Abstract
Grass grub (Costelytra zealandica) larval growth under controlled environment conditions was significantly lower on Trifolium arvense, T. affine, T. semipilosum, and T. uniflorum than on white clover (T. repens). There was no difference in growth of larvae feeding on T. repens and T. ambiguum. Weight gains on two seedlines of T. uniflorum were not significantly different from those on Lotus pedunculatus, a known resistant plant, but larval growth was significantly lower on a third seedline. Pooled data for T. repens × T. uniflorum hybrids suggested that although larval growth on hybrids backcrossed twice with T. repens (H5) was not significantly lower than on T. repens, hybrids backcrossed with T. uniflorum (H2) and T. repens (H3) supported significantly lower larval growth. However, this was not confirmed for two H3 genotypes (U12 and U18) in a repeated experiment with a higher replication. In a field experiment the number of larvae surviving under H5 hybrids was not significantly different from that under L. pedunculatus but was significantly fewer than the number surviving under T. repens. However, the root weight of both H5 hybrids and T. repens grown in plots with 400 larvae/m2 was lower than in plots without larvae whereas L. pedunculatus was not affected, and the final larval weights under both T. repens and H5 hybrids were higher than under L. pedunculatus.