Observations on the spatial distribution and on the measurement of pre-imaginal populations of the European fir budworm, Choristoneura murinana Hb. (Lep., Tortricidae). 2. Larval populations. Conclusion

Samples consisted of 1 m- or 1.50 m-branch tips. Mean distribution of the overwintering larvae along these tips has remained constant, or nearly so, between plots and between years (for 1.50 m-branch tips: about 20 per cent of the larvae on the apical third, 40 per cent on each of the median and basal thirds), this suggests that the mean distribution of the larvae along branches would only be a function of the total length of these latters. Within-tree variation of the larvae number was higher than among-tree variation, sometimes nearly equal to this one, both have not much varied between plots and b1750etween years. In comparison with their respective values for the eggs, the latter had strongly decreased while the former was sometimes higher, sometimes much lower. Intracyclic correlations between population levels per tree show (1) the overwintering larvae number was sometimes correlated significantly, but weakly, with the eggs number, sometimes independent from this latter, (2) the defoliation percentage (which is an estimator of the later-stages larvae number) was independent, or nearly so, from the overwintering larvae number and from the eggs number. Correlations between successive generations per tree were rather strong for the eggs, weak or very weak for the overwintering larvae, non-existent for the defoliation. These facts indicate that the pre-imaginal populations are subject, in summer and then in spring, to large-extent spatial redistibutions, within and between the trees. These redistributions are essentially caused by the passive dispersal, first of the L1, then of the post-hibernating L2. They prove to be strongly variable according to the stand and to the year. The aleatory nature of the factors (particularly the wind) influencing the passive dispersal makes apparently unpredictable the overwintering larvae distribution, so that it seems impossible to accurately sample these la