Urban green space, street tree and heritage large tree assessment in Bangkok, Thailand

ARTICLE IN PRESS

Urban Forestry & Urban Greening 7 (2008) 219–229 www.elsevier.de/ufug

Urban green space, street tree and heritage large tree assessment in Bangkok, Thailand Bunvong Thaiutsaa, Ladawan Puangchita, Roger Kjelgrenb,, Wanchai Arunpraparuta a

Faculty of Forestry, Kasetsart University, Bangkok 10900, Thailand Department of Plants, Soils, and Biometeorology, Utah State University, Logan, UT 84322, USA

b

Abstract Three inventories were conducted to quantify Bangkok’s green infrastructure for future planning and improvement in the context of a seasonal monsoonal dry climate. Total green space was quantified by extracting surface cover areas from remotely sensed data in a geographical information system (GIS) environment, and this information was used to designate suitable sites for future green spaces such as parks. Street trees were inventoried for species identification and size. Trees of heritage value were identified through a public awareness campaign, and then were subsequently surveyed for species identification, height and trunk diameter. GIS green space analysis showed that per capita park space was approximately 1.8 m2, but a master plan proposed increasing per capita park area to 4 m2 within 25 years. The increased park area will be largely in the form of lower cost, semi-naturalized tree parks. The inventoried street tree population, approximately 200,000, was skewed somewhat towards a monoculture, as 42% were the facultative evergreen Pterocarpus indicus Wild. By contrast, none of the other species exceeded 7% of the total. That most of the other species consisted of smaller trunk diameters than P. indicus, and therefore younger, suggests that Bangkok’s street tree plantings are becoming more diverse. The heritage large tree inventory was dominated by evergreen tree species, particularly exceptionally large Ficus species, found largely on Buddhist temple grounds, followed by Albizia saman (Jacq.) Merr. The slower growing evergreen heritage species are worth careful appraisal and preservation because they are less likely to be commonly planted. Careful species selection balancing drought deciduous and dry evergreen species can achieve adaptation to the monsoonal dry season with diverse aesthetic quality in both Bangkok’s street tree population and in its semi-naturalized tree parks. r 2008 Elsevier GmbH. All rights reserved. Keywords: GIS; Monsoonal climate; Per capita green space; Semi-naturalized forest park; Species selection; Tropics

Introduction Urban green space that includes contiguous vegetated areas such as parks or forest stands, and isolated trees growing along streets, in street medians, or private property, is a critical foundation for both a healthy Corresponding author.

E-mail address: [email protected] (R. Kjelgren). 1618-8667/$ - see front matter r 2008 Elsevier GmbH. All rights reserved. doi:10.1016/j.ufug.2008.03.002

population and healthy economy in any city. Indeed, the UN-World Health Organization recommends at least 9 m2 of urban green space per capita to mitigate a number of undesirable environmental effects and provide other benefits (Deloya, 1993). Urban greening and urban forests are particularly critical to healthy cities in developing countries that contain some of the world’s largest metropolitan areas. Green space and urban trees will become increasingly important because

ARTICLE IN PRESS 220

B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

the rate of urbanization is also greatest in developing countries, mostly in smaller cities of approximately 500,000 in Asia and Africa (UN-ESA, 2003). In order to understand and plan for the benefits of urban green spaces, cities in tropical and subtropical developing countries seeking to improve their quality of life need to inventory their green infrastructure to determine existing resources, and to set targets for future improvements (Miller, 1996). Identifying appropriate areas for greening requires a careful process of identification and of assessing suitability, and feasibility (Gul et al., 2006). This process is especially important in dense developing country cities (Cy, 2004) linking local to regional green space (Li et al., 2005). Successful greening of tropical urban spaces hinges on financial and ecological sustainability (Randal et al., 2003). Financial and ecological sustainability are in turn strongly affected by green infrastructure allocation between expensive, and typically less environmentally adapted, turf-flower ornamental landscapes, versus lower cost trees that offer greater selection potential for environmental adaptation. Since most cities in developing countries are in regions of native tropical forest, a large diverse pool of adapted species is available (Cy and Liu, 2001). Selection of tree species best suited for tropical urban conditions will depend on above and belowground space (Cy, 2001), and the actual climate of the city. For example, where a tropical city falls along the seasonality gradient of annual rainfall between a wet and a monsoonal climate, alternating dry and wet seasons in part govern foliage evergreeness versus deciduousness (Santiago et al., 2004), and will determine the importance of drought tolerance in species selection. Finally, inventorying, assessing and publicizing large, existing trees that may be of social and cultural heritage value is important for the character of a city and helps turn the general public into stakeholders in their green infrastructure (Cy, 2006). Bangkok is the capital of Thailand and is among the larger cities in Asia, with an estimated unofficial population well in excess of 10 million people. As an economic magnet, Bangkok’s population is continually increasing through in-migration from the Thai countryside. Adding to its native population, large numbers of tourists visit Bangkok as a destination or linger several days en route to other destinations. Bangkok has a monsoonal climate, and during the dry period from December through April the permanent and tourist populations are subject to sunny, hot conditions exacerbated by the heat island effect from pavement and buildings characteristic in urban monsoonal climates (Weng and Yang, 2004). Bangkok’s green areas and urban forest are critically important in improving its livability, particularly during the dry season. Since the early 1990s the city

government, the Bangkok Metropolitan Authority (BMA), has engaged in an active street tree-planting campaign. By the late 1990s, the absence of an inventory of its overall green area as well as its street trees hindered the BMA’s ability to manage and improve Bangkok’s green spaces and urban forest for the benefit of locals and visitors. This paper reports on inventories of Bangkok’s green space, street trees, and large heritage trees in terms of how information on tree diversity can influence future species selection for a city in a monsoonal tropical climate seeking to improve it green infrastructure.

Methods Climate Bangkok has a seasonal monsoonal climate where average daily high temperature remains relatively constant over the year, largely fluctuating within a range of 31–34 1C (Fig. 1). Like many tropical cities in the latitudinal belt between 151 and 251, Bangkok has a 6-month monsoonal wet season from May through October that ameliorates the heat. December–April is the dry, sunny season, somewhat cooler the first 3 months, but March and April have high solar intensity and longer days and thus can become quite hot. Evapotranspiration, calculated using the Hargreaves equation (RefET version 2.01, University of Idaho), showed a narrow range of variation similar to daily high air temperature, highest in April and lowest in February, varying between 4 and 6 mm on a daily basis.

Urban greenspace analysis In 2001, the BMA contracted with the Faculty of Forestry at Kasetsart University (KUFF), Thailand’s

Fig. 1. Monthly rainfall, evapotranspiration (Hargreaves equation) and daily high temperature of Bangkok based on 30-year historical average data.

ARTICLE IN PRESS B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

leading agricultural and natural resource institution, to develop a green space master plan for Bangkok using remote sensing and geographical information system (GIS) analysis. KUFF fused, for cost purposes, a Landsat three-band (RGB) image at 25  25 m2 resolution and a IRS-1D black and white satellite image at 5  5 m2 resolution, registered to each other, in ERDAS Imagine 8.4 (Leica Geosystems, Norcross Georgia) using the band overlay method (Phol and van Genderen, 1998). The resulting high-resolution (5  5 m2) fused color image was rectified to ground coordinates and imported into a GIS environment (ArcInfo ver. 8.0, ESRI Inc., Redlands CA). The image was then classified, automated and unsupervised, and pixels assigned to 11 land use classes: parks, sports fields, golf courses, urbanized, trees, open water, seasonally flooded, agriculture, fish farms, abandoned, and other. Ground area for each class was determined by summing, through GIS analysis, the number of pixels for each land use class. Total area for each land use class was summed for each of the 50 districts into which Bangkok is divided. Then each land use class was summed by grouping the 50 districts into six larger contiguous aggregate administrative district ‘‘district groups’’ as defined by the BMA. The amount of developed green space (golf, sports, park) and tree area were summed and then divided by the population (2001 census) within each district group to get the total green space (m2) per inhabitant, then the total park (excluding golf courses and sports fields) area per capita was also calculated. Total green space and park area per capita was averaged over the six district groups to arrive at the aggregate green space per capita for Bangkok for comparison to other cities and to assess the per capita needs for additional park space over the next 25 years. The classified GIS layer with parcel boundaries was then analyzed for potential development as park space. The objective was to identify suitable land parcels for conversion to park area, allowing BMA to later select financially and politically feasible areas (Gul et al., 2006). Specifically, these parcels were identified from unoccupied or abandoned privately or publicly owned space as measured by land use falling in the ‘‘other’’ category (vegetation that was not tree or park cover). Managers of each district were given the classified image with ‘‘other’’ areas designated and they verified that the land was unoccupied. Parcels with sufficiently viable area not near an existing park were identified as conversion candidates. These candidate areas were classified as either a potential maintained park, composed of flowers and turfgrass with embedded trees, or a potential semi-naturalized, minimal maintenance forest park with no turf or flowers. An area was classified as a potential future maintained park if there were no other maintained parks in that district. Otherwise, it was

221

classified as a potential future semi-naturalized forest park. Installation costs for each type of green space was estimated based on historical data obtained from the BMA, where installation (labor and supplies) was approximately $500,000US/ha for a flower and turfgrass park while a man-made forest park cost $25,000 US/ha (1 USD ¼ 35 Baht).

Street tree inventory In 1999 the BMA contracted with the Department of Silviculture of Kasetsart University to conduct an inventory of street trees along its main streets. The inventory was started in December 1999 and completed in March 2000. The inventory was carried out by Silviculture undergraduates that collected, for each tree, species name and diameter at breast height (DBH) with measuring tape. Finally, the location of each tree was recorded with a measuring tape from the nearest corner for later mapping by the BMA. Where species identification was uncertain, leaf samples were collected and then taken to the national herbarium for proper keying. Trunk diameter of the top five most common species or genera (four Lagerstroemia species were grouped), and grouped values for the remaining species, were distributed into 50 mm increments from 0 to 500 and plotted as a frequency distribution histogram. Each species was assigned to a growth rate class relative to one another based on accumulated anecdotal knowledge in KUFF. Finally, total shaded area was estimated based on the allometeric relationship between total leaf area and trunk diameter given by Nowak (1996), using a shading factor of 0.83 averaged over the values given in the paper.

Registered heritage large tree survey BMA offered a competition in 1999 asking the public to nominate the largest trees of heritage value on public or private property growing in their neighborhood. Nominations were sorted by species and the largest 53 different species were identified, and the person nominating the largest tree of each species was awarded a certificate. In 2001 the Silviculture Department at Kasetsart University was contracted to survey the 53 trees, as well as any adjacent large trees on the same property, which resulted in a total number of 261 surveyed. Data collected included DBH (stem area summed and diameter back-calculated for multistemmed specimens) and total tree height, visually estimated by scaled approximation of a known height next to each tree, as well as species identification. Species identities were confirmed following the same process as the street tree inventory, but a large number of

ARTICLE IN PRESS

33 25 22 9 19 8 13 57 61 66 61 78 68 65 0.585 1.343 1.857 6.798 1.731 4.500 16.82 1.75 0.61 2.29 3.08 0.40 0.62 1.82 3.99 1.73 12.09 43.74 0.72 8.74 11.8 858,205 979,726 1,219,522 831,390 776,277 1,024,083 568,9203 Park, tree, golf course, sports fields. Of total green area. c Official census. d As proposed in Bangkok Metropolitan Authority master plan. e Low maintenance naturalized tree plantings.

Parks (percentb)

38 48 60 76 30 91 72 3.43 1.69 14.74 36.37 0.56 8.95 66

Parks (m2/person) Total green (m2/person) Populationc (Yr. 2000) Total area (km2)

Trees (percentb)

Demographics Existing areas

44 35 19 7 55 7 16 b

1.2 3.0 14.3 26.2 50.5 4.8 100.0

a

19 47 225 411 792 75 1569

Central N Central NE East West Southwest

Parks, sports field, and golf courses Trees Water, seasonally flooded Agriculture/fish farms Developed Other, not used Total

Ratannakosin Chaopraya Burapa Srinakharindra North Thonburi South Thonburi Total

Percent

Location

km2

District group

Land use

Land area, population and greenery endowment of district groups in Bangkok

Table 1. Area of main land uses in Bangkok derived from GIS analysis of satellite imagery

Table 2.

Bangkok has a large surface area, over 1500 km2, but includes a large amount of non-constructed area within its boundaries, including agriculture, a modest amount of total green space, but a relatively small amount of park space (Table 1). Slightly over 50% of Bangkok is building, roads and other constructed surfaces, juxtaposed against a surprisingly high 26% of land in food production. The vast majority of this agricultural area is actual farmland and the balance shrimp farms on the periphery of the contiguous city boundaries. Over 14% of Bangkok’s total area is seasonally flooded during the wet season. These are critical areas that receive runoff from the 50% constructed area, helping to alleviate flooding in streets, but are generally unusable for other purposes and present management challenges during the dry season. Only 4.2% of Bangkok’s total area is green space (excluding agricultural land), mostly tree cover that includes street trees and naturalized areas, and 1.2% in developed green space: parks (including trees), sports field and golf courses. This 1.2% developed green space is almost evenly split between actual parks accessible to the entire population, and to golf courses which are not readily accessible to average citizens of Bangkok, and the remainder is in athletic fields. Land use is very different in the central part of the city compared to its eastern and western edges as defined by the aggregate administrative districts (Table 2, Fig. 2). The Srinakharindra (eastern) and South Thonburi (southwestern) aggregate district groups of Bangkok are contiguous with food producing areas and forest cover in adjacent provinces, thus have the highest percent of tree and food producing areas within Bangkok. Conversely, the Rattanakosin and Chaopraya district groups encompass central Bangkok and have the

Proposed park areasd

Green space analysis

62 119 274 665 163 287 1569

Results

Total park area (km2)

specimens in the genus Ficus could not be conclusively identified, and thus were simply identified as Ficus sp.

Tree plantinge (percent)

Turf, flowers (percent)

B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

Total green areaa (km2)

222

ARTICLE IN PRESS B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

223

Fig. 2. Map of current parks and new parks proposed in the green area master plan for the city of Bangkok, where circle size is proportional to current (closed circles) and proposed (open circles) park areas.

highest population density, 14,000/km2 and 8000/km2, respectively, 2–4 times greater than the other four district groups. Variation in population density skews the per capita green space values. Per capita green space averages only 2.8 m2 in the central Bangkok district groups, Ratanakosin and Chaopraya, versus 11.8 m2 for the city as a whole. While the per capita total green space of central

Bangkok, Ratanakosin and Chaopraya is lower than several other large cities in Asia, on a city-wide basis its per capita green space compares favorably to large cities in other developing countries. For example, Seoul has 14 m2 (Kuchelmeister, 1998), Singapore 10 m2 (Chow and Roth, 2006) and Beijing has 6 m2 (Dembner, 1993) of green space per capita, but greater than other capital cities in tropical and subtropical developing countries

ARTICLE IN PRESS 224

B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

such as in Mexico City, at 1.9 m2 (Deloya, 1993), and New Delhi at 0.12 m2 (Kuchelmeister, 1998). City-wide average per capita publicly accessible park space (excluding the other developed green space: sports fields and golf courses that have limited public access) was about 30% higher than the central Bangkok aggregate districts. At approximately 1.2 m2/person, about 40% of the green space in central Bangkok is actual park space, the rest in tree cover such as street trees. The master plan recommendation to the BMA was that city wide per capita park space was low and should be increased to 2.5 m2 in 15 years and ultimately to 4 m2. Specifically, the master plan proposed to the BMA that park space in Bangkok be doubled, ultimately adding 16.8 km2 to the existing 8.4 km2 to achieve the recommended 4 m2 per capita park space over 25 years (accounting for population growth). Of this park area, only 13% of the total is proposed to be high maintenance conventional turfgrass and flowerbeds. Conventional park area has high installation costs, due to high unit and labor costs of sod, flowers, and irrigation. While the BMA does not have direct numbers, maintenance costs are likely to also be proportionally as high due to labor, equipment, and supply expenses from mowing, flower replacement, and irrigation maintenance. Instead, the master plan recommends 65% of the new park space be semi-naturalized forest area that will be publicly accessible through paths and trails. Seminaturalized forest park area is much cheaper to install than conventional park area, largely due to low tree production costs that result in landscape-size (E50 mm DBH) trees that wholesale from roughly $5–$8 USD each. Similarly, maintenance of semi-naturalized forested parks is likely to be proportionally lower than conventional turfgrass and flowerbed parks, largely limited to preventative pruning and weed control but no irrigation during the monsoonal dry season. The remaining proposed park area (22%) is seasonally flooded area that currently receives wet season runoff that the master plan recommends be partially dredged to reduce management challenges during the dry season. The utility of these areas can be increased by creating elevated areas from the dredged soil that can be developed as semi-naturalized forested or conventional park space, and the dredged area would then be low enough to create permanent small lakes and ponds.

Street tree inventory Bangkok has a large street tree population encompassing a wide range of species. The nearly 200,000 trees surveyed were distributed in 36 families and 127 species, of which the 22 most numerous (more than 1000

individual trees) are presented in Table 3. Many of the tree species (57% of total) were in the legume (Fabaceae) family, although split amongst three subgroups within the family. The loosestrife (Lythraceae) family was the next most numerous, four species, all Lagerstroemia, at 12% of the total number of trees inventoried. The most common tree species was Pterocarpus indicus Wild, accounting for 42% of the total number of trees inventoried. P. indicus is a fast growing, mostly evergreen but frequently facultative deciduous, species native to the wet evergreen tropics of SE Asia, although not Thailand. About 70% of the P. indicus trees were in the 5–15 cm trunk diameter range, reflecting an aggressive tree-planting program initiated by BMA in the 1990s based on its desirability as an appropriately sized species that provides shade and has few management problems in Bangkok (Fig. 3). Tabebuia rosea (Bertol.) DC. and Cassia fistula L. were the next most common species, each with about 6.4% of the total. The size range of C. fistula, another SE Asia species native and national Thai tree, was smaller than P. indicus, indicating that it has become more popular since the late-1990s (Fig. 3). C. fistula is an obligate drought deciduous species, defoliating at the start of the monsoonal dry season, with a relatively small mature size suitable for Bangkok streets that produces a large floral display of yellow racemes before leafing out towards the end of the monsoonal dry season, depending on rainfall. Its increased use is due in part to the association of the flowers with the king of Thailand, whose royal color is yellow, the display of which is promoted on his birthday and anniversary of ascension to the throne. T. rosea is no longer planted as a street tree because it tends towards surface rooting that breaks and heaves paving in Bangkok. Consequently, Bangkok’s T. rosea population is aging, as nearly half of the trees fall within the 20–30 cm trunk diameter classes (Fig. 3). Swietenia macrophylla King, mahogany, a slow growing obligate evergreen species, maintaining foliage during the monsoonal dry season and native to central and South America, was the fourth most common species. S. macrophylla size distribution was similar to P. indicus, but because it is slower growing than P. indicus, the population probably reflects greater age in planting (Fig. 3). Interestingly, all four Lagerstroemia species were in the top 13 most numerous species. These species are fairly similar in appearance and floral display, blooming profusely at the end of their drought deciduous phase, and their size distribution was similar to P. indicus and S. macrophylla (Fig. 3). Ficus benjamina L., 1% of total, is the official symbolic tree of Bangkok. Bangkok has approximately 1200 km of main and sub streets, which translates to approximate 168 trees/km.

ARTICLE IN PRESS B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

225

Table 3. Distribution of main street tree species by growth rate, seasonality and geographical provenance for the most numerous (count X1000 trees) street trees inventoried in Bangkok in 2001 Percent of total

Seasonalityb

Geographical provenance

79,365

41.9

F

SE Asia.

4 3

12,792 1731

6.8 6.7

D D

Cent, S America SE Asia

3 2 3 2 1 2 2 3

9026 8293 7841 5829 5484 5270 4908 4901

4.8 4.4 4.1 3.1 2.9 2.8 2.6 2.6

E D E E E D D D

Cent, S America SE Asia SE Asia S Asia Africa, Asia SE Asia SE Asia SE Asia

2

4736

2.5

D

Africa

2

3714

2.0

D

SE Asia.

2 3

3444 3348

1.8 1.8

F E

Asia Africa

1 2 1

2455 1961 1917

1.3 1.0 1.0

E F F

SE Asia Asia Cent, S America.

5 3 3 4

1757 1399 1251 1157 189,409

0.9 0.7 0.7 0.6

E E E E

Asia. Asia SE Asia Australia

Scientific name

Family

Growth rate classa

Pterocarpus indicus

Fabaceae (Papilionoideae) Bignoniaceae Fabaceae (Caesalpinioideae) Meliaceae Lythraceae Sapotaceae Annonaceae Guttiferae Lythraceae Lythraceae Fabaceae (Caesalpinioideae) Fabaceae (Caesalpinioideae) Lythraceae

4

Tabebuia rosea Cassia fistula Swietenia macrophylla Lagerstroemia speciosa Mimusops elengi Polyalthia longifolia Calophyllum inophyllum Lagerstroemia loudonii. Lagerstroemia floribunda Peltophorum pterocarpum Delonix regia Lagerstroemia macrocarpa Azadirachta indica Tamarindus indica Millingtonia hortensis Terminalia catappa Cassia spectabilis Ficus benjamina Alstonia scholaris Casuarina junghuhniana Acacia auriculiformis Total a

Meliaceae Fabaceae (Caesalpinioideae) Bignoniaceae Combretaceae Fabaceae (Caesalpinioideae) Moraceae Apocynaceae Casuarinaceae Fabaceae

Total number

Empirical assessment of overall species growth rate relative to one another, where 1 ¼ slowest and 5 ¼ fastest. D ¼ deciduous, F ¼ facultative deciduous, E ¼ evergreen.

b

Assuming that street trees in Bangkok were planted in three rows (left, right and median strip), there would be an approximate average of 18 m spacing between trees. This spacing is very similar to the spacing recommended in urban forestry texts to provide shade yet avoid overcrowding (Miller, 1996). In a related fashion, shade is critical during Bangkok’s dry monsoonal period, as even the deciduous species are not leafless during the entire dry period, varying in periodicity and duration of deciduousness depending on rainfall (Elliot et al., 2006), thus contribute to shade to some extent. The rough calculation of crown area for the population of inventoried street trees (Nowak, 1996) showed that Bangkok’s street trees would provide approximately 0.87 km2 of cover and shade. This is a small fraction of the total Bangkok tree cover (which includes street trees but not park trees), and only 0.15 m2/person based on the official census but street trees probably constitute

the majority of the tree cover in the central Bangkok aggregate district groups.

Registered heritage large tree survey The heritage large trees surveyed were distributed among 65 species in 23 families, with 22 species having three or more individuals (Table 4). The average trunk diameter was 120 cm and average height was 15 m, but the modes were 80 cm and 12 m, respectively, due to several particularly tall and exceptionally wide species that skewed the averages. Like the street tree inventory, the greatest number of species (29%) was in the legume family, followed by 8% in the mulberry (Moracea) family, exclusively in the genus Ficus. Two species, Ficus religiosa L. and Albizia saman (Jacq.) Merr., had the most individual trees, 15% each of the total, followed by

ARTICLE IN PRESS 226

B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

Fig. 3. Trunk diameter frequency distribution for the four most common species, Cassia fistula, Pterocarpus indicus, Lagerstromia spp., Swietenia macrophylla, and Tabebuia rosea, and all other species combined.

the inconclusively identified Ficus sp. at 6% and P. indicus at 5%. Many of the surveyed trees were on the grounds of Buddhist temples because they have been relatively undisturbed over time compared to other properties due to religious tradition prohibiting the cutting of certain tree species. Buddhist scripture has that the Buddha found enlightenment under F. religiosa and thus the species was widely planted at Buddhist temples, but the other Ficus species surveyed were likely planted under the assumption that they were similar to F. religiosa, all native to south Asia. Consequently, of the total number of trees surveyed, 28% were Ficus species. Because they are large trees tending towards multiple trunks, all five Ficus species had an average diameter between 175 and 300 cm, approximately 2–4 times that of nearly all the other species surveyed. Several other heritage species were associated with Buddhist temples, including Crudia chrysantha (Pierre) K. Schum. and Couroupita guianensis Aubl. The Buddha was said to have been born under a species, Shorea robusta Gaertner F., native to south Asia, but its common name and that of C. guianensis are very similar in the Thai language, so C. guianensis has been misidentified and planted as S. robusta. Notable for size and number among the non-Ficus species was A. saman, a South American species in the Fabaceae. A. saman is a single-trunk species that averaged 114 cm in diameter, and the largest individual

was a massive 210 cm. The tallest tree on average was Dipterocarpus alatus Roxb., a SE Asia native found in riparian areas, notable for its exceptionally long unbranched trunks and remotely high crowns. Hopea ordorata Roxb. and Tamarindus indica L. had the most overall massive trees, both with average trunk diameter above 100 cm and height greater than 19 m. Hopea odorata and Sonneratia caseolaris (L.) Engl., a coastal species found adjacent to, and related to, mangroves on mudflats, are senile populations. H. odorata was widely planted during Bangkok’s early establishment phase as a source of boat-building wood, but has performed poorly in Bangkok’s urban conditions, and the nine measured trees were not in good health. S. caseolaris was the dominant tree in the low-lying area now occupied by Bangkok, but has not survived urbanization well and these trees appeared to be remnants of those stands. Comparing species frequency between the street tree inventory and the large tree survey, the number of common species was minimal. There were 19 species found in both the survey and inventory, but of the top 22 most numerous species in both the survey and inventory, there were only four in common C. fistula, Mimusops elengi L., P indicus, and Alstonia scholaris (L.) R. Br. The large C. fistula and Mimusops elengi surveyed were both below the mode in diameter and height, M. elengi being the second shortest of the top 22, suggesting even when mature as street trees they will less

ARTICLE IN PRESS B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

227

Table 4. Distribution of heritage tree species by tree dimensions, seasonality and geographical provenance of 22 most numerous (count X3 individual trees) large heritage tree species surveyed in Bangkok Scientific name

Family name

Total tree numb.

Diameter (mma)

Height (ma)

Native region

High Low Average High Low Average Foliage typeb Ficus religiosa Samanea saman Ficus sp. Pterocarpus indicus Mimusops elengi Diospyros decandra Hopea odorata Alstonia scholaris Dipterocarpus alatus Ficus glabella var. concinna Tamarindus indica Couroupita guianensis Ficus altissima Syzygium cumini Crudia chrysantha Diospyros malabarica var. siamensis Diospyros mollis Ficus elastica Cassia fistula Erythrina subrumbrans Nauclea orientalis Sonneratia caseolaris Total a

Moraceae Fabaceae (Mimosaoideae) Moraceae Fabaceae (Papilionoideae) Sapotaceae Ebenaceae Dipterocarpaceae Apocynaceae Dipterocarpaceae Moraceae

38 38

297 210

67 72

175 114

25 25

8 8

14.6 15.5

E E

16 14

665 121

67 42

235 88

28 22

11 10

15.5 16.2

E F

Asia Cent., S. America Asia SE Asia

11 10 9 8 8 7

80 165 140 150 99 450

48 54 66 54 45 150

67 101 106 92 73 277

14 20 28 25 23 30

5 6 15 10 17 15

9.5 12.9 19.2 17.5 19.8 18.9

E E E E F E

SE Asia SE Asia Asia Asia SE Asia Asia

Fabaceae (Caesalpiniodeae) Lecythidaceae

6

151

59

103

25

13

19.7

E

Africa

5

52

31

43

11

6

8.4

E

5 5 4

560 99 131

136 43 71

302 66 99

25 13 18

13 9 15

17.6 11.0 16.0

E E E

Cent., S. America Asia SE Asia SE Asia

4

110

55

79

15

10

12.5

E

SE Asia

4 4 3

110 210 52

53 144 40

86 190 48

19 18 12

8 10 11

12.5 14.0 11.7

E E D

SE Asia Asia SE Asia

3

110

50

83

10

10

10.0

D

SE Asia

3 3

110 151

69 67

95 99

16 22

15 12

15.3 15.7

E E

Australia Coastal Asia

Moraceae Myrtaceae Fabaceae (Caesalpiniodeae) Ebenaceae Ebenaceae Moraceae Fabaceae (Caesalpiniodeae) Fabaceae (Papilionoideae) Rubiaceae Sonneratiaceae

261

High ¼ maximum recorded size within a species, low ¼ minimum recorded size within a species. D ¼ deciduous, F ¼ facultative deciduous, E ¼ evergreen.

b

likely outgrow their space. Diameter and height of A. scholaris and P. indicus, were both above average, suggesting that they are more likely to outgrow their space as street trees. Interestingly, the heritage tree species were nearly all evergreen, while the street trees species were about half deciduous. This is probably due to tropical evergreen trees being typically slower growing and less floriferous than the deciduous trees selected for ornamental qualities.

Discussion Because the native vegetation of much of Thailand, like much of the topics, is forest, tropical forest tree

selection is key to successfully improving its urban green infrastructure. Information gleaned from the street tree and heritage tree inventories regarding species performance and native habitat can suggest general guidelines for appropriate species selection and management practices when put in the context of how the monsoonal climate affects native forest habitats. The tropical monsoonal dry season creates deciduous forests of well-adapted drought avoiders dormant for varying durations during dry season (Miles et al., 2006). Drought deciduous forests are typically found on more fertile sites where defoliation does not exact a nutrient cost, and very high transpiration rates advantageously maximize gas exchange and growth during the wet season (Ishida et al., 2006). High transpiration rates of

ARTICLE IN PRESS 228

B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

the deciduous species during the dry season would ostensibly be a concern for water stress. However, many deciduous species rapidly enter water stress as soil water is depleted at the start of the dry season (Choat et al., 2005). Rapid water stress allows these species to senesce their foliage at a relatively high soil water content (Yoshifuji et al, 2006), thus avoid damaging drought, and in turn allows many deciduous species to leaf out during mid dry season (Elliot et al., 2006). Many of the deciduous species have attractive flowers, so drought actually promotes flowering. Monsoonal dry evergreen forests are typically found on soils of lower fertility (Wright et al., 2002), and are generally drought avoiders with lower transpiration rates and slower growth (Choat et al., 2005) and less floriferous. The distinction between deciduous and dry evergreen species was evident in the heritage tree inventory, where more of the species were either slow growing evergreens, or the very fast growing, large evergreen Ficus species that are generally not used in the urban landscape outside the parks and large institutional grounds. Many of the slower growing evergreen species, such as H. odorata, are less likely to be produced in a nursery and planted in urban landscapes because of their growth rate. These species then take on a greater ecological and educational value to the general population which can be incorporated into appropriate appraisal (Cy, 2006) and educational schemes. Important species distinctions are also evident in the street tree inventory. Deciduous trees are often favored for street plantings because they grow fast and have more attractive floral displays that are enhanced by their drought deciduousness. The richness of the monsoonal deciduous forests (Balvanera and Aguirre, 2006), could be tapped into for more diverse species selection than P. indicus and possibly C. fistula in future plantings to offset the problems associated with tree monocultures. Monocultures are more prone to diseases, insects, and problems such as the root issues of T. rosea. That the less common street tree species fall into small trunk diameter is a healthy indication that BMA may already be increasing its species diversity in its street tree plantings. Careful species selection for tree size is also important, as Bangkok generally has narrow streets and extensive, low-hanging, aboveground wiring that causes space conflicts. While ‘‘V’’ pruning can potentially allow room for wires and allow relatively healthy and attractive tree crowns, smaller tree species such as M. elengi, can be used that may be less likely to cause conflicts with wires. Also, air pollution has to be considered in species selection, as Bangkok air has significant, but improving, pollution issues that have been shown to limit growth in P. indicus (Royampaeng, 1995). Tree species and habitat-type selection criteria for the proposed semi-naturalized tree parks will be particularly

important, but should generally be similar to that for street trees. Aesthetic and utility factors will likely drive a balance between an evergreen canopy, particularly for shade during the dry season, with enough colorful, flowering deciduous species such as Lagerstroemia species and the native Erythrina variegate L. (coral tree). Evergreen versus deciduous foliage will affect some management issues, as drought deciduous species may require a more rigorous fertilization program to avoid mining soil for nitrogen and other nutrients due to leaf loss and removal (Kjelgren and Clark, 1993), depending on how the ground layer is managed. Based on the GIS analysis and recommendations, BMA is moving towards more naturalized tree sparks. Indeed, of the recommended increase in park area based on GIS analysis, eight new parks have been added, all naturalized tree parks. One of them was planted to D. alatus, the favorite tree of the Thai king, in honor of his 80th birthday. A significantly challenging management issue regarding species selection for both semi-naturalized parks and street trees is the potential for water stress during the monsoonal dry period. Some administrative districts within BMA, such as Bang Kaen where Kasetsart University is located, have the policy of daily irrigation of its street trees, a resource intensive effort, particularly if applied to the semi-naturalized parks. This effort may not be necessary with careful selection of drought tolerant monsoonal forest species. However, longer leafless periods and slower growth, ecologically acceptable drought adaptations in drought deciduous species, could be avoided in the semi-naturalized parks simply by wider plant spacing to create a larger soil water reservoir and greater drought avoidance (Cy, 2001).

Acknowledgments This paper was written from the modified data of Thaiutsa et al. (2000) and Kasetsart University (2003). The authors would like to acknowledge Dr. Monton Jamroenprucksa, Dr. Damrong Pipattanawattanakul, Dr. Duangjai Sookchaloem, and Working Committee of the Green Space Master Plan for BMA for their primary data collection and useful information. We also want to express our appreciation to the BMA’s Office of City Planning and Office of Environment for financial support of the original investigation.

References Balvanera, P., Aguirre, E., 2006. Tree diversity, environmental heterogeneity, and productivity in a Mexican tropical dry forest. Biotropica 38, 479–491. Choat, B., Ball, M.C., Luly, J.G., Holtum, J.A., 2005. Hydraulic architecture of deciduous and evergreen dry

ARTICLE IN PRESS B. Thaiutsa et al. / Urban Forestry & Urban Greening 7 (2008) 219–229

rainforest tree species in north-eastern Australia. Trees: Structure and Function 19, 305–311. Chow, W.T.L., Roth, M., 2006. Temporal dynamics of the urban heat island of Singapore. International Journal of Climatology 26, 2243–2260. Cy, J., 2001. Managing urban trees and their soil envelopes in a contiguously developed city environment. Environmental Management 28, 819–832. Cy, J., 2004. Green-space preservation and allocation for sustainable greening of compact cities. Cities 21, 311–320. Cy, J., 2006. Formulaic expert method to integrate evaluation and valuation of heritage trees in compact city. Environmental Monitoring and Assessment 116, 53–80. Cy, J., Liu, H.T., 2001. Patterns and dynamics of urban forests in relation to land use and development history in Guangzhou City, China. Geographical Journal 167, 358–375. Deloya, M.C., 1993. Urban forestry activities in Mexico. Unasylva 173 (44), 28–32. Dembner, S.A., 1993. Urban forestry in Beijing. Unasylva 173 (44), 13–18. Elliot, S., Baker, P.J., Borchert, R., 2006. Leaf flushing during the dry season; the paradox of Asian monsoon forests. Global Ecology and Biogeography 15, 248–257. Gul, A., Gezer, A., Kane, B., 2006. Multi-criteria analysis for locating new urban forests: an example from Isparta, Turkey. Urban Forestry & Urban Greening 5, 57–71. Ishida, A., Diloksumpun, S., Ladpala, P., Staporn, D., Panuthai, S., Gamo, M., Yazaki, K., Ishizuka, M., Puangchit, L., 2006. Contrasting seasonal leaf habits of canopy trees between tropical dry-deciduous and evergreen forests. Tree Physiology 26, 643–656. Kasetsart University, 2003. Green space master plan for Bangkok Metropolitan Authority. A final report by the Faculty of Forestry submitted to the Office of City Planning, BMA, Bangkok. Kjelgren, R., Clark, J., 1993. growth and water relations of liquidambar styraciflua l in an urban park and plaza. Trees: Structure and Function 7, 195–201. Kuchelmeister, G., 1998. Urban forestry in the Asia-Pacific Region – situation and prospects. Document APFSOS/WP/ 44, FAO, Rome/Bangkok. Li, F., Wang, R.S., Paulussen, J., Liu, X.S., 2005. Comprehensive concept planning of urban greening based on

229

ecological principles: a case study in Beijing, China. Landscape and Urban Planning 72, 325–336. Miles, L., Newton, A.C., DeFries, R.S., Ravilious, C., May, I., Blyth, S., Kapos, V., Gordon, J., 2006. A global overview of the conservation status of tropical dry forests. Journal of Biogeography 33, 491–505. Miller, R.W., 1996. Urban Forestry: Planning and Managing Urban Green Spaces. Prentice-Hall, Englewood Cliffs, NJ. Nowak, D.J., 1996. Estimating leaf area and leaf biomass of open-grown deciduous urban trees. Forest Science 42, 504–507. Phol, C., van Genderen, J.L., 1998. Multisensor image fusion in remote sensing: concepts, methods and applications. International Journal of Remote Sensing 19 (5), 823–854. Randal, T.A., Chruchill, C.J., Baetz, B.W., 2003. A GIS-based decision support system for neighbourhood greening. Environment and Planning 30, 541–563. Royampaeng, S., 1995. Effects of air pollution on some physiological and morphological characteristics of Pterocarpus indicus Wild. MS Thesis. Kasetsart University, Bangkok. Santiago, L.S., Kitajima, K., Wright, S.J., Mulkey, S.S., 2004. Coordinated changes in photosynthesis, water relations and leaf nutritional traits of canopy trees along a precipitation gradient in lowland tropical forest. Oecologia 139, 495–502. Thaiutsa, B., Jamroenprucksa, M., Pipattanawattanakul, D., Sookchaloem, D., 2000. Street Trees of Bangkok. Aksorn Siam Printing, Bangkok. United Nations Department for Economic and Social Affairs, 2003. Long range population projections. Proc: UN Technical Working Group on Long-Range Population Projections, 30 June. New York. Weng, Q.H., Yang, S.H., 2004. Managing the adverse thermal effects of urban development in a densely populated Chinese city. Journal of Environmental Management 70, 145–156. Wright, I.J., Westoby, M., Reich, P.B., 2002. Convergence towards higher leaf mass per unit area in dry and nutrient poor habitats has different consequences for leaf life span. Journal of Ecology 90, 534–543. Yoshifuji, N., Kumagai, T., Tanaka, K., Tanaka, N., Komatsu, H., Suzuki, M., Tantasirin, C., 2006. Interannual variation in growing season length of a tropical seasonal forest in northern Thailand. Forest Ecology and Management 229, 333–339.