Effect of strain-specific hypersensitive resistance on spatial patterns of virus spread

Spatial patterns of spread were compared between strains of Bean yellow mosaic virus (BYMV) that differ in causing systemic necrotic (hypersensitive) or non-necrotic symptoms in narrow-leafed lupin (Lupinus angustifolius). Both types of BYMV were spread naturally by aphids from adjacent infected pasture into a large lupin block ('natural spread site'), or from clover plants introduced as virus sources into two field experiments with lupin. Cumulative spatial data for plants with disease symptoms from a range of times in the growing period were assessed using Spatial Analysis by Distance IndicEs (SADIE). At the 'natural spread site', with non-necrotic BYMV, the extent of clustering of plants with symptoms increased gradually over time, while with necrotic BYMV there was less clustering and no increase over time. In both experiments, for the type of BYMV that was introduced into a plot, there was a gradual increase in clustering, but with this being greater with non-necrotic BYMV. In the second experiment, there was also significant clustering of plants with symptoms of non-necrotic BYMV in plots without introduced non-necrotic foci but not for necrotic BYMV in plots without introduced necrotic foci. When clustering data for plants with newly recorded symptoms was tested for spatial association between successive assessment dates, association was positive for both BYMV types though stronger for the non-necrotic type, declining as the temporal lag increased. Generally, association was strongest for assessments 2-3 wk apart, corresponding approximately to the period for BYMV to move systemically in plants and for obvious symptoms to appear in shoot tips. Contour maps for local association between dates showed that the strongest spatial associations were from coincidence of infection gaps rather than infection patches. The combination of information from clustering and association analysis showed that spread of non-necrotic BYMV is less diffuse, with considerably more localised infection surrounding the infection sources. This work demonstrates how spatial virus spread can be diminished when hypersensitive (necrotic) resistance is deployed, and the limitations associated with employing hypersensitivity that is strain specific.