Effects of Variety and Nitrogen Levels on the Performance of Pearl Millet: Pennisetum Glaucum (L.) R. BR

The trials were conducted in 2010 and 2011 rainy seasons at Usmanu Danfodiyo University, Sokoto Teaching and Research Dry Land Farm, Dundaye, Nigeria to determine the effects of variety and nitrogen levels on performance of (SOSAT c-88 and Zango millet varieties). Data collected were on Stand count, tiller count, plant height and panicle length. Others included panicle weight and grain yield. All data collected were subjected to analysis of variance (ANOVA). The results SOSAT c-88 and Zango millet indicated that millet varieties have no significant effect on both stand and tiller production throughout the sampling periods though nitrogen levels has significant on tiller production at 6 weeks after planting of both 2010 and 2011 cropping seasons. Panicle length was affected by millet varieties and nitrogen levels in both cropping seasons. Improve millet variety Sosat c-88 gave higher grain yield while 80 kg ha and 120 kg ha produced significant higher millet grain yield hence should be practiced.


Introduction
Pearl millet "Pennisetum glaucum (L.)" belongs to the family poaceae, subfamily panicoideae, genus pennisetum and tribe paniceae (ICRISAT, 2006) [1][2][3]. It is largely grown for grain and fodder purpose under those situations where other crops generally fail. Pearl millet as a food crop is limited to the developing countries in Asia, and particularly in Africa and ranked sixth in the world following rice, wheat, corn, barley and sorghum [4]. Millet ranks as the sixth most important grain crop in the world, sustains one-third of the world population and is a significant part of the diet in Northern China, Japan, Manchuria and various areas of former Soviet Union, Africa, India, and Egypt [5]. It is one of the most important cereal crop grown in over 40 countries, predominantly in Africa and Asia, as a staple food, source of feed, fodder, fuel and construction material in hottest, driest, semi-arid and areas where rain fed agriculture is practiced [1].
The average grain yield of millet in Nigeria is about 1.55t ha -1 (FAOSTAT, 2006) which is considered very low compared to the global yield of 3.2t ha -1 [5]. The reasons attributed to this low yield include biotic and abiotic constraints, limited research on development and unfavorable agricultural policies and lack of trained manpower [6]. The nutrient content of pearl millet compares very well with other cereals and millets. It has high protein content with slightly superior amino acid profile. Pearl millet grain contains 13-14 percent protein, 5-6 percent fat, 74 percent carbohydrate and 1-2 percent minerals. It also contains higher amount of carotene, riboflavin (Vitamin B2) and niacin Vitamin B4 [7]. According to Gruhn et al. (2000) [8] insufficient application of nutrients and poor soil management, along with harsh climatic conditions and other factors have contributed to the degradation of soils in sub-saharan Africa which cause plants to fall over and die before harvest.
To achieve optimum grain production, suitable fertilizer program are essential. Until 1982, very little information was available on fertilizer requirement of pearl millet for the savanna region [9]. Use of manure is constrained by its inadequate supply which could be too little to meet the farmers' requirement for all his cropped land and in addition, manure is bulky to handle [10]. Soils in the savanna region of Nigeria are low in nutrient and organic matter content, soil degradation have increased and become a serious threat to agricultural production [11]. Continuous and intensive crop cultivation with near absence of fallow period long enough for restoring soil fertility and low use of external source of inputs are common practice among many small scale farmers in Nigeria [11,12]. Maintenance of soil fertility on farm lands has become very difficult as many of them have limited access to fertilizer (inorganic fertilizer) due to government withdrawal of subsidy on the input which has led to its high cost and unavailability [13].
The soils in this region have shown a continuous decline in fertility and productivity with attendant poor yield of grain crops when fertilizers are not applied [11]. Soils in the tropics are very low in N, P, and K. content and N is the most limiting factor for cereal production [14]. The major reason for declining soil fertility is because of the need to crop the land more intensively due to increased human population and low application of inorganic fertilizers and manures and therefore nutrient extracted by crops are not adequately replaced in the soil [5]. Appropriate selection of good variety is one of the pre-requisite to successful crop production in relation to vigour, growth and yield. Good variety may have highly likely better efficiency in utilizing both above and below ground require resources thereby enhancing it potential economic yield, and in the case of bad variety the reverse is the case. Millet varieties are classified based on the shape of the panicle which includes spreading, loose and compact and erect. Early maturing millet seed contains only small food reserves, fewer leaves, fibrous root systems, the lateral roots are strong and penetrating the soil down 200 cm and the aerial root are well developed [15]. FAO and ICRISAT (1996) [16], reported that the largest variability in maturity period, head size and shape and grain size is found in West Africa, the main centre of origin. Among the land races the Sauna (early, 70 to 90 days), and Sanio (Late, 120 to 180 days) are the most common.  The annual rainfall of the area during the period of trials were 1,157.4 mm and 558.4 mm, respectively [18]. The soil type of the area is predominantly sandy and has been classified as Ustipsamment [19]. Two varieties of millet, improved (SOSAT C-88) and local (Zango), four levels of Nitrogen (0, 40, 80 and 120 kg/ha) were factorially combined and laid out in a Randomized Complete Block Design (RCBD) replicated three times. Data collected were stand count, tiller count, plant height and panicle length. Other data taken include panicle weight Stover and grain yield. Data collected were subjected to analysis of variance (ANOVA) procedure for randomized complete block design (RCBD) using statistical analysis system (SAS 2003) [20]. Duncan's Multiple Range Test (DMRT) was used to separate treatment means where 'F' test shows significant difference between the treatments as described by Gomez and Gomez (1984) [21].

Stand Count
There was no significant difference among the two varieties SOSAT c-88 and Zango millet in terms of stand count in both 2010 and 2011 cropping seasons (Table 1) The varieties were statistically the same in terms of stand count (P > 0.05) 112345, 111851, and 12592, 111974 was obtained in SOSAT C-88 and Zango Millet variety in 2010 and 2011 respectfully. This is in conformity with what was reported by Egharevba et al. (1984) [22] that in cereals (millet and sorghum) stand count hardly affected significantly by variety if both environment and soil condition are favorable. Significant effect (P<0.05) of Nitrogen fertilizer application was found on millet variety stand count in 2011 rainy season when 40 kg/ha Nitrogen was applied. Stand count in both cropping seasons was not significantly affected by interaction between variety and nitrogen fertilizer application (Table 1). Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) ns= not significant, * = significant at 5% level.

Tiller Count
Tiller count of millet at 3WAS and 6WAS as influenced by Variety and Nitrogen levels in 2010 and 2011 cropping seasons are presented in (Table 2). There was no significant difference (P> 0.05) between the varieties on their tiller count at 3WAS and 6WAS in both 2010 and 2011 cropping seasons. This is in line with what was reported by Egharevba et al. (1984) [22] that cereals (millet and sorghum) tiller counts hardly affected significantly by variety if both environment and soil condition are favourable. Results showed no significant effect of nitrogen levels on tiller count at 3 WAS in both 2010 and 2011 seasons. Conversely, tiller count at 6 WAS recorded significant effect of nitrogen fertilizer application in both seasons. All the levels of nitrogen were the same (P>0.05) and differed only from the control. This could be due to split application of fertilizer (nitrogen) at 4 WAS that recorded significantly higher tiller count than the application of nitrogen at 2 WAS and it may probably be that tillering initiation started later after application of first dose of nitrogen fertilizer (2 WAS). This is in line with the findings by Pandey and Sinha (2010) [23] that the application of nitrogen fertilizer at right time (Split) and at higher level enhances many aspects of plant physiological stages involving photosynthesis, root growth and development, tallness or elongation of structural tissues such as stalk in cereals. There was no interaction effect of variety and nitrogen fertilizer application on tiller number (Table 2). Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) Ns= not significant, * = significant at 5% level.

Plant Height
Effects of variety, nitrogen levels and their interactions on plant height at maturity in millet are presented in Table  3. The result showed that there was no significant effect (p>0.05) on variety on plant height in both 2010 and 2011 cropping seasons. But in comparing their means, Zango Millet recorded significantly taller plant than SOSAT C-88 in both 2010 and 2011 cropping seasons. This result is in line with Raemaekers (2001) [24] who reported that local varieties are taller than the improved varieties. Results showed significant differences as influenced by nitrogen fertilizer application on plant height at maturity in both 2010 and 2011 cropping seasons. Nitrogen fertilizer application rate at 120 kgha -1 revealed significantly tallest plants than the remaining nitrogen levels in both cropping seasons (Table 3). This is similar with findings by Pandey and Sinha (2010) [23], that the application of nitrogen fertilizer at right time (split) and higher level enhances many aspect of plant physiological stages involving photosynthesis, root growth and development, tallness or elongation of structural tissues such as stalk in cereal crops. Regarding to variety and nitrogen fertilizer application interaction, no statistical significant (P > 0.05) effects were observed on plant height at maturity. Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) ns= not significant, * = significant at 5% level

Panicle length
There was significant effect (P <0.05) of variety on panicle length in both 2010 and 2011 cropping seasons (Table  4). Zango millet in both trials recorded the longer panicle which was significantly higher than what was recorded in the improved (SOSAT C-88). This finding is in line with Raemaekers (2001) [24] who reported that local varieties are taller than the improved varieties. This also agreed with Jennis (2006) [26], who reported that millet plant vary in panicle length, seed size, seed colour and plant height depending on the cultivars and environment. Nitrogen levels had significant effect on panicle length in both seasons. From 0-40KgNha -1 panicle length increases but at 80-120 KgNha -1 , panicle length increase is at maximum. This is similar with the findings of Pandey and Sinha (2010) [23] that the application of nitrogen fertilizer at right time enhances many aspects of physiological stages involving photosynthesis, root growth and development and elongation of structural tissues such as stalk in cereals. Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) ns = not significant, * = significant at 5% level.
There was interaction effect of variety and nitrogen fertilizer application on panicle length in both 2010 and 2011 cropping seasons as indicated in Table 5. The results revealed that the two varieties of millet showed significant differences of panicle length with levels of nitrogen fertilizer application. With application of 0-40 KgNha -1 , panicle length increases in both variety but at 80-120 KgNha -1 , panicle length is at maximum for Zango millet but not in SOSAT C-88 as shown in Table 5. This finding is in line with what was reported by Hassan and Bibinu (2010) [27] that the interaction effect of Ex-Borno millet variety and 90 kgNha -1 gave the highest panicle length than SOSAT-C-88 and LC-IC9702 millet varieties. Mean with the same letter(s) within a set of interaction are not significantly different using DMRT at 5% level of significance.

Panicle Weight
Effect of variety, Nitrogen and their interactions on panicle weight in 2010 and 2011 cropping seasons were presented in Table 6. There was no significant difference observed on panicle weight as influenced by variety in both 2010 and 2011 cropping seasons. This might implies that both local and improved varieties of millet produced comparable weight of their panicles as shown in table 6. Although, SOSAT C-88 recorded higher mean weight than Zango millet. Panicle weight has been significant affected by Nitrogen fertilizer levels applied in both 2010 and 2011 cropping seasons. From Table 6, 120 kgNha -1 recorded the highest panicle weight in both cropping seasons. This may be attributed to the application of fertilizer in split doses of two halves. Half of the dose as top dressing at two weeks after sowing and other half at four weeks after sowing as recommended by Onwueme and Sinha (1991) [29]. The interaction of variety and Nitrogen fertilizer application revealed no significant effect on millet panicle weight. This is in conformity with what was reported by Egharevba (1978) [25], that there is great intra-plant competition for nutrients and moisture as more heads were produced per stand and this intra-plant competition is detrimental to high grain yields. Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) ns= not significant, * = significant at 5% level.

Stover Yield
Stover yield in 2010 and 2011 cropping seasons had significantly affected by variety as shown in Table 7: However, in comparing their means, the improved variety (SOSAT C-88) had higher Stover weight (yield) than the local variety (Zango millet). This is not in line with the findings of Raemaekers (2001) [24] who reported that improved varieties, has low Stover weight compared to the land races. Significantly Stover yield in 2010 and 2011 cropping season were recorded as influenced by nitrogen fertilizer application as shown in Table 7. The three levels of nitrogen application were at par in both cropping season in term of Stover yield. The levels were however differed significantly from the control (0 kgha -1 ). This may be attributed to the application of fertilizer in split doses of two halves. Half of the dose as top dressing at two weeks after sowing and other half at four weeks after sowing as recommended by Onwueme and Sinha (1991) [28,29]. Higher level of nitrogen fertilizer, 80 kgha -1 and 120kgha -1 produced more Stover yield than 40 kgha -1 and 0 kgha -1 (the control). This may be attributed to the importance of fertilizer application at appropriate time (split) two weeks after sowing (2 WAS) and four weeks after sowing (4 WAS) for increased fertilizer use efficiency and yield (Stover). There were no significant interaction effect (P>0.05) among the factors on Stover yield in both cropping season. Means in a column followed by similar letter (s) are not significantly different at 5% level of significance using Duncan's Multiple Range Test (DMRT) ns= not significant, * = significant at 5% level.

Grain Yield
The result indicated there was significant effect of variety on grain yield from both 2010 and 2011 cropping seasons. SOSAT C-88 from both trials recorded the higher significant differences than the local (Zango millet) Table  8. This variation in grain yield is in accordance with the report by Raemaekers (2001) [24] that improved varieties have higher grain yield per unit area than the local varieties. There were significant effect of nitrogen fertilizer levels on grain yields in both 2010 and 2011 cropping seasons as shown in Table 8. Nitrogen application had significant effect (P<0.05) on grain yield of millet in both seasons. Application of 80 kgha -1 and that of 120kgha -1 were not significant in 2010 and 2011 cropping seasons. Three levels were statistically differed from 40 kgha -1 and the control (0kgha -1 ) in 2010 and 2011 cropping seasons. The control 0 kgha -1 and 40 kgha -1 differed significantly in both seasons. At 0-40 KgNha -1 , there was increase in grain yield but at 80-120 KgNha -1 , there was maximum yield increase. Similar results were reported by Hassan and Bibinu (2010) [27], in their studies that application of 90 kgNha -1 gave grain yield of 2467 kgha -1 , 2153 kgha -1 and 1700 kgha -1 for SOSAT C-88, LC-IC9702 and Ex-Borno varieties respectively. The interaction of variety and nitrogen, fertilizer application showed no significant effect on grain yield in both experiments.

Conclusion
The millet varieties did not varied in tiller production and height variation occurred in panicle length, panicle weight, stover yield and grain yield. There were significant effects of fertilizer rates on all the millet attributes. The improve millet variety couple with higher dose of fertilizer gave better millet performances in all the yield attributes. Therefore improved millet variety and NPK fertilizer application at 120kgNha was recommended.

Data Availability Statement
We declared that this trial was conducted in 2010 and 2011 rainy seasons at Usmanu Danfodiyo University, Sokoto Teaching and Research Dry Land Farm, Dundaye located on latitude 13 o 01 1 N; longitude 5 o 15 1 'E and at an altitude of about 350m above sea level in the Sudan savanna agro-ecological zone of Nigeria. The results written on this paper is originated from the data took on this trial.