It was clear that the district underwent serious land cover changes: deforestation between 1968 and 1992, with a simultaneous increase in rice land; a rapid decline in rice acreage from
Trang 1LAND COVER CHANGES BETWEEN 1968 AND 2003 IN CAI NUOC,
CA MAU PENINSULA, VIETNAM
T.N.K.D BINH1,*, NICO VROMANT2, NGUYEN THANH HUNG1,
LUC HENS3and E.K BOON3
1
National Centre for Natural Science and Technology, Department of Geography Information System and
Brussel, Human Ecology Department, Laarbeeklaan 103, B-1090 Brussels, Belgium
(*author for correspondence, e-mail: pvpialy@hcm,vnn.vn; fax: +84-8-8299618;
tel.: +84-8-8299618) (Received 19 January 2004; accepted 5 November 2004)
Abstract Since 2000, the shrimp industry expands at a fast rate in the coastal areas of the Mekong Delta, Vietnam Shrimp farming is known for its negative impact on the coastal environment However, other human interventions like agriculture and urbanization also deteriorate the coastal environment The land cover changes between 1968 and 2003 were determined and analyzed for the Cai Nuoc dis-trict, Ca Mau Province, Vietnam, using photos from 1968, 1992 (aerial photographs), 1997/98 (Spot) and 2003 (Landsat) It was clear that the district underwent serious land cover changes: deforestation between 1968 and 1992, with a simultaneous increase in rice land; a rapid decline in rice acreage from
1997 onwards, and, simultaneously, a blitz-increase in shrimp farming area The forest area declined by 75% between 1968 and 2003 About 40% of this loss could be attributed to shrimp farming, while the remaining 60% was attributed to needs for agricultural land Still, at present, shrimp farming is the major source of mangrove loss in the district In 1999 shrimp farms covered 6.374 ha, in 2000 they cov-ered 61.049 ha of the Cai Nuoc area The swap from rice cultivation to shrimp farming was most-prob-ably driven by households’ hopes for a higher income It must be feared that the shrimp industry will have a negative impact on the environment (e.g salinization) and on the livelihood of the district’s
Key words: mangrove, land cover change, remote sensing, shrimp, Vietnam.
1 Introduction About 100 years ago Vietnamese farmers started extensive shrimp farming (Nhuong et al., 2002) Since the 1980s at the start of the Doi Moi economic liberalization policy, modern intensive monoculture shrimp farming gradu-ally replaced the extensive farming systems as the Vietnamese Government explicitly encouraged the development of the shrimp farming industry
DOI 10.1007/s10668-004-6001-z
Readers should send their comments on this paper to: BhaskarNath@aol.com within 3 months of publication of this issue.
Trang 2(Nhuong et al., 2002) Indeed, the Vietnamese ‘‘National Master Plan’’ for aquaculture demands for an area expansion and intensification of shrimp aquaculture (Nhuong et al., 2002) Since 2000, when the government issued the resolution 09/NQ-CP allowing farmers to transform coastal saline rice fields into shrimp farms, the shrimp sector expanded at a fast rate (Nhuong et al., 2002): in 2001, 478,000 ha were used for shrimp farming in Vietnam, an increase by 42.6% as compared to the year 2000 (Mofi, 2002)
In the same year the area used for shrimp farming in Ca Mau was 202,00
ha, or 42.25% of the total shrimp area in Viet Nam (RIA2, 2003)
Worldwide, the shrimp industry is often blamed for its negative impact
on the environment (pollution, disease outbreaks, salt water intrusion), and more specifically on the mangrove forests (Wolanski et al., 2000; Dahdouh-Guebas et al., 2002; EJF, 2003b) But not only aquaculture can be blamed for mangrove loss, as other human interventions like agriculture, urbaniza-tion, and even forestry and warfare have been blamed for recent losses of mangrove habitat (Fortes, 1988; Marshall, 1994) Indeed, anthropogenic processes exert enormous pressure on tropical environments (Validea et al., 2001) In the Mekong Delta, the coastal complex used to be covered by natural vegetation, but, after the Vietnam war, land reclamation for rice cropping and shrimp culture began (Tanaka, 1995) As a result the coastal area is considered to be the region that has experienced the greatest changes in land use in the last few decades in the Mekong Delta (Tanaka, 1995) In this paper these changes are investigated Besides that, the paper tries to determine the impact of shrimp farming on land cover in the Ca Mau Peninsula, in general, and on mangrove forests, in particular The land cover changes between 1968 and 2003 were determined and analyzed for the Cai Nuoc district, Ca Mau Province, Mekong Delta, Vietnam Pho-tos from 1968, 1992, 1997/98 and 2003 were used to analyze the changes
2 Material and methods 2.1 Area Description
The 835 km2 Cai Nuoc district (Figure 1) is situated in the South-West of
Ca Mau province, Mekong Delta, Vietnam Its geographic coordinates are 806¢30¢¢ to 908¢00¢¢ latitude and 10447¢00¢¢ to 10509¢15¢¢ longitude The average annual rainfall ranges from 2000 to 2400 mm (Sub-Institute of Geography at Ho Chi Minh City, 2001) The area has two main seasons: (i) the wet season (from May to October) gets on average more than
2100 mm of rainfall, making up 90% of the total annual rainfall, and (ii) the dry season (from September to April) receives on average 200 mm of rainfall The average daily temperature ranges from 26.5 to 27.3 C The minimum daily temperature is always above 15 C The relative humidity
Trang 3varies between 87.0 and 89.4% during the wet season, and between 80.4 and 85.5% during the dry season (Sub-Institute of Geography at Ho Chi Minh City, 2001) The number of hours of sunshine ranges from 4.7 to 5.0 h day)1 during the rainy season and 7.4 to 8.1 h day)1 during the dry season (Sub-Institute of Geography at Ho Chi Minh City, 2001)
The Cai Nuoc area is a lowland delta plain with an average elevation of 0.2 m The area has three soil groups: (i) fluvisols (alluvial land) cover 44234.04 ha (54.0%), (ii) thionic fluvisols (acid sulphate soils) cover 33309.20 ha (40.69%) and (iii) salic fluvisols (alkaline soils) cover 4318.78 ha (5.28%) (Sub-Institute of Geography at Ho Chi Minh City, 2001) The water flow regime in the study area is complicated as it is affected by the Gulf of Thailand and the East sea (Sub-Institute of Geography at Ho Chi Minh City, 2001) The district is affected by salinity for over 6 months (Brennan et al., 2002)
2.2 Data Sources and Treatment
Three kinds of data sources were used to setup the Cai Nuoc land cover maps for the years 1968, 1992, 1997/98 and 2003: (1) a common topo-graphic base map, (2) aerial photographs of the year 1968 and 1992, and (3) satellite images for the years 1997/98 and 2003
The topographic base map (prepared under the direction of the Depart-ment of Defense and published by the U.S Army Topographic Command, Washington D.C) has the following specifications: scale 1/50.000, spheroid Everest, grid 1000 m UTM zone 48, vertical datum mean sea level at Ha
Trang 4Tien, and horizontal datum Indian Datum 1960 The map was compiled on
3 December 1969 (3rd edition) The map sheets was digitized and geo-refer-enced utilizing Drawing board III
For the 1968 land cover map, 58 aerial photographs (scale 1/25.000) were used For the 1992 land cover map, 154 pictures (scale 1/22.000) were used For both sets of photographs the horizontal covering was 60% while the vertical covering was 30% The 1968 photos were made by the U.S army, while the 1992 photos were obtained from the Vietnamese Ministry
of Defense Stereoscopy was used to analyze the aerial photos Once the different land cover categories were defined, the aerial photos were used to draw the contours of the land cover pattern on the base map A Cal comp digitizer (Drawing board III) using AutoCAD R12 software was used to digitize the map
The 1997/98 land cover map was derived from two Spot 4 images The
2003 land cover map is based on two Landsat 7 ETM+ images The satel-lite images were interpreted using the ENVI software package (Research Systems Inc.) Pre-processing of the images included geometric correction and image enhancing The images were analyzed with the Normalized Dif-ference Vegetation Index (NDVI) and the Principal Component Analysis (PCA) The NDVI was used to find out the vegetation distribution PCA was used to produce uncorrelated output bands, to sort out noise compo-nents, and to reduce the dimensionality of the data set (Richards, 1994) Plotting the NVDI output versus the PCA corrected pixels of band 3 allowed discrimination between several kinds of land cover First, an unsu-pervised classification (to recoginize the general type of land use or land cover) was obtained with the Isodata clustering algorithm (Tou and Gonz-alez, 1974) Second, the maximum likelihood method was used to obtain a supervised classification (Richards, 1994) With the Kappa coefficient method (McMorrow, 2002), the accuracy of the land cover classification was accessed as 99.5% for the Spot image and 98.7 for the Landsat image The land cover map was printed and regions with uncertain classification were marked for a ground truth check Following the ground truth check, the data were transformed to a vector format
3 Results 3.1 Land Cover Categories
The study of the maps enabled us to identify 15 land cover categories:
1 Rice monocropping, 1 crop per year: Farmers use traditional rice varie-ties with a growth period of 180–210 days The rice crop is grown from May to December-January This cropping system is practiced on both
Trang 5acid sulphate and alkaline soils Rice is transplanted The cropping sys-tem is characterized by a low input (Sanh et al., 1998)
2 Rice monocropping, 2 rice crops per year: Farmers use short duration (95–110 days) high yielding rice varieties: the first rice crop is grown from May/June to August/September; the second rice crop is grown from September/October to November/December Some farmers opt for a short duration crop followed by a long duration crop In this case the short-duration rice is grown from April to the middle of July and the long duration crop from the end of July to the middle of January Farm inputs are high and typically include fertilizers and pesticides (Sanh et al., 1998)
3 Rice cropping followed by vegetable cropping: The rice crop is grown from May to December/January and the vegetable crop (e.g sweet potato, spinach, onions) is grown from February to April
4 Vegetable farming: Farmers grow several short-duration crops such as watercress, ladyfingers, onions, shallots and chilli
5 Settlement and gardens: residential houses with adjoining fruit gardens (coconut, banana tree or other kinds of fruit)
6 Village: residential houses without gardens
7 Village and vegetables: residential houses with adjoining gardens used for horticulture activities
8 Nypa palm: Nypa fructicans is a mangrove species of the Old World In
Ca Mau, Nypa is typically found along canals and creeks
9 Cypress forests are dominated by Cypress trees
10 Rhizophora forest: forest dominated by Rhizophora spp
11 Mangrove forests are dominated by a mixture of mangrove trees like Rhizophora spp., Sonneratia sp and Avicennia sp
12 Avicennia forests are dominated by Avicennia species
13 Wasteland: is land that cannot be exploited for agricultural production due to high salinity
14 Mangrove–shrimp farming systems: farmers grow mangrove and shrimp (Penaeus monodon Fabricius) simultaneously The area dedi-cated to mangrove is good for 50 to 70% of the total surface covered
by the system The remaining area is used for shrimp farming
15 Shrimp monoculture: farmers grow shrimp (Penaeus monodon Fabri-cius) in ponds
To enhance the reading of the land cover maps, these 15 categories were regrouped into 6 major land cover groups: agriculture, rural settlement, wasteland, forest, mixed mangrove–shrimp farming systems and shrimp monoculture (Table I)
Trang 6compared to
compared to
compared to
compared to
compared to
Rural settlement Settlement/ gardens
Rural settlement
Rural settlement Village- Vegetable
Rhizophora forest
Mangrove forest
Mixed mangrove– shrimp farming systems Mangrove– Shrimp
Shrimp monoculture Shrimp farming
Trang 73.2 The 1968 Land Cover Map
The 1968 picture revealed 11 land cover categories (Table I) further grouped into 4 major categories (Figure 2) Rice monocropping with 1 rice crop per year covered more than half of the total land area The rice monocropping area was situated in the Eastern part of the district,
seven major land cover groups: (1) Agriculture including: 1 rice crop per year, 2 rice crops per year, 1 rice crop and 1 vegetable crop per year, and vegetable farming; (2) Rural settlement including: village fruit gardens, village, village mixed vegetable gardening; (3) Forest including: Nypa palm, cypress, Rhizophora forest, mangrove forest and Avicennia sp., (4) wasteland; (5) mixed mangrove–shrimp farming systems; (6) monoculture shrimp farming; and (7) water bodies such as canals and rivers.
Trang 8away from the sea The Western and Southern part of the district is covered by forest (19,509.74 ha or nearly 25% of the total area) Settle-ments and garden were mainly situated along thecanals Mangrove– shrimp farming systems and shrimp monoculture were non-existent
3.3 The 1992 Land Cover Map
The 1992 picture revealed 11 land cover categories (Table I) further grouped into 6 major categories (Figure 2) The major land cover was the rice monocropping system with 1 rice crop per year, mainly situated in the North-eastern part of the district Two other important farming systems are the settlement–garden system and the rice–vegetable system, good for respectively 23.76 and 18.01% of the total area A strip of mangrove forest borders the coastal area Mixed mangrove–shrimp farming systems are good for nearly 10% of the total land cover, mainly situated in the Wes-tern and Southern part of the district Intensive shrimp monoculture is marginal Settlements and gardens were aligned along the rivers, canals and main roads
3.4 The 1997/98 Land Cover Map
The 1997/98 picture revealed 9 land cover categories (Table I) further grouped into 6 major categories (Figure 2) Rice monoculture systems with two rice crops per year (mainly found in the North-eastern part of the Cai Nuoc district) formed the main land use system, followed by price mono-culture systems with one rice crop per year Together, both systems are good for slightly more than 60% of the total land area A small strip of mangrove forest bordered the coast line Behind this forest an area of
7639 ha of mangrove–shrimp farming systems can be found Along the Southern and Eastern part of the district borders shrimp monoculture sys-tems developed Both shrimp syssys-tems together are good for 15.43% of the Cai Nuoc land Settlement and gardens followed the canals and main roads
3.5 The 2003 Land Cover Map
The 2003 picture revealed 6 land cover categories (Table I) further grouped into 4 major categories (Figure 2) Shrimp monoculture systems (mainly situated in the North-eastern part of Cai Nuoc district) are good for 72.66% of the total land cover area Along the coastlines, the up to 300-m wide mangrove forest can be found, immediately followed by mangrove– shrimp farming systems People live in small villages or along canals or roads
Trang 93.6 Land Cover Changes
From 1968 to 2003, the Cai Nuoc district underwent serious changes in land cover (Table I) While agricultural (rice)-based farming systems cov-ered more than 50,000 ha in 1968, they had completely disappeared by the year 2003 Shrimp-based farming systems were non-existent in 1968, but dominant in 2003 (Figure 3) In the same period the forest area declined drastically (Figure 3 and Table I) The ratio ‘‘forested area/total/land cover’’ declined from 0.243 in 1968, to 0.109, 0.099 and 0.069 in 1992, 1997/98 and 2003, respectively (Table I) Accounting for the fact that mixed mangrove–shrimp systems have an area of 50 to 70% mangrove, the total forested area drops from 24.26% in 1968 to 5.18–6.42% in 2003 (Table II), a decrease by about 75% Between 1968 and 1992, the forest-type diversity was seriously reduced: While there were 5 kinds of forests in
1968 (Nypa palm, Cypress forest, Rhizophora forest, mangrove forest and Avicennia forest) only two remained in 1992 (Nypa and mangrove forest) Table III, which uses the data from Table II as input, shows that the biggest loss in mangrove area occurred between 1968 and 1992, when an estimated 63.2% of the original mangrove area was lost While the area lost between 1997/98 and 2003 was relatively small, a sharp increase in per-centage loss in mangrove area (4.18%) was noted
The total land area of the Cai Nuoc district increased from 80,420 to 82,142 ha between 1968 and 2003 New land is covered by mangrove Between 1968 and 2003, the area occupied by villages increased from 7.99
to 302.99 ha
Ca Mau province, Vietnam.
Trang 104 Discussion The Cai Nuoc district is a fast developing area In the period under study, the changes in land cover could be summarized as follows: (1) deforesta-tion between 1968 and 1992, with a simultaneous (2) increase in rice land, (3) a rapid decline in rice acreage from 1997 onwards, and, simultaneously,
a (4) blitz-increase in shrimp farming area
TABLE II The total forested areas in the Cai Nuoc area consists of two fractions: (1) Pure forest stands, and (2) Mixed mangrove–shrimp systems In these systems mangrove makes up 50–70% of the land area The table tabulates the total forested area for minimum and maximum possible area in for-est The real value lies probably somewhere in between.
Forest area
(pure stands)
Mixed shrimp–
mangrove area
Of which 70%
forested
Of which 50%
forested
Total forested
area
Pure stand
+ 70% shrimp–
mangrove
Pure stand
+ 50%
shrimp-mangrove
Total forested
area (average values)
per-centage of original area lost per year in Cai Nuoc district, Ca Mau province, Vietnam, for the periods 1968–1992, 1992–1997/98, 1997/98–2003 and the overall period 1968–2003 The average forest area data from Table II were used as input for this table.
Considered
period
Mangrove area
at onset of
period (ha)
Mangrove area
at end of period (ha)
Percentage loss between onset and end of considered period
Annual rate
Percentage of area lost per year in th considered period