Effect of tillage systems and herbicides on weed emergence, weed growth, and grain yield in dry-seeded rice systems

Field Crops Research 137 (2012) 56–69

Contents lists available at SciVerse ScienceDirect

Field Crops Research journal homepage: www.elsevier.com/locate/fcr

Effect of tillage systems and herbicides on weed emergence, weed growth, and grain yield in dry-seeded rice systems ˜ B.S. Chauhan ∗ , J. Opena International Rice Research Institute, Los Ba˜ nos, Philippines

a r t i c l e

i n f o

Article history: Received 27 June 2012 Received in revised form 23 August 2012 Accepted 24 August 2012 Keywords: Direct seeding Weed management Herbicide Asia

a b s t r a c t Farmers in many Asian countries are moving from puddled-transplanted rice to dry-seeded rice systems. Dry-seeded rice can be sown after land preparation or under zero-till conditions. Weeds, however, are the major constraint to the production of dry-seeded rice. A study was conducted during the wet season of 2011 and dry season of 2012 at the farm of the International Rice Research Institute to evaluate the effect of tillage systems (zero-till and conventional tillage) and herbicides on weed emergence, weed growth, and grain yield in dry-seeded rice systems. In the zero-till system, the densities of Digitaria ciliaris, Eleusine indica, Eclipta prostrata, and Ludwigia octovalvis increased many-fold from the first season to the second. The efficacy of herbicides (oxadiazon followed by fenoxaprop + ethoxysulfuron and oxadiazon followed by penoxsulam + cyhalofop) was lower in the zero-till system than in the conventional tillage system. Grain yield in the herbicide-treated and weed-free plots was similar between the tillage systems and this response was consistent during both the seasons. However, grain yield in the zero-till-control (one handweeded) plots was lower (0.9–1.5 t ha−1 ) than in the conventional tillage-control plots. The information gained from this study suggests that yields in zero-till systems similar to those in conventional-tilled systems can be achieved if weeds are effectively controlled. In the absence of effective weed control, zero-till systems may result in poor grain yield. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Rice is a principal source of food for more than half of the world’s population. More than 90% of this rice is grown and consumed in Asia. There it is commonly grown by manual transplanting of 3–6-week-old seedlings in puddled soils, resulting from intensive cultivation in wet soil conditions. Puddling creates a hard pan below the plow layer, leads to high water losses through surface evaporation, and influences soil health because of the dispersion of soil particles and compacting soil (Singh et al., 2008; Chauhan, 2012; Chauhan et al., 2012a). Puddling and transplanting operations require a huge amount of water and labor. In the future, many rice farmers in Asia may have limited access to irrigation water. Around one decade ago, it was predicted that, by 2025, 13 Mha of irrigated wetland rice in Asia may experience physical water scarcity and 22 Mha of irrigated dry-season rice may suffer from economic water scarcity (Tuong and Bouman, 2003). In addition, there is also a concern about labor shortage because of the increasing wage resulting from the migration of rural labor to

∗ Corresponding author at: Crop and Environmental Sciences Division, International Rice Research Institute, DAPO Box 7777, Metro Manila 4031, Laguna, Philippines. E-mail address: [email protected] (B.S. Chauhan). 0378-4290/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2012.08.016

the cities (Pandey and Velasco, 2005; Chauhan, 2012). It is difficult for farmers to acquire labor at the critical time of seedling transplanting. These studies indicate that there is a need to evaluate alternatives to puddled-transplanted rice before water and labor scarcities become more critical. In recent years, manual transplanting of rice in some Asian countries has been, or is being, replaced by dry-seeded rice as farmers respond to the decreased availability of labor or water or their increased costs (Pandey and Velasco, 2005). Dry-seeded rice crop has several advantages over transplanting rice. They are conducive to mechanization, are more rapidly and easily planted, are less labor intensive, use less water, and have fewer methane emissions (Farooq et al., 2011; Chauhan, 2012; Chauhan et al., 2012b). In dry-seeded systems, dry rice seeds are sown under zero-till (ZT) conditions or after tillage in a well-prepared seedbed. In addition to reducing labor and fuel costs, ZT systems may improve soil physical and chemical properties, conserve soil moisture, and reduce soil erosion (Triplett and VanDoren, 1977; Chauhan et al., 2007). Weeds, however, are the main constraint to the production of dry-seeded rice (Chauhan and Johnson, 2010; Chauhan, 2012; Chauhan et al., 2012a,b). The main reasons for the weed problem in dry-seeded rice systems are the absence of standing water at crop emergence to suppress weeds and the absence of a seedling size advantage between rice and weed seedlings as both emerge simultaneously in these production systems.

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Different tillage systems influence vertical weed seed distribution differently in the soil and this distribution of weed seed may influence the relative abundance of weed species in the field (Froud-Williams et al., 1981; Chauhan et al., 2006c; Chauhan and Johnson, 2009b). ZT systems, for example, leave most of the weed seeds near the soil surface after planting. The conditions for germination are favorable for weed seeds present on the soil surface. On the other hand, weed seeds present on the soil surface under ZT systems are prone to rapid desiccation and seed predation. With conventional tillage (CONT) systems, however, weed seedling emergence depends on the effect the tillage has on weed seed burial as seed buried too deep may not be able to emerge in these tillage systems. In a previous study in rainfed rice, seedling emergence of Ageratum conyzoides, Digitaria ciliaris, Echinochloa colona, Eclipta prostrata, Eleusine indica, and Portulaca oleracea was greater in a ZT system than in a CONT system (Chauhan and Johnson, 2009b); however, an artificial weed seed bank was used in this study. Information on the influence of tillage systems on the emergence pattern of different weeds from a natural seed bank is limited in the literature. Weeding in Asia is commonly done by labor; however, this is becoming less common because of the non-availability of labor at the critical time of weeding and high labor costs. In the absence of manual weeding, herbicides are widely used to control weeds in dry-seeded rice systems. The use of a single herbicide does not give effective weed control and may result in shifts in problematic weed species. The trend toward dry-seeded rice systems, either sown under ZT or after tillage, is likely to continue and herbicides are an important tool of weed management in these systems. Therefore, there is a need to evaluate the performance of different herbicides that can provide effective weed control under different tillage systems. The information gained from such studies would help in deciding on the best weed management options for these systems. A study was conducted at the farm of the International Rice Research Institute to evaluate the influence of tillage systems and herbicides on weed emergence, weed growth, and grain yield in dry-seeded rice systems.

2. Materials and methods 2.1. Description of the experiment A field study was conducted during the wet season of 2011 (WS11) and dry season of 2012 (DS12) at the Research Farm of ˜ the International Rice Research Institute, Los Banos, Philippines. The soil at the site had a pH of 6.5, sand of 29%, silt of 44%, and clay of 27%. The study had two tillage systems: CONT and ZT. Two pre-sowing cultivations to a depth of 10–15 cm were given to the CONT plots, while soil disturbance in the ZT plots was limited to the sowing operation only. The ZT plots were sprayed 3–4 days before crop sowing with glyphosphate at 0.75 kg a.e. ha−1 . The ZT crop was sown into anchored rice stubbles of 15–20-cm height and there was no loose residue in the field. Rice (variety Rc222) was dry-seeded at 50 kg ha−1 in both tillage systems with a combine drill fitted with knife-point openers. Rice was sown in rows 20 cm apart on July 1, 2011, and January 7, 2012. A complete fertilizer (14:14:14 of N:P2 O5 :K2 O) was drilled with the seed at 30 kg ha−1 for each nutrient. After crop emergence, N was applied as urea at 150 kg ha−1 in four split doses: 45 kg N ha−1 at 28 days after sowing (DAS), 30 kg N ha−1 at 42 DAS, 45 kg N ha−1 at 60 DAS, and 30 kg N ha−1 at 90 DAS. The field was surface-irrigated immediately after sowing. Each tillage system had four weed control treatments: oxadiazon (0.75 kg a.i. ha−1 ) applied at 2 DAS followed by

57

fenoxaprop + ethoxysulfuron (0.045 kg a.i. ha−1 ) applied at 28 DAS, oxadiazon (0.75 kg a.i. ha−1 ) applied at 2 DAS followed by penoxsulam + cyhalofop (0.072 kg a.i. ha−1 ) applied at 28 DAS, control (one hand-weeding), and weed-free. In the weed-free plots, weeds were removed by hand every week until 70 DAS. In the other three weed control treatments, a hand weeding was performed at 42 DAS, including in the control plots, where weeds were allowed to grow before and after the hand weeding. In completely weedy plots, yield losses in dry-seeded rice systems are greater than 90% (Chauhan and Johnson, 2011b). In addition, it is not common for farmers to leave their rice fields infested with weeds in irrigated areas. The herbicides were applied with a knapsack sprayer that delivered around 320 L ha−1 spray solution through flat fan nozzles at a spray pressure of 140 kPa. In both seasons, the area of each subplot was 44 m2 (10.0 m × 4.4 m). The experiments in both seasons were arranged in a split-plot design, with tillage system as the main plots and weed control methods as the subplots. There were four replications in each season, but one replication in the WS11 had very poor crop emergence due to prolonged standing water after crop sowing. Therefore, data from only three replications were used in the WS11. 2.2. Effect of tillage systems on weed seedling emergence pattern In the control plots of each tillage system, two quadrats of 20 cm × 20 cm were prepared immediately after crop sowing. In these quadrats, seedling emergence of important weed species was recorded at 7, 14, 21, 28, and 35 DAS and expressed as plants m−2 . 2.3. Effect of tillage systems and herbicides on weed density, weed biomass, and grain yield The efficacy of herbicides on different weeds was evaluated at 26 DAS (2 days before the spray of post-emergence herbicides) and 42 DAS (14 d after the spray of the post-emergence herbicides and before the hand weeding). At each sampling time, two quadrats of 40 cm × 40 cm were placed in each plot at random to determine the density and biomass of different weeds. At crop harvest, only total weed biomass was determined from the same number and size of quadrats. Biomass was determined after drying the samples at 70 ◦ C for 48 h. The crop was harvested in the last week of October 2011 and first week of May 2012. The harvested area for grain yield was 24 m2 in WS11 and 10 m2 in DS12. Grain yield was converted to t ha−1 at 14% moisture content. 2.4. Statistical analyses Data were analyzed using ANOVA to evaluate differences between treatments, and the means were separated using LSD at P = 0.05 (GenStat 8.0, 2005). Weed density data were square-root √ transformed ( x + 0.5) due to high variance, whereas transformation did not improve the homogeneity of weed biomass and grain yield data. The data on weed seedling emergence pattern in the control plots were presented using standard error of means. 3. Results 3.1. Effect of tillage systems on weed seedling emergence pattern The experimental plots contained many weed species: Cyperus iria, C. rotundus, D. ciliaris, E. colona, E. indica, E. prostrata, Ludwigia octovalvis, P. oleracea, Trianthema portulacastrum, Dactyloctenium aegyptium, Murdannia nudiflora, Lindernia anagallis, Boerhavia

58

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Conventional tillage 400 C. iria (WS11)

Zero-till C. iria (DS12)

300 200 100 0 80

D. ciliaris (WS11)

D. ciliaris (DS12)

E. colona (WS11)

E. colona (DS12)

E. indica (WS11)

E. indica (DS12)

60

Density (plants m-2)

40 20 0 140 120 100 80 60 40 20 0 200 150 100 50 0 0

7

14

21

28

35 0

7

14

21

28

35

Days after sowing Fig. 1. Effect of tillage systems (conventional tillage and zero-till) on emergence pattern of different weeds in the wet season of 2011 (WS11) and dry season of 2012 (DS12). The vertical bars represent standard error of means.

erecta, Cleome rutidosperma, Leptochloa chinensis, Fimbristylis miliacea, Phyllanthus niruri, Panicum brevifolium, Hedyotis corymbosa, H. biflora, Euphorbia spp., and Chloris barbata. However, the results are presented only for the first eight weed species as they were dominant in the samples. The effect of tillage systems on the emergence of different weed species was different in different seasons (Figs. 1 and 2). The emergence of C. iria, for example, was greater in the CONT system in the WS11, but, in the DS12, emergence was similar between the tillage systems (Fig. 1). D. ciliaris and E. indica emergence was very low in the WS11; however, the emergence of both species increased in the DS12 (Fig. 1). Similar results were found for E. prostrata and L. octovalvis (Fig. 2). Another interesting observation for these two species was that they did not occur or

occurred at a very low density in the first season in the ZT plots, but their densities increased to significant numbers in the second season. P. oleracea densities were greater in the CONT plots, but, because of very high variability, the differences were statistically nonsignificant between the tillage systems during DS12 (Fig. 2). The density of T. portulacastrum was similar between the tillage systems in the WS11; however, in the DS12, the density was greater in the CONT plots than in the ZT plots (Fig. 2). The results of seedling emergence over time (DAS) also showed that not all emerged seedlings resulted in established plants. Some seedlings die after emergence. This is clearly indicated by the results of C. iria, D. ciliaris, E. colona, and P. oleracea, in which seedling emergence reached a maximum density and then declined (Figs. 1 and 2).

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

59

Conventional tillage

Zero-till

E. prostrata (DS12)

300 E. prostrata (WS11) 250 200 150 100 50 0 L. octovalvis (WS11)

L. octovalvis (DS12)

P. oleracea (WS11)

P. oleracea (DS12)

150

-2 ) Density (plants m

100 50 0 100 80 60 40 20 0 350 T. portulacastrum (WS11)

T. portulacastrum (DS12)

300 250 200 150 100 50 0 0

7

14

21

28

350

7

14

21

28

35

Days after sowing Fig. 2. Effect of tillage systems (conventional tillage and zero-till) on emergence pattern of different weeds in the wet season of 2011 (WS11) and dry season of 2012 (DS12). The vertical bars represent standard error of means.

3.2. Effect of tillage systems and herbicides on weed density and biomass More than 20 weed species were present in the samples; however, results are presented only for the 12 dominant weed species (Tables 1–8). The interaction effects of tillage and weed control treatments were nonsignificant for some of the weed species; however, for clarity and simplicity, the results for all weed species are presented in a two-way form (tillage systems × weed control treatments). The interaction effect of tillage systems and weed control treatments before the application of post-emergence herbicides on individual weed density was significant for only a few weed species and the response was different in different seasons (Tables 1 and 2).

However, weed density was greatly influenced by the weed control treatments. In the WS11, irrespective of the tillage systems, the application of pre-emergence herbicide (i.e., oxadiazon) provided excellent control of C. iria and T. portulacastrum (Table 1). The results were more evident in the second season, in which the pre-emergence herbicide application provided effective control of many individual weed species, including C. iria, L. chinensis, P. oleracea, and T. portulacastrum (Table 2). A few weed species were not controlled well in the ZT system. They were mainly D. aegyptium, D. ciliaris, E. prostrata, and E. indica. In the WS11, total weed density was similar between the tillage systems, suggesting that oxadiazon had a similar effect on weed density under CONT and ZT systems (Table 3). However, in the DS12, there was an interaction between tillage and weed control treatments (Table 3). In

60 Table 1 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed density at 26 days after sowing (before the application of post-emergence herbicides) in the wet season of 2011. Weed √ density data were subjected to square-root ( [x + 0.5]) transformation before analysis and original values of weed emergence are shown in parentheses. Weed density (number m−2 )

CONT

ZT

CONT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

2.70 (15.0)

1.30 (2.0)

3.53 (12.5)

1.34 (2.1)

1.34 (2.1)

4.65 (39.6)

1.90 (6.0)

0.70 (0.0)

4.62 (22.9)

3.88 (25.0)

0.71 (0.0)

2.15 (8.3)

4.70 (25.0)

0.71 (0.0)

1.34 (2.1) 4.28 4.11

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

11.60 (200.0) 9.30 (98.0) 7.62 8.93

2.15 (8.3)

CONT

5.82 3.23

ZT

CONT

ZT

CONT

Echinochloa colona

Eclipta prostrata

0.71 (0.0)

0.71 (0.0)

2.35 (10.4)

3.25 (22.9)

1.93 (6.2)

0.71 (0.0)

0.71 (0.0)

2.56 (8.3)

2.78 (10.4)

5.62 (31.3)

0.71 (0.0)

1.34 (2.1)

3.32 (14.6)

5.99 (35.4)

6.16 (39.6)

2.53 (12.5)

3.95 (16.7) 2.43 2.86

ZT

CONT

5.04 3.85

ZT

3.16 (12.5) 3.02 3.63

Weed density (number m−2 )

CONT

Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

ZT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Eleusine indica

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portuzlacastrum

0.71 (0.0) 0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.34 (2.1)

0.71 (0.0)

1.34 (2.1)

2.60 (6.3)

2.30 (6.3)

1.34 (2.1)

0.71 (0.0)

1.34 (2.1) 1.93 (6.3)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.34 (2.1)

2.93 (8.3)

2.56 (8.3)

0.71 (0.0)

3.80 (18.8)

2.15 (8.3) 3.49 (16.7) 3.27 3.92

1.34 (2.08)

0.71 (0.0)

2.30 (6.3)

6.21 (58.3)

2.64 (8.3)

1.93 (6.2)

8.03 (70.8)

0.82 0.82

4.56 (29.2) 2.85 3.46

3.27 (10.4) 4.20 4.47

2.62 2.88

9.11 (83.3) 3.51 3.55

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Weed control treatment (WCT)

Table 2 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed density at 26 days after sowing (before the application of post-emergence herbicides) in the dry season of 2012. Weed √ density data were subjected to square-root ( [x + 0.5]) transformation before analysis and original values of weed emergence are shown in parentheses. Weed density (number m−2 )

CONT

ZT

Weed control treatment (WCT)

CONT

ZT

CONT

ZT

CONT

ZT

ZT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

5.96 (45.3)

5.46 (34.4)

0.71 (0.0)

3.61 (14.1)

0.71 (0.0)

2.74 (9.4)

0.71 (0.0)

2.41 (9.4)

7.20 (66.0)

12.50 (205.0)

1.43 (3.1)

0.71 (0.0)

6.94 (48.4)

2.66 (10.9)

0.71 (0.0)

1.43 (3.1)

0.71 (0.0)

5.68 (48.4)

0.71 (0.0)

3.86 (15.6)

2.80 (14.0)

22.10 (628.0)

7.25 (60.9)

6.74 (46.9) 4.34 3.97

6.67 (46.9)

1.63 (4.7) 3.60 4.04

0.71 (0.0)

3.30 (10.9) 1.61 0.88

5.50 (40.6)

Echinochloa colona

CONT

1.18 (1.6)

4.86 (34.4) 4.59 5.14

Eclipta prostrata

11.86 (143.8) 8.03 (70.3) 2.95 2.17

10.70 (114.0)

14.50 (214.0) 9.84 10.45

Weed density (number m−2 )

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

0.71 (0.0)

6.64 (58.0)

0.71 (0.0)

1.18 (1.6)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.18 (1.6)

1.43 (3.1)

4.93 (31.2)

1.18 (1.6)

0.71 (0.0)

0.71 (0.0)

5.14 (28.0)

0.71 (0.0)

1.90 (4.7)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.18 (1.6)

2.99 (14.1)

1.63 (4.7)

1.18 (1.6)

Eleusine indica Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

ZT

3.26 (29.7)

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

8.66 (80.0)

10.40 (130.0) 4.58 5.46

4.27 (25.0)

6.31 (42.2) 2.45 2.72

1.18 (1.6)

3.60 (15.6) 1.61 1.42

5.93 (50.0)

5.41 (34.4) 3.37 3.48

1.18 (1.6)

2.63 (7.8) 3.28 2.97

Trianthema portulacastrum

10.35 (109.4)

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Weed control treatment (WCT)

4.75 (23.4) 2.74 1.41

61

62

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Table 3 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on total weed density at 26 and 42 days after sowing in the wet and dry √ seasons. Weed density data were subjected to square-root ( [x + 0.5]) transformation before analysis and original values of weed emergence are shown in parentheses. Total weed density (number m−2 )

Weed control treatments (WCT)

CONT

ZT

CONT

Wet season 26 days after sowing (2 days before post-emergence spray) Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

ZT

Dry season

7.2 (54) 6.9 (50) 19.2 (381)

7.9 (77) 10.0 (110) 19.5 (396)

11.6 (145) 8.4 (73) 26.4 (698)

18.2 (366) 25.4 (758) 27.1 (758)

5.6 6.8

42 days after sowing (14 days after post-emergence spray) Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

9.3 10.9

6.7 (46) 8.2 (71) 27.3 (758)

11.1 (127) 14.7 (219) 18.3 (344)

11.4 (148) 10.0 (100) 34.7 (1234)

9.6 (97) 13.7 (195) 35.3 (1259)

5.0 5.8

6.7 6.0

the control (one hand-weeded) plots, total weed density was similar between the tillage systems, whereas total weed density in the herbicide-treated plots was greater in the ZT system than in the CONT system. The results for the effect of tillage and weed control treatments on biomass of individual weed species at 26 DAS are shown in Tables 4 and 5. The results for biomass were similar to those observed for weed density. Oxadiazon, for example, greatly suppressed growth of C. iria and T. portulacastrum compared with the control treatment (Table 4). In the DS12, the biomass of D. aegyptium, D. ciliaris, E. colona, E. prostrata, and E. indica was not suppressed in the ZT system by the application of pre-emergence herbicide (Table 5). The interaction effect of tillage systems and weed control treatments was nonsignificant on total weed biomass in the WS11, whereas the interaction effect was significant in the DS12 (Table 6). In the WS11, oxadiazon provided similar weed control between the tillage systems; however, in the DS12, weed suppression (represented by weed biomass) was greater in the CONT system (82–88%) than in the ZT system (41–42%).

Overall, at 42 DAS (14 days after the application of postemergence herbicides), the densities of individual weed species were similar in the two different post-emergence herbicide-treated plots (Tables 7 and 8). Irrespective of the weed control treatments, some of the weed species had greater densities in the ZT system than in the CONT system. There was no interaction effect of tillage and weed control treatments on total weed density at 42 DAS (Table 3). However, the results clearly showed that total weed density in the control plots increased in the second season and this was consistent in both tillage systems. Total weed density was similar between the tillage systems when weeds received different post-emergence herbicides. Because of the presence of many weed species, the interaction effect of tillage and weed control treatments on biomass was nonsignificant for most of the individual weed species (Tables 9 and 10). However, the effect of post-emergence herbicides was very apparent on some of the weed species and not on others. In both seasons, for example, herbicides in both tillage systems provided effective

Table 4 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed biomass at 26 days after sowing (before the application of postemergence herbicides) in the wet season of 2011. Weed control treatment (WCT)

Weed biomass (g m−2 )

CONT

ZT

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

Echinochloa colona

Eclipta prostrata

2.85

0.21

0.62

0.02

0.10

8.60

0.00

0.00

1.48

2.42

0.29

0.00

0.15

0.00

1.00

1.88

0.00

1.30

0.00

0.77

0.31

2.81

0.00

0.06

4.25

0.38

1.98

0.00

0.00

1.58

0.52

3.44

4.23

0.21

0.23 0.46 0.56

ZT

CONT

5.94 5.50 6.54

3.36 2.28

10.92 9.89

1.75 2.08

5.47 3.91

Weed biomass (g m−2 )

CONT

Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Eleusine indica

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portulacastrum

0.00

0.00

0.00

0.00

0.00

0.08

0.00

0.23

1.10

0.35

0.04

0.00

0.02

0.71

0.00

0.00

0.00

0.00

0.00

0.06

0.15

0.17

0.00

0.42

0.77

0.02

0.00

0.06

0.42

0.29

1.06

0.33

0.12

6.85

6.27 5.06 5.27

0.17 1.29 1.52

0.03 0.03

0.34 0.40

0.90 1.00

0.78 0.92

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

63

Table 5 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed biomass at 26 days after sowing (before the application of postemergence herbicides) in the dry season of 2012. Weed control treatment (WCT)

Weed biomass (g m−2 )

CONT

ZT

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

ZT

CONT

ZT

CONT

ZT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

CONT

ZT

CONT

Echinochloa colona

ZT

Eclipta prostrata

1.58

0.16

2.70

3.91

0.00

8.91

0.00

1.60

0.00

2.60

2.00

7.10

0.27

0.00

3.61

2.59

0.00

1.25

0.00

10.90

0.00

5.10

0.70

13.70

3.66

3.47

0.56

0.00

1.31

2.20

15.30 8.80 12.53 9.10

4.20

2.61 4.41 4.15

4.83 4.63

6.41 5.40

2.60 9.65 10.64

9.60 9.61 9.97

Weed biomass (g m−2 )

CONT

Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Eleusine indica

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portulacastrum

0.00

10.50

0.00

1.45

0.00

0.00

0.00

0.31

1.00

3.80

0.02

0.00

0.00

5.60

0.00

0.55

0.00

0.00

0.00

0.00

0.05

0.66

0.05

0.06

8.70

1.00

3.58

0.06

0.50

2.08

4.47

0.06

1.44

5.97

2.44 1.18 1.08

2.00 10.83 12.05

2.09 2.44

0.36 0.34

control of C. iria, E. indica, and P. brevifolium. Biomass of D. aegyptium in both seasons was not suppressed by the application of penoxsulam + cyhalofop in the ZT system. Similarly, E. colona under the ZT system was not well controlled by the post-emergence herbicides and this was consistent in both seasons (Tables 9 and 10). Another interesting observation was made in the control plots in different tillage systems, i.e., biomass of L. chinensis was greater in the ZT system, whereas biomass of T. portulacastrum was greater in the CONT system. The efficacy of post-emergence herbicides was consistent in both seasons. In the WS11, for example, weed control (represented by weed biomass) by herbicide application was greater in the CONT system (80–96%) than in the ZT system (50–61%) (Table 6). A similar trend was observed in the DS12; however, weed control in both tillage systems was better in this season compared to WS11. In the CONT system, weed biomass was 94–95% lower in the herbicide-treated plots than in the control plots, whereas weed biomass in the ZT plots was 73–84% lower in the herbicide-treated plots than in the control plots.

2.08 1.53

3.75 3.58

Pre- and post-emergence herbicide combinations did not control all weeds effectively in dry-seeded rice systems; therefore, one hand weeding was performed at 42 DAS. Even after one hand weeding, the herbicide-treated plots produced a significant amount of weed biomass (Table 11). Compared with the control plots, the biomass in the herbicide-treated plots at crop harvest was 55–67% and 67–82% lower in the WS11 and DS12, respectively. 3.3. Effect of tillage and weed control treatments on rice grain yield Grain yield was significantly influenced by the interaction between tillage systems and weed control treatments in both seasons (Table 12). Grain yield in the weed-free and herbicide-treated plots was similar between tillage systems, whereas grain yield in the control plots was greater in the CONT system than in the ZT system. In the WS11, yield in the weed-free plots in both tillage systems was 3.63 t ha−1 , which was statistically similar to the yield

Table 6 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on total weed biomass at 26 and 42 days after sowing in the wet and dry seasons. The values in parentheses represent percentage decrease in total weed biomass in the herbicide-treated plots relative to the control plots. Weed control treatments (WCT)

Total weed biomass (g m−2 ) CONT

ZT

Wet season 26 days after sowing (2 days before post-emergence spray) Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) 42 days after sowing (14 days after post-emergence spray) Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

ZT

Dry season

6.5 (66) 1.6 (92) 19.3

1 2.0 (42) 8.2 (61) 20.8

7.3 (82) 5.0 (88) 41.4

14.9 14.6 5.6 (96) 13.8 (80) 137.0

28.2 31.9 40.7 (61) 52.4 (50) 103.8

65.6 71.6

40.3 (42) 41.0 (41) 69.5

8.3 (95) 9.1 (94) 155.5

38.6 (84) 65.5 (73) 245.6 73.5 69.8

64 Table 7 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed density at 42 days after sowing (14 d after the application of post-emergence herbicides) in the wet season of 2011. Weed √ density data were subjected to square-root ( [x + 0.5]) transformation before analysis and original values of weed emergence are shown in parentheses. Weed density (number m−2 )

CONT

ZT

CONT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

1.34 (2.0)

0.71 (0.0)

2.78 (10.4)

3.80 (18.7)

0.71 (0.0)

2.53 (12.5)

0.71 (0.0)

1.93 (6.3)

0.71 (0.0)

2.35 (10.0)

2.56 (8.3)

4.29 (25.0)

1.34 (2.1)

2.78 (10.4)

1.34 (2.1)

0.71 (0.0)

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

16.57 (304.0) 4.23 (31.0) 5.85 7.11

4.35 (25.0) 4.02 4.79

CONT

3.27 (10.4)

ZT

1.93 (6.2) 3.69 3.25

CONT

5.14 (29.2)

ZT

CONT

ZT

CONT

2.30 (6.0)

6.83 (48.0)

2.30 (6.3)

1.34 (2.1)

4.98 (25.0)

7.82 (65.0)

2.69 (14.6)

2.15 (8.3)

11.88 (154.0) 6.01 (50.0) 5.33 6.31

4.94 (25.0)

3.75 (14.6) 3.59 4.32

Eclipta prostrata

Echinochloa colona

4.35 (18.8) 3.37 2.39

ZT

Weed density (number m−2 )

ZT

CONT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portulacastrum

1.34 (2.1)

1.34 (2.1)

0.71 (0.0)

1.34 (2.1)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.34 (2.1)

3.79 (14.6)

4.94 (25.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.93 (6.3)

0.71 (0.0)

2.56 (8.3)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.34 (2.1)

3.87 (14.6)

7.08 (68.8)

1.34 (2.1)

1.67 (4.2)

4.84 (29.2)

5.89 (35.4)

4.12 (18.7)

6.34 (41.7)

4.51 (27.1)

5.08 (43.8)

5.78 (33.3) 3.92 4.66

Eleusine indica Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

3.58 (16.7)

ZT

6.03 (37.5)

4.94 (25.0) 2.66 2.71

2.56 (8.3)

4.63 (29.2) 3.60 3.76

5.60 (35.4) 2.66 3.09

2.35 2.10

4.92 5.07

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Weed control treatment (WCT)

Table 8 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed density at 42 days after sowing (14 d after the application of post-emergence herbicides) in the dry season of 2012. Weed √ density data were subjected to square-root ( [x + 0.5]) transformation before analysis and original values of weed emergence are shown in parentheses. Weed density (number m−2 )

CONT

ZT

Weed control treatment (WCT)

CONT

ZT

CONT

ZT

CONT

ZT

CONT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

6.90 (50.0)

2.16 (6.3)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

3.90 (23.4)

1.20 (2.0)

2.10 (5.0)

7.20 (77.0)

2.90 (12.0)

0.71 (0.0)

0.71 (0.0)

5.95 (35.9)

3.79 (23.4)

0.71 (0.0)

1.90 (4.7)

0.71 (0.0)

2.19 (10.9)

0.71 (0.0)

7.10 (80.0)

5.40 (37.0)

3.70 (25.0)

10.96 (138.0) 4.24 4.69

7.03 (50.0)

10.40 (109.0)

10.40 (116.0)

ZT

CONT

7.81 (84.0)

2.52 (9.4) 3.70 3.24

2.99 (10.9)

3.35 (10.9) 1.35 1.35

6.68 (68.8)

Echinochloa colona

ZT

0.71 (0.0)

3.60 (21.9) 5.59 5.26

14.00 (295.0)

Eclipta prostrata

9.1 8.3

14.10 (336.0) 8.26 9.67

Weed density (number m−2 )

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portulacastrum

0.71 (0.0)

3.90 (19.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.43 (3.1)

1.43 (3.1)

1.18 (1.6)

3.50 (15.6)

1.43 (3.1)

2.19 (10.9)

0.71 (0.0)

2.27 (8.0)

0.71 (0.0)

0.71 (0.0)

0.71 (0.0)

1.43 (3.1)

2.51 (15.6)

1.43 (3.1)

1.63 (4.7)

4.27 (28.1)

1.43 (3.1)

1.43 (3.1)

2.74 (9.4)

5.42 (37.5)

Eleusine indica Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

ZT

0.71 (0.0)

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

9.10 (109.0)

10.39 (108.0) 4.22 4.50

5.22 (42.2)

7.32 (68.8) 3.57 4.05

3.39 (15.6) 2.09 2.30

2.86 (12.5) 3.82 4.67

6.18 (46.9)

3.46 (18.8) 3.77 4.40

5.98 (54.7)

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

Weed control treatment (WCT)

4.37 (25.0) 5.63 3.94

65

66

Table 9 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed biomass at 42 days after sowing (14 d after the application of post-emergence herbicides) in the wet season of 2011. Weed control treatment (WCT)

Weed biomass (g m−2 )

ZT

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

ZT

CONT

ZT

CONT

ZT

CONT

ZT

ZT

Cyperus rotundus

Dactyloctenium aegyptium

Digitaria ciliaris

0.00

0.56

1.23

0.00

2.40

0.00

1.23

1.50

7.50

0.96

0.04

0.00

2.30

0.48

1.87

1.40

10.70

2.71

0.00

6.10

11.10

0.75

0.40

11.30

1.85

1.83

2.80

6.00

4.21

1.54

48.90

27.50

3.10

31.40 23.58 28.56

2.87 2.62

15.72 16.67

Echinochloa colona

CONT

0.80

4.63 5.32

Eclipta prostrata

27.00 31.90

0.79 2.72 2.70

Weed biomass (g m−2 )

CONT

ZT

Eleusine indica Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

CONT

ZT

Trianthema portulacastrum

0.96

1.44

0.00

3.40

0.00

0.00

0.00

0.17

0.80

3.70

0.00

0.00

0.00

0.21

0.00

10.30

0.00

0.00

0.00

0.04

2.20

20.90

0.00

0.10

6.60

1.00

21.80

1.98

3.12

1.65

1.40

2.70

5.60

30.90

4.10 6.98 5.91

32.50 27.00

3.42 3.88

1.83 1.61

24.98 26.90

8.40 41.11 41.25

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

CONT

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

67

Table 10 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on weed biomass at 42 days after sowing (14 d after the application of post-emergence herbicides) in the dry season of 2012. Weed control treatment (WCT)

Weed biomass (g m−2 )

CONT

ZT

Cyperus iria Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage) Weed control treatment (WCT)

ZT

CONT

ZT

CONT

Cyperus rotundus

Dactyloctenium aegyptium

ZT

CONT

Digitaria ciliaris

ZT

CONT

Echinochloa colona

ZT

Eclipta prostrata

0.00

0.00

2.11

0.73

0.00

0.00

0.00

10.80

0.00

8.70

5.10

1.60

0.00

0.00

3.38

3.02

0.00

15.00

0.00

5.10

0.00

32.20

2.90

4.10

23.90

4.03

1.88

2.60

5.40

13.60

6.10

60.90

43.20

14.20

27.60 16.74 19.91

10.50 10.71 12.54

3.78 3.82

19.21 19.95

19.33 17.80

54.13 53.02

Weed biomass (g m−2 )

CONT

ZT

Eleusine indica Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control LSD (same level of WCT) LSD (same level of tillage)

CONT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

CONT

ZT

Leptochloa chinensis

Panicum brevifolium

Portulaca oleracea

Murdannia nudiflora

Trianthema portulacastrum

0.00

6.70

0.00

0.00

0.00

0.00

0.03

2.56

0.14

3.28

0.20

4.10

0.00

0.80

0.00

0.00

0.00

0.31

1.06

0.86

1.41

3.14

0.10

0.10

43.70

7.60

33.80

0.69

0.62

2.66

2.25

4.39

1.20

9.60

2.90 9.34 8.41

11.80 19.76 22.87

17.63 18.06

0.75 0.84

4.29 5.13

4.73 5.60

Table 11 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on total weed biomass at crop harvest in the wet and dry seasons. The values in parentheses represent percentage decrease in total weed biomass in the herbicide-treated plots relative to the control plots. Weed control treatments (WCT)

Weed biomass (kg ha−1 ) CONT

ZT

Wet season Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control Weed-free LSD (same level of WCT) LSD (same level of tillage)

CONT

ZT

Dry season

2030 (67) 2750 (55) 6090 100

2170 (66) 2510 (60) 6320 110

620 (75) 450 (82) 2470 100

2120 1930

1610 (69) 1740 (67) 5260 120 1910 1840

Table 12 Effect of tillage systems (conventional tillage, CONT; zero-till, ZT) and weed control treatments on rice grain yield in the wet and dry seasons. Weed control treatments (WCT)

Grain yield (t ha−1 ) CONT

ZT

Wet season Oxadiazon fb fenoxaprop + ethoxysulfuron Oxadiazon fb penoxsulam + cyhalofop Control Weed-free LSD (same level of WCT) LSD (same level of tillage)

ZT

Dry season

3.24 3.23 2.87 3.63

produced by the herbicide-treated plots. The trend was similar in the DS12; however, yield in all treatments improved in the dry season, when the weed-free plots produced 5.1 and 4.8 t ha−1 of rice yield in the CONT and ZT systems, respectively. This yield was statistically similar to the yield produced by the herbicide-treated plots in their respective tillage systems. Both herbicide-treated plots provided a similar yield advantage over the control plots, indicating a similar effect on yield; however, the yield advantage was greater in the ZT system than in the CONT system. In the WS11, for example, the herbicide-treated plots provided a 13% yield advantage over the control treatment (2.9 t ha−1 ) in the CONT system. In the ZT

CONT

3.41 3.51 2.02 3.63 0.45 0.32

5.35 5.17 3.89 5.07

4.85 5.12 2.39 4.83 0.97 0.73

system, on the other hand, the yield advantage was 69–74% over the control plots (2.0 t ha−1 ). The yield advantage in the herbicidetreated plots was greater in the DS12. The herbicide-treated plots provided 33–38% higher yield than the control plots in the CONT system, whereas these values in the ZT system were 103–114%.

4. Discussion Our study on mechanized dry-seeded rice sown in rows showed that weeds are an important biotic constraint if not contained. More

68

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

than 20 weed species occurred in the experiments over two seasons. The influence of tillage systems on weed emergence pattern was different for different weeds and the results were inconsistent over the two seasons. Seedling emergence of C. iria, for example, was greater in the CONT system in the first season, whereas, in the second season, seedling emergence was similar between the tillage systems. These results are different from the predictions made in a previous study (Chauhan and Johnson, 2009a). In that study, seeds of C. iria were very small in size, they required light to germinate, and their seedlings could not emerge from a burial depth of 1 cm. With these results, the authors suggested that this weed could become a problematic weed in ZT or reduced-till systems. In another study, seedling emergence of D. ciliaris, E. colona, and P. oleracea was greater in a CONT system than in a ZT system (Chauhan and Johnson, 2009b). In another study, too, D. ciliaris was a more dominant weed in no-till soybean fields than in tilled fields (Kobayashi and Oyanagi, 2005). Such an obvious trend, however, was not observed in our study. Although seedling emergence of D. ciliaris was greater in the ZT system than in the CONT system, the difference was statistically nonsignificant. Seedling emergence of E. colona and P. oleracea was similar between the tillage systems. There are several possible reasons for such different responses between the studies. First of all, the influence of tillage systems on weed seedling emergence pattern is mainly due to effects on vertical weed seed distribution in the soil (Chauhan et al., 2006a; Chauhan and Johnson, 2009b). Vertical weed seed distribution can be influenced by the implements used for tillage, the tines used for sowing, and the speed of the implements used for tillage and sowing operations. Deeply buried seeds may not be able to emerge because of their requirement for light to germinate or small seed size. On the other hand, seeds present on the soil surface may be prone to rapid desiccation and predation by ants (Chauhan et al., 2010). This information suggests that the effect of tillage systems on weed emergence is difficult to understand. Another possible reason for the different results between different studies is the type and quantity of the weed seed bank. In Chauhan and Johnson (2009b), for example, weed emergence was studied using an artificial seed bank, i.e., after broadcasting a known and equal amount of weed seeds. In our study, we studied a natural weed seed bank, which also explains the large variation observed within the tillage systems. A natural weed seed bank may have large spatial variability compared with an artificial weed seed bank. Irrigation may also influence seedling emergence of different weeds. Our study was conducted in an irrigated environment, in which irrigation was given when required. On the other hand, Chauhan and Johnson (2009b) conducted their study in a rainfed environment. Most of the published studies evaluated only weed emergence over a period of time. Our study, however, evaluated weed survival over a 35-day period after crop sowing. The results clearly showed that not all emerged weed seedlings survived over the sampling period. Seedlings of C. iria, D. ciliaris, and E. colona in the DS12, for example, reached a maximum level and then declined. A similar response was observed for P. oleracea in both seasons. Despite the inconsistent results between seasons and tillage systems, there was a clear indication that some weed species increased in the ZT system to a significant number in the second season. D. ciliaris, E. indica, E. prostrata, and L. octovalvis densities, for example, were minimal in the ZT system in the first season. In the second season, however, their densities increased by many-fold. These results indicate that problems of some weed species may increase in ZT systems if not effectively controlled. In a field study in India, for example, there was more than four times the abundance of E. colona after 4 years of ZT rice than when the soil had been cultivated (Singh et al., 2005). The results of weed seedling emergence pattern under different tillage systems also suggest that weed seedlings in any tillage

system emerging faster and earlier than the crop or other weed species have the potential to be more competitive with the crop and to cause greater yield loss (O’Donovan et al., 1985; Chauhan et al., 2006b). These results could also mean that late-emerging weed seedlings may escape the application of early post-emergence herbicides. The effect of tillage and weed control treatments was not clear on the weed density of many of the individual weed species; however, the effect was clearer on weed biomass. These results are not surprising as weed density is not the perfect parameter for evaluating the efficacy of herbicides. Weed biomass, on the other hand, should be used to evaluate herbicide efficacy on different weed species. Many weed species, including C. rotundus and M. nudiflora, were poorly controlled by the evaluated herbicide treatments. On the other hand, the herbicides provided excellent control of other weeds, including C. iria and P. brevifolium. The results also clearly showed that D. aegyptium, E. colona, and a few other weeds were poorly controlled by the herbicides in the ZT system. These responses resulted in lower efficacy of herbicides in the ZT system than in the CONT system. In the WS11, for example, herbicide-treated plots in the ZT system had 50–60% less weed biomass than the control plots, whereas these values for the CONT system were 80–96%. Lower efficacy of herbicides in ZT systems could be due to the presence of crop residues on the soil surface in ZT systems, which could intercept a significant amount of herbicides, and thereby reduce efficacy (Chauhan et al., 2006c, 2012b; Chauhan and Johnson, 2010; Chauhan, 2012). Although there was no loose residue on the soil surface under the ZT system in our study, anchored residue of 15–20-cm height may also intercept the applied herbicides. In previous studies, crop residues in the ZT system were found to intercept from 15 to 80% of the applied herbicide (Banks and Robinson, 1982; Buhler, 1995). Herbicide retention by organic residue and soil components at the soil surface may also increase photodecomposition or other processes and thereby reduce persistence in the soil (Mills et al., 1989; Jones et al., 1990). Another possible reason for lower herbicide efficacy in the ZT system is the large size of weed plants, which could be difficult to control by the application of post-emergence herbicides. A recent study found that fenoxaprop + ethoxysulfuron and penoxsulam + cyhalofop applied at the eight-leaf stage of D. ciliaris and E. colona did not provide effective control of these weed species (Chauhan and Abugho, 2012). The previous studies suggested that the timing of post-emergence herbicides is very critical to managing weeds effectively (Gopal et al., 2010; Chauhan and Abugho, 2012). Even after the applications of pre- and post-emergence herbicides, there was a need for one hand weeding at 42 DAS. The herbicide-treated plots still provided a significant amount of weed biomass at harvest. In the DS12, for example, the herbicide-treated plots in the CONT system produced 18–25% of the weed biomass of the control plots. Similarly, the herbicide-treated plots in the ZT system produced 31–33% of the weed biomass of the control plots. Grain yield, however, was not influenced by the weed biomass produced by the weeds emerging after 42 DAS in the herbicide-treated plots. A previous investigation on aerobic rice systems reported the critical periods for weed control, to obtain 95% of a weed-free yield, as between 17 and 56 DAS for crops in 15-cm rows (Chauhan and Johnson, 2011b). Weeds emerging later in the season (e.g., at 42 DAS in our study) may add seeds to the soil and may result in heavy infestation in the subsequent seasons (Bridgemohan et al., 1991). Therefore, there is a need to evaluate the efficacy of different postemergence herbicides so that the addition of weed seeds to the soil seed bank can be minimized. Overall, oxadiazon provided effective weed control, which is consistent with previous reports on its efficacy (Ishaya et al., 2007). Oxadiazon is used as a pre-emergence herbicide in dry-seeded rice

B.S. Chauhan, J. Ope˜ na / Field Crops Research 137 (2012) 56–69

to control annual grasses and broadleaf weeds (Ampong-Nyarko and De Datta, 1991). High soil moisture can increase the phytotoxicity to rice emergence, for example, because of heavy rain after a spray (Chauhan and Johnson, 2011a). These results suggest that further research is needed to improve understanding of the physiology of phytotoxicity to minimize the effects of herbicides on the performance of dry-seeded rice (Chauhan, 2012). The results also indicate that early pre-emergence herbicides should be evaluated to control weeds effectively in dry-seeded systems, especially where preemergence herbicide cannot be sprayed. In a tropical environment, the timing of rain is uncertain, especially in the wet season. If rain occurs immediately after the spray of a post-emergence herbicide, its efficacy declines. There is a need to evaluate the performance of different herbicides with adjuvants or stickers, such that the rainfastness period can be reduced. Grain yield in the weed-free plots was similar between the tillage systems and this was consistent during both seasons. These results suggest that the adoption of ZT rice systems may increase in farmers’ fields if effective weed control methods are adopted. In our study, yield in the herbicide-treated and weed-free plots was significantly similar; however, the yield in the herbicide-treated plots was achieved with one additional hand weeding. In the future, labor may become even more expensive than now. In addition, hand-weeding operations may result in the uprooting of some rice seedlings, and ultimately a non-uniform plant population. During hand weeding, it is hard to make the whole field weed-free, especially within rows. Therefore, there is a need to evaluate different weed management strategies in dry-seeded rice systems so that the need for hand weedings can be eliminated. The yield advantage in herbicide-treated plots (relative to control plots) was greater in the ZT system than in the CONT system. The reason for this response was the lower yield of control plots in the ZT system than in the CONT system. Grain yield in the ZT control plots was 0.9 and 1.5 t ha−1 less than in the CONT control plots in the WS11 and DS12, respectively. These results strongly suggest that, in the absence of effective weed control programs, farmers should not adopt ZT rice systems. The benefits of ZT systems or conservation agriculture can be fully exploited only if effective herbicides that are suitable for use in these systems can be identified. Acknowledgments The authors are grateful to Bill Hardy, International Rice Research Institute, Philippines, for providing comments on the manuscript. The authors would also like to thank Guido Ramos and Osmundo Bondad for providing excellent technical assistance. References Ampong-Nyarko, K., De Datta, S.K., 1991. A Handbook for Weed Control in Rice. ˜ (Philippines). International Rice Research Institute, Los Banos Banks, P.A., Robinson, E.L., 1982. The influence of straw mulch on the soil reception and persistence of metribuzin. Weed Sci. 30, 164–168. Bridgemohan, P., Brathwaite, R.A.I., McDavid, C.R., 1991. Seed survival and patterns of seedling emergence studies of Rottboellia cochinchinensis (Lour.) W.D. Clayton in cultivated soils. Weed Res. 31, 265–272. Buhler, D.D., 1995. Influence of tillage systems on weed population dynamics and management in corn and soybean in the central USA. Crop Sci. 35, 1247–1258. Chauhan, B.S., 2012. Weed ecology and weed management strategies for dry-seeded rice in Asia. Weed Technol. 26, 1–13.

69

Chauhan, B.S., Abugho, S.B., 2012. Effect of growth stage on the efficacy of postemergence herbicides on four weed species of direct-seeded rice. Scientific World J., Article ID 123071, 7 pages. Chauhan, B.S., Gill, G., Preston, C., 2006a. Influence of tillage systems on vertical distribution, seedling recruitment and persistence of rigid ryegrass (Lolium rigidum) seed bank. Weed Sci. 54, 669–676. Chauhan, B.S., Gill, G., Preston, C., 2006b. Seedling recruitment pattern and depth of recruitment of 10 weed species in minimum tillage and no-till seeding systems. Weed Sci. 54, 658–668. Chauhan, B.S., Gill, G., Preston, C., 2006c. Tillage system effects on weed ecology, herbicide activity and persistence: a review. Aust. J. Exp. Agric. 46, 1557–1570. Chauhan, B.S., Gill, G., Preston, C., 2007. Effect of seeding systems and dinitroaniline herbicides on emergence and control of rigid ryegrass (Lolium rigidum) in wheat. Weed Technol. 21, 53–58. Chauhan, B.S., Johnson, D.E., 2009a. Ecological studies on Cyperus difformis, C. iria and Fimbristylis miliacea: three troublesome annual sedge weeds of rice. Ann. Appl. Biol. 155, 103–112. Chauhan, B.S., Johnson, D.E., 2009b. Influence of tillage systems on weed seedling emergence pattern in rainfed rice. Soil Till. Res. 106, 15–21. Chauhan, B.S., Johnson, D.E., 2010. The role of seed ecology in improving weed management strategies in the tropics. Adv. Agron. 105, 221–262. Chauhan, B.S., Johnson, D.E., 2011a. Growth response of direct-seeded rice to oxadiazon and bispyribac-sodium in aerobic and saturated soils. Weed Sci. 59, 119–122. Chauhan, B.S., Johnson, D.E., 2011b. Row spacing and weed control timing affect yield of aerobic rice. Field Crops Res. 121, 226–231. Chauhan, B.S., Mahajan, G., Sardana, V., Timsina, J., Jat, M.L., 2012a. Productivity and sustainability of the rice-wheat cropping system in the Indo-Gangetic Plains of the Indian subcontinent: problems, opportunities, and strategies. Adv. Agron. 117, 315–369. Chauhan, B.S., Migo, T., Westerman, P.R., Johnson, D.E., 2010. Post-dispersal predation of weed seeds in rice fields. Weed Res. 50, 553–560. Chauhan, B.S., Singh, R.G., Mahajan, G., 2012b. Ecology and management of weeds under conservation agriculture: a review. Crop Prot. 38, 57–65. Farooq, M., Siddique, K.H.M., Rehman, H.M.U., Aziz, T., Lee, D., Wahid, A., 2011. Rice direct seeding: experiences, challenges and opportunities. Soil Till. Res. 111, 87–98. Froud-Williams, R.J., Chancellor, R.J., Drennan, D.S.H., 1981. Potential changes in weed floras associated with reduced-cultivation systems for cereal production in temperate regions. Weed Res. 21, 99–109. GenStat 8.0, 2005. GenStat Release 8 Reference Manual. VSN International, Oxford, UK. Gopal, R., Jat, R.K., Malik, R.K., Kumar, V., Alam, M.M., Jat, M.L., Mazid, M.A., Saharawat, Y.S., McDonald, A., Gupta, R., 2010. Direct dry seeded rice production technology and weed management in rice based systems. Technical Bulletin. International Maize and Wheat Improvement Center, New Delhi, India, p. 28. Ishaya, D.B., Dadari, S.A., Shebayan, J.A.Y., 2007. Evaluation of herbicides for weed control in three varieties of upland rice (Oryza sativa L.) in the Nigerian Savannah. Crop Prot. 26, 1490–1495. Jones, R.E., Banks, P.A., Radcliffe, D.E., 1990. Alachlor and metribuzin movement and dissipation in a soil profile as influenced by soil surface condition. Weed Sci. 38, 589–597. Kobayashi, H., Oyanagi, A., 2005. Digitaria ciliaris seed banks in untilled and tilled soybean fields. Weed Biol. Manage. 5, 53–61. Mills, J.A., Witt, W.W., Barrett, M., 1989. Effects of tillage on the efficacy and persistence of clomazone in soybean (Glycine max). Weed Sci. 37, 217–222. O’Donovan, J.T., de St Remy, E.A., O’Sullivan, P.A., Dew, D.A., Sharma, A.K., 1985. Influence of the relative time of emergence of wild oat (Avena fatua) on yield loss of barley (Hordeum vulgare) and wheat (Triticum aestivum). Weed Sci. 33, 498–503. Pandey, S., Velasco, L., 2005. Trends in crop establishment methods in Asia and research issues. In: Toriyama, K., Heong, K.L., Hardy, B. (Eds.), Rice is Life: Scientific Perspectives for the 21st Century. International Rice Research Institute and Tsukuba, Japan: Japan International Research Center for Agricultural Sciences, ˜ Los Banos, Philippines, pp. 178–181. Singh, G., Singh, Y., Singh, V.P., Johnson, D.E., Mortimer, M., 2005. System level effects in weed management in rice-wheat cropping in India. In: BCPC International Congress on Crop Science and Technology, SECC. Glasgow, UK. Singh, S., Ladha, J.K., Gupta, R.K., Bhusan, L., Rao, A.N., 2008. Weed management in aerobic rice systems under varying establishment methods. Crop Prot. 27, 660–671. Triplett Jr., G.B., VanDoren Jr., D.M., 1977. Agriculture without tillage. Scientific Am. 236, 28–33. Tuong, T.P., Bouman, B.A.M., 2003. Rice production in water-scarce environments. In: Kijne, J.W., Barker, R., Molden, D. (Eds.), Water Productivity in Agriculture: Limits and Opportunities for Improvements. CABI Publishing, UK, pp. 53–67.