Reduction of Bacterial Pathogens in a Single-Stage Steel Biodigester Co-Digesting Saw Dust and Pig Manure at Psychrophilic Temperature

Abstract

The experiment was conducted using a single-stage steel biodigester of 75 L working volume, charged with 75% pig manure and 25% pine wood sawdust and operated under batch mode at a psychrophilic temperature range (13.16–24.69 °C). The total viable count was determined via the spread plate method on selective microbiological media to determine viable numbers of the selected bacterial pathogens in samples collected from the biodigesting chamber every seven or fourteen days over the duration of study. Multiple linear regression models including the log bacterial counts (response) and number of days, pH, and average daily temperature as predictors were developed using Matlab for each bacterium. The reduction (90–99.9%) in numbers of isolates of E. coli, Salmonella, Yersinia, Campylobacter, and Listeria varied with time (days) from their initial respective counts of 2 × 10⁶, 7 × 10⁴, 3 × 10⁵, 9 × 10⁵, and 1 × 10⁴ cfu/g to concentrations lower than the detection limit (DL = 10² cfu/g substrate). E. coli demonstrated the least resistance to the environmental conditions in the biodigester and survived only for 77 days, unlike L. monocytogenes that lasted for 175 days and was the most resistant bacterium. From the models, the number of days and temperature were directly and inversely related to log Listeria counts, respectively, contrary to the others. The predictors, number of days, pH, and average daily temperature, were described as either primary or secondary factors based on the bacteria via the reliefF test.

1. Introduction

Environmental sustainability is a key quest for every society, especially now that human, commercial, industrial, and agricultural activities are on the rise, creating significant quantities of wastes that need to be appropriately managed. On farms, manure management includes handling, storage, and application [1]. The traditional management of animal manure involves the application on agricultural fields. However, the major limitation to the extensive use of manure as a fertilizer is the availability of land to receive it in an area close to the farm; therefore, extra land used for other agricultural purposes might be an option [2]. Clearly, direct land application of animal manure results in an improvement of the soil structure, increase in water retention capacity, soil root penetration, and microbial activities, which eventually affects the ability of the crop to absorb nutrients, causing an increase in growth and yield [3,4,5]. Such application is vulnerable to the dissemination of manure into water bodies and soil, causing contamination and pollution as it contains a number of pathogens and great organic matter. In detail, applying animal manure on the soil causes air pollution, the release of greenhouse gases [6], perturbance of carbon, nitrogen, phosphorus, and other nutrient cycles owing to the imbalance of nutrients in manure [7], trace element enrichment of the soil [8], as well as eutrophication of water bodies [6], which causes death of aquatic life due to oxygen depletion added to the pollution of the ground and spring water with microbial contents, causing infections and diseases [4,9].

On the other hand, sawdust encompasses small discontinuous chips or fine particles of wood, formed as a by-product from a series of timber manufacturing processes, including sawing, routing, drilling, and furniture manufacturing and joinery [10,11]. These wastes (lignocellulosic biomass) are disposed and managed either by dumping on roadsides, drainages, water bodies, or by incineration or by composting, resulting in the emission of greenhouse gases (GHG) and noxious gases, causing global warming and its release into the surroundings, leading to environmental hazards [12]. Furthermore, Adegoke and colleagues [13] emphasised that extra labour and cost are involved to transfer the sawdust to the site for regular burning, and that sawdust and other wood wastes constitute flammable materials or oftentimes assist in the spread of fire outbreaks at the mills. Moreover, sawdust has been reported to be responsible for creating a surface for adsorption of pathogenic microbes, and the blocking of sunlight into water bodies, thereby disrupting the oxygen flow in water [13]. Despite the adverse effects of sawdust, several authors have highlighted that to evade the aforementioned environmental effects, sawdust can become a valuable resource utilised in manufacturing, energy, and agricultural processes, generating products including briquette, fertiliser, particle board, animal bedding, mulching, and household energy (producer gas and oil [12,14]), consequently mitigating sawdust environmental pollution, conserving energy, as well as reducing disposal costs [11].

It is worth mentioning that South Africa is a water-scarce country, and on a global scale, it is projected to be amongst the vast majority contributing to the total increase in meat consumption, owing to rising urbanisation, increasing income, and rapid population growth [15]. This may suggest increases in the abundance of the animal wastes that will need proper treatment or management to prevent pollution of the environment and the water bodies in the country as well as to shirk the above-mentioned adverse effects. In order to obtain improved manure management, treatment and resource recovery can be designed to boost food security, even as protecting water bodies and providing revenue streams for small-scale farmers in order to lessen poverty. Most of the population experiencing poverty tend to live in rural areas and depend on agriculture (small-scale farming) for sustenance and these people can produce the majority of the food in developing countries, yet they are usually the most susceptible and deprived individuals lacking resources, provision, and appropriate technologies to mitigate the environmental impact of their activities [16]. Nevertheless, small-scale farmers might not demonstrate economic flexibility to procure advanced treatment options, including sequencing batch reactors, trickling filters, or engineered wetlands. Therefore, in developing regions, the usual animal manure management practice at small-scale farms is direct land application of manure on fields without treatment or treatment via oxidation lagoons [6]. Apparently, at small farms in developing communities, appropriate technologies for treatment should utilise locally available resources, be economically affordable, and should, in turn, safely provide resources to the farmer [17]. Microbial anaerobic degradation of these wastes taking place within a chamber devoid of oxygen or air (biodigester) is considered a good option to properly manage these wastes because it produces biogas (a renewable energy source) and biofertilizer, as well as causes decontamination of the wastes [18]. Notwithstanding, the generation of biogas or biofertiliser (as resources) does not only diminish the use of fuel and synthetic fertiliser but can likewise ease the adverse environmental influences of GHG emissions and eutrophication potential (caused by microbial and chemical contents of the manure) [19].

The microbial makeup of animal manure varies from one farm to the other, depending on the producing animals (growth stage of the animals), the feeding regime, and the waste management practices [20]. It is also clear that the biodigester type (design) and operational factors (temperature) of the process and manure characteristics affect the overall effectiveness of the anaerobic digestion process. The organic loading rate (OLR) is an important design parameter, which depends on the biodigester design (type) and substrate characteristics, as well as indicates the quantity of the volatile solids (VSs) to be fed per unit volume of the biodigester. The factor is crucial as the overloading of substrates can cause digester failure in little time [21]. The OLR in conjunction with temperature are regarded as key parameters for digestion and their suitable application influences the outcome in relation to energy feasibility [22].

The composition of these pathogens occurring in animal manure is observed as one amongst the major restrictions to the use of raw animal manure for soil fertilization in order to enhance food production in the quest to alleviate hunger and poverty [18]. The inactivation of pathogens (belonging to the bacteria domain) in animal manure serves as a prerequisite if manure is intended for subsequent uses which include recycling, disposal, or land application [6]. The decontamination of the wastes is attributed to several factors, including ammonia, pH, temperature, feedstock characteristics, proteins, volatile fatty acids, and carbohydrates [18]. Nevertheless, temperature is the leading cause of pathogen inactivation, causing prominent decontamination of effluent slurry. Based on temperature, a biodigester can be operated at psychrophilic (15–20 °C or less than 30 °C), mesophilic (30–40 °C), or thermophilic temperatures (50–60 °C) ranges and inactivation rates occur faster with increasing temperature from the psychrophilic to the thermophilic range [22]. However, the potential of a biodigester to cause pathogen inactivation relies on the digester temperature, the concentration of volatile fatty acids, the retention or treatment time, the type of feedstock, as well as the type of pathogen [18]. According to Lin et al. [23], for such a long time, the central focus in anaerobic digestion has been on the efficiency of the process to enable maximum biogas yield, and less attention has been given to the control of pathogens; however, in the current anaerobic digestion process, it is a challenge to maintain an efficient methanogenesis and an effective decontamination, simultaneously. Nonetheless, the majority of the studies that investigated bacterial inactivation via anaerobic digesters noted their findings obtained at mesophilic and thermophilic temperatures [23,24], explaining that rising temperature (37 °C to 70 °C) causes an increase in the fluidity and permeability of the cell membrane, thus permitting a more rapid diffusion of toxic chemicals into the cytoplasm and inhibiting cell growth [25]. At this juncture, it is worth mentioning that pathogens survive longer at mesophilic than thermophilic temperatures (higher bacterial inactivation), but thermophilic anaerobic digestion is associated with process instability and higher odour potential, and requires extra power/energy [25,26]. Moreover, Jiang et al. [18] remarked that most of the studies performed on the inactivation of pathogens during anaerobic digestion focused on a single and a particular pathogen and/or operational condition, and the pathogen inactivation was studied under ideal laboratory conditions. Following the literature, it is worth mentioning that the results revealed in this study via the use of a single-stage steel anaerobic digester operated at psychrophilic temperature differ from previously conducted studies. However, there is a dearth of information on the inactivation of bacterial pathogens via psychrophilic anaerobic digestion [25], and the present study is especially amongst the few studies conducted in South Africa and the world at large performing anaerobic digestion on several bacterial pathogens (both Gram-negative and Gram-positive) at psychrophilic temperature (without the consumption of extra energy), which resulted in the reduction in zoonotic pathogens.

Accordingly, different bacterial pathogens respond differently to changes in the environmental and nutritional conditions occurring within the closed digester system. Therefore, they have different rates of elimination or decay [27]. Thus, there is a continuous search of ways to manage the wastes as water and environmental policies are continuously regulated in a bid to ensure proper management of the ever-increasing wastes emanating from the rising activities in animal farming and agricultural activities [28].

Against this background, our study was undertaken to assess the ability of the single-stage biodigester to inactivate bacterial pathogens during anaerobic co-digestion of substrates occurring in a psychrophilic temperature range, which equally corresponded to the survival period of these pathogens. Viability studies (in terms of total viable count) were conducted while employing a range of microbiological media via the spread plate technique. Additionally, mathematical models were developed to demonstrate the log bacterial counts as a function of average daily temperature, number of days, and pH as predictors. Furthermore, the reliefF test of the Simulink embedded in Matlab was employed to describe the influence of the different predictors as primary or secondary on log bacterial counts.

2. Materials and Methods

2.1. Sources of Samples

Undiluted, undecomposed, and fresh pig manure was collected from the Piggery Farm in a sterile bucket with the aid of a spade, and pine wood sawdust was procured from a sawmill in Melani village, Hogsback, located some kilometres from Fort Hare University, Alice Campus. Traditionally, the pig manure on the farm is managed by separation and screening into liquid and solid portions, which are utilised for irrigation and as a substrate for composting, respectively. Similarly, the sawdust is traditionally treated by composting and incineration in the open (Local informant). With inference from findings obtained from studies focussed on environmental assessment and monitoring, these activities engaged in the management of these wastes are associated with environmental and public health hazards, as presented by other authors elsewhere [1,6].

Finally, the biomass was pre-treated by air-drying, mechanical grinding, and screening to remove objects such as sticks, stones, and plastics as the case may indicate. A slurry of 75% pig manure and 25% Pinewood sawdust in water was prepared in the ratio of 1:1, which was used to charge the biodigester as shown in Figure 1 [29].

2.2. Anaerobic Co-Digestion Process

The experimental setup is as shown below:

Figure 1. A display of the biodigesting chamber and other components.

Figure 1. A display of the biodigesting chamber and other components.

Anaerobic co-digestion of the combined matrix was conducted in a single-stage steel biodigester, which was constructed and charged with the slurry to about three quarters of its original volume, leaving a headspace for the biogas generated [30]. Anaerobic degradation occurred at a psychrophilic temperature regime of 13.16–26 °C as a batch operation over a period of seven (7) months without fresh introduction of raw substrate [21]. The process was batch-operated as it is a simple way to evaluate the biogas-producing potential of the organic wastes and the destruction of bacterial pathogens, as all the substrates were loaded at once and discharged only after the retention time was completed, indicating that the bacteria were retained (giving them sufficient time) in the digester throughout the process. Meanwhile, in continuous feeding, the substrate is constantly loaded into the digester and digested materials are constantly removed, thus affecting the reduction in the pathogens because they are not given considerable time to spend in the digester [30].

For the efficient and proper functioning of the biodigester, the temperature and pH of the system were continuously monitored using a temperature sensor embedded inside the slurry and attached externally to a Hobo U12 data logger, which was configured to log every one-minute interval, creating a daily slurry temperature profile. In addition, the pH of the digesting medium that was withdrawn based on the stipulated intervals from the biodigester was measured immediately using a portable PHSCAN 30 pH meter. The biodigester was enclosed in a wooden box and completely filled with compacted sawdust for insulation.

For analysis purposes, the prepared digester slurry was immediately sampled and referred to as the day 0 sample, and subsequently, samples were withdrawn at 7- or 14-day intervals via the stop valve (22 mm) located on one of the ports drilled on the biodigester.

2.3. Determination of Total Viable Count

The total viable count, also known as the standard plate count or the aerobic plate count, was investigated to determine the microbiological quality (in terms of number of growing/viable bacterial cells) of the samples withdrawn from the biodigester [31]. In this study, the spread plate method was employed to evaluate the number of viable bacterial cells of the samples collected from the biodigester that were increasingly serially diluted. This was performed as per the procedure reported by Gonzalez [32]. Selective microbiological media were used to restrict the growth of nontarget bacteria, including Salmonella/Shigella agar (SSA), E. coli Chromogenic agar, Blood-free Campylobacter agar base (CCDA), Listeria-selective agar base (LSAB), and Yersinia-selective agar base (YSA) for the growth of strains of Salmonella, E. coli, Campylobacter, Listeria monocytogenes, and Yersinia enterocolitica, respectively. For the growth of the last three bacteria, the microbiological media were made selective by introducing CCDA supplements, Oxford Listeria-selective supplements, and Yersinia (C.I.N) supplements, respectively. All the microbiological media were purchased from CONDA, Madrid, Spain and the selective supplements were from Oxoid, UK. Into a sterile falcon tube, 10 mL of the slurry was extracted from the digester, which was 10-fold serially diluted in bacteria-free normal physiological saline as previously described. Concisely, 1 g of the sample was dissolved in 9 mL of 0.9% physiological saline contained in sterile test tubes.

Each medium plate was inoculated with 100 μL of each serially diluted sample in triplicates. The inoculated plates were incubated at different temperatures and time intervals according to the specific medium and the bacterial pathogen. In relation to the growth of Salmonella and Escherichia coli, the inoculated plates were incubated at 37 °C for a period between 18 and 24 h and for 24–48 h for the growth and isolation of Listeria. Colonies with dark blue or crystal violet on the E. coli Chromogenic agar were counted as E. coli [33], pink colonies with a black centre produced on SSA were considered as Salmonella sp. [34], and grey colonies surrounded by black halos (esculin hydrolysis) were enumerated as L. monocytogenes [35]. In addition, incubation was performed at 30 °C for 24–48 h for the growth and isolation of Yersinia, while Campylobacter species were grown at 42 °C under microaerophilic conditions (CampyGen^TM, Oxoid, Basingstoke, Hants) in an anaerobic jar for 24–48 h. Colourless colonies with a red centre or which appeared with red bull’s eyes were counted as Y. enterocolitica [36], while tiny, smooth, convex, and colourless translucent-to-grey-coloured colonies were quantified as Campylobacter species [37].

After incubation, the mean of the developed colonies obtained from the triplicate inoculation was determined and recorded on respective tables as colony-forming units per gram in 1 mL of initial material.

2.4. Modelling of the Inactivation of Bacterial Pathogens

The modelling of the logarithmic inactivation of the bacterial pathogen as a function of temperature, number of days, and pH to investigate the influence of these parameters on the decimation reduction of viable bacterial cells was performed. The output or response variable was log bacterial counts, while the average daily temperature, number of days, and pH were considered as the predictor variables. Furthermore, the reliefF test was performed to categorise these predictors as primary or secondary factors, by weight, of the contribution in the logarithmic counts of the bacteria. This was performed according to the procedure of Tangwe et al. [38]. The models were developed and the reliefF test was performed for each bacterial species to categorise the predictors into primary or secondary factors. In addition, a multiple comparison test using an ANOVA test was employed to determine if there is any significant difference in the group means of log bacterial counts between the bacterial species.

2.5. Data Analysis

The experimental data were processed and analysed using Matlab software (vR2013a).

3. Results

3.1. Total Viable Count (TVC)

Due to the occurrence of bacterial growth on specifically enriched media, along with the cultural and morphological characteristics of the colonies observed, the following bacterial pathogens: E. coli, Salmonella sp., Y. enterocolitica, Campylobacter sp., and L. monocytogenes, were clearly identified and confirmed in the wastes under investigation. The initial counts were as follows: 2.0 × 10⁶, 7.0 × 10⁴, 3 × 10⁵, 9.0 × 10⁵, and 1.0 × 10⁴ with logarithmic values of 6.3, 4.8, 5.48, 5.9, and 4, respectively. The aforementioned counts of the respective bacterial pathogens were reduced to undetectable levels (<detection limit, 10² cfu/g) following the psychrophilic anaerobic digestion. An overall 1 log bacterial count reduction (T₉₀) was demonstrated by all the bacterial pathogens, although over a different number of days. Taking into consideration the time (in days) over which the bacteria survived, the decimation occurred as follows: E. coli (77 days) < Salmonella (84 days) < Y. enterocolitica (98 days) < Campylobacter (112 days) < L. monocytogenes (175 days). The decay rates/survival patterns of the viable bacteria cells were as displayed in Figure 2. L. monocytogenes displayed the longest survival time of 175 days, unlike E. coli, which experienced the least survivability (77 days) in the hostile environment of the biodigester.

3.2. Building of Mathematical Model

Regarding the building of the mathematical models,

Table 1. Multiple linear regression models of log bacterial counts with number of days, pH, and average daily temperature as predictors.

Bacteria Species	Log of Counts	A0	A1(t)	A2(PH)	A3 (T)	R2	p-Values
Escherichia coli	Y	3.822	−0.057	0.264	0.040	0.928	0.956
Salmonella sp.	Y	5.822	−0.044	−0.280	0.043	0.919	0.967
Y. enterocolitica	Y	−16.022	−0.016	3.261	0.011	0.978	0.970
Campylobacter sp.	Y	−3.833	−0.030	1.300	0.072	0.949	0.954
L. monocytogenes	Y	−3.920	0.002	1.251	−0.017	0.984	0.979

Y, modelled log bacterial counts (response); A0, forcing constant; A1, scaling constant of the variable, number of days; A2, scaling constant of the variable, PH; A3, scaling constant of the variable, average daily temperature.

shows the association between the log bacterial counts (output/response) and the predictors, including average daily temperature, number of days, and pH. In developing the regression models, the last datum/measurement of the whole data (<100 cfu/g, detection limit) was avoided in order to ensure data integrity. These models revealed the type of relationships that existed and are described as multiple linear regression models. These multiple regression models were employed to determine the strength of the relationship between the response and predictor variables as well as to forecast changes in the response variable, should a specific predictor variable experienced change while the others were kept constant. In addition, the ANOVA plot of the group means of log bacterial counts of the different bacteria was developed (Figure 3) and the plot indicated that there was a significant difference in the group mean log bacterial counts amongst the bacteria, but it was not clear between which bacteria.

For each of the groups, the major blue plot represents the ANOVA plots for each bacterium, the horizontal red line in each ANOVA plot represents the median for the group, the lower and upper black horizontal lines in each group represent the lower and upper confidence level in the group, respectively, and the red cross marker(s) in each group represent the outliers.

To identify which bacterial species were involved, we then employed a multiple comparison test to determine between which bacterial species there existed a difference in significance in the group means of log bacterial counts. The results are as shown in Figure 4. In Figure 4, each of the horizontal lines represents the full range of the data of log bacterial counts for each group. The circle marker at the centre of each horizontal line represents the mean of the group. The colours are produced by default with the coloured horizontal line (E. coli group) representing the reference used in comparing to the rest of the groups. The two broken vertical lines (cyan colour) represent the interval estimates for the reference group (E. coli). If none of the other horizontal lines (different groups) overlap with either or both broken vertical lines, that implies that no mean significant difference existed between the reference group and those groups. Therefore, as the group with the red-coloured horizontal line (Yersinia) did not overlap with the interval estimates, there existed a mean significant difference between the E. coli group and the Yersinia group. Alternatively, when there is an overlap of the compared groups (i.e., groups represented by cyan colour: Salmonella, Campylobacter, and Listeria) with either of the interval estimates for the reference group, that confirms that no mean significant difference existed between the reference group and the groups under consideration. Hence, there were no mean significant differences between 3 of the 4 groups with reference to E. coli (i.e., no mean significant difference existed between the following group pairs (Salmonella and E. coli, Campylobacter and E. coli, and Listeria and E. coli).

Furthermore, the reliefF algorithm of the Matlab statistical software (vR2013a) was applied to determine the weight contribution of the predictors (number of days, pH, and average daily temperature) to the response/output to describe the influence as either primary or secondary. The values of the weight contribution of the predictors are as shown in

Table 2. Weight contribution of the predictors.

Bacteria Species	Predictors
	Weight Contribution by Number of Days	Weight Contribution by PH	Weight Contribution by Average Daily Temperature
Escherichia coli	0.174	0.077	−0.040
Salmonella sp.	0.152	0.056	−0.034
Campylobacter sp.	0.109	0.101	−0.028
Y. enterocolitica	0.193	0.123	0.024
L. monoctogenes	0.311	0.061	0.003

. For all the bacterial species, the predictors (number days and pH) were considered as the primary factors that influenced log bacterial counts, while the average daily temperature was regarded as either a primary or secondary factor based on the particular bacteria.

4. Discussion

Although the microbial composition of animal manure varies appreciably with the type of animals, the geographical locations, and the individual practices regarding treatment facilities, the bacterial group is often considered for assessment [23]. This is attributed to the relative ease of analysis and quantification as well as the epidemiological significance of the category of organisms; therefore, they are well characterised and attract great interest [31]. In this light, we determined the survival of five selected bacterial pathogens of known environmental and public health concern during the psychrophilic anaerobic degradation of pig manure and sawdust in a batch reactor. In addition, most of these bacteria are zoonotic pathogens causing infections in animals as well as humans, hence attracting great attention [18]. The data obtained in this study revealed the initial viable counts of the bacterial pathogens in magnitude ≥ 10⁴ ≤ 10⁷ cfu/g, which indicated the potential of sanitary, ecological safety, and microbiological risk. In addition, the presence of Enterobacteriaceae plus the total number of Gram-negative bacteria suggested an accurate reflection of the hygienic status of the biowaste [18]. Therefore, should these wastes in their initial form with such a level of bacterial counts (microbial and physiochemical) be discharged into the environment, the aforementioned deleterious effects would be unavoidable [6,12]. In this light, these wastes via deliberate or accidental discharge occurring in the presence of hydrological drivers, including storm, rainfall, or irrigation and pathways, could directly enter the soil and aquatic ecosystems, having implications for food safety, environmental security, and trade, as well as human and ecosystem health [39], hence the significance of the employment of anaerobic digestion technology to help reduce the microbial load of these zoonotic pathogens in the wastes that will eventually end up in agricultural lands.

Moreover, the effectiveness of the anaerobic digestion process for ecological purposes is associated with the level of microbial decontamination measured in terms of faecal coliforms and indicator bacterium [18]. Specifically, bacterial counts less than the detection limit (DL = 10²) was noted on certain days depending on the particular pathogen and that specific day was described as the survival time of that bacterial pathogen. Accordingly, the survival time represented that specific day in which no occurrence of bacterial colonies or growth was observed on the agar plate inoculated with the substance from the very first dilution (10⁻¹ or 1:10). Initially, only 100 μL of the 10⁻¹ dilution (100/1000 μL) was plated and not 1 mL. Therefore, to complete a millilitre, an additional 10⁻¹ dilution was needed. Finally, the effective dilution factor for the very first dilution of the 10-fold series was 10⁻¹ × 10⁻¹ = 10⁻². Remarkably, Sutton [40] mentioned that the bacterial counts on such plates should be considered to be less than the DL = 10².

It is obvious that over a specified period, which depended on the bacteria, viable bacterial cells were appreciably inactivated during anaerobic digestion, as presented in Figure 2. Overall, a 1 log reduction in viable bacterial cells occurred with each bacterial pathogen and could also be described as a 90% reduction in viable bacterial counts. This indicates a direct reduction in the bacterial load and contamination potential of the co-digested substrates, making the co-digestate better than the original substrates in terms of enteric bacterial load. The rate and efficiency of transport of these pathogens found in manure into water bodies relies on the actual number of the pathogens shed and the release rate of the pathogens from manure [41]. The release of the pathogens from manure is a crucial factor, as it determines the availability of the pathogen for transport into environmental matrices and relates to the manure composition [42]. Therefore, owing to the 90% reduction in the enteric bacteria, the potential of the co-digestate to reach humans via water contamination and subsequently cause infections has been lessened to a greater degree as it depends on the infective dose of each bacterium; the lowest infective dose differs from one pathogen to the next. More challenging is the viability and the survival of many of the enteric pathogens outside the host organisms [43]. Accordingly, Black et al. [27] emphasised that the management of pathogen load via either anaerobic digestion or composting is a potential route to reduce and prevent outbreaks of infection with E. coli, Salmonella, Campylobacter, and Listeria; however, such studies should be considered as region-specific because of the variable parameters affecting the pathogen content and the survival in manure.

In addition, it is emphasized that a reduction of one logarithmic unit or the time during which viable counts of a population of microorganisms are reduced by 90% is termed the decimation reduction time (T₉₀). Vinnerås [44] reiterated further that the time at T₉₀ can be measured in hours and days, which is consistent with thermophilic and mesophilic digestion, respectively, and designates the variations in inactivation of bacterial pathogens in anaerobic digestion. The T₉₀ values for E. coli, Salmonella sp., Y. enterocolitica, Campylobacter sp., and L. monocytogenes were 77 days, 84 days, 98 days, 112 days, and 175 days, respectively. The time (days) related to the percentage of decay of the bacterium and the disparity in the number of days could be attributed to the differences in bacterial biology, nutrient availability, and characteristics of the biological process [18,23].

The data showed that L. monocytogenes, a Gram-positive bacteria belonging to the phylum Firmicutes, displayed the longest survival period as opposed to the other four Gram-negative bacterial pathogens, which belong to the phylum Proteobacteria. This substantiates the findings of Jiang et al. [18], which indicated that the inactivation of Gram-negative bacteria occurred rapidly during psychrophilic anaerobic degradation of pig manure as opposed to Gram-positive bacteria. Clearly, the peptidoglycan cell wall of bacteria offers protection and resistance to stress and hostile environmental conditions; however, the bacteria belonging to the Gram-positive category are walled by several layers of peptidoglycan and are, therefore, much thicker than those available in Gram-negative bacteria and also consist of teichoic acids [23]. The manifestation of such tendency by L. monocytogenes was also highlighted by Sibanda and Buys [45] who remarked on the potential of the bacterium to thrive in vast environmental conditions (pH, water, and temperature) resulting in a greater survival ability.

Furthermore, within the Phylum Proteobacteria, Campylobacter species (class Epsilonproteobacteria; family Campylobacteriaceae) had a prolonged survival time as opposed to the other three bacteria (class Gammaproteobacteria; family Enterobacteriaceae) from the same phylum. This may suggest that Campylobacter species were exposed to a lesser degree of competition regarding nutrients as they utilise amino acids, vitamins, and citric acid cycle intermediates that are not consumed by other bacteria, whereas E. coli, Salmonella sp., and Y. enterocolitica faced stiff competition from other nonmethanogenic bacteria (indigenous microflora) for carbohydrates [46]. In addition, Campylobacter species are able to acquire several transition metals necessary for metabolic activities via diverse mechanisms [47].

It is well known that the inactivation of bacteria is temperature-dependent; likewise, the microbial activities conducted by anaerobic microbes during anaerobic digestion are temperature-dependent. This is so because enzymes secreted by the anaerobes are responsible for the degradation process, especially at the hydrolysis stage [25]. Accordingly, Jiang et al. [18] reported a higher viable cell reduction caused by anaerobic digestion occurring at thermophilic rather than mesophilic conditions. Orzi et al. [46] reported that the temperature may cause a direct impact on the pathogen or enhance the effects of other chemicals such as volatile fatty acids toxicity and ammonia, although not investigated in this study. Nevertheless, in the present study, the inactivation of the aforementioned bacterial pathogens occurred in the psychrophilic temperature range (lower temperatures), suggesting that factors other than temperature, including pH, volatile fatty acids, and treatment retention time, could play a crucial role added to the progress of the anaerobic digestion process [18,23]. In a nutshell, the inactivation of pathogens via psychrophilic anaerobic digestion cannot be categorically ascribed to a specific reason but might probably be due to a combination of factors. However, the contribution to pathogen reduction by specific parameters during anaerobic digestion has been deliberated elsewhere [18,46].

Following the findings of Orzi et al. [46] which noted the various factors influencing the inactivation of bacteria during anaerobic digestion, we decided to develop multiple linear regression models with the number of days, average daily temperature, and pH as predictor variables to shed light on the impact of these variables on the response (log bacterial counts). Mathematical modelling is a process whereby real-life situations are simulated with mathematical equations to describe relationships between predictors and responses, as well as to forecast the future behaviour of the parameters involved in the process. From Table 2, A0 denotes the forcing constant, which caters for the contribution from other factors/parameters, though not investigated and incorporated in the development of the mathematical models but would equally have affected the bacterial count. From mathematical induction, a negative scaling constant of the predictor variable indicates an inverse relationship with the response, while a positive scaling constant of the variable demonstrates a direct relationship with the response.

By inference to Table 2, overall, the predictor variable, number of days, was inversely related to log bacterial counts for the models built for all the bacterial species except for L. monocytogenes, which indicated that an increase in the number of days could cause a decrease in the log bacterial counts. This finding is congruent with the work of Biswas et al. [48] who reported the complete inactivation of E. coli in slurry manure stored in tanks operated under lagoon conditions at 14 weeks while Listeria survived until the end of the experiment (29 weeks), although in relatively low levels. Similarly, with the exception of Salmonella sp., an increase in pH caused a corresponding increase in log bacterial counts. It is known that bacteria are well suited to live and thrive in an alkaline to neutral milieu [49]. In addition, the temperature was directly related to log bacterial counts according to the models developed for most of the bacterial species. Hence, an increase in temperature led to a corresponding rise in log bacterial counts. Clearly, metabolic processes in microorganisms are enzyme-mediated and, therefore, are influenced by changes in temperature [25]. This finding supports the study of Scofield et al. [50] which noted changes in bacterial metabolism in relation to a rise in the temperature of water in the lagoon. However, this was not the case with L. monocytogenes, a situation established by the study of Bevilacqua et al. [51] in which an increase in temperature resulted in a decrease in the survival rate of the pathogen from 18 days to 3 days in brine. In addition, the finding supports the view that L. monocytogenes, a psychrotolerant organism, survives and multiplies at low temperatures, reaching high levels in contaminated foods kept under refrigeration for prolonged times [25]. Concisely, the impact of the predictor variables on the log bacterial counts depended on the particular bacteria and the highest p-value of 0.979 for Listeria indicated the statistical significance of its model. Data on the multiple comparison test demonstrated a significant difference in the group means of log E. coli and Yersinia counts, Campylobacter and Yersinia counts, and E. coli and Campylobacter and Yersinia counts; however, no group means exhibited a significant difference with Salmonella and Listeria.

Following the reliefFtest, the influence of each predictor variable on the log bacterial counts or the strength of the relationships between the predictor variables and the response variable was determined per bacterial species. According to Tangwe et al. [38], the rerliefF test ranked the predictor variables by weight of importance to the response. The weight rank was between −1 and 1. If the weight rank obtained for a predictor was 1, it suggested that there existed a strong correlation between the predictor and response; however, if −1, it followed that no correlation existed between the predictor and response. Furthermore, a negative value associated with the weight indicated that the predictor variable was a secondary factor, but when positive, it was a primary factor influencing the response. The results for the reliefF test were as shown in Table 2. It is clear from Table 2 that average daily temperature, number of days, and pH were primary factors that influenced the log of Yersinia counts and log of Listeria counts, while the number of days and pH were primary factors affecting the log of E. coli, log of Campylobacter, and log of Salmonella counts.

Nonetheless, following the findings unravelled in this study, future studies are anticipated to determine the quantity/level of N, P, and K in the discharged effluent or digestate to further ascertain the hygienic status of the co-digested substrates, which can now be employed for land fertilisation. Further studies are needed to determine the presence of antibiotics and antibiotic resistance genes via molecular-based methods in order to halt the dissemination of antibiotic resistance to the environment and through food chains and water to humans.

5. Conclusions

The findings of this study reiterated the prevalence of notorious bacterial pathogens of great environmental and public health significance in levels between 10⁴ and 10⁶ cfu/g in biowastes, thus emphasising the necessity of appropriate management practices to minimise the transfer of bacterial pathogens via the food chain to humans. The reduction (90–99.9%) in numbers of isolates of E. coli, Salmonella, Yersinia, Campylobacter, and Listeria via anaerobic digestion over a varied period from 77 to 175 days greatly improved on the microbiological quality of the digested substrates. Therefore, considering the different number of days taken for the different bacterial pathogens in the animal manure to be inactivated, it is glaring that regardless of the temperature, anaerobic digestion of biomass in a single-stage steel biodigester is a sure and effective way of bacterial pathogen control. Summarily, psychrophilic anaerobic digestion could be considered a more economical approach to the management of animal wastes as Akindolire et al. [25] affirmed that it lessens or abolishes the energy expenditure and cost associated with heating digesters in cold climatic conditions, offers a more stable process in the treatment of ammonia-rich waste such as animal manure, creates the likelihood of extending anaerobic digestion technology to areas of cold climates wherein unheated digesters are prevalent, and serves an integrated-based solution to agricultural waste management for farmers operating ambient-temperature digesters.

6. Limitations of the Study

• It should be stressed that even if the level of indicator pathogens is below the detection limit, this cannot be directly translated to the absence of potential pathogenic risk, due to the occurrence of other pathogens [52].
• The study demonstrates its findings on the inactivation of viable and culturable bacterial pathogens, whereas other bacteria, e.g., spore-forming pathogens (Bacillus, Clostridium, etc.), viable but nonculturable pathogens, as well as pathogenic fungi, which also have significant environmental and public health consequences, were not evaluated. It is challenging and time-consuming to detect all the pathogens in the manure. Nevertheless, zoonotic pathogens should be given more attention as they can be transferred between animals and humans, causing threats to animals and human health [23]. Therefore, this study focussed on a host of zoonotic pathogens commonly used as indicator bacteria.
• Following the aforementioned information, molecular-based techniques should be employed alongside cultural methods with viable bacterial populations via anaerobic digestion so that a wide range of microorganisms are detected as well to give the actual level/concentration (including viable but nonculturable bacteria) of the microorganism occurring in the biodigester.
• The levels of volatile fatty acids and ammonia regarded as inactivation factors were not investigated.