The DnaSP v6 program allows to estimate statistical significance for Tajima's D by two ways:

1. assuming that it follows the beta distribution

2. by coalescent simulations

When I do the Tajima's test in standard method, I get “Not significant, 0.10 > P > 0.05” result.

But when I do coalescent simulations provided Standard Neutral Model, this Tajima's D value appears to be significant (p=0,011*).

How could it be and how should I interpret this?

I might have used incorrect parameters for coalescent simulations, but all the parameters were estimated for my sample in the same program.

Please, help me to understand.

Thank you in advance!