I’m working on a soil microbiome dataset using QIIME2 (ASVs via DADA2) and planning to evaluate common alpha diversity indices such as Shannon, Simpson, Observed Features, Pielou’s Evenness, Chao1, and Faith’s PD.

However, there’s a concern that our dataset may be highly saturated (very high richness with many rare taxa), where standard indices might plateau and fail to differentiate samples.

Given this:

How common is this issue in soil microbiome data? Are there better approaches for handling hyperdiverse datasets (e.g., Hill numbers, phylogenetic diversity, or composite indices)? Does it still make sense to first evaluate standard indices (e.g., vs organic matter or management groups) before moving to alternatives?

Any guidance or references would be really helpful.