Evaluation of herbicide mixtures and manual weed control method in maize (Zea mays L.) production in the Southern Guinea agro-ecology of Nigeria

Abstract: Field trials were conducted in 2015 and 2016 cropping seasons to evaluate some herbicide mixtures and manual weed control method in the production of maize in the southern Guinea savanna of Nigeria. The experiment consisted of 10 treatments as follows: Metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha, metolachlor + atrazine at 2.0 + 2.5 kg a.i./ha, metolachlor + atrazine at 3.0 + 3.0 kg a.i./ ha, pendimethlin + atrazine at 1.0 + 2.0 kg a.i./ha, pendimethlin + atrazine at 2.0 + 2.5 kg a.i./ha, pendimethlin + atrazine at 3.0 + 3.0 kg a.i./ha, metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha plus one supplementary hoe weeding (SHW) at 6 weeks after sowing (WAS) and pendimethlin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS, hand weeding at 3 and 6 WAS and a weedy check. These treatments were laid out in randomized complete block design with three replicates. Results showed that metolachlor + atrazine and pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS significantly reduced weed infestation and gave higher maize grain yield and economic returns. These methods are therefore recommended to farmers as alternative to two hand weeding at 3 and 6 WAS.


ABOUT THE AUTHOR
The author's name is E.O. Imoloame. He is a senior lecturer and a weed scientist in the Department of Crop Production, College of Agriculture, Kwara State University, Malete, Kwara State, Nigeria. His research area is Weed management in crops. He has done a lot of work on weed management in Nigeria and he is an advocate of Integrated Weed Management as the most effective and efficient solution to all weed problems. He has also conducted some research in weed ecology, economics and extension. He has plans to carry out research on the use of plant parts with allelopathic potentials to control weeds. This research report will be useful to Africans, Asians and South Americans and the tropical countries in the world that are still using manual weed control methods for weed control in maize.

PUBLIC INTEREST STATEMENT
My research seeks to find a method of weed control that can serve as an alternative to two hoe weeding at 3 and 6 weeks after planting which will significantly reduce weeds and promote higher maize yield. The manual weeding is very strenuous, unreliable, expensive and causes drudgery. These, de-motivate most of the youth population from venturing into farming which has been left for the aged and unproductive population of Nigeria.
The result of this research has revealed that a herbicide mixture of metolachlor and atrazine at 1.0 + 2.0 kg active ingredient per one hectare integrated with a supplementary hoe weeding at 6 weeks after planting is a suitable alternative to two hoe weeding (manual weeding). The aforementioned herbicide mixture is not only environmentally friendly, but will reduce drudgery and give higher yield and economic returns to farmers.

Introduction
Maize (Zea mays L.) is the third most important cereal crop in the world after wheat and rice (MINFAL, 2003). Currently more of this crop is produced annually than any other grain and is the most important cereal crop in sub-Saharan Africa and Latin America (IITA, 2012). Maize provides staple food to large number of human population in the world. In the developing countries it is a major source of income to many farmers (Tagne, Fenjic, & Sonna, 2008).
According to FAO (2011), 822.7 million metric tonnes of maize were produced worldwide in 2008. Out of this figure, Africa produced 53.2 million metric tonnes, while Nigeria produced 7.3 million metric tonnes in 2009. Despite its importance, the yield of maize obtained in Nigeria is far below expectation due to numerous factors which include weed infestation, low soil fertility and availability of labour. Yield losses of between 60-80% have been attributed to uncontrolled weed infestation in maize (Lagoke, Adeosun, Elemo, Chude, & Shebayan, 1998) and this finding was confirmed by Imoloame and Omolaiye (2016), who reported 89% yield loss in maize as a result of uncontrolled weed infestation.
Manual weeding is the commonest method of weed control in Nigeria. The traditional method is back-breaking which offer little hope for expanding the present farm size. Hoe weeding is labour intensive, expensive and strenuous. Also, labour availability to carry out hand weeding is uncertain, thus making timeliness of weeding difficult to attain. This has resulted in the loss of yield (Adigun & Lagoke, 2003). It is estimated that about 40-60% of production cost is spent on manual weeding (Remison, 1979) which is similar to the report of Ekeleme (2009) that 25-55% of the total cost of production is spent on labour and weeding operations.
Chemical weed control is a practical and economic, alternative to hand weeding. If herbicide is applied appropriately it could prevent weed infestation from planting to harvesting and promote higher yields by allowing closer crop spacing and therefore higher plant population. The efficiency of chemical weed control in increasing the yield of maize and other crops and reducing labour cost in the tropics especially in Nigeria have been documented (Akobundu, 1987;Imoloame, 2014;Joshua & Oni, 2002;Ogungbile & Lagoke, 1986).
Though chemical weed control has many advantages over hoe weeding, there is the possibility of reducing the herbicide rates in order to cut cost and mitigate the problem of environmental build up of herbicide residues and herbicide resistant weeds. This calls for Integrated Weed Management (IWM) strategy which is the combination of two or more weed control methods for more effective and efficient weed control than the a single method. This approach considers the use of cultural, mechanical and chemical control options that are both feasible in specific cropping systems and permitted by socio-economic conditions (Ganie, Singh, & Singh, 2014;Norsworthy et al., 2012;Vencill et al., 2012) Most of the available research carried out on methods of weed control in maize have been in the northern Guinea savanna of Nigeria. Also, the high cost of weed control coupled with the high labour demand of hoe weeding and the need to protect the environment has driven the desire for a method of weed control that will not only be safe, effective and efficient in minimizing weed density, but will also lead to higher maize grain yield.
The objectives of this research are to determine the weed control method that will result in effective and efficient weed control and also give higher grain yield of maize.

Materials and methods
A field experiment was conducted during the 2015 and 2016 rainy seasons at the Teaching and Research (T&R) Farm of Kwara State University, Malete (lat. 08°, 71′N; long.04°44′E) at 360 m above sea level. The soil at experimental site was sandy loam and slightly acidic. The nitrogen and available phosphorus content of the soil was low and inadequate (Table 1). The experiment consisted of 10 treatments as follows: Metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha, metolachlor + atrazine at 2.0+2.5 kg a.i./ha, metolachlor + atrazine at 3.0 + 3.0 kg a.i./ha, pendimethlin + atrazine at 1.0+2.0 kg a.i./ha, pendimethlin + atrazine at 2.0 + 2.5 kg a.i./ha, pendimethlin + atrazine at 3.0 + 3.0 kg a.i./ha, metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha plus one supplementary hoe weeding (SHW) at 6 weeks after sowing (WAS) and pendimethlin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS, hand weeding at 3 and 6 WAS and a weedy check. These treatments were laid out in a randomized complete block design (RCBD) and replicated three times. The maize variety that was used was SUWAN-1-SR which was sown on the 11th and 14th of July, 2015 and2016 respectively. The crop was spaced at 75 cm × 25 cm to give a plant population of 53,333/ha. Herbicides were applied a day after planting with a CP15 knapsack sprayer and a green nozzle which were calibrated to deliver a spray volume of 250 l/ha. Karate insecticide containing 2.5% lamdacyhalothrin as active ingredient was applied at the rate of 30 ml in 10 l of water three times beginning from 6 WAS, to control army worm. Fertilizer was applied at the rate of 120 kg N, 60 kg P 2 O 5 and 60 kg k 2 O. These were provided with a compound fertilizer 15:15:15. Harvesting of maize was done on a net plot of 9 m 2 after the row at the edges at both sides of the plots were discarded to reduce error. The parameters measured included the following:

Weed density
This was determined at 6 and 12 WAS, by counting the total number of weed species per unit area (quadrat) in each plot.

Weed cover score
This was visually assessed at 6 and 12 WAS using a scale of 1-10, where 1 represented no weed cover and 10 complete weed cover.

Weed dry weight (kg/ha)
This was obtained by taking weed samples at random from a 1 m 2 quadrat placed randomly in 3 locations in each plot at 6 and 12 WAS. The weeds were gathered together and put in a polythene bag and later oven-dried at a temperate of 80°C for 2 days to a constant weight. The oven-dried weight in gramme was converted to kilograme/ha for each plot.

Relative importance value
The Relative importance value (RIV%) of each species infesting the experimental plots was determined after the weeds were collected from the quadrat and before they were oven dried. The RIV was computed as the mean of the percentage of relative frequency and relative density for each species as indicated in the formula below (Wentworth, Conn, Skroch, & Mrozek, 1984).
Relative density (RD) was determined by dividing the total number of individuals of a weed species in all the quadrats by the total number of individual of all the weed species in all the quadrats multiplied by 100. The percentage relative frequency was calculated as the number of occurrence of a species in all the quadrats divided by the total of occurrence of all species in all the quadrats multiplied by 100 (Das, 2011).

Plant height
This was determined by measuring the height of 5 randomly selected plant per plot at 6, 9 and 12 WAS, using a meter rule from the soil level to the apical bud of the plant.

Leaf area (cm 2 )
The leaf area was determined at 9 and 12 WAS by measuring the length and breath of the top, middle and bottom leaves of five randomly selected plants from each plot and the average of these measurements was multiplied by a factor 0.75 to give the leaf area/plant (Moll & Kamprath, 1977).

Grain yield
This was determined by weighing the grains harvested from each net plot which was converted to kilograms per hectare using the formula below:

Data analysis
The data collected was subjected to analysis of variance using Assistat 7.7, 2017 version Statistical Package and were F value was significant, the means were separated using the Duncan's Multiple Range Test (DMRT) at 5% level of probability.

Economic analysis
Information on the cost of all the cultural practices from land preparation to harvesting and processing was collected from Kwara State Agricultural Development Programme (KWASADP), Ilorin, an agency responsible for extension services in Kwara State, Nigeria. The price of 1 kilograme of maize was obtained from the open market to calculate the income/total revenue. The economic analysis was carried out using partial budgeting (Okoruwa, Obadaki, & Ibrahim, 2005) to calculate the gross marging (profit) as follows: Grain yield kg/ha = Grain yield/Net plot × 10, 000 Net plot (m 2 ) where GM = Gross margin/ ha for each treatment; TR = Total revenue, Naira/ha for each treatment; VC = Variable cost, Naira/ha for each treatment; Ys = maize grain yield (kg// ha) for each treatment; Ps = Price of maize per kg; M = Value of material input (seeds, fertilizer, insecticide, herbicides); L = Value of Labour (land preparation, planting, insecticide and herbicide, fertilizer application, harvesting, processing and packaging).

Rainfall
The total amount of rainfall recorded in 2015 was 1,010.5 mm, with the month of September having the highest rainfall, while January, February, April and August had low rainfall. In 2016, higher rainfall of 1,493.4 mm was recorded which was evenly distributed (Figure 1).

RIV% at the experimental site
The relative importance value of weed species infesting the maize crop under each treatment is presented in Table 2. Paspalum scrobiculatum was the most dominant weed species both within and across all the treatments at 6 WAS in 2015. This weed species, followed by Rottboellia cochinchinensis and Mariscus alternifolius in the descending order were the most prominent under pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha. A total of 7 weed species were recorded under this treatment at 6 WAS. This increased to 11 including Paspalum scrobiculatum, Setaria barbata, Hyptis suaveolens and Commelina benghalensis in the descending order as the most prominent at 12 WAS (Table 3). At higher application rates of pendimethalin + atrazine, Paspalum scrobiculatum and Maniscus alternifolius were the most prominent and important weed species growing in maize plot at 6 WAS. However at 12 WAS other weed species emerged, the most prominent among them were Kyllinga erecta, Paspalum scrobiculatum, Hyptis suaveolens and Cyperus esculentus in the descending order in the plots treated with pendimethalin + atrazine at 2.0 + 2.5 kg a.i./ha and Kyllinga erecta Paspahum scrobiculatum and Commehlina benghalesis under pendimethalin + atrazine at 3.0 + 3.0 kg a.i./ha. The total number of weed species under these treatments increased from 6 and 9 at 6 WAS ( Table 2) to 8 and 10 respectively at 12 WAS (Table 3).
The most dominant weed species under metolachlor + atrazine at all the rates were Paspalum scrobiculatum, Setaria barbata and Rottboellia cochinchinensis at 6 WAS, while at 12 WAS, Paspalum scrobiculatum and Setaria barbata maintained their dominance across these treatments. Other weed species that were predominant under metolachlor + atrazine at 1.0 + 2.0 and 2.0 + 2.5 kg a.i./ ha included Digitaria horizontalis and Gomphrena celosoides and Hyptis suaveolens under   metolachlor + atrazine at 3.0 + 3.0 (Table 3). Also there was an increase in the number of weed species infesting the maize at 12WAS (Table 3). The weed species that were most prominent under the plot treated with pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6WAS were Paspalum, scrobiculatum, and Vernonia galamensis at 6 WAS. However at 12 WAS Paspalum scrobiculatum, Vernonia galamensis and Cyperus rotundus were more prominent. Paspalum, scrobiculatum, Setaria barbata and Digitaria horizontalis were the most prevalent weeds species under metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha + one SHW at 6 WAS, however at 12 WAS, the dominant weeds were Paspalum scrobiculatum, Gomphrena celosiodes and hyptis suaveolens (Table 3).
Paspalum scrobiculatum, Pycreus lanceolatus and Chloris pilosa were the most important weed species under two hand weeding at 6 WAS, however at 12 WAS, Paspalum scrobiculatum and Setaria barbata were predominant. Under the weedy plot, the most important weeds were Paspalum scrobiculatum Ludwigia decurrens, Pycreus lanceolatus and Mariscus alternifolius, however at 12 WAS, Setaria barbata, Digitaria horizontalis and Paspalum scrobiculatum constituted the most dominant weeds (Table 3).
At 6 WAS, Paspalum scrobiculatum was the most dominant weeds across the treatments followed by Mariscus altenifoluis, Rottboelia cochinchinensis and Setaria barbata in descending order (Table 2). However at 12 WAS, the same trend was recorded with Paspalum scrobiculatum occurring as the most dominant weeds followed by Setaria barbata, Gomphrena celosoides and hyptis suaveolens. (Table 3).

Effect of herbicide mixtures and manual weed control on weed dry matter and density in maize
Pre-emergence application of pendimethalin + atrazine and metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS caused a significant reduction in weed dry matter than the other methods of weed control, however, this effect was comparable to two hand weeding at 3 and 6 WAS and the other treatments at the mean, except weedy check which supported significantly higher weed dry matter (Table 4). Later in the season at 12 WAS, the two herbicide mixtures plus one SHW at 6 WAS and hand weeding at 3 and 6 WAS maintained significantly lower weed dry matter than the other treatments in both years and their mean (Table 4). Metolachlor + atrazine and pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS and hand weeding at 3 and 6 WAS were more effective in significantly reducing both weed density and weed cover compared to the other treatments (Table 5).

Effect of herbicide mixtures and manual weed control method on the growth of maize
All the herbicide treatments increased plant height significantly than the weedy check at 6 WAS, however, with time at 12 WAS, hand weeding at 3 and 6 WAS gave significantly taller plants than the other treatments except metolachlor + atrazine and pendimethalin at 1.0 + 2.0 kg a.i./ha and metolachlor + atrazine at 3.0 + 3.0 kg a.i./ha, which produced comparable taller plants at the mean (Table  6). Pre-emergence application of pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW resulted in leaf area significantly larger than the weedy check but was comparable to the other weed control methods including two hand weeding at 9 WAS (Table 7). However, pendimethalin + atrazine and metholachlor + atrazine at 1.0 + 2.0 plus one SHW, all the rates of metolachlor + atrazine and two hoe weeding gave significantly larger leaves in both years and their mean at 12 WAS.

Effect of herbicide mixtures and manual weeding on grain yield of maize
Two hoe weeding at 3 and 6 WAS produced grain yields that were comparable to metolachlor + atrazine and pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS, but was significantly higher than the rest of the treatments and the weedy check in both years and their means (Table 8).

Table 4. Effect of herbicide mixtures and manual weed control on weed dry matter (kg/ha)
Notes: Data presented as means ± SD. P + A = Pendimethalin + Atrazine; M + A = Metolachlor + Atrazine; SHW = Supplementary hoe weeding; Means ± standard deviation.

Economic assessment of the use of different methods of weed control
The highest grain yield of maize (2,814.8 kg/ha) was obtained from plots treated with pre-emergence application of metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS followed by pendimethalin + atrazine at 1.0 + 2.0 kg a.i.ha /ha plus one SHW at 6 WAS, while the least yield (1,862.9 kg/ha) was produced by metholachlor + atrazine at 2.0 + 2.5 kg a.i./ha in 2015 (Table 9). However in 2016 and the mean, hand weeding at 3 and 6 WAS resulted in the highest maize yield (3,028.3 and 2,782.7 kg/ha) followed by metolachlor + atrazine at 1.0 + 2.0 kg a.i./ha plus one supplementary hoe weeding at 6 WAS (1,956.6 and 2,385.7 kg/ha).

Table 8. Effect of herbicide mixtures and manual weed control on grain yield
Notes: Data presented as means ± SD. P + A = Pendimethalin + Atrazine; M + A = Metolachlor + Atrazine; SHW = Supplementary hoe weeding; Means ± standard deviation. https://doi.org/10.1080/23311932.2017.1375378 Table 9. Economic assessment of herbicide mixtures and manual weed control in the production of maize, 2015 and 2016 Notes: Calculation of total revenue is based on

Effect of herbicide mixtures and manual weeding on relative importance value
Paspalum scrobiculatum appears to be the most predominant weed species infesting maize within and across treatments at 6 and 12 WAS in maize plot. This could be as a result of the inability of the treatments to fully control this weed species which were also well adapted to the environment. The adaptive capacity of this weed species made it more persistent and competitive with the maize crop. This is in line with the findings of Imoloame and Omolaiye (2016), that weed species with the highest relative importance value in maize were Paspalum scobiculatum and Digitaria horizontalis. Since grass weeds have been reported to be more competitive and damaging in grass crops (Anonymous, 2007) the significant reduction in the yield of maize in the weedy check could have resulted from the predominance of Paspalum scrobiculatum. There was an increase in the number of weeds Species at 12 WAS under each treatment. This could have resulted from the germination of more weed species with time as the effect of the herbicides expired. Also, the appearance of broadleaved weeds as dominant weed species at 12 WAS suggest that broadleaved weed flushes comes up later in the season probably because they are buried at a greater depth of the soil. Deat, Sement, and Fontenay (1980) reported that 60-75% of total grassy weeds as against only 30-35% broadleaved weeds emerged during first 15 days of an intensively cultivated field in Ivory Coast.

Effect of herbicide mixtures and manual weed control method on weed infestation
The ability of metolachlor + atrazine and pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW and two hand weeding at 3 and 6 WAS to significantly reduce weed dry matter, weed density and weed cover proves the effectiveness and efficacy of these weed control methods. These different herbicide mixtures plus one SHW can be used in rotation for effective weed control in maize. The integration of herbicides with one supplementary hoe weeding have been found to be very effective in the control of weeds and promoting higher yields in various crops (Imoloame, 2014;Peer et al., 2013;Veeramani, Palchamy, Ramasamy, & Rangaraju, 2001).

Effect of herbicide mixtures and manual weed control method on the growth of maize
Two hand weeding at 3 and 6 WAS and the two different herbicide mixtures integrated with one SHW at 6 WAS resulted in significantly taller plants than the other treatments. This was probably due to their ability to significantly reduce weed infestation than the other treatments which could have minimized weed competition and made sufficient growth resources (moisture, plant nutrients, light) available for utilization and better performance by maize crop. Also, the larger leaf area of the maize plants produced from plots treated with metolachlor + atrazine and pendimethalin + atrazine plus one SHW is an additional proof of their efficacy to promote effective weed control and the utilization of growth resources for better growth. The larger leaf area confer advantage to maize as it provides a larger surface for the capture of more solar radiation for increased photosynthesis and higher yield.

Effect of herbicide mixtures and manual weeding on grain yield
Higher maize grain yield was produced by metolaclor + atrazine and pendimethalin + atrazine at 1.0 + 2.0 kg a.i./ha plus one SHW at 6 WAS and two hand weeding at 3 and 6 WAS compared to the other weed control methods probably as a result of better weed control provided by these treatments which gave rise to better growth, development and higher grain yields. The weedy check gave significantly lower yields due to the promotion of significantly higher weed dry matter, weed density and weed cover which led to more intense weed competition for sunlight moisture and plant nutrient which resulted in lower grain yield. This result is similar to that of Imoloame (2014) and Veeramani et al. (2001), who reported increase in grain yield as a result of the use of herbicide application plus one SHW.