xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002029664
007 cr mnu|||uuuuu
008 090918s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002964
Alegria, Victor Eduardo.
Land-based sources of pollutants to coastal waters of southern Belize :
b comparison of predictive model with empirical data
h [electronic resource] /
by Victor Eduardo Alegria.
[Tampa, Fla] :
University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 100 pages.
Thesis (M.S.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: A Level III fugacity-type model was applied to southern Belize (Stann Creek and Toledo Districts) to predict which of the pesticides most heavily used in banana and citrus farms are most likely to end up in streams and coastal waters via surface runoff. Concentrations of all target pesticides in coastal waters of southern Belize were then measured during two sampling campaigns (dry season and rainy season) in 2008. Several pesticides were measured in significant levels, including some as far out as waters overlying coral reefs. The presence of these pesticides in the coastal waters indicates that agricultural activities in southern Belize may have a potential impact on coral reefs offshore. Results of the predictive model were compared with the empirical data to determine how well the model works in a tropical region such as southern Belize. Overall, there is considerable agreement between the two, indicating that the model employed herein can be applied to other tropical areas. Concentrations of mercury and lead were also measured in the same study area. Mercury levels were uniform and low, suggesting natural sources. Lead levels varied and are most likely explained by the presence of unregulated and illegal waste dumps near streams in the region. An analysis was carried out to examine government policy on pesticide use. Findings indicate a lack of a coherent governmental policy on the sale and use of pesticides, most likely because of too many agencies/ministries being involved and the lack of national standards for these pesticides in the environment.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Joseph Dorsey, Ph.D.
x Environmental Science, Policy and Geography
t USF Electronic Theses and Dissertations.
Land Â–Based Sources of Pollutants to Co astal Waters of Southern Belize Â– Comparison of Predictive Model with Empirical Data by Victor Eduardo Alegria A thesis submitted in partial fulfillment of the requirements for the degree of MasterÂ’s in Science Department of Environmental Science, Policy and Geography College of Arts and Sciences University of South Florida Major Professor: Joseph Dorsey, Ph.D. Kathleen Carvalho-Knighton, Ph.D. Ambe Njoh, Ph.D. Date of Approval: April 3, 2009 Keywords: pesticides, metals, co ast, coral reef monitoring Copyright 2009, Victor Eduardo Alegria
Acknowledgements Many professors and individuals cont ributed valuable information, advice; comments and support which helped me accomplis h this degree. Recognition is due to all. Special thanks to my committee advi sors: Dr. Ambe Njoh, Dr. Kathleen Carvalho-Knighton and my Major Professor Dr Joseph Dorsey. You were right all along, the relationship between a graduate student and his/he r committee is really based on TLC (Trust, Loyalty and Care). Thanks for the trust, patience and commitment for the past two years of mentoring. To my wife Bertha and daughters Alisha and Britney, thank you for being there for me through my three and a half years of studies. I am thankful for everything you guys have done. I never woul d have finished this w ithout your moral support and patience. This is it, no more aw ay from you guys, thatÂ’s a promise. Finally, to my brothers Dr. Henry and Mar tin Alegria, thanks for really putting up with my constant harassment for the past three and a half years. Without you guys I wouldnÂ’t be where I am and be who I am. Than ks for keeping me in line with invaluable advice and encouragement that has made me mature during the last years.
i Table of Contents List of Tables ii List of Figures iv Abstract vi Chapter 1 1 Background 1 Research Project 4 Research Setting 5 Research Methodology 6 Significance of the Project 8 Conclusion 8 Chapter 2 10 Background 10 The Belize Situation 11 Research Project 12 Environmental Fate Modeling of Pesticides in Belize 12 Defining Equilibrium 12 The Concept of Fugacity 15 Degradation Processes 21 Biodegradation 21 Hydrolysis 23 Oxidation 24 Photolysis 25 Using Partitioning Data to Identify Key Half Lives 26 Application of Level 1 and 11 Modeling to Pesticides commonly used 28 in Belize Conclusion 55 Chapter 3 59 Introduction 59 Research Projects and Objectives 60 Research Area 61 Materials and Methods 63 Sampling Sites 63 Cleanup of Sampling Equipment and Material 65 Sampling 66 Processing 66
ii Extraction 68 Analysis 68 Paraquat and Glyphosate 69 Metals 72 Quality Control 72 Results 73 Metals 73 Glyphosate and Paraquat 74 Other Pesticides 75 Comparison of Empirical Data and Modeling Results 80 Conclusion 81 Chapter 4 83 Background 83 BelizeÂ’s Legislative Framework on Pesticides 84 The Pesticides Control Act (PCA) 84 The Environmental Protection Act (EPA) 87 Government Policy on and Mana gement of Pesticides 89 Rationalizing Legislative Amendments to Address Chemicals in Belize 91 BelizeÂ’s Institutional Framework on Pesticides 93 Pesticides Control Board 93 Conclusion and Recommendations 95 References 98
iii List of Tables Table 1: Physical-chemical pr operty data for selected CUPs used in Belize. 29 Data are taken from various sources, in cluding Mackay et al. (2006) and e stimated using the EPIWIN software package (Meylan, 1999). Table 2: Estimated reactivity data based on Mackay et al. (2006) and the 33 EPIWIN software package, mass fractions in air, water, and octanol and overall environmen tal half-life. Table 3: Summary of results from Level III calculations regarding fractions 54 in air, water, and soil, as well as Leve l III overall half-life. Note: assumed that emissions are 80% to soil, 10% to water, and 10% to air. Table 4. Coordinates of Sampling Sites. 64 Table 5. Metal concentrations in coas tal waters of southern Belize. 74 Table 6. Pesticide levels in coas tal southern Belize (pg). 78 Table 7. Pesticide levels in coas tal southern Belize (pg). 79 Table 8. Comparison of empirical and modeling results. 80
iv List of Figures Figure 1: Environmental partitioning of CUPs listed in Table 1 in an 31 environmental system that is at steady-state and equilibrium, consistent with the model environment desc ribed by Gouin et al. (2000). Figure 2: Environmental parameters us ed in Level III calculations. 34 Figure 3: Level III results for mancozeb 35 Figure 4: Level III results for Ethoprop 36 Figure 5: Level III results fo r chlorothalonil 37 Figure 6: Level III results for carbofuran 38 Figure 7: Level III results for glyphosate 39 Figure 8: Level III results for terbufos 40 Figure 9: Level III results for paraquat. 41 Figure 10: Level III results for cadusafos. 42 Figure 11: Level III results for oxamyl. 43 Figure 12: Level III results for fenamiphos. 44 Figure 13: Level III results acetochlor. 45 Figure 14: Level III results for bitertanol. 46 Figure 15: Level III results for dfenoconazole. 47 Figure 16: Level III results for aldicarb. 48 Figure 17: Level III results for chlorpyrifos. 49 Figure 18: Level III results for malathion. 50 Figure 19: Level III results for dacthal. 51
v Figure 20: Level III results for diazinon. 52 Figure 21: Level III results for azinphosmethyl. 53 Figure 22. Sampling region and sampling stations 62
vi Land-bases Sources of Pollutants to C oastal Waters of Southern Belize Â– Comparison of Predictive Modes with Empirical Data Victor Eduardo Alegria ABSTRACT A Level III fugacity-type m odel was applied to southern Belize (Stann Creek and Toledo Districts) to predict which of the pe sticides most heavily used in banana and citrus farms are most likely to end up in st reams and coastal waters via surface runoff. Concentrations of all target pesticides in co astal waters of southern Belize were then measured during two sampling campaigns (dry season and rainy season) in 2008. Several pesticides were measured in significant le vels, including some as far out as waters overlying coral reefs. The presence of these pes ticides in the coastal waters indicates that agricultural activities in southern Belize may have a potential impact on coral reefs offshore. Results of the predictive model we re compared with the empirical data to determine how well the model works in a tropical region such as southern Belize. Overall, there is considerable agreement between the two, indicating that the model employed herein can be applied to other trop ical areas. Concentra tions of mercury and lead were also measured in the same study ar ea. Mercury levels we re uniform and low, suggesting natural sources. Lead levels va ried and are most likely explained by the presence of unregulated and illegal waste dumps near streams in the region. An analysis was carried out to examine government policy on pesticide use. Findings indicate a lack of a coherent governmental policy on the sale and use of pesticides, most likely because
vii of too many agencies/ministries being involve d and the lack of national standards for these pesticides in the environment.
1 Chapter One Introduction Background Pesticides are used in huge quantities globally in order to ensure sufficient agricultural productivity. As a result, the issue of environmental contamination from these chemicals is still an active area of research. Once applied, pesticides may be quickly degraded microbiologically, physically, or chemically. In most cases, however, pesticides are persistent enough that they are subject to m ovement away from application sites. Pesticides have been shown to be transported away from application sites by volatilization into the atmo sphere, leaching and movement through soil and into groundwater, and surface runoff to streams and coastal waters. The predominant transport mechanism is determined by a variety of factors, including the chemicalÂ’s physical-chemical properties, the quantitie s employed in the field, the method of application, prevailing weather conditions, and the type of so il in which pesticides are employed. Consequently, determination of fa te and transport of pe sticides is a very complex endeavor. As a result, most studi es in this field concentrate on potential movement of pesticides in one media or via one mechanism at a time. That is, studies focus on soil-air exchange and atmospheric transport, or percol ation through soil into underground aquifers or laterally to surface waters, or surface runoff into surface waters, etc. Fewer studies attempt to look at movement of pesticides in all media simultaneously.
2 Regardless of approach, it has been shown that significant quantities of pesticides can be removed from sites of application via all mechanisms into all media. For example, volatilization of pesticides into th e atmosphere after application (as opposed to spray drift during application) can remove large fractions of pesticides from agricultural fields (Qiu et al., 2004). One study documen ted loss of over 50 percent of DDT applied to a field within one month. Sieber et al (1996) have documented significant pesticide loss via spray drift during aeria l application. Other studies have documented significant loss of pesticides from agricultural fields vi a surface runoff, especially with rainfall. Studies have calculated pesticide loss from runoff ranging from <0.01 to >10 percent of the pesticide applied (Southwick et al., 2003; Senseman et al., 1997). Although these fractions may seem small, if very large qua ntities are actually a pplied the amounts lost via surface runoff can be significant. Wh en rainfall occurs soon after pesticide application, losses may be more significant (Senseman et al., 1997). Ultimately, the concentrations of pesticides from runoff may reach concentrations in surface waters that are detrimental to flora and fauna that inha bit those waters. In a similar manner, many studies have documented the leaching of pes ticides from soils and contamination of groundwater (Geisler et al., 2004 and ref 1-4 therein). In areas with farms near coastal areas, surface runoff of pesticides into coastal waters is a major concern (Hapeman et al., 200 2; Scott et al., 2002; Al egria et al., 1999). Coastal waters constitute a very sensitive ecosystem, with important habitats that can be adversely affected by pesticides. Loss of im portant species in coastal waters has both ecological as well as economic repercussions since fisheries as well as tourism and recreation are adversely affected.
3 In general, there are two strategies to determine the transport and fate of pesticides from agricultural fi elds to the environment. On e is through fiel d studies and direct measurements of pesticides in the vari ous compartments available (that is, soil, air, water). The second is via the use of models that may predict the transport and fate of pesticides into these various compartments. Direct measurement of pesticides in the environment is costand time-intensive. Ther efore, this strategy is often not feasible. This is especially true in developing countries, which usually lack the finance, equipment and trained personnel to carry out such studies. Even when some direct measurements are feasible, models are useful to screen pesticides so as to re fine the potential target list. This approach will enable limited resources to be focused on the most likely transport mechanism of pesticides as well as on the pesticides most likely to present environmental problems. Many different types of models have b een developed to assess movement of pesticides from application sites. Some fo cus on particular transport pathways while some attempt to provide an ove rall picture of the movement of pesticides into all possible compartments. Models such as the Pest icide Root Zone Model (PRZM), Groundwater Loading Effects of Agricultural Manage ment Systems (GLEAMS), and Chemical, Runoff, and Erosion from Agricultural Ma nagement Systems (CREAMS) have been extensively used to simulate pesticide transp ort at field scale (Ram anarayan et al., 2005; Carsel et al., 1985; Leonar d et al., 1987; Knisel, 1980). Other models have been designed to study transport of pesticides from soils into selected compartments. For example, some models have been developed to simulate the vola tilization of volatile pesticides from soil into air (Reichman et al ., 2000; Jury et al., 1990; Chen et al., 1995;
4 Baker et al., 1995; Wang et al., 1998; Woodrow et al., 1997). Other models simulate pesticide loss via runoff into surface waters (Guo et al., 2004; Ve rro et al., 2002). Models differ in their complexity and th ey differ on how well they simulate what really happens in the field. The simpler the model, the less input pa rameters needed to run it. However, simpler models are less like ly to simulate field conditions accurately. There is still a great need for developing models to predict the fate of pesticides in the environment. Recently, the Canadian Envir onmental Modeling Center (CEMC) released a new model called Level III which attempts to predict the fate of pesticides in the environment. Research Project Detailed herein are the results of a resear ch project we have carried out that had as its main objectives: (1) application of th e CEMC Level III Model to predict the environmental fate of pesticides used in citrus, bananas and aquaculture farms in southern Belize, with a special focus on predicting whic h pesticides are more likely to end up in coastal waters via surface runoff; (2) meas uring pesticide and selected heavy metal concentrations in coastal waters of sout hern Belize to determine the extent of contamination from agriculture and garbage dumps; (3) comparing the empirical results with those predicted by the CEMC Level III m odel in order to validate the model; and (4) analysis of pesticide legisl ation in Belize. The project also had as an important component the generation of recommendations on pesticide usage to farmers as deemed appropriate by the results of the empirical data.
5 Research Setting The southern part of Belize, specifically the southern Stann Creek District, is home to the citrus and banana industries, as we ll as most of the major aquaculture farms. Most citrus and banana farms are located in coastal areas, and for obvious reasons all aquaculture farms are in coastal waters. All three industries ar e known to be heavily dependent on the use of pesticides (I. Fabro, per. comm.). The Toledo District is also home to some citrus farms. The entire coast of Belize is lined with the worldÂ’s second-longest system of coral reefs (generally referred to as the Barrier Reef). This barrie r reef is located close to the coast, anywhere from ~5km to ~20 km offshor e. There is therefore great concern that pesticides from agriculture/aquaculture ma y be transported to coastal waters and adversely affect the coral reefs, especially since coastal waters a nd the coral reefs are important sources of income for the country (from fisheries and tourism). There are several rivers that flow through banana and citrus farms and empty into the southern coast of Belize. There are also some rivers that flow through areas with less intensive agricultural practices prevalen t (subsistence farming mostl y, although there is one area with some significant area under rice cultivati on). These rivers can potentially transport pesticides into coastal waters in the region. Previous studi es have documented the input of pesticides from coastal agricultural fields via rivers/streams to coastal waters in other regions of the world (Alegria et al., 2000a,b; Hapeman et al., 2002; Scott et al., 2002). In recognition of these and other similar problem s, the United Nations, through its United Nations Environmental Programme (UNEP), has identified land-based sources of pollutants as a research focus in th e wider Caribbean, including Belize.
6 Research Methodology Chapters 2 Â– 4 will discuss in more detail the exact approaches used to carry out the research project. Briefly, we documented the types and amounts of pesticides used in citrus and banana farms and application prot ocols/strategies used. This information allowed us to rank pesticides in terms of quantities used in southern Belize. From the list of pesticides used in citr us, banana and aquaculture we selected those for which 250 kg or more are used annually in southern Belize as target pesticides. We applied the CEMC (Canadian Environmental Modeling Centre) Level III model to this list of pesticides. This model require s input parameters on pesticide name, pesticide properties, application rates, types of soil, amount of rainfa ll, and soil type. Chapter 2 will detail the results of applying this pr edictive model to the study region. We were especially concerned with identif ying those pesticides predicte d to be most susceptible to surface runoff since these would be the ones most likely to be found in coastal waters. To a lesser extent, those that ar e most likely to volatilize into air have the potential to also impact coastal waters. Previous studies have shown that pesticides that volatilize into air may be deposited into coastal waters by dry or wet deposition (Ale gria et al., 1999). During 2008 we carried out two campaigns to measure the co ncentrations of pesticides in coastal waters of southern Be lize (Stann Creek and Toledo Districts). This allowed us to compare the results of the pr edictive model with actual empirical data. Chapter 3 will detail the results of this part of the project. Briefly, we selected 8 rivers in the study region. Four were selected becau se they flow through citrus and banana farmlands, two were selected because they fl ow through areas that are used for other lessintensive agricultural activities (with the po ssible exception of some rice farms), one was
7 selected as a reference site since it flows through protecte d areas, and one was selected because it forms the boundary between Belize an d Guatemala and we were interested in how similar or different its si gnature would be compared to the other rivers (Figure 1). We collected water samples from transects laid out from the river mouths in all eight cases and moving out offshore until we reac hed waters overlying coral reefs. In a couple of cases we could not sample in a transe ct parallel to the co ast or all the way out to where the major coral reefs are located because of maritime borders between Belize, Guatemala and Honduras or because the reefs ar e located too far offshore (see Figure 1). Water samples were processed on-shore and tr ansported to our laboratories at USFSP for analysis. Most pesticides were measured using gas chromatography Â– mass spectrometry (GC-MS), but a couple had to be m easured using high-performance liquid chromatography (HPLC). The results of the empirical study have been compared to those of the predictive model, as detailed in chapter 3. Chapter 3 al so discusses levels of mercury and lead in coastal waters of southern Belize. While carrying out our first sampling campaign in 2008, our local partners in Beliz e mentioned concerns about th e presence of heavy metals in coastal waters, especially mercury and l ead. As a result, du ring the second sampling campaign we incorporated sampling for hea vy metals. Mercury in these samples was measured using graphite furnace Â– atomic absorption spectrophotometry (GFAAS) while lead was measured using inductively-coupl ed plasma mass spectrometry (ICPMS). Results will be discussed in detail in chap ter 3, including potentia l sources of these metals.
8 Chapter 4 will discuss the results of an analysis of pesticide policy in Belize. This chapter will also include an analysis of potential flaws in the way the government of Belize oversees the sale, use and appli cation of pesticides in the country. Significance of the Project There are several reasons why the results of this research project are important. First, we have documented for the first time th e extent to which pesticides and metals are present in coastal waters of Belize. This in formation is important in the countryÂ’s efforts to protect sensitive coastal ecosystems, especi ally coral reefs, from degradation due to agriculture and unauthorized wa ste dumps. Second, we have applied the relatively new Level III Model to a tropical area for validation for the first time, to our knowledge. This will add significantly to the body of knowledge on the use of such models. Finally, we will generate the first analysis of polic ies governing chemicals, and specifically pesticides, in Belize. This analysis will be very useful to help the relevant authorities develop improved chemical/pesti cide management strategies. Conclusions The results detailed in this thesis show conclusively that I have achieved the objectives set out for my M.S. degree in Envi ronmental Science and Policy. The results contribute to increasing the body of knowledge re garding land-bases sources of pollutants to coastal waters in tropical regions in gene ral, and southern Belize in particular. The validation of the Level III model is also im portant and also represents a significant contribution to science. Finally, the resu lts have been used to generate useful
9 recommendations to farmers in the citrus a nd banana industries in Belize, which should benefit all concerned stakeholders.
10 Chapter Two Application of Level III Fugacit y Model to Southern Belize Background Understanding the fate of pesticides is extremely important because they are designed to have a biological e ffect (i.e. they are toxic to at least some organisms) and because they are released into the environm ent deliberately and in significantly large quantities (Muir et al., 2004). Th ere is particular interest in understanding the impact of pesticides on non-target organisms and non-agricultural ecosystems. In general, there are two approaches to studying the fate and transport of pesticides in the environment. One is to carry out empirical studies to trace the movement and eventual fate of pesticides onc e they are applied. These studies must be long-term and in-depth. As such, they re quire funds, trained pe rsonnel and analytical laboratory capabilities to accomplish. Th e second approach is to use a modeling approach. This approach involves essentia lly computer simulations that mimic what occurs in the environment. As such, the ex tent to which they mimic reality is governed by input parameters. Some models are simple with relatively simple input parameters. These are easier to use a nd understand, but their realism is limited. More complex models require more and more complex input parameters. They are more difficult to manipulate and understand but their re sults more closely match reality. The agrochemical industry is a multi -billion one and new pesticides are introduced on a yearly basis. As a result, it is very difficult to study the fate and transport
11 of every pesticide in existence. This is where modeling becomes an extremely valuable tool. Pesticides may be divided by properties su ch as structure, mode of action, etc. If the fate and transport of members of a given fa mily of pesticides is known, then it may be possible to predict the fate and transport of other me mbers of that family of pesticides via modeling. The process may actually be applie d in the reverse whereby the results of modeling may be used to prioritize which pest icides should be empiri cally studied in the environment. One approach would be to a pply sophisticated models to pesticides and then develop a tiered system where prioriti es are based on the lik elihood of a pesticide having a significant impact on the environment, as indicated by the modeling results. Another approach is to determine via m odeling which of the various environmental media may be most impacted by pe sticides; that is, is a given pesticide more likely to be volatilized into air and be atmospherically tr ansported and deposited, or is it more likely to be transported via surface runoff into n earby streams, etc? Thus, depending on which ecosystem is prioritized in terms of protec tion, only the relevant pesticides may be empirically studied. The Belizean Situation Belize, as is the case in most developi ng countries, does not po ssess the analytical capabilities to carry out research on the fa te and transport of pesticides in the environment. In addition, ther e is a lack of funding for such research. As a result, there has never been, to our knowledge, any compre hensive study on the fate and transport of pesticides in the environment, only a few limite d studies on specific media (Alegria et al.,
12 2000; Wu et al., 2006). There is concern, however, that some ecosystems may be adversely affected by the presence of pesticides in them. In Belize, the banana and citrus industri es are concentrated in the southern Stann District and to a lesser extent the Toledo District (Figure 1). These industries are heavilydependent on pesticide usage. Lying offs hore in Belize is the worldÂ’s second longest barrier reef. This system of coral reefs lies an average of approximately 15 km offshore. The health of the barrier reef has been in decline in recent years and there exists the possibility that agrochemicals may be partly responsible for this. Research Project We have carried out modeling studies in which we have applied a fugacity model developed by the Canadian Environmental M odeling Centre (CEMC) to southern Belize with the objective of determining its effectiv eness at predicting which pesticides are most likely to end up in surface waters in the region. Environmental Fate Modeling of Pesticides in Belize Defining Equilibrium The environmental fate of a chemical, as determined by the Mackay Level I and Level II models, describes the behaviour of a su bstance in a steady-state system that is at equilibrium. This implies that conditions w ithin the system do not change with time, and that inputs and outputs are equal (Mackay et al., 1996). The environment is a closed
13 system that is composed of different envir onmental media, such as air, water, soil, sediment, aerosols, fish, etc., which have fixed volumes (Mackay, 2001). Thus the partitioning behaviour of a chemical that is in troduced to such an environment will have a mass balance expression in which the total amount of chemical present, M will equal the sum of amounts in each comp artment, as described by: M = CiVi (1) where C is the concentration in units of mol m-3, V is the volume in units of m3, and the subscript i is the environmental compartment in question (Mackay, 2001). Assuming that the volumes in a closed system are consta nt, and are not affected by the input of a chemical, it becomes necessary to determine the concentration of a substance in the various compartments in order to calculate a chemical mass balance within that system. This can be accomplished by considering the concentration ratio at equilibrium, as described by NernstÂ’s Distribution Law, K12 = C1/ C2 (2) where K12 is a constant referred to as the part ition coefficient. Thus, by knowing the concentration of a substance in one comp artment, it is possible to calculate the concentration of a substance in another co mpartment by using the appropriate partition coefficient (Mackay, 2001). Generally there are two partition coefficients that can be determined with relative ease, experiment ally, for most organic compounds. These partition coefficients are Kow and Kaw, where, Kow = Co/ Cw (3) and Kaw = Ca/ Cw (4)
14 where the subscripts a o and w refer to air, octanol and water respectively. The partition coefficient described by equation 3 is commonly referred to as the octanol-water partition coefficient. It is perhaps one of the most frequently used descriptors of chemical behavi our in the environment, and is a measure of a chemicalÂ’s hydrophobicity, or degree to which a chemical partitions out of water (Mackay, 2001). As the value for Kow increases the tendency for the chemical to partition out of water also increases. Meanwhile, the partition coefficient desc ribed by equation 4 is known as the airwater partition coefficient and is essentially the ratio of vapour pressure to water solubility, which is also referred to as the HenryÂ’s law constant, KH. KH = Ps/ RTCs w (5) where the superscript s denotes saturation, Ps is the vapour pressure (Pa), and Cs w is the solubility in water (mol m-3) (Mackay, 2001). Values for the constants given in equation 3 and 5 for a number of organic chemicals can be obtained from a variety of sources, including the handbooks of Mackay et al. (2006), Howard (1990), and Lyman (1990), or they may be estimated as described by Boet hling and Mackay (1999) or through the use of the EPIWIN suite of programs (Meylan, 1999). Although the partition coefficients defined in equations 3 and 5 have been studied for a large number of substances, there are ot her partition coefficients that need to be determined if one wishes to calculate the mass balance of a substance in an environmental system. Given the number of environmental compartments that might exist in a particular system, it could prove to be difficult, if not impossible, to define partition coefficients between all of the pairs of media that might exist. For instance a
15 system with 6 compartments will require 30 different partition coefficients to determine the concentrations in each compartment at equilibrium. However, it is possible to calculate the concentration of an organic chemical in various environmental media by utilizing another approach that has been de scribed extensively by Mackay (2001). This approach uses the concept of fugacity as the criterion for equilibrium. The Concept of Fugacity In a system that is in thermodynamic equilibrium, the laws of thermodynamics must be obeyed. For a closed system of constant composition there exists the thermodynamic properties of temperature T internal energy, U and entropy, S as introduced by the zeroth, first, and sec ond laws, respectively (Mackay, 2001). In addition, there is the property of Gibbs free energy, G which is defined in terms of the thermodynamic properties listed above. For a re versible process in a closed system of constant composition that can only perform pressure-volum e work, the first and second laws of thermodynamics may be combined to yield dU = T dS Â– P dV (6) This equation is often referred to as the fundamental equation for a closed system, and may be expressed in terms of Gibbs free energy as, G = U + PV Â– TS (7) The differential to this equation is, dG = dU + P dV + V dP Â– T dS Â– S dT (8) Substitution of equation 6 into equations 8 leads to dG = -S dT + V dP (9)
16 Equation 9 shows that a change in Gibbs free energy is proportional to changes in pressure and temperature, and thus suggests that G may be best regarded as a function of P and T Since environmental systems at equ ilibrium are found to be at constant temperature and pressure, minimization of the Gibbs free energy is the equilibrium criterion that the system is striving to wards (Mackay, 2001). The derivative of the differential shown in equation 9 is, ( G / T )P = -S (10) ( G / P)T = V (11) These relations show how the Gibbs energy va ries with temperature and pressure. For a pure substance, at constant temper ature and pressure, the molar Gibbs energy, Gm, is equivalent to the chemical potential, for that substance, and is defined as = ( G / n )T,P (12) As it happens, at equilibrium, the chemical potential of a substance is the same throughout a system, regardless of how many compartments there are (Mackay, 2001). To see the validity of this statement, consider a system in which the chemical potential of a substance is 1 at one location and 2 at another location. When an amount dn of the substance is transferred from one location to the other, the Gibbs energy of the system changes by 1dn when material is removed from location 1, and changes by + 1dn when that material is added to location 2. The overall change is therefore dG = ( 2 1) dn If the chemical potential at location 1 is highe r than that at location 2, the transfer is accompanied by a decrease in G and so has a spontaneous te ndency to occur. Only if 1 = 2 is there no change in G and only then is the system at equilibrium. Thus chemical
17 potential could be used as a criterion for equilibrium in determining the direction of mass diffusion (Mackay, 2001). However, since chemical potentials are loga rithmically related to pressure, as seen in the equation, = + RT ln ( P / P) (13) and since chemical potentials are difficult to measure, their use in calculating the concentration of a substance in various media is limited (Mackay, 2001). It is possible to adapt equation 13, which represents the chemical potential for an ideal gas, by replacing P by an effective pressure, referred to as fugacity, f and rewrite equation 13 as, = + RT ln ( f / P) (14) The term Â‘fugacityÂ’ comes from the Latin fo r Â‘fleetnessÂ’ in the sense of Â‘escaping tendencyÂ’ (Mackay, 2001). Fugacity has the sa me units as pressure, and at low partial pressures, under ideal conditions, fugacity is eq ual to the partial pressure of a substance, and is therefore linearly rela ted to concentration (Mackay, 20 01). In addition, fugacity is logarithmically related to chemical potential thus, it is a measure of the molar Gibbs energy, and as such can replace chemical potential as a criterion for equilibrium (Mackay, 2001). Given that fugacity is linearly related to concentration, the following relation can be used Ci = Zif (15) where Z is a proportionality constant, referred to as the fugacity cap acity, having units of mol m-3 Pa-1, and is analogous to heat capacity (Mackay, 2001). Substitution and rearrangement of equation 15 into equation 1 yields,
18 f = M / ViZi (16) thus the fugacity of a substance can be read ily obtained given that the total mass of a substance, and the volume of individua l environmental compartments are known constants. It is possible to determine the Z value for a particular environment by recalling the relationship that exists for partition coefficients, as seen in equation 2. Substitution and rearrangement of equation 2 into equation 15 thus leads to (12) K12 = C1/ C2 = Z1f / Z2f = Z1 / Z2 (17) and Z1 = Z2K12 (18) Since the Z value for air is equal to 1/ RT for systems in which the ideal gas law applies, Z values for other compartments can theref ore be readily obtained. For instance, the Z value for water is equal to ZaKaw, since it can be seen from equation 18 that Kaw = Za/ Zw (Mackay, 2001). In order to determine the overall persis tence of a substance it is important to recognize that the calculations shown above merely describe how a substance will partition in an environmental system, essent ially describing a Level I approach. To evaluate the length of time a substance will persist in that system it is necessary to consider processes that are responsible for removing it from the system. Generally there are two removal processes that are consid ered: removal by advection, and removal by degrading reactions. Since overall persisten ce is a measure of how long a chemical will remain unchanged in an environment, pro cesses involving removal by advection are not considered. Thus, advection residence times can be made to be infinity, implying that the substance never flows out of the closed system.
19 In a steady-state system at equilibrium, in which a chemical is being discharged at a constant rate into the system, the rate at which a chemical is being input to the system must equal the output rate (Mackay et al, 1996 ). For a first-order reaction of the form A B (19) the rate equation is d [A] / dt = k [A] (20) where k is the first order rate coe fficient, having units of L mol-1 h-1. Rearrangement and integration of equation 20 yields ln [A]0 / [A]n = k ( tn Â– t0) (21) Since t0 = 0 h, equation 21 can be rewritten as ln [A]0 / [A] = kt (22) The rate at which a substance is removed from an environmental system is related to its half-life. The half-life ( t1/2) of a reactant is defined as the time required for the concentration of the reactant to decrease to halfway between its initial and final values. Thus, at the half-life, where [A] = [A]0/2, equation 22 can be rewritten as ln [A]0 / ([A]0/2) = kt1/2 (23) or t1/2 = ln2/ k (24) The residence time ( ) of a reactant in a system is defined as = M / E (25) where E is the efflux of the reactan t, which describes the rate of removal of a chemical from a system. In a perfectly mixed clos ed system the following relationship exists between residence time and half-life
20 t1/2 = ln2 (26) For simplifying purposes, environmental degrada tion processes are considered to proceed following pseudo-first order reaction kinetics. Thus the relations described in equations 24 and 26 are referred to in carrying out Leve l II type calculations, which incorporate a degradation D value, having units of mo l/Pa.h, and defined as (12) Di = ViZiki (27) Values for Zi are obtainable from equa tion 18, while values for ki are obtainable from equation 24 using media specific half-lives. When a D value is multiplied by the fugacity of a substance a transport rate is obtained in terms of mol h-1, and is therefore similar to the rate constant seen in equation 20 (Mackay, 2001). The fugacity for a substance in which degradation processes are bein g considered can be written as f = I / Di (28) where I is the input rate a nd has units of mol h-1. Thus, using equation 15, it is possible to determine the concentration of chemical in an environmental compartment while taking degradation processes into consideration. From the concentrations present in each compartment, a value for the total amount of chemical present in the system can be determined, from which the residence time, using equation 25 can also be evaluated. Finally, the overall persistence, or overall half-life, of a substance is determined by equation 26.
21 Degradation Processes The persistence of a chemical in an eval uative environmental system is essentially described in terms of its half-life. The half-life is the time it takes for half of the amount of chemical to be removed from the environm ent. The actual rate of disappearance of a chemical from the environment will depend on the processes available for removing it. These processes, which will have different importance for different environmental compartments and in different parts of the globe, determine the ove rall persistence, and thus the persistence of the chemical. Ther efore a brief discussi on pertaining to the various mechanisms involved in removing a ch emical from an environmental system is warranted. Typical removal processes for a substa nce are associated with biological and chemical degradation processes. The most important environmen tal reaction processes are typically associated with biodegradati on, hydrolysis, oxidation, and photolysis. Biodegradation Biodegradation refers to the transfor mation of an organic compound into other compounds through microbial action, whic h can occur in the environmental compartments of water and/or soil and se diment. The agents of biodegradation are primarily bacteria and fungi. Each group is remarkably diverse, although the metabolic capabilities of bacteria as a group tend to be greater (Hemond and Fechner, 1994). Bacteria are active under both aerobic and anaerobic conditio ns, while fungi are active under only aerobic conditions.
22 Generally, the chemical transformations for which bacteria and fungi are capable can be described by two important principles. First of all, microbes generally mediate biotransformations that are energetically favourable (Hemond and Fechner, 1994). In other words, reactions involvi ng microbial activity result in a net decrease in the Gibbs free energy of the chemical system, with th e microbes utilizing the released energy for their own use. This can result in a signifi cant increase in the microbial population, which could dramatically affect the rate at whic h a chemical compound is transformed (Mackay, 2001). This principle also suggest s that if a chemical is presen t at a concentration that is too low to provide sufficient energy to the mi crobes, that it may essentially be ignored, and not transformed (Mackay, 2001). Secondly, most chemical transformations are accomplished by means of enzymes, which are proteins synthesized by organisms that act as a catalyst in the biotransformation of an orga nic compound (Hemond and Fechner, 1994). The role of an enzyme is to bind reactants and hold them in such an orientation that they are more readily available to react. The products of the reaction are then released, making the catalyst available to facilitate another tran sformation. Individual organisms will produce different enzymes, suggesting that some microbes can accomplish a certain chemical transformation, while other microbes cannot (Hemond and Fechner, 1994). Generally, enzymes are well adapted to chemically tr ansforming most naturally occurring organic compounds. However, most microbes have ye t to evolve the capability of transforming many of todayÂ’s synthetic organic compounds which do not naturally occur (Mackay, 2001). These include compounds containing a large number of branched carbon chains,
23 ether linkages, meta-substituted benzene rings, chlorine, amines, methoxy groups, sulfonates, and nitro groups. Given these factors, the rate at which a chemical is transformed will depend on the nature of the chemical compound, on th e amount and condition of enzymes which may be present in various organisms in va rious states of activation, and which are available to perform the chemical transformati on, on the availability of nutrients such as nitrogen, phosphorous, and oxygen, as well as temperature and the presence of other substances which might help or hinder th e transformation (Macka y, 2001). Thus the biotransformation of organic compounds can be extremely complex, and difficult to predict. In order to assign a rate constant fo r reactions involving microbes a number of simplifying assumptions have been made. These include the assumption that biodegradation occurs following first-order ki netics, which, given the number of factors which are necessary to facilita te the biodegradation of a co mpound is highly inaccurate. Thus by making estimates of what the first orde r rate constant or half-life is based on experiment and experience is the biode gradation rate of a compound assigned. Hydrolysis The process of hydrolysis involves the addi tion of water to a chemical species as a result of reaction with water, hydrogen i on, or hydroxyl ion (Mackay, 2001). There are two classes of organic compounds that are lik ely to undergo hydrolysis. The first class includes alkyl halides, which are straight-c hain or branched hydrocarbons in which a
24 hydrogen has been replaced by a chlorine, fluorine, bromine, or iodine atom. The reaction with water proceeds as follows: H2O + R-X RÂ–OH + H+ + XThe second class of compounds that ma y undergo hydrolysis includes esters and ester analogs (hemond and Fechner, 1994). Es ters are compounds containing a modified carboxylic acid group, in which the acidic hydrogen atom has been replaced by some other organic functional group. Hydrolysis conv erts esters into th e parent organic acid, plus an alcohol. For instance ethyl acet ate hydrolyzes to acetic acid and ethanol according to the overall reaction: H2O + CH3COOC2H5 CH3COOH + C2H5OH The intermediate steps that occur during the hydrolysis of alkyl halides and esters may involve a number of chemical species, including H2O, OHand H+. Where hydroxyl and hydrogen ions are involved as part of th e rate-limiting step, the hydrolysis reaction rate will be very sens itive to the pH of th e water (Mackay, 2001). Thus, one method of testing if a compound might be subject to hydroly sis is to subject the chemical to pHs of 3, 7 and 11, and observe the rate of decay (Mack ay, 2001). From this test it will then be possible to assign rate constants for acid, base and neutral hydrolysis, which can be combined to give an expression for th e rate at any given pH (Mackay, 2001). Oxidation A chemical may react with oxygen, an activ ated form of oxygen, such as singlet oxygen, with ozone, with hydrogen peroxide, or with various free radicals, most notably with hydroxyl and nitrate radicals (Mackay, 2001). Th e rapid reaction of hydroxyl
25 radicals with many trace gases in the atmos phere places it at the cen ter of much of the daytime atmospheric chemistry (24). This is be cause sunlight plays a critical role in the formation of hydroxyl radicals as s een in the following reactions: (1) O3 + h O(1D) + O2 (2) O(1D) + H2O 2OH with the hydroxyl radical going on to react w ith a large number of chemical species, including organic compounds. At night, atmospheric chemis try is dominated by the nitr ate radical, which can be formed due to reaction with ozone as follows: NO2 + O3 NO3 + O2 The nitrate radical then acts as a hydrogen at om abstractor in much the same way as the hydroxyl radical. CH4 + NO3 CH3 + HNO3 It should be noted that the atmosphere is a highly reactive medium, and that the examples shown above only represent a small fraction of the number of reactions that might be possible. Rates involving thes e reactions can be estimated by performing kinetic experiments in which the rate of decay at which a substance in contact with an oxidant is monitored, and a rate constant establis hed (Mackay, 2001). Photolysis A molecule may absorb electromagnetic radiation and, in the process, break down into its atomic or molecular components. Such chemical reactions are referred to as photochemical, and the process by which a phot ochemical reaction occurs is called
26 photolysis. The rate constant for a photochemical reaction may be of second order. For instance the reaction rate for NO2 can be written as Â–d[NO2] / d t = kÂ” [ h][NO2]. However, this expression is not very useful since the second-order rate constant would vary significantly with the energy of the photon involved in the reaction. Thus, it is simpler if a pseudo first-order process could be used to describe the rate of reaction. This can be done by assuming that th ere is a constant flux of phot ons with a fixed distribution with respect to wavelength. Thus, the rate expression gi ven above would become d[NO2] / d t = j [NO2], where j is the pseudo first-order rate coefficient that accounts for the absorption coefficient of the reactant, the quantum efficiency of the reaction in question and the solar spectrum and intensit y at the altitude and latitude under consideration. With a little information on the spectral characteristics in which the molecule absorbs light and the amount of inco ming radiation, it is relatively simple to make estimates of j for a number of compounds, from whic h the rate constant or half-life can be determined. Using Partitioning Data to Identify Key Half-lives Recognizing the need to minimize phys ical-chemical and reactivity data requirements, Gouin et al. (2000) have proposed an appro ach that would use partition coefficients as a means of establishing the compartment to which a substance is most likely to partition. This method is similar to the Level I approach, in that it uses an equilibrium, steady-state mass balance model to determine the overall persistence of a substance by first identifying the relevant half -lives. The advantage of this approach is that the method focuses on the mass fractions of a chemical in each medium, which can
27 be useful in assessing to which compartmen t a particular substance is most likely to partition. To this end, Gouin et al. (2000) have categorized the environment into compartments of air, water and octanol, where the octanol compartment represents the organic carbon content associat ed with both the soil and sediment compartments. An assumption has been made that suggests that the half-life in soil for a substance is equivalent to its half-life in octanol. This has been justified by de monstrating that 97.8% of the equivalent volume of octanol come s from soil, with th e remaining 2.2% coming from sediment. Since the soil compartmen t represents the medium from which the octanol is largely derived, it can be assumed that the degr adation processes a substance undergoes in the soil will be equivalent to those that the subs tance undergoes in the organic fraction of the soil. Knowing the volume ratios of each environmental compartment, the air-water partition coefficient, KAW, and the octanol-water partition coefficient, KOW, it is possible to determine the mass fraction of a chemical in each medium by, Fi = ( ViKiw / KwwVw + KAWVA + KOWVO) (36) where the subscript i is air, water, or octanol and Kww is 1.0 (29). If a substance is found to have a mass fraction in a particular compar tment that is greater than 99%, then it is unlikely that the half-lives for the other compartments will be required in determining the overall half-life. Using equation 36 to determine the partitioning behaviour of 233 chemicals, Gouin et al. (2001) observed that 60% of the substances were found to require a minimum of 2 half-lives, with the rema ining 40% being identified as multimedia. Chemicals that are found to be multimedia w ould require information regarding mode of entry, which could affect the overall half-life of these substances, more so than those
28 partitioning primarily to one environmental medi um. Using the identified key half-lives of the remaining substances, the overa ll half-life is then determined by, 1/ R = FA / A + FW /W + FO / O (37) Thus, the highlight of this method is its abil ity to identify multimedia substances, which are known to require mode of entry data in asse ssing their persistence. In addition to this, the identification of key half-lives also lead s to a reduction in reactivity data needed to assess a chemicalsÂ’ persistence in a steady st ate system that is at equilibrium. The accuracy of the results, however, is ultimat ely limited by the accuracy of the partition coefficients used, as well as the assumptions made pertaining to the rates of degradation in octanol. Generally this method is simple to understand, easy to use, requires minimal data input and produces reliable results. Given these attribut es, it is believed that this method can provide a valuable screening t ool for assessing how substances might partition in an environmental system. This in formation can be useful in defining the fate of a chemical, and can furthermore be used to better define which parts of an environmental system that the chemical is most likely to be found. Application of Level I and II modeling to pesticides commonly used in Belize Table 1 shows the main currently-used pest icides (CUPs) in the banana and citrus industries in Belize. Physical-chemical pr operty data for these chemicals have been obtained from Mackay et al.( 2006), or have been estimated using the EPIWIN suite of programs (Meylan, 1999), largely due to a pauc ity of property data available for these compounds. In many instances, where data are av ailable, a wide range of values exist, and where data are not available the estim ation methods used may or may not be
29 appropriate. Thus, caution is recommended when interpreting results. Nevertheless, it is believed that the property data should be su fficient to qualitatively predict the likely environmental fate of the pesticides listed in Table 1. Table 1: Physical-chemical property data for selected CUPs used in Belize. Data are taken from various sources, including Mackay et al. (2006) and estimated using the EPIWIN software package (Meylan, 1999). CAS No Chemical Name MW g/mol MP deg C Input Data PS Pa S g/m3 log KOW (298 K) 13194-48-4 Ethoprop 242.34 20 0.0465 7.00E+02 3.59 1897-45-6 Chlorothalonil 265.911 250 0.133 C6.00E-01 2.64 1563-66-2 Carbofuran 221.252 151 8.00E-05 3.51E+02 2.32 1071-83-6 Glyphosate 169.074 230 4.00E-05 C1.20E+04 <0D 13071-79-9 Terbufos 288.431 -29.2 0.0427 5.00E+00 4.48 1910-42-5 Paraquat A 257.16 300 1.34E-05 6.20E+05 <0 D 95465-99-9 Cadusafos A 270.39 20 0.120 2.48E+02 3.9 23135-22-0 Oxamyl 219.261 109 0.0306 2.82E+05 <0 D 22224-92-6 fenamiphos A303.36 49 0.000133 3.29E+02 3.23 34256-82-1 acetochlor A 269.77 0 0.00373 2.23E+02 3.03 70585-38-5 Bitertanol 337.415 118 1.00E-06 5 4.1 215934-32-0 azoxystrobin B 119446-68-3 difenoconazole A406.27 76 2.43E-06 1.50E+01 4.3 116-06-3 Aldicarb 190.25 99 0.004 6.00E+03 1.1 133855-98-8 epoxiconazole B 141517-21-7 trifloxystrobin B 2921-88-2 Chloropyrifos 350.6 41 0.00227 7.30E-01 4.92 121-75-5 Malathion 330.36 2.9 0.001 1.45E+02 2.8 1861-32-1 Dacthal 332 156 0.0066 2.92E-02 4.24 333-41-5 Diazinon 304.36 25 8.00E-03 6.00E+01 3.3 86-50-0 Azinphosmethyl 317.324 73 3.00E-05 3.00E+01 2.7 A Data obtained from EPIWIN software B No data available for these compounds C Data reported have large variability, selected value to be used with caution D Values of log KOW are negative, a value of 0 is assumed for illustrative purposes for Figure 1.
30 Using the method described by Gouin et al. (2000) the Leve l I partitioning behaviour of each of the chemicals listed in Table 1 has been calculated. Results are illustrated in Figure 1. Figure 1 is a plot of log KAW vs log KOW including points representing the partitioning properties of the target pesticides in this study. The 45 diagonals are lines of constant log KOA, the octanol-air partition coefficient, because KOA is KOW/ KAW or log KOA is log KOW log KAW. Lines of constant FA, FW, and FO are drawn in Figure 1 using the above volume ratios. Th e lines corresponding to one-third in each compartment converge at the point where log KOW is 3.1; log KAW is -2.74; and VW, KAWVA, and KOWVO are equal, i.e., KAW is VW/ VA or 1300/650 000 or 0.002 and KOW is VW/ VO or 1300/1. If the ratios of the volumes ch ange, for example, if the water volume is increased, the location of this central poi nt will move upward along the 45 diagonal changing the location of the lines of constant Fi The 1% and 99% lines divide the KAW/ KOW space into regions in which partitioning is predominately into one medium and in which it is likely that degr adation in that medium is mo st important. In the region to the upper left, where more than 99% is in ai r, the air half-life probably controls the persistence, and it is unlikely that half-lives in water, soil, or sediment are required. To the lower right of the 99% octanol line, substa nces are strongly sorbed, and only data for soils and sediments are likely to be neede d. To the lower left, water is the dominant medium of partitioning. Half-liv es in air are generally shorter than those in water, soil, and sediment, largely because of relatively rapid hydroxyl radical r eactions. As a result, even 0.5% partitioning to air can represent an appreciable fraction of the overall degradation. The 0.1% air line (a lso shown) may be a better li mit. Substances with more than 1% in each medium are classified as Â“multimedia.Â” For these substances, all half-
31 lives will be required. Similarly there are regions where only two half-lives may be required. -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 0123456789Log KOWLog KAW33% octanol 33% air 0.1% air 33% waterWaterAir OctanolWater & Octanol A & W & O Air & Octanol1% water1% octanol1% air18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 -1 -2 -3Log K OA Air & Water ChlorothalonilDacthal Terbufos Chorpyrifos Cadusafos Diazinon Ethoprop Mancozeb A cetochlo r Malathion A zinfosmethyl A ldicarb Fenamiphos difenoconazole bitertanol Carbofuran Oxamyl Glyphosate Figure 1: Environmental partitioning of CUPs lis ted in Table 1 in an environmental system that is at steady-state and equilibrium, consistent with the model environment described by Gouin et al. (2000).
32 Oxamyl, glyphosate, paraquat, and aldi carb all show a strong propensity for partitioning to the aquatic environment, whereas mancozeb, acetochlor, malathion, ethoprop, azinfosmethyl, fenamiphos, carbofuran bitertanol, and difenoconazole are found to largely partition equally between octanol and water (Table 2). For these substances their mode-of-entry into the envir onment may influence their overall fate. For instance, if emitted to soils they may likely remain largely sorbed to the organic carbon content found in the soil, and if degradation in soil is relatively fast in comparison to advection from soil to water, they would remain in the soil compartment. For the remaining substances the mass fraction in the air becomes increasingly important, suggesting the potential for transport away from sites of use due to surface-air exchange processes. For these substances mode-of-entry will also likely be important. In Table 2 estimates regarding reactivity processes in air, water, soil, and sediment have been assessed, and the overall half -life of each of the substances calculated based on the partitioning behaviour of the substance, using equation 37. The results suggest that the majority of substances will ha ve an overall environmental half-life that is <1000 h or about 6 weeks. The data reporte d in Table 2, however, is based on Level II calculations, which assume the mo del environment is steady-stat e and at equilibrium. To assess the fate of these CUPs in an environm ent that is steady-state but non-equilibrium, a Level III model calculation was carried out. For simplification, an assumption has been made that 80% of the substance is emitted to soil, 10% to water, and 10% to air. The model environment has also been parame terized to more closely resemble the environment of Belize. Environmental para meters used and results for each of the substances are illustrated in Figures 3 Â– 21.
33 Table 2: Estimated reactivity data based on Mackay et al. (2006) and the EPIWIN software package, mass fractions in ai r, water, and octanol and overall environmental half-life. CAS No Chemical Name HL air hours HL water hours HL soil hours HL sed Hours FA FW FO Overall Half-life (hours) 8018-01-7 Mancozeb 1.21 480 1680 3360 0.08% 49.42% 50.49% 4.95E+02 13194-48-4 Ethoprop 3.72 360 720 3240 0.08% 24.89% 75.03% 5.13E+02 1897-45-6 Chlorothalonil 170 170 550 1700 89.88% 7.56% 2.56% 1.73E+02 1563-66-2 Carbofuran 5 170 550 1700 0.00% 86.06% 13.94% 1.88E+02 1071-83-6 Glyphosate 170 1700 1700 5500 0.00% 99.92% 0.08% 1.70E+03 13071-79-9 Terbufos 1.06 900 1800 8100 2.00% 4.02% 93.99% 5.16E+01 1910-42-5 Paraquat 12.1 900 1800 8100 0.00% 99.92% 0.08% 9.00E+02 95465-99-9 Cadusafos 2.14 360 720 3240 0.37% 13.92% 85.71% 3.04E+02 23135-22-0 Oxamyl 11.4 900 1800 8100 0.00% 99.92% 0.08% 9.00E+02 22224-92-6 fenamiphos 3.3 900 1800 8100 0.00% 43.17% 56.82% 1.25E+03 34256-82-1 acetochlor 5.18 1440 2880 13000 0.05% 54.60% 45.35% 1.58E+03 70585-38-5 bitertanol 5.79 900 1800 8100 0.00% 9.30% 90.70% 1.65E+03 215934-32-0 azoxystrobin 119446-68-3 difenoconazole 11.9 4320 8640 38900 0.00% 6.07% 93.93% 8.14E+03 116-06-3 Aldicarb 5 550 1700 17000 0.00% 99.03% 0.97% 5.52E+02 133855-98-8 epoxiconazole 141517-21-7 trifloxystrobin 2921-88-2 Chloropyrifos 17 170 170 1700 0.33% 1.52% 98.14% 1.65E+02 121-75-5 Malathion 17 55 55 550 0.03% 67.14% 32.83% 5.50E+01 1861-32-1 Dacthal 291 206 412 1338 51.13% 3.38% 45.49% 3.31E+02 333-41-5 Diazinon 550 1700 1700 5500 0.32% 39.15% 60.53% 1.69E+03 86-50-0 azinphosmethyl 1.7 900 1800 8100 0.00% 72.02% 27.97% 1.02E+03
34 Figure 2: Environmental parameters used in Level III calculations.
35 Figure 3: Level III results for mancozeb.
36 Figure 4: Level III results for Ethoprop
37 Figure 5: Level III results for chlorothalonil
38 Figure 6: Level III results for carbofuran
39 Figure 7: Level III results for glyphosate
40 Figure 8: Level III results for terbufos
41 Figure 9: Level III results for paraquat.
42 Figure 10: Level III results for cadusafos.
43 Figure 11: Level III results for oxamyl.
44 Figure 12: Level III results for fenamiphos.
45 Figure 13: Level III results acetochlor
46 Figure 14: Level III results for bitertanol
47 Figure 15: Level III results for difenoconazole.
48 Figure 16: Level III results for aldicarb.
49 Figure 17: Level III results for chlorpyrifos.
50 Figure 18: Level III results for malathion.
51 Figure 19: Level III results for dacthal.
52 Figure 20: Level III results for diazinon.
53 Figure 21: Level III results for azinphosmethyl.
54 Table 3: Summary of results from Level III calcul ations regarding fractions in air, water, and soil, as well as Level III overall half-life. Note: assumed that emissions are 80% to soil, 10% to water, and 10% to air. CAS No Chemical Name FA FW Fsoil Overall Half-life (hours) 8018-01-7 Mancozeb 0.263.6295.7 1964 13194-48-4 Ethoprop 0.564.294.3 919 1897-45-6 Chlorothalonil 6.722.8190.4 512 1563-66-2 Carbofuran 0.87.192.0 625 1071-83-6 Glyphosate 0.5157.142.0 974 13071-79-9 Terbufos 0.281.5693.7 2374 1910-42-5 Paraquat 0.4951.447.7 874 95465-99-9 Cadusafos 0.553.8494.0 929 23135-22-0 Oxamyl 0.5851.447.6 868 22224-92-6 Fenamiphos 0.234.2694.8 2152 34256-82-1 Acetochlor 0.185.1694.1 3110 70585-38-5 Bitertanol 0.222.694.1 2311 215934-32-0 Azoxystrobin 119446-68-3 Difenoconazole 0.050.9595.5 10668 116-06-3 Aldicarb 0.637.561.6 864 133855-98-8 Epoxiconazole 141517-21-7 Trifloxystrobin 2921-88-2 Chloropyrifos 0.070.3293.9 11384 121-75-5 Malathion 6.281083.6 79.9 1861-32-1 Dacthal 1.731.6895.9 537 333-41-5 Diazinon 0.275.193.8 2074 86-50-0 Azinphosmethyl 0.277.9691.4 1924
55 Conclusions The Level III calculations for this group of CUPs indicates that the majority of substances when emitted 80% to soil, 10% to wa ter, and 10% to air, will remain primarily in the soil compartment to which they are emitted, with the overall environmental halflife being largely influenced by estimates rega rding their biodegradati on in soil. Similar to the Level II calcula tions, four of the CUPs, oxamyl, glyphosate, paraquat, and aldicarb all show a strong propensity for partitioning to the aquatic environment, however, not to the same extent as the Level II calculation. In this case, because the substances are emitted largely to soils a significant fracti on of the pesticide will remain in that compartment, although almost half will pa rtition to the water phase. For these substances, both the water and soil half-life will influence both their environmental fate and overall environmental half-lif e. It is also notable that in the Level III calculations both malathion and chlorothalonil show si gnificant fractions in the air phase, approximately 6% for each compound. For th ese substances assumptions regarding the air half-life become increasingly important for assessing their overall environmental halflife and environmental fate. For malathion, the relatively short ai r half-life (17 h), strongly influences its overall half-life, whereas the longer ai r half-life for chlorothalonil (170 h) will likely influence its ability to undergo long-range transport, while similarly influencing the overall half-life. Both chlorpyrifos and difenoconazole have mass fractions <1% in water, thus their environm ental fate and overall persistence will be strongly influenced by their reac tivity in soils and sediments. In general, it is likely that the majority of CUPs investigat ed here will have detectable amounts in surface and ground water n ear to where they are being used. This
56 can be quickly assessed by examining the mo lecular structure of these compounds, which tend to have polar functional gr oups associated with them. The Level I, II, and III model calculations presented here he lp to better quantify their e nvironmental fate. Combined with empirical measurements, model output can be compared with field data, and sources of error in the model calculations better defi ned. Currently, in the absence of empirical data, the model output should be interpreted with some caution, particularly given the largely unknown uncertainties a ssociated with the physical-c hemical properties of each the CUPs, errors associated with assumptions regarding their environmental degradation and mode-of-entry information, and assumptions relating to the parameterization of the physical environment used in the Level III model, which has been adjusted to better resemble the Belize environment. Because mobility of pesticides thro ugh soil profiles is dependent upon water movement it is essential that the hydrology be simulated accuratel y. Having reliable data from the field on hydrological characteristics of soil is obv iously of great importance. Having parameterised the model it is then n ecessary to develop a realistic simulated water balance (distribution of incoming precipitation between evapotranspiration, surface run-off, plant uptake and groundwater rechar ge). Once a realistic water balance is obtained it is then necessary to consider th e behaviour of the ch emical within this hydrological regime. Generally, the most sensit ive environmental fate parameters are the rate of degradation (which a ttenuates the concentration of pesticide in the soil profile) and the sorption coefficient (which controls the relative rate of transport through the soil). For the CUPs studied these va riables are indeed highly uncer tain. Thus, where sufficient hydrological data are available and it is possi ble to duplicate water ba lances and transport
57 of water through soil with reasonable accuracy there is a much greater chance of correctly predicting chemical transport. It is ther efore essential that characterization and parameterization of soil hydrology be carried out with great care. Numerous exercises have demonstrated that where it is possible to calibrate against observed hydrology data the accuracy of pesticide leaching simulations can be significantly increased. It must be recognized that each model has its own limitations a chromatographic flow model, for instance, canno t be used to accurately simulate leaching in cracking clay soils where preferential flow dominates. It is the recognition of model limitations that is one of the keys to adva ncing higher quality m odeling essentially a matter of choosing the right tools for the job. In this modeling exercise we have used evaluative models to assess th e likelihood that the CUPs liste d will accumulate in surface waters in Belize. Simplified assumptions have been made regarding mode-of-entry, assuming that the majority of the chemical is applied directly onto th e soil surface, with a fraction (10%) being emitted to air and surface waters as a result of spray drift, and model output has been based largely on estim ated physical-chemical property data, which are likely to be reasonable but not accurate. It is thus impor tant to appreciate how these assumptions influence the output of the model. It is suggested that the modelling wo rk with the Mackay-type models thus provides a first tier, of a tiered approach, with respect to assessing the environmental fate of CUPs in Belize. In this instance, the results suggest that a small fraction of the chemical applied to agricultural lands will migrate to surface waters. The next step would be an attempt to accumulate field data, reporting levels in the air, water, and soil, as well as the collect ion of other important paramete rs, such as soil organic carbon
58 content, information pertaining to soil hydr ology, application rate s, and meteorological information such as temperature and rain rate s. This information can then be used to assess the performance of the model, and hope fully to improve parameterisation. If the field and model data show good agreement, th e third step in the process would be an attempt to better assess environmental fate and transport, providing insight regarding environmental risk.
59 Chapter Three Levels of Pollutants in Coasta l Waters in Southern Belize Introduction Previous studies have shown that coasta l waters are susceptible to contamination from land-bases sources (Saison et al ., 2008; Hapeman et al., 2002: Leonard, 1990; Wauchope, 1978). Pollutants in coastal wate rs may originate from agricultural areas (pesticides, excessive nutrient s, pathogens), urban areas (pes ticides, polycyclic aromatic hydrocarbons, metals, polychlorinated biphenyl s, flame retardants, hydrocarbons, etc.), industrial parks (org anic solvents, flame retardants, fuel, polycyclic aromatic hydrocarbons, metals, etc.), vehicles (hydrocar bons, oils, etc.) and a myriad other sources (Jeong et al., 2008; Hou et al., 2006: Southwick et al., 2002 ; Dietrich and Gallagher, 2002). Coastal areas are also known for their tremendous value, both ecologically and economically (Cooper et al., 2009; Burke et al 2008). They are important areas for spawning of many valuable species of fish and also serve an important function for recreation and tourism (Cooper et al., 2009: Burke et al., 2008). As a result, protection of coastal areas is at the top of the environmental agenda of all countries with coastlines. In fact, the United Nations, through its United Nations Environmental Programme (UNEP), has made coastal protection one of its key ini tiatives. In the Wide r Caribbean countries it has set up a programme to fund research on La nd-Based Sources of Pollutants to Coastal Waters.
60 In countries such as those in the wide r Caribbean, there is special concern about the presence of pollutants from land-based s ources because many of these countries have coral reefs in their coastal waters. The heal th of coral reefs has been in decline for several years, and although coral bleaching due to warming waters has been implicated as the main culprit, there exists the distinct possibility that pollutants from land-based sources may at the very least be c ontributing to coral reef decline. Unfortunately, in most countries of th e Caribbean very few studies have been carried out to document the extent of pollution in coastal areas. This is due in part to scarce resources for scientific research and a lack of analytical facilities and trained personnel to carry out su ch studies. Such is the case in Belize. Research Project and Objectives As an initial effort to remedy the situation in Belize, we have carried out a campaign to document the extent of pollution in coastal waters of southern Belize with respect to pesticides and sele cted heavy metals. Our majo r objectives were: (1) Measure levels of selected agricultural pesticides in coastal waters. (2) Determine if any pesticides are transported as far out as areas containing coral reefs. (3) Compare the results of the sampling campaign with the results from the predictive Level III fugacity model. (4) Measure levels of mercury and lead in coas tal waters and determine potential sources.
61 Research Area Since both the banana and citrus industri es in Belize are concentrated primarily on the fertile, flat lands along th e southern coastal areas of Be lize and, in the case of citrus, on the Stann Creek Valley, Geographic Informa tion Systems (GIS) was used to visualize and focus our research area. Using the in formation gathered during interviews with stakeholders (chapter 2) a nd ArcGis 9.0, the major watershe ds that were potentially affected by these industries we re identified and targeted fo r sampling. A total of eight rivers were selected for sampling, starting with North Stann Creek River which empties directly in front of Dangriga Town on the north and going as far south as the Sarstoon River, which borders Belize with Guatemala (F igure 4). The three northernmost rivers, North Stann Creek River, Sittee River and Sout h Stann Creek River, were identified as the primary water source for the citrus industr y with a couple banana farms also using the latter river. In the case of North Stann Creek River, several incidences of fish kills have been reported to the Department of Envi ronment in the past, with claims that agrochemicals were to blame (Mai, pers. comm.). Identified further south were Mango Cr eek and Big Creek which were categorized as one for purposes of forming a transect they empty within 200 m of each other within a lagoon and Monkey River, both of which drain ar eas dominated by banana farms. Over 60% of the banana plantations use as thei r only water source the Swasey and Bladden Rivers, which join to form Monkey River (see Figure 4). These ri vers are intensively used by the banana plantations for a variety of purposes including chemical preparation, irrigation and processing. Further south was our reference river, Swasey River, whose
62 watershed has protected status as a biological corridor ma naged by YaxÂ’che Conservation Group and TIDE. The last two rivers in th e research area include the Rio Grande and the Sarstoon Rivers. Rio Grande River was selected b ecause it flows through ar eas characterized by low impact agriculture such as small scale rice and citrus plantations and subsistence farming. A particular point of interest with this river is that it has a dump site only a couple miles from the river mouth. Punta Go rda Town (the main urban centre in the Toledo District) and neighboring communities dump all categories of waste in this site. The southernmost river, Sars toon River, was selected si nce it borders Belize with Guatemala, which has large-scale liv estock rearing and agriculture. Figure 22. Sampling region and sampling stations.
63 Materials and Methods Sampling Sites Obtaining both spatial and temporal data during the project lifespan was the major consideration in choosing the sampling scheme and sampling campaign times. Meteorological data, especially rainfall, was an alyzed to select spec ific sampling periods for the rain and dry season. For purposes of this thesis two sampling campaigns were undertaken, one in late Februa ry Â– early March coinciding wi th the dry season and one in June at the start of the rai ny season. This will allow us to determine any differences in surface runoff between the two seasons. In order to obtain spatial resolution within the constraints of the project we decided to sample along transect s parallel to the coastline st arting from the mouths of all the rivers chosen for the study out to the areas containing cora l reefs. Using a hand-held Global Positioning System unit transects were laid out from each river mouth and sampling sites were chosen to make them as e qui-distant as possible. Most worked out to 2.5 Â– 3 miles apart. Table 4 has the exact co ordinates of each station. Figure 22 indicates that it was not possible to always obtain nice transects parallel to the coast.or to run the transects all the way to the ar eas containing coral reefs. Th e most extreme case of this was with the Sarstoon River. Due to the dist ance of the coral reef areas from the coast it was not possible to run a transe ct all the way out there. Because of the maritime borders existing between Belize, Guatemala and Honduras it was necessary to run the transect so as not to violate any border. Despite th ese limitations, Figure 22 shows fairly robust coverage of the study region.
64 Table 4. Coordinates of Sampling Sites. Site Name Latitude Longitude Site Name Latitude Longitude NSC 1 16 58 116 088 13 258 MR 1 16 21 922 088 29 143 NSC 2 16 58 146 088 10 512 MR 2 16 21 733 088 26 430 NSC 3 16 57 891 088 07 772 MR 3 16 21 347 088 23 741 NSC 4 16 57 641 088 05 032 MR 4 16 21 376 088 21 015 NSC 5 16 57 474 088 02 771 MR 5 16 21 594 088 18 296 SR 1 16 48 519 088 15 417 GS 1 16 13 513 088 44 053 SR 2 16 48 646 088 12 727 GS 2 16 13 494 088 41 338 SR 3 16 48 752 088 10 041 GS 3 16 13 420 088 38 600 SR 4 16 48 701 088 07 292 GS 4 16 13 048 088 35 910 SR 5 16 48 560 088 04 978 GS 5 16 12 733 088 33 210 SSC 1 16 43 427 088 18 067 RG 1 16 08 535 088 45 551 SSC 2 16 43 219 088 15 316 RG 2 16 08 355 088 42 799 SSC 3 16 43 173 088 12 578 RG 3 16 08 290 088 40 513 SSC 4 16 42 878 088 09 838 RG 4 16 08 173 088 38 253 SSC 5 16 42 478 088 07 114 RG 5 16 08 157 088 35 969 MC 1 16 32 865 088 24 666 MC 2 16 32 369 088 23 714 SAR 1 15 53 668 088 54 951 BC 3 16 30 664 088 24 039 SAR 2 15 55 398 088 52 885 MBC 4 16 29 884 088 21 413 SAR 3 15 57 517 088 51 214 MBC 5 16 29 646 088 18 629 SAR 4 15 59 197 088 49 137 MBC 6 16 28 854 088 13 238 SAR 5 16 00 581 088 46 842
65 Cleanup of Sampling Equipment and Material Prior to each sampling campaign all equipment and reagents were thoroughly cleaned to prevent sampling artifacts. Stainless steel canisters were thoroughly washed with soap and warm water, followed by several rinses with Ultrapure wate r. Each canister was sealed and triplewrapped in plastic bags. Th e plastic bottles for metal determination were washed thoroughly with soap and warm water, rinsed several times with deionized water, then washed with an acidic solution made by dilu ting ultrapure nitric acid with deionized water. The stainless steel filter holder and the stainless st eel columns for XAD-2 resin were thoroughly washed with soap and warm water, rinsed with Ultrapure water followed with pesticide-grade acetone. They were wr apped in solvent-cleaned Al foil and placed in a stainless steel case. Glass fibre filters were baked at 500 oC in an oven overnight, wrapped in solventcleaned Al foil and stored in Ziploc ba gs. XAD-2 resin was cleaned by sequential Soxhlet extractions as follows: 24-h extractio ns in pesticide-grade methanol, followed by acetone, hexane, and dichloromethane. This is followed by sequential 4-h Soxhlet extractions with hexane, followed by acetone, and finally methanol. The methanol was displaced by several rinses w ith Ultrapure water. Finally, the resin was stored in an amber bottle under Ultrapure water. Amber bottles were washed with soap and warm water, rinsed with distilled water, soaked in an acid bath for 3 da ys, and finally baked in a furnace at 450 oC.
66 Glass wool was Soxhlet-extracted overnight with pesticide-grade dichloromethane followed by petroleum ether. Sampling Water samples were collected in pre-clean ed stainless steel canisters from a small boat. Once a sampling was identified by GPS the boat was positioned so as to face the direction of the current and the engine was turned off. A 5gal stainless steel canister was then dipped into the water from the bow of the boat, ensuring that water was collected from the surface (to account for any surface-m icrolayer artifact) and from a depth of approximately 1 m. Once full, the caniste r was immediately pullout capped and stored in the shadiest portion of the boat. At each station a water probe was used to measure temperature, pH, and salinity. During the June sampling campaign samples were collected to measure concentrations of mercury and lead. At each station a pre-cleaned and pre-acidified 250mL plastic bottles was dipped quickly from th e bow of the boat from the side opposite the one where the stainless steel canister was dipped. Ultrapure concentrat ed nitric acid was added drop-wise to the water to take the pH to ~1. Each bottle was immediately placed in an ice cooler with ice. Processing Once on-shore, water for pesticide dete rmination was filtered through glass fibre filters and XAD-2 resin as follows: Teflon-lin ed tubing from the st ainless steel canister
67 to the top of a stainless steel filter holder containing a 135-mm glass fibre filter; the same type of tubing was run from the bottom of the stainless steel filter holder to the top of a stainless steel tube containing XAD-2 resin; the same type of tubing was run to a peristaltic pump. The peristaltic pump pu lled water through the assembly. The glass fibre filter is designed to trap particulate ma tter with any associated pesticides while the XAD-2 resin is designed to trap dissolved-phase pesticides. The filtration rate was set to 300 mL/min and was monitored frequently to adjust if needed to keep the rate as constant as possible, thus allowing the calculation of volume processed based on processing time. Samples collected from closer to shore often needed more than one glass fibre filter; in such cases all the filters used in a given site were combined. The steel column with XAD-2 resin wa s prepared as follows just before processing each sample: a plug of clean glass wool was added at the bottom of the tube; distilled water was added until it reached a height of approximately 20 cm; XAD-2 resin was added until the slurry reached approximate ly 25 cm; another plug of glass wool was added and the top cover of the column was secured. Once processed, the glass fibre filters we re wrapped in solvent-cleaned Al foil, placed in a Ziploc bag and stor ed in a freezer until transported for analysis. The XAD-2 resin slurry was poured in small amber bottle s with Teflon-lined lid s and refrigerated until transported for analysis. Both were tr ansported to USFSP for analysis in an icecooler with ice-packs.
68 Extraction XAD-2 resin and glass fibre filters were Soxhlet-extracted overnight (16-18 h) using 200 mL of 25% DCM/hexane. Resin a nd filters for each sample were extracted together since our objective in this projec t was to obtain overall concentrations of pesticides and not to determine partitioning be tween the dissolved and particulate phases. Extracts were concentrated using a rotary evaporator followed by a gentle stream of ultrapure nitrogen to a final volume of approxi mately 1 mL after solvent-exchanging into pure hexane. The concentrated extract was subjected to column chromatography using Florisil. A column was prepared by placing a plug of pr e-cleaned glass wool at the bottom of the column, adding 8 g of Florisil (pre-baked at 450 oC) deactivated with 200 L distilled water and overlaying with 1 cm pre-cleaned an hydrous sodium sulfate. The column was pre-eluted with 100 mL DCM followed by 100 mL hexane. The sample was placed on the top of the column and then eluted with 100 mL hexane followed by 100 mL DCM. Both fractions were concentrated and solvent-exchanged into isooctane using a rotary evaporator followed by a gentle stream of nitrogen. Analysis Pesticides were analyzed in two groups. The first group consisted of acetochlor, cadusafos, carbofuran, azoxystrobin, ethopropho s, fenamiphos, bitertanol and oxamyl. The second group consisted of dacthal, chlorpyrifos, diazino n, chlorothalonil,
69 pendimethalin, azinphosmethyl, trifluralin, carbary l, metribuzin, terbufos, dimethoate and malathion. Analytical details for the first group ar e as follows: Instrument Â– Shimadzu; detector type Â– mass spectrometer, quadrapole type; transfer line temperature Â– 290 oC; injection temperature Â– 250oC; carrier gas Â– helium; injector type Â– split/splitless set at splitless mode; injection volume Â– 3uL; colu mn Â– RTX-5MS from Restek Â– 15 meters long, 0.25um ID; detector settings Â– analyzing for ions 35 to 550; oven program initially at 90 oC for 2.0 minutes, ramp 15 oC/minute to 250 oC, hold for 3.0 minutes; instrument was run in selective ion monitoring (SIM) mode to enhance sensitivity. Analytical details for the first group are as follows: Instrument Â– Agilent 6890 GC Â– 5973; detector type Â– mass sp ectrometer, quadrapole type operated in electron capture negative ion mass spectrometry (GC-ECN I-MS); transfer line temperature Â– 250 oC; injection temperature Â– 250oC; reagent gas Â– methane; injector type Â– split/splitless set at splitless mode; injection volume Â– 2uL; column Â– DB5 Â– 30 meters long, 0.25um ID; oven program -initially at 90 oC for 1.0 minute, ramp 20 oC/minute to 160 oC, ramp 2 oC/minute to 200 oC, ramp 20 oC/min and hold for 15 minutes ; instrument was run in selective ion monitoring (SIM) m ode to enhance sensitivity. Paraquat and Glyphosate These herbicides are too polar to be sampled using the methodology detailed above. To sample for these herbicides we employed method-specific solid-phase extraction (SPE) cartridges. For paraquat we employed Ultraquat cartridges and for glyphosate we employed SAX (strong anion exch ange), quaternary amine ion-exchange
70 cartridges (both purchased from Restek). We had the Ultraquat cartridges custom-made to hold 1g of adsorbent. For paraquat, we collected 2L of wate r in pre-cleaned PVC bottles (following recommendations of EPA method 549.2). Bottles were stored in an ice-chest until further processing on-shore. Once on-shore, the Ultraquat SPE cartridges were conditioned by passing 4 mL ultrapure acetonitrile followed by 4 mL of deionized water. 1L of water was then filtered per cartridge so that 2 cartridges were used per sampling site. Filtration at a 25 mL/min was done using a six-position manifold attached to a vacuum pump. Cartridges were wrapped in pre-cleaned aluminum foil and refrigerated. For glyphosate sampling, we collected 1L of water in pre-cleaned PVC bottles. Bottles were stored in an ice-chest during sampling. On-shore, the SAX cartridges were conditioned by passing through 12 mL of a pH 6 solution made by diluting ultrapure nitric acid with HPLC-grade water to the re quired pH. 1L of sample water was then filtered through the cartridge at 5 mL/min us ing a six-position manifold attached to a vacuum pump. Cartridges were wrapped in pre-cleaned aluminum foil and refrigerated. Once each sampling campaign was completed, SPE cartridges were transported to our laboratories in a cooler with ice packs for analysis. For paraquat, an acidic solution for elution was prepared by diluting 1 mL of 85% phosphoric acid to 1L with deionized HPLC-gra de water. 2 mL of this solution was added to each cartridge and allowed to soak into the adsorbent bed for ~ 1 min. Then 4 mL of the solution was passed through the cart ridge slowly (dropwise) into glass testtubes. All test-tubes were previously deact ivated with dichlorodimethylsilane as per instructions on the reagent. The pH of the eluent was checked and if it was acidic it was
71 neutralized with drops of concentrated ammonium hydroxide; then deionized HPLCgrade water was added to adjust the final vol ume to 5 mL. The extracts from the two cartridges per site were combined into one final extract. For glyphosate, a pH 5 solution was prep ared using ultrapur e nitric acid and deionized HPLC-grade water. 2 mL of the pH 5 solution was added to each cartridge and allowed to soak into the adsorbent bed for ~ 1min. Then 13 mL of the pH 5 solution was added and slowly (dropwise) passed through the cartridge and collected in deactivated glass test-tubes. We initially planned to carry out th e analysis for paraquat and glyphosate ourselves but our instrument is not equipped wi th the appropriate detector, so we had to have those samples analyzed by a commerc ial laboratory. Both herbicides were measured by HPLC, using a photodiode array det ector with an absorbance wavelength of 257 nm for paraquat and de rivatization followed by fluorescence detection for glyphosate. As part of our quality contro l, we spiked three PVC bottles containing deionized HPLC-grade water with glyphosate an d three with paraquat to make solutions of known concentrations. These were filte red through the appropriate cartridges and processed and extracted as normal samples. Th ey were also analyzed by the commercial laboratory to determine percent recovery. We also had solutions of both herbicides of known concentrations analyzed by the co mmercial laboratory for quality control purposes.
72 Metals Due to a lack of instrumentation availabl e we had to contract out the samples for mercury and lead analysis. Mercury and l ead were measured in the water samples following EPA Method SW-846 and su itable procedures therein. Quality Control As part of our quality control, we spiked three PVC bottles containing HPLCgrade water with glyphosate and three with paraquat to make solutions of known concentrations. These were filtered through the appropriate cartri dges and processed and extracted as normal samples. They were also analyzed by the commercial laboratory to determine percent recovery. Results were unsatisfactory. For one paraquat and one glyphosate solution percent recovery was in ex cess of 90%. However, for two paraquat solutions percent recoveries were below 25% and for two glyphosate solutions results indicated below detection limits. We also ha d two solutions each of both herbicides of known concentrations prepared in HPLC-g rade water analyzed by the commercial laboratory for quality control purposes. Per cent difference between laboratory values and known concentrations were 56.2% for paraqu at and 65.5% for glyphosate. Limits of detection reported by the cont ract laboratory were 0.001 pp m for paraquat and 0.01 ppm for glyphosate. For other pesticides we prepared solu tions of known concentrations of labeled standards of azinphosmethyl, malathion, diazi non, carbofuran, and ethoprop. We spiked 6 samples from each sampling campaign each with 100 L of each of the 5 pesticides and treated as samples from extr action through analysis. Pe rcent recoveries of all 5 pesticides averaged over 90% and standa rd deviations were under 5%, indicating
73 excellent results. In addition, when running samples on the GC-MS, a solution of known concentration containing the target pesticides was run for every 10 samples to check that the instrument was working well. We also extracted four XAD-2 blanks and treated as samples. In all cases, blanks were below detection limits. The contract laboratory that analyzed for metals also followed the QA/QC protocol set out by the method th ey employed (EPA Method SW-846). Results Metals Mercury was found in all sampling sites (see Table 5 below). Concentrations were uniformly close to the method dete ction limit (0.001 ppm) suggesting natural sources for this metal. Lead was found in al l samples (see table below). Concentrations varied at different rivers, s uggesting point sources. An ongoing problem in Belize is the existence of illegal garbage dumps wher e people dispose of household goods, many containing metals. Of particular concern is the disposal of car batteries containing lead. We believe that the presence of lead is co rrelated with the pres ence of such garbage dumps. This is supported by the fact that le vels are higher at rive r mouths and higher at the mouths of rivers that drain municipa l areas (e.g. North Stann Creek traverses Dangriga Town, population ~ 10 000 and Monke y River passes near Monkey River Village, a village with a year-round population of around 4 000 people). Surprisingly, lead levels were significant at the mouth of Golden Stream, which drains mostly protected lands. One possibility is the pr esence of illegal dumpsites near this river.
74 Glyphosate and Paraquat Glyphosate and paraquat were below detection limits for all samples. This is despite that these two herbicides are by far the most heavily used in citrus and banana farms. There are two possible explanations for these results. Firs t, both paraquat and glyphosate are known to degrade very quickly in the environment. By the time water that flows through farms reaches the coast it is possible that enough time has elapsed to degrade all of these herbicides. However, a second explanation is that the methodology employed in this study was not suitable for the extraction and measurement of glyphosate and paraquat. This is supported by the lack of satisfactory results with regards to the samples and standards submitted to the contr act laboratory for analysis. As discussed previously, recovery studies were poor a nd the results for the calibration solutions submitted were significantly different from the true values. As a result, we are unable to make definitive statements regarding the potential impact of these herbicides on offshore coral reefs. Until a reliable, reproducible method is used to produce reliable results results are uncertain. Further studies are certainly necessary in this area. Table 5. Metal concentrations in coastal waters of southern Belize. Sample ID Mercury ppms Lead ppms NSC1 0.0010.28 NSC2 0.0010.18 NSC3 0.0010.05 NSC4 0.0010.02 NSC5 0.0010.01 SR1 0.001 0.24 SR2 0.001 0.13 SR30.001 0.11 SR4 0.001 0.08 SR5 0.001 0.04 SSC1 0.0010.15
75SSC2 0.0010.13 SSC30.0010.09 SSC4 0.0010.05 SSC5 0.0010.01 MC1 0.001 0.13 MBC3 0.001 0.15 MC4 0.001 0.09 MC5 0.001 0.08 MC6 0.001 0.05 MR1 0.0010.29 MR2 0.0010.19 MR3 0.0010.13 MR4 0.0010.1 MR5 0.0010.06 GS1 0.001 0.14 GS 0.001 0.14 GS3 0.001 0.14 GS4 0.001 0.11 GS5 0.001 0.09 RG1 0.0010.17 RG2 0.0010.09 RG3 0.0010.08 RG4 0.0010.07 RG5 0.0010.02 SAR1 0.001 0.14 SAR2 0.001 0.14 SAR3 0.001 0.12 SAR4 0.001 0.09 SAR5 0.001 0.12 Other Pesticides Tables 6 and 7 below summarise the resu lts for several pesticides that were detected in coastal waters of southern Beli ze. Due to analytical difficulties we were unable to analyse samples for all the pesticides targeted initially, but we were able to measure a sufficient number to draw preliminary conclusions and to compare with modeling results presented in Chapter 2 of this thesis. The first observation made from the data is that generally levels of pesticides were higher in June than in February. Fe bruary falls during the dry season in Belize
76 while June is right at the be ginning of the rainy season, which generally runs from July to November but can start earlier. A study by Burke and Sugg (2006) indicates that total riverine discharge into coastal Belize is ove r four times higher in June compared to February. Thus, one would expect increased riverine input of pes ticides in June, as observed in general in this st udy. In addition, in late May and early June at least two tropical storms affected Belize, Tropical St orms Alma and Arthur. Precipitation from these twin tropical storms cau sed the most severe flooding se en in Belize, especially southern Belize, in at least 50 years. As a result, higher concentrations of pesticides during June reflect increase d input from flooded rivers flowing through agricultural lands. The results indicate that the pesticides detected in at least one sample were cadusafos, ethoprop, acetochlor, fenamiphos, oxa myl, carbofuran, chlorpyrifos, dacthal, chlorothalonil, trifluralin, a nd malathion. Frequency of detection differed among these pesticides, as did levels. In terms of levels the first six pesticides listed above were measured in significantly higher levels (several orders of magnitude) compared to the last five. However, there was no correlation be tween levels and frequency of detection. Thus, for example, carbofuran was detected only in six samples in February and five samples in May, but the concentrations in t hose samples were 6-7 orders of magnitude higher than levels of chlorpyrifos, a pest icide detected in almost all samples. For some pesticides, namely oxamyl, cadusafos and chlorpyrifos, there is a general trend of decreasing c oncentrations from the river mouths extending offshore in each transect. This supports the hypothesis th at the sources of these pesticides are the citrus and banana farms in southern Belize. However, this trend does not hold perfectly
77 and does not hold for all pesticides. An intere sting observation is that the middle stations in most transects seem to have elevated levels of pesticides, breaking the trend of decreasing levels as one moves offshore. An examination of the current systems in coastal Belize might explain this observation. It has been re ported that in southern Belize there is a flow south along the coast until th e currents meet in the Gulf of Honduras the western flow coming from the Caribbean Sea. This creates a current moving north some miles offshore near the coral reef. It is plau sible to envision that th is creates a mixing of pollutants in the coastal area, with some c oncentration in the middle of the area, which would match our results. Due to analytical delays it was impossibl e for us to analyse the samples for all the target pesticides we chose at the beginning of the project. Th erefore, it is possible that there still other pesticides th at are being discharged into coastal waters of southern Belize. Further studies should help determin e this. The data also shows that some pesticides are indeed transported far e nough offshore that they are found in waters overlying the coral reefs. Thus, there is the potential that such pesticides may be adversely affecting coral reefs. As with meta ls, it should be a priority to measure levels of pesticides in the co ral reefs to determine if there is any accumulation. This has been noted in previous studie s (Glynn et al., 1989).
78 Table 6. Pesticide levels in coastal southern Belize (pg). TrifluralinTrifluralinChlorothalonilChlorothalonilDacthalDacthalMalathionMalathionChlorpyrifoschlorpyrifos FebMayFebMayFebMayFebMayFebMay NSC171 148 BD428891134904BD20483908 NSC2488BDBDBD78914BDBD29841409 NSC31054136297BD39651171618BD19587582 NSC482BD331BD49BDBDBD194BD NSC561109128BD209302BDBD67918017 SR110361113BD3366BDBD63950082 SR286BD119BDBD216BDBD63426452 SR310152164BD17077BDBD39419563 SR4382BD31BD730167BDBD5909538 SR5785579753385BDBD5625075 SSC151109BDBD8874BDBD92902512 SSC213661BDBDBD300551032409BD45941868 SSC39077106BD252189BDBD1077432 SSC483BD19770BD130BDBD697552 SSC5103343802323240BDBD9264234 MBC1BDBD244162187263BDBD10415409 MBC2131BD138BD35822BDBD82371228 MC34467067BDBD43156944035699BD66023415 MC4216BD200BD274BDBDBD902BD MC5286120BD281513223BDBD71713662 MC6BD110BDBDBD234BD941BD5331 MR184198BDBD576130BDBD52458014 MR225930BD1596843294000BD117802953 MR37754275BD9090BDBD32172366 MR462100428BD32128BDBD37389507 MR5BD233BDBDBDBDBDBDBD44853 RG1BD113BDBDBD102BDBDBD13431 RG230024252BD910210BDBD8182161 RG398229144BD69158BDBD2346816903 RG4219BDBDBD82557BDBD47292351 RG593 166 151222176140BDBD175244554 GS1126108BDBD28078BD 8308 212227686 GS2498832042241122449BDBD106810342 GS369110125BD2273BD12402404205 GS412350BD685520BDBDBD100368883 GS5BD75BDBDBD18BD434BD368 SAR198BD186BD49179BDBD125011821 SAR2307BD400BD783346BD35932013019865 SAR31265113358289304106BDBD148543575 SAR4228BD 297 BD63849BDBD6452278 SAR579164110BD110287BD6849862686
79 Table 7. Pesticide levels in coastal southern Belize (pg). cadusafoscadusafosethoprophosethoprophosacetochloracetochlorfenamiphosfenamiphosoxamyloxamylcarbofuran FebJuneFebJuneFebJuneFebJuneFebJuneFeb NSC17.13E+099.13E+09BDBD2.92E+075.40E+07BDBD1.44E+122.03E+12BD NSC29.78E+081.54E+09BDBD7.33E+06BDBDBD1.79E+12BDBD NSC3BDBDBDBDBDBDBDBD9.45E+111.03E+12BD NSC4BDBDBDBDBDBDBDBD4.62E+118.10E+11BD NSC5BDBDBDBDBDBDBDBDBDBDBD SR11.97E+093.08E+09BDBD1.91E+073.60E+076.67E+072.70E+082.09E+124.16E+12BD SR21.02E+092.21E+09BDBD1.01E+07BD2.07E+06BD1.83E+122.67E+12BD SR39.43E+081.03E+09BDBDBDBDBDBD1.14E+121.60E+12BD SR47.67E+088.98E+08BDBDBDBDBDBD1.86E+111.89E+11BD SR54.45E+087.74E+08BDBDBDBDBDBD1.77E+114.37E+11BD SSC15.13E+097.06E+096.73E+089.90E+082.22E+075.40E+07BDBD3.03E+119.07E+114.44E+07 SSC23.97E+095.55E+092.81E+08BD3.31E+05BDBDBD7.17E+10BDBD SSC31.23E+094.01E+095.48E+077.20E+07BDBDBDBDBDBDBD SSC48.91E+082.66E+09BDBDBDBDBDBDBDBDBD SSC53.43E+081.13E+09BDBDBDBDBDBDBDBDBD MBC1BDBDBDBDBDBDBDBDBDBDBD MBC2BDBDBDBDBDBDBDBDBDBDBD MC31.89E+094.90E+093.32E+075.40E+07BDBDBDBD8.89E+112.85E+12BD MC41.01E+091.44E+093.31E+075.40E+07BDBDBDBD1.14E+112.50E+11BD MC5BDBD5.67E+077.20E+07BDBDBDBD8.88E+101.66E+11BD MC6BDBDBDBDBDBDBDBDBDBDBD MR18.61E+091.45E+101.11E+082.34E+086.47E+071.80E+083.56E+071.31E+098.98E+111.53E+12BD MR26.67E+091.38E+101.66E+081.94E+083.06E+07BDBDBD7.11E+119.03E+12BD MR33.72E+091.22E+109.77E+071.44E+08BDBDBDBD5.55E+118.03E+11BD MR41.11E+096.56E+098.34E+079.89E+07BDBDBDBD1.22E+116.63E+11BD MR58.84E+081.56E+095.44E+077.20E+07BDBDBDBD9.87E+101.78E+11BD RG12.09E+093.06E+096.78E+071.44E+08BDBD4.32E+099.90E+084.44E+114.86E+11BD RG29.34E+089.88E+084.13E+077.87E+07BDBDBDBD4.23E+114.71E+11BD RG34.67E+088.28E+082.59E+077.20E+07BDBDBDBD2.12E+114.32E+11BD RG4BDBDBDBDBDBDBDBD9.87E+103.03E+11BD RG5BDBDBDBDBDBDBDBD7.88E+101.44E+11BD GS12.12E+092.74E+091.16E+081.62E+08BDBDBDBD8.23E+111.35E+122.22E+08 GS22.55E+092.61E+097.65E+071.40E+08BDBDBDBD6.67E+111.43E+128.81E+07 GS32.68E+092.79E+094.89E+071.26E+083.08E+083.78E+08BDBD4.43E+111.97E+123.34E+07 GS43.23E+093.43E+09BDBD2.77E+08BDBDBD1.12E+111.01E+123.37E+06 GS54.17E+095.60E+09BDBD1.11E+082.34E+08BDBD8.67E+101.98E+110 SAR1BDBD4.09E+079.00E+07BDBDBDBD1.97E+123.16E+126.35E+07 SAR2BDBDBDBDBDBDBDBD7.23E+111.99E+12BD SAR3BDBDBDBDBDBDBDBD3.33E+119.70E+11BD SAR47.76E+08BDBDBDBDBDBDBD1.15E+119.89E+11BD SAR51.22E+092.11E+09BDBDBDBDBDBD7.21E+101.32E+12BD
80 Comparison of Empirical Data and Modeling Results Table 8 below ranks the pesticides that c ould be measured in this study in terms of their fractions predicted by the model to partition into water, their actual concentrations measured, and by their predicted overall half-lives. Table 8. Comparison of empi rical and modeling results. Pesticide Rank by fraction predicted in water Rank by concentrations measured in water Rank by overall halflife (1=shortest) Glyphosate Paraquat Oxamyl 1 1 5 Malathion 2 8 1 Carbofuran 3 3 4 Acetochlor 4 4 9 Fenamiphos 5 6 8 Ethoprop 6 5 6 Cadusafos 7 2 7 Chlorothalonil 8 10 2 Dachthal 9 9 3 Chlorpyrifos 10 7 10 The Level III fugacity model employed to predict which pesticides are most likely to partition into water correla tes quite closely with empiri cal data. Assuming that the more a pesticide partitions into water the more likely it will be susceptible to runoff into nearby streams and into coastal waters, th en the ranking done by the modeling and the empirical levels should correlate. This is ind eed the case in general. Oxamyl is predicted to be the pesticide that partitions to the grea test extent into water and it is the pesticide measured in highest concentrations. Carbofur an is predicted to be third in partitioning into water and its measured concentrations were on average third hi ghest. Acetochlor
81 was predicted fourth, and measured four th. Fenamiphos was predicted fifth and measured sixth. Ethoprop was predicted sixt h and measured fifth. Chlorothalonil was predicted eighth and measured tenth. Dactha l was predicted ninth and measured ninth. Chlorpyrifos was predicted tenth and measur ed seventh. Only two pesticides did not seem to correlate well, malathion and cadus fafos. Malathion was predicted second and measured eighth and cadusafos was predicted seventh but measured second. Half-lives may help explain these anomalies between model and empirical data, especially for malathion. Even though it is predicted to parti tion heavily into water it is also predicted to have the shortest half-life. Thus, even if it partitions into water it may be degraded very quickly so that it is not found in high levels in coas tal waters. In the case of cadusafos, its high concentrations may be due to usage patterns (that is, it may simply be that our sampling campaign coincided with th e application time of this pesticide). Conclusions We have successfully carried out the firs t, to our knowledge, su rvey of levels of pesticides and metals in coastal waters of southern Belize. Our results indicate that there are some pesticides used in the banana and citrus industries that can be measured in coastal waters, and in some cases all the way out to waters overlyi ng the barrier reef. Thus, there is the need for further studies to determine if any pesticides or metals are causing adverse effects on coral reefs. This would help in determining if alternative pesticide usage patterns are necessary in the ba nana and citrus industries. In the interim, it is recommended that care be exercised in th e use of those pesticid es measured in very
82 high levels in this study (for example, oxamyl, cadusafos, chlorpyrifos, dacthal, chlorothalonil). Empirical results correlate very well with those predicted by a Level III fugacity model applied to the study region. This indicat es that this model ma ybe applied to other areas in Belize and probabl y other tropical areas.
83 Chapter Four Analysis of Pesticide Legislation in Belize Background Environmental management authority in Belize falls under several government ministries, quasi-governmental authorities and NGO institutions. Although the portfolio for environment currently rests with the Mini stry of Environment, major responsibilities are held by the Ministries of Health (water and sanitation), Agriculture and Fisheries (fisheries, coastal zone management, pesticid es control), and Natura l Resources (forestry, national parks and protected areas, wildlif e, water resource management, land use planning). Two quasi-government entities wi th specific responsibilities include Solid Waste Management Authority and Land Utili zation Authority and several conservation NGOs managing natural resources on beha lf of the Government of Belize. This proliferation of institutions with e nvironmental responsibilities gives rise to a number of implementation problems. Fo r instance the current basis for resource allocation among institutions is not clear, and financial resources and technical capacity remain a problem. The primary institutions involved in environmental protection and natural resources management may be conveniently classified into environmental protection institutions (DOE, PCB and Public Health Bureau); resource management institutions (Forest Department, Fisheries De partment); and land use planning institutions (Land Utilization Authority, Central Building Authority).
84 BelizeÂ’s Legislative Fram ework on Pesticides In Belize, the legislativ e jurisdictions unde r which chemicals are managed lie within control of four princi pal organizations. These are: (i) Pesticides Control Board (PCB), (ii) Department of the Environmen t (DOE), (iii) Public Health Bureau (PHB)/Ministry of Health and, (iv) Belize Ag riculture Health Authority (BAHA). These organizations are all government agencies. Ho wever, the main legal instruments used for controlling the use of chemicals, and speci fically the pesticide chemicals, are the Pesticide Control Act (PCA) and the Environmen tal Protection Act (EPA). Both of these Acts would fall under the broad legal fram ework of Environmental Legislation. The Pesticide Control Act (PCA) The Pesticides Control Act (PCA) of 1985 (Chapter 216, Revised Edition 2000) provides authority to control the manufactur e, importation, sale, storage and use of pesticides. The Pesticides Control Board (P CB) established by the Act carries out this function. To date prohibited, restricted and re gistered regulations have been completed. The Act itself identifies specific pesticides that are prohibited, restri cted or registered. Registered pesticides are those pesticides th at have been approved by the PCB for use in Belize. Sections 6 and 7 of the Act provide that the Board may grant a license for the manufacture or importation of any pesticides. Section 2 of the PCA provides for a Â“restricted use pesticide.Â” This refers to a pesticide which, if used in accordance with a widespread and commonly rec ognized practice, may genera lly cause, without additional regulatory action, unreasonable adverse eff ects on the environment, including the applicator and other people.
85 Several regulations have also been made for the proper use of registered pesticides. These regulations have been c onsolidated in the Pesticides Control Act, Chapter 216, R.E. 2003, showing the Subsidia ry Laws as at October 2003. These regulations include: (a) Regist ered and Restricted Pesticid es (manufacture, import and sale) Regulations, S.I 8 of 1989; (b) Registered and Restricted Pes ticides (registration) Regulations, S.I. 77 of 1995; (c). Registered and Restricted Pesticides (manufacture, import and sale) Amendment Regulations, S.I. 30 of 1996; (d) Registered and Restricted Pesticides (certified user) Regulations, S. I. 112 0f 1996 and (e) Pesticide Control (sale and confiscation) Regula tions, S.I. 71 of 1998 These regulations provide the legal regiments for registration, labeling, importation, sale and use of pesticides. Additionally, regulations were recently enacted in order to expand the legal requirem ents of the PCA. These include the Restricted Pesticides (Certified User) Regulations, and the Pesticides Control (Sale and Confiscation) Regulations. These regulations demonstrate some concerns for the health and well being of workers. For example, S.I. no. 112 of 1996 ( Restricted Pesticides (Certified User) Regulations) required that formal training of farmers, applicators and retailers be conducted on safe and efficient pesticide manageme nt. The enactment of S.I. No. 112, of 1996, required the pesticide user to pass a written or oral exam. Schedule III of this legislation required that the trainee comply with a num ber of stipulations including the ability to read and understand labels, sa fely and adequately prepare mixtures of pesticides, the proper calibration and use of e quipment, among others Additionally, with the enactment of S.I. No. 71 of 1998, Pesticides Control (Sale and Confiscation) Regulations the PCB was granted further powers to enforce S.I. no. 112 of 1996. This
86 S.I. required that establishments maintain a re gister of sales of rest ricted pesticides, and may only sell Restricted Use Pesticides (RUP Â’s) to persons in possession of a certified userÂ’s license. An application for registration of a pestic ide should be submitted to the PCB prior to importation and should be accompanied by chemical, toxicological and environmental impact data. Any person wishing to register any pesticide must also submit details of the labels of packaging. This application is analyzed and a recommendation is made by the PCBÂ’s Registration Committee and if approved, the conditions for such importation are detailed. This committee has for some time now been chaired by the Department of the Environment. In terms of Â“Prohibited PesticidesÂ”, th e PCA defines these as any pesticide of which the possible effects on the environmen t, plant, animal or human being are considered by the Minister to be too dangerous to justify its use. Section 13 (1) provides that pesticides not registered as required in section 6 or listed unde r section 8, should be prohibited pesticides and accordingly shall not be brought into or used in Belize. Section 6 (1) provides that any person may, subject to the provisions of this Act, manufacture, import, advertise or sell a pesticide which is decl ared to be a registered pesticide. Section 6 (2) provides a list of registered pesticides which is given in the Second Schedule to this Act, and this schedule may be amended or re placed from time to time by order made by the Minister, in consultation with the Board. Section 8 (1) provides that no person shall sell a rest ricted pesticide unless, (a) he is authorized in the prescribed manner to do so; (b) the premises in which the sale is carried out have been registered in the pres cribed manner for the purpose; and (c) the sale
87 is carried out in accordance with such other requirements as may be prescribed. Section 8 (2) provides another list of restricted pesticides and is given in the Third Schedule to this Act. This Schedule may be amended or replaced from time to time by order, by the Minister in consultation with the Board. The Act provides sufficient authority for enforcement if its provisions, authorizing a pe nalty for violation of the Act or regulations of a fine not exceeding five thousand dollars or to imprisonment for a term not exceeding five years or to both such fine and imprisonment. The Environmental Protection Act (EPA) Until the enactment of the Environmental Protection Act (EPA) in 1992, (Chapter 328, Revised Edition 2000), Belize had no co mprehensive environmental protection legislation. Authority to prev ent and control environmenta l pollution was contained in provisions of the Public Health Act, the Pesticides Control Act and the older Dumping at Sea, and Water and Sewerage Acts These Acts, however, were never effectively used to provide environmental protection, as the necessary supporting regulations, such as establishing environmental quality criteria and pollution control standards, were not established. The EPA makes provisions for guiding the rational use of natural resources, for controlling environmental po llution, and for overall protec tion of the environment. Part I of the Act deals with preliminary matters of interpretation. It defines key terms including: Â“EnvironmentÂ”, Â“Environmental PollutionÂ” Â“Environmental PollutantÂ”, Â“Hazardous SubstanceÂ”, Â“WasteÂ”
88 Part II of the Act legally established the Department of Environmental (DOE), giving it the responsibility and authority to enforce the Act an d its subsidiary legislation. Part III of the Act provides for the preven tion and control of environmental pollution and for the making of regulations. Unde r this Part and in section II (1) it is stated that: Â“No persons shall emit, import, discharge, deposit, dispose of or dump any waste that might directly or indirectly pollute water reso urces or damage or destroy marine life.Â” Part IV provides for the prohibition on Dumpin g. In particular section 13(1) states that: Â“No person shall dump or dispose of or deposit any garbage, refuse, toxic substances or hazardous wastes in any place th at may directly or indirectly damage or destroy flora or fauna, or pollute wa ter resources or the environment.Â” Part V provides for Environmental Impact Assessment (EIA) and regulations made there under. Section 20 (1) states:Â“Any person intending to under take any project, programme or activity which may significantly affect the environment shall cause an environmental impact assessment to be carried out by a suitably qualified person, and shall submit the same to the Depa rtment of Environment for evaluation and recommendations.Â” Part V Section 20 of the Pollution Regulations ad dresses the processing, storage, use and transportation of organic solvents and other volatile compounds. Through the Pollution Regulations, SI # 56 of 1996, mechanis ms have been developed to establish the prohibition of industries operating and emitti ng contaminants into the environment, without a permit from the DOE (Regulation 4). These include emissions into the air from (1) industry (Regulation 4), (2) power generating installa tions, (Regulation 8), (3)
89 burning of refuse in urban areas (Regulati on 13), (4) processing i ndustries (Regulation 15), (5) gasoline or petroleum storage unit (R egulation 18), (6) storage containers for solvents, pesticides and other volatile compounds (Regulation 21), and (7) combustion engines such as motor vehicles (Regulation 25); and provide DOE with powers to require owners, occupiers and other agents to cl ean-up (Regulation 50) and abate pollution (Regulation 51). In order to encourage voluntary compliance, DOE is empowered to develop an environmental incentive programme (Regulation 58 (1)), as well as a "facility environmental audit programme" as a compre hensive investigation and evaluation system designed for detecting and preventing viola tions of environmental requirements or the commission of pollution related offences (Regulation 58 (3)). Governmental Policy on and Ma nagement of Pesticides Authority for environmental management as a whole in Belize is shared by a number of ministries, departments, quasigovernmental authoritie s and non-governmental institutions. The portfolio for environm ent now rests with the Ministry of the Environment (environmental planning, po llution), but major environmental responsibilities are al so held by the Ministries of Health (water and sanitation), Agriculture and Fisheries (fis heries, coastal zone manageme nt, pesticides control), and Natural Resources (forestry, national parks a nd protected areas, wild life, water resource management, land use planning). Quasi-governme ntal statutory authorities, such as the Solid Waste Management Authority and the Land Utilization Authority, were created to carry out specific responsibi lities. Finally, several nongovernmental organizations
90 (NGOs) are responsible for cer tain natural resources mana gement functions for the Government of Belize. This proliferation of institutions with e nvironmental responsibilities gives rise to a number of implementation problems. For exam ple, the current lack of clearly delineated roles and responsibilities results in unnece ssary confusion and wasteful duplication of effort. Inter-ministerial coordination on environmental matters has not been fully institutionalized and needs ur gent improvement. The current basis for resource allocation among institutions is not clear, and financia l resources and technica l capacity remain a problem. The primary institutions involved in environmental protection and natural resources management may be convenient ly classified into three categories: environmental protection institut ions (Department of the Envi ronment, Pesticides Control Board, and Public Health Bureau); resource management institutions (Forest Department, Fisheries Department, Office of Geology and Petroleum, National Hydrological Service, and Land and Surveys Department); and land use planning institutions (Land Utilization Authority, Central Building Authority, North Ambergris Caye Development Corporation and the various Town Boards that ha ve been assigned planning functions). It should be noted that there are a numb er of overlapping responsibilities between the various institutions and this problem has been compounded by the variety of statutory provisions. Part of the problem of overlapp ing responsibilities stems from the fact that some institutions (e.g. Forestry and Fisher ies Departments) are comparatively old institutions while some (e.g. the Department of the Environment) are relatively new.
91 Another reason is that the various Department s and agencies have somewhat shifted their roles either to meet perceived needs or to deal with the realities of international funding. Rationalizing Legislative Amendments to Address Chemicals in Belize From an assessment of the above legisl ation, it is apparent that Belize has no comprehensive chemical management legislati on. As reflected above there are in fact two major legislations that significantly govern environmental Â“pollutionÂ” management. These are the Pesticides Control Act (PCA ), and the Environmental Protection Act (EPA). The PCB Act governs agricultural chem icals including the POPs pesticides; and the Environmental Protection Act which, with it s broad mandate to control, among other things, the volume, types, constituents and e ffects of waste, emission and discharges into the environment. The EPA can effectiv ely control and minimize the trans-boundary movement of toxic and hazardous waste in cluding the movement of PCBs and the unintentional releases of dioxins and furans into the environment. None of these Acts, however, make any reference to the control of exports of chem icals. Indeed, none of the Acts, or any other legislation address any of the following: Processes or procedures adopt ed to safely transport or export banned or obsolete chemicals; Environmental standards for dioxins or furans; Standards or procedures governing the bur ning of agricultural and other waste that would result in managing or reduci ng the release of chemical pollutants; Procedures for managing medical, pharmaceutic al, veterinary or other biomedical waste;
92 Provide capabilities to monitor source em issions of environmental and chemical pollutants; Determine acceptable standards of ambient chemical pollutants in air, soil, or water samples; and most importantly Provide responsibilities a nd capabilities to monitor the fate or impacts to pesticides being used in Belize upon its environment. In an effort to address these issues th e Department of the Environment recently developed new environmenta l legislations that have been submitted to BelizeÂ’s National Assembly for their approval. The new legislations include a Hazardous Waste Management Regulation, the Transporte rs of Hazardous Waste Regulations and, amendments to existing legislation such as Solid Waste Management Authority Amendment Act, Environmental Impact A ssessment (Amendment) Regulations and the Effluent Limitations (Amendment ) Regulations. It is antici pated that these regulations will become the formal basis for many of the global issues aimed at environmental protection. The Department has also recen tly initiated the development of BelizeÂ’s National Plan of Action (NPA) to address Land Based Sources of Marine Pollution. Prioritization of the major environmental issu es has been done and a report has been prepared for cabinetÂ’s endorsement. While there is recognition of the valu e of assessing and re-evaluating the relevance of the legislation, as it relates to the management of chemicals, the outcome of previous initiatives are less than stellar. The initiatives bo ldly unfolded by the
93 Department of the Environment are yet to ma terialize as a result of constraints resulting from lack of funding and other resources. For these reasons, the assertion is that clear outcomes will result only when an unambiguous strategy is develope d to chart a pathway that w ill lead to legal instruments that will fill all the regulatory gaps and inc onsistencies defined; and that will provide key government agencies with the power necessary to require Â– when needed Â– information from producers, importers, distributors, industr ial consumers of chemicals and generators of waste. BelizeÂ’s Institutional Framework to Address Pesticides Pesticides Control Board The Pesticides Control Board, establis hed in 1985 under the Pesticides Control Act, (Ch. 216 of the Rev. Laws of Belize 2000) is responsible for the management of the use, importation, manufacturing and sale of Pes ticides. The functions of the Board are to control imports or manufacturing of pest icides through a regi stration and licensing process, that also authorizes sale of these restricted pesticides and the process to follow (Registered & Restricted Pes ticides (Manufacturing, Import & Sale Regulations, SI # 8 of 1988) and as amended in 1996 through SI #30 of 1996; as well as Pesticide Control (Registration) Regulations, SI # 77 of 1995; to register premises in which a restricted pesticide may be sold establishing the conditi ons that these establishments must comply with, as well as the cases where confiscation of pesticides are possi ble (Pesticide Control (Sale & Confiscation) Regulati ons, SI # 71 of 1998); to author ize pesticide applicators to use restricted pesticides (Pesticides Contro l (Certified Users) Regulations, SI # 112 of
94 1996); to classify any pesticid e as a registered pesticide, restricted pesticide or a prohibited pesticide (Pesticides Control (Replacement Orde r), SI # 72 of 1988); and to deal with all aspects of the importation, ma nufacture, packaging, preparation for sale, sale, disposal and use of pesticides (Ch. 216, Rev. Laws of Belize, 2000). The Schedules to the Act contain a list of Re gistered, Restricted and Prohibi ted Pesticides and these have been recently revised by the Pesticides C ontrol (Replacement of Schedules) Order, 1995 (S.I. # 100 of 1995). Regulations made under th e Pesticides Control Act outline the mechanism for registering pesticides. Regulatio ns made under the Act indicate that in the application for the registration of pesticides applicants must complete a section that provides ecological information on the particular pesticide. This includes information on the persistence and al terations of the product in soil, water and air; the effects in terrestrial flora and fauna; the effects in aqua tic flora and fauna and data on toxicity to beneficial insects; and data on mobility of the pesticide in the ecosystem. Information must also be included on the possible toxic eff ect of the pesticide oil human beings (S.I. # 77 of 1995, Schedule 1). The Act makes provision for the appointment of a qualified technical person as an inspector who will have the power to enter any land or premises registered for the sale or manufacture of pesticide in order to carry out his functions which include inspecting, copying and examining relevant records or other documents; making examinations or inquiries and seizing and detain ing any article which is relate d to the contra vention of the Act (Sec. 17, Ch. 216, Rev. Laws of Belize, 2000). The Pesticides Control (Amendment) Act on 2002, redefines Â“PesticidesÂ” to incorporate and differentiate between restricted and prohibi ted pesticides; it grants the
95 Pesticides Control Board the power to establis h Committees to assist in detailed technical works required by the Board; it repealed the pos t of Secretary in the PCB Secretariat and converts it into the Â“Registr arÂ” of the Secretariat; it in creased the Penalties for contravention of this Act; and it amends the First Schedule of the main Act (of 1985). The Pesticides Control Board commi ssioned the Amendment of the Registration Regulations of 1995 (SI # 77 of 1995) in early 2003. This amendment to the Regulations was basically to differentiate between the various categories of registration of pesticides, with their corresponding fees; to provide for the BoardÂ’s ability to gran t Â“provisionalÂ” registration to a pesticide for one year ma ximum for experimental purposes; for the repealing of schedules I and II of the 1995 Regulations for implementation of these amendments; and also to provide for the Minist er of Agriculture to be granted with the power to Â“import a pesticide not registered in Belize, only during Emergency occasions.Â” Conclusions and Recommendations This studyÂ’s results indicate that the banana industry may be doing a better job with regard to being sensitive to environmental impacts of its pesticide activities compared to the citrus industry. The survey indicated clearly a grea ter knowledge of and interest in using environmen tally-friendly methods. Consid ering that a significant of citrus farms are owned by owners of banana farms, it is possible that the difference in attention to environmental issues is due to a greater international pressure from the European Union on the banana industry.
96 Despite differences in the levels of inte rest and knowledge, results indicate that overall there exists a great degree of knowle dge on environmental matters and interest on receiving training on such matters. An analysis of BelizeÂ’s attempts to properl y manage pesticides, years of strategies and media campaign focusing on Â‘safe pesticide managementÂ’ have failed to bring about improvements. Thus, the most effective wa y to reduce human and environmental health risks in developing countries lik e Belize is to reduce pesticid e use and/or switch to more environmentallyand health-friendly. This study clearly indicate s that the PCB is not activ ely engaged in stimulating interest in environmentally-friendly and other alternative pesticide management practices. There is also little on the way of training programmes for farmers. It is recommended as a first step to be tter manage chemicals, including pesticides, in Belize that the appropriate agency (i n our view DOE) carry out a comprehensive legislative study that would identify the st rengths and gaps among existing legislation, especially the PCA and the EPA. The result s of this study should lead to legislation encompassing comprehensive chemical manage ment. It should incorporate control of persistent organic pollutants (POP) and othe r hazardous chemicals and it should detail a monitoring plan for the fate of these in the environment. Also, since there is no one entity res ponsible for the management for chemical management in Belize, there is a need fo r the establishment of a single entity or established interagency with clear responsibil ities for the management of all aspects of chemicals. This entity or interagency m echanism needs to be provided with adequate human resources, material and equipment a nd financial support. The human resource
97 should in this entity should be have the form al necessary skills to perform all aspects of chemical management from enforcement, m onitoring and research, and translation of research findings to programmes, policies and finally to le gislation. There is an urgent need to enforce ex isting legislation on pe sticide use and its effects on water quality, human healt h, flora and fauna. Obtaining data on pesticides being used in Belize is challenging and to a large extend, what is made available can have huge discrepancies. It is therefore imperative that the regulatory body for chemical mana gement establish an accessible organized system of pesticides. It is definitely recommend that the Pesticide Control Board become active in setting up workshops and other training programmes for farmers to demonstrate alternative and more environm entally-friendly pesticide mana gement practices. Certainly the results indicate a hi gh degree of interest in further training on the part of farmers.
98 References Alegria, H.A.; dÂ’Autel, J.P.; Shaw, T.J. Â“O ffshore transport of pesticides in the South Atlantic Bight: Preliminary es timates of export budgets,Â” Mar. Pollut. Bull 2000a 40, 1178. Alegria, H.A.; Bidleman, T.F.; Shaw, T.J. Â“Ambient air levels of organochlorine pesticides in Belize, Central America,Â” Environ. Sci. Technol. 2000b 34, 1953. Alegria, H.A. and Shaw, T.J. Â“Rain deposition of pesticides in coastal waters of the South Atlantic Bight,Â” Environ. Sci. Technol 1999 33, 850. Baker, J. M.; Koskinen, W. C.; Dowdy, R. H. J. Environ. Qual. 1996 25, 169-177. Boethling, R.S.; Mackay, D., Handbook of property estimati on methods for chemicals: Environmental and health sciences CRC Press LLC: Boca Raton, FL, 2000; p 481. Burke, L.; Greenhalgh, S.; Prager, D.; Cooper, E. Â“Coastal capital Economic valuation of coral reefs in Tobago and St. Lucia.Â” Final Report, June 2008, World Resources Institute, Washington DC, 67 pp. Available online at http://www.wri.org/publications Carsel, R. F.; Mulkey, L. A.; Lorber, M. N.; Baskin, L. B. Ecol. Modeling 1985 30 49 69. Chen C.; Green, R. E.; Thomas, D. M.; Knuteson, J. A. Environ. Sci. Technol. 1995 29, 1818-1821. Cooper, E.; Burke, L.; Bood, N. Â„Coastal cap ital Â– Belize: The economic contribution of BelizeÂ’s coral reefs and mangroves.Â“ WRI Wo rking Paper, World Re sources Institute, Washington, DC, 53pp. Dietrich, A.M. and Gallagher, D. L. J. Agric. Food Chem 2002 50, 4409-4416. Eisler, G.; Hellweg, S.; Liech ti, S.; Hungerbuhler, K. Environ. Sci. Technol. 2004 38, 4457-4464. Glynn, P.W.; Szmant, A.M.; Corcora n, E.F.; Cofer-Shabica, S.V. Mar. Pollut. Bull 1989 20, 568-576. Gouin, T.; Mackay, D.; Webster, E.; Wania, F., Screening chemicals for persistence in the environment. Environ. Sci. Technol. 2000 34, 881-884.
99 Guo, L.; Nordmark, C.E.; Spurlock, F.C.; J ohnson, B.R.; Li, L.Y.; Marshall Lee, J.; Goh, K.S. Environ. Sci. Technol. 2004 38, 3842-3852 Hapeman, C.J.; Dionigi, C.P.; Zi mba, P.V.; McConnell, L.L. J. Agric. Food Chem 2002 50, 4382-4384. Hemond, H.F.; Fechner, E.J., Chemical fate and transport in the environment Academic Press: Toronto, ON, Canada, 1994. Howard, P.H., Handbook of environmental fate and exposure data for organic chemicals, volumes I-III Lewis: Chelsea, MI, USA, 1990. Hou, A.; Laws, E.A.; Gambrell, R. P.; Bae, H.-S.; Jan, M.; Delaune, R.D.; Li, Y.; Roberts, H. Environ. Sci. Technol 2006 40, 5904-5910. Jeong, Y.; Sanders, B.F.; McLaughlin, K.; Grant, S.B. Environ. Sci. Technol. 2008 42, 3609-3614. Jury, W.A.; Russo, D.; Streile, G.; Abd, H.E. Water Resour. Res. 1990 26, 13-20. Knisel, W. G CREAMS: A Fiel d Scale Model for Chemicals, Runoff, and Erosion from Agricultural Management Systems; Conserva tion Research Report 26: U.S. Department of Agriculture: Washington, DC, 1980. Leonard, R.A. Movement of pesticides into surface waters. In Pesticides in the Soil Environment: Processes, Impacts, and Modeling; Cheng, H.H., Ed.; Soil Science Society of America: Madison, WI, 1990; pp 303-349. Leonard, R. A.; Knisel, W. G.; Still, D. A. Trans. ASAE 1987 30, 1403-1428. Lyman, W.J., Handbook of chemical property estimation methods 2nd ed.; American Chemical Society: Washington, D.C., USA, 1990. Mackay, D., Multimedia environmental models: The fugacity approach Second Edition ed.; CRC Press LLC: Boca Raton, FL, 2001; p 261. Mackay, D.; Di Guardo, A.; Paterson, S.; Cowa n, C.E., Evaluating the environmental fate of a variety of types of chemicals using the EQC model. Environ. Toxicol. Chem. 1996 15, 1627-1637. Mackay, D.; Shiu, W.; Ma, K.-C.; Lee, S.C., Physical-chemical properties and environmental fate for organic chemicals 2nd ed.; CRC Press: Boca Raton, FL, 2006. Meylan, W., EPIWIN versio n 4.0(Computer programme). 1999 available at http://www.epa.gov/opptintr/e xposure/pubs/episuitedl.htm
100 Qiu, X.; Zhu, T.; Li, J.; Pan, H. ; Li, Q.; Miao, G.; Gong, J. Environ. Sci. Technol 2004 38, 1368-1374. Ramanayaranan, T.; Narasimhan, B.; Srinivasan, R. J. Agric. Food Chem 2005, 53, 8848-8858. Reichman, R.; Wallach, R.; Mahrer, Y. Environ. Sci. Technol 2000 34, 1313-1320. Scott, G.I.; Fulton, M.H.; Wirth, E.F.; Ch andler, G.T.; Key, P. B.; Daugomah, J.W.; Bearden, D.; Chung, K.W.; Strozier, E.D.; De Lorenzo, M.; Siverstsen, S.; Dias, A.; Sanders, M.; Macauley, J.M.; Goodman, L.R. ; LaCroix, M.W.; Thayer, G.W.; Kucklick, J. J. Agric. Food Chem. 2002, 50, 4400-4408. Senseman, S.A.; Lavy, T.L.; Mattice, J.D.; Gbur, E.E.; Skulman, B.W. Environ. Sci. Technol. 1997 31, 395-401. Southwick, L.M.; Grigg, B.C.; Fouss, J.L.; Kornecki, T.S. J. Agric. Food Chem. 2003, 51 5355-5361. Verro, R.; Calliera, M.; Maffioli, G.; Aute ri, D.; Sala, S.; Finizio, A.; Vighi, M. Environ. Sci. Technol. 2002 36, 1532-1538. Wang, D.; Yates, S.R.; Jury, W.A. J. Environ. Qual. 1998 27, 821-827. Wauchope, R.D. J. Environ. Qual. 1978, 7, 459-472. Woodrow, J.E.; Seiber, J.N.; Baker, L.W. Environ. Sci. Technol. 1997 31, 523-529.