Metarhizium Associated with Coffee Seedling Roots: Positive Effects on Plant Growth and Protection against Leucoptera coffeella

Abstract

Metarhizium species can be mutualistic symbionts of plants. They are able to colonize roots, promote plant growth and provide protection against pests. We previously found Metarhizium robertsii and M. brunneum associated with coffee roots in a diversified coffee system. Here, we investigated whether these fungi, when inoculated in coffee seedlings, can associate with roots, improve seedling growth and indirectly protect against the coffee leaf miner (CLM) Leucoptera coffeella (Lepidoptera: Lyonetiidae). We performed a greenhouse experiment with coffee seedlings using suspensions of each Metarhizium species applied as soil drenches to potted seedlings. We also challenged these plants with CLM infestation (two adult couples per plant). We recovered Metarhizium spp. from most of the seedling roots 43 days after fungal inoculation. Plants inoculated with M. robertsii showed a 30% leaf area increase compared to the control. Both isolates promoted protection against CLM in coffee seedlings, reducing the percentual of leaf area mined and prolonging CLM development time by two days versus controls. Besides this protection provided by Metarhizium, M. robertsii also improves seedling growth. Therefore, these Metarhizium species could be considered for the development of inoculants for coffee seedlings.

1. Introduction

The fungal genus Metarhizium (Hypocreales: Clavicipitaceae) has a multifunctional lifestyle. It can acquire nutrients by infecting insects, being considered an insect pathogen [1], and by colonizing plants to form a mutualistic symbiosis [2,3,4,5,6,7]. As insect-pathogenic fungi, Metarhizium species can infect and kill all life stages of economically important arthropod pests, such as termites, locusts, grasshoppers, cockroaches, whiteflies, thrips, mosquitoes and ticks [8,9], which have made them important tools for biological control [10,11]. The use of Metarhizium as a biological control agent is mainly in inundative control [8] and there are a number of commercial products based on Metarhizium registered around the world [9,12,13].

Metarhizium is also a symbiont of agriculturally and economically important plants, such as tomato, bean, corn, wheat and soybean [7]. When Metarhizium colonizes plants, it can confer benefits such as growth promotion [14,15,16], nutrient transfer [3] and protection against insect pests [15,16] and diseases [17,18]. In turn, the host plant provides the fungus with photosynthetic compounds [19]. This association of Metarhizium plus plant can be endophytic [3] and through rhizosphere competence [20]. Endophytic fungi develop within plant tissues without causing any noticeable symptoms of disease in the plant [21,22]. Meanwhile, rhizosphere-competent fungi grow in the rhizosphere without colonizing plant tissues [23]. Metarhizium species can associate with roots [16,24], stems and leaves [16,25]. However, they are most commonly found inhabiting roots [16,26,27,28].

The ability of Metarhizium to associate with plants can be considered in the context of it being phylogenetically closer to the grass endophytes Claviceps and Epichloë than to animal-pathogenic Hypocreales [29]. These authors also showed that Metarhizium harbors genes that codify plant-degrading enzymes in its genomes, indicating that Metarhizium may have evolved from a plant symbiont lifestyle, as was initially proposed by Spatafora et al. [30]. Our current understanding is that some species in the genus have maintained their role as plant symbionts but have more recently subsequently acquired the ability to infect and kill insects [31,32].

Colonization of roots by fungi increases the surface area from which plants can scavenge nutrients, facilitating the absorption of soil nutrients, which results in increased photosynthetic ability and enhanced plant growth [33]. The protection against herbivores can arise from the production of fungal secondary metabolites in plants. Metarhizium can produce destruxins [34], a secondary metabolite toxic to insects that was detected in cowpea plants [35], potato [36], and tomato leaves [37]. These responses in planta also can be activated by abscisic acid (ABA): during the early stages of fungal infection in bean plants, ABA reduces immune responses in plants during endophytic colonization by Metarhizium robertsii and increases immune responses to pathogenic colonization by Fusarium solani [38]. These results suggest that ABA plays a central role in differential responses to endophytic colonization in plants. Studies indicate that protection in plants can be caused by the induction of plant resistance since fungi colonizing plants can at first be recognized as potential invaders, triggering immune responses such as transcription factors involved in resistance against herbivores [16,39].

Several studies have reported enhanced plant growth and indirect negative effects of plant-associated Metarhizium on herbivores feeding on above-ground plant parts. Canassa et al. [40] found that root colonization of Phaseolus vulgaris by M. robertsii can suppress spider mites feeding on leaves and improve plant growth. Metarhizium brunneum inoculated on the roots of sweet pepper Capsicum annum increased plant growth while negatively affecting life history parameters of the aphid Myzus persicae (prolonged development time, delayed onset of reproduction and reduced birth rate [15]). Maize seeds treated with M. brunneum, M. anisopliae, and M. robertsii increase leaf collar formation, stalk length, average ear biomass and average stalk, and foliage biomass [14].

The coffee leaf miner (CLM) Leucoptera coffeella (Lepidoptera: Lyonetiidae) is a pest that damages coffee plants. It reduces seedling quality through reductions in photosynthetic leaf area and defoliation [41,42]. This pest often attains high population levels in many coffee-producing regions in Brazil, causing defoliation by up to 70% and reducing coffee yields by 50% [43]. Thus, it is important to produce healthy coffee seedlings in order to avoid planting CLM-infested seedlings. The use of pesticides is the most common measure for controlling CLM [44]. However, the rampant use of pesticides is damaging to the health and integrity of ecosystems [45]. Given the ability of Metarhizium species to affect herbivores negatively when associated with roots, their use as inoculants could protect coffee seedlings against CLM.

During field studies (MLF unpubl. data), M. robertsii and M. brunneum were isolated from coffee roots collected in a diversified coffee system in the municipality of Patrocínio–Cerrado (savannah like) biome of Minas Gerais, using the bait insect method [46,47]. Here, the potential of those two Metarhizium species, when in association with coffee roots to protect coffee seedlings from CLM damage, was evaluated. We hypothesized that Metarhizium species inoculated in coffee seedlings can (a) associate with roots and, as a consequence of this, (b) improve seedling growth and (c) indirectly promote protection against CLM.

2. Materials and Methods

2.1. Fungal Isolates, Plants and Insects

The isolates were RD-20.114 of M. robertsii and RD-20.120 of M. brunneum. These were obtained from coffee roots in 2020, using larvae of Tenebrio molitor (Coleoptera: Tenebrionidae) as a bait insect [46,47]. They were collected in a diversified coffee system at an Experimental Research Station of the Agriculture and Livestock Research Enterprise of Minas Gerais (EPAMIG) in Patrocínio, Savannah-like biome of Minas Gerais (18°9′48′′ S and 46°59′00′′ W). These isolates are maintained in tubes on slanted PDA at 5 ºC.

Coffee seedlings—Coffea arabica variety “IAC 44”—were obtained from the coffee nursery of the Plant Pathology Department at Federal University of Viçosa (UFV), Viçosa, MG, Brazil. They were three months old (“orelha de onça” stage, meaning “jaguar ear”: with the two cotyledons open), planted in soil. These seedlings were cultivated without use of pesticides and were visually inspected to ensure the absence of pests and diseases before setting up the experiment. The coffee seedlings were transplanted into 3 l pots (one seedling per pot) containing substrate MecPlant^® (commercial substratum based on Pinus bark) and kept in a greenhouse until fungal inoculation, totalizing 60 pots for all experiments. Each seedling was watered (30 mL) every two days. Monthly fertilization consisted of 30 mL of 4 g L⁻¹ ammonium sulfate per seedling. The experiment was set up three months after transplantation when plants had at least six pairs of true leaves.

CLM-infested leaves were collected from plants at Diogo Alves de Mello Experimental Station at UFV-Viçosa, MG, Brazil (20º45’14” S; 42º52’55” W). The rearing was kept in the Laboratory of Agroecology of EPAMIG-Viçosa, MG, Brazil, at 23 ± 1 ºC and 12:12 (L:D). Mined leaves were maintained in transparent acrylic cages (40 × 40 × 40 cm) with their petioles inserted into plastic boxes (20 × 10 cm) with flexible polyurethane foam sections soaked in tap water. When adults emerged from infested leaves, they were transferred to new cages with clean coffee leaves to ensure continuity of the CLM life cycle [48].

2.2. Fungal Suspensions

Conidia of M. robertsii and M. brunneum from stock cultures were plated on Petri dishes (9 cm diameter) containing potato dextrose agar (PDA) supplemented with 0.05 g L⁻¹ chloramphenicol and incubated in darkness at 26 °C for 15 days. Subsequently, Metarhizium conidia were harvested from the Petri dishes with a sterilized metal spatula and inoculated into plastic bags containing rice. For this, the bags with 100 g of rice (polished parboiled type 1) and 30 mL of sterile distilled water were autoclaved for 15 min at 1.0 atm pressure to 120 ºC. The bags were then kept in a laminar flow cabinet until completely cooled. Rice bags with Metarhizium conidia were kept in darkness at 26 °C for 7 days [49]. After that, 50 g of rice grains with conidia were suspended in 1 L of sterile Tween solution 0.05%. These suspensions were filtrated twice using bilayer sterile cheesecloth so as to remove rice grains and hyphal fragments and were then vortexed for 30 s.

The concentrations of suspensions were adjusted to 1 × 10⁸ conidia mL⁻¹ in sterile Tween solution 0.05%, with the aid of a Neubauer haemocytometer. We checked conidial germination by transferring 150 μL of the suspension onto Petri dishes (9 cm diameter) with PDA + chloramphenicol (0.05 g L⁻¹), incubating these at 26 °C for 24 h and counting germinated conidia. Suspension of both isolates presented germination rates higher than 98%.

2.3. Treatments

Seedling inoculation was carried out by soil drench placing 30 mL of suspension (1 × 10⁸ conidia mL⁻¹ in sterile Tween solution 0.05%) on the soil close to the stem. Controls received 30 mL of blank Tween 0.05%. The three treatments were M. robertsii, M. brunneum, and the control, with 20 coffee seedlings per treatment. The experiment was conducted in a completely randomized design. Each potted coffee seedling was kept in a cylindrical cage (30 cm diameter × 60 cm height) made with wire rods (3 mm diameter) and covered with gauze (Figure 1). The experiment was maintained in a greenhouse at the Laboratory of Insect–Microorganism Interactions, UFV.

2.4. Colonization of M. robertsii and M. brunneum in Roots of Coffee Seedlings

Roots from coffee seedlings were collected at the end of the experiment in order to evaluate the presence or absence of Metarhizium. Each root was shaken in a standardized manner to remove non-rhizosphere soil [50]. Subsequently, we ground 2 g of roots of each plant individually with a mortar and pestle and then suspended them in 5 mL of sterile 0.01% Tween, in a 15 mL Falcon^® tube. Tubes were rotated for one hour in a rotary shaker at 150 rpm [51]. The suspensions were then vortexed for 15 s and aliquots of 100 µL from each suspension were plated in three Petri dishes (9 cm diameter) with semi-selective medium and spread with a sterile Drigalski spatula. The semi-selective medium for hypocrealean fungi consisted of 10 g peptone, 20 g dextrose, 15 g agar and 1 L distilled water. After sterilizing the medium, we added 0.05 g L⁻¹ of cycloheximide and tetracycline, 0.6 g L⁻¹ streptomycin and 0.175 g L⁻¹ CTAB (cetyltrimethylammonium bromide) [10]. Petri dishes were incubated in darkness at 26 °C for 14 days. We then evaluated the frequency of fungal colonization in roots. Only plants that presented at least one Metarhizium colony were considered positive for fungal association.

2.5. Effects of M. robertsii and M. brunneum on Coffee Seedling Growth

To evaluate whether Metarhizium species have the potential to promote the growth of coffee seedlings, we evaluated the numbers of leaves, leaf areas, lengths of aerial parts and roots, stem diameters, and fresh and dry weights of aerial parts and roots. Since the seedlings had different initial sizes, the length of aerial parts and the numbers of leaves of each coffee seedling prior to treatments were measured. At the end of the experiment, 43 days after fungal inoculation, we repeated measurements of the aerial parts and leaf counts. We calculated increases in these values as proportional to the initial values for each plant. We used ImageJ software to measure leaf area [52], a tape measure to evaluate the length, digital calipers to measure the diameter and a precision balance to evaluate the weight.

2.6. Effects of M. robertsii and M. brunneum on L. coffeella

Eight days post-inoculation, each plant was infested with two males and two females of CLM. After 48 h, we removed the adults and counted the eggs laid by the females on each seedling with a pocket magnifier. From this time, CLM development time (from egg to adult) and the number of mines, pupae and adults were daily evaluated. The seedlings were evaluated until there was no more adult emergence, at ca. 40 days. On the last day of evaluation, all leaves from seedlings were removed and had pictures taken of them. We evaluated the percentage of leaf area that had been mined by measuring the total leaf area and mined area of the images with ImageJ software [52].

The survival and reproductive performance of CLM adults that emerged from seedlings were evaluated. For this, we used 26 CLM couples formed from adults that emerged from M. robertsii-inoculated plants and 30 from each of M. brunneum-inoculated plants and control plants. Each couple was placed inside a plastic pot (500 mL) covered with PVC, containing one clean untreated coffee leaf from coffee seedlings maintained in a greenhouse. The leaf petiole was inserted in a plastic container (3 mL) with water to maintain turgidity [48]. We evaluated the survival of males and females and numbers of eggs per female under a stereomicroscope daily until their death.

2.7. Statistical Analyses

Inoculation with M. robertsii, M. brunneum and controls were used as explanatory variables to investigate if they can associate with coffee roots and the effect on seedling growth and CLM parameters. To examine if Metarhizium species can associate with roots, the frequencies with which the fungi were recovered from coffee seedling roots were analyzed. The samples were scored as positive or negative for M. robertsii or M. brunneum. Based on the confirmation of the colonization by Metarhizium species, we excluded the inoculated coffee seedlings that were negative for fungus from all the analyses (effect on coffee seedling growth and CLM parameters). To examine coffee seedling growth variables (number of leaves, leaf area, stem diameter, length of roots and aerial part, fresh and dry mass of roots and aerial part) we used Analysis of Deviance (F-tests) assuming normal distribution. Pairwise comparisons were performed with the emmeans R-package (adjustment method: Tukey) [53]. Survival analyses with censored Weibull distributions were carried out to test CLM development times to adults and the survival of emerged males and females. We used generalized linear models (GLM) adjusted to a Poisson distribution to analyze CLM count data of number of eggs, mines, pupae and emerged adults per plant and the numbers of eggs per emerged CLM female. We used Analysis of Deviance with χ² tests and with pairwise comparisons as above. For percentage of leaf area mined, we used GLM adjusted to a Binomial distribution. We realized Analysis of Deviance with χ² tests and pairwise comparisons as above. R (R Core Team, 2018) was used to analyze all data.

3. Results

3.1. Colonization of M. robertsii and M. brunneum in Roots of Coffee Seedlings

Coffee seedlings of the control had no Metarhizium isolate recovered from roots (up to day 43). In contrast, both isolates were recovered from the roots of fungus-treated coffee seedlings. At 43 days after inoculation, M. robertsii was recovered from 18 of 20 coffee seedlings (90%) and M. brunneum from 15 of 20 seedlings (75%).

3.2. Effects of M. robertsii and M. brunneum on Growth of Coffee Seedlings

Only the inoculation of M. robertsii enhanced seedling growth. There were no differences among treatments in the numbers of leaves (F = 1.56, p = 0.22; Figure 2a), lengths of aerial parts (F = 0.37, p = 0.68; Figure 2b) and roots (F = 0.29, p = 0.743; Figure 2c), stem diameters (F = 1.23, p = 0.30; Figure 2d), fresh masses of roots (F = 0.19, p = 0.82; Figure 2e) and of aerial parts (F = 0.29, p = 0.74; Figure 2f) and dry masses of roots (F = 0.78, p = 0.46; Figure 2g) and of aerial parts (F = 0.37, p = 0.69; Figure 2h). However, the inoculation of M. robertsii in coffee seedlings increased the leaf area by 30% compared to the uninoculated control (t = 6.253, p < 0.001, df = 1; Figure 2i).

3.3. Effects of M. robertsii and M. brunneum on L. coffeella

Both isolates promoted protection against CLM. In coffee seedlings inoculated with M. robertsii (z = 2.39, p = 0.04) and M. brunneum (z = 5.68, p < 0.001), CLM presented delays of two days in the developmental times compared to uninoculated controls (Figure 3). There was no difference in the effects of inoculation with the two fungi on development times (z = 1.95, p = 0.12; Figure 3).

Fungal inoculation did not decrease the numbers of eggs (χ² = 1.95, p = 0.88; Figure 4a), mines (χ² = 24.43, p = 0.11; Figure 4b) or pupae (χ² = 26.59, p = 0.11; Figure 4c) of CLM. Nevertheless, inoculation of M. robertsii reduced the numbers of emerged CLM adults by a third compared to the control (z = 2.86, p = 0.01; Figure 4d).

Considering the damage caused by the pest insect, the percentage of mined leaf area was lower in plants inoculated with M. robertsii (0.08 ± 0.02%; t = 9.23, p < 0.001) and M. brunneum (0.28 ± 0.10%; z = 7.16, p < 0.001) when compared to the control (1.38 ± 0.17%) (Figure 5). The observed differences in mined leaf area in the two inoculated treatments were statistically significant (z = 3.58, p = 0.001; Figure 5).

The survival of emerged females (χ² = 3.37, p = 0.18) and males (χ² = 2.82, p = 0.24) were similar in all treatments (Figure 6). Females who emerged from plants inoculated with M. robertsii produced half the eggs that females who emerged from the control produced (z = 2.46, p = 0.03; Figure 7). The eggs of females that emerged from the M. brunneum treatment were similar to the control (z = 1.89, p = 0.13; Figure 7) and M. robertsii treatment (z = 0.77, p = 0.71; Figure 7).

4. Discussion

We showed that both Metarhizium isolates can successfully be applied as soil drenches to coffee seedlings where they establish associations with the roots. Metarhizium robertsii increases leaf area by a considerable 30%, an effect not seen from inoculation with M. brunneum. Furthermore, both Metarhizium isolates promoted some degree of protection against CLM in coffee seedlings as hypothesized, both reducing the percentage of mined leaf area and prolonging the insect developmental time. This effect was greater with M. robertsii with an additional reduction in the numbers of adults and eggs that emerged from inoculated plants.

Considering the colonization of coffee roots, Metarhizium robertsii appeared to be more efficient than M. brunneum, although a firm conclusion would require a study more specifically focused on this question and with appropriate sample sizes. However, studies with others plants show both species with similar root colonization capacity [14,54]. The ability of Metarhizium to colonize plant tissues varies by fungal species and strain, environmental conditions and host species [55]. Here, both Metarhizium isolates were in the same environmental conditions, and on the same host species. However, we used only one isolate for each Metarhizium species. Therefore, it is entirely possible that the difference found is due to the isolates rather than species. A study of these effects that considered a range of isolates of different species would be of great value, although this would present logistical difficulties.

Because we did not sterilize the surface of coffee seedling roots, it was not possible to distinguish between endophytic colonization by the fungi and growth only on the root surface. While such a point is important to understand how the fungi affect the plants, this was not the aim of the present study and so remains for future investigation. It is worth noting, however, that the initial stages of Metarhizium endophytic colonization involve rhizosphere colonization [56]. Likewise, in both associations, fungi can promote plant growth and protection against pests [3,4,15].

Inoculation with M. robertsii increased the total leaf area of coffee seedlings, indicating that this fungus acts as a growth promoter in coffee seedlings. EPF colonization may enhance plant growth by facilitating nutrient uptake through the plant root system or by translocating nitrogen from insect cadavers in the soil to the plant in exchange for carbon. [3,27]. We are aware of no previous study of Metarhizium inoculation in coffee plants. However, studies with other crops, such as sweet pepper (Capsicum annum) and sweet corn (Zea mays), have shown that M. brunneum inoculation can improve plant growth [14,15], a result we did not observe here.

We showed that plants inoculated with M. robertsii and M. brunneum reduced the total leaf area mined by 70% and prolonged CLM development time by two days compared to uninoculated controls. Furthermore, M. robertsii reduced the number of emerged CLM adults by a third and the females who emerged produced half the eggs produced by the control females. Numerically, the data do not show a great discrepancy between Metarhizium-inoculated and uninoculated plants (see Figure 3, Figure 4 and Figure 5 and Figure 7). However, this degree of protection against coffee pests in our study corresponds to the infestation couples of CLM for 48 h. Under natural conditions, coffee seedlings are subject to more than two CLM females, which live for about 13 days. Therefore, the effects that we found here will become more expressive in coffee nurseries.

The main aim of this study was to determine if inoculation with Metarhizium would provide a protective effect against an important herbivorous pest, in this case, CLM. Endophytic EPF can promote indirect protection against insects. Mutualist associations between plants and Metarhizium species may prolong the developmental time of herbivores, delay the onset of reproduction and also reduce birth rates [15,16,40]. It was suggested that the mechanisms of these systemic responses are fungal metabolites that could be produced and transported through the plant’s vascular system, either directly affecting herbivores or mediating indirect effects through the upregulation of plant defenses [35,36,57]. The fungal metabolites could be alkaloids, saponins, tannins, phenolic acids, steroids, quinones, and terpenoids, which are insect antagonists [58]. A further aspect to bear in mind is that endophytic fungi can induce changes in the emission of volatile compounds, which can, in turn, influence how insects choose plants for oviposition [59].

The insect’s feeding mode is an important feature of its interaction with the plant, which will affect any indirect protection of the plant by endophytic EPF. This is because chewing, sucking, mining and galling insects often respond differently to plant defenses [60]. Sucking insects are more strongly affected by endophytic EPF inoculation than the others [60] because fungal metabolites can be transported through plant vasculature [61]. It was argued that intracellular EPF growth is limited because of nutrient availability in the plant intercellular space and the absence of cell walls and cell membrane-degrading fungal enzymes [62]. Therefore, it was suggested that fewer effects of endophytic EPF occur in mining insects, as they do not feed directly on the plant vascular bundles, where EPF growth may be more extensive. In spite of this, we show here the negative effects of Metarhizium spp. against a mining insect. It would be interesting to examine the degree of intercellular growth in this case; meanwhile, we could expect even greater effects on sucking insects [60].

Brazil is the world’s largest producer and exporter of coffee [63]. However, climate-changing conditions and rising global temperatures are major threats to coffee production across the world [64]. There is strong evidence that rising temperatures and altered rainfall patterns are already affecting coffee yields, quality, pests and diseases [64]. Rising temperatures may promote pest insects by allowing them to produce more generations in a year [65,66]. CLM specifically is influenced by climatic conditions, with dry and hot areas being more favorable for this pest to reach high infestation levels [41]. In order to control heavy infestations of CLM, current management involves intensive insecticide use, such as chlorantraniliprol and organophosphate. As a consequence, problems with CLM resistance to conventional insecticides are frequent [44,67]. Therefore, the rising temperatures associated with indiscriminate insecticide use have increased the costs of CLM control, besides the negative impacts on the environment and human health.

The development of microbial inoculants that improve coffee seedling growth and have negative effects on CLM is a promising strategy to reduce costs (whether economical or in terms of environmental and human health) with pesticides. Our results are consistent with studies that report the indirect effect of EPF on reducing insect pest damage. The most common effects reported are delays in insect developmental time, feeding deterrence, retardation of insect growth, reduced survival and oviposition (reviewed in Bamisile et al. [27]). Studies of inoculation of entomopathogenic fungi in coffee are restricted to Beauveria bassiana [68,69], a biological control agent of coffee berry borer Hypothenemus hampei Ferrari (Coleoptera: Curculionidae), another key coffee pest worldwide [70]. Our study is therefore novel with respect to Metarhizium species and their colonization of coffee roots (here via soil drench), which improves plant growth and protection against CLM. Hence, we showed that M. brunneum and M. robertsii could be considered for the development of inoculants for coffee seedlings. Further studies are necessary to test the viability of using such a strategy in adult coffee plants in the field. Additionally, microscopical and molecular studies are also needed to elucidate the mechanisms behind the association of Metarhizium species and coffee roots as was performed with other plants (papers by Bidochka and St leger). Thereby, it will be possible to explain how the physiological processes occur inside the plant and may encourage new tests with other organisms.