Final Report for GNE13-059
Sustainable corn agriculture can be developed by promoting the efficient use of nutrients and water by the root systems, in combination with beneficial associations with soil microbes, and plant resistance to soil-borne pathogens. A promising root trait that increase rooting depth, and decrease the metabolic cost of soil exploration in maize, is the production of air spaces in the cortex, or root cortical aerenchyma (RCA). Plants that have more RCA have less living tissue to sustain belowground, and can therefore expand in bigger volumes or deeper depths in the soil profile. Research results indicate that plants with augmented RCA grow better under nutrient and water limitations, compared to plants that have low RCA. A possible tradeoff of selecting for plants with more RCA, is the reduction in microhabitats for beneficial microbes such as mycorrhizal forming fungi, or the facilitation of the spread of pathogens throughout the air spaces in the cortex. With this study we measured the RCA of commercial hybrid maize entries, and measured the effect of augmented RCA on the fungal colonization of field and greenhouse-grown plants. We found that plants with significantly augmented RCA, ranging 15 – 20% of aerenchyma in the cortex, did not showed a significant reduction in arbuscular mycorrhizal colonization (AMC), which values were 0 – 80%. High RCA plants did not show decreased root rots (RR) in field samples either, showing RR of 0 – 90 % of the root length. High RCA plants did not present augmented or decreased pathogen infection (PI) with Fusarium verticillioides in the greenhouse, with values 0 – 10 pg fungal DNA per microgram of root DNA and per gram of root. These results mean that plant breeding programs could target augmented RCA without major concerns on the interactions with arbuscular mycorrhizas, RR in the field, or PI by F. verticillioides. Other anatomical traits promising in maize agriculture, identified by our research group, and their effects on AMC, RR and PI need to be analyzed and studied in order to deliver integral solutions to plant breeders and ultimate, to farmers.
Climate change and excessive fertilizer inputs threaten maize production in the northeast of the United States. First, climate change is projected to cause more variable water availability in the future due to changes in rainfall distribution in time and space. Additionally, fertilizer inputs need to be controlled and minimized to stop the contamination of the Chesapeake Bay and its watershed. Predictions of climate change effects in the northeastern US anticipate increases in precipitation which will leach nutrients applied to crop fields causing increased nutrient loading of the Bay (Strzepek et al., 2010). Accordingly, farmers will need plants with greater drought/flooding tolerance and increased fertilizer uptake efficiency. To face these challenges, maize cultivars with better and more efficient root systems may be developed or selected through plant breeding programs.
The selection of plants with root anatomical traits that reduce the metabolic cost of soil exploration is a promising alternative for sustainable maize production. The general hypothesis is that metabolically “cheap” roots will result in more efficient performance of the plant under stressful conditions (Lynch, 2013). For example, high levels of root cortical aerenchyma (RCA)— air spaces in the root cortex— is a root trait that has conferred improvements to maize growth and biomass accumulation under low N and drought in Pennsylvania soil (Zhu et al., 2010). In fact, RCA is a promising trait under low N and P stress as demonstrated by greenhouse studies and modeling simulation approaches (Postma and Lynch, 2011b, a). These results suggest that plants with increased RCA may experience better growth or improved recovery under prolonged drought periods and are more efficient acquiring fertilizers.
Although these results with RCA are promising, we still need to understand the tradeoffs for the selection of “cheap” roots in maize. Since a cheap root system has lower carbon allocation to the roots, and more air spaces in the root tissue, some eco-physiological processes like root exudation, tissue accumulation, and root respiration may be markedly different in roots with more RCA. In particular, the ecological interaction of roots with soil microorganisms is an aspect that may be affected by a reduced substrate for the microbial growth in the rhizosphere or in the root cortex. Another possibility is that RCA could be an avenue through which microorganisms may easily spread inside the plant. These aspects of maize cultivation in the agronomical conditions of the northeast region are unknown and this project is the first attempt to study the anatomy combined with the microbiology of the maize root systems in Pennsylvania.
Results obtained in this study relevant for sustainable agriculture, which is based on the integration of plant and soil biology. First, we are providing information about the variation of several anatomical traits in the hybrids that are currently used in Pennsylvania. This information could be considered a baseline for the development of better cultivars that use fertilizers and water more efficiently. Second, we are complementing this information with the potential interactions anatomical root traits with soil microorganisms. Microbial communities in soils and their interaction with roots must be understood because the microbes are, together with plants, the main nutrient-transforming elements in soils. Here, we are delivering information about anatomical traits that influence ecological and physiological processes such as the interaction of maize roots with mycorrhizal forming fungi and with root rots pathogens.
The overall objective of this project was to evaluate the effect of the anatomical trait root cortical aerenchyma (RCA) of corn hybrids on the interaction of roots with soil microorganisms in Central Pennsylvania. The soil microbes studied are arbuscular mycorrhizae forming fungi, and root pathogens. The former was determined by measuring arbuscular mycorrhizal colonization (AMC), and the latter was quantified as root rots in the field (RR), and pathogen infection (PI) in the greenhouse. AMC and RR/PI will be used herein to denote the two soil microbe groups in reference to their interaction with roots.
We completed the activities planned per specific objective, as described below:
2.1. Evaluate the variation of Root Cortical Aerenchyma (RCA) of corn plants normally planted by farmers in Pennsylvania. Root samples were collected at three experimental sites of the 2014 Grain and Silage Hybrid Corn Test across Pennsylvania (Figure 1). The samples were analyzed for anatomical traits and the variation of RCA was measured as initially proposed. The method allowed for the measurements of other anatomical traits that are being used in follow-up analysis after the conclusion of this project. Although we used commercial maize entries usually planted in Pennsylvania for the field part, we reevaluated the use of these entries in the greenhouse experiments because significant variation was found in anatomical phenotypes among maize entries in the field, even within samples from the same seed company. Since we had difficulties finding untreated seeds of the exact same entries by greenhouse planting, we decided to use inbred lines for the greenhouse experiments in 2014 and 2015.
2.2. Correlate RCA levels with Arbuscular Mycorrhizal Colonization (AMC) of corn plants normally planted by farmers in Pennsylvania. Representative root subsamples from the plants described above were analyzed for AMC. Correlation analysis were performed to determine the effect of RCA on AMC, both in the field and greenhouse environments.
2.3. Correlate RCA levels with Root Rots (RR) and pathogen infection (PI) of corn plants normally planted by farmers in Pennsylvania. The field samples described in 2.1., were scored for RR in the second and third whorls (in the inner part of the crown root, Figure 2). Also, a greenhouse study was conducted in July-August 2015 to quantify PI in a more controlled fashion. Technical details such as fungal pathogen species, inoculation technique, growth media composition and plant growth time, were determined through three preliminary experiments conducted in 2014 and 2015. We used a mild-pathogenic strain, of Fusarium verticillioides, isolated in Pennsylvania, and kindly provided by Dr. Gretchen Kuldau, at the Plant Pathology Department at Penn State. A protocol for the molecular quantification of F. verticillioides with real time quantitative PCR (RT-qPCR) was developed and employed on the greenhouse root samples. Although we had originally proposed to submit our samples to the Genomic Core facility at Penn State for the pathogen quantification with qPCR, we decided to develop the technique in collaboration with Dr. Dawn Luthe Lab, at the Plant Science Department at Penn State.
General approach: Two field studies were used in this project. The experiments were part of the program “Pennsylvania Commercial Grain and Silage Hybrid Corn Test Report”, by the Outreach program and the Department of Plant Science of the College of Agricultural Sciences at Penn State University. First, during the first year (2013), root samples were collected at the Penn State Research farm in order to select commercial maize entries with contrasting RCA. The purpose of this first sampling was to select maize entries for subsequent greenhouse studies. Second, a more extensive sampling on three different commercial farms, was conducted during the second year (2014), in order to study the variation of RCA, AMC and PI in 30 genotypes, and to study the correlations among these variables. Plants that had shown contrasting levels of RCA were selected for the greenhouse experiment. Despite all our efforts to obtain untreated seeds of these genotypes, we learned that farmers rarely plant the exact same entries in consecutive years, and seed companies seldom release the same entries year after year. Consequently, we decided to use inbred lines available our laboratory and that had been previously phenotyped; this is, their RCA levels confirmed, for the greenhouse experiments in 2014 and 2015. In the greenhouse, inoculation trials with these maize lines and mycorrizal forming fungi, and with the maize root pathogen Fusarium verticillioides served to test the hypothesis of correlation between AMC and PI with RCA.
Data Analysis and Experimental Design: We employed mainly ANOVAs and correlation analysis to evaluate the variation in the RCA, AMC and PI, and the degree of relatedness of RCA with AMC and RPI respectively. Multivariate techniques such as principal component analysis and cluster analysis were performed to determine groups of samples according to their anatomical traits. Finally, we used the classification tree analysis, Random Forest, to determine the most important anatomical root traits for AMC and PI.
In a more specific way, the methods used to accomplish each activity, for each objective are described below:
Activity 1. Measure RCA of maize currently planted in Central Pennsylvania. The goal of this objective was to quantify the variation of RCA among hybrid plants under the regular growth conditions of Central Pennsylvanian farms. Three farms shown in Figure 1 were selected for sampling. Root crowns were collected from 30 - 40 entries, with three replicates per entry. We used the 2014 Grain and Silage Hybrid Corn Test across Pennsylvania. Three plants per plot (for a total of 900 crowns) were excavated and hose-washed, and root segments collected from two representative crown roots per plot, and preserved for anatomical analysis. Root segments from the crowns were obtained according to described methods (Burton et al., 2013). The segments were then laser-ablated in the laboratory, and images of the cross-root sections acquired, edited and analyzed with RootScan. This is the software in use for obtaining anatomical traits like RCA, developed by Burton et al. (2012) in our laboratory.
Activity 2. Correlate RCA levels with mycorrhizal colonization. Field samples were naturally colonized with mycorrrhizal fungus present in the soils. No inoculation was performed on these field trials. In the greenhouse, the selected maize lines were inoculated with 300 spores of Rhizophagus intraradices (MIKE spore suspension for research, Premiere Tech Ltd. Quebec, Canada). Mycorrhizal colonization was assessed in approximately 4 g (fresh weight) of roots. The tissue was cleared and stained according to described methods (Vierheilig et al., 1998). AM colonization was quantified as the percentage of root length colonized in three fragments comprising a total length of 30 cm of the root main axis. Counts were done following either the grid-line intercept, or the field-intercept method, both described in Brundrett (1991). We found that the field-intersect method gives more reliable results given the heterogeneity of fungal colonization in field root samples. Microscopic, instead of stereoscopic observations, were more accurate when the roots had fungal colonization by non-AMC species, as the observer has the option of bigger magnification to confirm the presence of arbuscules in the root cortex. For the greenhouse samples, the gridline-intersect method was still valid and easy to use.
Activity 3. Correlate RCA levels with root pathogen colonization. The field samples described above, were scored for RR in the second and third whorls (in the inner part of the crown root, Figure 2) at harvest. A greenhouse study was conducted in July-August 2015. Technical details such as inoculation technique, growth media composition and plant growth time were determined through three preliminary experiments conducted in 2014 and 2015. Mesocosm systems, consisting in 1.5 m pipelines filled with 28 L of media (35% vermiculite, 5% autoclaved soil, 55% sand, 5% perlite), were used for the definitive study. Each pot was inoculated with 10 g of fresh mycelia of Fusarium verticillioides provided by Dr. Gretchen Kuldau, from the Plant Pathology Department at Penn State. The isolate was originally obtained in Pennsylvania and is part of the Fusarium collection at Penn State. Maize lines with contrasting levels of RCA were grown from seeds, and sawn in the mesocosms, after heat sterilization (1 min at 65°C). Plants were daily fertigated with 250 mL of a water-soluble fertilizer (285 g of Plantex 40-0-40, by Master Plant-Prod Inc. Ontario, Canada., and 20 µM H2PO4, pH adjusted to 5.5±0.1). Roots were harvested seven weeks after planting, and samples were collected for RCA and fungal colonization quantification. A protocol for the molecular quantification of F. verticillioides with quantitative real-time PCR (qRT-PCR) was developed and used for the greenhouse root samples.
Detection of F. verticillioides on root tissue. Details of the procedure performed to validate primers and quantify fungal colonization in roots will be submitted to a peer-review journal for publication. For this reason, a general overview of the technique is provided here. Three sets of primers were designed based on previous publications that used qRT-PCR for F. vertillioides detection in maize tissue (Mulè et al., 2004; Nicolaisen et al., 2009; Kurtz et al., 2010; Faria et al., 2011). The primers were validated by regular PCR of targeted genes on F.verticilliodes DNA, and cloning and sequencing of the PCR products , used to confirm the amplification of the targeted genes. PCR on fungal DNA, fungal-free plant DNA and on plant DNA from control plants (non-inoculated plants) allowed to confirm specificity of the primers. The technique was suitable for quantification of 0.01 – 100 ng*µL og F.verticillioides in plant DNA samples. For the quantification of PI, 1-3 g (fresh weight) root samples, collected in the greenhouse, were flash-frozen in liquid N2 at harvest and preserved at -80 °C. Root DNA extractions were performed with a commercial DNA extraction kit, Plant Maxi Kit (QIAGEN, Hilden, Germany). qRT-PCR were performed directly on the extracted DNA samples with a Fast Start Universal SYBR Green Master Mix (Roche Applied Science, USA).
RCA and RR were measured in 30 hybrid maize lines, collected in three farms across the state of Pennsylvania in 2014 (Table 1, see Figure 1). RCA Values ranged between 0 – 30% for all the sites (Figure 3), with significant differences between hybrid entry and site. In overall, the RCA variation found here is similar to previously reported values for inbred maize populations (Burton et al., 2013). Among sites, Rocksprings showed the lowest RCA levels, with a total average of 10%, while England and Kulp had averages close to 20% RCA. There were significant differences of RCA among hybrid entries, but not significant differences were observed due to the companies producing the seeds. These results indicate plasticity in RCA due to environmental factors, represented here by the different farms. Also, finding variation in RCA among entries, indicates potential to target RCA as a trait to be targeted in plant breeding programs, because it shows enough variation to select plants that have higher RCA and that could potentially perform better under stress conditions, like drought and low fertilizer availability.
RR presented significant differences among sites and maize entries, ranging 0-100%, with the highest average values at Rocksprings (85%), and lowest at England and Kulp farms (50 % and 60 %, respectively. Figure 4). We observed more RR in the oldest whorls, with a decrease in average RR from the third to the second whorl of 25% for all the sites. The second whorl is the oldest, and new, and more external whorls emerge along the season. The fact that the third whorl had lower RR levels could be interpreted as an increase resistance of the plant as it ages, with stronger, thicker root epidermis and cortex, and possible primed defenses later in the season. Plant defenses are activated by pathogens and other microorganisms when they first contact the plant tissue, and they can provide protection for extended periods of time under field conditions (Vallad and Goodman, 2004).
AMC was measured in 10 entries presenting contrasting levels of RCA on samples collected at England and Kulp. Despite the optimum fertilization regimes in these fields, specially phosphoric; which is expected to cause a decrease in AMC, we found wide variation in AMC, with values of 0-80%, and averages of 30% and 50% for Kulp and England, respectively (Figure 5). These results diverge with the general assumption that AMC is repressed under high P conditions and leave open questions regarding the utility of the symbiosis under optimum P fertilization levels. For example, what could be the ecological role of mycorrhizae in agricultural systems besides P uptake and translocation? Is there any other relevant function that is favoring the symbiosis even under optimum phosphoric fertilization in these agroecosystems? Important processes such as soil aggregate formation and plant defense increase have participation of mycorrhizal forming fungi (Rillig and Mummey, 2006; Pozo and Azcon-Aguilar, 2007), and could explain the prevalence of the symbiosis in high P agroecosystems like the farms Kulp and England.
Analysis of variance (ANOVAS) with AMC or RR as a function of the RCA phenotypes, and linear correlation analysis between RCA and AMC and RR in the field trials were performed and the results are presented in Figure 6 and Figure 7. We found not significant correlations in any case, indicating that the microbial colonization by mycorrhizal forming fungi and pathogens are not significantly affected by an augmented or reduced RCA under the field conditions of England and Kulp. This means that high RCA plants could be developed through plant breeding programs, with no effects of the root anatomical phenotypes on the microbial interactions with mycorrhizae or pathogens. High RCA maize plants would still develop AMC and RR at similar levels as low RCA in agroecosystems similar to Kulp, England, and Rocksprings.
The results of the greenhouse experiments, and the relation between RCA, AMC and RR, are depicted in Figure 8. Three anatomical phenotypes were identified post-harvest; high, intermediate and low RCA levels developed 5- 10 %, 3-6% and 0-2.5% RCA, respectively. No significant differences were found in AMC or PI, measured as fungal DNA per gram of root, among the RCA phenotypes. However, a trend was observed with of lower maximum values of AMC and PI in high RCA lines, and higher maximum values of AMC and PI in high RCA lines. Although these results are in accordance with those found in the field trials, with no significant effects of RCA on AMC or PI, we can see that RCA is limiting the maximum extend of microbial colonization in the roots for both, AMC and PI. Low RCA plants have the potentiality to develop higher AMC and PI, which indicates that having more cells in the cortex, not replaced by aerenchyma lacunae, serves as better niche for fungal colonization. However, this anatomical factor by itself is not enough to control or determine the extent of this colonization, and other plant or environmental factors could be playing significant roles that overlap the effects of RCA on the interactions of roots and microbes.
Other anatomical traits affecting AMC and RR in maize
In order to investigate other anatomical traits that could be influencing the microbial colonization in maize roots, we have conducted multivariate analysis on the datasets obtained after phenotyping the plants used in this and other studies. The goal is to find the most significant anatomical variables for microbial root colonization. The research results will be submitted to peer-reviewed journals for publication, and for this reason, we are not including detailed findings in this report. We have presented preliminary results in a conference presentation poster, attached to this report. As can be seen in the poster, 23 % of the variability in RR and AMC is explained by anatomical traits, with cortical cell file number (CCFN) affecting RR and AMC in opposite ways. More CCFN had more RR, while the same phenotype produced less AMC in the field. Interestingly, there is significant correlations between xylem-related root traits and RR, as presented in more detail in the poster. These results will be corroborated with the greenhouse experiments and further field experiments conducted in 2015.
Drought, and low nutrient uptake efficiency are major problems in agriculture. Farmers need plants with better roots that capture water, nitrogen and phosphorous more efficiently but that are not compromised for pathogen resistance or beneficial symbiotic associations. Root anatomical traits, and more specifically, RCA, has been identified by the Roots Lab, at Penn State, as a promising target for plant breeding programs to develop lines with better roots to cope with stress. However, the root microbial relationships, which are key components of the agro ecosystems in the context of sustainable agriculture, had not been studied in relation to the root anatomical traits to our knowledge. This research is informing plant breeders about how the selection of maize plants with augmented RCA do not have their microbial root associations with mycorrhizae, RR causing microbes, and F. verticillioides, compromised. We are retrieving an integral study that considers ecological aspects of root biology, promoting sustainable agriculture.
We are also contributing the description of root anatomy of commercial hybrid maize lines in the U.S., which contrast with previous reports that have focused on inbred populations and research material. With the results presented here, there are insights about what is the anatomical state of the current commercial lines used by farmers in Pennsylvania. The contrasting levels observed here, from 0 – 30% demonstrate the potential that the current maize breeding programs have to target promising root traits such as RCA. Besides the anatomical traits, we are also informing farmers, extension specialists and scientist about the RR and AMC states of these hybrids. With a simple scale, farmers could use the method employed here to assess the incidence of RR in their fields.
Besides the new knowledge we are generating regarding root biology, we are also informing and demonstrating the application of a very precise technique such as RT-qPCR, to detect F. verticillioides in root tissue. This technique, that has been used for research purposes in previous reports (Nicolaisen et al., 2009; Kurtz et al., 2010; Faria et al., 2011), could be included in further studies about the presence of this fungal species in corn tissue in field specimens. The main concern with F. verticilliodes infection relies on the production of fuminosins, mycotoxins that harm animal health and have been classified as ‘possibly carcinogenic to humans’ (IARC, 1993 as in Scott (2012)). For this reason, keeping control over the fungal population in maize fields is a key aspect to guarantee food security. Environmental control agencies or universities could benefit of the application of this technique for diagnostic of this fungal species in commercial corn.
Last, this research will report a pioneer study of the variability in root rots and anatomical traits in commercial hybrids for the first time in the U.S. It will also help to understand the variation and possible root factors that make maize plants more likely to present root rots or colonization with F. verticillioides. Not only RCA but also other anatomical traits are being studied and their importance in microbial colonization will be reported in scientific papers within the next months. In summary with this proposal, an integrative, ecological perspective is included in our attempts to provide recommendations for plant breeding in maize for a more sustainable agriculture.
Education & Outreach Activities and Participation Summary
The field results of this study were presented at the 69th Northeastern Corn Improvement Conference, at the Penn Stater Hotel, University Park, PA, on February 12, 2015. The talk “Roots and microbes in hybrid corn planted in Pennsylvania” was presented to an audience of approximately 80 people comprising seed companies’ representatives, professors, extension specialists, researchers and students. The abstract of the talk is attached to this report, and it is part of the event proceedings. Further data analysis was also presented in a poster at the American Society of Plant Biologist meeting, on July 27, 2015, at the Minneapolis Convention Center, Minneapolis, MN. The poster abstract is on the abstract book available online (http://my.aspb.org/?page=ME_Index) and is attached to this report as well. Additionally, this project was broadcasted at the Roots Lab web page once the funding was notified in 2013 (http://plantscience.psu.edu/research/labs/roots/news/2013/northeast-sare-grant-awarded-for-root-microbe-work). This website received around 30,000 total visits in the last year, from 25,000 unique individuals. Final results will be submitted for publication in peer-review journals within the next year, and further dissemination of these results will continue through agronomy, crop science and plant biology regional and national meetings such as the 70th Northeastern Corn Improvement Conference, and the American Society of Plant Biology meeting.
Areas needing additional study
As noted before, additional analysis with other anatomical traits, besides RCA, and their effect on RR and AMC will complement the results obtained here. Such analyses are being conducted and will be part of the PhD dissertation research of the coordinator of the project, Tania Galindo. Root anatomical traits like epidermis thickness, number of cortical cells, cortical cell file number, cortical living area, and root diameter, are intuitively important in the process of microbial root colonization, and are being included in the analysis.
A more comprehensive study of the process of root colonization by microbes per root class, root depth, and over time, would help understanding the physiological and anatomical traits that play main roles in the microbial colonization. This knowledge is surprisingly absent in scientific literature, where authors have focused either on the plant, or the fungal colonization independently, and have rarely considered the spatial and temporal effect of the root as a niche for microbial colonization belowground.
- Brundrett, M., 1991. Mycorrhizas in Natural Ecosystems. Adv Ecol Res 21, 171-313.
- Burton, A., Lynch, J., Brown, K., 2013. Spatial distribution and phenotypic variation in root cortical aerenchyma of maize (Zea mays L.). Plant and Soil 367, 263–274.
- Faria, C.B., Abe, C.A.L., Silva, C.N.d., Tessmann, D.J., Barbosa-Tessmann, I.P., 2011. New PCR Assays for the Identification of Fusarium verticillioides, Fusarium subglutinans, and Other Species of the Gibberella fujikuroi Complex. International journal of molecular sciences 13, 115-132.
- Kurtz, B., Karlovsky, P., Vidal, S., 2010. Interaction between western corn rootworm (Coleoptera: Chrysomelidae) larvae and root-infecting Fusarium verticillioides. Environmental entomology 39, 1532-1538.
- Lynch, J.P., 2013. Steep, cheap and deep: an ideotype to optimize water and N acquisition by maize root systems. Ann Bot-London.
- Mulè, G., Susca, A., Stea, G., Moretti, A., 2004. A species-specific PCR assay based on the calmodulin partial gene for identification of Fusarium verticillioides, F. proliferatum and F. subglutinans. European Journal of Plant Pathology 110, 495-502.
- Nicolaisen, M., Supronien?, S., Nielsen, L.K., Lazzaro, I., Spliid, N.H., Justesen, A.F., 2009. Real-time PCR for quantification of eleven individual Fusarium species in cereals. Journal of Microbiological Methods 76, 234-240.
- Postma, J.A., Lynch, J.P., 2011a. Root Cortical Aerenchyma Enhances the Growth of Maize on Soils with Suboptimal Availability of Nitrogen, Phosphorus, and Potassium. Plant Physiol 156, 1190-1201.
- Postma, J.A., Lynch, J.P., 2011b. Theoretical evidence for the functional benefit of root cortical aerenchyma in soils with low phosphorus availability. Ann Bot-London 107, 829-841.
- Pozo, M.J., Azcon-Aguilar, C., 2007. Unraveling mycorrhiza-induced resistance. Current opinion in plant biology 10, 393-398.
- Rillig, M.C., Mummey, D.L., 2006. Mycorrhizas and soil structure. New Phytologist 171, 41-53.
- Scott, P.M., 2012. Recent research on fumonisins: a review. Food Additives & Contaminants. Part A: Chemistry, Analysis, Control, Exposure & Risk Assessment 29, 242-248.
- Strzepek, K., Yohe, G., Neumann, J., Boehlert, B., 2010. Characterizing changes in drought risk for the United States from climate change. Environ Res Lett 5.
- Vallad, G.E., Goodman, R.M., 2004. Systemic acquired resistance and induced systemic resistance in conventional agriculture. Crop Science 44, 1920-1934.
- Vierheilig, H., Coughlan, A.P., Wyss, U., Piche, Y., 1998. Ink and vinegar, a simple staining technique for arbuscular-mycorrhizal fungi. Appl Environ Microb 64, 5004-5007.
- Zhu, J.M., Brown, K.M., Lynch, J.P., 2010. Root cortical aerenchyma improves the drought tolerance of maize (Zea mays L.). Plant Cell Environ 33, 740-749.