The integration of agroecological principles into conventional agriculture is increasingly being championed as a means to sustainably increase agricultural production (Rillig et al., 2016; French, 2017; Begum et al., 2019). In the present study, we investigated the effect of intercropping on mycorrhizal root colonisation and plant shoot phosphorus.
Comparison of root colonisation in the field and glasshouse experiments suggests some level of host plant preference for specific AMF species, and vice versa. Chickpea and lentil had
similarly high levels of colonisation in the field (~ 60%), but in the glasshouse colonisation of chickpea roots was between 15–30%, less than both the glasshouse lentil and field chickpea and lentil. While plant-AM specificity was previously thought to be low (Mosse, 1975; Brundrett, 2009), a meta-analysis by Van Geel et al. (2016) found that symbiosis with different AM species produces different growth and nutrition responses in a given plant. Similarly, Xavier & Germida (2002) report that plant response to inoculation with AMF varies significantly depending on the AMF species. In our study, the field soil would have harboured a number of different mycorrhiza species (Vályi et al., 2016; Guzman et al., 2021) while the inoculum in the glasshouse experiment contained only Rhizophagus irregularis spores. The low colonisation of chickpea roots in the glasshouse but high colonisation in the field suggests that chickpea prefers to associate with AMF species other than R. irregularis, while lentil freely associates with R. irregularis and potentially other species. The literature reports colonisation of chickpea roots of ~ 60% by AMF species Funneliformis mosseae and Glomus intraradices (Tavasolee et al., 2011; Li et al., 2022), a similar level of colonisation to the chickpea in the field, supporting the notion of host-AMF species specificity. This suggests that farmers should potentially focus on cultivating diverse communities of mycorrhiza in the soil to ensure improved growth of multiple crops (Guzman et al., 2021).
Root mycorrhizal colonisation was only correlated with shoot phosphorus in the lentil in the glasshouse experiment, contrary to our first hypothesis. While a lack of correlation between mycorrhizal root colonisation and shoot phosphorus is to be expected for non-mycorrhizal canola (Fester & Sawers, 2011; French, 2017), it is especially surprising for linseed, which is known to be highly mycorrhizal (Thingstrup et al., 1998; McGonigle et al., 2011; Rahimzadeh & Pirzad, 2019). As in the literature, we also found high levels of mycorrhizal root colonisation in linseed in both the field and glasshouse experiments.
The lack of correlation between mycorrhizal colonisation and shoot phosphorus in the field experiment at H2020 could be explained by the adequate supply of background soil phosphorus (Table 1). The most significant benefits from AMF have been reported under phosphorus limitation, with adequate N, light, and water supply (Thingstrup et al., 1998; Smith & Smith; 2011; Ryan & Graham, 2018; Tran et al., 2019). The background Colwell-P value of the soil at H2020 was well above the critical range for most oilseeds (16–19 mg kg− 1) and legumes (20–29 mg kg− 1) in low to mid rainfall environments (Bell et al., 2013). Given ample phosphorus supply, plants would not have had to rely heavily on the AMF symbiosis to fulfil their growth requirements. At H2021 and in the glasshouse, however, the soil did not have sufficient Colwell P (10 mg kg− 1 and 14.5 mg kg− 1, respectively). In these scenarios, the lack of correlation between shoot and colonisation could be a function of the short duration of the experiment. Perhaps P limitation at later growth stages would have necessitated greater reliance on the AMF symbiosis to provide phosphorus as plant phosphorus demands change over time (Veneklaas et al., 2012). Field studies of longer duration and on P-limited soils are needed to investigate the role of AMF in intercrop phosphorus nutrition.
In contrast to mycorrhizal colonisation, intercropping did affect shoot phosphorus. This is in line with the literature, which suggests that intercrop component interactions alter P availability in the rhizosphere (Costa et al., 2014; Nie et al., 2016). In the glasshouse experiment, canola shoot phosphorus was highly positively correlated with chickpea intercropping, and was greater in the intercrop with chickpea than in either the sole crop or when intercropped with lentil. This suggests a phosphorus benefit of intercropping with some species combinations and not others. Enhanced resource acquisition in an intercrop can be explained by positive interspecific interactions such as resource partitioning or facilitation (Hinsinger et al., 2011; Li et al., 2018), or by competitive dominance wherein one crop component increases resource acquisition at the expense of the other (Loreau & Hector, 2001; Li et al., 2018). In the case of the chickpea-canola intercrop, the increased canola shoot phosphorus but similar chickpea shoot phosphorus relative to their respective sole crops suggests that the chickpea facilitates increased uptake in the canola, rather than the canola outcompeting the chickpea. Chickpea, along with field pea and faba bean, has been shown to have a superior ability to mobilise soil phosphorus compared with other legumes (Miheguli et al., 2018). It exudes large amounts of acid phosphates that hydrolyses organic phosphorus into plant available inorganic forms (Li et al., 2003; Liao et al., 2020). More research is needed to test compatible legume-oilseed intercrop combinations with regards to phosphorus acquisition. Research should focus on investigating the mechanisms behind the intercrop benefit, with selection for complementary and facilitative interactions that enhance the nutrient dynamics of the whole system, over competitive dominance, which enhances one crop at the expense of the other.
Contrary to our second hypothesis, mycorrhizal colonisation of plant roots was not affected by intercropping. The one exception to this was the lentil in the glasshouse experiment, which had decreased colonisation in the intercrop compared with the sole crop. Most surprisingly, being intercropped with the non-mycorrhizal canola did not affect lentil or chickpea root colonisation (excepting lentil in the glasshouse) relative to the sole crop or being intercropped with the highly mycorrhizal linseed. This is unexpected, as canola is known to release glucosinolates into the soil (Gimsing & Kirkegaard, 2009; Couëdel et al., 2019), that are toxic to fungi and are thought to contribute to the plant’s non-mycorrhizal status (Floc’h et al., 2022). Indeed, mycorrhizal plants, such as maize and linseed, had reduced mycorrhizal colonisation and yield in years where they follow canola in crop rotation, compared with another mycorrhizal crop (McGonigle et al., 2011; Higo et al., 2017).
In explanation of the nil canola effect, Trenbath (1993) and Boudreau (2013) suggest that sowing brassica species with a companion legume may negate any negative effects on soil biota, and it is possible this was the case with the mycorrhiza in our lentil-canola and chickpea-canola intercrops. Alternatively, it is possible that we harvested the plants before the canola could exude sufficient amounts of glucosinolates to affect the mycorrhizal colonisation of its companion, and that it would have released potentially more toxic levels when more established. Glucosinolate content in canola is reported to increase up until budding (~ 100 DAS, growth stage 2.2; Berkenkamp, 1973), with the majority of total glucosinolate content at this time in the plant roots (40–86%) (Clossais-Besnard & Larher, 1991; Sarwar & Kirkegaard, 1998). Further research involving whole-season experiments, with sampling at multiple growth stages (such as early flowering, podding and maturity) is needed to establish the longer-terms effects of intercropping with canola. An important aspect of this is break crop rotation research. If the yield of mycorrhizal cereal crops such as wheat is worse following canola, as much of the literature suggests (Owen et al., 2010; Bakhshandeh et al., 2017), could legume-canola intercrops take the place of sole canola? This would maintain the break crop and market relevance benefits, whilst also providing continuation of the mycorrhizal community for the following cereal crop.
Our study found lentil to be highly mycorrhizal, which is consistent with many previous studies (Khan et al., 1988; Amirnia et al., 2019). As discussion of sustainable farming practices gains momentum, farmers are increasingly focussing more attention on enhancing and maintaining abundant communities of mycorrhiza in the soil (Ryan & Graham, 2018; Kirkegaard & Condon, 2019). Although linseed is a highly mycorrhizal crop (Grant et al., 2009), it has a relatively small global market (FAOSTAT, 2022). By using lentil as their ‘mycorrhizal’ crop, farmers could ensure the maintenance of soil mycorrhizal populations with a species that has high market relevance. In Australia, for example, lentil has the third largest planted area of the pulse crops (‘000 ha; ABARES, 2022), and is already included as break crops in cereal dominant rotations (GRDC GrowNotes, 2018).