Look at the actual sequences and see how well they match with the query sequence; BLAST should provide that information.

Dear friend

First look at sequence identity, e-value and query coverage. Then you can perform phylogeny, the sequence very close to your sequence will be the best hit.

minimum e value and maximum score

Based on the result, all of the sequences seem homologous. It means all of the sequences are present in all of the hits (match). I think you need to take a longer sequence to figure out the best match.