gisforlandevaluationforshrimpfarmingin...
TRANSCRIPT
ARTICLE IN PRESS
Ocean & Coastal Management 48 (2005) 51–63
0964-5691/$ -
doi:10.1016/j
�CorrespoE-mail ad
www.elsevier.com/locate/ocecoaman
GIS for land evaluation for shrimp farming inHaiphong of Vietnam
Dao Huy Giap, Yang Yi�, Amararatne Yakupitiyage
Aquaculture and Aquatic Resources Management, School of Environment, Resources and Development,
Asian Institute of Technology, PO Box 4 Klong Luang, Pathumthani 12120, Thailand
Abstract
This study was conducted to identify appropriate sites for shrimp farming development in
Haiphong province of Vietnam using geographical information systems (GIS). Thirteen base
layers (thematic maps) were grouped into four main land use requisites for aquaculture,
namely, (1) potential for pond construction (slope, land use type, soil thickness, elevation), (2)
soil quality (soil type, soil texture, soil pH), (3) water availability (distance to sea, and water
source), and (4) infrastructure and socio-economical status (population density, distance to
roads, local markets, and hatcheries). A constraint layer was used to exclude areas from
suitability maps that were not allowed to implement shrimp farming. A series of GIS models
was developed to identify and prioritize the most suitable areas for shrimp farming.
This study shows that the land evaluation model is useful for identifying suitable areas for
shrimp farming and for allocating land for efficient income generation, effective conservation,
and sustainable land management. It was estimated that about 31% (2604 ha) of the total land
area (8281 ha) in Haiphong was highly suitable for shrimp farming. Since existing shrimp
farms cover only 1690 ha of land in the study area, the potential for expanding shrimp farms
should take into consideration further political and environmental issues.
r 2005 Elsevier Ltd. All rights reserved.
see front matter r 2005 Elsevier Ltd. All rights reserved.
.ocecoaman.2004.11.003
nding author. Tel.: +662 524 5454; fax: +66 2 524 6200.
dress: [email protected] (Y. Yi).
ARTICLE IN PRESS
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6352
1. Introduction
Aquaculture is a rapidly growing industry with high potential for environmentaldestruction in developing countries. In Vietnam, the land area used for shrimpfarming increased rapidly in recent years due to a steadily rising demand for shrimpin international markets with high economic returns. However, the expansion ofshrimp culture has changed ecological and environmental conditions in coastal areasused for other activities such as rice farming and salt production. There is an urgentneed for appropriate methodology to assist planners for site selection in shrimpfarming development, which should incorporate environmental and naturalresources, and socio-economic factors should be considered in site selections foraquaculture. Inappropriate land use for aquaculture without considering the abovefactors can lead to misuse of natural resources and degradation of the environment,breeding poverty and other social conflicts. To alleviate these potential problems,land evaluation can predict land performance based on its attributes using a varietyof analytical models. The structure of these models range from qualitative toquantitative, functional to mechanistic, or specific to general [1].
The traditional method in land use planning of addressing these issues is byindividual site assessment [2], a subjective and time-consuming procedure. With therecent development of computerized geographic information systems (GIS), it is nowpossible to assess the suitability of multiple sites in a rapid and systematic way toenhance the planned progression of land-based coastal aquaculture towards a moresustainable future [3]. GIS has been widely used for evaluating sites for variouscultured species and environments, e.g., assessments of fish farming potential on theAfrican continent [4], in the Red River Delta of Vietnam [5] and in Latin America[6]; environmental models for aquaculture development in Mexico [7], marine cageculture in Chile [8] and Norway [9]; land-based aquaculture planning in Australia [3];shrimp and/or crab culture in Bangladesh [10]. GIS was also used in site selection forland-based aquaculture around the mouth of the Logan River in south-eastQueensland Australia [3] and shellfish culture in the Indian River lagoon, Florida,USA [11]. This paper presents a quantitative land evaluation method using GIStechniques to identify appropriate sites for shrimp farming development inHaiphong, North Vietnam.
2. Description of the study area
This study was conducted in Doson and Kienthuy districts of Haiphong provincesituated in the eastern part of the Red River Delta, Vietnam. The study area covers8281 ha of land and 9767 ha of water bodies. It lies between 2014001000 and 2014803500
North and 10614202000 and 10614903500 East (Fig. 1).The climate of the study area is affected by two monsoon regimes, the southwest
monsoon during April–September, and the northeast monsoon during October–March. During the annual cycle, only southwest monsoon period is suitable forshrimp farming as it creates a wet and warm climate with temperatures ranging from
ARTICLE IN PRESS
Fig. 1. Map showing location of the study area.
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 53
27 to 34 1C. The northeast monsoon induces a dry and cold climate withtemperatures ranging from 13 to 27 1C which is not suitable for shrimp farming.However, the heavy rainfall during May–September (200–280mm) results in highturbidity and low salinity (5%) in coastal water. Tides play an important role inwater supply for shrimp culture [12]. The tidal regime in Haiphong is diurnal with a24 h and 50min cycle and 3.7–3.8m height.
Haiphong province is attractive to a variety of industrial and commercialdevelopments coupled with existing agricultural, housing and recreational establish-ments. These environmental burdens have a great impact on biological diversity andnatural resource conservation in the area [12,13]. Rapid increase in shrimp farmingin recent years expanded the aquaculture land area to 1690 ha at the end of 2000.More than 50% rice field, saline land, and inefficient salt farms were converted toshrimp farms integrated with crab, fish, and seaweed culture. Shrimp and crab arecommonly cultured during March–August and September–March, respectively. Fishand seaweeds are occasionally cultured.
3. Methodology
3.1. Remote sensing data, thematic maps, and computer software
The primary data sources used in this study include a multi-spectral band SPOTimage, digital and analog maps, statistical data from various reports, GPS data, and
ARTICLE IN PRESS
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6354
socio-economic surveys. According to the basic requisite for successful aquacultureoperation, thirteen base layers (thematic maps) were grouped into the following fourcategories: (1) potential areas for pond construction (slope, land use type, soilthickness and elevation requirements), (2) soil quality (suitability of soil type,texture, and pH), (3) water availability (distance to sea or/and other water sources),and (4) infrastructural and socio-economic status (population density, distance toroad, local markets, and hatcheries) [14]. A constraint layer was used to excludeareas that where shrimp farms are not allowed to be built [15,16].
Land use maps provide information on land use type, road network, source water,constraints (e.g. hill–land forest, road, urban and built-up area, or river), hatcheries,and market location. These maps were updated using a SPOT satellite imageacquired on 5 November, 2000 and primary data from field surveys. Layers forslope, elevation, and distance to sea were extracted from topographical maps [17].Soil thickness, type, texture and pH were elaborated from soil maps and the fieldsurveys, and extracted from Soil Information in Vietnam [18]. Population densitydata were collected from the Statistical Department of Haiphong. EnVi 3.4 (TheEnvironment for Visualizing Images, Research Systems Inc., USA), ArcView 3.2(The Environmental Systems Research Institute, USA), and MS Office 10.0 (TheMicrosoft Corporation, USA) software were used to analyze remote sensing images,GIS, and statistics, respectively.
3.2. Analytical procedure
Two criteria used for land evaluation are factors and constraints. A factor is ameasure of the suitability of the criterion relative to the activity under consideration[16]. A constraint is a restriction that limits the alternatives under consideration in abinary manner [16], including hilly lands, forests, roads, urban areas, built-up land,or river.
Suitability ratings were established according to FAO classification [2,19] on theappropriateness of land for defined uses. Weighing and suitability ratings were basedon the level of importance of a particular factor that influences aquaculture. Eachfactor was ranked and classified according to the following (Table 1):
�
Unsuitable (US ¼ 1): requires considerable time and cost or both that is notworthwhile for shrimp farming.�
Moderately suitable (MS ¼ 2): requires significant interventions before shrimpfarming can be operated.�
Suitable (S ¼ 3): requires modest investment and time period. � Highly suitable (HS ¼ 4): provides a situation in which a minimum time andinvestment are required to develop shrimp farming.
Constraints (hill–land forest, road, urban and built-up area, and river) were coded as0 to exclude constraint areas from suitability maps and the rest of the areas werecoded as 1 [16].
ARTICLE IN PRESS
Table 1
Land use requirement, land characteristics and suitability levels for shrimp farming
Land use requirement/
land characteristics
Suitability rating and score
Highly suitablea Suitableb Marginally
suitablecNot suitabled
(4) (3) (2) (1)
Potential for pond
construction
Land use typee,f Aquaculture pond Rangeland, salt
farm
Agricultural land Mangroves,
village mixed
orchard
Slope o2 2–5 45–10 410
Soil thickness (m) 41 0.5–1 o0.5 —
Elevation (m) 2.0–2.5 2.5–4 or 1–2 44–5 45 or o1
Soil quality
Soil type Gley fluvisols Eutric fluvisols Gleyic
solonchaks, gleyic
arenosols
Haplic calcisols
Soil pH 6–7 5–6 4–5 or 7–8 o4 or 48
Soil texture (% clay) 435 18–35 o18 —
Water availability
Distance to sea (km) o1 1–2 2–3 43
Water sources Tide Irrigation Rain fed —
Infrastructure and socio-
economic
Distance to road (km) o0.5 0.5–1 1–2 42
Population density
(people per km2)
o500 500–1000 1000–2000 42000
Distance to local
market (km)
1–2 o1 or 2–4 44 —
Distance to hatchery
(km)
o4 4–8 48 —
aThe highly suitable (HS ¼ 4) level provides a situation in which minimum time or investment is
required to develop shrimp farming.bFor an area classified as suitable (S ¼ 3), modest time and investment are required.cMarginally suitable (MS ¼ 2), significant interventions may be required before shrimp farming can be
operated.dThe area is not suitable (NS ¼ 1), the time or cost, or both, are too great to be worthwhile for fish
farming.eThe urban and built-up land, and forest land are marked as constraints. A constraint is a restriction
that limits the alternatives under consideration in a binary manner.fWater body is marked as the background.
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 55
The weight for each factor was determined by pair-wise comparisons in thecontext of a decision-making process known as the analytical hierarchy process [20].The suitability rating for each level of a factor was determined from the surveyresults and expert opinions (Table 2). Suitability scores were calculated using the
ARTICLE IN PRESS
Table 2
A pair-wise comparison matrix for assessing the relative importance of land characteristic factors for each
land use requirement (numbers show the ratings of the row factor relative to the column factor)
Land use type Slope Soil thickness Elevation Weight
Potential for pond
construction
Land use type 1 3/2 2 3 0.40
Slope 2/3 1 3/2 2 0.27
Soil thickness 1/2 2/3 1 2/3 0.16
Elevation 1/3 1/2 3/2 1 0.17
Consistency ratio (C.R.) ¼ 0.03
Soil type Soil pH Clay content Weight
Soil quality
Soil type 1 3/2 2 0.46
Soil pH 2/3 1 3/2 0.32
Clay content 1/2 2/3 1 0.22
Consistency ratio (C.R.) ¼ 0.00
Distance to sea Water sources Weight
Water availability
Distance to sea 1 2/3 0.40
Water source 3/2 1 0.60
Consistency ratio (C.R.) ¼ 0.00
Distance to
road
Population
density
Distance to
market
Distance to
hatchery
Weight
Infrastructure and
socio-economic
factors
Distance to road 1 2 3 4 0.46
Population density 1/2 1 2 3 0.28
Distance to market 1/3 1/2 1 2 0.16
Distance to
hatchery
1/4 1/3 1/2 1 0.10
Consistency ratio (C.R.) ¼ 0.01
Potential for
pond
construction
Soil quality Water
availability
Infrastructure
and social
economic
factors
Weight
Land use requirements for
assessment of site
suitability for shrimp
farming
Potential for pond
construction
1 2 3/4 2 0.31
Soil quality 1/2 1 1/2 1 0.17
Water availability 4/3 2 1 3/2 0.34
Infrastructure and socio-
economic factors
1/2 1 2/3 1 0.18
Consistency ratio (C.R.) ¼ 0.01
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6356
ARTICLE IN PRESS
DATA SOURCES
Remote sensing data:
SPOT image
Digital data:topographic and land use maps
Analogue data:various maps
Statistical data:various reports
Other data:GPS data
Field survey data
BASIC THEMES(Land
characteristics)
- Land use type- Slope- Elevation- Soil thickness
- Soil type- Soil texture- Soil pH
Constraints- urban and build-up area- hill-land forest
- Water sources- Distance to sea
- Distance to road- Population density- Distance to market- Distance to hatchery
DERIVED THEMES(Land use
requirements)
Potential for construction
Soil quality
Water availability
Infrastructure and socio-economic
Overall suitability map
OUTPUTData input
and analysisModels Models
Background- river and sea
Fig. 2. Schematic model of land evaluation for shrimp culture.
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 57
following equation:
Suitability score ¼ ðw1r1 þ w2r2 þ � � � þ wnrnÞ=n � 100
where wn and rn are weight and rating scores for each factor, respectively.The development procedure of a suitability map is presented in Fig. 2 and the
suitability score classification is shown in Table 3. A high range score (above 60)indicates more suitable areas for shrimp farming.
4. Results
The existing land use areas and suitability ranking for shrimp culture are presentedin Table 4. The land use classification shows that the dominant land use activities arecrop cultivation (2998 ha), aquaculture (1690 ha), and village mixed orchard
ARTICLE IN PRESS
Table 4
Land use patterns in 2000 and their suitability ranking for shrimp farming
Land use type General characteristics Area (ha) Ranking
Aquaculture pond Semi-intensive and extensive brackish
water aquaculture ponds.
1690 Highly suitable
Range land Consisted of natural grasslands, shrub
lands, and wet meadows.
406 Suitable
Salt farms This kind of land use was composed of
many small ponds and cemented ground
for salt production.
107 Suitable
Crop land Crop land was mostly paddy fields, and
some areas for vegetables.
2998 Marginally suitable
Village mixed
orchard
A combination of houses and gardens
was a regular feature in the countryside.
1462 Not suitable
Mangroves Vegetation located along river mouths
and tidal flats.
797 Not suitable
Urban and built-up
land
This land use type was composed of
urban land, and commercial, service and
industrial areas.
572 Constraint
Forest land Mainly pine and eucalyptus tree in hills
and mountains
249 Constraint
Total 8281
Sea and river 9767 Background
Table 3
Overall suitability rating scores for shrimp farming
Suitability rating Range of score (%)
Highly suitable 100–80
Suitable 80–60
Marginally suitable 60–40
Not suitable 40–0
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6358
(1462 ha), which accounts for 36, 20, and 18% of the total land area, respectively.The forest area covers 249 ha and urban and built-up land accounts for 572 ha; andthus only 821 ha (10% of the total land area) is considered as constrained areas forshrimp farming (Table 4).
The classified areas for four land requisites containing 13 land characteristics andconstraint layer are summarized in Table 5 and Fig. 3. The highly suitable areas forpond construction are found along the coastline with a low slope (o2) andappropriate elevation (2.0–2.5m). These areas cover the existing aquaculture areas,range land and salt farms. Around 83–87% of total land area is suitable for shrimpfarming in terms of slope, soil thickness, and elevation (Table 5). Suitable and
ARTICLE IN PRESS
Table 5
Areas (ha) and different suitability levels (%) of land for shrimp farming in the study area (total land area
is 8281 ha)
Land use
requirement/land
characteristics
Highly suitable Suitable Marginally
suitable
Not suitable Constraints
(ha) (%) (ha) (%) (ha) (%) (ha) (%) (ha) (%)
Potential for pond
construction
Land use type 1690 20 513 5 2998 37 2259 27 821 10
Slope 7186 87 507 6 189 2 399 5 0 0
Soil thickness 6900 83 528 6 853 10 0 0 0 0
Elevation 6907 83 280 3 506 6 588 7 0 0
Sub-overall 2372 29 2808 34 2107 25 173 2 821 10
Soil quality
Soil type 3383 41 1816 22 2404 29 678 8 0 0
Soil pH 7632 92 649 8 0 0 0 0 0 0
Soil texture 3692 45 1829 22 2760 33 0 0 0 0
Sub-overall 3239 39 1823 22 2398 29 0 0 821 10
Water availability
Distance to sea 3542 43 2122 26 1329 16 1288 16 0 0
Water source 3833 46 3433 41 1015 12 0 0 0 0
Sub-overall 3703 45 2403 29 1354 16 0 0 821 10
Infrastructure and
socio-economic
factors
Distance to road 5962 72 1898 23 421 5 0 0 0 0
Population density 2596 31 2633 32 1595 19 1457 18 0 0
Distance to local
market
3854 47 3989 48 437 5 0 0 0 0
Distance to
hatchery
6514 79 1767 21 0 0 0 0 0 0
Sub-overall 1610 19 4761 57 1053 13 36 0 821 10
Overall suitability of
site
2604 31 3280 40 1544 19 32 0 821 10
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 59
marginally suitable areas for pond construction are located in crop land and mixedorchard cultivation areas.
The area of highly suitable soil quality for shrimp farming is 39% of the total landarea (Table 5) located mostly in agricultural land in the western part of the studyarea (Fig. 3).
Nearly half of the study area (46%) can access brackish and sea water during hightides (3.7–3.8m) through extensive brackish water irrigation networks. However, theexpansion of brackish water area is restricted by the main road that prevents theintrusion of seawater to adjacent paddy fields. The freshwater irrigation areaaccounts for 45% of total area, mostly covering paddy field and other crop land.
ARTICLE IN PRESS
Fig. 3. Suitability maps of different land use requirements for shrimp farming.
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6360
Only 12% of the total area is not covered by irrigation networks. Areas within ashort distance to the sea (o 1 km) account for 43% of the total area. As a result, 45%of highly suitable land for shrimp farming in terms of water availability is locatedalong the coast and bordered with the sea and main roads. The area further from thesea was less suitable for shrimp farming due to less water availability (Fig. 3).
ARTICLE IN PRESS
Fig. 4. Overall potential land suitability map for shrimp farming.
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 61
Infrastructural and socio-economical factors greatly limit areas most suitable(19%) for shrimp farming (Table 5). This is mainly due to the high populationdensity in the study area. Domestic wastes in areas with dense population can havesignificant impacts on shrimp farming. Well-established and easily accessiblehatcheries favor shrimp farming development in the study area. More than half ofthe total area (57%) is classified as suitable for shrimp farming in terms ofinfrastructure and socio-economic factors.
The overall evaluation shows that about 31% of the total land area (2604 ha)located along the coastline and the river in the east and north-east parts of the studyarea is highly suitable for shrimp farming, (Table 5; Fig. 4). Existing aquacultureponds cover most of the highly suitable areas, while the rest are found in the rangeland, salt farms, and agricultural land. Most of the suitable areas (about 40% of thetotal area) for shrimp farming are agricultural land.
5. Discussion
Land evaluation using GIS techniques estimates that highly suitable land area forshrimp farming is 2604 ha in Haiphong province. However, existing shrimp farms
ARTICLE IN PRESS
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–6362
cover only 1690 ha of land in the study area. This indicates that shrimp farming canbe further expanded to other areas such as low-productivity salt farms and rangeland. However, the land suitability evaluation in this study is only based on the siteselection criteria for shrimp farms. Other important factors such as political andenvironmental issues should also be considered in the decision-making process underintegrated coastal zone management scheme.
The total area in the present study is relatively small, covering only 8281 ha of thecoastal land area. Several important factors such as climate and water quality are notconsidered in the present study due to little variation in climatic factors within thestudy area and difficulties in extrapolating from a small number of collected samples.These factors may be useful for land evaluation in larger areas.
The present study demonstrates the potential usefulness of GIS in combinationwith site selection criteria for aquaculture [2] development in land evaluationframework [21] to evaluate land suitability for shrimp farming. However, thisapproach requires up-to-date and accurate thematic information to make preciseestimations. Moreover, the selection of suitability ranks and identification ofimportant factors have direct influence on the results; thus, it is imperative to involveas many stakeholders as possible to make the evaluation criteria during thedevelopmental stage of land evaluation.
Incorporating key socio-economic and environmental factors into land evaluationfor shrimp farming will allow coastal zone planners to develop better land usepatterns and make decisions based on objective characteristics and requirement oroverall potential of land. This study shows that GIS databases of different formatsand sources can be used effectively to establish spatial models in land evaluation forshrimp farming. It is expected that the land evaluation model developed in this studyis useful to identify suitable areas for shrimp farming in terms of efficient incomegeneration, effective conservation, and sustainable land management.
Acknowledgements
The authors wish to express their gratitude to Prof. C.K. Lin (AIT) for hisinvaluable guidance and advice over the years, and Dr. Vu Dung and Mr. Tran VanMuu of Research Institute for Aquaculture No. 1 of Vietnam for their valuableassistance during the field survey. Many thanks are due to Mr. Tran Van Dien andMr. Nguyen Van Thao of Haiphong Institute of Oceanography of Vietnam for theirassistance and cooperation in satellite images classification and GIS analysis. Thisstudy is part of Mr. Dao Huy Giap’s doctoral dissertation research carried out at theAsian Institute of Technology.
References
[1] Rossiter DG. A theoretical framework for land evaluation (with discussion). Geoderma
1996;72:165–202.
ARTICLE IN PRESS
D.H. Giap et al. / Ocean & Coastal Management 48 (2005) 51–63 63
[2] Hajek BF, Boyd CE. Rating soil and water information for aquaculture. Aquaculture Engineering
1994;13:115–28.
[3] Mcleod I, Pantus F, Preston N. The use of a geographical information system for land-based
aquaculture planning. Aquaculture Research 2002;33:241–50.
[4] Aguilar-Manjarrez J, Nath SS. A strategic reassessment of fish farming potential in Africa. Rome:
Food and Agriculture Organization; 1998 100pp.
[5] Tran NT, Demaine H. Potentials for different models for freshwater aquaculture development in the
Red River Delta (Vietnam) using GIS analysis. Naga 1996;19(1):29–32.
[6] Kapetsky JM, Nath SS. A strategic assessment of the potential for freshwater fish farming in Latin
America. Rome: Food and Agriculture Organization; 1997 128pp.
[7] Aguilar-Manjarrez J, Ross LG. Geographical information system (GIS) environmental models for
aquaculture development in Sinaloa State, Mexico. Aquaculture International 1995;3:103–15.
[8] Krieger Y, Muslow S. GIS application in marine benthic resource management. In: GIS for the 1990s,
Proceedings National Conference (CISM), Ottawa, Canada, 1990.
[9] Ibrekk H, Kryvi H, Elvestad S. Nation-wide assessment of the suitability of Norwegian coastal zone
and rivers for aquaculture (LENKA). Coastal Management 1993;21(1):53–73.
[10] Salam MA, Ross LG, Beveridge CMM. A comparison of development opportunities for crab and
shrimp aquaculture in southern Bangladesh, using GIS modelling. Aquaculture 2003;220:477–94.
[11] Arnold WS, White MW, Norris HA, Berrigan ME. Hard clam (Mercenaria Spp.) aquaculture in
Florida, USA: geographic information system applications to lease site selection. Aquacultural
Engineering 2000;23:203–31.
[12] Pham MC. Sustainable coastal zone management using remote sensing and GIS: a case study in
Haiphong coastal area, Vietnam. M.Sc. Thesis, Bangkok, Thailand: Asian Institute of Technology,
1998. 128pp.
[13] Tran VB. Vietnam. In: Cruz DA, editor. Report of the APO seminar on aquaculture management
held in Taipei, Taiwan, 3–8 December 2001. Tokyo, Japan: Asian Productivity Organization; 2003.
p. 255–8.
[14] Rossiter DG. ALES: a framework for land evaluation using a microcomputer. Soil Use &
Management 1990;6:7–20.
[15] Malczewski J. GIS and multicriteria decision analysis. New York: Wiley; 1999 392pp.
[16] Nath SS, Bolte JP, Ross LG, Aguilar-Manjarrez J. Applications of geographical information systems
(GIS) for spatial decision support in aquaculture. Aquaculture Engineering 2000;23:233–78.
[17] Vietnam Cartographic Publishing House. Land use, and topographic maps of Haiphong. Hanoi,
Vietnam: Vietnam Cartographic Publishing House, 1999.
[18] Tran K. Soil of Vietnam. Hanoi, Vietnam: Agricultural Publishing House; 2000 441pp (in
Vietnamese).
[19] FAO. A framework for land evaluation. Rome: Food and Agriculture Organization of the United
Nations; 1977 87pp.
[20] Saaty TL. The analytic hierarchy process: planning, priority setting, resource allocation. Pittsburgh:
RWS Publications; 1990 287pp.
[21] Rossiter DG. Economic land evaluation: why and how. Soil Use & Management 1995;11:132–40.