Evaluation of the biocontrol potential of the fungus Botrytis euroamericana strain HZ-011 for herbicidal activity

Abstract

Background

Weeds in farmland seriously threaten crop yield and cause huge economic losses. Due to the extensive use of chemical herbicides, a series of problems have been arisen, such as environmental pollution, soil degradation and pesticide residues. To assess the herbicidal activity, crop safety, taxonomic identity and infection process of strain HZ-011, the methods of inoculation on detached leaves in vitro and pot plants in vivo, as well as scanning electron microscopy, were used in this study.

Results

The results indicated that strain HZ-011 had pathogenic effects on detached leaves of four weeds, including Amaranthus retroflexus Linn., Elsholtzia densa Benth., Malva crispa Linn. and Chenopodium album Linn. in vitro. Strain HZ-011 also showed high pathogenicity to C. album and A. retroflexus in vivo, in which the pathogenicity rates were 100%, meaning all plants died after 7 days in the pot test, while the pathogenicity rates for E. splendens and M. crispa were 60.00 and 29.60%, respectively. This strain was safe for local crops, including Vicia faba Linn., Pisum sativum Linn., Brassica napus Linn., Hordeum vulgare Linn. and Triticum aestivum Linn. Strain HZ-011 was identified as the fungus Botrytis euroamericana based on its morphology, molecular biology and a constructed phylogenetic tree. The infection process of B. euroamericana HZ-011 in C. album was studied by plant histopathological observations after pathogen infection. This procedure showed that the mycelium of strain HZ-011 invaded C. album tissues from the stomata, infected and propagated within the tissues, and the spores produced further damage in the C. album tissues and lesions occurred on the surfaces of C. album leaves.

Conclusion

These tests provide a basis for fungus B. euroamericana HZ-011 as a potential microbial herbicide.

Background

As one of major problems in the agricultural ecosystem, weed control is an important part of agricultural production (Zhu et al. 2020b). There are 67 species of farmland weeds in 25 families in Qinghai Province, China. The broadleaf weeds, specifically Chenopodium album and Elsholtzia densa, are the major dominant species constituting weed communities in conservation tillage fields throughout this region, where these weeds have shown strong negative effects on the growth and yield of crops, and weed control is becoming more and more challenging (Wei et al. 2013). The control of weeds throughout the world mostly depends on the use of chemical agents, but the lasting and extensive use of chemical agents are also the main contributors to many ecological and environmental problems against farmlands (Sardana et al. 2017), such as the occurrence of herbicide-resistant weed plants, soil pollution, water degradation and damage to non-target microorganisms, etc. (Mejri et al. 2013).

With the increasing emergence of safety problems associated with agricultural products and the new requirements of modern agricultural development, microbial herbicides have been a focus of relevant scientific research in major countries throughout the world for nearly 60 years. Many strains with biocontrol and herbicidal activity have been isolated and screened from a large number of diseased weeds (Bailey et al. 2009). A recent survey found that a total of 468 biological control agent species have been used to control 175 species of target weeds in 48 plant families in 90 countries, and across all countries and regions, a total of 65.7% of the weeds were targeted for biological experienced control to some extent (Schwarzländer et al. 2018). Current microorganisms with herbicidal activity that have been already screened represented 40 genera and included approximately 80 species of active microorganisms, which can be used for controlling more than 70 species of weeds, such as Exserohilum, Bipolaris, Phytophthora, Fusarium, Alternaria, Colletotrichum, Cercospora, Curvularia, Colleototrichum and Entyloma (Liu et al. 2013). Fungal phytotoxins are natural secondary metabolites produced by plant pathogenic fungi during host–pathogen interactions, and they are novel and environmentally friendly herbicides (Maurizio et al. 2017). Microbial herbicides have become a research hotspot in the field of biological weed control, and relevant research achievements have also been constantly emerging (James et al. 2018). With the continuous isolation of and research on weed pathogens, the research and development of microbial herbicides, especially fungal herbicides, has been markedly strengthened (Schwarzländer et al. 2018).

The Key Laboratory of Agricultural Integrated Pest Management in Qinghai Province has carried out research and development of biocontrol herbicides for over ten years, and a total of 26 species of fungi have been obtained from more than 40 species of diseased weed plants and used for the biocontrol of various weeds (Zhu et al. 2020a). In the present study, a fungal strain, which was isolated from naturally infected Rumex patientia leaves, was assessed using in vitro and in vivo inoculation to determine its herbicidal activity toward A. lividus, E. densa, M. crispa and C. album. Potted plant bioassays, used to evaluate the safety to major crops in Qinghai Province; morphological and DNA sequencing analyses, were used to identify the strain, and scanning electron microscopy was used to observe the infection process of inoculated strain HZ-011 in C. album at 1–7 days. All of these studies provided an important basis for the further research, development and utilization of new and efficient microbial herbicides.

Methods

Phytopathogenic fungus, weeds and crops

Strain HZ-011 was isolated from naturally infected R. patientia leaves and stored in the Laboratory Key Laboratory of Agricultural Integrated Pest Management in Qinghai Province. Common broadleaf weeds from farmlands in Qinghai Province tested were Amaranthus retroflexus, E. densa, Malva crispa and C. album, and the major crops in Qinghai Province included V. faba, P. sativum, B. napus, H. vulgare and T. aestivum.

Pathogenicity of strain HZ-011 to weed leaves in vitro

Leaves from typical weeds (E. densa, M. crispa, C. album and A. lividus) were collected from farmlands, and their surfaces were rinsed in the laboratory. Then, they were placed in medium plates (Φ = 90 mm) lined with filter paper (3–4 pieces for each culture dish). The filter paper was made wet with sterile water to create a humid environment; a purified strain plate was made with bacterial cakes (Φ = 8 mm) using a hole punch and inoculated on each leaf surface (repeated three times), with sterile PDA used as the control. The lesion area was measured after 7 days (Zhang 2005) and is expressed as: lesion area = 1/4 × length × width × 3.14.

Pathogenicity to weeds

The weeds E. densa, M. crispa, C. album and A. retroflexus in the 4–5 leaves stage and normal growth conditions were transplanted into the pots (Φ = 15 cm) and cultured at room temperature (25 ± 1 °C) for 1 week. The strain was cultured in PDA medium plates for 7 days, and then, fungal blocks were taken from the edge of the growing strain (Φ = 8 mm), inoculated into 250 ml/bottle PDB medium liquid (5 for each bottle) and cultured at 25 °C and 180 r·min⁻¹ for 7 days. The inoculated liquids were filtered with sterile gauze and diluted until the concentration of spore suspension reached 1.0 × 10⁵ CFU/ml to 1.0 × 10⁷ CFU/ml. The spore suspension was placed into a 500-ml spraying bottle that had been disinfected with 75% alcohol, and 2 ml of Tween-20 was added. Then, the suspension was inoculated onto transplanted healthy weed plants by spraying at 25 ml/pot, and plants inoculated with PDB medium liquid were used as the blank control. Each treatment was repeated three times. The experiment was repeated in 2020 and 2021. Inoculated weed plants were placed at room temperature, maintained at a relative humidity of 60% by a JBX-3.5 centrifugal industrial humidifier and cultured under alternating light and darkness for 12 h.

Determination of disease incidence

The incidence of inoculated weeds was observed, and the severity and grade distributions of weeds were investigated after 7 days, with the disease incidence and disease indices calculated, using the following criteria and formulas. The severity grade was assigned in accordance with the references (Li et al. 2014) below: grade 0: no lesions on the leaves; grade 1: a few lesions distributed on the leaves; grade 2: 1/3–2/3 of the leaf surface rotten to death; grade 3: 2/3 or more of the leaf surface rotten to death; and grade 4: the whole leaf surface rotten to death. Incidence = number of sick leaves/total number of leaves investigated × 100%; disease index = (number of sick leaves × number of corresponding grades)/ (total number of leaves investigated × number of leaves in high grade) ×  100%.

Safety to crops

Potential host plants: V. faba, P. sativum, B. napus, H. vulgare and T. aestivum, were cultivated in Φ = 15 cm pots and cultured in the laboratory. The preparation and inoculation of fermented liquids were carried out using the same method described above for the pathogenicity of weeds.

Investigation of disease severity: The disease onset in potential host crops was observed at 7 days post-inoculation and recorded in accordance with the following criteria: NS for no symptoms (no lesions, plants in normal growth conditions); LS for low symptoms (few lesions distributed on leaves, growth slightly inhibited); MS for moderate symptoms (lesions on 1/5–1/4 of leaf surface, growth inhibited); and SS for severe symptoms (lesions on 1/4 or more of leaf surface, growth inhibited seriously).

Morphological identification

The isolated strains were cultured on PDA medium, and the morphology of hyphae and spores was observed under the optical microscope. The fungus with a diameter of 8 mm was inoculated on plates of different media (Φ = 90 mm) and cultured at 25 °C, with regular observations on the colony growth rate, colony morphology and color changes. The colonies were initially identified according to the Manual of Fungus Identification (Pei et al. 2021).

Identification at the molecular level

The genomic DNA was extracted according to the instructions of the fungal genomic DNA extraction kit, and the ITS region of the 5.8S rDNA was amplified, using the ITS universal primers ITS 1: 5'-TCCGTAGGTGAACCTGCGG-3' and ITS 4: 5'-TCCTCCGCTTATTGATGTGC-3'. The G3PDH region was amplified, using the G3PDH universal primers G3PDH: 5'-ATTGACATCGTCGCTGTCAACGA-3' and G3PDHrev: 5'-ACCCCACTCGTTGTCGTACCA-3'. The HSP60 region was amplified, using the HSP60 universal primers HSP60: 5'-CAACAATTGAGATTTGCCCACAAG-3' and HSP60rev: 5'-GATGGATCCAGTGGTACCGAGCAT-3'. The RPB2 region was amplified, using the RPB2 universal primers RPB2: 5'-GATGATCGTGATCATTTCGG-3' and RPB2rev: 5'-CCCATAGCTTGCTTACCCAT-3'. The purity was detected by 1% agar sugar gel electrophoresis, and the products of PCR amplification were sent to Shanghai Sangon Biotech Services Co., Ltd., for sequencing. The obtained ITS, G3PDH, HSP60 and RPB2 sequences were subjected to homology sequence analysis in the nucleic acid sequence library of GenBank, using BLAST, and sequences with various similarities were selected for phylogenetic analysis, using MEGA 6.0 software. The neighbor-joining (NJ) method was used to construct the phylogenetic tree to determine the classification status of strain HZ-011.

Electron microscopy for the pathogenicity of strain HZ-011 to C. album

In sterilized Petri dishes (Φ = 90 mm) lined with filter paper and weed leaves, the filter paper was made wet with sterile water to provide a humid environment, and a cake of fungus obtained with a punch (Φ = 8 mm) was placed in the center of the leaf, and the sterile PDA, a block of fungus was used as the control. All of the treatments and control were repeated three times. The experiment was conducted at 25–28 °C. After inoculation, the samples were taken every day for 7 days to observe the invasion process of the mycelia.

Statistical analyses

Statistical analysis was carried out using Excel 2010 and DPS 9.01, and in single-factor statistical analysis, Duncan's new complex difference method was used for analysis of variance, and P value was used to describe the significance of the data.

Results

Pathogenicity of strain HZ-011 to weed leaves in vitro

The pathogenicity of strain HZ-011 to detached leaves is shown in Fig. 1 and Table 1. The target weeds were A. retroflexus, E. densa, M. crispa and C. album. The results indicated that HZ-011 had pathogenic effect on all the four weeds. The leaves of A. retroflexus and E. densa were heavily infected by the strain HZ-011, and the lesion areas were 3.567 cm² and 2.808 cm², respectively. The leaves of C. album and M. crispa were moderately infected, and the lesion areas were 1.803 cm² and 0.858 cm². The study demonstrated that the strain HZ-011 could infect the leaves of weeds such as A. retroflexus, E. densa, M. crispa and C. album in vitro. This strain had herbicidal activity against four different weeds in different extents.

Fig. 1

Pathogenicity of strain HZ-011 to detached leaves of four different species of weeds. a Amaranthus retroflexus, b Elsholtzia densa, c Malva crispa, d Chenopodium album

Table 1 Comparison of pathogenicity of strain HZ-011 on weeds in vitro

Pathogenicity of weeds

The herbicidal activity of the strain HZ-011 during two consecutive years is shown in Fig. 2 and Table 2. In 2020, this strain had the significant (p < 0.05) control effects on four weeds (A. retroflexus, E. densa, M. crispa and C. album). After inoculating for 7 days, there was exceeded 80% (Table 2) that the disease incidences of A. retroflexus, E. densa and C. album. The weed C. album had the highest disease incidence reaching 96.67%, followed by the weed A. retroflexus with the disease incidence of 92.78%. The weed M. crispa had the least disease incidence of over 50%. The disease incidences of these weeds were significantly different (p < 0.05).

Fig. 2

Pathogenicity of strain HZ-011 to potted weeds. a Amaranthus retroflexus, b Elsholtzia densa, c Malva crispa, d Chenopodium album

Table 2 Disease incidence of the strain HZ-011 on different potted weeds in 2020 and 2021

In 2021 (Fig. 2), this tested fungal strain was inoculated on four weeds such as A. retroflexus, C. album, M. crispa and E. densa. The results showed that the disease incidence of A. retroflexus and C. album reached 100%. But, the disease incidence of E. densa was 60.00%, while that of M. crispa was only 29.60%. Compared to 2020, the disease incidence decreased and indicated that the herbicidal activity of this strain on E. densa and M. crispa was unstable.

The herbicidal activity tested of the strain HZ-011 in 2020 and 2021 demonstrated that compared with M. crispa, the other three weeds A. retroflexus, C. album and E. densa were more sensitive to the strain HZ-011. Therefore, the strain HZ-011 can be selected as a potential biocontrol agent with herbicidal activity against the previous weeds.

Safety to crops

After crops were inoculated with the strain HZ-011, V. faba, P. sativum, B. napus, H. vulgare and T. aestivum indicated no pathogenicity, as shown in Fig. 3, compared to the control, the growth and plant height of crops were not affected and showed no symptoms according to visual observation, indicating that the strain HZ-011 had no pathogenicity to crops and was relatively safe. Similar trial results were obtained in 2020 and 2021.

Fig. 3

Safety of strain HZ-011 to crops. a Vicia faba; b Pisum sativum; c Brassica napus; d Hordeum vulgare; and e Triticum aestivum

Morphological identification

As shown in Fig. 4, the strain HZ-011 isolated from the leaves of naturally susceptible R. patientia grew rapidly on PDA culture medium. The surface of the colonies was white initially, and the aerial hypha was underdeveloped. After 5 days of growth, the center gradually turned dark yellow. The back surface was white initially and then changed to orange after 3 days. After 10 days of growth, the growth of granular matters was accompanied. Conidiophore was sporadic and colorless; the top of the cell was expanded into a sphere attached with many conidiophores. A single spore was oval, club-shaped or long elliptic and gathered like grape spike-stalk on conidiophores. Conidia reached 14.5–25.5 × 7.5–10.5 μm. Therefore, it was preliminarily identified as Botrytis.

Fig. 4

Micromorphology of strain HZ-011. a, d Culture on PDA; b, c, f conidiophores; e conidia; b, e bars = 2 μm; c, f bars = 10 μm

Molecular identification

PCR amplification was carried out for the DNA of the strain HZ-011, and 511 bp (ITS), 906 bp (G3PDH), 1030 bp (HSP60) and 1159 bp (RPB2) were selected for comparison after sequencing. The top 12 fungi with the similarity > 95% to ITS, G3PDH, HSP60 and RPB2 gene sequences of the strain HZ-011 were selected for the phylogenetic tree analysis by MEGA 6.0 software. In this test, the neighbor-joining method was used to construct the phylogenetic tree, as shown in Fig. 5. The results showed that the strain HZ-011 had the same evolutionary branch as Botrytis euroamericana (KC191680.1/KX266728.1/K266734.1/K366740.1), which proved strain HZ-011 was B. euroamericana. The pathogen fungus was identified as B. euroamericana by the amplification of the ITS, G3PDH, HSP60 and RPB2 genes and by phylogenetic analysis.

Fig. 5

Phylogenetic tree of B. euroamericana strain HZ-011 and Botrytis by neighbor-joining method. Note 0.01 represents the genetic distance between two nucleosides; only bootstrap value of > 40% is displayed in clades

Electron microscopy for the pathogenicity of B. euroamericana strain HZ-011 to C. album

Observed from scanning electron microscope, the tissue surface of controlled C. album was intact (Fig. 6-a). At 1 day after C. album was inoculated with the B. euroamericana strain HZ-011, hypha penetrated through the stoma, and the tissue was not damaged (Fig. 6-b). After 2 d, multiple hyphae were distributed around the stoma of C. album tissue, and a single spore was attached on the tissue surface (Fig. 6-c). After 3 days, as hyphae and spores parasitized on the surface of the tissue, they could be able to grow on the nutrients in the tissue, forming the fungal network on the surface of the tissue (Fig. 6-d). After 4–5 days, the tissue surface was damaged, the hyphae-parasitized tissue absorbed nutrients, and the plants had obvious symptoms (Fig. 6-e and f). After 6 days, the tissue was completely wrapped by hyphae growing vigorously (Fig. 6-g). After 7 days, the conidiophore produced spores, forming dense spores on the surface of the damaged tissue (Fig. 6-h).

Fig. 6

Scanning electron microscope observation of the characteristics of strain HZ-011 infecting Chenopodium album. Scale bars: (a, b, c, d, e, f) = 10 µm × 500, (g, h) = 10 µm × 1000

Discussion

More than 20 microbial-based herbicidal products have been successfully developed in the world (Schwarzländer et al. 2018), including COLLEGO ™, DeVine, Chontrol ™, BioMal ™, Velgo ™, Smolder ™, Biochon, Coltru ™, StumpOut ™, Casst ™, Sarritor, Myco-Tech Paste ™, Camperico, SolviNix^® LC, Fiesta^® and Bialaphos (Dayan et al. 2010), among which Biochon, Bialaphos, Camperico, Myco-TechPaste and Stump Out have been listed among international products (Duke 2014). These microbial herbicides are derived from fungi, bacteria, viruses and natural products. Although the proportion of microbial herbicides in the current weed control is still relatively low, the development and utilization of microorganisms and their metabolites to develop herbicides has become the focus of industrial development in this field (Zhuang 2014).

In the present study, the pathogenic strain HZ-011 was separated from the leaves naturally infected by R. patientia. This strain was identified as B. euroamericana by the morphology, gene sequences and GenBank’s nucleic acid sequence library combined with the development of phylogenetic tree. Botrytis fungus usually infects the plants in the form of pathogens, so fungus-based herbicide has the advantages including easy to be isolated and discovered with high diverse species and strong host specificity. It has been reported that B. cinerea has been used in biocontrol of weeds. Wang et al. (2005a, b) reported that compounds with strong herbicidal activity, such as abscisic acid, abscisic acid glucoside, abscisic acid glucoside ester, 10-cis-dihydride cinereus dialdehyde and dimethyl phthalate, were separated and identified from the metabolites of B. cinerea and others (Wang et al. 2005a, b). Li (2003) reported that B. cinerea was separated from the samples of B. cinerea in different regions producing metabolites having strong inhibitory effects on the leaves and seedlings of Amaranthus retroflexus, Pharbitis nil, Convolvulus arvensis, barnyard grass, Digitaria sanguinalis, Eleusine indica, Portulaca oleracea, Setaria viridis and C. album and had a wide spectrum of weed control. Zhang (2005) found that when the concentration of fermented filtrate of strain B. cinerea BC1 was 150 ml/l, it had significant inhibitory effects on the growth of potted A. retroflexus, and the effect of soil treatment was 83.4%, better than those of stem and leaf treatments, and the effect on S. viridis was less than 60%. Moparthi et al. (2020) firstly reported a gray mold of chickpea caused by B. euroamericana in the USA, and the pathogen could cause lesions on chickpea and lentil foliage. In this study, B. euroamericana strain HZ-011 had significant pathogenic effects on the detached leaves of common weeds C. album, E. densa, A. retroflexus and M. crispa. This strain has a wide spectrum of weed control, which provides a practical basis for further application in field biocontrol of weeds.

It is not enough for herbicide strains to show herbicidal activity in laboratory tests. Only biocontrol strains that still show good herbicidal effect in field applications have potential as biocontrol agents (Yao et al. 2021). In this study, pathogenicity in vivo indicated that B. euroamericana strain HZ-011 had 100% pathogenicity on potted weeds belonging to C. album and A. retroflexus, 60.00% on E. densa and 29.60% on M. crispa. The safety evaluation of potted crops including V. faba, P. sativum, B. napus, H. vulgare and T. aestivum showed no pathogenic effects and relatively safety to local crops in the potted environment.

Observed from the scanning electron microscope, the hyphae of B. euroamericana strain HZ-011 invaded the C. album tissue through the stoma, parasitized, produced spores in the tissue and gradually destroyed the tissue. At the same time, the leaf surface began to turn from green to yellow, accompanied by the formation of lesions, which eventually led to the death of the plants due to the failure of photosynthesis. Gao et al. (2021) reported that purine metabolism, secondary metabolite synthesis, amino acid metabolism and changes to carbohydrate metabolism in leaves, after Aureobasidium pullulans PA-2 infected C. album, may be the main mechanism of pathogenicity of C. album by PA-2 infection. The molecular mechanism of strain HZ-011 infecting C. album will be further investigated.

In conclusion, B. euroamericana strain HZ-011 had significant pathogenic effects on the weeds C. album, E. densa, A. retroflexus and M. crispa and was safe to the local crops in Qinghai Province, which is a potential weed biocontrol fungus with a broad spectrum of weed control. Scanning electron microscope observation indicated that hyphae of B. euroamericana strain HZ-011 invaded the C. album tissue to produce spores and destroyed the C. album tissue to death.

Conclusions

This research can lead to develop B. euroamericana strain HZ-011 as a microbial herbicide. However, other experiments have to be conducted with regard to field tests and assessment of the impact of comprehensive factors such as temperature, light and humidity on biocontrol agent. A large number of experiments are needed to explore its value in future.