Interaction of a bioherbicidal fungus with a phenoxy herbicide for controlling velvetleaf (Abutilon theophrasti)

ABSTRACT The fungus, Fusarium lateritium Nees ex Fr. (FL), has shown potential as a bioherbicide for velvetleaf (Abutilon theophrasti Medik.) control. However, to achieve optimal infection and weed mortality, bioherbicide applications must be applied to weeds that are in the cotyledonary growth stage, followed by a dew treatment of at least 16 h. Greenhouse and field experiments were established to discover possible synergistic disease interactions with a phenoxy herbicide to improve the bioherbicidal potential of this fungus, and to mitigate some of the physical and environmental restrictions that limit the utility of this fungus as a bioherbicide. In greenhouse experiments, mortalities of 94% and 94% were recorded from sequential applications of 2,4-DB [4-(2,4-dichlorophenoxy)butyric acid] (0.02 kg ae ha−1) followed by FL (1.5 × 106 spores ml−1) at 5 min or 7 days after herbicide treatment. Velvetleaf plants in the fifth-to-seventh leaf growth stage were infected and killed with < 8 hrs. of dew. Similar results occurred under field conditions. Infection and weed control were inhibited by tank mixtures of 2,4-DB and FL, and by sequential applications of FL followed by 2,4-DB. These results suggest that timely applications of 2,4-DB followed by F. lateritium may provide effective control of velvetleaf.


Introduction
Velvetleaf (Abutilon theophrasti Medik.), an invasive weed native to China (Sattin et al., 1992), was introduced to North America in the seventeenth century primarily for fibre production (Spencer, 1984).Velvetleaf is self-pollinating, and a single plant can produce up to 17,000 seeds that may persist up to 50 yr in the soil and emerge throughout the crop growing season when conditions are favourable (Lueschen & Andersen, 1980;Warwick & Black, 1988).Up to 48% viability of velvetleaf seeds after 39 yr of seed burial has been reported (Toole & Brown, 1946).It is one of the most problematic weeds in corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] in the eastern U.S. (Colton & Einhellig, 1980;Hagood et al., 1980).
The success of velvetleaf as a weed is due to inadequate control and 'weedy' characteristics such as seed dormancy, the ability to germinate from deep in the soil, and tolerance for many herbicides used in corn and soybeans.These factors have led to large velvetleaf seed reservoirs in many areas of the midwestern U.S. where corn and soybeans are grown (Spencer, 1984).Velvetleaf interference can significantly reduce the yield of most agronomic crops, including corn, soybean, and cotton.Several studies reported that corn yield loss from velvetleaf interference approached a maximum asymptote around the weed density of 10 plants m −2 , with further density increases resulting in minor increases in crop damage (Lindquist et al., 1996).
The fungus, Fusarium lateritium Nees ex Fr. (FL) (Figure 1), has shown potential as a bioherbicide for velvetleaf control, as well as for controlling spurred anoda (Anoda cristata L.) Schlecht.(Walker, 1981a), prickly sida (Sida spinosa L.) (Walker, 1981b).However, the utility of this fungus for velvetleaf control was limited due to the inability to control weeds larger than cotyledonary to first leaf stages of growth, and the requirement of at least 12 h of dew (Boyette & Walker, 1985).We hypothesised that it may be possible to apply this approach (i.e.applying a low rate of 2,4-DB and substituting a bioherbicidal agent for the herbicide) for controlling velvetleaf.

Pathogen culture and storage
Cultures of Fusarium lateritium were grown on modified vegetable juice agar (Miller, 1955) in petri dishes inverted on open-mesh wire shelves in an incubator (Precision Scientific Inc., Chicago, IL, USA) at 25°C.Photoperiods (12 h) were provided by 20-W cool-white, fluorescent lamps, positioned 12 cm above each shelf to provide 200 μEm −2 s −1 photosynthetic active radiation (PAR).Conidia from 10to 12-day-old cultures were harvested by rinsing spore lawns on the petri dishes with distilled water and filtering debris through 4 layers of cheesecloth.Freshly harvested spores were used as the inoculum for all experiments.Spore concentrations were estimated and adjusted using haemacytometers.Long-term storage of the fungus was achieved in screw-capped tubes containing sterilised soil, maintained at 4°C (Bakerspigel, 1953).

Greenhouse experiments
In all experiments, velvetleaf plants were grown from seed obtained from Azlin Seed Co. (Leland , MS, USA) in a commercial potting mix contained in peat strips.Each strip contained 12 plants.The potting mix was supplemented with a slow-release fertiliser (14:14:14, NPK).The plants were placed in sub-irrigated trays on benches in the greenhouse.Greenhouse temperatures ranged from 25°C to 30°C with 40% to 60% relative humidity (RH).The photoperiod was 12 h with 1650 µEm 2 s −1 PAR measured at midday.

Disease kinetics experiments -field
Field experiments were conducted on a Dundee very fine sandy loam (Aeric Ochraqualf) soil at the USDA-ARS, Crop Production Systems Research Unit at Stoneville, MS.Plots consisted of four rows of soybeans ('Centennial' cv.), 12.2 m long and 1 m apart, with the two centre rows receiving treatment.All rows were planted with scarified coffee senna seed at a density of about 100 seeds•m -1 of row.Treatments consisted of: 1) FL (at a concentration of 1.5 × 10 6 ml −1 ) in distilled water; 2) FL + 2,4-DB; 3) FL fb 2,4 -DB at 0.02 kg ae ha −1 (after 5 min); 4) 2,4-DB fb FL (after 5 min).Velvetleaf plants were in the 5-7 leaf growth stage (40 cm, avg.) and were sprayed until fully wetted (approximately 200 L•ha -1 ).Applications were made at midday with a hand-held pressurised sprayer.For the first year, environmental conditions at the time of inoculation and for 24 h following treatment were: temperature at inoculation, 33 ˚C with a RH of 74%.The high temperature for the 24 h period was 33 ˚C and the low temperature was 25 C.
Maximum RH was 94%, with a dew period of 8 h.For the second year, environmental conditions at the time of inoculation and for 24 h following treatment were: temperature at inoculation, 36 ˚C with a RH of 58%.The high temperature for the 24 h period was 34 ˚C, and the low temperature was 24 ˚C.Maximum RH was 92%, with a dew period of 7 h.The plants were monitored for disease development at 3 d intervals for 12 d as described previously.Weed control was determined in two randomly selected 1m 2 subplots from each treated plot, 12 d after treatment.A randomised complete block design was utilised, and the treatments were replicated four times.Data over the 2 years from the experiments were pooled following subjection to Bartlett's test for homogeneity and analyzed using analysis of variance.Significant differences were determined using Fisher's Protected Least Significant Difference (FLSD) at P = 0.05.

Quantification of interactions
Interactions between components in mixtures were analyzed according to Colby (1967), using the formula E = X A + Y B -(X A ) (Y B ) / 100, where X A and Y B represent plant survival herbicide X A (2,4-DB) and bioherbicide Y B (FL), respectively.The observed response is obtained by comparing the activity of the single components (FL and 2,4-DB) with mixtures containing the same rate of the components applied singly.A deviation from the expected response, as calculated from the level of interaction (R) between the expected and the observed response of the two components, could indicate synergism or antagonism.Interactions are classified as additive when R = 1.0, synergistic when R > 1.0 and antagonistic when R < 1.0.Due to the inherent biological variability of the test systems, synergistic and antagonistic interactions are considered statistically significant when R ≥ 1.5 and R ≤ 0.5, respectively and additive interactions are significant at R values between 0.5 and 1.5 (Gisi et al., 1985).

Statistical analysis
In all experiments, the treatments were replicated four times.The experiments were repeated over time, and data were averaged following Bartlett's test for homogeneity of variance (Steel et al., 1997).A randomised complete block experimental design with four replications was utilised.The mean percentage of plant mortalities and biomass reductions were calculated for each treatment and were subjected to arcsine transformation.The transformed data were statistically compared using analysis of variance (ANOVA) at the 5% probability level.Results were back-transformed to the original measurements (percentages) for presentation.Data were analyzed via standard mean errors and best-fit regression analysis.All data were analyzed using the PROC MIXED function of SAS v 9.3 (SAS Institute, Cary, NC, USA).

Greenhouse experiments
Fl and 2,4-DB effects on mortality and biomass Macroconida of FL (Figure 1) were used in these experiments to test the efficacy on velvetleaf plants when applied alone or in combination with 2,4-DB.Following an 8-h dew treatment, high levels of velvetleaf mortality (94%) and biomass reduction (98%) were achieved when 2,4-DB was applied prior to FL under greenhouse conditions (Table 1).The level of weed mortality achieved by FL alone was only 6%, and that of 2, 4 -DB alone was 22% and a mixture of FL and 2,4-DB also resulted in 22% mortality (Table 1).When FL was applied prior to 2,4-DB, mortality of only 22 -24% occurred (Table 1).

Effects of dew duration
Dew periods greater than 8 h provided no significant increase in mortality or biomass reduction of velvetleaf seedlings caused by FL (Figures 2 and 3).Plant mortality resulting Table 1.Biological control of Abutilon theophrasti under greenhouse conditions by an interaction with the herbicide 2,4-DB.94 a 98 a a Means followed by the same letter in columns do not differ significantly at P = 0.05 according to Fisher's LSD. from treatment of 2,4-DB followed by FL after 5 min was represented by a second-degree polynomial regression curve which provided the best fit to the data, that is best described by the equation: Y = 7.43 + 15.75 -0.70X 2 , R 2 = 0.96 (Figure 2).Plant biomass reductions followed a similar trend and are represented by a second-degree polynomial regression curve which provided the best fit to the data, with the relationship best described by the equation: Y = 7.97 + 16.75X -0.79X 2 , R 2 = 0.96 (Figure 3).Tan to dark-brown lesions were visible 3 days (not shown) following inoculation with an 8 h dew period, and severity increased 5 days after treatment with 2,4-DB followed 5 min later by an inoculative spray of FL conidia (Figure 4).More severe lesion development (lesion coalescence) occurred after an 8-day treatment (Figure 5).

Field experiments
Results from field experiments corroborated the findings from greenhouse experiments.The only significant velvetleaf mortality and biomass reduction was achieved in test plots treated with 2, 4 -DB followed 5 min later by an FL treatment in field experiments (Table 2).The bioherbicide treatment alone caused only minimal effects, whereas FL applied to plants prior to herbicide application resulted in levels of mortality and biomass reduction equal to that of the herbicide alone suggesting herbicide toxicity to the fungus.An analysis of the possible interactions of the herbicide and bioherbicide under greenhouse and field conditions revealed that a strong synergistic interaction occurred when 2,4-DB was applied prior to FL, and that only slight additive interactions occurred with the other herbicide:pathogen treatments (Table 3).

Disease kineticsgreenhouse
Analysis of plant injury over a 12day period indicated that the injury severity progressed rapidly when plants were sprayed with 2,4-DB followed by FL either 5 min or 7 d after the herbicide treatment.Because there was no significant difference in plant injury at either pathogen inoculation time interval, only data from the 2,4-DB followed by FL after 5 min are presented.Extensive leaf wilting occurred as leaf spotting lesions coalesced by 8-9 days after application.The leaves were extensively blighted, and large stem lesions were evident (Figure 5).The effects on biomass reduction from the treatment of 2,4-DB followed by FL after 5 min was represented by a second-degree polynomial regression curve which provided the best fit to the data, and the relationship was best described by the equation: Y = 0.31 + 1.44X -0.012X 2 , R 2 = 0.98 (Figure 6).Plant injury was significantly reduced when FL was applied alone, or prior to the herbicide treatment, either after 5 min or 7 d.No injury occurred to plants treated with water only, thus a regression equation is not shown.Figure 5.Further lesion development (denoted by arrow) incited by F. lateritium on a velvetleaf stem under greenhouse conditions 8 d after treatment with 2,4-DB (0.02 kg a.e.ha −1 ) followed 5 min later by an inoculative spray of F. lateritium conidia at 1.5 × 10 6 conidia ml −1 .

Disease kinetics -field
Because greenhouse results demonstrated that no significant differences in plant injury occurred at either pathogen inoculation time interval, only data from the 2,4-DB followed by FL after 5 min are presented.The relationship was best described by the equation: Y = 0.01 -0.44 X -0.01 X 3 , R 2 = 0.98 (Figure 7).No visible soybean damage was observed (data not shown).As in greenhouse experiments, a strong synergistic interaction occurred when 2,4-DB was applied prior to FL, and only slight additive interactions occurred with the other herbicide:pathogen treatments under field conditions (Table 3).
Table 3. Interaction of Fusarium lateritium and 2,4 -DB relative to plant survival of velvetleaf under greenhouse and field conditions as analyzed using Colby's method (Colby, 1967).Colby (1967) equation: E = (X + Y) -(X)(Y)/100; E is the expected plant survival, and components X and Y represent MV and GLY, respectively.c R is the ratio of expected control/observed control.R values ≥ 1.5 are considered as synergistic and additive interactions occur when R is between 0.5 and 1.5 (Gisi et al., 1985); *denotes additive interaction and **denotes synergistic interaction.
quinclorac were applied to plant tissues of hemp sesbania (Sesbania exaltata (Raf.)Cory.), sicklepod, and kudzu Pueraria montana var.lobata Willd.(Ohwii)] seedlings (Hoagland et al., 2011).Additive or synergistic interactions also resulted in improved control of hemp sesbania following combinations of glyphosate and the bioherbicidal fungus, Colletotrichum truncatum (Schw.)Andrus & Moore) (Boyette et al., 2008a).Synergistic or additive interactions of MV and glyphosate on kudzu, redvine [Brunnichia ovata (Walt.)Shinners], and trumpetcreeper (Campsis radicans L.) have been demonstrated (Boyette et al., 2006;Boyette et al., 2008b).Weaver et al. (2012) reported antagonistic interactions of several commercial glyphosate products and MV.The glyphosate product (Touchdown®, Syngenta Crop Protection, Greensboro, NC, USA) combined with MV spores exhibited additive and synergistic interactions for controlling kudzu in the greenhouse and field (Boyette et al., 2014).Phoma herbarum West.was found to be compatible with non-formulated 2,4-D (2, 4dichlorophenoxyacetic acid) and their combination exhibited synergistic interaction for the control of dandelion (Taraxacum officinal) under field conditions (Schnick & Boland, 2004).In other studies with phenoxy herbicides, combinations of Phoma proboscis conidia with a mixture of 2,4 -D and MCPP [2-(2-methyl-4-chlorophenoxy)propionic acid] caused high mortality of mature field bindweed (Convolvulus arvensis) when dew periods ranged from 8 to 10 h (Heiny, 1994).The results of the studies presented in this report demonstrate that there was a synergistic effect upon weed control of velvetleaf when a conidial suspension of FL was applied subsequent to a sub-lethal dose of commercially formulated 2,4-DB.These data also suggest that a fungicidal response may have occurred when FL and 2,4-DB are tankmixed, and that a fungistatic response may have occurred when FL was applied prior to 2,4-DB application.Herbicides may affect many secondary and tertiary pathways and/or enzymes in plants some of which are closely related to plant disease defenses.Furthermore, herbicides may also directly influence the biochemistry and infectivity mechanisms of the bioherbicide (Hoagland, 1996).Additional comprehensive investigations on plant and pathogen biochemistry and molecular biology with respect to a specific bioherbicide, herbicide and weed will be necessary to define the intricacies of such herbicide: bioherbicide interactions. .Disease progression of F. lateritium (at 1.5 × 10 6 conidia ml −1 ) and 2,4-DB (at 0.02 kg a.e ha −1 ) infecting hemp sesbania under field conditions based upon disease rating of 0to 5, where 0 = no disease and 5 = plant mortality.Spray application rates were approximately 200 L ha −1 .The relationship for these components is best described by the following equations: FL (open circle, short dash), Y = -0.02+ .007X, R 2 = 0.96; FL + 2,4-DB (open inverted triangle, small dots), Y = 0.007 + 0.44 X + 0.04 X 2 , R 2 = 0.98; FL fb 2,4-DB after 5 min (solid square, long dash + small dots), Y = 0.17 + 0.04 X -0.02 X 2 , R 2 = 0.98; 2,4-DB fb FL after 5 min (solid sphere, solid line), Y = 0.01 -0.44 X -0.01 X 3 , R 2 = 0.98.Error bars represent ± 1 SEM.

Table 2 .
Biological control of Abutilon theophrasti under field conditions by an interaction with the herbicide 2, 4 -DB.
Observed percent survival of velvetleaf plants treated with FL, 2-4 D B, or FL + 2,4 -DB (either tank-mixed, immediately, sequentially).b Expected values were determined using the a