54 © 2025 by the authors; licensee Asian Online Journal Publishing Group Agriculture and Food Sciences Research Vol. 12, No. 1, 54-68, 2025 ISSN(E) 2411-6653/ ISSN(P) 2518-0193 DOI: 10.20448/aesr.v12i1.6819 © 2025 by the authors; licensee Asian Online Journal Publishing Group On-farm evaluation of fertilizer recommendation methods: Impacts on rice yield and economic benefits in subtropical soils Md. Enamul Haque1 Richard W. Bell2 Miaomiao Cheng3 Md. Abdus Satter4 Md. Baktear Hossain5 Md. Jahiruddin6 Md. Mofizur Rahman Jahangir7 Md. Aminul Islam8 Sohela Akhter9 Md. Monayem Miah10 A. K. M. Anisur Rahman11 ( Corresponding Author) 1,2,3School of Agricultural Science, Murdoch University, Murdoch 6150, Australia. 1Email: Enamul.haque71@gmail.com 2Email: R.Bell@Murdoch.edu.au 3Email: Miaomiao.Cheng@dpird.wa.gov.au 4,5Bangladesh Agricultural Research Council, Farmgate, Dhaka, Bangladesh. 4Email: A.Satter1959@gmail.com 5Email: Baktear@gmail.com 6,7Department of Soil Science, Bangladesh Agricultural University, Mymensingh, Bangladesh. 6Email: M_jahiruddin@yahoo.com 7Email: Mmrjahangir@bau.edu.bd 8Soil Science Division, Bangladesh Rice Research Institute, Gazipur, Bangladesh. 8Email: Aminbrri@gmail.com 9Soil Science Division, Bangladesh Agricultural Research Institute, Gazipur, Bangladesh. 9Email: Sohela_akhter@yahoo.com 10Agricultural Economics Division, Bangladesh Agricultural Research Institute, Gazipur, Bangladesh. 10Email: Monayem09@yahoo.com 11Department of Medicine, Bangladesh Agricultural University, Mymensingh, Bangladesh. 11Email: Arahman_med@bau.edu.bd Abstract Bangladesh publishes fertilizer recommendations but most farmers do not use these recommendations. In this study, we tested the performance of the Fertilizer Recommendation Guide (FRG), Rice Crop Manager (RCM) and Soil Testing Kit (STK) for determining fertilizer requirements of monsoon rice (Oryza sativa L) against Farmers’ Fertilizer Practice (FFP). Rates of nitrogen, phosphorus, potassium, sulfur and zinc were determined by four methods and used on 72 farmer’s fields in five agro-ecological zones (AEZ) over two years. All methods produced significantly higher grain yield than FFP (4.12 t ha-1): they were 13.8%, 9.6% and 8.3% higher for STK, RCM and FRG, respectively. RCM didn’t perform consistently across the locations. Overall, the STK performed the best followed by FRG. The superiority of STK is attributed to its assessment of current soil status while the FRG recommendation is grounded on an older data set of soil analysis for each AEZ. The N dose was comparable among the methods while P dose was much higher for FFP, and RCM underestimated the K requirement of rice. Farmers declined to adopt the STK. By contrast, the supply of FRG information to the farmers by providing a simplified card with training is a simple and accessible technology. Keywords: Adoption, FRG card, Nitrogen, Phosphorus, Potassium, Rice Crop Manager, Soil Test Kit, Sulfur, Zinc. Contents 1. Introduction ...................................................................................................................................................................................... 55 2. Materials and Methods ................................................................................................................................................................... 55 3. Results ................................................................................................................................................................................................ 61 4. Discussion .......................................................................................................................................................................................... 64 5. Conclusion ......................................................................................................................................................................................... 66 References .............................................................................................................................................................................................. 66 mailto:Enamul.haque71@gmail.com mailto:R.Bell@Murdoch.edu.au mailto:Miaomiao.Cheng@dpird.wa.gov.au mailto:A.Satter1959@gmail.com mailto:Baktear@gmail.com mailto:M_jahiruddin@yahoo.com mailto:Mmrjahangir@bau.edu.bd mailto:Aminbrri@gmail.com mailto:Sohela_akhter@yahoo.com mailto:Monayem09@yahoo.com mailto:Arahman_med@bau.edu.bd https://www.doi.org/10.20448/aesr.v12i1.6819 https://orcid.org/0000-0002-6551-5174 https://orcid.org/0000-0002-7756-3755 https://orcid.org/0000-0002-2979-6976 https://orcid.org/0000-0002-3636-1057 https://orcid.org/0009-0009-6936-5634 https://orcid.org/0000-0002-7944-2156 https://orcid.org/0000-0002-9501-3239 https://orcid.org/0000-0003-0110-4488 https://orcid.org/0000-0002-8387-0727 https://orcid.org/0000-0002-1054-4437 https://orcid.org/0000-0001-9660-4949 Agriculture and Food Sciences Research, 2025, 12(1): 54-68 55 © 2025 by the authors; licensee Asian Online Journal Publishing Group Citation | Haque, M. E., Bell, R. W., Cheng, M., Satter, M. A., Hossain, M. B., Jahiruddin, M., Jahangir, M. M. R., Islam, M. A., Akhter, S., Miah, M. M., & Rahman, A. K. M. A. (2025). On-farm evaluation of fertilizer recommendation methods: Impacts on rice yield and economic benefits in subtropical soils. Agriculture and Food Sciences Research, 12(1), 54–68. 10.20448/aesr.v12i1.6819 History: Received: 7 April 2025 Revised: 15 May 2025 Accepted: 16 June 2025 Published: 25 June 2025 Licensed: This work is licensed under a Creative Commons Attribution 4.0 License Publisher: Asian Online Journal Publishing Group Funding: This research is supported by the Australian Centre for International Agricultural Research (ACIAR), Australia (Grant number: LWR/2016/136) and the Krishi Gobeshona Foundation (KGF), Bangladesh (Grant number: CN/FRPP: ICP-II). Institutional Review Board Statement: Not applicable. Transparency: The authors confirm that the manuscript is an honest, accurate, and transparent account of the study; that no vital features of the study have been omitted; and that any discrepancies from the study as planned have been explained. This study followed all ethical practices during writing. Competing Interests: The authors declare that they have no competing interests. Authors’ Contributions: All authors contributed equally to the conception and design of the study. All authors have read and agreed to the published version of the manuscript. Contribution of this paper to the literature: Tools have been developed to assist farmers in making better fertilizer decisions, but they are rarely tested against one another and demonstrated to be more effective than farmers' current fertilizer use. We demonstrated that an expert-driven fertilizer recommendation system was accessible to smallholder farmers and achieved an 8% average yield increase. 1. Introduction Rice (Oryza sativa L.) is the most dominant crop in Asia while about 40% people of the world consume rice as a major staple food. Rice crops in Bangladesh occupy about 75% of the crop area and rice-based cropping system is predominant [1]. The rice soils are commonly deficient in N, P, K, S and Zn. In the intensive, high yield cropping systems of south Asia, especially in India and Bangladesh, farmers generally apply excessive P fertilizer to rice but generally insufficient K, S or Zn [2]. Nitrogen applications relative to recommendations vary with farm size, cropping pattern and crop species [2]. However, improvements in N use efficiency can increase crop yields [3] since commonly less than 50% of urea-N is taken up by the crops and the rest is lost through leaching (NO3), volatilization of NH3 [4] and denitrification (N2O, N2) processes [5] and thereby contaminating environment [6]. Optimising the use of fertilizers continues to be a global concern since crop yield loss occurs with underuse while the risk of environmental loss increases with overuse. There are numerous fertilizer recommendation systems for crop cultivations globally. According to Beneduzzi, et al. [7] a literature search identified 12 methods for recommending N, eight methods for P, and seven methods for recommending K, in addition to five computer programs to make fertilizer recommendations at varying rates. Buresh, et al. [8] determined fertilizer K and P requirements for rice (Oryza sativa L.) cultivation by the site-specific nutrient management (SSNM) approach using estimated target yield, nutrient balances, and yield gains from added nutrients. That approach was based on model estimates without reliance on fresh soil analysis [9]. Dobermann and Cassman [10] argued that effective strategies for site-specific nutrient management should be based on a quantitative understanding of the relationships between nutrient supply and crop demand. Usually, soil analysis is considered as an essential step for evaluation of soil fertility and thus on the fertilizer recommendation. Soil tests provide a scientific basis for fertilizer recommendations. Understanding and measuring spatial variability regarding nutrient availability in the soil is crucial to defining site-specific fertilizer management strategies to increase production efficiency and sustainability of agricultural production [11]. However, soil chemical analysis is expensive and time consuming to the point of being economically unfeasible for site-specific nutrient application in low value-added crops [12]. For smallholder farms, spatial and temporal variability in nutrient status on many small fields adds to the cost of using soil testing for site specific nutrient management. There are three methods of fertilizer recommendation now available in Bangladesh – FRG [13] Rice Crop Manager (RCM) and Soil Testing Kit (STK). However, the majority of farmers do not use those methods [2] and employ their own calculation which in this study we have called Farmers’ Fertilizer Practice (FFP). The fertilizer rates proposed may vary among recommendation methods but their impact on crop yield and profitability is not known. With this understanding, we studied performances of three fertilizer recommendation methods relative to FFP, with two objectives: (i) to determine the fertilizer (urea, TSP, MoP, gypsum and zinc sulphate) requirement for transplanted aman rice, and (ii) to identify the best method of fertilizer recommendation for rice in terms of yield benefits, economic return and adoptability by smallholder farmers. 2. Materials and Methods 2.1. Experimental Site and Soil The field trials were conducted in six locations representing five agro-ecological zones (AEZ) of the country. In each location, the experiment was set up in six sites representing six farmer plots and it was done in two years (2018 and 2019) giving a total of 72 field trials (6 locations × 6 sites × 2 years) (Figure 1). All the experimental sites were on medium-high land. Bangladesh has a sub-tropical humid climate and is characterized by hot and humid summer and cool dry winter. The country experiences monsoon rainfall, 80% of which occurs between June and October when transplanted Aman rice is grown. General characteristics of the soils under the experiments are presented in Table 1. The methods for soil analysis were followed as outlined by Page, et al. [14]. https://creativecommons.org/licenses/by/4.0/ https://creativecommons.org/licenses/by/4.0/ Agriculture and Food Sciences Research, 2025, 12(1): 54-68 56 © 2025 by the authors; licensee Asian Online Journal Publishing Group Figure 1. Map of Bangladesh showing trial locations. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 57 © 2025 by the authors; licensee Asian Online Journal Publishing Group Table 1. General characteristics of soils in different locations and sites under experiments. Parameters Durgapur Godagari Mymensingh Sadar Thakurgaon Sadar Dacope Amtali Methods AEZ 11 26 9 1 13 13 GIS Texture Silt loam Silty clay loam Silt loam Sandy clay loam Silty clay Silty clay Hydrometer pH 6.5-7.2 5.7-6.9 6.0-7.3 5.6-6.1 6.6-7.9 6.6-7.9 Soil-water ratio 1:2.5 Organic matter % 1.45-2.25 1.45-2.25 1.45-2.25 1.45-2.55 1.45-2.25 1.45-2.25 Wet oxidation Total N (%) 0.045-0.085 0.045-0.085 0.045-0.085 0.045-0.085 0.045-0.085 0.045-0.085 Kjeldahl Extractable P (mg /kg) 3.75-11.3 11.3-18.8 3.75-11.3 11.3-18.8 3.75-11.3 3.75-11.3 NaHCO3 extraction Exch. K ((cmol/kg) 0.085-0.225 0.085-0.225 0.045-0.085 0.045-0.085 0.085-0.225 0.085-0.225 NH4OAc extraction Extractable S (mg /kg) 5.4-16.2 5.4-16.2 5.4-16.2 5.4-16.2 37.8-48.5 37.8-48.5 CaCl2 extraction Extractable Zn ((mg /kg) 0.27-0.81 0.27-0.81 0.81-1.36 0.27-0.81 1.36-1.89 1.36-1.89 DTPA extraction Note: AEZ 1 = Old Himalayan piedmont plain, 9 = Old Brahmaputra floodplain, 11 = High Ganges River floodplain, 13 = Ganges tidal floodplain, 26 = High barind tract. Source: FAO/UNDP [15]. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 58 © 2025 by the authors; licensee Asian Online Journal Publishing Group Three types of fertilizer recommendation methods - FRG, RCM and STK, and FFP were tested in 72 farmers’ (36 farmers each year) plots during two consecutive aman rice seasons of 2018 and 2019. In Bangladesh, soil fertility experts have developed a Fertilizer Recommendation Guide (FRG) for all 30 AEZs for the farmers to select a recommended dose of fertilizers in their crops. Here fertilizer recommendation for a crop or cropping pattern is given on the basis of general fertility level of an AEZ. The Bangladesh Agricultural University (BAU), Mymensingh has developed a low-cost rapid soil testing technology i.e. Soil Testing Kit (STK) based on an on-site soil test, from which a fertilizer requirement for a crop is estimated. The Rice Crop manager (RCM) is a web-based fertilizer recommendation system developed by IRRI for site-specific nutrient management (SSNM) in the specific rice fields [8]. Before initiating crop establishment with RCM (http://webapps.irri.org/in/od/rcm), each farmer is interviewed to collect essential data inputs. The data collected includes: the farmer’s field location, field size, rice variety, anticipated age of seedlings at transplanting, water management (irrigated or rainfed), rice yield in previous years with the same or similar variety, portion of above- ground residues from the previous crop retained in the field, and the farmer’s choice of fertilizer sources. RCM utilizes this data to calculate a field-specific fertilizer recommendation tailored to achieve a target yield set by RCM. The RCM recommendation includes rates and timings for the application of fertilizer sources selected by the farmer [16]. 2.2. Treatments and Design Each method of fertilizer estimation is considered as a treatment. The experiment consisted of four treatments: T1 = Farmer fertilizer practice (FFP), T2 = FRG based fertilizer dose, T3 = RCM based fertilizer dose and T4 = STK based fertilizer dose. Urea (46% N), triple superphosphate (TSP 20% P), muriate of potash (MoP 50% K), gypsum (18% S) and zinc sulphate heptahydrate (21% Zn) were used to supply N, P, K, S, & Zn, respectively. The doses of nutrients based on the four methods including FFP are given in Table 2 and their fertilizer conversions are shown in supplementary Table S1. Each experiment was laid out in a randomized complete block design (RCBD) with six dispersed replications, i.e. each farmer/site represented one replication. So, there were 12 plots in each location for both years; farmers/sites were different for the two years. Each plot size ranged from 115 to 334 m2 (Table 3). Table 2. Estimation of N, P, K, S and Zn rates based on four different methods for five locations for use in transplanted Aman rice crops (Average of 2 years). Location (Upazila) Fertilizer rate estimation methods Nitrogen (kg/ha) Phosphorus (kg/ha) Potassium (kg/ha) Sulphur (kg/ha) Zinc (kg/ha) Durgapur, Rajshahi FFP 93.8 26.1 39.5 5.2 1.89 FRG 79.9 8.5 38.0 9.9 1.61 STK 88.7 8.0 75.5 11.2 2.14 RCM 95.4 10.3 24.0 5.1 1.56 Mean ± SD 89.4 ± 7.0 13.2 ± 8.6 44.3 ± 22.0 7.9 ± 3.1 1.8±0.28 Godagari, Rajshahi FFP 106.0 15.3 36.8 12.0 0.91 FRG 88.9 10.7 62.5 8.5 0.49 STK 88.7 3.0 53.0 13.4 4.25 RCM 117.0 12.2 29.8 7.5 2.45 Mean ± SD 100 ± 13.9 10.3 ± 5.2 45.5 ±14.9 10.3 ±2.8 2.03±1.71 Mymensingh, Sadar FFP 50.0 12.6 29.5 0.5 0.00 FRG 89.1 9.1 43.0 7.7 1.28 STK 88.7 9.2 69.3 5.6 2.14 RCM 139.3 24.6 33.8 3.4 1.31 Mean ± SD 91.7 ±36.6 13.9 ± 7.3 43.9 ± 17.8 4.3 ±2.8 1.18±1.71 Thakurgaon, Sadar FFP 99.2 18.9 48.3 14.1 3.15 FRG 86.4 8.4 39.5 8.5 1.59 STK 86.2 10.1 64.8 8.7 3.17 RCM 70.2 9.7 35.8 6.3 2.38 Mean ± SD 92.0 ±11.9 12.3 ± 4.8 45.1 ±12.9 7.9 ± 3.3 1.86±0.75 Dacope, Khulna FFP 69.1 17.3 42.3 4.2 0.00 FRG 81.7 12.7 34.3 4.3 1.28 STK 88.9 11.0 44.5 6.8 2.64 RCM 73.4 9.7 24.3 8.7 0.79 Mean ± SD 85.5 ± 8.8 11.8 ±3.3 47.1 ± 9.2 9.4 ± 2.2 2.57 ± 1.1 Amtali, Barguna FFP 84.4 20.1 12.0 3.0 0.00 FRG 72.7 8.6 29.3 3.4 1.37 STK 89.1 13.0 65.5 4.5 2.15 RCM 84.4 9.0 24.0 1.2 0.88 Mean ± SD 82.6 ± 7.0 12.7 ± 5.3 32.7 ± 23.0 3.0 ± 1.4 1.10±0.90 Note: FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. 2.3. Crop Management Rice (varieties mentioned in Table 3) was grown as a rainfed crop during July to November (kharif season) in each site and location. Thirty-day-old rice seedlings were transplanted at three seedlings per hill with a spacing of 20 × 20 cm. Urea was applied in three splits - 50% urea during final land preparation, 25% urea at 30 days (tiller stage) and 25% urea at 50 days after transplanting (panicle initiation stage). The other fertilizers viz. TSP, MoP, gypsum and zinc sulphate were added during final land preparation before transplanting rice seedlings. Pre and post emergence herbicides were used to all plots as per requirement. Insecticides ‘Brifer 5 G’ and ‘Cidial 5 G’ were used to control insect attack principally stem borer of rice. Irrigation was required in some sites before final land preparation. The crop was harvested at maturity from four randomly selected quadrats of 10m2 from each plot to record their yields. The yields were reported at 14% grain moisture. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 59 © 2025 by the authors; licensee Asian Online Journal Publishing Group 2.4. Statistical Analysis Yield data were statistically analyzed by MSTAT-C statistical software program (Michigan State University, USA) based on a randomized complete block design (RCBD) and the mean differences were compared by Duncan’s Multiple Range Test (DMRT) [17]. Correlation statistics was done to examine the relationship between the methods of fertilizer estimation (FFP, FRG, RCM and STK) and pairwise treatment comparisons were done by Tukey’s HSD (Honest Significant Difference) method. The impact of nutrients or fertilizers (N, P, K, S & Zn) on rice yield were evaluated by mixed-effects linear regression models [18]. 2.5. Economic Analysis Total variable costs were calculated considering costs for land preparation, seedlings, fertilizers, irrigation, harvesting, and labor for all operations including threshing and drying. Gross return was calculated by multiplying the amount of produce by its corresponding price at harvest. Gross margin was calculated by subtracting variable cost from gross return. Each of the treatments was evaluated based on total variable cost, gross return, gross margin and benefit-cost ratio. The labor required to complete each operation (land preparation, irrigation, and herbicide or insecticide application) in a particular treatment plot was recorded and converted to person-days /ha considering 8h as equivalent to one person-day and the daily labor wage was Tk. 400 (1US$ = 85 Tk.) per person per day (Bangladesh Government wage rate). Prices of urea, TSP, MOP, gypsum and zinc sulphate were 16, 22, 15, 10 and 180 Tk./kg; urea, TSP and MoP prices are government subsidized. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 60 © 2025 by the authors; licensee Asian Online Journal Publishing Group Table 3. Locations, rice varieties, seedling age, and transplanting and harvesting dates of crops during 2018 and 2019. Year Location No. of field sites GPS location Individual plot size (m2) Rice varieties Seedling age (days) Seedling transplanting period Harvesting date Latitude Longitude 2018 Durgapur, Rajshahi 6 24°27'27"N - 24°28'8"N 88°45'23"E - 88°45'54"E 264 BRRIdhan72 24-25 29-31 Jul 2018 9-17 Nov 2018 Godagari, Rajshahi 6 24°23'46"N - 24°24'05"N 44°26'27"E - 44°26'57"E 238 BRRIdhan51 30-35 11-18 Jul 2018 9-17 Nov 2018 Sadar, Mymensingh 6 24°40'53"N - 24°41'14"N 90°26'6"E - 90°27'0"E 213 BRRIdhan49 30-35 5-13 Aug 2018 22-27 Nov 2018 Sadar, Thakurgaon 6 25°59'59"N - 26°0'0"N 88°28'27''E - 88°30'34''E 334 BRRIdhan51 25-30 30 Jul - 8 Aug 2018 13-19 Nov 2018 Dacope, Khulna 6 22˚34'59"N - 22˚36'30"N 89˚28'2"E - 89˚30'81"E 214 BR-23 30-35 3-15 Sep 2018 12-24 Dec 2018 Amtali, Barguna 6 22˚2'14"N –22˚2'26" N 90˚14'30"E- 90˚14'34" E 227 BR-23 30-35 9-16 Sep 2018 13-19 Dec 2018 2019 Durgapur, Rajshahi 6 24˚25'36"N - 24˚25'38"N 88˚44'38"E - 88˚44'39"E 119 BRRIdhan48 24-25 26-28 Aug 2019 16-Nov 2019 Godagari, Rajshahi 6 24˚23'42"N - 24˚43'05"N 88˚26'27"E - 88˚26'47"E 179 BRRIdhan51 24-27 16-19 Jul 2019 16-18 Nov 2019 Sadar, Mymensingh 6 24°40'56"N - 24°41'49"N 90°25'07"E - 90°26'49"E 115 BRRIdhan49 29-34 1-11 Aug 2019 12-26 Nov 2019 Sadar, Thakurgaon 6 26°01'17"N - 26°03'36"N 88°28'49"E - 88°30'40"E 175 BRRIdhan51 22-35 4-28 Aug 2019 17-25 Nov 2019 Dacope, Khulna 6 22°36'20"N - 22°36'38"N 89°18'36"E - 89°30'9"E 187 BR-23 30-40 30 Aug - 18 Sep 2019 8-22 Dec, 2019 Amtali, Barguna 6 22°2'16"N –22°2'30" N 90°14'32"E-90°14'51" E 131 BR-23 28-30 7-19 Sep 2019 9-18 Dec 2019 Agriculture and Food Sciences Research, 2025, 12(1): 54-68 61 © 2025 by the authors; licensee Asian Online Journal Publishing Group 3. Results 3.1. Variations in the Estimated Fertilizer Rates The nutrient or fertilizer rates varied with different methods of estimation. The average N rate ranged from 83.1 to 96.6 kg/ha, with the highest estimation by RCM and the lowest by FRG, while for FFP and STK rates were 83.7 and 88.4 kg N/ha, respectively (Figure 2). For P, the FFP had the highest rate, 18.4 kg/ha, while STK recommended the lowest, 9.1 kg/ha. The STK estimated the highest rate of K (62.1 kg/ha) and RCM proposed the lowest (27.3 kg/ha). For S and Zn also, STK estimation was the highest followed by RCM. The Zn rate was <1 kg/ha for FFP. Hence, the FFP rate of P was double the value of STK while in case of K the FFP rate was half of the STK rate. Figure 2. Nutrient rates (kg/ha) as estimated by different methods over the locations. Note: FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, RCM=Rice crop manager, STK=Soil testing kit. RCM = Rice crop manager. 3.2. Grain Yield The two years’ average yields of rice against each fertilizer assessment method for each location and sites are presented in Figure 3. The year-wise grain yield for 2018 and 2019 are shown in Table S2 (Supplementary). The grain yield varied with different fertilizer recommendation methods, locations and sites. The STK method always gave the highest grain yield and FFP gave the lowest. Overall, RCM and FRG methods produced an identical effect on rice yield. The result trends were STK > RCM = FRG > FFP. Figure 3. Effects of different methods of fertilizer estimation on rice yield at different locations. Note: FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, RCM = Rice crop manager, STK=Soil testing kit Agriculture and Food Sciences Research, 2025, 12(1): 54-68 62 © 2025 by the authors; licensee Asian Online Journal Publishing Group Considering location variations, the grain yield recorded at Durgapur (Rajshahi district) varied from 3.96 to 4.58 t/ha, Godagari (Rajshahi) from 5.23 to 6.16 t/ha, Mymensingh Sadar from 5.07 to 5.52 t/ha, Thakurgaon Sadar from 4.32 to 4.69 t/ha, Dacope (Khulna) from 4.38 to 5.07 t/ha and Amtali (Barguna) from 4.09 to 5.57 t/ha (Table 4). The mean grain yield of rice for the six locations over the treatments were 4.25, 5.68, 5.36, 4.52, 4.67, and 4.64 t/ha, respectively. It reveals that STK demonstrated the highest yield (4.69 t/ha) and FFP exhibited the lowest (4.12 t/ha). Comparing the results for RCM and FRG, similar results were noted for Durgapur, Mymensingh and Thakurgaon locations, but the yield was significantly higher for RCM over FRG at Dacope and Amtali while the reverse was true for Godagari i.e. FRG gave higher yield over RCM. The location average yield was found to be 5.09 t/ha for STK, 4.94 t/ha for RCM, 4.88 t/ha for FRG and 4.51 t/ha for FFP; the grain yields recorded with RCM and FRG were statistically identical. Table 4. Grain yield (t/ha) of rice in different locations as influenced by different fertilizer doses determined by various methods (Results are the average of 2 years). Fertilizer rate estimation methods Durgapur Godagari Mymensingh Sadar Thakurgaon Sadar Dacope Amtali Methods average FFP 3.96 b 5.23 b 5.07 b 4.32 b 4.38 b 4.09 c 4.51 c FRG 4.29 ab 5.83 a 5.38 b 4.51 b 4.38 b 4.91 b 4.88 b RCM 4.19 b 5.49 b 5.46 ab 4.58 ab 4.86 a 5.07 a 4.94 b STK 4.58 a 6.16 a 5.52 a 4.69 a 5.07 a 4.51 a 5.09 a F-test ** ** * * ** ** * Location average 4.25 5.68 5.36 4.52 4.67 4.64 Note: * Indicate p < 0.05. ** indicate p<0.01. Within a column, the mean values followed by the same letter are not significantly different at the 0.05 level of probability by DMRT. FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. 3.3. Economics of Fertilizer Use Total variable costs include all inputs costs (fertilizers, seedlings, pesticides, irrigation, etc.) and labor costs for operations from land preparation, transplanting etc. to harvest and then processing for marketable produce. Year- wise variable costs are displayed in Table S3 (Supplementary) and the two years’ average cost is presented in Table 5 which indicates that the variable costs for STK were US$ 817 followed by 799, 797 and 792 US$ in case of RCM, FRG and FFP, respectively. Fertilizer rates and likely costs varied with the different fertilizer assessment methods. It is noted that generally the fertilizer doses or costs were lower for RCM than STK particularly in saline (Dacope and Amtali) and calcareous soils (Durgapur), but the opposite was the case for the other three locations (Barind, non-calcareous and piedmont soils). However, the costs varied among the methods of fertilizer estimation and obviously between the doses of fertilizer applied for different methods. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 63 © 2025 by the authors; licensee Asian Online Journal Publishing Group Table 5. Total fertilizer costs, gross margin and benefit-cost ratio of crop in different locations as influenced by different fertilizer doses determined by various methods (Results are the average of 2 years). Location Total fertilizer costs (US$/ha) Gross margin (US$/ha) Benefit-cost ratio FFP FRG RCM STK FFP FRG RCM STK FFP FRG RCM STK Durgapur 93 98 86 102 587 722 672 793 1.63 1.81 1.75 1.86 Godagari 112 103 107 130 734 930 830 1034 1.88 2.13 2.00 2.22 Mymensingh Sadar 52 81 97 77 867 885 841 909 2.07 2.05 1.94 2.05 Thakurgaon Sadar 117 98 65 106 739 805 828 870 2.00 2.16 2.20 2.21 Dacope 77 84 78 98 531 528 678 688 1.70 1.69 1.89 1.87 Amtali 55 79 52 108 363 502 569 397 1.51 1.69 1.78 1.52 Note: FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, RCM = Rice crop manager, STK=Soil testing kit. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 64 © 2025 by the authors; licensee Asian Online Journal Publishing Group Gross return calculated from the price of rice grain and straw against each treatment is presented for year-wise data in Table S4 (Supplementary). Like crop yield, the gross return was relatively lower in saline zones (Dacope and Amtali) that was followed by piedmont plain (Thakurgaon), and gross return for the other three locations (Dur-gapur, Godagari and Mymensingh) was relatively higher. For all locations except Amtali, the STK recorded the highest gross return (US$ 1,598) and FFP had the lowest return (US$ 1,434). The gross return in Durgapur ranged from US$ 1,517 to 1,709, Godagari US$ 1,567-1,879, Mymensingh US$ 1,676-1,772, Dacope US$ 1,290- 1,479 and in Amtali the return was US$ 1,078- 1,294 per ha (Table 5). The year-wise values of gross margin for different fertilizer recommendation tools and locations are shown in Table S5 (Supplementary) and 2-year average results are displayed in Table 5. The gross margin followed the trend: STK > RCM = FRG > FP. For locations, the gross margin followed the order: Godagari > Mymensingh > Thakurgaon > Durgapur > Dacope > Amtali indicating that gross margin in saline regions was relatively lower which can be attributed to lower yield. The gross margin in Durgapur ranged between 587 and 793 US$, Godagari between 734 and 1034 US$, Mymensingh between 841 and 909 US$, Thakurgaon between 739 and 870 US$, Dacope between 531 and 688 US$ and in Amtali it was between 363 and 569 US$ per ha. 4. Discussion We tested the performance of FRG, RCM, and STK in determining the fertilizer requirements of monsoon rice compared to FFP. Additionally, we assessed the total fertilizer costs, gross margin, and benefit-cost ratio of the crop in different locations for various fertilizer doses determined by each method. Although performance of the STK method was better than FRG, the lack of a profitable business model for operators has prevented its widespread use by farmers. Consequently, we recommend the distribution of pertinent FRG information to farmers through the implementation of FRG Cards. These cards would contain information on fertilizer requirements for each crop, providing a simple and adoptable technology to enhance farmers' decision-making regarding fertilizer usage. 4.1. Effects of Methods of Fertilizer Rate Estimation on Rice Yield The different rates of application of fertilizers produced differences in rice yield but the STK method consistently had the highest grain yield while the FFP had the lowest. For the other two methods (FRG and RCM), the yield was comparable over the locations. Thus, the grain yield followed the order of STK > RCM = FRG > FFP (Table 4). Overall, the STK yield was 0.5-1.0 t/ha higher than the FFP yield. The potential yield (same as FRG target yield) of rice varieties used in this study is 5.0 ± 0.5 t/ha. Hence, the STK method enabled rice to reach the potential yield in all locations, the RCM in all places except Durgapur, FRG in all locales except Durgapur and Dacope, and FFP in only two locations, Godagari and Mymensingh. 4.2. Relationship Between Fertilizer Estimation Methods in Terms of Rice Yield All the methods of fertilizer rate assessment showed significant and positive correlation between each other indicating that yield trend for every method over the locations was similar. It is noted that STK is most correlated with the other methods and FFP is very less correlated (Table 6). The highest correlation occurs between FRG and STK (r=0.813) because both methods are soil analysis based; the STK value is from fresh soil sample analysis while the FRG from an average of historical soil test value of a whole AEZ. The yield differences among the methods were assessed by ‘t’ statistics. It reveals that the yield increase over FFP for all other methods was statistically significant. Specifically, STK showed significance at the 0.1% level, RCM at the 1% level, and FRG at the 5% level of significance while the increases as a percentage were 12.9%, 9.6% and 8.3% higher over FFP, respectively. The average yield differences of FRG versus RCM, FRG versus STK, and RCM versus STK across the locations were not significant (Table 6). Table 6. Correlation and ‘T’ statistics to determine the relationship between fertilizer estimation methods with respect to rice yield (n=72, 6 locations, 2 years). Fertilizer rate estimation methods ‘r’ value ‘t’ value+ FFP vs FRG 0.750*** 2.64* FFP vs RCM 0.745*** 3.08** FFP vs STK 0.780*** 4.10*** FRG vs RCM 0.756*** 0.43 FRG vs STK 0.813*** 1.46 RCM vs STK 0.771*** 1.05 Note: *, P<0.05; **, P<0.01; ***, p<0.001. + Pairwise treatment comparisons by Tukey’s HSD method. 4.2. Relationship Between Nutrient Rates Determined by Different Methods and Rice Yield The rate of different nutrients (N, P, K, S & Zn) for rice production was determined by four methods. Their indi-vidual contribution to rice yield is evaluated by mixed-effects linear regression models and a t-test that uses Sat-terthwaite’s method (Table 7). All the nutrients except P had positive contribution to rice yield in the decreasing order of K> Zn> N> S: rates of P had a negative effect on yield. Statistically, the effect of P, K and Zn on rice yield was significant. Noticeably the FFP-P dose was 2-3 times higher than that of other methods’ dose in all locations except Mymensingh where RCM-P dose was 2 times higher than FFP-P dose. On the other hand, the STK-K dose was 2-3 times higher than the RCM-K dose in all places except Godagari where it was about 10 kg/ha lower compared to RCM-K. Probably the low estimation of K rates accompanied with inconsistency yield performances across the locations is a weakness of RCM method. Additional research to improve the calibration of the RCM K rates may improve its utility as a recommendation tool. Like RCM, a major weakness of FRG method is the low valuation of K dose. The next revision of the FRG may need to increase the recommendations for K as suggested by Islam, et al. [2] and Islam, et al. [19]. Furthermore, nutrient balance studies show that negative K balance is quite large and that increases in K rate by 25 to 50% can increase crop yield [19, 20]. The FFP method Agriculture and Food Sciences Research, 2025, 12(1): 54-68 65 © 2025 by the authors; licensee Asian Online Journal Publishing Group has failed to offer potential yield presumably due to use of very high dose of P, low or no application of Zn and low or high rate of S application (Table 2). Interestingly, all the N doses were similar, with only a 10% variation among the methods. Table 7. Correlation and ‘T’ statistics to determine the relationship between nutrient rates with respect to rice yield (n=24, 6 locations, 4 fertilizer methods). Fertilizer rate ‘r’ value ‘t’ value (n=24) (Satterthwaite's method) Yield vs N rate 0.253 1.49 Yield vs P rate -0.380* -3.39** Yield vs K Rate 0.242 2.32* Yield vs S rate 0.178 0.39 Yield vs Zn rate 0.181 2.14* Note: *, P<0.05; **, P<0.01. The initial soil analysis indicated consistently low to very low levels of soil exchangeable K across all study locations [13]. A study estimated the relationship between grain yield and nutrient accumulation in dry matter of irrigated rice, especially focusing on harvest index values ≥0.4. Predicted reciprocal internal efficiencies (RIEs) at 60–70% of yield potential corresponded to specific nutrient accumulations. For irrigated rice with a harvest index ≥0.4, the estimated accumulation per ton of grain yield included 14.6 kg of N, 2.7 kg of P, and 15.9 kg of K [8]. Based on that consideration, we estimate that the uptake by rice grain could range from 66 to 74 kg N /ha, 12-14 kg P /ha and 72-81 kg K/ha. It is important to note that the K uptake for a rice crop may be considerably higher, given that the concentration of K in rice straw is generally more than four times that in the grain [21]. While a portion of the K uptake is anticipated to be supported by the soil (though the exact amount remains unknown due to the absence of a K control), irrigation water, and accretion from rainfall [19] the remaining requirement must be fulfilled through the application of K fertilizer (MoP). Inadequate application of K also restricts the N uptake which lowers N use efficiency and increases the leaching risk of soluble forms of N. Most of the K in soil is chemically bound in insoluble forms and is slowly available for plant growth. This is especially true in soils that have been depleted due to continuous farming [22]. While a rice crop removes a large amount of K from soil for growth of straw and less through grain, some of the straw K is recycled to soils. Increased crop residue retention as in Conservation Agriculture practice increases the K recycling to soil and decreases the magnitude of negative K balance. Panaullah, et al. [23] reported highly negative K balance for rice-based cropping system in Bangladesh. Recently Hasan, et al. [24] reported an area of 0.287-2.43 Mha (out of 8.87 Mha arable land) have very low to low K status. 4.4. Economic Benefits for Different Methods of Fertilizer Requirement Assessment The fertilizer cost in STK method was highest in four locations - Durgapur, Godagari, Dacope and Amtali – but not in Mymensingh and Thakurgaon. For the case of FFP, fertilizer cost was the lowest in Durgapur, Mymensingh, Dacope and Barguna. Across all locations, average fertilizer cost follows the order: STK>FRG>FFP>RCM, with an overall variation of 10-20%. So, although the cost variation from different fertilizer management strategies methods was not large, due to more effective balances of nutrients supplied, the impact on rice yield was substantial. Hence, the highest BCR was for STK (1.96) and the lowest for FFP (1.8), while BCR for both RCM and FRG was 1.92 (Table 5). The consistent benefit from the STK method of fertilizer requirement assessment was due to a combination of reasonable fertilizer costs with higher income from output sales. Improved economic performance of any practice is an important factor for farmer adoption [25]. Hence based on profitability, the STK method of fertilizer requirement assessment was expected to be most attractive to farmers but other factors as discussed below mitigated against its acceptance by farmers. 4.5. Adoption of Fertilizer Recommendation Information by Farmers To gain widespread adoption, a fertilizer recommendation system must effectively tackle the dual challenges of optimizing recommendations for both yield and profit, while also ensuring acceptance among farmers for various crops and cropping patterns. The endorsement of such a system by policies is crucial for its success. In Bangladesh, and in many other nations, the transfer of crop technologies primarily occurs through government-funded ag- ricultural extension agencies. Despite this established pathway, , a substantial percentage (60-85%) of Bangladeshi farmers do not adhere to fertilizer recommendations provided by the Fertilizer Recommendation Guide (FRG) [2]. Consequently, the current extension system appears to be falling short in effectively disseminating fertilizer management technology to farmers. Furthermore, while STK method demonstrated superior performance compared to both FRG and RCM, practical challenges hinder its widespread implementation. Operating the soil test involves continual procurement of chemicals and demands skilled personnel for handling. Unfortunately, establishing a profitable business model for making the STK widely accessible to smallholder farmers remains elusive. Inconsistent results of the current RCM method discourage its promotion to farmers. As discussed above, further research to improve its calibration for N, P, K, S and Zn fertilizer recommendations may improve its utility and adoptability. Hence, we explored the FRG as an alternative, effective and adoptable method of disseminating information. To address the issue, we developed FRG Cards taking information from the latest Fertilizer Recommendation Guide [13] and tested them with farmers. Across the ten project hubs, 33,237 (including 1,336 women farmers) farmers received FRG Cards followed by a short training session (about 20 mins each) for use in cropping on 21,160 ha of arable land. An evaluation was done to assess the effectiveness of the FRG card in disseminating FRG information to the farmers [26]. The key findings were as follows: (i) 100% of respondents were familiar with FRG Card, but less than 50% with STK and RCM, (ii) about 95% of the FRG Card holders followed the recommendation of Bangladesh Agricultural Research Council (BARC) [13] in T. Aman rice due to user-friendliness, and (iii) a small percentage of farmers used Soil Testing Kit (STK). Agriculture and Food Sciences Research, 2025, 12(1): 54-68 66 © 2025 by the authors; licensee Asian Online Journal Publishing Group Hence, FRG Card when provided with training to the farmers was accessible, while applying the FRG information in their crops increased aman rice grain yield by 84% and net profit increased by 143% over FFP. The next step is to scale-out the use of the FRG Card by larger numbers of farmers. This will require lower cost methods of distributing the FRG Card and training farmers in its use including mass media and social media. 5. Conclusion The study indicates that the STK (Soil Testing Kit developed by Bangladesh Agricultural University, Mymensingh) had the best performance with regard to rice yield and gross margin. Next to it was the Fertilizer Recommendation Guide (FRG) based on agroecological zone-specific recommendation. In contrast, the Rice Crop Manager (RCM) did not consistently yield reliable results. A notable weakness observed across all three methods was the consistent underestimation of potassium (K) requirements, except for the STK method, which directly assessed the current soil K availability. This underestimation of K needs is a common issue in subtropical Asia, where rice cropping systems deplete substantial amounts of K annually, leading to an increasing demand for K fertilizer over time. Although performance of the STK method was better than FRG, the lack of a profitable business model for operators has prevented its widespread use by farmers. Previous barriers to supplying relevant FRG information to farmers can be overcome by providing necessary packaged information on fertilizer requirements for each crop to farmers in the form of a FRG Card. The FRG Card is a simple and adoptable technology for improved fertilizer decision-making by the farmers. Abbreviations The following abbreviations are used in this manuscript: % Percentage AEZ Agro-ecological zones BAU Bangladesh agricultural university cmol Centimole DMRT Duncan’s multiple range test FAO Food and Agriculture Organization of United Nations FFP Farmer fertilizer practice FRG Fertilizer recommendation guide GPS Geo-graphical Positioning System ha Hectarage K Potassium kg Kilogram mg Milligram Mha Million hectarage MoP Muriate of Potash N Urea NUMAN Nutrient Management for Diversified Cropping in Bangladesh P Phosphorus RCBD Randomized complete block design RCM Rice crop manager S Sulphur (gypsum) SSNM Site-specific nutrient management STK Soil testing kit t Tonne t/ha Ton per hectarage Tk. Taka TSP Triple superphosphate UNDP United nations development program US$ US dollar Zn Zinc sulphate fertilizer References [1] M. Nasim et al., "Distribution of crops and cropping patterns in Bangladesh," Bangladesh rice journal, vol. 21, no. 2, pp. 1-55, 2017. [2] M. S. Islam, R. W. Bell, M. M. Miah, and M. J. Alam, "Unbalanced fertilizer use in the Eastern Gangetic Plain: The influence of Government recommendations, fertilizer type, farm size and cropping patterns," PLoS One, vol. 17, no. 7, p. e0272146, 2022. [3] M. A. Kader et al., "Long-term conservation agriculture increases nitrogen use efficiency by crops, land equivalent ratio and soil carbon stock in a subtropical rice-based cropping system," Field Crops Research, vol. 287, p. 108636, 2022. [4] S. Uddin et al., "Ammonia fluxes and emission factors under an intensively managed wetland rice ecosystem," Environmental Science: Processes & Impacts, vol. 23, no. 1, pp. 132-143, 2021. [5] M. Jahangir et al., "Reduced tillage with residue retention and nitrogen application rate increase N2O fluxes from irrigated wheat in a subtropical floodplain soil," Agriculture, Ecosystems & Environment, vol. 306, p. 107194, 2021. [6] S. Snapp et al., "Scientists urge shifting more nitrogen to low-input farms and better use on high-yield farms. Environmental Health and Biodiversity. CIMMYT," 2023. Retrieved: https://www.cimmyt.org/news/scientists-urge-shifting-more-nitrogen-to- low-input-farms-and-better-use-on-high-yield-farms/. 2023. [7] H. M. Beneduzzi, E. G. d. Souza, W. K. Moreira, R. Sobjak, C. L. Bazzi, and M. Rodrigues, "Fertilizer recommendation methods for precision agriculture–a systematic literature study," Engenharia Agrícola, vol. 42, p. e20210185, 2022. [8] R. J. Buresh, M. F. Pampolino, and C. Witt, "Field-specific potassium and phosphorus balances and fertilizer requirements for irrigated rice-based cropping systems," Plant and Soil, vol. 335, pp. 35-64, 2010. [9] R. J. Buresh and C. Witt, "Site-specific nutrient management," in Proceedings of the IFA International Workshop on Best Management Practices, 7–9 March 2007, Brussels, Belgium. International Fertilizer Industry Association, Paris, pp 47–55, 2007. [10] A. Dobermann and K. Cassman, "Plant nutrient management for enhanced productivity in intensive grain production systems of the United States and Asia," Plant and Soil, Vol. 247, pp. 153-175, 2002. [11] A. K. Shukla et al., "Spatial distribution and management zones for sulphur and micronutrients in Shiwalik Himalayan region of India," Land Degradation & Development, vol. 28, no. 3, pp. 959-969, 2017. https://www.cimmyt.org/news/scientists-urge-shifting-more-nitrogen-to-low-input-farms-and-better-use-on-high-yield-farms/ https://www.cimmyt.org/news/scientists-urge-shifting-more-nitrogen-to-low-input-farms-and-better-use-on-high-yield-farms/ Agriculture and Food Sciences Research, 2025, 12(1): 54-68 67 © 2025 by the authors; licensee Asian Online Journal Publishing Group [12] B. van Raij, H. Cantarella, and J. Quaggio, "Rationale of the economy of soil testing," Communications in Soil Science and Plant Analysis, vol. 33, no. 15-18, pp. 2521-2536, 2002. [13] Bangladesh Agricultural Research Council (BARC), Fertilizer recommendation guide (FRG-2018). Farmgate, Dhaka: BARC, 2018. [14] A. L. Page, R. H. Miller, and D. R. Keeney, Methods of soil analysis, 2nd ed. Madison, WI. USA: American Society of Agronomy, 1982. [15] FAO/UNDP, Land resources appraisal of Bangladesh for agricultural development: Agroecological regions of Bangladesh. Rome: Food and Agriculture Organization of the United Nations, 1988. [16] S. Sharma et al., "Field-specific nutrient management using Rice Crop Manager decision support tool in Odisha, India," Field Crops Research, vol. 241, p. 107578, 2019. [17] K. A. Gomez and A. A. Gomez, Statistical procedures for agricultural research. New York: John Wiley & Sons, 1984. [18] D. Bates, M. Mächler, B. M. Bolker, and S. C. Walker, "Fitting linear mixed-effects models using lme4," Journal of Statistical Software, vol. 67, no. 1, pp. 1-48, 2014. [19] M. Islam et al., "Conservation agriculture in intensive rice cropping reverses soil potassium depletion," Nutrient Cycling in Agroecosystems, vol. 125, no. 3, pp. 437-451, 2023. [20] M. Islam et al., "Conservation agriculture improves yield and potassium balance in intensive rice systems," Nutrient Cycling in Agroecosystems, vol. 128, no. 2, pp. 233-250, 2024. [21] P. Csathó, T. Arendas, N. Fodor, and T. Németh, "Evaluation of different fertilizer recommendation systems on various soils and crops in Hungary," Communications in Soil Science and Plant Analysis, vol. 40, no. 11-12, pp. 1689-1711, 2009. [22] C. Srinivasarao et al., "Soil potassium fertility and management strategies in South Asian agriculture," Advances in Agronomy, vol. 177, pp. 51-124, 2023. [23] G. Panaullah et al., "Nutrient uptake and apparent balances for rice-wheat sequences. III. Potassium," Journal of Plant Nutrition, vol. 29, no. 1, pp. 173-187, 2006. [24] M. N. Hasan, M. A. Bari, and M. R. Lutfar, Soil fertility trends in Bangladesh 2010 to 2020 (SRSRF Project). Dhaka, Bangladesh: Soil Resource Development Institute, Ministry of Agriculture, 2020. [25] M. Haque, R. Bell, M. Islam, and M. Rahman, "Minimum tillage unpuddled transplanting: An alternative crop establishment strategy for rice in conservation agriculture cropping systems," Field Crops Research, vol. 185, pp. 31-39, 2016. [26] M. W. Rahman, Evaluative research survey report. Dhaka, Bangladesh: NUMAN Project, 2022. Table S1. Estimation of rates of urea, TSP, MoP, gypsum and zinc sulphate fertilizers based on four different methods and locations for use in T. Aman rice cultivation in 2018 and 2019. Location Treatment Fertilizer dose (kg/ha) Urea TSP MoP Gypsum Zinc sulphate 2018 2019 2018 2019 2018 2019 2018 2019 2018 2019 Durgapur, Rajshahi FFR 76 341 151 110 76 82 0 58 3.3 7.5 FRG 197 158 50 35 71 81 76 34 2.7 6.5 STK 197 197 40 40 151 151 62 62 0.0 12.2 RCM 180 244 44 59 39 57 19 38 1.4 7.5 Godagari, Rajshahi FFR 221 250 72 81 72 75 83 50 4.0 1.2 FRG 199 196 57 50 100 150 50 44 2.8 0.0 STK 199 195 15 15 138 74 75 74 12.2 12.1 RCM 257 263 56 66 60 59 42 41 8.4 5.6 Sadar, Mymensingh FFR 97 125 57 69 38 80 0 6 0.0 0.0 FRG 198 198 51 40 71 101 64 22 2.8 4.5 STK 197 197 40 52 151 126 0 62 0.0 12.2 RCM 354 265 178 68 70 65 0 38 0.0 7.5 Sadar, Thakurgaon FFR 214 227 71 118 71 122 0 157 7.1 10.9 FRG 186 198 48 36 57 101 60 34 2.6 6.5 STK 186 197 37 64 104 155 47 50 11.6 6.5 RCM 157 155 63 34 98 45 36 34 7.1 6.5 Dacope, Khulna FFR 132 175 89 84 85 84 47 0 0.0 0.0 FRG 165 198 51 76 36 101 25 23 2.8 4.5 STK 198 197 40 70 78 100 25 50 2.8 12.3 RCM 163 163 43 54 43 54 43 54 0.0 4.5 Amtali, Barguna FFR 250 125 125 76 0 0 0 0 0.0 0.0 FRG 165 158 51 35 37 80 25 13 2.8 5.0 STK 198 198 70 60 151 111 0 50 0.0 12.3 RCM 250 125 62 28 36 30 0 13 0.0 5.0 Note: FFR=Farmer’s fertilizer rate, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM=Rice crop manager MoP=Muriate of potash, TSP = Triple superphosphate. Table S2. Grain yield (t/ha) of T. Aman rice in different locations and years as influenced by different fertilizer doses determined by various methods in 2018 and 2019. Fertilizer tools Durgapur, Rajshahi Godagari, Rajshahi Mymensingh, Sadar Thakurgaon Sadar Dacope, Khulna Amtali, Barguna 2018 FFP 3.76b 4.95b 4.56 3.55 3.48b 3.47b FRG 3.98ab 5.51a 4.68 3.6 3.57b 3.79ab STK 4.14a 5.75a 4.59 3.66 3.99a 3.67b RCM 3.95ab 5.29ab 4.38 3.81 4.19a 4.23a F-test ** ** NS NS ** ** 2019 FFP 3.50b 4.55b 4.86b 4.3 4.48b 3.99c FRG 3.81ab 5.09ab 5.1b 4.61 4.39b 5.14a STK 4.19a 5.46a 5.45a 4.87 5.24a 4.52b RCM 3.68b 4.69b 5.56a 4.52 4.65b 5.00a F-test ** ** ** NS ** ** Note: ** indicate p<0.01; NS = Not significant. Within a column, the mean values followed by the same letter are not significantly different at the 0.05 level of prob ability by DMRT. FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. Agriculture and Food Sciences Research, 2025, 12(1): 54-68 68 © 2025 by the authors; licensee Asian Online Journal Publishing Group Table S3. Variable costs (USD/ha) of T. Aman rice in different locations and years as influenced by different fertilizer doses determined by various methods in 2018 and 2019 Fertilizer tools Durgapur, Rajshahi Godagari, Rajshahi Mymensingh, Sadar Thakurgaon Sadar Dacope, Khulna Amtali, Barguna 2018 FFP 785 833 778 710 762 786 FRG 791 815 815 709 749 763 STK 795 840 810 731 761 789 RCM 767 829 891 717 748 773 2019 FFP 1074 833 840 767 756 643 FRG 1001 830 869 681 780 697 STK 1,38 851 917 704 821 743 RCM 1025 835 897 661 774 676 Note: NS = Not significant. Within a column, the mean values followed by the same letter are not significantly different at the 0.05 level of probability by DMRT. FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. Table S4. Gross return (USD/ha) of T. Aman rice in different locations and years as influenced by different fertilizer doses determined by various methods in 2018 and 2019 Fertilizer tools Durgapur, Rajshahi Godagari, Rajshahi Mymensingh, Sadar Thakurgaon, Sadar Dacope, Khulna Amtali, Barguna 2018 FFP 1.748 2.120 1.989 1.822 1.409 1.317 FRG 1.841 2.374 2.015 1.807 1.437 1.410 STK 1.897 2.545 2.001 1.902 1.601 1.384 . 1,799 2.269 1.933 1.865 1.668 1.559 2019 FFP 1.285 1.014 1.363 1.133 1.171 838 FRG 1.395 1.131 1.439 1.193 1.148 1.054 STK 1.521 1.213 1.543 1.273 1.356 942 RCM 1.337 1.054 1.537 1.169 1.209 1.028 Note: NS = Not significant. Within a column, the mean values followed by the same letter are not significantly different at the 0.05 level of probability by DMRT. FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. Table S5. Gross margin (USD/ha) of T. Aman rice in different locations and years as influenced by different fertilizer doses determined by various methods in 2018 and 2019 Fertilizer tools Durgapur, Rajshahi Godagari, Rajshahi Mymensingh Sadar Thakurgaon Sadar Dacope, Khulna Amtali, Barguna 2018 FFP 963 1287 1211 1112 647 531 FRG 1050 1559 1200 1098 688 647 STK 1102 1705 1191 1171 840 595 RCM 1032 1440 1042 1148 920 786 2019 FFP 211 181 523 366 415 195 FRG 394 301 570 512 368 357 STK 483 362 626 569 535 199 RCM 312 219 640 508 435 352 Note: NS = Not significant. Within a column, the mean values followed by the same letter are not significantly different at the 0.05 level of probability by DMRT. FFP = Farmer fertilizer practice, FRG=Fertilizer recommendation guide, STK=Soil testing kit, RCM = Rice crop manager. Asian Online Journal Publishing Group is not responsible or answerable for any loss, damage or liability, etc. caused in relation to/arising out of the use of the content. Any queries should be directed to the corresponding author of the article.