CHAPTER 2 Characteristics of Studies Reviewed The studies that are reviewed in this book have investigated pesticide occurrence in bed sediment or in one or more species of aquatic biota. The emphasis is on studies that investigated pesticides in bed sediment or aquatic biota in rivers and streams in the United States. The publications reviewed are summarized in Tables 2.1, 2.2, and 2.3 (located at the end of the book), respectively, according to three main categories: national and multistate monitoring studies, state and local monitoring studies, and process and matrix distribution studies. Within each of Tables 2.1, 2.2, and 2.3, individual studies are listed by publication date, and then alphabetically. Review papers (Table 1.1), laboratory studies, and selected papers on pesticides in lakes or in foreign countries are cited in the text as needed, but they are not included in Tables 2.1, 2.2, and 2.3. National and multistate monitoring studies (Table 2.1) are occurrence surveys for one or more classes of pesticides at several (or many) sites in multiple states. Table 2.1 includes published national and multistate studies of United States estuaries, as well as rivers. That is because these studies individually contribute a lot to our understanding of the national distribution of pesticide contaminants in sediment and aquatic biota. Results from some national programs were reported in a series of publications, either sequential reports that each present 1–4 years of national data (such as reports from the Fish and Wildlife Service’s [FWS] National Contaminant Biomonitoring Program [NCBP]), or regional reports that divided the national data set geographically (such as reports from the National Oceanic and Atmospheric Administration’s [NOAA] National Benthic Surveillance Project). For programs with multiple publications, Table 2.1 contains a data summary for each individual report published, and these report summaries are grouped by program. Under each program title, individual publications are ordered first by publication date, then by author. For some national programs, more than one federal agency participated in sediment or biota data collection; in these cases, both agencies are named in the program title in Table 2.1, for example, the U.S. Geological Survey’s (USGS)–U.S. Environmental Protection Agency’s (USEPA) Pesticide Monitoring Network. To the extent possible, each report in Table 2.1 represents a unique data set. Where sequential papers were published from a given program, later papers sometimes included a summary of data from earlier papers; in these cases, only the new (previously unpublished) data are summarized in Table 2.1. Occasionally, two publications contained identical data sets; in these cases, the two publications are listed together under the most recent publication date and authors. The location of sampling © 1999 by CRC Press LLC Pesticide Monitoring Network, 1975–1980 (USGS–USEPA) National Status and Trends Program, 1984–1989 (NOAA) EXPLANATION 0 500 MILES 0 500 KILOMETERS sites from the major national programs for bed sediment and aquatic biota are presented in Figures 2.1 and 2.2, respectively. State and local monitoring studies (Table 2.2) are occurrence and distribution surveys for specific compounds or compound classes, usually at multiple sites within a specific area. They include publications from statewide monitoring programs, one-time reconnaissance (screening) surveys for specific areas or basins, and longer-term monitoring efforts at a limited number of sites. Single basin studies with sites in more than one state were considered local (rather than regional) studies, and are included in Table 2.2. The only state and local studies reviewed were those that sampled United States rivers, although some of these studies included estuarine, coastal, or lake sites in addition to river sites. State- and local-scale studies of United States estuaries and coastal areas are not included in Table 2.2, unless these studies also included some riverine sites, because of the sheer volume of such studies and the focus of this book on United States rivers and streams. For studies that had riverine sites as well as estuarine or coastal sites, data for all sites are summarized in Table 2.2 (not just data for river sites). The geographic distribution of the state and local monitoring studies reviewed is shown in Figure 2.3. Process and matrix distribution studies (Table 2.3) generally measured the distribution of one or more pesticides among various hydrologic compartments, including bed sediment and Figure 2.1. Site locations for major national programs that sampled bed sediment. Abbreviations: NOAA, National Oceanic and Atmospheric Administration; USEPA, U.S. Environmental Protection Agency; USGS, U.S. Geological Survey © 1999 by CRC Press LLC National Contaminant Biomonitoring Program, 1967–1986 (FWS) National Status and Trends Program (Mussel Watch), 1984–1986 (NOAA) National Pesticide Monitoring Program, 1965–1972 (BofCF–USEPA) National Study of Chemical Residues in Fish, 1986–1987 (USEPA) 0 500 MILES 0 500 KILOMETERS EXPLANATION aquatic biota, to investigate environmental fate processes or to determine dissipation rates. These studies tended to fall into two categories: (1) dissipation studies in which a specified quantity of a given pesticide was applied to a treatment area and monitored for its movement away from the site of application or into various environmental compartments, and (2) transplantation studies in which nonresident biota or semipermeable membrane devices (SPMD) were deployed in a study area to measure pesticide bioconcentration. Most field experiments investigated the environmental fate and persistence of a specific pesticide, and involved relatively specialized sampling at one or several sites for several days, weeks, or months. The geographic distribution of the process and matrix distribution studies reviewed is shown in Figure 2.4. 2.1 GENERAL DESIGN FEATURES Table 2.4 summarizes selected characteristics of the studies reviewed. Several national or multistate monitoring studies generated multiple reports (such as the NCBP and the NOAA’s Figure 2.2. Site locations for major national programs that sampled aquatic biota. Abbreviations: BofCF, Bureau of Commercial Fisheries; FWS, Fish and Wildlife Service; NOAA, National Oceanic and Atmospheric Administration; USEPA, U.S. Environmental Protection Agency. © 1999 by CRC Press LLC National Status and Trends [NS&T] Program for Marine Environmental Quality). Therefore, although 49 reports from national and multistate monitoring studies are listed in Table 2.1, these reports represent only 25 separate studies. In computing study characteristics for Table 2.4, each national and multistate study was counted only once. There were far more state and local monitoring studies (318) than process and matrix distribution studies (47) or national and multistate monitoring studies (25). National and multistate monitoring studies had predictably more sites and samples than either state and local monitoring studies or process studies (as shown in Table 2.4). However, the maximum numbers of sites for state and local studies (229 and 303 sites for sediment and biota, respectively) were almost as great as those for national and multistate studies (285 and 388 sites for sediment and biota, respectively), indicating that some state and local studies collected samples at a large number of sites. Note that the state and local study category contains some statewide studies, some of which covered a large geographic area (such as California). The study duration also was longer for national and multistate studies than it was for the other study types. State and local studies and process studies ranged in duration from 1 month to about 11 years, and had a median duration of just under 1 year. National and multistate studies ranged from 3 months to 24 years, with the Food and Drug Administration’s (FDA) National Monitoring 0 0 600 KILOMETERS 400 MILES 16 16 14 5 24 13 10 1 1 1 8 8 7 7 10 8 8 16 2 2 28 24 38 5 16 15 3 3 3 4 4 4 3 3 3 3 12 5 5 4 4 1 1 4 4 4 2 6 Figure 2.3. Geographic distribution of state and local monitoring studies that were reviewed. These studies are listed in Table 2.2. The number in each state refers to the number of reports reviewed from state and local monitoring studies conducted within the state. No number in a state indicates that no reports of this type from this state were reviewed. Ontario, Canada, had 24 reviewed studies. © 1999 by CRC Press LLC Program for Food and Feed (NMPFF) responsible for the maximum value, and with a median duration of 2 years. Study design ranged from monitoring a single pesticide at a single site to national studies of multiple pesticide classes. There was little consistency in site selection strategy, sediment collection methods, species of organisms sampled, tissue type analyzed, or analytical detection limits. In contrast, there was great consistency in target analytes: the majority of all three types of studies focused on organochlorine insecticides (as shown in Table 2.4). This was particularly true of monitoring studies (from national to local in scale) as opposed to process studies, regardless of the decade of sampling, as will be discussed in more detail in the next section. Even among studies that measured organochlorine insecticides, there were considerable differences in analytical methods (such as use of packed-column versus capillary column gas chromatography, or quantitation of technical chlordane versus individual chlordane and nonachlor isomers), analytical detection limits, and data reporting and normalization. This is particularly important when comparing older studies to more recent ones, since improvements in analytical methodologies over time have tended to remove interferences, improve precision, and decrease method detection limits. A few studies did not report some characteristics of study design, such as analytes that were targeted but not detected, analytical method detection limits, species or tissue type sampled, number of sampling sites or samples, sampling site location, or even sampling dates. 4 4 6 1 1 1 1 1 4 1 1 1 2 8 2 2 2 2 5 5 3 3 1 1 0 0 600 KILOMETERS 400 MILES Figure 2.4. Geographic distribution of process and matrix distribution studies that were reviewed. These studies are listed in Table 2.3. The number in each state refers to the number of reports reviewed from process and matrix distribution studies conducted within the state. No number in a state indicates that no reports of this type from this state were reviewed. Ontario, Canada, had four reviewed studies. © 1999 by CRC Press LLC Study Characteristics Study Type National and Multistate Monitoring Studies (from Table 2.1) State and Local Monitoring Studies (from Table 2.2) Process and Matrix Distribution Studies (from Table 2.3) Number of studies: 25 (49 reports) 318 47 Number of sites sampled: — Sediment Range 17–285 1–229 1–43 Median 117 8 5.5 — Biota Range 9–388 1–303 1–156 Median 57 6 4.5 Number of samples taken: — Sediment Range 117–1,426 1–771 1–108 Median 783 14 22 — Biota Range 26–13,262 1–1,310 1–644 Median 85 30 68 Study duration (years): Range 0.25–24 0.08–11 0.08–10 Median 2 0.8 0.9 Matrix sampled (number of studies sampling each matrix): Bed sediment 6 219 26 Aquatic biota (any) 23 202 39 Fish 20 173 25 Mollusks 8 32 12 Other invertebrates 4 40 19 Plants 0 8 8 Algae 0 7 2 Amphibians 0 8 5 Other aquatic biota 0 21 4 Pesticide class (number of studies analyzing each class): OC insecticides 25 312 29 OP insecticides 6 52 6 Carbamate insecticides 0 8 1 Other insecticides 0 2 1 CP acid herbicides 1 37 5 Triazine herbicides 1 10 2 Other herbicides 4 40 8 Fungicides 4 24 3 Table 2.4. General study characteristics condensed from Tables 2.1, 2.2, and 2.3 [Abbreviations: CP, chlorophenoxy acid; OC, organochlorine; OP, organophosphate] © 1999 by CRC Press LLC 2.2 GEOGRAPHIC DISTRIBUTION Figures 2.1 and 2.2 show the location of sampling sites for major national programs that measured pesticides in sediment and aquatic biota, respectively. The distribution of study effort, by state, is shown for state and local monitoring studies in Figure 2.3, and for process and matrix distribution studies in Figure 2.4. Figures 2.1 and 2.2 show that national studies have provided good geographic coverage of the United States, particularly for aquatic biota sampling. National estuarine and coastal monitoring programs have provided extensive coverage along the Atlantic, Pacific, and Gulf coasts; freshwater monitoring programs have provided fair coverage in inland states. Monitoring efforts by the state and local studies reviewed (which sampled rivers and streams, rather than lakes or coastal areas) have been heavier in the Great Lakes area, along the Mississippi River, and on the west coast, especially in California (see Figure 2.3). The geographic distribution of process and matrix distribution studies that were reviewed is highly uneven, with no studies in many states, especially in the Great Plains and Rocky Mountain areas (see Figure 2.4). Again, there appears to have been greater process-type study efforts in the Great Lakes area, along the Mississippi River, and in a few other states, especially Florida and California. 2.3 TEMPORAL DISTRIBUTION Two national studies spanned almost the entire duration of time encompassed by this report. The FDA’s NMPFF has measured pesticide residues in fish and shellfish samples from 1963 to the present. However, these samples are fish and shellfish products in commerce in the United States, and are not necessarily representative of the water resources of the United States. Moreover, published FDA reports did not provide information on sampling location, species of organism, tissue type, or type of hydrological system (lake, river, or marine system) sampled. The FDA results are useful for an assessment of human exposure (discussed in Section 6.2), but do not contribute much to our understanding of pesticides in the hydrologic system and will not be discussed further in this context. In contrast, the FWS’s NCBP, which sampled whole fish nationwide from 1967 to 1986, provides a unique data set for assessing the distribution and trends in pesticide residues across the United States. For monitoring studies in general, there appeared to be an increase in study effort per decade. This is illustrated both in Figure 2.5, which shows the total study duration (in years) by decade of sampling and by pesticide class, and in Figure 2.6, which shows the percentage of studies conducted in each decade by study type. The NMPFF was not included in compiling data for these figures for the reasons given above. In Figure 2.6, national and multistate monitoring studies and state and local monitoring studies show similar patterns, with the number of studies conducted at these scales increasing with each decade. Between 5–15 percent of monitoring studies at both scales were conducted during the 1960s, 30–35 percent during the 1970s, and about 50 percent during the 1980s. In contrast, the distribution of study effort by decade for process and matrix distribution studies shows that most of these studies were conducted during the 1970s, with fewer studies conducted during the 1980s than during the 1960s. Many process and matrix distribution studies were field experiments designed to assess the environmental fate, © 1999 by CRC Press LLC distribution, and persistence of a target pesticide after application of a known quantity of that pesticide, sometimes to meet requirements for pesticide registration. Although there are exceptions (such as studies that transplanted mollusks or that used SPMDs), many process studies were exploratory in nature. The observed temporal patterns in monitoring studies versus process studies suggests that a decrease during the 1980s in field experiments designed to assess the general environmental fate and persistence of pesticides was accompanied by an increase in monitoring studies that focused on analytes known to be present in sediment and aquatic biota (namely, the organochlorine insecticides and a few other, moderately hydrophobic pesticides). Of the national monitoring programs listed in Table 2.1, only the NS&T Program (besides the NMPFF) has continued into the 1990s. Many state monitoring programs are also continuing during the 1990s. Although the NCBP has been discontinued, additional national-scale sampling 1960s 1970s 1980s early 1990s early 1990sTotal 100 0 200 300 400 500 600 700 800 Duration, in total study years Fungicides Other herbicides CP acid herbicides Other insecticides OP insecticides OC insecticides Decade of sampling 10 8 6 4 2 0 12 Duration, in total study years Figure 2.5. Distribution of study effort by decade of sampling and by pesticide class. Duration (total study years) of study effort was calculated from all the monitoring studies listed in Tables 2.1 and 2.2. Each study was assigned to the most appropriate decade of sampling. Inset: enlargement of data for the early 1990s (1990–1994). Abbreviations: CP, chlorophenoxy; OC, organochlorine; OP, organophosphate. © 1999 by CRC Press LLC National monitoring Local monitoring Process 10 0 20 30 40 50 60 Percent of studies in each decade Study type 1960s 1970s 1980s Early 1990s EXPLANATION of pesticides in fish during the 1990s will be done under the National Water Quality Assessment (NAWQA) Program (which also is analyzing bed sediment) and the USGS’s Biomonitoring of Environmental Status and Trends (BEST) Program (formerly a program of the FWS). 2.4 SAMPLING MATRICES The scope of this review is limited to studies that investigated pesticide occurrence in bed sediment and aquatic biota. A companion book (Larson and others, 1997) reviewed studies on pesticides in water-column matrices, including whole (unfiltered) water, filtered water, suspended solids (biotic or abiotic particles separated from water by filtration or centrifugation), colloidal or dissolved organic carbon, and the surface microlayer. Because a number of studies measured one or more water-column matrices as well as bed sediment or biota, there is understandably some overlap between studies in Larson and others (1997) and in this review. Specific matrices included in this review are bed sediment, fish, mollusks, other invertebrates, plants, algae, amphibians, and other aquatic biota. The number of studies that investigated pesticide occurrence in each of these matrices is listed in Table 2.4. Many studies also measured pesticide residues in other media (such as water, suspended sediment, mammals, or birds). This information is not included in Table 2.4, but is noted under the listing for each individual study in Tables 2.1–2.3. Table 2.4 shows that more national and multistate monitoring studies measured pesticides in aquatic biota (23 of 25 studies) than in sediment (6 of 25 studies). The focus of monitoring efforts on the state and local scale was much more evenly divided, with slightly more studies measuring residues in bed sediment (219 studies) than in aquatic biota (202 studies). More process and matrix distribution studies measured pesticide residues in aquatic biota (39 studies) than in bed sediment (26 studies). Figure 2.6. Temporal distribution of studies reviewed, by the decade of sampling and by the study type. Study type: National monitoring—national and multistate moni- toring studies; Local monitoring—state and local monitoring studies; Process—process and matrix distribution studies. Graph is based on all the studies listed in Tables 2.1–2.3. © 1999 by CRC Press LLC Of aquatic biota investigated, more studies of all three types measured residues in fish than in other aquatic organisms. Other invertebrates and mollusks placed a distant second and third, respectively, in state and local monitoring studies and in process studies. In national and multistate monitoring studies, mollusks placed second relative to fish, and other invertebrates a distant third. As shown in Table 2.4, a number of state and local monitoring studies and process studies measured pesticide residues in plants, algae, amphibians, or other aquatic biota, whereas no national or multistate monitoring studies did. This is particularly true for process studies, since this group includes matrix distribution studies specifically designed to determine the distribution of the target pesticide in various environmental compartments. For studies that measured pesticide residues in fish, there was little consistency in the type of tissue sampled: whole fish, fillets or muscle tissue, liver, or other organs. Information on species sampled is provided for each study reviewed, when available (see Tables 2.1, 2.2, and 2.3). Species generally are referred to by common name. The aquatic species named in this report are listed in Appendix D. 2.5 TARGET ANALYTES Chemical and common names of pesticide analytes discussed in this book are listed in Appendix E. As noted in Section 2.2, all three types of studies tended to focus on organochlorine insecticides: 100 percent of national and multistate monitoring studies, 98 percent of state and local monitoring studies, and 62 percent of process and matrix distribution studies analyzed for organochlorine insecticides. Proportionately more process studies than monitoring studies focused on pesticides other than the organochlorine insecticides. Such studies established the extent to which certain pesticides were expected to be found in various environmental media, including sediment and biota. In contrast, monitoring studies targeted compounds known to be found in sediment and biota. The pesticide analytes investigated in the studies listed in Tables 2.1, 2.2, and 2.3 are classified into six pesticide groups: organochlorine insecticides, organophosphate insecticides, other insecticides, chlorophenoxy acid herbicides, other herbicides, and fungicides. The distribution of monitoring effort devoted to these six pesticide groups as a function of sampling decade is displayed in Figure 2.5. A few studies that analyzed for pesticides also analyzed for residues of tetrachlorodibenzo- p -dioxin (TCDD), usually the 2,3,7,8-isomer, which is a contaminant in some pesticide formulations, as well as a byproduct of incineration and some chlorination processes. For such studies, either TCDD or 2,3,7,8-TCDD (as appropriate for a given study) is listed as an analyte in Tables 2.1–2.3, although this information is not included in Table 2.4 or Figure 2.5. In compiling the data for Figure 2.5, all the monitoring studies that were reviewed (i.e., those listed in Tables 2.1 and 2.2) were included if they reported the pesticide analytes and sampling dates. As stated previously, the NMPFF was not included in Figure 2.5 because this program sampled fish and shellfish in interstate commerce; also, the published reports from this program listed the pesticide analytes for all raw agricultural commodities, but not fish and shellfish alone (Duggan and others, 1971; Food and Drug Administration, 1980, 1981, 1988, 1989a, 1990a, 1991, 1992; Yess and others, 1991a,b). For Figure 2.5, each state or local study was assigned to a single decade on the basis of the decade with the most years in which samples were collected. For national studies published as a series of reports, the data from © 1999 by CRC Press LLC [...]... was computed by summing the durations of studies that sampled during the appropriate decade, regardless of whether the studies looked at pesticides in sediment, aquatic biota, or both The number of analytes, sampling sites, and samples were not taken into account in Figure 2. 5; nonetheless, this figure gives a general indication of monitoring efforts in bed sediment and aquatic biota over the last few... hydrophobic and persistent in the environment; they therefore tend to be associated with organic materials in soil and sediment, and to bioaccumulate in living organisms (as discussed in Chapter 4) Also, organochlorine compounds continue to be used in other countries, and have the potential for long-range atmospheric transport and for contamination of migratory wildlife spending parts of their lives in those... organophosphate insecticides (especially chlorpyrifos and diazinon) were occasionally analyzed in aquatic biota, also Chlorophenoxy acid herbicides (2, 4-D, 2, 4,5-T, and silvex) and other herbicides (such as dacthal and trifluralin) constituted 10 and 14 percent, respectively, of total study effort during the 1970s During the 1980s, the study effort increased overall, with the relative proportion devoted to pesticides. .. the sample In general, increasing the sample mass will result in a lowering of detection limits, provided that sampling and extraction efficiencies remain the same Certain changes in analytical methodology, such as a change from packed-column to capillary-column gas chromatography, will markedly lower method detection limits In general, improvements in analytical methods for the organochlorine compounds... concentrations in 1977 compared with the previous 12 months of sampling (19 72 or earlier for different estuaries) Specifically, Butler (1973b) reported that total © 1999 by CRC Press LLC DDT was detected in 64 percent of the 1965–19 72 mollusk samples; dieldrin was detected in 15 percent of samples; and toxaphene, endrin, and mirex were each detected in 0 2 percent of samples In contrast, total DDT was detected in. .. during the 1980s Because detection frequencies for herbicides, fungicides, and organophosphate insecticides never reached the levels in sediment or aquatic biota that have been observed for organochlorine insecticides, it is not surprising that they have not been targeted with the same consistency in sediment and biota monitoring studies At the time this review was conducted, relatively few data were... organochlorine insecticides (44 percent) remaining about the same as during the 1970s Fungicides (primarily pentachlorophenol and its metabolite, pentachloroanisole) became more important target analytes during the 1980s, making up 11 percent of study effort Study effort devoted to organophosphate insecticides and chlorophenoxy acid herbicides declined to 8 and 4 percent, respectively, during the 1980s... for total DDT in 1977 This means that concentrations between 5 and 10 µ g/kg would have been reported as detections during 1965–19 72, but not in 1977 Because the mean concentration of total DDT in mollusks also decreased between the 1965– 72 and 1977 studies (Butler, 1973b), it is likely that total DDT in estuarine mollusks did decline during this time However, the evidence for such a decline would be... detection limit values listed in Tables 2. 1 and 2. 2 are those concentrations below which data were reported as nondetections in the studies reviewed Depending on the study, the © 1999 by CRC Press LLC detection limit may be a method detection limit (the lowest concentration that can be detected by the analytical method in an ideal matrix following specified test procedures) or a reporting limit (a threshold... matrix interferences) Many of the individual studies reviewed did not specify which kind of detection limit was used A major problem in comparing results from different studies is that of unknown or variable detection limits In some cases, analytical detection limits were inferred from the reported data when less-than values were given In other studies, a range of detection limits was reported for a single . sampling each matrix): Bed sediment 6 21 9 26 Aquatic biota (any) 23 20 2 39 Fish 20 173 25 Mollusks 8 32 12 Other invertebrates 4 40 19 Plants 0 8 8 Algae 0 7 2 Amphibians 0 8 5 Other aquatic biota. studies that investigated pesticides in bed sediment or aquatic biota in rivers and streams in the United States. The publications reviewed are summarized in Tables 2. 1, 2. 2, and 2. 3 (located. national programs for bed sediment and aquatic biota are presented in Figures 2. 1 and 2. 2, respectively. State and local monitoring studies (Table 2. 2) are occurrence and distribution surveys