Liming and sulfur amendments improve growth and yields of maize in Rubona Ultisol and Nyamifumba Oxisol

Abstract

Aluminum toxicity is a major limitation to crop production on highly weathered and leached soils in Rwanda. Moreover, sulfur though widely deficient in Rwanda acidic soils has received little attention by soil fertility researchers. A field experiment on maize response and soil nutrients status to liming materials of travertines at 3.4 t ha⁻¹, ash wood 1.2 t ha⁻¹ of CaO equivalent and sulfur at 10 kg ha⁻¹ combined with NPK at 80, 60, and 45 kg ha⁻¹ respectively was conducted in Rubona Ultisol and Nyamifumba Oxisol. Results revealed that travertine and wood ash increased the soil pH from 4.7 to 5.8 or higher and decreased exchangeable Al³⁺ and H⁺ to near 0 cmolc kg⁻¹. Soil nutrients generally increased to high or medium ranges for crop production. Leaf dry biomass, plant height and maize grain yields were significantly higher in Rubona Ultisol than in Nyamifumba Oxisol. Plots that received wood ash, with NPKS or with NPK, generally had higher maize yields, followed by those which received travertines and NPKS or NPK which had maize growth response as compared to the control plots which received NPK only. Thereby, a combination of wood ash with NPKS or NPK, travertines with NPKS was found to neutralize soil aluminum toxicity, increase soil nutrients status to required levels for plant growth and increase maize yields significantly.

Keywords:

• travertines
• wood ash
• sulphur
• acidic soils
• maize yields

Introduction

Soil acidity is a major limitation to crop production in many areas of the world including Rwanda where 60% of the arable land is covered by acidic soils. In Rwanda Al toxicity is usually the main factor limiting crop growth in acid soils (Crawford et al. 2008). Several studies indicated that Rwanda has good sources of travertine that can be used as liming material for agricultural production (Van Straaten 2002; Sirikare et al. 2012; Nduwumuremyi et al. 2013). The liming effect of travertines is primarily attributable to their high pH and high calcium carbonate content (Crawford et al. 2008; Sirikare et al. 2012). Rwanda travertine deposits are located in the Northern Province (Musanze district) and in the Western Provinces (Karongi and Rusizi districts). These travertine deposits are found in the Rift Valley area in western Rwanda (Sirikare et al. 2012; Nduwumuremyi et al. 2013).

In many African regions, including Rwanda, fuel wood constitutes 61 –86% of primary energy consumption and generates large quantities of wood ash, which could be usefully utilized for agricultural production (Campbell 1990; Samir 1999; Avelino et al. 2011). After combustion most of the inorganic nutrients and trace elements from woody biomass are retained in the ash which could be a significant source of K, Mg, Ca, lime and could thus be used as a supplement to fertilizers (Naylor & Schmidt 1989; Bougnom et al. 2009; Bougnom & Insam 2009) in addition to ameliorating soil acidity. Travertine and wood ash from household cooking can thus be used as inputs to the Rwandan semi-subsistence farming.

Maize (Zea mays L.) is an important staple cereal crop in small-scale, rain-fed production systems in the central plateau (1600 m) and highland (1800 –2400 m asl) zones of Rwanda. The crop covers about 114,800 ha in the country (FAO 2009) but production is concentrated mainly in the highland zone. As a staple, it is consumed in several traditional food preparations (Nyirigira et al. 2005). However, annual production is often short of demand and grain has to be imported. In 2006, a total of 91,813 tons of maize grain was produced from approximately 114,836 ha but 29,076 tons had to be imported to supplement domestic production (FAO 2009). Farm productivity is therefore low, averaging about 0.8 t ha⁻¹. Furthermore, sulfur (S) fertilizer, combined with nitrogen (N), phosphorus (P) and potassium (K), has been shown to improve maize and wheat yields in sulfur deficient soils (Salvagiotti & Miralles 2008; O’ Leary & Rehm 2013). Grain yield response to NPKS fertilization combined with liming materials has not been studied in Rwandan acidic soils. This research was conducted to determine the effects of travetines, wood ash and S on maize grain yields and soil fertility improvement on Rubona Ultisol and Nyamifumba Oxisol in Rwanda.

Material and methods

Location of field experiments

The field experiment was conducted in Rubona (longitude: 2° 28.9′S; latitude 29° 45.6′E; altitude : 1 630 m), in Huye district and Nyamifumba (longitude: 2° 38.4′S; latitude : 29° 43.8′E; altitude : 2 012 m) in Nyamagabe district in the southern province of Rwanda. The mean annual rainfall is 1 200 mm in Rubona and 1 400 mm in Nyamifumba.

Soils of the experimental areas

The soils ware classified as Ultisol for Rubona site and Oxisol for Nyamifumba site (Birasa et al. 1992). Rubona Ultisol had a pH of 4.74, exchangeable hydrogen of 1.69 cmol_c kg⁻¹, exchangeable aluminum of 2.8 cmol_c kg⁻¹, total nitrogen of 0.30%, extractable phosphorus of 7.7 mg kg⁻¹, sulfur content of 1.4 mg kg⁻¹ and 47, 681, and 31.6 mg kg⁻¹ for exchangeable potassium, calcium and magnesium respectively while Nyamifumba Oxisol had a pH of 4.90, exchangeable hydrogen of 1.60 cmol_c kg⁻¹, exchangeable aluminum of 3.3 cmol_c kg⁻¹, total nitrogen of 0.33%, extractable phosphorus of 6.7 mg kg⁻¹, sulfur content of 1.6 mg kg⁻¹ and 35, 586, and 14.5 mg kg⁻¹ for exchangeable potassium, calcium and magnesium respectively. Most of the nutrients were found low to allow crop production (Okalebo et al. 2002; Horneck et al. 2011). The low pH and high exchangeable Al levels of these soils indicate that amelioration can be achieved by liming. These data were extracted from Sirikare et al. (2015).

The liming materials used

Travertines, as the liming materials used, were collected from Mashyuza, Gishyita, and Mpenge deposits and ground and sieved to pass a 2-mm size. A rate of 3.4 t ha⁻¹ of CaO equivalent was selected from the results of laboratory incubation experiments of all sources of travertine with the Rubona Ultisol and Nyamifumba Oxisol. This rate was sufficient to neutralize exchangeable aluminum to near zero cmol_c kg⁻¹and raise the pH somewhat above 5.6 (Sirikare et al. 2015). Wood ash was collected from a household in Ruboana using Calliandra and Leuceana agroforestry trees as the source of fuel wood, and the properties of the wood ash were 10.3 for pH, 0.22, 1.6 , 8.42 , 1.72, 2.21, and 0.01% for total nitrogen, extractable phosphorus, calcium, potassium, magnesium, and sulfur, respectively. These nutrients contents were found medium expect for sulfur which was low as source of soil amendments (Campbell 1990; FAO 2005; Yudani et al. 2011). These data were extracted from Sirikare et al. (2015).

The rate of 1.2 t ha⁻¹ CaO equivalent of wood ash was selected following the results of laboratory incubation experiments with the Rubona Ultisol and Nyamifumba Oxisol. This rate was sufficient to neutralize exchangeable aluminum to near 0 cmolc kg⁻¹ and raise the pH above 5.7 (Sirikare et al. 2015).

Field experiments

The experimental fields were cultivated manually at each site using a hoe. Plots of 3 m long × 3 m wide were set. Maize was planted in rows 75 cm apart with an intra-row spacing of 30 cm. The ground travertines from Mashyuza, Gishyita, and Mpenge were applied at the rate of 3.4 t ha⁻¹ and wood ash at the rate of 1.2 t ha⁻¹ of their CaO equivalent. The travertines and wood ash were applied in bands 15 cm wide along the rows and were incorporated to a depth of 10 cm using a hand hoe. Respective fert ilizers were applied at rates of 80, 60, and 45 kg ha⁻¹ of N, P, and K, respectively (Kelly & Murekezi 2000; Sallah et al. 2009) and ammonium sulfate at 10 kg ha⁻¹ of S (Ray & Spider 2000). This level of nitrogen in ammonium sulfate was taken into consideration for the total nitrogen applied. Phosphorus and potassium were broadcast over all plots and then harrowed manually using a hoe prior to sowing. The N fertilizer was applied/banded two weeks after planting. Sulfur was applied/banded as a 15-cm diameter ring around the planting hole at the time of sowing for those treatments that received S. Thus, the treatments are shown in Table 1.

Table 1. Treatments used in field experiment.

Treatment	Quantities of inputs applied
Only: NPK^a	Only			N80	P60	K45
Mashyuza travertine + NPK	MAT	3.4 t ha^−1	+	N80	P60	K45
Mashyuza travertine + NPKS	MAT	3.4 t ha^−1	+	N80	P60	K45	S10
Gishyita travertine + NPK	GIT	3.4 t ha^−1	+	N80	P60	K45
Gishyita travertine + NPKS	GIT	3.4 t ha^−1	+	N80	P60	K45	S10
Mpenge travertine + NPK	MPET	3.4 t ha^−1	+	N80	P60	K45
Mpenge travertine + NPKS	MPET	3.4 t ha^−1	+	N80	P60	K45	S10
Wood ash travertine + NPK	ASH	1.2 t ha^−1	+	N80	P60	K45
Wood ash travertine + NPKS	ASH	1.2 t ha^−1	+	N80	P60	K45	S10

^aThe control treatment is designed this way, that is control + NPK, because almost all farmers in Rwanda use N, P, and K fertilizers in maize production. MAT = Mashyuza travertine; MPET = Mpenge travertine; GIT = Gishyita travertine while N₈₀P₆₀K₄₅S₁₀ = 80 kg ha⁻¹ of N, 60 kg ha⁻¹ of P, 45 kg ha⁻¹ of K and 10 kg ha⁻¹of S

Seeds of maize cultivar ZM 607, an inbred line, purchased from the Rwanda Agriculture Board (RAB) were sown into the center of the lime treated bands. Urea was band-applied adjacent to plants two weeks after planting, during thinning. The urea was buried immediately after application to avoid loss due to volatilization (Sallah et al. 2009). Hand-weeding was done using a hoe when necessary to keep the plots free of weeds.

Maize data collection

Ten plants were selected randomly in every plot for biomass determination. Plant height was measured at tasseling (Ray & Spider 2000).

At maturity, net plots of 3.6 m² each were harvested and corn ears were removed by hand. The ears were dried to constant weight at 12% moisture content and then shelled to obtain grain weight in k ilog ram in each plot. The moisture content was measured using a portable moisture meter.

Soil sampling and analysis after harvest

Soil sampling was carried out after harvesting. Soil was randomly sampled in the plant row to depth of 20 cm using an auger (four samples per plot) and the samples from each plot were bulked to form a composite. Soil pH_water was determined in 1:2.5 soil suspensions in distilled water using a glass electrode. Exchangeable calcium, magnesium and potassium were extracted using 1 N of ammonium acetate (NH₄C₂H₃O₂) and determined using atomic absorption spectrophotometry. Exchangeable aluminum was extracted with 1 M KCl and determined using atomic absorption spectrophotometry (Moberg 2001). Exchangeable hydrogen was determined by the difference between total exchangeable acidity and exchangeable aluminum. Total exchangeable acidity was extracted with 1 M KCl and determined in the extract by titration with 0.020 M NaOH using phenolphthalein as an indicator. Extractable phosphorus was determined following extraction by 0.03 M ammonium fluoride and 0.025 M hydrochloric acid using the Bray 1 method (Okalebo et al. 2002) while sulfur was determined by the BaCl₂ turbi-dimetric method using a UV spectrophotometer (Moberg 2001).

Statistical analysis

The experimental design used was the completely randomized block design (CRBD) with three replications. The data were subjected to analysis of variance (ANOVA) using the Genstat XIV package and treatment means were compared using the least significant differences (LSD) at the 5% probability level. Regressions analysis was undertaken between the levels of yields and corresponding levels of exchangeable aluminum under the different treatments in the respective soils types.

Results and discussion

Liming potential of the different liming materials

Effects of liming on soil pH and soil exchangeable hydrogen

Figure 1(a) and 1(b) shows the effects of different types of liming travetines and wood ash on soil pH. Soil pH increased significantly (P ≤ 0.001) following addition of travertines from Mashyuza, Gishyita, and Mpenge from the pH of control soils to the highest levels above 5.92 in both the Rubona Ultisol and Nyamifumba Oxisol. Wood ash also significantly (P ≤ 0.001) increased the soil pH in the Rubona Ultisol and Nyamifumba Oxisol from the lower levels of the control soils to the highest level above 5.83 and 5.68, respectively.

Figure 1. Change in soil pH, exchangeable Al³⁺ and exchangeable H⁺ in the Rubona Ultisol (a, c, and e) and Nyamifumba Oxisol (b, d, and f) after maize harvesting. LSD (P ≤ 0.001) shown as a vertical bar. MAT = Mashyuza travertine, GIT = Gishyita travertine, MPE = Mpenge travetine, NPKS = nitrogen, phosphorus, potassium, and sulfur fertilizers. Treatments shown above are the ones presented in Table 1.

Figure 1(e) and 1(f) shows the effects of the liming travertines and wood ash on soil exchangeable hydrogen on the Rubona Ultisol and Nyamifumba Oxisol. Soil exchangeable hydrogen decreased significantly (P ≤ 0.001) by addition of travertines from Mashyuza, Gishyita, and Mpenge from the highest level of the control soil to the lowest levels below 0. 37 cmol_c kg⁻¹. Wood ash also significantly (P ≤ 0.001) decreased the soil exchangeable hydrogen in the Rubona Ultisol from the higher level of the control soils to the lowest levels below 0.25 cmolc kg⁻¹.

Brady et al. (1994), Charles et al. (1999), Mokolobate and Haynes (2002), Naramabuye and Haynes (2006), and Fageria and Baligar (2005) have shown that the presence of calcium carbonate (CaCO₃) in the liming materials was responsible for the rise in the soil pH and decrease of soil exchangeable hydrogen, according to the equations CaCO₃ + H⁺ → HCO₃⁻ + Ca²⁺; HCO₃⁻ + H⁺ → CO₂ + H₂O (Fageria & Baligar 2005). The neutralization of H⁺ then leads to the rise in soil pH. Therefore, the CaCO₃ in the travertines have contributed greatly to the rise of the soil pH and decrease soil exchangeable hydrogen in the Rubona Ultisol and Nyamifumba Oxisol. Rubona Ultisol and Nyamifumba Oxisol sites generally behaved similarly following application of all sources of travertine at 3.4 t ha⁻¹ of CaO equivalent, with the pH increasing in all soils to the highest levels above 5.92 relative to the control soil pH and soil exchangeable hydrogen decreasing from the highest level of the control soil to the lowest levels below 0. 37 cmol_c kg⁻¹. These field results confirm those of the laboratory incubation experiment undertaken in respect of these same soils (Sirikare et al. 2015).

Naylor and Schmidt (1989), Bougnom et al. (2009), and Bougnom and Insam (2009) found that the calcium or magnesium oxide or carbonate present in ash were responsible for raising the pH and decreasing exchangeable hydrogen of acidic soil according to the equation CaO + H⁺ → Ca²⁺ + OH⁻; MgO + H⁺ → Mg²⁺ + OH⁻ (Fageria & Baligar 2005). Thus, the contents CaCO₃, CaO, and MgO in wood ash contributed to raise the soil pH and decrease soil exchangeable hydrogen in the Rubona Ultisol and Nyamifumba Oxisol. All soil sites generally behaved similarly after application of wood ash at 1.2 t ha⁻¹ of CaO equivalent with pH increasing in all soils to the highest levels above 5.83 relative to the control soil and soil exchangeable hydrogen decreasing from the highest level of the control soil to the lowest levels below 0. 25 cmol_c kg⁻¹. These field results confirm those of an earlier incubation study (Sirikare et al. 2015).

Effects of liming on soil exchangeable Al³⁺

Figure 1(c) and 1(d) shows the effects of the different types of liming travertines and wood ash on soil exchangeable aluminum in the Rubona Ultisol and Nyamifumba Oxisol. Travertines from Mashyuza, Gishyita, and Mpenge decreased soil exchangeable aluminum significantly (P ≤ 0.001) from the highest levels of the control soils (2.8–3.3 cmol_c kg⁻¹) to the lowest levels below 0.22 cmol_c kg⁻¹. Wood ash also significantly (P ≤ 0.001) decreased the soil exchangeable aluminum in the Rubona Ultisol and Nyamifumba Oxisol from the higher level of the control soil to the lowest levels below 0.20 cmol_c kg⁻¹.

In their studies, Kochian (1995), Panda and Matsumoto (2007), and Sónia (2012) found that as the level of calcium increased due to the liming in acidic soils, aluminum ions held by the negative sites on soil particle surfaces were displaced into soil solution where they were hydrolyzed to become insoluble sesquioxides Al(OH)₃ according to the equation Al³⁺ + 3H₂O → Al(OH)₃ + 3H⁺. Hence, the presence of calcium cations in travertines, once added to soil, displaced the aluminum cations held on negatively charged soil particles into soil solution, thereby forming insoluble sesquioxide. Rubona Ultisol and Nyamifumba Oxisol sites generally behaved similarly following application of all sources of travertine at 3.4 t ha⁻¹ of CaO equivalent, with the exchangeable aluminum decreasing in all soils to levels below 0.22 cmol_c kg⁻¹ relative to those in the control soils.

The decrease in soil exchangeable aluminum where wood ash was applied was due to presence of calcium, magnesium, and potassium in wood ash which, once added to the soil, displaced the aluminum cations held on negatively charged soil particles into soil solution, thereby forming insoluble sesqui-oxide (Campbell 1990; Panda & Matsumoto 2007; Sónia 2012) as shown by the chemical equation presented above. The Rubona Ultisol and Nyamifumba Oxisol behaved similarly upon application of wood ash at 1.2 t ha⁻¹ of CaO equivalent. The exchangeable aluminum decreased in all soils sites to the lowest levels below 0.19 cmol_c kg⁻¹ relatives to the control. The above observations show the efficacy of the applied rates of the liming materials in neutralizing exchangeable Al³⁺ to low, harmless levels, as was similarly observed in the laboratory incubation experiments (Sirikare et al. 2015).

It can be seen that these liming materials were effective in raising the soil pH to the acceptable levels, as well as reducing the exchangeable H⁺ and the toxic exchangeable Al³⁺ to levels that would no longer pose a toxicity threat to plant growth. It may be proposed here that subsequent research should be undertaken to determine whether or not the rates of liming materials could be further reduced but still achieve comparable and/or acceptable changes in soil pH, exchangeable Al³⁺ and H⁺, for economic reasons.

It may also be further stated that in addition to resulting in favorable soil pH, exchangeable Al³⁺, and H⁺, the liming materials also improved other soil characteristics, as now subsequently discussed, which could contribute to improved plant growth and crop yields. These are presented in the following sections.

Effects of liming materials and inorganic fertilizers on soil nutrients status in the Rubona Ultisol and Nyamifumba Oxisolss

Table 2 shows the effects of travetines from Mashyuza, Gishita, and Mpenge, wood ash and inorganic fertilizers on levels of selected nutrients in the Rubona Ultisol and Nyamifumba Oxisol. Following these treatments, total nitrogen was generally moderate to high in the soils. Extractable phosphorus generally increased significantly in the amended plots (P ≤ 0.001) except the control although it received the same level of phosphorus. However, in both soils, the levels of extractable phosphorus were significantly (P ≤ 0.001) lower in the plots that received NPKS as compared to the plots that received NPK. Sulfur generally increased significantly (P ≤ 0.001) in all plots who received sulfur as fertilizer with the highest in the plot amended with wood ash. Exchangeable potassium levels generally increased significantly (P ≤ 0.001) in all plots which received muriate of potash as potassium with the highest level in plots amended with wood ash. Exchangeable calcium and magnesium levels generally increased significantly (P ≤ 0.001) in all plots which received travertine from Mashyuza, Gishyita, and Mpenge and wood ash respectively with the highest levels in plots amended with travertines.

Table 2. Effects of liming material and inorganic fertilizer on Rubona Ultisol and Nyamifumba Oxisol nutrients status after harvesting.

Soil	Treatments	N (%)	P (mg kg^−1)	Mg (mg kg^−1)	K (mg kg^−1)	S (mg kg^−1)	Ca
Rubona Ultisol	Control NPK	0.32H	6.32L	46.20L	1.57L	648L	35.57L
	MAT + NPK	0.26H	28.13M	74.33M	2.07L	2281H	80.65H
	MAT + NPKS	0.28H	20.67M	78.93M	7.25M	2087H	85.60H
	GIT + NPK	0.24M	32.47M	83.72M	2.01L	2195H	96.69H
	GIT + NPKS	0.26H	18.77L	70.03M	5.70M	1928H	105.20H
	MPET + NPK	0.23M	23.57M	87.87M	1.82L	2145H	51.89M
	MPET + NPKS	0.26H	16.67L	71.90M	7.40M	2111H	57.62M
	ASH + NPK	0.21M	22.47M	121.33M	4.61L	1162M	74.16M
	ASH + NPKS	0.24M	21.00M	153.40M	9.10M	1121M	71.51M
	LSD (P ≤ 0.001)	0.06	3.8	26.22	1.53	212.40	14.19
Nyamifumba Oxisol
	Control NPK	0.33H	6.10L	33.46L	1.53M	544L	16.75L
	MAT + NPK	0.32H	32.20M	77.30M	1.83M	2441H	80.22H
	MAT + NPKS	0.34H	22.83M	74.23M	6.76M	2205H	79.33M
	GIT + NPK	0.28H	31.17M	79.03M	1.63M	2638H	93.75H
	GIT + NPKS	0.31H	21.70M	74.48M	7.53M	2498H	99.13H
	MPET + NPK	0.33H	32.07M	80.00M	1.59L	2768H	51.97M
	MPET + NPKS	0.35H	17.77L	83.60M	7.16M	2524H	50.90M
	ASH + NPK	0.26H	23.13M	152.37M	4.26L	1051M	80.67H
	ASH + NPKS	0.29H	19.03L	169.23M	8.46M	1163M	83.90H
	LSD (P ≤ 0.001)	0.06	3.84	16.88	1.95	362.81	15.67

H = high; M = medium; L = low, according to Horneck et al. (2011) and Okalebo et al. (2002). Treatments shown above are the ones presented in Table 1.

Liming is the most widely used long-term method of soil acidity amelioration, and its success is well documented. Application of lime at an appropriate rate brought several chemical changes in the soil which was beneficial in improving soil nutrients on acid soils. Adequate liming eliminated aluminum toxicity and H⁺ and improved availabilities of calcium, phosphorus, and magnesium (Scott et al. 2001; Kaitibie et al. 2002; Fageria & Baligar 2009). Studies have shown that the liming materials combined with inorganic fertilizer resulted in increases of soil nutrients such as N, P, and K (Lwkin et al. 1994; Motavalli & Miles 2002; Wenyi et al. 2012). Therefore, the increased level of extractable phosphorus and potassium in Rubona Ultisol and Nyamifumba Oxisol was likely due to the combine sources of P and K fertilizer, travertine and wood ash.

Oxisol and Ultisol soils are naturally deficient in total and plant-available phosphorus and significant portions of applied P are immobilized due either to precipitation of P as insoluble iron or aluminum phosphates (Fageria & Baligar 2009; Jianbo et al. 2011). Edmeades and Perrott (2004) and Fageria and Santos (2008) reported that the primary benefit of liming occurs through an increase in the availability of P by decreasing P adsorption and/or precipitation according to the equation AlPO₄ (P fixed) + 3OH⁻ → Al(OH)₃ + PO₄³⁻ (P released). Thus, the difference in increasing extractable phosphorus observed between plots that benefited from use of travertine and wood ash in relation to the control plot which received only NPK fertilizer into the Rubona Ultisol and Nyamifuba Oxisol was due to the increase of soil pH, due to liming, thereby reducing precipitation/adsorption of phosphorus by iron and aluminum. Further, Taalab et al. (2008) and Deshbhratar et al. (2010) found that the combination of phosphorus sources and/or nitrogen with S resulted in better effect on N, P, and K uptake by corn and pigeon pea than when phosphorus or nitrogen are applied alone. Thus, the low levels of extractable phosphorus in the plots where P was combined with S were probably due to high P uptake by the maize since the extractable P status in the soil was evaluated after harvesting and yields were significantly higher in plots that received S as compared to the plots which did not receive S.

Clark (1984) and Marschner (1995) found that liming materials rich in calcium and magnesium carbonate increased soil-exchangeable Ca and Mg. Thus, the increase of exchangeable calcium and magnesium in the plots that received travertines and wood ash was mainly due the presence of calcium carbonate and magnesium carbonate in travertines (Sirikare et al. 2015) and calcium and magnesium carbonate or hydroxide present in wood ash (Adriano et al. 2002; Campbell 1990).

Plots which received wood ash resulted in high level of exchangeable potassium; this increase of potassium was likely due to high amount of potassium present in the wood ash (Campbell 1990). Plots that received sulfur have shown an increase in soil sulfur content derived from the application of ammonium sulfate (Ray & Spider 2000).

Effects of liming material and inorganic fertilizer on maize growth and yields

Table 3 shows the effects of travetines from Mashyuza, Gishita, and Mpenge and wood ash combined with inorganic fertilizer (NPKS) on maize growth and yields in Rubona Ultisol and Nyamifumba Oxisol. Germination rate was not negatively affected by the treatments and in all plots was above 88.67%. All plots which received travertines or wood ash plus NPKS or NPK had significantly (P ≤ 0.001) increased plants heights in relation to the control plots which received NPK fertilizer only. Plots which received travertines or wood ash combined with NPKS had increased leaf dry biomass and maize grain yields significantly (P ≤ 0.001) relative to the control plots and the plots which received travertines plus NPK fertilizers. Plots amended with wood ash combined with NPK with or without S fertilizers had significantly (P ≤ 0.001) increased the maize grain yields in relation to the plots which received travertine combine with NPKS fertilizers.

Table 3. Effects of liming material and inorganic fertilizer on maize growth and yields on Rubona Ultisol and Nyamifumba Oxisol.

Soil	Treatments	GR (%)	Leaf dry biomass (kg ha^−1)	Plant height (cm)	Grain yield (t ha^−1)
Rubona Ultisol	Control NPK	91.25	41.3	78.2	0.83
	MAT + NPK	92.42	133.9	150.5	2.60
	MAT + NPKS	93.40	147.0	164.1	3.90
	GIT + NPK	88.67	134.9	158.3	3.60
	GIT + NPKS	89.65	162.3	163.3	4.20
	MPET + NPK	93.83	128.0	156.0	3.40
	MPET + NPKS	93.33	161.1	166.8	4.20
	ASH + NPK	97.00	176.3	168.7	5.30
	ASH + NPKS	93.75	182.3	179.8	5.60
	LSD(P ≤ 0.001)	7.60	23.71	23.92	0.64
Nyamifumba Oxisol	Control NPK	93.33	32.7	35.5	0.35
	MAT + NPK	93.82	83.5	113.3	2.60
	MAT + NPKS	96.67	104.2	127.8	3.36
	GIT + NPK	97.08	76.1	116.7	2.46
	GIT + NPKS	95.67	128.1	129.3	3.17
	MPET + NPK	94.50	79.7	120.1	2.63
	MPET + NPKS	96.25	127.9	133.4	3.03
	ASH + NPK	97.92	133.9	135.3	3.40
	ASH + NPKS	96.00	151.4	150.0	3.90
	LSD (P ≤ 0.001)	4.50	11.81	17.88	0.30

Note: Treatments shown absove are the ones presented in Table 1. GR = germination rate.s

Several studies have shown the influence on sulfur fertilizer when combined with nitrogen and phosphorus on maize yields. Rahman et al. (2011), Ray and Spider (2000), and Varin et al. (2010) have shown that the combination of sulfur with nitrogen and phosphorus fertilizers improved the whole plant dry biomass and maize grain yields. Therefore, the increase in grain yields and dry leaf biomass in plots which received sulfur fertilizer combined with nitrogen, phosphorus, and potassium and liming material was due to the effect of sulfur which is recognized to limit the full yield potential of the maize once deficient in the soil (Tandon 1989; Muhammad et al. 2004). It should be noted that the soils of the present study were low in S.

Plots which received wood ash combined with NPKS or without S fertilizers had the highest grain yields. Wood ash beside being rich in the liming material such as CaCO₃, MgCO₃, CaOH₂, MgOH is also rich in other crop nutrients such nitrogen, phosphorus, sulfur, and potassium, hence its contribution to the maize grain yield could greatly be attributed to the presence of those crops nutrients that travertines do not have (Naylor & Schmidt 1989; Bougnom et al. 2009; Bougnom & Insam 2009).

By undertaking an overall ANOVA, it was revealed that maize in the Rubona site overall had significant (P ≤ 0.001) better performance 140.8 kg ha⁻¹, 153.9 cm 3.7 t ha⁻¹ for leaf biomass, plant height and grain yield respectively than that of Nyamifumba 102.0 kg ha⁻¹, 117.9 cm, and 2.8 t ha⁻¹ for leaf biomass, plant height and grain yield, respectively.

Cultivar ZM 607 is described as a late maturing maize cultivar, adapted to mid-altitude (800– 1600 meters above sea level) environments, according to Magorokosho and Pixley (1997), Magorokosho et al. (2003), and Weiwei et al. (2011). Cultivar ZM 607 did not performed in Nyamifumba Oxisol and yet the nutrients there were found to be sufficient. Nyamifumba is located in the highlands of Crete Congo Nile divide, above 2000 meters above sea level. Therefore, the highland area, in which ZM 607 cultivar was not adapted to, could be the main reason the cultivar overall failed to perform as it did in Rubona Ultisol.

Relationships between maize yields and soil exchangeable Al³⁺

Figure 2 shows the regressions between exchangeable aluminum and maize growth/yields. A strong negative relationship was found between leaf dry biomass, plant height and maize grain yields with soil exchangeable aluminum levels both in Rubona Ultisol and in Nyamifumba Oxisol. A strong coefficient of determination, R², generally higher than 0.90 (P ≤ 0.001) in both sites, illustrates how leaf dry biomass, plant height, and maize grain yields were dependent of the level of exchangeable Al³⁺ in the soil.

Figure 2. Relationships between soil esxchangeable aluminum and (a) leaves biomass, (b) plants height, (c) maize grain yield in Rubona Ultisol, (d) dry leaves biomass, (e) plants height, and (f) maize grain yields in Nyamifumba Oxisol. Regression equations, lines of best fit, and highly significant coefficients of determination, R², are shown.

In their studies on maize performance in acidic soil, The et al. (2006), and Ermias et al. (2013) found that neutralization of aluminum toxicity was associated with a significant increase in plant height, in dry matter of the aerial part of the maize plants and a significant (P ≤ 0.05) (139–208%) increase in corn grain yields. Hence, the improved growth and yields of maize in the present study was due to neutralization of aluminum toxicity in the soils by the travertines and wood ash.

Aluminum toxicity is a major limitation to crop production on highly weathered and leached soils in Rwanda. Moreover, sulfur though widely deficient in Rwanda acidic soils has received little attention by soil fertility researchers. Results of the present studies revealed that the travertines from all sources, and wood ash, increased the soil pH from 4.7 to higher than 5.8, and decreased exchangeable Al³⁺ and H⁺ to near 0 cmol_c kg⁻¹. Soil nutrients generally increased to higher or their medium ranges for crop production. Leaf dry biomass, plant height, and maize grain yields were significantly (P ≤ 0.001) higher in Rubona Ultisol than in Nyamifumba Oxisol. Generally, plots that received wood ash, with NPKS or with NPK, had higher maize yields, followed by those which received travertines and NPKS or NPK which had higher maize growth response as compare to the control plots which received NPK only. This may be because maize cultivar ZM 607 was more adapted in the Rubona Ultisol located in mid-altitude as compared to the high altitude in the Nyamifumba Oxisol.

Acknowledgments

We are grateful to the Africa Green Revolution (AGRA) Soil Health Program for funding this work, Rwanda Agriculture Board and University of Rwanda for laboratory facilitation.

Disclosure statement

No potential conflict of interest was reported by the authors.