Today, we are sharing a recent publication from the Noyes lab investigating a very popular sampling technique, pooling and the impact it has on resulting microbiome profiles. The article is available in open access on Microbiology Ecology’s website.
Methods
- Samples were taken at a commercial farrow-to-wean facility.
- Ten litters of day-1 piglets and ten litters of 20 day-old piglets were enrolled in the study.
- The sows had not received any antibiotic treatment starting one week prior to farrowing and piglets had not received any antibiotic treatment.
- A rectal swab was collected for every piglet, totaling 258 swabs.
- Pen-level samples were composed of 3 swabs collected from three different areas of the farrowing stall.
- Fecal material from 4 random swabs in a litter were pooled together to create the “fecal pools”.
- Another set of pools was created by pooling samples after the DNA extraction step was completed. The samples from the same 4 individuals were pooled to make a “DNA pool”.
Results
- The average number of raw sequences was similar across individual samples, fecal pools, and DNA pools for both day-old and 20-day-old piglets.
- The fecal microbiome recovered from both types of pools (fecal and DNA pools) was similar to the one recovered in individual samples.
- Conversely, the fecal microbiome recovered from pen-level samples was found to be very different, suggesting that environmental samples would be inadequate to characterize piglet fecal microbiome.
Abstract
Pooling individual samples could be an efficient approach for large-scale population-based microbiome studies. However, it is unknown whether pooled samples accurately reflect the microbiome composition and diversity obtained from individual samples. This study investigated the impact of various pooling methods on the observed fecal microbiome of preweaned piglets. Individual fecal samples were collected from 10 litters of day-old piglets ( N = 137) and 10 litters of 20-day-old piglets ( N = 121), as well as pen-floor samples from the same litters. The individually collected samples were processed individually and also used to create pools of both raw feces and extracted DNA. Individual samples, raw feces pools, DNA pools, and pen-floor samples were subjected to 16S rRNA gene sequencing. The microbial profile in pen-floor samples from litters of preweaned piglets was very different from individual piglet samples within the pen; thus, they may not be suitable for litter-level piglet microbiome studies. However, overall microbial diversity and composition from DNA and feces pools were comparable to individual samples, despite potentially underestimating some low-abundance or low-prevalence taxa. These results suggest that pooling can be used as an efficient and cost-effective approach to characterize litter-level microbial profiles for current and future population-level microbiome research in preweaned piglet populations.