Characterizing interactions between Fusarium graminearum and Fusarium virguliforme on early soybean growth and development

Abstract

Fusarium graminearum and Fusarium virguliforme are two fungal pathogens that can infect soybean [Glycine max (L.) Merr.] early in the growing season. A greenhouse experiment was conducted to determine if the presence of both pathogens resulted in an increased negative impact on early soybean growth compared with each pathogen alone. Treatment factors consisted of four inoculation treatments: a non-inoculated control, inoculation with F. graminearum only (FG), inoculation with F. virguliforme only (FV), and a combination of both pathogens (FG + FV). As well, three soybean cultivars (CH2105R2, P92Y51 and ‘Sloan’), and three seed treatments (a control, mefenoxam + fludioxonil, and fluxapyroxad) were evaluated. Results indicated FG + FV did not result in increased root disease, a decrease in shoot growth, or reduced root growth compared with FG and FV. The FV inoculation treatment exhibited more root disease severity compared with FG, but there were no differences between the two inoculation treatments for the other below-ground measurements, as well as all above-ground measurements. Fungicide seed treatment did not reduce root disease severity or lead to increased early soybean growth compared with the non-treated control. Cultivar differences were the primary factor which influenced early-season soybean growth. Because there was no evidence of an increased effect between presence of F. graminearum and F. virguliforme, these results suggest management decisions for both pathogens can remain separate for soybean growers in Wisconsin, but management of F. virguliforme should take priority over F. graminearum due to its prevalent cause of soybean sudden death syndrome.

Résumé

Fusarium graminearum et Fusarium virguliforme sont deux agents pathogènes fongiques qui peuvent infecter le soya (Glycine max [L.] Merr.) tôt durant la saison de croissance. Une expérience réalisée en serre a été menée pour déterminer si l’occurrence combinée des deux agents avait un effet négatif plus grand sur la croissance des jeunes plants de soya que chaque agent pathogène individuel. Les variables de traitement consistaient en quatre traitements d’inoculation: un témoin non inoculé, une inoculation avec F. graminearum (FG) seulement, une inoculation avec F. virguliforme (FV) seulement et une combinaison des deux agents pathogènes (FG + FV). De plus, trois cultivars de soya (CH2105R2, P92Y51 et ‘Sloan’) et trois traitements de semences (un témoin, méfénoxam + fludioxonil et fluxapyroxade) ont été évalués. Les résultats ont indiqué que l’action combinée de FG + FV n’a pas accru l’incidence de la maladie racinaire ni réduit la croissance des plantules ni réduit la croissance des racines, comparativement à FG et à FV. L’inoculation avec FV a engendré une maladie racinaire plus grave qu’avec FG, mais il n’y avait pas de différence entre les deux traitements quant aux autres mesures souterraines et de surface. Le traitement des semences avec les fongicides n’a pas réduit la gravité de la maladie racinaire ni contribué à accroître la croissance des premiers stades, comparativement au témoin non traité. Les différences entre les cultivars ont été les principaux facteurs qui ont influencé la croissance du soya en début de saison. Parce qu’il n’y a eu aucune preuve d’accroissement de l’effet résultant de l’action combinée de F. graminearum et de F. virguliforme, ces résultats suggèrent que les décisions relatives à la gestion des deux agents pathogènes peuvent être appliquées indépendamment par les producteurs du Wisconsin, mais que la gestion de F. virguliforme devrait primer celle de F. graminearum étant donné qu’il est la cause la plus fréquente du syndrome de la mort subite chez le soya.

Keywords:

• cultivar
• Fusarium graminearum
• Fusarium virguliforme
• seed treatment
• soybean
• sudden death syndrome

Mots clés:

• Cultivar
• Fusarium graminearum
• Fusarium virguliforme
• traitement des semences
• soya
• syndrome de la mort subite

Introduction

Soybean diseases represent one of the largest yield-limiting factors in the USA and particularly in Wisconsin. Those diseases specifically caused by Fusarium spp. have continued to be some of the most economically damaging (Koenning & Wrather, 2010). Fusarium graminearum Schwabe is primarily regarded as an economically important pathogen on corn (Zea mays L.) and small grains, but more recently, this species was found to be pathogenic on soybean (Pioli et al., 2004). Since then, F. graminearum has been shown to cause pod blight, seed and root rot, and pre- and post-emergence damping-off on soybean (Ellis et al. 2011). Fusarium virguliforme O’Donnell & T. Aoki is the causal agent of soybean sudden death syndrome (SDS) on soybean in the USA (Aoki et al., 2003). Both of these Fusarium species can be found in most of the major crop growing regions of the USA

A survey by Leslie et al. (1990) found F. graminearum to be present within corn, sorghum [Sorghum bicolor (L.) Moench], and soybean fields in northern states (IL, IN, MO, OH and WV) and in southern states (AL, AR, FL, GA, MS, NC and SC) of the USA. A 3-year study in Iowa showed F. graminearum was isolated from soybean roots in 23–50% of fields sampled and in 23–60% of counties sampled during the course of the study (Días Arias et al., 2013b). For F. virguliforme, SDS has been confirmed in 21 states throughout the soybean growing regions of the USA (Navi & Yang, 2014). In Wisconsin, F. graminearum is widespread due to its ability to cause Fusarium head blight (FHB) on wheat and ear and stalk rot on corn (authors, personal observation), and F. virguliforme has spread to areas further north, south, and east of the south central part of Wisconsin from where it was first found (Marburger et al., 2013).

Sudden death syndrome has emerged as one of the top soybean diseases in the USA. Yield loss estimates by Wrather & Koenning (2009) from 1996 to 2007 ranked SDS as the second to fifth top yield-suppressing disease during those years. Further work by Koenning & Wrather (2010) showed yield loss attributed to SDS totalled 1.5 million metric tons from 2008 to 2009, making it the fifth most yield-suppressing soybean disease across those two years. Yield loss of up to 80% in individual fields has been attributed to SDS, but yield loss of 5–15% is more common (Roy et al., 1997). Some research has been done to elucidate soybean yield loss due to F. graminearum. Koenning & Wrather (2010) estimated soybean yield loss due to the Fusarium root rot complex from 2006 to 2009 ranged from 158 000 to 289 000 metric tons each year. While F. graminearum is considered part of this complex, the amount of yield loss specifically due to F. graminearum has yet to be fully determined. Díaz Arias et al. (2013a) conducted three experiments at two locations in Iowa from 2008 to 2010 using microplots (0.37–1.0 m²) to determine soybean yield loss due to several Fusarium spp. Inoculating with F. graminearum did not significantly reduce soybean yield or seed size for any of the three experiments for both locations. In addition, Marburger et al. (2017) did not find evidence of soybean yield loss due to F. graminearum under field conditions in Wisconsin.

Research has examined interactions between and among members of the Fusarium genus. For ear rot diseases on corn, F. moniliforme J. Sheldon [= F. verticillioides (Saccardo) Nirenberg] was shown to suppress F. graminearum growth, along with suppressing another fungal corn ear rot caused by Aspergillus flavus Link:Fr. (Zummo & Scott, 1990, 1992). Reid et al. (1999) found corn ears inoculated with F. graminearum alone resulted in more disease and mycotoxin production compared with ears inoculated with F. graminearum + F. moniliforme and F. moniliforme alone. Other research has also shown negative associations between F. moniliforme with both F. graminearum and F. subglutinans [(Wollenweb. & Reinking) P.E. Nelson, T.A. Toussoun, & Marasas] (Marasas et al., 1979; Rheeder et al., 1990). At the present time, little is known about the interactions between or among Fusarium spp. on soybean. Because of their widespread distribution and ability to infect soybean, understanding the interactions between F. graminearum and F. virguliforme would be valuable information in order to help improve the management decisions for both species. Therefore, the objective of this study was to characterize interactions between F. graminearum and F. virguliforme on early soybean growth.

Materials and methods

Isolate source and identification

Single-spore cultures from three F. graminearum isolates (85T8, 35t10 and 253L4) and three F. virguliforme isolates (00–11-183, NRRL22823 and Soy-1) were grown on potato dextrose agar [Potato Dextrose Agar (Dehydrated); Fisher Scientific, Pittsburgh, PA] for 14 days at 23 ± 2°C with 12 h diurnal light. All three F. graminearum isolates originated in Iowa and were recovered from soybean plants. These isolates were identified by EF1-α sequences to be F. graminearum sensu stricto (lineage 7) (Ellis & Munkvold, 2014). Pure isolates of F. graminearum from Wisconsin were not available at the time of this study. Only one F. virguliforme isolate (Soy-1) which originated in Wisconsin was available during this study. The Soy-1 isolate was recovered from soybean, and PCR was performed on extracted DNA using the ITS5 and ITS4 primers. Resulting products were then sequenced, and sequenced products had 98% homology with F. virguliforme sequences in GenBank. The remaining two F. virguliforme isolates originated in Indiana and were recovered from soybean plants as well. Polymerase chain reaction was performed on extracted DNA from the isolates using primers developed by Wang et al. (2015). The resulting products were sequenced and compared with sequences in GenBank to confirm the identity as F. virguliforme. The pathogenicity of each isolate was not confirmed prior to the greenhouse experiment establishment. However, there is published evidence that two of the three F. graminearum isolates used, 85T8 and 253L4, are pathogenic to soybean (Díaz Arias et al., 2013a). There is also published evidence that all three of the F. virguliforme isolates used are pathogenic to soybean (Marburger et al., 2016; Marburger et al., 2018).

Inoculum preparation

A cornmeal-sand medium was used as the carrier and was prepared 30 days prior to the establishment of the greenhouse experiment. Sand (950 g), yellow cornmeal (50 g) and distilled water (150 mL) were added to an autoclavable bag and mixed. This step was repeated two times for each isolate and 12 times for a non-infested control to achieve the desired amount of carrier. All media were autoclaved on two consecutive days for 60 min at 121°C. The sterilized cornmeal-sand medium was infested by adding 25, 1 cm² blocks cut from a colony on a 9 cm diameter Petri dish to one autoclavable bag containing 950 g sand and 50 g cornmeal. Infesting the cornmeal-sand medium was performed individually for each isolate. All infested and non-infested bags were incubated at 22 ± 2°C for 30 days. After the 30 days and immediately prior to use in establishing the greenhouse experiment, infested cornmeal-sand media from all three isolates of each species was mixed in a larger bag, and this was performed separately for each species in order to avoid contamination across species and isolates. At this time, two separate 10 g subsamples were collected from the combined cornmeal-sand medium prepared for F. graminearum, F. virguliforme and the non-infested control to estimate the amount of inoculum present. This estimation was performed after the greenhouse experiment was established, as described below.

Greenhouse experiment

Soybean cultivars used

A greenhouse experiment was conducted at the University of Wisconsin-Madison West Madison Agricultural Research Station from February to April in 2014. Two separate runs (i.e. replications) of the experiment were performed. For each run, the experimental design was a completely randomized design with six replications of treatments organized in a split-plot arrangement. The whole plot factor consisted of four inoculation treatments: a non-inoculated control, inoculation with F. graminearum only (FG), inoculation with F. virguliforme only (FV), and a combination of F. graminearum and F. virguliforme (FG + FV). Subplots were arranged in a 3 × 3 factorial consisting of three soybean cultivars and three seed treatments. The cultivars used were: ‘Sloan’ (publicly available cultivar), CH2105R2 (Channel brand; Monsanto Co., St. Louis, MO) and P92Y51 (Pioneer brand; Pioneer Hi-Bred International, Inc., Johnston, IA). The cultivar ‘Sloan’ was chosen due to its susceptibility to F. virguliforme (Ziems et al., 2006; Tande et al., 2014) and F. graminearum (Broders et al., 2007). Cultivars CH2105R2 and P92Y51 were chosen because of their popularity due to high-yield potential and adaptability to Wisconsin. Each cultivar is described as partially resistant to F. virguliforme. The brand SDS rating for CH2105R2 is reported as a 2 on a 1–9 scale with 1 indicating most resistant. The brand SDS rating for P92Y51 is reported as a 7 on a 1–9 scale with 1 indicating most susceptible. While the cultivar ‘Sloan’ is also susceptible to F. graminearum (Broders et al., 2007), the resistance of cultivars CH2105R2 and P92Y51 to F. graminearum is unknown.

Seed treatments used

The three seed treatments used were: a control (non-treated seed), ApronMaxx® RTA (mefenoxam {(R,S)-2-[(2,6-dimethylphenyl)-methoxyacetylamino]-propionic acid methyl ester} and fludioxonil {(4–2,2-difluoro-1,3-benzodioxol-4-yl)-1H-pyrrole-3-carbonitrile)} [0.0101 mg a.i. seed⁻¹]; Syngenta AG, Basel, Switzerland), and Acceleron® DX-612 (fluxapyroxad {3-(difluoromethyl)-1-methyl-N-[2-(3ʹ,4ʹ,5ʹ-trifluorophenyl)phenyl] pyrazole-4-carboxamide} [0.0082 mg a.i. seed⁻¹]; Monsanto Co., St. Louis, MO). The two fungicide seed treatment products were selected based on their widespread use by soybean producers in Wisconsin and their product label information in regards to Fusarium control. The ApronMaxx® RTA label states the product protects against damping-off and seed rots due to Fusarium spp., and the Acceleron DX-612 label states suppression of seed and seedling disease caused by Fusarium spp. Neither product label specifies which Fusarium spp. are controlled or suppressed; however, it is widely recognized that both products do not control F. virguliforme. At the time of this study, no commercially available fungicide seed treatments were labelled for F. virguliforme control.

Inoculation

At the time of establishment, the FG and FV inoculation treatments were each prepared by combining the cornmeal-sand medium with Metro Mix™ potting soil (Sun Gro Horticulture, Agawam, MA) in a 1 part cornmeal-sand medium to 3 parts potting soil ratio. For the FG + FV treatment, the ratio was 3 parts Metro Mix® potting soil, 0.5 parts F. graminearum cornmeal-sand medium, and 0.5 parts F. virguliforme cornmeal-sand medium. Again, two separate 10 g subsamples were collected at this time from each of the four inoculation treatments. After the subsample collection, the cornmeal-sand-potting soil mixture was used to fill individual Cone-tainers™ (Stuewe and Sons, Inc., Tangent, OR). Cone-tainers™ measured 3.8 cm in diameter and 21 cm in depth. One seed was planted per Cone-tainer™. For set-up in the greenhouse, nine Cone-tainers™, representing the 3 × 3 factorial structure within each inoculation treatment, were placed in a 15 cm diameter pot. Each pot was placed in an aluminium tray. The 24 pots with aluminium trays (i.e. the four inoculation treatments × six replications) were randomly assigned to positions on a greenhouse bench.

For each run of the experiment, plants were grown for 30 days after the time of potting. During that time, greenhouse air temperature was maintained at 24 ± 5°C. Natural light was supplemented with grow lights (1000 watts) set for a photoperiod of 16 h of light day⁻¹. Plants were watered every day for the first 7 days after potting by adding ~200 mL of tap water to each Cone-tainer™, as well as filling the aluminium trays with tap water. After that, plants were watered every other day by filling each aluminium tray with tap water. In addition, aluminium trays were rotated every 7 days to mitigate potential greenhouse effects.

Inoculation treatment quantification

From each of the 10 g subsamples collected as described above, 0.5 g was used to estimate the amount of inoculum present. This estimation was done using the quantitative polymerase chain reaction (qPCR) methods outlined by Marburger et al. (2015). Because the qPCR methods used do not distinguish between mycelia and spores, the results provided in Table 1 are reported as ‘spore-equivalents’.

Table 1. Population estimates (i.e. spore-equivalents) for Fusarium graminearum and Fusarium virguliforme for each inoculation treatment within each run (i.e. replication) of the experiment using quantitative polymerase chain reaction (qPCR) analysis based on methods outlined by Marburger et al. (2015).

			F. graminearum			F. virguliforme
Run	Inoculation treatment^a	Media^b	Cq average^c	Cq S. D.^d	Estimated population^e –spore-equivalents g^−1 –	Cq^c	Cq S. D. ^d	Estimated population^e –spore-equivalents g^−1 –
FG positive control^f	-	-	22.91	0.58	354,770	N/A^g	N/A	0
FV positive control^f	-	-	N/A	N/A	0	28.30	0.11	272,540
Negative control^h	-	-	N/A	N/A	0	N/A	N/A	0
1	NTC	Sand	N/A	N/A	0	N/A	N/A	0
1	NTC	MM	N/A	N/A	0	N/A	N/A	0
1	FG	Sand	24.95	0.37	83,010	N/A	N/A	0
1	FG	MM	24.73	0.45	97,440	N/A	N/A	0
1	FV	Sand	N/A	N/A	0	23.36	0.44	14,492,300
1	FV	MM	N/A	N/A	0	24.67	0.85	3,591,880
1	FG + FV	Sand	25.48	2.45	56,920	23.21	0.52	8,948,570
1	FG + FV	MM	30.21	1.26	1,960	24.03	0.38	5,319,860
2	NTC	Sand	N/A	N/A	0	N/A	N/A	0
2	NTC	MM	N/A	N/A	0	N/A	N/A	0
2	FG	Sand	30.71	0.78	1,380	N/A	N/A	0
2	FG	MM	27.87	0.84	10,380	N/A	N/A	0
2	FV	Sand	N/A	N/A	0	26.24	0.43	1,176,900
2	FV	MM	N/A	N/A	0	27.78	1.86	596,070
2	FG + FV	Sand	32.34	0.37	P^i	34.05	1.41	P
2	FG + FV	MM	28.86	0.98	5,130	27.33	0.32	539,170

^aNTC: non-inoculated control; FG: inoculated with a mixture of three isolates of F. graminearum only; FV: inoculated with a mixture of three isolates of F. virguliforme only; FG + FV: inoculated with a combination of a mixture of three isolates of F. graminearum and a mixture of three isolates of F. virguliforme.

^bSand: mixture of the cornmeal-sand growth media used to grow each isolate; MM: a combination of the cornmeal-sand medium prepared for each inoculation treatment with Metro Mix™ potting soil (Sun Gro Horticulture, Agawam, MA) in a 1 part cornmeal-sand medium to 3 parts potting soil ratio. For the FG + FV treatment, the ratio was 3 parts Metro Mix® potting soil, 0.5 parts F. graminearum cornmeal-sand medium, and 0.5 parts F. virguliforme cornmeal-sand medium.

^cCq: quantification cycle (i.e. the number of polymerase chain reaction cycles at which the target amplicon was quantified on based on the parameters outlined by Marburger et al. 2015). The average represents two replications from each inoculation treatment and three technical replications used for the qPCR (n = 6).

^dCq S.D.: quantification cycle standard deviation.

^ePopulation estimate given in spore-equivalents per g of media^b. Equations used to estimate are outlined by Marburger et al. (2015).

^fPositive control: DNA template used in the qPCR reaction was extracted from the F. graminearum and F. virguliforme isolates used in this study.

^gN/A: not detected and quantified by the qPCR methods used.

^hNegative control: autoclaved Milli-Q purified water used in place of DNA template.

ⁱP: present but not quantifiable under the conditions outlined by Marburger et al. (2015).

Data collection

At the end of the 30 days, plants were measured for height. Plants were then removed from each Cone-tainer™, and the shoots and roots were separated. Roots were washed to remove particulate matter. The root system from each plant was placed in a clear plastic tray, immersed in water, and scanned using a flatbed scanner (Epson Perfection V700 Photo Scanner) at 400 dpi with a pixel size of 0.063 mm. The resulting images were analysed using WinRhizo 2013 software (Regent Instruments, Inc., Quebec, Canada). For the image analysis, colour classes were manually designated for the root system and the background. Root disease symptoms appeared as dark brown to black lesions. Therefore, within the designation for the root system, colour classes were also manually chosen to distinguish ‘healthy’ and ‘diseased’ root tissue. The data parameters collected using the image analysis included total root length, root surface area, root volume and per cent disease severity (i.e. the amount of diseased tissue in relation to the amount of total root system). The software calculated the root length, surface area and volume measurements based on Tennant’s (1975) statistical line intersect method (Ortiz-Ribbing & Eastburn, 2003). For determining disease severity, a healthy:disease ratio was calculated by the imaging software. Using this ratio, the length (cm) of ‘diseased’ root was estimated by the following equation: total root length/(‘healthy:disease ratio’ + 1). Per cent disease severity was then calculated by dividing the length of ‘diseased’ root by the total root length and multiplying by 100.

After the plants were scanned from each run of the experiment, two randomly chosen plants from each inoculation treatment were used to re-isolate each Fusarium species. Isolations were also performed on ‘healthy’ and ‘diseased’ root tissue. Root sections ~0.5 cm long were surface disinfested with 95% ethanol and grown on potato dextrose agar for 14 days at 23 ± 2°C with 12 hours diurnal light. Morphological characteristics were then used to identify each Fusarium species (Leslie et al., 2006). The appropriate species was re-isolated from each inoculation treatment, and neither species was recovered from the non-inoculated control (data not shown). Furthermore, the Fusarium spp. were recovered from ‘diseased’ root tissue and not from the ‘healthy’ root tissue (data not shown). Therefore, the per cent disease severity value calculated from the image analysis was assumed to be due to the Fusarium spp. Dry root and shoot mass data were also collected after drying the plant parts at 60°C for 7 days.

Statistical analysis

Mixed-model analysis of variance (ANOVA) was conducted using PROC MIXED within SAS Version 9.3 (SAS Institute, Cary, NC). For each growth characteristic measured, each data point was compared to the global average of non-inoculated control treatment within its respective growth characteristic to determine the effect relative to the control. This was done in order to remove inherent variability associated with the control inoculation treatment, and it was performed for each run of the experiment. Initial analyses were then conducted to investigate differences between each run of the experiment. Statistical differences were found between the two runs for a majority of the growth characteristics measured; however, further examination revealed these results were due to differences in magnitude between the two runs for those growth characteristics. Therefore, the data from both runs of the experiment were combined for analysis. Models were constructed and analysed individually for each growth characteristic measured. For all analyses, inoculation, cultivar, seed treatment and their interactions were considered fixed effects. Run, replication × inoculation (run), and the overall error term were considered random effects. For all analyses, fixed effects were tested for significance at the 5% level (α = 0.05), and means comparisons were calculated based on Fisher’s protected LSD. Degrees of freedom were calculated using the Kenward–Rogers method, and the SLICE option in SAS was used to compare means of significant interactions (Littell et al., 2006).

Results

Greenhouse experiment

Foliar measurements

The above-ground plant observations included plant height and shoot weight. Soybean plants were also monitored for foliar disease symptoms. During both replications of the experiment, foliar disease symptoms were not observed on any plants. The inoculation main effect and its interactions with cultivar and seed treatment did not significantly affect plant height (Table 2), but a cultivar × seed treatment interaction was found. No differences among the seed treatments were detected for CH2105R2 and P92Y51 (Table 3). For ‘Sloan’, decreased plant height was found for the mefenoxam + fludioxonil seed treatment compared with the control and fluxapyroxad treatments.

Table 2. Analysis of variance (ANOVA) results for soybean above- and below-ground growth characteristics as influenced by four inoculation treatments (non-inoculated control, inoculation with Fusarium graminearum only (FG), inoculation with Fusarium virguliforme only (FV), and a combination of F. graminearum and F. virguliforme (FG + FV)), three cultivars (Sloan, CH2105R2 and P92Y51), and three fungicide seed treatments (non-treated control, mefenoxam + fludioxonil, and fluxapyroxad).

		Above-ground measurements		Below-ground measurements^a
Source of variation	df	Plant height	Shoot weight	Disease severity^b	Root weight^c	Root length^c	Root surface area^c	Root volume^c
Inoculation (I)	2	0.1820	0.9907	<0.0001	0.2364	0.1308	0.2195	0.2834
Cultivar (C)	2	<0.0001	0.0002	0.1152	0.0002	0.0004	0.0002	0.0002
I × C	4	0.6272	0.8579	0.4258	0.9830	0.9704	0.9802	0.9722
Seed treatment (ST)	2	0.0002	0.0966	0.2426	0.1354	0.2243	0.2622	0.3279
I × ST	4	0.2663	0.3729	0.1797	0.1824	0.0804	0.1700	0.3515
C × ST	4	0.0011	0.0172	0.3112	0.0089	0.0084	0.0189	0.0500
I × C × ST	8	0.4677	0.7466	0.2724	0.6476	0.5668	0.5801	0.6326

^aThe root system from each plant was placed in a clear plastic tray, immersed in water, and scanned using a flatbed scanner (Epson Perfection V700 Photo Scanner) at 400 dpi with a pixel size of 0.063 mm. The resulting images were analysed using WinRhizo 2013 software (Regent Instruments, Inc., Quebec, Canada). For the image analysis, colour classes were manually designated for the root system and the background. Within the designation for the root system, colour classes were also manually chosen to distinguish ‘healthy’ and ‘diseased’ root tissue.

^bA healthy tissue:disease tissue ratio was calculated by the imaging software. Using this ratio, the length (cm) of ‘diseased’ root was estimated by the following equation: total root length/(‘healthy tissue:disease tissue ratio’ + 1). Per cent disease severity (i.e. the amount of diseased tissue in relation to the amount of total root system) was then calculated by dividing the length of ‘diseased’ root by the total root length and multiplying by 100.

^cData parameter was calculated by the imaging software using Tennant’s (1975) statistical line intersect method (Ortiz-Ribbing & Eastburn, 2003).

Table 3. Cultivar × seed treatment results for soybean plant height, shoot weight, root weight, root length, root surface area and root volume.

		Seed treatment^b
Variable	Cultivar^a	Control	Mefenoxam + fludioxonil	Fluxapyroxad
		^_______________________ cm ^________________________
Plant height	CH2105R2	−1.58 acd	−1.24 a	−0.98 a
	P92Y51	−0.52 a	−0.97 a	−0.34 a
	Sloan	−1.91 a	−4.65 b	−2.31 a
		^_______________________ mg ^_______________________
Shoot weight	CH2105R2	−87.9 acd	−23.5 a	−20.3 a
	P92Y51	−25.3 a	−25.4 a	+4.8 a
	Sloan	−63.1 a	−298.3 b	−122.5 a
		^_______________________ mg ^_______________________
Root weight	CH2105R2	+5.7 acd	+37.2 a	+30.7 a
	P92Y51	+4.6 a	+0.4 a	−2.4 a
	Sloan	+10.9 a	−113.4 b	−42.8 ab
		^_______________________ cm ^_______________________
Root length^e	CH2105R2	+36.8 acd	+98.2 a	+67.8 a
	P92Y51	−7.7 a	−2.6 a	−8.3 a
	Sloan	+70.6 a	−45.5 b	−138.2 b
		^_______________________ cm^2 _______________________
Root surface area^e	CH2105R2	+8.4 acd	+22.7 a	+19.0 a
	P92Y51	−1.7 a	−2.0 a	−3.8 a
	Sloan	+13.3 a	−35.6 b	−14.4 ab
		^_______________________ cm^3 _______________________
Root volume^e	CH2105R2	+0.13 acd	+0.39 a	+0.41 a
	P92Y51	−0.06 a	−0.11 a	−0.14 a
	Sloan	0.17 a	−0.75 b	−0.36 ab

^aCH2105R2, Channel Brand (Monsanto Co., St. Louis, MO); P92Y51, Pioneer Brand (Pioneer Hi-Bred International, Inc., Johnston, IA); Sloan, public cultivar.

^bControl, non-treated seed; mefenoxam + fludioxonil, ApronMaxx RTA (0.0101 mg a.i. seed⁻¹; Syngenta AG, Basel, Switzerland); fluxapyroxad, Acceleron DX-612 (0.0082 mg a.i. seed⁻¹; Monsanto Co., St. Louis, MO).

^cValues followed by the same letter for each seed treatment within a row (i.e. the cultivar) are not significantly different at P ≤ 0.05. Means comparisons were calculated based on Fisher’s protected LSD. The SLICE option in SAS (v. 9.3) was used to compare means of this significant interaction (Littell et al., 2006).

^dEach data point was compared with the global average of the non-inoculated control treatment within each run (i.e. replication) of the experiment in order to remove inherent variability associated with the non-inoculated control. Values represent the average difference compared with the non-inoculated control across both runs (i.e. replication) of the experiment.

^eRoot length, root surface area and root volume values were obtained from digital image analysis using a WinRhizo 2013 root scanner system (Regent Instruments, Inc., Quebec, Canada).

The inoculation main effect and its interactions with cultivar and seed treatment also did not significantly impact shoot weight (Table 2). A cultivar × seed treatment interaction was observed, and the results were similar to plant height. No differences in shoot weight among the three seed treatments were found for CH2105R2 and P92Y51 (Table 3). Decreased shoot weight for the mefenoxam + fludioxonil seed treatment compared with the control and fluxapyroxad treatments was observed for the cultivar ‘Sloan’.

Root measurements

The below-ground observations included root disease severity, weight, length, surface area and volume. For root disease severity, the inoculation main effect was the only significant effect identified (Table 2). Because each data point was compared with the global average of the non-inoculated control, per cent disease severity estimates for each inoculation treatment resulted as −1.3% for FG, +10.8% for FV, and +7.6% for FG + FV. There was no statistical difference between the FV and FG + FV treatments, but both inoculation treatments were greater than the FG treatment.

There was evidence of a cultivar × seed treatment interaction for each of the remaining below-ground measurements (Table 2). For CH2105R2 and P92Y51, there were no differences detected among the three seed treatments for root weight, root length, root surface area and root volume (Table 3). Seed treatments differences were observed for the cultivar ‘Sloan’. For this cultivar, the mefenoxam + fludioxonil seed treatment resulted in decreased root weight, root surface area, and root volume compared with the control, but no difference between fluxapyroxad and the control was found for each of these three root characteristics (Table 3). Root length measurements showed mefenoxam + fludioxonil and fluxapyroxad use decreased root length compared with the control for ‘Sloan’ (Table 3). While these differences were found for the ‘Sloan’ cultivar, no visual phytotoxic effects from seed treatments were observed.

Discussion

This study is the first to quantify interactions between F. graminearum and F. virguliforme on early soybean growth. The presence of both Fusarium spp. examined (i.e. FG + FV) did not result in increased root disease symptom development compared with FG or FV. In addition, the FV inoculation treatment displayed more root disease severity than FG, suggesting management of F. virguliforme should take priority over F. graminearum in Wisconsin. Furthermore, the FG treatment had little impact on disease severity compared with the non-inoculated control. While these results suggested FG contributed little to disease symptom development, they are consistent with findings by Marburger et al. (2017). In that study, inoculating with F. graminearum did not negatively impact early-season soybean plants (i.e. V1; Fehr and Caviness 1977) under field conditions in Wisconsin. The +10.8% root disease severity for the FV inoculation treatment is close to the 16.1% root disease severity observed by Marburger et al. (2018) in which the same F. virguliforme isolates and inoculum preparation methods were used. Additionally, Marburger et al. (2018) only observed several plants which developed foliar SDS symptoms, and that finding is consistent with the lack of plants with foliar SDS symptoms in the current study.

The presence of both Fusarium spp. also did not result in decreased soybean shoot and root growth compared with FG or FV. In addition, the FG and FV treatments did not significantly impact early season soybean above- and below-ground growth characteristics. This latter result agrees with Marburger et al. (2018) who also found that inoculating with F. virguliforme did not significantly affect early season above- and below-ground growth characteristics. However, this latter result contrasts with Díaz Arias et al. (2013a). In that study, F. graminearum and F. virguliforme were highly aggressive (i.e. the relative ability to colonize and cause damage) under greenhouse conditions on soybean plants sampled at V3. Both species caused high root disease severity (~90%) and negatively affected the same above- and below-ground growth measurements examined in this study. The contrasting results between Díaz Arias et al. (2013a) and the current study could be due to the sampling time (V1 versus V3) and/or differences between the Fusarium isolates used. Furthermore, the lack of decreased root and shoot growth despite root disease symptoms in the current study may have been due to superficial or cortical root decay and not colonization and decay occurring near or within the vascular tissue (Díaz Arias et al., 2013a). However, this distinction was not made in the present study.

While the inoculum preparation was consistent during each run of the experiment with no observable variability, the observed results for the inoculation treatments may have been affected by the amount of inoculum present for each species, the variability among inoculation treatments, and the variability between the replications of the experiment. Furthermore, the differences in inoculum amounts for each of the inoculation treatments may have been affected by the inability of the qPCR method outlined by Marburger et al. (2015) to distinguish between mycelium and spores. Since mycelium and spores were both present during the inoculum preparation, the estimated population was reported as ‘spore-equivalents’. The estimated population (i.e. spore-equivalents g⁻¹ media) for F. virguliforme using qPCR methods was significantly higher than the estimated population for F. graminearum during each run of the experiment (Table 1). Ellis et al. (2011) reported 2.5 × 10⁴ macroconidia mL⁻¹ or greater were needed for optimum disease development by F. graminearum in a rolled-towel assay. The population estimates for F. graminearum in this study are greater than the minimum described by Ellis et al. (2011) in the first run but are slightly below it for the second run. However, the qPCR methods used to quantify the amount of inoculum have not been correlated to root disease symptoms caused by F. graminearum. For F. virguliforme, populations of 10²–10³ colony forming units (CFU) g⁻¹ of soil isolated from dilution plating methods have been reported from symptomatic soybean plants (Rupe et al., 1997; Scherm et al., 1998; Mbofung et al., 2011). While the population estimates for F. virguliforme in this study were well above the numbers reported in those previous studies, they are similar to the inoculum quantification results by Marburger et al. (2018) who used similar inoculation preparation and quantification methods. Similar to F. graminearum, the qPCR methods described by Marburger et al. (2015) have not been correlated to root and foliar SDS disease symptoms.

The lack of a detectable effect on above- and below-ground characteristics for the CH2105R2 and P92Y51 cultivars when inoculated with F. virguliforme was hypothesized to be due to their resistance. Marburger et al. (2018) also observed similar results for CH2105R2. In a field study conducted in Wisconsin, the CH2105R2 and P92Y51 cultivars were the only two of 10 cultivars examined which consistently exhibited the least amount of SDS foliar symptom development at the R5 to R7 (Fehr & Caviness, 1977) growth stages. Their resistance to F. graminearum is unknown, however. Therefore, attributing cultivar resistance as the cause for the lack of above- and below-ground growth impact when inoculated with F. graminearum has to be verified. Xue et al. (2007) reported significant differences in root-rot severity, per cent reduction in shoot length and per cent reduction in dry mass among three cultivars inoculated with F. graminearum under growth chamber conditions. The isolates and soybean cultivars used in that study were from Canada, and differed from the results from the current study. The ‘Sloan’ cultivar used in this study has demonstrated susceptibility to F. graminearum (Broders et al., 2007) and F. virguliforme. Although the F. graminearum population estimates were similar overall with those described by Ellis et al. (2011), the amount of inoculum present still may not have been enough to cause optimum disease development, even on the ‘Sloan’ cultivar. Inoculating with F. virguliforme can result in foliar symptoms on young susceptible plants (Tande et al., 2014). Even with the population estimates in this study, foliar SDS symptoms were not observed. Marburger et al. (2016) used the same isolates in a field study in Wisconsin. In that study, visible foliar SDS symptoms were also not observed on young plants (V1–V3), but foliar SDS symptoms did develop later in the growing season (R5–R7). The absence of foliar symptoms is again fairly consistent with Marburger et al. (2018) who only observed a few plants with symptoms. A loss of pathogenicity of the isolates is not considered to be a plausible contributing factor.

Studies have demonstrated fungicide seed treatments have efficacy on Fusarium spp. (e.g. fludioxonil), but efficacy differs by Fusarium sp. (Munkvold & O’Mara, 2002; Broders et al., 2007; Ellis et al., 2011). The fungicide seed treatments used in this study did not translate to increased early soybean growth. In addition, mefenoxam + fludioxonil use led to decreased plant growth for the cultivar ‘Sloan’. These results support some of the findings by Marburger et al. (2017) who investigated whether mefenoxam + fludioxonil and fluxapyroxad would mitigate potential soybean yield loss caused by F. graminearum under field conditions. In that study, both fungicide seed treatments resulted in decreased disease severity compared with the non-treated control in 2013, but no difference among the seed treatments was found in 2014. Despite the result from 2013, both fungicide seed treatments did not display increased root weight, length, surface area, or volume compared to the non-treated control each year. Even though mefenoxam + fludioxonil use showed decreased plant growth for the cultivar ‘Sloan’ in the current study, mefenoxam + fludioxonil and fluxapyroxad use did not negatively affect soybean growth for CH2105R2 and P92Y51. This result is consistent with Marburger et al. (2017) who also used CH2105R2 and P92Y51 in that experiment. In using seed treatments to control F. virguliforme, commercially available seed treatments during this study had shown inconsistent or no effects for reducing SDS symptoms under field conditions (Chilvers et al., 2011; Weems et al., 2015). However, in December 2014, a new fungicide seed treatment, fluopyram, was made commercially available to soybean growers. Since then, this new fungicide labelled for protection against F. virguliforme has been shown to reduce SDS severity and protect against yield reductions (Kandel et al., 2016a, 2016b; Gaspar et al., 2017; Vosberg et al., 2017).

Because there was no evidence of an increased effect on soybean disease development or early growth in the presence of both F. graminearum and F. virguliforme, the results from the current study suggest management decisions for both pathogens in Wisconsin can remain separate. Soybean growers should continue selecting the highest-yielding and best adapted cultivars for their region and should consider other management practices and seed treatments for controlling these two Fusarium spp.

Acknowledgements

We thank Dr Carol Groves, John Gaska, Dr Adam Gaspar, and Adam Roth for their technical assistance in conducting this research.

Additional information

Funding

We thank the Wisconsin Soybean Marketing Board (Grant #161389) for funding this research.