Nitrogen response of two high yielding rice varieties as influenced by nitrogen levels and growing seasons

ABSTRACT

Proper nitrogen (N) management of high yielding rice varieties is crucial for obtaining their potential yield benefit. In this study, we compared the effect of different N rates on rice growth, yield, and N use efficiency of two high yielding rice varieties (BRRI dhan58 and BRRI dhan75) in a Boro (dry season)-fallow-T. Aman (wet season) cropping pattern. The N rates were: 0, 25, 50, 75, 100, 125, 150 and 175 kg ha⁻¹ which were assigned in randomized complete block design with three replications. In Boro season, BRRI dhan58 showed higher response to applied N compared to BRRI dhan75 in T. Aman season. The calculated optimum N dose to maximize yield of BRRI dhan58 and BRRI dhan75 were 142 kg ha⁻¹ and 82 kg ha⁻¹, respectively. The agronomic N use efficiency and fertilizer N recovery efficiency were comparatively higher in BRRI dhan58 grown in Boro season than BRRI dhan75 grown in T. Aman season, while in both seasons, the N use efficiencies decreased with increased N rates. This study results suggest that variety-specific N fertilization with respect to growing season is the best N management practice to maximize rice yield avoiding the excess use of N fertilizer.

KEYWORDS:

• Nitrogen uptake
• agronomic nitrogen use efficiency
• fertilizer nitrogen recovery efficiency
• grain yield
• straw yield

1. Introduction

Rice is one of the most important cereal crops of the world which feeds half of the world’s population providing 35-60% of the total calorie (Tayefe et al., 2014). Rice is grown in a wide range of climatic conditions covering one-third of the world’s total-cropped area. In Bangladesh rice is the staple food of about 17 million people which covers about 75% of the total-cropped area (Nasim et al., 2017). The major rice-growing seasons in Bangladesh are Aus, Aman, and Boro of which the transplanted Aman (T. Aman) and Boro in combined occupy more than 80% of the total rice-cultivated land. In recent years, rice farmers in Bangladesh are extensively cultivating modern high yielding rice varieties. Therefore, nutrient management of the high yielding rice varieties is crucial to obtain the desired yield.

Among the major nutrient elements, nitrogen (N) is the most limiting nutrient for rice crop growth and yield which is required in higher amounts compared to other nutrients (Djaman et al., 2018). N influences rice yield by playing major role in the photosynthesis, biomass accumulation, effective tillering, and spikelets formation (Yoshida et al., 2006).

Most of the agricultural soils in Bangladesh are deficient in N (P. K. Saha et al., 2012; Sarkar et al., 2016). Therefore, N fertilization is imperative for modern rice varieties in order to exploit their full yield potential (Chamely et al. 2015). High yielding modern rice varieties show a greater response to applied nitrogen, while they differ in N demand depending on their genotype and agronomic traits under different climatic conditions (Rahman et al., 2007). On the other hand, excessive N application can lead to ground water pollution, increased production cost, reduced yield and environmental pollution (Djaman et al., 2018). Therefore, variety-specific N fertilizer recommendation could be an effective option for better N management. Considering these, the present study was carried out to study the effect of different N rates on two high yielding rice varieties of Bangladesh. The objectives of the investigation were to determine the optimum N rates of two newly released high yielding rice varieties and estimate their N use efficiency.

2. Materials and methods

2.1. Experiment location

The experiment was conducted at the experimental farm of Bangladesh Rice Research Institute in Boro (dry) and T. Aman (wet) seasons in 2017 and between these two seasons there was a fallow period. The climate of the experimental location is subtropical in nature and experiences periodic southwestern monsoon with an average annual rainfall of 2000 mm. The 80% of the rainfall occurs from mid-June to end of September. The lowest mean temperature (15°C) prevails in January and highest (30°C) in May. The soil of the experimental site belongs to the order Inceptisols in USDA soil classification system. Before the initiation of the experiment, soil samples from the top layer (20 cm) were collected and analyzed. The soil was silty clay loam in texture having pH 6.8 (1:2.5 soil: water), organic carbon 12 g kg⁻¹ (wet oxidation method), total Nitrogen 1.1 g kg⁻¹ (Kjeldahl digestion method), Olsen available phosphorus 8.2 mg kg⁻¹ (0.5 M NaHCO₃ extraction method), exchangeable potassium 39 mg kg⁻¹ (1 N ammonium acetate extraction method) and available sulfur 19 mg kg⁻¹ [0.16 M Ca(H₂PO₄) extraction method].

2.2. Experimental design and treatments

Eight rates (0, 25, 50, 75, 100, 125, 150, and 175 kg ha⁻¹) of N were tested on two high yielding rice varieties (BRRI dhan58 and BRRI dhan75) in a double rice cropping system (Boro-Fallow-T. Aman). BRRI dhan58 was grown in Boro season, while BRRI dhan75 was grown in T. Aman season. The N levels were assigned following randomized complete block design. The N was applied as urea in three equal splits: the first one at the time of final land preparation, the second at maximum tillering stage and the third at the time of panicle initiation. All the treatments received an equal amount of phosphorus (P), potassium (K), and sulfur (S) fertilizers and applied as basal. In Boro season the rate of P-K-S was 20-65-10 kg ha⁻¹, while in T. Aman season it was 10-50-10 kg ha^−1. The unit plot size was 6 m × 7 m. In Boro season, 35-day-old seedlings of BRRI dhan58 were transplanted in the first week of January and harvested in last week of May. In T. Aman season, 25-day-old seedlings of BRRI dhan75 were transplanted in the first week of August and harvested in the last week of November.

2.3. Data and plant sample collection

Plant height was recorded at maturity form 3 hills per plot from the soil surface to the tip of the tallest panicle of each hill. Number of tillers and panicles per m² were counted from 16 hills in each plot. Number of filled and unfilled grains per panicle were counted from five panicles in each plot. Panicle length (cm) was measured from the panicle neck to the apex of the panicle from five panicles. Grain yield was calculated from a harvest area of 5 m² in each plot and adjusted to 14% moisture content. Straw yield was calculated as oven-dry basis from randomly collected 16 hills. The grain and straw samples were analyzed for their N content by micro-Kjeldahl method (Nelson & Sommers, 1973) and the crop N uptake was calculated from dry biomass (grain + straw) weight and N concentrations (Sarkar et al., 2016).

2.4. N use efficiency

The agronomic N use efficiency (AE_N) and fertilizer N recovery efficiency (RE_N) was calculated for different N rates using the following equation as described by Ahmed et al. (2016):

$\mathrm{A}{\mathrm{E}}_{\mathrm{N}}=\frac{\begin{array}{c}\left(\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{N}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{l}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\right.\\ \left.-\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{y}\mathrm{i}\mathrm{e}\mathrm{l}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{N}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\right)\mathrm{k}\mathrm{g}\mathrm{h}{\mathrm{a}}^{-1}\end{array}}{\mathrm{A}\mathrm{p}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{d}\mathrm{N}\mathrm{i}\mathrm{n}\mathrm{N}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\left(\mathrm{k}\mathrm{g}\mathrm{h}{\mathrm{a}}^{-1}\right)}$

$\mathrm{R}{\mathrm{E}}_{\mathrm{N}}=\frac{\begin{array}{c}\left(\mathrm{N}\mathrm{u}\mathrm{p}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{N}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{l}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\right.\\ \left.-\mathrm{N}\mathrm{u}\mathrm{p}\mathrm{t}\mathrm{a}\mathrm{k}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{N}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\right)\mathrm{k}\mathrm{g}\mathrm{h}{\mathrm{a}}^{-1}\end{array}}{\mathrm{A}\mathrm{p}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{d}\mathrm{N}\mathrm{i}\mathrm{n}\mathrm{N}\mathrm{f}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}\left(\mathrm{k}\mathrm{g}\mathrm{h}{\mathrm{a}}^{-1}\right)}$

2.5. Statistical analysis

The data were analyzed using analysis of variance. The mean comparison among the treatments was done following Tukey’s Honest Significance Difference Test (HSD) at 5% level of significance. The analyses were carried out using the STAR (Statistical Tool for Agricultural Research) software version 2.0.1.

3. Results

3.1. Plant height

In Boro season, plant height of BRRI dhan58 significantly increased with increased N rates and the highest plant height (98.20 cm) was observed with 175 kg N ha⁻¹ which was statistically identical to 100, 125, and 150 kg N ha⁻¹ (Table 1). The plant height response of BRRI dhan58 to N rates was quadratic (Figure 1). In T. Aman season, the plant height of BRRI dhan75 did not differ significantly with N rates (Table 1).

Table 1. Effect of nitrogen rates on plant height, tiller and panicle of Boro and T. Aman rice

	Plant height (cm)		Tiller m^−2		Panicle m^−2
N rates	Boro (BRRI dhan58)	T. Aman (BRRI dha75)	Boro (BRRI dhan58)	T. Aman (BRRI dhan75)	Boro (BRRI dhan58)	T. Aman (BRRI dhan75)
N0	70.13 c	110.67 a	167 f	244 d	157 e	229 c
N25	75.4 c	113.73 a	200 e	255 cd	189 d	239 bc
N50	83.27 b	113.07 a	230 d	270bcd	223 c	256 abc
N75	85.47 b	116.53 a	268 c	292abc	251 b	284 a
N100	92.8 a	115.13 a	268 c	295 ab	258 b	295 a
N125	93.73 a	116.00 a	317 b	297 ab	300 a	280 ab
N150	93.93 a	112.93 a	333 ab	301 ab	312 a	277 ab
N175	98.20 a	116.00 a	347 a	308 a	302 a	277 ab
CV (%)	2.62	2.42	3.44	4.57	2.91	5.85

Means followed by the same lowercase letter in a column do not differ from each other by Tukey’s HSD test at 5% level of significance.

Figure 1. Plant height response to N rates of Boro rice (BRRI dhan58).

3.2. Tiller number

Tiller m⁻² of BRRI dhan58 and BRRI dhan75 in Boro and T. Aman seasons, respectively, were significantly influenced by N rates and increased linearly with the increase of N rates (Table 1 and Figure 2(a)). In case of BRRI dhan58, the highest tiller m⁻² was observed with N₁₇₅ which was similar to N₁₅₀. In case of BRRI dhan75, the highest tiller m⁻² (308) was found with the highest N rate (N₁₇₅) which was similar to N₇₅, N₁₀₀, N₁₂₅, and N₁₅₀ (Table 1).

Figure 2. Number of tillers m⁻² (a) and number of panicle m⁻² (b) response to N rates of Boro and T. Aman rice.

3.3. Panicle number and panicle length

In both Boro and T. Aman seasons, panicle number m⁻² increased significantly with the increase of N rates (Table 1) and the regression analysis showed a quadratic response of panicle number m⁻² to N rates (Figure 2(b)). In Boro season, the highest number (312) of panicle m⁻² in BRRI dhan58 was obtained with 150 kg N ha⁻¹ which was statistically similar to 125 and 175 kg N ha⁻¹. In T. Aman season, the highest number (295) of panicle m⁻² was found in 100 kg N ha⁻¹ which was statistically similar with 50, 75, 125,150, and 175 kg N ha⁻¹ (Table 1).

In both Boro and T. Aman seasons, panicle length was significantly affected by N rates. In Boro season, the highest panicle length (22.87 cm) was observed with 150 kg N ha⁻¹ which was statistically similar to other N rates except N₀ and N₅₀. In T. Aman season, the highest panicle length (24.07 cm) of BRRI dhan75 was recorded with 75 kg N ha⁻¹ which was statistically similar with other N rates except N₀ and N₁₇₅ (Table 2).

Table 2. Effect of nitrogen rates on panicle length, filled grain and unfilled grain of Boro and T. Aman rice

	Panicle length (cm)		Filled grain panicle^−1		Unfilled grain panicle^−1
N rates	Boro (BRRI dhan58)	T. Aman (BRRI dha75)	Boro (BRRI dhan58)	T. Aman (BRRI dha75)	Boro (BRRI dhan58)	T. Aman (BRRI dha75)
N0	20.45 c	20.63 c	110 c	117 a	9.10 d	32.30 c
N25	21.68 abc	22.27 abc	124bc	128 a	8.87 d	35.63 bc
N50	21.49 bc	22.53 abc	136 b	134 a	11.97 cd	34.87 bc
N75	22.72 ab	24.07 a	142 ab	138 a	19.10 c	36.07 abc
N100	21.79 ab	23.12 ab	147 ab	132 a	30.87 b	39.47 abc
N125	21.78 ab	23.54 ab	164 a	124 a	38.43 b	42.94 abc
N150	22.87 a	22.30 abc	145 ab	125 a	52.77 a	51.31 ab
N175	22.23 ab	22.03 bc	123bc	118 a	53.27 a	53.073 a
CV (%)	2.09	3.05	6.67	9.19	11.04	14.82

Means followed by the same lowercase letter in a column do not differ from each other by Tukey’s HSD test at 5% level of significance.

3.4. Filled and unfilled grain

In Boro season, filled grain panicle⁻¹ significantly increased with increased N rates up to 150 kg N ha⁻¹ than it significantly reduced at 175 kg N ha⁻¹. The highest number of filled grain panicle⁻¹ in BRRI dhan58 was found with 125 kg N ha⁻¹ which was statistically similar to 75, 100 and 150 kg N ha⁻¹. However, in T. Aman season, BRRI dhan75 did not respond significantly to N rates in case of filled grain panicle⁻¹ (Table 2). Both in Boro and T. Aman seasons, the unfilled grain panicle⁻¹ of BRRI dhan58 and BRRI dhan75, respectively, significantly increased with the increase of N rates and the lowest number of unfilled grain panicle⁻¹ was found in N₀ treatment (Table 2).

3.5. Grain and straw yield

Grain yield is determined by the combined action of different yield contributing factors like number of panicle, panicle length, filled grain, and 1000 grain weight (B. Saha. et al., 2017). Results of this study showed remarkable influence of different N rates on rice grain yield. In Boro season, grain yield of BRRI dhan58 with different N rates increased by 55-348%. The highest grain yield (7.30 t ha⁻¹) of BRRI dhan58 was obtained with 125 kg N ha⁻¹ which was 348% higher than N₀. In T. Aman season, the grain yield response of BRRI dhan75 to N fertilization at varying rates was comparatively lower than BRRI dhan58 in Boro season. The highest yield of BRRI dhan75 was obtained with 75 kg N ha⁻¹ which was 18% higher than N₀. However, this increase was similar to application of 50, 100 and 125 kg N ha⁻¹ (Table 3). BRRI dhan58 and BRRI dhan75 in Boro and T. Aman, respectively, showed quadratic relation between grain yield and applied N (Figure 3). The calculated optimum nitrogen doses obtained from differentiating the quadratic equations were 142 and 82 kg ha⁻¹ for BRRI dhan58 and BRRI dhan75, respectively.

Table 3. Effect of nitrogen rates on rice yield of Boro and T. Aman rice

	Boro (BRRI dhan58)			T. Aman (BRRI dhan75)
N rates	Grain yield (t ha^−1)	Yield increase over control (%)	Straw yield (t ha^−1)	Grain yield (t ha^−1)	Yield increase over control (%)	Straw yield (t ha^−1)
N0	1.63 g		1.49 f	4.10 d		5.98
N25	2.53 f	55.21	1.96 f	4.28 cd	4.39	6.17
N50	4.51 e	176.69	3.55 e	4.65 ab	13.41	5.59
N75	5.01 d	207.36	4.45 d	4.82 a	17.56	5.32
N100	6.39 bc	292.02	5.33 c	4.75 a	15.85	6.44
N125	7.30 a	347.85	6.65 b	4.65 ab	13.41	5.52
N150	6.71 b	311.66	7.54 a	4.42 bc	7.80	5.69
N175	6.16 c	277.91	7.44 a	4.20cd	2.44	6.30
CV (%)	3.32		4.35	3.95		14.59

Means followed by the same lowercase letter in a column do not differ from each other by Tukey’s HSD test at 5% level of significance.

Figure 3. Response behavior of rice grain yield to applied N in Boro and T. Aman seasons.

The straw yields in Boro and T. Aman seasons are presented in Table 3. The straw yield of BRRI dhan58 in Boro season significantly responded to the applied N rates. The straw yield of BRRI 58 increased with the increase of N rates and was highest (7.54 kg ha⁻¹) with N₁₅₀ which was statistically identical to N₁₇₅. However, in T. Aman season the straw yield of BRRI dhan75 did not differ significantly with applied N rates.

3.6. Nitrogen uptake

The applied N rates significantly affected nutrient uptake of BRRI dhan58 and BRRI dhan75 in Boro and T. Aman seasons, respectively (Table 4). In Boro season, the grain N uptake by BRRI dhan58 significantly increased with N rates up to 125 kg N ha⁻¹. The straw N uptake by BRRI dhan58 increased with the increase of N rates. The total N uptake by BRRI dhan58 ranged between 24.52 and 138.58 kg ha⁻¹. The highest total N uptake (138.58 kg ha⁻¹) by BRRI dhan58 in Boro season was observed with 125 kg N ha⁻¹ which was similar to N₁₅₀. In T. Aman season, the grain and straw N uptake of BRRI dhan75 varied from 42.30 to 64.60 kg ha⁻¹ and 24.45–50.05 kg ha⁻¹, respectively. The highest grain and straw N uptakes by BRRI dhan75 were found with 75 and 100 kg N ha⁻¹, respectively. The total N uptake by BRRI dhan75 varied from 75.53 to 114.66 kg ha⁻¹ and the highest was with 100 kg N ha⁻¹.

Table 4. Effect of nitrogen rates on nitrogen uptake in Boro and T. Aman seasons

N rates	Boro (BRRI dhan58)			T. Aman (BRRI dhan75)
	Grain	Straw	Total	Grain	Straw	Total
	kg ha^−1
N0	16.64 f	7.87 e	24.51 f	42.3 d	25.45c	70.53 e
N25	29.07 e	11.76 e	40.83 f	47.15 cd	35.74 bc	82.88 d
N50	51.73 d	20.47 d	72.21 e	55.01 bc	38.17 abc	93.18 cd
N75	63.06 c	26.15 cd	89.21 d	64.87 a	37.92 abc	102.79 bc
N100	80.94 b	30.64 c	111.58 c	64.60 a	50.05 a	114.66 a
N125	99.19 a	39.39 b	138.58 a	63.68 ab	46.4 ab	110.08 ab
N150	85.58 b	46.95 a	132.53 ab	61.09ab	44.26 ab	105.35 ab
N175	78.94 b	49.76 a	128.71 b	59.37 ab	46.39 ab	105.76 ab
CV (%)	4.66	7.43	2.73	5.62	11.08	4.08

Means followed by the same lowercase letter in a column do not differ from each other by Tukey’s HSD test at 5% level of significance.

3.7. Nitrogen use efficiency

In this study, N rates significantly influenced nitrogen use efficiency of BRRI dhan58 and BRRI dhan75 in Boro and T. Aman seasons, respectively. However, in T. Aman season the nitrogen use efficiency of BRRI dhan75 was much lower compared to BRRI dhan58 in Boro season (Table 5). Agronomic N use efficiency (AE_N) indicates the yield increasing effect of the applied N (Dobermann, 2005). In Boro season, the AE_N of BRRI dhan58 varied between 26 and 58 kg kg⁻¹, while in T. Aman season, the AE_N of BRRI dhan75 ranged between 0.56 and 11.02 kg kg⁻¹ with different N rates. In both BRRI dhan58 and BRRI dha75, the highest AE_N was found with the 50 kg N ha⁻¹ and the lowest was found with the highest N rate (175 kg N ha⁻¹).

Table 5. Effect of nitrogen rates on nitrogen use efficiency in Boro and T. Aman seasons

N rates	Boro (BRRI dhan58)		T. Aman (BRRI dhan75)
	AEN (kg kg^−1)	REN (%)	AEN (kg kg^−1)	REN (%)
N0
N25	36.00 cd	65.26 c	7.31 ab	49.43 a
N50	57.45 a	95.39 a	11.02 a	45.32 ab
N75	45.01 bc	86.27 ab	9.60 a	43.02 ab
N100	47.64 ab	87.07 ab	6.58 abc	44.13 ab
N125	45.34 bc	91.25 a	4.42 bcd	31.64 ab
N150	33.85 d	72.01 bc	2.14 cd	23.22 ab
N175	25.88 d	59.54 c	0.56 d	20.14 b
CV (%)	8.58	6.75	30.27	26.78

†Means followed by the same lowercase letter in a column do not differ from each other by Tukey’s HSD test at 5% level of significance.

‡ AE_N = agronomic nitrogen use efficiency, RE_N = nitrogen recovery efficiency.

The RE_N reflects the percentage recovery of applied N in the above-ground biomass (Dobermann, 2005). In Boro season, the RE_N efficiency of BRRI dhan58 varied between 60-95%, while in T. Aman season, the RE_N of BRRI dhan75 was 20–49% with different N rates. The RE_N of BRRI dhan58 was highest (95.39%) with 50 kg N ha⁻¹, while the highest RE_N of BRRI dhan75 was found with 25 kg N ha⁻¹. In case of both varieties, the lowest RE_N was observed with the highest rate (175 kg N ha⁻¹).

4. Discussion

4.1 Plant height

In Boro season, the plant height response of BRRI dhan58 was linear which was supported by previous studies (Gewaily et al., 2018; Moro et al., 2015). However, Contreras et al. (2017) reported quadratic response of rice plant height to N rates. The higher plant height with N application might be attributed to the role of N in improving rice growth, internodes elongation, photosynthesis and metabolism and assimilated production. In T. Aman season, soil’s inherent nutrients availability increases due to high temperature and rainfall which supports the rice growth. This might be the reason of no significant response in plant height of BRRI dhan75 to applied N rates. Tayefe et al. (2014) also reported no significant variation in plant height with N fertilization.

4.2 Tiller number

The tiller number of BRRI dhan58 and BRRI dhan75 significantly increased with the increased N rates. The increased tiller with N fertilization was due to the increased availability of N which played a vital role in cell division. In our study, tiller number m⁻² showed a linear response to applied N. However, Contreras et al. (2017) found a quadratic relation of rice tillering to applied N rates.

4.3. Panicle number and length

Panicle number is a major yield determining factor of rice. In this study, rice panicle number and panicle length significantly increased with the increase of N rates. Nitrogen contributes to rice panicle formation by stimulating cell division in the reproductive stage of crop growth. The increase in panicle number and panicle length with N fertilization was reported by Gewaily et al. (2018); Yoseftabar (2013) which supports our results.

4.5. Filled and unfilled grain

The filled and unfilled grains in per panicle were significantly influenced by N rates. The increase in the number of filled grain with the increase in N rates indicates that N fertilization is important for both sources and sinks development. On the other hand, when a single nutrient element increases, the availability of other nutrients and genetic potentiality of the crops become more determining factors for the yield component (Yesuf & Balcha, 2014). This might be the reason for low response or decline in filled grain per panicle with increased N rates in case of BRRI dhan75 in T. Aman season. Yoshida et al. (2006) reported more production of spikelet and less translocation of photoassimilates from leaves to spikelet with increased N rates which explains the increased unfilled grain with higher N rates in our study.

4.6. Grain and straw yield

Our results showed that rice yield increased with applied N to a certain level and after that it decreased with the increased N rates. Tayefe et al. (2014) and Moro et al. (2015) reported quadratic response of grain yield to N fertilization which supports our result. In T. Aman season, the lower increase of rice grain yield with applied N rates compared to N₀ was due to the availability of soil indigenous nutrients. In T. Aman season, due to high rainfall and temperature compared to Boro season, the soil indigenous nutrients become more available leading to low response to the applied N. The optimum N rates for maximum yield also differed with rice varieties and growing seasons indicating that N fertilization should be based on considering rice cultivar and climatic conditions. Nitrogen fertilization improved rice vegetative growth in terms of plant height and tiller number leading to increased straw yield of BRRI dhan58 in Boro season.

4.7. Nitrogen uptake and use efficiency

Our results suggest that in both seasons total N uptake by rice plant increased with increased N rates up to a certain level than it decreased. Yesuf and Balcha (2014) also reported similar findings. However, BRRI dhan58 in Boro season showed more response to higher N rates than BRRI dhan75 in T. Aman season which indicates that BRRI dhan75 has low N requirement. Nitrogen use efficiency is largely influenced by grain yield, N fertilizer input and N uptake (Qiao et al., 2012). According to Dobermann (2005) under optimum management practices, AE_N should be above 25 kg kg⁻¹. In Boro season, most of the AE_N values were above the Doberman standard with different N rates, while in T. Aman season, the AE_N values of BRRI dhan75 were below the standard (Table 5). Islam et al. (2015) reported 16–36 kg kg⁻¹ AE_N in Boro rice, while Hussain et al. (2016) found AE_N between 0.52 and 17 kg kg⁻¹ in T. Aman rice which supports our results. Our study also indicates that AE_N decreased with the increase of N inputs. The much lower AE_N of BRRI dhan75 in T. Aman season is due to the increased availability of soil inherent nutrients. In both Boro and T. Aman seasons, RE_N significantly decreased with the increase of N rates (Table 5). The RE_N of BRRI dhan58 in Boro season was much higher than that of BRRI dhan75 in T. Aman season which clearly indicates rice response to applied N varies with rice genotypes and growing seasons. Doberman, 2005 reported that RE_N is 50-80% in well-managed system or low level of N use or low soil N supply which explains the variation in RE_N in Boro and T. Aman seasons.

4. Conclusion

The growth and yield response of BRRI dhan58 and BRRI dhan75 grown in Boro and T. Aman seasons, respectively, to applied N varied significantly. BRRI dhan58 responded to higher N rates, while BRRI dhan75 showed less response to applied N doses. BRRI dhan58 showed a yield increase up to 348% with 125 kg N ha⁻¹ in Boro season, whereas BRRI dhan75 resulted in a yield increase up to 18% with 75 kg N ha⁻¹ in T. Aman season. However, the optimum N rates calculated from the quadratic regression equations were 142 kg ha⁻¹ and 82 kg ha⁻¹ for BRRI dhan58 and BRRI dha75, respectively. The AE_N and RE_N for applied N were much higher in BRRI dhan58 compared to BRRI dhan75 and were significantly low with higher N rates irrespective of growing seasons and variety. Our study results suggest that N fertilizer requirement of rice crops should be based on variety and climatic conditions in order to improve rice yield as well as to minimize the excessive use of chemical fertilizer.

Acknowledgments

The authors are thankful to Soil Science Division, Bangladesh Rice Research Institute for providing supports to conduct field experiments and laboratory facilities. This research did not receive any specific funding.

Disclosure statement

The authors have declared that no competing interests exist.