EVALUATION OF PCDDs AND PCDFs IN WILD GAME TAKEN FROM THE FLOODPLAIN ALONG THE TITTABAWASSEE RIVER

Transcription

1 EVALUATION OF PCDDs AND PCDFs IN WILD GAME TAKEN FROM THE FLOODPLAIN ALONG THE TITTABAWASSEE RIVER Prepared by: ENTRIX, Inc. East Lansing, Michigan Prepared for: The Dow Chemical Company Midland, Michigan June 2004

2 Table of Contents 1.0 INTRODUCTION GENERAL AREA PCDDS, PCDFS AND CALCULATION OF TEQS STUDY OBJECTIVES REPORT ORGANIZATION METHODS STUDY SITE SELECTION COLLECTION OF WILD GAME ANIMALS PROCESSING Deer tissue processing Turkey tissue processing Squirrel processing Tissue homogenization QUALITY ASSURANCE AND QUALITY CONTROL SAMPLES CHEMICAL ANALYSIS CALCULATION OF TOXIC EQUIVALENCY CONCENTRATIONS (TEQ) USING PROXY VALUES FOR NON-DETECTS STATISTICAL ANALYSES CONGENER PATTERN ANALYSIS RESULTS LABORATORY COMPARISON CONCENTRATIONS OF PCDDS AND PCDFS IN WILD GAME ANIMALS COMPARISON OF TEQS IN DEER MUSCLE (VENISON) TO NATIONWIDE MARKET BASKET FOODS STATISTICAL COMPARISONS OF TEQ CONCENTRATIONS White-tailed Deer Wild Turkeys Squirrels Comparison of TEQ Concentrations in White-tailed Deer, Wild Turkeys, and Squirrels PERCENT CONTRIBUTION OF PCDD, PCDF, AND PCB CONGENERS TO THE TOTAL TEQ PRINCIPAL COMPONENTS ANALYSIS CONCLUSIONS REFERENCES Appendix A: Wild Game Sample Information Appendix B: Congener-Specific Analytical Data Appendix C: Laboratory Comparison Data Appendix D: Congener-Specific and Total TEQ Descriptive Statistics for Wild Game Samples Appendix E: Detailed Statistical Comparisons of Wild Game Data By Sex, Age, and Location Appendix F: Percent Contribution of PCDD, PCDF, and PCB Congeners to the Total TEQ Appendix G: Principal Component Analysis (PCA) of Wild Game Samples WGS Report ii June, 2004

3 Table of Tables Table 1-1. Example TEQ calculation Table 2-1. Number of wild game animals collected from the three sampling locations Table 3-1. Mean concentrations of TEQ MID 1 (ng/kg, wet weight) ± standard deviation in wild game animals collected from the reference sampling location and two sampling locations downstream of Midland, MI Table 3-2. Average concentrations of TEQ MID (ng/kg, wet weight) in muscle and skin from turkeys collected from the reference and Smith s Crossing sampling locations WGS Report iii June, 2004

4 Table of Figures Figure 2-1. Wild game sampling locations along the Tittabawassee River Figure 3-1. Comparison of concentrations of TEQ MID in market basket food items and venison harvested from the floodplain Figure 3-2. Percent contribution of PCDD and PCDF congeners to the total TEQ concentration in wild game animals WGS Report iv June, 2004

5 Definitions and Acronyms AhR ANOVA GPS IP HSD LOQ MANOVA MDEQ MDNR MSUATL ND PCA PCB PCDD PCDF RME SC TCDD TDI TEF TEQ TEQ DF TEQ MIN TEQ MID TEQ MAX U.S. USDA WHO Aryl hydrocarbon Receptor Analysis of Variance Global Positioning System near Imerman Park (study site) Honestly Significant Difference Limit of Quantitation Multivariate Analysis of Variance Michigan Department of Environmental Quality Michigan Department of Natural Resources Michigan State University Aquatic Toxicology Laboratory Non-detect Principal Components Analysis Polychlorinated biphenyl Polychlorinated dibenzo-p-dioxin Polychlorinated dibenzofuran Reasonable Maximum Exposure Smith s Crossing (study site) 2,3,7,8-tetrachlorodibenzo-p-dioxin Tolerable Daily Intake Toxic Equivalency Factor TCDD Equivalents TCDD equivalent concentration due to PCDDs and PCDFs TCDD equivalent concentration in which congeners below the method detection limit are assigned a proxy value of zero TCDD equivalent concentration in which congeners below the method detection limit are assigned a proxy value of ½ the method detection limit. TCDD equivalent concentration in which congeners below the method detection limit are assigned a proxy value equal to the method detection limit. United States United States Department of Agriculture World Health Organization WGS Report v June, 2004

6 1.0 INTRODUCTION 1.1 General Area The Tittabawassee River is located in central Michigan, and begins at the confluence of the Tobacco and Molasses Rivers. The river flows from the north through Midland Michigan in a southeastern direction to the confluence with Saginaw River and eventually to Saginaw Bay. Specifically, the general area being investigated as part of the remedial investigation includes the upstream boundary of The Dow Chemical Company to the confluence of the Tittabawassee and Shiawassee Rivers downstream of Greenpoint Island, as defined in the Hazardous Waste Management Facility Operating License, which was issued on June 12, 2003 by Michigan Department of Environmental Quality (MDEQ) to The Dow Chemical Company (Dow). Previous studies have reported that in this stretch of the Tittabawassee River, elevated concentrations of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) have been measured in sediments, floodplain soils, and fish (Hilscherova et al., 2003, MDEQ, 2003). 1.2 PCDDs, PCDFs and Calculation of TEQs PCDDs and PCDFs are mixtures of as many as 210 individual chemical congeners, 75 PCDDs and 135 PCDFs. The toxic potency of individual PCDD and PCDF congeners varies over several orders of magnitude. Thus, it is not possible to determine the toxic potency of these mixtures solely by determining the total PCDD and PCDF concentrations as a sum of congener concentrations. To combine concentrations of PCDD and PCDFs into a single aggregated measure, the toxic potency of each congener is calculated relative to the most toxic congener, which is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Toxic Equivalency Factors (TEFs) are congener-specific values that indicate the relative toxic potency of that congener compared to TCDD. There are 7 PCDDs and 10 PCDFs for which the World Health Organization (WHO) has determined TEF values (Van den Berg et al., 1998). The TEF value for each congener is multiplied by the measured concentration of that congener to calculate TCDD equivalents (TEQ). Because PCDDs, PCDFs, and related chemicals have the same mechanism of action, the concentrations of individual PCDD and PCDF congeners can be converted to TCDD equivalents (TEQs) for each sample (Equation 1-1,). A sample TEQ calculation is provided (Table 1-1). TotalTEQ = [( Congeneri TEFi ) +...( Congenern TEFn )] (Equation 1-1) i n In this report, concentrations of TEQs are expressed in units of nanogram TEQs per kilogram of sample (ng TEQs/kg sample, or ppt) on a wet weight basis. Lipid-normalized data are presented in the appendices.

7 Table 1-1. Example TEQ calculation 1 Chemical Concentration (ng/kg, wet weight) WHO TEF 2 TEQ (ng/kg, wet weight) 3 2,3,7,8-TCDD ,2,3,7,8-PeCDD ,2,3,4,7,8-HxCDD ,2,3,6,7,8-HxCDD ,2,3,7,8,9-HxCDD ,2,3,4,6,7,8-HpCDD OCDD ,3,7,8-TCDF ,2,3,7,8-PeCDF ,3,4,7,8-PeCDF ,2,3,7,8,9-HxCDF ,3,4,6,7,8-HxCDF ,2,3,4,7,8-HxCDF ,2,3,6,7,8-HxCDF ,2,3,4,6,7,8-HpCDF ,2,3,4,7,8,9-HpCDF OCDF Total TEQ = Congeners below the method detection limit were assigned a proxy value equal to ½ the method detection limit. 2 Toxic equivalency factor is relative to 2,3,7,8-TCDD. TEFs are based on Van den Berg et al., Product of the concentration of each congener and its TEF value. 4 Sum of the TEQ for each individual congener 1.3 Study Objectives Since some concentrations of PCDDs and PCDFs in soils from the floodplain along the Tittabawassee River (Floodplain) are greater downstream of Midland compared to those from reference locations (MDEQ, 2003; Hilscherova et al., 2003), questions have been raised regarding human consumption of wild game that reside and/or forage within this area. White-tailed deer (Odocoileus virginianus), wild turkeys (Meleagris gallopavo), and fox squirrels (Sciurus niger) were collected in this study because these game animals are commonly hunted in Michigan. Concentrations of PCDDs and PCDFs were measured in the edible portions of these game species. The purpose of this study was to answer the following questions: 1. Are concentrations of PCDDs and PCDFs in wild game collected from the Floodplain downstream of Midland statistically significantly greater than concentrations of PCDDs and PCDFs in samples of the same tissues from the same species collected at an upstream reference location? 2. Are wild game taken from the Tittabawassee River area downstream of Midland, MI acceptable for human consumption? (Addressed in a separate report) WGS Report 1-2 June, 2004

8 1.4 Report Organization The remainder of this WGS Report is organized into the following sections and appendices as follows: Section 2.0. Methods This section provides details concerning the selection of study sites, collection of wild game animals, tissue processing, quality assurance and quality control samples, trace residue analysis methodology, calculation of TEQs, statistical methodology, and congener pattern evaluation methodology. Section 3.0. Results This section provides details concerning the comparison of laboratory performance on split samples, summary tables of analytical results, comparison of concentrations of TEQs in deer muscle to other meats in the US food supply, an overview of statistical comparisons and congener pattern evaluations, and the relative contribution of PCDD, PCDF, and (for some samples) PCB congeners to the concentrations of total TEQ. Section 4.0. Conclusions This section provides the main conclusions from this study. Section 5.0. References Appendix A. Wild Game Sample Information This section provides the GPS coordinates, sample locations plotted on aerial photos, age and sex information for each of the wild game animals harvested. Appendix B. Congener-Specific Analytical Data This section provides the congener-specific analytical data, including the 17 co-planar PCDD and PCDF concentrations, for each wild game sample. Appendix C. Laboratory Comparison Data This section provides tables that compare the total TEQ concentrations and the relative percent differences for wild game sample splits sent to two different analytical laboratories. Appendix D. Congener-Specific and Total TEQ Descriptive Statistics for Wild Game Samples This section provides tables of descriptive statistics for concentrations of PCDD and PCDF congeners and total TEQ concentrations for each wild game species and sampling location. Wet weight and lipid normalized concentrations are summarized in this section. Appendix E. Detailed Statistical Comparisons of Wild Game Data By Sex, Age, and Location This section provides the detailed statistical tests that were conducted on the wild game data sets. Tests for normality, variance homogeneity, and comparisons between sexes, ages, and among sampling locations were conducted for the wild game species. Statistical power, when appropriate, was also evaluated in this section. Appendix F. Percent Contribution of PCDD, PCDF, and PCB Congeners to the Total TEQ This section provides spreadsheets that delineate the percent contribution of PCDD, PCDF, and PCB congeners to total TEQ concentrations on a sample-specific basis. WGS Report 1-3 June, 2004

9 Appendix G. Principal Component Analysis (PCA) of Wild Game Samples This section provides principal component analyses that were conducted for each wild game species in order to evaluate congener pattern differences among the three sampling locations. WGS Report 1-4 June, 2004

10 2.0 METHODS 2.1 Study Site Selection All study sites were located on the floodplain along the Tittabawassee River (Floodplain; Figure 2-1). Two collection sites were located downstream of Midland, MI and one reference site was located upstream of Midland, MI. Samples from the downstream locations were compared to samples collected from the upstream reference location. The reference site (RF) used in this study was located on privately owned land along the Tittabawassee River upstream of Midland, MI. This site was selected based on location, land use, and known concentrations of PCDDs and PCDFs in surrounding areas. The concentrations of PCDDs and PCDFs in the sediments and floodplain soils near the reference area are consistent with statewide background concentrations of TEQs (MDEQ, 2003). The reference area is approximately 9 km upstream of the Dow dam and is composed of mixed agricultural fields and forested areas. This reference location is reasonably far enough removed (15 km) from the nearest downstream study site that wild game animals were not expected to traverse between study sites. The first downstream site, referred to as Smith s Crossing (SC), is located approximately 6 km downstream of the Dow dam, and is owned by the Dow Chemical Company. The second downstream site is located on private property near Imerman Park (IP), approximately 21 km downstream of the Dow dam. The land use at both downstream sites can be described as primarily mixed agriculture and forest. Concentrations of TEQs in floodplain soil samples collected in and near Imerman Park ranged up to 2400 ng/kg (on a dry weight basis). Thus, the available data suggest that this area contains concentrations of PCDDs and PCDFs that are elevated above background concentrations. Reference Area Dow Dam Near Smith s Crossing Near Imerman Park Figure 2-1. Wild game sampling locations along the Tittabawassee River WGS Report 2-1 June, 2004

11 2.2 Collection of Wild Game Animals Wild game animals were harvested just prior to, during, and after the hunting season, from November 2003 to January Thus, the sampling effort coincided with normal hunting activities and the samples are thus representative of what hunters would be expected to harvest from the Floodplain. White-tailed deer, turkeys, and squirrels were collected from areas along the Tittabawassee River at the three areas that were described previously (Figure 2-1). The number of each species harvested is presented separately for each location (Table 2-1) and individual sample locations are shown (Appendix A). White-tailed deer (Odocoileus virginianus), wild turkey (Meleagris gallopavo), and fox squirrels (Sciurus niger) were collected in compliance with state collecting permits. United States Department of Agriculture (USDA) Wildlife Services personnel harvested all deer and turkey for this study, while squirrels were harvested by ENTRIX personnel. Table 2-1. Number of wild game animals collected from the three sampling locations Study Site # White-tailed deer collected # Wild turkeys collected # Fox squirrels collected Reference (private property) Smith s Crossing (Dow Property) Near Imerman Park (private property) Total The wild game sampling effort took place from November 10, 2003 until January 14, The sampling methods involved standard firearm hunting practices. In addition, spotting lights were used for sample collection of deer at night. After the animals were harvested, ENTRIX field personnel completed documentation as specified in the associated standard operating procedures to include information such as: sample identification (ID) number, date, time, temperature, weather conditions, GPS coordinates, and samplers initials. Sample information (e.g., sample identification number, location, and date) for each harvested animal is presented in Appendix A. After documentation was completed, each deer was wrapped in plastic sheeting and each turkey or squirrel was placed in a separate plastic bag and transported to the field laboratory. Deer were preliminarily aged in the field. Deer that were determined to be less than 1.5 years old were not submitted for analysis of PCDDs and PCDFs in accordance with the Wild Game Sampling Workplan (ENTRIX, 2003). A more detailed age determination by the cementum annuli method was conducted by Dr. Paul Friederich of the MDNR at the Rose Lake Research Laboratory (Appendix A). 2.3 Processing All wild game animals were dressed according to standard hunting practices. However, to avoid crosscontamination, care was taken to rinse all surfaces and instrumentation with analytical grade acetone and hexane between samples. After dressing, tissue samples were collected from wild game animals. Both liver and muscle tissues were collected from white-tailed deer, while only muscle tissues were collected from wild turkeys and squirrels. The tissue processing steps are described below for each species Deer tissue processing Only one deer was processed at a time. Each deer was elevated using a gambrel, then the hide was removed. In addition, the head was removed and stored for later aging. If antlers greater than 3 were present, they were removed and tagged with sample identification and given to the DNR in accordance with permit requirements. The internal organs of the deer were then removed and disposed of with the WGS Report 2-2 June, 2004

12 exception of the liver. The liver was collected and rinsed with HPLC grade water to remove any foreign debris and/or fur that may have come in contact with the tissue during field dressing. The liver was weighed, cut into approximately 1 cubes, and then g of liver was transferred into a chemically clean 1-L I-Chem glass jar that was labeled with sample identification (ID) number, date, time, and contents. The inside of the carcass was also rinsed with HPLC grade water to remove any possible debris. Rump roast (500 g), back strap (250 g), and tenderloin (250 g) were cut into approximately 1 cubes, the weights of each tissue recorded, and then the three portions of meat were combined into a 1-L chemically clean, I-Chem glass jar. Each sample jar was labeled with sample identification (ID) number, date, time, and contents. The meat samples were then placed in a -20 C freezer until transfer to the Michigan State University Aquatic Toxicology Laboratory (MSU-ATL). The samples were packaged and shipped in accordance with SOP #214 Documentation, Preservation, Handling, And Tracking of Samples For Analysis to Michigan State University s Aquatic Toxicology Laboratory (MSU-ATL). The remainders of the carcasses were stored frozen until the field sampling portion of the project was completed. The samples were then delivered to a butcher to be processed and the meat frozen until analytical results have been reviewed Turkey tissue processing Only one turkey was processed at a time. Each turkey was weighed, sexed, and aged. Then, each turkey was plucked, the neck and wings were removed, and the body was field dressed. The outside of the skin and the inside of the body cavity were rinsed with HPLC grade water to remove any foreign debris and/or feathers that may have come in contact with the tissue during field dressing. Skin-on muscle tissue samples were taken from the breast (700 g) and the legs and thighs (300 g). Reasonable efforts were made to avoid areas that were obviously affected by gunshot. The white and dark meat was cut into approximately 1 cubes, weights recorded, and then the two portions of meat were combined into a 1-L chemically clean, I-Chem glass jar. Each sample jar was labeled with sample identification (ID) number, date, time, and contents. The meat samples were then placed in a 20 C freezer until transport to the MSU-ATL University Research Containment Facility (UCRF). The samples were packaged and shipped in accordance with SOP #214 Documentation, Preservation, Handling, And Tracking of Samples For Analysis to Michigan State University s Aquatic Toxicology Laboratory (MSU-ATL) Squirrel processing Only one squirrel was processed at a time. Each squirrel was weighed, sexed, and aged. Then, each squirrel was skinned and the paws and head were removed. The body was then field dressed. The outside and inside of the body were rinsed with HPLC grade water. Muscle tissue samples were taken from all parts of the body. The meat was cut into approximately 1 cubes. The portions of the meat were combined into a chemically clean 250 ml I-Chem jar. Each sample jar was labeled with sample identification (ID), date, time, and contents. The meat samples were then placed in a 20 C freezer until transport to the MSU-ATL University Research Containment Facility (UCRF). The samples were packaged and shipped in accordance with SOP #214 Documentation, Preservation, Handling, And Tracking of Samples For Analysis to Michigan State University s Aquatic Toxicology Laboratory (MSU-ATL) Tissue homogenization Tissue samples were homogenized in stainless steel blenders. Between samples, the blenders were washed with Liquinox soap, rinsed with nanopure water followed by reagent grade acetone and hexane. Samples homogenates were aliquoted into six separate chemically clean I-CHEM jars. One jar was shipped to the analytical laboratory, and remaining jars were archived at Michigan State University s WGS Report 2-3 June, 2004

13 Aquatic Toxicology Lab. All tissue homogenates were stored at -20 C until time of shipment to the analytical laboratory. 2.4 Quality Assurance and Quality Control Samples In addition to tissue samples, atmospheric blanks from the field and laboratory were collected and submitted for analysis. Surfaces, dissecting equipment, and homogenization equipment that came into contact with meat were solvent rinsed at the beginning and end of each day and between each sample in accordance with the Quality Assurance Project Plan (QAPP). Solvent rinsate blanks were submitted for analysis as pairs (e.g. beginning and end of the day) for both the tissue processing and homogenization activities. 2.5 Chemical Analysis Analyses of PCDD, PCDF, and PCB concentrations in wild game samples were conducted at Agriquality Limited (Lower Hutt, New Zealand) using EPA method In addition, some duplicate samples were sent to Alta Analytical (El Dorado Hills, CA). Deer, turkey and squirrel samples and corresponding duplicate blind samples were sent to both laboratories so that accuracy and precision of the chemical analyses could be evaluated at both laboratories. 2.6 Calculation of Toxic Equivalency Concentrations (TEQ) Using Proxy Values for Non-Detects TEQ concentrations were calculated several different ways based on substituting various proxy values for congeners that were present at concentrations less than the method detection limit (MDL). The proxy values that were used to calculate TEQ concentrations were: Non-detects (ND) = 0, ND = ½ MDL, and ND = MDL. TEQ MIN concentrations were calculated when congeners below the MDL were assigned a proxy value of 0.0. TEQ MID concentrations were calculated when congeners below the MDL were assigned a proxy value of ½ MDL. TEQ MAX concentrations were calculated when congeners below the MDL were assigned a proxy value equal to the MDL. Therefore, the TEQ MIN is the lower bound estimate of TEQ concentrations, while the TEQ MAX is the upper bound estimate of total TEQ concentrations. While 4 non-ortho- (co-planar) and 8 mono-ortho-substituted polychlorinated biphenyls (PCBs) also exhibit AhR activity, only PCDDs and PCDFs were used to calculate concentrations of TEQs in the majority of collected wild game animals. Concentrations of PCBs were measured in 25% of the deer muscle tissues, and total TEQ DFP (including PCDDs, PCDFs, and PCBs) were calculated for these samples. 2.7 Statistical Analyses In all cases, statistical comparisons were conducted with concentrations of total TEQs derived from the concentrations and relative potencies of the 7 PCDDs and 10 PCDFs that exhibit AhR agonist activity. TEQ MID concentrations were used in all statistical evaluations. Separate analyses were performed for each species to test for differences in concentrations of total TEQ MID among sample locations. The Shapiro-Wilk s test was used to test the data sets for normality. To test for homogeneity of the variances, Levene s test was used. If the data did not meet the criteria for normality, the TEQ data were log transformed and then reanalyzed. If log-transformed data still did not meet the criteria for normality and the variances among locations were not determined to be homogeneous, non-parametric statistics were used for subsequent comparisons. WGS Report 2-4 June, 2004

14 When data from wild game samples met the requirements for parametric tests, then a Student s t-test (equal sample sizes) or the tabled t-test (unequal sample sizes) was used to compare TEQs between two locations. When all three study sites were compared, an ANOVA with Tukey s Honestly Significant Difference (HSD) was used to compare TEQs among locations (Wilkinson, 2000). If the data did not meet the criteria for being normally distributed and/or did not meet the criteria for variance homogeneity, non-parametric statistical tests were used to evaluate differences among locations. The Mann-Whitney U test was used to evaluate differences between locations when the comparison involved either two study sites or sample types. When comparisons were made among all three study sites, the Kruskal Wallis test was used for statistical analyses (Wilkinson, 2000). For comparisons between locations, one-tailed statistical tests were used to evaluate potential differences in TEQ concentrations between the upstream reference location and the downstream locations. For comparisons between male and female TEQ concentrations, two-tail statistical tests were used to evaluate potential differences. The significance level (α) for all statistical tests, was set at 0.05 or 5%. Alpha (α) indicates the probability of committing a Type I error, that is, rejecting a null hypothesis (mean1 =mean2) when in fact it is true. Thus, 95% of the time our statistical test would correctly reject the null hypothesis with a 5% chance that it would arrive at an incorrect conclusion. For the probability of committing a Type II error, a Beta (β) of 0.2 was used in all analyses. β is the error rate for committing a Type II error where one accepts the null hypothesis when in fact the alternative hypothesis is true. That is, the analyst is willing to accept 20% error rate in which a Type II error may be committed. The power of a statistical test is calculated as 1-β. Generally, a statistical test is considered to have sufficient power if 1- β is equal to or greater than 80% (Salsburg, 1986). For power analyses, the Type I error (α) was set to 0.05 and the relative effect s distance (difference between locations) was selected as 3-fold the TEQ concentration found in comparable upstream reference samples. Power analyses were conducted with PASS 2002 (NCSS, Kaysville, UT). 2.8 Congener Pattern Analysis When statistical differences in total TEQs were detected among upstream and downstream locations, analyses of the patterns of relative concentrations (frequency and magnitude) of congeners were evaluated. Significant differences in congener patterns were evaluated with Principal Components Analysis (PCA). If the concentration data were not normally distributed, congener data was log transformed, and the PCA utilized the log-transformed data, with a correlation-based PCA using pair-wise deletion of missing data and Varimax rotation. WGS Report 2-5 June, 2004

15 3.0 RESULTS 3.1 Laboratory Comparison A subset of deer, turkey, and squirrel samples and blind duplicates were sent to Agriquality (Lower Hutt, New Zealand) and Alta Analytical (El Dorado Hills, CA) to assess and compare the accuracy and precision in analytical measurements at both laboratories. Results from this laboratory comparison indicated that both laboratories produced similar high quality data. Total TEQ concentrations reported in a subset of deer muscle and liver tissues were not significantly different between labs, suggesting that the results at both laboratories were accurate. To assess laboratory precision, concentrations of total TEQs in wild game samples were compared to corresponding blind duplicate samples, and the relative percent difference (RPD) in total TEQ concentrations was calculated for each sample and averaged across all samples to give a mean RPD value for each laboratory. The mean RPD for samples analyzed at Alta (n= 11) was 28.7%, while the mean RPD for samples analyzed at Agriquality (n=11) was 11.5%. Based on this assessment of laboratory precision, it was decided that Agriquality labs would analyze the remaining wild game samples. Congener specific analytical results for each wild game sample are listed in Appendix B. Analytical laboratory comparison data are presented in Appendix C. 3.2 Concentrations of PCDDs and PCDFs in Wild Game Animals In white-tailed deer, wild turkey, and squirrels, edible portions of muscle tissue were collected and analyzed for concentrations of PCDDs and PCDFs. In addition, deer livers were collected and analyzed for concentrations of PCDDs and PCDFs. Concentrations of PCBs were analyzed in a subset of deer muscle tissue samples. Descriptive statistics for PCDD and PCDF congeners as well as TEQ concentrations are available for each wild game species at each sampling locations (Appendix D). Concentrations of total TEQs in deer muscle and liver, wild turkey muscle and squirrel muscle are presented (Table 3-1). To examine the distribution of PCDDs and PCDFs in turkey tissue, separate analyses were conducted on a subset of 3 turkeys from each of the reference and Smith s Crossing sampling locations. For these six birds, concentrations of PCDDs and PCDFs were measured in skin-on muscle tissue, skin-off muscle tissue, and in the skin (Table 3-2.). Results indicated that turkey skin at both sampling locations contained greater concentrations of PCDDs and PCDFs when compared to muscle tissue, and the TEQ concentrations in skin-on muscle tissue were approximately 2-fold greater than skin-off muscle tissue. In other words, concentrations of TEQs can be reduced in half by removing skin and associated fat. WGS Report 3-1 June, 2004

16 Table 3-1. Mean concentrations of TEQ MID 1 (ng/kg, wet weight) ± standard deviation in wild game animals collected from the reference sampling location and two sampling locations downstream of Midland, MI Reference Smith s Crossing Near Imerman Park a Deer Muscle ± (n= 13) Deer Liver a ± (n=13) Turkey Muscle 2 a ± (n=12) Squirrel Muscle a ± (n=12) b ± (n=10) b 11.1 ± 5.49 (n=10) b ± (n=10) c ± (n=10) c 64.0 ± 56.8 (n=10) b 10.2 ± 6.98 (n=12) b 1.32 ± 1.58 (n=10) 1 TEQ MID: Congeners below the method detection limit were assigned a proxy value equal to ½ the method detection limit. Concentrations reported as mean + standard deviation (ng/kg, wet weight) 2 Turkey muscle data were combined for Smith s Crossing and near Imerman Park since there was only a single sample harvested near Imerman Park. Different superscript letters indicate a significant difference at p < WGS Report 3-2 June, 2004

17 Table 3-2. Average concentrations of TEQ MID (ng/kg, wet weight) in muscle and skin from turkeys collected from the reference and Smith s Crossing sampling locations Reference Skin-on meat (ng/kg, wet weight) ± 0.12 (n= 3) Skin-off meat (ng/kg, wet weight) ± 0.03 (n=3) Skin (ng/kg, wet weight) ± 0.71 (n=3) Ratio of TEQ concentrations in skin-off meat 0.52 ± 0.24 to skin-on meat (unitless) 2 (n=3) Smith s Crossing 12.0 ± 7.93 (n=3) 6.49 ± 5.90 (n=3) 120 ± 125 (n=3) 0.49 ± 0.12 (n=3) 1 TEQ MID: Congeners below the method detection limit were assigned a proxy value equal to ½ the method detection limit. Concentrations reported as mean + standard deviation (ng/kg, wet weight). Note also that the analyses were conducted on the same sample of wild turkey (3 per location) that were split and analyzed as skin-on meat, skin-off meat, and skin. 2 Ratio of TEQ concentrations in skin-off meat to skin-on meat was calculated for each individual sample before conducting statistics. WGS Report 3-3 June, 2004

18 3.3 Comparison of TEQs in Deer Muscle (Venison) to Nationwide Market Basket Foods Concentrations of TEQs in venison harvested from the Floodplain were compared to TEQ concentrations in Nationwide Market Basket (NMB) surveys of commercially available foodstuffs (Institute of Medicine, 2003; Fiedler et al., 1997; Focant et al., 2003; Jensen et al., 2000; Ryan et al, 1997). TEQ concentrations in commercially available meats were comparable to TEQ concentrations measured in venison collected from the floodplain along the Tittabawassee River (Figure 3-1). TEQ concentrations in beef, pork, poultry, and fish were sometimes greater than TEQ concentrations in deer from the floodplain, including deer collected near Imerman Park. 2.0 TEQ MID (ng/kg) Beef Fish Pork Poultry Reference Smith s Crossing Near Imerman Park Market Basket Foods Venison from the Floodplain Figure 3-1. Comparison of concentrations of TEQ MID in market basket food items and venison harvested from the floodplain. WGS Report 3-4 June, 2004

19 3.4 Statistical Comparisons of TEQ Concentrations White-tailed Deer Concentrations of TEQs in muscle and liver tissues were not statistically different between male and female deer [Student s t-test, p=0.148 (muscle); Mann-Whitney U, p=0.236 (liver)]. Therefore, male and female deer data were combined for subsequent statistical analyses among sampling locations. TEQ concentrations in deer sampled from Smith s Crossing and near Imerman Park were statistically greater than concentrations measured in deer muscle sampled from the reference location [Tukey s HSD, p=0.004(rf vs. SC), p< (RF vs. IP)]. Furthermore, a comparison of TEQ concentrations in deer sampled at Smith s Crossing and near Imerman Park indicated that TEQ concentrations in deer from near Imerman Park were statistically greater than concentrations measured in deer sampled from Smith s Crossing (Tukey s HSD, p< ). The same results were observed among the deer liver samples. Concentrations of TEQs in deer liver were significantly greater at Smith s Crossing and near Imerman Park compared to the reference sampling location [Mann-Whitney U, p< (RF vs. SC), p< (RF vs. IP)]. Concentrations of TEQ were also greater in deer livers collected from near Imerman Park compared to Smith s Crossing (Mann-Whitney U, p= 0.002). Detailed statistical procedures and comparisons for deer muscle and liver samples are presented in Appendix E Wild Turkeys Since only a single turkey was collected from the sampling location near Imerman Park, and because the TEQ value in this individual was within the range from those in the Smith s Crossing sampling location, turkeys collected from both locations downstream of Midland, MI were pooled for statistical analyses. Mean concentrations of TEQs were greater in turkeys collected at downstream locations compared to the reference sampling location (Student s t-test, p<0.0001). Detailed statistical procedures and comparisons for turkey muscle samples are presented in Appendix E Squirrels TEQ concentrations in squirrel muscle were found to be significantly different between reference and downstream sampling locations [p= (RF vs. SC), p= (RF vs. IP)]. Even though mean TEQ concentrations in squirrel muscle near Imerman Park were approximately 3-fold greater than mean TEQ concentrations at Smith s Crossing, this difference was not statistically significant (p= 0.257). Detailed statistical procedures and comparisons for squirrel muscle samples are presented in Appendix E Comparison of TEQ Concentrations in White-tailed Deer, Wild Turkeys, and Squirrels At each location, the rank order of average TEQ concentrations was deer meat (venison) < squirrel meat < turkey meat < deer liver. The likely reasons for the lower concentrations of TEQs in deer relative to wild turkeys is probably related to soil ingestion. For example, incidental soil ingestion by deer has been estimated to be less than 2% whereas wild turkeys consume up to 9% soil (Beyer et al., 1994). Animals consume soil as part of their diet in several ways: by grooming dirt off of their fur, through accidental ingestion during feeding, or actively seeking to increase the amount of trace elements in their diet or in the case of turkeys, ingestion of grit (small rocks and gravel) as an aid in digestion. While the amount of incidental soil ingestion is unknown for fox squirrels, it is likely to be intermediate to that of deer and wild turkey. The greater concentrations of TEQs in deer liver compared to deer muscle is likely explained by the ability of the liver to sequester PCDDs and PCDFs through mechanisms such as WGS Report 3-5 June, 2004

20 induction of the cytochrome P450 enzyme, CYP1A2, and subsequent binding to this same enzyme (Santostefano et al., 1996). 3.5 Percent Contribution of PCDD, PCDF, and PCB Congeners to the Total TEQ The percent contribution of PCDDs and PCDFs to the total TEQ concentration in wild game animals was different at the various sampling locations. At the reference location (RF), concentrations of 2,3,7,8- TCDD, 1,2,3,7,8-PeCDD, and 2,3,4,7,8-PeCDF, contributed the largest percentages to the total TEQ. At the two sampling locations downstream of Midland, MI, the percent contributions to the total TEQ were different such that the percent contribution of TCDD and other PCDDs decreased and the percent contribution of PCDFs, especially 2,3,4,7,8-PeCDF, increased (Figure 3-2). Total PCBs were assessed in a total of nine deer (three from each location). At the reference location, total PCBs contributed an average of 28.7% to the total TEQ (which equates to ng TEQ/kg). At Smith s Crossing and near Imerman Park, the average percent contribution of PCBs to the total TEQ was 8.7 and 6.1 %, respectively (which equates to and ng TEQ/kg, respectively). These tissue residue concentrations of PCBs are consistent with other data that suggest that PCBs do not contribute significantly to total TEQs in floodplain soils along the Tittabawassee River downstream of Midland. More detailed information on the percent contribution of PCDD, PCDF, and PCB congeners to the total TEQ are available in Appendix F. 3.6 Principal Components Analysis Congener pattern analysis was conducted on turkey, squirrel and deer muscle congener data, as well as deer liver congener data since statistically significant differences in total TEQ concentrations were observed among locations for these various wild game matrices. Congener profiles were distinctly different among sampling locations for both deer muscle and liver as well as turkey muscle samples. Congener profiles overlapped to a greater extent among squirrel muscle samples collected at the various sample locations. Detailed principal components analysis (PCA) for each of the wild game species is presented in Appendix G. WGS Report 3-6 June, 2004

21 100% 80% 60% 40% 20% Other congeners 1,2,3,4,7,8-HxCDF 2,3,4,7,8-PeCDF 1,2,3,7,8-PeCDF 2,3,7,8-TCDF 1,2,3,7,8-PeCDD 2,3,7,8-TCDD 0% Deer Muscle Deer Liver Turkey Muscle Squirrel Muscle Deer Muscle Deer Liver Turkey Muscle Squirrel Muscle Deer Muscle Deer Liver Turkey Muscle Squirrel Muscle Reference Smith s Crossing Near Imerman Park Figure 3-2. Percent contribution of PCDD and PCDF congeners to the total TEQ concentration in wild game animals (Other congeners included 1,2,3,4,7,8-HxCDD, 1,2,3,6,7,8-HxCDD, 1,2,3,7,8,9-HxCDD, 1,2,3,4,6,7,8-HpCDD, OCDD, 1,2,3,6,7,8-HxCDF, 2,3,4,6,7,8-HxCDF, 1,2,3,7,8,9-HxCDF, 1,2,3,4,6,7,8-HpCDF, 1,2,3,4,7,8,9-HpCDF, and OCDF, which individually constituted < 5% of the total TEQ.) WGS Report 3-7 June, 2004

22 4.0 CONCLUSIONS In conclusion, concentrations of TEQs in edible portions of wild game taken downstream of Midland were greater than the concentrations of TEQs from upstream of Midland. Average concentrations of TEQs followed the order of deer muscle (venison) < squirrel meat < turkey meat < deer liver. Concentrations of TEQs in deer muscle averaged 0.07 ng TEQ/kg (parts per trillion, ppt) from the upstream reference location, while concentrations at the two locations downstream of Midland averaged 0.17 and 0.52 ppt. These values are within the range of TEQs concentrations in meats, dairy, and eggs ( ppt) as determined by nationwide market basket surveys. The greatest concentrations of TEQs were measured in deer liver, which is understandable, since one of the functions of liver is to retain such materials. Concentrations of TEQs in deer liver averaged 0.57 ppt from the upstream reference location, while concentrations of TEQs at the two locations downstream of Midland averaged 11 and 64 ppt. Wild turkeys were collected and analyzed in two ways, skin-on breast with leg meat and skin-off breast with leg meat. Concentrations of TEQs in skin-on wild turkey meat averaged 0.18 ppt from the upstream reference location, while concentrations for the two combined locations downstream of Midland averaged 10 ppt. Results indicated that removal of skin and fat, a common practice in preparing wild turkey in order to avoid plucking the feathers, reduced concentrations of TEQs in wild turkey meat by approximately 50% at both sampling locations. The greater concentrations of TEQs in turkey compared to deer meat is likely due to differences in incidental soil ingestion (e.g., wild turkeys have an incidental soil ingestion rate of approximately 9% compared to less than 2% for white-tailed deer) and dietary differences. Concentrations of TEQs in squirrel meat averaged 0.07 ppt from the upstream reference location, while concentrations at the two locations downstream of Midland averaged 0.40 and 1.3 ppt. WGS Report 4-1 June, 2004

23 5.0 REFERENCES Beyer, W.N., Connor E., and S. Gerould Survey of soil ingestion by wildlife. Journal of Wildlife Management 58: Entrix, Inc Draft Work Plan for the Interim Response Activity of Evaluating Wild Game Taken from the Tittabawassee River Floodplain. Fiedler, H., Cooper, K., Bergek, S., Hjelt, M., Rappe, C Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans (PCDD/PCDF) in food samples collected in southern Mississippi, USA. Chemosphere 34: Focant, J., Pirard, C., De Pauw, E Levels of PCDDs, PCDFs, and PCBs in international fast food samples. Organohalogen Compounds 64: Hilsherova, K., Kannan, K., Nakata, H., Yamashita, N., Bradley, P., McCabe, J., Taylor, A.B., Giesy, J. P., Polychlorinated Dibenzo-p-dioxin and Dibenzofuran Concentration Profiles in Sediments and Flood-Plain Soils of the Tittabawassee River, Michigan. Environ. Sci. Technol. 37: Institute of Medicine Dioxin and Dioxin-like Compounds in the Food Supply: Strategies to Decrease Exposure. The National Academic Press, Washington D.C. Jensen, E., Canady, R., Bolger, P.M Exposure assessment for dioxins and furans in seafood and dairy foods in the United States, Organohalogen Compounds 47: MDEQ Phase II Tittabawassee/Saginaw River Dioxin Floodplain Sampling Study June Michigan Department of Environmental Quality Remediation and Redevelopment Division. Ryan, J.J., Beaudoin N., Mills, P., Patry, B Dioxin-like compounds in total diet food, Canada Organohalogen Compounds 32: Salsburg, D.S Statistics for Toxicologists. Marcel Dekker, Inc, New York. Santostefano, M.J., Johnson K.L., Whisnant N.A., Richardson V.M., DeVito M.J., Diliberto J.J., Birnbaum L.S Subcellular localization of TCDD differs between the liver, lungs, and kidneys after acute and subchronic exposure: species/dose comparisons and possible mechanism. Fundam Appl Toxicol. 34(2): Van den Berg, M., Birnbaum, L. Bosveld, B.T.C., Brunstrom, B., Cook, P., Feeley, M., Giesy, J.P., Hanberg, A., Hasegawa, R., Kennedy, S.W., Kubiak, T., Carsen, J.C., van Leeuwen, F.X.R., Liem, A.K.D., Nolt, C., Peterson, R.E., Poellinger, L., Safe, S., Schrenk, D., Tillitt, D., Tysklind, M., Younes, M., Waern, F., and Zacharewski, T Toxic Equivalency Factors (TEFs) for PCBs, PCDDs, PCDFs, for Humans and Wildlife. Environ. Health Perspec. 106: Wilkinson, L Nonparametric Statistics in SYSTAT R 10. Statistics II. Chicago, Ill: SPSS Inc., Chapter 5, WGS Report 5-1 June, 2004

24 APPENDIX A: WILD GAME SAMPLE INFORMATION A- i

25 Table of Tables Table A -1: White-tailed deer sample information...a-1 Table A -2: Wild turkey sample information...a-2 Table A -3: Squirrel sample information...a-3 Table of Figures Figure A -1: White-tailed deer sampling locations at Site 1 (upstream reference)...a-4 Figure A -2: White-tailed deer sampling locations at Site 2 (near Smith's Crossing)...A-5 Figure A -3: White-tailed deer sampling locations at Site 3 (near Imerman Park)...A-6 Figure A -4: Wild turkey sampling locations at Site 1 (upstream reference)...a-7 Figure A -5: Wild turkey sampling locations at Site 2 (near Smith's Crossing)...A-8 Figure A -6: Wild turkey sampling locations at Site 3 (near Imerman Park)...A-9 Figure A -7: Squirrel sampling locations at Site 1 (upstream reference)...a-10 Figure A -8: Squirrel sampling locations at Site 2 (near Smith's Crossing)...A-11 Figure A -9: Squirrel sampling locations at Site 3 (near Imerman Park)... A-12 A-ii

26 Table A -1: White-tailed deer sample information Field ID Description Location Northing Westing Collection Date Age 1 (years) Sex TR01DR01 DEER RF N ' W ' 11-Nov FEMALE TR01DR02 DEER RF N ' W ' 13-Nov FEMALE TR01DR03 DEER RF N ' W ' 19-Nov FEMALE TR01DR04 DEER RF N ' W ' 19-Nov MALE TR01DR05 DEER RF N ' W ' 19-Nov-03 ND MALE TR01DR06 DEER RF N ' W ' 20-Nov FEMALE TR01DR07 DEER RF N ' W ' 20-Nov MALE TR01DR08 DEER RF N ' W ' 12-Jan FEMALE TR01DR09 DEER RF N ' W ' 12-Jan FEMALE TR01DR10 DEER RF N ' W ' 12-Jan FEMALE TR01DR11 DEER RF N ' W ' 12-Jan-04 < 1.5 FEMALE TR01DR12 DEER RF N ' W ' 13-Jan FEMALE TR01DR13 DEER RF N ' W ' 13-Jan FEMALE TR01DR14 DEER RF N ' W ' 13-Jan FEMALE TR02DR01 DEER SC N ' W ' 12-Nov-03 < 1.5 MALE TR02DR02 DEER SC N ' W ' 12-Nov-03 < 1.5 FEMALE TR02DR03 DEER SC N ' W ' 12-Nov MALE TR02DR04 DEER SC N ' W ' 17-Nov MALE TR02DR05 DEER SC N ' W ' 19-Nov-03 < 1.5 MALE TR02DR06 DEER SC N ' W ' 19-Nov-03 < 1.5 MALE TR02DR07 DEER SC N ' W ' 19-Nov MALE TR02DR08 DEER SC N ' W ' 20-Nov FEMALE TR02DR09 DEER SC N ' W ' 20-Nov FEMALE TR02DR10 DEER SC N ' W ' 01-Dec FEMALE TR02DR11 DEER SC N ' W ' 01-Dec FEMALE TR02DR12 DEER SC N ' W ' 04-Dec FEMALE TR02DR13 DEER SC N ' W ' 04-Dec FEMALE TR02DR14 DEER SC N ' W ' 16-Dec FEMALE TR03DR01 DEER IP N ' W ' 12-Nov MALE TR03DR02 DEER IP N ' W ' 12-Nov FEMALE TR03DR03 DEER IP N ' W ' 12-Nov MALE TR03DR04 DEER IP N ' W ' 12-Nov MALE TR03DR05 DEER IP N ' W ' 12-Nov FEMALE TR03DR06 DEER IP N ' W ' 12-Nov FEMALE TR03DR07 DEER IP N ' W ' 12-Nov MALE TR03DR08 DEER IP N ' W ' 12-Nov MALE TR03DR09 DEER IP N ' W ' 13-Nov FEMALE TR03DR10 DEER IP N ' W ' 18-Nov FEMALE RF: Reference sampling location; SC: Downstream sampling location near Smith s Crossing; IP: downstream sampling location near Imerman Park, ND: Not Determined 1 Age for deer 1.5 years and younger was determined by the jaw wear method. For older deer, the cementum annuli method was used to age deer. A-1

27 Table A -2: Wild turkey sample information Collection Date Body Weight (kg) Field ID Description Location Northing Westing TR01TY01 TURKEY RF N ' W ' 10-Nov JUVENILE FEMALE TR01TY02 TURKEY RF N ' W ' 10-Nov JUVENILE FEMALE TR01TY03 TURKEY RF N ' W ' 11-Nov MATURE FEMALE TR01TY04 TURKEY RF N ' W ' 11-Nov ADULT FEMALE TR01TY05 TURKEY RF N ' W ' 11-Nov MATURE FEMALE TR01TY06 TURKEY RF N ' W ' 11-Nov MATURE FEMALE TR01TY07 TURKEY RF N ' W ' 11-Nov MATURE FEMALE TR01TY08 TURKEY RF N ' W ' 11-Nov MATURE FEMALE TR01TY09 TURKEY RF N ' W ' 17-Dec IMMATURE MALE TR01TY10 TURKEY RF N ' W ' 17-Dec IMMATURE MALE TR01TY11 TURKEY RF N ' W ' 18-Dec MATURE MALE TR01TY12 TURKEY RF N ' W ' 18-Dec MATURE MALE TR02TY01 TURKEY SC N ' W ' 18-Nov IMMATURE FEMALE TR02TY02 TURKEY SC N ' W ' 18-Nov IMMATURE ND TR02TY03 TURKEY SC N ' W ' 18-Nov MATURE FEMALE TR02TY04 TURKEY SC N ' W ' 18-Nov ND MALE TR02TY05 TURKEY SC N ' W ' 18-Nov ADULT MALE TR02TY06 TURKEY SC N ' W ' 18-Nov ND FEMALE TR02TY07 TURKEY SC N ' W ' 18-Nov JUVENILE MALE TR02TY08 TURKEY SC N ' W ' 19-Nov JUVENILE FEMALE TR02TY09 TURKEY SC N ' W ' 19-Nov JUVENILE FEMALE TR02TY10 TURKEY SC N ' W ' 19-Nov JUVENILE FEMALE TR02TY11 TURKEY SC N ' W ' 19-Nov JUVENILE FEMALE TR03TY01 TURKEY IP N ' W ' 10-Dec MATURE MALE RF: Reference sampling location; SC: Downstream sampling location near Smith s Crossing; IP: downstream sampling location near Imerman Park, ND: Not determined Age Sex A-2

28 Table A -3: Squirrel sample information Collection Date Body Weight (kg) Field ID Description Location Northing Westing TR01SL01 SQUIRREL RF N ' W ' 10-Dec IMMATURE FEMALE TR01SL02 SQUIRREL RF N ' W ' 10-Dec IMMATURE MALE TR01SL03 SQUIRREL RF N ' W ' 10-Dec IMMATURE MALE TR01SL04 SQUIRREL RF N ' W ' 10-Dec IMMATURE FEMALE 12-Dec SPRING MALE TR01SL05 SQUIRREL RF N ' W ' BORN TR01SL06 SQUIRREL RF N ' W ' 12-Dec IMMATURE FEMALE TR01SL07 SQUIRREL RF N ' W ' 12-Dec IMMATURE FEMALE TR01SL08 SQUIRREL RF N ' W ' 12-Dec ND ND TR01SL09 SQUIRREL RF N ' W ' 12-Dec IMMATURE FEMALE TR01SL10 SQUIRREL RF N ' W ' 22-Dec IMMATURE FEMALE 22-Dec SUMMER MALE TR01SL11 SQUIRREL RF N ' W ' BORN 22-Dec SPRING MALE TR01SL12 SQUIRREL RF N ' W ' BORN TR02SL01 SQUIRREL SC N ' W ' 21-Nov IMMATURE FEMALE TR02SL02 SQUIRREL SC N ' W ' 21-Nov IMMATURE FEMALE TR02SL03 SQUIRREL SC N ' W ' 09-Dec SUBADULT FEMALE TR02SL04 SQUIRREL SC N ' W ' 09-Dec IMMATURE MALE TR02SL05 SQUIRREL SC N ' W ' 09-Dec IMMATURE MALE 27-Dec SPRING MALE TR02SL06 SQUIRREL SC N ' W ' BORN TR02SL07 SQUIRREL SC N ' W ' 27-Dec IMMATURE FEMALE TR02SL08 SQUIRREL SC N ' W ' 27-Dec IMMATURE FEMALE TR02SL09 SQUIRREL SC N ' W ' 12-Jan IMMATURE FEMALE TR02SL10 SQUIRREL SC N ' W ' 14-Jan SUBADULT FEMALE TR03SL01 SQUIRREL IP N ' W ' 04-Dec IMMATURE FEMALE TR03SL02 SQUIRREL IP N ' W ' 04-Dec MATURE MALE TR03SL03 SQUIRREL IP N ' W ' 04-Dec IMMATURE MALE TR03SL04 SQUIRREL IP N ' W ' 04-Dec ADULT MALE TR03SL05 SQUIRREL IP N ' W ' 04-Dec IMMATURE FEMALE TR03SL06 SQUIRREL IP N ' W ' 04-Dec IMMATURE FEMALE TR03SL07 SQUIRREL IP N ' W ' 05-Dec IMMATURE MALE 17-Dec EARLY MALE TR03SL08 SQUIRREL IP N ' W ' SUMMER TR03SL09 SQUIRREL IP N ' W ' 17-Dec IMMATURE FEMALE TR03SL10 SQUIRREL IP N ' W ' 22-Dec IMMATURE FEMALE RF: Reference sampling location; SC: Downstream sampling location near Smith s Crossing; IP: downstream sampling location near Imerman Park, ND: Not Determined Age Sex A-3

29 Bay Midland Saginaw Legend ,000 Feet Deer Collection Locations Figure A-1. White-tailed deer sampling locations at site 1 (upstream reference). A-4

30 Bay Midland Saginaw Legend ,500 Feet Deer Collection locations Figure A-2. White-tailed deer sampling locations at site 2 (near Smith's Crossing). A-5

31 Bay Midland Saginaw Legend ,000 Feet Deer Collection locations Figure A-3. White-tailed deer sampling locations at site 3 (near Imerman Park). A-6

32 Bay Midland Saginaw Legend Feet Turkey Collection Locations Figure A-4. Wild turkey sampling locations at site 1 (upstream reference). A-7

33 Bay Midland Saginaw Legend Turkey Collection Locations 0 1,000 2,000 Feet Figure A-5. Wild turkey sampling locations at site 2 (near Smith's Crossing). A-8

34 Bay Midland Saginaw 01 Legend ,000 Feet Turkey Collection Locations Figure A-6. Wild turkey sampling locations at site 3 (near Imerman Park). A-9

35 Bay Midland Saginaw Legend Feet Squirrel Collection Locations Figure A-7. Squirrel sampling locations at site 1 (upstream reference). A-10

36 Bay Midland Saginaw Legend ,500 Feet Squirrel Collection Locations Figure A-8. Squirrel sampling locations at site 2 (near Smith's Crossing). A-11

37 Bay Midland Saginaw Legend ,000 Feet Squirrel Collection Locations Figure A-9. Squirrel sampling locations at site 3 (near Imerman Park). A-12