Back

Digital imaging outperforms traditional scoring methods of spittlebug tolerance in Urochloa humidicola hybrids

Hernandez, L. M.; Espitia, P.; Castiblanco, V.; Cardoso, J. A.

2020-10-14 plant biology
10.1101/2020.10.13.338186 bioRxiv
Show abstract

American spittlebug complex (Hemiptera: Cercopidae) is a critical pest for existing Urochloa humidicola cultivars in the neotropical savannas. The U. humidicola breeding program of the International Center for Tropical Agriculture aims to increase tolerance to spittlebugs. To develop U. humidicola genotypes with superior tolerance to spittlebugs than existing cultivars, adequate screening methods ought to be deployed. Currently, visual scores of plant damage by spittlebugs is the standard method to screen for variation in plant tolerance. However, visual scoring is prone to human bias, is of medium throughput and relies of the expertise of well-trained personnel. In this study, we compared estimations of plant damage from two alternative methods (SPAD measurements and digital images) and visual scoring from an inexpert evaluator with the plant damage estimated from an expert. This information should instruct if different methods could be implemented in the U. humidicola breeding program. Time needed to evaluate damage was recorded for each method. Lins correlation coefficient, Pearsons correlation coefficient and broad sense heritability values were also calculated. Overall, damage estimated from digital images showed the highest throughput (twice as fast as visual scoring from an expert); high correlations with visual scoring (r > 0.80, p < 0.0001); and heritability values for plant damage as good or better (> 0.7) than those obtained by visual scoring from an expert. Our results indicate that digital imaging is a phenotyping method that might improve the efficiency of breeding for increased tolerance to spittlebugs in U. humidicola. HighlightDigital imaging outperformed standard scoring method of spittlebug tolerance in Urochloa humidicola, suggesting that this method might improve the efficiency of breeding for such stress.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.