Crop domestication impacts on rhizosphere interactions and nitrogen acquisition

Abstract

Synthetic nitrogen (N) fertilizer is an essential pillar of modern industrial agriculture. Production and application of synthetic N fertilizer, however, are two of the most expensive, energy intensive, and environmentally deleterious processes in agriculture. Therefore, alternative means of providing N in an agroecosystem are of great interest in sustainable agriculture. While many solutions – from cover cropping to intercropping – have been suggested over time it remains unclear if the modern high-yielding crops can thrive in these alternative N conditions. Decades of breeding under high synthetic N input as well as the inherently annual nature of these modern cereal crops may prevent them from fully taking advantage of these alternative N sources. In this dissertation, I explored the impact of domestication on crop rhizosphere interactions and N acquisition, in both retrospective and prospective terms. First, I investigated how modern maize (Zea mays subsp. mays), and its wild relative Teosinte (Zea mays subsp. parviglumis) differed in their ability to adapt to, and take up, cover crop residue N and synthetic N inputs. We designed a 13C (carbon)/15N dual isotope labeling experiment in which we compared the C allocation patterns of modern maize and teosinte in response to synthetic (urea) and organic (cover crop residue) forms of N. Teosinte responded to organic N by increasing its biomass root-to-shoot (R:S) ratio by 50% compared to synthetic N, while modern maize maintained the same biomass R:S ratios in both N treatments. Recent photosynthate R:S ratio (measured using 13C-CO2, 7 weeks after establishment) was greater in organic N than in synthetic N treatments for both modern maize and teosinte (91% and 37%; respectively). Label-derived dissolved organic C (DOC), representing recent rhizodeposits, was 2.5 times greater in the organic N treatments for both genotypes. Modern maize took up a similar amount of organic N as teosinte using different C allocation strategies. Our findings suggest that intensive breeding under high N input conditions has not affected this modern maize hybrid's access to organic N sources while improving its ability to take up synthetic N. Next, I shifted my focus to the novel perennial grains Kernza and perennial wheat. Kernza® is a domesticated intermediate wheatgrass (IWG, Thinopyrum intermedium). Perennial wheat is a hybrid between Kernza/IWG and modern annual durum wheat (Triticum turgidum subsp. durum). Kernza, in addition to being a perennial, may still possess beneficial belowground traits that may have been lost in modern cereals through millennia of aboveground-focused plant breeding. If so, such traits may be passed down to perennial wheat. To characterize root architecture, exudate profiles, and microbial communities of Kernza and perennial wheat in relation to annual wheat, I conducted a greenhouse experiment. We grew three genotypes/species (Kernza, perennial wheat, annual wheat) and collected their root exudates after 8 weeks of growth. The exudates were analyzed via LC-MS/MS for their chemical composition. We extracted DNA from rhizosphere soils and sequenced them for 16S and ITS profiles. Lastly, we scanned the roots to analyze root distribution across different diameter classes. We found that perennial wheat invested more heavily into very fine (< 250 µm) roots compared to annual wheat and Kernza. Perennial wheat also exuded at a greater rate of exudates per amount of root biomass. We suspect that the greater proportion of very fine roots in perennial wheat led to greater surface area and greater specific exudation rate, and that this may be related to hybrid vigor. We did not find evidence of a genotype effect on root exudate or microbial community composition. However, root exudates (overall metabolite profiles) significantly correlated with root architecture (distribution of root volume over different diameter classes) and the microbial community composition. These interactions represent a potential pathway through which plants can exert influence over the rhizosphere microbial community. Overall, these results emphasize the importance of root architecture in mediating belowground interactions. Understanding rhizosphere dynamics and the response to domestication and hybridization can guide further development of robust perennial cereal crops. In a third experiment, I studied how Kernza, perennial wheat, and annual responded to cereal-legume intercropping (biculture) in the field. To do so, we planted each of the three genotypes in monoculture or in biculture with alfalfa (Medicago sativa). We sampled their rhizosphere over two growing seasons and extracted soil DNA to construct rhizosphere 16S and ITS profiles. We hypothesized that 1) rhizosphere microbial community composition of annual wheat and Kernza will be most dissimilar from each other with perennial wheat intermediate, and 2) microbial community composition will shift in biculture, with the greatest change in Kernza and the smallest in annual wheat. We found that the rhizosphere 16S profiles differed significantly from the other two genotypes but the 16S profile of perennial wheat did not differ from that of annual wheat. Perennial wheat seemingly inherited microbial recruitment traits of its annual parent more so than its perennial parent's. Interestingly, inclusion of legumes led to the convergence, rather than divergence, of 16S profiles among genotypes. We postulate that the competitive pressure of alfalfa may have led to this convergence of 16S profiles across genotypes. The fungal community did not show evidence of genotype effect. However, the fungal community composition changed over two years in monoculture but not in biculture. This result implies that fungal community may become distinct over time if it is influenced by only one genotype (i.e., monoculture) rather than two (i.e., biculture). In conclusion, we found evidence of genotype-driven microbial community assembly that changed with legume's competitive pressure. The inheritability of microbial assembly was present but skewed towards the annual parent. Our study demonstrates the importance of including rhizosphere interactions in our evaluation of novel cereal crops in and out of cereal-legume biculture. In a final study, I investigated how rhizosphere microbial ecology of these three genotypes (Kernza, annual wheat, and perennial wheat) could be linked to their ability to acquire N from neighboring alfalfa plants. We designed a greenhouse study in which we planted all three cereals in monoculture or in biculture with alfalfa and used 15N leaf feeding technique to track the movement of N from alfalfa to cereals. In addition, we also extracted DNA from the soil and sequenced it for 16S rRNA profiles. Arbuscular mycorrhizal fungi (AMF) infection rate was also measured on all cereals and legumes. We hypothesized that: 1) annual wheat would produce the greatest biomass but Kernza would have highest proportion of legume derived N in its biomass, 2) all microbial communities will shift in biculture, with the greatest change in Kernza and the smallest in annual wheat, and 3) Kernza would have the highest rate of infection from AMF, especially in biculture. Surprisingly, we found no evidence of genotype or cropping system (monoculture or biculture) effect on either proportion or absolute amount of N derived from legume. We did find, however, that DOC concentration was higher in cereal rhizosphere grown in biculture than in monoculture, suggesting greater belowground investment in exudates when the grasses are grown with a legume. Despite this trend, annual wheat had much lower microbial biomass carbon (MBC) level in its rhizosphere compared to the perennials, in biculture. We contended that this may be due to substrate suitability of annual wheat's rhizodeposit. We also found that AMF infection rate was in fact the lowest in Kernza. Lastly, we found that 16S profiles of all three cereals shifted towards that of alfalfa in biculture. This trend might suggest microbial spillover, wherein rhizosphere microbial community of one genotype colonizes that of a neighboring plant, from alfalfa rhizosphere. Overall, we demonstrated that quantifying the N transfer in the rhizosphere can provide important insight into how these genotypes may be inducing changes in soil biogeochemistry in response to neighboring legumes. In summation, this dissertation provides links between crop genotype, root exudate chemistry and rate, microbial community assembly, and their biogeochemical consequences, in alternative N environments. Deepened understanding of how complex rhizosphere interactions may affect internal N cycling could be leveraged to further optimize these unique systems such as perennial cereal-legume biculture. In doing so, we will be one step closer to a more sustainable future, that is less reliant on synthetic N fertilizers.