Persistence of Mycobacterium avium subsp. paratuberculosis in soil, crops, and ensiled feed following manure spreading on infected dairy farms

Abstract

The goal of this study was to determine the persistence of Mycobacterium avium subsp. paratuberculosis (MAP) in soil, crops, and ensiled feeds following manure spreading. This bacterium was often found in soil samples, but less frequently in harvested feeds and silage. Spreading of manure on fields used for crop harvest is preferred to spreading on grazing pastures.

Johne’s disease (JD) is a costly, progressive, and ultimately fatal infection of cattle and other ruminants caused by Mycobacterium avium subsp. paratuberculosis (MAP). In cattle, the infection is characterized by a protein-losing enteropathy leading to chronic diarrhea and weight loss in the later stages of the disease (1). Infected cattle can shed billions of MAP organisms in their feces every day, contributing to considerable environmental contamination, which leads to the spread of the disease within the herd (2). This bacterium is frequently found in manure lagoons on infected farms, and studies have shown that MAP survives under various environmental conditions for up to 1 y (3–4). These conditions include manure storage systems, such as liquid slurry and manure, soil, and water (5). Therefore, disposal of manure on infected farms can affect the potential for contamination of pasture and/or crops intended for feed. In fact, the impact of manure spreading on the transmission of JD has been identified by the National Academy of Science, and by the United States Animal Health Association Johne’s Disease Committee as one of the most important “knowledge gaps” facing the industry (6–7). It is suspected that MAP is more likely to contaminate feed when manure is deposited directly on the plant, for example in grazing pastures, hay, or “green chop” compared with spread on cropland prior to planting. In one study performed in a laboratory setting using a small-scale fermentation system, fermentation was shown to adequately inhibit MAP growth (8). In another study performed under controlled field conditions, and looking at the impact of soil type and the amount of rainfall on the recovery of MAP in grass and various soil layers following the application of MAP-spiked manure, MAP was found to only remain on the grass and the upper layers of pasture soil, and never within the deeper soil layers, representing a clear infection hazard for grazing livestock (9). To the authors’ knowledge, the survival of MAP in crops, soil, and ensiled feed after spreading of manure on dairy farms naturally infected with JD has not been tested in a field setting.

The objective of this study was to determine the persistence of MAP in samples of soil, crops, and ensiled feeds following manure spreading on farms with JD. We hypothesized that on farms naturally infected with JD, MAP will persist on crops grown where manure has been spread. We also hypothesized that ensiled feeds will have less persistence of MAP than unpreserved forages such as grass, “green chop,” or hay.

This study was conducted on 10 Pennsylvania dairy farms, located primarily in southeastern and central Pennsylvania, and for which whole herd mycobacterial fecal cultures and on-farm environmental cultures from the previous year were available. Farms were enrolled if they were JD-positive, and if they spread manure on fields used for the production of cattle feed. On-farm samples were obtained at 4 time periods: 1 — manure samples from the storage area before it was spread onto the fields; 2 — soil samples approximately 1 mo following manure spreading; 3 — crops (freshly cut forage) sampled approximately 2 to 4 mo following manure spreading; and 4 — silage samples following the ensiling period (3 to 9 mo). Samples were obtained by one of the authors, or by farmers or their veterinarians at each time period and shipped to the laboratory overnight for processing. All samples were frozen at −70°C upon arrival at the laboratory, and processed later in batches. All samples were processed for MAP detection using 2 techniques: standard Herrold’s Egg Yolk medium (HEYM) culture, and real-time polymerase chain reaction (RT-PCR).

Mycobacterial cultures were processed according to standard 3-day double incubation technique as recommended by the National Veterinary Services Laboratory (NVSL) (10). For the manure and soil samples, 2 g of the sample were placed into 35 mL of sterile water in a conical tube, rocked for 30 min, and allowed to settle for 30 min. Forage and feed samples (2 g) were placed into a sterile container and 35 mL of sterile water were added prior to stomaching for 2 min. The liquid from the processed forage and feed sample was poured into a sterile 50 mL conical tube and allowed to settle for 30 min. For all sample types, 20 mL of supernatant were transferred into a sterile container and saved for RT-PCR testing. Also for all sample types, 5 mL of supernatant were transferred to a 50 mL conical tube containing 25 mL of brain heart infusion (BHI)/hexadecylpyridinium chloride (HPC) and incubated at 37°C overnight. On the second day each tube was spun for 30 min at 900 × g and then decanted. The pellet was resuspended in 1 mL of antibiotic solution (amphotericin B, nalidixic acid, and vancomycin) and incubated at 37°C overnight. On the third day the tube was vortexed and 100 μL were added to 4 HEYM tubes. Cultures were read every 2 wk for 16 wk, at which time the number of MAP colony-forming units (CFU) per tube was recorded. The mean CFU/tube was then converted to a mean CFU/g of sample.

For RT-PCR, a commercially available RT-PCR kit (Tetracore VetAlert; Tetracore, Rockville, Maryland, USA) was used. Manure samples were processed according to the manufacturer’s recommendations for DNA extraction and amplification (Smart Cycler; Cepheid, Sunnyvale, Caliornia, USA) as described elsewhere (11). For feed and soil samples, the MAP suspension from the stomacher phase was subjected to the DNA extraction protocol and DNA samples were run in duplicate wells in the thermal cycler. The number of cycles to positive threshold (Ct) was recorded, with a Ct value > 42.0 indicating a negative result. A sample was considered positive if at least 1 well was positive. If both wells were positive, the mean of the Ct of each well was used for the analysis. Culture and RT-PCR results were analyzed using descriptive statistics, classifying samples as either positive or negative for MAP to determine the frequency with which MAP contamination could be found in each sample type.

Among the 10 dairy herds included in the study, 9 employed free-stall and 1 employed tie-stall housing. All of the herds had liquid or semi-liquid slurry systems for handling the manure generated in the barns. Four of the herds utilized pasture or “loafing lots” for the lactating herd for at least part of the day when climatic conditions were conducive to having the cows outside. The number of milking cows/herd ranged from 55 to 2200, with a median of 375 milking cows/herd. At the time of the study, the MAP seroprevalence in the herds ranged from 0.6% to 6.5%.

A total of 275 samples were analyzed. Of the 10 farms in the study, 7 had at least 1 manure sample positive for MAP on both culture and RT-PCR, and 3 had manure that was positive on RT-PCR only. The MAP load in the manure samples was variable between dairies and ranged between 83 MAP CFU/g feces to 1579 MAP CFU/g feces, with Ct values ranging from 29.0 to 40.9. Because farmers or their veterinarians were sometimes responsible for the sample collection, not all sample categories were collected for all the farms in this study. Of the 10 farms, 7 had soil samples tested, 6 farms had freshly harvested crops tested, and 8 farms had silage samples tested. None of the soil or feed samples tested was positive on culture. Table 1 shows the RT-PCR results for each of the sample categories, broken down by herd (A–J). Six of the 7 (86%) farms with soil samples tested were positive on RT-PCR; and 3 of those 6 farms had freshly harvested crop samples with positive RT-PCR results. Of the 8 farms with silage samples available, 3 had positive RT-PCR results. One of those farms also had RT-PCR positive soil and crop samples, 1 had all negative soil and crop samples, and 1 did not have soil or crop samples available for analysis.

Table 1.

Real-time polymerase chain reaction (RT-PCR) results [in cycles to positive threshold (Ct)] for each sample type

Herd	Manure	Soil	Crop	Silage
A	29.0	37.0	39.5	Neg
B	32.6	39.4	Nega	Neg
C	36.5	39.4	Neg	Neg
D	40.9	38.6	40.8	Neg
E	39.8	38.6	N/A	N/A
F	29.8	39.2	39.2	40.4
G	34.0	N/A	N/A	40.7
H	39.1	N/A	N/A	Neg
I	32.8	N/A	Neg	N/A
J	33.4	Neg	N/A	38.2

^aNegative.

N/A — Not available for testing.

The main objective of the present study was to evaluate the persistence of MAP in samples of soil, crops, and ensiled feeds following manure spreading on farms with JD. Although viable MAP were recovered by HEYM culture from the manure storage system of 7/10 farms sampled in this study, MAP was never detected by HEYM culture in the subsequent samples, which included soil, freshly harvested crops, and silage. However, MAP DNA was recovered in all types of samples, including silage. There are several possible explanations for the finding of MAP DNA in the soil, crop, and feed samples, with negative HEYM culture results. First, this may represent a difference in sensitivity for detection of MAP by culture versus RT-PCR. The vast dilution that occurs by spreading the manure over a wide area, combined with attrition from time and precipitation could result in low concentrations of MAP. Second, it is possible that MAP was present in the samples but its viability may have waned in the environment between the time of manure spreading and the time of sample collection, leaving DNA from non-viable organisms to be detected by RT-PCR.

In this study, manure was spread on fields prior to the emergence of plants, and MAP was found in plant material less frequently and at a lower concentration than in soil. In contrast, in another field study conducted on a JD-positive farm where yearling cattle were allowed to follow the adult herd on pasture, most of the yearling cattle became contaminated with MAP, demonstrating the importance of direct deposition of MAP on plant material before it is consumed by susceptible animals (12). Recent studies on MAP-infected farms demonstrated that MAP may be present in dust samples that settle on various surfaces (13). This raises the possibility that the finding of MAP DNA in silage samples could be post-ensiling contamination by dust or other sources of MAP on the farm.

In conclusion, on farms positive for JD, following the spread of manure onto fields used for cattle feed production, MAP was often found in soil samples, but was less frequently detected in crops and silage samples. In all cases, MAP present in soil and feed samples was in a non-viable state, or in concentrations below the limit of detection by culture. When faced with the decision about where best to spread manure, fields used for production of crops intended for harvest are preferred to grazing pastures. In general, it appears that spreading of manure on fields prior to emergence of crops represents a low risk practice for the spread of infection on the farm.