GIS for Coastal Zone Management - Chapter 17

CHAPTER SEVENTEEN GIS for Assessing Land-Based Activities that Pollute Coastal Environments J.I. Euán-Avila, M.A. Liceaga-Correa, and H. Rodríguez-Sánchez 17.1 INTRODUCTION According to Heathcote (1998), the development of workable management options (structural and non-structural actions) in a watershed requires the identification of all point and non-point sources of pollution. Discharge of effluents from industrial, urban and sewage treatment plants where a pipe or diffuser outfalls into a water body are called “point sources” of contamination. Another type of source, called a “non-point source” (NPS), is described as the diffuse drainage of rainwater from urban, industrial and agricultural lands that can introduce nutrients, pesticides, and metals to water bodies. Non-point sources are some of the more serious forms of pollution and the effects are often less obvious than those of point sources (Abel, 1998). Clapman et al. (1998) indicated that agricultural impacts on ground and surface water quality are more significant than other land use impacts because of their large aerial extent compared to other human land uses. Lack of information related to agricultural practices may lead to overuse of fertilizers and herbicides, and deforestation with serious impacts on soil erosion and water quality. The Yucatan Peninsula is blessed with large freshwater reserves, bays and coastal lagoons and an exclusive economic zone of 200,000 km2. However, the maintenance of the water quality seems to be an enormous challenge for the State and other interested groups considering the unfulfilled basic needs of a large part of the population, rates of population growth, immigration, and lack of an integrated approach to the management of the resources and coastal areas. In June 1996 a massive kill of 20,000 fish of the species Arius felis was reported in the Bay of Chetumal, State of Quintana Roo (SEMARNAP, 1996). Studies reported harm to their organs and accumulation of PCBs, organochlorine insecticides and polyaromatic hydrocarbon (Noreña-Barroso, 1998). Concerns exist that increasing loss of water quality may have adverse effects on mangroves and coral reefs in an area where the second most important barrier reef, the Mesoamerican reef, and the sanctuary of the manatee Trichechus manatus are drawing international attention. © 2005 by CRC Press LLC Agricultural activities introduce diffuse pollution to watercourses, aquifers, lagoons and estuaries in the form of sediments, nutrients, pesticides, viruses, salt and other toxins which affect aquatic organisms. Pollution cause-and-effect relationships are complex and the need to find practical tools to generate useful information for decision-making may be addressed with models that introduce expert knowledge. Within this framework, this exercise attempts to rank the agricultural lands according to several factors that may contribute to water contamination on the Mexican side of the Rio Hondo watershed near the Othon P. Blanco municipality border with Belize. 17.2 GEOGRAPHICAL SETTING 17.2.1 The Yucatan Peninsula The Yucatan Peninsula in México has mainly sub-surficial water dynamics driven by its geological and topographic nature: there are few rivers or lakes in the area. Exceptions are the Candelaria and Champotón rivers in the zone of the Términos Lagoon in Campeche, and the Hondo River in the State of Quintana Roo near the border with Belize. Yucatán has no rivers; however underground discharge in the coastal zone has been estimated at 9.7 million m3 per year (CNA, 1998). The main features of the coastal areas are bays and lagoons, which are distributed throughout the three states: Términos Lagoon in Campeche; Celestún, Dzilam and Rio Lagartos in Yucatán; and Chetumal, Ascensión and Spiritu Santo bays in Quintana Roo. Surrounding these water bodies, other wetlands cover an area of approximately 8000 km², thus forming an important part of the coastal ecosystem (CNA-UNU/RIAMAS, 2000). Due to its geographical location between the Caribbean Sea and the Gulf of Mexico, the region is influenced by severe hydrometeorological phenomena. The climate of the region is semi-arid in the coastal zone of the north part of the Peninsula and warm with variation of dry to humid in the rest of the peninsula. The mean annual temperature is 26ºC. There are two main seasons in the regional climate: the “rainy season,” including extreme phenomena such as hurricanes and tropical storms from May to October; and the "winds of the north” season, from November to April. The region receives abundant but uneven rainfall with mean annual precipitation ranging from 1600 mm in the southeast to 500 mm in the north. Mean annual evaporation is around 1.78 mm (CNA-NU/RIAMAS, 2000). Human activities in the Yucatán Peninsula are related to agriculture, livestock production, tourism, fishing, oil production, and transportation, and recently to a large number of maquiladoras. Fertilizers, pesticides and metal residues have been found in coastal waters and the aquifer. These negative effects on the environment have been reported in several locations (Pacheco and Cabrera, 1996, Benitez and Bárcenas, 1996, Ortiz and Sáenz, 1997, Noreña-Barroso et al., 1998, CAN, 1998, Herrera-Silveira et al., 1998). Coastal resources in the Yucatan Peninsula provide increasing opportunities for economic development in a large number of traditional fishing communities. Lack of awareness of the deleterious effects of land-based activities on water quality may lead to a reduction in © 2005 by CRC Press LLC biodiversity and esthetical quality of the landscape, thus putting at risk the continuity of fishing, tourism, and recreational activities. 17.2.2 The Municipality of Othon P. Blanco Othon P. Blanco is a municipality located in the State of Quintana Roo in the southeast part of the Yucatan Peninsula (Figure 17.1). It has an area of 619,799 hectares, 161,226 (26%) of which are occupied by agricultural use, 129,396 (20.9%) by natural and cultivated grass, 325,155 (52.5%) by forest and 4,021 (0.6%) by other uses. Figure 17.1 Study area in the Yucatan peninsula, Mexico. Municipality of Othon P. Blanco in the state of Quintana Roo. Main hydrological features are the Hondo River and the Bay of Chetumal. The population in 1990 was 172,563 with 53.5% less than 19 years of age. The municipality has 437 rural and urban centers with 45% of the population living in 436 towns with less than 5,000 inhabitants. Close to 50% of the population are immigrants. Of the population 15 years or older, 17.4% completed elementary school (INEGI, 1991a). Main crops by cultivated area include: corn (19,196 hectares), sugar cane (16,000), chili (5,151), and beans (1, 391) (INEGI, 1993). Production in 1991 was estimated at 12,000 tonnes for corn, 785,000 for sugar © 2005 by CRC Press LLC cane, 26,000 for chili, and 84 for beans. Other crops cultivated in the area are orange, coconut, and banana (INEGI, 1994). Some of these agricultural lands can be found in the watershed of Chetumal Bay. The Bay is 67 km in length and 20 km wide, and receives freshwater from the Hondo River and Guerrero Lagoon, causing it to exhibit estuarine characteristics. Man-made channels and other tributaries close to agricultural lands are linked to the Hondo River. The most dominant soil types, Rendzinas and Litosoles, cover 70% of the Peninsula and 85% the study area (INEGI, 1985). 17.3 MODEL AND DATA LAYERS 17.3.1 Model The selected model attempts to rank agricultural lands according to the potential menace they represent to water quality. The model is a multi-criteria evaluation method provided by the IDRISI software, which combines several layers as the criteria to form an index of evaluation (Eastman, 1999). Relevant factors in the process of NPS pollution assessment (being those for which an estimate was available) were: amount of agrochemical inputs, slope, proximity to surface water, and distance to aquifer. Given the natural continuity in factors, ratio layers and a weighted linear combination of them can provide an index for ranking. Their mathematical representation is as follows: S = ( 6 wi* xi ) * 3 cj (17.1) where s= NPS index, wi= weighting factor i, xi= factor i, cj= constraint j Factors must be normalized according to the accuracy of our knowledge of their respective ranges of impact, as well as the ways in which they behave at different scales. The procedure used, which is based on fuzzy sets, is also provided by the IDRISI program. Finally, the model allows a weight to be assigned for the relative contribution of each factor. Criteria called Analytical Hierarchy Process (AHP), based on a pair-wise comparison of factors along a continuous rating scale, can derive weighs by calculating the principal eigenvector of the created matrix (Eastman, 1999). 17.3.2 Data layers 17.3.2.1 Geographic location of agricultural lands A good estimation of the location of agricultural land is a basic prerequisite for assessing NPS pollution. A two-band WIFS image with 180 x 180 m spatial resolution acquired in 1999 was used to estimate the location and extent of the © 2005 by CRC Press LLC agricultural lands in the Othon P. Blanco municipality (Figure 17.2a). A supervised classification was conducted to estimate the total cultivated land in the area. Census data from the AGROS system elaborated by the INEGI provided crop information for areas called Basic Geo-statistical Area (AGEB in Spanish) (INEGI, 1996). Each AGEB is a well defined polygonal (vector) area with a link to a data base (Figure 17.2b). These two layers were used to estimate the location of cultivated areas by AGEB. 17.3.2.2 Agrochemical practices The quantification of chemicals used per unit area and frequency of use was another factor in the analysis. A survey was conducted in September 1999 for the purpose of estimating the quantities of fertilizer and pesticides used per crop each year. A total of 97 farmers were interviewed in five agricultural towns: 1) Nicolás Bravo, 2) Palmar, 3) Pucté, 4) Sergio Butron, and 5) Morocoy. The questionnaire was divided into three sections to learn about fertilizer, insecticide, and herbicide practices per crop. Each section focuses on names, dosage, and frequency of applications of agrochemicals by crop per hectare per year. Substances were normalized according to thresholds determined through consultation with experts in the field. 17.3.2.3 Digital elevation model and slope Slope is a well known factor that favours the movement of the substances on the terrain surface. Slope and aspect are the main factors that determine velocity and direction of the overland flow during storms. In areas with long slopes, the capacity of vegetation to reduce erosion is diminished. Slope was computed from a DEM of the area using the TNT software (Figure 17.2c). The DEM was produced by INEGI in a scale of 1: 250,000. Two archives, E1604 and E1607, were mosaiced to cover the studied area. 17.3.2.4 Proximity to surface water Distance from agricultural activities to water bodies is also a factor that may facilitate contaminants reaching water courses, ponds, or estuaries. Proximity maps were constructed based on water features identified on band 2 of the WIFS image as well as those features found on topographic maps E16-4-7 at a scale of 1:250,000 produced by INEGI. These two layers were combined to generate a reference object for computing proximity to surface water (Figure 17.2d). 17.3.2.5 Proximity to groundwater In karst formations (carbonated rocks) such as the Yucatan Peninsula in Mexico, precipitation tends to infiltrate rapidly because of the high number of fractures and solution cavities in the massif. Depth to the aquifer was estimated from a regression model using 15 well depths provided by the Regional Office of the Comision Nacional del Agua (CNA) and their corresponding elevation in the DEM (Figure 17.2e). © 2005 by CRC Press LLC ... - tailieumienphi.vn