The response of weedy sunflower ( Helianthus annuus L . ) to nicosulfuron : an examination of vegetative parameters and acetolactate synthase activity

Genetic and morphological interpopulation variability of weed species is often responsible for variable responses to herbicides. As weedy sunflower, an invasive form of Helianthus annuus L., possesses high morphological and genetic variability, very different responses of its populations to herbicides can be expected. This species is one of the dominant weed species in row crops, including maize, in many European countries in which nicosulfuron is intensively used for weed control. There are little available data about the response of this sunflower form to nicosulfuron or of the interpopulation variability of its response to other herbicides. The responses of three weedy sunflower populations to nicosulfuron were studied in field dose-response experiments, and acetolactate synthase (ALS) enzyme activity at different herbicide concentrations was determined in vitro. Interpopulation variability in the response to nicosulfuron was confirmed. Populations WS2 and WS3 were more that 20-fold and 30-fold less susceptible to nicosulfuron, respectively, than population WS1, based on fresh weight, whereas the differences were not so prominent based on other parameters, including plant height, leaf area and ALS activity, and ranged from 2 to 12-fold.


INTRODUCTION
Helianthus annuus L. (family Asteraceae) is a species that occurs in many different forms: as "normal" crop plants, atypical plants known as "off-type" crops (which are the result of crossing sunflower hybrids with wild plants during seed production), wild sunflower (present in the area of origin of sunflower in America), volunteer plants (present in those areas where the cultivated sunflower was grown over the last one or two years), and weedy plants.The origin of weedy populations can be different: (i) the seeds of wild forms could be introduced unintentionally; (ii) during sunflower seed production, wild sunflowers could hybridize with crops, making crop-wild hybrids; (iii) weedy sunflower could result from the spontaneous evolution of its volunteer populations; and (iv) volunteer populations could hybridize with ornamental sunflowers grown in gardens [1].Wild, weedy and volunteer sunflower populations are of concern in several regions of the world because of their invasive capacity and crop interference [2][3][4][5][6].In America, where common sunflower is native, weedy forms of sunflower strongly affect the yield of crops such as soybean and maize [7], while volunteer plants have never been reported to constitute self-perpetuating populations, nor to cause serious agronomic problems [8].In Europe, where sunflower is not native, volunteers are commonly present in the fields [8].Furthermore, weedy populations are also present in some countries, including France [9], Spain [9,10], Hungary [11], Serbia [6,12], Croatia, Romania, etc. Muller et al. [1] have documented that the infestation of sunflower fields with weedy sunflower was between 13 and 27% in Spain, while the density of this weed in France reached 15 plants m -2 .The largest populations in Spain covered an area of about 1500 m 2 [10], while in Serbia prominent larger populations were recorded in the southern Srem (around 1000 ha of crop and noncrop fields) and southern Banat (around 7-8000 ha of crop and non-crop fields) areas in Vojvodina Province (northern Serbia) [6,12].These populations show typical wild traits in combination with domesticated traits, and they are morphologically clearly different from the volunteers from which they originate [9].Owing to their highly competitive ability, these populations can cause considerable yield losses.Studying the impacts of weedy populations on sunflower cultivation, Muller et al. [9] found that weedy sunflowers greatly decrease crop yield, causing losses of more than 50%.Also, they have been documented to decrease the yield of soybean by up to 97% [13] and of maize up to 64% [14].An additional problem with these populations can be the potential flow of the genes responsible for the tolerance to ALS-inhibiting herbicides from herbicide-tolerant sunflower hybrids to the weedy populations [15,16], leading to the development of resistant populations.
Nicosulfuron is a selective sulfonylurea herbicide for post-emergence control of sensitive perennial and annual grasses and some broadleaf weeds in corn [17,18].The mode of action of this herbicide is the inhibition of the enzyme acetolactate synthase (ALS), also known as acetohydroxyacid synthase (AHAS), which controls the synthesis of amino acids (leucine, isoleucine and valine) needed for protein biosynthesis.The amino acid substitution in ALS enzyme, represented by 28 possible substitutions identified in different weed species, is the main mechanism of target-site-based resistance to this group of herbicides [19,20,21].Also, a non-target-site-based mechanism of resistance is possible [22].Herbicides kill sensitive plants while resistant plants survive, leading to the development of resistant populations due to continuous and intensive herbicide use.Currently, weed resistance to nicosulfuron has been reported in 52 unique cases, with 23 species (12 dicots and 11 monocots) [19].
In America, many cases of common sunflower resistance to herbicides have been confirmed [23][24][25], but the response of European weedy sunflower populations to herbicides has rarely been studied [26,27].Although weedy sunflower is present in some European countries, including France, Spain, Hungary, Croatia, Romania and Serbia, there are little data about its populations and their response to herbicides.
Since weedy sunflower is one of the dominant weeds in many maize fields in Serbia, as well as in many other European countries with intensive sunflower production where the application of nicosulfuron for weed control is very common, it is useful to know its response to this herbicide.Božić et al. [27] have confirmed a reduction in the vegetative and generative production of this weed by the recommended rate of nicosulfuron application (40 g active ingredient (a.i.) ha -1 ).Also, Zollinger [28] has documented a 50-70% sunflower crop injury by nicosulfuron.Contrary to this, Brighenti et al. [29] found that nicosulfuron had no phytotoxic effect on sunflower crops.It may not be possible to extrapolate the results of these studies to the populations of weedy sunflower due to the high morphological and genetic variability of different sunflower forms [30][31][32][33].Owing to the distinct morphological and genetic variations of weedy sunflower, which include different proportions of cultivated and wild traits, a very different response of its populations to herbicides can be expected.Also, the popularity of herbicide-tolerant sunflower hybrids included in the Clearfield ® , Clearfield Plus ® and ExpresSun® systems [34,35] further increases the potential for obtaining very variable responses of populations due to the gene flow between hybrids and weedy populations [16,36].
Therefore, the objective of this study was to study the response of weedy sunflower to nicosulfuron, with special emphasis on the interpopulation variability of that response.

Seed collection
Seeds of three randomly selected weedy sunflower populations (WS1, WS2, WS3) were collected from maize fields near Belgrade (localities WS1 -Padinska Skela; WS2 -Surčin; WS3 -Surčin), Serbia.Seeds were collected at maturity at the beginning of September 2007.Conditions at all localities were similar, with sunny and dry weather.The precise history of herbicide application in the fields is unknown, but at all localities weed control usually includes herbicide application, with the main group of used herbicides being ALS inhibitors.The collected seeds were cleaned and stored at room temperature until use.

Dose response
A field dose-response experiment was performed in Padinska Skela in two consecutive years.The experimental layout was a completely randomized design, with four replicates.Young seedlings of three weedy sunflower populations were transplanted to the field from containers at the end of April.The plot size was 5×4.2 m, with a plant density of 5.7 m -2 (interrow spacing of 24 cm and the distance between rows being 70 cm).All the relevant information about the experiments is summarized in Table S1 and the meteorological conditions during the experiment are shown in Table S2.
Four-leaf-stage plants were treated with different rates of nicosulfuron (Motivell, 40 g ai L -1 , SC, BASF, Germany).Five herbicide rates (plus the untreated control) were used as follows: 10, 20, 40, 60 and 80 g a.i.ha -1 , with the recommended rate 40 g a.i.ha -1 .The herbicides were applied using a Neptune 15, Kwazar® knapsack sprayer, delivering 300 L ha -1 , equipped with RS-MM 110°/04 nozzles.Plant height, fresh weight and leaf area were measured 30 and 33 days after treatment (DAT) in the 1 st and 2 nd years, respectively.Plant height and fresh weight of the aboveground part of plants were measured in the field immediately after sampling, while leaf area was measured using a Delta-T leaf area meter (Delta-T Devices, Burwell, Cambridge, UK) in the laboratory about 4-6 h after sampling.

ALS assay
For the ALS assay, seeds were sown in pots (38 cm 2 surface area) containing a commercial potting mix (Flora Gard TKS1, Germany).The plants were grown in a controlled environment chamber with a 16/8 h (day/night) photoperiod of light 300 μEm -2 s -1 and temperature of 28°C, and were irrigated manually when needed.The in vitro ALS activity study was conducted according to a procedure described by Ray [38], with some modifications described by Božić et al. [39].This assay detects the production of acetoin from acetolactate, which is the ALS enzyme product.All procedures were performed at 4°C.Young leaves from two-pair-leaf-stage seedlings were harvested for ALS enzyme extraction.Two g of the plant tissue was pulverized using a cold mortar and pestle in 6 mL of extraction buffer (0.1 M potassium phosphate (pH 7.5) containing 1 mM sodium pyruvate, 0.5 mM thiamine pyrophosphate, 0.5 mM MgCl 2 , 10 μM flavin adenine dinucleotide (FAD) and 100 mL L -1 glycerol).The homogenates were filtered through a layer of cheese cloth and centrifuged at 14000 x g for 25 min.ALS contained in the supernatant was precipitated with ammonium sulfate and the solution was centrifuged at 14000 x g for 25 min.The supernatant was discarded and the pellets were resuspended in the resuspension buffer (0.12 M potassium phosphate (pH 7.0) containing 20 mM sodium pyruvate, and 0.5 mM MgCl 2 ).Protein concentrations of the crude enzyme extracts were determined by the method of Bradford [40], using bovine serum albumin for the standard curve.
ALS activity was assayed by adding 0.1 mL of enzyme preparation to 0.5 mL of the reaction mixture (20 mM potassium phosphate (pH 7.0) containing 20 mM sodium pyruvate, 0.5 mM thiamine pyrophosphate, 0.5 mM MgCl 2 , 10 μM FAD), containing increasing concentrations of the tested herbicide.Nicosulfuron was added at rates of 0, 0.01, 0.1, 1, 10, 100 μM to all populations.The reaction mixture was incubated at 30°C for 1 h and the reaction was terminated by the addition of 0.05 mL H 2 SO 4 (6 N).Then, 0.5 mL of creatine (5 g L -1 in water) was added and the solution was incubated for 15 min at 60°C.The added sulfuric acid terminated the ALS reaction and decarboxylated the enzyme product acetolactate to acetoin.Acetoin was detected as a colored complex (A 525nm ) formed after the addition of 0.5 mL α-naphthol (50 g L -1 freshly prepared in 2.5 N NaOH) and incubation at 60°C for 15 min.A standard curve was constructed using commercial acetoin.The experiment was conducted as a completely randomized design.Experiments for each combination of hybrid and herbicide were performed twice, with three replicates.

Data analysis
Dose response was used to analyze the following response parameters: plant height, fresh weight and leaf area from the field trials, as well as acetoin levels from the ALS assay.More specifically, dose-response analysis was based on the three-parameter log-logistic model: y=d/(1+exp(b(log(x) -e))), where x and y denote the rate applied and the resulting observed response, respectively.Parameter d denotes the average response for rate 0, whereas parameter e corresponds to the rate which reduces the response by 50%; parameter b is proportional to the slope of the dose-response curve at a rate equal to e, i.e. the larger the value of b, the steeper the dose-response curve [41,42].For each response parameter, simultaneous models for all three populations were considered.Field trials repeated in two consecutive years were analyzed using separate models for the two years due to differences in meteorological conditions (Table S2).Parameter estimates with corresponding estimated standard errors are summarized in tables and the fitted dose-response curves are shown with average response values in graphs.The delta method was used to calculate the standard errors of the ratios of GR 50 (the dose required to reduce growth by 50%, i.e.I 50 , which is the concentration required to inhibit activity by 50%) [41].The statistical analyses and visualization of the fitted curves were carried out using the statistical environment R and the extension package drc [41].

Response of weedy sunflower to nicosulfuron in the field
Dose-response experiments showed that the WS1, WS2 and WS3 weedy sunflower populations displayed very different responses to nicosulfuron (Fig. 1).Approximately 0.77-65.02g nicosulfuron ha -1 was required to reduce the vegetative parameters (plant height, fresh weight, leaf area) by 50%, depending on the parameter, population and year.Generally, the GR 50 values were less than the recommended rate, with the exception of plant height in populations WS2 and WS3 in both years.The differences between the populations were more pronounced for fresh weight.Namely, in the 1 st year it was observed that for a 50% reduction in fresh weight in the WS1 population, a rate of 0.77 g a.i.ha -1 of nicosulfuron was required.This rate is about 25-and 37-fold less than the required rates for the   1).In the 2 nd year, rates for all three populations were slightly higher: 0.91 g a.i.ha -1 for WS1, 21.18 g a.i.ha -1 for WS2 and 29.89 g a.i.ha -1 for WS3.The least sensitive parameter was plant height and the rates of nicosulfuron needed to reduce this parameter by half varied from 30.73 g a.i.ha -1 (WS1) to 62.59 g a.i.ha -1 (WS3), and from 34.88 (WS1) to 63.69 (WS3) in the 1 st and 2 nd years, respectively (Table 1).The rates that caused a 50% reduction in leaf area were from 1.52 g a.i.ha -1 (WS1) to17.99 g a.i.ha -1 (WS3), and from 1.68 g a.i.ha -1 (WS1) to 18.68 g a.i.ha -1 (WS3) in the1 st and 2 nd years, respectively.The ratio calculated from the GR 50 values (Table 2) revealed that the WS1 population was more sensitive than the other two populations, which was most noticeable in the fresh weight parameter.
In addition to the GR 50 values, the slopes of the dose-response curves (parameter b) varied considerably between parameters and populations (Table 1), but were significantly (P=0.0001)lower for WS1 when compared with WS2 and WS3 for all vegetative parameters in both experimental years.The differences between WS2 and WS3 were mainly insignificant (P>0.05), with the exception of fresh weight in the 2 nd year (P=0.0093).The average response for rate 0 depended on the parameter, as indicated by parameter d (Table 1).Thus, the untreated control did not differ (P>0.05) between WS1 and WS2, while significant differences (P=0.0001) were confirmed between WS1 and WS3 and between WS2 and WS3.

Response of weedy sunflower to nicosulfuron based on ALS activity
In vitro incubation of the ALS enzyme (Fig. 2) extracted from the leaves of three weedy sunflower populations, with nicosulfuron concentrations ranging from 0.01 to 100 µM, caused 40-97%, 28-96% and 19-97% inhibition of enzyme activity in the WS1, WS2 and WS3 populations, respectively.Nicosulfuron required for ALS I 50 (herbicide concentration required to inhibit the enzyme activity by 50%) was 0.03 µM, 0.11 µM and 0.17 µM for WS1, WS2 and WS3, respectively (Table 1).The level of ALS activity in the control was similar in all three populations (parameter d in Table 1).Also, the slopes of the dose-response curves (parameter b) were similar for WS2 and WS3, while for WS1 the slope was slightly lower (Fig. 2).Based on the ratio between I 50 values, ALS from WS1 was about 4-and 6-fold more sensitive to nicosulfuron than ALS from WS2 and WS3, respectively.Meanwhile, the sensitivity of ALS from WS1 and WS2 was similar (Table 2).

DISCUSSION
The results presented herein reflect a large range in the growth responses of weedy sunflower populations to nicosulfuron depending on population and growth parameter.Generally, nicosulfuron GR 50 values were greater for WS2 and WS3 than for WS1, and mostly greater for WS3 than for WS2.The most pronounced differences were obtained for fresh weight (1 st year:  25-fold (WS2/WS1) and 38-fold (WS3/WS1); 2 nd year: 23-fold (WS2/WS1) and 33-fold (WS3/WS1)), indicating that the WS1 population was much more susceptible to nicosulfuron than the other populations.Only 0.77 and 0.91 g a.i.ha -1 GR 50 values were estimated for this population based on fresh weight in two consecutive year experiments.The susceptibility of this population to nicosulfuron was a little lower than the susceptibility of Setaria faberi, whose GR 50 value based on plant dry weight was <0.5 g nicosulfuron ha -1 [43].However, this population was more susceptible to nicosulfuron than the population of Amaranthus retroflexus (GR 50 = 4.11 g a.i.ha -1 ) [44] and Xanthium strumarium (8 g a.i.ha -1 ) [45].Unlike WS1, the populations WS2 (GR 50 : 25 g a.i.ha -1 and 23 g a.i.ha -1 in 1 st and 2 nd year, respectively) and WS3 (GR 50 : 38 g a.i.ha -1 and 33 g a.i.ha -1 in 1 st and 2 nd year, respectively) were considerably less susceptible to nicosulfuron than the populations of the abovementioned S. faberi and A. retroflexus, while their susceptibility was similar to the one recorded for X. strumarium populations, studied by Božić et al. [45].Similarly, significant differences in herbicide (thifensulfuron, also an ALS inhibitor) responses were identified among the populations of Amaranthus rudis and A. tuberculatus [46].Explanations for such herbicide response variability were the genetic variability of the studied species due to their dioecious nature and the development of herbicide resistance.In addition to the reported differences between populations, differences in herbicide response were also visible between years, which was probably due to differences in environmental conditions, especially precipitation, during the period between sowing and sampling (the 1 st year was characterized by slightly higher precipitation than the 2 nd year).This concurs with previous studies demonstrating that environmental conditions significantly affect the activity and effects of foliageapplied herbicides [39,45,47].
The obtained results regarding in vitro ALS enzyme activity with I 50 values of WS2 (0.11 µM) and WS3 (0.17 µM) were similar to those for sensitive populations of Ambrosia artemisiifolia (I 50 =0.137µM), Avena fatua (I 50 =0.104µM), Panicum miliaceum (I 50 =0.135µM), A. retroflexus (I 50 =0.104µM) [48] and X. strumarium (I 50 =0.11µM and I 50 =0.16,depending on the population) [41] incubated with nicosulfuron, while this value for the WS1 population (0.03 µM) was significantly lower.These results prove that the differences in susceptibility of weedy sunflower populations to nicosulfuron are the result of an altered susceptibility of the ALS enzyme, which is a confirmed mechanism of weed resistance [49][50][51] or crop tolerance to ALS-inhibiting herbicides in many cases [39,52].Differences in response to nicosulfuron between populations WS1, WS2 and WS3 were most evident when plant fresh weight was compared.Plant weight is the most suitable parameter for studies of plant responses to herbicides and resistance confirmation [22,39,53], although in some cases resistance was confirmed based on plant height or leaf area [39,54].However, vegetative parameters cannot provide an explanation for the mechanism of resistance, and ALS enzyme activity studies are necessary.Given that the investigation of ALS activity is a more sensitive method than vegetative parameters, the ratio between the GR 50 / I 50 values for a resistant and sensitive population is usually higher for ALS activity [39,50].Contrary to this, the ratios between I 50 values of WS2 and WS1 and WS3 and WS1 were lower than the ratios between GR 50 values for the same populations based on fresh weight.Similarly, Hanson [49] obtained a much higher ratio between the population of Camelina macrocarpa resistant to chlorsulfuron and a sensitive population.A lower ratio for ALS activity points to a possible combination of two mechanisms of resistance -altered ALS enzyme and a metabolic mechanism; to confirm this, studies of nicosulfuron metabolism and molecular studies of ALS gene are necessary [22].
Previous research has shown that weed populations can vary greatly in their susceptibility to herbicides [46,55,56].A possible reason for variability in the herbicide response can be the geographical origin of populations, which might be related to the differences in herbicide use or genetic variation [55].On the other hand, populations from geographically close locations can show variability in their herbicide responses [56,57], which is in accordance with our results.Namely, determination of the origin of weedy sunflower is specific in comparison to many other weed species because the population genesis can be very different [1].Claerhout et al. [58] found a weak correlation between the morphological/genetic variations and herbicide sensitivity of Echinochloa crus-galli and E. muricata accessions, where interpopulation variability at the morphological and genetic levels could potentially promote variation in the herbicide (nicosulfuron and others) response [46].Also, the high ratios between WS1 and WS2 and WS3, based on GR 50 values for fresh weight, suggest the possibility of herbicide-resistance development.In order to clarify the reasons for variability in herbicide responses of weedy sunflower populations, studies of morphological and genetic variability, population origin and herbicide use history are necessary.

Fig. 2 .
Fig. 2. Response of weedy sunflower populations to nicosulfuron based on in vitro ALS activity.

Table 1 .
Parameters (±SE) of the log-logistic equation † used to calculate the rate of nicosulfuron application required for 50% reduction of plant height, fresh weight and leaf area (GR 50 ; g a.i.ha -1 ) and ALS activity (I 50 ) of weedy sunflower populations.

Table 2 .
GR 50 (growth reduction by 50%) ratio of vegetative parameters and ALS activity of three weedy sunflower populations in response to nicosulfuron.