Rotavirus A (RVA) is among the leading etiological agents of porcine gastroenteritis and causes stunted growth among piglets. Moreover, there is increasing evidence for zoonosis of RVA, which is also the principal cause of diarrhea in children. In the absence of rigorous animal health monitoring in Philippine backyard farms, there is, therefore, a strong need for RVA surveillance.
Backyard farms in Davao City, Philippines were subjected to 12-month inspection on RVA. Monthly farm-level detection rate ranged from 0 to 52%, with an overall average of 23%. RVA was most prevalent in non-diarrheic stools, indicating asymptomatic circulation of the virus. Spatiotemporal analysis demonstrated that the viral distribution exhibits a seasonal pattern that peaks and forms geographical clusters during the colder months of the year, suggesting farm-to-farm transmission.
Risk factor analysis identified specific farm conditions that increase the likelihood of RVA circulation: the presence of finishers and gilts, larger herd size, the presence of other animals, and abiotic factors such as low relative humidity and low altitude. The same analysis also revealed three dominant management practices that can help reduce the pressure of infection in these farms: sanitation and waste disposal, animal grouping, and diet and nutrition. This new perspective on porcine RVA circulation will benefit the underprivileged backyard farmers and help empower them to protect both animal and public health.
Keywords: porcine rotavirus, surveillance, spatial autocorrelation, risk factors, backyard farms
This project was funded by the University of the Philippines Balik-PhD Program. The authors would like to thank the local government officials and farmers who granted permission to conduct the study in the area.
Swine production is the second largest agricultural industry in the Philippines after rice and is dominated by backyard farms, which produce around 70% of the total pork in the country (Stanton, Emms & Sia, 2010). However, backyard pig farming faces significant challenges such as widespread malnutrition, low reproductive performance, and high mortality, mainly due to lack of accessible information on essential pig management and husbandry (Riedel et al., 2012; Matabane et al., 2015).
One of the economically critical porcine diseases is enteritis, which causes morbidity and mortality in young piglets. Rotavirus A (RVA) is among the leading etiological agents of pig enteritis (Vlasova et al., 2017). The virus belongs to the family Reoviridae and has a genome consisting of 11 segments of double-stranded RNA. Its genome encodes six structural (VP1 to VP4, VP6, and VP7) and six non-structural proteins (NSP1 to NSP 6) (Estes and Cohen, 1989). RVA infections are characterized by viral replications in small intestinal enterocytes causing villous blunting and thus, leading to poor nutrient absorption and incomplete digestion (Vlasova et al., 2017).
It is also shed in the feces following intestinal infection (Janke et al., 1988). The virus is transmitted through the fecal-oral route, exhibits a high degree of stability, and can maintain infectivity despite prolonged exposure to harsh environmental conditions (Murphy et al., 1983; Ramos et al., 1998). The virus also exhibits a high degree of diversity due to frequent genetic reassortment events (Estes and Cohen, 1989; Vlasova et al., 2017). Rotaviruses are host-specific but have been shown to cross-species barrier among avian, bovine, porcine, and rodent species, including humans (Vlasova et al., 2017). Potential zoonotic transmission of RVA from pigs to humans has been reported (Midgley et al., 2012). Cook et al. (2004) speculated that there is the intermittent but low-level introduction of animal rotaviruses into the human population that can still lead to infection.
RVA circulation in underprivileged backyard pig farms may be problematic as there is no awareness and no means to address the build-up of infection pressure. In the absence of rigorous animal health systems in Philippine backyard farms, there is, therefore, a need for robust surveillance and control of RVA. Thus, this study aimed to investigate the temporal and spatial circulation of RVA in backyard pig farms in Davao City, Philippines and determine the drivers of infection.
This will benefit small-scale animal farmers in the area by providing them with a local perspective on RVA distribution and spread while empowering them with control strategies against the virus. This is also consistent with promoting public health interest as rotaviruses remain to be the leading cause of diarrhea-associated deaths in children below five years of age (Parashar et al., 2009).
Materials And Methods
- Farm Selection and Survey
A total of 30 backyard pig farms were selected through stratified random sampling from 831 farms in 21 barangays of the districts Calinan, Tugbok, Toril and Bunawan in Davao City. The selection criteria used was: 1) herd size not beyond 20 heads, excluding piglets, to be considered as a backyard farm; 2) no prior use of rotavirus vaccine; 3) cemented or wooden flooring to eliminate soil contamination of fecal samples, and 4) located within the 15 km radius of the Philippine Atmospheric Geophysical and Astronomical Services Administration (PAGASA) stations.
Permission was requested from each barangay, and informed consent was secured from the farm owners before the conduct of the study. Data on farm management practices and environmental characteristics were collected on a monthly basis from November 2016 to October 2017, while data on weather variables for the same period were obtained from PAGASA stations in Davao City. The GIS coordinates of the farms were also collected.
- Fecal Sampling
A total of 275 freshly voided swine fecal samples were collected from the farms between November 2016 and October 2017. In case a farm had more than one pig pen, the fecal sample was obtained from the pen with the most pigs. The total number of pigs per farm (herd size) and some pigs in the sampled pen (pen size) were noted. The age group of the pigs was also recorded. Stool consistency was determined as solid or hard (regular), semi-solid or soft (loose), and watery (diarrheic). All feces from a pen were pooled to represent farm-level detection. Fecal samples were labeled accordingly with the farm ID and date of acquisition and stored at -80°C.
- RNA Extraction
A 10% (w/v) fecal suspension was prepared in DNA/RNA Shield (Zymo Research, California, USA). Viral RNA was extracted from the fecal suspension using the QIAmp viral RNA extraction kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions.
- Reverse Transcription Nested PCR
Reverse transcription was performed using the QuantiNova Reverse Transcription Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. Nested PCR of the VP6 gene was performed for the cDNA using the 2X Taq Master Mix (Vivantis, Subang Jaya, Malaysia) and RVA-specific primers which have been tested against various RVA strains, including porcine RVA (Elschner et al., 2002). The first round of PCR was performed using five μL 2X Taq mix, 0.5 μL of 10 μM each of RVA VP6 F2 (5’ AAGATGCTAGAGACAAAATTGT 3’) and RVA VP6 R2 (5’ AATCAGATTGTGGTGCTATTCC 3’), and four μL of cDNA for a total of 10 μL per reaction. The cycling conditions were: 2 min at 94°C, 35 cycles of 30 s at 94°C, 30 s at 51°C, 30 s at 72°C, and 5 min at 72°C. Nested PCR was conducted using the previously mentioned protocol for inner primers RVA VP6 nF1 (5’GACAAAATTGTCGAAGGCACATTATA 3’) and RVA VP6 nR1 (5’ TCGGTAGATTAC CAATTCCTCCAG 3’) with the following cycling conditions: 2 min at 94°C; 35 cycles of 30 s at 94°C, 30 s at 54°C, and 30 s at 72°C; and 5 min at 72°C. PCR products were visualized using agarose gel electrophoresis, and a couple of the amplicons were verified by DNA sequencing (Online Resource 1).
- Descriptive Statistics
The rate of farm-level RVA detection was determined according to month, stool consistency, pen size, and age class relative to the sample size of each category.
- Mapping and Spatial Analysis
GIS coordinates were used in ArcGIS 10.3 (Esri, Redlands, California) to generate maps. Moran’s Index (Moran’s I), a global spatial autocorrelation method, was employed to evaluate the pattern of distribution of RVA-positive farms using the Spatial Statistics Tools of the ArcGIS software. Inverse distance, fixed distance (with threshold distances of 2,000 meters, 2,500 meters, and 3,000 meters), Delaunay triangulation, and K-nearest neighbors (k=5, k=6, and k=8) were used to impose structure on the spatial relationships of the RVA detection.
The Inverse distance method assumes that the impact of one feature on another feature decreases with distance. The Fixed distance method assumes that all other elements within a specified critical distance of each component are included in the analysis bearing the same impact, and all features outside the critical distance are excluded. The K-nearest neighbors method chooses the closest k features to include in the report where k is a specified numeric parameter.
The Delaunay triangulation is a model for a mesh of nonoverlapping triangles created from feature centroids and those features associated with triangle nodes that share edges are neighbors. Spatial weight matrices were obtained using these methods and were used to evaluate Moran’s I. These parameters were chosen so as to follow the rule of thumb which aims for a spatial weight matrix where every feature has at least one neighbor, most have about eight neighbors, and that every element do not have all the other parts as its neighbors (Er et al., 2010). The distribution is dispersed if the value of Moran’s I am -1 or close to -1, clustered if 1 or close to 1, and random if 0. Spatial autocorrelation test was done for each of the twelve months.
- Risk Factor Analysis
Farm-level circulation of the virus was scored based on RVA detection in pen, wherein detection of RVA = 1 and non-detection of RVA = 0. The effect of socio-demographic variables, management practices, environmental characteristics, and weather variables on the likelihood of RVA incidence were estimated using Stata 13 (StataCorp LLC, Texas, USA) via three regression models: 1) standard logit and standard probit, 2) panel logit and panel probit random effects, and 3) panel logit fixed effects.