Bemisia tabaci B-Arabidopsis Interactions Examined by Electrical Penetration Graphs

In the absence of strong resistance mechanisms to control the world-wide pest Bemisia tabaci B, new methods for control must be derived from understanding the plant innate immune response to whiteflies. Using four Arabidopsis defense-signaling mutants, transcriptome and hormone treatment studies, it was shown that B. tabaci B activates SA-regulated defenses, suppresses JA-regulated defenses, and the JA-defenses are effective in slowing whitefly nymphal development. Here we used the Electrical Penetration Graph (EPG) technique to determine the tissue location of resistance traits that influence B. tabaci B adult and nymph feeding in wild-type plants and in four defense mutants (cim10, cev1, coi1, and npr1). EPG studies measured 90 variables reflecting adult whitefly feeding behaviors. However, few variables differed significantly among the five genotypes and between pooled data of fast-development genotypes (cim10 and coi1) and slow nymph development genotypes (npr1 and cev1). These data indicated that the suite of variables that are most strongly associated with host plant acceptance/suitability were not consistently associated with how each genotype affected nymphal development. For most of the feeding variables where the ANOVA detected significant differences among the five genotypes, the significant differences were due to cim10 differing significantly from one or more of the other genotypes. While evaluating adult whitefly feeding behaviors, EPG classical and three new types of E2 waveforms were identified. All four E2 waveforms were associated with the proxy for ingestion (honeydew production). Interestingly, the second type of E2 waveform, which resembled the E1 (salivation) waveform, was also associated with honeydew production. Finally, whiteflies ingesting phloem sap occasionally entered pathway phase for a brief period, and re-inserted stylets into the same sieve element to reinitiate E2, without an E1 waveform indicating that E1 was not always a prerequisite for E2. Whitefly nymphs generated two phloem-phase waveforms (H and L) in Arabidopsis. Only waveform H was correlated with honeydew production. When nymph feeding behavior was compared among the four defense mutants or between pooled data of fast- and slow-nymphal development genotypes, no differences were found. Collectively these data indicated that phloem sap quality was the reason for differences in nymphal developmental rate