Fertilizer Sector Improvement (FSI+) in Burma

Transcription

1 Fertilizer Sector Improvement (FSI+) in Burma REPORT OF 2017 WET SEASON TRIALS Trials Using the Urea Deep Placement Technique on Transplanted Rice, Broadcast Rice, and the Rice-Gram System Submitted to USAID Burma Agreement Number BFS-IO May 2018 INTERNATIONAL FERTILIZER DEVELOPMENT CENTER PO BOX 2040 MUSCLE SHOALS, AL USA

2 Table of Contents Introduction... 1 Trials Tested in the 2017 Wet Season... 3 Materials and Methods... 4 Site Selection... 4 Trial Failures... 5 Varieties and Farmers... 5 Experimental Design... 6 Basal Fertilizers... 6 Treatments... 7 Trial No. 1 UDP on Direct Seeded Rice... 7 Trial No. 2 Balanced Fertilization With Compound Fertilizer on TPR... 7 Trial No. 3 UDP and Prilled Urea Deep Placement Trial on TPR... 8 Trial No. 4 Different N Rates With UDP and Prilled Urea on TPR... 9 Trial No. 5 Different P Rates on TPR... 9 Trial No. 6 Different K Rates on TPR Trial No. 7 Different S Rates on TPR Trial No. 8 Fertilizer Management in the Rice-Gram System Trial No. 9 Fertilizer Management in the Rice-Gram System With Starter N Trial No. 10 UDP Evaluation Under Submerged Conditions Trial Plot Size and Spacing Results and Discussion Trial No. 1 UDP on Direct Seeded Rice Trial No. 2 Balanced Fertilization With Compound Fertilizer Trial Trial No. 3 UDP and PU Deep Placement Trials Trial No. 4 Different N Rates Trial Using UDP and PU Trial No. 5 Different P Rates Trial Trial No. 6 Different K Rates Trial Trial No. 7 Different S Rates Trial Trial No. 8 Fertilizer Management Trials on the Rice-Gram Cropping System Trial No. 9 Fertilizer Management Trial on the Rice-Gram System With Starter N on Gram Trial No. 10 UDP Evaluation Under Submerged Conditions Trial Conclusion UDP on Direct Seeded Rice Trials Balanced Fertilization With Compound Fertilizer Trial Different N Rates Trial With UDP and PU FSI+ Report of 2017 Wet Season Trials i

3 Different P Rates Trial Different K Rates Trial Different S Rates Trial Fertilizer Management Trials on the Rice-Gram Cropping System Fertilizer Management Trial on the Rice-Gram System With Starter N on Gram UDP Evaluation Under Submerged Conditions Trial Appendix 1. Data Sheets Data of UDP on DSR Trial at Bawine, Pantanaw Data of UDP on DSR Trial at Magyi Kan, Kawhmu Data of Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya Data of UDP and PU Deep Placement Trial at Kan Gyi, Letpadan Data of UDP and PU Deep Placement Trial at The Yet Kone, Nyaunglaybin Data of Different N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi Data of Different P Rates Trial at Gyoe Phyu, Taikkyi Data of Different K Rates Trial at Gyoe Phyu, Taikkyi Data of Different S Rates Trial at Gyoe Phyu, Taikkyi Data of Fertilizer Management Trial on the Rice-Gram System at Thar Kwin, Einme Data of Fertilizer Management Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa Data of UDP Evaluation Under Submerged Conditions at Kwin Yar, Kangyidaunt Data of UDP Evaluation Under Submerged Conditions at Mayan, Kunchangone Photos FSI+ Report of 2017 Wet Season Trials ii

4 Tables Table 1. Locations and Coordinates of Field Trials, Wet Season Table 2. Varieties, Trial Types, and Collaborating Farmers for the 2017 Wet Season Trials... 6 Table 3. Experimental Designs Used for 2017 Wet Season Trials... 7 Table 4. Plot Size and Spacing Used for Specific Trials, 2017 Wet Season Table 5 Locations, Varieties, and Dates of All Trials, 2017 Wet Season Table 6 Yield and Component Characteristics of UDP on DSR Trial at Test Locations Table 7 Yield and Component Characteristics of Compound Fertilizer Trial at The Pyay Chaung, Myaungmya Table 8 Yield and Component Traits of UDP and PU Deep Placement Trials at Test Locations Table 9 Yield and Component Traits of N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi Table 10. Yield and Component Traits of Different S Rates Trial at Gyoe Phyu, Taikkyi Table 11 Component Traits of UDP Evaluation Under Submerged Conditions Trial at Test Locations Table 12 Treatment Yields of UDP Evaluation Under Submerged Conditions Trial at Test Locations Table 13 Yield and Yield Increase of Different Treatments Over Control from Rate Trials at Gyoe Phyu, Taikkyi Figures Figure 1. Treatment Yields of UDP on Direct Seeded Rice Trials by Location Figure 2. Rice Yield Response (t/ha) to Different N Rates of UDP and PU at Gyoe Phyu, Taikkyi Figure 3. Rice Yield Response (t/ha) to Different P Rates at Gyoe Phyu, Taikkyi Figure 4. Rice Yield Response (t/ha) to Different K Rates at Gyoe Phyu, Taikkyi FSI+ Report of 2017 Wet Season Trials iii

5 Photos Photo 1. Photo 2. Photo 3. Photo 4. Photo 5 Layout and Bunds of Fertilizer Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa, Yangon Region, on June 16, UDP Application during the UDP and PU Deep Placement Trial at The Yet Kone, Nyaunglaybin, Bago Region, on July 13, Transplanting UDP Evaluation Under Submerged Conditions Trial at Mayan, Kunchangone, Yangon Region, on July 14, First Topdressing of UDP on DSR Trial at Bawine, Pantanaw, Ayeyarwady Region, on August 22, Harvesting Crop Cuts during the Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya, on November 8, FSI+ Report of 2017 Wet Season Trials iv

6 Acronyms and Abbreviations ANOVA Ca cm DAS DAT DSR FP FSI ft g ha HYV IFDC K kg lb LSD MOP N P PU RCB S t TPR TSP UDP USAID Analysis of Variance Calcium centimeter Days after Sowing Days after Transplanting Direct Seeded Rice Farmers Practice Fertilizer Sector Improvement foot gram hectare High-Yielding Variety International Fertilizer Development Center Potassium kilogram pound Least Significant Difference Muriate of Potash Nitrogen Phosphorus Prilled Urea Randomized Complete Block Sulfur ton Transplanted Rice Triple Superphosphate Urea Deep Placement United States Agency for International Development FSI+ Report of 2017 Wet Season Trials v

7 Conversions To Convert To Multiply by acre hectare hectare acre U.S. ton/acre t/ha 2.24 lb/acre kg/ha 1.12 kg/ha lb/acre 0.89 K2O K 0.83 K K2O P2O5 P P P2O Urea N 46% TSP P2O5 45% TSP P 20% MOP K2O 60% MOP K 50% Gypsum S 18% Gypsum Ca 23% FSI+ Report of 2017 Wet Season Trials vi

8 Fertilizer Sector Improvement Project (FSI+) Trials Using the Urea Deep Placement Technique on Transplanted Rice, Broadcast Rice, and the Rice-Gram System in the 2017 Wet Season, Myanmar Introduction Urea deep placement (UDP) technology is a proven technology that can increase the yield of transplanted lowland rice by 15-20% with less use of urea (up to 40%) compared to broadcast application of urea. This has been proven in Bangladesh and in sub-saharan African countries. A number of UDP adaptation trials on transplanted rice (TPR) were conducted at selected locations in the Yangon, Bago, and Ayeyarwady regions of Myanmar during the 2014 wet season to the 2017 dry season by the International Fertilizer Development Center (IFDC) as part of the Fertilizer Sector Improvement (FSI) project, funded by the United States Agency for International Development (USAID). The results prove that UDP technology, using the same or a lesser rate of nitrogen (N), is superior to urea broadcasting at rates recommended by the Department of Agriculture and at rates used by local farmers (farmers practice [FP]). The yield increase of UDP with a more or less similar rate of N to FP ranged from 4.3% (2014 Wet Season) to 41.7% (2016 Dry Season). The yield increase of UDP with the same rate of N and basal fertilizer as urea surface broadcasting ranged from 3.4% to 46.9% (2015 Wet Season). Nutrient use efficiency of UDP was found to be double that of the urea broadcasting practice. Other UDP trials were conducted to explore the possibility of UDP application on broadcast rice, which farmers currently practice because it is cost-effective and requires less labor. The results show that UDP application on broadcast rice is also effective. It can be applied just before seed broadcasting or days after sowing (DAS). It can be applied on either earlymaturing varieties or medium-maturing varieties. Application later than 25 days DAS can be difficult due to plant height and thickness of seedlings. UDP can be applied on broadcast rice at the same spacing as TPR of 40 centimeters (cm) x 40 cm. FSI+ Report of 2017 Wet Season Trials 1

9 Other trials, such as potassium (K) x N rates using UDP, plant spacing and N rates with UDP using a larger briquette size, and long-term fertilizer management trials on the rice-gram system, have been conducted. The results show that the K requirement was site-specific, but N was found deficient at test locations and requires a high rate to reach potential yields. There was no need for plant spacing closer than 20 cm x 20 cm and no need to apply larger briquettes in wet season rice. However, there was a need to apply a 2.7-gram (g) briquette with 20 cm x 20 cm spacing in dry season rice. No significant results were observed with fertilizer management in the rice-gram system. There was no clear carryover effect of P and K on the subsequent crop. Plots with phosphorus (P) and K application gave a higher yield in the season applied. There was no residual effect of UDP on the gram crop. Balanced fertilization with compound fertilizers was also tested, since farmers apply compound fertilizer as a basal application and a topdressing of prilled urea (PU). The rate applied by farmers is 50 kilograms (kg) per acre of compound fertilizer, which appeared inadequate. Compound fertilizer with additional P and N was evaluated. The results indicate that compound fertilizer only was insufficient, giving the lowest yield. Balanced fertilization with triple superphosphate (TSP), muriate of potash (MOP), and gypsum with UDP was the best treatment. Balanced fertilization using compound fertilizer with PU was the second best. The difference between the two treatments was due to deep placement of the urea. This again proves that UDP is a more effective application than PU. Farmers practice of compound fertilizer and PU did not achieve the yield of balanced fertilizer application. UDP on direct seeded rice (DSR) was tested over two seasons to simulate mechanized application of UDP. This shows that UDP applications both at sowing and at 25 DAS are significant. Therefore, machine application of UDP with DSR may. improve rice yield with UDP and solve the labor problem at the same time. The trial will continue for one more season. After completing the first cycle of long-term trials on fertilizer management in the rice-gram system, results from application of starter N on the gram crop, carryover effect of P and K, and the N effect of UDP on the gram crop were not clear. Results were confounded by germination and poor crop establishment for both rice and gram. The trial will continue in the 2017 wet season (for rice). UDP was also evaluated under saline conditions in the dry season and submerged conditions in the wet season. The results often were not significant. But a salinity-tolerant improved highyielding variety (HYV) was better than the local variety, and a submergence-tolerant improved FSI+ Report of 2017 Wet Season Trials 2

10 variety was better than a normal HYV. UDP had good results under both conditions. These trials will also continue in the upcoming wet and dry seasons. To introduce a site-specific nutrient management practice, soils from selected farms in the project areas were collected and tested using omission pot trials. This was followed by omission trials in the fields where the soils were collected to confirm nutrient deficiency and rate trials to determine response curves for specific elements. Different rates of deficient elements were determined for specific sites. More omission pot trials will be conducted for soils at sites selected for their dominant rice production. In this 2017 wet season trials with UDP on direct seeded rice, balanced fertilization with compound fertilizer, and UDP evaluation under submerged conditions will be conducted at new sites with new farmers. Long-term fertilizer management on the rice-gram system will continue on the same plots at the same sites, and rate trials for N, P, K, and sulfur (S) will be conducted where those elements were identified as deficient in the omission pot trials. All trials are conducted in the project regions. The trials established are listed below. Trials Tested in the 2017 Wet Season There are 10 trials designed for the 2017 wet season. Some were a continuation from previous seasons, such as the long-term trials in the rice-gram system, and some were trials to confirm previous tests. Nutrient rate trials were conducted at locations where those elements were identified as deficient in the omission pot trials. With replications of some trials, a total of 16 trials were conducted. FSI+ Report of 2017 Wet Season Trials 3

11 Trial No. Trial Names Number of Trials 1. UDP on direct seeded rice 2 2. Balanced fertilization with compound fertilizer 2 3. UDP and prilled urea deep placement 2 4. Different N rates using UDP on TPR 1 5. Different P rates using UDP on TPR 1 6. Different K rates using UDP on TPR 1 7. Different S rates using UDP on TPR 1 8. Fertilizer management on the rice-gram system 2 9. Fertilizer management on the rice-gram system with starter N UDP evaluation on submerged rice 3 TOTAL 16 Materials and Methods Site Selection During the 2017 wet season, most trial sites and farmers were newly selected, except for the long-term fertilizer management trials in the rice-gram system, which continued at the same locations. Different nutrient rate trials were conducted at the sites on the land where those elements were identified as deficient in the omission pot trials. UDP evaluation in submerged conditions and other trials were also conducted in new locations with new farmers. The coordinates of the selected sites with the number of trials are given in Table 1. Soil samples from each location were dried and stored for later testing to identify initial soil conditions. FSI+ Report of 2017 Wet Season Trials 4

12 Table 1. Locations and Coordinates of Field Trials, Wet Season 2017 Region Township Village Latitude Longitude Elevation Trial No Thonegwa Anauk Ywa ' N ' E 27 ft 9 Kawhmu Magyi Kan ' N ' E 32 ft 1 Kunchangone Mayan ' N ' E 21 ft 10 Yangon Gyoe Phyu ' N ' E 61 ft 4 Taikkyi Gyoe Phyu ' N ' E 61 ft 5 Gyoe Phyu ' N ' E 61 ft 6 Gyoe Phyu ' N ' E 61 ft 7 Nyaunglaybin The Yet Kone ' N ' E 67 ft 3 Bago Letpadan Kan Gyi ' N ' E 51 ft 3 Thayarwaddy Thonese ' N ' E 81 ft 2 Einme Thar Kwin ' N ' E 36 ft 8 Einme Parami Daunt ' N ' E 35 ft 8 Myaungmya Kyar Phoo Ngon ' N ' E 26 ft 10 Ayeyarwady Myaungmya The Pyay Chaung ' N ' E 9 ft 2 Pantanaw Bawine ' N ' E 27 ft 1 Kangyidaunt Kwin Yar ' N ' E 24 ft 10 Shaded rows indicate failed trials. Trial Failures Out of 16 trials, 13 could be continued through harvesting. Three trials had to be discontinued before harvest. One trial on UDP evaluation under submerged conditions at Kyar Phoo Ngon, Myaungmya, failed at nursery stage due to severe flooding. Seedlings died and not enough remained for transplanting. One trial in the rice-gram system at Parami Daunt, Einme, also had to be discontinued at nursery stage when farm ownership changed. The new owner of the land was not willing to continue with a research trial on his land. The balanced fertilization with compound fertilizer trial at Thonese, Thayarwaddy, was poorly managed in the vegetative stage. The trial was extremely weedy due to a lack of water in the field. This trial was abandoned at the vegetative stage. Failed trials are indicated in Table 1 with shading. Varieties and Farmers Varieties and farmers for all locations are provided in Table 2. FSI+ Report of 2017 Wet Season Trials 5

13 Table 2. Varieties, Trial Types, and Collaborating Farmers for the 2017 Wet Season Trials Township Village Farmer Variety Trial No. Thonegwa Anauk Ywa U Zaw Wate Nang Myaing HYV 9 Kawhmu Magyi Kan U Zaw Htay Sin Thu Kha HYV 1 Kunchangone Mayan U Zaw Min Oo 2 varieties HYV + Swarna 10 Taikkyi Gyoe Phyu U Kyaw Lin Thu Sin Thu Kha HYV 4 Gyoe Phyu U Kyaw Lin Thu Sin Thu Kha HYV 5 Gyoe Phyu U Kyaw Lin Thu Sin Thu Kha HYV 6 Gyoe Phyu U Kyaw Lin Thu Sin Thu Kha HYV 7 Nyaunglaybin The Yet Kone U Myo Lwin Soe Sin Thu Kha HYV 3 Letpadan Kan Gyi U Aye Thwin Sin Thu Kha HYV 3 Einme Thar Kwin U Aung Htay Win Sin Thu Kha HYV 8 Myaungmya The Pyay Chaung U Win Thein Sin Thu Kha HYV 2 Pantanaw Bawine U Sein Win Sin Thu Kha HYV 1 Kangyidaunt Kwin Yar U San Oo Sin Thu Kha HYV + Swarna 10 Note: Failed trials are not included. Experimental Design A randomized complete block (RCB) design, with four to six treatments and three replications, was used in all trials except one. A split block design was used for the different N rates trial with UDP and PU (Table 3). A detailed plan for each trial is available in the protocols, retained in a separate document. Basal Fertilizers In all the trials, a basal fertilizer of TSP at 80 kilograms per hectare (kg/ha), MOP at 40 kg/ha, and gypsum at 25 kg/ha was used as the source of P, K, and S. The 15:15:15 compound fertilizer was used in the balanced fertilization with compound fertilizer trial. The amount of compound fertilizser was calculated based on the K rate, and additional amounts of TSP and urea were calculated and applied with the compound fertilizer to increase the rates of P and N. Compound fertilizer was also used in the farmers practice treatment for most trials. In calculating the rates, it is assumed: TSP = 45% P2O5 or 20% P MOP = 60% K2O or 50% K Gypsum = 18% S and 23% Ca FSI+ Report of 2017 Wet Season Trials 6

14 Table 3. Experimental Designs Used for 2017 Wet Season Trials No. Trial Name Treats x Reps Design 1 UDP on direct seeded rice 4 x 3 RCB 2 Balanced fertilization with compound fertilizer 4 x 3 RCB 3 UDP and prilled urea deep placement 2 x 5 x 3 Split block 4 Different N rates using UDP and PU on TPR 5 x 3 RCB 5 Different P rates using UDP on TPR 5 x 3 RCB 6 Different K rates using UDP on TPR 5 x 3 RCB 7 Different S rates using UDP on TPR 5 x 3 RCB 8 Fertilizer management on the rice-gram cropping system 4 x 3 RCB 9 Fertilizer management on the rice-gram system with starter N 6 x 3 RCB 10 UDP evaluation on submerged rice 4 x 3 RCB Treatments Trial No. 1 UDP on Direct Seeded Rice There were four treatments in this trial. UDP was applied at two different times (at seeding time and 21 DAS) as two treatments. This was compared with PU surface broadcasting and a control plot. The trial used a manual direct seed placement method (placing seeds with 20 cm x 20 cm spacing). This trial was established at two locations (Table 1). Treatment 1 = Zero N (Control) Treatment 2 = UDP application at sowing time (51.75 kg N/ha) Treatment 3 = UDP application at DAS (51.75 kg N/ha) Treatment 4 = PU surface broadcasting (51.75 kg N/ha) A basal fertilizer application of TSP at 80 kg/ha, MOP at 40 kg/ha, and gypsum at 25 kg/ha was applied on all treatments. UDP was applied only once at sowing time for Treatment 2 and 25 DAS for Treatment 3 by deep-placing a 1.8-g briquette at the center of four alternate sowing points. For Treatment 4, PU was applied in three split applications of equal amounts. The first application was at 25 DAS, the second application was at 60 DAS, and the third application was at 90 DAS. Dates of sowing, transplanting, UDP, topdressing, and harvesting are given in Table 5. Trial No. 2 Balanced Fertilization With Compound Fertilizer on TPR This is a repeat trial with the same four treatments and three replications. Although it was planned for two locations, only one could be harvested. A 15:15:15 compound fertilizer was FSI+ Report of 2017 Wet Season Trials 7

15 used, and the rate was based on K requirement (24 kg K2O/ha), with additional TSP and urea to make up the balance (36 kg P2O5/ha and 51 kg N/ha). Treatments are as follows: Treatment 1 = Compound (15:15:15) 160 kg/ha, gypsum 25 kg/ha, and PU 60 kg/ha Treatment 2 = Balanced with TSP 80 kg/ha, MOP 40 kg/ha, gypsum 25 kg/ha, and UDP kg/ha Treatment 3 = Balanced with compound (15:15:15) 160 kg/ha, TSP 27 kg/ha, gypsum 25 kg/ha, and PU 60 kg/ha Treatment 4 = Compound (15:15:15) 247 kg/ha and gypsum 25 kg/ha Treatment 1 applies compound fertilizer at 64.8 kg/acre, in which the K rate is equal to that in the balanced application. PU was applied in three split applications. Treatment 2 applies balanced fertilization with UDP using straight fertilizers for each element. Urea briquettes of 1.8-g size were applied. Treatment 3 also applies balanced fertilization but uses compound fertilizer and PU. The rates of all nutrients were the same as in Treatment 2. PU was applied in three split applications. Treatment 4 applies compound fertilizer (at the farmer s rate) and gypsum only. Trial No. 3 UDP and Prilled Urea Deep Placement Trial on TPR This is a repeat trial to confirm the 2017 dry season trial in order to study the response of prilled urea deep placement compared with UDP and the surface broadcast PU practice. There were five treatments: 1. Zero N 2. Urea briquette deep placement (UDP) kg N/ha 3. Prilled urea point deep placement kg N/ha 4. Prilled urea band deep placement kg N/ha 5. Prilled urea surface broadcast kg N/ha TSP at 80 kg/ha (36 kg P2O5/ha or 16 kg P/ha), MOP at 40 kg/ha (24 kg K2O/ha or 20 kg K/ha), and gypsum at 25 kg/ha (4.5 kg S/ha) were applied as a basal fertilizer for all treatments just before seed broadcasting. FSI+ Report of 2017 Wet Season Trials 8

16 Urea fertilizers, both UDP and PU, for deep placement treatments (Treatments 2, 3, and 4) were applied only one time at 7 days after transplanting (DAT). A briquette size of 1.8 g was used for UDP. Point placement (Treatment 3) and band placement (Treatment 4) were used for PU. There were three split applications for PU surface broadcast treatment (Treatment 5). It was applied in equal amounts at 7 DAT the first time, 40 DAT the second time, and at flowering the third time. The urea rate was the same for all urea application treatments. Trial No. 4 Different N Rates With UDP and Prilled Urea on TPR This trial was conducted on soil deficient in N, as identified in the omission pot trials. There were five treatments. Four different N rates were tested along with a control (0 N). Both UDP and PU were used as sources of N. The experiment was designed as 2 x 5 x 3 split block with three replications. 1. Control (0 N) kg N/ha kg N/ha kg N/ha kg N/ha TSP at 50 kg/acre (56 kg P2O5/ha or 25 kg P/ha), MOP at 25 kg/acre (37 kg K2O/ha or 30 kg K/ha), and gypsum at 10 kg/acre (4.5 kg S/ha) were applied as basal fertilizer for all treatments just before seed broadcasting. UDP was applied only one time at 7 DAT by placing one 1.8-g urea briquette 3-4 inches deep in the soil. PU was applied as three splits in equal amounts. Trial No. 5 Different P Rates on TPR This trial was conducted on the soil where P was identified as deficient according to an omission pot trial. There were five treatments including the control plot with 0 P. Treatments were as follows: 1. Control (0 P) kg P2O5/ha kg P2O5/ha kg P2O5/ha kg P2O5/ha MOP at 25 kg/acre (37 kg K2O/ha or 30 kg K/ha) and gypsum at 20 kg/acre (9.0 kg S/ha) were applied as basal fertilizer at plot layout time just before transplanting. FSI+ Report of 2017 Wet Season Trials 9

17 Nitrogen was applied as UDP only one time at 7 DAT by placing 1.8-g briquettes at a depth of 3-4 inches in at the center of four alternating rice hills. This gave an N rate of 51.8 kg/ha. Trial No. 6 Different K Rates on TPR This trial was conducted on the soil where K was identified deficient in an omission pot trial. There were four different rates of K in this experiment as well as a control plot. Altogether, five treatments were tested with three replications. Treatments were as follows: 1. Control (0 K) kg K2O/ha kg K2O/ha kg K2O/ha kg K2O/ha TSP at 50 kg/acre (56 kg P2O5/ha or 25 kg P/ha) and gypsum at 20 kg/acre (9.0 kg S/ha) were applied on all plots as basal fertilizer at plot layout time just before transplanting. N was applied as UDP at 51.8 kg N/ha by deep-placing one 1.8-g briquette at the center of four alternating rice hills. Trial No. 7 Different S Rates on TPR This trial was conducted on the soil where S was found to be deficient according to an omission pot trial. There were five treatments including the control plot (0 S). Treatments were as follows: 1. Control (0 S) kg S/ha kg S/ha kg S/ha kg S/ha Gypsum was applied as the S source. TSP at 50 kg/acre (56 kg P2O5/ha or 25 kg P/ha) and MOP at 25 kg/acre (37 kg K2O/ha or 30 kg K/ha) were applied on all plots as basal fertilizer at plot layout time just before transplanting. N was applied as UDP at 51.8 kg N/ha by deepplacing one 1.8-g briquette at the center of four alternating rice hills. FSI+ Report of 2017 Wet Season Trials 10

18 Trial No. 8 Fertilizer Management in the Rice-Gram System This is a long-term trial that extends across the rice crop and the gram crop at the same location. The rice is a wet season crop, and the gram is a dry season crop. It is a 4 x 3 RCB design with four treatments. Because the results from the prior year showed no carryover effect of the fertilizer applied on the first crop on the subsequent crop, this trial was conducted with fertilizer rates 1.5 times higher than before. Treatments were as follows: Treatments on Rice Treatments on Gram 1. P2O5 60 kg/ha + K2O 60 kg/ha P2O5 0 kg/ha + K2O 0 kg/ha 2. P2O5 30 kg/ha + K2O 30 kg/ha P2O5 30 kg/ha + K2O 30 kg/ha 3. P2O5 0 kg/ha + K2O 0 kg/ha P2O5 60 kg/ha + K2O 60 kg/ha (no residue) 4. P2O5 0 kg/ha + K2O 0 kg/ha P2O5 60 kg/ha + K2O 60 kg/ha (+crop residue) In this season, the trial was conducted with the rice crop. N was applied as UDP on all plots. The briquette size used in the trial was 1.8 g, which gave an N rate of 51.8 kg/ha. In this year, there was only one trial at Thar Kwin, Einme. The other two trials from prior years could not be continued because the farmer was unwilling to proceed at Kyauktan and the land was sold to a new farmer at Parami Daunt, Einme. Trial No. 9 Fertilizer Management in the Rice-Gram System With Starter N This is also a long-term trial on the rice-gram system using starter N on the gram crop. This is the second cycle with the second rice crop. This trial differs from Trial No. 8 in that N is applied to the rice crop as either UDP or broadcast PU and the gram crop has an N application as broadcast PU. This results in six treatments. An RCB design was used with three replications. Treatments were as follows: Treatments on Rice (per hectare) Treatments on Gram (per hectare) 1. UDP 51.8 kg N + 60 kg P2O kg K2O 0 kg N + 0 kg P2O5 + 0 kg K2O 2. UDP 51.8 kg N + 0 kg P2O5 + 0 kg K2O 0 kg N + 60 kg P2O kg K2O 3. PU 51.8 kg N + 60 kg P2O kg K2O 0 kg N + 0 kg P2O5 + 0 kg K2O 4. PU 51.8 kg N + 0 kg P2O5 + 0 kg K2O 0 kg N + 60 kg P2O kg K2O 5. UDP 51.8 kg N + 60 kg P2O kg K2O PU 14 kg N + 0 kg P2O5 + 0 kg K2O 6. PU 51.8 kg N + 60 kg P2O kg K2O PU 14 kg N + 0 kg P2O5 + 0 kg K2O FSI+ Report of 2017 Wet Season Trials 11

19 Trial No. 10 UDP Evaluation Under Submerged Conditions Trial This trial is part of the Soil Fertility Technology Adoption, Policy Reform and Knowledge Management Project implemented by IFDC. Two fertilizer practices farmers practice (F1) and UDP (F2) were applied on two varieties a local popular variety (V1) and a submergence-tolerant variety (V2). Varieties were sown in blocks, and fertilizer treatments were applied randomly across each block. The experimental designed was a 4 x 3 RCB design. Treatments were: 1. F1 V1 (farmers practice on the popular variety) 2. F1 V2 (farmers practice on the submergence-tolerant variety) 3. F2 V1 (UDP on the popular variety) 4. F2 V2 (UDP on the submergence-tolerant variety) The trial was conducted at two locations, Kunchangone and Kangyidaunt. The farmers practice fertilizer was the same at both locations 50 kg/acre of compound (15:15:15) fertilizer as basal and 50 kg/acre of prilled urea in three split applications in equal amounts. Bay Gyar Lay was used as the local variety at Kunchangone, and Sin Thu Kha was used as the local variety at Kangyidaunt. The submergence-tolerant variety was the same for both locations Swarna sub1. Plot Size and Spacing There are variations in plot size among the trials according to the type of trial and the kind of treatment. Many trials were conducted with a plot size of 16 ft x 16 ft. Plot sizes in each trial are given Table 4. A detailed plan can be seen in the protocols, under separate cover. Table 4. Plot Size and Spacing Used for Specific Trials, 2017 Wet Season No. Trial Name Plot Size TPR Spacing 1 UDP on direct seeded rice 16 x 16 ft 8 x 8 in 2 Balanced fertilization with compound fertilizer trial 16 x 16 ft 8 x 8 in 3 UDP and PU deep placement trial 16 x 16 ft 8 x 8 in 4 Different N rate trial with UDP and PU 24 x 10.7 ft 8 x 8 in 5 Different P rate trial 16 x 16 ft 8 x 8 in 6 Different K ratetrial 16 x 16 ft 8 x 8 in 7 Different S rates trial 16 x 16 ft 8 x 8 in 8 Fertilizer management trial on the rice-gram cropping system 16 x 16 ft 8 x 8 in 9 Fertilizer management trial on rice-gram with starter N 20 x 20 ft Broadcast 10 UDP evaluation on submerged rice trial 20 x 13.2 ft 8 x 8 in FSI+ Report of 2017 Wet Season Trials 12

20 Trial descriptions with varieties used, sowing dates, UDP dates, topdressing dates, and harvest dates are given in Table 5. Table 5 Locations, Varieties, and Dates of All Trials, 2017 Wet Season Date of: Region/ Trans Top Top No. Township Village Variety Sowing Layout plant UDP dress 1 dress 2 Harvest Yangon 1 Thonegwa Anauk Ywa Nang Myaing 17-Jun 16-Jun - 11-Jul 21-Aug 20-Sep 30-Oct 2 Kawhmu Magyi Kan Sin Thu Kha 22-Jul 22-Jul - 17-Aug 17-Sep 23-Oct 23-Nov 3 Kunchangone Mayan Swarna sub1 17-Jun 14-Jul 14-Jul 22-Jul 17-Aug 23-Oct 24-Nov 4 Gyoe Phyu Sin Thu Kha 6-Jun 30-Jun 1-Jul 9-Jul 3-Aug 13-Sep 27-Oct 5 Gyoe Phyu Sin Thu Kha 6-Jun 30-Jun 1-Jul 9-Jul Oct Taikkyi 6 Gyoe Phyu Sin Thu Kha 6-Jun 30-Jun 1-Jul 8-Jul Oct 7 Gyoe Phyu Sin Thu Kha 6-Jun 30-Jun 1-Jul 8-Jul Oct Bago 8 Nyaunglaybin The Yet Kone Sin Thu Kha 10-Jun 5-Jul 6-Jul 13-Jul 9-Aug 19-Sep 2-Nov 9 Letpadan Kan Gyi Sin Thu Kha 8-Jun 7-Jul 8-Jul 15-Jul 11-Aug 22-Sep 28-Oct 10 Thayarwaddy Thonese Sin Thu Kha 12-Jun 9-Jul 10-Jul 17-Jul 11-Aug Abandoned Ayeyarwady 11 Thar Kwin Sin Thu Kha 14-Jun 24-Jul 24-Jul 31-Jul Nov Einme 12 Parami Daunt Sin Thu Kha 28-Jun Stopped due to land sold 13 Kyar Phoo Ngon Sin Thu Kha 9-Jul Stop due to flooded nursery Myaungmya The Pyay 14 Chaung Sin Thu Kha 7-Jun 3-Jul 4-Jul 12-Jul 8-Aug 14-Sep 8-Nov 15 Pantanaw Bawine Sin Thu Kha 23-Jun 23-Jun - 18-Jul 22-Aug 26-Sep 6-Nov 16 Kangyidaunt Kwin Yar Swarna sub1 26-Jun 20-Jul 21-Jul 28-Jul 24-Aug 2-Oct 10-Nov Results and Discussion Trial No. 1 UDP on Direct Seeded Rice As shown in Table 1 and Table 2, the trial was conducted at two locations Bawine village, Pantanaw township, Ayeyarwady region, and Magyi Kan, Kawhmu township, Yangon region. Rice was sown by manual direct seeding with 20 cm x 20 cm spacing. UDP was applied at two different times once at planting and once 25 days after planting. The UDP method was compared with PU broadcasting with the same rate. There was a control treatment with 0 N. Bawine, Pantanaw Table 6 shows that a significant result was found only with plant height and panicle length. Other parameters, including yield, were not significantly different. (Refer to Data of UDP on DSR Trial at Bawine, Pantanaw.) Plant height was the highest (129 cm) with Treatment 2 (UDP at seeding treatment), which was not statistically different from the FSI+ Report of 2017 Wet Season Trials 13

21 plant height (126 cm) of Treatment 3 (UDP at 25 DAS). Plant height of the control treatment and PU broadcast treatment were the same at 113 cm. Panicle length was the longest (25.7 cm) with Treatment 3 (UDP at 25 DAS). It was not significantly different from Treatment 2 (UDP at seeding) with a panicle length of 25.0 cm or Treatment 4 (PU broadcast) with 24.6 cm. The control treatment gave the shortest panicle length at 23.6 cm, which was not statistically different from Treatment 4 (PU broadcast). Although not significant, biomass weights and yield gave their highest values with Treatment 2 (UDP at seeding). The highest (6.55 t/ha) yield was with UDP at seeding. The second highest yield (6.27 t/ha) was with UDP at 25 DAS, followed by 5.38 t/ha with broadcast PU. The control plot produced the lowest yield at 5.08 t/ha. However, these were not significantly different. Biomass weights were also the highest with UDP at seeding, followed by UDP at 25 DAS. The control plot produced the lowest biomass weights. (Refer to Data of UDP on DSR Trial at Bawine, Pantanaw.) Magyi Kan, Kawhmu Table 6 shows the significant differences found in yield and number of panicles per hill. The yield was significant at P(0.01) and plant height at P(0.05). Other traits were not different among treatments. (Refer to Data of UDP on DSR Trial at Magyi Kan, Kawhmu.) Number of panicles per hill (15.5) was the highest with Treatment 2 (UDP at seeding). The second highest number of panicles per hill (13.2 ) was produced by Treatment 3 (UDP at 25 DAS). Treatment 4 (PU broadcast) was third highest at 11.0 panicles per hill, and the control plot produced the lowest number at 10.8 panicles per hill. However, the three treatments with the lowest number of panicles per hill were not significantly different. Treatment 3 (UDP at 25 DAS) produced the highest yield at 9.08 t/ha, which was significantly higher than all other treatments. Treatment 2 (UDP at seeding) produced the second highest yield (6.44 t/ha), which was not statistically different from the third highest yield (5.53 t/ha) produced by Treatment 4 (broadcast PU). The control plot produced the significantly lowest yield of 2.60 t/ha. Although not significant, biomass yields were similar to the yield results; the highest yield was produced by Treatment 3 (UDP at 25 DAS). Plant height and panicle length followed the same trend. (Refer to Data of UDP on DSR Trial at Magyi Kan, Kawhmu.) The results from both locations show that UDP technology can be applied on direct seeded rice with spacing of 20 cm x 20 cm. This gave a better yield than broadcast PU application with an equal N rate. The results indicate that UDP could be applied either at seeding time or at 25 DAS. UDP application at seeding time gave a higher yield than UDP application at 25 DAS at Bawine, Pantanaw. The result was opposite at Magyi Kan, Kawhmu (Figure 1). The soil at Bawine seemed more fertile, with more N, than that at Magyi Kan, since the yield with 0 N at FSI+ Report of 2017 Wet Season Trials 14

22 Bawine was high (5.08 t/ha), while that at Magyi Kan was low (2.60 t/ha). Applying N fertilizer with either UDP or PU significantly improved rice yield at Magyi Kan, Kawhmu. In the trial at Magyi Kan, UDP applied at 25 DAS was noted to give an extremely high yield (9.08 t/ha) compared to other treatments ( Table 6). This may be due not only to UDP. Two replications of that treatment gave very high yields, which may be due to the soil fertility gradient of the test plot. (Refer to Data of UDP on DSR Trial at Magyi Kan, Kawhmu.) Although the rice yield was not significantly improved with N fertilizer application at Bawine, Pantanaw, the UDP treatments did produce higher yields than PU and control. In this trial, direct seeding was done manually using a rope and markers to get the exact spacing of 20 cm x 20 cm. UDP application was done simultaneously with seeding. Manual seeding currently is not practiced by farmers due to a lack of labor. Manual direct seeding with simultaneous manual application of UDP would be difficult and impractical. UDP application just before seeding or just after germination can be considered. UDP could be applied more precisely if done just after germination, but this should be done manually. This trial was a simulation of mechanized application of UDP and direct seeding using a combine seeder/udp applicator. The results show a mechanized combine planter that sows seed with UDP application is an acceptable approach for best yield. Table 6 Yield and Component Characteristics of UDP on DSR Trial at Test Locations Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Location Treatment Bawine, Pantanaw Control (0 N) 113b 9.7 ab , UDP at seeding 129a 11.5 ab , UDP at 25 DAS 126a 11.7 ab , PU Broadcast 113b 11.3 ab , Significant level ** ns ns ns ns ns ns ns ns Magyi Kan, Kawhmu Control (0 N) ba c UDP at seeding aa b UDP at 25 DAS ab a PU Broadcast ba b Significant level ns * ns ns ns ns ns ns ** Yield (t/ha) FSI+ Report of 2017 Wet Season Trials 15

23 Figure 1. Treatment Yields of UDP on Direct Seeded Rice Trials by Location Trial No. 2 Balanced Fertilization With Compound Fertilizer Trial The trial was conducted at two locations, but one was abandoned at the vegetative stage as it was too weedy and had poor crop establishment (see Trial Failures). The trial was designed to determine whether applying compound fertilizer provides enough nutrients for rice cultivation, since farmers are using compound fertilizers more often. The trial at The Pyay Chaung, Myaungmya, showed significant results for nearly half of the traits, including yield (Table 7). Plant height and biomass dry grain weight were significant at P(0.05), and yield and biomass straw weights were significant at P(0.01). Number of panicles per hill, panicle length, number of spikelets and grains per panicle, biomass wet grain weight, and 1,000 grain weight did not show significant results. Plant height was the highest (122 cm) with balanced fertilization using TSP, MOP, gypsum, and UDP; it was significantly higher than with all other treatments. The other treatments were not significantly different, with plant heights ranging from 112 cm to 114 cm. Fresh biomass straw weight and biomass grain weights (both fresh and dry) had the same trend as plant height. The highest values were found with balanced fertilization using TSP, MOP, gypsum, and UDP. They were significantly different than with all other treatments. FSI+ Report of 2017 Wet Season Trials 16

24 The yield was also the highest (9.94 t/ha) with Treatment 2 (balanced fertilization using TSP, MOP, gypsum, and UDP); it was significantly higher than all other treatments. The second highest yield (8.25 t/ha) was given by Treatment 3 (balanced fertilization using compound fertilizer, TSP, gypsum, and PU), followed by Treatment 1 (compound+gypsum+pu), which produced 7.86 t/ha. The treatment using compound and gypsum only (Treatment 4) produced the lowest yield at 7.76 t/ha. However, those treatments were not significantly different according to analysis of variance (ANOVA, Table 7). (Refer to Data of Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya.) Results indicate that the rice yield could be improved by using balanced fertilization. Balanced fertilizers, either with compound or a combination of straight fertilizers, produced higher yields than unbalanced fertilization based on compound fertilizer. Both balanced fertilizer treatments (Treatment 2 and Treatment 3) had the same nutrient amounts. The difference was that Treatment 3 was balanced using compound fertilizer based on the K rate with additional TSP and used broadcast PU for N application, while Treatment 2 was balanced with TSP and MOP and used UDP for N application. The nutrient rates were the same. Therefore, the yield difference between the balanced fertilization treatments is assumed to be due to the difference between UDP and PU. The UDP technology again proved to be an efficient method of N application, producing a significantly higher yield. Application of compound fertilizer at the farmers rate (Treatment 4) was inadequate and gave a low yield. Nitrogen also seems limiting in the test soil. With unbalanced fertilization with compound fertilizer, the yield was higher with compound and PU than compound only. Compound fertilizer + gypsum gave the lowest yield. The rice yield may be improved by applying balanced nutrient application, either compound or straight fertilizer, and further improvement can be obtained by using UDP. Table 7 Yield and Component Characteristics of Compound Fertilizer Trial at The Pyay Chaung, Myaungmya Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Yield (t/ha) Treatment Compound + gypsum + PU 114b ,848 b b 386b 7.86b Balanced with UDP 122a ,209a a 481a 9.94a Balanced with compound + 114b ,751b b 343b 8.25b PU Compound + gypsum only 112b ,704b b 361b 7.76b Significant level * ns ns ns ** ns ** * ** FSI+ Report of 2017 Wet Season Trials 17

25 Trial No. 3 UDP and PU Deep Placement Trials This trial was designed to study the response of deep placement of PU (both point and band placement) compared to UDP. PU was deep-placed using both point and band placement methods at the same rate and the same time as UDP, which was applied seven DAT. The trial was conducted at two locations in the Bago region one at Letpadan township and one at Nyaunglaybin township. Kan Gyi, Letpadan According to ANOVA, significant differences at P(0.05) were found in fresh biomass straw weight, dry biomass grain weight, and yield. Number of grains per panicle was significant at P(0.01). The other traits were not significant (Table 8). (Refer to Data of UDP and PU Deep Placement Trial at Kan Gyi, Letpadan.) The UDP treatment gave the highest dry biomass grain weight (275 g). PU point placement gave nearly the same dry biomass grain weight at 273 g. However, PU point deep placement gave the highest values for the other significant traits, including yield. Number of grains per panicle was not statistically different among N treatments except PU band placement. There was no statistical difference between any of the N treatments for fresh biomass straw weight, but the deep-placed treatments were statistically different from the control. There was no statistical yield difference between any of the N treatments, and all were higher than the control. The control treatment gave the lowest values for all traits except plant height. The yield was significantly lowest (3.31 t/ha) with 0 N (Table 8). The Yet Kone, Nyaunglaybin ANOVA showed a significant result at P(0.01) in yield only. All other traits were not significantly different among treatments. (Refer to Data of UDP and PU Deep Placement Trial at The Yet Kone, Nyaunglaybin.) The control plot produced the lowest yield at 2.91 t/ha. According to the Least Significant Difference (LSD) comparison, it was not statistically different from the PU surface broadcast treatment, which produced 3.91 t/ha. UDP treatment gave the highest yield at 5.53 t/ha, which was not different from the yield of 5.43 t/ha produced by PU point placement. PU band placement produced a significantly lower yield than the UDP and PU point placement treatments at 4.07 t/ha. This was not different from the yield of 3.91 t/ha produced by the broadcast PU treatment but was significantly higher than the control plot. Variation of results was observed in component traits. Number of panicles per hill (8.9) and panicle length (25.2 cm) were the highest with the UDP treatment. Broadcast PU treatment produced the highest fresh biomass straw weight at 2,223 g. Plant height was the highest at 106 cm with the control treatment. Biomass weights, number of spikelets, and FSI+ Report of 2017 Wet Season Trials 18

26 number of grains per panicle were the highest with PU point placement. However, none of the traits were statistically different among treatments (Table 8). According to the results from both locations, PU application was as effective as UDP in rice cultivation when applied deep into the soil. In Kan Gyi, Letpadan, PU placement (both point and band) had even better yields than UDP, although the yields of plots with N application were not significantly different. In The Yet Kone, Nyaunglaybin, UDP gave the highest yield, but this was not statistically different from the yield with PU point placement. PU band placement gave the third highest yield, which was not statistically different from the PU broadcast treatment. The control plot gave the lowest yield at both locations, indicating that N is limiting in both. Therefore, urea application can increase rice yield, and UDP is the most effective application method for N application. PU also can be as effective as UDP if it is applied deep into the soil. However, deep placement of PU manually is difficult to apply in consistent amounts, whether by point or band placement. An applicator is needed for it to be practical in the rice fields. Table 8 Yield and Component Traits of UDP and PU Deep Placement Trials at Test Locations Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Location Treatment Kan Gyi, Letpadan Control (0 N) cz bb cb 3.31b UDP (51.75 kg N/ha) ab ,484ab ab 5.36a PU point az ,655ab ab 6.52a PU band bc ,280ab bc 6.18a PU Broadcast ab ,328ab ab 5.62a The Yet Kone, Nyaunglaybin Control (0 N) , cb UDP (51.75 kg N/ha) , ab PU point , ab PU band , bb PU Broadcast , bc Yield (t/ha) Trial No. 4 Different N Rates Trial Using UDP and PU This trial was conducted on the soil identified as N-deficient according to an omission pot trial. It was designed to determine how much N is needed to produce the maximum yield. N was FSI+ Report of 2017 Wet Season Trials 19

27 applied as UDP and broadcast PU at five different rates, including a 0 N control. A split block design with three replications was used. ANOVA showed no difference in either factor form of urea or rate. No interaction effect of the form of urea and rate was observed. Major yield-determining traits showed no differences among treatments. However, the yield was the lowest with 0 N for both UDP and PU application. Maximum yield with UDP (4.94 t/ha) was found with the 50 kg N/ha rate, which was the lowest N application rate in this trial. When the N rate was increased from 50 kg/ha to 150 kg/ha, the yield decreased but then increased again at an N rate of 200 kg/ha. Rice grain yield and yield-contributing traits, such as number of panicles per hill and number of grains per panicle, were the highest with 50 kg N/ha; there were 11.7 panicles per hill and 165 grains per panicle. Plant height and biomass grain weights also were the highest with 50 kg N/ha. However, the maximum biomass straw weights were obtained with an N rate of 150 kg/ha. (Refer to Data of Different N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi and Table 9.) With PU application, the yield increased as the N rate was increased up to 150 kg N/ha, which produced the maximum yield at 4.49 t/ha. This indicates more fertilizer is needed with PU application to obtain maximum yield compared with UDP. The yield decreased when the N rate was increased from 150 kg/ha to 200 kg/ha. An N rate of 200 kg/ha produced 3.90 t/ha. The control plot produced the lowest yield at 3.28 t/ha. Biomass straw and grain weights and component characteristics, such as number of spikelets and number of grains per panicle, showed the same response. An N rate of 150 kg/ha gave maximum values. However, these were not significantly different according to ANOVA. (Refer to Data of Different N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi, and Table 9.) The results clearly indicate that N is required in this soil. The control plot always produced the lowest yield. N application can improve the rice yield. UDP was found to be a more efficient application method than PU. As an average, UDP produced 4.03 t/ha, while PU produced 3.82 t/ha (Table 9). Figure 2 shows that UDP was a more efficient method of application than PU. An N rate of 50 kg/ha was needed to obtain maximum yield with UDP, while 150 kg N/ha was needed for the PU application. Increasing the N rate beyond those amounts decreased yields for both UDP and PU application. UDP could, therefore, improve rice yield using less urea than PU. With UDP, major yield-contributing characteristics, such as number of panicles FSI+ Report of 2017 Wet Season Trials 20

28 per hill and number of grains per panicle, were the highest with 50 kg N/ha. Biomass grain weight also was the highest at this rate. However, biomass straw weights were the highest with a higher N rate 150 kg N/ha. These results indicate that excess N with UDP promotes more vegetative growth rather than production of grain. With PU, the amount of N required to obtain maximum yield was more than with UDP. The results were the same for other characteristics (Table 9). Table 9 Yield and Component Traits of N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Location Treatment With UDP Control (0 N) kg N/ha kg N/ha kg N/ha kg N/ha Average 4.03 With Prilled Urea Control (0 N) kg N/ha kg N/ha kg N/ha kg N/ha Average 3.82 Yield (t/ha) FSI+ Report of 2017 Wet Season Trials 21

29 Figure 2. Rice Yield Response (t/ha) to Different N Rates of UDP and PU at Gyoe Phyu, Taikkyi Trial No. 5 Different P Rates Trial This trial was conducted at the same location, Gyoe Phyu, Taikkyi, where the soil was identified as P-deficient according to an omission pot trial. The main objective was to determine how much P is needed to produce the maximum yield and to give a recommendation for P application on that soil. There were five different P rates, including control. The lowest P rate was 50 kg P2O5/ha and the highest was 200 kg P2O5/ha, with 50-kg increments. A 5 x 3 RCB design was used. ANOVA showed no difference among treatments on all traits and yield. Variabilities were observed in component traits. Many traits showed their highest values with 50 kg P2O5/ha and 100 kg P2O5/ha. Number of grains per panicle, biomass straw weights, and 1,000 grain weight were highest with 100 kg P2O5/ha. A total of 1,942 g of fresh biomass straw weight, 497 g of dry biomass straw weight, and 16.7 g of 1,000 grain weight were produced with a P rate of 100 kg P2O5/ha. Plant height and biomass grain weights were highest with 50 kg P2O5/ha. Plant height (105 cm) and biomass grain weights (523 g for fresh grain weight and 460 g for dry grain weight) were highest with 50 kg P2O5/ha. A few traits (number of panicles per hill and number of spikelets per panicle) had their highest values with the control plot (0 kg P2O5/ha). FSI+ Report of 2017 Wet Season Trials 22

30 The yield, unlike other traits, was the highest with the 150 kg P2O5/ha treatment at 9.26 t/ha. Other plots with P application produced from 8.18 t/ha to 8.38 t/ha. The control plot produced the lowest yield at 7.70 t/ha. The rice yield increased as the P rate was increased up to 150 kg P2O5/ha. When the P rate was increased from 150 kg P2O5/ha to 200 kg P2O5/ha, the yield decreased. The rice yield with a P rate of 200 kg P2O5/ha was 8.23 t/ha (Figure 3). However, it was noted that the yield and all other traits were not significantly different among different P rates. (Refer to Data of Different P Rates Trial at Gyoe Phyu, Taikkyi.) Although not significant, the results suggested that the test soil is deficient in P. The control plot with 0 P gave the lowest yield (7.70 t/ha). Most traits showed their highest values with P application treatments at 50 kg P2O5/ha or at 100 kg P2O5/ha. The highest yield of 9.26 t/ha was produced by 150 kg P2O5/ha. This yield was very high among the P application treatments. The highest P rate did not give the highest yield; it was 8.23 t/ha with 200 kg P2O5/ha. The yields of other P application treatments ranged from 8.18 t/ha to 8.38 t/ha. As there were no significant differences in yield, P fertilizer at 50 kg P2O5/ha is the recommended application for the test soil at Gyoe Phyu, Taikkyi. Figure 3. Rice Yield Response (t/ha) to Different P Rates at Gyoe Phyu, Taikkyi FSI+ Report of 2017 Wet Season Trials 23

31 Trial No. 6 Different K Rates Trial A K rate trial was also conducted at Gyoe Phyu, Taikkyi, where K was observed to be deficient in the omission pot trial. There were five different K rates, including a control plot with 0 K. The trial used a 5 x 3 RCB design. According to ANOVA, a significant result was found in plant height and number of grains per panicle only. Other traits, including rice yield, were not significant. Plant height was significant at P(0.01), and the tallest plant (103 cm) was observed with 150 kg K2O/ha. It was significantly taller than with other treatments according to P(0.05). Number of grains per panicle was significant at P(0.05), and the maximum of 184 grains was observed with the highest K rate of 200 kg K2O/ha. Number of panicles per hill, number of spikelets per panicle, and fresh biomass straw weight showed their highest values with the highest K rate. The other biomass weights and 1,000 grain weight were the highest with the second highest K rate of 150 kg K2O/ha. (Refer to Data of Different K Rates Trial at Gyoe Phyu, Taikkyi.) The yields were not significantly different. However, the control plot gave the lowest yield at 6.78 t/ha. The yield increased as the K rate was increased up to 150 kg K2O/ha. The K rate of 150 kg K2O/ha produced the highest yield at t/ha. The yield then decreased when the K rate was increased from 150 kg to 200 kg K2O/ha. The K rate of 200 kg K2O/ha produced only 8.84 t/ha (Figure 4). The results indicate that the soil was K-deficient. By applying 50 kg K2O/ha, the yield increased by 0.71 t/ha, which was 10% more than the control plot. By applying 100 kg K2O/ha, the yield increased by 1.57 t/ha, which was 20% more than with the K rate of 50 kg K2O/ha. Another yield increase of 1.12 t/ha, or 12%, was attained by increasing the K application rate from 100 kg to 150 kg K2O/ha. Therefore, the recommended K rate should be based on affordability by the farmer. Farmers with available cash can make maximum production and profit by applying a K rate of 150 kg K2O/ha. Poorer farmers can still make a return by applying lower K rates. FSI+ Report of 2017 Wet Season Trials 24

32 Figure 4. Rice Yield Response (t/ha) to Different K Rates at Gyoe Phyu, Taikkyi Trial No. 7 Different S Rates Trial The soil of Gyoe Phyu, Taikkyi, was identified as S-deficient in an omission pot trial. This field trial was conducted in that location with five different S rates from 0 S to 40 kg S/ha at 10-kg intervals. The trial was conducted to determine the requirement of S fertilizer for normal growth and maximum production. ANOVA showed no differences among treatments in all traits except fresh biomass straw weight. Fresh biomass straw weight was the only trait significant at P(0.05). Four characteristics showed their highest values with an S rate of 10 kg/ha, which was the lowest rate of S application. Plant heights ranged from 96 cm to 101 cm, which was not significant. The control plot produced a 96 cm plant height, which was the shortest, and the 10 kg S/ha treatment produced a 101 cm plant height, which was the tallest. Dry biomass straw weight and biomass grain weights (both fresh and dry) had their highest values with 10 kg S/ha. Five characteristics, including yield, showed their highest values with 20 kg S/ha. These included the number of panicles per hill, number of spikelets and grains per panicle, fresh biomass straw weight, and yield. Fresh biomass straw weight was significant at P(0.05), and the maximum weight of 2,275 g was observed with 20 kg S/ha. This was not statistically different from the second highest weight of 2,181 g given by 40 kg S/ha. The weights of other treatments ranged from 1,734 g to 1,812 g and were not statistically different. The yields were not significantly different. The highest yield of t/ha was produced with 20 kg S/ha. When the S rate was increased from FSI+ Report of 2017 Wet Season Trials 25

33 20 kg/ha to 40 kg/ha, the yield decreased. The highest S rate of 40 kg/ha gave the lowest yield (7.56 t/ha). The control plot gave the second highest yield at 9.02 t/ha. However, these were not statistically different (Table 10). Later, it was learned that the TSP applied as basal for the P element also contained 3% S according to lab analysis. Therefore, the results of this trial were due to both S from TSP and S application treatments. The control plot (0 S) also received some amount of S from the TSP. Table 10. Yield and Component Traits of Different S Rates Trial at Gyoe Phyu, Taikkyi Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Yield (t/ha) Treatment Control (0 S) ,799b kg S/ha ,734b kg S/ha ,275a kg S/ha ,812b kg S/ha ,181a Significant level ns ns ns ns * ns ns ns ns Trial No. 8 Fertilizer Management Trials on the Rice- Gram Cropping System This was the third rice crop season for the long-term trial on the rice-gram system studying the carryover effect of P and K on the subsequent crop with UDP applied on the rice crop. Four treatments were tested with three replications. This was also the second rice season testing the P2O5 and K2O rates of 60 kg/ha. The long-term trial could be continued at only one location, Thar Kwin, Einme. The two trials at other locations were discontinued due to unwillingness of the farmer to continue and an ownership change. Significant results at P(0.05) were found in yield and fresh biomass straw weight. The other characteristics were not significant. The response of the characteristics to treatments was not consistent. Plant height, number of panicles per hill, and biomass straw weights showed their highest values with Treatment 1, in which the full amount of P and K (60 kg/ha each) was applied. The biomass grain weights and 1,000 grain weight had their highest values with Treatment 2, in which half the amount of P and K (30 kg/ha each) was applied. Panicle length FSI+ Report of 2017 Wet Season Trials 26

34 and number of spikelets and grains per panicle were at their maximum with Treatment 4, in which no P or K was applied on rice but the full amount of P and K (60 kg/ha each) had been applied on the gram crop plus crop residue from the gram crop. The effects of P and K application on those traits were inconsistent and difficult to explain. The explanation for one trait may not agree with that for another trait. (Refer to Data of Fertilizer Management Trial on the Rice-Gram System at Thar Kwin, Einme.) The yield was significant, and the highest yield (5.63 t/ha) was produced by Treatment 3, in which no P or K fertilizer and no gram residue was applied on the rice crop but the full amount of P (60 kg P2O5/ha) and K (60 kg K2O/ha) had been applied on the previous gram crop. The highest yield was not statistically different from the second highest yield (5.60 t/ha) produced by Treatment 1, in which 60 kg/ha each of P2O5 and K2O were applied, or the third highest yield (5.57 t/ha) produced by Treatment 4, in which no P or K was applied but there was residue from the gram crop in which the full amount of P and K was applied. Treatment 2, in which half the amount of P and K was applied on the rice crop, gave a significantly low yield of 5.45 t/ha; this was not significantly different from the second lowest yield of 5.57 t/ha with Treatment 4. (Refer to Data of Fertilizer Management Trial on the Rice-Gram System at Thar Kwin, Einme.) The results indicate that P and K can come either from carryover of the gram crop or from application directly on the crop. This means that there was a carryover effect of P and K applied to the gram crop on the subsequent rice crop. This effect was revealed this rice season, which is the third cropping cycle. In the second cycle, this effect was not seen, possibly because a lower amount of P and K was applied in the earlier cycles that could have been consumed by the gram crop with a negligible amount remaining for the rice crop. This time the higher rate of P and K was applied and possibly remained in the soil for the rice crop. Residual moisture content in the gram season is often inadequate to dissolve all P and K to make it available to the gram crop. The third highest yield (5.57 t/ha) was not statistically different from the highest (5.63 t/ha), and the second highest yield (5.60 t/ha) also indicated that incorporation of residue from the gram crop with full P and K application gave a certain benefit to the subsequent rice crop. This effect was not clear until the end of second cropping cycle. Therefore, the benefit of crop residue application is expected to increase if it can be practiced continuously. Quicker or slower response to crop residue depends on the biomass of the gram crop, which depends on residual moisture to produce a good yield with more biomass. FSI+ Report of 2017 Wet Season Trials 27

35 Trial No. 9 Fertilizer Management Trial on the Rice- Gram System With Starter N on Gram This is the second rice crop of the rice-gram system testing the effect of starter N on the gram crop. It is a long-term trial on the rice-gram system conducted on broadcast rice at Anauk Ywa in Thonegwa township. ANOVA showed no significant results in any trait, including yield. Variability in results was observed. Plant heights from plots with UDP application (110 cm, 112 cm, and 108 cm) were taller than those from plots with PU application (108 cm, 104 cm, and 104 cm). UDP clearly could provide more N to rice plants than PU. The number of panicles per square meter was highest (279/m 2 ) with UDP and full P and K application treatment. UDP with 0 P and 0 K had the same number of panicles per square meter (244/m 2 ) as PU with 0 P and 0 K. UDP without P and K did not seem to improve the number of panicles per unit area. However, that is a difficult conclusion to make because the seed broadcasting practice presented an uneven population. Uneven germination often occurred with broadcast rice. The number of spikelets and number of grains per panicle were high with UDP + 0 P and 0 K and PU with 60 kg/ha each of P2O5 and K2O. UDP + 0 P and 0 K gave 217 spikelets and 193 grains per panicle and PU + 60 kg/ha each of P2O5 and K2O on which starter N was applied to the previous gram crop gave 211 spikelets and 196 grains per panicle. According to these results, both urea and basal fertilizer are important to produce fully developed panicles. Biomass weights were the highest with Treatment 6, in which PU was applied with the full amount of P and K, and with Treatment 5, in which UDP was applied with the full amount of P and K. Both treatments used starter N at 14 kg urea/acre on the gram crop. Thus, high biomass weights could result from the N applied on both crops. Some of the N applied on the gram crop may have had carryover effect on the rice crop. (Refer to Data of Fertilizer Management Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa.) However, no significant differences were noted in any of the traits. The yield was also not different among treatments. PU + no P and K basal fertilizer gave the highest yield at 4.29 t/ha, while PU with full basal fertilizer produced 3.90 t/ha, which was difficult to explain. This could suggest that P and K are not important in this soil. The second highest yield of 4.14 t/ha was produced by Treatment 5, in which UDP and the full amount of P and K were applied and starter N had been applied on the previous gram crop. Treatment 1, in which UDP and full basal fertilizers were applied but no starter N on the gram crop, gave the third highest yield at 3.92 t/ha. This was similar to 3.90 t/ha given by PU and full basal FSI+ Report of 2017 Wet Season Trials 28

36 treatment with no starter N. As an average, UDP produced a lower yield (3.84 t/ha) than PU (4.01 t/ha). (Refer to Data of Fertilizer Management Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa.) P and K seem to be important somehow. The results indicate inconsistency in yield. Variable and inconsistent results like these are often found in trials conducted with broadcast rice. Broadcast rice needs good land leveling, with uniform broadcasting and germination, which does not typically occur with broadcast practice, especially in the farmers fields in the Delta regions. Trial No. 10 UDP Evaluation Under Submerged Conditions Trial UDP technology was evaluated under submerged conditions for the second time this wet season. The trial was planned for three submergence-prone areas, but one location failed (see Trial Failures). Two locations were harvested. UDP technology was studied against the farmers practice of fertilizer management with two different varieties, i.e., farmers preferred variety and a submergence-tolerant variety (Swarna sub1). The farmer at Kwin Yar, Kangyidaunt, used Sin Thu Kha, which is the most popular high-yielding variety, and the farmer at Mayan, Kunchangone, used Bay Gyar Lay, a local tall variety. Kwin Yar, Kangyidaunt ANOVA showed significant differences in most traits, including yield. Only four traits panicle length, number of spikelets and grains per panicle, and 1,000 grain weight were not significant. The yield and fresh biomass straw and grain weights were significant at P(0.01), and the dry biomass straw and grain weights, plant height, and number of panicles per hill were significant at P(0.05). (Refer to Data of UDP Evaluation Under Submerged Conditions at Kwin Yar, Kangyidaunt.) Plant height was taller with UDP than FP for both varieties. Swarna sub1 was significantly taller with UDP than with FP at 114 cm and 109 cm, respectively. Sin Thu Kha was also taller with UDP than with FP, but 114 cm with UDP was not statistically different than 112 cm with FP. The number of tillers per hill had the same type of response to fertilizer practices for both varieties. More tillers were produced with UDP than with FP. Swarna sub1 had 14.2 and 12.7 panicles/hill with UDP and FP, respectively, which was not statistically different. Sin Thu Kha had 13.3 panicles per hill with UDP, which was significantly higher than 10.7 panicles per hill produced with FP. Biomass straw and grain weights were greater with UDP than with FP for both varieties. However, within the same variety, almost all biomass weights, except for fresh straw, were not significantly different between UDP and FP. Fresh straw weight with UDP was significantly higher than that with FP FSI+ Report of 2017 Wet Season Trials 29

37 on Sin Thu Kha. Swarna sub1 had a significantly higher biomass than Sin Thu Kha with the same fertilizer practice (Table 11). The yields of both varieties were not significantly different between fertilizer practices. Sin Thu Kha produced the same yield (5.15 t/ha) with both fertilizer practices (FP and UDP). Swarna sub1 produced a slightly higher yield with FP than with UDP at 6.57 t/ha and 6.38 t/ha, respectively. A significant difference was found between varieties only. Swarna sub1 produced more yield than Sin Thu Kha with both fertilizer practices. As an average across fertilizer practices, Swarna sub1 produced 6.48 t/ha, which was significantly higher than 5.15 t/ha produced by Sin Thu Kha (Table 12). Therefore, UDP could improve some important traits and biomass weights for both varieties but not yield. Submerged conditions were not noted at Kwin Yar, Kangyidaunt, during the season. Both varieties performed well. However, standing water was deeper than 30 cm during the whole season. Sin Thu Kha, which gave less yield than Swarna sub1, may be affected by such deep water conditions, though it survived. In this trial, UDP yield (5.77 t/ha) was not better than FP yield (5.86 t/ha). However, improvement in other significant traits, such as plant height and biomass weights, was observed (Table 11 and Table 12). Mayan, Kunchangone ANOVA showed significant differences in five parameters and no differences in six parameters. Plant height and dry biomass straw weight were significant at P(0.01), and panicle length and fresh and dry biomass grain were significant at P(0.05). The yield, number of panicles per hill, number of spikelets and grains per panicle, 1,000 grain weight, and fresh biomass straw weight were not significantly different among treatments. (Refer to Data of UDP Evaluation Under Submerged Conditions at Mayan, Kunchangone.) According to LSD(0.05), there was no difference in plant height between the two fertilizer practices, but there was a significant difference between the two varieties. The plant height of Bay Gyar Lay was significantly taller than that of Swarna sub1. This is because Bay Gyar Lay is a local tall variety and Swarna sub1 is a semi-dwarf high-yielding variety. Although not significant, both varieties were taller with UDP than with FP. Bay Gyar Lay was 139 cm high with UDP and 129 cm high with FP. Swarna sub1 was 98 cm high with UDP and 91 cm high with FP (Table 11). Dry biomass straw weight was significantly different, and the Bay Gyar Lay variety with UDP gave the significantly highest dry straw weight (539 g). The second highest dry straw weight (376 g) was also produced by Bay Gyar Lay with FP, but it was not FSI+ Report of 2017 Wet Season Trials 30

38 different from the dry straw weight of Swanar sub1 with UDP (341 g) or Swarna sub1 with FP (265 g). Therefore, UDP significantly improved the dry biomass straw weight of Bay Gyar Lay because it is a taller plant with large vegetation. UDP did not improve the dry biomass straw weight of Swarna sub1. Fresh and dry biomass grain weights of Swarna sub1 were greater than those of Bay Gyar Lay. The fresh grain weight of Swarna sub1 with UDP was 363 g and the dry weight was 324 g. Bay Gyar Lay with UDP had a fresh grain weight of 206 g and a dry grain weight of 172 g. With FP, Swarna sub1 had a fresh grain weight of 280 g and a dry grain weight of 255 g, while those of Bay Gyar Lay were 220 g and 201 g, respectively. Dry grain weights of Swarna sub1 were significantly higher than those of Bay Gyar Lay for both fertilizer practices. Fresh grain weight of Swarna sub1 with UDP was significantly higher than those of Bay Gyar Lay with both UDP and FP. The weight of Swarna sub1 with FP was not significantly different than those of Bay Gyar Lay. According to the LSD(0.05) calculation, there was no difference due to fertilizer practice for either variety (Table 11). The difference was between varieties only. This is because Swarna sub1 is a high-yielding improved variety giving more grain weight than Bay Gyar Lay. Bay Gyar Lay is a local tall variety giving more straw weight than Swarna sub1. The yield was not significantly different at Mayan, Kunchangone. However, Swarna sub1 gave a higher yield than Bay Gyar Lay with both fertilizer practices. With FP, Swarna sub1 produced 2.92 t/ha and Bay Gyar Lay produced 2.45 t/ha. With UDP, Swarna sub1 produced 4.49 t/ha and Bay Gyar Lay produced 3.24 t/ha. As an average across fertilizer practices, Swarna sub1 produced 3.71 t/ha and Bay Gyar Lay produced 2.84 t/ha. The high-yielding improved variety produces a greater yield than the local variety. Comparing fertilizer practices, UDP was more effective than FP. UDP gave a greater yield than FP for both varieties. Bay Gyar Lay gave 2.45 t/ha and 3.24 t/ha with FP and UDP, respectively. Swarna sub1 gave 2.92 t/ha and 4.49 t/ha with FP and UDP, respectively. As an average across varieties, UDP produced 3.87 t/ha and FP produced 2.68 t/ha (Table 12). But there was no significant difference between varieties or between fertilizer practices. Submerged conditions were noted for five days at Mayan, Kunchangone. Swarna sub1 was totally underwater for five days but Bay Gyar Lay was not, since it is a tall variety. During the season, there was deep water most of the time with no way for it to drain out. Transplanting had to be done in deep water at Mayan, Kunchangone (Photo 3). Layout, bunds, and basal fertilizer application were difficult to accomplish. Bay Gyar Lay is a tall variety that does not go completely underwater if minor flooding occurs. Farmers in the Delta regions are still growing local tall varieties, such as Paw San, Mee Done, Nga Kywe, etc., in such deep water areas. High production is not expected. FSI+ Report of 2017 Wet Season Trials 31

39 Table 11 Component Traits of UDP Evaluation Under Submerged Conditions Trial at Test Locations Location Plant Height (cm) No. of Panicles per Hill No. of Grains per Panicle 1,000-Grain Weight (g) Fresh Straw Weight (g) Fresh Grain Weight (g) Dry Straw Weight (g) Dry Grain Weight (g) Treatment Kwin Yar, Kangyidaunt (Sin Thu Kha) Sin Thu Kha with FP 112ab 10.7ba ,018ca 349b 626ca 558b Sin Thu Kha with UDP 114aa 13.3aa ,091ba 359b 670bc 568b Swarna sub1 with FP 109ba 12.7ab ,241ab 462a 741ab 708a Swarna sub1 with UDP 114aa 14.2aa ,264aa 433a 771aa 668a Significant level * * ns ns ** ** * * Mayan, Kunchangone (Bay Gyar Lay) Bay Gyar Lay with FP 129a ba 376b 201bc Bay Gyar Lay with UDP 139a , ba 539a 172ca Swarna sub1 with FP 91b ab 265b 255ab Swarna sub1 with UDP 98b aa 341b 324aa Significant level ** ns ns ns ns * ** * Table 12 Treatment Yields of UDP Evaluation Under Submerged Conditions Trial at Test Locations Treatment Kwin Yar, Kangyidaunt Yield (t/ha) Variety Average Fertilizer Average Yield (t/ha) Mayan, Kunchangone Variety Average Fertilizer Average Farmers variety with FP 5.15b Farmers variety with UDP 5.15b Swarna sub1 with FP 6.57a Swarna sub1 with UDP 6.38a 4.49 Significant level ** ns According to results from the submergence trials, UDP can be applied effectively in an area of submerged or deep water. The variety can be chosen by the farmer. They can grow highyielding varieties, such as Sin Thu Kha, a variety widely grown in areas where water is not too deep, or local tall varieties in areas where the water depth is always more than 1.5 feet. The submergence-tolerant variety Swarna sub1 is still not popular in the region but is suitable for submerged conditions for up to two weeks. UDP application may be difficult in deep water. In such deep water areas, farmers still use tall varieties by transplanting method. They use a transplanting stick (fork) when transplanting seedlings of at least 45 days old. Topdressing with urea fertilizer is never practiced. FSI+ Report of 2017 Wet Season Trials 32

40 Conclusion UDP on Direct Seeded Rice Trials Of two locations, the trial at Magyi Kan, Kawhmu, showed a significant yield difference. The trial at Bawine, Pantanaw, did not show a significant yield difference among treatments, but plant height and panicle length were significant. The insignificant result may be due to initial soil N content. The control plot with 0 N treatment gave a high yield (5.08 t/ha) at Pantanaw, and the yield improvement with N fertilizer application was not significant. At Magyi Kan, Kawhmu, the number of panicles per hill was significant at P(0.05). The other traits were not significantly different. (Refer to Data of UDP on DSR Trial at Bawine, Pantanaw, and Data of UDP on DSR Trial at Magyi Kan, Kawhmu.) UDP application, whether at seeding time or 25 DAS, gave more yield than urea broadcasting and the control plot. The control plot gave the lowest yield at both locations. UDP at seeding time gave a higher yield than UDP at 25 DAS at Bawine, Pantanaw, but this was not significant. At Magyi Kan, Kawhmu, the result was the opposite; UDP at 25 DAS produced more yield than UDP at seeding, and this was significant. UDP at 25 DAS had an extremely high yield (Table 6). According to the results, UDP can improve the rice yield of direct seeded rice with 20 cm x 20 cm spacing. It can be applied at seeding time or 25 DAS. However, manually spaced seeding and sequential application of UDP at seeding time is impractical. UDP should be applied just before seeding or at germination. Application of UDP at 25 DAS is possible and can be precisely applied. The direct seeding practice is no longer used in farmers fields due to a lack of labor. Another direct seeding method, like drum seeding, should be considered, but it is not currently practiced in the Delta region because of a lack of skilled labor. Balanced Fertilization With Compound Fertilizer Trial This trial was designed to determine the effectiveness of compound fertilizer that farmers apply on rice. Farmers practice of compound fertilizer + urea with additional gypsum (Treatment 1) and compound + gypsum only (Treatment 4) were compared to balanced fertilization using compound fertilizer with PU, TSP, and gypsum (Treatment 3) or a balance of single element fertilizer, TSP, MOP, and gypsum with N applied as UDP (Treatment 2). The trial was conducted at one location in The Pyay Chaung, Myaungmya. ANOVA showed significant differences in nearly half of the recorded parameters, including yield. (Refer to Data of Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya.) The results reveal that balanced fertilization with TSP, MOP, gypsum, FSI+ Report of 2017 Wet Season Trials 33

41 and UDP is the best fertilizer practice. It gave the highest yield (9.94 t/ha) and the highest values in almost all traits. The highest yield (9.94 t/ha) was significantly higher than with all other treatments. Plant height (122 cm), panicles per hill (10.0), grains per panicle (181), and biomass straw and grain weights were highest with balanced fertilization with UDP. All significant characteristics were statistically higher than with all other treatments (Table 7). The second highest yield (8.25 t/ha) was produced with balanced fertilization with compound fertilizer and PU. Farmers practice of compound and urea gave the third highest yield at 7.86 t/ha, which was lower than both balanced fertilization treatments. Compound fertilizer with gypsum gave the lowest yield at 7.76 t/ha (Table 7). Therefore, farmers practice using compound fertilizer and urea could not supply enough nutrients, especially P and K. Balanced fertilization with straight or compound fertilizer could improve rice yield. Applying compound fertilizer was more costly than applying straight P and K fertilizer. The difference between the balanced treatments (Treatments 2 and 3) was between UDP and PU. UDP technology, therefore, is found to be the most effective application for N. Different N Rates Trial With UDP and PU This trial was conducted on the soil at Gyoe Phyu, Taikkyi, in which N was found to be deficient according to the omission pot trial. It was designed to determine how much N fertilizer is needed to produce the maximum yield on transplanted rice. ANOVA using a split block design showed no significant differences in any recorded traits among the N rates. The interaction of urea formulation and the rate also was not significant. However, the control plot (0 N) gave the lowest yield with both UDP and PU. The yield of the control plot was 3.69 t/ha with UDP and 3.28 t/ha with PU. Applying N fertilizer improved the rice yield. With UDP, only 50 kg N/ha was needed to produce the maximum yield of 4.94 t/ha. The yield decreased when the N rate was increased from 50 kg N/ha (Figure 2). The other yield-contributing characteristics, except for biomass straw weight, showed their highest values with the N rate of 50 kg/ha. Biomass straw weight continued to increase with up to 150 kg N/ha (Table 9). Therefore, if the N rate was increased from 50 kg/ha with UDP, only vegetative growth was enhanced, but grain production was not. With PU, the yield increased when the N rate was increased from 50 kg/ha. An N rate of 150 kg/ha produced the maximum yield (4.49 t/ha), and the yield decreased when N was applied at 200 kg/ha (Figure 2). The number of grains per panicle and biomass straw and grain weights also were the highest with the 150 kg N/ha rate. But plant height and panicle length showed their highest values with 200 kg N/ha (Table 8). Although not significant, these results indicate that the soil was deficient in N since FSI+ Report of 2017 Wet Season Trials 34

42 the lowest yield was given with 0 N. Less urea is needed with UDP than with PU to get the highest yield. To produce the highest yield, UDP was needed at 50 kg N/ha, which is close to the urea rate of FP using PU. But the PU treatment was needed at 150 kg N/ha, which is a very high rate for farmers. Urea application should be performed along with balanced fertilization of P, K, and S. The results confirm the higher nutrient efficiency of N application with UDP than with PU. Farmers should use this more efficient application method to obtain maximum yield with less urea. Different P Rates Trial This trial was established in the same location at Gyoe Phyu, Taikkyi, where the N rate trial was conducted. The soil was also deficient in P according to the omission pot trial. ANOVA found there were no significant differences among treatments in any recorded traits. There were variations among the traits. Plant height and biomass grain weights were the highest with a P rate of 50 kg P2O5/ha, which was the lowest P rate applied except for control. Number of grains per panicle, biomass straw weights, and 1,000 grain weight were the highest with 100 kg P2O5/ha. Panicle length was longest with 150 kg P2O5/ha. The control plot had the highest number of panicles per hill (11.3) and spikelets per panicle (201). (Refer to Data of Different N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi.) The yield was not significant among treatments. However, the control plot with 0 kg P2O5/ha produced the lowest yield (7.70 t/ha). The yield increased with increasing the P rate up to 150 kg P2O5/ha; it was 8.18 t/ha with 50 kg P2O5/ha, 8.38 t/ha with 100 kg P2O5/ha, and 9.26 t/ha with 150 kg P2O5/ha. The yield decreased to 8.23 t/ha when the P rate was increased to 200 kg P2O5/ha (Figure 3). The results indicate that P also is deficient in the soil at Gyoe Phyu, Taikkyi. It is more critical than N. The yield could be improved by nearly 0.5 t/ha with P fertilizer application at 50 kg P2O5/ha. The soil needs 150 kg P2O5/ha for maximum production. Less than that amount may be more economical with optimum production. Different K Rates Trial The K rate trial was on the same soil at Gyoe Phyu, Taikkyi, where K was found to be deficient according to the omission pot trial. FSI+ Report of 2017 Wet Season Trials 35

43 ANOVA showed that plant height was highly significant between treatments at P(0.01), and number of grains per panicle was significant at P(0.05). The other traits, including yield, were not significant. The tallest plant (103 cm) was produced with the application of 150 kg K2O/ha. It was significantly taller than with other treatments. Plant heights with other treatments ranged from 98 cm to 100 cm, which were not statistically different. Number of grains per panicle was greatest at 184 with 200 kg K2O/ha; this was significantly higher than with other treatments. (Refer to Data of Different K Rates Trial at Gyoe Phyu, Taikkyi.) The yield was not significant. But the yield increased with increasing the rate of K2O. The highest yield (10.18 t/ha) was produced with a K rate of 150 kg K2O /ha. When the K rate was increased from 150 kg K2O/ha to 200 kg K2O/ha, the yield decreased. A K rate of 200 kg K2O/ha produced 8.84 t/ha, which was lower than 9.06 t/ha produced by 100 kg K2O/ha and higher than 7.49 t/ha produced by 50 kg K2O/ha. The control plot produced the lowest yield at 6.78 t/ha (Figure 4). The results indicate that K was deficient in the test soil. The yield was improved by applying K fertilizer. According to yield results, although not significant, 150 kg K2O/ha is needed for maximum production. Different S Rates Trial ANOVA showed only one trait, fresh straw weight, was significant at P(0.05). Other traits were not significantly different among treatments. Biomass straw weight was greatest with an S rate of 20 kg S/ha. It was not significantly different from the second highest fresh straw weight of 2,181 g produced with the highest S rate (40 kg S/ha). The other biomass straw weights were not significantly different, ranging from 1,734 g to 1,832 g. Dry biomass straw weight and biomass grain weights were highest with the lowest S application rate of 10 kg S/ha. Number of panicles per hill and number of spikelets and grains per panicle also were the highest with 20 kg S/ha. (Refer to Data of Different S Rates Trial at Gyoe Phyu, Taikkyi.) The yield also was the highest (10.08 t/ha) with 20 kg S/ha. The yield decreased when the S rate was increased to 30 kg S/ha and 40 kg S/ha. Even though the control plot gave the second highest yield (9.02 t/ha), the test soil could be deficient in sulfur since application of S at 20 kg S/ha improved the rice yield. A rate of up to 20 kg S/ha, but no higher, should be recommended to improve rice yield. Later, TSP was found to contain a certain amount of S due to its production process. Therefore, the control plot in this trial received S from the TSP, resulting in a high yield. FSI+ Report of 2017 Wet Season Trials 36

44 All of the rate trials for N, P, K, and S were conducted on the soil at Gyoe Phyu, Taikkyi, where these elements were identified as deficient according to the omission pot trials. All rate trials showed the lowest yield with the control plots except for the S rates trial. N is assumed to be the most limiting element, since the 0 N plot had a very low yield at 3.69 t/ha in the UDP trial and 3.28 t/ha in the PU trial. The second most limiting element was K with a yield at 0 K of 6.78 t/ha, followed by P and S. The yield could be improved with application of N, P, K, and S. Generally, the rice yield could be improved by around 35% with N fertilizer application, and less N is needed with UDP than with PU. With UDP and PU, 50 kg N/ha (Rate 1) and 150 kg N/ha, respectively, were required for this soil. With application of P and S, the yield could be improved by about 20%. P at 150 kg P2O5/ha and S at 20 kg S/ha would be needed. Application of K could improve the rice yield by 50%, and a rate of 150 kg K2O/ha is needed (Table 13). Based on yield improvement, K is shown to be the responsive element, followed by N and P or S. Table 13 Yield and Yield Increase of Different Treatments Over Control from Rate Trials at Gyoe Phyu, Taikkyi Yield (t/ha) % Yield Increase Over Control Yield (t/ha) % Yield Increase Over Control Yield (t/ha) % Yield Increase Over Control Yield (t/ha) % Yield Increase Over Control Yield (t/ha) % Yield Increase Over Control Treatment UDP Treatment PU Treatment P Treatment K Treatment S Treatment Control Rate Rate Rate Rate Note: Rates 1-4 are 50 kg, 100 kg, 150 kg, and 200 kg for N, P 2O 5, and K 2O and 10 kg, 20 kg, 30 kg, and 40 kg for S. Fertilizer Management Trials on the Rice-Gram Cropping System This was the beginning of the third cycle of the rice-gram system with the third rice crop. After the first cycle, there was no significant carryover effect of P and K on the subsequent crop. That was assumed to be due to inadequate rates of P and K application; 40 kg P2O5/ha and 40 kg K2O/ha were applied on both crops. The second cycle used higher rates of P and K (60 kg/ha each). The second cycle also showed no carryover effect on the subsequent crop. However, FSI+ Report of 2017 Wet Season Trials 37

45 application of the full amount of P and K on each crop gave high yields for both the rice and the gram crop. This season was the beginning of second cycle with P and K rates of 60 kg/ha each. ANOVA showed significant results at P(0.05) on yield and fresh biomass straw weight. The other characteristics were not significant. The rice with half the amount of P and K (30 kg/ha each) gave the lowest yield at 5.45 t/ha. The plot received the same amount of P and K on the gram crop. The results indicate that P and K applied on the gram crop was used by the gram crop, and no significant amount remained for the rice crop. P and K applications of 30 kg/ha each on the rice crop seemed inadequate to produce a high yield. This yield was statistically not different from the second lowest yield (5.57 t/ha) produced by no P and K but with crop residue applied on the rice after the full amount of P and K had been applied on the gram crop. The carryover effect of P and K from the gram crop may be equivalent to 30 kg/ha. All P and K may not have been absorbed by the gram crop. An appreciable amount may have remained for the rice crop. Crop residue still did not show an effect on the rice, since the yield (5.57 t/ha) was similar to or a little lower than the yield (5.63 t/ha) of the same P and K treatment without crop residue, which was the highest yield. The second highest yield (5.60 t/ha) was given by the full P and K application treatment. The results indicate that, if there were leftover P and K from the gram crop, the rice did not need the full amount (60 kg/ha) of P and K application. However, availability of nutrients to the gram crop largely depends on soil moisture content, which is low in the dry season. If there was an irrigation system in the dry season, this amount of P and K could have been taken up by the gram crop with no carryover effect shown on the rice crop. The effect of gram crop residue on rice yield is not significant due to less biomass of gram crop within two crop cycles. Longer term application of crop residue is needed to improve soil physical properties and rice yield. Fertilizer Management Trial on the Rice-Gram System With Starter N on Gram This was the second rice crop of the second cycle in the rice-gram system. Unlike the first rice crop of the first cycle, there was no significant difference in any traits, and this was compounded by the results being variable among traits. In this season as an average across different basal fertilizer applications, UDP yield was not better than PU. UDP produced 3.84 t/ha and PU produced 4.01 t/ha. That was because the yield (3.44 t/ha) of UDP with 0 P and 0 K was much lower than that (4.29 t/ha) of PU with the same basal. UDP with the full amount (60 kg/ha) of P and K produced more yield than PU with the same basal. But FSI+ Report of 2017 Wet Season Trials 38

46 differences were very small. Among UDP application treatments, plots with basal fertilizer application gave more yield than those without. However, among PU applied treatments, the yield was better without basal than with basal. The results make it difficult to reach a conclusion on the best fertilizer application method. The trial was conducted on broadcast rice and seeds were broadcast on uneven land, leading to uneven germination. Plant population in any given area varied. Harvesting broadcast rice can be difficult, and crop cut area was smaller than that specified by the protocol. UDP Evaluation Under Submerged Conditions Trial Two out of three locations could be harvested one at Kangyidaunt and one at Kunchangone. Both showed significant differences in many traits. The Kangyidaunt harvest showed that all biomass weights, plant height, number of panicle per hill, and yield were significant. Two yield-contributing characteristics number of grains per panicle and 1,000 grain weight were not significant. The Kunchangone harvest showed plant height, panicle length, dry straw weight, and biomass grain weights were significant. The yield was not significant and neither were some yield-determining traits, such as number of panicles per hill, number of grains per panicle, and 1,000 grain weight (Table 11 and Table 12). Both varieties gave higher values with UDP than that with FP in many traits at Kwin Yar, Kangyidaunt. Significant traits, such as plant height, number of panicles per hill, and most biomass weights of both varieties, were higher with UDP than that with FP. However, the number of grains per panicle and biomass grain weights of Swarna sub1 were higher with FP than with UDP (Table 11). The significant yield difference at Kwin Yar, Kangyidaunt, was due to variety. The yields of both varieties with different fertilizer practices were not significantly different. Sin Thu Kha gave the same yield (5.15 t/ha) with UDP and FP. Swarna sub1 produced a yield of 6.57 t/ha with FP and 6.38 t/ha with UDP, which was not significantly different although FP was slightly higher than UDP. However, Swarna sub1 gave a significantly higher yield than Sin Thu Kha. As an average across fertilizer practices, Swarna sub1 produced 6.48 t/ha and Sin Thu Kha produced 5.15 t/ha. At Mayan, Kunchangone, ANOVA showed five significant traits and six traits that were not significant. Plant height, panicle length, dry biomass straw weight, and biomass grains weight were significant. The other traits and yield were not significant. (Refer to Data of UDP FSI+ Report of 2017 Wet Season Trials 39

47 Evaluation Under Submerged Conditions at Mayan, Kunchangone.) Bay Gyar Lay, a tall variety, was significantly taller than Swarna sub1. The same trend was observed in biomass straw weights; some were significant and some were not. But biomass grain weights of Swarna sub1 were higher than those of Bay Gyar Lay (Table 11). The yields of Swarna sub1 with both fertilizer practices were higher than those of Bay Gyar Lay though not significant. As an average across fertilizer practices, Swarna sub1 gave 3.71 t/ha and Bay Gyar Lay gave 2.84 t/ha. Both varieties produced higher yields with UDP than with FP. As an average, UDP produced 3.87 t/ha and FP produced 2.68 t/ha. The results indicate that both varieties were more productive with UDP than with FP (Table 12). This trial encountered submerged conditions for five days during the season. However, Bay Gyar Lay was not completely underwater. Swarna sub1, a submergence-tolerant variety, was not affected by five days underwater. Therefore, Swarna sub1, as a high-yielding variety, produced a higher yield than Bay Gyar Lay, a local non-hyv. Both trials showed UDP application is more efficient than broadcast PU. But manual UDP application in such deep water areas is not practical. Water depth at transplanting is often around 1.5 feet. Farmers in the area are growing local tall varieties, which are less responsive to fertilizer, especially urea. Farmers transplant the seedling using a transplanting stick with a fork on the tip in deep water and apply little to no fertilizer. FSI+ Report of 2017 Wet Season Trials 40

48 Appendix 1. Data Sheets Data of UDP on DSR Trial at Bawine, Pantanaw Exp. Design 4 x 3 RCB Treatments Location Bawine, Pantanaw 1 Control (0 N) Year/season 2017 wet Season 2 UDP at seeding time Farmer U Sein Win 3 UDP at 25 DAS Variety Sinthukha 4 PU broadcast Plot size 256 sq ft. UDP date 18-Jul Harvested area 100 sq. ft. Sowing Date 23-Jun Harvest date 6-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 N 6-Oct UDP(seeding) 6-Oct UDP at 25DAS 6-Oct PU B'st 6-Oct N 6-Oct UDP(seeding) 6-Oct UDP at 25DAS 6-Oct PU B'st 6-Oct N 6-Oct UDP(seeding) 6-Oct UDP at 25DAS 6-Oct PU B'st 6-Oct Average values 1 0 N UDP(seeding) UDP at 25DAS PU B'st F test ** ns * ns ns ns ns ns ns ns ns CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 41

49 Data of UDP on DSR Trial at Magyi Kan, Kawhmu Exp. Design 4 x 3 RCB Location Magyi Kan, Kawhmu Treatments Year/season 2017 Wet Season 1 Control (0 N) Farmer U Zaw Htay 2 UDP at seeding time Variety Sin Thu Kha 3 UDP at 25 DAS Plot size 256 sq ft. 4 PU broadcast Sowing Date 22-Jul UDP date 17-Aug Harvested area 100 sq. ft. Harvest date 23-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 N 4-Nov UDP(seeding) 4-Nov UDP at 25DAS 4-Nov PU B'st 4-Nov N 4-Nov UDP(seeding) 4-Nov UDP at 25DAS 4-Nov PU B'st 4-Nov N 4-Nov UDP(seeding) 4-Nov UDP at 25DAS 4-Nov PU B'st 4-Nov Average values 1 0 N UDP(seeding) UDP at 25DAS PU B'st F test ns * ns ns ns ns ns ns ns ns ** CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 42

50 Data of Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya Exp. Design 4 x 3 RCB Location The Pyay Chung, Myaungmya Year/season 2017 Wet Season F1 = Compd + Gyp + PU Farmer U Win Thein F2 = Balanced with TSP, MOP, Gyp + UDP Variety Sinthukha F3 = Balanced with Compound used PU Plot size 256 sq ft. F4 = Compd + Gypsum only Sowing Date 7-Jun UDP date 12-Jul Harvested area 100 sq. ft. Transplant Date 4-Jul Harvest date 8-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 F1 26-Sep F2 26-Sep F3 26-Sep F4 26-Sep F1 26-Sep F2 26-Sep F3 26-Sep F4 26-Sep F1 26-Sep F2 26-Sep F3 26-Sep F4 26-Sep Average values 1 F F F F F test * ns ns ns ns ** ns ** * ns ** CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 43

51 Data of UDP and PU Deep Placement Trial at Kan Gyi, Letpadan Exp. Design 5 x 3 RCB Location Kan Gyi, Letpadan Year/season 2017 Wet Season Farmer U Aye Thwin Variety Sinthukha Plot size 256 sq ft. Sowing Date 8-Jun UDP date 15-Jul Harvested area 100 sq. ft. Transplant Date 8-Jul Harvest date 28-Oct Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 Control 24-Sep UDP 24-Sep PU point 24-Sep PU band 24-Sep PU B'cast 24-Sep Control 24-Sep UDP 24-Sep PU point 24-Sep PU band 24-Sep PU B'cast 24-Sep Control 24-Sep UDP 24-Sep PU point 24-Sep PU band 24-Sep PU B'cast 24-Sep Average values 1 Control UDP PU point PU band PU B'cast F test ns bs ns ns ** * ns ns * ns * CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 44

52 Data of UDP and PU Deep Placement Trial at The Yet Kone, Nyaunglaybin Exp. Design 5 x 3 RCB Location The Yet Kone, Nyaunglebin Year/season 2017 Wet Season Farmer U Myo Lwin Soe Variety Sinthukha Plot size 256 sq ft. Sowing Date 10-Jun UDP date 13-Jul Harvested area 100 sq. ft. Transplant Date 6-Jul Harvest date 2-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 Control 27-Sep UDP 27-Sep PU point 27-Sep PU band 27-Sep PU B'cast 27-Sep Control 27-Sep UDP 27-Sep PU point 27-Sep PU band 27-Sep PU B'cast 27-Sep Control 27-Sep UDP 27-Sep PU point 27-Sep PU band 27-Sep PU B'cast 27-Sep Average values 1 Control UDP PU point PU band PU B'cast F test ns ns ns ns ns ns ns ns ns ns ** CV (%) 13.5 LSD (0.05) 1.1 FSI+ Report of 2017 Wet Season Trials 45

53 Data of Different N Rates Trial Using UDP and PU at Gyoe Phyu, Taikkyi Exp. Design 2 x 5 x 3 split block Location Gyoe Phyu, Taikkyi Year/season 2017 Wet Season Farmer U Kyaw Lin Thu Variety Sin Thu Kha Plot size 256 sq ft. Sowing Date 6-Jun UDP date 9-Jul Harvested area 100 sq. ft. Transplant Date 1-Jul Harvest date 27-Oct Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UP 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UP 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UDP 24-Sep N UP 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep N PU 24-Sep Average values 1 0 N UDP N UDP N UDP N UDP N UDP N UP N PU N PU N PU N PU FSI+ Report of 2017 Wet Season Trials 46

54 Data of Different P Rates Trial at Gyoe Phyu, Taikkyi Exp. Design 5 x 3 RCB Location Gyoe Phyu, Taikkyi Year/season 2017 Wet Season Farmer U Kyaw Lin Thu Variety Sin Thu Kha Plot size 256 sq ft. Sowing Date 6-Jun UDP date 9-Jul Harvested area 100 sq. ft. Transplant Date 1-Jul Harvest date 26-Oct Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep P 2 O 5 23-Sep Average values 1 0 P 2 O P 2 O P 2 O P 2 O P 2 O F test ns ns ns ns ns ns ns ns ns ns ns FSI+ Report of 2017 Wet Season Trials 47

55 Data of Different K Rates Trial at Gyoe Phyu, Taikkyi Exp. Design 5 x 3 RCB Location Gyoe Phyu, Taikkyi Year/season 2017 Wet Season Farmer U Kyaw Lin Thu Variety Sin Thu Kha Plot size 256 sq ft. Sowing Date 6-Jun UDP date 8-Jul Harvested area 100 sq. ft. Transplant Date 1-Jul Harvest date 25-Oct Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep K 2 O 24-Sep Average values 1 0 K 2 O K 2 O K 2 O K 2 O K 2 O F test ** ns ns ns * ns ns ns ns ns ns CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 48

56 Data of Different S Rates Trial at Gyoe Phyu, Taikkyi Exp. Design 5 x 3 RCB Location Gyoe Phyu, Taikkyi Year/season 2017 Wet Season Farmer U Kyaw Lin Thu Variety Sin Thu Kha Plot size 256 sq ft. Sowing Date 6-Jun UDP date 8-Jul Harvested area 100 sq. ft. Transplant Date 1-Jul Harvest date 24-Oct Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 0 kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep kg S/ha 24-Sep Average values 1 0 kg S/ha kg S/ha kg S/ha kg S/ha kg S/ha F test ns ns ns ns ns * ns ns ns ns ns CV (%) 8.7 LSD (0.05) FSI+ Report of 2017 Wet Season Trials 49

57 Data of Fertilizer Management Trial on the Rice-Gram System at Thar Kwin, Einme Exp. Design 4 x 3 RCB 1 basket = kg Location Thar Kwin, Einme 1 hectare = acre Year/season 2017, Wet Season T1 60 kg P2O kg K2O Farmer U Aung Htay Win T2 30 kg P2O kg K2O Variety Sin Thukha T3 0 kg P2O5 + 0 kg K2O + no residue Plot size 256 ft2 T4 0 kg P2O5 + 0 kg K2O + Crop Residue Sowing Date 14-Jun UDP date 31-Jul 146 Harvest area 100 ft2 Transplant Da 24-Jul Harvest date 7-Nov Harvest plants 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield Rep ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 1 T1 1-Oct T2 1-Oct T3 1-Oct T4 1-Oct T1 1-Oct T2 1-Oct T3 1-Oct T4 1-Oct T1 1-Oct T2 1-Oct T3 1-Oct T4 1-Oct Average values 1 T T T T F test ns ns ns ns ns * ns ns ns ns * CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 50

58 Data of Fertilizer Management Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa Exp. Design 6 x 3 RCB Location Anauk Ywa, Thongwa 1 basket = kg Year/season 2017, Wet Season 1 hectare = acre Farmer U Zaw Wate Variety Nang Myaing Plot size 256 ft2 Sowing Date 17-Jun UDP date 11-Jul (24 DAS) Harvest area 100 ft2 Harvest date 30-Oct Rice crop per meter square Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/100 ft2 Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per m2 (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 UDP 60P 60K 30-Sep UDP 0P 0K 30-Sep PU 60P 60K 30-Sep PU 0P 0K 30-Sep UDP 60P 60K 30-Sep PU 60P 60K 30-Sep UDP 60P 60K 30-Sep UDP 0P 0K 30-Sep PU 60P 60K 30-Sep PU 0P 0K 30-Sep UDP 60P 60K 30-Sep PU 60P 60K 30-Sep UDP 60P 60K 30-Sep UDP 0P 0K 30-Sep PU 60P 60K 30-Sep PU 0P 0K 30-Sep UDP 60P 60K 30-Sep PU 60P 60K 30-Sep AVERAGE 1 UDP 60P 60K UDP 0P 0K PU 60P 60K PU 0P 0K UDP 60P 60K PU 60P 60K F test ns ns ns ns ns ns ns ns ns ns ns FSI+ Report of 2017 Wet Season Trials 51

59 Data of UDP Evaluation Under Submerged Conditions at Kwin Yar, Kangyidaunt Exp. Design 4 x 3 RCB Location Kwin Yar, Kangyidaunt Year/season 2017 Wet Season Variety Fertilizer Farmer U San Oo V1 Local (STK) F1 FP (15:15:15 50 kg and Urea 50 kg/ac)urea 3 splits Variety Sin Thu Kha (STK) V2 Swarna sub1 F2 UDP Plot size 256 sq ft. Sowing Date 26-Jun UDP date 28-Jul Harvested area 100 sq. ft. Transplant Date 21-Jul Harvest date 10-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 V1 F1 9-Oct V1 F2 9-Oct V2 F1 9-Oct V2 F2 9-Oct V1 F1 9-Oct V1 F2 9-Oct V2 F1 9-Oct V2 F2 9-Oct V1 F1 9-Oct V1 F2 9-Oct V2 F1 9-Oct V2 F2 9-Oct Average values 1 V1 F V1 F V2 F V2 F F test * * ns ns ns ** ** * * ** CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 52

60 Data of UDP Evaluation Under Submerged Conditions at Mayan, Kunchangone Exp. Design 4 x 3 RCB Location Mayan, Kungyangon Year/season 2017 Wet Season Variety Fertilizer Farmer U Zaw Min Oo V1 Local (Bay Gyar) F1 FP (15:15:15 50 kg and Urea 50 kg/ac)urea 3 splits Variety Bay Gyar Lay V2 Swarna sub1 F2 UDP Plot size 256 sq ft. Sowing Date 17-Jun UDP date 22-Jul Harvested area 100 sq. ft. Transplant Date 14-Jul Harvest date 24-Nov Harvested hills 225 Treat Date Days Plant No of Panicle No of No of Spikelet Fresh Biomass Dry Biomass 1000 Yield/225 hills Yield ment Treatment of to height panicles Length spikelets grains per Fertility Straw Grain Straw Grain grain Wet Calculat MC No Flowering Maturity (cm) per hill (cm) /panicle panicle % wt. (g) wt. (g) wt. (g) wt. (g) wt. (g) (g) (g) (%) t/ha 1 V1 F1 10-Oct V1 F2 10-Oct V2 F1 10-Oct V2 F2 10-Oct V1 F1 10-Oct V1 F2 10-Oct V2 F1 10-Oct V2 F2 10-Oct V1 F1 10-Oct V1 F2 10-Oct V2 F1 10-Oct V2 F2 10-Oct Average values 1 V1 F V1 F V2 F V2 F F test ** ns * ns ns ns * ** * ns ns CV (%) LSD (0.05) FSI+ Report of 2017 Wet Season Trials 53

61 Photos Photo 1. Layout and Bunds of Fertilizer Trial on the Rice-Gram System With Starter N at Anauk Ywa, Thonegwa, Yangon Region, on June 16, 2017 Photo 2. UDP Application during the UDP and PU Deep Placement Trial at The Yet Kone, Nyaunglaybin, Bago Region, on July 13, 2017 FSI+ Report of 2017 Wet Season Trials 54

62 Photo 3. Transplanting UDP Evaluation Under Submerged Conditions Trial at Mayan, Kunchangone, Yangon Region, on July 14, 2017 (Note water depth at transplanting time in flood- or submergenceprone area) Photo 4. First Topdressing of UDP on DSR Trial at Bawine, Pantanaw, Ayeyarwady Region, on August 22, 2017 FSI+ Report of 2017 Wet Season Trials 55

63 Photo 5 Harvesting Crop Cuts during the Balanced Fertilization With Compound Fertilizer Trial at The Pyay Chaung, Myaungmya, on November 8, 2017 FSI+ Report of 2017 Wet Season Trials 56