Abstract
Tomato crops are attacked by several pests, including mites. While the main predatory mites are not effective enough to control mite pests due to problems with plant dispersal and establishment (associated with glandular trichomes - GT - on leaves and stems), recent studies have shown encouraging results with the European endemic phytoseiid Typhlodromus (Anthoseius) recki. The first objective of the present study was to assess the ability of this species to disperse along the tomato stem, considering six genotypes of Solanum lycopersicum, S. peruvianum and S. cheesmaniae with contrasted trichome numbers and types of stem trichomes, accuratetly characterised in a previous study. As morphological variation in body size has been observed within the species T. (A.) recki, the second objective was to determine how predator morphological traits can explain dispersal along the tomato stem. For this, ambulatory dispersal ability of females was tested in lab conditions on the eight Solanum genotypes. Then, the females were mounted on slides and body dimensions measured. No effect of the tomato genotypes was observed on the dispersal ability of the predator. However, specimens that succeeded in crossing the stem, had a higher mobility time than those that failed. Furthermore, body width at midbody and dorsal shield length were negatively correlated with dispersal ability. This suggests that the more slender and relatively small the specimens, the more are mobile and able to successfully cross the stem, regardless of the plant genotype considered. The number of glandular trichomes type (GT) VI and to a lesser extent GT I and IV, and non-glandular trichomes (NGT) II&III appear to limit dispersal. The GT VI seems to have a repellent effect, inducing mite escape in some cases. On the opposite, the number of NGT V were positively correlated with high mobility and stem crossing rates. Assuming that the main barrier to biological control efficiency is dispersal along tomato stems, the results obtained here should have implications for biological control success. The proportion of mites with 'optimal dimensions' appears to be low and further studies should be undertaken to better assess the proportion of mites with such ideal dimensions in different populations and also to determine whether these morphological traits are associated with different feeding abilities and/or abiotic conditions.