Bacterial Community and Diversity
After cleaning and filtering of sequence data, a total of 12,173 bacterial amplicon sequence variants (ASVs) were detected across the 96 samples (Table S1). Sample reads were rarefied to the lowest read count of 23,073 reads (from #30 sample – KSN 00:00 week 6; Table S1). The effective number of ASVs was calculated to compare alpha diversity between sites, land cover, and times. Alpha diversity at Mark Oliphant CP was higher than Kenneth Stirling CP (mean ± SD, Mark Oliphant CP: 554 ± 132 effective number of ASVs; Kenneth Stirling CP: 479 ± 149 effective number of ASVs; chi-squared = 6.28, p = 0.012; Figure 2a). Alpha diversity was higher at the cleared land cover type at Kenneth Stirling CP (cleared: 519 ± 123 effective number of ASVs; native: 439 ± 164 effective number of ASVs; chi-squared = 4.42, p = 0.035) and at Mark Oliphant CP (cleared: 603 ± 104 effective number of ASVs; native: 504 ± 139 effective number of ASVs; chi-squared = 8.82, p = 0.003; Figure 2a).
Time did not have an effect on alpha diversity in Kenneth Stirling CP at either the cleared (mean ± SD; 00:00: 505 ± 109 effective number of ASVs; 06:00: 513 ± 132 effective number of ASVs; 12:00: 529 ± 161 effective number of ASVs; 18:00: 529 ± 119 effective number of ASVs; chi-squared = 0.313, p = 0.957; Figure 2b) or native land cover plot (00:00: 520 ± 140 effective number of ASVs; 06:00: 339 ± 180 effective number of ASVs; 12:00: 462 ± 203 effective number of ASVs; 18:00: 436 ± 99.4 effective number of ASVs; chi-squared = 4.547,p = 0.208; Figure 2b). Time also had no effect on alpha diversity at Mark Oliphant CP at the cleared (00:00: 602 ± 134 effective number of ASVs; 06:00: 589 ± 88.3 effective number of ASVs; 12:00: 591 ± 82 effective number of ASVs; 18:00: 631 ± 127 effective number of ASVs; chi-squared = 0.393, p = 0.942; Figure 2c) or native land cover type (00:00: 591 ± 188 effective number of ASVs; 06:00: 518 ± 103 effective number of ASVs; 12:00: 450 ± 105 effective number of ASVs; 18:00: 457 ± 129 effective number of ASVs; chi-squared = 2.61, p= 0.456; Figure 2c).
The principal coordinate analysis and PERMANOVA based on Bray-Curtis distance showed that land cover (p = 0.001) had a stronger and significant effect on beta diversity than site (p = 0.001) and time (p = 1.0) (Figure 2d). The site effect was stronger in the cleared land cover types, with Kenneth Stirling CP and Mark Oliphant CP samples separating more distinctly than in the native land cover (Figure 2d). There was also a greater bacterial community heterogeneity measured by distance to the centroid in the native land cover plot compared to the cleared land cover plot in Kenneth Stirling CP (mean ± SD; cleared: 0.477 ± 0.062 distance to centroid; native: 0.534 ± 0.0456 distance to centroid; P-adj = 0.001, 95% CI [-0.096, -0.018]; Figure 3). The same trend was seen in land cover at Mark Oliphant CP though this was not significant (cleared: 0.483 ± 0.046; native: 0.508 ± 0.040; P-adj = 0.351, 95% CI [-0.063, - 0.014]; Figure 3).