Zai Technology and Integrated Nutrient Management for Improved Soil Fertility and Increased Sorghum Yields in Kitui County, Kenya

Deteriorating soil fertility, low unreliable rainfall and soil moisture stress has resulted to low crop yields among farmers of sub-Saharan Africa (SSA), necessitating a search for more sustainable production practices. Zai technology has the ability to promote soil moisture retention and enhances soil fertility. A four-seasons field experiment was conducted to assess the impact of Zai technology combined with cattle manure and inorganic fertilizer on selected soil properties and sorghum yields in Kabati, Kitui County. The experiment was set up in a Randomized Complete Block Design (RCBD) with eight treatments replicated thrice with sorghum Gadam as the test crop. Soil sampling was done at the beginning of the first season and at the end of the fourth season at a dept of 0–15 cm across each plot for laboratory analyses. From the results, the increase in electrical conductivity was significant at p < 0.05 in all the treatments after four cropping seasons. Total organic carbon significantly increased in Zai with cattle manure (p = 0.045), conventional with no input (p = 0.038) and conventional with cattle manure (p = 0.045). Available phosphorous significantly (p < 0.05) increased in treatments under Zai technology while total nitrogen significantly (p < 0.05) reduced after the four cropping seasons. There was a significant (p < 0.05) interactive effect of the tested factors on soil pH, electrical conductivity, total nitrogen, and available phosphorous at the end of the experiment. Moreover, there was significant (p < 0.05) interactive effects on grain yields (SR18 and SR19 seasons) and stover yields (SR18, LR19, and SR19 seasons), with higher yields being recorded in treatments under Zai technology. This study demonstrates the importance of Zai technology in increasing crop yield by trapping water and enhancing its retention and infiltration into the soil for uptake by plants. This study concluded that positive impacts on important soil properties and crop yield could be realized when Zai technology is utilized alongside either sole inorganics or a combination of organic and inorganic amendments and this could be used as a strategy to improve crop production in eastern Kenya and other similar areas.


INTRODUCTION
According to FAO (2010), rain-fed agriculture contributes to about 60% of the world crop production. Despite this, low nutrient, inadequate water supply in quality and quantity (Hengsdijk and Langeveld, 2009;Nelly, 2010) and soil moisture stress (Rockstrom and Karlberg, 2010) have been found to be the major drawbacks that limit crop productivity and are major restraints to agricultural development in many areas of the world. Water scarcity and the resultant food insecurity are of particular concern in most tropical regions of the world, where many countries are less developed (Nelly, 2010).
For most countries in Sub-Saharan Africa (SSA), rainfed agriculture is a significant sector of their economy with close to 70% of the total population living in rural areas and largely depending on small-scale subsistence agriculture for their livelihood security (Rockstrom, 2000). In this region, rain-fed agriculture, which is characterized by droughts in most parts (Rockstrom and Karlberg, 2010), has threatened the food demand projected by 2050 (Grafton et al., 2015). With such projections, rain-fed agriculture is one of the most vulnerable sectors to climate variability and change as reported by (Intergovernmental Panel on Climate Change, 2007).
The uncertainty and risk for current farm-level production associated with variability of rainfall amount and distribution within and between the seasons have downplayed the likely impact of innovations aimed at improving farm productivity (Biazin et al., 2012). This is because the uncertainty discourages the decision to invest in improved farming practices by the farming communities and the wide range of stakeholders in agriculture (Cooper and Coe, 2011). The uncertainty of rainfall amounts and distribution and the high evapotranspiration makes crop production a risky enterprise in semi-arid areas, as a result of agricultural drought and intra-seasonal dry spells (ISDS) leading to reduced yield in rain-fed agriculture (Kahinda et al., 2007).
Despite this, it is imperative to recognize that the situations in many African countries are manageable. The global crop yield growth average of the major cereals in the world varies between 0.9 and 1.6% per year and the increase rates have fallen in the past two decades (Grafton et al., 2015). Rosegrant et al. (2002) reported that grain (maize, millet, sorghum, and teff) yields from rain-fed agriculture were ∼1.5 t ha −1 in developing countries. There is need to produce more to sustainably meet the food needs of the ever-growing human population. Thus, this necessitates the need to develop new strategies for agricultural production in sub-Saharan Africa. These situations suggest that the challenges of limited yields in rain-fed farming systems in SSA might be overcome with soil water conservation techniques combined with fertility management technologies (Rockstrom and Karlberg, 2010). These combinations generate systems that further increase water efficiency and yields in smallholder farms (Winterbottom et al., 2013).
Studies have shown that Kenya is one of the most vulnerable countries to climate variability and changes due to its low adaptive capacity and over-dependence on climate sensitive sectors such as rain-fed agriculture (Kurukulasuriya et al., 2006;Herrero et al., 2010;FAO, 2011). Like many other countries in SSA, Kenya has had significant drought events with the 2004-2006 one affecting food availability for 2.5 million people and 2008-2010 affecting 10 million people (Rarieya and Fortun, 2009). This can be attributed to low soil fertility and moisture deficits inhibiting crop production in the arid and semi-arid areas (Gichangi et al., 2006). Use of suitable water and soil management technologies such as Zai pit (Evett and Tolk, 2009) increases rainfall use efficiency and bridge intra-seasonal dry spells (Dile et al., 2013).
Zai pit system is one of the successful interventions that ensure reduction in runoff and evaporation, improved precipitation capture, and improved agricultural productivity (Evett and Tolk, 2009). Since most lands under cultivation are characterized by compaction, reduced permeability, limited plant root development, and inadequate aeration (Zougmoré et al., 2014), pitting facilitates more water infiltration as runoff water is harvested due to the earthen bund formed downslope of the pits (Kabore and Reij, 2004). Despite this, Zai pits have been found to be efficient when combined with other soil moisture conservation methods and organic and inorganic soil inputs (Burpee et al., 2015).
The improvement of water status in the soil and the increased decomposition and nutrient release results in a beneficial impact of the Zai systems on crop performance under semi-arid conditions (Zougmoré et al., 2014). Zai pit is a form of ancient dryland farming technique that was first initiated in Burkina Faso although some literature points it to Dogon in Northern Mali (Danjuma and Mohammed, 2015). It involves the utilization of basins or holes of 20-120 cm and about 10-60 cm in depth for agricultural activities (Sawadogo, 2011). Their utilization has been found to minimize the effects of droughts since they ensure soil maintenance, soil erosion control and water preservation (Figure 2). Zai have been found to be capable of collecting up to 25% or more of a run-off coming from 5 times its area (Malesu et al., 2006). Zai pits are known to allow crops do well in areas with high risk of crop failure as a result of harsh climatic conditions (Critchley and Gowing, 2012). Water stored in the Zai delay the onset and occurrence of severe water stress thereby buffering the crop against damage caused by water deficits during dry periods (Nyamadzawo et al., 2013). Zai pits increase the amount of water stored in the soil profile by trapping or holding rainwater where it falls (Mutunga, 2001). Besides enhancing water storage, Zai pits increases water infiltration and reduces run-off for plant uptake during the dry periods (Drechsel et al., 2005). The pits play a key water harvesting role.
The application of soil fertility management inputs enhances soil nutrient availability and improves nutrient uptake by crops from soil reserves (Kar et al., 2013). Positive impacts of Zai pit combined with integrated soil fertility management (ISFM) inputs on agricultural productions of different crops have been recorded in other studies (Kabore and Reij, 2004;Sawadogo, 2011). Zai pits combined with cattle manure has been known to promote soil fertility (Fatondji et al., 2006;Sawadogo, 2011) and increase organic matter content (Barry et al., 2009) in soil as compared to use of mineral fertilizer. Other studies have recorded increased yields in maize (Recha et al., 2014), pearl millet grain (Fatondji, 2002), and sorghum (Sawadogo, 2011).
Most farmers in eastern Kenya are entirely depended on rain-fed agriculture as a source of livelihood sustenance and to meet their daily food demands. The study area, which is located in eastern Kenya, is characterized by low, erratic and unreliable rainfall, high temperatures and thus high evaporation rates. These, coupled with low soil fertility has resulted to low, insufficient crop yields among farmers thus resulting to food insecurity. This necessitates the need of research aimed at improving crop production for food security in the area. The present study therefore sought to investigate how selected soil parameters (Soil pH, Electrical conductivity, Total nitrogen, Total organic carbon and Available phosphorous) and sorghum yields are affected by the use of Zai pit technology combined with integrated nutrient management amendments (cattle manure and mineral fertilizer) in Kabati, Kitui County, Kenya.

Profile of the Study Area
The study was carried out in Kabati area, located in Kitui County, Kenya (Figure 1). Based on the stratification of the agroecological zones, Kabati represents the semi-arid areas of Kitui County. It lies between 400 m and 1,830 m above sea level and generally slopes eastwards with the highest regions being Mutitu Hills and Kitui Central (KCIDP, 2018). The County lies between Frontiers in Sustainable Food Systems | www.frontiersin.org latitudes 0010 ′ and 300 ′ South and longitudes 37,050 ′ and 3,900 ′ East (Figure 1).
The region has a bimodal rainfall distribution in a year, the long rains (LR) which are experienced between March and May and the short rains (SR) between October and December. The total annual rainfall average is between 500 and 1,150 mm with a mean annual temperature of 24 • C (Jaetzold et al., 2006). February and September are the hottest months of the year with temperatures ranging to a maximum of 26-34 • C as compared to temperature ranges of 14-22 • C experienced in the other months. Agriculture, both livestock keeping and crop farming, is the backbone of people's economy in the area and a major source of most households' income (Republic of Kenya, 2005;KCIDP, 2018).

Experimental Design
The field experiments were laid out in a randomized Compete block design (RCBD). Each experimental plot measured 6 m by 4.5 m with a 1 m wide alley separating plots within a block and 2 m wide alley left between blocks. The field experiment ran for four consecutive cropping seasons; short rains 2018 (SR18) and 2019 (SR19) and long rains 2019 (LR19) and 2020 (LR20). There was a total of eight treatments ( Table 1). The treatments comprised of two systems, either Zai pit or convention system (no Zai pits), both with four levels of fertilization (no inputcontrol), sole cattle manure, sole mineral fertilizer and both cattle manure and mineral fertilizer). The Zai pits measured 60 cm by 60 cm with a 30 cm depth and was spaced at 70 and 75 cm inter and intra-row, respectively (Figure 2). Dryland Sorghum Gadam variety was used as the experimental test crop with three seeds planted per hill, at a spacing of 75 and 20 cm inter and intrarow, respectively and later thinned out to two seedlings per hill two weeks after emergence to achieve the recommended plant density per ha −1 . External nutrient amendments were applied, based on the Phosphorous and Nitrogen contents as determined through a lab analysis of soil samples, at the beginning of every season to give an equivalent amount of 60 kg N ha −1 and 60 kg P ha −1 , the recommended Nitrogen and Phosphorous rates in the study area (FURP, 1994;Karanja et al., 2014). Cattle manure was broadcasted before plowing for conventional plots and placed in the prepared pits for Zai plots where it was thoroughly mixed with soil. Mineral fertilizers were pre-weighed for each plot before going to the field and applied using dollop cups to ensure uniform distribution within the plot. Weeding was done twice using hand hoe weeding.

Soil Sampling and Laboratory Analysis
The initial sampling was done beginning of SR18 season while the final sampling was done end of LR20 season. Soil samples were collected using an alderman auger at soil depth of 0-15 cm. Soil auguring was done at five spots following a W pattern across each plot and thoroughly mixed to make a composite sample for each plot. Before sampling at every subsequent plot, the soil auger and the collection buckets were thoroughly cleaned to avoid sample contamination. A representative sample was taken from each plot's composite sample, packed in bags with marked tags indicating the plot number and transported to the laboratory for analyses. The fresh-field moist soils were then air-dried, grounded and sieved before subjecting them chemical analysis in the laboratory. All the laboratory analyses on the chemical properties of the soil were done using the standard methods for analyzing soils by Motsara (2008). Soil pH was determined at a 1:1 soil/water ratio using an electronic pH while electrical conductivity was measured using an electrical conductivity probe. Total nitrogen was determined using Kjeldahl method as recently used by Bolo et al. (2021), Soil organic carbon was determined using the Walkley and Black (1934) wet oxidation method and available phosphorus was determined using the Bray and Kurtz (1945) P-1 soil test phosphorus (P) method.

Harvesting and Yield Determination
Sorghum grain yield and stover yield were harvested at physiological maturity from the net plot. The heads were manually separated from the stover. All heads from each plot were collected and separately placed in different sacks which were well-labeled. Heads were then sun-dried, hand-threshed and weighed at 12.5 % moisture content. Maize stover was cut at ground level and total above-ground fresh weight determined. Both grain and Stover yields were then converted from kilograms per net area to tons per hectare (t/ha).

Statistical Analysis
Both sorghum yield and soil data were subjected to a two-way analysis of variance (ANOVA) in R-studio program to obtain an F-value of the effect of the model. Pair-wise comparison of selected soil parameters differences between the start and the end of the experiment were analyzed using t-test. Differences between treatment means were examined using least significance difference (LSD) at p = 0.05.

Rainfall Distribution During the Four Experimental Seasons
There was seasonal rainfall variability over the four seasons of the study (Figure 3). The four consecutive cropping seasons SR18, LR19, SR19, and LR20 were characterized by different rainfall patterns, with higher rainfall occurring in the short rain seasons (SR18-938.1 mm and SR19-442.71 mm) compared to the long rainy seasons (LR19-116.3 mm and LR20-228 mm). The total number of rainy days varied in the four growing seasons with SR19 having the highest number of rainy days (47 days) followed by SR18 (31 days) ( Table 2). LR19 and LR20 had the lowest number of rainy days i.e., 10 and 11 days, respectively. Much of the rainfall during the SR18 and SR19 seasons was received during the first 2 months with a prolonged dry spell period occurring from day 64 after planting and prolonging toward the end of the planting seasons.
In Contrast, the long rains seasons (LR19 and LR20) experienced interchanging periods of rainy days and consecutive dry days during the 1st month. An intra-seasonal dry spell of 73 and 69 days was experienced during the LR19 and LR20 respectively, whereas during the SR18 and SR19 drought periods of 43 and 14 days respectively, were experienced ( Table 2). The rainfall in Eastern Kenya is distributed within two seasons annually, the long rains which are experienced between March and May and the short rains between October and December. Much of the total annual rainfall is received during the short rains season and thus the short rains are considered more reliable. It is during this season that higher crop yields are realized as compared to the long rains season. In this study, highest total season rainfall amounts were experienced during the short rains and this had a positive impact on sorghum yield as compared to the long rains' seasons.

Effects of Zai Pits in Combination With Selected ISFM Options on Selected Soil Chemical Properties
Generally, there was observed variation in pH levels across the treatments with all treatments that had manure application whether solely or in combination with mineral fertilizer increasing while those with sole application of mineral fertilizer decreasing at the end of the experiment. There was a significant (p < 0.05)interactive effect on soil pH at the end of the experiment. The increase in soil pH was significant at p < 0.05 in Zai with manure (p = 0.02234), conventional with manure (p = 0.04697), Zai pit with manure and half rate mineral fertilizer (p = 0.02597), and conventional with manure and half rate mineral fertilizer (p = 0.04697).
Electrical conductivity significantly increased at p < 0.05 in all the treatments at the end of the experiment ( Table 3); Zai pit with no input (p = 0.015), Zai pit with manure (p = 0.022), Zai pit with full rate fertilizer (p = 0.0058), Zai pit with both manure and half rate mineral fertilizer (p = 0.013), conventional with no input (p = 0.001), conventional with manure (p = 0.006), conventional with full rate mineral fertilizer (p = 0.019), and conventional with both manure and half rate mineral fertilizer (p = 0.001).
All the treatment under Zai pit system had a significant (p < 0.05) positive effect on electrical conductivity of soil as compared to their counterparts under conventional system. Zai with full rate mineral fertilizer and Zai with cattle manure and half rate mineral fertilizer treatments significantly (p = 0.021) increased electrical conductivity as compared to conventional with full rate mineral fertilizer and conventional with cattle manure and half rate mineral fertilizer. Zai with no input and Zai with sole cattle manure recorded 27 and 6% higher electrical conductivity values compared to Conventional with no input and Conventional with sole cattle manure, respectivelyAt the end of the experiment, total nitrogen significantly decreased (Table 3) at p < 0.05 in Zai pit with no input (p = 0.0003), Zai pit with sole cattle manure (p = 0.012), Zai pit with manure and half rate mineral fertilizer (p = 0.047), Conventional with sole cattle manure (p = 0.027), conventional with full rate mineral fertilizer (p = 0.0005) and conventional with cattle manure and half rate mineral fertilizer (p = 0.018). Zai with no input significantly (p = 0.014) increased total nitrogen as compared to conventional with no input treatment. Conventional with cattle manure and half rate fertilizer had a significant positive effect on total nitrogen than Zai with cattle manure and half rate mineral fertilizer.
Significant (p < 0.05) increases in Total organic carbon between the beginning and end of the experiment were recorded under Zai pit with sole cattle manure treatment (p = 0.045), conventional with sole cattle manure treatment (p = 0.045) and conventional with no input treatment (p = 0.038). Under the Zai pit system, Zai with sole cattle manure increased Total organic carbon by 36, 22, and 19% higher as compared to Zai with full rate mineral fertilizer, Zai with no input and Zai with cattle manure and half rate mineral fertilizer respectively ( Table 3).
The amount of available phosphorous significantly (p < 0.05) increased at the end of the long rains 2020 season in Zai with sole cattle manure (p = 0.022), Zai with full rate mineral fertilizer (p = 0.013), Zai with cattle manure and half rate mineral fertilizer (0.021), and conventional treatment with cattle manure and half rate mineral fertilizer (p = 0.039). At the end of the experiment, Zai with full rate mineral fertilizer and Zai with cattle manure and half rate mineral fertilizer treatments significantly (p = 0.018) increased phosphorous as compared to Conventional with full rate mineral fertilizer and Conventional with cattle manure and half rate mineral fertilizer. Zai with no input and Zai with sole cattle manure recorded 3 and 8% higher available phosphorous amounts compared to Conventional with no input and Conventional with sole cattle manure, respectively ( Table 3).

Effects of Zai Pits in Combination With Selected ISFM Options on Sorghum Yields
Generally, higher grain and stover yields were recorded during the short rains' seasons (SR18 and SR19) as compared to the long rains' seasons (LR19 and LR20) ( Table 4). Even though treatments under Zai system recorded higher grain yields in LR19 and LR20, the difference was not significant at P = 0.05 (Table 4). Zai treatments recorded significantly (p < 0.05) higher grain yields during the SR18, SR19 and LR19 seasons as compared to the treatments under the conventional system. In SR18, Zai with no input, Zai with sole cattle manure, Zai with full rate mineral fertilizer and Zai with cattle manure and half rate mineral fertilizer recorded 33, 25, 42, and 28% significantly higher grain yield than conventional with no input, conventional with sole cattle manure, conventional with full rate mineral fertilizer and conventional with cattle manure and half rate mineral fertilizer, respectively (Figure 5).
The highest grain yields (4.85 t ha −1 in SR19 and was recorded under the Zai pit system with cattle manure and half rate mineral fertilizer and this was significantly (p < 0.05) higher than the similar treatment under conventional system ( Table 4). During the SR19 season, Zai pit with cattle manure and half rate mineral fertilizer, Zai with sole cattle manure, Zai with full rate mineral fertilizer and Zai with no input recorded 48, 58, 39, and 40% significantly (p < 0.05) higher yields than their counterpart treatments under conventional system (Figure 4). The interaction effect on Sorghum grain yield between water management and nutrient amendment options was significant at p = 0.05 in SR18 and SR19 (Figure 5). In the LR19 and LR20 seasons, all the treatments under Zai system recorded higher  grain yield than their conventional counterpart treatments, with the highest (1.39 t ha −1 ) grain yield recorded under Zai with sole cattle manure in LR20 being 21% higher than that of conventional with sole cattle manure ( Table 4).
Highest stover yields were recorded during the SR19 season while the lowest was recorded in the LR19 season. Significant (p < 0.05) positive interaction effects on stover yields were recorded in treatments under Zai system in the SR18, LR19, and SR19 seasons as compared to the treatments under conventional system (Figure 6). In SR18, Zai with cattle manure and half rate mineral fertilizer recorded 53% significantly (p < 0.05) higher stover yields than conventional with cattle manure and half rate mineral fertilizer. During the LR19 season, Zai with no input, Zai with sole cattle manure, Zai with full rate mineral fertilizer and Zai with cattle manure and half rate mineral fertilizer recorded 35, 50, 40, and 21% significantly (p < 0.05) higher stover yields than their conventional counterparts. In SR19 season, significantly (p < 0.05) higher stover yields were recorded under Zai treatments as compared to similar treatments under conventional system, with the highest (13.14 t ha −1 ) under Zai with cattle manure and half rate mineral fertilizer being 35% higher than the similar treatment under conventional system (Figure 6).
Generally, across all the treatments in both systems in the four study seasons, treatments with fertility inputs recorded higher sorghum grain and stover yields as compared to both Zai and conventional with no input. Furthermore, treatments with cattle manure application either solely or in combination with mineral fertilizer recorded significantly (p < 0.05) higher yields as compared to the treatments with sole mineral fertilizer FIGURE 4 | Interaction effect of Zai technology and selected soil nutrient amendment options on Soil pH at the end of the LR20 season. Beg, Beginning of experiment SR18; End, End of experiment LR20; ZNO, Zai with no inputs; ZM60, Zai + Manure; CM60, Conventional + Manure; CNO, Conventional with no inputs; ZF60, Zai + 60 kg N ha −1 ; CF60, Conventional + 60 kg N ha −1 ; ZMF30, Zai + Cattle manure + 30 kg N ha −1 ; CMF30, Conventional + Cattle manure + 30 kg N ha −1 . The error bar denote the least significant difference (lsd) per season at p = 0.05. Means followed by same letters in the same column denote no significant difference between treatments at p = 0.05. application in both systems. All the treatments with sole cattle manure application or with combined manure and mineral fertilizer application recorded significantly higher yields as compared to Zai pit treatments without cattle manure application. The effect of interaction on Sorghum grain yield between the selected water management strategy and the nutrient amendment options was significant in the SR18 and the SR19. The two seasons recorded highest grain yield with Zai treatments recording higher yields than the treatments under conventional system (Figure 5).

DISCUSSION
Generally, there was observed variation in pH levels across the treatments with all treatments that had manure application whether solely or in combination with mineral fertilizer increasing while those with sole application of mineral fertilizer decreasing at the end of the experiment (Figure 4).
The decrease in soil pH in mineral fertilizer-amended treatments could be as a result of the H + ions, which are added on the cation exchange complex of soils from the mineral fertilizer. Nitrogen fertilization acidifies the soil by the oxidation of dry-deposited compounds, loss of basic cations through ion exchange, and plant uptake and nitrification of ammonium (Cai et al., 2019). Synthetic nitrogen application significantly reduces exchangeable base cations in soils, which lead to declined soil pH. A decline in base saturation is symptomatic of soil acidification (Stevens et al., 2009). Additionally, synthetic nitrogen application shifts soils into the Al 3+ buffering stage. Aluminum is released into solution at a pH below 5 by the hydrolysis of both Alhydroxides and silicates on clay mineral surfaces (Cai et al., 2019). Decrease in soil pH increases the availability of potentially toxic heavy metals and also contribute to the reduction of the microbial organisms beneficial in root functions (Stevens et al., 2009). The increase in soil pH in manure-amended soils could be attributed to the reduction of exchangeable aluminum in acidic soils (pH 4.5-5.5) which is considered to occur through aluminum precipitation or chelation on organic colloids (Hue and Amien, 1989;Mucheru-Muna et al., 2007) or by complexation of soluble aluminum by organic molecules, especially organic acids (Hue and Amien, 1989). Generally, the ash-alkalinity nature of manure is associated with protons to neutralize soil acidity (Rukshana et al., 2013). The alkalinity of organic materials following the decarboxylation of organic anions and the ammonification of organic N (Cai et al., 2019) are the major causes of the recorded increase in soil pH in manure-amended soils.
The observed pH decreases in fertilizer amended soils corroborate with the findings by Cai et al. (2019) where lowered soil pH was recorded in treatments with sole application of N fertilizers which resulted in no yield after 12 years of study. Accordingly, other studies reported that synthetic fertilizer application could significantly decrease soil pH (Mucheru-Muna et al., 2007;Cai et al., 2014;Zhu et al., 2018). The increase in soil pH recorded in manure-amended soils correspond with the findings by Wildemeersch et al. (2015) where increment in soil pH to 5.0 from the initial 4.2 was recorded following manure application. A number of other studies have also reported significant increases in pH with manure treatment (Eghball, 2002;Bayu et al., 2005;Mucheru-Muna et al., 2007Mugwe et al., 2009;Mutegi et al., 2012).
Higher increases in electrical conductivity recorded in manure amended soils (Table 3) could be as a result of presence of dissolved salts in the cattle manure used as a result of feed additives which when mixed with water, they dissolve and break into tiny electrically charged ions, hence increasing the ability of the soil solution to conduct electricity. The increase could also be attributed to the resultant increases in exchangeable potassium ions and organic matter, which supplies a pool of nutrients and ions that can be released in the soil solution. This observation corresponds with the findings by Ozlu and Kumar (2018) who reported increased Electrical conductivity by 2.2 times in manure-amended soils as compared to that of fertilizer for 0-10 cm depth. The higher increase in electrical conductivity in the treatments with manure application as compared to nonmanure amendments is consistent with the findings of several other studies (Dong et al., 2001;Eghball, 2002;Eigenberg et al., 2002;Cai et al., 2014;Zhao et al., 2014;Carmo et al., 2016;Miller et al., 2017;Zhu et al., 2018).
Generally, there was reduction in Total Nitrogen in all the treatments at the end of the experiment ( Table 3). The observed decrease in total nitrogen content could be attributed to uptake for utilization by sorghum plants. The recorded nitrogen losses could also as a result of erosion by run off since the experiments were carried out during rainy seasons and the soils in the study area, being predominantly sandy to loamy sand in texture, are poorly drained, susceptible to erosion and are limited in their capacity to retain water and nutrients. Furthermore, the reduction recorded in manure-amended treatments could be as a result of increased microbial biomass which could utilize the nitrogen and losses through leaching, denitrification and ammonia volatilization. These results correspond with the findings of other studies where reduction in total nitrogen were recorded and the losses attributed to erosion (Jaetzold et al., 2006), utilization through microbial activity (Van Diepeningen et al., 2006) and losses through denitrification and ammonia volatilization (Ju et al., 2009).
The recorded significant increases in total organic carbon in treatments with sole application of manure (Table 3) could be as a result of addition of more carbon source through manure application. Manure directly increases carbon inputs into the soil and also influences crop residues, which potentially sequestrates agricultural soil organic carbon and nutrient release. These finding is corroborated by the finding by Hati et al. (2007) who attributed the observed increase in soil organic carbon among different fertility treatments after 28 years of study to the addition of carbon source through crop residues, root biomass, and farmyard manure. Similarly, other studies have reported a significant increase in soil organic carbon in treatments that had sole cattle manure application (Dunjana et al., 2012;Bedada et al., 2014) and in those with manure and half rate fertilizer (Kaur et al., 2007;Bedada et al., 2014). Several other studies have reported increases in organic carbon content in soil as a result of manure application as compared to use of mineral fertilizer (Mucheru-Muna et al., 2007;Lal, 2008;Bandyopadhyay et al., 2010;Dunjana et al., 2012;Liang et al., 2012;Kuzyakov and Blagodatskaya, 2015;Xin et al., 2016;Zhou et al., 2016).
There was increase in available phosphorous in all the treatment at the end of the experiment ( Table 3). The higher levels of phosphorous recorded in treatments with sole fertilizer or fertilizer combined with manure application was as a result of phosphorous available in the NPK fertilizer used. Zai treatments with either sole cattle manure application or manure combined with mineral fertilizer had higher available phosphorous levels at the end of the experiment as compared to similar treatments under conventional system. This could be as a result of the composition of cattle manure, which contains both organic and inorganic phosphorous. Plants take up phosphorous in form of inorganic orthophosphates and it generally constitutes 45-90% of the phosphorous in manure (Buckley and Makorto, 2004), making manure an important source of phosphorous. Despite the effect of nutrient amendment inputs, all the treatments under Zai system recorded higher phosphorous levels as compared to conventional treatments. This could be due to the ability of the Zai pit system to hold nutrients in place by preventing nutrient loss through run-off by erosion. The current trend of results is in consonance with the findings of studies by Mucheru-Muna et al. (2007), Tirol-Padre et al. (2007, and Muui et al. (2013). Cattle manure supplies macronutrients and micronutrients to the soil (Gemechu, 2020) and has been known to provide higher phosphorous levels to soil as compared to other organic inputs (Opala et al., 2012).
Generally, the two short rains seasons (SR18 and SR19) recorded higher grain and stover yields as compared to the two long rains seasons (LR19 and LR20). This could be as a result of the high rainfall amounts (442.71 and 938.1 mm for SR18 and SR19, respectively) recorded in the short rains seasons (Figure 1) which were higher than those recorded in the long rain seasons (166.3 and 228 mm for LR19 and LR20, respectively). The amount of rainfall received during the planting season has an effect on yield. The number of rainy days were also more during these two short seasons ( Table 2) with more rainy days occurring during the first 62 days after planting (critical growth and developmental stages). According to Hansen (2004), agricultural activities in eastern Kenya depends on the OND (October, November, and December) short rainfall seasons which have been found to be more predictable and reliable. Thus, better annual yields are received during the OND seasons in most agriculture-depended households (Amissah-Arthur et al., 2002;Barron et al., 2003). The lower yields recorded during the LR19 and the LR20 seasons was as a result of the lower unreliable rainfall amounts, fewer number of rainy days and poor distribution during the season (McCarl et al., 2008). These two seasons experienced long intra-seasonal dry spells (73 and 69 days for LR19 and LR20, respectively) which resulted to soil moisture stress, hence the low recorded yields. Intra-seasonal dry spells are a characteristic feature of semi-arid areas in East Africa (Barron et al., 2003) and South Africa (Usman and Reason, 2004). The occurrence of these dry spells has had major detrimental effects on crop yields. In their study, Craufurd and Peacock (1993) found out that sorghum grain yields were reduced by 87% as a result of water stress which occurred during critical crop developmental stages such as flowering and stem swelling (McCarl et al., 2008). According to Bewket (2009) and Mulat et al. (2004), the amount and temporal distribution of rainfall is generally the major determinant of inter seasonal fluctuations observed in crop production levels.
Zai pit is a form of ancient dryland farming technique which involves the utilization of basins or holes for agricultural activities so as to minimize the effects of droughts since they ensure soil maintenance, soil erosion control and water preservation (Sawadogo, 2011;Danjuma and Mohammed, 2015). In Kenya, Zai pits technology has been promoted as a water harvesting technique for maize production in eastern region (Recha et al., 2014). The results of this study indicated that treatments under Zai pit technology with the selected soil nutrient amendment options had a significant (p < 0.05) interaction effect on sorghum grain and stover yields as compared to similar treatments under conventional system. Furthermore, higher yields were realized in Zai treatments containing either sole cattle manure or manure combined with mineral fertilizer.
Despite the fertility inputs, the higher yields recorded under Zai treatments could be attributed to the potential of Zai pits in increasing the amount of water stored in the soil profile by trapping or holding rainwater where it falls (Mutunga, 2001), therefore delaying the onset and occurrence of severe water stress thus buffering the crop against damage caused by water deficits during dry periods (Nyamadzawo et al., 2013). Besides enhancing water storage, Zai pits increases water infiltration and reduces run-off for plant uptake (Drechsel et al., 2005) thereby, boosting crop yield. The results of this study correspond with other studies whose results indicated a positive effect of Zai pit utilization on crop yield. Fatondji et al. (2006) found out that the use of Zai pits in Burkina Faso led to increased harvests of pearl grain Millet.
They attributed this increase to the use of Zai pits as well as the use of organic fertility inputs. A study by Sawadogo (2011) reported increased yields variations from 300 to 400 kg ha −1 by the Zai system in degraded land. Sawadogo (2011) reported substantial sorghum grain yield increases in farmer's fields from 319 to 642 kg ha −1 without Zai pit system to 975 to 1,600 kg ha −1 with Zai pit system. The use of potential of Zai pit system in increasing crop yield has also been reported in Zambia (Haggblade and Tembo, 2003;Thierfelder and Wall, 2009), South Africa (Magombeyi and Taigbenu, 2008), Niger , Ethiopia (Amede et al., 2011), and in Zimbabwe (Thierfelder and Wall, 2009;Gumbo et al., 2012). Besides its positive effect on soil moisture, Zai pits have a good potential to increase soil biological activities and to promote nutrient use and agronomic efficiency (Fatondji, 2002).

CONCLUSION
Significant differences in soil properties and sorghum grain and stover yields among different fertilization treatments under Zai and conventional systems were recorded in this study. Application of sole manure or combined with mineral fertilizer significantly increased soil pH, electrical conductivity, soil organic carbon and available phosphorous as compared to treatments with sole mineral fertilizer at the end of the experiment.
There was significant (p < 0.05) interaction effect of Zai technology and the selected soil nutrient amendment options on the selected soil properties and sorghum grain and stover yields. Generally, treatments under Zai system recorded higher pH, electrical conductivity, organic carbon, phosphorous, and nitrogen levels as compared to conventional treatments and this could be due to the ability of the Zai pit system to hold nutrients in place by preventing nutrient loss through run-off by erosion. The significant effects resulting from the interaction of Zai system and the selected soil nutrient amendment options on sorghum grain and stover yields as compared to the conventional system could be attributed to moisture retention and increased nutrient use efficiency enhanced by the Zai pits. The results of this study demonstrates that significant positive interaction effects on crop yields could be achieved when Zai technology is combined with inorganic nutrient amendments either solely or in combination with organic inputs. The ability of Zai technology to improve crop yield is largely attributed to its pit design. Apart from capturing water, Zai pits enhances soil moisture retention and infiltration thereby increasing the amount of water in the soil that is available for uptake by plants.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.