Introduction
Indaziflam is a Group 29 herbicide (cellulose biosynthesis inhibitor [CBI]) that was originally developed and marketed for the vine and tree nut market. Indaziflam was introduced in 2010 and since that time there has been significant label expansion that now includes vines, tree nuts, sugarcane (Saccharum officinarum L.), forestry, rights-of-way, turf, natural areas, and most recently rangeland. Rangeland use focuses on the control of several winter annual grasses that are highly invasive on western rangeland. These invasive winter annual grasses include cheatgrass (Bromus tectorum L.), Japanese brome (Bromus japonicus Thunb. ex Murr.; syn.: Bromus arvensis L.), cereal rye (Secale cereale L), jointed goatgrass (Aegilops cylindrica Host.), ventenata [aka African wiregrass, Ventenata dubia (Leers) Coss.], and medusahead [Taeniatherum caput-medusae (L.) Nevski]. Rangelands are increasingly threatened by these winter annual grasses that disrupt native plant communities, alter fire regimes, and reduce forage quality and biodiversity. Before the introduction of indaziflam, imazapic (Group 2, acetolactate synthase inhibitor) was the standard herbicide treatment for managing invasive annual grasses on rangeland. Both herbicides are now widely used on western rangelands, often in combination.
Indaziflam differs in selectivity from other CBIs because it is extremely active on monocots at low rates (Sebastian et al. Reference Sebastian, Fleming, Patterson, Sebastian and Nissen2017a), while dicots are more sensitive to other CBIs (i.e., isoxaben, dichlobenil) (Sabba and Vaughn Reference Sabba and Vaughn1999). Indaziflam provides preemergence control of annual grasses and must be applied well before germination to allow enough time for moisture to incorporate the herbicide into the soil solution (Clark et al. Reference Clark, da Silva, Dayan, Nissen and Sebastian2019). In perennial systems like native grasslands, indaziflam provides multiple years of annual grass control, allowing native perennial grasses and forbs to grow without competition from winter annual grasses (Clark et al. Reference Clark, Sebastian, Nissen and Sebastian2020; Courkamp et al. Reference Courkamp, Meiman and Nissen2022a; Mangold et al. Reference Mangold, Sencenbaugh, Harvey, Rebis and Rew2024; Sebastian et al. Reference Sebastian, Sebastian, Nissen and Beck2016b). Indaziflam’s selectivity is due primarily to its low water solubility (2.8 mg L−1), keeping the herbicide close to the soil surface, where it can contact germinating weed seeds (Shaner Reference Shaner2014). Bromus tectorum control can exceed 3 yr, providing the opportunity to deplete the soil seedbank (Clark et al. Reference Clark, Sebastian, Nissen and Sebastian2020; Courkamp et al. Reference Courkamp, Meiman and Nissen2022a; Sebastian et al. Reference Sebastian, Nissen, Sebastian and Beck2017b).
Imazapic is very water soluble (2,200 mg L−1 at pH 5, solubility increases as pH increases) compared with indaziflam (Shaner Reference Shaner2014). It provides short-term annual grass control, typically a single growing season, but under ideal conditions control can extend for a second season (Clark et al. Reference Clark, Sebastian, Nissen and Sebastian2020; Mangold et al. Reference Mangold, Parkinson, Duncan, Rice, Davis and Menalled2013; Sebastian et al. Reference Sebastian, Sebastian, Nissen and Beck2016b, Reference Sebastian, Fleming, Patterson, Sebastian and Nissen2017a). Imazapic has both preemergence and early postemergence activity on winter annual grasses, so applications can be made shortly before germination to soon after plants germinate, generally before tillering (Mangold et al. Reference Mangold, Parkinson, Duncan, Rice, Davis and Menalled2013). Annual grasses are usually more susceptible to imazapic than perennial grasses, although injury to desirable species can occur, especially during periods of plant stress such as drought (Shinn and Thill Reference Shinn and Thill2004). Unlike indaziflam, imazapic would require yearly applications to deplete the annual grass seedbank.
Sensitivity differences among species likely contribute to the difference in control longevity between indaziflam and imazapic. A root bioassay showed that monocots were more susceptible to indaziflam than dicots, with indaziflam concentrations resulting in a 50% reduction in root growth for B. tectorum approximately 3-fold lower than for kochia [Bassia scoparia (L.) A.J. Scott] (Sebastian et al. Reference Sebastian, Fleming, Patterson, Sebastian and Nissen2017a). A sequential greenhouse screening also showed the indaziflam concentration resulting in a 50% reduction in biomass (GR50) of B. tectorum was three times lower than for B. scoparia (Sebastian et al. Reference Sebastian, Fleming, Patterson, Sebastian and Nissen2017a). Indaziflam was significantly more active than imazapic on all annual grass species tested, except for A. cylindrica (Sebastian et al. Reference Sebastian, Nissen and Rodrigues2016a). For B. tectorum, the GR50 for imazapic was 11.8 times higher than indaziflam’s GR50 (2.71 and 0.23 g ai ha−1, respectively). Aegilops cylindrica was 32 times less susceptible to indaziflam than B. tectorum (Sebastian et al. Reference Sebastian, Sebastian, Nissen and Beck2016b).
There are also differences in the chemical characteristics between imazapic and indaziflam that may be contributing to the variance in length of annual grass control. Indaziflam is lipophilic (log Kow 2.8) and binds strongly to soil organic matter (OM) (Koc 605 to 1,094 ml g−1), limiting its movement in the soil (Shaner Reference Shaner2014). Imazapic is more hydrophilic (log Kow 0.393) and is weakly adsorbed to soil (Koc 137 ml g−1), with a higher potential for leaching with rainfall or in sandy soils (Shaner Reference Shaner2014; Tu et al. Reference Tu, Hurd and Randall2001). Both herbicides have moderate soil persistence (half-life: imazapic ∼120 d; indaziflam >150 d) that varies based on soil type and climate, with imazapic’s persistence increasing in soils with lower pH (Shaner Reference Shaner2014). Considering the soil half-lives are similar for both herbicides, the difference in solubility, adsorption to soil, and sensitivity of annual grass species may explain why indaziflam provides much longer control in the field. Additionally, direct soil analysis for indaziflam residues confirms that after 37 mo, only trace amounts of indaziflam were detected in the soil (<24 ppb), and B. tectorum seeds planted into the same soil showed no adverse effects (Courkamp et al Reference Courkamp, Meiman and Paschke2022b). This indicates that additional factors beyond soil residual are most likely contributing to the three or more years of annual grass control observed in the field.
The long-term annual grass control provided by indaziflam may also in part have to do with the herbicide’s adsorption to annual grass seeds still on the plant or on the soil surface, when applications are made. Clark et al. (Reference Clark, da Silva, Dayan, Nissen and Sebastian2019) found that indaziflam readily adsorbed to plant litter and was not easily washed off with simulated rainfall, while most of the applied imazapic could be removed with simulated rainfall. Very little published research has examined the impact of herbicide applications made directly to seeds. Young et al. (Reference Young, Gealy and Morrow1984) applied several herbicides directly to seeds of winter annual grasses and then placed the treated seeds on non-treated, moist soil. Paraquat reduced but did not eliminate germination of three weedy grass species (B. tectorum, A. cylindrica, and S. cereale), while glyphosate reduced B. tectorum and S. cereale germination. Pronamide and propham also impacted germination of some of the grass species (Young et al. Reference Young, Gealy and Morrow1984). Other research has evaluated treatments made to exposed seed on soil and found similar impacts to germination (Egley and Williams Reference Egley and Williams1978; Klingman and Murray Reference Klingman and Murray1976; Salazar and Appleby Reference Salazar and Appleby1982).
Indaziflam has provided multiyear winter annual grass control over a wide range of climatic and soil conditions. The current study was designed to determine whether one factor contributing to long-term winter annual grass control could result from seeds being exposed to high indaziflam concentrations at the time of application. The main objective of this research was to compare the impact of imazapic and indaziflam applied to B. tectorum seeds and A. cylindrica spikelets on seedling establishment. We hypothesized that indaziflam would adsorb more to the annual grass seed than imazapic, subsequently having a significant impact on seedling establishment.
Materials and Methods
Sources of Seed
Aegilops cylindrica spikelets and B. tectorum seeds were collected from field populations located at the Agricultural Research and Development Center (ARDEC; 40.6525°N, 104.9986°W; elevation: 1,552 m), north of Fort Collins, CO, in late summer 2020. ARDEC is the main field research center for the College of Agricultural Sciences at Colorado State University (CSU). Seeds were air-dried and stored under ambient conditions until greenhouse experiments were conducted from December 2020 until March 2021. Germination for both species was >70% at the time they were collected, and germination remained at that level during the course of these studies.
Soil
A clay loam soil (Nunn clay loam: fine smectic, mesic Aridic Agriustolls) was collected from an organic production field at ARDEC South Horticulture Research Facility (40.6106°N, 104.9967°W; elevation: 1,522 m). This soil was selected to ensure that there would be no residual herbicide present. Soil was passed through a 2-mm sieve and stored in 20-L containers in the Plant Growth Facility at CSU until needed. The soil was 38% sand, 28% silt, and 34% clay with 2.8% OM and a 7.8 pH.
Greenhouse Dose Response
Bromus tectorum seeds and A. cylindrica spikelets were counted out or weighed out depending on the species; for B. tectorum, 50 seeds were counted out, while 1.8 g of A. cylindrica spikelets were weighed out, each representing one experimental unit for that species for a total of three replicates per treatment. The seeds or spikelets were placed in trays made of wire screen measuring 6 by 6 by 1 cm. A single-nozzle track sprayer (Generation III Research Sprayer, DeVries Manufacturing, Hollandale, MN 56045) was used to apply indaziflam (Rejuvra® 200 SC, Bayer CropScience, Research Triangle Park, NC 27709) or imazapic (Plateau®, BASF, Research Triangle Park, NC 27709) to the seeds. The track sprayer was calibrated to deliver 187 L ha−1 at 172 kPa using a 11002 EVS flat-fan nozzle (TeeJet® Technologies, Springfield, IL 62703). Seeds were treated with 0, 5.5, 11, 22, 43, 73, 87, 102, and 174 g ha−1 of indaziflam or 0, 22, 43, 73, 87, 102, and 174 g ha−1 of imazapic plus one non-treated control. All treatment solutions contained 1% methylated seed oil. After herbicide applications, the seeds were allowed to dry for 1 h before being planted.
Plastic pots measuring 8 by 8 by 7.5 cm were used for the experiments. Pots were filled with approximately 380 g of soil, and treated seeds were placed on top and gently pressed into the soil. An additional 180 g of soil was used to cover the seeds, resulting in a planting depth of approximately 1 cm. The rainfall function of the track sprayer was used to apply 12 mm of simulated rainfall to wet the soils and start the germination process. Plants were allowed to grow under greenhouse conditions with a 25/20 C day/night temperature regime at an approximate 60% relative humidity. Natural light was supplemented with 400-W sodium-halide light bulbs (approximately 200 μmol m−2 s−1) to give a 15-h photoperiod. During the course of the experiment, pots were watered from the top with a fine mist as needed to avoid soil crusting, and twice per week all pots were subirrigated.
After 21 d, aboveground biomass was harvested. Plant biomass was dried at 60 C for 2 d to determine total dry biomass per pot. The pots were arranged as a randomized complete block design; blocks were arranged to account for any temperature gradients associated with the opening of vents to cool the greenhouse. The experiment was repeated.
Seed Rain-Off
To determine whether rainfall would impact the potential for indaziflam to inhibit A. cylindrica spikelet establishment, a second set of experiments was established. The same procedures as previously described were followed, with the exception that treated seeds were allowed to dry for 1 h and 24 h before they were subjected to a 12-mm simulated rainfall event using the single-nozzle track sprayer equipped with a 9504EV TeeJet® nozzle. Once treated seeds were subject to the rainfall event, treated and untreated seeds were planted and watered as previously described and allowed to grow for 21 d. Biomass was collected as previously described, and the experiment was repeated.
Retention of Indaziflam and Imazapic on Bromus tectorum Seeds and Aegilops cylindrica Spikelets
Five A. cylindrica spikelets and five B. tectorum seeds were placed in the same wire baskets previously described. They were then treated with 102 g ai ha−1 of indaziflam or imazapic using the single-nozzle, track sprayer at the same volume and pressure previously described. Seeds were allowed to dry for 24 h and were then subjected to simulated rainfall amounts of 0, 3, 6, 12, 24 mm as previously described. Following the rainfall events, seeds and spikelets were transferred to 12-ml glass test tubes filled with 2 ml of HPLC-grade acetonitrile. The test tubes were placed horizontally on an oscillating shaker moving forward and backward at a rate of 180 strokes min−1. Seeds/spikelets were shaken for 1 h, then 2 ml of the acetonitrile solution was transferred to 2-ml autosampler vials for liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis (Shimadzu LCMS 8040, Shimadzu Scientific Instruments, Columbia, MD 21046). The LC-MS/MS conditions are listed in Table 1.
To determine the amount of herbicide desorbed from A. cylindrica spikelets and B. tectorum seeds, the total herbicide adsorbed before rainfall was established first. A four-point calibration curve was used and 1 μl of the 2 ml extract was injected to calculate 100% retention from samples before rainfall. Subsequent samples were compared with the no-rainfall samples to determine desorption, and these values were expressed as a percent retained following simulated rainfall.
Table 1. Liquid chromatography–tandem mass spectrometry (LC-MS/MS) conditions for imazapic and indaziflam analyses.
a Material source: Phenomenex, Torrance, CA 90501.
b Material source: Millipore Sigma, St Louis, MO 63118.
c Material source: Thermo Fisher Scientific, Waltham, MA 02451.
Statistical Analyses
After failing to reject the null hypothesis of a Levene’s test that experimental variances are equal, repeated studies for all experiments were combined for analysis (R Core Team 2023). Nonlinear regression using the drc package in R v. 4.3.2 was used to determine indaziflam and imazapic rates required to reduce winter annual grass biomass by 50% and 90% (ED50 and ED90, respectively) when applied directly to seed, as well as to determine the amount of simulated rainfall required to remove 50% of herbicide applied to seed (r50) (R Core Team 2023). Dose–response curves were compared with log-logistic regression models, and best fit was determined using Akaike information criterion values. Lack-of-fit F-tests were conducted to explore whether the models were appropriate to describe these data. A two-parameter log-logistic regression model (Equation 1) was fit to the B. tectorum biomass data. The equation used to regress herbicide rate with percent reduction in B. tectorum biomass was:
$$f\left( x \right) = {1}\over{{1 + {\rm{exp}}\left\{ {b\left( {\log \left( x \right) - \log \left( {{\rm{ED}}50} \right)} \right)} \right\}}}$$
where b is the slope of the curve, and ED50 is the herbicide dose resulting in 50% reduction in response (biomass).
A three-parameter log-logistic regression model (Equation 2) was fit to the A. cylindrica biomass data. The equation used to regress herbicide rate with percent reduction in A. cylindrica biomass was:
$$f\left( x \right)=0 +{{{d - 0}}\over{{1 + {\rm{exp}}\left\{ {b\left( {\log \left( x \right) - \log \left( {{\rm{ED}}50} \right)} \right)} \right\}}}}$$
where d is the upper response limit, b is the slope of the curve, and ED50 is the herbicide dose resulting in 50% reduction in response (biomass). A four-parameter log-logistic regression model (Equation 3) was fit to the seed rain-off experiment biomass data, with slope, upper limit, and lower limit held the same for the curves. The equation used to regress herbicide rate with percent reduction in biomass after simulated rain was:
$$f\left( x \right) = c + {{{d - c}}\over{{1 + {\rm{exp}}\left\{ {b\left( {\log \left( x \right) - \log \left( {{\rm{ED}}50} \right)} \right)} \right\}}}}$$
where d is the upper response limit, c is the lower response limit, b is the slope of the curve, and ED50 is the herbicide dose resulting in 50% reduction in response (biomass). Herbicide desorption data from B. tectorum and A. cylindrica seed were fit to a three-parameter log-logistic regression model with upper limit constrained to 100 (Equation 4). The equation used to regress rainfall rate with percent herbicide desorption from seed was:
$$f\left( x \right) = c + {{{100 - c}}\over{{1 + {\rm{exp}}\left\{ {b\left( {\log \left( x \right) - \log \left( {{\rm{r}}50} \right)} \right)} \right\}}}}$$
where c is the lower response limit, b is the slope of the curve, and r50 is the rainfall rate resulting in 50% herbicide desorption from seed (Clark et al. Reference Clark, da Silva, Dayan, Nissen and Sebastian2019). For all regression models, an F-test of the curves was conducted to determine whether the difference between the ED50/ED90 and r50 values were statistically significant at the 5% level of probability. A likelihood ratio test was conducted to compare the parameters statistically among herbicides, wait time before simulated rainfall, and simulated rainfall rate.
Figure 1. Representative replicate from the dose-response experiment showing (A) Bromus tectorum and (B) Aegilops cylindrica at 21 d after treatment with imazapic and indaziflam. Visual differences in growth reflect herbicide rate effects on seedling establishment and biomass accumulation.
Results and Discussion
Greenhouse Dose Response
Indaziflam completely controlled B. tectorum at all application rates tested (Figure 1A), precluding the need for statistical analysis. Although imazapic provided B. tectorum control at rates ranging from 22 to 174 g ai ha−1 (Figure 1A), its efficacy was lower than that of indaziflam. Regression analysis established an ED90 of 67 g ai ha−1 imazapic (Figure 2A).
To our knowledge, this is the first report for imazapic or indaziflam applications directly to winter annual grass seeds for the purpose of determining their impacts on seedling establishment. The finding that imazapic applications did control B. tectorum was interesting, as it could be reasoned that grass seeds contain some free branched-chain amino acids that would allow seedlings to grow for some period without a functioning biosynthetic pathway. The herbicide concentration at the time of application is significantly higher than the concentration that seeds would be exposed to once the herbicide reaches the soil surface and is incorporated with rainfall. For example, a rate of 104 g ai ha−1 applied in 187 L ha−1 water is a concentration of 627 ppm or 2.3 mM, while that same application rate spread over 1 ha of soil to a 2-cm depth would be a concentration of only 0.34 ppm. Due to differences in water solubility and sorption characteristics, soil concentrations of imazapic are likely to be significantly lower than those of indaziflam.
Because of B. tectorum’s high sensitivity to indaziflam, A. cylindrica was included in this research to further understand how herbicides adsorbed to seeds could be a factor in providing long-term winter annual grass control. Aegilops cylindrica is approximately 32-fold less sensitive to indaziflam than B. tectorum (Sebestian et al. Reference Sebastian, Nissen and Rodrigues2016a). In similar dose–response experiments with A. cylindrica, we were able to calculate ED values for both herbicides (Figure 1B). Imazapic did not sufficiently decrease A. cylindrica growth to calculate an ED90 value, but the ED50 was 175 g ai ha−1. For indaziflam, the ED90 was 7.4 g ai ha−1 (Figure 2B), which is 14-fold lower than the 102 g ai ha−1 recommended broadcast application rate for indaziflam on A. cylindrica.
Figure 2. Greenhouse dose response of (A) Bromus tectorum and (B) Aegilops cylindrica to a range of seed-applied imazapic and indaziflam treatments. Dry weight was measured at 21 d after treatment and expressed as a percentage of the non-treated control. Symbols represent treatment means; bars indicate SE (n = 6).
Herbicide Rain-Off
In greenhouse dose–response experiments, B. tectorum seeds and A. cylindrica spikelets were planted 1 h after treatment (HAT), once the herbicide had dried on the seed. An obvious question was, does rainfall after herbicide application impact indaziflam’s herbicidal activity? To address this, three scenarios were compared: no rainfall, rainfall at 1 HAT, and rainfall at 24 HAT; results were not significantly different between treatments (Figure 3). Indaziflam’s low water solubility suggests that it is rapidly adsorbed to the surface of A. cylindrica spikelets impacting radical elongation even when simulated rainfall is applied just 1 HAT. As rainfall at 1 HAT did not reduce indaziflam efficacy, it is logical that waiting 24 HAT would also have no impact.
Figure 3. Impact of simulated rainfall on indaziflam efficacy following application to Aegilops cylindrica spikelets. Rainfall (12 mm) was applied either 1 h after treatment (1 HAT) or 24 h after treatment (24 HAT), after which spikelets were planted. Dry weight was measured at 21 d after treatment and expressed as a percentage of the non-treated control. Symbols represent treatment means; bars indicate SE (n = 6).
This experiment established that once indaziflam is applied to the seed, it is not easily removed by subsequent rainfall. This characteristic could provide control of many generations of seed, those still on the plant and older generations on the soil surface. Previous research has established that indaziflam and imazapic were adsorbed to B. tectorum litter (Clark et al. Reference Clark, da Silva, Dayan, Nissen and Sebastian2019). Imazapic was significantly easier to remove than indaziflam, 70% versus 38% desorbed with rainfall 24 HAT, respectively (Clark et al. Reference Clark, da Silva, Dayan, Nissen and Sebastian2019). Because both litter and seed coats are primarily composed of cellulose and lignin, it is reasonable to infer that these herbicides can similarly adsorb to the seed coat, contributing to prolonged herbicidal activity.
It is important to note that rainfall in natural environments is often variable in both timing and magnitude. Consequently, the efficacy of indaziflam observed in controlled conditions may be even more pronounced under field conditions, particularly in arid or semiarid regions. This is especially relevant in the western United States, where dry conditions and limited precipitation are common during the periods when winter annual grass seeds are dispersed and herbicides are typically applied.
Retention of Indaziflam and Imazapic on Bromus tectorum Seeds and Aegilops cylindrica Spikelets
The herbicide rain-off experiments demonstrated that indaziflam, and to a lesser extent imazapic, can adsorb to seeds and that 12 mm of rainfall applied 1 HAT did not impact indaziflam’s inhibition of A. cylindrica (Figure 3). Figure 4 illustrates how varying rainfall amounts (3 to 24 mm) influenced the desorption of indaziflam and imazapic from B. tectorum seeds (Figure 4A) and A. cylindrica spikelets (Figure 4B). Imazapic’s high water solubility resulted in less rainfall being required to remove imazapic from both B. tectorum seeds and A. cylindrica spikelets. The model estimated that for imazapic, 1.6 mm of rainfall was sufficient to achieve 50% desorption (r50) from B. tectorum seeds, and nearly all imazapic was removed at 24 mm (Figure 4A). In contrast, indaziflam was more strongly retained. For B. tectorum, an estimated 12.6 mm of rainfall was required to achieve r50. Even with 24 mm of rainfall, B. tectorum seeds still retained 37.2 ± 2.2% (mean ± SE) of applied indaziflam (Figure 4A).
Figure 4. Desorption of indaziflam and imazapic from (A) Bromus tectorum seeds and (B) Aegilops cylindrica spikelets following simulated rainfall event of 0, 3, 6, 12, and 24 mm at 24 h after treatment. Herbicide retention was measured after simulated rainfall and expressed as a percentage of total herbicide applied. Symbols represent treatment means; bars indicate SE (n = 6).
Desorption from A. cylindrica spikelets was generally lower than from B. tectorum seeds for both herbicides. For imazapic, r50 was estimated at 6 mm, and 82.2 ± 0.9% was removed at 24 mm (Figure 4B). Indaziflam was even more tightly adsorbed to A. cylindrica spikelets, with 56.0 ± 3.4% of the applied herbicide being retained at the highest rainfall amount (24 mm) (Figure 4B). Lower-limit model comparisons determined significant differences in maximum desorption between imazapic and indaziflam for both B. tectorum and A. cylindrica (P < 0.001 and P = 0.002, respectively) (Figure 4).
As an absolute amount per spikelet or seed, A. cylindrica adsorbed approximately 6 times more indaziflam compared with B. tectorum, 129 ± 10.6 ng per spikelet compared with 29 ± 4 ng per seed, respectively. However, the indaziflam concentration was actually higher for B. tectorum seeds (12 ng mg−1 per seed) compared with A. cylindrica spikelets (2.1 ng mg−1 per spikelet). This reflects the difference in size, the average A. cylindrica spikelet weighing 62 ± 4.4 mg compared with 2.4 ± 0.5 mg for B. tectorum seeds. For imazapic, absolute values were significantly lower compared with indaziflam. Aegilops cylindrica adsorbed 43 ± 3.3 ng per spikelet, approximately 5 times more than the 8 ± 0.4 ng per B. tectorum seed. As with indaziflam, B. tectorum had a higher imazapic concentration, 3.6 ng mg−1 per seed compared with 0.69 ng mg−1 per A. cylindrica spikelet. Comparing the indaziflam to imazapic ratio within species is interesting, because the ratio was basically the same: five times more indaziflam was adsorbed compared with imazapic. Because the herbicides were applied to dry seeds and spikelets at the same rate, this suggests that indaziflam’s low water solubility allowed it to concentrate on the surface of seeds and spikelets.
Final Considerations
There is limited published literature on herbicide retention by weed seeds or the impacts of direct herbicide applications to weed seeds on germination or seedling establishment. However, the concept of using herbicides to suppress seedling establishment has been explored for winter annual grass management in reduced tillage wheat (Triticum aestivum L.) production in the Pacific Northwest. Our findings support previous research (Young et al. Reference Young, Gealy and Morrow1984) that herbicides applied directly to annual grass seed can impact germination and establishment, and we also observed variations by herbicide.
Indaziflam applications for winter annual grass control represented a new paradigm in weed control on rangeland. It is the only cellulose biosynthesis-inhibiting herbicide currently registered for rangeland applications and offers long-term winter annual grass control while having a soil half-life similar to that of imazapic (Shaner Reference Shaner2014). Indaziflam’s mode of action is uniquely suited to work in this manner, because as soon as there is sufficient moisture to support germination, there would be sufficient herbicide in contact with the primary root to stop seedling establishment. This is most likely not the only factor contributing to long-term winter annual grass control, but it appears to be a major contributor, keeping in mind that weed seeds (propagules) will be exposed to highest herbicide concentrations at the time of application.
One concern that may arise from this research is the impact to native seed. Field trials and operational treatments have shown native annual, biennial, and perennial species recolonizing sites after indaziflam applications (Jones et al. Reference Jones, Harrison and Prather2024; Sebastian et al. Reference Sebastian, Swanson and Sebastian2023; Swanson Reference Swanson2024). There are a few explanations for why annual grass seeds are being impacted more than native seeds: monocots are more susceptible to indaziflam than dicots (Sebastian et al. Reference Sebastian, Sebastian, Nissen and Beck2016b, Reference Sebastian, Fleming, Patterson, Sebastian and Nissen2017a), so native dicots would be more tolerant. Additionally, native species do not have seed recruitment events every year, reducing the change of exposure when native seed was present, and native species are germinating from deeper in the soil seedbank (Saatkamp et al. Reference Saatkamp, Poschlod, Venable and Gallagher2014). For the first time, there is a management tool providing long-term control that could ultimately result in eliminating winter annual grasses from the soil seedbank.
Acknowledgments
The authors would like to thank Franck Dayan for his assistance with LC-MS/MS method development.
Funding statement
This research received no specific grant from any funding agency or the commercial or not-for-profit sectors.
Competing interests
SJN has received grants from Bayer CropScience. SLC was a contract employee at Bayer CropScience at the time of the research.
Open access
