Microbial Contamination in North Carolina Waters: Microbial Source Tracking Database Development and Method Validation

<p>NON-TECHNICAL SUMMARY:<br/> According to North Carolina's 2004 Draft Integrated Report, about 8,760 acres of Class A primary-contact-recreation marine waters are listed as impaired for overall use. Eighty-one percent (462) of closing/advisory days in 2003 were due to monitoring that revealed elevated bacteria levels from "unknown" sources. About 30 miles and 35,560 acres of estuarine waters are listed as impaired for shellfish harvesting and are permanently closed due to elevated levels of fecal coliforms. Furthermore, the 2000 Census ranked North Carolina fourth in the nation for an increase in population growth. Urbanization along the land-seawater interface is considered to have significant environmental consequences. A NC coastal study found that of five estuaries in urbanizing watersheds in NC, the most important anthropogenic factor associated with fecal coliform
abundance was percentage watershed impervious surface coverage, which consists of roofs, roads, driveways, sidewalks, and parking lots. Many inland areas in North Carolina are greatly impacted by host sources found in agriculture. North Carolina ranks second in the nation for both swine and turkey livestock inventories and fourth in the nation for broilers and other meat-type chicken livestock inventories. If the manure from these industries isn't managed properly, pathogens from the manure may be disseminated in the environment potentially contaminating surface water, groundwater and crops. Contamination of crops may occur directly during land application or indirectly by contact with irrigation water contaminated with manure. Irrigation water may be contaminated following recharge to groundwater, run-off from land to rivers or streams, and through other groundwater/surface water
interactions by water contaminated with manure. Without specific host source information the ability for regulatory agencies and managers to protect public health and remediate pollution sources is compromised. Microbial source tracking (MST) methods have been developed to discriminate between animal and human fecal pollution sources. Thus far there has been a lack of integrated study approaches with the purpose of combining quantitative MST with watershed hydrology and traditional monitoring parameters that can be associated with food safety. The proposed research will entail validation testing on a range of MST methodologies using various environmental samples from North Carolina coastal and inland areas. The fundamental purpose of this study is to develop a practical strategy involving validated methods for determining the source(s) of fecal pollution in coastal and inland areas while
also assessing factors associated with water-quality and produce safety. Results should contribute to quantitative guidelines for science-based good agricultural practices and best management practices to protect surface water from pathogen contamination from different sources within the watershed.
<p>APPROACH:<br/> Fecal indicator bacteria (FIB) from host sources in the watersheds will be collected for antibiotic resistance analysis (ARA) library development, speciation and evaluation for specific genetic markers. A series of blind tests will be conducted on the FIB to determine repeatability (isolates already in the library or database for each method), accuracy (isolates from different sewage or scat samples not previously tested), and robustness (isolates from sources outside the watershed). Field tests associated with validation will involve a waterbody suspected of having human contamination and one suspected of livestock contamination, but the actual sources are unknown. Water and sediment samples will be collected quarterly for a period of three years from the two representative waterbodies. In addition to ARA and fluorometry as tools for source
identification, the human-specific BacH and the ruminant-specific BacR methods will be used for source determination, as well as, markers for Methanobrevibacter smithii, Ent. faecium and human polyomaviruses. The prevalence of pathogens on a concentrated animal feeding operation (CAFO) farm will be determined using water samples collected from irrigation water distribution lines and lagoons quarterly for three years for determination of FIB, Enterococcus Spp., E. coli O157 and non- O157. Survival in soil studies will involve the collection of soil samples from six locations on the farm to identify localized isolates and determine their distributions at the study site. Isolates will be characterized for genes responsible for resistance to different antibiotics in FIB using multiplex PCR for three beta lactamase encoding genes, 14 TE resistance-encoding alleles [tet(A), tet(B), tet(C),
tet(D), tet(E), tet(G), tet(K), tet(L), tet(M), tet(O), tet(S), tetA(P), tet(Q), and tet(X)], and 3 sul alleles (sul1, sul2, and sul3) etc. Repetitive element PCR will be conducted to assess the genetic diversity of E. coli and Enterococcus species and to analyze strain transmission from host to the environment. Statistical modeling approaches will be used to determine whether there are specific patterns of distribution for the individual markers that are located in the farm environment what variables are associated with the possible modes of movement, survival, and distribution around the source (feedlot, animal production system, manure handling facility) and compare this to the distribution of pathogens. Validated MST methods coupled with the knowledge of persistence, and variables associated with the possible modes of movement of pathogens will be useful to producers, managers of
watersheds, and regulators in the development of science-based good agricultural practices (GAPs) and best management practices (BMPs) to protect fresh produce and surface water from pathogen contamination from different sources within the watershed.
<p>PROGRESS: 2011/10 TO 2012/09<br/>OUTPUTS: Microbial source tracking and fate/transport research conducted during this project have been disseminated to the scientific community through journal publications listed below, presentations at the 2012 Land Grant/Sea Grant National Water Conference, and the 2012 American Society for Microbiology General Meeting. PARTICIPANTS: Nothing significant to report during this reporting period. TARGET AUDIENCES: Provides information for use by other scientists, state and local policymakers, students, and special interest groups. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.
<p>PROGRESS: 2010/10/01 TO 2011/09/30OUTPUTS: Outreach activities involved presenting information to program participants at organizational meetings and training workshops. For example, on November 4, 2011, I presented "Fecal TMDL: Pathogen Indicators, Present and Future" to the Municipal Association of South Carolina (MASC): South Carolina Association of Stormwater Managers (SCASM) during their Fourth Quarter Meeting in Columbia, SC. I also presented "Using Harvested Rainwater-Real vs. Perceived Health Risks" to participants of the Rainwater Harvesting Workshop - 2011: Moving our Communities toward More Sustainable Water Use. This event was sponsored by the NC Coastal Training Program and NC Cooperative Extension on October 18, 2011 in Jacksonville, NC. Copies of presentation materials were distributed to the participants on site during the meeting. PARTICIPANTS:
Mark Ibekwe, a microbiologist with the USDA Agricultural Research Service has collaborated in efforts to determine the relationships between manure management, populations of human pathogens and antibiotic resistant bacteria in soil and water environments. TARGET AUDIENCES: Nothing significant to report during this reporting period. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.
PROGRESS: 2009/10/01 TO 2010/09/30<br/>OUTPUTS: Four MST methods (ARA, rep-PCR, optical brightener detection, and human specific Bacteroides HF183 marker detection) for identifying sources of fecal pollution were assessed in Oyster Creek, located in Hyde County, NC. This restricted shellfish harvesting area is impaired by levels of bacteria exceeding North Carolina's water quality standards for fecal coliform, which has resulted in closure of the waterbodies to shellfish harvesting. A variety of data at the watershed scale, including shoreline sanitary survey data and Geographic Information Systems (GIS) data coverage, were also used to identify potential fecal coliform contributions. There are no permitted point source facilities in the restricted shellfish area addressed in this report. Antibiotic resistance analysis was performed on 5,088 host-origin E. coli
isolates and 10,368 E. coli isolates from 216 surface water samples. Host-origin isolates were collected throughout the study for library development. The application of stringent statistical measures for library development resulted in a final ARA library composed of 1,065 E. coli isolates with a 90% average rate of correct classification (ARCC). According to ARA, isolates identified as wildlife (54-96%) represented the predominant host source of E. coli. The 1,065 isolates used in the ARA library were subjected to repetitive element polymerase chain reaction (rep-PCR) as a means for cross validation of the ARA results. A subset of E. coli isolates (1,728) from the surface waters of Oyster Creek were evaluated by rep-PCR. Similar to the ARA method, rep-PCR implicated wildlife (57-100%) as the predominant host source E. coli isolates in Oyster Creek. Four MST methods (ARA, rep-PCR,
optical brightener detection, and human specific Bacteroides HF183 marker detection) implicated a human contribution at eight of the 22 sampling stations. The combined Bayesian statistical method and finite difference transport modeling approaches was used to estimate fecal coliform load from watersheds and to develop TMDLs. The TMDLs developed for the restricted shellfish harvesting area of Oyster Creek for fecal coliform load are as follows: The fecal coliform TMDL = 2.3 X 10 11 counts per day. The load reduction needed to meet the shellfish criterion is 92.6%. The reduction established based on the 90th percentile criterion ensures that the water body will meet water quality standards 90% of the time. Management strategies to meet the proposed reduction will be implemented on a daily basis, to achieve the control of fecal loads for all but the most extreme 10% of events (i.e. ensure
that 90% of the concentrations are at or below the 90th percentile criterion). These extreme events are often caused due to hydrologic variability, storm water management, change of land use practices, and change of wildlife activities during the previous ten year period. Source reductions can be assigned by first managing controllable sources (human, livestock, and pets) and then determining if the TMDL could be achieved. If the total required reduction was not achieved, then the wildlife source can be considered to be reduced. PARTICIPANTS: Jay Levine, Population Health and Pathobiology Department, North Carolina State University College of Veterinary Medicine: Jian Shen, Virginia Institute of Marine Science, College of William and Mary: Yuan Zhao, Virginia Institute of Marine Science, College of William and Mary. TARGET AUDIENCES: Nothing significant to report during this reporting
period. PROJECT MODIFICATIONS: Nothing significant to report during this reporting period.
<p>PROGRESS: 2008/10/01 TO 2009/09/30<br/>OUTPUTS: Three source tracking methods for identifying sources of fecal pollution were assessed in a segment of the North River, Carteret County, North Carolina. In all, 306 water samples were collected from 34 stream sites (representing estuary, mixing and tidal drainage ditch locations) and the abundance of fecal indicator bacteria exceeded regulatory standards for designated recreational waters in all samples. Antibiotic resistance analysis was performed on 14,688 E. coli isolates from the 306 water samples and 4,320 E. coli isolates from 90 sediment samples. Overall, 47% of the 6,480 E. coli isolates from water collected from estuary and mixing locations (combined) were identified as human origin, while 61% of the 1,584 E. coli isolates from those sediments were identified as human origin. On average, 48% of the 8,208 E. coli
isolates from water collected from tidal ditch locations were identified as human origin, while 51% of the 2,736 E. coli isolates from the ditch sediments were identified as human. Water samples (n=120) were evaluated by fluorometry for optical brighteners (OBs). On average, 30 of the 34 sampling locations were positive for OBs with ?100 fluorometric units (FU) before UV light exposure. After 4 hours of UV exposure, 26 of the 30 sampling locations were confirmed as positive for OBs with a ? 30% relative decrease in FU. Water samples (n=120) were assayed by PCR analysis for the human-specific Bacteroides HF183 marker. The HF183 marker was detected in five samples. The low incidences of the human marker were surprising given the ARA and OB data suggested a significant human contribution. However, the low incidences may be explained by PCR inhibition. The results from this study
suggest a chronic, persistent allocation of isolates of human origin. The high human source identifications of E. coli in sediment samples suggest that sediments are a reservoir of these organisms and may contribute to the findings in the water column. On a swine farm in Sampson County, NC, Escherichia coli isolates were recovered from swine manure and water samples by basic microbiological culture and IDEXX Colilert methods, respectively. Antibiotic resistance genes were identified from the isolates using the polymerase chain reaction (PCR) method. On average, E.coli counts in the surface water were 272.1 CFU/100 mL, 10 fold higher than that in ground water (21.1 CFU/100 mL). A total of 1,208 E. coli isolates from swine feces, lagoon effluent, cattle, deer, dog, bird and nearby ground and surface waters (n=238, 234, 192, 48, 48, 48, 200, and 200, respectively) were evaluated for ARGs. A
total of 5909 E. coli isolates were evaluated for phenotypic expression of resistance to various concentrations of eight antibiotics. Genotypic evaluation indicated the presence of aadA, strA, strB, tetA, tetB, tetC, sul1, sul2, sul3, and Aac(3)IV ARGs in all the sources of isolates. The database developed from antibiotic resistance patterns (ARP) and antibiotic resistant genes (ARGs) of E. coli isolates from the known sources were able to associate fecal bacteria in ground and surface water to lagoon effluent, livestock and wildlife. PARTICIPANTS: Dr. Daniel Israel, Participant in evaluation of swine lagoon effluent. NC WRRI, financial support for swine lagoon effluent study. Dr. David Lindbo, Participant in North River Study. Z. Smith Reynolds Foundation, financial support for North River Study. TARGET AUDIENCES: Not relevant to this project. PROJECT MODIFICATIONS: Not relevant to this
project.
<p>PROGRESS: 2007/10/01 TO 2008/09/30<br/>OUTPUTS: The use of antibiotics in animals is suspected to be a major route for transfer of antibiotic resistance genes (ARGs) to human pathogens, although different antibiotics are used in animals than in humans. North Carolina has the second largest swine production industry in the USA which is concentrated in a small geographical area in the southeastern part. This may increase the risk of antibiotic resistant bacteria reaching nearby surface waters. The goal of this year's work was to identify and quantify ARGs in Escherichia coli isolates from swine feces, lagoon effluent, and ground and surface waters from a commercial swine facility. Escherichia coli isolates were recovered from manure and water samples by basic microbiological culture and IDEXX Colilert methods, respectively. Antibiotic resistance genes were identified
from the isolates using the polymerase chain reaction (PCR) method. A total of 848 E. coli isolates from swine feces, lagoon effluent, and nearby ground and surface waters (n=238, 234, 193, and 183, respectively) were recovered. All the E. coli isolates were evaluated for phenotypic expression of resistance to various concentrations of the following antibiotics: erythromycin, neomycin, oxytetracycline, streptomycin, tetracycline, cephalothin, apramycin, trimethoprim, and rifampicin. All the isolates displayed multiple antibiotic resistances. Genotypic evaluation indicated the presence of aadA, strA, strB, tetA, tetB, tetC, sul1, sul2, sul3, and aac(3)IV ARGs in all the sources of isolates. The number of ARGs in E. coli isolated from swine feces and lagoon effluent were not significantly different, while those from lagoon effluent were significantly (p?0.05) fewer than from ground
and surface water. The host source of the E. coli recovered from ground and surface waters will be identified by the microbial source tracking method, antibiotic resistance analysis (ARA) and related to the lagoon effluent application onto nearby fields. If the source proves to be lagoon effluent, this would support the need for improved mitigation strategies. In an effort to evaluate the targeted sampling approach to identifying sources of fecal pollution in surface water, a study was conducted on the North River, Carteret County, NC. The community along the North River developed on soils that are not suitable for conventional on-site systems, thus these systems may be a potential source of fecal coliforms in the North River. Surface water was collected monthly for a period of nine months. Well water sampling occurred 4 times over the duration of the study. Over 20,000 E. coli isolates
from the North River have been evaluated by antibiotic resistance analysis. The data analyzed to date suggest that human and wildlife are the predominant source of fecal pollution. PARTICIPANTS: Dr. Daniel Israel, Participant in evaluation of swine lagoon effluent. NC WRRI, financial support for swine lagoon effluent study. Dr. David Lindbo, Participant in North River Study. Z. Smith Reynolds Foundation, financial support for North River Study. TARGET AUDIENCES: Not relevant to this project. PROJECT MODIFICATIONS: Not relevant to this project.
<p>PROGRESS: 2006/10/01 TO 2007/09/30<br/>The E. coli and Enterococcus isolates previously evaluated by antibiotic reistance analysis have also been evaluated by PCR based on the 16S-23S rDNA Intergenic Spacer Region of Enteric Bacteria. Ribosomal DNA (rDNA) are sequences encoding ribosomal RNA (rRNA). The rRNA genetic locus, rrn, is sufficiently conservative and can be used in a universal organization of evolutionary relationships. PCR analysis of the 16S-23S rDNA intergenic spacer region (ISR) has been used to evaluate E. coli and Enterococcus isolates and is emerging as a tool in microbial source tracking research. The banding patterns of isolates from different sample sources were similar, but not always identical. Molecular weight of each band was analyzed and the percentages of isolates having bands at specific molecular weight regions were recorded The molecular
weight was divided into 16 regions, 50 bp increments from 100 bp fragments to 800 bp fragments. This was done because there were very few bands smaller than 100 bp and larger than 800 bp, and most 50 bp regions within 100 to 800 bp had only one band within them. Band lengths were quantified and converted to binary data in each region. Discriminant analysis of the binary PCR data generated from the analysis of E. coli revealed considerable similarities among banding patterns from the four sources (swine nursery feces, swine nursery lagoon liquid, swine finishing feces, and swine finishing lagoon liquid). The percentage of isolates which was correctly classified as finishing feces was high, 89%. But the percentages of the other three sources were low, all less than 70%. However, the results of classification of isolates by broad categories of nursery and finishing source categories were
better with a correct classification rate greater than 75%. Discriminant analysis of the binary PCR data generated from the analysis of Enterococcus also revealed considerable overlap among sources. The percentage of correct classification for nursery feces was large, close to 90%; the percentages of correct classification for other three sources were low, especially for the finishing lagoon, which was only 38.67%. The percentages of correct classification became larger if combine feces and lagoon together as broad source categories, which were > 80% for both nursery and finishing source category. The isolates evaluated by PCR are currently being evaluated for the presence of antibiotic resistance genes and virulence genes.
<p>PROGRESS: 2005/10/01 TO 2006/09/30Manure, lagoon and soil samples have been collected from swine nursery and finishing facilities. There were two sampling events, one in December 2005 and the other in July 2006 to represent a cool and warm season, respectively. The objectives are 1) to determine the effects of lagoon treatment on the persistence of antibiotic resistance in enteric bacteria; 2) to determine if there are genetic differences among the enteric bacteria and how it relates to the persistence of antibiotic resistance and 3) to determine if there is a seasonal variation of antibiotic resistance in enteric bacteria. Interestingly, neither E. coli nor Enterococcus was isolated from soil samples. Isolates analyzed for antibiotic resistance were collected from 13 samples in total (3 nursery manure samples, 3 nursery lagoon liquid samples, 4 finishing manure
samples, and 3 finishing lagoon liquid samples). Each manure sample represents a composite sample of manure collected from a number of swine in a particular swine house. For example, 3 nursery manure samples refers to 3 different swine houses at the nursery facility and each house represents a different stage of development (age) of swine. For each sample, 192 E. coli isolates and 192 Enterococcus isolates were collected for a total of 2496 E. coli isolates and 2496 Enterococcus isolates. The antibiotic treatments included those that have been commonly reported in microbial source tracking studies. These antibiotics include Amoxicillin, Cephalothin, Chlortetracycline, Erythromycin, Neomycin, Oxytetracycline, Streptomycin, Tetracycline, Rifampicin, and Vancomycin. E .coli and Enterococcus isolates displayed multiple antibiotic resistances in all the manure and lagoon samples. Isolates
displayed different antibiotic resistance patterns for manure samples collected from different swine houses. For nursery facility samples, there were lower percent of isolates that showed antibiotic resistance in lagoons than in the manure. For finishing facility samples, no such trend was apparent. For both E. coli and Enterococcus, percentages of resistant isolates were higher in nursery manure than in finishing manure. The resistance patterns were similar for the antibiotics which are in the same family. The percent of isolates resistant to tetracycline family were higher than other antibiotics. Three hundred isolates have been randomly selected for molecular evaluation using repetitive element polymerase chain reaction. Future work will involve data analysis to determine whether there is a relationship between the phenotypic and genotypic patterns of resistance. Additionally, the
data will be evaluated to determine if there is a seasonal impact on resistance.
<p>PROGRESS: 2004/10/01 TO 2005/09/30<br/>Manure and lagoon samples have been collected from a swine nursery and finishing facilities. In addition, soil samples have been collected from the spray fields at each facility. Enterococcus and Escherichia coli isolates have been isolated from the manure and lagoon samples. All isolates will be evaluated using antibiotic resistance analysis. The phylogenic diversity of the antibiotic resistant bacteria isolated from manure, lagoon, and soil samples will be determined using amplified ribosomal DNA restriction analysis. Additionally, microorganisms that are indigenous to the soil will be evaluated for antibiotic resistance to antibiotics that are commonly used in the swine industry.