A new method for the analysis of access period experiments, illustrated with whitefly-borne cassava mosaic begomovirus

Reports of low transmission efficiency, of cassava mosaic geminivirus (CMB) in Bemisia tabaci whitefly, diminished the perceived importance of whitefly in CMB epidemics. Studies indicating synergies between B. tabaci and CMB prompt a reconsideration of this assessment. In this paper, we analysed the retention period and infectiousness of CMB-carrying B. tabaci as well as B. tabaci susceptibility to CMB. We assessed the role of low laboratory insect survival in historic reports of a 9d virus retention period. To do this, we introduced Bayesian analyses to an important class of experiment in plant pathology. We were unable to reject a null hypothesis of life-long CMB retention when we accounted for low insect survival. Our analysis confirmed low insect survival, with insects surviving on average for around three days of transfers from the original infected plant to subsequent test plants. Use of the new analysis to account for insect death may lead to re-calibration of retention periods for other important insect-borne plant pathogens. In addition, we showed that B. tabaci susceptibility to CMB is substantially higher than previously thought. We also introduced a technique for high resolution analysis of retention period, showing that B. tabaci infectiousness with CMB was increasing over the first five days of infection. Author summaryOne of the major crops for food security is cassava. Diseases caused by whitefly-borne viruses, such as cassava mosaic begomoviruses (CMBs; causative agent of cassava mosaic disease), are among the major constraints of cassava production in sub-Saharan Africa. Reports of relatively low transmission efficiency of CMB in Bemisia tabaci whitefly helped to diminish the perceived importance of whitefly in CMB epidemics. But what role has the rate of insect laboratory survival played in these reports? Here we introduced Bayesian methods of statistical analysis to account for insect laboratory survival in the analysis of an important class of experiment in plant pathology. In essence, we show, for CMB, that published laboratory data is consistent with high insect transmission efficiency and low insect laboratory survival. We also show for CMB that there is no evidence that the retention period-the length of time the pathogen-bearing insect vector is infectious-is less than the insect's lifespan. Our methods can be used to analyse published laboratory data for a range of additional insect-borne plant pathogens, and, in particular, to analyse viral retention period-a central parameter in the classification of viruses according to their transmission type.