|
African Crop Science Journal, Vol. 7. No. 4, pp. 397-406, 1999 Response of maize to phosphorus fertilisation at selected sites in western Kenya N.A. NjuI and A.A.O. Musandu1 Code Number: CS99030 ABSTRACT Maize (Zea mays L.) yields are insufficient to insure smallholder food security in Siaya District, Kenya. Phosphorus (P) deficiency is the main limitation to maize production but little is known about the response of maize to phosphorus application in this region. An on-farm P rate trial was conducted to determine how band P placement and P fertility levels affect maize production. The trial was carried out in two sites. In the experiment, five P rates (0, 10, 20, 30 and 60 kg P2O5 ha-1) were combined with two rates (12 and 40 kg N ha-1) of nitrogen (N). The ten treatments were arranged as a split plot with N as the main plot and P in the sub-plots. Nitrogen at 12 kg ha-1 was a basal application whereas at 40 kg ha-1 it was a combination of both basal application and topdressing (28 kg ha-1). Economic analysis was performed on the grain yields obtained in order to determine the treatment with the most profitable returns. From the results, P application had a significant effect (P<0.05) on maize grain yield. In addition, P and N application significantly (P<0.05) influenced dry matter yield production. The combined grain and dry matter yield effect of N and P application was much better than basal N only. The highest grain yield of 3.75 t ha-1 was obtained with the highest rates of N (40 kg N ha-1) and P (60 kg P2O5 ha-1). The linear coefficient in regression models for grain yield response to P application with and without N topdressing was significant for both sites. The results indicate that N topdressing is necessary for effective P response. It is therefore, recommended that at least 20-30 kg N ha-1 be topdressed. Significant (P<0.05) response to P at 10 kg P2O5 ha-1 was observed. From economic analysis, this low rate of P proved profitable to the farmers. Therefore, it is recommended that 10-20 kg P2O5 ha-1 be applied during planting. Key Words: Maize production, nitrogen, phosphorus placement, Siaya district, Zea mays RÉSUMÉ Les rendements du maïs (Zea mays L.) sont insuffisants pour assurer la sécurité alimentaire des petits fermiers dans le destrict de Siaya au Kenya. La déficience en phosphore (P) est lobstacle principal de la production du maïs, mais peu est connu sur la réponse du maïs à lapplication du phosphore dans cette région. Lessai du taux de phosphore en champ dagriculteur a été mené pour déterminer comment le placement en bande du P et les niveaux de fertilité du phosphore affectent la production du maïs. Lessai a été cenduit en deux sites. Lessai comprenait 5 taux de P (0, 10, 20, 30 et 60 kg P2O5 ha-1) qui étaient combinés avec deux taux dazote (12 et 40 kg N ha-1). Les dix traitements étaient arrangés comme split plot avec N étant la parcelle principale et P la sous-parcelle. Lazote à 12 kg ha-1 était lapplication de base alors que lazote à 40 kg ha-1 était une combinaison dapplication de base et dapplication à la surface (28 kg ha-1). Lanalyse économique a été faite sur les rendements de grains en vue de déterminer les traitements avec les revenus les plus profitables. Des résultats, il a été montré que lapplication du P avait un effect significatif (P<0.05) sur le rendement en grains du maïs. En plus, lapplication de lN et P a influencé significativement (P<0.05) la production du rendement en matiére sèche. Leffet combiné de lapplication dN et du P sur le rendement en matière sèche et le rendement en grains était plus meilleur plus que lapplication de base seul de lN. Le plus haut rendement en grains de 3.75 t ha-1 a été obtenu avec les plus haut taux d N (40 kg ha-1) et de P (60 kg P2O5 ha-1). Le coefficient de régression de la reponse du rendement des graines à lapplication du P avec ou sans lapplication à la surface de lazote était significatif pour les deaux sites. Les résultats indiquent que lN de surface est nécessaire pour une reponse effective de lazote. Il a été ainsi recommandé qu' au moins 20-30 kg N ha-1 soit une fumure de surface. Une reponse significative (P<0.05) au P au taux de 10 kg P2O5 ha-1a été observée. De lanalyse économique, ce faible taux du P a été prouvé bénéfique aux agriculteurs. Il a été alors recommandé que 10-20 kg P2O5 ha-1 soit appliqué durant la plantation. Mots Clés: Production du maïs, azote, placement du phosphore, district de Siaya, Zea mays Introduction Maize (Zea mays L.) is the most important food staple in Kenya. It is widely distributed and the leading food crop of the 30 million Kenyans. It is estimated to contribute more than 20% of total agricultural production, 25% of agricultural employment, 78% of total cereal consumption, 44% of total food energy needs and 32% of the total protein requirements in the country (Ruto, 1992). The Kenya Agricultural Research Institute (KARI) has recognised the economic importance of maize by ranking it fourth among commodity programmes and first among cereals in its 1991 priority setting exercise. Maize is now extensively grown in Kenya and occupies some 1.5 million hectares which is a large proportion of the countrys limited arable land. Small-scale farmers currently produce more than 90% of the maize (Mwenda,1985). Yields on farmers fields are low, averaging only 1.5 t ha-1 compared to research centre yields of 7-8 t ha-1. The farmers grow both local and improved varieties, with more use of improved varieties due to their higher yield potentials. The high-potential agriculture in Siaya District occurs in the northern part of the District (Ukwala, Boro, Ugunja, and Wagai divisions), which was the study area. In 6 out of 10 years, this region receives 1300-1700 mm per year of rainfall, that is bimodal and allows two crops of maize to be grown per year (Jaetzold and Schmidt, 1983). The most common soils are Ferralsols and Acrisols. They are moderately deep and well drained. They are acidic and considered to be of low fertility potential due to their highly weathered and leached nature (Jaetzold and Schmidt, 1983). According to FURP (1987), 20% of the Kenyan soils are acidic. This includes soils in Aberdares, Kwale, Siaya, and some parts of the Rift Valley. Crop production on acidic soils suffers from deficiencies of essential elements (N, P and Ca) and toxicity of Al, Fe and Mn (Ta et al., 1989). Deleterious effects of acidic soils on crops include impairment of root development, which is later manifested in poor growth and delayed maturity (Yamoah et al., 1996). Coupled to that, continuous cropping with little nutrient input depresses the fertility even more. As a consequence, soils are deficient in one or more nutrients, leading to very low yields. The small-scale farmer is therefore locked in the poverty trap, in which the low yields result in low income turnover and hence low or no fertiliser input use. With population pressure resulting in 1-2 ha per household, agricultural intensification is a prerequisite to improved food security. The main hindrances to agricultural intensification in this region include the low purchasing power of the small scale farmer; the inadequate infrastructure for the supply of fertiliser inputs; cost of farm inputs that are beyond the means of small-scale holders and the inefficient use of these limited inputs. If farmers are to break out of this poverty trap, these hindrances need to be addressed. Low soil available phosphorus contents have been identified as a factor which limits crop production in most of Africa (Buresh et al., 1998). The aim of this study was to identify optimal phosphate fertiliser rates under modest nitrogen fertilisation for use by farmers in western Kenya. Materials and Methods Site description and experimental layout. On-farm field trials were conducted at two sites in Siaya District; Yenga and Wagai. The soils were rhodicFerralsols and orthic Acrisols. The soils in the trial sites were red to yellowish/brownish in colour and friable when moist. These soils were acidic with low amounts of total N, organic carbon, phosphorus and exchangeable bases (Table 1). The reasons for this are not clear, but probably the continuous cropping of land with little nutrient returns contributes to the decline in soil fertility characterised by the low levels of organic carbon, nitrogen and phosphorus (Okalebo et al., 1990; 1993). Also, while conducting fertiliser trials in Siaya District, FURP (1987), attributed the very low organic carbon contents (0.63%) to low inherent soil fertility. The trial conducted was a phosphorus rate experiment for the sites in Yenga (rhodic Ferralsols) and Wagai (orthic Acrisols). Table 1. Mean values of the determined soil chemical properties
at the field trial sites
The trial was designed to determine P response and the N and P combination that would maximise maize yields. The experimental design was a randomised complete block design (RCBD) with a split-plot arrangement with two and three replications in Wagai and Yenga, respectively. The main plots were nitrogen at two levels, basal N at 12 kg ha-1 and basal N (12 kg ha-1) plus topdressing (28 kg ha-1) giving 40 kg N ha-1. The split-plots were five rates of P2O5 at 0, 10, 20, 30, and 60 kg ha-1. Maize hybrid H512 was the test crop. The trials were conducted in 1997. Soil and plant sampling. Soil samples for assessing initial fertility status were collected randomly across each experimental site using a 5 cm diameter auger from 0-20 cm depths. Each sample was a mixed composite collected from five spots in each trial site. The composite samples were analysed for various chemical and physical properties (Table 1). The second soil sampling was done after maize harvest following the procedure described above for nutrient analysis especially in relation to phosphorus use efficiency by maize. Plant sampling. Dry matter production and grain yields at harvest were measured. Maize cobs were harvested from an area measuring 10.8 m2 end to end in the three middle rows for grain yield assessment at 15 % moisture content. Dry matter production was measured by cutting all the stover. From the harvested maize grain and stover, sub-samples were taken from each for nutrient P analysis. Other parameters that were determined included plant count at 20 days after emergence (DAE), at harvest and sterile ear formation. Statistical analysis. Data obtained from all experimental variables were subjected to analysis of variance (ANOVA) using MSTATC programme (Bricker, 1989) based on the statistical models used. Means were separated by LSD procedures at P<0.05. Regression analysis using treatment means was performed to estimate the relationship between applied phosphate with and without N topdressing and grain yield. Economic analysis to determine the economic returns for each treatment in the phophorus rate trial was done using the Partial Budget Approach (CIMMYT, 1988). Results and Discussions Maize grain yield with fertiliser application. Table 2 presents the means of maize grain yield as affected by N topdressing and the different rates of P application on the two sites. The average maize yield for the various P treatments ranged from 0.74 to 3.75 t ha-1 when topdressed and 0.85 to 2.47 t ha-1 when not topdressed, with the two soil types. The different rates of P application had a significant effect (P<0.05) on maize grain yield in both sites. On the rhodic Ferralsols, the separation of means showed that with the application of 30 kg P2O5 ha-1 without N topdressing there was a significant effect on the grain yield. However, with N topdressing, significantly higher grain yields were obtained with the application of a minimum of only 20 kg P2O5 ha-1. The highest yield of 3.75 t ha-1 was obtained with 60 kg P2O5 ha-1 and 40 kg N ha-1 topdressing. On the orthic Acrisol, the separation of means showed that the application of 10 kg P2O5 ha-1 with and without N topdressing had a significant effect on the grain yield. Application of P improved maize yield significantly (P<0.01) above the control. The average maize yields in the area without fertiliser use is below 0.5 t ha-1. Information on the yields obtained using fertiliser were scarce and inconsistent but generally lower than 1 t ha-1 since most of the small-scale farmers use small amounts of combined inorganic and organic fertilisers. The increased maize yields observed in this study could be attributed to the inorganic fertiliser applied. Table 2. Effect of N topdressing and P application on grain yield
(t ha-1) of maize for the two sites
Means followed by the same letter in a row or in a column within each site are not significantly different (P<0.05) A comparison of the yields obtained with and without N topdressing on the rhodic Ferralsols at Yenga (Fig. 1 ) indicate that, without P, yields with basal N were higher than those with N topdressing. With 10 kg P2O5 ha-1, the yields obtained with and without 40 kg N ha-1 were the same but thereafter, the yield benefit from applied P was greater with N topdressing than with basal N. Maize responded better to a combination of split N and P application. At 30 kg P2O5 ha-1, the yield obtained with N topdressing was equivalent to that of 60 kg P2O5 ha-1 without topdressing. A saving of 30 kg P2O5 ha-1 when topdressing is done. Consequently, to improve the efficiency of P utilisation on rhodic Ferralsols, nitrogen topdressing is required. This finding is supported by similar studies done by Chardas et al. (1989) and Nguu (1987), who emphasised the need for adequate N and P applications in order to obtain optimum maize yield. Yields with and without N topdressing on the orthic Acrisol at Wagai (Fig. 2) indicate that, with increase in P rates, the difference in grain yields becomes smaller and smaller. At 40 kg P2O5 ha-1, the yields obtained with and without topdressing are the same and thereafter the yield benefit obtained from topdressing becomes slightly higher. These differences are however not significant. Generally topdressing at 28 kg ha-1 seems to have no effect on this particular soil. This could be an indication that, the low rate used for topdressing was inadequate for this soil. Probably if a higher rate of N topdressing was used on the orthic Acrisols it would improve the yields. The results indicate that, treatments which received P and were later topdressed had higher yields. The increase in yields in the topdressed treatments could be related to enhanced P uptake due to N application. An enhancement effect of N on P uptake occurs particularly when N is applied as ammonium-N, due to the stimulation of anion uptake in response to cation (NH4+) uptake. According to Fan and Mackenzie (1995), banding TSP with urea reduces soil pH increases from urea hydrolysis and depresses NH4+ concentrations. In the acidic soils used in this study, acidification by NH4+-N is not likely to be the explanation for enhanced P uptake but probably the stimulation of NH4+-N on root growth. This is because high P supply is particularly important on root growth. Le Mare et al. (1987), observed that application of rapidly decomposable organic material to soils could lead to a supply of N and K which enhances plant availability and uptake of P and also reduces permanent adsorption of P (Iyamuremye and Dick, 1996; Ohno and Cranwell, 1996). Maize dry matter production (DM) with fertiliser. Table 3 shows the means of dry matter production in the phosphorus rate trial as affected by N topdressing and different rates of fertiliser P application on the two soils. Table 3. Effect of N topdressing and P application on dry matter production
(t ha-1) at the two sites
Means followed by the same letter in a row or in a column within each site are not significantly different (P<0.05) Nitrogen and phosphate application had a significant effect (P<0.05) on dry matter yield. The average dry matter yields ranged from 3.10 to 11.2 and 2.69 to 9.58 t ha-1 with and without topdressing, respectively, for the two soil types. The increase in dry matter on both soil types was not consistent with increasing P rates except on rhodic Ferralsol where N topdressing was done. The treatments receiving 40/60 and 40/0 kg N/P2O5 ha-1 produced the highest and lowest dry matter, respectively, on the rhodic Ferralsols whereas highest and lowest dry matter on the orthic Acrisol were produced by 40/30 and 12/20 kg N/P2O5 ha-1 (Table 3). There were significant differences between and within soil types. The treatments that received phosphorus had higher dry matter yields than the control except where 20 kg P2O5 ha-1 with basal nitrogen was applied at the site in Wagai. With N topdressing, the DM yields were generally higher than those with basal N only. Of the two sites, rhodic Ferralsols had higher DM yields as compared to orthic Acrisols. On both soil types, the significantly higher production in dry matter yield where P had been applied as compared to the control (0 P205 kg ha-1) is as a result of phosphorus deprivation. P deprivation decreases fresh weight and chlorophyll content of shoots and induces the early initiation and accelerated remobilisation of N from old leaf blades (Usuda, 1995). Since this phenomenon is a critical feature of leaf senescence, it is suggested that premature leaf fall could have led to the low dry matter yield. In response to N topdressing, increase in dry matter production in the treatments that received P could be attributed to the synergistic effect resulting from the combination of N topdressing and P application. According to Tripathi and Singh (1982), the combination of nitrogen and phosphorus leads to an increase in dry matter accumulation. In this trial, pH (KCl) comparison (data not presented) of the soils at the sites where the experiments were conducted revealed that the orthic Acrisols were more acidic than the rhodic Ferralsols. This could have contributed to the lower DM produced on the orthic Acrisols since acidity has deleterious effects on root growth and function and hence is a constraint to plant growth. The combination of poor root growth and low mobility of P may lead to deficiencies even if extractable P is present in what would otherwise be adequate amounts (Rowell, 1995). Economic returns to maize fertilisation. Partial budgets performed on maize grain yield for both soil types in the phosphorus rate trial are presented in Tables 4 and 5. The factors that determine the profitability of a fertiliser use are its cost, transportation and application, and the market value of the production. Actual market data were obtained following a survey of the prices in the different shops within the study area and used in this analysis (Table 3). Table 4. Partial budget for the fertiliser experiment on H512 maize on orthic
Acrisol at Wagai in 1997
1 kg of grain threshing cost @ Kshs. 0.40 Cost of TSP fertiliser Kshs 960.00 for 50 kg 1 kg of grain transport cost @ Kshs. 0.60 Cost of KCl fertiliser Kshs 800.00 for 50 kg Transporting 50 kg fertiliser bag =Kshs. 60.00 1 kg of Furadan @ 320.00 Wage rate for 1 manday = Kshs. 70.00 1 kg of Hybrid seed @ 69.00 1 kg of grain harvest cost @ Kshs. 0.20 1 kg of local seed @ 20.00 1 kg of grain bagging cost @ Kshs. 0.30 1 kg of Dipterex @ 90.00 Yield adjustment 20% Table 5. Partial budget for the fertiliser experiment on H512
maize on rhodic Ferralsol at Yenga in 1997
*Maize prices and cost of inputs as in Table 3 Table 6 shows the return ratio on investment (RR) in the P rate trial for both sites.The return ratio reveals how the net benefits from an investment increase as the amount invested increases. This comparison is important to farmers because they are interested on seeing the increase in costs required to obtain a given increase in net benefits. The percentage shows the RR. In changing from the conventional practice (Treatment 1, Control) encountered in this region to use of P fertiliser at the rate of 10 kg P2O5 ha-1 (Treatment 2), farmers would get a RR of approximately 28%. This means that for every Kshs. 1.00 invested in phosphorus and its application, farmers can expect to recover the Kshs. 1.00, and obtain an additional Kshs. 28. Experience and empirical evidence have shown that for the majority of the situations the minimum acceptable rate of return to farmers will be between 50 and 100 %. If the technology is new to the farmers, and requires that they learn some skills, a 100 % minimum rate of return is a reasonable estimate. If a change in technology offers a rate of return above 100 % (which is equivalent to a 2 to 1 return) it would seem safe to recommend it. All the above percentages offer an acceptable rate of return. The change from treatment 1 to 2 offers earnings at an attractive rate of return and is the best recommendation to farmers given that they have not been using P fertilisers. With time the farmers can target further earnings at a higher rate of P fertilisation as indicated by the dominating treatments. A dominating treatment is any treatment that has net benefits higher than those of a treatment with lower variable costs. Table 6. Return ratio on investment
I assumes baseline addition of 12 kg N ha-1 This recommended rate of return is sufficiently profitable and attractive to farmers whose minimum return is only 100%. At the point of unit marginal return, the profit/ha is the largest, so that any move from there in the direction of a more efficient use of external inputs is inevitably at the expense of net profits. It would be unrealistic to assume a static price relationship between input and output. Sometimes farmers desperate for cash are forced to sell at nearly half the official price in years of abundant harvest. Assuming the field price falls by half from Kshs. 18.50 to Kshs. 9.25, the minimum yield response would increase from 62 kg ha-1 to 124 kg ha-1 (1147/9.25). It would thus reduce the average marginal rate of return by half. This rate of return is still acceptable to farmers given the minimum acceptable return of 100%. It is therefore apparent that, the profitability of this new practice is stable over a wide range of output price changes. Conclusions Hybrid maize varieties have not been widely adopted in Siaya District because farmers consider their own local varieties to be superior to hybrids under the production system practised and due to taste preference and storage factors. The grain yield response to phosphorus application with nitrogen topdressing resulted in improved yields. For farmers in this region, it is recommended that the use of hybrid seed under the suggested 12 (basal) and 28 (topdressing) kg N ha-1 and 10 kg P2O5 ha-1 fertilisation conditions be practiced. The findings of this study also indicate that a wide variety of factors influence the rate of fertiliser use and that the importance of these factors vary from site to site, thus calling for implementation of specific solution to each site. The large gap between farmer yields and what is feasible with the available technology can be reduced by a more intensified extension effort, focused on crop husbandry, reduction of losses during storage, and appropriate policy actions to motivate farmers. Acknowledgement This research was funded by The Rockefeller Foundation's Forum for Agricultural Resource Husbandry. We thank Ms. Carol Maina and Dr. Paul Woomer for their assistance with the manuscript preparation. References
©1999, African Crop Science Society The following images related to this document are available:Photo images[cs99030a.jpg] [cs99030b.jpg] |
|