Seed Collection

Four genetically uniform C. difformis biotypes were employed in germination studies. Two ALS-S biotypes, namely AS (American susceptible) and IS (Italian susceptible), and two ALS-R biotypes (WA and IR [Italian resistant]) were selected. IR and IS were mass collected in rice farms located close by in northern Italy, whereas biotypes WA and AS were collected in rice fields located less than 10 km apart from each other in California’s northern Sacramento Valley, USA. All biotypes except IS were part of earlier resistance gene-flow studies done by Merotto et al. (Reference Merotto, Jasieniuk, Osuna, Vidotto, Ferrero and Fischer2009).

Twenty plants from each biotype were placed in individual greenhouses and selfed during the 2010 growing season. Mature seeds were hand harvested in September 2010 by gentle shaking into paper bags. Preliminary studies showed that seeds must be sterilized before germination to avoid fungal growth. Thus, immediately following harvest, seeds were soaked for 3 min in a solution containing sodium chloride at 3%, and then rinsed with de-ionized water (DI H₂O) for an extra 3 min. Seeds were then soaked in a solution containing 0.4% captan (N-trichloromethylthio-4-cyclohexene-1,2-dicarboximide) fungicide (50% wettable powder) for 3 min, rinsed with DI H₂O for the same amount of time, and dry-stored at 7 C.

Germination Conditions

Germination tests were carried out on a one-dimensional thermogradient table, in which light was provided by four cool white fluorescent tubes with a 14-h photoperiod. Illuminance was measured using a light quantum meter (Spectrum Technologies, Plainfield, IL) and equaled a photosynthetic photon-flux density of 18 µmol m⁻² s⁻¹, which is considered appropriate for germination studies (Baskin and Baskin Reference Baskin and Baskin1998).

Starting in April 2011, seeds were subjected to a 3-mo-long cold, wet stratification period to simulate winter fallow conditions and overcome possible dormancy status, as stratification has been shown to be required for other weed species (Boddy et al. Reference Boddy, Bradford and Fischer2012). To this end, seeds of each biotype were placed into separate water-filled containers, which were further sealed to allow for dark conditions during stratification. Dark conditions were needed to avoid germination during the stratification process, as C. difformis is a positively photoblastic species, that is, its seeds require a light treatment to germinate (Sanders Reference Sanders1994). Containers were then stored at 10 C and kept under these conditions until laboratory germination experiments were carried out in July and August 2011 for the elucidation of cardinal temperatures and base water potential, respectively.

Experimental units consisted of covered 3.5-cm-diameter petri dishes with two Whatman No. 1 filter paper disks. Dishes were randomized within isothermal lanes on the thermogradient table. DI H₂O was added until the paper was saturated (nearly 2.0 cm³). Afterward, the dish was held on its side to drain excess water and avoid formation of a film of water around the seeds, thus allowing for an aerobic environment for seed germination (Baskin and Baskin Reference Baskin and Baskin1998). Dishes were sealed with tightly fitting covers to prevent evaporation.

Due to their minute size—C. difformis seeds measure only 300 by 600 µm (Sanders Reference Sanders1994)—germinating seeds were examined under a microscope (StereoZoom 6 Plus^®, Leica Mycrosystems, Buffalo Grove, IL) at 10× magnification. In preliminary studies, around 5% of the germinating seed population showed abnormal germination, as they developed radicles but no further coleoptile development was noticed. Therefore, coleoptile protrusion of approximately 0.5 mm was taken as the end point of germination instead of radicle appearance in order to avoid scoring abnormal seeds as germinated ones.

Data Analysis and Modeling

Thermal times for germination of the 50th percentile (θ_T(50)) are reported and represent the median of the seed population. Biotypes’ θ_T(50) values and final germination percentages, determined in the temperature-response experiment, were subjected to ANOVA using JMP v. 8.0 software (SAS Institute, Cary, NC), after Box-Cox transformation to meet assumptions.

Time (days) to median (50%) germination (t ₅₀) was often not observed directly; hence it was estimated using the probit procedure in SAS (SAS Institute) as discussed by Steinmaus et al. (Reference Steinmaus, Prather and Holt2000). Calculated t ₅₀ values were then converted into germination rates (GR ₅₀, 1/t ₅₀, expressed in days,) and plotted against temperature to estimate the optimum temperature at which the greatest GR was observed. Median values, that is, number of days until 50% of the seeds had germinated, were used because they typically exhibit less variation in developmental time than any other seed fraction (Dahal et al. Reference Dahal, Bradford and Jones1990).

Parameters for the hydrotime and thermal time models were inferred from germination data using probit regression analysis (Dahal and Bradford Reference Dahal and Bradford1990, Reference Dahal and Bradford1994; Dahal et al. Reference Dahal, Bradford and Jones1990). Base temperature was determined as follows:

([5]) $${\rm{probit}}(g) = \{ [{\rm{log}} \; (T - {T_{\rm{b}}})\;{t_{\rm{g}}}] - {\rm{log}}\;{{\rm{\theta }}_{{\rm{T}}({\rm{50}})}}\} /{\sigma _\theta }_{\rm{T}}$$

where probit(g) is the probit transformation of cumulative germination percentage g, θ_T(50) is median thermal time to germination, and σ_θT is the standard deviation in log thermal times, calculated as the inverse of the probit regression line. Other variables were defined following Equation 1.

Original germination time courses were normalized on a thermal time basis and reproduced as cumulative normal curves of the function:

([6]) $$G = \{ {\rm{log }} \; {t_{\rm{g}}}-[{\rm{log }} \; {{\rm{\theta }}_{{\rm{T}}({\rm{50}})}}-{\rm{log }} \; (T-{T_{\rm{b}}})]\} /{\sigma _{{\rm{\theta T}}}}$$

where G represents cumulative percentage germination, as derived from Equation 1 (Dahal et al. Reference Dahal, Bradford and Jones1990).

Parameter estimations for this function were obtained by using the Solver tool in Microsoft Excel^® (2003–2010) to minimize the root mean-square error (RMSE) between observed and simulated germination data (Huarte and Benech-Arnold Reference Huarte and Benech-Arnold2010), as follows:

([7]) $${\rm{RMSE}} = {\rm{ }}\sqrt {(1/n)\sum\limits_{i = 1}^n {{{({\gamma _{{\rm{obs}}}} - {\gamma _{{\rm{pred}}}})}^2}} {\rm{ }}} $$

in which γ_obs and γ_pred are the observed and predicted germination values, respectively. For each biotype, three sets of model parameters were derived by replication. The average of the three sets of parameters per biotype is presented; protected LSD (P < 0.05) values were obtained from an ANOVA of each parameter.

The distribution of Ψ_b (g) and other parameters associated with the hydrotime model were determined using the following adjusted repeated probit analysis:

([8]) $${\rm{probit}}(g) = [\Psi - ({{\rm{\theta }}_{\rm{H}}}{\rm{/}}{t_g}) - {\Psi _{{\rm{b(50)}}}}]/{\sigma _{{\rm{\Psi b}}({\rm{50}})}}$$

where σ_Ψb(50) is the standard deviation of the base water potential within a seed population, and Ψ is the actual water potential of the imbibition medium. The model assumes that σ_Ψb, Ψ_b(50) (base water potential for the median cohort), and θ_H are constants (Dahal and Bradford Reference Dahal and Bradford1990). Germination data from a range of water potentials were combined, and Equation 8 was then used to predict germination as a function of thermal time. The Solver tool was used as described earlier, yielding four sets of model parameters by replication. Parameter averages were obtained from an ANOVA of each parameter, and means were separated by Student’s t-test at the 95% confidence level.

To test our thermal time unit model on field-level seedling emergence, time to 50% emergence expressed in GDD at each water regime and soil was calculated using PROC PROBIT in SAS across observed values (SAS Institute) (Steinmaus et al. Reference Steinmaus, Prather and Holt2000). The base temperature for growth was assumed to be the same as the base temperature for germination (Bradford Reference Bradford2002). Parameters necessary for developing such model (θ_T(50), T _b, and σ_θT(50)) were chosen from the ALS-R biotype WA, given that ALS-inhibitor resistance is widespread throughout rice fields from which soils were collected. Emergence prediction of ALS-R biotypes, moreover, is necessary for optimizing stale seedbed systems for C. difformis management, given that this system is designed to address the increasing issue of herbicide resistance in California. Time to 50% emergence values (θ_T(50)E, expressed in GDD) were subtracted from time to 50% germination (θ_T(50)) of the WA biotype. The remaining GDD values were then added to the germination prediction line to yield an emergence prediction model in each water regime and soil in order to explain differences in emergence patterns as a function of differences in germination times among the seed population.

Seed Germination Experiments

Seed viability was high across all C. difformis biotypes (Table 1), as also reported for C. difformis at the population level (Pedroso et al. Reference Pedroso, Dourado Neto, Victoria Filho, Fischer and Al-Khatib2019). Final percentage germination averaged above 92% for all biotypes when the test temperature was ≥22.3 C (Figure 1), values slightly larger than those reported by Chauhan and Johnson (Reference Chauhan and Johnson2009) and Ismail et al. (Reference Ismail, Mansor and Rahman2007).

Table 1.

Final germination (G, %), base temperature for germination (T _b, expressed in C), thermal time constant to 50% germination (θ_T(50), GDD), and standard deviation in thermal time within the seed population (σ_θT(50)) of four Cyperus difformis biotypes, expressed in growing degree days (GDD).

^a AS (American susceptible) and IS (Italian susceptible) are ALS-inhibitor susceptible smallflower umbrella sedge biotypes, whereas WA (Washington) and IR (Italian resistant) are ALS-inhibitor resistant biotypes.

^b Average percent of total germinated seeds placed at temperatures ≥22.3 C.

^c Parameters are derived from the equation probit(g) = [(log(T − T _b)t_g) − log_θ50]/σ_θT and θ_T(50) is calculated as 10^θ50.

^d n = 15.

Figure 1.

Cumulative germination percentages of two acetolactate synthase–resistant (ALS-R; WA and IR [Italian resistant]) and two acetolactate synthase–susceptible (ALS-S; AS [American susceptible] and IS [Italian susceptible]) Cyperus difformis biotypes, plotted over time (days after seeding). Data points (symbols) are averages based on three replicates of ~100 seeds each.

Characterizing Germination Responses to Temperature

Differences in germination among biotypes were noticed at cold test temperatures (i.e., those ≤19.3 C) (Figure 1), as the ALS-R biotypes IR and WA germinated earlier than all S biotypes under these conditions. At a constant temperature of 13.25 C, neither S nor R biotypes were able to germinate throughout the course of the experiment.

Final germination percentages (Table 1) varied significantly among biotypes, with the ALS-S biotype IS having the greatest final germination among all biotypes. There was no clear pattern between ALS-R and ALS-S biotypes in this regard; base temperatures (T _b) also did not vary across C. difformis biotypes (Table 1). Estimated T _b values averaged 16.96 C, similar to those obtained by Pedroso et al. (Reference Pedroso, Dourado Neto, Victoria Filho, Fischer and Al-Khatib2019). However, this value is 2.15 C larger than that reported by Derakhshan and Gherekhloo (Reference Derakhshan and Gherekhloo2013), which could be related to the model presented in this work being developed using a wider range of test temperatures (Bradford Reference Bradford2002). Variations within a species due to genetic diversity of ecotypes from separate geographic regions can also be expected (Baskin and Baskin Reference Baskin and Baskin1998).

ALS-R biotypes had significantly lower thermal time to 50% germination (θ_T(50)) values relative to ALS-S (Table 1), and hence need to accumulate fewer thermal units to complete their germination process, allowing them to germinate faster than S biotypes given similar T _b values. Interestingly, enhanced germination under cold temperatures and ability to germinate faster than S biotypes have been linked with mutations in the ALS gene (Dyer et al. Reference Dyer, Chee and Fay1993; Eberlein et al. Reference Eberlein, Guttieri and Berger1999; Park et al. Reference Park, Mallory-Smith, Ball and Mueller-Warrant2004; Purrington and Bergelson Reference Purrington and Bergelson1999). The standard deviation in thermal time within the seed population (σ_θT(50)) averaged 0.21 GDD and was larger in ALS-R relative to ALS-S biotypes.

ALS-R biotypes (WA and IR) had higher median germination rates (GR ₅₀, 1/t ₅₀) and thus completed the germination processes more rapidly than their S counterparts (Figure 2). GR ₅₀ increased linearly with test temperature, and therefore no thermoinhibition effect was noticed (i.e., no decline in germinability was found as test temperatures increased). Regardless of biotype, GR ₅₀ was thus greatest at the warmest temperature tested (33.7 C), indicating that T _o was ≥33.7 C. Original germination time courses (Figure 1) were normalized on a thermal-time basis and reproduced as cumulative normal curves (Figure 3). Prediction lines fit observed germination well, although slight deviations from the model were found at 19.3 C, which can be attributed to the proximity to the species’ estimated T _b (Steinmaus et al. Reference Steinmaus, Prather and Holt2000).

Figure 2.

Median germination rates (symbols) of four Cyperus difformis biotypes across a temperature range from 19 to 33.7 C. Bars represent standard errors based on three replicates of ~100 seeds. Coefficient of determination (R²) is >0.95 for all lines. See Figure 1 for biotype descriptions.

Figure 3.

Thermal-time model germination curves of four Cyperus difformis biotypes across six constant temperature regimes at 0 MPa. Cumulative observed (symbols) and predicted (lines) germination are plotted over growing degree days calculated for each biotype according to parameters in Table 1, and using the equation G = {log t _g −[log θ_T(50) − log(T − T _b)]/σ_θT(50)}. Bars represent SEs based on three replicates of ~100 seeds. See Figure 1 for biotype descriptions.

Germination and Osmotic Stress

Figure 4 shows observed germination time courses at a range of constant Ψ levels and test temperatures. A delay in germination was observed when seeds were placed at reduced Ψ relative to the time course in water, and germination of all biotypes was greatly affected by water stress. Temperature affected model parameters (Table 2) differently across biotypes, and a statistically significant interaction (P < 0.0109) was found between test temperature and median base water potential (Ψ_b(50)) estimations. Similar results have been reported by Kebreab and Murdoch (Reference Kebreab and Murdoch1999), in which Ψ_b values of Orobanche spp. seeds varied significantly with temperature. Importantly, unlike the present work, several other studies have reported relatively constant Ψ_{b (50)} values in the suboptimal range of T for a given plant species, that is, base water potential values do not commonly vary across a range of suboptimal test temperatures (Alvarado and Bradford Reference Alvarado and Bradford2002; Masin et al. Reference Masin, Loddo, Benvenuti, Zuin, Macchia and Zanin2010; Roman et al. Reference Roman, Thomas, Murphy and Swanton1999).

Figure 4.

Hydrotime-model germination curves of four Cyperus difformis biotypes across four constant moisture regimes at three test temperatures: 22.3 C, 26.45 C, and 29.8 C. Time courses of observed (symbols) and predicted germination (lines); predicted lines were obtained according to parameters in Table 2 and the population-based threshold model: probit(g) = [Ψ − θ_H/t _g(Ψ) − Ψ_b]/σ_Ψb, which predicts the germination time courses at each Ψ based upon a normal distribution of Ψ_b (g), and the hydrotime constant θ_H. Bars represent SEs based on four replicates of 100 seeds. See Figure 1 for biotype descriptions.

Table 2.

Base water potential (Ψ_b(50), MPa), hydrotime constant (θ_H, days), and standard deviation in base water potential within the seed population (σ_Ψb, MPa) of four Cyperus difformis biotypes, estimated using the probit method.

^a AS (American susceptible) and IS (Italian susceptible) are ALS-inhibitor susceptible smallflower umbrella sedge biotypes, whereas WA (Washington) and IR (Italian resistant) are ALS-inhibitor resistant biotypes.

^b Parameters derived from the equation probit(g) = [Ψ − θ_H/t _g − Ψ_b(50)]/σ_Ψb(50). The same lowercase letters within a column indicate means are not statistically different at P < 0.05. The same capital letters within a row indicate means for a given parameter are not statistically different at P < 0.05. Data were log transformed.

The ALS-R biotype WA had consistently lower (more negative) Ψ_{b (50)} values regardless of the test temperature, whereas IR had a significantly lower base water potential value in comparison with ALS-S biotypes at 26.5 and 29.8 C, suggesting greater capacity of ALS-R biotypes for germinating under drier conditions. General trends indicate C. difformis biotypes had a more negative Ψ_b(50) at 26.4 C, indicating greater ability to take up water and hence germinate at drier conditions around such a temperature. High temperatures are known to reduce water viscosity and therefore increase its diffusion, which could account for lower (more negative) Ψ_b values at 26.5 C rather than at 22.3 C.

The standard deviation in Ψ_b among seeds (σ_Ψb) was relatively constant with T (Table 2). Interestingly, ALS-R biotypes had significantly higher σ_Ψb values at 22.3 and 26.5 C, indicating greater variability in base water potential within the seed population, as was also found in regard to σ_θT(50). Both findings indicate less synchronous germination of ALS-R seeds. This constitutes an undesirable trait from a weed control point of view, because it is correlated with less uniform seedling emergence (Forcella et al. Reference Forcella, Arnold, Sanchez and Ghersa2000), which could hamper control using POST herbicides.

Model Validation and Emergence in Soils

Experimental temperature and moisture conditions recorded throughout the course of outdoor experiments are presented in Figures 5 and 6. Even though calculated Ψ_{b (50)} values for germination (Table 2) indicate seed germination can take place in somewhat dry conditions of up to −1 MPa, emergence of seedlings in both soils (HR and RD) was poor across treatments subjected to moisture stress (Figure 7). Across both soils, total C. difformis seedling density was higher at the flooded and saturated water regimes, which represents useful information from a C. difformis management point of view, because the goal of the stale seedbed technique is to maximize weed emergence before rice sowing.

Figure 5.

Environmental conditions during field soil emergence experiments expressed in days after watering (DAW). Water potential values (A) calculated as the average of moisture readings taken at 8 AM and 8 PM daily at the daily and 3-d flush treatments at each soil. Temperatures (C; chart B) represent 24-h averages of five replicates pooled for both soils, as they did not vary significantly between soils.

Figure 6.

Observed cumulative emergence (symbols) of Cyperus difformis seedlings expressed in days after watering (DAW), as affected by four water regimes in soil HR (A) and soil RD (B). Bars represent SEs based on five replicates. Soil HR = fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents with 36% clay and 1.8% organic matter; soil RD = Castro clay, fine, thermic Typic Calciaquolls with 28% clay and 2.8% organic matter.

Figure 7.

Cumulative observed emergence of Cyperus difformis seedlings, expressed in growing degree days for three water regimes in two soils: HR (A–C) and RD (D–F). Symbols indicate observed emergence; solid lines are the predicted germination; and dashed lines are predicted emergence. Emergence expressed as a fraction of final emergence within the treatment to better observe differences in emergence timing. Bars represent SEs based on five replicates. Soil HR = fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents with 36% clay and 1.8% organic matter; soil RD = Castro clay, fine, thermic Typic Calciaquolls with 28% clay and 2.8% organic matter.

Cyperus difformis seedling emergence under the daily water treatment was inconsistent across soils, indicating low tolerance to water stress, as expected for an aquatic species. Such results cannot be explained by differences in soil water potential values, which were similar between both soils throughout the course of the experiment (Figure 5A). Temperatures were also not significantly different between soils (Figure 5B). Furthermore, there was no C. difformis emergence at the 3-d flush treatment regardless of soil (Figure 6), because it produced dry conditions that prevented seed germination and emergence, as indicated by solid lines dropping well below −2 MPa on Figure 5.

Field test results (Figure 6A and B) indicated C. difformis seedling emergence is affected by water potential at a much greater magnitude than seed germination itself, thus not allowing for the use of the hydrotime model, because variations in soil Ψ greatly affected emergence of this species. Therefore, only data from water regimes not affected by or under low water stress (daily water, saturated, and flooded) were used to develop and validate a thermal time model capable of predicting C. difformis emergence in soils, given these allowed for successful emergence and establishment of this species. Flooded and water-saturated treatments were not subjected to any moisture stress whatsoever; under the daily water regime, water potential in both soils remained within 0.1 MPa of field capacity (Figure 5A), creating low moisture stress conditions (Singer and Munns Reference Singer and Munns2002).

In the absence of soil moisture stress, early growth phases of emergence are dependent on thermal time exclusive of hydrotime (Roman et al. Reference Roman, Thomas, Murphy and Swanton1999). Figure 7 shows cumulative emergence recorded throughout the course of 25 d from May to June 2009. Water regimes significantly differed in GDD until 50% emergence of seedlings (Table 3), and a significant interaction between water regime and soil was determined by ANOVA (P < 0.001). There was an average lapse of 64.5, 93.5, and 100.3 GDD from the start of the experiment until emergence (2.0-cm seedling height) for the water regimes: daily water, saturated, and flooded, respectively (Figure 7). Differences in emergence patterns and thermal times between anaerobic and aerobic water regimes were also found by Boddy et al. (Reference Boddy, Bradford and Fischer2012) while studying herbicide-resistant biotypes of the rice weed rice barnyardgrass [Echinochloa phyllopogon (Stapf) Koso-Pol.]. Emergence of C. difformis from soils subjected to anaerobic stress imposed by flooding required a greater number of GDD to take place relative to nonflooded, aerobic soils (Figure 7; Table 3). Indeed, previous studies found a significant reduction in shoot elongation in E. phyllopogon in response to anoxia due to reduced growth rates (Fox et al. Reference Fox, Kennedy and Rumpho1994); in turn, the daily water regime allowed for the lowest thermal time to emergence across all water regimes, which is expected, due to aerobic conditions for germination that favor growth. Moreover, differences in germination times due to the flooded condition could also play a role in delaying seedling emergence, which our germination experiments in the laboratory could not account for, because germination was performed under aerobic conditions.

Table 3.

Time to 50% emergence (growing degree days [GDD]) of Cyperus difformis seedlings across three water regimes, observed for soils HR and RD.

^a The same lowercase letters within a column indicate means are not statistically different at P < 0.05. The same capital letters within a row indicate means are not statistically different at P < 0.05.

^b Soil HR: fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents with 36% clay and 1.8% organic matter; soil RD: Castro clay, fine, thermic Typic Calciaquolls with 28% clay and 2.8% organic matter.

Model predictions scored well against observed emergence data in all moisture treatments (Figure 7; Table 4), although a better fit was observed for daily water and saturated treatments, which bear a closer resemblance to the aerobic conditions used for developing the germination model in the laboratory. Our models were more successful at predicting emergence in the upper fractions of the seed population at saturated and flooded treatments (Figure 7).

Table 4.

Validation measurements of predicted Cyperus difformis emergence in two field soils (HR and Rd) subjected to three irrigation regimes.

^a Soil HR: fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents with 36% clay and 1.8% organic matter; soil RD: Castro clay, fine, thermic Typic Calciaquolls with 28% clay and 2.8% organic matter.

^b RMSE is expressed in the same units as the observed quantities, and is calculated as {[Σ(O _i − P _i)²]/n}^0.5, where O _i is an observation and P _i the corresponding prediction (Mayer and Butler Reference Mayer and Butler1993; Spokas and Forcella Reference Spokas and Forcella2006).

^c For the validation measurement EF, 1.0 is the optimal value, while a score below 0 indicates the observed mean is a better predictor than the model; calculated as 1 − Σ(O _i − P _i)² / Σ(O _i − Ō)², where Ō is the mean of observed values (Mayer and Butler Reference Mayer and Butler1993; Legates and McCabe Reference Legates and McCabe1999; Spokas and Forcella Reference Spokas and Forcella2006).

^d This metric varies from 0.0 to 1.0, with higher values indicating greater agreement between predicted and observed values; calculated as 1 − Σ(O _i − P _i)² / Σ [(P _i − Ō) + (O _i − Ō)]² (Legates and McCabe Reference Legates and McCabe1999; Spokas and Forcella Reference Spokas and Forcella2006).

Overall, results indicate a greater ability of ALS-R C. difformis biotypes used in this study to successfully complete seed germination under drier conditions, as shown by their more negative base water potential values relative to ALS-S seeds (Table 2; Figure 4). In addition, ALS-R C. difformis was also shown to germinate more quickly at colder test temperatures, as indicated by greater germination rates (Figure 2), while also displaying significantly lower θ_T(50), therefore germinating faster than ALS-S biotypes due to the need for fewer GDD. Such findings indicate that the repeated use of ALS-inhibiting herbicides has selected for more aggressive and noxious ecotypes of C. difformis that bear no fitness penalty in this regard. Although specific mutations of the ALS gene are yet to be determined for the ALS-R C. difformis biotypes employed in the present study, alterations at their ALS-inhibitor target site have been positively confirmed (Pedroso et al. Reference Pedroso, Al-Khatib, Hanson and Fischer2017). Field experiments showed poor and inconsistent C. difformis emergence under water-deficit irrigation, suggesting that to optimize springtime seedling emergence for this species, soil moisture should be kept around field capacity. Furthermore, the values of modeling efficiency coefficients obtained are good indicators of the model’s goodness of fit, and suggest that growers could rely on the proposed mathematical model for decision-making guidance regarding the best moment for C. difformis control in rice fields through a better understanding of the dynamics of its germination and emergence.

In conclusion, this study is the first to report clear differences in base water potential values for germination between ALS-R and ALS-S biotypes and the germination process being completed more rapidly in ALS-R biotypes—a fact attributed mainly to their lower thermal-time requirements and greater germination rates relative to ALS-S C. difformis biotypes. This has important implications for developing management strategies to control this weed.