Nitrogen acquisition and its relation to growth and yield in recent high-yielding cultivars of rice (Oryza sativa L.) in Japan

Abstract

During the last two decades, high-yielding cultivars for multipurpose use of rice have been bred and released in Japan. Some of them have repeatedly recorded high yields of over 9 t ha⁻¹ of brown rice (about 11.25 t ha⁻¹ of rough rice). Here, characteristic features of nitrogen (N) acquisition and its relation to formation of yield components, dry matter production and grain yield at yield levels over 9 t ha⁻¹ of brown rice in recent high-yielding cultivars, a large grain type of japonica variety, “Akita 63,” extra-panicle weight types of indica variety, “Takanari” and “Saikai 198,” and a panicle weight type of japonica variety, “Fukuhibiki,” are described as compared with those in the standard japonica cultivars, “Toyonishiki” and “Nipponbare.” The grain yield of the recent high-yielding cultivars was 9.4 to 11.6 t ha⁻¹ of brown rice; that is 1.2−1.7 times greater than those of the standard cultivars. Sink capacity (1000-grain weight × spikelet number per unit land area) was 47−62% greater in the recent high-yielding cultivars, largely due to their 1.3−1.5 times greater N-use efficiency for sink formation (sink capacity per unit amount of total plant N in the aboveground part at maturity), although major component(s) responsible for their greater sink capacity differ among the cultivars. The ratio of grain yield to total dry matter was 1.1−1.4 times greater in the recent high-yielding cultivars than in the standard cultivars, indicating that the former efficiently translocate dry matter into spikelets during the grain-filling period. N-use efficiency for dry matter production (total dry matter per unit amount of total plant N) was comparable between “Akita 63,” “Fukuhibiki” and “Toyonishiki,” and slightly greater in “Takanari” and “Saikai 198” than in “Nipponbare.”

These results indicate that greater N-use efficiency for sink formation and efficient translocation of dry matter into spikelets contribute greatly to the high-yielding potential of the recent high-yielding cultivars.

Keywords:

• high yield, N-use efficiency, rice, sink capacity, yield components.

Introduction

The world's population will increase to a level 1.4−1.5 times the present population by 2025, which will require that rice (Oryza sativa L.) production be increased by 50−60% (IRRI 1995). It is, therefore, crucial to increase rice production within a relatively short period. The increase in rice production must be achieved by an increase in yield per unit land area because there is little scope for expanding land area. In addition, this increase needs to be achieved through efficient use of nitrogen (N) fertilizer to conserve the natural environment (Cassman et al. 1998; Ladha et al. 1998). One solution to this problem is to use a high-yielding cultivar exhibiting a higher grain yield per unit amount of N acquired, namely, higher N-use efficiency for grain production (Mae et al. 2006).

Rice production in Japan more than doubled during the 10 years after World War II to meet the strong demand for a staple food grain. The yield per unit land area increased by about 50% from the mid-1950s to the mid-1970s (Honya 1989). Increased production, however, resulted in a surplus from the late 1960s. In addition, the eating habits of the Japanese changed with people eating a wider variety of food than before, and with demand shifting from quantity to high quality of rice. As the result of such trends together with the decrease in the nation's population and increase in the proportion of older people, the nation's rice consumption has decreased year by year, and the present consumption of rice per person is almost half that in the 1960s. Now, around 40% of the total land area for rice cultivation is excessive in Japan from the standpoint of supply and demand. It has, therefore, become a serious problem how to effectively use the excess paddy fields. On the other hand, today's self-sufficiency ratio of Japan for food is only around 40% (calorie basis), a level which is especially low among advanced countries. Therefore, it is strongly desired for Japan to reduce the amount of imported agricultural products and to increase the amount of domestically grown agricultural products.

One way to solve the above problem might be to grow super high-yielding rice cultivars in excess paddy fields, their products being used for multiple purposes other than as a staple cereal, such as materials for rice flour, alcohol production, animal feed, etc. For such purposes, super high-yielding cultivars will be required when the cost performance is taken into account. High productivity would become the most important target for rice breeding rather than grain quality and eating quality, both of which have been the primary required traits for rice breeding in Japan.

To respond to such demand, a “super high-yielding rice project” was initiated by the Japanese Ministry of Agriculture, Forestry and Fisheries and by some prefectural agricultural experimental stations in 1981 (Horie et al. 2005; Masaki 2006). Since then, many new cultivars have been bred and released.

Challenges to high yields at yield-potential levels employing new rice cultivars, and analysis of the factors responsible for their high yields, are important for further improvement of productivity in rice cultivation.

Undoubtedly, management of N nutrition is one of the factors essential to achieve a high yield of rice, because the formation of yield-determining factors such as panicle number per unit land area, spikelet number per panicle, leaf area index, leaf N contents, and canopy structure are closely related to the amount of N absorbed by the plants at crucial stages (Wada et al. 1986). N supply from soil and fertilizer to rice plants and its absorption by the plants throughout their life spans are the key issues to achieve a high yield of rice. However, N acquisition and its relation to growth and yield at near yield-potential levels over 9 tons (t) ha⁻¹ of brown rice have seldom been studied in recent high-yielding cultivars of rice (Mae et al. 2006; Nagata et al. 2009, 2010, Teshirogi et al. 1995; Ying et al. 1998a,b).

In this review, although available data are limited and fragmental, N acquisition and its relation to formation of yield components, dry matter production, and grain production in some representative cultivars (“Akita 63,” “Takanari,” “Saikai 198,” and “Fukuhibiki”) of recent high-yielding rice in Japan are described as compared with those in old high-yielding cultivars and standard japonica cultivars. In addition, a systematic study of N-use efficiencies for sink formation, dry matter production, and grain production in “Akita 63” is described.

High Yields if Rice in Japan

High yields at near yield-potential levels over 9 t ha⁻¹ of brown rice were not often achieved until the 1980s in Japan. In a rice-yield competition among farmers from 1949 to 1968, there were only 10 examples of high yields over 9 t ha⁻¹ of brown rice: 4 in Akita Prefecture located in northeastern Japan; 3 in Nagano Prefecture and 2 in Toyama Prefecture in midland Japan, and 1 in Kagawa Prefecture in southern Japan (Honya 1989). The highest yield was 10.5 t ha⁻¹ of brown rice (about 13.2 t ha⁻¹ of rough rice) with the japonica cultivar “Ohtori” in Akita Prefecture in 1960. That was almost triple the national average yield at that time and nearly double the current yield in Japan. The contest was, however, discontinued with the advent of overproduction of rice in the late 1960s. The challenge to grow high-yielding rice has been continued only by some national and regional agricultural experimental stations and universities. Two remarkably high yields were recorded in northeastern Japan during the 25 years following the end of the Japanese farmers’ yield contest, i.e., 10.11 t ha⁻¹ of brown rice in Akita Prefecture with the japonica cultivar “Akihikari” in 1975 (Kamata et al. 1978), and 10.25 t ha⁻¹ of brown rice in Yamagata Prefecture with the japonica cultivar “Yukigesyou” in 1986 (Jinbo et al. 1987).

Since the start of the “super high-yielding rice project” in 1981, a number of new high-yielding cultivars for multipurpose use have been released and their yield potentials have been examined. Successful cultivars repeatedly exhibited high grain yields of over 9 t ha⁻¹ of brown rice in different years and/or at different locations (Nagata 2009). “Fukuhibiki,” a japonica panicle weight type cultivar, recorded high yields of 9.38, 9.78, 9.48, 9.59, and 10.00 t ha⁻¹ of brown rice in Fukushima Prefecture in northeastern Japan in the period 1990−1994 (Higashi et al. 1994; Teshirogi et al. 1995). A large-grain japonica cultivar, “Akita 63,” recorded 9.83 t ha⁻¹ and 9.54 t ha⁻¹ of brown rice in Akita Prefecture in 2000 and 2001, respectively (Mae et al. 2006). The indica extra-panicle weight-type cultivar “Takanari” recorded 9.71 t ha⁻¹ of brown rice in Tokyo in 2001 (San-oh et al. 2004) and 11.61 t ha⁻¹ and 9.46 t ha⁻¹ of brown rice in Hiroshima in southern Japan in 2008 and 2009, respectively (Nagata et al. 2009, 2010). The indica extra-panicle weight-type cultivar “Saikai 198” recorded 10.82 t ha⁻¹ of brown rice in Hiroshima Prefecture in 2008. The indica extra-panicle weight-type cultivar “Hokuriku 193” recorded 11.22 t ha⁻¹ in 2008 and 10.02 t ha⁻¹ in 2009 and the extra-panicle weight type cultivar of crossbred japonica-indica “Saikai 203” recorded 10.07 t ha⁻¹ in 2008 in Hiroshima Prefecture (Nagata et al. 2009, 2010).

Grain Yield, Dry Matter Production, Ratio of Grain Yield to Total Dry Matter, and N Acquisition in Recent High-Yielding Cultivars

Available data on total plant N, grain yield, total dry matter and yield components at high-yield levels over 9 t ha⁻¹ of brown rice in some recent and old high-yielding cultivars in Japan were collected from the literature (Honya 1989; Jinbo et al. 1987; Kamata et al. 1978; Mae et al. 2006; Nagata et al. 2009, 2010, Teshirogi et al. 1995) and their N-use efficiencies for sink formation, dry matter production and grain production were compared with those of standard japonica cultivars (Tables 1–4).

Table 1 Grain yield, total dry matter, ratio of grain yield to total dry matter, and total plant nitrogen (N) content in recent and old high-yielding cultivars and reference cultivars of rice plants grown with high levels of N supply

Year	Cultivar	Grain yield (brown rice) (t ha^−1)	Total dry matter (t ha^−1)	Ratio of grain yield to total dry matter	Total plant N (kg ha^−1)	Location
2000	Akita 63	9.83 (129)	19.12 (95)	0.51 (134)	161 (101)	Akita Pref.
	Yukigesyou	7.63 (100)	20.03 (100)	0.38 (100)	159 (100)	Akita Pref.
2001	Akita 63	9.39 (123)	20.17 (113)	0.47 (109)	184 (123)	Akita Pref.
	Toyonishiki	7.65 (100)	17.92 (100)	0.43 (100)	149 (100)	Akita Pref.
2008	Takanari	11.61 (169)	23.51 (120)	0.49 (140)	245 (106)	Hiroshima Pref.
	Saikai 198	10.82 (157)	23.90 (122)	0.45 (129)	254 (110)	Hiroshima Pref.
	Nipponbare	6.89 (100)	19.65 (100)	0.35 (100)	231 (100)	Hiroshima Pref.
1994	Fukuhibiki	10.00 (131)^†	21.20 (118)^†	0.47 (109)^†	150 (101)^†	Fukushima Pref.
1975	Akihikari	10.11 (132)^†	20.00 (112)^†	0.51 (119)^†	214 (144)^†	Akita Pre.
1986	Yukigeshou	10.25 (134)^†	25.32 (141)^†	0.41 (95)^†	289 (194)^†	Yamagata Pref.
1960	Ohtori	10.52 (138)^†			177 (119)^†	Akita Pref.
Recent high-yielding cultivars are “Akita 63” (Mae et al. 2006), “Takanari,” “Saikai 198” (Nagata et al. 2009), and “Fukuhibiki” (Teshirogi et al. 1995). Old high-yielding cultivars are “Akihikari” (Kamata et al. 1978), “Yukigesyou” (Jinbo et al. 1987) and “Ohtori” (Honya 1989). Reference cultivars are “Yukigesyou” (2000), “Toyonishiki” (2001) and “Nipponbare” (2008).
Values in parentheses show the percentage of the value of Yukigesyou (2000), Toyonishiki (2001) or Nipponbare (2008) in respective years. Values in parentheses with ^†show the percentage of the value of Toyonishiki (2001).

Table 2 Yield components and sink capacity in recent and old high-yielding cultivars and reference cultivars of rice plants grown with high levels of nitrogen (N) supply

Year	Cultivar	No. of panicles (m^−2)	No. of spikelets per panicle	Total No. of spikelets (×10^3m^−2)	1000-grain wt (g)	Filled spikelets (%)	Sink capacity^† (g m^−2)
2000	Akita 63	456	91	41.5	31.2	78.5	1293
	Yukigesyou	432	80	34.6	23.1	92.5	800
2001	Akita 63	455	94	42.6	30.6	72.1	1304
	Toyonishiki	436	79	34.1	23.6	95.0	805
2008	Takanari	305	197	60.2	21.2	91.0	1276
	Saikai 198	344	158	54.3	23.4	85.1	1272
	Nipponbare	432	83	35.8	22.4	85.9	802
1994	Fukuhibiki	453	107	48.7	24.3	85.7	1183
1975	Akihikari	603	98	58.9	21.6	86.7	1272
1986	Yukigeshou	515	114	58.7	21.3	82.0	1250
1960	Ohtori	446			24.7	83.3
Recent high-yielding cultivars are “Akita 63” (Mae et al. 2006), “Takanari,” “Saikai 198” (Nagata et al. 2009), and “Fukuhibiki” (Teshirogi et al. 1995). Old high-yielding cultivars are “Akihikari” (Kamata et al. 1978), “Yukigesyou” (Jinbo et al. 1987) and “Ohtori” (Honya 1989). Reference cultivars are “Yukigesyou” (2000), “Toyonishiki” (2001) and “Nipponbare” (2008).
^†Sink capacity = 1000-grain weight (g) × total No. of spikelets per unit land area (m^−2)

Table 3 Nitrogen (N)-use efficiencies for sink formation, dry matter production, and grain production in recent and old high-yielding cultivars and reference cultivars of rice plants grown with high levels of N supply

		N-use efficiency for
Year	Cultivar	Sink formation (kg   kg-N^−1)	Dry matter production (kg kg-N^−1)	Grain production (kg kg-N^−1)
2000	Akita 63	80.3	119	61.1
	Yukigesyou	50.3	126	48.0
2001	Akita 63	70.9	110	51.0
	Toyonishiki	54.0	120	51.3
2008	Takanari	52.1	96	47.4
	Saikai 198	50.1	94	42.6
	Nipponbare	34.7	85	29.8
1994	Fukuhibiki	78.9	141	66.7
1975	Akihikari	59.4	93	47.2
1986	Yukigeshou	43.3	88	35.5
1960	Ohtori			59.4
Recent high-yielding cultivars are “Akita 63” (Mae et al. 2006), “Takanari,” “Saikai 198” (Nagata et al. 2009), and “Fukuhibiki” (Teshirogi et al. 1995). Old high-yielding cultivars are “Akihikari” (Kamata et al. 1978), “Yukigesyou” (Jinbo et al. 1987) and “Ohtori” (Honya 1989). Reference cultivars are “Yukigesyou” (2000), “Toyonishiki” (2001) and “Nipponbare” (2008).
N-use efficiencies for sink formation, dry matter production or grain production were defined as sink capacity, total dry matter or grain yield per unit amount of total plant N, respectively.

Table 4 Total plant nitrogen (N) content, rates of N application, and partial factor productivity of applied N in recent and old high-yielding cultivars and reference cultivars of rice plants grown with high levels of N supply

			Fertilizer-N applied			PFPN ^‡ (kg kg^−1)
Year	Cultivar	Total plant N acquired (kg-N ha^−1)	Basal application (kg-N ha^−1)	Topdressing (kg-N ha^−1)	Total (kg-N ha^−1)
2000	Akita 63	161	20 + 60^†	20 × 4 times	160	61.4
	Yukigesyou	159	20 + 60^†	20 × 4 times	160	47.7
2001	Akita 63	184	40 + 70^†	20 × 2 times	150	62.6
	Toyonishiki	149	40 + 70^†	20 × 2 times	150	51.0
2008	Takanari	245	50	20–30 × 5 times	180	64.5
	Saikai 198	254	50	20–30 × 5 times	180	60.1
	Nippon bare	231	50	20–30 × 5 times	180	38.3
1994	Fukuhibiki	150	80	20 × 2 times	120	88.3
1975	Akihikari	214	70	20 × 3 times	130	77.8
1986	Yukigeshou	289	30 + 70^†	15–20 × 4 times	170	60.3
1960	Ohtori	177	107	0	107	98.3
Recent high-yielding cultivars are “Akita 63” (Mae et al. 2006), “Takanari,” “Saikai 198” (Nagata et al. 2009), and “Fukuhibiki” (Teshirogi et al. 1995). Old high-yielding cultivars are “Akihikari” (Kamata et al. 1978), “Yukigesyou” (Jinbo et al. 1987) and “Ohtori” (Honya 1989). Reference cultivars are “Yukigesyou” (2000), “Toyonishiki” (2001) and “Nipponbare” (2008).
^†The amount of N applied as controlled release fertilizer LP100 Type (polyolefin-coated urea, Chisso Co., Japan). ^‡Partial factor productivity of applied N (kg grain per kg N applied).

Table 1 shows the grain yield (brown rice), total dry matter and total plant N in the aboveground part at maturity, and the ratio of grain yield to total dry matter in recent and old high-yielding cultivars and reference cultivars. A high-yielding cultivar, “Akita 63,” and a reference cultivar, “Yukigesyou” (2000) or “Toyonishiki” (2001), were grown with the same high levels of N supply (Mae et al. 2006). High-yielding cultivars, “Takanari” and “Saikai 198,” and a reference cultivar, “Nipponbare,” were also grown with the same high level of N supply in 2008 (Nagata et al. 2009). “Akihikari,” “Yugigesyou,” and “Ohtori” are old high-yielding japonica cultivars (Honya 1989; Jinbo et al. 1987; Kamata et al. 1978). Here, “Toyonishiki” (2001) and “Nipponbare” (2008) are regarded as the standard japonica cultivars in northern Japan and southern Japan, respectively.

The yield of “Akita 63” in 2000 was 9.83 t ha⁻¹, 29% greater than that of the reference cultivar, “Yukigesyou.” The amounts of total dry matter and the total plant N did not differ between “Akita 63” and “Yukigesyou,” but the ratio of grain yield to total dry matter was much larger (34%) in “Akita 63.” The yield of “Akita 63” in 2001 was 9.39 t ha⁻¹, 23% greater than that of the reference cultivar, “Toyonishiki.” Both total dry matter and total plant N were greater in “Akita 63.” The ratio of grain yield to total dry matter was slightly larger (9%) in “Akita 63.”

The yield of “Takanari” in 2008 was very high; 11.61 t ha⁻¹ of brown rice, 69% greater than that of the reference cultivar, “Nipponbare.” Total dry matter was 20% greater and the total plant N was slightly larger (6%) in “Takanari.” The ratio of grain yield to total dry matter was much larger (40%) in “Takanari.”

“Saikai 198” recorded results similar to those of “Takanari.” The yield of “Saikai 198” was 10.82 t ha⁻¹, 57% greater than that of “Nipponbare.” Total dry matter was 20% greater and the total plant N was 10% greater in “Saikai 198.” The ratio of grain yield to total dry matter was much greater (29%) in “Saikai 198.”

“Fukuhibiki” yielded 10.00 t ha⁻¹ of brown rice with relatively smaller amounts of total dry matter and total plant N, and a greater ratio of grain yield to total dry matter as compared with those of other high-yielding cultivars and the reference cultivars.

“Akihikari” yielded 10.11 t ha⁻¹ of brown rice in 1975 with almost the same amounts of total dry matter and a greater amount of total plant N as compared with those of “Akita 63.” The ratio of grain yield to total dry matter in “Akihikari” was as large as those of “Akita 63” and “Takanari.” Yukigeshou recorded 10.25 t ha⁻¹ of brown rice in 1986 at Yamagata Prefecture with a greater amount of dry matter production, a greater amount of total plant N, and a relatively smaller ratio of grain yield to total dry matter. “Ohtori” recorded 10.52 t ha⁻¹ of brown rice in 1960 with a relatively smaller amount of total plant N.

Yield Components and Sink Capacity in Recent High-Yielding Cultivars

Sink capacity

High-yielding cultivars can be roughly classified into three types based on the difference in the components most responsible for sink formation, namely, large-grain type, (extra-) panicle-weight type, and panicle-number type.

“Akita 63” is a large-grain type. The weight of 1000 grains (brown rice) of “Akita 63” is about 30−31 g, which is 30−50% greater than those (20−23.5 g) of common japonica cultivars cultured in Japan (Table 2) (Hoshikawa 1975). The total number of panicles per unit land area was slightly larger, and the number of spikelets per panicle 14−19% greater, in “Akita 63” than in the reference cultivars. Thus, the total number of spikelets per unit land area was 20−25% larger in “Akita 63” than in the reference cultivars (Table 2). Due to the much larger grain size and the greater number of spikelets, the sink capacity, defined as 1000-grain weight × number of spikelets per unit land area, was 62% larger in “Akita 63” than in the reference cultivars in both years, 2000 and 2001.

“Takanari” and “Saikai 198” are extra-panicle weight-type cultivars of high-yielding rice having a large number of grains per panicle. The number of spikelets per panicle was 197 in “Takanari,” 158 in “Saikai 198,” and 83 in Nipponnbare. The number of spikelets per panicle was 1.9−2.4 times greater in “Takanari” and Saikai 198 than in “Nipponbare.” Although the number of panicles per unit land area was 20−29% fewer in “Takanari” and “Saikai 198,” the sink capacity of “Takanari” and “Saikai 198” was 59% greater than that of “Nipponbare” due to a much greater number of spikelets per panicle (Table 2). The 1000-grain weight was not very different between “Takanari,” “Saikai 198” and “Nipponbare.”

“Fukuhibiki” is a panicle-weight type cultivar of high-yielding rice. “Fukuhibiki” had a 32% greater number of spikelets per panicle and a slightly larger 1000-grain weight than did “Toyonishiki.” Its sink capacity was 47% greater than that of “Toyonishiki” (Table 2).

“Akihikari” is a panicle-number type japonica cultivar. The number of panicles per unit land area was 603 in “Akihikari,” about 40% greater than in “Toyonishiki.” The number of spikelets per panicle was about 20% larger in “Akihikari.” Thus, the sink capacity of “Akihikari” was 59% greater than that of “Toyonishiki” (Table 2).

“Yukigeshou” in 1986 had a 41% greater number of spikelets per panicle and a 19% greater number of panicles per unit land area than “Toyonishiki.” Sink capacity was 56% larger in “Yukigesyou” than in “Toyonishiki” (Table 2).

From the data in Table 2 it is clear that the ability for sink formation of the recent high-yielding cultivars is superior to those of common japonica cultivars under conditions with high levels of N supply.

Proportion of filled spikelets

The proportions of filled spikelets of “Akita 63” were 72−79% (Table 2) with higher ratios of grain yield to total dry matter (Table 1), while the proportions of the reference cultivars were 93−95% with lower ratios of grain yield to total dry matter (Table 1). This indicates that “Akita 63” still had some space to further accumulate dry matter into the spikelets, while the reference cultivars had almost no space to further accumulate dry matter into the spikelets.

The proportion of filled spikelets of “Takanari” was 91%, an especially high level among the recent high-yielding cultivars. The ratio of grain yield to total dry matter was higher in “Takanari” than in “Nipponbare” (Table 1).

The proportion of filled spikelets was 85.1% in “Saikai 198,” almost same as that of “Nipponbare,” but with a higher ratio of grain yield to total dry matter (Tables 1 and 2).

The proportions of filled spikelets of “Fukuhibiki” and “Akihikari” were 86−87% with higher ratios of grain yield to total dry matter than that of “Toyonishiki.”

These results indicate that the translocation of dry matter into spikelets proceeds more efficiently in the high-yielding cultivars than in the standard japonica cultivars during the grain-filling period.

N-Use Efficiencies for Sink Formation, Dry Matter Production, and Grain Production in Recent High-Yielding Cultivars

N-use efficiency for sink formation (sink capacity per unit amount of total plant N) was much greater in “Akita 63,” “Takanari,” “Saikai 198” and “Fukuhibiki” and slightly greater in “Akihikari” than in the standard cultivars (Table 3).

The N-use efficiency for sink formation was greater in “Akita 63,” “Fukuhibiki,” and “Akihikari” than in “Takanari,” “Saikai 198” and “Nipponbare.” “Akita 63,” “Fukuhibiki,” and “Akihikari” were grown in northeastern Japan, while “Takanari,” “Saikai 198” and “Nipponbare” were cultured in southern Japan. It has been previously shown that the spikelet formation efficiency of rice plants for a given amount of total plant N is greater in colder regions than in warmer regions (Murayama 1969; Wada et al. 1986). The difference in N-use efficiencies for sink formation might be due to the differences in the location of their cultivation.

The N-use efficiency for sink formation in “Yukigeshou” in 1986 was smaller than that in “Yukigeshou” in 2000. N-use efficiency for sink formation becomes smaller with an increase in total plant N (Yoshinaga et al. 2009). The lower N-use efficiency for sink formation in 1986 might be due to the greater amount of total plant N in 1986 than in 2000.

The N-use efficiency for dry matter production was not different between “Akita 63” and “Yukigesyou” in 2000, but the N-use efficiency for grain production was greater in “Akita 63” (Table 3). The N-use efficiency for dry matter production was slightly lower in “Akita 63” than in “Toyonishiki” in 2001, and the N-use efficiency for grain production was not different between the two cultivars (Table 3). N-use efficiencies for dry matter production and for grain production are generally lower with an increase in the amount of total plant N (Wada et al. 1986; Jinbo et al. 1987). The amount of total plant N was 184 kg-N per ha in “Akita 63” and 149 kg-N per ha⁻¹ in “Toyonishiki” in 2001. The total plant N was lower in “Toyonishiki.” Therefore, it can be speculated that N-use efficiency for grain production would be greater in “Akita 63” if comparison were made for the same amount of total plant N. This will be clarified later (Figs 1−4).

Figure 1. Relationship between total dry weight and total plant nitrogen (N) content per unit land area at harvest in Akita 63 and the reference cultivars, Yukigeshou, Toyonishiki and Akitakomachi. (Mae et al. 2006, with permission from Elsevier) ▴: Akita 63 in 2000, •: Akita 63 in 2001, ▪: Akita 63 in 2002, ○: Yukigeshou in 2000, □: Toyonishiki in 2001, ▵: Akitakomachi in 2000 and ⋄: Akitakomachi in 2002.

The N-use efficiency for dry matter production was slightly greater in “Takanari” (13%) and “Saikai 198” (11%) than in “Nipponbare,” in spite of their larger amounts of total plant N (Table 3). This indicates that both cultivars would have greater N-use efficiency for dry matter production for a given amount of total plant N. The N-use efficiency for grain production was much greater in “Takanari” (59%) and “Saikai 198” (43%) than in “Nipponbare.” These results indicate that the potential abilities for dry matter production and grain production in “Takanari” and “Saikai 198” are superior to those in the standard japonica cultivars.

The N-use efficiencies for dry matter and grain production in “Fukuhibiki” were especially greater among the high-yielding cultivars (Table 3).

The N-use efficiencies for dry matter production and grain production in “Akihikari” were comparable to those of “Takanari” and “Saikai 198.” The N-use efficiency for grain production in “Yukigesyou” in 1986 was not as great as that for other high-yielding cultivars, but the amount of total plant N (289 kg-N ha⁻¹) was significantly greater in “Yukigeshou” than in others.

N Fertilization in Recent High-Yielding Cultivars

The total amount of fertilizer-N applied to high-yielding rice ranged from 107 to 180 kg-N ha⁻¹ (Table 4). This amount is about 1.5−3 times greater than the amount of fertilizer-N applied to rice plants producing an average yield of 5−6 t ha⁻¹ of brown rice in Japan. Partial factor productivity of applied N defined as grain yield per unit amount of applied fertilizer-N was greater in the recent high-yielding cultivars than in the standard japonica cultivars.

In the case of “Akita 63,” 20 kg-N ha⁻¹ of ammonium sulfate and 60 kg-N ha⁻¹ of controlled release fertilizer (LP100 Type polyolefin-coated urea, Chisso Co., Japan) were applied as a basal application and 80 kg-N ha⁻¹ of ammonium sulfate was applied as topdressing in 4 splits with 20 kg-N ha⁻¹ each at the early and middle tillering stages, the panicle initiation stage, and the reduction division stage in 2000. In 2001, 40 kg-N ha⁻¹ of ammonium sulfate and 70 kg-N ha⁻¹ of controlled release fertilizer (the same as above) were applied as a basal application and 40 kg-N ha⁻¹ was applied as topdressing in 2 splits with 20 kg-N ha⁻¹ each at the middle tillering stage and the reduction division stage). As a single application, 100 kg-N ha⁻¹ of the controlled release fertilizer, LP100 Type, and 40 kg-N ha⁻¹ of controlled fertilizer, LPS 100 Type, were applied as a basal application and topdressing was neglected (grain yield; 9.54 t ha⁻¹ of brown rice, Data not shown (Mae et al. 2006)). Both LP100 Type and LPS100 Type fertilizers release 80% of their total N content by 100 days after their application at any temperature between 20°C and 30°C. The average recovery of N from conventional N-fertilizer by rice plants is around 30% for a basal application and 55% for topdressing after the panicle initiation stage (Mae and Shoji 1984) and the recovery of LP100 Type and LPS-100 Type is around 60% in northeastern Japan (Shoji and Gandeza 1992). By using these recovery ratios, the amounts of fertilizer-N acquired by “Akita 63” were estimated. The amount of N recovered from the ammonium sulfate applied as a basal application was 6 kg-N ha⁻¹ in 2000 and 12 kg-N ha⁻¹ in 2001. The amount of N recovered from the controlled-release fertilizer was 36 kg-N ha⁻¹ in 2000 and 42 kg-N ha⁻¹ in 2001. The amount of N recovered from the ammonium sulfate applied as topdressing was 44 kg-N ha⁻¹ in 2000 and 22 kg-N ha⁻¹ in 2001. The total amount of fertilizer-N acquired by “Akita 63” was 86 kg-N ha⁻¹ in 2000 and 76 kg-N ha⁻¹ in 2001. From these estimated amounts of fertilizer-derived N, the amount of soil-derived N acquired by “Akita 63” was estimated to be 75 kg-N ha⁻¹ in 2000 and 108 kg-N ha⁻¹ in 2001. The ratio of fertilizer-derived N to total plant N was 53% in 2000 and 41% in 2001. These ratios are 10-20% larger than the ratio in rice plants producing a standard level of grain yield (5-6 t ha⁻¹ of brown rice) in northeastern Japan (Mae and Shoji 1984).

In “Takanari” and “Saikai 198” in 2008, 50 kg-N ha⁻¹ was applied as a basal application (NPK compound fertilizer), and 130 kg-N ha⁻¹ was applied as topdressing in 5 splits at appropriate growth stages (20 kg-N ha⁻¹ at the middle tillering stage (NPK compound fertilizer); 30 kg-N ha⁻¹ each at the late tillering stage and the panicle initiation stage (NK compound fertilizer); 20 kg-N ha⁻¹ at reduction division stage (NK compound fertilizer); and 30 kg-N ha⁻¹ at the full-heading stage (ammonium sulfate) (Nagata et al. 2009). In Fukuhibiki, 80 kg-N ha⁻¹ was applied as a basal application, and 40 kg-N ha⁻¹ as topdressing in 2 splits (20 kg-N ha⁻¹ each at the panicle initiation and the reduction division stages) (Teshirogi et al. 1995). In “Akihikari,” 70 kg-N ha⁻¹ was applied as a basal application and 60 kg-N ha⁻¹ as topdressing in 3 splits of 20 kg-N ha⁻¹ each (Kamata et al. 1978). In “Yukigesyou” in 1986, 110 kg-N ha⁻¹ was applied as a basal application (80 kg-N ha⁻¹ as a controlled release fertilizer LP100) and 60 kg-N ha⁻¹ as topdressing in 4 splits (Jinbo et al. 1987)).

The amount of fertilizer-N applied as a basal application ranged from 50 to 110 kg-N ha⁻¹. In “Akita 63,” “Toyonishiki” and “Yukigeshou,” however, a large part of the applied fertilizers was in the form of controlled release fertilizer. Therefore, the amount of fertilizer-N cannot be simply compared between them. The amount of conventional fertilizer-N applied as a basal application was 50 kg-N ha⁻¹ for “Takanari” and “Saikai 198,” which were grown in Hiroshima Prefecture in southern Japan, while the amount was 80 kg-N ha⁻¹ for “Fukuhibiki” and 70 kg-N ha⁻¹ for “Akihikari,” which were grown in Fukushima Prefecture and Akita Prefecture in northeastern Japan, respectively. The amount was greater in northern Japan than in southern Japan. It has been pointed out that in warmer regions more N is released from the soil during the vegetative stages than in the cold regions but this trend is reversed after panicle initiation (Suzuki 1997; Yamamuro 1991). Therefore, the difference in the amount of fertilizer-N applied among the cultivars would possibly be due to the difference in the locations where the cultivars were grown. It would be necessary for “Fukuhibiki” and “Akihikari” to be supplied with more fertilizer-N to support the growth at their young stages. The total amount of N for topdressing ranged from 40 kg-N ha⁻¹ to 130 kg-N ha⁻¹ in 2-5 splits and the amount was less in northern Japan than in southern Japan.

N-Use Efficiencies in “Akita 63”

The N-use efficiencies shown in Table 3 certainly provide valuable information about the relationships between N nutrition, growth, and grain yield in the high-yielding cultivars. However, due to the limited number of available data, the N-use efficiencies obtained were only from one or two data set(s) for each cultivar, and the amount of total plant N differed between the cultivars. The N-use efficiency for grain production is often smaller with an increase of total plant N (Jinbo et al. 1987; Wada et al. 1986). Wada et al. (1986) noted that the N requirement of rice plants to produce 100 kg of paddy is about 1.5 to 1.7 kg, while in the case of high-yield rice cultivation at least 1.9 kg of N is absorbed for the production of 100 kg of paddy. Therefore, strictly speaking, when the magnitude of N-use efficiency is compared among different cultivars, the comparison should be made for the same amount of total plant N. Unfortunately, no experiment has been conducted to strictly compare the N-use efficiencies between the high-yielding cultivars and common japonica cultivars, except for “Akita 63.”

A systematic experiment was conducted by Mae and colleagues (Mae et al. 2006) to examine the relationship between total plant N and the number of spikelets per unit land area, sink capacity, leaf area index, total dry matter, and grain yield in “Akita 63” and the reference cultivars, “Yukigesyou,” “Toyonishiki,” and “Akitakomachi.” The rice plants were grown with different levels of N supply for three years.

There was no difference in the relationship between the total plant N and the total dry matter in “Akita 63” and the reference cultivars (Fig. 1). This means that N-use efficiency for dry matter production for a given amount of total plant N is the same between “Akita 63” and the reference cultivars.

Similar results were also obtained for the relationship between the total plant N and the spikelet number per unit land area. There was no difference between “Akita 63” and the reference cultivars with regard to their relationship (Fig. 2).

Figure 2. Relationship between number of spikelets and total plant nitrogen (N) content per unit land area at harvest in Akita 63 and the reference cultivars, Yukigeshou, Toyonishiki and Akitakomachi. (Mae et al. 2006, with permission from Elsevier) Symbols are the same as those in Fig. 1.

Figure 3 shows the relationship between the total plant N and sink capacity. There was a clear difference in the relationship between “Akita 63” and the reference cultivars. The N-use efficiency for sink formation was clearly greater in “Akita 63” than in the reference cultivars for a given amount of total plant N. From the relationships shown in Figs 2 and 3, it is evident that the greater N-use efficiency for sink formation in “Akita 63” is simply due to its larger grain size.

Figure 3. Relationship between sink capacity^† and total plant nitrogen (N) content per unit land area at harvest in Akita 63 and the reference cultivars, Yukigeshou, Toyonishiki and Akitakomachi. (Mae et al. 2006, with permission from Elsevier) Symbols are the same as those in Fig. 1. ^†Sink capacity: number of total spikelets per unit land area (m2) × 1000-grain weight of each cultivar.

Figure 4 shows the relationship between total plant N and grain yield. Grain yield for a given amount of total plant N was greater in “Akita 63” than in the reference cultivars. Thus, the N-use efficiency for grain production is greater in “Akita 63,” indicating that the translocation of dry matter into spikelets proceeded more efficiently in “Akita 63” than in the reference cultivars during grain filling. In fact, photosynthetically-assimilated ¹³C was partitioned into panicles at higher ratios in “Akita 63” than in the reference cultivars throughout the grain-filling period (Mae et al. 2006).

Figure 4. Relationship between grain yield (brown rice) and total plant nitrogen (N) content per unit land area at harvest in Akita 63 and the reference cultivars, Yukigeshou, Toyonishiki and Akitakomachi. (Mae et al. 2006, with permission from Elsevier) Symbols are the same as those in Fig. 1.

The N-use efficiency for formation of leaf area index at the full-heading stage was almost the same between “Akita 63” and the reference cultivars for a given amount of total plant N (data not shown). Thus, the sink capacity or grain yield for a given unit of leaf area was much greater in “Akita 63” than in the reference cultivars.

Discussion

Sink capacity is the primary determinant of grain yield in cereal crops grown in high-yield environments without stress. Sink capacity in rice can be increased either by increasing the spikelet number per unit land area or grain size or both. The spikelet number per unit land area can be increased either by increasing the spikelet number per panicle or panicle number per unit land area. Because a strong compensation mechanism exists among these three components (Takita 1988; Ying et al. 1998a), it has often been experienced that an increase in one component does not necessarily result in an overall increase in sink capacity (Masaki 2006; Takeda et al. 1987). The sink capacity would be increased only if the compensation mechanism were weakened or cancelled.

As shown in Table 2, the sink capacities of the recent high-yielding cultivars, “Akita 63,” “Takanari,” “Saikai 198” and “Fukuhibiki,” were 47-62% greater than those of the standard japonica cultivars although a major yield component responsible for their greater sink capacity is different among the cultivars. The larger grain size contributed to its greater sink capacity in “Akita 63” (Figs 2 and 3), while the greater number of spikelets per panicle contributed to the larger sink capacity in “Takanari” and “Saikai 198” (Table 2). Both the larger grain size and the greater number of spikelets per panicle contributed to its greater sink capacity in “Fukuhibiki.”

More importantly, the N-use efficiency for sink formation was 1.3–1.5 times greater in the recent high-yielding cultivars than in the standard cultivars (Table 3). This characteristic is greatly advantageous for the high-yielding cultivars to achieve a high yield. For a high yield, a larger sink capacity is required. Achievement of a larger sink capacity is generally realized by a larger amount of N application and always accompanied by an increase in leaf area index, which often causes heavy self-shading and results in lodging. As the sink capacity per unit amount of plant N was greater in the recent high-yielding cultivars than in the standard japonica cultivars, the amount of total plant N required for achieving a sink capacity necessary for a high yield would be much less in the recent high-yielding cultivars than in common japonica cultivars (superior in physiological N-use efficiency for sink formation). Thus, the amount of N supplied can be much reduced in the recent high-yielding cultivars.

There were considerable differences in dry matter production among the high-yielding cultivars (Table 1). The amount of total dry matter ranged from 19.12 t ha⁻¹ (“Akita 63”) to 25.32 t ha⁻¹ (“Yukigesyou” in 1986). The difference among the cultivars can be attributed to the differences in the amount of total plant N acquired and/or N-use efficiency for dry matter production. The amount of total plant N was different among the high-yielding cultivars. It ranged from 150 kg-N ha⁻¹ (“Fukuhibiki”) to 289 kg-N ha⁻¹ (“Yukigesyou”). There was no difference in the N-use efficiency for dry matter production between “Akita 63” and the reference cultivars (Fig. 1). N-use efficiency for dry matter production was greater in “Takanari” and “Saikai 198” than in “Nipponbare.” The superiority of “Takanari” for dry matter production has been previously proved elsewhere (Xu et al. 1997). The greater dry matter production in “Yukigesyou” in 1986 could be attributed to the greater amount of total plant N because the N-use efficiency for dry matter production for a given amount of total plant N did not differ between “Akita 63” and “Yukigesyou” (Fig. 1).

The ratio of grain yield to total dry matter was 1.1–1.4 times greater in the recent high-yielding cultivars than in the reference cultivars under the conditions of high levels of N supply (Table 1). There seem to be two reasons for that. First, the recent high-yielding cultivars are able to produce sufficient sink capacity to achieve a high yield with a smaller amount of total plant N. The sink capacity was clearly insufficient for achieving a high yield in the standard japonica cultivars even with high levels of N supply (Table 2). Second, the high-yielding cultivars are able to efficiently translocate dry matter into spikelets during grain filling. In “Nipponbare,” the ratio of grain yield to total dry matter was not so great as those of the high-yielding cultivars (Table 1). Translocation of dry matter into spikelets seems to have ceased during grain filling for some reason, because the proportion of filled spikelets remained at 86% and the spikelets were likely to still have had some space to further accumulate dry matter (Table 2).

Concluding Remarks

The potential for grain production and the consistency of high yields have certainly been improved in the recent high-yielding cultivars, the upper threshold of their grain yield approaching a level over 11 t ha⁻¹ of brown rice (almost equal to 13.75 t ha⁻¹ of rough rice). Such achievements can be largely attributed to their superiority of N-use efficiency for sink formation and an efficient translocation of dry matter into spikelets.

Acknowledgments

The author wishes to thank Dr. Kenji Nagata and his colleagues for permitting the use of their data and Dr. Amane Makino for critical reading.