Responsiveness of Soil Series in Southwestern Nigeria to Single Super Phosphate Fertilization

Application of Single Super Phosphate (SSP) fertilizer in soils with inadequate phosphorus (P) should elicit crop response. However, there is an indication of inconsistent response behaviour to SSP fertilizer in some Nigeria soils despite their low available P. This study identified major soil types in Southwestern Nigeria that are unresponsive to P fertilizer application under screenhouse condition and soil properties that are related to this unresponsive behaviour were determined. Soils from twenty locations in Southwestern Nigeria were selected to evaluate their response to four phosphorus fertilizer rates with a control replicated three times in a factorial arrangement laid out in randomized complete block design. Results indicated that most of the locations responded to initial P application except for Fashola 2 which produced significantly higher maize shoot yield in control soils. There were five other locations where no-response was observed. Multiple regression analysis showed that phosphorus sorption index was related to sand, organic matter and clay. Phosphorus accumulation in soil P reserves may explain the behaviour of the P-unresponsive soils irrespective of their available P levels.


Introduction
Increasing demand for food crop production for the ever rising human population still remains a challenge of modern agriculture.However, external P application remains a key to sustainable production under tropical conditions (Iyamuremye et al., 1996;Zhenoy & Kalagudi, 2005).Studies have shown that many Nigeria soils have available phosphorus (P) levels below critical level of < 8 mg kg -1 and hence require P fertilizer application (Agbenin, 2003;Egwu et al., 2010).However, soils with high available P levels are mainly associated with high soil organic matter due to cropping system and management practices (Potarzycki et al., 2004).The application of P fertilizer has been shown to have tremendous effect on the soils especially the highly weathered tropical soils (Brady & Weil, 2008).FAO (2008) has reported significant crop response with application of native phosphorus fertilizer from rock phosphate in acid soils.Similarly, the use of Single Super Phosphate (SSP) has also produced significant crop response albeit with cases of inconsistent response (Njoku & Enwezor, 1988).Response to P fertiliser and ultimately maize yield are dependent on factors that influence plant growth such as soil.Different soils based on the variation in their physicochemical properties (Federal Fertilizer Department Agriculture [FFDA], 2012) lead to different crop response potentials.Differential soil responses are largely due to inherent soil characteristics, P forms and management practices (Valizadeh et al., 2002), hence, fertiliser recommendations should be designed to take cognizance of sorption characteristics of the soil.
Although, quite a number of information has been documented on responses to P in major food crops in Africa (Cottenie, 1981;Njoku & Enwezor, 1988;Okalebo et al., 1996;FAO, 2008), responses observed in Nigeria were attributed to cropping season, soil heterogeneity, and crop residue removal after clearing (Kang & Osiname, 1979).In some Southeastern Nigeria soils (Osodeke et al., 2004), no response to P was attributed to high soil P levels in the locations studied while other locations with low soil P content, required unusually higher P application rates to elicit response to P fertilizer (Enwezor, 1977).Similarly, pockets of soils with inconsistent soil behaviour have been observed in some Western Nigeria savanna and forest soils (Adepetu & Corey, 1977;Kang & Osiname, 1979;Ayodele, 1986).fertiliser recommendation for this group of soils (Ayodele, 1980).Little or no study in the past has focused on identifying P-unresponsive soils and little or no attempt has been made to identify the soil properties responsible of unresponsiveness to P as well as the appropriate management practices.Generalization on the soil properties that influence its unresponsiveness can not best explain its unusual behaviour.Very few or no studies have taken into consideration variations in soil parent materials as well as soil characteristics.There is a possibility that, P released from sorption sites during cropping may influence this soil unresponsiveness.The inability of soil test extractants to measure residual P in these unresponsive soils and/or the presence of numerous sorption sites that exist in these soils may also contribute to soil unresponsiveness to applied P. The lack of crop response to P application could also be related to geology i.e. soil parent material and soil management practices.Hence, the need to study soils representative of various locations and their parent materials, the agro-ecology and soil management practices.It is believed that soils with inconsistent behaviour constitute a considerable part of cropland in Nigeria.Hence, identifying these soils with low response potentials would elicit suggestions on how to improve P release and P uptake by crops.Information provided from this study, will also improve existing P calibration i.e standard soil P test value at which we expect response to P fertilizer recommendations.
The objectives of this study under screenhouse conditions using maize as a test crop are (i) to identify soils in Southwestern Nigeria that are unresponsive to P fertilizer and (ii) to determine soil properties that are related to their unresponsiveness.

Soil Sample Collection
A total of seventeen soil samples of different soil series (from twenty locations, belonging to a variety of parent materials, vegetation and soil management practices were collected from various locations in Lagos, Oyo, Ogun, Osun, Edo and Ondo states of Nigeria (Table 1).Soils were sampled from farmers' fields which have been cropped for two years up to eight years, in fallows, and other ones in the national Agricultural Research Institute (ARI) fields established on coastal swamp, forest and derived savannah vegetations assembling the coastal plain sands, basement complex and sedimentary parent materials.The locations of the selected sites are shown in Figure 1.Bulk top soil (0-20 cm) samples were collected from 5-6 points in each location.Soil samples were air dried, sieved through 2 mm sieve.Composite samples were taken from the bulked soil samples for physical and chemical analysis.The bulk soil samples were labeled and stored for laboratory analysis and filling of plastic pots for screenhouse work.

Soil Analyses
Soil analyses were carried out in the laboratory of the Department of Agronomy, University of Ibadan, Nigeria.Soil physical and chemical parameters were analysed as described by Udo and Ogunwale (1986) in the laboratory Manual for the analysis of soil, plant and water.Particle size analysis into sand, silt and clay was done using the hydrometer method.Soil pH was carried out on a 1:2 (soil:water) ratio after 15 minutes equilibration period using a glass electrode calibrated to buffers pH 4, 7 and 9. Organic carbon was determined by the dichromate wet oxidation method.Exchangeable acidity was extracted with 1 N KCl and titrated against 0.01 N NaOH.The Effective cation exchange capacity (ECEC) was obtained by summing up of the total exchangeable base (TEB) and total exchangeable acidity (TEA).Total nitrogen was determined by the Kjeldahl method.Exchangeable cations (Ca, Mg, K and Na) were extracted with I N NH 4 OAc (pH 7) at a soil: extraction solution ratio of 1:10 for 15 minutes.Trace elements i.e.Fe, Mn, Cu and Zn were extracted with 0.1 N HCl.and quantified using Atomic Absorption Spectrophotometer (Model 210/211 and bulk scientific, USA).Bray P-1 was determined by the Molybdenum blue colour method.

Determination of Free Fe and Al Oxides
The extraction method with 0.2 M ammonium oxalate acid and oxalic acid solution (Blakemore et al., 1987) was used to extract ammonium oxalate extractable iron and aluminium (Fe o and Al o. ).They were quantified using inductively coupled plasma -atomic emission spectrum at 258.8 and 396.1 nm wavelength, respectively.Dithionite citrate bicarbonate (DCB) extraction method (Blakemore et al., 1987) was used to extract dithionite citrate bicarbonate extractable iron and aluminium (Fe d and Al d ) in the extracts and quantification done by inductively coupled plasma -atomic emission spectrum.

Determination of Phosphorus Sorption Index (PSI)
Modified single point adsorption (SPA) procedure (Bauche & Williams, 1971;Sims, 2000) was carried out to provide a quick estimate of the phosphorus sorption capacity of the soils screened and the information on the phosphorus sorption index (PSI) of the soils.Phosphorus sorption index was calculated by dividing the P sorbed with the log of P in the equilibrium solution as follows: (1) Where c = P concentration at 75 mg P l -1 .

Screenhouse Studies with Single Super Phosphate
The rates of single super phosphate (SSP) applied per hectare were 0, 30, 60, 90, and 120 kg P 2 O 5 equivalent to 0.25, 0.5, 0.75 and 1 g per 6cm-diameter plastic pot containing 3 kg of soil with drainage holes.Each rate was comprised of 20 plastic pots arranged in completely randomized design with three replications i.e. 60 plastic pots for the three replications.A total of three hundred experimental pots were used for all the soils.Urea was applied at the rate of 120 kg N ha -1 i.e. 0.37 g/plastic pot to supply adequate N to the maize crop in addition to SSP rates and thoroughly mixed with the soil.Pottasium was applied at 30 kg K20 ha-1 i.e 0.06 g/plastic pot and Zn was applied at 4 kg Zn ha-1 i.e 0.75 ml/plastic pot.The experiment was a factorial laid out in randomized complete block design.The soils were sown with four maize seeds (TZE comp 3DT) after watering to field capacity (0.53-0.80 L).Maize plants were thinned to two seedlings, five days after sowing and subsequently grown for six weeks.The harvesting of maize shoots was done after six weeks of growth and oven dried at 65 o C until constant weight and stored in labeled envelopes.The second cropping of maize was carried out after removal of the roots from the plastic pot soil.Maize shoot dry matter, PSI and soil properties data were subsequently analyzed statistically using the analysis of variance (ANOVA) and multiple regression analysis in Genstat (Genstat version 10.3DE).

The Physical and Chemical Characteristics of the Soils Studied
Soils were analysed prior to using them to fill the plastic pots.Results indicated that total nitrogen (N) was below critical limit of 1.5 g kg -1 .K, Ca and Mg were above FFDA's, (2012) critical limits of 0.15, 0.2, and 0.2 cmol kg -1 , respectively.The physical and chemical characteristics of the top soil (0−20 cm) are shown in Table 2.The soil properties varied with the soil type.Soil pH (H 2 O) ranged from 5.0 to 6.4 with an average of 5.6 ± 0.54 for top soil while pH (KCl) ranged from 3.7 to 5.8 with an average of 4.7 ± 0.62.Therefore, the soils ranged from strongly acid to slightly acid.Mean organic carbon was 19 ± 9.6 g kg -1 ranging from 11 to 54 g kg -1 for 0-20 cm top soil while the total N varied from 0.1 to 1.26 g kg -1 with an average value of 0.47 ± 0.3 g kg -1 .The classification of exchangeable bases in the decreasing order of presence in the top soil was Ca, Mg, K, Na.Mean values for Ca, Mg, K, and Na were 2.47 ± 1.6, 1.2 ± 0.02, 0.85 ± 0.24, and 0.2 ± 0.02 cmol kg -1 .The ECEC ranged from 3.3 to 9.1 cmol kg -1 (mean = 5.6 ± 1.5 cmol kg -1 ) and the total exchangeable acidity from 0.2 to 1.4 cmol kg -1 (mean = 0.54) for 0−20 cm top soil.The sand fraction was dominant in all soils and had a mean value of 81.5 ± 7.8% with a range of 69.2 to 92.0%.The silt fraction ranged from 3.4 to 19.4% with a mean of 9.3 ± 4.5% while the clay fraction had a mean value of 9.2 ± 3.8% and the corresponding range values of 4.0 to 15.4%.The soils were broadly of sand, loamy sand and sandy loam textures.

Correlation between PSI and Some Soil Properties
PSI was negatively correlated (-0.77, -0.84, -0.60, p ≤ 0.05) with Fe o , Al o and Fe d respectively, but positively correlated (sand 0.60, p ≤ 0.05) with sand in the top soil.

Screenhouse Experiments
Figure 2 and 3 shows maize response to SSP in the first cropping.Significant difference was obtained in maize shoot dry matter response to P levels x soil types interaction (p <= 0.001, p = 0.001 respectively).Except with no P application, maize shoot dry matter in the first cropping was generally higher than the second.At both croppings, Egbeda 2 soil produced the highest maize shoot dry matter with no P application while the lowest shoot dry matter was obtained at NIFOR 1 soil in the first crop and Ikoyi soil in the second.Maize plants of the first crop with 30 kg P 2 O 5 ha -1 , produced the lowest in Owena 2 soil while Iwo 1 had the highest.However, in the second maize crop, Fashola 2 produced the lowest while Egbeda 2 had the highest.Maize plants with SSP application of 60, 90 and 120 kg P 2 O 5 ha -1 , produced the lowest and highest maize shoot dry matter in the second maize crop at Ikoyi and Egbeda 2 soils, respectively.Controls or soils which didn't receive SSP produced significantly higher maize shoot dry matter comparable to higher phosphorus of 30 to 90 kg P 2 O 5 ha -1 at Fashola 2 and Ikenne 2 soils in the first crop and in Fashola 2, Iwo 2, Ikenne 1, Owo 2 and Egbeda soils in the second maize crop.Owena 1 and Owena 2 soils with 120 kg P 2 O 5 ha -1 produced significantly higher maize shoot dry matter in the first cropping however, it was not significantly different from 60 and 90 kg P 2 O 5 ha -1 for Owena 1 soil and 90 kg P 2 O 5 ha -1 for Owena 2 soil in the second cropping.Based on the PSI of 51% and 52%, moderate P application may be required for Owena 1 and Owena 2 soils.NIFOR 1, NIFOR 2, Fashola 1, Fashola 2, Ikoyi, Iwo 1, Iwo 2, Auchi, Ilaro and Epe soils with PSI levels between 60 -90%, may require high P application.However, with Egbeda 1 soil, significantly higher maize shoot dry matter was obtained from 120 kg P 2 O 5 ha -1 in the first cropping while 30 kg P 2 O 5 ha -1 produced significant higher maize shoot dry matter in the second cropping.

Discussion
The high maize shoot dry matter in response to 120 kg P 2 O 5 ha -1 observed at Egbeda 1 soil in the first cropping (Egbeda soil series) is in line with the result of Kang and Osiname (1979) whose top soil texture appears to be the same with the soil in this study.Even though national fertilizer recommendation for low P soils in Nigeria is 60 kg P 2 O 5 ha -1 (FFDA, 2012), Egbeda soil series may require high P application to build up and maintain available P levels of 14 mg P kg -1 (Kang & Osiname, 1979).However, there is need for caution in applying P fertilizer in Egbeda soil based on the PSI (53%) and the residual P effect obtained at 30 kg P 2 O 5 ha -1 .Using the maize shoot dry matter production as a criterion for selecting P-unresponsive soils, higher maize shoot dry matter from control soils over soils with P applications was considered as P-unresponsive and this formed the basis for identifying P-unresponsive soils.From the first cropping, it was observed that NIFOR 2, Fashola 2 and Ikenne 2 soil produced higher maize shoot dry matter over control soil.Fashola 2 and NIFOR 2 soils had moderate P levels (Cottenie, 1980;FFDA, 2012).The high maize shoot dry matter from soils with no SSP addition for Fashola 2 confirms previous observation of a lack of P response in Fashola soils by Sanginga et al. (1995) and Sanginga et al. (2000).Higher maize shoot dry matter from soils with zero SSP application compared with lower maize shoot dry matters from soils where SSP was applied was also observed at Epe soil with predominant sand texture in the second cropping (Brady & Weil, 2008).Previous study (Adepetu & Corey, 1977) has indicated no P response at Iwo 1 soil (Iwo soil series).
Three of the selected P-unresponsive soils (Fashola 2, Owo 2 and Egbeda 2) are major agricultural soils derived from basement complex parent materials which are very variable in soil physical and chemical properties.Ikenne 1 and Ikenne 2 soils are derived from sedimentary parent material which is moderately variable, while NIFOR 2 soil derived from coastal plain sands parent material is least variable.The six selected P-unresponsive soils fell into the medium and high P retention class.Four of the six selected P-unresponsive soils produced higher maize shoot dry matter in the residual cropping.This may suggest that residual/sorbed P may influence P-unresponsiveness.
There was positive and significant relationship between P sorbed and sand particles.Free Fe oxides can exist as coatings on sand soils (Bolland et al., 1996) and as such P was not directly sorbed onto Fe and Al oxide in the soils of this study.Earlier studies by Ojanuga et al. (1975) and Uyovbisere and Chude (1995) have also reported negative relationship between sorption parameters and free Fe oxides in some Nigerian soils.The multiple regression equation relating percent PSI to soil parameters is given as: PSI (%) = -42.9+ Sand (0.1207) + organic matter (0.288) + clay (0.1712).
Multiple regression analysis performed to look for the correlation between soil properties and phosphorus sorption index indicated that sand fraction, organic matter and clay content were the most associated variables.
(R 2 = 0.88, p= 0.001)The relative contribution of the soil variables were sand (31%), organic matter (28%), clay (27%), Bray P-1 (8%) and pH (H 2 O) (5%).Criteria used was based on the F probability of each individual variable (soil properties) added to the model.This study ranked clay as third contributor to P sorption unlike that of Axt & Walbridge (1999) who identified clay and silt as best predictors of PSI using stepwise regression for some upland soils.

Conclusion
These selected P-unresponsive soils may have P accumulation in soil P sorption sites that is made available in the residual cropping.Thus, residual/sorbed P may influence the response potentials of these soils irrespective of their native available P levels.Efforts to improve P availability in these soils can either mask P deficiency or accumulate P and consequently, produce irregular response to P. For NIFOR 2, Fashola 2, Ikenne 1, Ikenne 2, , Owo 2, and Egbeda 2 where inconsistent and no response to P fertilizer was observed, there is need for detailed study of the chemistry of these P-unresponsive soils.

Figure 1 .
Figure 1.Locations of soils sampled

Figure 2 .Figure 3 .
Figure 2. Effects of different levels of phosphorus on maize shoot dry matter of twenty different locations under screenhouse conditions for first cropping (Extensions on bars represent SED AT P < 0.05)

Table 2 .
Physical and chemical properties of the top soils (0-20 cm)

Table 3 .
Phosphorus sorption index (PSI) and some soil properties of the top soils screened for unresponsiveness to phosphorus Dithionite extractable Fe; Al d -Dithionite extractable Al; Fe o -Ammonium oxalate extractable Fe; Al o -Ammonium oxalate extractable Al; SED -Standard Error of difference of Mean.