Cherry leaf roll virus in the female gametophyte and seed of birch and its relevance to vertical virus transmission

Abstract

SUMMARYELISA detected cherry leaf roll virus (CLRV) in infected birch leaf sap diluted 1/320 in buffer, in extracts of one infected leaf with nine healthy leaves and in leaf sap frozen for 2 wk. Similarly, ELISA detected CLRV in mixtures of one infected bud with four virus‐free buds. Intensive bioassays and ELISA showed that in some trees CLRV was restricted to only a few branches whereas in others it occurred throughout the tree. The prevalence of CLRV in unmanaged birch populations in Britain was less (3%) than in Midlands street trees (17%).In CLRV‐free birch trees that received pollen from infected ones, ELISA indicated that antigen was introduced into, and multiplied within, the embryos but not the seed coat or the pericarp/wings. In one instance, antigen was detected in a branch of an experimentally pollinated tree but not in those parts of the crown that had been exposed to open pollination. The proportion of seed germinating after crosses in which both parents were CLRV‐free was greater than when either or both parents were infected but the largest difference occurred with infection in the female parent. Few embryos seemed to escape invasion with CLRV when the maternal tissue was naturally infected. Overall, seed transmission ranged between 0 and 38% (mean 17%) when only females in a cross were infected, and between 11 and 75% (mean 30%) when only the males were infected. Assuming no selective advantage that would help infected plants to achieve reproductive age, we found that CLRV would be lost from a birch population within two generations if transmitted only through seed.Embryos in seeds from CLRV‐infected birch that received CLRV‐free pollen differed from their healthy counterparts in being shrunken and suspended in a loosely fibrillar matrix that contained numerous virus‐like particles in tubular inclusions. In two trees, CLRV‐free pollen tended to fertilise a greater percentage of ovules than did CLRV‐infected pollen. Seedlings derived from infected seed and cuttings from naturally‐infected trees grew less rapidly than their healthy counterparts.In still air, most birch pollen liberated from a height of 3·5 m fell within 3 m of the drop zone and none was detected 10 m from the source. Field observations on the patterns of virus spread as measured by seedling infection were consistent; about 3% of seedlings from a tree 6·9 m away from the nearest source of inoculum were infected but no infected seedlings were detected in more distant trees, even though each was experimentally infectible with CLRV and pollen from the infected tree germinated on their stigmas.