The use of weather data and counts of aphids in field to predict incidence of yellowing viruses of sugar-beet crops in England in relation to use of insecticides

Partial regression equations were calculated that relate the mean percentage of plants infected with yellowing viruses (beet yellows and beet mild yellowing viruses) in sugar‐beet crops at the end of August to the number of days during January, February and March when temperatures fell below – 0.3 °C (31‐5 °F) and the mean temperatures in April, for the 21 yr, 1951–71, using weather records from Rothamsted Experimental Station. Regression analyses were also made to find the effect of other factors including mean and minimum temperatures for the same months, and also mean counts of ‘green aphids’, mainly of the vector Myzus persicae, on sugar‐beet plants during May and June. Significant relationships were established with all factors, but ‘frost‐days’ and April mean temperatures accounted for the greatest percentage of the variance in yellows incidence. The calculations were made separately for the years from 1951 to 1958, when no routine advice was given to farmers about aphid control, and 1959–71 when a ‘spray‐warning scheme’ was in operation, and many crops were sprayed at critical times to prevent aphid‐ and virus‐spread. Weather factors had the same effects in both periods, but for any particular weather less virus was spread in the second period than in the first, although there were sufficient aphids, i.e. the numbers expected from the prevailing weather conditions. There was no evidence that insecticide treatment used in any one year affected aphid‐incidence in subsequent years. Regression analyses on weather variables were also calculated separately for each of seventeen beet‐sugar factory collection‐areas, using weather records from local weather stations, and also the Rothamsted weather records. Unexpectedly, the fit of the regressions was always better with Rothamsted weather data than with local weather records. Mean yellows‐incidence for the different factory areas declined from south to north, and there was a linear relationship with the square root of the latitude above 50 °C. At the same time the correlation coefficients relating yellows‐incidence to ‘frost‐days’ became smaller and less significant, and those showing dependenc