White clover (Trifolium repens) is naturally polymorphic for cyanogenesis (hydrogen cyanide release following tissue damage). The ecological factors favouring cyanogenic and acyanogenic plants have been examined in numerous studies over the last half century, making this one of the best-documented examples of an adaptive polymorphism in plants. White clover cyanogenesis is controlled by two, independently segregating Mendelian genes: Ac/ac controls the presence/absence of cyanogenic glucosides; and Li/li controls the presence/absence of their hydrolysing enzyme, linamarase. In this study, we examine the molecular evolution and population genetics of Li as it relates to the cyanogenesis polymorphism. We report here that Li exists as a single-copy gene in plants possessing linamarase activity, and that the absence of enzyme activity in li/li plants is correlated with the absence of much or all of the gene from the white clover genome. Consistent with this finding, we confirm by reverse transcription-polymerase chain reaction that Li gene expression is absent in plants lacking enzyme activity. In a molecular population genetic analysis of Li and three unlinked genes using a worldwide sample of clover plants, we find an absence of nucleotide variation and statistically significant deviations from neutrality at Li; these findings are consistent with recent positive directional selection at this cyanogenesis locus.

The discovery in 1981 that ryegrass endophyte is responsible for ryegrass staggers has required a total revision of our understanding of ryegrassbased pastures. The effects on grazing livestock range from obvious clinical disorders to chronic impairment of productivity. On the other hand, endophyte infection is essential to ryegrass persistence in most of New Zealand, protecting the plant from some invertebrate pests and from over-grazing, and perhaps promoting its tolerance of moisture stress. Vigorous, endophyte-infected pastures sometimes include less clover than pastures relatively free of endophyte. Active compounds produced by the endophyte have been identified, and conditions causing high levels of them have been documented in part. Endophyte strains which do not produce the compounds responsible for livestock toxicoses offer exciting new possibilities in grassland agriculture. Keywords: invertebrate pests, mycotoxicoses, Neotyphodium lolii, perennial ryegrass, ryegrass staggers

Ryegrass seed lines with (+E) and without (-E) the fungal endophyte Acremonium lolii were sown with white clover. Fewer clover seedlings survived under +E ryegrasses irrespective of whether the plots were mown or grazed. The dominant contrast was an inverse competitive relationship between grass yield and clover seedling survival. A further component unexplained by effects of defoliation or competition is interpreted as an allelopathic effect. Two pot trials, using wheat, were established to determine it there was any residual effect of +E wards on following crops. Wheat grown in soil with a history of +E or-E ryegrass with clover produced higher yields from the-E than the +E soil. However, wheat grown on soil from previously clover-free pastures showed no difference between +E and -E. Residual effects on wheat crops are therefore interpreted as being related to lower nitrogen fertility, as a result of reduced clover growth in +E pastures. Keywords: Ryegrass, endophyte, white clover. competition, vigour. allelopathy, soil nitrogen, wheat yield, Acremonium lolii

Two glasshouse experiments were conducted to evaluate possible allelopathic effects of perennial ryegrass/endophyte (Lolium perennel Neotyphodium lolii) associations on white clover (Trifolium repens). The influence of betweenspecies competition, environmental stress, and soil fauna was eliminated by the use of a sterile sand nutrient culture technique to sustain potted clover seedlings, to which aqueous extracts from moisturestressed perennial ryegrass pseudostem were applied. In Experiment 1, extracts from two ryegrass cultivars ('Grasslands Pacific' and 'Grasslands Nui'), each hosting in separate seed lines two endophyte strains (ES, WT), suppressed clover growth (mean suppression 22% at 100% concentration) relative to extracts from endophytefree lines of the same cultivars. The degree of suppression increased with increasing extract concentration, differed between cultivars, but did not differ between endophyte strains. In Experiment 2, extracts at 100% concentration from a third ryegrass cultivar ('Grasslands Ruanui'), hosting in separate seed lines three endophyte strains (AR4, ES, WT), suppressed clover growth by a mean of 27% relative to extracts from an endophyte-free line *Author for correspondence A98022 Received 2 June 1998; accepted I December 1998 of this cultivar (range 11 to 47%), with significant differences between strains. The endophyte strains differed in three known endophyte alkaloids, viz peramine, ergovaline, lolitrem B, but in this study none of these alkaloids could solely account for the allelopathic response. The suppression of clover by all ryegrass cultivar/endophyte strain associations tested indicates that clover suppression is not confined to the specific associations used in previous studies, and that further testing of new associations is warranted as variation in the degree of allelopathy existed, being dependant on endophyte strain and host ryegrass cultivar.

These results indicated that management approaches are needed to overcome potential vulnerabilities to heat stress for sheep grazing AR6 perennial ryegrass. Furthermore, following grazing such pasture, lambs will need to graze endophyte-free ryegrass for >18 days, to completely clear ergot alkaloids from their vasculature, assuming that complete clearance can be achieved.