Biomass, root distribution and overyielding potential of faba bean/wheat and white clover/ryegrass mixtures

Abstract

Mixtures of legumes and non-legumes are often characterized by higher grain and biomass yields compared to their pure stands. Complementarity between plant species is assumed to be the major driver behind this aboveground overyielding. Cultivar characteristics can affect mixture performance. Nevertheless, novel legume cultivars/genotypes are primarily bred and tested for pure stand purposes. However, well-performing genotypes in pure stands do not necessarily perform similarly well in mixtures. To fully understand mixed cropping systems, it is necessary to investigate their underlying spatiotemporal above- and belowground processes. Roots are of particular importance for the plant, as they acquire water and nutrients. Nonetheless, little is known about differences in root biomass and distribution between pure stands and mixtures. So far, the lack of a simple and time-efficient method has hampered the analysis of root species proportions in mixtures. In the present study, novel legume genotypes of arable land and grassland were sown as pure stands and mixtures with non-legumes. Two different field experiments were conducted at the experimental station ‘Reinshof’ of the Georg-August-University of Goettingen (Germany) to investigate the biomass, root distribution and overyielding potential of these pure and mixed stands. In the arable land experiment, eight genotypes of winter faba bean (Vicia faba L.) and one cultivar of winter wheat (Triticum aestivum L., cv. Genius) were sown in pure stands and in substitutive 50/50 mixtures. The intra- and interspecific variation of shoot and root biomass, the horizontal and vertical root distribution and the overyielding potential were investigated in all crop stands at full flowering of faba bean. Aboveground biomass of 1 m² was harvested and roots were sampled in May 2015 and May 2016. Root samples were taken on and between rows with a root auger down to 60 cm soil depth. Fourier transform infrared (FTIR) spectroscopy was used to quantify species-specific root biomasses in mixtures. The vertical root distribution was evaluated using the equation y = 1 - βd (Gale and Grigal 1987). To assess above- and belowground mixture overyielding, the relative yield total (RYT) was calculated for shoot and root biomass. The results showed that all FTIR quantification models performed well in the prediction of root species proportions. Roots of both species proliferated into the soil space between the rows and under the mixture partner’s row to a similar extent. In mixtures, faba bean and wheat on their own row produced higher root fractions in shallower soil layers than in pure stands, while simultaneously both species had more roots in deeper soil layers under the partner’s row than on their own row. Overyielding of faba bean/wheat mixtures was more pronounced for belowground biomass than for aboveground biomass. In mixtures, faba bean genotypes differed significantly in root biomass, root:shoot ratio, overyielding potential and vertical root distribution on wheat rows but not in shoot biomass. In the grassland experiment, the root biomass of eight genotypes of white clover (Trifolium repens L.) and one perennial ryegrass genotype (Lolium perenne L., Elp 060687) were investigated. Four different crop stands were established in May 2014: (i) unfertilized clover pure stand of each clover genotype, (ii) unfertilized ryegrass pure stand, (iii) N-fertilized ryegrass pure stand and (iv) unfertilized mixture of each clover genotype with ryegrass. Similar to the first experiment, root sampling was conducted from 0 to 60 cm soil depth in June 2015. Clover and ryegrass root proportion in mixtures was determined via FTIR spectroscopy. Belowground RYT was calculated for each mixture. The results showed that FTIR models demonstrated a satisfactory residual predictive deviation. In pure stands as well as in mixtures, clover produced significantly lower root biomasses than ryegrass. Nitrogen fertilization did not affect the root biomass of ryegrass. In pure stands, clover root biomass differed significantly between genotypes. Furthermore, root RYT was higher than one in all the eight clover/ryegrass mixtures but differed between the genotypes. This belowground overyielding was mainly caused by the high relative root biomass of ryegrass. The present study showed that FTIR spectroscopy is a suitable tool for the identification and quantification of root species in legume/non-legume mixtures. From the two experiments, it can be concluded that both faba bean/wheat and clover/ryegrass mixtures overyield with regard to root biomass. Root overyielding in legume/non-legume mixtures compared to the pure stand equivalents might lead to better resource utilization and enhanced aboveground yields of these systems. The fact that genotypes performed differently in pure and mixed stands shows the potential of legume breeding for mixture purposes. In both arable land (Vf5) and grassland (Tr6), one legume genotype was identified for further breeding in mixed cropping systems. The results of the present study suggest that investigations of root properties should be included in mixture breeding processes.