Response of vetch , lentil , chickpea and red pea to pre-or post-emergence applied herbicides

Broad-leaved weeds constitute a serious problem in the production of winter legumes, but few selective herbicides controlling these weeds have been registered in Europe. Four field experiments were conducted in 2009/10 and repeated in 2010/11 in Greece to study the response of common vetch (Vicia sativa L.), lentil (Lens culinaris Medik.), chickpea (Cicer arietinum L.) and red pea (Lathyrus cicera L.) to several rates of the herbicides pendimethalin, S-metolachlor, S-metolachlor plus terbuthylazine and flumioxazin applied pre-emergence, as well as imazamox applied post-emergence. Phytotoxicity, crop height, total weight and seed yield were evaluated during the experiments. The results of this study suggest that common vetch, lentil, chickpea and red pea differed in their responses to the herbicides tested. Pendimethalin at 1.30 kg ha, S-metolachlor at 0.96 kg ha and flumioxazine at 0.11 kg ha used as pre-emergence applied herbicides provided the least phytotoxicity to legumes. Pendimethalin at 1.98 kg ha and both rates of Smetolachlor plus terbuthylazine provided the greatest common lambsquarters (Chenopodium album L.) control. Imazamox at 0.03 to 0.04 kg ha could also be used as early post-emergence applied herbicide in common vetch and red pea without any significant detrimental effect. Additional key words: Cicer arietinum L.; flumioxazin; imazamox; Lathyrus cicera L.; Lens culinaris Medik.; phytotoxicity; Vicia sativa L. * Corresponding author: vasilakoglou@teilar.gr Received: 10-02-13. Accepted: 30-09-13. Abbreviations used: AHAS (acetohydroxyacid synthase); ALS (acetolactate synthase); POST (post-emergence); PPO (protoporphyrinogen oxidase); PRE (pre-emergence); WAS (weeks after seeding). Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA) Spanish Journal of Agricultural Research 2013 11(4): 1101-1111 Available online at www.inia.es/sjar ISSN: 1695-971-X http://dx.doi.org/10.5424/sjar/2013114-4083 eISSN: 2171-9292


Introduction
Weeds constitute a serious problem in the production of winter legumes because they can compete for resources like light, nutrients, water and space, directly influencing legumes yield and standability.In some legume crops, which are poor competitors, weed competition can result in def icient establishment (Fraser et al., 2004;Fedoruk et al., 2011).Especially for broad-leaved weeds, their control is a major production problem in winter legumes due to the absence of selective herbicides for these crops (Malik & Waddington, 1989;Fraser et al., 2003).
Herbicides of the chloroacetamides (such as metolachlor), dinitroanilines (such as trifluralin and ethalfluralin) or imidazolinones (such as imazethapyr) have been evaluated for possible use in legumes (Friesen & Wall, 1986;Wilson & Miller, 1991;Wall, 1996;Fraser et al., 2003).However, these herbicides, which are not moreover registered in Europe for legumes, due to the European agri-environmental policy and the reduced interest of the chemical companies for these herbicides, did not control all key weeds or significantly injured legumes.
Only few selective herbicides are registered in Europe in winter legumes.Nevertheless, some herbicides have been shown to control several weeds that are problematic in these crops (Senseman, 2007).Many Greek legume producers have also expressed interest in using new herbicides (Dr.D. Vlachostergios, pers.comm.), but research reports on selective herbicides for legume crops are relatively limited in the literature.So, more research is needed to identify herbicides with an acceptable margin of crop safety that can be used for weed management in legumes production.Therefore, the objective of this study was to evaluate, under field conditions, the response (phytotoxicity, height, total weight and seed yield) of four winter legumes (common vetch, lentil, chickpea and red pea), widely cultivated in Europe, to different rates of the herbicides pendimethalin, S-metolachlor, S-metolachlor plus terbuthylazine, flumioxazin and imazamox.

Experimental site
Four field experiments were conducted in 2009/10 (season 1) and were repeated in 2010/11 (season 2) to determine the effects of application rate of pendime-

Treatments and experimental design
Nitrogen and phosphorous at 30 and 60 kg ha -1 , respectively, were incorporated before legumes seeding.The four legumes were seeded on 25 December, 2009 (season 1) and on 1 December, 2010 (season 2).Varieties Tempi of common vetch, Samos of lentil, Amorgos of chickpea and Rhodos of red pea were seeded at 180, 100, 160 and 150 kg ha -1 , respectively.The seeder (model D8-30 SPECIAL, Amazonen-Werke, Hasbergen-Gaste, Germany) was equipped with 15-cm diam disc openers and adjusted to seed at 2-3 cm depth in 16 cm rows.Legumes were seeded at 4-to 6-cm intervals in rows, reflecting the common practice on Greek legume fields.So, the total seed number per m 2 was about 133, 167, 111 and 133 for common vetch, lentil, chickpea and red pea, respectively.The previous crop was barley harvested in mid-June.Straw bales were removed after harvest and land was ploughed and left undisturbed during summer.The experimental area was cultivated twice (in late October and in mid November) with a harrow disk to prepare the soil for legumes seeding and to incorporate the fertilisers into the soil.Legumes seeds were not inoculated with rhizobia to encourage biological N 2 -fixation, because common vetch had been grown in the field recently.The experimental area was naturally infested with wild mustard, common field poppy (Papaver rhoeas L.) and common lambsquarters, as confirmed by visual assessments made during the previous growing season.
In each experiment (legume crop), a randomized complete block design was used with four replicates for each herbicide rate and for each control treatment (untreated weedy and untreated weed-free, as well as diuron).Fifty six plots were included in each experiment, with plot size of 2.5 × 4.0 m including 16 rows of crop (Fig. 2).All blocks were separated by 3-m buffer zone.Table 1 shows that all herbicide treatments consisted of two rates of pendimethalin, S-metolachlor, S-metolachlor plus terbuthylazine or flumioxazin applied in pre-emergence (PRE), as well as three rates of imazamox applied in post-emergence (POST).Diuron was applied PRE at 1.0 kg ha -1 .The PRE treatments were applied four days after seeding, while imazamox was applied about 10 weeks after seeding (WAS), when legumes were at the six-to eight-compound leaf growth stage.Also, the two untreated (weedy and weed-free) controls were included to assist in rating injury and yield loss due to herbicide treatments.In the weed-free plots, weeds were twice handremoved at 4 and 10 WAS.Herbicide applications were made by an air-pressurized hand-f ield plot sprayer (AZO-Sprayers, Ede, The Netherlands), with a 2.4 m wide boom fitted with six 8002 flat fan nozzles (Teejet Spray System Co., Wheaton, IL, USA), which was calibrated to deliver 300 L ha -1 of water at 280 kPa pressure.
100% (complete plant death) at 4, 12 and 16 WAS.The final height of the four crops was assessed at 18 WAS.During the first two weeks of June each year, when legumes were at the physiological maturity stage, plants located in the four centre rows of each plot (0.64 m × 4 m) were harvested by hand.Then, total fresh and dry weights, as well as the seed yield of each legume were recorded.For dry weight determination, plant fractions (1 kg) were air-dried in the shade for 3 days and oven-dried at 65°C for 24 h to a constant weight.

Statistical analysis
For each legume crop separately, plant number, % injury, total fresh and dry weights, as well as seed yield data were subjected to a combined over-season ANOVA.The MSTAT program (MSTAT-C, 1988) was used to analyze variances.Fisher's Protected LSD and Tukey's Honestly Significant Difference test procedures were used to detect and separate mean treatment differences at p = 0.05.

Weeds presence
In all experimental plots, weeds emerged at very low densities.The weed species with the greatest density, especially in lentil and chickpea, was common lambsquarters emerged in late winter.Wild mustard and common field poppy emerged at very low densities (1-to 2-plants m -2 ) (data not shown).All herbicide treatments significantly reduced common lamsquarters density, as compared with the untreated weedy control (Fig. 3).Pendimethalin at 1.98 kg ha -1 and both rates of S-metolachlor plus terbuthylazine provided the greatest common lambsquarters control.Pendimethalin at 1.30 kg ha -1 , both rates of metolachlor and flumioxazin, imazamox at 0.03 and 0.04 kg -1 and diuron slightly provided a lower weed control.However, imazamox at 0.02 kg ha -1 achieved partial common lambsquarters control.

Crops response
For the four legumes, crop injury and their yield components were affected by growing season (p < 0.001), herbi-cide treatments (p < 0.001) and growing season × herbicide treatments interaction (p < 0.001).So, for each crop separately, the growing season × herbicide treatments interaction means are presented in Tables 1, 2, 3 and 4.

Common vetch
In season 1, the PRE-applied herbicides had not any phytotoxic effect on common vetch.However, imazamox treatments caused slight crop injury which ranged from 5.0 to 16.3% at 12 WAS, without plant number reduction (data not shown).Visible injury caused by imazamox was present as signs of stem and leaf chlorosis (especially in the apex), as well as standing.These phytotoxicity symptoms decreased with time due to plant regrowth, so that at 16 WAS phytotoxicity ranged only from 0.0 to 7.5% (Table 1).Higher herbicide effects in season 2 were observed, in particular, for the greater rates of pendimethalin and S-metolachlor plus terbuthylazine (48.8-50%) at 4 WAS.At 16 WAS, crop injury caused by the greater rates of pendimethalin, S-metolachlor plus terbuthyazine and imazamox was lower, ranging from 17.5 to 21.3%.The other herbicide treatments caused crop injury which ranged from 0.0 to 7.5%.
In both growing seasons, common vetch height was slightly decreased (14 to 16% as compared with the weed-free control) by the two greater rates of imazamox (Table 1).The other herbicide treatments did not significantly affect common vetch height.
At harvest, common vetch total fresh (data not shown) and dry weights were not significantly affected by the herbicide treatments (Table 1).Weedy and weedfree treatments did not differ according to total fresh or dry weight.In season 1, seed yield did not differ among treatments.However, in season 2, common vetch seed yield in weedy treatment was slightly lower than that in weed-free and pendimethalin treatments.Generally, common vetch seed yield was greater in season 2 than in season 1.

Lentil
In season 1, the PRE herbicide treatments had not any phytotoxic effect on lentil (Table 2).On the contrary, imazamox treatments caused crop injury which  caused 21 to 27% lentil height reduction (Table 2).The other herbicide treatments did not significantly affect lentil height.At harvest in season 1, lentil total fresh and dry weights were not significantly affected by the herbicide treatments (Table 2).Weedy and weed-free treatments again did not differ, according to total fresh or dry weight.However, in season 2, total fresh and dry weights of lentil treated with the greater rates of S-metolachlor plus terbuthylazine or imazamox were significantly decreased as compared with weights achieved by the other treat-ments.In season 1, seed yield did not differ among treatments.However, in season 2, lentil seed yield in imazamox treatments of 0.03 and 0.04 kg ha -1 was lower by 41 to 55% than that in weed-free treatment.Lentil seed yield was greater in season 2 than in season 1.

Chickpea
In season 1, the PRE-applied herbicides had not any phytotoxic effect on chickpea (Table 3).On the contra-  In both growing seasons, imazamox rates of 0.03 and 0.04 kg ha -1 caused 19 to 23% chickpea height reduction (Table 3).The other herbicide treatments did not significantly affect chickpea height.Also, chickpea height was similar to both weedy and weed-free controls.
At harvest in season 1, chickpea total fresh and dry weights were not significantly affected by the herbicide treatments (Table 3).Fresh and dry weights in weedy treatment were slightly lower than those in weed-free treatment.In season 2, total fresh and dry weights of chickpea treated with imazamox were significantly decreased compared to other treatments.In both growing seasons, chickpea seed yield was significantly reduced by imazamox ranging from 27 to 47% and 51 to 63%, in season 1 and season 2, respectively.

Red pea
In season 1, all herbicide treatments had not any phytotoxic effect on red pea (Table 4).However, in season 2, all PRE herbicide treatments caused significant injury to red pea plants.In particular, the greatest crop injury (from 28.8 to 60.0% at 4 WAS) was caused by the higher rates of pendimethalin, S-metolachlor and S-metolachlor plus terbuthylazine.However, crop injury decreased with time and at 16 WAS was 28.8, 8.8 and 42.5%, respectively.The other herbicide PRE treatments caused crop injury which ranged from 0.0 to 25.0%.Imazamox treatments did not affect red pea growth.
In season 1, red pea height was not affected by herbicide treatments (Table 4).However, in season 2, the higher rates of S-metolachlor plus terbuthylazine and flumiozaxin, as well as diuron caused height reduction of red pea, which ranged from 27 to 42%, as compared with the weed-free control.
At harvest in season 1, red pea total fresh and dry weights were not significantly affected by the herbicide treatments (Table 4).Weedy and weed-free treatments did not differ according to total fresh or dry weight.However, in season 2, red pea total fresh and dry weights were significantly reduced by S-metolachlor plus terbuthylazine and diuron treatments.In season 1, red pea seed yield was not affected by herbicide treatments, while seed yield in weedy treatment was slightly lower than that in weed-free treatment.In season 2, S-metolachlor plus terbuthylazine and diuron treatments caused the greatest red pea seed yield reduction which ranged from 42 to 58%, as compared with that in weed-free control.

Weeds presence
In all experimental plots, weeds emerged at very low densities, maybe due to the late soil cultivation which had as result the control of annual winter weeds emerged until early November.The greatest rate of pendimethalin provided excellent common lambsquarters control, while S-metolachlor achieved lower control of this weed.Similarly, Chomas & Kells (2004) found that pendimethalin provided greater (98%) and more consistent control of common lambsquarters in maize (Zea mays L.) than metolachlor (66%).Soltani et al. (2007) also found poor control of common lambsquarters by the chloroacetamide dimethenamid applied in kidney beans, while Kantar et al. (1999) reported satisfactory control of this weed by the urea linuron or by the imidazolinone imazethapyr applied in preemergence in chickpea.

Crops response
In season 1, PRE herbicide treatments had not any phytotoxic effect on legumes.Friesen & Wall (1986) reported that lentil was tolerant to preplant soil incorporated metolachlor.However, significant plant injury was caused by the PRE-applied herbicides in season 2, maybe due to increased soil humidity during herbicide application and the increased rainfall during December and January after herbicide application (Fig. 1), resulting in reduced herbicide adsorption by soil colloids and increased herbicide leaching to crop root system depth (Monaco et al., 2002).This crop injury was reduced with time.Wall (1996) found that lentil tolerance to PRE-applied imazethapyr differed among years and crop injury tented to decrease between 2 and 4 weeks after treatment.According to the same researcher, the greater herbicide uptake by crop under conditions of increased and excessive soil moisture, as well as the higher precipitation for the week following herbicide application may account for the observed greater lentil injury.
The POST-applied recommended rate (0.04 kg ha -1 ) of imazamox caused significant legumes injury, especially to lentil and chickpea, visible as signs of stem and leaf chlorosis, as well as standing, but the phytotoxicity symptoms decreased with time due to plant regrowth.Similar phytotoxicity symptoms have been reported in lentil to imazethapyr (Wall, 1996).According to Bukun et al. (2012), the increased lentil susceptibility to imazamox, as compared with the other legumes, could be attributed to the combined effects of increased herbicide absorption and reduced metabolism.Chickpea injury caused by imazamox is in agreement with other studies reporting chickpea tolerance to imazamox and imazethapyr (Taran et al., 2010).
Generally, common vetch and lentil seed yields were greater in season 2 than in season 1, maybe due to greater rainfall recorded in April (period of legumes anthesis and fruit development) the second season (Fig. 1).For all legumes, weedy and weed-free treatments did not differ in most cases in terms of fresh or dry weight, due to low weed densities.
The results of this study suggest that common vetch, lentil, chickpea and red pea differ in their responses to herbicides tested.Pendimethalin at 1.30 kg ha -1 , S-metolachlor at 0.96 kg ha -1 and flumioxazin at 0.11 kg ha -1 in pre-emergence caused the least phytotoxicity to common vetch, lentil, chickpea or red pea.Also, imazamox at 0.03 to 0.04 kg ha -1 in post-emergence was slightly phytotoxic to common vetch or red pea.These treatments could be a chemical option to control serious weeds in legumes, without any yield lost.

Figure 1 .
Figure 1.Total monthly rainfall and mean monthly temperature during the experiments.

Table 2 .
Visual injury caused by various herbicide treatments, as well as height, total fresh weight, total dry weight and seed yield of lentil (Lens culinaris) grown in 2009/10 (season 1) and 2010/11 (season 2)Response of vetch, lentil, chickpea and red pea to herbicides 1107 with time due to plant regrowth, so that at 16 WAS phytotoxicity ranged only from 3.5 to 17.5% (Table2).In season 2, all herbicide treatments caused significant injury to lentil plants.In particular, the greater rate of flumioxazin, both rates of pendimethalin and S-metolachlor plus terbuthylazine, as well as diuron caused

Table 3 .
Visual injury caused by various herbicide treatments, as well as height, total fresh weight, total dry weight and seed yield of chickpea (Cicer arietinum) grown in 2009/10 (season 1) and 2010/11 (season 2) WAS: weeks after seeding.PRE: pre-emergence.POST: post-emergence.

Table 4 .
Visual injury caused by various herbicide treatments, as well as height, total fresh weight, total dry weight and seed yield of red pea (Lathyrus cicera) grown in 2009/10 (season 1) and 2010/11 (season 2)