Seed Dressing Maize with Imazapyr to Control Striga hermonthica in Farmers’ Fields in the Savannas of Nigeria

Use of small doses of imazapyr and pyrithiobac for seed coatings of imazapyr-resistant maize hybrids (IR-Maize) offers an effective means to control Striga hermonthica. Field trials were conducted in Bauchi and Kano States of Nigeria in 2014 and 2015 under heavy Striga infestation to evaluate the potential effectiveness of herbicide coated hybrids maize on Striga control in farmers’ field. Results showed that herbicide coated seeds reduced number of emerged Striga per m2 and Striga damage symptoms in farmers’ fields in all the locations. In Kano the number of emerged Striga was 4.9 to 7.9 times less in herbicide treated hybrids in comparison with those of the same hybrids planted without herbicide treatment. The Striga-resistant open pollinated variety (OPV) (TZL COMP1 SYN) had 6.7 to 8.0 times more Striga than the treated hybrids. In Bauchi, the number of emerged Striga on the untreated IR-maize hybrids were over four-times higher on the treated IR-maize hybrids than on the untreated hybrids. The Striga-resistant OPV check had four-times more Striga than the treated IR-maize hybrids and twice more than the untreated IR-maize hybrids across the two years. However, the effects of herbicide seed coating on grain yields were not consistent because of strong seasonal effects. The result revealed that coating of imazapyr-resistant hybrid maize seeds with imazapyr was effective in reducing Striga infestation in farmers’ fields. Although herbicide seed coating did not give consistent yield advantages of the hybrids over the untreated checks, a combination of herbicide seed treatment and genetic resistance to Striga would serve as an effective integrated approach that could significantly reduce the parasite seed bank from the soil and prevent production of new seeds. The IR-hybrids and the OPV checks contained Striga resistance/tolerant genes that protected them against drastic yield loss in the Striga infested fields in both Bauchi and Kano.


Introduction
Striga hermonthica constitutes one of the most severe constraints to cereal production in the semi-arid and sub-humid areas of sub-Saharan Africa [1]. An estimated 21 million ha of land is infested with S. hermonthica in Africa [2]. In Nigeria, over 40% of the 93 million ha arable land in the Savannas is already moderately or severely infested [3]. Surveys in the Northern Guinea Savannah of Nigeria (NGS) showed that S. hermonthica has remained a serious problem, attacking millet, sorghum (Sorghum bicolor L. Moench), maize (Zea mays L.) and upland rice [4][5][6]. In northeast Nigeria, over 85% of the fields planted to maize and sorghum were infested with Striga [7]. Field studies conducted in northern Nigeria showed that Striga incidence range from 0% to 100% in farmers' maize fields [8].
The increasing incidence of Striga has been attributed to poor soil fertility and structure, intensification of land-use through continuous cultivation and an expansion of cereal production [9]. It was reported that Striga population in northern Nigeria was negatively related to total nitrogen, exchangeable potassium and clay content [8]. They also reported that up to 75% of the variations in maize grain yields in farmers' fields could be explained by Striga population and soil organic carbon. The extent to which Striga reduces the growth of its host is highly variable and depends on factors such as host plant genotype, parasite infestation level, and environment [10]. Striga infestation of cereal crops has impacted negatively on the livelihoods of small-holder farmers in northern Nigeria because of its impacts on crop yield and systems productivity. Grain yield losses due to S. hermonthica infestation range from 10% to 100% for these crops, forcing farmers to abandon their cereal fields [3,11].
Crop rotation [12,13], intercropping [14], organic [15] and inorganic [16] fertilizers have all been recommended for Striga management in farmers' fields. However, these approaches require several seasons of repeated use before they begin to produce yield benefits [17]. Much of the Striga-infested area of Africa has very high levels of Striga seeds in the soil due to years of neglect (Ekeleme et al. 2014). Thus, there is an urgent need for cost-effective approaches that allow achievement of adequate crop yields while at the same time depleting the Striga seed bank in the soil for subsequent planting of cereal crops such as maize [18].
Studies have showed that the use of small doses of imazapyr and pyrithiobac for seed coatings offers an effective means to controlling Striga [19][20][21]. These herbicides inhibit the biosynthesis of branch-chained amino acids [22]. The use of herbicide seed coating reduces yield loss due to Striga, and depletes the Striga seed bank in the soil, so subsequent Striga numbers are less the following year [23]. The application of imazapyr and pyrithiobac as drenches or coatings to maize seeds possessing target-site resistance has been reported to effectively control Striga early in the season [20,22,24].
In order to use phytotoxic herbicides to control Striga and prevent crop injury, a gene for resistance to these herbicides was incorporated into tropically adapted maize germplasm [25] that already have natural resistance to S. hermonthica at the International Institute of Tropical Agriculture. Combining herbicide resistance with natural resistance to agronomic performance of IR-maize (imazapyr-resistant maize hybrids) minimized the risk of Striga damage and severe yield loss when the herbicide is washed away due to excess rain [22,25]. Imazapyr-coated seeds of Striga-resistant hybrids planted under S. hermonthica infestation sustained less than 20% yield loss, showed less Striga damage symptoms and supported very few emerged parasites than without seed coating [22,25]. These IR-maize hybrids can be planted in Striga-infested areas without imazapyr seed coating at certain intervals to delay the development of high levels of resistance to imazapyr and prolong the effectiveness of the herbicides [25]. However, IR-maize hybrids having natural resistance to Striga have been tested on-stations under artificial infestation with Striga. Large-scale testing of treated seeds of these hybrids on farmers' fields has not been reported for West Africa. Therefore, the objectives of this study were to evaluate Striga-resistant IR-maize hybrid treated with imazapyr under natural infestation with S. hermonthica in farmers' fields in two states in Nigeria.

Study Area
The study was conducted in two States Bauchi (10 • 18 N, 09 • 50 E) and Kano (12 • 00 N, 08 • 31 E) States in northern Nigeria, where Striga is endemic and infestation is high. Farmer-managed trials were conducted in 30 villages (Table 1) each in Bauchi and Kano States in Nigeria in 2014 and 2015. Participatory research and extension approaches were used as a basis for farmer involvement [26].
Each year, 30 farms on basis of their severe infestation based on observations were chosen in each State in the preceding season. The area is characterized by a mono-modal rainfall distribution with an average precipitation of 900-1300 mm and growing period of 150 to 180 days (May-October). Figure 1 shows the average monthly rainfall distribution in the two study areas during the experimental years. The predominant soil types are Alfisols of moderate to low fertility. Important crops are maize (Zea mays), sorghum (Sorghum bicolor), millet (Pennisetum typhoides), soybean (Glycine max), cowpea (Vigna unguiculata) and groundnut (Arachis hypogaea) with both sole and intercropping of legume and cereal being practiced.  Figure 1 shows the average monthly rainfall distribution in the two study areas during the experimental years. The predominant soil types are Alfisols of moderate to low fertility. Important crops are maize (Zea mays), sorghum (Sorghum bicolor), millet (Pennisetum typhoides), soybean (Glycine max), cowpea (Vigna unguiculata) and groundnut (Arachis hypogaea) with both sole and intercropping of legume and cereal being practiced.

Planting Materials and Coating
Two promising IR-maize (IR hybrid 1 and IR hybrid 4) top cross identified in on-station trials in 2013 and two non-IR OPVs (open pollinated variety) (TZL COMP1 SYN and DT STR SYN W) with known polygenic remittance to Striga were included in these trials. Imazapyr solution was prepared by dissolving solid imazapyr acid in distilled water and by gradually adding potassium hydroxide to raise the pH of the solution to 6 and 8. Polyvinylpyrollidone (1.8%) was also added as a binding agent to 21 mM of the prepared K-salt of imazapyr. Seeds of each maize hybrid were thoroughly mixed with the imazapyr solution. The seeds were allowed to soak for 24 hours to give a coating of 0.4 mg a.e imazapyr per seed (22). The treated seeds were dried and distributed to extension agents for planting within a week.

Planting Materials and Coating
Two promising IR-maize (IR hybrid 1 and IR hybrid 4) top cross identified in on-station trials in 2013 and two non-IR OPVs (open pollinated variety) (TZL COMP1 SYN and DT STR SYN W) with known polygenic remittance to Striga were included in these trials. Imazapyr solution was prepared by dissolving solid imazapyr acid in distilled water and by gradually adding potassium hydroxide to raise the pH of the solution to 6 and 8. Polyvinylpyrollidone (1.8%) was also added as a binding agent to 21 mM of the prepared K-salt of imazapyr. Seeds of each maize hybrid were thoroughly mixed with the imazapyr solution. The seeds were allowed to soak for 24 hours to give a coating of 0.4 mg a.e imazapyr per seed (22). The treated seeds were dried and distributed to extension agents for planting within a week.

Cultural Practices, Measurements and Statistical Analysis
Each of the 30 selected farmers' fields represented a replicate in each State in each year. In 2014, each on-farm trial was composed of two IR-maize hybrids (1 set treated and 1 set un-treated) along with an untreated-OPV. In 2015, each on-farm trial consisted of the same set of IR hybrids (1 set treated and 1 set untreated) along with two untreated OPVs. In each field, each hybrid or OPV was planted in a 10 × 10 m plot in the previously marked Striga infested fields. Four maize seeds were planted in rows 75 cm spaced apart at a spacing of 50 cm. At two weeks after sowing (WAS) maize seedling were thinned to two plants per hill. Other weeds were carefully removed using a hoe at two, six and eight WAS. Two weeks after sowing, fertilizer was applied at the rate of 50 Kg N ha −1 , 50 Kg P 2 O 5 ha −1 and 50 Kg K 2 O ha −1 using NPK (15:15:15). At five WAS, 50 Kg N ha −1 was again applied using urea. Data collected included the number of emerged Striga, Striga damage severity rating on a 1-9 scale at full silking (10 weeks after planting maize). The whole plots were harvested when the crop was fully mature and completely dry and cobs weighed. Grain yield was computed using 80% shelling of the maize cobs. As the fields were not artificially and uniformly infested and the trials were established on new plots every year, data recorded from all communities in each State were subjected to separate analysis of variance for each year in SAS using the model shown below: where Y ij is the observed yield at each location: µ is the overall mean for grain yield; L i is the effect of the ith location; H j is the effect of the jth hybrid; and e ij is the residual effect. Treatment means were compared using the LSD at 5% level of probability [27].

Results
Herbicide seed coating reduced number of emerged Striga per plot in farmers' fields in all the locations. In Kano in 2014, the number of emerged Striga was 4.9 to 7.9 times less in herbicide treated hybrids in comparison with those of the same hybrids planted without herbicide treatment. The Striga-resistant OPV (TZL COMP1 SYN) had 6.7 to 8.0 times more Striga than the treated hybrids ( Figure 2). When treated with herbicides in 2015, number of emerged Striga was 0.67 and 1.13 per m 2 for hybrid 1 and hybrid 4, respectively ( Figure 3). When untreated, the hybrids and the OPVs had number of Striga that were significantly higher than those of the treated hybrids. Number of emerged Striga was 5.19 per m 2 for hybrid 1 and 5.35 for hybrid 4. Numbers on the OPVs were 3.11 Striga per m 2 for DT STR SYN W and 2.03 Striga per m 2 for TZL COMP1 SYN. The numbers of Striga on the untreated hybrids were also higher than those on the two Striga-resistant OPVs (DT STR SYN W and TZL COMP1 SYN) (Figure 3).
In Bauchi in 2014, the number of emerged Striga on the untreated IR-maize hybrids were twice those on the treated IR-maize hybrids though the figures were not statistically different. The Striga-resistant OPV check had 4 times more Striga than the treated IR-maize hybrids and twice more than the untreated IR-maize hybrids (Figure 2). Similarly, in 2015, herbicide treated IR-maize hybrids had number of emerged Striga that was significantly lower than that on the untreated hybrids and the OPV checks ( Figure 3). When untreated, the number of emerged Striga plants was 7.3 times higher on hybrid 1 and 4.6 times higher on hybrid 4 than on the corresponding treated IR-maize hybrids. The OPV DT STR SYN W had 4.7 times more Striga than herbicide-treated hybrid 1 and 2.6 times more Striga than hybrid 4, while OPV TLZ COMP1 SYN had 3 times more Striga than the treated hybrid 1 and 1.6 times more Striga than the treated hybrid 4. When untreated the hybrids had number of Striga that were significantly higher than that of DT STR SYN W and TZL COMP 1 SYN. Striga damage was scored in both locations only in 2015. Striga damage rating on farmers' fields ranged from 4 to 5 in Kano and 3 to 5 in Bauchi (Figure 4). Herbicide seed treatment significantly reduced Striga damage scores in both locations. Untreated hybrids recorded damage scores that were similar to those of the untreated OPV checks in Kano. In Bauchi, the hybrid IR hybrid 4 had damage score that was significantly higher than the two Striga-resistant open-pollinated varieties (DT STR SYN and TZL Comp 1).
for hybrid 1 and hybrid 4, respectively (Figure 3). When untreated, the hybrids and the OPVs had number of Striga that were significantly higher than those of the treated hybrids. Number of emerged Striga was 5.19 per m 2 for hybrid 1 and 5.35 for hybrid 4. Numbers on the OPVs were 3.11 Striga per m 2 for DT STR SYN W and 2.03 Striga per m 2 for TZL COMP1 SYN. The numbers of Striga on the untreated hybrids were also higher than those on the two Striga-resistant OPVs (DT STR SYN W and TZL COMP1 SYN) (Figure 3). In Bauchi in 2014, the number of emerged Striga on the untreated IR-maize hybrids were twice those on the treated IR-maize hybrids though the figures were not statistically different. The Strigaresistant OPV check had 4 times more Striga than the treated IR-maize hybrids and twice more than the untreated IR-maize hybrids (Figure 2). Similarly, in 2015, herbicide treated IR-maize hybrids had number of emerged Striga that was significantly lower than that on the untreated hybrids and the OPV checks (Figure 3). When untreated, the number of emerged Striga plants was 7.3 times higher on hybrid 1 and 4.6 times higher on hybrid 4 than on the corresponding treated IR-maize hybrids. The OPV DT STR SYN W had 4.7 times more Striga than herbicide-treated hybrid 1 and 2.6 times more Striga than hybrid 4, while OPV TLZ COMP1 SYN had 3 times more Striga than the treated hybrid 1 and 1.6 times more Striga than the treated hybrid 4. When untreated the hybrids had number of Striga that were significantly higher than that of DT STR SYN W and TZL COMP 1 SYN. Striga damage was scored in both locations only in 2015. Striga damage rating on farmers' fields ranged from 4 to 5 in Kano and 3 to 5 in Bauchi (Figure 4). Herbicide seed treatment significantly reduced Striga damage scores in both locations. Untreated hybrids recorded damage scores that were similar to those of the untreated OPV checks in Kano. In Bauchi, the hybrid IR hybrid 4 had damage score that was significantly higher than the two Striga-resistant open-pollinated varieties (DT STR SYN and TZL Comp 1).  In 2014, IR-Hybrid 1 treated with imazapyr produced grain yield that was not significantly different from the grain yield of the untreated hybrid in Kano ( Figure 5). Herbicide treatment of IR-hybrid 4 however, produced grain yield that was significantly higher than that of the untreated hybrid. Grain yield of the OPV TZL COMP 1 SYN was significantly lower than those of the hybrids irrespective of herbicide treatment except for untreated hybrid 4. Grain yield of TZL COMP 1 SYN did not significantly differ from that of untreated hybrid 4. Despite the significant differences in number of Striga among the treated and untreated hybrids, differences in grain yield among these hybrids were not significant irrespective of hybrid treatment in Kano in 2015 ( Figure 6) probably due to poor rainfall (Figure 1) Additionally, both the treated and untreated IR-maize hybrids produced grain yields that were significantly lower than the OPVs (Figure 6). In 2014, IR-Hybrid 1 treated with imazapyr produced grain yield that was not significantly different from the grain yield of the untreated hybrid in Kano ( Figure 5). Herbicide treatment of IRhybrid 4 however, produced grain yield that was significantly higher than that of the untreated hybrid. Grain yield of the OPV TZL COMP 1 SYN was significantly lower than those of the hybrids irrespective of herbicide treatment except for untreated hybrid 4. Grain yield of TZL COMP 1 SYN did not significantly differ from that of untreated hybrid 4. Despite the significant differences in number of Striga among the treated and untreated hybrids, differences in grain yield among these hybrids were not significant irrespective of hybrid treatment in Kano in 2015 ( Figure 6) probably due to poor rainfall (Figure 1) Additionally, both the treated and untreated IR-maize hybrids produced grain yields that were significantly lower than the OPVs (Figure 6). In 2014, differences in grain yield among the maize hybrids were not significant in Bauchi irrespective of herbicide treatment ( Figure 5). Except for herbicide-treated hybrid 4 that produced grain yield that was not statistically different from the untreated OPV check TZL COMP1 SYN, all the hybrids produced grain yields that were higher than that of TZL Comp1 SYN. In 2015, the treated IR-maize hybrids produced grain yields that were significantly higher than those of the untreated IRmaize hybrids in Bauchi. Grain yields of the treated hybrids did not however, significantly differ from those of the untreated two OPV checks ( Figure 6).

Discussions
Although, crop rotation and intercropping involving legumes [13,28,29], application of organic In 2014, differences in grain yield among the maize hybrids were not significant in Bauchi irrespective of herbicide treatment ( Figure 5). Except for herbicide-treated hybrid 4 that produced grain yield that was not statistically different from the untreated OPV check TZL COMP1 SYN, all the hybrids produced grain yields that were higher than that of TZL Comp1 SYN. In 2015, the treated IR-maize hybrids produced grain yields that were significantly higher than those of the untreated IR-maize hybrids in Bauchi. Grain yields of the treated hybrids did not however, significantly differ from those of the untreated two OPV checks ( Figure 6).

Discussions
Although, crop rotation and intercropping involving legumes [13,28,29], application of organic and inorganic fertilizers [30], and the use of Striga-resistant cultivars [31,32] can partially reduce the Striga problem, no short-term control measure has been developed that subsistence farmers could use within their financial resources or that fits well into their traditional cropping systems. The use of imazapyr-resistant maize offers an opportunity to (i) control Striga itself so that adequate crop yields can be achieved each cropping season, (ii) deplete the Striga seed bank in the soil, (iii) is cost-effective; and (iv) is compatible with existing small-holder cropping systems [33]. Imazapyr seed coating provide an effective measure for controlling Striga [34] as a short-term measure until crop varieties with adequate levels of genetic resistance become available [18]. Results from our field trials confirmed the efficacy of imazapyr to reduce Striga infestation in maize. Untreated maize hybrids had over 4.4 Striga plants per m 2 across the two years in Bauchi and 4.7 in Kano. Number of emerged Striga per plot was over 2 to 7 times greater on untreated hybrids in Bauchi in 2014 and 2015. These results are in agreement with earlier results [22,25] who reported that herbicide treatments of seeds of the IR-maize hybrids considerably reduced Striga infestation in experimental plots in on-station trials in northern Nigeria. Similar reports of reduction in Striga infestation with herbicide seed coating have been obtained in East and Southern Africa [34,35].
In general, all IR-maize hybrids produced comparable yields regardless of herbicide treatment in Kano in 2014, possibly because the hybrids also possess field resistance to Striga. Unlike the results in 2014, maize hybrids produced lower grain yields than the OPV checks in Kano in 2015 mainly due to the drought stress that occurred in the area. Rainfall in the zones was low ( Figure 4) and its distribution was poor which adversely affected the performance of the hybrids. The heterogeneous nature of the two OPVs used as checks in 2015 may better impact resistance under variable growing conditions with resistance to Striga [36]. In Bauchi, response of maize grain yield to herbicide treatment was not consistent. In 2014, IR-maize hybrids produced similar grain yields regardless of the use of herbicide seed treatment, confirming resistance of these hybrids to Striga infestation. In 2015, hybrid treated with herbicides produced grain yields that were higher than those of the untreated hybrids. This result shows that when rainfall distribution during the cropping season is good, herbicide seed treatment will give yield advantage over untreated hybrid. This was the case of seed treatment in Bauchi in 2015. Rainfall (Figure 1) during the cropping period of July-September was good and well distributed to support maize growth. Although there was no consistent yield advantage across locations and years associated with the herbicide seed treatment, the number of emerged Striga was significantly lower in plots planted to herbicide-coated hybrid suggesting that planting herbicide-coated seed will contribute to reduce Striga infestation and seed bank of Striga. The lower dose requirement of the herbicide seed treatment makes it ecologically sensible to use in reducing Striga infestation on farmers' fields. De Groote [37] reported a very good marginal rate of return for the IR technology when the cost of herbicide was 4 US dollars per ha −1 .

Conclusions
The results of this study showed that coating IR maize hybrid seeds with imazapyr was effective for Striga control. There was however no consistent yield advantage of herbicide seed coating over the untreated hybrids because the hybrids were tolerant to Striga. Combination of herbicide seed treatment and genetic resistance to Striga would serve as an effective integrated approach that would reduce the parasite seed bank from the soil and prevent production of new seeds. The IR-maize hybrids and the OPV checks containing Striga resistance genes did not suffer from drastic yield losses in Striga infested fields in both Bauchi and Kano.