Test | Data | Salmon vs SalmoSim | Salmon vs inoculum | SalmoSim vs inoculum |
|---|
UniFrac | Unweighted (0%) | All | 0.001 *** | 0.002 ** | 0.002 ** |
Subset | 0.001 *** | 0.002 ** | 0.001 *** |
Core OTUs | 60% | 0.04 * | 0.032 * | 0.143 |
50% | 0.001 *** | 0.001 *** | 0.033 * |
40% | 0.001 *** | 0.003 ** | 0.244 |
30% | 0.001 *** | 0.001 *** | 0.005 ** |
Balanced (50%) | All | 0.001 *** | 0.003 ** | 0.001 *** |
Subset | 0.001 *** | 0.001 *** | 0.003 ** |
Core OTUs | 60% | 0.138 | 0.059 | 0.12 |
50% | 0.002 ** | 0.019 * | 0.041 * |
40% | 0.002 ** | 0.062 | 0.132 |
30% | 0.001 *** | 0.005 ** | 0.008 ** |
Weighted (100%) | All | 0.012 * | 0.007 ** | 0.003 ** |
Subset | 0.012 * | 0.007 ** | 0.004 ** |
Core OTUs | 60% | 0.381 | 0.063 | 0.125 |
50% | 0.008 ** | 0.217 | 0.078 |
40% | 0.023 * | 0.467 | 0.122 |
30% | 0.021 * | 0.014 * | 0.06 |
Bray–Curtis | All | 0.001 *** | 0.001 *** | 0.23 |
Subset | 0.001 *** | 0.001 *** | 0.273 |
Core OTUs | 60% | 0.009 ** | 0.004 ** | 0.079 |
50% | 0.001 *** | 0.008 ** | 0.394 |
40% | 0.001 *** | 0.002 ** | 0.327 |
30% | 0.001 *** | 0.001 *** | 0.388 |
Jaccard | All | 0.001 *** | 0.001 *** | 0.147 |
Subset | 0.001 *** | 0.001 *** | 0.161 |
Core OTUs | 60% | 0.002 ** | 0.003 ** | 0.073 |
50% | 0.001 *** | 0.002 ** | 0.386 |
40% | 0.001 *** | 0.002 ** | 0.22 |
30% | 0.001 *** | 0.001 *** | 0.254 |
| Â | Salmon | SalmoSim | Salmon | Inoculum | SalmoSim | Inoculum |
Number of samples | 18 | 54 | 18 | 9 | 54 | 9 |
Number of OTUs | 192 | 508 | 192 | 388 | 508 | 388 |
Number of shared OTUs | 139 | 139 | 131 | 131 | 291 | 291 |
Number of reads | 78,400 | 1,004,494 | 78,400 | 192,429 | 1,004,494 | 192,429 |
Shared OTU reads | 77,123 | 707,199 | 76,963 | 134,984 | 989,884 | 187,569 |
% shared OTU reads | 98.37% | 70.40% | 98.17% | 70.15% | 98.55% | 97.47% |
- The table summarises different beta diversity analysis outputs calculated by using different distances: phylogenetic (unweighted, balanced, and weighted UniFrac) and ecological (Bray–Curtis and Jaccard), between different samples (data from all gut compartments combined): real salmon (Salmon), SalmoSim inoculum from the real salmon (Inoculum), and SalmoSim (only stable time points: 16, 18, and 20 fed on Fishmeal diet and 36, 38, and 40 fed on Fishmeal-free diet). A permutational multivariate analysis of variance (PERMANOVA) by using phylogenetic and ecological distances was performed to determine if the separation of selected groups is significant as a whole and in pairs. Numbers represent p values, with p values < 0.05 identifying statistically significant differences between compared groups. The comparisons are shown for 3 different datasets: all (completed data set containing all the OTUs sequenced: 978 OTUs in total), subset (containing OTUs that appear only in more than 3 samples and contribute to 99.9% of abundance within each sample: 374 OTUs in total), and core OTUs (containing OTUs that appear in 60% (6 OTUs in total), 50% (13 OTUs in total), 40% (34 OTUs in total) and 30% (65 OTUs in total) of the samples). The asterisks flag the levels of significance: one asterisk (*) for p values between 0.05 and 0.01, two asterisks (**) for p values between 0.01 and 0.001, and three asterisks (***) for p values below 0.001. Finally, the bottom of the table compares number of samples, OTUs, reads for each sample group, as well as number of shared OTUs and their reads within each sample within compared groups. It also summarises what percentage of a given group of samples’ total reads came from the shared OTUs. The SalmoSim samples used for this test consist of stable SalmoSim time points: days 16, 18, and 20 (Fishmeal diet—once bacterial communities adapted to the SalmoSim environment) and days 36, 38, and 40 (Fishmeal-free diet—once bacterial communities adapted to feed change). For non-inoculum real salmon, all samples were included (fed on both Fishmeal and Fishmeal-free diets), and for inoculum real salmon, all samples were included (fed on Fishmeal diet)
