STRUCTURAL FEATURES OF FIELD BOUNDARIES WHICH INFLUENCE THE OVERWINTERING DENSITIES OF BENEFICIAL ARTHROPOD PREDATORS

1. Vegetation cover was manipulated within enclosures on a field bound ary in southern England to test experimentally the effect on the overw intering of Tachyporus hypnorum and Demetrias atricapillus populations , species that use grassy boundaries of arable fields as refugia durin g winter. 2. Winter survival was lowest for beetles enclosed on bare e arth and highest for those enclosed on tussocks of Dactylis glomerata. The contrast in structural complexity of the experimental treatments caused a 44%, 43% and 36% variation in the final densities of beetles during three successive winters. 3. The densities of T. hypnorum on ad jacent field boundaries were estimated from the composition of their v egetation cover and the survival rates of the beetles on different tre atments. There was no significant difference between the number of T. hypnorum allocated to seven field boundaries by the model and the numb er of beetles sampled from soil and vegetation of those boundaries at the end of winter. 4. The cover of boundaries by non-tussock grass spe cies accounted for 91% of the predicted overwintering numbers of T hyp norum because tussock grasses and bare soil were not common on existin g boundaries. We therefore considered the influence of boundary struct ure on the overwintering of the beetles. 5. Higher winter densities of T. hypnorum were sampled from boundaries with deeper soil and greater vegetation height, that were wider and higher above the field level, with an east to west orientation, warmer mean daytime temperature and lower soil moisture. 6. A quadrat survey was carried out in Norway on the equivalent group to T. hypnorum that comprised T. hypnorum, T. chr ysomelinus and T. obtusus. With multiple regression, more individuals of Tachyporus spp. and other beneficial arthropods occurred in samplin g units from boundaries raised higher above the field level that compr ised grass cover with a high proportion of tussock grass. 7. Other fac tors such as pre-winter crop husbandry, food supply and parasitism may affect the dispersal power, habitat selection and cold hardiness of b eneficial arthropod species within available boundaries and account fo r the observed variation in beetle numbers.