Soil nutrient management is an important plant growth factor to increase crop production. In Sub-Saharan Africa, soil fertility is declining especially in areas occupied by smallholder farmers. In Uganda, rice consumption demand is increasing and rice farmers are experiencing yields  less   than   2.5 t ha^-1,    the    impressive   current production increase is mainly due to extension of cultivated area (Ahmed, 2012). According to Oonyu (2011), paddy rice growing areas significantly improved rural farmer household income when yields of up to 2.5 t ha^-1 were attained in the 1980s. This is because for the last decade,  rice  has become  the  most  important  food resource in homes for small scale communities (Balasubramanian et al., 2007) and changing a traditional norm of most rice consumption to be associated with urbanization (Africa Rice Center, 2008). However, the incomes declined primarily due to declining soil fertility levels, farmers’ inability to control water use, low yielding varieties and the poor farming practices (Oonyu, 2011; Nhamo et al., 2014).

Smallholder farmers of rice need to have potential to increase production and productivity of rice yields through using improved rice cultivars and agronomic management techniques (Rodenburg et al., 2014), enhance nutrient and water availability and use efficiencies (Balasubramanian et al., 2007) and to control weeds (Africa Rice Center, 2008). Ecosystem is vulnerable to degradation when farmers do not change their farming practices and indiscriminately develop skills for the sole purpose of agricultural production. Rodenburg et al. (2014) estimates that, employing improved water and weed management practices, production of rice from a small percentage of inland valley areas could mount to total current demand in Africa. The rice production indices to increase, the farmers in Sub Sahara African region need to orient their outs to the markets and strategies required to meet the market demands. This leaves farmers with no choice but to improve productivity and a significant amount of land will be safeguarded for other tenacities (Mwaseba et al., 2006).

One of the major causes of low rice production is the decline in soil fertility, most especially nitrogen (N) and phosphorus (P) (Oonyu, 2011). Nitrogen is largely limiting nutrient in most soils (Bekunda et al., 2002) for rice production. Application of N influences plant length, leaf area, panicle number, spikelet number, ripening ratio and hence grain yield increase (Dobermann and Fairhurst, 2000). On the other hand, P has developmental activity in plant’s root growth (Rose et al., 2013). Unfortunately, this nutrient is highly deficient in soils around the country (IFDC, 2006). The introduction of New Rice for Africa (NERICA) varieties in the western highlands agro-ecological zone of Uganda left farmers using blanket fertilizer application for other crops or not even single fertilizer rates are applied (Kijima et al., 2011). Rice cultivation fields of smallholder farmers are experiencing nutrient deficiencies and farmers lack knowledge on the fertilizer rates required to be applied to boost  rice  growth and improve yields. Therefore, an appropriate application and management of these nutrients is essential for optimizing rice grain yields. The purpose of this research is therefore, to examine NERICA 4 performance on fertilizer rates and establish an optimal fertilizer amount of N and P and their influence on plant growth, yield and yield components in rice production for smallholder farmers.

Site characterization

The study area is located on the eastern central of Japan, with an average altitude of 578 meters above sea level, an annual precipitation of 1200 mm and maximum temperature of 19.0^oC and minimum temperature of 12.0^oC through the year. The soil type is an oxisol with chemical and soil pH composition as shown in table 1. The study was conducted with the aim of investigating the effect of fertilizer application of nitrogen and phosphorus on growth and yield of NERICA 4 rice variety at Tsukuba International Center, Japan from April to September 2014 cropping season. This variety is characterized with high panicle weight compared to others, high yielding potential, early to medium maturing, adaptable upland conditions and it of medium resistant to leaf blast. The variety was investigated in an upland condition for various parameters like plant height, tiller number, SPAD value, shoot dry matter, panicle dry matter, nitrogen uptake, panicle number, spikelet number, panicle length, ripening ratio, thousand grain weight, harvest index and paddy yield.

Statistical analysis

Analysis of Variance (ANOVA) was used to compare means of growth, yield and yield components data. Fisher’s protected Least Significant Difference (LSD) at P ≤ 0.05 level of significance was used to separate means. Correlation analysis was used to analyze the relationship between fertilizer application rates and crop yield.  It was also used to estimate the optimal amount required for maximum yield of NERICA 4.

Growth observation parameters

Tiller number

There was a general increase in the number of tillers  per square meter in all treatments (Figure 1). However, a significant difference among treatment combinations was observed only for P70:N80 (Phosphorus: 70 kg ha^-1; Nitrogen: 80 kg ha^-1), with the highest number of tillers through the growing season.

Leaf area index

Figure 2 indicates the leaf area index performance of rice‘NERICA 4’ for the different N and P treatments. There was a general increase in the leaf area in all treatments, however, treatments with no application of N performed significantly lower than those with N application. In the months of August, maximum area was registered in all treatments but treatment P0:N80 mainatained the performance to september.

SPAD readings

Makino (2011) and  Figueiredo  et al. (2014) stated that a SPAD value of 35 in any rice genotype is about the critical leaf N status to attain optimal plant growth and then yield (Figure 3). Nitrogen is known to increase the photosynthetic rate of plant and this is the basis for enhanced biomass production of a crop. Therefore, the application of certain amount of N stimulated plant growth. There was a significant difference between P70:N0 performing lower than other treatments in SPAD values until August (Figure 3), however, towards the ending of the month all treatments were insignificantly different and SPAD values started declining to harvest time. Maximum N  application  alone  was  not  significant until early July, however, later in the months and the rest of other days it was not significantly different from other treatments (Figure 3).

Plant dry matter

In Figure 4 there was a general trend in all treatments for the dry matter to increase through the growth period. N and P treatment combinations did not have significant differences among them until July, but thereafter dry matter  increased  significantly  above P0:N0 and P70:N0 (not significantly different from each other), from July 10 to September 18. The increasing dry matter through the growth period is because of accumulation of nutrients in the plant body and resulting from photosynthesis, as demonstrated by SPAD values (Figure 3).

Rice yield

There was significant difference among the N levels applied to ‘NERICA 4’ (Figure 5). However, the response of ‘NERICA 4’ was relatively high to lower N dose. The findings however, indicate that an increase in N above 60 kg N ha^-1 had similar results in terms of yield (t ha^-1). The results  is   in  agreement  with  Oikeh  et  al.  (2008)  who reported a strong response to fertilizers by NERICA and considered application of 60 kg N ha^-1 and 13 kg P ha^-1 appropriate for low to moderate input smallholder farmers has proved double grain yield as compared to zero fertilizer application. Furthermore, Mghase et al. (2011) suggest that NERICA can be grown in low fertility upland soils in Africa where soil fertility amendment options are not practical due to farmer’s poor access to fertilizers resulting in their low purchasing power.

Yield components

Application of fertilizers generally improved the yield components  performance  of  the  ‘NERICA 4’  (Table 2).

Application of nitrogen with or without phosphorus had no significant effect on the yield obtained. When 80 kg N ha^-1 was applied with 70 kg P ha^-1, the combination produced the highest number of panicle per square meter (378), however, it had less yield when compared with 80 kg N ha^-1 without P applied (Table 2). Application of 80 kg N ha^-1 and 70 kg P ha^-1 produced a significant amount of straw weight (9.4 t ha^-1) compared to 40 kg N ha^-1 with and without phosphorus applied. Nitrogen application with or without P did not have significant effect on both ripening ratio (%) and number of spikelets m^-2. The soils of the experimental field used had sufficient potassium (Table 1). Dobermann et al. (1998) and Dobermann and Fairhurst (2000) reported that K increases the yield components in rice especially the number of grains per panicle, percentage of filled grains and grain weight. On the other hand, the high dose of N applied produced more straw weight compared to lower dose applied (Table 2). The harvest index value obtained when applied 80 kg N ha^-1 was low compared to 40 N kg ha^-1 dose applied (with no phosphorus effect).

Nitrogen use efficiency (NUE) and production efficiency (E_p)

The results indicate  that Nitrogen  Use  Efficiency  (NUE) after application of N fertilizer at different levels of P was high through the growth stages (Table 3). There was relatively high NUE when 40 kg N ha^-1 and 70 kg P ha^-1 fertilizers were applied through different growth stages of rice NERICA 4. The plant when compared to nutritional application level 80 kg N ha-1 and 70 kg P ha-1 at some stages of plant growth (Table 3) obtained a significant amount of N. The findings clearly indicate that nitrogen dose 40 kg ha^-1 was highly recovered and utilized by the plant. In this case, it appears that relatively lower amount of N when applied to the plant, the plant utilization capacity is high than the rate of loss of N into the soil. The increase in the amount of N application decreases production efficiency with linear relation regardless of phosphorus application (Table 4). However, there was a sharp decline in treatments with no P application compared to when P was applied. The likelihood that increasing plant N concentration can induce sterility, this was even experienced in yield reduction when number of spikelets m^-2 was relatively lower yet a high number of panicle was obtained (Table 4).

Generally, there was an increased performance of NERICA  4  when   N   and   P   fertilizers   were   applied compared to no N and P fertilizers application in the upland conditions. Applying N and P fertilizers at 80 and 70 kg ha^-1 respectively, as well as applying N fertilizer 80 kg ha^-1 increase growth parameters and yield compared to other rates of N and P fertilizers applied. However, production efficiency is high with lower rates of N fertilizer application; in fact, it increases further with 0 kg N fertilizer application. Further investigation should be made under uniform and lower soil N content.

The authors have not declared any conflict of interests.

The author would like to express his gratitude to Tsukuba International Center (TBIC) for Technical and material assistance for this experiment. The author is also grateful for the financial assistance rendered by the Japan International Cooperation Agency (JICA).