Intercropping organic broccoli with Rhododendron tomentosum and Fagopyrum esculentum: a test of bottom-up and top-down strategies for reducing herbivory

Abstract

Brassicaceous plants are attacked by a wealth of specialist herbivores that include the Diamondback moth (DBM) Plutella xylostella L. (Lepidoptera: Plutellidae), control of which requires novel biocontrol strategies. DBM is a cosmopolitan pest causing damage that varies yearly in Finland depending on the timing and extent of their migration. Intercropping with companion plants can hamper host location by herbivores or attract their natural enemies. We tested two sustainable companion plant-based protection strategies on field-grown broccoli (Brassica oleracea var. italica), which comprised 1) aromatic and repellent-releasing Rhododendron tomentosum (RT) (bottom-up strategy), and 2) nectar-producing buckwheat Fagopyrum esculentum (FE) (top-down strategy) combined with an early-season floating row cover (mechanical control) (MC + FE). In addition, a control (no companion plant or cover) and mechanical control (MC) without FE were included. DBM adults on yellow sticky traps and larvae on plants were counted, and feeding damage on leaves was quantified. Volatile organic compounds emitted by broccoli plants in control and RT plots, and emitted by boundary RT plants in RT plots, were identified and quantified. There was a mass outbreak of DBM in early summer with a population peak in mid-July when the second adult generation emerged. DBM adult densities were significantly lower in RT and MC + FE than in control plots. Broccoli in RT plots had a lower larval density and lower damage intensity than in control plots in the early-season. Larval densities and damage intensities were the highest in the latter half of July. Control plots had the highest number of larvae followed by RT, MC, and MC + FE plots. Damage intensities in control plots were significantly higher than in all other treatments throughout the season. Damage intensity was lower in MC and MC + FE plots than in RT plots at the end-of-July. R. tomentosum emitted two characteristic sesquiterpene alcohols, palustrol and ledol, but no evidence of adsorption and re-release of these compounds from Brassica plants grown in RT plots was found. We conclude that incorporating RT as a boundary plant and using a mechanical row cover reduces DBM damage on broccoli, but further multi-year trials under varying degrees of pest pressure are needed.