H.—34.

Lucerne-flea. In the North Island the main area where this insect occurs is an eastern region south of Dannevirke to Pirinoa in lower Wairarapa, and a western region from Palmerston North to Bulls, with two isolated areas —one at Maraekakaho in Hawke's Bay and the other at Pokeno, just south of Auckland. It) the South Island the insect has not been found outside the region south of the Rangitata River from Temuka to Gore. Periodic inspections have been made of the infestation on subterranean clover at Maraekakaho and during the autumn, winter, and spring of 1938 a heavy infestation, which at its peak numbered ] ,300 fleas per square foot, developed on this area. The loss of leaf-weight attributable to the flea was 38 per cent., and there was evidence that both cattle and sheep tended to discrimination against the affected pasture. The heavy infestation which developed here was undoubtedly due to the unusual climatic conditions prevailing during the summer of 1937-38 in that the normal period of summer dryness was much shortened and temperatures in summer and autumn were much above the average. The result was that adequate soil moisture and favourable temperatures extended over a longer period than usual. Owing to the extended period of drought during the past summer 1938-39, conditions have not been conducive to a heavy infestation, and it is thus anticipated that the population level on this area during the coming winter and spring will be much lower than last year. The lucerne-flea, though it is widely distributed and has apparently been present for some years on areas of mixed pastures in the North Island, has given no evidence of its being a serious pest on this type of pasture under normal conditions with regard to grazing. Similarly, there is no evidence of a serious threat to the lucerne. The potential seriousness of the pest lies in its relation to subterranean clover in the low rainfall areas, particularly in Hawke's Bay, in which this plant finds its greatest use. The evidence so far accumulated points to the insect being a serious pest on subterranean clover in this area in years of exceptional climatic conditions. In normal years it seems likely that the depredations of the pest will not be sufficiently great to warrant chemical control measures. Apart from chemical control by the use of lime-sulphur sprays, the bdellid mite (Biscirus lapidarius) is considered in Australia to be a valuable agent in the control of the flea. This mite has been found to be naturally distributed in several widely separated localities in the North Island. Consideration was given by the Department of Agriculture and the Entomology Division to the practicability of eradication of the flea on the isolated area at Maraekakaho, but no action was taken. A detailed report on the insect was published in the Journal of Science and Technology. Insect Pests of Wheat. With the object of determining whether there was any marked diflerence in varietal resistance, a series of wheat-variety trials was carried out in regard to the incidence of Hessian fly (Mayetiola destructor) and the Argentine stem-weevil (Hpyerodes griseus), Hunters 1, Cross 7, Dreadnought, Bencubbin, Variety 31-05, and Solid-straw Tuscan being the wheat varieties used. The results reveal that Solid-straw Tuscan, Variety 31-05, and Bencubbin, this year show a degree of low Hessian-fly infestation (3-8 per cent., 3-9 per cent., and 7-4 per cent.) in marked contrast with the other varieties (28-9 per cent, to 60-7 per cent), while in the case of the stem-weevil the infestation shows no such marked contrasts. Further, the data based on damage due to stems broken down by the weevil larvse shows no great extremes, but the study is being extended this coming season. Cocksfoot-seed Midge. A study of the life-history of this midge (Stenodiplosis sp.) reveals that hibernation occurs as fullgrown larvae in seed-heads ; these larvae pupate in October, and the first generation of midges emerges from about the end of October to mid-November, at which period eggs and young larvae occur. This generation of larvae pupates in December and gives rise to midges during the same month. Several generations follow and overlap so much that they cannot be distinguished under field conditions, though there is a peak emergence of midges in January ; the midges are to be found to the end of March. Experiments to ascertain whether meadow-foxtail midge would infest cocksfoot and the cocksfoot-midge-meadow foxtail proved inconclusive. In an attempt to delay flowering as a means of control it was found that after cutting the crop at different periods flowering occurred at the normal time and the flowering was reduced ; in the latest cutting which was done during November, however, the grass failed to head normally. Some spraying experiments were carried out on a small scale, but with negative results. It will be necessary to extend this research. Sheep-dipping Trials. This recently inaugurated research is an activity of the Canterbury Agricultural College, where a member of the Division of Entomology is co-operating in regard to the biology of the insect parasites and influence upon them of the various fluids experimented with. Insect Life of Tussock Grasslands. In relation to the deterioration of tussock grasslands, an ecological study of the insect factor was commenced in February. Certain species of moth larvae and a species of beetle larva have been found attacking the foliage and roots respectively. The species of moths, at least, are widely distributed. However, no conclusions can be reached at this early stage of the investigation, though there are indications that insects are not a primary factor in deterioration.

5—H. 34.

33