Comparison of metatranscriptomic datasets from day and night samples in January, April and August.

**Number of protein clusters found when combining all datasets together: therefore, protein clusters from combined dataset, not a sum of clusters from individual datasets.
***filtering out any sequence with >10% Ns; <75 bp; >60% any single base; absolute duplicates. Coverage  =  (1−(number of singletons/number of resampled sequences)) x 100. Nr – not recorded.