Oilseed rape (Brassica napus L., OSR) is a globally important crop, both for the production of high-value vegetable oil and protein-rich animal feed (as exhausted rapeseed meal, a by-product from oil extraction), and as a break crop in cereal-dominated cropping systems (Angus et al., 2015). A range of pests, pathogens and weeds constrain OSR production, resulting in intensive pesticide use in conventional farming systems and yield instability in organic systems (Valantin-Morison et al., 2007; Williams, 2010; Lutman, 2018; Zheng et al., 2020). Crop protection practices face challenges from stricter pesticide regulations and increased pesticide resistance (Duke, 2012; Brandes and Heimbach, 2018; Arthey, 2020; Zheng et al., 2020). New ways to reach profitable OSR yield levels by using more sustainable crop protection methods are urgently needed.

Diversification of OSR production by increasing the crop and non-crop plant diversity on the local (field) and landscape scale holds promise for the development of more sustainable crop protection strategies (Isbell et al., 2017). We focus on the potential of intercropping (growing a mixture of two or more species), as it may offer mechanisms for simultaneous control of pests, pathogens and weeds (Malézieux et al., 2009). Legumes are of special interest for intercropping with OSR since their capacity for symbiotic nitrogen fixation increases overall nitrogen-use efficiency and minimizes competition for soil and fertilizer-derived nitrogen (Verret et al., 2017; Dowling et al., 2021). Intercropping OSR with legumes is relatively understudied (but see Dowling et al., 2021), especially when it comes to the implications for crop protection. In addition, intercropping studies rarely evaluate crop protection strategies against multiple crop antagonists simultaneously.

Seminal intercropping studies, like Root (1973), have been conducted on vegetable Brassicaceae crops closely related to OSR (e.g., cabbage, collards, broccoli). Intercropping can reduce insect pest infestation and crop damage by disrupting visual or olfactory host plant location, reducing host plant quality or increasing natural enemy activity (Root, 1973; Trenbath, 1993; Hooks and Johnson, 2003; Finch and Collier, 2012). In the few intercropping studies exploring insect pest responses in OSR, legume intercrops reduced damage from cabbage stem weevil (Ceutorhynchus pallidactylus) (Paulsen et al., 2006) and rape winter stem weevil (C. picitarsis) (Cadoux et al., 2015), but not flea beetle (Phyllotreta cruciferae) (Weiss et al., 1994). Intercropping OSR with red clover (Trifolium pratense) did not affect the population of pollen beetles (Brassicogethes aeneus) the same season, but decreased the number of emerging pollen beetles in the next generation (Hokkanen, 2004).

There is to our knowledge no published research on the effects of intercropping on slugs in OSR. Research on the effects of crop residue on slug abundance and crop damage has been inconsistent and mostly done in tillage experiments (Voss et al., 1998; Glen and Symondson, 2003; Douglas and Tooker, 2012; Howlett, 2012; Rowen et al., 2020). A microclimatic effect from the intercrop similar to the effect of crop residue that is beneficial for slugs has been assumed (Gaba et al., 2015), yet studies have found a reduction in slug damage for the main crop in intercropped systems (Glen, 2000; Brooks et al., 2005; Choi et al., 2006). The mechanisms for damage reduction are not well-understood, but dilution effects, based on the generalist feeding requirements of slugs, antifeedant effects and benefits for slug predators have been documented (Cook et al., 1997; Dodds et al., 1999; Le Gall and Tooker, 2017).

Intercropping has been shown to reduce some plant diseases, especially foliar diseases (Boudreau, 2013). Several mechanisms may be involved, such as resource dilution, disruption of spatial or temporal disease cycles, changed microclimate, allelopathic effects, effects on the host morphology or physiology, or changes in the suppressiveness of plant or soil-associated microorganisms (Ratnadass et al., 2012; Boudreau, 2013). Despite having a denser canopy resulting in a microclimate favorable for foliar disease development, intercropping OSR and peas produced generally healthy crops (Soetedjo et al., 2003; VanKoughnet, 2016; Dowling et al., 2021). The effects of intercropping on disease susceptibility and mechanisms for resistance, however, are still relatively understudied in legume-OSR intercrops.

Growing crop species mixtures has been shown to result in lower weed density and weed biomass compared to sole crops (Liebman and Dyck, 1993). Weed suppression is particularly effective in grain legume and cereal intercrops, in which weed biomass is reduced compared to grain legume sole crops (Bedoussac et al., 2015). Weed suppression in grain legume-OSR intercrops has been less studied. Research on mixtures of winter OSR and legume service crops where the legume is sown simultaneously with oilseed rape and killed by frost during winter, has shown promising effects on weed suppression during autumn (Cadoux et al., 2015; Lorin et al., 2015). The suggested mechanism is that the frost-sensitive legumes increase the competition for light by occupying surfaces that are not covered by the small winter OSR plants during autumn (Cadoux et al., 2015).

In summary, intercropping OSR with legumes has the potential to make crop stands more competitive against weeds and reduce pathogen, insect pest and slug damage through physical, visual and olfactory disruption, habitat alteration and promotion of biological control. The empirical evidence base for such varied intercropping benefits for OSR is limited with few attempts to simultaneously measure crop protection against weeds, pests and pathogens using a multifunctional approach. Our aims were to assess the multifunctional crop protection implications of intercropping OSR with both frost-sensitive and overwintering legumes.

Slug damage to OSR plants in autumn 2019 did not differ between treatments in the September or November assessments, though the cage trial showed lower slug damage to OSR in the OSR + Sfaba treatment (Table 1). CSFB damage was significantly lower in OSR + Clover compared to other treatments in both September and November; CSFB larval abundance was significantly lower in both OSR + Clover and OSR + Sfaba compared to OSR alone (Table 1). In June 2020, damage from PB did not differ between treatments. Damage levels from BPM were low overall, but they were higher in OSR + Clover compared to OSR + Wpeas (Table 1). No symptoms of Sclerotinia root rot were detected in any OSR treatment. The assessment of faba bean rust (U. fabae) showed a significantly lower disease severity index on the leaves of winter faba bean in the OSR + Wfaba intercropped plots (17.86 ± 3.57) compared to the Wfaba sole crop (46.43 ± 2.70, F = 40.7, df = (1, 6), p <0.001). The November 2019 weed assessment showed dicots were predominant while monocots, including cereals, did not exceed 5% of the total weed cover. Weed biomass from May 2020 was lowest in Wfaba and Wpeas and highest in OSR + Wfaba (Table 2). OSR biomass was similar across treatments while legume biomass was highest in the legume sole crops (Table 2). Grain OSR yield was lower in OSR + Wpeas compared to other treatments and winter peas gave a higher legume yield than winter faba beans when intercropped with OSR (Table 2). The land equivalent ratio was 1.13 for OSR + Wfaba. No LER was calculated for OSR + Wpeas since the winter peas sole crop was severely lodged and minimal harvest was possible only in one block. The multifunctionality of intercrops is shown relative to the OSR sole crop in Figure 1.

Our field experiment and lab study supported the idea that intercropping OSR with legumes can improve crop protection in several cases and present a framework for multifunctional metrics to assess intercropping systems.

In the field experiment, slug damage in both September and November was low and similar in all treatments perhaps indicating the drawback of plot experiments in being able to evaluate the impacts of intercropping for mobile organisms. In contrast, the supplementary feeding experiment with cages showed reduced slug damage to OSR plants when intercropped with spring faba bean. Larger field studies are an important next step, but these findings indicate that intercropping does not increase slug damage to OSR plants and the cage study illustrates the potential for intercropping to reduce slug damage to OSR. Further experimentation is needed to explore any mechanisms involved in the potential of intercropping to reduce slug herbivory.

Cabbage stem flea beetle feeding damage to OSR seedlings was reduced in both September and November when intercropped with berseem clover and in OSR + Wpeas for November damage. CSFB larval abundances were reduced by both berseem clover and spring faba bean intercropping, as compared to the OSR sole crop indicating an effect on oviposition which was measurable despite the small plot size. These results are in line with Swiss field experiments where intercropping of OSR with mixtures that included berseem clover reduced adult flea beetle damage to OSR seedlings, though they found no effects on larval abundances (Breitenmoser et al., 2020). Our results suggest that annual legumes planted early in the season decrease oviposition by the early-season pest CSFB. The earlier sowing date, higher biomass and ground coverage of these legumes in autumn might increase their visual and olfactory interference with host plant location by CSFB. The late-planted winter grain legume intercrops (OSR + Wfaba and OSR + Wpeas) are, on the other hand, not present in the field early in the season when most damage from CSFB and slugs occurs.

Pollen beetle damage did not differ between treatments, while intercropping OSR with berseem clover increased the percentage of pods damaged by brassica pod midge, compared to when OSR was intercropped with winter peas. The level of BPM damage was low across treatments. Brassica pod midge damage occurs in late spring when most of the berseem clover intercrop is no longer present in the field though intercropping species-driven variation in OSR nutritional content, which affects insect herbivory in OSR, could also play a role (Veromann et al., 2013; Blake et al., 2014; Skellern and Cook, 2018).

Effects of intercropping on pathogen control for OSR could not be evaluated due to an absence of Sclerotinia stem rot in the experiment. The reduction of faba bean rust in intercropping is in line with a previous study on intercropping maize and faba bean, where the severity of chocolate spot disease (Botrytis fabae) was reduced compared to the faba bean sole crop (Stoltz et al., 2018).

Weeds were more successfully controlled in OSR when intercropped with spring faba bean or clover compared to OSR intercropped with winter faba bean (Table 2). This is in line with earlier findings and could be explained by the ability of spring faba bean to quickly produce biomass that competed with weeds (Lorin et al., 2015; Verret et al., 2017). Intercropping with frost sensitive legumes also facilitated mechanical weed control in spring, which was not possible for grain legumes sown in between OSR rows in late autumn and kept for harvest.

Nitrogen fertilizer use was reduced by 25% in the intercropping treatments without any OSR yield penalty, except when OSR was intercropped with winter peas. OSR yield was reduced by ~50% when intercropped with winter peas, which could be linked to the high biomass of the winter peas that were maintained until OSR harvest, but this treatment also resulted in a pea grain yield that was higher than the OSR yield and the most consistently stable pest control across the species measured in this study (Figure 1). The relatively low OSR yield in all treatments in this experiment, relative to the southern Swedish (Skåne) 2016–2020 average of 3,384 kg/ha in conventional production (Swedish Board of Agriculture., 2021), is attributed to the low N fertilization and lack of pesticide inputs.

Technical limitations when it comes to managing, harvesting and separating mixed crop species constitute an important barrier to the uptake of OSR intercropping (Dowling et al., 2021). This makes intercropping of OSR with service-crops that are not harvested, such as spring faba bean or berseem clover sown in autumn, of interest, as they can provide intercropping benefits without the need for harvesting and processing mixed crops. Intercropping with a grain legume intended for harvest also provides benefits with potential increased legume yield. This was the case with winter peas, where the OSR provided physical support until maturity in the OSR + Wpeas intercrop, facilitating pea grain harvest. Our experiment shows that intercrops not maintained for the full crop cycle of oilseed rape are less likely to provide crop protection benefits against pests that attack the crop at late growth stages, such as brassica pod midge, though they might be effective against early season pests like cabbage stem flea beetle thus improving overwintering capacity.

There are several potential benefits of intercropping OSR with legumes, some of which we could confirm in the field experiment. The concerns that intercropping would facilitate increased slug and pathogen problems due to changes in the microclimate were not supported in this study. Each legume intercrop, however, had its own benefits and drawbacks. Our results suggest that the choice of intercrop will depend on what aspects of pest, pathogen and weed control are rated as most important. This pilot study provides the foundation for intercropping studies on the field scale, which would overcome the limitation of using small plots to evaluate effects on mobile organisms like insects. Multi-year assessments would better capture the long-term consistency of multifunctional crop protection provided by different OSR and legume intercropping mixes. Due to the necessity of cropping systems to provide control against multiple crop antagonists, we propose a multifunctional perspective to be adopted more often to quantify how intercropping facilitates or creates resilience against pests, pathogens and weeds.

The datasets generated for this study can be found in the Swedish National Data Service at https://doi.org/10.5878/ppwm-4826.

SE, PA, GC, HF, ML, A-CW, and OL conceived the ideas and designed methodology. SE, GC, ML, PA, A-CW, and OL collected data. SE analyzed the data and led the writing of the manuscript. All authors contributed critically to the drafts and gave final approval for publication.

Funding was provided by Platform Plant Protection at the Swedish University of Agricultural Sciences and Swedish farmers' foundation for agricultural research, Grant O-18-20-165.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.