Table 1. OTU numbers in the raw and filtered data sets.
| Clustering threshold | Region (n) |
OTU richness |
Removed OTUs (%) | |||||
|---|---|---|---|---|---|---|---|---|
|
Raw data set |
Filtered data set |
|||||||
| Min | Median (±s.d.) | Max | Min | Median (±s.d.) | Max | |||
| 80% | Burgundy (109) | 36 | 68 (±16.6) | 115 | 20 | 39 (±11.5) | 70 | 43 |
| Brittany (131) | 45 | 92 (±17.9) | 152 | 29 | 57 (±13.5) | 125 | 38 | |
| Landes (54) | 43 | 60 (±19.3) | 120 | 27 | 39 (±12.7) | 90 | 34 | |
| South-East (108) | 20 | 86 (±21.0) | 171 | 10 | 48 (±17.2) | 126 | 45 | |
| 85% | Burgundy (109) | 83 | 156 (±37.4) | 241 | 44 | 88 (±25.2) | 153 | 44 |
| Brittany (131) | 105 | 211 (±39.3) | 360 | 52 | 123 (±29.9) | 274 | 42 | |
| Landes (54) | 94 | 144 (±61.8) | 435 | 62 | 90 (±43.8) | 321 | 38 | |
| South-East (108) | 33 | 211 (±42.6) | 377 | 15 | 120 (±33.6) | 263 | 43 | |
| 90% | Burgundy (109) | 183 | 344 (±84.9) | 577 | 91 | 181 (±55.0) | 367 | 47 |
| Brittany (131) | 217 | 488 (±99.9) | 869 | 125 | 260 (±62.7) | 478 | 47 | |
| Landes (54) | 201 | 343 (±188.5) | 1361 | 136 | 223 (±133.8) | 935 | 35 | |
| South-East (108) | 56 | 485 (±95.2) | 835 | 22 | 277 (±71.6) | 483 | 43 | |
| 95% | Burgundy (109) | 428 | 841 (±188.8) | 1376 | 245 | 435 (±110.0) | 828 | 48 |
| Brittany (131) | 473 | 1171 (±227.0) | 1862 | 316 | 628 (±127.2) | 1106 | 46 | |
| Landes (54) | 475 | 875 (±496.1) | 3423 | 356 | 570 (±247.4) | 1622 | 35 | |
| South-East (108) | 93 | 1165 (±209.6) | 1638 | 33 | 669 (±128.9) | 997 | 43 | |
| 97% | Burgundy (109) | 669 | 1273 (±267.2) | 2061 | 415 | 666 (±147.7) | 1131 | 48 |
| Brittany (131) | 758 | 1752 (±306.4) | 2620 | 494 | 904 (±143.7) | 1392 | 48 | |
| Landes (54) | 712 | 1360 (±691.1) | 4687 | 541 | 832 (±249.4) | 1641 | 39 | |
| South-East (108) | 106 | 1725 (±294.4) | 2236 | 44 | 959 (±163.7) | 1295 | 44 | |
Abbreviation: OTU, operational taxonomic unit.
For the raw and the filtered data sets, the median number of OTUs identified per site, its minimum and its maximum were determined per region at each clustering threshold considered in this study. The s.d. of the mean is given in brackets. The percentage of OTUs removed by filtering steps was estimated by dividing the median number of OTUs in the filtered data set by the median number of OTUs in the raw data set.