The IRRI-CIMMYT joint project "Sustainable intensification of rice-maize (R-M) production systems in Bangladesh" started in November 2008 after an inception workshop on 9-10 November at the Bangladesh Rice Research Institute (BRRI), Gazipur. This 5-year (1 July 2008 to 30 June 2013) project involves multiple partnerships for research and technology transfer on R-M systems in three districts (Comilla, Rajshahi, and Rangpur). The national partners in the project include both government organizations (BRRI; Bangladesh Agricultural Research Institute, BARI; Bangladesh Academy for Rural Development, BARD) and nongovernment organizations (Bangladesh Rural Advancement Committee, Rangpur-Dinajpur Rural Services).The specific objectives of the project are to (1) assess and prioritize constraints to, and opportunities for, the uptake of improved management options in R-M systems; (2) evaluate and identify elite maize germplasm tolerant of excess moisture during rabi and kharif-1 seasons; (3) develop locally adapted management solutions for high-yielding, profitable, resource-efficient, and sustainable R-M systems through on-farm and on-station trials; and (4) build capacity and disseminate and deliver key technologies for R-M systems.
Of the three activities under Objective 1, rapid surveys in 18 villages as envisaged helped generate a rapid insight into the maize sector in each of the project areas through semi-structured focus-group discussions with farmers and different stakeholders. Data from these village surveys are now being entered into predefined statistical packages for social sciences (SPSS) data entry sheets. Information from these diagnostic village surveys will be used for designing a questionnaire for household surveys in August. Under the first activity of Objective 2, two sets of elite hybrids, consisting of 38 from the public and private sector in Bangladesh and 70 BARI test-crosses, were screened for seedling-stage waterlogging tolerance using a "cup" screening method during the rabi season in Gazipur. Concurrently, a third set of 50 hybrids from the public and private sector in India was also screened in Hyderabad, India. In kharif-1, a set of 30 hybrids (28 single- and 3-way crosses between normal and QPM hybrids from CIMMYT, and 2 local hybrids from Bangladesh) was screened for late-vegetative-stage (V8-V10) waterlogging at the BARI research stations in Gazipur and Jamalpur. Only six hybrids from the third set screened in Hyderabad were identified with good tolerance of early-stage excessive moisture. Data are being processed for the first two sets. In kharif-1, only a few plants of only two hybrids survived until maturity in Gazipur but none of the hybrids survived in Jamalpur.
Under Activity 1 of Objective 3, ten adaptive trials with conservation agriculture (CA)-based resource-conserving technologies (RCTs) and another ten site-specific nutrient management (SSNM) trials, both for rabi maize under a rice-maize-mungbean system, were conducted in two upazillas in both Rangpur and Rajshahi districts. Also, six SSNM trials on kharif-1 maize in each of the three districts have now been conducted and the crops are near maturity. Sample processing and data scrutiny for grain and stover yield and for plant nutrient analysis for rabi maize will be completed by the end of May while that for kharif-1 maize will be completed by the end of June. Under Activity 2, on-station experiments on kharif-1 maize with CA-RCTs under a rice-potato-maize system and on mungbean under a rice-maize-mungbean system started at the BARI research station in Comilla. Under Activity 3, existing versions of a power-tiller-operated seeder (PTOS) and bed planter with a fluted-type seed-metering device were tested for planting maize and mungbean during rabi and kharif-1 in 13 farmers' fields in Rangpur and Rajshahi. Based on this, we are now in the process of modifying metering devices for rice and maize.
The first activity, "Build consensus across stakeholders through a Technical Working Group (TWG) and National Working Group," under Objective 4 was not achieved due to other high-priority work in the project. The TWG will be formed and its first meeting will be held by September. Under the second activity, ten researchers, field assistants, and machine operators from the project and another ten operators from outside the project were given demonstrations and hands-on training on the use of a PTOS, bed former/planter, zero-till drill, and strip-till drills at the Wheat Research Centre, Rajshahi. Likewise, 12 researchers and extension workers from the project were trained on the principles and practices of nutrient management in rice and maize. Under the fifth activity, two farmers' field days in the rabi season in Rajshahi and Rangpur and one during the kharif-1 season in Comilla were conducted. Farmers' field days and demonstrations, and training activities for researchers and field assistants, were found to be useful for the dissemination of CA-RCTs and SSNM technologies. These activities have also begun to build capacity among the project partners and stakeholders.

The IRRI-CIMMYT joint project involves multiple partnerships for research and technology transfer in four districts in Bangladesh. In addition to the original five national partners, the International Plant Nutrition Institute (IPNI) signed a Letter of Agreement with IRRI in December 2009 to collaborate on site-specific nutrient management (SSNM) under the overall umbrella of the project. The project is also formally linked to the IRRI-CIMMYT-IFPRI-ILRI-led Cereal Systems Initiative for South Asia (CSISA) project.
The project has four objectives. Under Objective 1, data from diagnostic village surveys from 18 villages in six upazillas of Comilla, Rangpur, and Rajshahi were analyzed and a draft synthesis report prepared. Information from these surveys was used to design questionnaires for individual household surveys. The research teams received feedback from farmers and other stakeholders about the adaptive research and demonstration trials, which was used in refining and redesigning the trials. Under Objective 2, 20 hybrids from the public and private sector were compared under three conservation agriculture (CA)-based tillage systems during kharif-1 in 2009. Five hybrids were found to be less susceptible to waterlogging during 8- to 10-leaf stages. During rabi 2009-10, 120 hybrids were evaluated using "cup" screening and 66 hybrids using field screening in Gazipur. Promising hybrids selected from the previous year's research were evaluated under CA (zero tillage, ZT; strip tillage, ST; minimum tillage, MT; and raised beds, RB) at the three project sites during rabi 2009-10. A locally made farmers' implement was used to make shallow slits for sowing in ZT, a power-tiller-operated seeder (PTOS) for making strips for ST and MT, and a bed planter for sowing on RB.
Under Activity 1 of Objective 3, adaptive trials on CA with dry direct-seeded rice (DSR) and unpuddled transplanted rice (TPR) were conducted under ZT, ST, MT, and RB in aman 2009 in 41 farmers' fields. Likewise, researcher-managed Nutrient Manager (NM) for Rice evaluation trials with three treatments (farmers' fertilizer practice, BRRI recommendation, NM-based recommendation) were conducted on puddled TPR in 49 farmers' fields. Yields were not significantly different among the tillage treatments but cost of production of both DSR and unpuddled TPR was lower than the conventional puddled TPR. Likewise, in NM trials, the BRRI recommendation gave slightly greater yield but the NM recommendation was more profitable in most cases. During rabi 2009-10, CA-based long-term trials on rice-maize-mungbean (R-M-MB) systems began in 44 farmers' fields and on rice-potato-maize (R-P-M) systems in 11 farmers' fields. Likewise, SSNM trials on rabi maize were conducted in 36 farmers' fields and on kharif-1 maize in 29 farmers' fields. Data processing and analysis of the rabi trials is under way and kharif-1 trials are currently at the reproductive stage.
Under Activity 2, an on-station replicated CA experiment with six tillage treatments (ZT, ST, and MT, fresh RB and permanent RB, and conventional tillage [CT] or farmers' practice [FP]) on R-M-MB systems and another with two tillage treatments on R-P-M systems began at the BRRI research station in Rajshahi starting with 2009-10 rabi maize and potato. Likewise, long-term demonstration plots with different CA practices also began at the WRC station in Rajshahi. In addition, unreplicated SSNM trials with seven treatments (ranging from nutrient omission plots to reduced to full levels of N, P, and K) were conducted at both stations. At the BARI station in Comilla and RDRS farm in Rangpur, CA trials (R-M-MB and R-P-M) with different tillage treatments that started in kharif-1 2009 were continued in aman 2009 and rabi 2009-10, and now with kharif-1 maize and mungbean. At the BRRI station in Rangpur, the long-term R-M-MB experiment was continued during aman and rabi, and an additional plant population experiment with six populations was conducted during rabi 2009-10. During aman 2009 at the BRRI station in Rangpur, grain yields were significantly greater in ZTDSR (6.68 t/ha) than for CTDSR (5.79 t/ha) and TPR (5.57 t/ha), and yield losses to weeds were least with one herbicide application followed by one hand weeding. Under Activity 3, a PTOS and a bed planter with a fluted-type seed-metering device with 8 and 16 grooves were extensively tested and used for sowing DSR, maize, and mungbean in different trials.
Under Objective 4, three technical working groups were formed in each district. More than 100 researchers, field assistants, machine manufacturers and operators, service providers, and farmers from both within and outside the project, and field-level extension workers from the Department of Agriculture Extension (DAE) and NGOs were trained on the use of different machinery for sowing DSR, unpuddled TPR, and maize at all project sites. More than 540 farmers, NGO and local DAE workers, and representatives from the partner organizations from both within and outside the project attended eight farmers' field days during aman and rabi seasons. Thirty-two researchers and their supervisors involved in the project participated in two week-long traveling seminars at the project sites conducted during aman and rabi seasons. All these events were found to be very useful for the dissemination of technologies generated from within and outside the project. These activities also helped build capacity among the project partners and stakeholders.

The IRRI-CIMMYT joint project, "Sustainable intensification of rice-maize production systems in Bangladesh" involves five national partners - Bangladesh Agricultural Research Institute (BARI), Bangladesh Rice Research Institute (BRRI), Bangladesh Academy of Rural Development (BARD), Rangpur-Dinajpur Rural Services (RDRS), and Bangladesh Rural Advancement Committee (BRAC), and one international partner - International Plant Nutrition Institute (IPNI). The project is formally linked to the IRRI-CIMMYT-IFPRI-ILRI-led Cereal Systems Initiative for South Asia (CSISA) project and the IRRI-CIMMYT-World Fish Centre-led CSISA expansion project in Bangladesh.
The project has four objectives - socio-economic studies of rice-maize (R-M) systems; evaluation of elite maize germplasm for tolerance of excess moisture; development of locally adapted integrated management for high-yielding and sustainable R-M systems; and capacity building and dissemination of management solutions for R-M systems.
Under objective 1, the earlier conducted baseline village surveys and household surveys were revisited, data further analysed, and results synthesized. The village surveys report has been completed and the household surveys data are being reanalysed and synthesized. An adoption survey of key technologies promoted by the project is underway. Under objective 2, screening of 180 public and private sector maize inbreds and hybrids using 'cup' screening methods was conducted against excess moisture at seedling stage during early 2011. Five out of thirty-one tested hybrids were found to be the best in kharif-1 trial conducted at Gazipur in 2010. Evaluation trial of those five hybrids is currently underway in five districts. None of the 90 inbred lines screened for excess moisture tolerance during late vegetative stage (V8-V10) were found to be tolerant under field conditions.
Under objective 3, several on-station and on-farm trials on rice-maize-mungbean and rice-potato-maize systems with the principles of conservation agriculture (CA) are being continued in the three project districts. On-farm adaptive trials with different tillage options in rice-maize-mungbean cropping systems conducted during 2009-10 rabi and aman seasons demonstrated significant site - tillage interaction effect (p< 0.05) on grain yields of maize and rice. Maize performed generally better under reduced tillage, strip tillage, and raised beds as compared to farmer's practice. Though rice yields in farmer's practice did not differ significantly from any of the tillage practices in 2010, yields under reduced tillage and raised beds were more profitable than the farmer's practice.
The objective 3 also includes site-specific nutrient management (SSNM) for R-M systems. Omission nutrient trials and evaluation of Nutrient Manager, a software for estimating fertilizer need, were conducted under activity 1 of objective 3. Omission plot trials involve 28 fields in Boro 2009-2010, 32 farmers' fields in rabi maize 2009-2010, and 27 fields in kharif-1 maize in 2010. In boro, the NPKSZn treatment gave 5.89, 5.97 and 5.37 t/ha in Comilla, Rangpur, and Rajshahi, respectively. Omission of N reduced rice yield by 0.71, 2.13 and 1.34 t/ha, respectively, in those sites. Yield reduction due to omission of P and K was significant in Rangpur only. In 2009-2010 rabi maize, the NPK treatment in omission plot trials gave a mean yield of 8.34, 9.27 and 8.27 t/ha in Comilla, Rajshahi and Rangpur, respectively. Omission of each of N, P and K individually compared to NPK significantly reduced grain yield in all the three locations. Based on the lower yield level observed in the - N plots, the site-specific N doses for rabi maize would be 200-220 kg/ha in all the three locations. The mean P dose for rabi maize Comilla and Rangpur sites would be 16 kg/ha and that for Rajshahi site would be 22 kg/ha. The mean K dose for those sites would be 136, 65 and 36 kg/ha, respectively. The omission plot trials in Kharif-1 maize demonstrated negligible response to applied N, P, and K because of residual effect of large amounts of fertilizers applied to previous potato crop.
The Nutrient manager (NM) evaluation trial includes 43 farmers' fields in aman 2010 season and 30 fields in 2009-2010 boro season. The NM demonstrated the higher partial factor productivity (PFP) in both aman and boro seasons in all sites. In boro, the NM dose did not result in better yield in comparison to BRRI recommended dose but inclusion of S and Zn along with NM dose gave comparable yield to that obtained with BRRI dose. The lower yield in NM+S than in NM+SZn in all three sites explains that Zn was the key factor to achieve higher yield. However, the NM needs to be further improved to include Zn along with the N, P and K dose determination for rice.
Refinements and testing of seeding machinery is underway. In collaboration with National Agro-industries, a machinery manufacturing company in Ludhiana, Punjab, our collaborator at BARI in Rajshahi has fabricated and tested the 'inclined plate' seed metering device by attaching with the existing 2-wheel tractors (also called power tiller operated seeder-PTOS) for sowing rice, maize, mungbean and other seeds precisely.
The project organized several farmers' field days, farmers' motivational visits, travelling seminars, demonstration plots, and trainings for farmers, service providers, extension workers, and researchers in the project sites as part of capacity building and dissemination activities. Two of the travelling seminars were conducted in Nepal. The project has made significant capacity building of the researchers and field-level staff of participating institutions as well as made significant community impacts.