Extraction of genetic test measurements from LiDAR cloud data

Abstract Forest measurements, including genetic trials, have relied on traditional measurement methods, an approach affected by different types of errors. To assess genetic trials, Terrestrial Laser Scanning (TLS) devices offer potential to improve accuracy. This study aimed to implement an approach for analyzing forest genetics trial measurements using TLS data. A 15-year-old Pinus taeda L. progeny test in North Carolina USA was assessed using both TLS data and traditional field measurements. Accuracy was assessed using adjusted R 2 , bias, percent bias, and RMSE. Genetic parameters were estimated via BLUP for diameter at breast height (DBH). The $${R}_{\text{adj}}^{2}$$ R adj 2 values were 0.56 for DBH and 0.29 for total height (HT). Field-measured DBH had higher heritability ( h 2 = 0.32) than raw TLS data ( h 2 = 0.17). However, “cleaned” TLS estimates (DBH R ) improved heritability ( h 2 = 0.27) and showed stronger phenotypic correlation with DBH F ( R = 0.84) than DBH L ( R = 0.75). GCA predictions using BLUP showed high correlation ( R = 0.92) between field and TLS DBH estimates. Estimated gains using DBH F were 11.3% and 12.1% for selecting the top 1st progeny (30 families) and the top 1st and 2nd progenies (15 families), respectively. Estimated gains using DBH F were 11.3% and 12.1% for selecting the top 1st progeny (30 families) and the top 1st and 2nd progenies (15 families), respectively. Corresponding gains from DBH L were 6.9% and 9.6%, and from DBH R , 8.5% and 10.3%. The results demonstrate that TLS, combined with the proposed methodology, is a reliable alternative for genetic analysis in forest trials.