Reduced sugarcane acreages in Hawaii provide an opportunity to produce locally-demanded broccoli using heat-tolerant varieties. Recent medical findings indicate that broccoli consumption may inhibit the development of human colorectal cancer. Only 5% of locally consumed broccoli is produced within the state. Major broccoli pests include the diamondback moth (DBM), Plutella xylostella (L.), cabbage looper (CL), Trichoplusia ni (Hübner) imported cabbageworm (ICW), Artogeia rapae (L.), and the cabbage webworm (CW), Hellula undalis Fabricius. Most Hawaii farmers apply agro-chemicals to prevent crop losses that result from larval feeding on broccoli leaves and heads. However, the non-target effects of pesticides (e.g., development of pesticide resistance, destruction of natural enemies, environmental impacts) make alternative control strategies necessary for sustainable broccoli production. Research is limited with respect to non-chemical management tactics for lepidopterous broccoli pests. Crop diversification may provide such an alternative, but has received little attention relative to lepidopteran pest control in broccoli.

  Recent studies examined the effects of intercropping broccoli with either a locally-bred, jalapeno-type, pungent, perennial chili pepper, Capsicum annuum L. (Solanaceae), variety ‘Wailua’ (Broccoli-pepper) or yellow sweetclover (YSC), Melilotus officinalis L (Photo 1). Broccoli plants in broccoli-YSC environments sustained the lowest proportion of plants injured by cabbage webworm by 39 days after broccoli planting in the untreated plots and monoculture broccoli contained the highest percentage of CW damaged plants. A significantly greater percentage of plants was damaged in monoculture broccoli compared with broccoli dicultures and broccoli-pepper intercroppings compared with broccoli-YSC habitats, respectively. Of great practical interest is the fact that the percent CW damaged plants in insecticide sprayed and unsprayed broccoli-YSC subplots were identical, suggesting that intercropping with YSC may be a feasible alternative control for CW. Factors responsible for lowering CW densities are unclear. However, given that YSC completely surrounds young broccoli plants and emits a vanilla-odor, the YSC may behave as a physical and chemical deterrent, thus lowering the rate of colonization. No broccoli heads in the pesticide-treated subplots contained insects or associated contaminants.  Mean number of insects per broccoli head in the pesticide-free plots were 0.125, 0.325, and 0.15 in monoculture, broccoli-pepper and broccoli-YSC habitats, respectively.  A small proportion of broccoli heads were contaminated by the DBM and CW in the monoculture and broccoli-pepper habitats. Broccoli plants undersown in the yellow sweetclover living mulch were smallest and suffered the greatest yield reductions compared with broccoli grown in bareground or intercropped with pepper. This was probably due in part to competition between the broccoli plants and the YSC.

Unexpectedly, the only predators observed feeding on lepidopterous larvae in broccoli plots were spiders, mostly Nesticodes = (Theridion) rufipes (Araneae: Theridiidae) (87%), and orb-weavers (Araneidae), lynx (Oxyopidae) (Photo 2), and some unidentified families made up the remaining 13%. Spider densities were lowest on broccoli plants in the broccoli-YSC habitat from early to midseason. Spider populations were generally greater in the broccoli-pepper intercropping compared with broccoli-YSC habitat. Spider populations were more frequently found in monoculture broccoli contrasted with broccoli dicultures in untreated plots. Spider densities were similar among broccoli habitats during the late season. During this study, the YSC did not flower and, therefore, the only obvious potential offerings to predators such as spiders would have been additional shelter and prey. The presence of YSC may have adversely impacted neighboring broccoli plants because spiders that normally resided on broccoli plants may have dispersed to the YSC, resulting in lower populations on broccoli plants in the YSC habitats. These results suggest that the complexity of the entire system must be understood before accurate predictions can be made about natural enemies.