Abstract
The absence of R. crispus on a shingle bank may be a result of grazing by a chrysomelid beetle, G. viridula. Sixty R. crispus plants are grown from seed in the glasshouse and planted in a regular grid on the shingle, interspersed with R. obtusifolius, an alternative food plant for G. viridula. Thirty R. crispus were kept free of G. viridula grazing by removal of the beetle by hand. Dispersal of G. viridula from the R. obtusifolius was permitted on the other 30 R. crispus plants. G. viridula may have up to 3 generations in a year on the shingle. The experimental plants of R. crispus were harvested at the end of each G. viridula generation and shoot and root dry weights were measured. In 1978 there was a significant reduction in shoot and root dry weights of R. crispus when grazed by G. viridula compared with ungrazed controls. In 1980, shoot and root dry weights were significantly reduced at the end of the 1st generation of the beetle but reductions in the weight of grazed plants after the 2nd generation were not significant. None of the R. crispus plants was killed by grazing. During the winter and spring following grazing by G. viridula, the experimental R. crispus plants were flooded. Survival of previously grazed plants was .apprx. 1/3 that of plants previously protected from grazing. The reduction in rooting volume due to grazing appeared to be the main cause. The virtual absence of R. crispus on the shingle bank is attributed to the combined effects of grazing by G. viridula and flooding. Neither factor would by itself eliminate R. crispus. The presence of the beetle in sufficient numbers depends on the availability of R. obtusifolius as an alternative food.