Grantee Research Project Results
Final Report: Prevalence and Distribution of Genotypes of Cryptosporidium Parvum in Feedlots in the Western United States
EPA Grant Number: R828038Title: Prevalence and Distribution of Genotypes of Cryptosporidium Parvum in Feedlots in the Western United States
Investigators: Atwill, Edward R. , Epperson, William B. , Grotelueschen, D. M. , Carpenter, L. V. , Smith, B. , Hoar, Bruce , Danaye-Elmi, Cyrus , Brewster, D.
Institution: University of California - Davis , University of Nebraska at Lincoln , South Dakota State University
Current Institution: University of California - Davis , South Dakota State University , University of Nebraska at Lincoln
EPA Project Officer: Hahn, Intaek
Project Period: April 1, 2000 through March 31, 2002 (Extended to February 28, 2002)
Project Amount: $248,461
RFA: Drinking Water (1999) RFA Text | Recipients Lists
Research Category: Drinking Water , Water
Objective:
The overall objectives of this research project were to: (1) quantify the fecal concentration of genotypes of Cryptosporidium parvum in United States feedlot cattle to estimate the oocyst loading rate for this sector of animal agriculture; (2) establish the prevalence and distribution of various genotypes of C.parvum in populations of feedlot cattle; and (3) identify management and animal risk factors that are associated with feedlot cattle shedding one or more genotypes of C. parvum.
Summary/Accomplishments (Outputs/Outcomes):
We sampled 22 feedlots across the Central and Western United States, resulting in 5,274 cattle being examined for C. parvum oocysts. States that participated in our study were California, Washington, Colorado, Oklahoma, Texas, Nebraska, and South Dakota (see Table 1). We stratified our feedlot sampling across season so that cattle were sampled at different climate regimes. Within each feedlot, we targeted about 240 cattle for sampling, stratified across two to six pens of cattle, with one pen being recent arrivals (nutrition primarily forages), one pen of cattle nearing their slaughter date (nutrition primarily concentrates), and the remaining pens randomly selected across the feedlot. This sampling scheme would allow us to identify, if present, possible environmental and management risk factors for C. parvum infection in feedlot cattle.
From our sample of 5,274 cattle, only 9 cattle (0.17 percent) were shedding detectable levels of oocysts as determined by direct immunofluorescent microscopy (DFA) (see Table 1) (Pereira, 1999). A primary concern we had with this low apparent prevalence was that DFA might not have reliably detected cattle shedding at very low concentrations of oocysts. The sensitivity of this assay was determined by using statistical methods outlined in Atwill, et al. (2003). Briefly, two negative fecal samples from every other feedlot were spiked with purified field-strain dairy calf C. parvum (bovine genotype A). Poisson regression was used as the statistical model, with animal identification as a Gaussian random effect for adjusting standard errors and coefficients for intracow dependencies on DFA performance (Hardin and Hilbe, 2001). Negative binomial regression did not provide a significantly better fit compared to the Poisson model for these data. Therefore, t est sensitivity for a sample with c oocysts/gram feces, S(c), was defined as the probability of detecting one or more oocysts per assay given that oocysts were present. This is equivalent to suggesting that sensitivity is equal to 1 minus the probability that zero oocysts were detected. For the Poisson regression model, S(c) = 1 - Q - rcW , where r is the percent recovery of the assay, c is the concentration of oocysts in the fecal sample, and W is the mass of the fecal smear (Hoar, 2000; Atwill, et al., 2003). We observed a mean percent recovery (r) of 42.2 percent (95 percent confidence interval, 39.0-45.6 percent) and a mean fecal smear weight (W) of 11.7 mg. The concentration of oocysts detected with a 90-percent probability (DT 90) was 465 oocysts/g feces, and the DT 50 was 140 oocysts/g feces. This suggests that cattle shedding fecal concentrations of oocysts in excess of 400 oocysts/g were reliably detected, while cattle shedding oocyst concentrations of less than 100 were likely missed by DFA (false negative) .
To determine an estimate of the number of false negatives in our data due to infected cattle shedding low concentrations of oocysts, 10 randomly chosen negative fecal samples from each feedlot (n = 219) were screened for C. parvum using immunomagnetic separation followed by direct immunofluorescent microscopy (IMS-DFA) (Pereira, 1999; Atwill and Periera, 2003a). To our knowledge, this method is the most sensitive method published to date on detecting C. parvum oocysts in adult bovine fecal material. The DT 90 was 2.4 oocysts/g feces; DT 50 was 0.7 oocysts/g feces; and there was a 64 percent probability of detecting 1 oocyst/ g feces. Only 2 out of 219 (0.91 percent) DFA-negative fecal samples contained oocysts, suggesting a low false negative proportion in our data (see Table 1).
Using these diagnostic results, we can estimate the true-point prevalence of C. parvum shedding in our study population of feedlot cattle from the Central and Western United States. The apparent-point prevalence based on DFA was 0.17 percent. The false negative proportion for DFA (using IMS-DFA as the gold standard) was 0.91 percent; the estimated total number of false negatives in our sample of 5,265 DFA negative cattle would be 48. Therefore, the true number of infected cattle in our study population would be 57 (9 + 48), leading to a true-point prevalence of 1.1 percent (57/5,274). This estimate for the true-point prevalence assumes that the specificity of our method for DFA was approximately 100 percent. This assumption seems reasonable, given the following two observations:
- IMS-DFA is approximately 2-log 10 more sensitive than DFA (Pereira, 1999), making the probability of a fecal sample to be positive for DFA yet negative for IMS-DFA P(DFA+|IMS-DFA - ) very unlikely (P » 0). Specificity (Sp) is defined as the probability of being test negative when the sample is truly negative, P(T - |D - ). We are using IMS-DFA as our gold standard for the true shedding status at a point in time (point prevalence). Hence, Sp = P(T - |D - ), so 1-Sp = P(T+|D - ), analogous in our study to P(DFA+|IMS-DFA - ). The P(DFA+|IMS-DFA - ) is approximately 0, leading to Sp approximately 1.
- Out of the nine DFA+ animals observed in this study, four had sufficient oocysts to allow molecular confirmation as C. parvum. This leaves five DFA+ samples without molecular confirmation, and therefore, possible false positives (T+|D - ). Using the binomial distribution to calculate probabilities of occurrence (Atwill, et al., 2003), the probability of observing five or less false positive samples if the Sp were £ 99.6 percent was P £ 0.0003. Hence, the Sp was likely to be greater than 99.6 percent. In fact, the probability of observing five or less false positive samples if the Sp was 99.8 percent or 99.9 percent was P = 0.049 and 0.52, respectively. These results indicate that Sp was very likely in excess of 99.9 percent, or effectively 100 percent.
Knowing the true-point prevalence of 1.1 percent and the percent recovery of our DFA and IMS-DFA methods, we estimated the environmental loading rate of C. parvum per feedlot animal. The arithmetic and geometric mean concentration of C. parvum oocysts for the nine DFA positive cattle was 1.40 ´10 6 and 4.46 ´10 5 oocysts/kg feces, respectively. The arithmetic and geometric mean concentration of C. parvum oocysts for the two IMS-DFA positive cattle was 1.29 ´10 5 and 1.69 ´10 4 oocysts/kg feces, respectively. The overall weighted arithmetic mean from both diagnostic methods was 3.29 ´10 5 oocysts/kg feces for positive cattle and 3.55 ´10 3 oocysts/kg feces for all 5,274 cattle in the study. The overall weighted geometric mean from both diagnostic methods was 8.47 ´10 4 oocysts/kg feces for the positive cattle and 915 oocysts/kg feces for all 5,274 cattle in the study. If we assume that feedlot steers produce between 20 and 40 kg of feces per day (American Society of Agricultural Engineers, 1992), then the arithmetic mean environmental loading rate of C. parvum oocysts by feedlot steers in this study would range from 7.11 ´10 4 to 1.42 ´10 5 oocysts/steer/day (see Table 2). The geometric mean environmental loading rate of C. parvum oocysts by feedlot steers would range from 1.83 ´10 4 to 3.66 ´10 4 oocysts/steer/day.
Of the nine DFA+ and two IMS-DFA+ samples, four isolates had sufficient numbers of oocysts to allow molecular confirmation as C. parvum. Using a nested polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique by Xiao, et al. (1999), the genotype of these isolates was bovine genotype A.
Table 1. Number of Feedlot Cattle Shedding Detectable Levels of C. parvum Oocysts From Central and Western United States
State | No. of Feedlots | No. of Cattle Shedding Detectable C. parvum a | No. of False Negative Cattle b | PCR-RFLP Genotype c |
CA |
4 |
0/960 |
0/39 |
|
WA |
1 |
2/240 (0.8 %) |
0/10 |
Bovine A |
TX |
3 |
0/711 |
0/30 |
|
OK |
3 |
0/722 |
1/30 |
|
CO |
4 |
6/957 (0.6 %) |
0/40 |
Bovine A |
SD |
4 |
1/964 (0.1 %) |
1/40 |
Not done |
NE |
3 |
0/720 |
0/30 |
|
TOTAL |
22 |
9/5,274 (0.2 %) |
2/219 (0.9 %) |
a DFA used as the diagnostic procedure (Pereria, 1999).
b Ten fecal samples per feedlot that were negative on DFA were reexamined using immunomagnetic separation followed by DFA (Periera, 1999).
c Nested PCR followed by RFLP on the 18S SSU rRNA gene (Xiao, et al., 1999).
Table 2. Estimated Mean Environmental Loading Rate a of C. parvum Oocysts by Feedlot Cattle
in the Central and Western United States (oocysts/animal/day)
Kg Feces Produced/Animal/Day |
|||
20 |
30 |
40 |
|
Arithmetic Mean |
71,085 |
106,628 |
142,171 |
Geometric Mean |
18,309 |
27,464 |
36,618 |
aEnvironmental loading rate (Atwill, et al., 2003) calculated as the mean intensity of oocysts observed in positive cattle (number of oocysts/kg feces), multiplied by the true-point prevalence of infection (using IMS-DFA as the gold standard for shedding status), multiplied by the estimated daily production of feces per feedlot steer (ASAE, 1992).
C. parvum has emerged as a common waterborne microbial pathogen, with specific strains or genotypes readily transmitted between livestock and humans. One of the first steps in designing a watershed management program for minimizing the occurrence of C. parvum in drinking water supplies is to identify significant quantitative sources of this parasite. Adult cattle often are considered to be potential sources of environmental contamination for C. parvum, but there is disagreement over the relative importance that adult cattle have in loading watersheds with significant amounts of C. parvum oocysts.
The objective for this project was to quantify the rate of environmental loading of C. parvum oocysts by feedlot cattle from across Central and Western United States. Whether one is designing a fecal waste management system for sufficient log reduction of C. parvum or calculating a total maximum daily load for C. parvum on a watershed, we must have reliable estimates of quantitative shedding of C. parvum from biological sources if our management objectives for C. parvum are to be achieved. We found a point prevalence of 1.1 percent of feedlot cattle to be shedding C. parvum oocysts in their feces. This low prevalence was adjusted for the possibility of false negatives. This study population was from a wide range of climates and feedlot management systems, none of which apparently functioned as risk factors for elevated infection. Feedlot cattle produced about 3.55 ´10 3 oocysts/kg feces, or about 7.11 ´10 4 to 1.42 ´10 5 oocysts/animal/day. Although these loading rates are far less than some other biological sources of oocysts on a per-animal basis, the high spatial densities of cattle on feedlot systems in combination with the low infectious dose of bovine C. parvum for humans ( Okyhusen, 1999) suggests that it would be prudent for feedlot operators to insure that off-site waterborne transport of infective oocysts be minimized using beneficial management practices designed to reduce overland and subsurface transport of oocysts and their survivability.
References:das Graças CPM, Atwill ER, Jones T. Comparison of sensitivity of immuno-fluorescent microscopy to that of a combination of immunomagnetic separation and immuno-fluorescent microscopy for detection of Cryptosporidium parvum oocysts in adult bovine feces. Applied and Environmental Microbiology 1999;65:3236-3239.
Atwill ER, Hoar B, das Graças CPM, Tate KW, Rulofson F, Nader G. Improved quantitative estimates of low environmental loading and sporadic periparturient shedding of Cryptosporidium parvum in adult beef cattle. Applied and Environmental Microbiology 2003;69:4604-4610.
Hardin J, Hilbe J. Clustered data. In: Generalized Linear Models and Extensions. Stata Press: College Station, TX, 2001, pp.193-204.
Atwill ER, das Graças CPM. A lack of detectable environmental loading of Cryptosporidium parvum in periparturient dairy cattle. Journal of Parasitology (in press, 2003).
American Society of Agricultural Engineers. Manure production and characteristics. In: American Society of Agricultural Engineers Standards D384.1. Agricultural Engineering Yearbook, St. Joseph, MI, 1992, pp. 485-487.
Xiao L, Morgan UM, Limor J, Escalante A, Arrowwood M, Shulaw W, Thompson RCA, Fayer R, Lal AA. Genetic diversity within Cryptosporidium parvum and related Cryptosporidium species. Applied and Environmental Microbiology 1999;65:3386-3391.
Atwill ER, Hou L, Karle BM, Harter T, Tate KW, Dahlgren RA. Transport of Cryptosporidium parvum through vegetated buffer strips and estimated filtration efficiency. Applied and Environmental Microbiology 2002;68:5517-5527.
Journal Articles on this Report : 1 Displayed | Download in RIS Format
Other project views: | All 9 publications | 2 publications in selected types | All 2 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Atwill ER, Pereira MDGC, Alonso LH, Elmi C, Epperson WB, Smith R, Riggs W, Carpenter LV, Dargatz DA, Hoar B. Environmental load of Cryptosporidium parvum oocysts from cattle manure in feedlots from the central and western United States. Journal of Environmental Quality 2006;35(1):200-206. |
R828038 (Final) |
Exit Exit |
Supplemental Keywords:
protozoa, bovine, agricultural environment, agriculturally impacted watershed, anthropogenic stress, aquatic ecosystem, bacteria, comprehensive nutrient management program, Cryptosporidium parvum, oocysts, dairy farms, drinking water system, drinking water contaminants, drinking water distribution system, environmental monitoring, exposure, exposure and effects, fecal contamination, feedlot cattle, genotype distribution, immunofluorescent assay, microbial contamination, microbial risk assessment, monitoring, nutrient management, public health, water quality, water quality parameters, waterborne disease., RFA, Scientific Discipline, Water, Nutrients, Environmental Chemistry, Health Risk Assessment, Biochemistry, Environmental Monitoring, Drinking Water, cryptosporidium parvum oocysts, Safe Drinking Water, microbial contamination, aquatic ecosystem, agriculturally impacted watershed, anthropogenic stress, monitoring, microbial risk assessment, feedlot cattle, water quality parameters, waterborne disease, genotype distribution, bacteria, exposure and effects, DNA sequencing, fecal contamination, zoonotic parasite, other - risk assessment, exposure, drinking water distribution system, public health, comprehensive nutrient management program, dairy farms, agricultrual environment, water quality, drinking water contaminants, immunofluorescent assay, drinking water systemProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.