Native root-associated bacteria rescue a plant from a sudden-wilt disease that emerged during continuous cropping

Abstract

Plants maintain microbial associations whose functions remain largely unknown. For the past 15 y, we have planted the annual postfire tobacco Nicotiana attenuata into an experimental field plot in the plant's native habitat, and for the last 8 y the number of plants dying from a sudden wilt disease has increased, leading to crop failure. Inadvertently we had recapitulated the common agricultural dilemma of pathogen buildup associated with continuous cropping for this native plant. Plants suffered sudden tissue collapse and black roots, symptoms similar to a Fusarium-Alternaria disease complex, recently characterized in a nearby native population and developed into an in vitro pathosystem for N. attenuata. With this in vitro disease system, different protection strategies (fungicide and inoculations with native root-associated bacterial and fungal isolates), together with a biochar soil amendment, were tested further in the field. A field trial with more than 900 plants in two field plots revealed that inoculation with a mixture of native bacterial isolates significantly reduced disease incidence and mortality in the infected field plot without influencing growth, herbivore resistance, or 32 defense and signaling metabolites known to mediate resistance against native herbivores. Tests in a subsequent year revealed that a core consortium of five bacteria was essential for disease reduction. This consortium, but not individual members of the root-associated bacteria community which this plant normally recruits during germination from native seed banks, provides enduring resistance against fungal diseases, demonstrating that native plants develop opportunistic mutualisms with prokaryotes that solve context-dependent ecological problems.

Keywords: Alternaria; Fusarium; Nicotiana attenuata; microbiome function; plant disease resistance.

Fig. 1.

Abundance of bacteria and fungi isolated from the roots of diseased N. attenuata plants. Abundance of culturable bacteria and fungi isolated from native field-grown plants exhibiting the sudden wilt disease symptoms. Potential plant pathogens are in red font. (A) Only two potential bacterial pathogens (C. flaccumfaciens and P. tremae) were found in the 70 members of the bacterial community retrieved from the roots of seven diseased plants. (B) In contrast, potential fungal pathogens (Alternaria and Fusarium) were abundant among the 36 culturable isolates of the fungal community from the roots of eight diseased plants. Isolates, which were found in two or more or four or more plants are indicated by (°) and (°°), respectively. Bacterial genus acronyms: B, Bacillus; Br, Brevibacterium; Bu, Budvicia; C, Chryseobacterium; Ci, Citrobacter; Cu, Curtobacterium; E, Enterobacter; Er, Erwinia; L, Leifsonia; M, Microbacterium; P, Pseudomonas; Pa, Pantoea; Pae, Paenibacillus; R, Rhizobium. Fungal genus acronyms: A, Alternaria; Ap, Aspergillus; F, Fusarium; H, Hypocrea. (C) Symptoms of the sudden wilt disease in field-grown N. attenuata plants included black coloration of the roots. For details see Fig. S1.

Fig. S1.

Symptoms of the sudden wilt disease. (A) Disease symptoms first occurred only sporadically in our field plot in elongated and flowering plants and later were also observable in rosette-stage plants. Regular field experiments on the Old plot were ended in 2012 because of the unacceptably high plant mortality. Total plant mortality was recorded during the last three field seasons. (B) Sudden wilt disease symptoms characterized by dry or flaccid leaves developed within 1 or 2 d in N. attenuata plants. (C) The wilting was specific to N. attenuata; surrounding plants were not affected, and the surrounding soil usually was still moist. (D–F) The signature characteristic of a diseased plant was the development of black roots; the discoloration was visible on the outside as well as in longitudinal sections. The occurrence of wilting together with the black roots was diagnostic of a plant being affected by the sudden wilt disease. (G) Plants with differently pronounced disease levels were observed during the 2014 field season and illustrate the course of the disease. Wilting first was observed only in elongated plants (here with mild symptoms and mainly white roots) but during the last three field seasons appeared in younger plants also (here with marked symptoms and completely black roots).

Fig. S2.

Inoculation with a mixture of native bacteria, a fungicide, and two native fungal isolates reduced seedling mortality under in vitro conditions. (A) Schematic of the in vitro experimental setup. Seedling mortality was observed in separate infection assays using two fungal pathogens that previously had been isolated from diseased N. attenuata plants in a native population and characterized: Fusarium sp. U3 and Alternaria sp. U10 (3). (B and C). Evaluation of six native bacterial isolates for potential biocontrol abilities. The seeds were inoculated with individual cultures of P. azotoformans A70, P. frederiksbergensis A176, B. megaterium B55, B. cereus CN2, A. nitroguajacolicus E46, or B. mojavensis K1 or with a mixture of all strains (SI Materials and Methods). The mixed inoculation of all six strains had the strongest effects against Fusarium sp. U3 and Alternaria sp. U10 and was selected as the treatment for the field experiments. (D and E) The fungicide seed treatment (Landor; Syngenta) significantly reduced seedling mortality of N. attenuata seedlings infected with fungal pathogens. (F and G) Native fungal isolates were tested for possible biocontrol abilities. The seeds were inoculated with Chaetomium sp. (C16, C39, and C72) or Oidodendron sp. (Oi3) before being infected with fungal pathogens. The C72 and Oi3 treatments were chosen for field experiments because they reduced the mortality of seedlings inoculated with Fusarium sp. U3. Bars represent mean seedling mortality (± SEM, n = 10 plates); the different letters above the bars indicate significant differences in a one-way ANOVA with Fisher’s protected least significant difference (PLSD) test; P < 0.05.

Fig. 2.

Workflow of the three main strategies and the treatments used for 2013 field experiment. Of the three main strategies pursued to curb the spread of the disease in the field, the inoculation with microbes (bacteria or fungi) and fungicide treatment were first evaluated under in vitro conditions in the laboratory (Fig. S2). The mixed inoculation with six bacterial isolates, two fungal isolates, and the treatment with a commercially available fungicide in vitro reduced the mortality of N. attenuata seedlings infected with native isolates of fungal pathogens (Fusarium sp. and Alternaria sp.), and these treatments were selected for the field experiments. The combination of the strategies resulted in seven different treatments (including control treatment) that were deployed for the 2013 field experiment. All treatments were applied to N. attenuata seedlings before or during their planting into the field. The repeated fungicide treatment (fungicide 5×) was reapplied four times at 1-wk intervals after planting.

Fig. S3.

Planting of the treatment groups during the 2013 field season. (A) Before planting, charcoal was added to the charcoal and combined treatment groups. (B) Jiffy pots were treated before planting; the applied fungicide solution can be seen by the red color. (C and D) Treatment groups were planted in a randomized design on the lower section of the Old field plot (N 37.1463 W 114.0198), and, as a control experiment, also on the New plot (N 37.1412 W 114.0275). (E) Google Maps view of Old and New field plots ∼900 m apart.

Fig. S4.

Overview of the Old field plot and the rate of change in plant mortality during the 2013 field season. (A) The Old field plot during the 2013 season. The rectangle indicates the lower area of the plot where plants had been planted. This portion of the plot was selected because the greatest number of diseased plants was found in this area during the 2012 growing season. (B) Schematic illustrating the spatial distribution of plants (from the rectangle shown in A), distinguishing plants with the sudden wilt symptoms (wilting and black roots) and plants with only wilting symptoms. The occurrence of the sudden wilt disease was not distributed equally throughout the plot. (C) Mortality rate of N. obtusifolia planted together with N. attenuata in an adjacent block of a separate experiment. (D) The development of plant mortality is shown as the rate of change (percentage dead plants at one observation minus dead plants at previous observation). The error bars reflect the differences among the treatments. Plants with only wilting symptoms were observed mainly at the early time points. (E) Percentage of dead plants at 34 dpp showing black roots or only wilting symptoms. Most plants with only wilting symptoms were found in the fungi, charcoal, and combined (charcoal + fungicide) treatments.

Fig. 3.

Efficiency of the different treatments in reducing the mortality of field-grown N. attenuata plants (2013 field season). Plants in the different treatment groups (fungi, charcoal, fungicide 5×, fungicide 1×, bacteria, and combined charcoal + fungicide) were planted together with control plants in a randomized design on the Old (diseased) field plot (see Materials and Methods). (A) Plant mortality at 15 dpp was significantly reduced in the bacterially treated group compared with the control plants (G test: P < 0.05, n = 105 plants per group). (B) The increase in plant mortality was observed every 3 or 4 d for a 22-d observation period. The plants receiving the bacterial treatment had the lowest overall mortality rate. For details of the spatial distribution of plants and the rate of change in mortality see Fig. S4.

Fig. 4.

Growth parameters of plants in the different treatment groups in two field plots. N. attenuata plants from the different treatment groups (bacteria, charcoal, fungicide 1×, fungicide 5×, fungi, and combined treatment with charcoal + fungicide) were planted together with control plants in 2013 into two field plots (Old and New), and their growth parameters (rosette diameter and stalk height) were quantified. (A and B) Mean rosette diameter and stalk height of the different treatment groups compared with control plants (dotted line) grown in the Old (diseased) plot (± SEM; n = 105 plants per group). (C and D) Mean rosette diameter and stalk height of plants from the different treatment groups compared with control plants (dotted line) grown in the New plot (± SEM; n ≥ 29 plants per group). A comprehensive characterization of 32 traits known to be important for insect resistance and general ecological performance, including hormone levels and defense parameters (Table S1), was conducted on plants grown in the New plot to evaluate the effect of bacterial inoculation on traits not directly related to fungal pathogen resistance.

Fig. 5.

Reproducibility of the disease-suppression effect of bacterial consortia in the 2014 field season. Based on the results of in vitro tests (Fig. S5) we parsed the bacterial consortia into two groups of two (B55 + A70) or three (K1 + A176+ E46) bacteria and compared these groups with the mixture of the five isolates (K1 + A176 + E46 + B55 + A70) in protecting inoculated seedlings from the sudden wilt when planted into the Old plot. Preinoculated plants were planted together with control plants in a randomized design on the Old (diseased) plot (2014 field season; see Materials and Methods). Inoculation with three bacteria (K1 + A176 + E46) or five bacteria (K1 + A176 + E46 + B55 + A70) reduced plant mortality by 36% and 52%, respectively, compared with control plants at 40 dpp (n = 45 plants per group). The inoculation with two bacterial strains (B55 + A70) had no significant effect in reducing the rate of death compared with noninoculated control plants.

Fig. S5.

In comparison with the effects of seedling inoculation with a mixed bacterial consortium consisting of five taxa, the individual absences of three bacterial isolates (K1, E46, and A176) significantly increased seedling mortality under in vitro conditions. Different mixed bacterial consortia were evaluated for potential biocontrol abilities against Alternaria sp. U10. Mixed bacterial consortia lacking one bacterial isolate (e.g., mix minus K1, mix minus E46, and mix minus A176) significantly increased seedling morality. Based on these results, the three-bacteria mixed consortium (K1 + A176 + E46) and two-bacteria mixed consortium (B55 + A70), along with the mixture of all five bacteria, were selected as treatments for the 2014 field experiments. Bars represent mean seedling mortality (± SE; n = 7–10); the different letters above the bars indicate significant differences in a one-way ANOVA with Fisher’s PLSD test; P < 0.05. For the experimental set-up see Fig. S2A.

Fig. S6.

Reisolation of the bacteria from healthy field-grown plants at the flowering stage demonstrated the persistence of the inoculated bacterial taxa. In 2013, healthy plants from the control and bacterial treatment groups were harvested at the early flowering stage from both field plots, and the culturable bacterial consortium were isolated. (A and B) Four of the six native bacterial taxa used in the bacterial mix (seed and Jiffy treatment) persisted throughout growth under field conditions and were reisolated from the roots of plants from both field plots. The inoculated roots showed strong colonization by B. cereus CN2 and B. mojavensis K1. The bars represent the mean number of isolates (± SEM; Old field plot: n = 7 roots, 70 isolates; New field plot: n = 3 roots, 30 isolates). (C and D) The control plants showed natural colonization by P. frederiksbergensis, which was also used in the bacterial mixture as A176. Bacterial genus acronyms: A, Acinetobacter; B, Bacillus; Br, Brevibacterium; E, Escherichia; P, Pseudomonas; S, Stenotrophomonas.

Fig. S7.

Reproducibility of the reisolation of the mixed bacterial consortium from Utah and Arizona genotypes in the 2014 field season and the protection effect of the consortium for the Arizona ecotype. Bacterially treated Utah and Arizona genotypes were harvested at the early flowering stage from the New field plot, and the culturable bacterial consortium were isolated. (A and B) All five native bacterial taxa used in the bacterial inoculation treatment (seed treatment only) persisted throughout growth under field conditions of both ecotypes. The persistence of the inoculated bacteria taxa within both ecotypes demonstrates the consistency of the bacterial association with N. attenuata. The bars represent the mean number of isolates (± SEM; n = 5 roots, 50 isolates). Bacterial genus acronyms: A, Achromobacter; B, Bacillus; C, Ciceribacter; E, Enterobacter; P, Pseudomonas; Pa, Pantoea. (C) Under in vitro conditions, the consortium of five mixed bacterial isolates significantly reduced the mortality of the N. attenuata Arizona ecotype inoculated with Alternaria sp. U10 (± SEM; n = 11, P < 0.001, t-test).