. A. 1974. Vegetation heterogeneity during a primary dune succession. Canadian Journal of Botany 52: 397-410. Nakanishi, H. 1982. Coastal vegetation on the shingle spits of southwestern Japan. Phytocoenologia 10: 57-71. 254 Daisuke HAYASAKA and Kazue FUJIWARA

Nakanishi, H. 1987. Distribution and dispersal of Ipomoea pes-caprae and Canavalia lineate in Japanese mainland. Journal of Phytogeography & Taxonomy 35: 21-26 (in Japanese). Nakanishi, H. & Fukumoto, H. 1987. Coastal vegetation and topography in northern Hokkaido, Japan. Hikobia 10: 1-12. Nakanishi, H. & Fukumoto, H. 1993. Zonation of coastal vegetation and depositional topography of Noto Peninsula, central Honsyu, Japan. Journal of Phytogeography & Taxonomy 41: 95-101 (in Japanese with English summary). Nakanishi, H. & Fukumoto, H. 1994. Zonation of coastal vegetation and depositional topography in Aomori Prefecture, northern Honsyu, Japan. Hikobia 11: 575-586 (in Japanese with English summary). National Research Council. 1992. Restoration of Aquatic Ecosystems: Science, Technology, and Public Policy. National Academy Press, Washington D.C.. Numata, M. 1947. Ecological judgement of grassland condition and trend: I. Judgement by biological spectra. Grassland Science 11: 20-33 (in Japanese with English summary). Ohba, T., Miyawaki, A. & Tüxen, R. 1973. Pflanzengesellshaften der Japanischen Dünen-Küsten. Vegetatio 26: 1-143. Olff, H., Huisman, J. & VAN Tooren, B. F. 1993. Species dynamics and nutrient accumulation during early primary succession in coastal sand dunes. Journal of Ecology 81: 693-706. Omura, R., Muranaka, T., Michikawa, M. & Washitani, I. 1999. Vegetation developed on the dredged mud from Lake Kasumigaura. Japanese Journal of conservation Ecology 4: 1-19 (in Japanese with English summary). Randall, R. E. 1970. Vegetation and environment on the Barbados coast. Journal of Ecology 58: 135-172. Raunkiaer, C., Egerton, F. N., Gilbert, C. H. ＆ Fausboll. 1934. Life Forms of Plants and Statistical Plant Geography. Clarendon Press, Oxford. 632 pp. Richmond, T. DE A. & Muller-Dombois, D. 1972. Coastline ecosystem on Oahu, Hawaii. Vegetatio 25: 365-400. Ridley, H. N. 1930. The dispersal of plants throughout the world. L. Reeve, Ashford, Kent. Sasaki, S., Kondo, T. & Matsushima, H. 2002. The influence of vehicles running on the vegetation and soil in Ishikari coast, Hokkaido. Journal of the Japanese Society of Revegetation Technology 28: 342-352. Seelinger, U. 1992. Coastal fore-dunes of southern Brazil: Physiography, habitats and vegetation. Pp. 367-381 in Seelinger, U. (ed.). Coastal plant communities of Latin America. Academic Press, New York. 392 pp. ter Braak, C. J. F. 1986. Canonical correspondence analysis: A new eigenvector technique for multivariate direct gradient analysis. Ecology 67: 1167-1179. ter Braak, C. J. F. 1994. Canonical community ordination. Part I: Basic theory and linear methods. Ecoscience1 1: 127-140. ter Braak, C. J. F. 1995. Canonical correspondence analysis and related multivariate methods in aquatic ecology. Aquatic Sciences 57: 255-289. Tatewaki, M. & Ro. T. 1960. Dune and meadow communities along the Okhotsk Sea near Abasiri, Hokkaido, Japan. Biological Journal, Nara Women’s University 10: 84-90. van der Valk, A. G., Pederson, R. L. & Davis, C. B. 1992. Restoration and creation of freshwater wetlands using seed banks. Wetland Ecology and Management 1: 191-197. van der Veen, A., Grootjans, A. P., de Jong, J. & Rozema, J. 1997. Reconstruction of an interrupted primary beach plain succession using a Geographical Information System. Journal of Coastal Conservation 3: 71-78. Received 11st Oct. 2004 Accepted 10th Feb. 2005