Effect of weaning age on piglet weight and occurrence of diarrhea
Forty-eight Large White piglets (23 females and 25 males) were divided into four groups of 12 animals that were weaned at different ages: 14 days (W14), 21 days (W21), 28 days (W29), and 42 days (W42). These groups are hereafter referred to as the weaning groups. Animals presenting diarrhea were unevenly distributed across groups, with a strong reduction in the proportion of affected animals in the groups W28 and W42: 3/10 (30%) in the W14 group, 5/12 (41%) in the W21 group, 1/12 (8%) in the W28 group and 0/11 (0%) in the W42 group. A Chi-square test confirmed that these differences were significant (p<0.05).
To characterize piglet growth, we monitored the weight of pigs in each weaning group from birth (day 0) to 62 days of age (weight was measured at 5, 12, 20, 27, 33, 48, 55, and 62 days of age). Using ANOVAs, we found that the weaning groups differed in weight across time and that patterns of differences varied (Table S1). In general, after weaning, the mean weight for the W14 group was consistently lower than the mean weights for the other groups (Figure 1). In addition, piglets in the groups W14 (at day 20), W21 (at day 27), and W28 (at day 33) lost weight immediately after weaning. Indeed, three animals from the W14 group were euthanized because they were lethargic and failed to grow (decision made in accordance with the project’s established ethical guidelines). On day 62, the mean weights for the groups W21, W28, and W42 were statistically similar to each other, and they all differed from the weight for the W14 group (p<0.05).
Figure 1: Growth curves for piglets weaned at 14 days of age (W14), 21 days of age (W21), 28 days of age (W28), and 42 days of age (W42). The solid and dashed lines show each group’s mean and standard deviation, respectively. The initial sample sizes for each group were as follows: W14: 10 animals, W21: 12 animals, W28: 12 animals, and W42: 10 animals. The samples sizes for each group after weaning were as follows: W14: 4 animals, W21: 6 animals, W28: 6 animals, and W42: 5 animals. Any statistical differences between groups are indicated by different letters in each time point, and further details can be found in Table S1.
Fecal microbiota sequencing, OTU identification and annotation
The piglets’ fecal microbiota were analysed by sequencing the bacterial 16S rRNA gene using an Illumina MiSeq Sequencer. Samples with fewer than 10,000 reads following quality control procedures were removed from the analysis, resulting in sample sizes of 3–12 piglets per sampling point (see the Methods section). After performing quality control, a mean of 63,716 reads were available for each sample. Sequences from the whole sample set were successfully clustered into 1,121 operational taxonomic units (OTUs), and only 0.26% of the OTUs could not be assigned to a given phylum. Overall, 539 of the 1,121 OTUs (48%) were assigned to a genus. The phyla Firmicutes (700/1,121) and Bacteroidetes (340/1,121) represented 62% and 30% of the OTUs, respectively. Within the phylum Firmicutes, 95% (665/700) of the OTUs were assigned to the order Clostridiales, 40% (265/665) to the family Ruminococcaceae, and 23% (153/665) to the family Lachnospiraceae. Within the phylum Bacteroidetes, 53% (179/340) were assigned to the genus Prevotella. Other phyla were also represented, but they were less common (e.g., Proteobacteria: 5%, Spirochaetes: 0.45%, Fusobacteria: 0.45%, Actinobacteria: 0.35%, Deferribacteres: 0.27%, and Tenericutes: 0.01%; Figure 2A). At the phylum (Figure 2A) and genus (Figure 2B) levels, the overall abundance of diverse OTUs varied based on weaning age and among sampling points within weaning groups (see the following sections). When we examined the 75% most prevalent taxa in each group at the three sampling points, we found that, out of the 1,121 OTUs observed overall, 760 OTUs were present in the W14 group, 807 OTUs were present in the W21 group, 882 OTUs were present in the W28 group, and 933 OTUs were present in the W42 group. This result illustrates that OTU richness increased with age at weaning.
Figure 2: Relative abundance of the different microbial phyla (A) and genera (B) at each sampling point for every individual pig in each weaning group. Only genera present in at least 20% of the piglets are shown.
Effect of weaning age on fecal microbiota diversity and composition before and after weaning
lpha diversity, beta diversity, and richness were calculated using the rarefied OTU counts for each group and then compared among weaning groups and sampling points (Figure 3). ANOVAs and Tukey’s honest significant difference (HSD) tests were used to assess any resulting differences (Table S2). Overall, there were significant differences (p<0.05) in alpha diversity and richness among sampling points within all the weaning groups except W42. In the W42 group, only beta diversity differed significantly among sampling points. The results for alpha diversity and richness reflect the diversification that takes place in the gut microbiota during and after weaning. The results for beta diversity fit with the idea that microbiota heterogeneity declines as animals grow older. The Tukey’s HSD tests highlighted that the significant differences mainly post-weaning sampling points. Moreover, we observed that beta diversity declined between 7 days post weaning and 60 days of age, except in the W14 group (Figure 3B).
Non-metric multidimensional scaling (NMDS) analyses were carried out using Bray-Curtis dissimilarity values quantifying overall differences in gut microbiota composition between samples collected before weaning, 7 days after weaning, and at 60 days of age for piglets in each weaning group (Figure 4). For the groups W14, W21, and W28, there were clear differences between the results for the three sampling points. For the group W42, in contrast, the centroid for the pre-weaning data was distinct from the centroids for the data from 7 days post weaning and 60 days of age, which overlapped.
We used the metagenomeSeq package in R to identify differentially abundant (DA) OTUs within the full dataset (1,121 OTUs) for each weaning group; we specifically compared the pre-weaning data and the data obtained 7 days after weaning. In the W14 group, there were 224 DA OTUs (Table S3). In the W21 group, this number increased to 484 (Table S4). In W28 and W42, there were 395 DA OTUs (Table S5) and 461 OTUs (Table S6), respectively. There was some degree of overlap among the DA OTUs (Figure S1), although there were unique OTUs in all the weaning groups (W14: 44, W21: 106, W28: 71, and W42: 107). Overall, Bacteroides, Ruminococcus, Oscillospira, and Clostridium were more abundant before weaning and Succinivibrio, Prevotella, and Campylobacter were more abundant 7 days after weaning. Interestingly, Faecalibacterium prausnitzii was found to be highly abundant after weaning in all the weaning groups.
Figure 3: Boxplots for alpha diversity (A), beta diversity (B) and richness (C) for each sampling time point in animals weaned groups at 14 days of age (W14), 21 days of age (W21), 28 days of age (W28) and 42 days of age (W42). Statistical differences are included in the figure. Significative values are reported as follows: * (p<0.05); ** (p<0.01); *** (p<0.001).
Effect of weaning age on F. prausnitzii abundance before and after weaning
In the full dataset, three OTUs were annotated as F. prausnitzii (OTU IDs 851865, 350121, and 525215). Since at least one of these OTUs was DA in most comparisons, we decided to explore the overall abundance of F. prausnitzii by summing the abundances of the three OTUs for each sample. We had previously normalized these data by log scaling the cumulative sum scaling (CSS) values obtained in metagenomeSeq. For each weaning group, there was a clear increase in F. prausnitzii abundance over time, and the highest abundances were observed in the W42 group (Figure 5). In the groups W14 and W21, there was a marked increase in abundance between weaning and 60 days of age; in the groups W28 and W42, abundance tended to be more stable 7 days post weaning. At weaning, F. prausnitzii was most abundant in the W42 group, equivalently abundant at lower levels in the W21 and W28 groups, and least abundant in the W14 group. There were significant differences among the four weaning groups (ANOVA: p<0.05), and F. prausnitzii was more abundant before weaning in piglets weaned at a later age (Table S7). Indeed, piglets weaned at 14 days of age had the lowest abundance of F. prausnitzii before weaning, a pattern that persisted until 60 days of age. Post-hoc analysis found differences in the abundance of F. prausnitzii between the groups W14 and W42 before weaning and between various combinations of the weaning groups at 7 days post weaning and 60 days of age (Table S8).
Table of contents :
1. Pig production
a. Pig data in the World
b. Data on pig production in Europe, Italy and France
c. Pig production systems
d. Antibiotic practises in pig herds and the antibiotic resistance issue
i. Antibiotics consumption data in Europe
2. The gut microbiota
a. The gut microbiota in pigs
b. Gut microbiota and gut health: what is a healthy gut microbiota?
i. Eubiosis, dysbiosis and symbiosis
ii. Probiotics, Prebiotics and Synbiotics
iii. Next-generation probiotics from microbiota studies: the example of Faecalibacterium prausnitzii
3. Health and disease
a. Diseases in pig production
b. Weaning enteric diseases
i. The Enterotoxigenic Escherichia coli infection in piglets
ii. The MUC4 and FUT1 candidate genes in piglets
9. Supplementary information
3. Materials and Methods
8. Supplementary information
1. On microbiota’s role in the post-weaning diarrhoea of piglets
2. On technical choices for the study of gut microbiota in our work
3. Summary of challenges and opportunities for the prevention of diarrhoea at weaning