Why are there error bars in some of the measures to calculate alpha-diversity?

These bars basically represent the Standard Error (SE) in calculating the alpha-diversity. Methods that calculate a SE include ACE and Chao1, which are used as similarity estimators, allowing for rigorous comparison of two or more similarity index values. SEs are computed by a bootstrap procedure, which requires resampling the observed data and recomputing the estimators many times.

An example output from MicrobiomeAnalysis Alpha diversion profiling is shown below: