“…We screened out 16 variation hotspots with higher Pi values (Pi >1.5%), including 11 in LSC region ( rpl22‐rps19 , ndhC‐trnV‐UAC , ycf3‐trnS‐GGA , trnH‐GUG‐psbA , rps14‐psaB , trnR‐UCU‐atpA , psbC‐trnS‐UGA , trnG‐UCC‐trnR‐UCU , 5'‐trnK‐UUU‐rps16 , rps2‐rpoC2 , rps3‐rpl22 ), four in SSC region ( ψycf1‐ndhF , ccsA‐ndhD , trnL‐UAG‐ccsA , rps15‐ycf1 ) and one in IRb/SSC boundary ( rps19‐rpl2 ). Among them, eight sequences ( ndhC‐trnV‐UAC , ccsA‐ndhD , ycf3‐trnS‐GGA , trnH‐GUG‐psbA , trnR‐UCU‐atpA , psbC‐trnS‐UGA , rps15‐ycf1 , 5'‐trnK‐UUU‐rps16 ) with lengths more than 200 bp were chosen as excellent candidate markers following Li et al (2021) and were suitable for developing specific barcodes of Sanguisorba . In agreement with Park et al (2021), trnH‐GUG‐psbA , trnR‐UCU‐atpA and ccsA‐ndhD were selected as areas with high divergence; however, in disagreement with Park et al (2021), petB‐petD was not selected as candidate markers of Sanguisorba in our study considering that the Pi value of petB‐petD was only 0.2116%.…”