Soil & Tillage Research 55 (2000) 99±106

Soil & Tillage Research 55 (2000) 99±106 Evaluation of non-puddling under shallow water tables and alternative tillage methods on soil and crop parameters in a rice±wheat system in Uttar Pradesh R.K. Bajpai a,*, R.P. Tripathi a,b a Department of Soil Science and Agricultural Chemistry, Indira Gandhi Agricultural University, Raipur 492012, India b Department of Soil Science, G.B. Pant University of Agriculture and Technology, Pantnagar 263145, Uttar Pradesh, India Received 8 July 1997; received in revised form 15 April 1998; accepted 16 March 2000 Abstract The existence of a shallow water table (surface to 0.54 m from June to October) is common phenomenon in Tarai (foothills of the Himalaya) of Uttar Pradesh, India. Puddled rice (Oryza sativa L.) crop followed by conventional land preparation for the succeeding wheat (Triticum aestivum L.) crop is normal cultivation practice in the region. This shallow water table can be effectively utilised to avoid puddling operations for the seeding of rice and reduce the degree of tillage required for the following wheat crop. An investigation was made in silty clay loam (Chernozem), for two consecutive years (1992±1993 and 1993±1994) at Pantnagar, India, in a rice±wheat cropping system. The treatments for rice were puddling and non-puddling with two fertility levels (NPK: 120:40:40 and 180:60:60) and for wheat two tillage systems (conventional and zero tillage) in puddled and non-puddled rice eld with two fertility levels. signi cantly reduced the bulk density of the surface (0± 0.06 m) soil at the tillering stage of rice, compared to non-puddling, whereas it was signi cantly higher after harvest. The hydraulic conductivity of the 0±0.06 m soil depth also reduced to one-sixth and one-half due to puddling at tillering and harvesting stages, respectively. In ltration rate was decreased from 0.68 to 0.46 mm h 1 at tillering and 1.78 to 0.94 mm h 1 at harvest due to puddling. The puddling only in rice enhanced the root length density by 12% but affected adversely the wheat crop and minimised the root length density by 28%. Both puddling and non-puddling were found to be equally effective for grain yield of rice. However, non-puddling of rice produced signi cantly higher wheat grain yield than that of wheat followed by puddled rice. Conventional tillage of wheat produced signi cantly higher (25%) grain yield than that of zero tillage. This study indicated that in shallow water table conditions, direct drilling of rice in place of puddled rice and conventional tillage for wheat is an alternative cultivation practice in a rice±wheat system. # 2000 Published by Elsevier Science B.V. All rights reserved. Keywords: Bulk density; Hydraulic conductivity; In ltration rate; Root growth; Rice±wheat system 1. Introduction * Corresponding author. Tel.: 91-771-422-149; fax: 91-771-424-532. E-mail address: pankaj.oudhia@usa.net (R.K. Bajpai) In India, rice and wheat occupy an area of 42 and 23 million hectares, respectively, and require different 0167-1987/00/$ ± see front matter # 2000 Published by Elsevier Science B.V. All rights reserved. PII: S 0167-1987(00)00111-2

100 R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 soil physical environments. is a common eld preparation practice to maintain wetland conditions for rice. is not only time consuming and capital intensive, but also alters the soil physical condition so that it is not conducive for the succeeding wheat. Rice grown after minimum tillage can produce yields similar to that under conventional puddling with minimised expenses on eld preparation (Sharma et al., 1988). for rice often causes subsurface compaction which may adversely affect the yield of the succeeding crop due to reduction in root growth and its distribution under poor soil physical environment (Oussible et al., 1992). Reduced root growth limits water uptake and consequently plants may experience water stress. Amelioration of the soil compaction requires additional tillage and energy for succeeding crops. Minimum tillage was shown to have an advantage over puddling in a clay loam soil for maintaining physical condition and saving eld preparation time (Brown and Quantrill, 1973). Excessive wetness in puddled rice soil can delay the planting of the following wheat and result in yield reductions of 35±40 kg ha 1 per day by a delay in planting after November 20 (Randhawa et al., 1981; Hobbs, 1987). Traditional land preparation in India for wheat after rice consists of 5±7 cultivator operations followed by levelling with heavy wooden plank. Land is usually left for few days to dry after ploughing and then irrigated to allow rapid decomposition of residues to obtain a good tilth. In heavy soils, Majid et al. (1987) compared the traditional method of sowing of wheat with direct drilling in between the rice stubbles and found no signi cant difference in grain yield and biomass production. The water table in foothill soils may rise almost to the ground surface (to less than 0.3 m depth) in the rainy season and hence rice plants may directly utilise the water (Choudhary, 1979). This provided the basis for investigating possibilities of avoiding puddling operation by direct seeding of rice, which in turn would ease the sowing of the succeeding wheat crop. Objectives of the present investigation were to study the effect of different tillage practices on physical properties of soil, rooting pattern and grain yield of rice and wheat in rice±wheat system. 2. Materials and methods 2.1. Field experiments The eld experiments were conducted at G.B. Pant University of Agriculture and Technology, Pantnagar, India, for 2 years (1992±1993 and 1993±1994). The mean annual rainfall of the area is 1364 mm. The total rainfall during the rice growing season from June to October was 772 mm in 1992 and 1327 mm in 1993. During the wheat growing season, the rainfall from November to April was 80 mm in 1992±1993 and 71 mm in 1993±1994. The soil is a silty clay loam mixed hyperthermic Aquic Hapludoll (Haplic Chernozem). The soil contained 150 g kg 1 sand, 530 g kg 1 silt and 320 g kg 1 clay, with a high concentration of organic carbon (28 g kg 1 ) (Walkley and Black, 1934), a medium concentration of KMnO 4 extractable nitrogen (250 kg ha 1 ), 0.5 M sodium bicarbonate (NaHCO 3 ) extractable phosphorus (15.3 kg ha 1 ), and available potassium (225 kg ha 1 ). A split plot design was used for the experiments on rice while a split±split plot design was used for the wheat experiments. For rice, puddling (P) for transplanted rice and non-puddling (NP) for direct drilled rice were taken as main plot treatments and two fertility levels as sub-plot: (NPK) 120:40:40 kg ha 1 (F 1 ) and 180:60:60 kg ha 1 (F 2 ). For wheat, the same plots of P and NP for rice were the main plot, while the conventional tillage (C) for wheat and zero tillage (Z) was the sub-plots and the two fertility levels (F 1 and F 2 ) were the sub±sub plot treatments. The treatments were repeated in the same plots in both years. Each treatment was replicated four times and sub-plot size was 6 m5.4 m. was conducted using a tractor drawn puddler and paddy was sown in line by a tractor drawn seeder with a fertiliser applicator. For wheat, conventional tillage (C) consisted of harrowing ve times about 0.10±0.12 m depth followed by drilling of wheat at 0.03±0.05 m depth. In the zero tillage (Z) treatment, direct drilling of wheat was conducted using a tractor drawn New Zealand zero till seed drill with a fertiliser applicator. The fertility level F 1 represented the recommended application of fertiliser and F 2 as 1.5F 1 both in the rice and wheat crop. The 35 days old rice seedlings 2±3 plants per hill (cv. PD-4) were transplanted in rainy season. Full rate of P as single super phosphate and K were applied at the time of land preparation in non-puddled direct

R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 101 drilled plots, and at the time of transplanting in puddled plots. Nitrogen was applied in three equal splits (40 and 60 kg ha 1 ) at transplanting, and at 3 and 6 weeks after transplanting. Nitrogen in nonpuddled plots was applied at the time of eld preparation and at tillering and panicle initiation stages of crop growth. Anilofos was sprayed two days after sowing at the rate of 1.5 kg active ingredient ha 1 in non-puddled plots to control weeds. Wheat (cv.``hd 2329'') was sown during 29 November to 2 December in the 2 years. Full rate of P and K was applied at the time of sowing and N was applied in two equal splits one at sowing and the other at 4 weeks later. Wheat was irrigated (three irrigations each of 60 mm were required) according to irrigation schedule recommended for wheat crop in the Tarai region. 2.2. Groundwater and soil measurements To determine the uctuation in groundwater table under natural conditions, piezometers were installed at six representative sites in the eld by drilling holes equal to the internal diameter of pipes with the bucket auger specially designed for this purpose. The pipes were driven 1.5 m into the soil with the help of heavy wooden blocks and hammer leaving 0.15 m above the soil surface. The measurements were made on alternate days. Soil bulk density and saturated hydraulic conductivity were each determined as described by Smith and Mullins (1991) on intact soil cores (three soil cores per plot). In the rice treatments, observations were taken at tillering stage and 2 days after harvest. For wheat, bulk density was determined at crown root initiation stage and at harvest. These observations were taken in both years. For saturated hydraulic conductivity and bulk density, aluminium cores (78 mm diameter, 58 mm high) were driven to a depth of 0±0.06 and 0.12± 0.18 m (compact horizon as per pro le study). A constant water head was maintained on top of each core in the laboratory and the rate of water ow through the soil was measured at steady state. Darcy's law was utilised to calculate the saturated hydraulic conductivity. The soil cores were then oven-dried to calculate bulk density. In ltration was measured in situ with a double-ring (three rings per treatment) in ltrometer (Mishra and Ahmad, 1990). The inside ring, from which measurements were taken, was 300 mm in diameter and the outer guard ring was 500 mm. In ltration rate was measured at tillering stage and at harvest of rice crop. 2.3. Root measurements Rice root samples (three samples per plot) were taken at penicle initiation and at ripening stage. Wheat root samples were taken at 105 days after sowing, with a core sampler of 100 mm diameter and 150 mm height, from 0 to 0.10, 0.10 to 0.20, 0.20 to 0.30, 0.30 to 0.40, and 0.40 to 0.50 m depths in sequence. The soil from each depth was washed over a 0.1 mm screen and the separated roots were stored in bottles containing 5% formalin solution. Root length was measured by the Newman (1966) method modi ed by Tennant (1975) using 0.01 m0.01 m size grid as follows: Root length R ˆ0:786 number of intersections grid unit (1) The root length density of rice and of wheat were each calculated from the total root length and the sample volume. 2.4. Crop yield The net sub±sub-plot area (5 m4.6 m) was harvested after removing the border rows and threshed. The grain yield was recorded after cleaning and drying at 140 g kg 1 moisture and expressed in kg ha 1. 2.5. Statistical methods Analysis of variance (ANOVA) was performed to study the effect of puddling, tillage, and fertility levels on soil physical properties, root growth and grain parameters. The ANOVA was done in split and split± split plot design as described by Cohran and Cox (1957). 3. Results and discussion 3.1. Groundwater The mean water table during the rice crop season was 0.22 m in 1992 (0.12 m) and 0.16 m in 1993 (0.16 m).

102 R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 Table 1 Effect of puddling and fertility levels on soil bulk density (mg m 3 ) at two soil depths during rice crop period Treatment Soil depth (0±0.06 m) Soil depth (0.12±0.18 m) 1992 1993 Mean 1992 1993 Mean Tillering stage 1.30 1.33 1.31 1.64 1.61 1.62 Non-puddling 1.42 1.41 1.42 1.61 1.58 1.59 LSD (0.05) 0.02 0.05 0.02 NS NS NS F 1 1.35 1.38 1.36 1.62 1.59 1.60 F 2 1.36 1.37 1.36 1.63 1.60 1.61 LSD (0.05) NS NS NS NS NS NS After harvest 1.48 1.50 1.49 1.68 1.70 1.69 Non-puddling 1.45 1.44 1.44 1.66 1.64 1.65 LSD (0.05) NS 0.03 0.03 NS 0.05 0.03 F 1 1.46 1.45 1.46 1.66 1.65 1.65 F 2 1.47 1.49 1.48 1.68 1.68 1.68 LSD (0.05) NS NS NS NS NS NS 3.2. Soil bulk density performed for planting of rice signi cantly reduced bulk density of surface soil (0± 0.06 m) only at tillering stage (Table 1). At harvest, bulk density of puddled plots increased and was found to be signi cantly higher than that of the non-puddled plot at both depths (0±0.06 and 0.12±0.18 m) during second year. The two fertility levels had no effect on bulk density at either depth or growth stage. The puddling and non-puddling operation in uenced signi cantly the bulk density at soil depth (interaction of puddling and soil depth). The puddling operation produced the highest bulk density in 0.12±0.18 m soil depth at both the stages, whereas the lowest bulk density was observed at 0±0.06 m soil depth under puddled plots at tillering and in non-puddled plot at harvest of the rice. Ghildyal (1982) and Rahman (1991) have also reported an increase in soil density below the puddled layer due to physical compaction during the puddling process. The interaction effect of puddling and non-puddling with different years indicated that puddling and non-puddling operations performed during both the years differed in response at the tillering stage of rice. The bulk density was maximum (1.55 mg m 3 ) under non-puddled condition during 1992, whereas it decreased (1.51 mg m 3 ) signi cantly during the subsequent year. Bulk density remained unchanged due to puddling over these years. The puddling performed on rice also in uenced bulk density at the 0±0.06 m soil surface layer during the wheat growing period (Table 2). After harvest of wheat, bulk density was signi cantly higher under puddled plot than non-puddled plots. The increase in bulk density may have been due to settling of soil particles. Similar observations were recorded by Sur et al. (1981) and Sawhney and Sehgal (1989). The tillage operation performed in wheat signi cantly reduced soil bulk density. Conventional tillage bulk density was signi cantly lower than zero tillage. 3.3. Hydraulic conductivity and in ltration rate The hydraulic conductivity of non-puddled plots in both soil depths (0.0±0.06 and 0.12±0.18 m) was signi cantly higher than that of puddled plot at both the tillering and at harvest stage of rice (Table 3). The decrease in hydraulic conductivity by puddling was probably due to destruction of soil aggregates and reduction of non-capillary pores (Sharma and De

R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 103 Table 2 Effect of puddling, fertility levels and tillage on bulk density (mg m 3 ) of soil in 0±0.06 m depth during wheat crop period Treatment CRI a stage After harvest 1992±1993 1993±1994 Mean 1992±1993 1993±1994 Mean 1.44 1.45 1.45 1.49 1.49 1.49 Non-puddling 1.42 1.39 1.41 1.45 1.44 1.45 LSD (0.05) NS 0.04 0.02 0.03 0.02 0.02 F 1 1.43 1.41 1.42 1.46 1.47 1.47 F 2 1.43 1.43 1.43 1.48 1.47 1.47 LSD (0.05) NS NS NS NS NS NS Tillage Conventional 1.40 1.38 1.39 1.46 1.44 1.45 Zero 1.47 1.46 1.47 1.49 1.50 1.50 LSD (0.05) 0.02 0.02 0.02 0.02 0.04 0.02 a Crown root initiation stage. Dutta, 1985; Mambani et al., 1989). Two fertility levels had no effect on hydraulic conductivity. The interaction of puddling and non-puddling operations with depths revealed that non-puddled plots recorded the highest hydraulic conductivity in the 0±0.06 m soil depth at tillering (3.3 mm h 1 )as well as at harvest (2.11 mm h 1 ) stage of rice. The lowest hydraulic conductivity (0.45 mm h 1 ) was noted in the 0.12±0.18 m soil depth at tillering stage under the puddled plot. The in ltration rate of the nonpuddled plot was signi cantly higher than that of the puddled plot (Table 4). The interaction effect of treatments with the years was signi cant for in ltration rate at tillering as well as harvest stage. 3.4. Root growth Root length density (RLD) of rice in the puddled treatment was signi cantly higher than in the nonpuddled treatment. The major portion of roots was concentrated in 0±0.10 m soil depth and hence recorded signi cantly higher RLD than lower depth (Table 5). Ghildyal and Satyanarayana (1969) also reported that rice roots were mainly restricted to 0± 0.07 m layer in a sandy clay loam. Similarly, RLD was signi cantly higher in 1.5 recommended fertiliser rate than in the recommended rate. The third order interaction of puddling and nonpuddling operations with fertility levels and depth of Table 3 Effect of puddling on hydraulic conductivity (mm h 1 ) of soil at two soil depths during rice crop period Treatment Soil depth (0±0.06 m) Soil depth (0.12±0.18 m) 1992 1993 Mean 1992 1993 Mean Tillering stage 0.61 4.57 0.59 0.46 0.43 0.45 Non-puddling 3.35 3.24 3.30 1.24 1.40 1.32 LSD (0.05) 0.13 0.04 0.08 0.04 0.05 0.08 After harvest 1.04 1.03 1.04 0.97 1.16 1.06 Non-puddling 2.13 2.17 2.15 1.78 1.84 1.81 LSD (0.05) 0.16 0.05 0.06 0.13 0.11 0.0

104 R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 Table 4 Effect of puddling on in ltration rate (mm h 1 ) of soil during rice crop period Treatment Tillering stage After harvest 1992 1993 Mean 1992 1993 Mean 0.47 0.45 0.46 0.96 0.92 0.94 Non-puddling 0.64 0.75 0.69 1.75 1.81 1.78 LSD (0.05) 0.05 0.07 0.06 0.06 0.03 0.06 soil was signi cant for RLD at panicle initiation and ripening stage of rice (Table 6). The RLD was higher in puddled plot than that of the non-puddled plot. The maximum RLD was obtained at 0±0.10 m depth fertilised with 1.5 recommended rate of fertiliser. In comparison, minimum RLD was observed at 0.40± 0.50 m soil depth fertilised with recommended rate of fertiliser in puddled plots. Almost similar interaction was observed at ripening stage. Higher root growth in the surface layer might have been due to lower bulk density in the puddled plot. The wheat grown on non-puddled rice plot signi cantly increased the RLD (3.03 mm mm 3 ) at ripening stage in 1993±1994 in 0±0.50 m soil depth (Table 7). The lower bulk density of non-puddled rice plot promoted the root growth of wheat. The conventional tillage produced a signi cantly higher RLD (3.11 mm mm 3 ) than did zero tillage (2.30 mm mm 3 ) treatment. The interaction between tillage and depths was signi cant (Table 7). RLD of 0±0.50 m soil depth under conventional tillage maintained its superiority over zero tillage, but at 0.40±0.50 m depth the differences in RLD under these tillage operations were similar. 3.5. Grain yield The grain yields of rice from puddled or nonpuddled treatments were statistically similar indicating successful cultivation of rice even under nonpuddled condition (Table 8). The variation of soil bulk density, hydraulic conductivity and in ltration rate due to non-puddling and puddling operation did not Table 5 Effect of puddling and fertility levels on RLD (mm mm 3 ) of rice at two growth stages Treatment Panicle initiation stage Ripening stage 1992 1993 Mean 1992 1993 Mean 4.22 4.98 4.60 6.32 6.81 6.57 Non-puddling 4.03 4.23 4.13 5.60 5.97 5.79 LSD (0.05) 0.48 0.75 0.25 0.710 0.80 0.34 F 1 3.70 4.28 4.00 5.42 6.01 5.71 F 2 4.46 4.93 4.70 6.48 6.78 6.64 LSD (0.05) 0.22 0.60 0.26 0.46 0.23 0.18 Depth (m) 0±0.10 14.42 14.81 14.61 21.67 23.00 22.34 0.10±0.20 2.20 2.74 2.47 3.47 3.56 3.52 0.20±0.30 1.95 2.50 2.23 2.20 2.54 2.37 0.30±0.40 1.30 1.70 1.50 1.48 1.74 1.61 0.40±0.50 0.77 1.12 0.95 0.97 1.13 1.05 LSD (0.05) 0.36 0.64 0.36 1.08 0.98 1.03

R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 105 Table 6 Interaction effect of puddling (P), fertiliser (F) and soil depth (D) on RLD (mm mm 3 ) of rice at two growth stages Soil depth (m) Panicle initiation stage Ripening stage F 1 F 2 F 1 F 2 0.0±0.10 14.45 18.10 22.93 25.83 0.10±0.20 2.36 2.66 3.53 3.88 0.20±0.30 1.90 2.10 2.05 2.15 0.30±0.40 1.14 1.40 1.21 1.47 0.40±0.50 0.74 0.89 0.72 1.07 Non-puddling 0.0±0.10 12.13 13.77 18.13 22.45 0.10±0.20 2.40 2.47 3.25 3.41 0.20±0.30 2.34 2.58 2.48 2.80 0.30±0.40 1.59 1.83 1.78 1.98 0.40±0.50 0.94 1.21 1.07 1.34 LSD (0.05) 2 fertiliser at PD 0.49 0.67 2 depth at PF 0.36 0.70 affect the grain yield of rice. The RLD of rice due to puddling was mainly concentrated in surface layer but under non-puddled treatment its concentration was more lower in the pro le (Table 6). This shows a potential possibility of raising direct seeded rice in non-puddled elds at these conditions. Grain yield of wheat was signi cantly lower in puddled rice plot than in non-puddled rice plots in both the years (Table 9). This may be due to subsurface compaction. Similar results have also been reported in a clay loam soil by Oussible et al. Table 7 RLD (mm mm 3 ) of wheat in different soil depths (D) at ripening stage as in uenced by tillage (T) practices of wheat Soil depth (m) Tillage Conventional Zero Mean 0.00±0.10 6.80 5.60 6.20 0.10±0.20 3.40 2.30 2.80 0.20±0.30 2.30 1.50 1.90 0.30±0.40 1.80 1.30 1.50 0.40±0.50 1.20 0.80 1.00 LSD (0.05) D 0.26 DT (between D) 0.50 DT (between T) 0.40 Table 8 Effect of puddling and fertility levels on grain yield (kg ha 1 )of rice over 2 years Treatment 1992 1993 Mean 5880 5890 5895 Non-puddling 5340 5420 5380 LSD (0.05) NS NS NS F 1 5190 5310 5250 F 2 6030 6000 6015 LSD (0.05) 280 300 341 Table 9 Effect of puddling, fertility levels and tillage on grain yield (kg ha 1 ) of wheat over 2 years Treatment 1992±1993 1993±1994 Mean 3075 4046 3560 Non-puddling 3558 4600 4079 LSD (0.05) 320 460 200 F 1 3220 4180 3700 F 2 3420 4470 3945 LSD (0.05) NS 230 140 Tillage Conventional 3550 4990 4245 Zero 3130 3650 3390 LSD (0.05) 210 440 160 (1992). Conventional tillage for wheat produced signi cantly higher grain yield than did zero tillage. The higher fertility level produced signi cantly higher grain yield than did lower fertility levels. The puddling and tillage interaction was found to be signi cant. The zero tillage for wheat in non-puddled rice produced 23% higher grain yield than did the zero tillage in puddled rice. Conventional tillage in non-puddled rice produced 9% higher grain yield than conventional tillage in the puddled rice. 4. Conclusions The tillage practices carried out for planting/seeding in a rice±wheat system under a shallow ground water condition revealed that rice grown under non-

106 R.K. Bajpai, R.P. Tripathi / Soil & Tillage Research 55 (2000) 99±106 puddled conditions produced similar grain yield to that of rice grown under puddled conditions. increased bulk density of surface and subsurface soil at harvest and caused a decrease in in ltration rate and hydraulic conductivity in subsequent years. Increased soil bulk density adversely affected root growth and grain yield of the following wheat crop. The conventional tillage was superior than zero tillage for grain yield of wheat. More interestingly, the conventional tillage utilised for wheat after non-puddled rice gave a higher yield than that under puddled rice. Acknowledgements The rst author is grateful to the Council of Scienti c and Industrial Research, New Delhi, India, for awarding a senior research fellowship and admissible nancial support to carry out this study. References Brown, I.A., Quantrill, R.A., 1973. The role of minimum tillage in rice with particular reference to Japan. Outlook Agric. 7, 179± 183. Choudhary, B.C., 1979. Field water balance studies in rice. Ph.D. Thesis. G.B. Pant University of Agriculture and Technology, Pantnagar, Uttar Pradesh, India. Cohran, W.G., Cox, G., 1957. Experimental Designs, 2nd Edition. Wiley, New York, pp. 293±316. Ghildyal, B.P., 1982. Nature, physical properties and management of submerged rice soil. In: Vertisols and Rice Soils of the Tropics. Symposium Paper II. Proceedings 12th International Congress on Soil Science. Indian Society of Soil Science, New Delhi, India, pp. 121±142. Ghildyal, B.P., Satyanarayana, T., 1969. In uence of soil compaction on shoot and root growth rate of rice (Oryza sativa L.). Indian J. Agron. 14, 187±192. Hobbs, P.R., 1987. A perspective on research needs for the rice± wheat rotation. In: Klatt, A.R. (Ed.), Wheat Production Constraints in Tropical Environment. Proceedings of the International Conference, January 19±23, 1987. Chiang Mai, Thailand, pp. 197±211. Majid, A., Astam, M., Hashmi, N.J., 1987. Potential use of minimum tillage in wheat after rice. In: Klatt, A.R. (Ed.), Wheat Production Constraints in Tropical Environment. Proceedings of the International Conference, January 19±23, 1987. Chiang Mai, Thailand, pp. 71±77. Mambani, B., De Datta, S.K., Redulla, C.A., 1989. Land preparation requirements for rainfed rice as affected by climatic water balance and tillage properties of lowland soils. Soil Till. Res. 14, 219±230. Mishra, R.D., Ahmad, M., 1990. Manual on Irrigation Agronomy. Oxford and IBH Publishing, New Delhi, 61 pp. Newman, J., 1966. A method of estimating the total length of root in a sample. J. Appl. Ecol. 3, 139±145. Oussible, M., Crookstan, R.K., Lorson, W.E., 1992. Subsurface compaction reduces the root and shoot growth and grain yield of wheat. Agron. J. 84, 34±38. Rahman, S.M., 1991. Tillage effect on some soil physical properties. Ann. Agric. Res. 12, 196±199. Randhawa, A.S., Dillon, S.S., Singh, D., 1981. Productivity of wheat varieties as in uenced by the time of sowing. J. Res. 18, 227±233. Sawhney, J.S., Sehgal, J.L., 1989. Effect of rice±wheat and maize± wheat crop rotations on aggregation bulk density and in ltration characteristics of some alluvium-derived soil. J. Indian Soc. Soil Sci. 37, 235±244. Sharma, P.K., De Dutta, S.K., 1985. in uence on soil, rice development and yield. Soil Sci. Soc. Am. J. 49, 1451±1457. Sharma, P.K., De Datta, S.K., Redulla, C.A., 1988. Tillage effects on soil physical properties and wet land rice yield. Agron. J. 80, 34±39. Smith, K.A., Mullins, C.E. (Eds.), 1991. Soil Analysis Physical Methods. Marcel Dekker, New York. Sur, H.S., Prihar, S.S., Jalota, S.K., 1981. Effect of rice±wheat and maize±wheat rotation on water transmission and wheat root development in a sandy loam soil of Punjab, India. Soil Till. Res. 1, 361±371. Tennant, D., 1975. A test of modi ed line intersect method of estimating root length. J. Ecol. 63, 995±1001. Walkley, A., Black, C.A., 1934. An examination of the method of determination of organic matter and a proposed modi cation of the chromic acid titration method. Soil Sci. 37, 29±34.