Sequential sampling to estimate Malacosoma neustrium (L.) population density

The spatial distribution of Malacosoma neustrium (L.) egg masses was studied in order to develop a sequential sampling plan for estimating population density. The observations were conducted in Calabria (in four localities during the years 1998-2000) and in Sardinia (in nine localities during the years 2000-2001) where infestations of the insect occurred in Quercus ilex L. and Q. suber L. forests, respectively. The cut-twig method was used to obtain egg mass counts: from the outer portions of the canopy of each tree 5 branch tips, each about 30 centimetres long, were sheared (unit of sampling). Taylor's power law and mean crowding regression have been used to evaluate dispersion of egg masses within and among trees. Patterns of aggregation, as measured by the two mean-variance methods, did not differ between Sardinia and Calabria. A two-stage sampling plan for estimating the mean density at two precision levels (D = 0.10; D = 0.25) is here discussed and the sequential stop lines for egg masses were calculated by using the parameters of Iwao's patchiness regression. For instance, at D = 0.25 the sequential plan would require to examine 85 branch tips, collected from 17 trees, at a mean density near 0.5 egg masses per branch tip. At the same density but at the higher precision level, it would be necessary to collect a minimum of 540 branches from 108 trees.