In general, the levels of indigenous E. coli decreased rapidly in the manure-fertilized soils (Table 2). Statistical comparisons of the intervals between the time of manure application and the time at which one or both plots for each treatment were analyzed by enrichment showed that the mean time intervals did not differ significantly between soils (P > 0.05). For three of four treatments on loamy sand, the first plot was analyzed by enrichment after E. coli survival may have resulted from the more rapid drainage and drying in the loamy sand or from differences in the frequency and amount of rainfall and irrigation (Table 4). Interestingly, the silty clay loam empirically differed from the other two soils in the length of time between manure application and the first enrichment-negative result (Table 2) despite the lack of a statistically significant difference between soils. For silty clay loam, no enrichment-negative results were obtained for five of the six treatments. In contrast, enrichment-negative results for the silt loam were obtained in ≤90 days for one treatment, in 110 to 128 days for two treatments, and not at all for three treatments. For loamy sand, the first enrichment-negative results were obtained for one treatment in 100 days, the first enrichment-negative results were obtained for two treatments in 118 to 126 days, and no enrichment-negative results were obtained for three treatments. Low levels of E. coli generally persisted in manure-fertilized soil for well over 100 days and were detected in enriched soil from all three sites 132 to 168 days after manure application.
There was considerable variability with no statistically significant differences between soils or between application-to-tilling intervals in terms of first enrichment, second enrichment, and first enrichment-negative times. The addition of chopped oat straw to the manure applied had no apparent effect on indigenous E. coli survival for any soil type. Because this practice resulted in more difficult tilling and increased weed growth, it cannot be recommended. For the sake of brevity, results from added-straw treatments are not shown here.
Low-level contamination (direct plating negative, enrichment positive) with indigenous E. coli occurred sporadically for washed carrots, lettuce, and radishes regardless of whether manure had been applied. Nevertheless, enrichment-negative results were common for carrots and lettuce harvested 3). For four of six, two of four, and four of six treatments, respectively, on the three soils, enrichment-negative results were obtained for washed carrots harvested ≤100 days after manure application. Occasional enrichment-positive results were obtained for washed carrots harvested 128 and 133 days after manure application for samples from loamy sand and silty clay loam soils, respectively.
Contamination of lettuce was less likely to occur than contamination of carrots. For most treatments for the three soils, enrichment-negative results were obtained for lettuce harvested within 100 days after manure application; the exceptions were two loamy sand treatments in which the lettuce bolted soon after enrichment-positive results at 78 to 80 days after manure application (Table 3). Most lettuce bolted before the 120-day application-to-harvest limit could be tested, but sporadic enrichment-positive results were obtained for lettuce harvested from the silty clay loam 120 to 121 days after manure application.
The results for radishes were more ambiguous because of rapid radish maturation. On loamy sand, only one of six treatments resulted in enrichment-negative results for washed radishes harvested ≤77 days after manure application (Table 3). The five other treatments did not yield enrichment-negative results by the time of the final harvest (61 to 82 days after manure application). Similar results were obtained for radishes grown in silt loam soil. For radishes grown on silty clay loam, five of six treatments resulted in enrichment-negative results within ≤69 days after manure application. Thus, an application-to-harvest interval of ≥100 or 120 days for radishes could not be effectively evaluated. There was no statistically significant difference between soils or between manure application-to-tilling intervals in terms of time before enrichment-negative results were obtained for any type of vegetable.
For all of the field sites combined, indigenous E. coli was detected by direct plating of washed vegetables for only 5 of 38, 1 of 48, and 6 of 24 samples throughout the study for carrots, lettuce, and radishes, respectively. For the 12 direct plating-positive results, the mean level of indigenous E. coli detected was less than 1.0 log CFU/g for all but one treatment for carrots (loamy sand treatment with June manure application and harvest 89 days later) and two treatments for radishes (loamy sand with May manure application and harvest 41 days later and loamy sand with June manure application and harvest 58 days later). The mean levels of indigenous E. coli in these cases were 1.1 to 1.2 log CFU/g.