Previous work has shown that, under natural conditions in the Lake District, larvae of Phyllopertha horticola (L.) hatch in early July and feed actively on the roots of pasture plants during the next 3½–4 months, undergoing two moults. They then empty the gut and enter hibernation, pupating the following spring. Stores of organic material in grass roots are at their highest level during this autumn feeding period. It has also been shown that the egg production depends almost entirely on the weight of the hibernating larva or of the pupa, which thus plays an important part in determining the reproductive rate of the population, and studies were accordingly made on larval growth and certain factors affecting it.

Newly hatched larvae were cultured at 15°C. in moistened plaster-of-paris containers filled with a mixture of soil and germinating grass seeds. They moulted at about 20 and 45 days after hatching, and stopped feeding and entered hibernation at about 100 days. When the seeds were scattered on the soil surface, larval growth was slower. The mean larval weight, plotted against age, gave a sigmoid curve; in the. first instar and most of the second, the rate of increase in weight was proportional to the weight, but thereafter, up to the time of hibernation, it was more or less constant. The rate of growth of the individual larva was irregular, being slower at the moults and variable even in the middle of the instar.

Larvae cultured under semi-natural conditions in pots of growing grass in the open moulted about 3–4 and 7–9 weeks after hatching and entered hibernation at 100–120 days. Growth is possible on a wide variety of food plants, larvae cultured on 13 species of pasture plants grown in pure stands surviving to the hibernation stage on all but two of them.

During hibernation, the larva loses 20–25 per cent, of its weight, mostly in the first few weeks. The pupal weight is almost constant and does not appear to be affected by the temperature treatment of the hibernating larva. It is thus a useful index of effective larval growth.

The mean and (in brackets) range of the weights of all pupae collected in two fields in the Lake District between 1950 and 1953 were 139·3 mg. (65–242) for males and 171·3 mg. (72–310) for females. Field samples of hibernating larvae and of pupae show considerable variation in weight from place to place, from year to year and within apparently homogeneous areas.

Variation in the time at which larval growth takes place is a major cause of variation in pupal weight. The growth period of larvae in a field at Buttermere was three weeks earlier in 1952 than in 1950, though of the same duration, and the resulting pupae in 1953 were heavier than those in 1951. Two lots of larvae of similar parentage, grown in adjacent plots of grass out of doors, one of which both hatched and entered hibernation three weeks before the other, likewise showed a difference in weight at hibernation, the earlier lot being the heavier. A series of weighed samples of larvae taken from part of a field at Ambleside in 1953 at weekly intervals during the period when they were entering hibernation showed that heavier individuals did so before lighter ones, and males before females. Factors inducing mortality during this period thus operate selectively against females, because these are exposed to them for longer.

Field-collected larvae fed in the third instar on roots of lettuce produced pupae the following spring that were significantly heavier than those from larvae fed on roots of either ryegrass or clover.

There is no evidence to show that population density affects the weight of the hibernating larva or the pupa. On the other hand, larvae from soil from which the damaged turf had been stripped by birds were significantly lighter than those from the surrounding undisturbed sward.

When moving through the soil, larvae may meet and fatally injure each other by an undirected “ snapping ” reaction. This mechanism may limit population density. In an experiment in which larvae were reared in loose soil on grass roots, the mortality rate was seen to increase with the size and activity of the larvae, and also with the larval density.