Introduction

Maize (Zea mays L.) is an important staple crop in most of Africa and accounts for more than 50 percent of total cereal production in the region. Ragasa et al. (2013) reports that the bulk of maize produced goes into food consumption in sub-Saharan Africa (SSA) making it an important crop for food security. The development and productivity of the livestock and poultry sectors also depend on the maize value chain since maize is a major component of poultry and livestock feed (Ragasa 2014). Despite the availability of improved germplasm, average maize yield in SSA remains one of the lowest (1.6 t ha−1) in the world (FAOSTAT 2010). In most parts of SSA, soil fertility decline is a major reason for the low yields of major food crops including maize. Unsustainable farming activities have severely depleted soil nutrients throughout much of the farming regions. Although fertilizer consumption increased steadily in recent times in SSA (Sommer et al. 2013), average fertilizer use rates are still considered too low and ineffective for sustaining crop production and maintaining soil fertility (Gruhn et al. 2000; Ragasa et al. 2013). A recent survey identified factors including high costs of fertilizers due to removal of subsidies, lack of access to fertilizers and inefficient marketing systems as the major constraints to the sub-optimal application of fertilizers (Chapoto and Ragasa 2013; FAO 2012).

With the limited availability of conventional fertilizers, the use of organic materials is considered a viable alternative to increase the productive capacity of soils. In SSA, the use of organic resources is identified as a mainstream opportunity for agricultural development in the region due to their relative availability (Partey and Thevathasan 2013). The most common organic resources employed in soil fertility programs in SSA include plant residues, green manure sources of leguminous crops, animal manure, mulches and tree/shrub prunings from agroforestry practices (Partey et al. 2011). Animal manures are bulky, have unpleasant scent and have high potential for harbouring pathogens (Bernal et al. 2009; Crutzen et al. 2008) making the use of plant residue sources the preferable option. Recent research of high quality plant residues (with relatively high N concentration and low C/N ratio) applied to agricultural fields indicated tremendous yield increase for major food crops in SSA (Beedy et al. 2010). The addition of such organic resources to the soil reportedly improves soil temperature, enhances soil structure, maintains high soil nutrient status and improves soil moisture retention (Chivenge et al. 2011; Makumba et al. 2006).

In Ghana, the use of organic resources is included in soil management practices but many potential plant residues remain underutilized due to limited research and knowledge transfer (Partey et al. 2011; Partey and Thevathasan 2013). Their utilization could complement and contribute to solving nutrient deficiencies due to limited application of inorganic fertilizers. Some of these organic resources are the leaf biomasses of Acacia auriculiformis, Baphia nitida, Albizia zygia, Azadirachta indica, Senna siamea, Senna spectabilis, Tithonia diversifolia, Gliricidia sepium and Leucaena leucocephala. Despite having limited use in Ghanaian cropping systems, studies in Kenya, Nigeria, Uganda and other parts of SSA have shown that leaf biomass application increase crop yields even on depleted soils (Beedy et al. 2010; Ikerra et al. 2007; Mucheru-Muna et al. 2014; Nziguheba et al. 2000). Biomass transfers from these species could therefore contribute to the development of sustainable soil improvement practices that improve and sustain crop production for smallholder farmers in Ghana. However, the development and adoption of such soil fertility improvement practices will be influenced by research results that reveal the viability of organic sources for improving soil fertility (particularly N availability) and crop yields. It was therefore the objective of this study to determine the N supplying capabilities of rarely-used plant residues for maize production in Ghana. The selection of the plant materials were based on their relative abundance at the study location, experimental evidence of their use as organic fertilizers in SSA (Gachengo et al. 1999; Jama et al. 2000; Partey et al. 2011), their residue N concentrations and underutilized status for soil management at the study location (Partey and Thevathasan 2013). The research was based on the hypothesis that with increased net N mineralization from the leaf biomass of the tested species, the biological yield of maize will increase because of a resultant increase in N availability and uptake.

Materials and methods

Study site

The study was conducted at the agroforestry demonstration field of the Faculty of Renewable Natural Resources (FRNR), Kwame Nkrumah University of Science and Technology, Kumasi (KNUST), Ghana, located at Lat 01 43°N and Long 01 36°W. The research area had been fallowed for five years prior to the execution of this study. The area falls within the moist semi-deciduous forest zone of Ghana and is characterized by a bimodal rainfall pattern, with the major wet season between May and July. This area also experiences a short dry season in August and a long one between December and March. The annual rainfall of the area ranges between 1250 and 1500 mm. The area is characterized by a mean annual temperature of 26.6 °C. Precipitation data recorded during the research period is shown in Fig. 1. Soil type at study site is a ferric acrisol.

Fig. 1
figure 1

Monthly rainfall distribution recorded during the experimental periods in 2013 and 2014. Data points are the means of three replicates

Full size image

Initial soil characterization

Prior to establishing the field experiment, soil samples were randomly collected from the surface 20 cm from 16 locations at the site for characterization using a stainless steel soil auger (25 inch. in diameter). The samples were composited and homogenized into one sample. They were then air-dried and sieved to 2-mm before being sub-divided into four sub-samples for analysis. Soil pH was measured with a glass electrode (1: 1 H2O), particle size was determined using the hydrometer method, total N was determined by dry combustion using a LECO TruSpec™ CN autoanalyzer (LECO Corporation), organic carbon was determined by the dichromate oxidation method (Motsara and Roy 2008), cation exchange capacity was measured using flame photometry of ammonium acetate extracts, available P by the ammonium phosphomolybdate method and available K by flame photometry (Toth and Prince 1949). The initial physicochemical properties of the soil at the study location were: pH (4.6), total N (0.42 g kg−1), available P (2.1 mg kg−1), available K (224.0 mg kg−1), organic C (13.8 g kg−1), CEC (5.8 cmol kg−1), sand (67.6 %), silt (28.4 %) and clay (4.0 %).

Plant residue quality and N mineralization study

Plant residue characterization

Organic resources used in the experiment were the leaf biomass of A. auriculiformis, B. nitida, A. zygia, A. indica, S. siamea, S. spectabilis, T. diversifolia, G. sepium, Z. mays and L. leucocephala. Table 1 presents a summary of the botany, growth habits and general uses of the species used in the experiment. In order to characterize the plant residues for quality parameters, portions of their leaf biomass including soft stems were oven dried at 65 °C for 72 h, ground with a grinder and sieved to 0.5 mm. The sieved plant materials were analyzed for total N, P, K, Ca, Mg and C in four replicates. For all analyses, total N and C were determined simultaneously by dry combustion using a LECO TruSpec™ CN autoanalyzer (LECO Corporation) while total K, Ca, and Mg were determined by the dry ashing and atomic absorption spectrophotometry as described by Eneji et al. (2005). Phosphorus was also determined in an ash solution by the ammonium phosphomolybdate method (Motsara and Roy 2008). The general chemical characteristics of all plant materials used are reported in Table 2.

Table 1 Botanical information and general uses of the species used in the experiment
Full size table
Table 2 Chemical characteristics of plant materials used in the experiment
Full size table

Quantification of N mineralization with plant residue application

An incubation experiment was performed under laboratory-controlled conditions to determine N mineralization from the selected plant residues. Briefly, 125 mg (equivalent to 5 t ha−1) of 0.5 mm sieved dried and ground leaf biomass of each plant were mixed with 50 g of 2 mm sieved sandy-loam soil in 250 ml beakers and incubated in the dark at 28 °C for 84 days. Unamended soil was used as a control. There were 24 beakers for every treatment. Prior to amending the soil, the soil was preconditioned by moistening to 50 % water holding capacity for 5 days. This was done to stabilize microbial activities (Xiang et al. 2008). The beakers were covered with aluminium foil to prevent rapid loss of water due to evaporation. Soil moisture content was checked by weighing every other day and the weight loss was replaced by addition of distilled water. The moisture content was kept constant at 50 % water holding capacity of the soil throughout the experiment. Nitrogen mineralization was determined by measuring the production of mineral N (NH4 + + NO3 −) at 3, 7, 14, 28, 56, and 84 days of incubation. Ammonium and nitrate were determined by extracting 25 g of moist soil with 2 M KCl at a 1: 4 soil and extractant ratio. Ammonium and nitrate in the KCl extract were determined by the indophenol blue and phenoldisulphonic acid methods respectively (Motsara and Roy 2008). All measurements were done by sampling four beakers per treatment on every sampling period. Analysis was done separately for each soil sample in a beaker. Net cumulative N mineralized from the different treatments was calculated by subtracting the inorganic N of the unamended control from amended soils at each sampling time (Abbasi and Khizar 2012; Sistani et al. 2008).

Field experiment

The field experiment was conducted in four continuous planting seasons: major rainy season of 2013, minor rainy season of 2013, major rainy season of 2014 and minor rainy season of 2014. For both years, major rainy season experiments were conducted between June and August while minor rainy season experiments were conducted between September and November. The experiment was first set up during the major rainy season of 2013 using a randomized complete block design with four replicates. The treatments included a control; leaf biomass sources of each plant (all applied at 5 t dry matter ha−1); and mineral fertilizer applied at a recommended rate of 90 kg N ha−1 (Partey et al. 2014a). Fresh leaf biomasses of the species (including soft stems) were harvested from nearby fields at the study location for the planting seasons in which leaf biomass treatments were imposed. Each treatment was allocated to a plot size of 3.2 m × 3.2 m. There were 44 plots in all. The tree leaf biomass was generally applied in whole with random cuts into smaller pieces where necessary. The biomass was surface applied on designated plots by hand and incorporated by hoeing a week before planting. The 90 kg N ha−1 inorganic fertilizer treatment (in the form of urea) was split applied on the designated plots at 7 days after planting (DAP) and 30 DAP using 40 and 60 % of the total fertilizer respectively. To reduce the effects of P deficiency, all 44 plots received one time basal P application (in the form of triple superphosphate) at a rate of 60 kg ha−1 7 DAP. The experimental treatments were applied in only the first three seasons: major rainy season of 2013, minor rainy season of 2013; and major rainy season of 2014. In the fourth cropping season (minor rainy season of 2014), the residual effects of the treatments were evaluated (no plant residue treatments or fertilizer applied). During planting, four maize seeds (of a local variety named ‘obatampa’) were sown per hill at 0.4 × 0.8 m spacing and thinned to two plants per hill within 2 weeks. Thinning was done to ensure that plants left in the field had uniform growth. As much as possible, confounding effects of crop residues were controlled by removing all maize biomass after every trial including the roots.

Determination of maize productivity and nutrient uptake

At physiological maturity, all maize plants within 4-m2 were sampled. To determine stover yield, the plants were uprooted from the soil after watering the surface soil. Uprooting plants was necessary to minimize confounding effects of crop residues. The above-ground residues were separated from the roots and oven dried in the laboratory at 65 °C for 72 h. To determine nutrient uptake, samples of the oven-dried above-ground residue were ground to pass through a 0.5-mm sieve and analysed for N concentration. Nitrogen was determined using LECO TruSpec™ CN autoanalyzer (LECO Corporation). Nitrogen uptake was determined by multiplying the dry-matter yields by the N nutrient concentration of the above-ground biomass. Grain yield was determined by collecting cobs into perforated harvesting bags and sun drying over two weeks until the grain reached 12.5 % moisture content (the acceptable moisture content in most African markets) (Kurwakumire et al. 2014).

Statistical analysis

Data on maize agronomic performance, nutrient uptake, and N mineralization were analysed using the analysis of variance (ANOVA) test. Repeated measures analysis was used to determine seasonal effects and the effect of season and treatment interaction on maize biological yield. Where test results were significant, the least significant difference method was used for mean comparison at a 5 % probability level. Correlation and regression analyses were used to establish significant relationships among measured parameters. All statistical analyses were conducted with Genstat 12 software (VSN International).

Results

Plant residue quality and N mineralization patterns

The C and nutrient content of the ten plant materials used in the study are reported in Table 2. Nitrogen ranged from 12.2 g kg−1 in maize stover to 39.2 g kg−1 in the leaf biomass of B. nitida. The C/N ratio also ranged from 12.1 in B. nitida to 34.4 in maize stover. Among the plant materials, leaf biomass of T. diversifolia recorded the greatest level of P and K. Calcium concentration was lowest in maize stover while Mg content was significantly (p ≤ 0.05) higher in T. diversifolia.

Figure 2 shows the N mineralization patterns of the plant materials used in the experiment. Immobilization occurred in soil amended with leave from A. auriculiformis, A. zygia, B. nitida, maize stover and S. siamea during the first 13 days of the incubation. Nitrogen immobilization also occurred in A. indica for 3 days. The highest levels of net N mineralization were observed for soil amended with T. diversifolia, G. sepium, L. leucocephala and S. spectabilis. Cumulative net N mineralization was significantly higher with T. diversifolia (93.5 mg N kg−1) at the end of the incubation period (84 days). The net N mineralization rate in T. diversifolia was approximately 1.1 mg N kg−1 day−1. Cumulative net N mineralization was comparable between G. sepium and L. leucocephala and lowest for A. auriculiformis, A. zygia, B. nitida, maize stover and S. siamea. A correlation and regression analysis indicated that cumulative net N mineralization of the plant materials were not significantly related to their initial N concentrations (r2 = 0.14, p = 0.28) or C/N ratios (r2 = 0.26, p = 0.14).

Fig. 2
figure 2

Cumulative net N mineralization of ten plant materials over 84 days of incubation under laboratory controlled conditions. Data points are the means of four replicates. Error bars are standard error of means. Aa Acacia auriculiformis, Az Albizia zygia, Ai Azadirachta indica, Bn Baphia nitida, GS Gliricidia sepium, Ll Leucaena leucocephala, Ms maize stover, Ssi Senna siamea, Ssp Senna spectabilis, Td Tithonia diversifolia

Full size image

Effects of treatments on N uptake and biological yield of maize

Table 3 shows the nitrogen uptake of maize as influenced by the treatments. In general, the application of plant residues increased N uptake on all amended plots. However, among treatments, N uptake was significantly (p < 0.001) higher on plots that received either inorganic fertilizer or T. diversifolia leaf biomass application. With respect to the biological yield of maize, repeated measures analysis showed significant (P < 0.001) seasonal effects and interaction with treatments (Table 4). Both grain and stover yields of maize where greater during the major rainy season trials. During the major rainy season of 2013, grain yield ranged from 1.0 to 3.1 t ha−1. Among treatments, grain yield was significantly higher on plots that received inorganic fertilizer or T. diversifolia leaf biomass (Table 5). Relative to the control, grain yield in the major season of 2013 was about 300 % greater with the application of either inorganic fertilizer or T. diversifolia leaf biomass. Comparable results were also obtained during the 2013 major season for A. auriculiformis, A. zygia, B. nitida, maize stover, and S. siamea. The effect of G. sepium leaf biomass aplication was comparable to that of S. spectabilis, A. indica and L. leucocephala. In the minor season of 2013, grain yield ranged from 0.8 t ha−1 for the control to 2.2 t ha−1 with the application of T. diversifolia leaf biomass. Statistically, the greatest effects were obtained from plots that received inorganic fertilizer, G. sepium or T. diversifolia treatments. The effects of A. auriculiformis, A. zygia, B. nitida, maize stover, and S. siamea were comparable and intermediate. Maize grain yield was also significantly (p < 0.001) higher with inorganic fertilizer, G. sepium and T. diversifolia treatments during the major rainy season trial of 2014. On average, maize grain yield was about twice that of the control when maize received either inorganic fertilizer, G. sepium or T. diversifolia leaf biomass. During the minor rainy season of 2014, the residual effects of T. diversifolia and inorganic fertilizer (although significantly higher than the control) were comparatively lower than was observed in previous trials. Generally, there were comparable results among treatments between growing seasons. Total cumulative grain yield obtained from the four experimental trials showed T. diversifolia, G. sepium and inorganic fertilizer recorded the highest grain yield of maize (Table 5). Among the cropping seasons, trials in the major rainy seasons of 2013 and 2014 recorded the highest mean grain yield (approximately 2.2 t ha−1).

Table 3 Nitrogen uptake (kg ha−1) of maize as affected by inorganic fertilizer and plant residue treatments
Full size table
Table 4 Repeated measures analysis for the effect of treatment and time on maize grain and stover yields
Full size table
Table 5 Grain yield (t ha−1) of maize as affected by inorganic fertilizer and plant residue treatments during the minor and major rainy seasons of 2013 and 2014
Full size table

Similar to the results on grain yield, the application of treatments increased maize stover yield in all four crop growing seasons (Table 6). Maize stover yield was significantly (p < 0.001) higher on plots that received inorganic fertilizer or T. diversifolia treatments. In 2014, maize stover yield ranged from 3.1 to 6.1 t ha−1 in the major rainy season and 2.1 to 4.7 t ha−1 in the minor rainy season. Similar to the results in 2013, G. sepium, T. diversifolia and inorganic fertilizer recorded the highest effects in the 2014 major season trial. The residual effects were generally comparable among treatments. At the end of the four experimental trials, total stover yield ranged from 10.1 t ha−1 for the control to 22.0 t ha−1 for T. diversifolia.

Table 6 Maize stover yield (t dry matter ha−1) as affected by inorganic fertilizer and plant residue treatments during the minor and major rainy seasons of 2013 and 2014
Full size table

Discussion

Considering that crop production is a soil-based industry, agricultural technologies that improve soil fertility in agroecosystems have major implications for reducing hunger by enhancing crop yields. The use of organic amendments is strongly recommended for the highly weathered tropical soils of SSA that are normally low in organic matter (Vanhie et al. 2015). In this study, the nitrogen supplying capabilities of ten rarely-used plant residues were tested on maize production in comparison with inorganic fertilizer. The N supplying capabilities of the species were determined based on their residue chemistry (N content and C/N ratio); N mineralization and effect on maize performance. The use of substrate N concentration and C/N ratio as indices for determining the decomposability and N mineralization of plant residues in agroforestry systems is well documented (e.g. Constantinides and Fownes 1994; Kumar and Goh 1999; Partey et al. 2014b; Gentile et al. 2008). According to Troeh and Thompson (2005), the break-even point for decomposing and increased net N mineralization of organic materials within a few weeks is a C: N ratio of about 32: 1. Whilst this assertion was consistent with the pattern of N mineralization recorded for T. diversifolia, G. sepium, L. leucocephala and S. spectabilis (Fig. 2), it contradicted that of B. nitida, A. zygia, and A. indica which had C/N ratios narrower than the critical maximum of 32: 1. However, the results are consistent with previous studies that demonstrated that the C/N ratio of plant residues may not be a reliable indicator of organic matter decomposition and N mineralization in both temperate and tropical regions (Ostrowska and Porebska 2015; Palm and Sanchez 1990; Partey et al. 2012). While these contrasting results do not undermine the applicability of substrate initial N concentrations and C/N ratio as plant litter quality indicators, they imply the necessity for prudent decision making in selecting plant residues for soil fertility improvement based on multiple plant and soil factors. It is therefore reasonable to assume that unlike the initial N concentrations of the plant residues and their C/N ratios, their decomposition and N mineralization may be related to: (1) different decomposer communities that may have developed on the plant residues based on their intrinsic qualities (Cobo et al. 2002); (2) other plant quality variables such as lignin, polyphenol, hemicellulose concentrations and their ratios with N; (3) C/N ratio of soil; as well as (4) the water retention capacity of plant residues which although unassessed in this study have been shown to influence plant residue decomposition and N mineralization (Palm et al. 2001; Iqbal et al. 2013; Makkonen et al. 2013). Based on the results on N mineralization, leaf biomass application of T. diversifolia, G. sepium, L. leucocephala and S. spectabilis are expected to improve soil N availability and subsequent uptake by crops for increased biological yield. The accelerated mineralization of N in T. diversifolia and G. sepium biomass may limit their use for long term sustenance of soil fertility and soil organic matter. Farmers may have to apply these materials every cropping season which may have significant economic implications.

The differential effects of the plant residues on the biological yield of maize reflected their differences in quality and N supplying capabilities. Apart from nutrient supply, plant residue quality has implications for other soil properties such as soil moisture, pH and cation exchange capacity which are intrinsically linked to soil organic matter content and quality (Bhupinderpal-Singh and Rengel 2007). The overall effects of the treatments on maize may therefore be a combination of factors beyond just N supply (which was the emphasis of this study). Generally, the results showed increased biomass yield, grain yield and N uptake of maize in all treatments compared with the control. However, the greatest effects occurred on plots that received either G. sepium or T. diversifolia treatments especially during the major rainy season; possibly because of increased water availability. These observations are consistent with the results of previous studies (Gachengo et al. 1999; Nziguheba et al. 2000; Partey and Thevathasan 2013). In Western Kenya, field trials conducted by Nziguheba et al. (2000) showed that the addition of T. diversifolia green manure tripled total maize yields after six seasons compared to the control and inorganic fertilizer treatments. Experimental trials under similar tropical conditions in Malawi, Brazil and other parts of SSA reported multiple increments in maize grain yield with the application of G. sepium prunings (Barreto et al. 2012; Beedy et al. 2010; Makumba et al. 2006). Further, the results on maize yield showed the nitrogen supplying capabilities of the plant residues may be more closely correlated with their N mineralization patterns than their N composition. It was evident that even with the application of 196 kg N ha−1 from B. nitida leaf biomass (Fig. 3), plots amended with B. nitida leaf biomass produced some of the smallest effects on maize grain and stover yield. Considering the significantly low net N mineralization of soil amended with B. nitida leaf biomass, the release of N may not have synchronized with crop N demand. According to Salas et al. (2003), organic resources contain significant concentrations of organic nutrients that undergo biological decomposition and mineralization processes to become available for crop use. It is therefore reasonable to assume that with high N mineralization, high N uptake could be expected, which may consequently result in high crop yield. This assertion would explain why the plant materials with high cumulative net N mineralization such as G. sepium and T. diversifolia recorded the greatest impact on the biological yield of maize. The argument is further supported by the significant (p < 0.001) positive correlation obtained between the amount of N mineralized and the biological yield and N uptake of maize (Table 7). However, the collection of large amounts of biomass and the resultant labour demands may limit large-scale adoption of T. diversifolia and G. sepium biomass for maize production. In addition, it is evident (Tables 5, 6) that, compared with the species with least N mineralization rates, T. diversifolia and G. sepium may have low residual impacts due to accelerated decomposition and N release. This may necessitate regular application of their biomass during crop growing seasons which may pose greater financial burden on resource-poor farmers; especially where labour requirements may be high for biomass collection.

Fig. 3
figure 3

Amount of N applied from inorganic fertilizer and plant residues used in the experiment. Data points are the means of four replicates. Error bars are standard error of means. Aa Acacia auriculiformis, Az Albizia zygia, Ai Azadirachta indica, Bn Baphia nitida, C control, Fert inorganic fertilizer, GS Gliricidia sepium, Ll Leucaena leucocephala, Ms maize stover, Ssi Senna siamea, Ssp Senna spectabilis, Td Tithonia diversifolia. All plant materials were applied at 5 t dry matter ha−1, inorganic fertilizer was applied at 90 kg N ha−1

Full size image
Table 7 Pearson correlation coefficients for the relationship between N uptake, N mineralization and the total biological yield of maize
Full size table

Conclusions

From the results on plant residue quality, most of the plant residues recorded relatively high N concentration and low C/N ratio although these properties were not always significantly related to their N mineralization patterns. Application of B. nitida, A. auriculiformis, A. zygia leaf biomass and maize stover resulted in an initial net N immobilization that lasted for 14 days. On the effect of the treatments on maize, the results confirmed that all the treatments could increase maize yield in the study area. However, the effect will be greater with either inorganic fertilizer, G. sepium or T. diversifolia leaf biomass application. Relative to the control, total grain yield after four cropping seasons was found to be comparable between inorganic fertilizer (9.2 t ha−1), G. sepium (8.8 t ha−1) and T. diversifolia (9.4 t ha−1) treatments. The results showed that differential effects of the species on maize biological yield were attributed to their differences in N mineralization. It was evident that with high N mineralization, high N uptake could be expected, which may consequently result in high crop yield. We therefore suggests that in places where inorganic fertilizers are limited, leaf biomass from G. sepium and T. diversifolia could offer the most suitable option in comparison with the other species used in this study. This notwithstanding, labour requirements and cost implications for harvesting leaf biomass should be considered in adopting this practice.