Cumulative pathogen reduction during aerobic composting of compost mixtures prepared with fresh dairy manure and different carbon

Two separate studies were conducted to address the condition and the type of feedstocks used during composting of dairy manure. In each study, physical (temperature), chemical (ammonia, volatile acids, and pH), and biological (Salmonella, Listeria monocytogenes, and Escherichia coli 0157:H7) parameters were monitored during composting in bioreactors to assess the degree to which they were affected by the experimental variables and, ultimately, the ability of the chemical and physical parameters to predict the fate of pathogens during composting. Compost mixtures that contained either aged dairy manure or pine needles had reduced heat generation; therefore, pathogen reduction took longer than if fresh manure or carbon amendments of wheat straw or peanut hulls were used. Based on regression models derived from these results, ammonia concentration, in addition to heat, were the primary factors affecting the degree of pathogen inactivation in compost mixtures formulated to an initial carbon-nitrogen (C:N) ratio of 40:1, whereas, the pH of the compost mixture along with the amount of heat exposure were most influential in compost mixtures formulated to an initial C:N ratio of 30:1. Further studies are needed to validate these models so that additional criteria in addition to time and temperature can be used to evaluate the microbiological safety of composted manures. Dairy cattle production is a large global industry, with approximately 115 million dairy farms in operation worldwide in 2005 (8). Although milk production practices and management systems vary between countries, one challenge that is faced at every facility is the production of manure. For example, the Environmental Protection Agency documented that in 2007 within the United States, there were 12.5 million dairy cows that produced 190.8 million tons of manure (24). Although land application has been one m ethod for disposal o f this manure, the size of many operations and the amount o f manure generated precludes direct land application for disposal, as it exceeds the loading capacity of the soils, leads to nutrient losses by surface runoff and leaching, and contributes to eutrophication of surface water bodies or contamination o f groundwater (10). Two common practices to handle the excess quantities o f manure are anaerobic digestion that generates methane for energy and thermophilic aerobic com posting that generates soil amendments for improving soil structure and providing stabilized plant nutrients. In the latter process, nitrogen-rich manures are mixed with one or more carbon amendments to produce initial carbon-nitrogen (C:N) ratios (20:1 to 40:1) that are favorable for the metabolism of therm ophilic microorganisms. Under these conditions, indigenous m icroflora break down the com post materials, o f which volatile acids, ammonia, and heat are by-products * Author for correspondence. Tel: + 1 770-412-4742; Fax: + 1 770-2293216; E-mail: mericks@uga.edu. that serve in various capacities to inactivate zoonotic pathogens that are often present in the manure (15). A wide assortment o f carbon amendments may be added to manures; however, even though two amendments could have the same C:N ratio, the carbon availability may vary, depending on the surface area or particle size o f the am endment and the extent o f lignification o f the material (4). Examples o f amendments that are more resistant to degradation are those that contain a sizeable fraction of lignin and cellulose (e.g., wood chips). Availability o f these amendments to thermophilic microbes could, therefore, affect the organism s’ level of metabolic activity and, thereby, the am ount of heat generated within a given time frame. Such was the case when a variety of organic wastes were cocomposted with dairy cattle feces in a study by Hanajim a et al. (7). In that study, enhancement o f the thennophilic phase of com posting was related to the initial amount of easily digestible carbon present in the mixture. Quantitative and qualitative differences in products gener­ ated during com posting, however, may also occur, depend­ ing on the carbon source. For example, volatile fatty acids are com mon intermediates in the metabolism of carbohy­ drates, whereas phenolics are released through the decom ­ position of lignin (9). Such variations in metabolic activity and outputs in response to the incorporation of different carbon amendments could, therefore, have an impact on the survival o f zoonotic pathogens residing in the manure. Another variable that can affect the thermal profile during com posting is the age of the manure. In general, 1912 ERICKSON ET AL. J. Food Prot., Vol. 77, No. 11 manure that has been stockpiled undergoes some decom­ position and, when mixed with a carbon amendment, either takes longer to heat or its maximal temperatures are lower than the fresh manure (3,11). The age of the manure could also be a contributing factor to the different responses (heat generation and pathogen inactivation) exhibited in bioreac­ tors and field compost piles when initial C:N ratios of materials are varied between 20:1 and 40:1 (5, 6, 20). More specifically, in bioreactor systems where fresh manure was used, pathogens were eliminated more rapidly in compost mixtures formulated to C:N ratios of 20:1 than in compost mixtures formulated to 30:1 or 40:1, yet the cumulative heat exposure was least in the 20:1 systems (5, 6). In contrast, in field compost pile systems where aged manure was used, pathogens were eliminated more rapidly and maximal temperatures were greater in compost mixtures formulated to 30:1 than in those formulated to 25:1 or 20:1 (20). Based on these studies, it would appear that aging of manure could impact not only quantitatively the degree of indigenous microbial activity but also qualitatively the type of breakdown products, including the production of antimi­ crobials (i.e., volatile acids or ammonia). Given that the bioreactor and field studies were conducted under different conditions and with different dairy manure sources, it was the objective of this study to examine the influence of aging of dairy manure on inactivation of Salmonella during composting in bioreactors using mixtures formulated to an initial C:N ratio of 40:1. Another objective of this study was to compare the inactivation of Salmonella to the inactivation of Escherichia coli 0157:H7 in bioreactors using compost mixture formulated with fresh manure only. In a separate bioreactor study designed to address the influence of carbon amendments (wheat straw, pine needles, or peanut hulls) on pathogen survival, Salmonella and Listeria monocytogenes inactivation was related to the generation of heat, pH, volatile acids, and ammonia in dairy manure-based compost materials formulated to an initial C:N ratio of 30:1. MATERIALS AND METHODS Pathogen strains. Strains of Salmonella Enteritidis (M E-18, H4639, and H3353), Salmonella Newport (1 1590-K), L. monocy­ togenes (101M, 12443, F6854, G3982, and H7550), and E. coli 0157:H7 (C7927, E0143, K262, C0083, and E0139) were used and were obtained from the culture collection housed at the Center for Food Safety, University of Georgia, Griffin. All sfrains were labeled with a green fluorescent protein (GFP) plasmid to aid in detection and enumeration of pathogens in compost materials. In addition, to reduce interference from background microbial flora, the plasmid also contained an ampicillin-resistant marker for GFPlabeled Salmonella and E. coli 0157:H7 strains, whereas the plasmid contained an erythromycin-resistant marker for GFPlabeled L. monocytogenes strains. Plasmid stability for most of these strains was reported previously and ranged from 0 to 77% plasmid loss after 20 generations (12). Inoculum preparation and enumeration. The protocol used to culture Salmonella, L. monocytogenes, and E. coli 0157:H7 was similar to that described for previous bioreactors studies (5, 6). Stock solutions of each pathogen strain were prepared by suspending cells in 0.1% peptone water for concentrations of ca. 109 CFU/ml that corresponded to an absorbance at 630 nm of 0.5. Mixtures to be used for spraying of compost materials were then prepared by combining equivalent volumes of individual strains of each pathogen and diluting 10-fold with deionized water to give cell concentrations of 10s CFU/ml. Cell populations of these mixtures were determined by plating on either tryptic soy agar (Difco, BD, Sparks, MD) and 100 pg/ml ampicillin for Salmonella and E. coli 0157:H 7 or modified Oxford medium (Acumedia Manufacturers, Lansing, MI) containing 10 mg/ml buffered colistin methanesulfonate, 20 mg/ml buffered moxalactam solution, and 8 pg/ml erythromycin for L. monocytogenes. On tryptic soy agar and 100 pg/ml ampicillin plates, fluorescent colonies of the pathogens were visualized using a handheld UV light (365 nm) Small fluorescent colonies o f L. monocytogenes on plates containing 10 mg/ml buffered colistin methanesulfonate, 20 mg/ ml buffered moxalactam solution, and 8 pg/ml erythromycin required visualization using a Leica MZ16 FA stereo fluorescent microscrope (Leica Microsystems, Bannockburn, IL). Source and chemical analysis of composting ingredients. Fresh cow manure was collected from a dairy farm near Griffin, GA, for each replicate trial. For the study comparing fresh versus aged manure as the nitrogen source in composting mixtures, the manure was split into two large portions and one small portion. One large portion was immediately frozen to kill insect eggs, whereas the other large portion was spread on a plastic tarp and held in a greenhouse for 3 weeks during September and October. The small portion was used for determination of its carbon, nitrogen, and moisture contents. Wheat straw and cottonseed meal, purchased from a local feed store, served as the carbon amendment in that study. In a separate study comparing different carbon amendments in composting mixtures, wheat str