Soil Maize Cultivar-related Challenges on Striga hermonthica Infested Fields in Western Kenya

Maize production in Western Kenya is constrained by Striga hermonthica and declining soil fertility. Integrated Striga Management (ISM) packages have been proposed. An ISM field experiment assessed combination of 4 maize varieties with 5 levels of soil fertility amendments. Imazapyr Resistant (IR) maize and local yellow seed Shipindi had highest germination percentages of 90% and 81% respectively, compared to commercial white seed Duma and local white seed Rachar. Duma had significantly large plants in terms of leave size and plant height; and taking least time to silking and tasseling while producing heaviest cobs and grains per plant. Synthetic fertilizer (DAP+CAN) was associated with the least germination percentage, but produced the largest plants with many leaves, took the shortest time to silking, and produced highest cob weight and grain weight, with very low S. hermonthica infestations regardless of the maize varieties. Cattle manure (CM) and water hyacinth compost containing cattle manure culture (HCM) and Effective MicrobesTM (HEM) had the highest S. hermonthica population per unit area. Maize grown with water hyacinth compost containing Effective MicrobesTM (HEM) positively influenced cob weight than those receiving cattle manure (CM) and the controls; while being associated with the highest numerical increase in grain yield/area. Alternative soil fertility interventions based on these observations are therefore proposed.

Unfortunately, no single weed control measure has been successful in controlling S. hermonthica when applied in isolation (Marley, Aba, Shebayan, Musa, & Sanni, 2004). Integrated Striga Management (ISM) systems are therefore being encouraged to control S. hermonthica in Africa (Schulz, Hussaini, Kling, Berner, & Ikie, 2003;Avedi et al., 2014). The use of IR maize, or yellow maize that is tolerant to S. hermonthica, in combination with water hyacinth composts, could be beneficial for ISM in Western Kenya. Evaluating this ISM approach was the objective of the present study. Our goal is to share information that could create awareness and help mobilize research resources for addressing S. hermonthica and soil fertility related problems affecting farmers in Western Kenya.

Experimental Design
The experiment consisted of a 4 × 5 factorial treatment design comprising four maize seed types and five types of soil fertility amendments, on a piece of land measuring 32.7 m × 15.9 m. The resulting 20 treatment combinations comprised of plots in the form of 8.7 m long lines spaced at 0.7 m; each having 30 planting holes spaced at 0.3 m. These were laid out in a completely randomized block design of 3 blocks (replicates). The blocks that measured 13.3 m × 8.7 m were spaced at 2 m from each other and 1.3 m from the field boundary. The experiment was conducted during the long rain season (May to July 2014) and repeated once in time, during the short rains (August to November 2014). Climatic data (rainfall) for this region during the study period can be found in Midega et al. (2015a).

Maize and Striga Seeds
Seeds of Imazapyr Resistant (IR) maize hybrid (FRC 425-IR) (Freshco Seeds, Nairobi, Kenya), pre-treated with the imidazolinone herbicide (De Groote et al., 2007), and those of Duma (SC Duma 43) white seed maize variety (Agri-Seed Company Ltd, Nairobi-Kenya), were purchased from Agrovet shops in Busia town, Kenya. The local yellow seed Shipindi variety and the white seed Rachar variety were purchased from the Busia town market in Kenya. Agronomic performances of these local maize varieties have been reported by Ojiem et al. (1996) and Hassan (1998). S. hermonthica seeds (0.5 kg) were obtained from Striga harvest stocks at KALRO-Kibos station in Kisumu, and formulated into Striga seed-sand mixture (1:4) (Berner et al., 1997).

Planting and other Agronomic Practices
Land was prepared by hand-digging, and planting holes (~10 cm diameter and 5 cm deep) excavated using a hoe. The HCM, HEM and CM composts were applied volumetrically using containers of 150 mL per hole as per the respective treatments and mixed with soil (Naluyange et al., 2014). The water hyacinth composts therefore supplied approximately 0.9 g N, 0.02 g P, and 0.99 g N, 0.03 g P, for HEM and HCM, respectively (Naluyange et al., 2016). DAP fertilizer was applied at the rate of 2 level teaspoonfuls per planting hole (~10 g) and mixed with soil; supplying approximately 2.1 g N and 2.3 g P. One maize seed per hole was planted at a depth of ~3 cm, together with a table spoonful of S. hermonthica seed-sand mixture (approx. 1000 seeds) in every planting hole as described by Berner et al., (1997). The DAP treated plants were later top-dressed with CAN once as per the respective treatments at the rate of 2 level teaspoonfuls per maize plant (~10 g), 42 days after planting, supplying ~1.3 g N. Hand weeding was done after every two weeks for all weeds, except S. hermonthica (Avedi et al., 2014).

Data Collection
Number of days taken for each maize seed per planting hole to emerge was recorded. The number of emerged maize seedlings out of the total planted seed population per treatment was used to calculate germination percentage. Maize plant growth-related parameters that included plant height, leaf length, leaf width and number of leaves were scored every two weeks until tasseling (10 weeks). Maize plant height from the base of the stem to the apex of the youngest leaf was recorded. The number of open leaves per maize plant, as well as length (base to apex) and width (widest part) of the youngest open leaves were recorded. The number of days from planting to silking, tasseling and physiological maturity of each maize plant was recorded. Cob weights per plant were recorded. Yield-related data that included average grain weight per plant was recorded and expressed as grain weight per unit area.
Number of days from planting to the emergence of the first S. hermonthica seedling per host plant was recorded. Population of the parasitic weed per maize plant was recorded daily and used to calculated Striga weed population per unit area (counts per m 2 ) (Berner et al., 1997); the number of days to flowering of S. hermonthica was recorded when the first flower appeared per host plant.

Statistical Analysis
Data was analyzed using SAS 9.1 software (SAS Institute Inc.) at P < 0.05 significance level. Proc Means was used in generating means and standard errors for maize germination percentage, developmental time, plant size and yield-related parameters, as well as S. hermonthica developmental time, population per maize plant and per unit area. The two-week scores for plant height, leaf number and size were averaged. Mean germination percentage for maize was computed from individual percentages of the three treatment replicates per season generated by Proc Freq. S. hermonthica population was expressed as mean counts/m 2 using sums from the three treatment replicates per season, divided by the 6.3 m 2 plot size (0.7 m × 0.3 m × 30 planting holes). Analysis of variance (ANOVA) between treatments and between seasons was done using Proc GLM with mean separation by Tukey's Studentized Range (HSD) test when there were significant differences.
Number of days to maize seed emergence did not vary between the four maize seed varieties and the five soil fertility amendments (P > 0.05) ( Table 1). DAP+CAN treated maize plants took the shortest period in days to silking, while the controls took the longest period (P < 0.0001) ( Table 1). Duration to silking (days) was shortest in Duma variety and longest in IR maize (P < 0.0001) ( Table 1). Total days to maturity did not vary between the maize varieties and the soil fertility amendments (P > 0.05) ( Table 1).

Maize Plant Size and Grain Yields
DAP+CAN treated maize plants were the largest in terms of height, number and size of leaves; the controls produced the smallest plants (P < 0.05) ( Table 1). There was no difference in number of leaves per plant between the four maize varieties (P > 0.05) (Table 1); with the maximum number of leaves at 10 weeks being 17, 16, 17 and 15, for IR maize, local yellow Shipindi, local white Rachar and improved white Duma respectively. Duma maize variety had the largest leaves in terms of length and width, followed by the local white Rachar, while IR maize and the local yellow Shipindi had the smallest leaves (P < 0.0001) ( Table 1). The maximum size of leaves (length/width) at 10 weeks being 64/10, 69/10.5, 75/9 and 78.3/8.5 for IR maize, local yellow Shipindi, local white Rachar and improved white Duma respectively. IR maize variety had the shortest plants while the other three varieties were of similar height (P < 0.0001) (Table 1); with the maximum height at 10 weeks being 164.5, 220, 220 and 180 centimeters, for IR maize, local yellow Shipindi, local white Rachar and Improved white Duma respectively. Plants in the second season were larger than those in the first season in terms of height, number and size of leaves (P < 0.05) ( Table 1).  (Table 1). However, these differences in seed weight between soil fertility amendments were not reflected per unit area (P > 0.05; Table 1).
Duma maize variety produced the heaviest cobs compared to the cob weights of the remaining three varieties, which were similar (P = 0.0006) ( Table 1). Grain weight per plant was highest in Duma variety and lowest in IR maize, while the two local varieties were intermediate (P = 0.0005) (Table 1). However, these differences in grain weight between maize varieties did not reflect in weights per unit area (P > 0.05; Table 1).

Striga Hermonthica Growth and Population
Per unit area, S. hermonthica population was low on plots that received DAP+CAN and high when the CM, HCM and HEM were applied, regardless of the maize variety (P < 0.0001) (Figure 1). S. hermonthica population per unit area did not vary between seasons (P > 0.05). Duration between planting and Striga emergence (54.3 ± 1.8 days) did not vary between the treatments (P > 0.05), but was shorter in the first season (44.1 ± 1.9 days) than in the second season (61.5 ± 0.9 days) (P ˂ 0.0001). Duration between Striga emergence and flowering (31.3 ± 0.7 days) did not vary between treatments and seasons (P > 0.05).

Discussion
Maize varieties in the present study yielded quite below their inherent potential. For instance, IR maize only attained 1.1 ton/ha, yet under ideal soil fertility conditions on farmers' fields, this variety has been found to yield 2.5 ton/ha (Kanampiu & Friesen, 2004). The application of DAP+CAN fertilizers showed potential of improving the size of maize plants. For example, DAP+CAN treated maize plants were 62cm tall, had 8 leaves, with leaf size of 48cm long and 8cm wide; while the controls had the smallest and shortest plants. However, the relevance of the stimulative effect of the DAP+CAN was not reflected in the yield per unit area. The water hyacinth composts and the farmer-produced cattle compost did not offer immediate effect on yield per unit area. Water hyacinth composts have been found to improve maize production (Osoro et al., 2014).
Options for improving maize production on Striga-infested fields in Western Kenya are still being sought (Woomer et al., 2016). Despite the fact that the four maize varieties have their varying advantages, IR maize seeds remain ideal for suppressing S. hermonthica in Western Kenya (Kanampiu et al., 2003;De Groote et al., 2007). This is because the herbicide kills S. hermonthica hence reducing the seed bank of this parasitic weed in the soil (Kanampiu & Friesen, 2004). However, the success of IR maize will greatly depend on the mitigation of the soil fertility problem (Jamil et al., 2012;Larsson, 2012). Alternative soil fertility interventions based on these observations are therefore proposed.

Conclusion
IR maize seeds remain ideal for suppressing S. hermonthica in Western Kenya. Combined use organic and inorganic fertilizers such as DAP+CAN with water hyacinth compost enhanced with effective microbes (HEM) or cattle manure culture (HCM) could be a viable option for farmers. While DAP+CAN fertilizers offer instant nutrients for maize plant growth, water hyacinth composts on the other hand will improve soil carbon, microbial colonization and diversity, soil fertility and reduce soil acidity in the long run.