- Email: [email protected]
Integrating best management practices for rice with farmers’ crop management techniques: A potential option for minimizing rice yield gap
Field Crops Research 144 (2013) 62–68
Contents lists available at SciVerse ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Integrating best management practices for rice with farmers’ crop management techniques: A potential option for minimizing rice yield gap M. Murshedul Alam a , Md. Rezaul Karim a , J.K. Ladha b,∗ a b
International Rice Research Institute (IRRI), Bangladesh Office, House-9, Road 2/2, Banani, Dhaka 1213, Bangladesh IRRI India Office, 1st Floor, CG Block, NASC Complex, DPS Marg, New Delhi 110012, India
a r t i c l e
i n f o
Article history: Received 5 October 2012 Received in revised form 16 January 2013 Accepted 16 January 2013 Keywords: Rice–rice cropping system Yield gap Productivity Net profit Leaf color chart
a b s t r a c t The major cereal cropping system in Bangladesh is rice (Oryza sativa L.) in the dry season (Boro) – rice in the wet season (Aman). The average productivity (7 t ha−1 ) of this system is far below attainable yields (14 t ha−1 ) in farmers’ fields, resulting in a large yield gap mainly due to farmers’ traditional management practices. We evaluated a set of selected best management practices (BMP) along with two N management options in the farmers’ crop management practices for rice in numerous farmers’ fields across 24 villages over 5 contiguous seasons during 2006–2008. Across years, BMP and two N management options increased grain yields compared with the farmers’ practice (FP) by 0.73 t ha−1 in both Aman and Boro seasons. The higher yield response (24.6% in Aman and 8.6% in Boro season) occurred with BMP in combination with leaf color chart (LCC) aided N management (BMP–LCC) than BMP with Urea Super Granule as N source (BMP–USG). Best management practices and two N management options reduced the yield gap of FP by 45%, with an average of 1.5 Mg ha−1 . The average added net returns with BMP were US$22 to US$120 ha−1 in the wet season and US$93 to US$115 ha−1 in the dry season. Our study shows that the integration of BMP and either LCC-aided N management or USG as an N source with the farmers’ management techniques and its adoption has the potential to boost rice yield and profit and total rice production in Bangladesh. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Rice is a staple food and accounts for more than 40% of the calorie supply of most Asians (IRRI, 2008). About 90% of the world’s rice is grown and produced (142 million ha area with production of 622 million tons) in Asia (FAO, 2010). Economically disadvantaged people spend as much as 30–40% of their income just to buy rice (IRRI, 2008). Increasing the amount of rice production and keeping the rice price low and affordable to them are crucial for poverty reduction. Rice yield must continue to increase at an annual rate of 1.5% compared with the current rate of 0.8% to keep pace with the expected demand (Aureus, 2011). Like in many other Asian countries, rice is the staple food in Bangladesh and it contributes one-half of the agricultural GDP and 55% of the total labor employment (Bangladesh Economic Review, 2009). During the last 20 years, Bangladesh has increased rice production 1.8 times with almost no increase in its land area of about 11.7 million ha (FAO, 2012). However, because of the continuous increase in population growth, rice demand in 2050 is projected to be 56% higher than in 2001 (Mukherjee et al., 2011). This is
∗ Corresponding author. Tel.: +91 11 25843802; fax: +91 11 25841801. E-mail address: [email protected] (J.K. Ladha). 0378-4290/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2013.01.010
going to be doubly challenging in light of the diversion of better quality land, water, and labor to other sectors of the national economy. Evidence is now appearing that the productivity of rice–rice and rice–wheat systems is plateauing because of a fatigued natural resource base (Ladha et al., 2003; Pathak et al., 2003). The productivity and sustainability of rice-based cropping systems are threatened because of (i) the inefficient use of fertilizer, water, and labor; (ii) increasing scarcity of water and labor; (iii) changing climate; (iv) emerging energy crisis and rising fuel prices; and (v) emerging socioeconomic changes such as urbanization, migration of labor, and preference for nonagricultural work (Ladha et al., 2009). In addition, recent increases in the prices of farm inputs in relation to outputs, fewer off-farm work opportunities for supplementing farm income, reduced remittances from relatives working outside villages, and declining income and purchasing power of poor consumers have threatened the existence of rice producers and consumers (Ladha et al., 2009). Bangladesh produced about 50.1 million tons of rough rice from 11.7 m ha of land in 2010 with a productivity of 4.3 t ha−1 (FAO, 2012). The present productivity is far below the attainable yield of 8–10 t ha−1 in the dry season (Boro) and 5–6 t ha−1 in the wet season (Transplanted Aman) in farmers’ field experiments (BRRI, 2007, 2008, 2009, 2010). This difference in yield in farmers’ fields between farmer-managed and researcher-managed trials is
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
mainly due to the differences in management practices adapted by researchers and farmers. The causes of such a yield gap are classified into two broad categories: (i) biotic factors such as poor-quality seeds and seedlings, insects, diseases, weeds, and rodents; and (ii) abiotic factors such as soil, nutrients, and water. However, a large portion of this yield gap remains unexplained. Given the continuously increasing population, coupled with scarcity of land, water, labor, and energy, Bangladesh is under tremendous pressure to increase, by all possible means, rice production. Since the possibility of expanding rice area is limited, the extra rice production must come from productivity increases. The major challenge is to minimize or bridge the yield gap in farmers’ fields and increase productivity with less water, less agro-chemicals, and less labor, thereby making rice farming more profitable and sustainable in the long term. An adequate understanding and the development of appropriate practical technologies to minimize the causes of the yield gap are critical for meeting the challenges of continued gains in rice production, without degrading the natural resource base. Minimizing the yield gap and increasing profit and product quality are becoming increasingly difficult to achieve by using a single-technology-centric approach. The use of a component technology in isolation also has limited widespread adoption. Combining and simultaneously applying a number of the best compatible individual or component technologies is crucial for maximizing overall benefits to farmers. Depending on the need and profitability of new technologies, farmers generally integrate new technology with the cultural practices being practiced by them on their farms. This process of integrating new technologies with existing farmers’ practice has been referred to as integrated crop and resource management (ICRM) or best management practices (Ladha et al., 2009). The choice or selection of the best individual or component technologies in an integrated manner is crucial for achieving the full benefit. The integration of best component technologies should (i) ensure timely crop establishment and
63
uniform crop stands, resulting in higher crop yields; (ii) enhance resource- or input-use efficiency; (iii) provide immediate, identifiable, and demonstrable economic benefits; and (iv) reduce adverse effects on the environment and build up soil fertility over the long term. This study therefore evaluated a set of selected best management practices integrated with the farmers’ management techniques in rice for productivity and profitability in many farmers’ fields over five seasons during three years.
2. Materials and methods 2.1. Experimental sites and seasons On-farm trials were conducted in a continuous rice–rice cropping system for five consecutive seasons during 2006–2008 at 24 villages located within a radius of 50 km in central Bangladesh. Number of selected villages were 14 in Kapasia Upazila (23◦ 55 to 24◦ 12 N and 90◦ 29 to 90◦ 43 E) of Gazipur District, 4 in Pakundia Upazila (24◦ 16 to 24◦ 0 N and 90◦ 46 E) of Kishoreganj District, and 6 in Monohardi Upazila (24◦ 03 to 24◦ 16 N and 90◦ 38 to 90◦ 49 E) of Narshingdi District (Table 1). The villages which represent common cropping systems, farmers’ management practices, land types, and soil types were selected. Each village had 800–1500 farm families. All the villages in Gazipur District belong to Agro-Ecological Zone (AEZ) 28 (Madhupur Tract), AEZ 8 (Young Brahmaputra and Jamuna Floodplains), and AEZ 9 (Old Brahmaputra Floodplain), whereas the villages in Kishoreganj District belong to AEZ 8 and AEZ 9 and those in Narshingdi District belong to AEZ 9. The soils are (a) well drained friable clay loam to clay or heavy clay, strongly acidic, mainly phosphate fixing, low in P, K, S, and B in AEZ – 28, (b) permeable silt loam to silty clay loam, neutral to slightly acidic, deficient in N, P and S but medium in K and Zn status in AEZ – 8, and (c) silt loam to silty clay loam, moderately acidic to neutral, with low K and medium P status in AEZ – 9 (FRG, 2005).
Table 1 Locations and number of farmers’ fields where the trials were conducted in central Bangladesh in Aman and Boro seasons during 2006–2008. Location District
Number of farmers’ fields Upazila
Village
Agro-ecological zone (AEZ) – 28, AEZ – 8 and AEZ – 9 Kapasia Saluateki Gazipur Sultanpur Tokenagar Borjapur Shahartoke Bagua Baraigao Barapusia Pabur Roynanda Sonaroa Uttarkhamer Gosherkandi Aralia AEZ – 9 Narshingdi
Monohardi
Harordia Kachikata Kathalia Hatirdia Kocherchar Dosdona
AEZ – 8 and AEZ – 9 Kishoreganj
Pakundia
Kodalia Egarasindur Motkhola Meratala
Total
Aman 2006 9 29
Boro 2007 14 32 29 31 23
Aman 2007 33 28 20 15 12
Boro 2008 19 49 34 17 13 7 7 7 7 7 7 7
Aman 2008 15 48 42
31 25 20 20
10 6 16 10 10 10 10 10 15 5
38
129
108
221
263
64
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
Transplanted rice was grown on puddled soil under rainfed conditions in the wet season or Aman (June–July to November–December) and under irrigated conditions in the dry season or Boro (December–January to April–May). In both seasons, soils were well prepared through 4–5 passes with a power tiller followed by 2–3 ladderings. High-yielding semi-dwarf rice cultivars BR 11 (145 d seed to seed) and BRRI dhan 28 (140 d seed to seed) were grown in the Aman and Boro seasons, respectively.
(BMP–USG). The BMP consisted of options for managing farmers’ yield-limiting constraints identified from a farmers’ survey carried out prior to the start of the trials. The identified constraints were poor quality of seeds and seedlings, sub-optimal planting method, and sub-optimal fertilizer application and N timing and N source. In BMP–LCC and BMP–USG treatments, crop management practices other than BMP coupled with N management were done as practiced by the farmers in FP.
2.2. Experimental design and treatments
2.3. Treatment details
The trials were established in farmers’ fields in a randomized complete block design with a set of three treatments in each farmer’s field. Farmers’ fields were considered as replications. The number of replicate farmers varied among villages and seasons (Table 1). In a village, the farmers’ fields were selected in a contiguous fashion forming a block and farmers of the selected fields belong to marginal, small, and medium farmer groups. Out of three treatments, treatment number 1 was farmers’ management practices (FP), treatment number 2 was best management practices (BMP) coupled with leaf color chart (LCC)-aided fertilizer N management (BMP–LCC) and treatment number 3 was BMP coupled with urea super granules (USG) as source of fertilizer N
2.3.1. Treatment number 1 – FP The farmers’ management practice of seed and seedling raising, planting method, and fertilization varied considerably among farmers and seasons (Table 2). All farmers used seeds from their own source having 80–90% germination capacity without any seed treatment. They prepared a flat seedbed without any organic matter and used a very high seed rate (135–165 g m−2 ) on the seedbed. Five to eight 28–45-day-old rice seedlings per hill in Aman and five to seven 28–61-day-old rice seedlings in the Boro season were transplanted at random spacing. Across seasons and years, all farmers applied N fertilizer at early tillering [(10–20 days after transplanting (DAT)] and late tillering
Table 2 Farmers’ management practices for transplanted rice in central Bangladesh in Aman and Boro seasons during 2006–2008a .
a b
Parameterb
Aman 2006
Aman 2007
Aman 2008
Boro 2007
Boro 2008
Seed rate on seed bed, range (g m−2 ) Seed rate to transplant 1 ha main field (kg) Seedling age, range (days) Hill spacing at transplanting (cm) Seedling number (hill−1 )
145–165 65 ± 9 28–39 Random 5–6
135–150 48 ± 8 30–45 Random 5–6
138–149 33 ± 6 33–34 Random 5–8
155–165 48 ± 6 29–61 Random 5–7
154–160 39 ± 4 28–45 Random 5–6
Fertilizer management N applied at basal Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
0 0
0 0
N applied at early tillering Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
100 10 ± 2 46 ± 4
100 16 ± 3 31 ± 3
100 18 ± 4 49 ± 3
100 16 ± 5 53 ± 14
100 20 ± 3 57 ± 4
N applied at late tillering Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
100 26 ± 3 69 ± 18
100 36 ± 3 41 ± 14
100 38 ± 4 55 ± 4
100 28 ± 3 55 ± 13
100 36 ± 3 47 ± 6
N applied at panicle initiation Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
0 0 0
35 51 ± 2 21 ± 2
0 0 0
93 40 ± 6 52 ± 13
100 45 ± 4 47 ± 6
Fraction of total N applied At basal (%) At early tillering (%) At late tillering (%) At panicle initiation (%)
0 40 60 0
0 33 44 23
0 47 53 0
0 33 34 33
0 38 31 31
P applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
93 23 ± 7
100 17 ± 4
K applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
79 38 ± 13
100 37 ± 5
S applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
59 13 ± 5
97 12 ± 2
Zn applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
15 4±1
52 2±2
In all cases, farmers’ own seed was the source, seeds were not treated, seedbed had a flat surface, no decomposed cowdung was used, and hill spacing at TP was random. Mean ± standard deviation (S.D.). All reported means for timing and amount of fertilizers are based on farmers applying fertilizers.
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
(26–36 DAT) and 0–100% of the farmers applied at panicle initiation (40–51 DAT). The mean total N dose was 93–160 kg ha−1 . Farmers applied on average 67–100% of fertilizer N between early tillering and late tillering (Table 2). In the Aman season, no farmer applied P, K, S, and Zn fertilizer. In the Boro season, 93–100% of the farmers applied P fertilizer, 79–100% applied K fertilizer, 59–97% applied S fertilizer, and 15–52% applied Zn fertilizer. The average P, K, S, and Zn application rates for farmers ranged 17–23 kg ha−1 , 37–38 kg ha−1 , 12–13 kg ha−1 and 2–4 kg ha−1 , respectively. 2.3.2. Treatment number 2 – BMP–LCC A set of five BMP integrated with existing farmers’ practices were (i) certified seeds, (ii) healthy young seedlings, (iii) hill spacing of 20 cm × 20 cm at transplanting with 3–4 seedlings per hill, (iv) AEZ-based recommended P, K, S, and Zn fertilization, and (v) N application by using the LCC in a real-time N management approach. The management practices recommended by BRRI (2004) were used for raising healthy seedlings (Table 3). Across the years, seedling age ranged from 25 to 32 days in Aman and from 30 to 45 days in Boro. The P, K, S, and Zn fertilizers were applied at the rates and time recommended by BRRI (2004) based on AEZ-based recommendations. The P and K rates varied between locations and between seasons while the S and Zn rates were the same across seasons and locations (Table 3). Nitrogen as urea was applied by using the LCC in a real-time N management approach (Alam et al., 2004). The N rate varied among farmers’ fields and seasons. The mean total N dose ranged from 77 to 104 kg ha−1 in Aman and from 97 to 120 kg ha−1 in Boro (Table 3). 2.3.3. Treatment number 3 – BMP–USG The management practices in BMP–USG were same as BMP–LCC except N fertilizer management. Nitrogen as USG was deep-placed within 7–10 DAT at 7–10-cm soil depth at the rate of one USG (1.8 g) in Aman and one USG (2.7 g) in Boro in the middle of 4 hills. The total N dose was 52 kg ha−1 in Aman and 78 kg ha−1 in Boro (Table 3). 2.4. Data collection and analysis Grain yields (rough rice) were obtained from two central 6 m2 harvest areas in each plot at harvestable maturity and reported at 0.14 g H2 O g−1 fresh weight of grain. Human labor used for seedling raising, transplanting, and fertilizer application and wage rate were recorded. The time required to complete each field operation was expressed as person-days ha−1 , considering 8 h to be equivalent to 1 person-day. The farm-gate price of rough rice and prices of seed, fertilizers, and seed-treating chemicals were also recorded. Analysis of variance (ANOVA) was performed with the PROC MIXED procedure of the SAS/STAT software (SAS Institute, 1999) to determine the effects of villages and treatments on yield, input uses, and added net return. Duncan’s multiple range test (DMRT) was used at the P < 0.05 level to test the differences between the treatment means. Unless indicated otherwise, differences were considered significant only when P ≤ 0.05. Descriptive statistics such as means, standard deviation, range, 25% quartile, and 75% quartile were used to determine the variability of parameters. Treatments were evaluated based on added net return relative to the farmers’ practice, which is the difference between added gross return and added cost for a treatment as compared with FP. Added gross return equaled additional yield as rough rice (yield of treatment − yield of FP) multiplied by the price of yield. Added cost equaled the sum of costs for differences in labor determined as [(labor for seedling raising, transplanting, and fertilizer application of the treatment − labor for seedling raising, transplanting, and fertilizer application of FP) × wage rate] and costs for differences in fertilizer and seed-treating chemicals determined as (fertilizer and seed-treating chemical costs of treatment − fertilizer and
65
seed-treating chemical costs of FP). The prices of seeds, fertilizer, seed-treating chemical, rough rice, and labor wage were as follows: seed = US$0.29 to 0.51 kg−1 , P fertilizer = US$1.61 to 3.65 kg−1 P, K fertilizer = US$0.41 to 1.31 kg−1 K, S fertilizer = US$0.49 to 0.97 kg−1 S, Zn fertilizer = US$2.64 to 3.65 kg−1 Zn, urea = US$0.22 to 0.38 kg−1 N, USG = US$0.25 to 0.41 kg−1 N, seed-treating chemical = US$36.40 to 38.70 kg−1 , rough rice = US$0.15 to 0.26 kg−1 , and labor wage = US$1.46 to 2.63 person-day−1 (US$1 = Bangladesh Taka 68.5). 3. Results and discussion 3.1. Grain yield The treatment of best management practices along with two N management treatments, either LCC or USG, had a significant response in rice grain yield in both Aman and Boro seasons in all the villages and all three years (Tables 4 and 5). On average, the yields in Boro were 65% higher than the yields in Aman, with the highest yield of 8.4 Mg ha−1 in Boro and 5.3 Mg ha−1 in Aman. Compared with the farmers’ practice, the yield increases with BMP and two N management options ranged from 12 to 28% in Aman and from 3 to 22% in Boro, with an average of 0.73 Mg ha−1 in both seasons (Tables 4 and 5). BMP with LCC on average had a higher response than BMP with USG by 24.6% in Aman and by 8.6% in Boro. Across villages and years, the average yield for the system (combined yield of Aman and Boro seasons within a cropping year) for FP (10.7 Mg ha−1 ) and for BMP along with N management options (12.2 Mg ha−1 ) resulted in a yield gap of 3.3 Mg ha−1 in FP and 1.8 Mg ha−1 in BMP based on the BRRI-reported (BRRI, 2007, 2008, 2009, 2010) average attainable yield of 14 Mg ha−1 for the system. Thus, in this study, BMP along with any of two N management options reduced the yield gap of the farmers’ practice by 1.5 Mg ha−1 (45%). In the 2007 and 2008 Aman seasons and the 2008 Boro season, when more villages were included, the interaction effect between treatment and village was significant. In these seasons, the yields obtained in BMP along with either LCC or USG were higher by 0.5–1.1 Mg ha−1 than the yields in FP across villages (Tables 4 and 5). The yields of BMP–LCC and BMP–USG were always similar in the range 4.3–7.8 Mg ha−1 except in 4 villages in Aman 2008 and in 7 villages in Boro 2008. In all 4 villages in Aman and 5 of 7 villages in Boro, BMP–LCC produced the highest yields (5.0–8.4 Mg ha−1 ) and in 2 villages in Boro BMP–USG produced the highest yield (7.8 Mg ha−1 ). In the fields where BMP–USG produced the highest yields, the soils were always either saturated or had standing water, suggesting a better performance of USG under an assured water source. The positive response of BMP along with two N management options was due to the combined effects of certified seeds with seed treatment, healthy seedlings, optimum planting density, balanced fertilization, and better synchronization of N fertilization with plant needs. These findings suggest potential to increase farmers’ rice yields through BMP in an integrated way. Other studies in India and Bangladesh rice fields also reported comparatively higher yields in the range 0.8–2.0 Mg ha−1 with the use of improved methods of more than one crop production factor in an integrated way (Regmi et al., 2009; Alam et al., 2005, 2006). 3.2. Economic analysis Added costs and added net returns for BMP relative to the FP were used to assess profitability for the use of BMP along with either LCC or USG (Table 6). The average added costs for BMP in Aman and Boro across years ranged from $55 to 99 ha−1 and from $33 to
66
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
Table 3 Best management practices (BMP) for selected crop production factors integrated with farmers’ management techniques for transplanted rice and applied by farmers in central Bangladesh in Aman and Boro seasons during 2006–2008. Crop production factors and parametera
Best management practices (BMP)
Seed and seedlings Seed source Seed germination (%) Seed treatment
Certified seeds from BRRI and Bangladesh Agricultural Development Corporation (BADC) 90–95 Seeds treated with bavistine at 3 g per kg seeds. Seeds were soaked in water mixed with bavistine for 24 h 30 kg seeds used to transplant 1 ha main field Land having loamy or clay loam soils, sufficient sunlight, irrigation, and drainage facility was selected for seedbed preparation Decomposed cowdung at 2.0–2.8 kg m−2 was spread uniformly on the surface of the seedbed before or at land preparation Land plowed 4–5 times by power tiller followed by laddering and allowed to decompose for 7–12 days to make the soil soft and muddy When soils became sufficiently soft and muddy, 1.0-m-wide raised beds (length as long as the length of the land) were made surrounded by a 25-cm-wide and 10–15-cm deep canal. The soils from the canal were put on the bed. The number of beds made depended on the land size and requirement Surface of the raised bed was leveled by a plane and flat wood Soils at the surface of the raised bed were allowed to settle for 2–4 h before seed sowing 90–110 Canals surrounding the raised beds always kept full with water to prevent soil cracking at the surface of the raised bed. Soil surface of the raised bed was not allowed to crack 25–32 in Aman and 30–45 in Boro
Seed rate (kg ha−1 ) Land for seedbed preparation Seedbed preparation
Seed rate sown on seedbed, range (g m−2 ) Water management Seedling age, range (days) Plant population at planting Hill spacing at transplanting (cm) Number of seedlings per hill (no.) P, K, S, and Zn fertilizer management
20 × 20 3–4 Total amounts of P as triple superphosphate, S as gypsum, Zn as zinc sulfate, and two-thirds of K as KCl were applied basally immediately before transplanting rice in both Boro and Aman seasons. The remaining one-third of K as KCl was applied at panicle initiation. The fertilizer rates (kg ha−1 ) are given below for different upazilas Boro season: 17 P, 62 K, 10 S, and 2 Zn; Aman season: 12 P, 42 K, 10 S, and 2 Zn Boro season: 14 P, 62 K, 8 S, and 2 Zn; Aman season: 10 P, 42 K, 8 S, and 2 Zn With USG option, the rates were 52 kg N ha−1 in Aman and 78 kg N ha−1 in Boro The N rates (kg ha−1 ) for the LCC option in different upazilas are given below: Aman 2006: 104 ± 6 in 4 splits; Aman 2007: 77 ± 11 in 3–4 splits; Aman 2008: 78 ± 9 in 3–4 splits; Boro 2007: 97 ± 13 in 3–4 splits; Boro 2008: 108 ± 15 in 3–4 splits Aman 2008: 78 in 3 splits; Boro 2008: 120 in 4 splits Aman 2008: 78 in 3 splits
Kapasia upazila Monohardi upazila N fertilizer management Kapasia upazila, mean ± S.D.
Monohardi upazila, mean ± S.D. Pakundia upazila, mean ± S.D. a
S.D., standard deviation of the mean.
75 ha−1 , respectively. The added costs associated with BMP in at least 75% of the farmers’ fields were ≥$57–107 ha−1 in Aman and ≥ $46–80 ha−1 in Boro. These added costs for BMP were mainly because of the additional application of P and K and additional labor requirement with BMP. The large ranges in added costs suggest
substantial variation in the amounts of additional P, K, and labor use. The average added net returns with BMP were always positive, ranging from US$22 to 120 ha−1 in Aman and from US$93 to 115 ha−1 in Boro across years (Table 6). The added net returns
Table 4 Rice grain yield with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Aman seasons in farmers’ fields of central Bangladesh, 2006–2008. Grain yield (Mg ha−1 )
Treatment
Aman 2006
FP BMP–USG BMP–LCC
Aman 2007
3.4 b 4.1 a 4.2 a
Borjapur
Saluateki
Shahartoke
Sultanpur
Tokenagar
3.8 b 4.6 a 4.7 a
3.9 b 4.4 a 4.5 a
3.7 b 4.5 a 4.6 a
3.8 b 4.3 a 4.4 a
3.7 b 4.5 a 4.6 a
Aman 2008
FP BMP–USG BMP–LCC
FP BMP–USG BMP–LCC a
Sultanpur
Saluateki
Aralia
Gosherkandi
Tokenagar
Meratala
Motkhola
Egarasindur
4.2 b 4.7 a 4.8 a
4.1 c 4.7 b 5.1 a
4.0 c 4.6 b 5.0 a
4.0 c 4.7 b 5.1 a
3.9 b 4.6 a 4.7 a
4.3 b 5.0 a 5.1 a
3.9 b 4.4 a 4.5 a
4.1 b 5.0 a 5.0 a
Kodalia
Kathalia
Araihazar
Harordia
Hatirdia
Kocherchar
Kachikata
Dosdona
3.9 b 4.4 a 4.5 a
3.8 b 4.4 a 4.6 a
3.8 b 4.3 a 4.4 a
3.9 c 4.7 b 5.0 a
3.9 b 4.6 a 4.7 a
4.4 b 5.1 a 5.3 a
4.0 b 4.7 a 4.8 a
4.1 b 4.7 a 4.9 a
In a column, means followed by the same letter are not significantly different at the 5% level by Duncan’s multiple range test.
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
67
Table 5 Rice grain yield with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Boro seasons in farmers’ fields of central Bangladesh, 2006–2008. Grain yield (Mg ha−1 )a
Treatment Boro 2007
FP BMP–USG BMP–LCC
Boro 2008
5.6 b 6.4 a 6.5 a
FP BMP–USG BMP–LCC a
Baraigao
Borjapur
Saluateki
Shahartoke
Sultanpur
Tokenagar
Barapusia
6.9 c 7.8 a 7.4 b
7.4 c 7.7 b 7.9 a
6.9 b 7.1 a 7.3 a
6.8 b 7.1 a 7.3 a
7.3 c 7.5 b 7.8 a
7.7 c 8.0 b 8.3 a
7.0 c 7.8 a 7.5 b
Harordia
Kachikata
Pabur
Roynanda
Sonaroa
Uttarkhamer
Bagua
6.3 c 7.1 b 7.4 a
6.3 b 7.3 a 7.4 a
6.8 b 7.8 a 7.7 a
6.9 c 8.1 b 8.4 a
7.2 b 7.7 a 7.6 a
6.6 b 7.8 a 7.8 a
6.8 b 7.8 a 7.6 a
In a column, means followed by the same letter are not significantly different at the 5% level by Duncan’s multiple range test.
Table 6 Added costs and added net returns with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Aman and Boro seasons in farmers’ fields of central Bangladesh, 2006–2008. Best management practices
Added cost (US$ ha−1 )
Added net return (US$ ha−1 )
Mean
Range
25% quartile
Aman 2006 BMP–LCC BMP–USG
66 55
50 to 283 45 to 64
57 51
Aman 2007 BMP–LCC BMP–USG
93 99
60 to 112 66 to 110
Aman 2008 BMP–LCC BMP–USG
82 74
Boro 2007 BMP–LCC BMP–USG Boro 2008 BMP–LCC BMP–USG
Mean
Range
63 57
92 59
−2 to 1039 −54 to 179
89 95
97 103
43 22
1 to 139 −7 to 131
77 69
107 99
33 39
−19 to 95 −11 to 103
17 22
70 75
−42 to 106 −41 to 113
66 66
associated with BMP in at least 75% of the farmers’ fields were ≥$13 to 107 ha−1 in Aman and ≥$130 to 160 ha−1 in Boro. The average added net returns were always higher with BMP–LCC than with BMP–USG by 48–55.9% in Aman and by 16–17% in Boro, suggesting LCC-aided N management as a more profitable option than USG. The negative added net returns in at least 25% of the farmers’ fields were mainly due to high added costs ($74–99 ha−1 ) compared with added returns from increased yield (0.61–0.8 t ha−1 ) in Aman 2007, Aman 2008, and Boro 2008. The large ranges in added net returns ($−2 to 1039 ha−1 ) indicate large farm-to-farm variability in the costs and gains associated with inputs and outputs for using BMP along with either LCC or USG across seasons and years. The positive added net returns with the use of BMP in the vast majority of farmers’ fields confirmed that BMP are ready for wide-scale evaluation and promotion in Bangladesh. Our results suggest that BMP along with either LCC or USG can be promoted through integration with the farmers’ own portfolio of practices for increased profitability and the economic benefit is more with the LCC-aided N management option than with USG. The markedly increased yield and profit for rice in Bangladesh reported in another study (Alam et al., 2005) arise from the use of LCC-based N management combined with improved management of other nutrients and improved weed control. The added costs associated with BMP arising from increased fertilizer use and more labor required suggest that the adoption of BMP depends on fertilizer availability and price, availability of labor, wage rate,
75% quartile
25% quartile
75% quartile
35 24
107 95
−317 to 481 −290 to 25
−24 −43
89 48
120 81
−127 to 658 −388 to 300
−45 −74
80 13
46 52
115 99
−165 to 511 −185 to 542
46 27
160 146
75 80
109 93
−355 to 647 −259 to 531
17 −32
154 130
and opportunity costs of farmers’ time. Such labor-demanding practices could, however, become less attractive in the future as wages and the opportunity cost of labor increase with progress in economic development (Dawe et al., 2004). 4. Conclusions The data from numerous farmers’ fields show that the integration of improved crop management practices with the farmers’ management practices has potential to boost rice productivity and production in Bangladesh. Farmers tend to use sub-optimal crop and nutrient management, especially N fertilizer. Our study shows 3–28% grain yield increases with BMP resulting in farmers’ net profit increase of US$22 to120 ha−1 . Adoption of BMP in an integrated way is crucial to attain the productivity and profit increases. A massive and effective program for wider demonstration, refinement, and dissemination of BMP in an integrated way throughout the country may be taken to boost the productivity and profit from rice farming. The inclusion of labor-saving technologies in BMP may make the program more attractive to farmers in the future. Training for extension workers to prepare a highly qualified professional workforce and public–private partnerships will also be very important for a successful program. Acknowledgments We acknowledge the “Accelerating Adoption of ResourceConserving Technologies in South Asia” Project funded by the
68
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
USAID famine fund for supporting this study. This work was made possible through technical inputs from the Rice-Wheat Consortium. We thank the Bangladesh Rice Research Institute and Department of Agricultural Extension, Bangladesh, for their collaboration in conducting the study. References Alam, M.M., Ladha, J.K., Foyjunnessa Rahman, Z., Khan, S.R., Rashid, H., Khan, A.H., Buresh, R.J., 2006. Nutrient management for increased productivity of ricewheat cropping system in Bangladesh. Field Crops Res. 96, 374–386. Alam, M.M., Ladha, J.K., Khan, S.R., Foyjunnessa Rashid, H., Khan, A.H., Buresh, R.J., 2005. Leaf color chart for managing nitrogen fertilizer in lowland rice in Bangladesh. Agron. J. 97, 949–959. Alam, M.M., Buresh, R.J., Ladha, J.K., Khan, A.H., 2004. Site-Specific Nutrient Management (SSNM) for Rice. Bangladesh Rice Research Institute, Gazipur, pp. 13–17. Aureus, M., 2011. The Rise of Rice. Rice Today, Jan–Mar. International Rice Research ˜ Philippines, p. 16. Institute (IRRI), Los Banos, Bangladesh Economic Review, 2009. Government of the People’s Republic of Bangladesh. Ministry of Finance, Dhaka. BRRI, 2010. Annual Internal Review Report for 2009–10. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2009. Annual Internal Review Report for 2008–09. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2008. Annual Internal Review Report for 2007–08. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2007. Annual Internal Review Report for 2006–07. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2004. Adhunik Daner Chas. Bangladesh Rice Research Institute, Gazipur, Bangladesh (A Bangali Publication). Dawe, D., Dobermann, A., Witt, C., Abdulrachman, S., Gines, H.C., Nagarajan, R., Satawathananont, S., Son, T.T., Tan, P.S., Wang, G.H., 2004. Nutrient management in the rice soils of Asia and the potential of site-specific nutrient management. In: Dobermann, et al. (Eds.), Increasing productivity of intensive rice systems through site-specific nutrient management. Science Publishers and IRRI, Enfield, ˜ NH, and Los Banos, Philippines., pp. 337–358.
FAO, 2012. FAOSTAT. Food and Agriculture Organization of the United Nations – for a world without hunger. Statistics Division. http://faostat.fao.org/ FAO, 2010. FAOSTAT. Food and Agriculture Organization of the United Nations – for a world without hunger. Statistics Division. http://faostat.fao.org/ FRG, 2005. Fertilizer recommendation guide – 2005. Bangladesh Agricultural Research Council (BARC), Dhaka, Bangladesh. I.R.R.I., 2008. Responding to the Rice Crisis: How IRRI Can Work with its Part˜ Philippines, ners. International Rice Research Institute (IRRI), Los Banos, pp. 1–14. Ladha, J.K., Kumar, V., Alam, M.M., Sharma, S., Gathala, M.K., Chandna, P., Saharawat, Y.P., Balasubramanian, V., 2009. Integrating crop and resource management technologies for enhanced productivity, profitability and sustainability of the rice-wheat system in South Asia. In: Ladha, J.K., Singh Yadvinder, Erenstein, O., Hardy, B. (Eds.), Integrated Crop and Resource Management in the Rice˜ Wheat System of South Asia. International Rice Research Institute, Los Banos, Philippines, p. 69-108. Ladha, J.K., Pathak, H., Padre, A.T., Dawe, D., Gupta, R.K., 2003. Productivity trends in intensive rice-wheat cropping systems in Asia. In: Ladha, J.K., Hill, J., Gupta, R.K., Duxbury, J., Buresh, R.J. (Eds.), Improving the productivity and sustainability of rice-wheat systems: issues and impacts. ASA Special Publication 65. American Society of Agronomy, Madison, WI, pp. 45–76. Mukherjee, N., Choudhury, G.A., Khan, M.F.A., Islam, A.K.M.S., 2011. Implication of changing consumption pattern on food security and water resources in Bangladesh. In: 3rd International Conference on Water and Flood Management (ICWFM-2011). Pathak, H., Ladha, J.K., Aggarwal, P.K., Peng, S., Das, S., Singh, Y., Singh, B., Kamara, S.K., Mishra, B., Sastri, A.S.R.A.S., Aggarwal, H.P., Das, D.K., Gupta, R.K., 2003. Climatic potential and on-farm yield trends of rice and wheat in the Indo-Gangetic plains. Field Crops Res. 80, 223–234. Regmi, A.P., Tripathi, J., Giri, G.S., Bhatta, M.R., Sherchan, D.P., Karki, K.B., Tripathi, B.P., Kumar, V., Ladha, J.K., 2009. Improving food security through integrated crop and resource management in the rice-wheat system in Nepal. In: Ladha, J.K., Singh, Yadvinder, Erenstein, O., Hardy, B. (Eds.), Integrated crop and resource management in the rice-wheat system of South Asia. International Rice Research ˜ Philippines, pp. 69–108. Institute, Los Banos, SAS Institute, 1999. SAS System Version 8. SAS Institute, Cary, NC.
Contents lists available at SciVerse ScienceDirect
Field Crops Research journal homepage: www.elsevier.com/locate/fcr
Integrating best management practices for rice with farmers’ crop management techniques: A potential option for minimizing rice yield gap M. Murshedul Alam a , Md. Rezaul Karim a , J.K. Ladha b,∗ a b
International Rice Research Institute (IRRI), Bangladesh Office, House-9, Road 2/2, Banani, Dhaka 1213, Bangladesh IRRI India Office, 1st Floor, CG Block, NASC Complex, DPS Marg, New Delhi 110012, India
a r t i c l e
i n f o
Article history: Received 5 October 2012 Received in revised form 16 January 2013 Accepted 16 January 2013 Keywords: Rice–rice cropping system Yield gap Productivity Net profit Leaf color chart
a b s t r a c t The major cereal cropping system in Bangladesh is rice (Oryza sativa L.) in the dry season (Boro) – rice in the wet season (Aman). The average productivity (7 t ha−1 ) of this system is far below attainable yields (14 t ha−1 ) in farmers’ fields, resulting in a large yield gap mainly due to farmers’ traditional management practices. We evaluated a set of selected best management practices (BMP) along with two N management options in the farmers’ crop management practices for rice in numerous farmers’ fields across 24 villages over 5 contiguous seasons during 2006–2008. Across years, BMP and two N management options increased grain yields compared with the farmers’ practice (FP) by 0.73 t ha−1 in both Aman and Boro seasons. The higher yield response (24.6% in Aman and 8.6% in Boro season) occurred with BMP in combination with leaf color chart (LCC) aided N management (BMP–LCC) than BMP with Urea Super Granule as N source (BMP–USG). Best management practices and two N management options reduced the yield gap of FP by 45%, with an average of 1.5 Mg ha−1 . The average added net returns with BMP were US$22 to US$120 ha−1 in the wet season and US$93 to US$115 ha−1 in the dry season. Our study shows that the integration of BMP and either LCC-aided N management or USG as an N source with the farmers’ management techniques and its adoption has the potential to boost rice yield and profit and total rice production in Bangladesh. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Rice is a staple food and accounts for more than 40% of the calorie supply of most Asians (IRRI, 2008). About 90% of the world’s rice is grown and produced (142 million ha area with production of 622 million tons) in Asia (FAO, 2010). Economically disadvantaged people spend as much as 30–40% of their income just to buy rice (IRRI, 2008). Increasing the amount of rice production and keeping the rice price low and affordable to them are crucial for poverty reduction. Rice yield must continue to increase at an annual rate of 1.5% compared with the current rate of 0.8% to keep pace with the expected demand (Aureus, 2011). Like in many other Asian countries, rice is the staple food in Bangladesh and it contributes one-half of the agricultural GDP and 55% of the total labor employment (Bangladesh Economic Review, 2009). During the last 20 years, Bangladesh has increased rice production 1.8 times with almost no increase in its land area of about 11.7 million ha (FAO, 2012). However, because of the continuous increase in population growth, rice demand in 2050 is projected to be 56% higher than in 2001 (Mukherjee et al., 2011). This is
∗ Corresponding author. Tel.: +91 11 25843802; fax: +91 11 25841801. E-mail address: [email protected] (J.K. Ladha). 0378-4290/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2013.01.010
going to be doubly challenging in light of the diversion of better quality land, water, and labor to other sectors of the national economy. Evidence is now appearing that the productivity of rice–rice and rice–wheat systems is plateauing because of a fatigued natural resource base (Ladha et al., 2003; Pathak et al., 2003). The productivity and sustainability of rice-based cropping systems are threatened because of (i) the inefficient use of fertilizer, water, and labor; (ii) increasing scarcity of water and labor; (iii) changing climate; (iv) emerging energy crisis and rising fuel prices; and (v) emerging socioeconomic changes such as urbanization, migration of labor, and preference for nonagricultural work (Ladha et al., 2009). In addition, recent increases in the prices of farm inputs in relation to outputs, fewer off-farm work opportunities for supplementing farm income, reduced remittances from relatives working outside villages, and declining income and purchasing power of poor consumers have threatened the existence of rice producers and consumers (Ladha et al., 2009). Bangladesh produced about 50.1 million tons of rough rice from 11.7 m ha of land in 2010 with a productivity of 4.3 t ha−1 (FAO, 2012). The present productivity is far below the attainable yield of 8–10 t ha−1 in the dry season (Boro) and 5–6 t ha−1 in the wet season (Transplanted Aman) in farmers’ field experiments (BRRI, 2007, 2008, 2009, 2010). This difference in yield in farmers’ fields between farmer-managed and researcher-managed trials is
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
mainly due to the differences in management practices adapted by researchers and farmers. The causes of such a yield gap are classified into two broad categories: (i) biotic factors such as poor-quality seeds and seedlings, insects, diseases, weeds, and rodents; and (ii) abiotic factors such as soil, nutrients, and water. However, a large portion of this yield gap remains unexplained. Given the continuously increasing population, coupled with scarcity of land, water, labor, and energy, Bangladesh is under tremendous pressure to increase, by all possible means, rice production. Since the possibility of expanding rice area is limited, the extra rice production must come from productivity increases. The major challenge is to minimize or bridge the yield gap in farmers’ fields and increase productivity with less water, less agro-chemicals, and less labor, thereby making rice farming more profitable and sustainable in the long term. An adequate understanding and the development of appropriate practical technologies to minimize the causes of the yield gap are critical for meeting the challenges of continued gains in rice production, without degrading the natural resource base. Minimizing the yield gap and increasing profit and product quality are becoming increasingly difficult to achieve by using a single-technology-centric approach. The use of a component technology in isolation also has limited widespread adoption. Combining and simultaneously applying a number of the best compatible individual or component technologies is crucial for maximizing overall benefits to farmers. Depending on the need and profitability of new technologies, farmers generally integrate new technology with the cultural practices being practiced by them on their farms. This process of integrating new technologies with existing farmers’ practice has been referred to as integrated crop and resource management (ICRM) or best management practices (Ladha et al., 2009). The choice or selection of the best individual or component technologies in an integrated manner is crucial for achieving the full benefit. The integration of best component technologies should (i) ensure timely crop establishment and
63
uniform crop stands, resulting in higher crop yields; (ii) enhance resource- or input-use efficiency; (iii) provide immediate, identifiable, and demonstrable economic benefits; and (iv) reduce adverse effects on the environment and build up soil fertility over the long term. This study therefore evaluated a set of selected best management practices integrated with the farmers’ management techniques in rice for productivity and profitability in many farmers’ fields over five seasons during three years.
2. Materials and methods 2.1. Experimental sites and seasons On-farm trials were conducted in a continuous rice–rice cropping system for five consecutive seasons during 2006–2008 at 24 villages located within a radius of 50 km in central Bangladesh. Number of selected villages were 14 in Kapasia Upazila (23◦ 55 to 24◦ 12 N and 90◦ 29 to 90◦ 43 E) of Gazipur District, 4 in Pakundia Upazila (24◦ 16 to 24◦ 0 N and 90◦ 46 E) of Kishoreganj District, and 6 in Monohardi Upazila (24◦ 03 to 24◦ 16 N and 90◦ 38 to 90◦ 49 E) of Narshingdi District (Table 1). The villages which represent common cropping systems, farmers’ management practices, land types, and soil types were selected. Each village had 800–1500 farm families. All the villages in Gazipur District belong to Agro-Ecological Zone (AEZ) 28 (Madhupur Tract), AEZ 8 (Young Brahmaputra and Jamuna Floodplains), and AEZ 9 (Old Brahmaputra Floodplain), whereas the villages in Kishoreganj District belong to AEZ 8 and AEZ 9 and those in Narshingdi District belong to AEZ 9. The soils are (a) well drained friable clay loam to clay or heavy clay, strongly acidic, mainly phosphate fixing, low in P, K, S, and B in AEZ – 28, (b) permeable silt loam to silty clay loam, neutral to slightly acidic, deficient in N, P and S but medium in K and Zn status in AEZ – 8, and (c) silt loam to silty clay loam, moderately acidic to neutral, with low K and medium P status in AEZ – 9 (FRG, 2005).
Table 1 Locations and number of farmers’ fields where the trials were conducted in central Bangladesh in Aman and Boro seasons during 2006–2008. Location District
Number of farmers’ fields Upazila
Village
Agro-ecological zone (AEZ) – 28, AEZ – 8 and AEZ – 9 Kapasia Saluateki Gazipur Sultanpur Tokenagar Borjapur Shahartoke Bagua Baraigao Barapusia Pabur Roynanda Sonaroa Uttarkhamer Gosherkandi Aralia AEZ – 9 Narshingdi
Monohardi
Harordia Kachikata Kathalia Hatirdia Kocherchar Dosdona
AEZ – 8 and AEZ – 9 Kishoreganj
Pakundia
Kodalia Egarasindur Motkhola Meratala
Total
Aman 2006 9 29
Boro 2007 14 32 29 31 23
Aman 2007 33 28 20 15 12
Boro 2008 19 49 34 17 13 7 7 7 7 7 7 7
Aman 2008 15 48 42
31 25 20 20
10 6 16 10 10 10 10 10 15 5
38
129
108
221
263
64
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
Transplanted rice was grown on puddled soil under rainfed conditions in the wet season or Aman (June–July to November–December) and under irrigated conditions in the dry season or Boro (December–January to April–May). In both seasons, soils were well prepared through 4–5 passes with a power tiller followed by 2–3 ladderings. High-yielding semi-dwarf rice cultivars BR 11 (145 d seed to seed) and BRRI dhan 28 (140 d seed to seed) were grown in the Aman and Boro seasons, respectively.
(BMP–USG). The BMP consisted of options for managing farmers’ yield-limiting constraints identified from a farmers’ survey carried out prior to the start of the trials. The identified constraints were poor quality of seeds and seedlings, sub-optimal planting method, and sub-optimal fertilizer application and N timing and N source. In BMP–LCC and BMP–USG treatments, crop management practices other than BMP coupled with N management were done as practiced by the farmers in FP.
2.2. Experimental design and treatments
2.3. Treatment details
The trials were established in farmers’ fields in a randomized complete block design with a set of three treatments in each farmer’s field. Farmers’ fields were considered as replications. The number of replicate farmers varied among villages and seasons (Table 1). In a village, the farmers’ fields were selected in a contiguous fashion forming a block and farmers of the selected fields belong to marginal, small, and medium farmer groups. Out of three treatments, treatment number 1 was farmers’ management practices (FP), treatment number 2 was best management practices (BMP) coupled with leaf color chart (LCC)-aided fertilizer N management (BMP–LCC) and treatment number 3 was BMP coupled with urea super granules (USG) as source of fertilizer N
2.3.1. Treatment number 1 – FP The farmers’ management practice of seed and seedling raising, planting method, and fertilization varied considerably among farmers and seasons (Table 2). All farmers used seeds from their own source having 80–90% germination capacity without any seed treatment. They prepared a flat seedbed without any organic matter and used a very high seed rate (135–165 g m−2 ) on the seedbed. Five to eight 28–45-day-old rice seedlings per hill in Aman and five to seven 28–61-day-old rice seedlings in the Boro season were transplanted at random spacing. Across seasons and years, all farmers applied N fertilizer at early tillering [(10–20 days after transplanting (DAT)] and late tillering
Table 2 Farmers’ management practices for transplanted rice in central Bangladesh in Aman and Boro seasons during 2006–2008a .
a b
Parameterb
Aman 2006
Aman 2007
Aman 2008
Boro 2007
Boro 2008
Seed rate on seed bed, range (g m−2 ) Seed rate to transplant 1 ha main field (kg) Seedling age, range (days) Hill spacing at transplanting (cm) Seedling number (hill−1 )
145–165 65 ± 9 28–39 Random 5–6
135–150 48 ± 8 30–45 Random 5–6
138–149 33 ± 6 33–34 Random 5–8
155–165 48 ± 6 29–61 Random 5–7
154–160 39 ± 4 28–45 Random 5–6
Fertilizer management N applied at basal Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
0 0
0 0
N applied at early tillering Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
100 10 ± 2 46 ± 4
100 16 ± 3 31 ± 3
100 18 ± 4 49 ± 3
100 16 ± 5 53 ± 14
100 20 ± 3 57 ± 4
N applied at late tillering Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
100 26 ± 3 69 ± 18
100 36 ± 3 41 ± 14
100 38 ± 4 55 ± 4
100 28 ± 3 55 ± 13
100 36 ± 3 47 ± 6
N applied at panicle initiation Farmer (%) Time ± S.D. (DAT) Amount ± S.D. (kg ha−1 )
0 0 0
35 51 ± 2 21 ± 2
0 0 0
93 40 ± 6 52 ± 13
100 45 ± 4 47 ± 6
Fraction of total N applied At basal (%) At early tillering (%) At late tillering (%) At panicle initiation (%)
0 40 60 0
0 33 44 23
0 47 53 0
0 33 34 33
0 38 31 31
P applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
93 23 ± 7
100 17 ± 4
K applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
79 38 ± 13
100 37 ± 5
S applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
59 13 ± 5
97 12 ± 2
Zn applied Farmer (%) Amount ± S.D. (kg ha−1 )
0 0
0 0
0 0
15 4±1
52 2±2
In all cases, farmers’ own seed was the source, seeds were not treated, seedbed had a flat surface, no decomposed cowdung was used, and hill spacing at TP was random. Mean ± standard deviation (S.D.). All reported means for timing and amount of fertilizers are based on farmers applying fertilizers.
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
(26–36 DAT) and 0–100% of the farmers applied at panicle initiation (40–51 DAT). The mean total N dose was 93–160 kg ha−1 . Farmers applied on average 67–100% of fertilizer N between early tillering and late tillering (Table 2). In the Aman season, no farmer applied P, K, S, and Zn fertilizer. In the Boro season, 93–100% of the farmers applied P fertilizer, 79–100% applied K fertilizer, 59–97% applied S fertilizer, and 15–52% applied Zn fertilizer. The average P, K, S, and Zn application rates for farmers ranged 17–23 kg ha−1 , 37–38 kg ha−1 , 12–13 kg ha−1 and 2–4 kg ha−1 , respectively. 2.3.2. Treatment number 2 – BMP–LCC A set of five BMP integrated with existing farmers’ practices were (i) certified seeds, (ii) healthy young seedlings, (iii) hill spacing of 20 cm × 20 cm at transplanting with 3–4 seedlings per hill, (iv) AEZ-based recommended P, K, S, and Zn fertilization, and (v) N application by using the LCC in a real-time N management approach. The management practices recommended by BRRI (2004) were used for raising healthy seedlings (Table 3). Across the years, seedling age ranged from 25 to 32 days in Aman and from 30 to 45 days in Boro. The P, K, S, and Zn fertilizers were applied at the rates and time recommended by BRRI (2004) based on AEZ-based recommendations. The P and K rates varied between locations and between seasons while the S and Zn rates were the same across seasons and locations (Table 3). Nitrogen as urea was applied by using the LCC in a real-time N management approach (Alam et al., 2004). The N rate varied among farmers’ fields and seasons. The mean total N dose ranged from 77 to 104 kg ha−1 in Aman and from 97 to 120 kg ha−1 in Boro (Table 3). 2.3.3. Treatment number 3 – BMP–USG The management practices in BMP–USG were same as BMP–LCC except N fertilizer management. Nitrogen as USG was deep-placed within 7–10 DAT at 7–10-cm soil depth at the rate of one USG (1.8 g) in Aman and one USG (2.7 g) in Boro in the middle of 4 hills. The total N dose was 52 kg ha−1 in Aman and 78 kg ha−1 in Boro (Table 3). 2.4. Data collection and analysis Grain yields (rough rice) were obtained from two central 6 m2 harvest areas in each plot at harvestable maturity and reported at 0.14 g H2 O g−1 fresh weight of grain. Human labor used for seedling raising, transplanting, and fertilizer application and wage rate were recorded. The time required to complete each field operation was expressed as person-days ha−1 , considering 8 h to be equivalent to 1 person-day. The farm-gate price of rough rice and prices of seed, fertilizers, and seed-treating chemicals were also recorded. Analysis of variance (ANOVA) was performed with the PROC MIXED procedure of the SAS/STAT software (SAS Institute, 1999) to determine the effects of villages and treatments on yield, input uses, and added net return. Duncan’s multiple range test (DMRT) was used at the P < 0.05 level to test the differences between the treatment means. Unless indicated otherwise, differences were considered significant only when P ≤ 0.05. Descriptive statistics such as means, standard deviation, range, 25% quartile, and 75% quartile were used to determine the variability of parameters. Treatments were evaluated based on added net return relative to the farmers’ practice, which is the difference between added gross return and added cost for a treatment as compared with FP. Added gross return equaled additional yield as rough rice (yield of treatment − yield of FP) multiplied by the price of yield. Added cost equaled the sum of costs for differences in labor determined as [(labor for seedling raising, transplanting, and fertilizer application of the treatment − labor for seedling raising, transplanting, and fertilizer application of FP) × wage rate] and costs for differences in fertilizer and seed-treating chemicals determined as (fertilizer and seed-treating chemical costs of treatment − fertilizer and
65
seed-treating chemical costs of FP). The prices of seeds, fertilizer, seed-treating chemical, rough rice, and labor wage were as follows: seed = US$0.29 to 0.51 kg−1 , P fertilizer = US$1.61 to 3.65 kg−1 P, K fertilizer = US$0.41 to 1.31 kg−1 K, S fertilizer = US$0.49 to 0.97 kg−1 S, Zn fertilizer = US$2.64 to 3.65 kg−1 Zn, urea = US$0.22 to 0.38 kg−1 N, USG = US$0.25 to 0.41 kg−1 N, seed-treating chemical = US$36.40 to 38.70 kg−1 , rough rice = US$0.15 to 0.26 kg−1 , and labor wage = US$1.46 to 2.63 person-day−1 (US$1 = Bangladesh Taka 68.5). 3. Results and discussion 3.1. Grain yield The treatment of best management practices along with two N management treatments, either LCC or USG, had a significant response in rice grain yield in both Aman and Boro seasons in all the villages and all three years (Tables 4 and 5). On average, the yields in Boro were 65% higher than the yields in Aman, with the highest yield of 8.4 Mg ha−1 in Boro and 5.3 Mg ha−1 in Aman. Compared with the farmers’ practice, the yield increases with BMP and two N management options ranged from 12 to 28% in Aman and from 3 to 22% in Boro, with an average of 0.73 Mg ha−1 in both seasons (Tables 4 and 5). BMP with LCC on average had a higher response than BMP with USG by 24.6% in Aman and by 8.6% in Boro. Across villages and years, the average yield for the system (combined yield of Aman and Boro seasons within a cropping year) for FP (10.7 Mg ha−1 ) and for BMP along with N management options (12.2 Mg ha−1 ) resulted in a yield gap of 3.3 Mg ha−1 in FP and 1.8 Mg ha−1 in BMP based on the BRRI-reported (BRRI, 2007, 2008, 2009, 2010) average attainable yield of 14 Mg ha−1 for the system. Thus, in this study, BMP along with any of two N management options reduced the yield gap of the farmers’ practice by 1.5 Mg ha−1 (45%). In the 2007 and 2008 Aman seasons and the 2008 Boro season, when more villages were included, the interaction effect between treatment and village was significant. In these seasons, the yields obtained in BMP along with either LCC or USG were higher by 0.5–1.1 Mg ha−1 than the yields in FP across villages (Tables 4 and 5). The yields of BMP–LCC and BMP–USG were always similar in the range 4.3–7.8 Mg ha−1 except in 4 villages in Aman 2008 and in 7 villages in Boro 2008. In all 4 villages in Aman and 5 of 7 villages in Boro, BMP–LCC produced the highest yields (5.0–8.4 Mg ha−1 ) and in 2 villages in Boro BMP–USG produced the highest yield (7.8 Mg ha−1 ). In the fields where BMP–USG produced the highest yields, the soils were always either saturated or had standing water, suggesting a better performance of USG under an assured water source. The positive response of BMP along with two N management options was due to the combined effects of certified seeds with seed treatment, healthy seedlings, optimum planting density, balanced fertilization, and better synchronization of N fertilization with plant needs. These findings suggest potential to increase farmers’ rice yields through BMP in an integrated way. Other studies in India and Bangladesh rice fields also reported comparatively higher yields in the range 0.8–2.0 Mg ha−1 with the use of improved methods of more than one crop production factor in an integrated way (Regmi et al., 2009; Alam et al., 2005, 2006). 3.2. Economic analysis Added costs and added net returns for BMP relative to the FP were used to assess profitability for the use of BMP along with either LCC or USG (Table 6). The average added costs for BMP in Aman and Boro across years ranged from $55 to 99 ha−1 and from $33 to
66
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
Table 3 Best management practices (BMP) for selected crop production factors integrated with farmers’ management techniques for transplanted rice and applied by farmers in central Bangladesh in Aman and Boro seasons during 2006–2008. Crop production factors and parametera
Best management practices (BMP)
Seed and seedlings Seed source Seed germination (%) Seed treatment
Certified seeds from BRRI and Bangladesh Agricultural Development Corporation (BADC) 90–95 Seeds treated with bavistine at 3 g per kg seeds. Seeds were soaked in water mixed with bavistine for 24 h 30 kg seeds used to transplant 1 ha main field Land having loamy or clay loam soils, sufficient sunlight, irrigation, and drainage facility was selected for seedbed preparation Decomposed cowdung at 2.0–2.8 kg m−2 was spread uniformly on the surface of the seedbed before or at land preparation Land plowed 4–5 times by power tiller followed by laddering and allowed to decompose for 7–12 days to make the soil soft and muddy When soils became sufficiently soft and muddy, 1.0-m-wide raised beds (length as long as the length of the land) were made surrounded by a 25-cm-wide and 10–15-cm deep canal. The soils from the canal were put on the bed. The number of beds made depended on the land size and requirement Surface of the raised bed was leveled by a plane and flat wood Soils at the surface of the raised bed were allowed to settle for 2–4 h before seed sowing 90–110 Canals surrounding the raised beds always kept full with water to prevent soil cracking at the surface of the raised bed. Soil surface of the raised bed was not allowed to crack 25–32 in Aman and 30–45 in Boro
Seed rate (kg ha−1 ) Land for seedbed preparation Seedbed preparation
Seed rate sown on seedbed, range (g m−2 ) Water management Seedling age, range (days) Plant population at planting Hill spacing at transplanting (cm) Number of seedlings per hill (no.) P, K, S, and Zn fertilizer management
20 × 20 3–4 Total amounts of P as triple superphosphate, S as gypsum, Zn as zinc sulfate, and two-thirds of K as KCl were applied basally immediately before transplanting rice in both Boro and Aman seasons. The remaining one-third of K as KCl was applied at panicle initiation. The fertilizer rates (kg ha−1 ) are given below for different upazilas Boro season: 17 P, 62 K, 10 S, and 2 Zn; Aman season: 12 P, 42 K, 10 S, and 2 Zn Boro season: 14 P, 62 K, 8 S, and 2 Zn; Aman season: 10 P, 42 K, 8 S, and 2 Zn With USG option, the rates were 52 kg N ha−1 in Aman and 78 kg N ha−1 in Boro The N rates (kg ha−1 ) for the LCC option in different upazilas are given below: Aman 2006: 104 ± 6 in 4 splits; Aman 2007: 77 ± 11 in 3–4 splits; Aman 2008: 78 ± 9 in 3–4 splits; Boro 2007: 97 ± 13 in 3–4 splits; Boro 2008: 108 ± 15 in 3–4 splits Aman 2008: 78 in 3 splits; Boro 2008: 120 in 4 splits Aman 2008: 78 in 3 splits
Kapasia upazila Monohardi upazila N fertilizer management Kapasia upazila, mean ± S.D.
Monohardi upazila, mean ± S.D. Pakundia upazila, mean ± S.D. a
S.D., standard deviation of the mean.
75 ha−1 , respectively. The added costs associated with BMP in at least 75% of the farmers’ fields were ≥$57–107 ha−1 in Aman and ≥ $46–80 ha−1 in Boro. These added costs for BMP were mainly because of the additional application of P and K and additional labor requirement with BMP. The large ranges in added costs suggest
substantial variation in the amounts of additional P, K, and labor use. The average added net returns with BMP were always positive, ranging from US$22 to 120 ha−1 in Aman and from US$93 to 115 ha−1 in Boro across years (Table 6). The added net returns
Table 4 Rice grain yield with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Aman seasons in farmers’ fields of central Bangladesh, 2006–2008. Grain yield (Mg ha−1 )
Treatment
Aman 2006
FP BMP–USG BMP–LCC
Aman 2007
3.4 b 4.1 a 4.2 a
Borjapur
Saluateki
Shahartoke
Sultanpur
Tokenagar
3.8 b 4.6 a 4.7 a
3.9 b 4.4 a 4.5 a
3.7 b 4.5 a 4.6 a
3.8 b 4.3 a 4.4 a
3.7 b 4.5 a 4.6 a
Aman 2008
FP BMP–USG BMP–LCC
FP BMP–USG BMP–LCC a
Sultanpur
Saluateki
Aralia
Gosherkandi
Tokenagar
Meratala
Motkhola
Egarasindur
4.2 b 4.7 a 4.8 a
4.1 c 4.7 b 5.1 a
4.0 c 4.6 b 5.0 a
4.0 c 4.7 b 5.1 a
3.9 b 4.6 a 4.7 a
4.3 b 5.0 a 5.1 a
3.9 b 4.4 a 4.5 a
4.1 b 5.0 a 5.0 a
Kodalia
Kathalia
Araihazar
Harordia
Hatirdia
Kocherchar
Kachikata
Dosdona
3.9 b 4.4 a 4.5 a
3.8 b 4.4 a 4.6 a
3.8 b 4.3 a 4.4 a
3.9 c 4.7 b 5.0 a
3.9 b 4.6 a 4.7 a
4.4 b 5.1 a 5.3 a
4.0 b 4.7 a 4.8 a
4.1 b 4.7 a 4.9 a
In a column, means followed by the same letter are not significantly different at the 5% level by Duncan’s multiple range test.
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
67
Table 5 Rice grain yield with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Boro seasons in farmers’ fields of central Bangladesh, 2006–2008. Grain yield (Mg ha−1 )a
Treatment Boro 2007
FP BMP–USG BMP–LCC
Boro 2008
5.6 b 6.4 a 6.5 a
FP BMP–USG BMP–LCC a
Baraigao
Borjapur
Saluateki
Shahartoke
Sultanpur
Tokenagar
Barapusia
6.9 c 7.8 a 7.4 b
7.4 c 7.7 b 7.9 a
6.9 b 7.1 a 7.3 a
6.8 b 7.1 a 7.3 a
7.3 c 7.5 b 7.8 a
7.7 c 8.0 b 8.3 a
7.0 c 7.8 a 7.5 b
Harordia
Kachikata
Pabur
Roynanda
Sonaroa
Uttarkhamer
Bagua
6.3 c 7.1 b 7.4 a
6.3 b 7.3 a 7.4 a
6.8 b 7.8 a 7.7 a
6.9 c 8.1 b 8.4 a
7.2 b 7.7 a 7.6 a
6.6 b 7.8 a 7.8 a
6.8 b 7.8 a 7.6 a
In a column, means followed by the same letter are not significantly different at the 5% level by Duncan’s multiple range test.
Table 6 Added costs and added net returns with farmers’ management practices (FP), best management practices (BMP) coupled with leaf color chart (LCC) aided nitrogen (N) management (BMP–LCC) and BMP with urea super granules (USG) as fertilizer N (BMP–USG) during Aman and Boro seasons in farmers’ fields of central Bangladesh, 2006–2008. Best management practices
Added cost (US$ ha−1 )
Added net return (US$ ha−1 )
Mean
Range
25% quartile
Aman 2006 BMP–LCC BMP–USG
66 55
50 to 283 45 to 64
57 51
Aman 2007 BMP–LCC BMP–USG
93 99
60 to 112 66 to 110
Aman 2008 BMP–LCC BMP–USG
82 74
Boro 2007 BMP–LCC BMP–USG Boro 2008 BMP–LCC BMP–USG
Mean
Range
63 57
92 59
−2 to 1039 −54 to 179
89 95
97 103
43 22
1 to 139 −7 to 131
77 69
107 99
33 39
−19 to 95 −11 to 103
17 22
70 75
−42 to 106 −41 to 113
66 66
associated with BMP in at least 75% of the farmers’ fields were ≥$13 to 107 ha−1 in Aman and ≥$130 to 160 ha−1 in Boro. The average added net returns were always higher with BMP–LCC than with BMP–USG by 48–55.9% in Aman and by 16–17% in Boro, suggesting LCC-aided N management as a more profitable option than USG. The negative added net returns in at least 25% of the farmers’ fields were mainly due to high added costs ($74–99 ha−1 ) compared with added returns from increased yield (0.61–0.8 t ha−1 ) in Aman 2007, Aman 2008, and Boro 2008. The large ranges in added net returns ($−2 to 1039 ha−1 ) indicate large farm-to-farm variability in the costs and gains associated with inputs and outputs for using BMP along with either LCC or USG across seasons and years. The positive added net returns with the use of BMP in the vast majority of farmers’ fields confirmed that BMP are ready for wide-scale evaluation and promotion in Bangladesh. Our results suggest that BMP along with either LCC or USG can be promoted through integration with the farmers’ own portfolio of practices for increased profitability and the economic benefit is more with the LCC-aided N management option than with USG. The markedly increased yield and profit for rice in Bangladesh reported in another study (Alam et al., 2005) arise from the use of LCC-based N management combined with improved management of other nutrients and improved weed control. The added costs associated with BMP arising from increased fertilizer use and more labor required suggest that the adoption of BMP depends on fertilizer availability and price, availability of labor, wage rate,
75% quartile
25% quartile
75% quartile
35 24
107 95
−317 to 481 −290 to 25
−24 −43
89 48
120 81
−127 to 658 −388 to 300
−45 −74
80 13
46 52
115 99
−165 to 511 −185 to 542
46 27
160 146
75 80
109 93
−355 to 647 −259 to 531
17 −32
154 130
and opportunity costs of farmers’ time. Such labor-demanding practices could, however, become less attractive in the future as wages and the opportunity cost of labor increase with progress in economic development (Dawe et al., 2004). 4. Conclusions The data from numerous farmers’ fields show that the integration of improved crop management practices with the farmers’ management practices has potential to boost rice productivity and production in Bangladesh. Farmers tend to use sub-optimal crop and nutrient management, especially N fertilizer. Our study shows 3–28% grain yield increases with BMP resulting in farmers’ net profit increase of US$22 to120 ha−1 . Adoption of BMP in an integrated way is crucial to attain the productivity and profit increases. A massive and effective program for wider demonstration, refinement, and dissemination of BMP in an integrated way throughout the country may be taken to boost the productivity and profit from rice farming. The inclusion of labor-saving technologies in BMP may make the program more attractive to farmers in the future. Training for extension workers to prepare a highly qualified professional workforce and public–private partnerships will also be very important for a successful program. Acknowledgments We acknowledge the “Accelerating Adoption of ResourceConserving Technologies in South Asia” Project funded by the
68
M.M. Alam et al. / Field Crops Research 144 (2013) 62–68
USAID famine fund for supporting this study. This work was made possible through technical inputs from the Rice-Wheat Consortium. We thank the Bangladesh Rice Research Institute and Department of Agricultural Extension, Bangladesh, for their collaboration in conducting the study. References Alam, M.M., Ladha, J.K., Foyjunnessa Rahman, Z., Khan, S.R., Rashid, H., Khan, A.H., Buresh, R.J., 2006. Nutrient management for increased productivity of ricewheat cropping system in Bangladesh. Field Crops Res. 96, 374–386. Alam, M.M., Ladha, J.K., Khan, S.R., Foyjunnessa Rashid, H., Khan, A.H., Buresh, R.J., 2005. Leaf color chart for managing nitrogen fertilizer in lowland rice in Bangladesh. Agron. J. 97, 949–959. Alam, M.M., Buresh, R.J., Ladha, J.K., Khan, A.H., 2004. Site-Specific Nutrient Management (SSNM) for Rice. Bangladesh Rice Research Institute, Gazipur, pp. 13–17. Aureus, M., 2011. The Rise of Rice. Rice Today, Jan–Mar. International Rice Research ˜ Philippines, p. 16. Institute (IRRI), Los Banos, Bangladesh Economic Review, 2009. Government of the People’s Republic of Bangladesh. Ministry of Finance, Dhaka. BRRI, 2010. Annual Internal Review Report for 2009–10. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2009. Annual Internal Review Report for 2008–09. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2008. Annual Internal Review Report for 2007–08. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2007. Annual Internal Review Report for 2006–07. Bangladesh Rice Research Institute, Gazipur, Bangladesh. BRRI, 2004. Adhunik Daner Chas. Bangladesh Rice Research Institute, Gazipur, Bangladesh (A Bangali Publication). Dawe, D., Dobermann, A., Witt, C., Abdulrachman, S., Gines, H.C., Nagarajan, R., Satawathananont, S., Son, T.T., Tan, P.S., Wang, G.H., 2004. Nutrient management in the rice soils of Asia and the potential of site-specific nutrient management. In: Dobermann, et al. (Eds.), Increasing productivity of intensive rice systems through site-specific nutrient management. Science Publishers and IRRI, Enfield, ˜ NH, and Los Banos, Philippines., pp. 337–358.
FAO, 2012. FAOSTAT. Food and Agriculture Organization of the United Nations – for a world without hunger. Statistics Division. http://faostat.fao.org/ FAO, 2010. FAOSTAT. Food and Agriculture Organization of the United Nations – for a world without hunger. Statistics Division. http://faostat.fao.org/ FRG, 2005. Fertilizer recommendation guide – 2005. Bangladesh Agricultural Research Council (BARC), Dhaka, Bangladesh. I.R.R.I., 2008. Responding to the Rice Crisis: How IRRI Can Work with its Part˜ Philippines, ners. International Rice Research Institute (IRRI), Los Banos, pp. 1–14. Ladha, J.K., Kumar, V., Alam, M.M., Sharma, S., Gathala, M.K., Chandna, P., Saharawat, Y.P., Balasubramanian, V., 2009. Integrating crop and resource management technologies for enhanced productivity, profitability and sustainability of the rice-wheat system in South Asia. In: Ladha, J.K., Singh Yadvinder, Erenstein, O., Hardy, B. (Eds.), Integrated Crop and Resource Management in the Rice˜ Wheat System of South Asia. International Rice Research Institute, Los Banos, Philippines, p. 69-108. Ladha, J.K., Pathak, H., Padre, A.T., Dawe, D., Gupta, R.K., 2003. Productivity trends in intensive rice-wheat cropping systems in Asia. In: Ladha, J.K., Hill, J., Gupta, R.K., Duxbury, J., Buresh, R.J. (Eds.), Improving the productivity and sustainability of rice-wheat systems: issues and impacts. ASA Special Publication 65. American Society of Agronomy, Madison, WI, pp. 45–76. Mukherjee, N., Choudhury, G.A., Khan, M.F.A., Islam, A.K.M.S., 2011. Implication of changing consumption pattern on food security and water resources in Bangladesh. In: 3rd International Conference on Water and Flood Management (ICWFM-2011). Pathak, H., Ladha, J.K., Aggarwal, P.K., Peng, S., Das, S., Singh, Y., Singh, B., Kamara, S.K., Mishra, B., Sastri, A.S.R.A.S., Aggarwal, H.P., Das, D.K., Gupta, R.K., 2003. Climatic potential and on-farm yield trends of rice and wheat in the Indo-Gangetic plains. Field Crops Res. 80, 223–234. Regmi, A.P., Tripathi, J., Giri, G.S., Bhatta, M.R., Sherchan, D.P., Karki, K.B., Tripathi, B.P., Kumar, V., Ladha, J.K., 2009. Improving food security through integrated crop and resource management in the rice-wheat system in Nepal. In: Ladha, J.K., Singh, Yadvinder, Erenstein, O., Hardy, B. (Eds.), Integrated crop and resource management in the rice-wheat system of South Asia. International Rice Research ˜ Philippines, pp. 69–108. Institute, Los Banos, SAS Institute, 1999. SAS System Version 8. SAS Institute, Cary, NC.