(Please note that tables and figures for this paper are not yet included in this Internet version and will be scanned in at a later date).
PRINCIPAL CONSIDERATIONS FACING BELIZE
Increasingly over the last decade, environmental quality has begun to play a major role in the economic well-being of Belize, formerly known as British Honduras. Tourism is now the single largest source of national income, largely because of the extensive sports diving industry associated with the offshore barrier reef system, and the ecotourism boom associated with Belize's largely undeveloped tropical wildlands, Mayan archaeology, raftable scenic rivers and extensive aquatic and terrestrial biodiversity.
Owing to its small population (which has only recently reached 200,000), its low population density, its colonial status (it gained independence from Britain only in 1981), its relative richness in natural resources, including water and lumber, and its lack of extensive urban and industrial development, environmental concerns and pressures have historically been low priorities on the domestic agenda. However, environmental monitoring has recently become a high priority due to changing national opinion and subsequent government initiatives, often supported by external funding agencies, to put more focus on conservation programs.
In the field of environmental management, water quality and other forms of environmental monitoring in the country have been characterized by fragmentation of responsibilities, duplication of services and facilities, and a lack of focus on issues of national concern in the water resources sector. A review of existing water quality databases and collection efforts in 1994 showed an understandable lack of integration since the early 1980s. Six principal sources of contamination assessment, each using different techniques and protocols, were identified:
Notwithstanding these efforts, it was acknowledged in the Belize report to the United Nations Conference on Environment and Development (GOB, 1992b) that the government still had insufficient baseline and trend data with which changes in water resource environmental quality could be quantified. In order to address this situation, an initiative was begun in 1993 by the USAID/Belize Government's Natural Resources Management and Protection (NARMAP) Project to develop a comprehensive national water quality monitoring program (GOB, 1995) and a set of enforceable industrial effluent quality standards (Miller and Miller, 1994). The NARMAP project seeks to control and prevent aquatic environmental degradation with a twin strategy of pollution control regulations and an environmental water quality monitoring program for the systematic assessment and control of land-based sources of contaminants through more effective monitoring and regulation. In this way, NARMAP hopes to safeguard the valuable and sensitive inland riverine, nearshore mangrove and offshore barrier reef ecosystems from contamination.
ESTABLISHING AN INTEGRATED AND EFFECTIVE WATER QUALITY MONITORING PROGRAM
In general terms, the establishment of an integrated and effective water quality monitoring program requires: knowledge of the hydrological system to be monitored; an understanding of the temporal and spatial variation of water quality indicators being monitored, and its causes; and clear objectives concerning how and why the data will subsequently be used, for example in regulatory enforcement and the design of corrective actions. One of the most useful steps in the development of any water quality monitoring program is the establishment of descriptive profiles of the river systems to be monitored and their contributing catchments. Such profiles integrate the available knowledge of natural and anthropogenic factors that do or could give rise to natural and induced water quality characteristics and hence may influence the choice of water quality variables to measure. The selection of water quality variables, sampling sites and sampling frequencies must reflect the sensitivities appropriate to each environment or the range of environmental conditions and the likely contaminants and their levels of occurrence. Because of the complexity of influences within natural aquatic environments that determine water quality and discharge characteristics, there are no universal standards that can be adopted for all water bodies as norms (Chapman, 1993). Selection of variables is a balance between the desire to adequately assess the range of potentially important environmental factors and the desire to keep the monitoring process logistically and financially manageable.
There are 16 major river catchments (Figure 1) in Belize which drain to the coastal waters of the Caribbean Sea from the Mayan Mountain Range, which forms the topographic and geological backbone of the country. Many of these rivers have a highly developed branching network of tributaries, several of which are significant rivers in their own right, specifically the Bladen and Swasey branches of the Monkey River, the Macal River tributary to the Belize River, and the Booths River and Rio Escondido (Mexico) tributaries to the Rio Hondo (GOB, 1995). Five of these catchments extend into or out of the neighboring countries of Mexico and Guatemala. No systematic database existed for these 16 major catchments and their characteristics prior to the NARMAP project. In order to provide a sound basis to prioritize sampling sites and critical water quality variables, and to focus national efforts, a preliminary survey was thus established for the major catchment systems in Belize using the following categories (modified from Chapman, 1993):
A summary of the complete profiles has been taken from GOB (1995) and presented in Table 1. For the construction of these profiles, the major activities or characteristics with potential impacts on water quality were identified as best as possible from diverse public and private sources including: land resources assessments (King et al, 1986, 1992, 1993); pollution studies (Archer, 1994. Newell, 1993. Haskoning, 1994); environmental reviews (Hartshorne et al, 1984); land use (LIC, 1994) and hydrological databases; selected field visits; aerial reconnaissance from the Project Lighthawk aircraft; and general literature sources (see GOB, 1995).
METHODOLOGY FOR COMPARING AND CONTRASTING BELIZE RIVER CATCHMENTS
As a guide to help direct and prioritize monitoring efforts to maximize limited resources, an attempt was made to compare and contrast river catchments based on an assessment of their land-based environmental risks as a way of understanding the expected dynamics of river contamination in Belize prior to any full-scale field monitoring study. A preliminary risk index, based principally on land-use statistics, was developed. From data provided by the Belize Land Information Centre's geographical information system database (LIC, 1994), a breakdown of land-use for the whole country and by major catchment was developed and plotted as 1:300,000 maps. Using 1989-92 survey data, around 90% of the land was classified as savannah rangeland, forest and other wooded areas (broadleaf forest, pine forest, thickets, shrubland, riparian zone, coastal forest, mangrove or swamp), or unproductive land (bare or water). Around 7% of this area was sub-classified as being secondary regrowth or regenerated land following prior farming or clearing. Less than 0.5% of the landscape was classified as having urban usage, although the majority of this is located next to the rivers. Depending on the extent of double counting of fallow areas as both wooded areas and cultivated zones, agriculture in the form of herbaceous annual crops, bananas, sugar-cane, tree crops, shifting cultivation (maize and bean Milpa), pasture and aquaculture was calculated to make up around 10% or 220,000 hectares of Belize's land surface.
To calculate the risk index, four categories were selected for each catchment based on available data: total agricultural activity, orchard land-use, urban land-use and industrial activity. Each catchment was assigned a rating from one to five based on the relative magnitude of these categories. While fairly subjective, the boundaries chosen for the rating provided a useful indication of the broad magnitude of relative risk. Based on this very preliminary rating system, the majority of rivers were, in relative terms, classed in the low and very-low risk categories. Using these scores based on broad attributes of Belize catchments, there was a considerable range of risk from very low scores of 4-6, which signify hardly modified catchments, to a high of 18, which signifies relatively extensive land-use and environmental pressures that already or could in the future lead to water quality impacts. The developed data indicated that the higher risk catchments were in the North of the country, with the lowest risks occurring in the South (Figure 1). The three most important catchments, all things considered, were the Belize River, New River and North Stann Creek. Given the relatively large size of the Rio Hondo catchment and that much of the land-use in the larger Mexican and Guatemalan portion of the basin is unknown, it was also considered a likely key source of contaminants, even though it had a moderate risk score. Corollary factors accentuating the significance of the environmental risk posed by discharge into the marine environment from the higher ranking rivers included the fact that:
- average water depths are low;
- there is a general north-south movement of nearshore coastal waters that brings contaminants from the west and north of the country down to join the contaminants flushed out from the central heartland by the Belize River;
- the coastal margin with the lowest assimilative capacity receives the largest volumes of runoff water and masses of contaminants; and
- the shallow depths of the northern nearshore zone encourage greater and more extensive penetration of a given volume of discharge out across the shallow coastal belt into the marine environment and towards the cayes.
LIMITATIONS OF THE INDEX
While a useful tool to focus attention on the question of spatial variations in the accumulation of contaminants in surface water systems, the risk index is clearly limited by a number of factors. One of the clearest limitations is the lack of data on the total land-use of the shared catchments with Mexico and Guatemala. While the impact on the ratings for the Belize and Hondo rivers, which already receive high risk ratings, are likely to be negligible, this may not be true in the cases of the Moho, Temash and Sarstoon rivers to the South, which can be expected to exhibit sizeable areas of land devoted to Milpa and non-mechanized annual crop production in their Guatemalan extension. Another weakness is that the index is based broadly on the concept of the mass of possible contaminants available (a function of area and land-use activity) rather than their concentration or transport processes within the catchment. The latter could not be considered because of the lack of adequate data on the discharge from all catchments. According to King et al (1993), the transport processes that could carry polluted river discharge out over the Belize barrier reef are not sufficiently understood to identify the critical catchments for nearshore water quality protection.
The risk index could be strengthened by several other indicators that were considered as risk modifiers during the study but which could not be included at that time due to a lack of developed data. These include: drainage density and average slope gradient, important in affecting whether non-point land-based contaminants would reach river channels and how fast; distance to the cayes and depth of water offshore, important when considering the fate of contaminants and potential for contaminant plumes to reach the important coral ecosystems; and total average annual rainfall input to the catchment, which in the absence of runoff data could broadly indicate discharge potential. Total and relative application of agrochemicals by catchment would also provide a noteworthy improvement to the index. Records from the Central Statistics office (Contreras, 1992) showed a considerable overall increase in pesticide imports, particularly insecticides, over the period 1982-1990. Ministry of Agriculture data show that in 1993, a total of 2.3 million pounds of these chemicals were imported into Belize. Fertilizer import figures varied considerably from 1989 to 1993, ranging between 17,519 tons in 1989 to 99,503 tons in 1992, with 20,526 tons imported in 1993. However, no developed data exists as to where, when and how much agrochemicals are used by major catchment, only at the national scale.
The risk assessment exercise carried out in the study was a somewhat subjective, but nevertheless useful tool for focusing attention on contaminants and locations of key concern in the absence of detailed data to analyze on contaminant loading and runoff production by specific areas. As indicated in Stednick et al (1995), the three major water quality parameters of importance to Belize, i.e. present in locally significant concentrations, are pesticides and agrochemicals, sediments and fecal coliforms. In the absence of primary water quality data, the index attempts to get to these parameters through reasoned use of secondary sources known to be important in the generation of these contaminants of ecological concern. As a complementary exercise to develop more specific information, a qualitative characterization of water quality or contaminant loading hot spots was made based on a review of known, strong potentials for point and non-point pollution (GOB, 1995).
PRINCIPAL RECOMMENDATIONS FOR THE IMPLEMENTATION OF ENVIRONMENTAL WATER QUALITY MONITORING IN BELIZE
Following an extensive evaluation of the environmental, institutional, and technical context within Belize, a number of principal recommendations for the monitoring aspects of a national program to control and prevent aquatic environmental degradation were made. Firstly, an inventory of monitoring sites were selected. A major objective of the proposed water quality monitoring program is to create baseline water quality conditions for each major catchment within Belize, allowing statistical comparisons with later water data to assess trends and influence land use management decisions as well as providing information on the general state of the river ecosystems at the coastal margin and inland. While there is no "correct" combination of sampling sites or form of network, only different levels of investment and density of coverage, a guiding principal for site selection was to concentrate on quantifying the overall volume of flow and the nature and mass of contaminants leaving the land and entering the marine environment. Thus sites were prioritized to produce a rational and workable inventory that takes into account the interests of multiple agencies for water quantity and quality data and attempts to maximize the use of the existing framework of hydrological monitoring stations. The proposed inventory establishes or consolidates sites to determine volumetric runoff and quality constituents (types present and mass output) at the most feasible and closest locations inland from the coastal margins. It also establishes a series of upstream water quality sampling locations for those catchments with extensive developments and land-use conversions that can provide an assessment of the magnitude of water quality changes taking place within the river system as a result of anthropogenic inputs. It consolidates existing or planned additional water quantity monitoring sites operated by the Hydrology Service to add paired water quality samples and to document the nature of major sub-catchment responses throughout Belize's hydrological system. And finally, it establishes receiving water assimilative capabilities in the immediate locations of the two major point pollution zones, that of Tower Hill to Libertad on the New River, and Alta Vista to Pomona on the North Stann Creek (GOB, 1995). Since it would take some time to create a fully functioning monitoring system, emphasis was placed in the short-term on quantifying the overall characteristics of the Belize river systems by sampling at lower gauging stations closest to the coastline and downstream of river border crossings within shared catchments with Guatemala and Mexico, designed to meet the resource management decision-making needs of at least three agencies concerned with water quality and quantity at the land-sea interface: the Department of Environment, the National Hydrological Service, and the Coastal Zone Management Unit. More detailed analysis of water quality from different elements of the catchments can be added at a later date based on the mobilization of greater resources.
Secondly, emphasis was placed on starting out relatively slowly, stressing the need to obtain representative samples that are properly collected, adequately documented in the field at the time of collection, with the exercising of proper care and custody practices. It was thus proposed that two specially trained teams made up of environmental and hydrological technicians from the Department of Environment and the National Hydrological Service begin a sampling program based on a regular bimonthly cycle, with an attempt to take up to two more samples each year at the extremes of high and low flows to more accurately identify variability in water quality concentrations. One team would sample at all sites south of Belmopan (Southern Highway Team) and the other would sample all sites along the Western Highway and Northern Highway, allowing the respective technicians to develop a good feel for the individual rivers they are monitoring. With regard to laboratory options, Belize had a relatively broad range of analytical capability at the time of the study, although much of it was sub-utilized and some of it required improved maintenance and care. The majority of the analytical equipment needed to carry out even the most detailed of laboratory analysis - such as atomic absorption spectrophotometry and gas chromatography - were available, with the most straightforward option being to consolidate it all into a national accredited water quality or pollution control laboratory. However, it was recognized that since this cuts across budgets and responsibilities, rationalizing laboratory facilities would require political consensus to reallocate these physical and human resources. Since political decisions normally work best when those involved have a range of alternatives to consider, several options were presented, all of which could reasonably meet national needs for water quality data including: establishing a wholly new accredited national laboratory; expanding and/or reconditioning an existing laboratory or laboratories; maximizing in-field analysis (a kit-based approach, e.g. Hach digital titration, etc.); and/or sending samples abroad for selected laboratory tests. Based on an assessment of the environmental profiles and hot spots (GOB, 1995), and through a review process carried out by a peer group of water quality and water sector professionals in Belize, a recommended list of constituents of concern were determined for use by the environmental water quality monitoring program. They represent those variables most likely to characterize and indicate overall water quality conditions and patterns of degradation throughout the major catchments and water bodies of the country. Moreover, they represent factors critical to environmental and resource management decisions such as the exploitation of a given water body for drinking or agricultural use, or the protection of receiving waters. For each constituent, a standard field sampling and analytical procedure was selected and documented in the form of a comprehensive protocol manual prepared for the Belize Department of Environment (Stednick and Gilbert, 1995). The constituents of concern are: volumetric discharge; temperature; pH; conductivity; dissolved oxygen; alkalinity; turbidity; total suspended solids; total dissolved solids; pesticides; herbicides; biochemical oxygen demand; chemical oxygen demand; nitrate nitrogen; total Kjeldahl nitrogen; soluble reactive phosphorous (orthophosphorous); chlorides; total organic carbon; total petroleum hydrocarbons; sulfate; metals; and coliform bacteria.
Given the size of the population of Belize, its level of urbanization, its degree of industrialization, the review of known land-based pollution sources and the apparent largely rural nature of the shared sections of catchments with Mexico and Guatemala, it was considered unlikely, at an overall national environmental level, that metals and pesticides would be found in significant proportions. However, since water monitoring must keep an eye to the future, the growing use and concentration of agro-chemicals in certain zones (particularly between the North Stann Creek and Monkey River), the value of the unique offshore marine environment to Belize, and the growing fear in Belize that even small concentrations of chemicals may cause damaging effects on the offshore ecosystems, it was decided that this assumption should be tested in the establishment of baseline conditions during the early phases of any monitoring plan. Decisions could be made later whether to continue with the full portfolio of contaminants of concern for all monitored catchments, depending on the observed results.
REFERENCES
Archer, A.B., 1994. Land-Based Sources of Marine Pollution Inventories: Belize & Cayman Islands. UNEP Regional Coordinating Unit Report, May.
Chapman, D. (ed), 1993. Water quality assessments: A guide to the use of biota, sediments and water in environmental monitoring. Chapman & Hall, London.
Contreras, M., 1992. Report on Pesticides Control Board. Zamorano, Honduras. RENARM Report, April.
Department of Environment, 1994. Mission Statement. Belmopan, Belize.
Government of Belize, 1992a. Environmental Protection Act. October.
Government of Belize, 1992b. Belize: The National Report. United Nations Conference on Environment and Development, Rio de Janeiro, Brazil.
Government of Belize, 1995. Environmental Water Quality Monitoring Program. Final Report to USAID-NARMAP by Lee M.D., Stednick J.S. and Gilbert, D.M., June 1995. 310 pages plus maps.
Hartshorne G. et al, 1984. Belize: Country Environmental Profile. A Field Study.
Haskoning, 1994. Citrus Waste Stabilization. Report to Government of Belize/UNIDO, November.
King R.B., Baillie I.C., Bissett P.G., Grimble R.J., Johnson M.S. and Silva G.L., 1986. Land Resource Survey of Toledo District, Belize. Land Resources Development Centre, p177.
King R.B., Baillie I.C., Abell T.M.B., Dunsmore J.R., Gray D.A., Pratt J.H., Versey H.R., Wright A.C.S. and Zisman, S.A., 1992. Land Resources Assessment of Northern Belize, Volumes 1 & 2. Natural Resources Institute Bulletin 43.
King, R.B., Pratt J.H., Warner M.P. and Zisman, S.A., 1993. Agricultural Development Prospects in Belize. Natural Resource Institute Bulletin No.48.
LIC, 1994. The Land Use of Belize 1989/92. Preliminary Report. Land Information Centre, May.
Miller and Miller, Ltd., 1994. Recommendations on Industrial Effluent Quality Standards. Draft Final Report to NARMAP.
Newell, P.J., 1993. Industrial Pollution Management in the Sugar Cane, Rum Distilleries and Other Industries in Belize. UNIDO Technical Report SI/bze/92/801.
Stednick, J.D. and Gilbert D.M., 1995. Environmental Water Quality Monitoring Program Water Quality Monitoring Protocol. Report to Government of Belize, Department of Environment.
Stednick J.D., Gilbert D.M. and Lee M.D., 1995. "Development of a National Water-Quality Monitoring Program for Belize". American Institute of Hydrology. Water Resources at Risk, May 14-18, Denver. Pages RA16-22.