1. Introduction

In sub-Saharan Africa, maize is one of the most widely grown staple food crops occupying more than 33 million hectare yearly [1]. The average annual grain production of maize in Sierra Leone was about 70 metric tonnes at a growth rate of 19.3% [2]. Maize production in Sierra Leone is very significant amongst resource poor farmers as it is a source of income, vitamins A, C and E, essential minerals, contain 9% protein, rich in dietary fibre and calories, which are a good source of energy [3, 4].

Sierra Leone is constrained by declining soil fertility, pest and diseases, but weeds are a constant source of concern for limiting self sufficiency in food production. Smallholder farmers are the main actors in food production and hardly adopt high input farming technologies, thus majority of farm households are managing plots which do not exceed two cropped hectares [5] due to weed problems. Various studies have confirmed that mulches provides moderate soil temperature, increases soil porosity, controls runoff and erosion as well as suppresses weed growth and maintained high crop yields [6-9]. Organic mulches are more popular in cropping systems as they can suppress weeds and minimized soil tillage for weed control [10]. Reference [11] reported that mulching with organic materials showed good results in weed control.

Gliricidia sepium is an extremely versatile nitrogen-fixing agroforestry tree that can be incorporated in diverse ways into many different smallholder-farming systems [12]. The application of G. sepium leaves mulch are widely used for mulching purposes, has allelopathic effects on weed seed germination and result to positive yield response of maize [13].

Maize is susceptible to weed competition in the early stages of growth, thus herbaceous legumes that serve the purpose of suppressing weeds as well as growth and yield of crops have not been widely adopted by maize farmers in this region. Therefore, the objective of the study was to evaluate the suppressive ability of G. sepium leaf mulch on weed growth and maize productivity.

2. Materials and Methods

2.1. Study Area

The present study was conducted at Njala Agricultural Research Centre experimental farm (08° 14' S, 12° 1' W), located in the savannah woodland agro-climatic zone of Sierra Leone, during the period from June to September 2016. The climatic condition experienced in the study area is not different from the rest of the country with two distinct seasons, rainy season (May-October) and dry season (November-April). The predominant vegetation of the study area was secondary bush. The mean annual rainfall was 2800mm; mean monthly temperature range from 33°C to 28°C and relative humidity 70% were observed during the experimental period. The soil texture of the trial site was gravely clay loam soil and its chemical properties (0-20cm depth) were pH = 3.91, organic carbon = 1.85%, available P = 8.56 mg/kg, Total % of Nitrogen = 0.08% and K = 0.14 cmol/kg.

2.2. Crop Management

The experimental site was manually cleared of vegetation, debris removed, ploughed and levelled using machetes and hoes. The trial area was marked out into plot sizes measured 5m x 3m with spacing of 1m x 1m between plots and replications respectively. The experimental area consisted of 15 plots in an area of 29m x 11m (319m² or 0.32ha). An improved maize variety (DMR-ESR-Yellow) was planted on the 10th of June on flat beds with spacing of 75cm x 50cm given a plant population of 23,333 plants/ha. Two seeds were sown and seedlings thinned to one per stand at two weeks after planting. The G. sepium trees were pruned, leaves separated from stem, sun dried for three days and applied as mulch at the rates of 0, 30, 60, 90 and 120kg ha^-1 respectively. At 2 weeks after planting (WAP), N-P-K (15:15:15) fertilizer was applied at the rate of 200kg ha^-1 using the side placement method, while urea (46% N) was top dressed at 8 WAP at the rate of 60kg N ha^-1. Manual weeding was done twice by hoeing at 4 and 8 WAP in the unmulched plots and hand pulling once in the mulched plots at 8 WAP.

2.3. Experimental Treatments and Design

The experimental treatments were five levels of Gliricidia sepium leaves mulch (0, 30, 60, 90 and 120kg ha^-1), evaluated using one variety of maize (DMR-ESR-Yellow). The treatments were laid out in a randomized complete block design with three replications.

2.4. Data Collection

Soil samples were transversely collected from ten different points at a depth of 0-20cm before trial establishment. The soil was put together and mixed to form composite sample, air dried, crushed, sieved and analysed for chemical properties. A 50cm x 50cm metal quadrat was thrown once in each plot to identify weed species harvested, counted, oven dried to a constant weight of 80°C and recorded to compute the relative frequency (RF), relative density (RD) and relative importance value (RIV) of each species according to [14].

(1)

(2)

(3)

Plants height (cm) was measured from 5 randomly selected plants in the middle rows of each plot from the base of the stem to the last emerged leaf using a graduated pole. The mean height from the 5 randomly selected plants at 1 and 2 MAP was taken as the score for each plot.

Stem girth (cm) was taken 10cm above ground level at the base of the maize plant from 5 randomly selected plants in the middle rows per plot and mean stem girth score at 1 and 2 MAP computed.

The number of leaves per plant from 5 randomly selected plants in the middle rows of each plot was determined by counting and scores computed at 1 and 2 MAP.

The leaf area of 5 randomly selected plants from the middle rows of each plot was measured using a graduated pole by calculating length by width (L x W) of the leaves, which was later determined based on the relationship proposed by [15]: Leaf area index = Population of plants per plot x Average number of leaves per plant x Average area per leaf x (Area of plot)^-1. (4)

At full maturity, a 2m² area of maize was harvested to determine the grain yield ha^-1, dry Stover yield, number of harvested cobs, weight of dry cobs + grains and 1000 kernel weight. The total grain yield of 35 plants per plot were harvested, sun dried for about 2 weeks and carefully threshed to compute the yield in tonnes/ha based on the plant population of 23,333 plants/ha used in this study. The total grain yield was estimated as per the relationship proposed by [15]: Grain yield/ha = Average grain yield/plant x Plant population/ha. (5)

The dry Stover yield was determined by randomly harvesting the leaves and stem of 5 maize plants from the middle rows of each plot at maturity and oven dried at 80°C for 4 days.

Economic analysis of the various Gliricidia sepium rates were measured using [16] formula: Revenue = Yield (t ha^-1) x Market price (Le ha^-1); Net revenue = Revenue - Production cost; Cost benefit ratio = Cost of production/Net revenue. (6)

2.5. Statistical Analysis

Data was subjected to analysis of variance using Proc GLM statement in Statistical Analysis System (SAS) 9.3 version. Mean separation was done using the Student Newman-Keuls (SNK) test at 5% level of probability. The relative frequency, relative density and relative importance value of weed species were analysed using descriptive analysis in excel.

3. Results and Discussion

3.1. Mulching Effects on Maize Growth and Development

Analysis of variance showed significant (P<0.05) effect of Gliricidia sepium leaves mulch on vegetative growth of maize (Table 1). Plots treated with 120kg ha^-1 mulch rate recorded the tallest plants followed by 90kg ha^-1 mulch rate, while the shortest plants were obtained in the control treatments (0kg ha^-1) in both months. Plant heights were not significantly different at 30kg ha^-1 and 60kg ha^-1 mulch rates for both months, but were significantly different (P<0.05) when compared to control treatment. As the rates of Gliricidia sepium leaf mulch increases plant heights maximized, this could be attributed to rapid decomposition of the G. sepium leaves required for plant growth. Similar result was reported by [6] that maize crop grew taller under greater G sepium leaf mulch levels. Furthermore, [17] and [18] confirms that Gliricidia sepium significantly increased the height of plants.

Plots treated with 120kg ha^-1 of Gliricidia sepium leaves mulch produced maximum number of leaves followed by 90kg ha^-1 mulch rate, and was significantly different (P<0.05) from those other treatments at all sampling periods. The significant increase of leaf number in the mulched plots over unmulched could imply that the mulch plots constituted higher mineral nutrients from decomposed mulched materials (Table 1). At 1 MAP, there were no significant difference between the application rates of 30kg ha^-1 and 60kg ha^-1 mulch rates on number of leaves plant^-1. The control plots however obtained significantly (P<0.05) the lowest number of leaves plant^-1 in both months (Table 1). This result agrees with the findings of [19] and [20] who reported that optimum rates of Gliricidia sepium leaf mulch will supply nitrogen and increase assimilation rate and building blocks of plant. Reference [21] similarly reported that the use of G. sepium mulches promotes vigorous foliage.

Stem girths observed at 30 and 60 kg ha^-1, and 90 and 120kg ha^-1 mulch rates were not significantly different at all sampling periods, though it shows variation in increased stem girth development of maize in both months. The highest stem girth was recorded with 120kg ha^-1 mulch rate while control plots produced the least at all sampling periods. This could be attributed to better nutrient uptake and development of the plants. This result agrees with [22] who found that G. sepium leaves mulch significantly improved plant growth in terms of leaf area, number of leaves plant^-1 and stem diameter.

Leaf area index was greatest at 120kg ha^-1 mulch rate (5.3 and 5.9), while the control plots produced the least leaf area index (1.7 and 3.4) respectively in both months. Leaf area index at 30kg ha^-1 and 60kg ha^-1 mulch rates were not statistically significant at 5% level of probability from each other compared to 1 MAP. Additionally, there was no significant difference (P>0.05) as indicated by Student Newman-Keuls multiple range test between 30, 60, 90, and 120kg ha^-1 of G. sepium rates at 2 MAP (Table 1).

Table 1. Effect of Gliricidia Sepium Leaves Mulch on Vegetative Growth and Development of Maize

3.2. Mulching Effects on Maize Yield and Yield Components

The results obtained in Table 2 shows that G. sepium leaf mulch applied at the rate of 120kg ha^-1 had significant (P<0.05) effect on number of cobs harvested and recorded the highest cob number (38.3) followed by 90kg ha^-1 (31.0), while 0kg ha^-1 produced the least cob number (10.0). This result may be due to adequate supply of nutrients from decomposed G. sepium leaf needed for proper cob development in maize. The result confirms the report of [23] that G. sepium leaf mulch could be as effective as commercial nitrogen fertilizer for yield response.

Table 2. Effect of Gliricidia Sepium Leaves Mulch on Yield and Yield Components of Maize

Plots treated with 120kg ha^-1 Gliricidia sepium leaf mulch significantly recorded the highest mean weight of dry cobs and grains (532.6g), while control plots (0kg ha^-1) gave the lowest (235.3g). Reference [24] reported that G. sepium leaf litter positively affect growth and yield of maize.

Analysis of variance showed that the application rate of 120kg ha^-1 Gliricidia sepium leaf mulch was statistically significant (P<0.05) and produced the highest total grain yield (8.5t ha^-1) over all other treatments. The non-mulched plots (0kg ha^-1) significantly (P<0.05) recorded lower grain yield (1.4t ha^-1). The prominent increase in the 30, 60, 90 and 120kg ha^-1 rates of Gliricidia sepium leaves mulch produced more than thrice the total grain yield obtained from 0kg ha^-1 treatments. The progressive increase of maize yield in Table 2 may be ascribed to the release of sufficient macro and micro nutrients during G. sepium leaf mulch decomposition. This result agrees with the reports of [25] and [26] that production of grain yield might be due to better growth, development and dry matter accumulation.

Table 2 shows that 1000 kernel weight under 120kg ha^-1 of Gliricidia sepium leaves mulch rate was maximum (195.6g) followed by 90kg ha^-1 (190.3g). While the minimum 1000 kernel weight (158.6g) was recorded under control treatments, which could be attributed to no input from Gliricidia sepium leaves mulch to improve on the yield. The observation from this study agrees with the findings of [16] and [27]. They reported that increase in 1000 grain weight was ascribed to maximum nitrogen use efficiency during grain filling, development and growth stages. Furthermore, decrease of 1000 grain weight in the control plots may be ascribed to low availability of nitrogen and other nutrients [26]. Earlier studies by [28] and [29] confirm that mulching practices are useful for the enhancement of maize grain quality and affects thousand grain weights.

Increasing the mulch rates from 0kg ha^-1 - 120kg ha^-1 resulted in corresponding increases in dry Stover yield (Table 2). The maximum dry Stover yield (27.3t ha^-1) was recorded at the rate of 120kg ha^-1 Gliricidia sepium leaves mulch. The 60kg ha^-1 and 90kg ha^-1 rates of Gliricidia sepium leaves mulch produced 21.3t ha^-1 and 23.3t ha^-1 dry Stover yield respectively, but were not statistically significant at 5% level of probability. Across the treatments, the 0kg ha^-1 Gliricidia sepium leaves mulch rate produced the least dry Stover yield 13.3t ha^-1. These observations were similarly explained by [30] that under field conditions, the rate of mulch decomposition and mulch quality may produce the highest dry matter yield.

Weed biomass was reduced in all mulched treated plots (Table 2). The application rates of 90kg ha^-1 and 120kg ha^-1 G. sepium leaves mulch recorded the highest weed suppressive ability than all other rates. This could be attributed to the thickness of the mulch which smoother weeds thus reduced the biomass of weeds. The highest weed biomass was produced by 0kg ha^-1 (24.0g) mulch rate indicating high competition between the crop and weed. Reference [8] reported similar result, that mulches have the ability to smoother weeds depending on their thickness.

3.3. Descriptive Variables of Weed Species

Seventeen different weed species were identified, with Ochthocosmus africanus, Newbouldia laevis, Ficus exasperata and Cyperus rotundus recorded the lowest relative frequency, relative density and relative importance value in the mulched plots compared to unmulched treatment (Table 3). Also, Spigelia anthelmia, Diodia scandens, Croton hirtus and Andropogum tectorum were more frequent and survived in all mulched treatments. While the unmulched plots produced the highest weed biomass and weed infestation (Table 3). This could be ascribed to the thick mulch layer of G. sepium leaves that retards weed seed germination and establishment of weeds by the potential allelopathic compounds found in Gliricidia sepium.

Table 3. Effect of Gliricidia Sepium Leaves Mulch on Relative Frequency, Relative Density and Relative Importance Value of Weed Species in Maize Production

3.4. Economic Evaluation of Gliricidia Sepium Leaf Mulch Rates on Weed Control in Maize

The partial budget in Table 4 shows cost and benefits ratio analysis of different rates of Gliricidia sepium leaf mulch as a weed control treatments in maize production. The total revenue from the different rates of G. sepium leaf mulch varied from Le 25,531,800.00 to Le 155,014,500.00 (USD 3.42 to USD 20.74). The application rate of 120kg ha^-1 G. sepium leaf mulch recorded the highest revenue Le 155,014,500.00 (USD 20.74) followed by 90 kg ha^-1 Le 116,716,800.00 (USD 15.62). The control treatment (0kg ha^-1) had the lowest revenue Le 25,531,800 (USD 3.42) compared to the other treatments. This could be attributed to the differences in yield ha^-1 with 120kg ha^-1 G, sepium mulch rate recording the highest yield.

Plots treated with 120kg ha^-1 G. sepium leaf mulch recorded the highest cost of production Le 940,000.00 (USD 125.77) followed by 90kg ha^-1 Le 915,000.00 (USD 122.42), while 0kg ha^-1 of G. sepium rate recorded the lowest cost of production Le 840,000.00 (USD 112.39). The highest net revenue Le 154,074,500.00 (USD 20.61) was obtained under 120kg ha^-1 of G sepium leaf mulch rate, while 0kg ha^-1 rate recorded the least Le 24,691,800.00 (USD 3.30).

In Table 4, the lowest cost benefit ratio (1:0.01) was recorded under 60, 90 and 120kg ha^-1 rates of G. sepium leaf mulch, while plots treated with 0kg ha^-1 (Control) rate obtained the highest (1:0.03). This indicates that the use of 60, 90 and 120kg ha^-1 rates of G. sepium leaf mulch were more profitable compared to other treatments in the production of maize.

Table 4. Partial Budget Analysis for Maize Production in 2016 Cropping Season

4. Conclusions

The study showed that Gliricidia sepium leaves mulch has the possibility of getting higher crop growth and yields as well as decreased weed growth in maize production. The results from this trial indicated that optimum growth and yield of maize would be achieved by applying Gliricidia sepium leaves mulch at 120kg ha^-1 rate during the rains. Thus, Gliricidia sepium leaves mulch could be a potential alternative to hoe weeding which is time consuming and costly. Additionally, the application of 60, 90, and 120kg ha^-1 rates of G. sepium leaf mulch were more profit table and economical for weed control in this study. However, 60kg ha^-1 of Gliricidia sepium leaf mulch is the most profitable and preferred due to its relatively low cost of production. This may be attractive to small scale farmers who may be ready to adopt it, since it does not involve technical grimness.

ACKNOWLEDGEMENTS

The authors of this paper acknowledge the contributions of Field Technicians at Njala Agricultural Research Centre (NARC), Njala, Sierra Leone.