Evaluation of Land Suitability for Crab Culture: A Methodological Study using GIS


Salam, M. A.
Assistant Professor, Department of Aquaculture,
Bangladesh Agricultural University, Mymensingh-2202, Bangladesh.
Phone: +88 091 55695-7/2282, Fax: +88 091 55810
E-mail: salam@royalten.net

Ross, Lindsay G and Beveridge, C. M. Malcolm
Institute of Aquaculture, Stirling University, FK9 4LA, Scotland, UK.

Introduction
Aquaculture has benefited significantly from the use of Geographical Information Systems (GIS) and Remote Sensing in recent years and these techniques have jointly showed their capabilities in the evaluation and assessment of suitable sites for a variety of culture systems. A GIS is a combination of computer hardware and specialised software, which is used to store, manipulate and analyse data of diverse kinds with a common geographical base. With such automation, reporting can be in the form of tabular data, graphics, and most importantly maps (Burrough, 1986).

GIS have been developed rapidly in the past decade and begin to be used for aquaculture and fisheries development studies. To date, GIS has been applied to regional, country wide or sector studies for aquaculture, where human resources, specific site, economics, markets and socio-cultural resources were used (Kapetsky et al.,1988; Meaden and Kapetsky,1991). On a large scale, a number of sector studies have been made, for example, in the African continent (Kapetsky, 1994), on the Gulf of Nicoya, Costa Rica (Kapetsky,1987), on Johor, Malaysia (Kapetsky,1989), on Louisiana state, USA (Kapetsky et al., 1990) and on Ghana (Kapetsky et al., 1991). Species related GIS for catfish culture was developed by Kapetsky et al., (1988), salmonid cage culture by Ross et al., (1993) and white prawn modelling by Scott et al., (1998).

Although there is considerable potential for shrimp culture in South-western Bangladesh, aquatic resources management in the region have not to date been fully integrated with the rural economy. These resources have great potential and could be further developed to help meet the increasing demand for fish and shrimp in the area. This is best done rationally, establishing a structured decision making and planning scheme, which would play a vital role as a promoter of efficient development.

In this paper focus has been given to using Geographical Information Systems to locate, inventory and compare areas suitable for crab culture at a regional level.

Materials and Methods

Study Area
The study area is located 300km Southwest of the capital city of Dhaka, Bangladesh, and bordered with India. The area is situated on the vast coastal plain at the apex of the Bay of Bengal and encompasses approximately 14,000 km2 (Figure 1). The area lies between 21030' and 23015' North longitudes and 89000' and 90000 East latitudes and includes the world’s largest continuous mangrove forest, the estuarine marshlands and numerous rivers, canals and their tributaries (Viju, 1995 and Giri and Shrestha 1996). The terrain is relatively flat with elevation ranging from sea level to 5 m above the mean sea level.

The region is one of the most promising areas for aquaculture due to two major factors. Firstly, it’s fresh and coastal water resources are the most abundant in the country. Secondly, the world’s largest continuous mangrove forest is situated in the region, which provides a food source and nursery for the offshore fishery, protection of the coasts from storm surges and cyclones, domestic and commercial products, recreation and tourist services, habitat for various fin fishes, crustaceans and molluscs. It is also the habitat for the Royal Bengal tiger and other national heritage. Culture of fresh and marine water fish, shrimp and other crustacean species are highly important as they can be easily integrated with other activities such as agriculture and livestock rearing.

Khulna region has around 104,624 hectares of shrimp farm in production (DOF, 1994-95). Additionally, there is one principal research and shrimp fry and 72 fish seed production centres at Government level and 12 shrimp fry production centres in the private sector. Likewise, there are 272 fish seed centres in the area in private hand. However, despite this actual aquaculture development, there is a lack of crab seed production centre in the region. The most common culture systems have minimum management, hence are considered to be a subsistence activity (Khan and Hossain,1996).

Figure 1 Location of the study area in Bangladesh and World context
The GIS software used in this study was IDRISI for Windows version 2.0, a raster based GIS, developed at Clark University, USA. TOSCA digitising system (Version 2.03), DIGI-EDIT (Version 1.050) and CalComp Drawing Board III were used to digitise and edit vector layers. The software operated on a Pentium Pro 200 MHz processor with 128Mb RAM, 10 Gb hard disk. Display was via an iiyama Vision MasterTM 21? colour monitor using a Matrox Millennium 8Gb video card.

Remotely sensed data sources
The primary data sources for this study were two adjacent Landsat TM satellite images containing three multispectral bands (2, 3 and 4), acquired on 9 and 16 February 1996. A cloud free area of (column 3676 by 4915 rows) 18067,540 pixels was subset from the original two full scene image using WINDOW module of IDRISI for windows (version 2.0). This sub-scene image was geometrically rectified to the BANGTM46 grid co-ordinate system using 140 reference points on the 1:50,000 scale map. The rectification was achieved at an accuracy of 12.26 m.

Thematic maps
Data were extracted from a range of sources. Soil texture and soil pH were extracted from soil reconnaissance survey maps prepared by the Soil Resources Development Institute (SRDI) during the period of 1972 to 1975. Land use, flood, storm surge and drought were taken from SRDI maps of 1997. Soil and water salinity was taken from SRDI soil salinity report 1997 and Surface Water Modelling Centre (SWMC) report 1996. Roads, railways, rivers, international boundary and other topographical information were taken from topographical maps produced by Survey Bangladesh, scale 1:50,000, prepared during 1972 and 1979. Population density, literacy, agricultural by-product, animal wastes were taken from Bangladesh Bureau of Statistics (BBS) 1996 and BBS pocket dictionary 1996. Other data were taken from different sources such as atlas, journals, periodicals and published papers.

Construction of Sub-models
Analysis was first made of essential criteria for crab farming potential, and the spatial data obtained for these criteria are described. The primary data were then grouped into series of sub-models.

Water sources: When evaluating a site for aquaculture potential, the total land and water resources of the site are the prime determining factors. However, growing demands for water from an expanding aquaculture industry competes with other water users such as irrigation, industry, navigation, drinking, and for meeting environmental needs day by day for this limited resources (Muir and Beveridge, 1987; Patricia, 1999; Phillips et al.1991). For this study, an indication of water availability for crab culture in the coastal zone was obtained by constructing a layer based on of perennial streams, tidal canals and river channels.

Available Land: All land in various categories, which would not be available for development was removed from consideration, such as rivers, roads, mangrove forest and urban development. The land, which was considered most suitable for this study, was that which is near the coastline and not intensively used for agriculture, in order to minimise the conflict with other activities.

Support: Supports services were designated as the extension services and other help which farmers can obtain from Non Governmental Organisations (NGO), Government offices, Research stations, Universities and by agglomeration with other farms. Using tabular data, a layer was compiled for support from NGO, Government offices, Research stations, Universities and agglomeration with other farms for crab farming in the region.

Soil chemistry: Soil quality is also an important land criterion. From the soil reconnaissance map of the Soil Resources Development Institute (SRDI) during the period of 1972 to 1975, a soil texture and pH layer were created by digitising the map and converted into raster format. Soils were classified according to their texture (Figure 3, Coche,1985). Data on salinity and pH collected during a 1998 field visit were also used to develop the soil sub-model.

Risk assessment sub-model: To compile the risk sub-model, different layer as winter rainfall, pollution from the industries and town centre, drought and elevation were added together from maps and tabular data.

Sources of crab seed: Crab seed have to be available for successful culture practice. If hatchery reared seeds are not available then the juvenile crab must be collected from the wild. Thus, there is an advantage in knowing where crab seed are most abundant so that they can be collected. In Khulna region, most of the seeds come from the environment, especially from the Sundarban mangrove forest.

Water chemistry: The water chemistry sub-model was prepared from different sources of thematic layer and tabular data. Water temperature was calculated from air temperature using regression equation (Kapetsky,1994) and likewise, dissolved oxygen was derived from temperature and salinity relationship describe by (Colt, 1984). Moreover, water salinity was taken from Surface Water Modelling Centre (SWMC) report (1996) and data on water pH collected during field visit (1998), were also used to develop the water sub-model.

Figure 2: Comparison of the best soils for pond construction with the soil texture classes (after Coche, 1985).
Infrastructure: Infrastructure was considered as an important criterion for crab farming. Therefore, roads were extracted by digitising from the map 1:50,000 scale of Survey Bangladesh and Landsat TM image. This sub-model also included distance from the markets.

Weighting procedure of factors and sub-models
In order to determine potential sites for brackish water crab farming, thirty environmental and economic criteria were selected and compiled. These criteria were of two types: factors and constraints. A factor is a criterion that enhances, or detracts from, the suitability of a specific alternative under consideration. It is therefore measured on a continuous scale. On the other hand, a constraint serves to limit the alternatives under consideration. Suitability ratings for each criterion were employed, and each factor was reclassified on a scale from 1 to 4. The next step was to establish the weighting for each of the factor according to the pair wise comparison matrix of Saaty (1977). Using the weighting procedure, eight sub-models were developed. Finally, a system-oriented model was generated for brackish water crab culture. To illustrate the approach, eight sub-models were used to the final matrix (Table 1). The weighting of each sub-model was derived using the weight module, based on Saaty’s matrix. The weights were then used in the Multi-Criteria Evaluation (MCE) module to show potential sites for crab culture
Table 1: Weighting derived by the pair-wise comparison matrix for assessing crab culture potential in south-western part of Bangladesh (numbers show the rating of the row factor relative to the column).

Factors	Land use	Infrastructure	Risk	Support	Water chemistry	Soil chemistry	Seed sources	Water sources	Weightings
Land use	1								.1845
Infrastructure	1/3	1							.0771
Risk	1/7	1/3	1						.0576
Support	1/4	1/2	1	1					.0406
Water chemistry	1	4	5	5	1				.2056
Soil chemistry	1	3	4	4	1/2	1			.1564
Seed sources	1/6	1/5	1	1/2	1/7	1/7	1		.0270
Water sources	2	5	7	6	2	2	7	1	.2572
Total									1.00

Results of GIS Study
The results for each criterion are presented firstly for each sub model and the sub models are then used in final analysis. The reclassified surface area in each sub model are summarised in Table 2 and the corresponding spatial distributions of the areas are shown as maps (Figures 3-10).

Areal distribution of sub-models
Based on actual intensity of rice and other crop farming system, about 64.4 % land is very suitable for crab culture, and 15.6%, 20.0%, and 0% are moderately, marginally and unsuitable respectively (Table 2 and Figure 3). The spatial distribution of perennial streams and rivers to show overall water availability for crab farming is shown in Figure 4.
Table 2: Summary of the results of the criteria combined into sub-models expressed as percentages of the total surface of the area.

Name of Sub-model	Very suitable (%)	Moderately suitable (%)	Marginally suitable (%)	Unsuitable (%)
Land use	64.4	15.6	20	0
Infrastructure	0.757435	16.4498	65.85322	16.93954
Risk	3.91	54.67	41.42	0
Support	1.02	22.03	68.48	8.47
Water chemistry	28.12	57.02	14.86	0
Soil chemistry	9.84	71.05	19.09	0.02
Seed sources	18.95	19.39	54.82	6.84
Water sources	30.5	23.5	24.6	21.4

Optimum suitability of soils occurs in the middle of the area where other physical factors are more or less favourable. There are large areas of moderately suitable soils found in Satkhira, Bagerhat, Khulna, Jessore and Narail districts (Figure 5). However, most of the areas in Gopalganj district and few patches in Khulna and Bagerhat districts fall into marginally suitable class due to presence of peat soils.

In the combined risk sub-model, it is apparent that much of the unsuitable area is in the lower part of the area (Figure 6), principally because this area is very low lying and extremely prone to flooding.

Areas with very suitable or moderately suitable water chemistry are widely distributed in the lower part of the region because of uniform salinity, temperature and dissolved oxygen (Figure 7). Similarly, areas with support potential for crab farming at an optimum level are patchily distributed except for a few areas in the Khulna and Bagerhat districts (Figure 8). This is because most of the NGO’s and Government offices are located in the town centre, which is far away from the farming sites.

A look at infrastructural potential for crab farming indicated that large parts of the area are unsuitable, but some small patches in the north-western and central region are very suitable to moderately suitable (Figure 9). Most of the area is marginal to unsuitable due to the absence of major roads and distance from the markets.

Overall, Figure10 clearly highlights the suitability of the lower parts of the region for crab farming, with its favoured areas of high quality water, of a uniform salinity and having a consistent flow of tidal water. These areas also have good access to road transport; high agglomeration potential and they are not too far from the market outlets.

It is important to note that some of the areas, which have already been developed for crab farming, only scored as moderately suitable. These include most of the Bagerhat and Satkhira region, and some parts of Khulna district. In these areas soil quality is lower because of high proportion of fine clay, though the major problem is the distance from most of the markets and processing plants. However, some areas showed suitable for crab culture in the lower part of Jessore district although these are not in culture at present due to low salinity of available water.

Discussion
The results predicted as suitable from this GIS evaluation are partly verified by the existing farms in the area. Comparison has been made with the satellite image and published data. GIS predicted that 228,111 hectares of land are suitable for crab farming in the region. However, there was some variation among the predicted and actual locations possibly due to weighting method used and the different factors employed in the models.

Much of the area ranked most suitable for brackish water crab culture, are situated near the riverbanks in the tidal zone. This ensures easy access to saline water. However, most of the areas were ranked as moderately suitable because of lack of other facilities, obtaining water and drainage difficulties. Large water bodies, which contain fresh water, were not considered of potential use because of problems with the low salinity. Natural forest and wild life refuges, Sundarban, roads, rivers and urban areas were not considered for culture activities.

Although about 30.5% of coastal land within the reach of water sources for shrimp farming, when risk factors were excluded the suitability of land for crab farming was about 13.28%. However, when the risk factors were included the area reduced to 3.17% of the total areas. This is due to a great deal of the suitable land being within one meter of mean sea level, which is vulnerable to cyclone or tidal bore. Additionally the area is far away from the markets. Moreover, although there is good access to river transport systems, access to roads is poor.

There are ample areas ranked as moderately suitable but not yet under culture. Crop agriculture, which is a traditional source of income, has not yet been affected by the development of brackish water aquaculture. Although new sources of income have been created by brackish water aquaculture in the region, there has been drastic reduction of livestock and poultry in the farming areas. (Rahman, et al., 1995; Khan and Hossain,1996).

It is interested to note that, unlike some other areas of Bangladesh only one crop of crab can be cultured in the region as the salinity of water falls to nil in the months of July-December. This is a suitable time for rice production and fresh water prawn and fish culture (Rahman, et al., 1995).

Assessing the outcomes in terms of money from these suitable areas predict by GIS we need to identify the probable area required for roadways, dikes, and inlet and outlet channels and for the guardrooms. If we allow 10% of land to be used for that purpose, still we will have 205,300 hectares of land, which will be the actual pond area available for culture. At present, average production is 240 kg /hectares and the price per kilogram of crab is US $3. From the above calculation total output based on a single crop could be 49,272MT/year and total earnings will be about 148 million US dollar per year (based on 1 US $=60 Taka) from this region.

Conclusion
The results of this study are indicative of the modelling power of GIS for this application and could be used to refine the models in future, particularly if supported by more field verification. There is considerable potential for further exploitation of GIS for optimisation of competing aquatic production activities and their interaction with land based farming systems. The GIS planning process has an important role, particularly where land use patterns are intensive, and where developments must be sustainable in terms of sensitive environment issues, as in most parts of Bangladesh.

Reference

• BBS. 1997. Statistical Pocketbook of Bangladesh, Bangladesh Bureau of Statistics, Statistics Division, Ministry of Planning, Government of the People Republic of Bangladesh, Dhaka, Bangladesh.
• ———. 1996. Statistical Yearbook of Bangladesh, Bangladesh Bureau of Statistics, Statistics Division, Ministry of Planning, Government of the People Republic of Bangladesh, Dhaka, Bangladesh.
• Burrough, P. A. 1986. Principles of Geographical Information Systems . Oxford, U.K.: Oxford University Press.
• Coche, A. G. 1985. Simple methods for aquaculture : Soil and freshwater fish culture. FAO Traning Series, 6. Rome-Italy: FAO.
• Colt, John. 1984. Computation of dissolved gas concentrations in water as functions of temperature, salinity, and pressure. Washington, D. C.: American Fisheries Society.
• Department of Fishreies (DOF). 1994-1995. Fish catch statistics of Bangladesh, Department of Fishreies, Government of Bangladesh, Dhaka, Bangladesh.
• Giri, C., and S. Shrestha. 1996. Land cover mapping and monitoring from NOAA AVHRR data in Bangladesh. International Journal of Remote Sensing 17, no. 14: 2749-59.
• Kapetsky, James M. 1989. "Malaysia. A geographical information system for aquaculture development in Johor State. Report prepared for the project Land and Water Use Planning for Aquaculture Development." FAO, Rome Italy.
• ———. 1985. "Mangroves, fisheries, and aquaculture." SELECTED WORKING PAPERS SUBMITTED TO THE ELEVENTH SESSION OF THE ADVISORY COMMITTEE OF EXPERTS ON MARINE RESOURCES RESEARCH, Suppliment, 338. FAO, Rome, Italy.
• ———. 1994. "A strategic assessment of warm-water fish farming potential in Africa ." CIFA TECHNICAL PAPER, Food And Agriculture Organization , Rome Italy.
• Kapetsky, James. M., John M. Hill, and Worthy L. Dorsey. 1988. A Geographical information systemfor Catfish Farming Development. Aquaculture 68: 311-20.
• Kapetsky, James M., John M. Hill, Worthy L. Dorsey , and David L. Evans. 1990. Assessing Potential for Aquaculture Development with a Geographic Information System. Journal of the World Aquaculture Society 21, no. 4: 241-49.
• Kapetsky, James. M., McGregor Lloyd, and Nanne E. Herbert. 1987. Ageographical information system and satellite remote sensing to plan for aquaculture development: A FAO-UNDP/GRID co-operative study in Costa Rica, Food and Agriculture Organization of the United Nations, Rome.
• Kapetsky, James M., U. N. Wijkstrom, N. MacPherson, M. M. J. Vincke, E. Ataman, and F. Caponera. 1991. "Where are the best opportunities for fish farming in Ghana? The Ghana geographical information system as a decision making tool for fish farming development." Technical assistance and investment framework for aquaculture in Ghana. Field technical report 5. FAO, Rome, Italy.
• Khan, Yusuf Sharif Ahmed, and M. Sahadat Hossain. 1996. Impact of shrimp culture on the coastal environment of Bangladesh. International Journal of Ecology and Environmental Sciences. 22, no. 2: 145-58.
• Meaden, G. J., and James M. Kapetsky. 1991. Geographical information systems and remote sensing in inland fisheries and aquaculture, FAO Fisheries Technical Pape.
• Muir, J.F. and Beveridge, M.C.M. (1987). Water resources and aquaculture development. Archives of Hydrobiology, 28: 321-324.
• Patricia, W. (1999). Keynote: Integrating our World; Rivers of discord, rivers of peace. In: GeoSolutions: Integrating Our World. Vancouver, B.C., Canada, Adams Media, Vol.13.
• Phillips, M.J., Beveridge, M.C.M. and Clarke (1991). Impact of aquaculture on water resources. Brune, D.E. and Tamasso, J.R., (Eds.). In: Advances in World Aquaculture, 3 pp. 568-591.
• Rahman, Atiur, M. Aminul Islam, Indrajit Roy, Lanin Azad, and K. Shafiqul Islam. 1995. Shrimp culture and environment in the coastal region , Bangladesh Institute of Development Studies.
• Ross, Lindsay Glenn, Q. M. E. A. Mendoza, and M. C. M. Beveridge. 1993. The application of geographical information systems to site selection for coastal aquaculture: An example based on salmonid cage culture. Aquaculture 112, no. 2-3: 165-78.
• Saaty, T. L. 1977. A scaling method for priorities in hierarchical structure. Journal of Mathematics Ans Physiology 15: 234-81.
• Scott, Philip C., Tavares Marcos, M. Abdus Salam, and Lindsay Glenn Ross. 1998. A Modelling white shrimp fishery distribution for Sebetiba Bay, Brazil. GIS Planet 98Lisbon, Portugal.
• SRDI. 1997. Soil Salinity, Soil Resource Development Institute, Dhaka, Bangladesh.
• SWMC. 1996. Surface water simulation modelling programme phase III, Surface water modelling centre, Dhaka, Bangladesh.
• Viju, Ipe C. 1995. Issues in the management of the environment and natural resources in Bangladesh. Journal of Environmental Management 45, no. 4: 319.