INTRODUCTION
It has long been recognized that soil acidity is one of the most serious challenges to agricultural production worldwide in general, developing countries in particular. Area affected by acidity is estimated at 4 billion ha, representing about 30% of the total ice-free land area of the world (Sumner and Noble, Reference Sumner, Noble and Rengel2003). It is mostly distributed in developing countries, where population growth is fast and demands for food and fibre is increasing. In Ethiopia, about 41% of the total land mass is affected by soil acidity and 33% of this area has Al-toxicity (Schlede, Reference Schlede1989). The highlands of Ethiopia (areas >1500 m above sea level) are the most affected region by soil acidity. The cause of soil acidity is high amount of precipitation that exceeds evapo-transpiration that leaches appreciable amounts of exchangeable bases from the soil surface. As a result, most of the soils have a pH range of 4.5 to 5.5, contain low organic matter (<20 g kg−1) and low nutrient availability (Temesgen et al., Reference Temesgen, Getachew, Tesfu, Tolessa and Mesfin2011). Increased soil acidity causes solubilization of Al, which is the primary source of toxicity to plants at pH below 5.5, and deficiencies of P, Ca, Mg, Mo, N, K and micronutrients (Ernani et al., Reference Ernani, Bayer and Maestri2002; Kariuki et al., Reference Kariuki, Zhang, Schroder, Edwards, Payton, Carver, Raun and Krenzer2007; Mesfin, Reference Mesfin2007). Theoretically, soil acidity is quantified on the basis of hydrogen (H+) and aluminium (Al3+) concentrations of soils. For crop production, however, soil acidity is a complex of numerous factors involving nutrient/element deficiencies and toxicities, low activities of beneficial microorganisms, and reduced plant root growth which limits absorption of nutrients and water (Fageria and Baligar, Reference Fageria, Baligar and Rengel2003b). However, Al3+ toxicity is one of the major limiting factors for crop production on acid soils by inhibiting root cell division and elongation, reducing water and nutrient uptake (Wang et al., Reference Wang, Harsh, Guo-Ping, Neville and Zhou2006), poor nodulation or mycorrhizal infections (Kochian et al., Reference Kochian, Hoekenga and Piñeros2004; Delhaize et al., Reference Delhaize, Gruber and Ryan2007), consequently leading to poor plant growth and yield of crops.
Several agricultural practices have been recommended to overcome the problem of tropical acid soil infertility worldwide. Amongst them, the most common and widely used method is liming, which is defined as the application of ground calcium and/or magnesium carbonates, hydroxides and oxides aiming at increasing the soil pH, modifying its physical, chemical and biological properties (Edmeades and Ridley, Reference Edmeades, Ridley and Rengel2003). Because of its great ameliorative effect, lime is commonly called the foundation of crop production or ‘workhorse’ in acid soils (Fageria and Baligar, Reference Fageria and Baligar2008a). Upon liming, numerous authors have reported decreases of Al in the soil solution as well as in the exchange complex (e.g., Álvarez et al., 2009; Delhaize et al., Reference Delhaize, Gruber and Ryan2007; Prado et al., Reference Prado, Natale and Rozane2007), improved soil structure (Crawford et al., Reference Crawford, Singh and Breman2008), significant yield increases (Buri et al., Reference Buri, Wakatsuki and Issaka2005; Fageria and Baligar, Reference Fageria and Baligar2008a), increases in phosphorus (P) uptake by plants (Fageria and Santos, Reference Fageria and Santos2008; Haynes, Reference Haynes1982), higher abundance and diversity of earthworms (Bishop, Reference Bishop2003); and improved organic matter decomposition and nutrient mineralization (Bradford et al., Reference Bradford, Tordoff, Eggers, Jones and Newington2002). Acid soils have high P fixing capacity and applications of both lime and P fertilizer is frequently required for successful crop production. In addition to Al3+ toxicity, low P availability to crops is also cited as another factor limiting crop production on acid soils (Barber, Reference Barber1995). Therefore, P deficiencies and Al3+ toxicities often occur simultaneously in many acid soils and are thought to be responsible for poor crop yields in acid soils. However, as the fixed P would be released for plant uptake after liming, the amount of additional P added has to be determined experimentally (Waigwa et al., Reference Waigwa, Othieno and Okalebo2003).
Tolerance to Al toxicity or acidic soils differs greatly amongst cereal species, and barley is usually considered the most susceptible member of the Poaceae (Garvin and Carver, Reference Garvin, Carver and Rengel2003). The Al tolerance order as reported is maize>rye>triticale>wheat>barley (Polle and Konzak, Reference Polle and Konzak1985), rye>oats>millet>bread wheat>barley>durum wheat (Bona et al., Reference Bona, Wright, Baligar and Matuz1993) and rice>maize>pea>barley (Ishikawa et al., Reference Ishikawa, Wagamatsu, Sasaki and Manu2000). The barely crop is more tolerant to alkaline soils, and a pH of 6–8.5 is generally acceptable. Fifty per cent yield reductions in barley were reported when grown in naturally acidic soils (pH 4.9) compared to that grown on pH 5.8 (Gallardo et al., Reference Gallardo, Borie, Alvear and Baer1999). In Ethiopia, barley is the most important cereal crop with total area coverage of 1 046 555 ha and total annual production of about 1.7 million tons during main growing season (CSA, 2011).
Even though, the problem of soil acidity is considered to be one of the major bottlenecks to barley production in the highlands of Ethiopia, it is still a problem that has not been addressed in depth and yet widespread decline in barley productivity has been observed in the country. Liming is not a practice used for barley production in the highlands of Ethiopia. To mitigate the negative effects soil acidity, and low soil fertility, traditionally farmers practice barley–fallow production system. Such practice seems to degrade the soil resource in the long-run and will not be sustainable due to severe soil erosion from bare fallows. In this production system, application of P fertilizer is not commonly practiced by farmers. To protect a potential loss of grain yield, at least a maintenance application of 10 kg P ha−1 is needed for responsive sites that had soil test P levels above the critical levels (Getachew and Berhane, Reference Getachew and Berhane2013). However, it varies depending on the soil type, preceding crop, barley variety used and the prevailing environmental conditions. Therefore, Al3+ toxicity and P deficiency are the two major factors limiting barley production on acid soils, and are partly responsible for the seasonal food shortages in some parts of the highlands of Ethiopia. Hence, liming and P fertilization appears to be amongst the most important operations required to boost barley productivity in the highlands of Ethiopia. However, there are no or little research results available in literature on the effect of lime and P fertilizer in barley–fallow-based production systems in the highlands of Ethiopia. In response to this, field experiments were conducted to fill the gap with the following objectives: (1) quantify changes that would occur on soil pH and exchangeable Al3+ as a consequence of applications of different rates of lime; (2) assess the effects of different levels of lime and P on grain yield and yield components of barley; (3) determine the optimum combinations of lime and P fertilizer required to increase barley productivity.
MATERIALS AND METHODS
Site descriptions
The study was carried out during 2010 and 2011 cropping seasons on an experimental field located at Watabecha Minjaro (9° 05′ 55″ N, 38° 36′ 21″ E, altitude 2565 m a.s.l) in the central highlands of Ethiopia (Figure S1). The site is typically characterized by flat plains with cool subtropical climate. Annually, about 1100 mm rainfall is received, and the rainfall pattern is bimodal in distribution (short and long rainy season). The experiment was conducted during long rainy season which extends from June to September. The soils are classified as Nitisols with deep, red, well-drained tropical soils (FAO, 2006). In these soils, soil acidity (pH < 5) and associated low nutrient availability are constraints to crop production. The major cereal crop grown in the area is food barley. The crop is stable food and source of cash income for the majority of the people living in higher altitudes in Ethiopia.
Field operations, experimental design and treatments
Before the start of the experiment, land preparation was uniformly performed across all plots by tractor mounted moldboard plough to 30 cm soil depth. Subsequent tilling operations were done by harrowing to about 10 cm depth by conventional tillage. Lime was evenly applied to treatment plots one month ahead of planting. A high quality limestone (98% CaCO3, 99.5% <250 µm in diameter) was used. All plots were hand weeded at 30 and 60 days after sowing. Faba bean (Vacia faba L.) was the preceding crop for barley. The amount of lime that was applied at each level was calculated on the basis Al concentration of the site (Kamprath, Reference Kamprath and Adams1984) assuming that one mole of exchangeable Al3+ would be neutralized by equivalent mole of CaCO3. The mean Al3+ of the study site was 1.32 Cmol kg−1, and the corresponding lime rates were 0, 0.55, 1.10, 1.65 and 2.2 t ha−1. The liming experiment consisted factorial combinations of five levels of lime (0.0x, 0.5x, 1.0x, 1.5x and 2.0x) based on exchangeable Al3+ content of the soils, and four levels of Phosphorous (0, 10, 20 and 30 kg P ha−1) in a randomized complete design. A total of 20 treatments in three replications were used in the experiment. P was broadcast applied in the form of TSP (triple superphosphate), and the recommended rate of nitrogen for barley i.e. 50 kg ha−1 was applied uniformly to all treatments including control plots. A high yielding barley variety named HB-1307 was used as a test crop at a seed rate of 125 kg ha−1. The plot size used was 4.5 × 5.1 m (22.95 m2). Data were collected on barley grain yield and yield components. At crop maturity, the whole plot area (22.95 m2) was hand harvested at ground level from each plot for determination of grain yield and biomass yield. Grain yield was adjusted to 12.5% moisture content. Data on yield components were determined from random sample measurements of each parameter following the standard procedures of Anderson et al. (2002). Soil samples were randomly collected prior to experimentation and after harvesting for analysis of pH and Al3+. Soil pH was determined by using a pH meter in a 1:2.5 soil/water suspension using pH metre, and exchangeable A1+3 was extracted by 1 M KCl (Mclean, Reference McLean and Black1965).
Statistical analysis
Analysis of variance was performed using SAS statistical software version 9.1 (SAS Institute, 2001). A Proc GLM model was constructed to compare the measured agronomic parameters for both years separately as well as combined over years. Significance differences were set at P < 0.05. When the effects were found significant, further analysis was made using Tukey multiple comparison test. Pearson correlation coefficients were also used to assess the significance of the relationships between yield and yield components.
The mathematical formulation of the model is
$$\begin{equation*}
{Y_{ijk}} = \mu + {\alpha _i} + {\beta _j} + {\gamma _k} + \beta {\gamma _{jk}} + {\varepsilon _{ijk}}
\end{equation*}$$
with i = 1, 2, 3 for the blocks, j = 1, . . . ,5 for the levels of lime and k = 1, . . . ,4 for the levels of P fertilizer, and being:
-
Yijk = observed value of the dependent variable for the level of lime j and level of P k in the block i.
-
μ = general mean effect; α i = main effect of the block i.; β j = main effect of the level of lime j; γ k = main effect of the level of P k; βγ jk = interaction effect of the level of lime j with the level of P k; ε ijk = random error in the dependent variable for the level of lime j and level of P k in the block i; The assumptions for the model were:
-
• ε ijk independent and identically distributed;
-
• ε ijk →N(0, σ2) with σ2 = random variance for errors.
-
RESULTS
Soil pH and exchangeable aluminium
Soil analysis results after 2 years of liming is depicted in Figure 1b and Figure 1c. Results indicated that soil pH was significantly (P < 0.05) increased and Al3+ was markedly reduced to a negligible level. Liming at the rate of 0.55, 1.1, 1.65 and 2.2 t ha−1 increased soil pH by 0.48, 0.71, 0.85 and 1.1 units, and decreased Al3+ by 0.88, 1.11, 1.20 and 1.19 mill equivalents per 100 g of soil, and respectively. Generally, with successive increase in the amounts of lime, soil pH values increased with a corresponding decrease in exchangeable Al3+ of the soil.
(a) Mean monthly maximum, minimum temperature and precipitation for the study site (average of 10 years), (b) soil PH as affected by successive applications of lime (after 2 years of application), (c) exchangeable aluminium as affected by successive applications of lime (after 2 years of application). Bars are standard error of the mean (n = 3), and similar letters above the vertical bars with the same letters denote no significant difference at P < 0.05.
Grain yield and yield components
Mean thousand seed weight (TSW), number of seeds per spike (NSPS) and hectolitre weight (HLW) as affected by different levels of lime and P fertilizer is presented in Table 1. Since the interaction effects of lime and P were not significant for yield components, their individual effects are presented. Analysis of variance showed that all limed treatments had higher mean values of TSW, NSPS and HLW relative to control (no lime and P) both in 2010 and 2011 cropping seasons. Amongst the liming treatments, the highest rate (2.2 t ha−1) of lime recorded the highest mean values NSPS during both years. However, applications of 2.2 t ha−1 lime were not statistically different from lime rate of 1.65 t ha−1 for TSW, NSPS and HLW. Generally, applications of P fertilizer have resulted in increased yield components of barley as compared to control.
Mean values of thousand seed weight (TSW), number of seeds per spike (NSPS) and hectolitre weight (HLW) of barley as affected by different rates of lime and phosphorus.
*Mean values within a column that share similar letters are not significantly different at P < 0.05.
Table 2 shows grain yield (2011), biomass yield, plant height at harvest and number of tillers (2010 and 2011) as affected by different levels and P. The highest mean grain yield, biomass yield, plant height and number of tillers were recorded in the lime amended plots. In 2011, grain yield obtained by applications of 1.65 and 2.2 t ha−1 lime was statistically comparable, and significantly superior to 0.55 and 1.1 t ha−1 lime rates. Similarly, grain yield obtained by application of 20 and 30 kg P ha−1 were also comparable, and significantly (P < 0.05) higher than the control and applications of 10 kg P ha−1. As expected, the lowest grain yield was recorded in control plots, with no lime, no P. During first year (2010 cropping season), interaction effects of lime and P were observed on grain yield of barley (Figure 2). Accordingly, the highest mean grain yield was obtained by applications of 1.65 t ha−1 and 30 kg P ha−1. An increasing response to applied P with increasing rates of added lime have been attributed to either an improved rate of P supply by the soil or an improved ability of the plant to absorb P when Al toxicity has been eliminated. However, successive applications of lime beyond 1.65 t ha−1 and P beyond 30 kg P ha−1 could not successively increase barley grain yield (Figure 2).
Grain yield, biomass yield, plant height and number of tillers as affected by different levels of lime and phosphorus fertilizer on acid soils.
*Mean values within a column that share similar letters are not significantly different at P < 0.05.
Barley grain yield (kg ha−1) in 2010: Interaction between rate of lime (t ha−1) and rate of P (kg ha−1) applied (P < 0.05). The perpendicular line drawn shows optimum lime and P rate where maximum grain yield was achieved.
The effects of lime and P on grain yield and yield components of barley combined over 2 years is presented in Table 3. Our results showed that the highest significant grain yield and biomass yield were recorded by application of 1.65 and 2.2 t ha−1 lime. Statistically, these two lime rates were comparable, and significantly (P < 0.05) superior to control (no lime), 0.55 and 1.1 t ha−1 of lime. P applications combined over 2 years also significantly (P < 0.05) affected yield and yield components of barley. Accordingly, the highest significant barley grain yield was obtained by application of 30 kg P ha−1. Additions of 10, 20 and 30 kg P ha−1 have also increased grain yield by about 29, 55 and 66% as compared to control without P. Year cultivation had also significant effect on the yield and yield components of barley in this study (Figure S2). Hence, grain yield of barley obtained during 2011 was significantly higher than the yield obtained in 2010.
Number of tillers m−2, number of seeds per spike (NSPS), thousand seed weight (TSW), hectolitre weight (HLW), grain yield, biomass yield and plant height as affected by lime and phosphorus fertilizer combined over 2 years.
Note: NSPS= number of seeds per spike; TSW= thousand seed weight; HLW= hectolitre weight; †Mean values within a column that share similar letters are not significantly different at P < 0.05; *, **Significant at P < 0.05 and P < 0.01 probability levels, respectively; NS = Not significant.
Relationships between grain yield and yield components
Grain yield was positively correlated with plant height, number of tillers m−2, TSW and biomass yield, and the correlation was significant at P < 0.0001 (Table S2). However, grain yield was correlated most strongly biomass yield (r = 0.76), followed by number of tillers m−2 (r = 0.71) and plant height (r = 0.71). Similarly, positive and significant correlation coefficients were observed between biomass yield plant height, number of tillers m−2, TSW and NSPS. However, there was no significant (P > 0.05) correlation between biomass yield and HLW.
DISCUSSION
Soil pH and exchangeable aluminium
Soil pH generally increased in a linear fashion with increasing lime rate. The increase was highest with applications of the maximum rate (2.2 t ha−1) of lime. When lime is added to acid soils that contain high Al3+ and H+ concentrations, it dissociates into Ca+2 and OH− ions. The hydroxyl ions will react with hydrogen and Al3+ ions forming Al3+ hydroxide and water, thereby increase soil pH in the soil solution. Meanwhile, applications of the highest rate of lime appreciably reduced soil exchangeable Al3+, which was 1.32 Cmol kg−1 at the start of the experiment to a negligible level of 0.12 Cmol kg−1 after 2 years of soil analysis. Many authors (e.g., Álvarez et al., 2009; Fageria, and Baligar, Reference Fageria and Baligar2008a; Fageria and Stone, Reference Fageria and Stone2004) have also reported that liming raises soil pH, base saturation, and Ca and Mg contents, and reduces Al3+ concentration.
Effect of P on grain yield and yield components
In our study successive applications of P increased grain yield and yield components, and counteracted Al toxicity by precipitating exchangeable Al3+ as AlPO4. This could be the reason why large applications of phosphate fertilizers to acid soils overcome the toxic effects of Al and thereby improve growth of plants. A major characteristic of Al toxicity is an inhibition of the uptake and translocation of P by plants (Foy and Fleming, Reference Foy, Fleming and Jung1978). Thus, liming acid soils often increases P uptake by plants by decreasing A1 toxicity rather than by an effect on soil P availability, per se (Haynes and Ludecke, Reference Haynes and Ludecke1981). After a reviewing of liming on phosphate availability, Haynes (Reference Haynes1982) concluded that large additions of phosphates to acid soils reduce the injurious effects of Al ions by precipitating it from the soil and supplying sufficient phosphate for plant metabolic activity. However, the classical explanation of increased phosphate availability following liming is that in the short-term, the increased pH results in the hydrolysis of strengite and variscite with the release of phosphate ions into soil solution (Negeri, Reference Negeri1984).
Effect of lime on grain yield and yield components
Liming induced favourable conditions for plant growth was the main reason for yield increment of barley in this study. Numerous authors (Álvarez et al., 2009; Farhoodi and Coventry, Reference Farhoodi and Coventry2008; Scott et al., Reference Scott, Conyers, Poile and Cullis1999) also reported that application of lime at an appropriate rate brings several chemical and biological changes in the soil, which is beneficial or helpful in improving crop yields in acid soils. Studies elsewhere (e.g., Farhoodi and Coventry, Reference Farhoodi and Coventry2008; Wang et al., Reference Wang, Xu and Li2011) reported that yield increase from liming is mainly associated with an increase in soil pH and a reduction in plant uptake of Al and Mn. Response of applied lime could be affected by many factors in soil. However, type of crop species, time of application and environmental variables such as moisture have subtle effect on applied lime. In this study, the higher grain yield observed in 2011 might be attributed to solubility and downward movement of lime as the time progresses, and normal rainfall with uniform distribution throughout growing season in 2011 as compared 2010. Similar yield increments with time in limed plots were reported by Meng et al. (Reference Meng, Xiaonan, Zhihong, Zhengyi and Wanzhu2004).
Mechanisms of yield increase due to lime and P applications
Several mechanisms are involved in increasing yield and yield components of barley when lime and P are used to ameliorate acid soils. Past laboratory and field studies conducted to determine how P availability responds to lime addition reported that that liming enhances P uptake by alleviating Al toxicity and thereby improving root growth (Bolan et al., 2003; Fageria and Santos, Reference Fageria and Santos2008; Haynes, Reference Haynes1982). The improved root growth would allow a great volume of soils to be explored. This in turn favours improvement of barley grain yield and yield components. Many authors (e.g., Meng et al., Reference Meng, Xiaonan, Zhihong, Zhengyi and Wanzhu2004, Moir and Moot, Reference Moir and Moot2010) also reported that liming increased soil pH and significantly reduced the concentrations of exchangeable Al3+ in the soil. The observed significant lime × P interaction is in the first year (2010) is typical of P deficient, highly weathered, acid soils (Freisen et al., Reference Freisen, Miller and Juo1980).
Correlations of yield and yield components
Grain yield in cereals is the product of its yield components. Consequently, yield components of barley such as number of tillers m−2, NSPS, TSW and biomass yield were highly correlated with its grain yield. Numerous authors (Abeledo et al., Reference Abeledo, Calderini and Slafer2003; Ortiz et al., Reference Ortiz, Nurminiemi, Madsen, Rognli and Bjornstad2002) have also reported significant associations of barley grain yield with its yield components. Results obtained in the present investigation on soil applications of lime and P fertilizer clearly showed that the remarkable increase in NSPS has greatly contributed to increase in grain yield barley during 2010 as well as 2011.Numerous authors (Baethgen et al., Reference Baethgen, Christianson and Garcı´a Lamothe1995; Pablo, et al., Reference Pablo, Savin and Slafer2004) also reported increase in grain yield due to number of grains per unit land area.
CONCLUSIONS
For long time, soils of the study area were considered less suitable for crop production. The results of this study, however, clearly demonstrated that these soils are responsive to lime and phosphate fertilizer applications. Overall, results showed that there were significant changes in soil pH and exchangeable Al3+ as a result of amendment through liming. Applications of 1.65 t ha−1 lime drastically decreased the exchangeable Al3+ to the minimum level, and raised soil pH close to the optimum pH requirement of barley. Hence, for sustainable and higher productivity, barley production in the highlands of Ethiopia should entail applications of 1.65 t ha−1, 30 kg P ha−1, and use of improved high yielding barley varieties. However, as soils vary from site to site, the amount of lime applied should be based on the concentrations of exchangeable Al3+ of site. Therefore, in light of this finding, short-fallow periods currently practiced by farmers to mitigate the negative effects of soil acidity would not be a long-term solution due to rapidly growing population, and prevalence of severe soil erosion from bare fallow fields. The lime and P rates obtained in this study could serve as a reference to boost barley production in the study area and areas with similar ago-ecology having soil acidity problems.
Acknowledgements
The field and laboratory works were funded by the Ethiopian Institute of Agricultural Research (EIAR) in Ethiopia. We would like to thank Asnakech Dubale and Dabata Midhekisa for their assistance in monitoring the field activity. We highly acknowledge two anonymous reviewers for their useful comments and critiques on the earlier versions of this paper.
SUPPLEMENTARY MATERIALS
For supplementary material for this article, please visit http://dx.doi.org/10.1017/S0014479716000491.
