However, the impact of animal waste-borne microbiomes on environment including soil, water, and plant is not well-understood. The United State Department of Agriculture estimates that there are approximately 450,000 animal feeding operations in the U.S., which include beef cattle, dairy, poultry and swine production industries. Annually, over 2 billion tons of animal manure are generated in the U.S. In California alone, 60 million tons of manure are produced annually by 5.2 million cattle and calves, and a considerable portion of the manure is applied onto cropland as fertilizers. While the use of manure as fertilizer in cropland has numerous benefits, such as reducing chemical fertilizer application, additional understanding of how animal waste-borne microbiomes could impact cropland and public health is needed to utilize the full potential of manure and to understand any consequential negative impacts of manure on cropland and environment. Elevated pathogen/pathogen indicator levels in surface and ground water and their potential linkages with animal waste have received considerable public attention because of associated public and animal health risks and produce contamination. In general, the use of fresh and untreated manure as fertilizers has a greater potential to increase pathogen loads in cropland, and subsequently these bacterial populations can be transported to rivers and streams during rainfall/runoff events. Furthermore, the use of untreated manure as fertilizer can facilitate the transfer of harmful bacterial population to ready-to-eat crop. To control the bacterial loads in manure used as fertilizer, grow rack several manure treatment practices such as composting, anaerobic lagoon systems, and anaerobic digestions are used.

Previous studies showed that pathogens such as E. coli, Salmonella, and Listeria in dairy manure are reduced through the application of these waste treatment processes, though the complete elimination of these pathogens by these processes is uncertain. Further, the existing knowledge is weak in terms of changes in microbial community in manure after various manure handling processes, such as solid-liquid separation, manure piling, and storage lagoon. In a typical, large California dairy, both liquid and solid manure are produced by flush manure management systems, which are common in California’s Central Valley. In such systems, a dairy barn is flushed with water, and flushed manure is passed through solid-manure separation systems, where liquid manure is separated from solid . Solid streams are stored in the form of piles, and liquid manure streams are stored in lagoon systems prior to the application of manure into the cropland as fertilizers. Although both lagoon systems and compost piles are used extensively to manage dairy manure in California, the efficacies and effectiveness of these manure handling processes for regulating microbial population are not well-understood. For these practical reasons that have considerable impact on agriculture, manure management in dairy farms, and its application in cropland, we hypothesized that microbial quality of dairy manure should change with on-farm manure handling/treatment processes. Moreover, this change in microbial population should be consistent from one farm to another. Such changes or shifts in microbial population of manure—and continuous use of that manure as fertilizer in a cropland for long period—have potential to impact the microbiome of cropland receiving manure as fertilizer. Therefore, the understanding of how the dominant bacterial community levels changes in typical dairy manure management practices in a farm environment is essential.Although numerous previous studies targeted investigation of the inactivation of selective bacterial pathogens such as E. coli, Salmonella, Listeria under specific conditions, these studies mainly focused on understanding of selective human pathogen inactivation in various treatment processes.

The insights of how various microbial populations at genus level change in particular processes are crucial, however, not well-reported. Further, having such information can help improve the currently-available manure management techniques, and support decision-making in terms of using manure as fertilizer in a specific cropland. Previous studies have used high-throughput microbial community profiling methods to gain insights into microbial community distribution in different environments. Amplicon-based community analysis has been used to determine the microbial communities in various samples, including food samples, anaerobic sludge, biosolids, natural environments, and agricultural grasslands. These methods have also been applied in raw dairy manure . However, the application of these methods to understand the microbial communities of manure fertilizers processed at various levels of treatment has not been explored yet. As a test of our hypothesis to determine the differences in microbiome of manure fertilizers, the goal of this study was to quantify the microbial population levels of various forms of dairy manure, such as liquid and solid produced in a typical dairy farm environment in California Central Valley. The objectives of the study are to: 1) determine the dominant microbial communities in solid and liquid forms of dairy manure fertilizer; and 2) understand the changes in microbiome of manure after solid-liquid separation, lagoon storage, and manure piling . We anticipate that the outcomes of this study will reveal greater understanding in terms of microbial quality of dairy manure fertilizers, and will help in making informed decisions. Further, improved insights will help in advancing dairy manure management, manure application, and understanding the environmental and public health risks associated with animal waste-borne microbial pathogens.The solid and liquid samples in dairy farms were collected from California Central Valley, which has the most densely populated dairy farms in California. Fig 2 shows county maps of Tulare, Glenn, and Merced, including the herd size in Tulare and Merced Counties.

For the current study, we collected 33 manure samples, which include solid and liquid manure. Solid manure samples were collected from manure piles located in dairy facilities, while liquid samples were collected from liquid manure storage ponds  as well as from flushed manure pits.Dairy facilities used for sample collections are located in three counties . From each dairy facility, we collected one liter of liquid manure sample in sterile bottles from each pond, and 600 g of solid manure in sterile bottles from each pile. Immediately after collection, samples were transported using a cooler and subsequently stored at -20˚C prior to analysis. For analysis, samples were thawed at room temperature. The solid samples collected from piles that were less than 2 weeks old were termed as Fresh Pile , while older piles were termed as Compost Pile . It is important to note that the studied CP does not necessarily mean the sample was subjected to standard composting processes, hydroponic rack where maintaining the thermophilic temperature and mixing is necessary. The liquid manure samples collected from Primary Lagoons and Secondary Lagoons were termed as PL and SL, respectively.The microbial diversity assessment of solid and liquid wastes using phylotype taxonomy resulted in a total taxa of 1818. Approximately 85% of 1818 taxa were classified at the genera level and 10% at the family level. In FP solid samples, sequence reads varied from 13,950 to 453,625 with an average of 153,316. In solid samples from CP pile, sequence reads varied from 15,798 to 1,092,032 with an average of 296,153. The average reads for FM were 333,450 with range from 5,242 to 989,040. In PL and SL liquid samples, the average sequence reads were 186,341 and 130,888 , respectively. FP samples, which were not dried and composted, showed the abundance of bacteria of genus Acinetobacter and Enterococcus; these bacteria were the most common and accounted for 3.5% – 39.53% and 4.8% – 11.86%, respectively. In CP samples, which were either dried or composted, the proportion of bacteria of genus Acinetobacter ranged from 18.3% to 19.2%. Other abundant species in CP samples were Flavobacteriaceae, Bacillaceae, Pseudoxanthomonas, Clostridia, and Sphingobacterium, accounting for 3.9–24.9%, 7.2–7.3%, 2.8–5.5%, 4.5–6%, 2.1–3.1%, respectively. Other abundant species in FP samples were Bacteriodetes, Trichococcus, Clostridiales, Flavobacterium, and Psychrobacter. A heat map of the top 50 taxa in total of 1818 taxa is shown in Fig 3. In FM samples, the most common species were Ruminococcaceae varying from 7.2 to 13.1%. Species such as Bacteroidetes and Clostridium varied 4.2–11.5% and 3.5– 9.7%. In lagoon samples , however, the most common species were Bacteroidetes, Flavobacteriaceae, and Psychrobacter accounting for 11.1–15.9%, 3.3–13.1%, and 19.3–29.6%, respectively. Dendrograms and PCA plots are shown in supplementary figures . The application of algorithm based on the abundance criteria resulted in 128 taxa, and the analysis of top 50 communities in 128 taxa is shown in Fig 4. The dendrogram and PCA of CP, FM, FP, PL, and SL are shown in Fig 4A and 4B, respectively. The heat map shows the distribution of microbial communities in CP, FM, FP, PL, and SL. In the dendrograms, the horizontal axis represents the distance of dissimilarity between clusters. The vertical axis represents objects and clusters. Results showed that FM is more similar to PL than SL. Furthermore, the similarity between FM and PL was greater than SL and CP. The frequency and abundance in 128 taxa of solid manure samples are shown in supplementary file , and these characteristics for liquid samples are shown in supplementary table .Sample grouping tendency in 128 taxa was evaluated using PCA. The PCA score plot shows a two-dimensional plot of 33 samples. The first two principal components explained 56.7% of the total variance in the microbial community composition. FP and CP samples were clustered together mostly in the upper and lower left corner of the plot, while FM, PL, and SL were clustered together in the lower left of the plot. As shown in the figure, CP and FP groups were similar to each other and distinct from PL, SL, and FM. Further, PL and SL were grouped together. The clear separation of CP and FP from PL, SL, and FM indicates that manure handling processes such as solid-liquid separation adapted in dairy farms have the potential to alter the microbial communities in the manure. Together, these results demonstrate that the forms of manure fertilizers will affect the microbial quality of manure fertilizers, which prove the central idea of our hypothesis.

The abundance of genera in each sample is shown in a heat map . In the heat map, the light blue indicates low abundance and dark red indicates high abundance. The results of top 50 genera present in samples indicate that species abundance differs among samples. InFP, the most abundant species such as Acinetobacter, Psychrobacter, and Enterococcus accounted for 9.9%, 2.5%, and 2.5%, respectively. The other unclassified bacteria in FP accounted for 31%. In CP, the most abundant species include Planifilum, Acinetobacter, and Flavobacteriaceae accounted for 6.4%, 6.0%, and 4.4%, respectively. The unclassified bacteria in CP were 25.6%.To understand the distinction among microbial communities in solid and liquid samples, the data of liquid samples and solid samples were grouped separately. The dendrogram plot showed clustering among solid and liquid samples . Results indicated that all the liquid samples were more similar to each other than solid samples. The PCA plot displayed that the solid samples were mostly clustered to the left side, while the liquid samples were clustered to the right side, indicating a clear separation among liquid and solid samples. The top 50 genera in 128 taxa are presented in heat map . Results showed that the top 22 species listed at the top of heat map were more abundant in solid samples compared to the liquid. These species include Bacteroidetes , Ruminococcaceae , Flavobacteriaceae , Clostridium , Cloacibacillus , Petrimonas , Psychrobacter , and Proteiniphilum . The bottom 28 species were more abundant in liquid samples compared to solid samples, and these microbial communities include Smithella , Pseudomonas , Sporobacter , Treponema , and Aminivibrio . Based on the canonical analysis , the genus Gp4, Nocardioides and Caryophanon were highly correlated with solid manure, while the genus Succiniclasticum, Porphyromonas, Methanospirillum, Anaeroplasma, Armatimonadetes, Eubacterium, Vampirovibrio, Anaerovorax and Lactonifactor, and the family Porphyromonadaceae were highly correlated with liquid manure . Canonical values from the discriminate analysis were also used to identify bacteria that were highly correlated and led to differentiation of FP vs CP, and FM vs LM . Bacteria genus Coraliomargarita was highly correlated with FP and genus Ruania and family Peptococcaceae were highly correlated with CP. From this analysis, we observed that the genus Bifidobacterium, Murdochiella, Nitrosomonas, Arcanobacterium, Gallicola, and Kurthia were highly correlated with FM. Overall, the comparison of FM and LM had a more similar microbial composition and diversity than the comparison of FP and CP.