| Cellular abnormalities induced by Empoasca kraemeri probing upon Phaseolus vulgaris L. were described. In susceptible genotypes, damage to the mesophyll consisted of cell emptying and increased intercellular spaces. In the main veins, Phloem crushing and xylem re-differentiation, cell hypertrophy and division were visible with enlarged nuclei and stained nucleoli. Tolerant genotypes showed less cellular damage and a compensatory response evidenced by more autofluorescent tracheary elements of larger size with higher relative flow rates 48 h after feeding.;Electronic monitoring of leafhopper feeding was compared with a field trial at CIAT's headquarters in Cali, Colombia, to determine whether genotypes could be similarly segragated by both techniques according to their tolerance. Three tolerant and two susceptible genotypes in unprotected and insecticide-protected plots were used. The tolerant genotypes had lower hopperburn scores, higher unprotected yields, higher pod setting and seed filling capacities, faster crop growth rates and larger partitioning coefficients than the susceptible control. Thirty females were electronically monitored on each genotype for 2 h 45 min. each. Three major waveforms were performed. Exhaustive univariate analysis of these individual waveforms did not reveal any diagnostic characteristic for genotype separation. Multivariate analysis supported the hypothesis that E. kraemeri combines waveforms into three probing methods: Pulsing laceration is the most frequently performed and most damaging method, responsible for most of hopperburn symptoms; cell-rupturing ingestion is less damaging, responsible for mesophyll cell emptying and leaf curling, and lancing-sap ingestion occurs in the vascular tissues and is composed of very long ingestion.;Tolerant genotypes induced significantly more cell-rupturing probing, intermediate Lancing sap ingestion and similar pulsing laceration as the susceptible control. Principal component analysis led to the development of a Probing Index that classified genotypes similarly to the Yield Index obtained from field data. 'BAT 41', 'EMP 385' and 'EMP 392' were correctly classified as susceptible and tolerant, respectively. 'Porrillo Sintetico' was misclassified as highly susceptible, and 'EMP 84' as highly tolerant, probably due to the opposite variations they stimulated in pulsing laceration. The importance and applications of the Probing Index, as well as future prospects for its role in a recurrent selection breeding program are discussed. |
