No abstract

Electrical penetration graph (EPG) monitoring is the most rigorous means of observation and quantification of feeding by piercing-sucking arthropods. Previous EPG studies with aphids and leafhoppers have demonstrated that the X wave identifies when the stylets of these phloem fluid-ingesting insects make contact with their preferred plant vascular cell, phloem sieve elements. This article presents the first direct evidence of an X wave identifying ingestion from a xylem tracheary element by a xylem fluid-ingesting type of leafhopper Homalodisca liturata Ball (Hemiptera: Cicadellidae: Cicadellinae), whose waveforms are nearly identical to those of the glassy-winged sharpshooter, Homalodisca vitripennis (Germar). We document consistent association of the sharpshooter X wave with salivary sheath termini in xylem, especially lignified secondary xylem cells, and absence of the X wave in the rare instance of ingestion from a nonxylem cell. The sharpshooter X wave is a complex, multicomponent waveform, composed of X wave-specific variants of waveform subtypes B1w (representing salivation), B1s (representing precibarial valve movement for tasting), types C1 (a new waveform type that may represent egestion) and C2 (a new designation for the waveform type representing ingestion/cibarial pumping). It is proposed that the sharpshooter X wave represents a blended suite of behaviors that function to 1) physically seal stylet tips into the cell via sheath salivation, 2) repeatedly taste then eject (egest) chemical constituents of the cell to determine acceptability, and 3) mechanically test the strength of the stylet seal via trial cibarial pumping (ingestion). It is further hypothesized that the X wave represents vector behaviors that control inoculation of the Pierce's disease bacterium, Xylella fastidiosa. The ingestion-(salivation and egestion) hypothesis is stated for the mechanism of transmission of X. fastidiosa.

A formant analyser is interpreted probabilistically via a noisy channel model. This leads to a robust method of incorporating formant features into hidden Markov models for automatic speech recognition. Recognition equations follow trivially, and Baum-Welch style re-estimation equations are derived. Experimental results are presented which provide empirical proof of convergence, and demonstrate the effectiveness of the technique in achieving recognition performance advantages by including formant features rather than only using cepstrum features.

No abstract

No abstract