Enhancing water use efficiency and nutritional quality of maize fodder under deficit irrigation with microbial and potassium amendments

Abstract

Optimizing maize fodder production under water limited conditions is essential for sustainable livestock systems facing climate induced water stress and declining soil fertility. In this two-year field study, we investigated how deficit irrigation at different maize growth stages (VT-R3, R3-R6, VE-VT) in combination with potassium fertilization (0, 50 kg ha⁻¹) and bacteria inoculation (B-, B+) influences water use efficiency and the nutritional quality of maize fodder. Data was recorded on acid detergent fiber (ADF), neutral detergent fiber (NDF), crude protein (CP), ash, hemicellulose, biological yield, water use efficiency and irrigation water use efficiency. Irrigation timing exerted the strongest influence on most traits. Late season deficit irrigation (R3-R6) increased fiber accumulation (ADF and NDF), whereas early season deficit irrigation (VE-VT) consistently reduced fiber accumulation while improving crude protein content and irrigation water use efficiency. Full irrigation produced the highest biological yield, followed by early season deficit irrigation. Potassium fertilization enhanced crude protein content and water use efficiency, particularly under water limited conditions. Bacterial inoculation significantly improved crude protein concentration, biological yield, and both water use efficiency and irrigation water use efficiency, with more pronounced effects under deficit irrigation. Overall, the results indicate that early season deficit irrigation (VE-VT) combined with potassium supplementation and bacterial inoculation represents an effective integrated strategy to improve forage biomass production and water productivity in maize; however, these practices may also increase fiber fractions, potentially reducing digestibility and voluntary intake, and therefore require careful balancing of yield gains with forage quality.

Introduction

Cereal crops, such as maize (Zea mays L.), wheat, and rice are used worldwide for grain and livestock feed. Therefore, the dietary content of maize stover and wheat straw must be accurately and promptly analyzed [1, 2]. The time of harvest, storage conditions, and processing procedure usually affect roughage composition [3]. Beneficial phytochemicals found in straw, such as NDF, are crucial for promoting rumen fermentation in ruminants. Thus, an assessment is necessary to understand the sources and objectives of straw [4].

Fertilizers and irrigation significantly impact the productivity and nutritional value of fodder [5]. However, excessive irrigation and fertilizer use increases production costs and can harm the ecosystem and soil, ultimately decreasing yield. Excessive and improper use of chemical fertilizers can result in significant yield losses [6, 7]. Limited irrigation can influence the leaf to stem ratio, affecting the quality of the feed [8]. Several studies have proven the detrimental effects of limited irrigation on various stages of crop growth [9–11].

Water scarcity is the most prominent abiotic threat to crop productivity in dry and semiarid regions and a key issue in addressing climate change [12]. Climate change is expected to negatively impact agricultural production in the coming years by causing more frequent and severe occurrences of low precipitation and insufficient moisture, as well as an increase in mean and maximum temperatures [13]. Water scarcity reduces photosynthesis by lowering chlorophyll content and leaf area. It also impairs metabolic activity and negatively impacts final yield [14]. Water scarcity decreases the relative water content of leaves, transpiration rate, stomatal conductance, and cell enlargement rate. Ultimately, these changes hinders plant growth, particularly during critical growth stages [15]. Therefore, effective crop management requires an understanding of how deficit irrigation affects maize growth and development. Beneficial soil microorganisms, such as plant growth-promoting rhizobacteria (PGPR), can enhance plant growth and productivity [16]. When introduced to seeds or roots, PGPR colonizes the root system and uses root exudates for energy, thereby promoting growth and increasing yield. PGPR also helps plants cope with abiotic stress, improves nutrient uptake, and regulates plant hormones, such as auxin, abscisic acid, cytokinin, ethylene, and gibberellins [17]. Numerous investigations have focused on several PGPR taxonomic groups, particularly native Bacillus species [18, 19].

Adequate irrigation significantly improves the nutritional value of maize straw. It increases crude protein (CP) content and reduces structural fiber components, such as acid detergent fiber (ADF) and neutral detergent fiber (NDF). Increasing irrigation levels in silage maize resulted in a 13% increase in CP content and a decrease in fiber fractions, which aligns with reported trends [20]. Similarly, study [21] examined the effects of irrigation technology and rate on forage yield and quality. They found that appropriate irrigation practices can enhance forage quality by increasing CP content and reducing fiber fractions. Another study [22] showed that irrigation and nitrogen management significantly affect these quality indices, underscoring the importance of water and nutrient management in forage production. Partial root zone drying and drip irrigation methods can maintain or improve CP content while lowering ADF and NDF values [23]. Collectively, these findings suggest that appropriate irrigation strategies enhance the feeding value of maize straw, whereas water stress deteriorates its nutritional quality by increasing fiber and reducing protein content. Recent studies have demonstrated that effects of deficit irrigation on forage yield in maize and other forage cereals, with crop responses largely dependent on the severity and timing of water stress [24–27]. Furthermore, optimized deficit irrigation strategies can improve water productivity and sustain forage yield by enhancing soil water extraction and root system efficiency [28, 29].

Accordingly [30], recently observed that reducing irrigation water by 26% during specific growth stages (VT, or tasseling stage; R1, or silking stage; and R2, or blister stage), as classified by [31], could be suitable for new commercial maize hybrids, while preserving a similar whole plant yield and forage nutritive values. Conversely [32], studied the effect of four irrigation levels (0, 153, 305, and 480 mm of total water) on maize and reported that decreased irrigation levels caused a linear decrease in yield, crude protein content and water-soluble carbohydrates, as well as an increase in fibrous fractions.

Similarly [33], tested different amounts of irrigation water (an average of 451 to 975 mm) and reported modifications to the chemical composition of maize grown in semi-arid conditions.

Due to the conflicting results regarding the effect of irrigation water on the nutritive value of maize silage, as well as the desire to reduce the amount of water used for irrigation, it is crucial to investigate how maize responds to a severe reduction in irrigation water in terms of its chemical composition, rumen dry matter, and fiber digestibility.

Although the effects of irrigation management and PGPR on maize have been widely studied [30–32], their combined influence with potassium nutrition under growth-stage-specific deficit irrigation remains poorly understood. Most studies have evaluated these factors independently or under uniform seasonal water stress. This creates uncertainty regarding how potassium and PGPR interact with early, mid, and late season water deficits to regulate stress physiology, nutrient assimilation, biomass partitioning, and the trade-offs between forage yield and quality. Growth stage specific deficit irrigation is biologically important because maize responds differently to water stress across developmental stages. Early season water limitation mainly restricts canopy growth, root development, and nitrogen assimilation, whereas stress during the reproductive stage enhances lignification and secondary cell wall formation [29]. Therefore, clarifying how potassium supply and PGPR interact with irrigation timing is essential to explaining variations in forage quality and water use efficiency under water-limited conditions.

This study aimed to determine the effects of potassium and bio-fertilizer application on water use efficiency, biomass yield, and quality characteristics of maize fodder under seasonal deficit irrigation conditions. Additionally, the study aimed to reveal the ability of these factors to tolerate seasonal water deficiency.

Materials and methods

The study was conducted during the years 2023 and 2024 at experimental research farm of the Isparta University of Applied sciences, Türkiye. The experiment was conducted in randomized complete blocks design (RCBD) with split split plot arrangement. The main plots of the experiment were allotted with irrigation as control and seasonal deficit irrigations based on Zadoks [34] growth scales as VT-R3 (deficit irrigation from tasseling to grain filling stage), R3-R6 (deficit irrigation from grain filling stage to physiological maturity), VE-VT (deficit irrigation from to emergence to tasseling stage) and a rainfed treatment. Although maize cultivation under fully rainfed conditions is not agronomically feasible in the study region, the rainfed treatment was intentionally included as a reference to quantify crop response under extreme water limitation and to enable the calculation and comparison of irrigation-related indices, particularly irrigation water use efficiency (IWUE), subplots with potassium doses (0 and 50 kg ha^{− 1}) and subsub plots of the experiment was assigned with Bacillus bacteria (B-/B+). Pre experiment soil analysis showed that soil samples taken from a depth of 0–90 cm in the experimental area had sufficient potassium (K) content: 1.94 cmol kg⁻¹ at 0–30 cm, 0.75 cmol kg⁻¹ at 30–60 cm, and 0.58 cmol kg⁻¹ at 60–90 cm. Therefore, the 50 kg ha⁻¹ rate was selected to evaluate the supplemental effect of potassium, particularly in combination with bacterial inoculation and deficit irrigation treatments. This approach allowed assessment of whether additional potassium could enhance crop performance and water-use efficiency beyond the existing soil K status.

For proper seedbed preparation the soil was cultivated with cultivated twice followed by rotavator to ensure proper sowing and emergence of the seeds. The length of the plot was maintained as 5 m with 70 cm row to row and 30 cm plant to plant distance. All the control plots (without Bacillus) were sowed at first followed by inoculated seeds. The seed were inoculated with recommended dose of Bacteria (36 g 100 kg^− 1 seed) prior to sowing and were properly mixed. The Bacillus inoculant used in this study was i30FP (Bacillus simplex strain − 1 × 10⁸ cfu/g). All experimental plots received equal applications of nitrogen and phosphorus at rates of 200 kg ha⁻¹ N and 80 kg ha⁻¹ P, supplied as urea and monoammonium phosphate (MAP), respectively. Phosphorus was applied entirely at sowing, whereas nitrogen was applied in two equal splits, with 50% at sowing and the remaining 50% at the 45-cm crop growth stage.

Irrigation water and evapotranspiration

Control irrigation was performed using drip irrigation system throughout the growing season of the plots following the moisture loss. During the experiment, soil water content was determined using the gravimetric method to monitor changes in soil moisture in the plant root zone. To verify and cross check the amount of irrigation to be applied the class A evaporation container was installed in the experiment area. The amount of irrigation water was calculated based on the daily evaporation values measured from the Class A evaporation pan.

The amount of the applied irrigation water to the treatments was calculated using Eq. 1. The amount of irrigation water determined as depth terms, was calculated in volume terms by multiplying with the parcel unit area and the percentage of the wetted area during the applications (Eq. 2).

1

2

Where;

I, irrigation water amount (mm); Epan, cumulative evaporation amount from the class A pan during the irrigation interval according to the treatments (mm); Kp, pan coefficient taken as 1 based on the principles given by [35]; V, irrigation water amount as volume (L); A, plot area (m²); WA, percentage of the wetted area (47%).

Plant water consumption (PWC) for experimental treatments was calculated on a water budget basis using Eq. 3 [36].

3

Where;

ET: Evapotranspiration (mm), P: Effective rainfall (mm), Dp: Deep percolation (mm), Cp: Capillary rise (mm), Rf : Surface runoff (mm), ΔS: Water content change in soil profile (mm).

Capillary rise was neglected because of no ground water problem in the experimental area. In addition to this, the infiltration rate of the soil was determined as a result of the infiltration tests and the appropriate dripper flow rate (1.5 L h^− 1) was selected, there were no surface runoff losses. For this reason, the formula was simplified (Eq. 4).

4

Equation (4) was used to calculate the irrigation requirement (I). In cases of excessive rainfall (P), the amount of deep percolation (DP) was deducted from the irrigation calculation. Throughout the crop growing seasons in both years, soil moisture changes (ΔS) were regularly monitored, and irrigation was scheduled when 30% of the available soil moisture was depleted from field capacity (5–7 days interval based on moisture depletion).

Water use efficiency (kg ha^− 1.mm)

The Eqs. 5 and 6 given by [37], was used to determine the water use efficiency (WUE) and irrigation water use efficiency (IWUE) realized in the trial.

5

In the equations; BY = Biological yield (kg ha^− 1); ET = Plant water consumption (mm).

Irrigation water use efficiency (kg ha^− 1.mm)

6

In the equations; BY = Biological yield (kg ha^− 1); IR = Irrigation water applied (mm), BY₀ = Biological yield (kg ha^− 1) taken from rainfed treatment.

The coverage ratio of the applied irrigation water to the evapotranspiration (IRc)

The irrigation coverage ratio (IRc) is used to evaluate the adequacy of irrigation in meeting the crop water requirements. It is calculated using the following formula 7 [38].

7

Where,

IRc= irrigation water to the evapotranspiration, I= irrigation water applied, ET = Plant water consumption.

Biological yield (kg ha^− 1)

For recording biological yield (BY), two central rows in each subplot was harvested along with cobs at maturity and was tied in bundles. The bundles were sun dried for two days and weighed for recording biological yield data then was converted into kg ha^− 1 by the following formula (Eq. 8).

8

Determination of nutritional traits in maize fodder

For determining the nutritional traits maize fodder biomass (only above ground biomass without ears (excluding grains)) samples were dried in an oven at 65 °C until they reached a constant weight. The dried samples were grinded to pass through a 1 mm sieve. Acid detergent fiber (ADF), neutral detergent fiber (NDF), crude protien CP, hemicellulose content (HC), and ash content (AC) were determined using Near Infrared Spectroscopy (NIRS) calibrated against traditional wet laboratory analyses. In these analyses, a FOSS Forage Analyzer 5000 (FOSS XDS RCA, Denmark) NIRS device equipped with the Win ISI II software package (version 1.5, Intra Soft International, LLC) was used. Outliers were detected during the analysis process, the relevant samples were re-scanned to obtain accurate data points. A basic NIRS calibration for maize fodder was developed and validated using calibration and validation sets, with a Global H value of 1, in accordance with conventional laboratory analyses. This basic NIRS equation was updated by analyzing 10% of new fodder samples after it was created using conventional analyses [39, 40].

Climatic characteristics of the experimental site (2023–2024)

Air temperature and precipitation patterns showed notable interannual variation between 2023 and 2024 across the months from May to October (Table 1).

Table 1.

Monthly averages of air temperature, precipitation, relative humidity, and solar radiation during crop growing season (2023–2024)

2023
Months	Air temperature (°C)			Precipitation (mm)	R. Humidity (%)	S. Radiation (W/m2)
	Min	Max	Mean	Total	Mean	Mean
May	8.08	20.73	14.11	52.2	77.69	189.95
June	12.01	25.87	18.85	0.8	67.62	286.5
July	11.81	32.08	23.16	0.4	42.3	346
August	13.73	35.05	25.14	60.2	42.53	306.43
Sept.	8.58	28.75	18.84	2.22	53.23	253.66
October	5.05	23.51	13.83	95	67.11	203.33

2024
Months	Air temperature (°C)			Precipitation (mm)	R. Humidity (%)	S. Radiation (W/m2)
	Min	Max	Mean	Total	Mean	Mean
May	6.42	23.09	14.87	41	60.9	250.25
June	13.12	33.85	24.65	6	38.99	335.93
July	14.69	32.69	24.28	34.2	48.63	312.38
August	13.15	33.06	24.13	11.4	43.68	298.77
Sept.	11.22	28.57	20.07	9.4	59.19	243
October	2.03	23.02	12.79	1.8	52.74	200.96

In 2023, mean air temperatures ranged from 13.83 °C in October to 25.14 °C in August. The highest maximum temperature recorded was 35.05 °C in August, while the lowest minimum was 5.05 °C in October. Precipitation was highest in October (95 mm) and lowest in July (0.4 mm), with particularly dry conditions observed from June through September. In contrast, 2024 displayed the mean values ranging from 12.79 °C in October to 24.65 °C in June. The highest recorded temperature (33.85 °C) occurred in June, and the lowest minimum (2.03 °C) in October. Precipitation levels in 2024 were generally lower, with the exception of May (41 mm) and July (34.2 mm), while October recorded the lowest total (1.8 mm), indicating a drier end to the growing season compared to the previous year.

To eliminate potential residual effects from previous treatments, particularly bacterial inoculants, the experimental area was changed annually to preserve the independence of treatments and uphold data reliability. Soil texture remained clay loam in both years. In 2023, bulk density values were 1.45 g/cm³ and 1.43 g/cm³, with field capacity of 24.3% (105.70 mm) and 24.2% (103.81 mm), respectively. In 2024, bulk density varied more, with 1.34 g/cm³ in the upper layer and 1.51 g/cm³ in the lower, while field capacity was 21.96% (88.30 mm) and 21.92% (99.15 mm). Total field capacity values was 209.51 mm in 2023 and 187.45 mm in 2024 (Table 2).

Table 2.

Soil characteristics of the experimental site (2023–2024)

2023
Soil depth (cm)	Structure class	Bulk density (gr/cm3)	Field Capacity		Wilting Point
			%	mm	%	mm
0–30	Clay Loam	1.45	24.3	105.70	14.1	61.33
30–60	Clay Loam	1.43	24.2	103.81	14	60.06
Total (0–60 cm)				209.51		121.39

2024
Soil depth (cm)	Structure class	Bulk density (gr/cm3)	Field Capacity		Wilting Point
			%	mm	%	mm
0–30	Clay Loam	1.34	21.96	88.3	11.55	46.43
30–60	Clay Loam	1.51	21.92	99.15	11.99	54.24
Total (0–60 cm)				187.45		100.67

Statistical analysis

The collected data were statistically examined using the analysis of variance (ANOVA) method in accordance with a randomized complete block design (RCBD) split split plot method by using Statistix software (version 8.1). When the F-test was significant, the least significant differences (LSD) test was used to link data means [41].

Results

Evaluation of crop water use and irrigation requirements

Over two growing seasons (2023 and 2024), the effects of different irrigation treatments on irrigation amount, precipitation, crop evapotranspiration (ETc), and irrigation requirement (IRc) were evaluated (Table 3). The total irrigation amount varied significantly among treatments and years. In both years, the control treatment received the most irrigation, at 318.2 mm in 2023 and 396 mm in 2024. Conversely, the VE-VT treatment applied the least amount of irrigation in 2023 (163.8 mm); in 2024, the R3-R6 treatment used the least (215 mm). The precipitation amount was 210 mm in 2023 and 98.4 mm in 2024. Seasonal ETc values ranged from 462.11 to 595.88 mm in 2023 and from 429.4 to 523.39 mm in 2024. The highest ETc values were recorded in the control plots in both years. The ratio of applied irrigation water to evapotranspiration (IRc) was highest in the control plots in both years (53.40% in 2023 and 75.77% in 2024), reflecting a greater dependency on supplemental irrigation. The R3-R6 and VE-VT treatments consistently exhibited lower IRc values, especially in 2023 (38.32% and 35.45%, respectively). Overall, these results suggest that strategically timing irrigation (e.g., limiting it to specific growth stages, such as VE-VT or R3-R6) can significantly reduce water usage while maintaining comparable ETc levels, contributing to more efficient water management.

Table 3.

Evaluation of crop water use and irrigation requirements

Treatments	Irrigation amount (mm)		Precipitation (mm)		Etc (mm)		IRc (%)
	2023	2024	2023	2024	2023	2024	2023	2024
Control	318	396	210	98	596	523	53	76
VT-R3	257	306	210	98	537	451	48	68
R3-R6	216	215	210	98	563	429	38	50
VE-VT	164	267	210	98	462	433	35	62
Rainfed	0	0	210	98

Acid detergent fiber (%)

Acid detergent fiber (ADF) was significantly affected by the seasonal deficit irrigation, potassium application, and bacterial inoculation in both years (Table 4). R3-R6 irrigation consistently produced the highest ADF across years (62.55% in 2023 and 64.91% in 2024), followed by VT-R3. Early season deficit irrigation VE-VT, on the other hand, consistently produced the lowest ADF (53.58% and 52.73%, respectively). Rainfed and control treatments showed intermediate values. Potassium application significantly influenced ADF only in 2024. Bacterial inoculation increased ADF in both years, with higher values observed in B+ than B- plots. The interaction between irrigation and bacteria was significant in both seasons (Fig. 1).

Table 4.

Nutritional parameters of maize fodder as affected by seasonal deficit irrigation potassium and bacteria

Treatments		Examined traits
	ADF (%)		NDF (%)		Crude Protein (%)
Irrigation Levels (IL)	2023	2024	2023	2024	2023	2024
Control	59.34b	60.59b	83.00b	78.51c	2.45c	2.52c
VT-R3	58.78b	63.92a	73.78d	77.13d	2.82b	2.62c
R3-R6	62.55a	64.91a	87.67a	90.53a	2.27d	2.36d
VE-VT	53.58d	52.73d	72.41d	68.79e	2.76b	3.65a
Rainfed	56.07c	58.93c	78.26c	80.66b	3.54a	3.20b
LSD	1.15	1.2	1.51	1.18	0.14	0.14
F-Value	37.08	254.06	94.54	429.39	96.61	103.58
Potassium (P) kg ha− 1
0	56.31b	58.34b	77.44b	77.15b	2.59b	2.64b
5	59.82b	62.09a	80.63a	81.10a	2.95a	3.11a
LSD	0.73	0.75	0.96	0.74	0.09	0.09
F-Value	82.99	145.63	47.01	60.32	78.81	276.06
Bacteria (B)
B-	54.75b	56.76b	75.52b	74.82b	2.40b	2.52b
B+	61.38a	63.68a	82.53a	83.43a	3.14a	3.23a
LSD	0.73	0.75	0.96	0.74	0.09	0.09
F-Value	539.16	252.3	268.23	870.89	231.38	199.75
Interaction (F value)
I×P	1.57	5.64*	0.99	1.6	21.99**	15.21**
I×B	8.36**	13.43***	9.51**	30.42***	7.46**	5.86**
P×B	4.58*	0.52	1.88	36.25***	13.69**	7.84**
I×P×B	2.7	0.25	3.20*	0.57	7.54***	0.57

The difference between the values shown with different letters in the same columns is significant at P ≤ 0.05 level. ns: p ≥ 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001

Fig. 1.

Effects of seasonal deficit irrigation and bacteria interaction on crop response across the 2023 and 2024 seasons

In 2023, the highest ADF occurred under R3-R6 irrigation with bacterial inoculation, while the lowest ADF occurred under VE-VT irrigation treatment without bacteria. In 2024, the maximum ADF was recorded under VT-R3 with bacteria, and the minimum values were recorded under VE-VT without bacterial application. Across both years, bacterial inoculation consistently increased ADF, particularly under VT-R3 and R3-R6 irrigation. These treatments generally produced higher ADF values than the control, VE-VT, and rainfed treatments (Fig. 1).

The potassium × bacteria interaction was significant in 2023, but not in 2024 (Fig. 2). Conversely, the irrigation × potassium interaction became significant in 2024 (Fig. 3). The three-way interaction (irrigation × potassium × bacteria) was not significant in either year. In 2023, the highest ADF occurred with 50 kg K ha⁻¹ combined with bacterial inoculation, while the lowest ADF occurred in the untreated control.

Fig. 2.

Interactive effects of potassium fertilization and bacterial inoculation on crop responses during the 2023 and 2024 growing seasons

Fig. 3.

Crop response to combined effects of seasonal deficit irrigation and potassium fertilization in 2023 and 2024

Neutral detergent fiber (%)

Neutral detergent fiber (NDF) was significantly affected by irrigation regime, potassium application, and bacterial inoculation (Table 4). The R3-R6 irrigation regime produced the highest NDF values (87.67% in 2023 and 90.53% in 2024) and the VE-VT regime produced the lowest values (72.41% and 68.79%, respectively). Control and rainfed treatments showed intermediate responses. Potassium application slightly increased NDF in both years. Bacterial inoculation markedly increased NDF, with higher values were observed in B+ than in B- plots. The interaction between irrigation and bacteria was significant in both seasons, indicating that the effects of bacteria on NDF varied with the irrigation regime (Fig. 1).

The highest NDF values in 2023 occurred under the R3-R6 irrigation with bacterial inoculation, while the lowest values occurred under VT-R3 and VE-VT without bacteria. A similar pattern emerged in 2024: maximum NDF occurred under R3-R6 with bacteria, and minimum NDF occurred under VE-VT without bacteria. Overall, bacterial inoculation markedly increased NDF under R3-R6 irrigation, whereas its effect was smaller under VE-VT and rainfed conditions. These results suggest that the effects of bacteria on fiber accumulation depends on irrigation timing and growth stage (Fig. 1).

The potassium × bacteria (P × B) interaction was highly significant in 2024 (Fig. 2), though it was not significant in 2023. This suggests a delayed or environmentally mediated synergistic effect. The maximum NDF value occurred under 50 kg K ha⁻¹ with bacterial inoculation, while the minimum value occurred in the non-fertilized, non-inoculated treatment. Applying bacteria alone resulted in higher NDF than applying potassium alone, indicating that bacterial inoculation has a comparatively stronger effect on cell wall fiber accumulation.

Crude protein (%)

Crude protein (CP) content was significantly influenced by irrigation regime, potassium application, and bacterial inoculation in both 2023 and 2024 (Table 4). The highest CP occurred under rainfed conditions in 2023 (3.54%) and VE-VT deficit irrigation in 2024 (3.20%). The lowest values were recorded under R3-R6 water deficit in both years. Potassium fertilization significantly increased CP in both seasons, and bacterial inoculation consistently enhanced protein content, resulting in higher CP in B+ than B- plots.

The interaction between irrigation and bacteria was significant in both years. In 2023, the highest CP was recorded under rainfed conditions with bacterial inoculation, while the lowest CP occurred under R3-R6 without bacteria. In 2024, the maximum CP was observed under VE-VT with bacteria, whereas minimum values again occurred under R3-R6 without inoculation. Overall, bacterial inoculation consistently increased CP within each irrigation regime, with rainfed and VE-VT treatments producing higher CP than R3-R6, particularly in the absence of bacteria (Fig. 1). The potassium × bacteria interaction was significant in both years. In 2023, crude protein was highest with 50 kg K ha⁻¹ combined with bacterial inoculation and lowest in the untreated control. The same trend was observed in 2024, confirming the consistent positive effect of combined potassium fertilization and bacterial inoculation on crude protein content (Fig. 2).

Ash content (%)

Seasonal deficit irrigation regimes had a statistically significant effect on ash content (Table 5). However, potassium fertilization and bacterial inoculation treatments did not significantly effect ash content except in 2024. According to 2023 and 2024 data, plots without irrigation during VE-VT period showed the highest ash content (6.13% and 6.90%). This result suggests that early water stress may have a stimulating effect on mineral matter accumulation or concentration. Conversely, in 2023, during the R3-R6 period, the non-irrigated treatment had the lowest ash value with 4.78%, while there was no statistical difference with the rainfed treatment (4.90%). In 2024, the rainfed treatment had the lowest ash content (3.41%). These findings suggest that late or continuous water stress may have limiting effects on mineral accumulation. The control (2023: 5.15%, 2024: 6.25%) and the treatment without irrigation during the VT-R3 period (2023: 5.51%, 2024: 4.29%) had moderate ash content. Potassium treatment slightly increased ash content (control: 4.85%; 50 kg ha^− 1: 5.28%), but this difference was not statistically significant during first year of study (2023). Similarly, no significant difference was observed between B- (5.30%) and B+ (5.29%) in the bacterial inoculation treatment. The interaction between the treatments also did not reveal any statistically significant change. These results indicate that ash content was mainly sensitive to irrigation timing and only slightly affected by potassium fertilization.

Table 5.

Nutritional parameters and biological yield of maize as affected by seasonal deficit irrigation potassium and bacteria

Treatments	Examined traits
	Ash content (%)		Hemicellulose (%)		Biological yield (kg ha− 1)
Irrigation Levels (IL)	2023	2024	2023	2024	2023	2024
Control	5.15bc	6.25b	23.66ab	18.50b	34583a	36528a
VT-R3	5.51b	4.29c	15.00d	23.77a	26240b	26872c
R3-R6	4.78c	4.46c	25.11a	25.12a	22618c	23990d
VE-VT	6.13a	6.90a	18.83c	15.75c	27135b	29399b
Rainfed	4.90c	3.41d	22.19b	17.68bc	14052d	15424e
LSD	0.46	0.57	1.77	2.05	1158.7	1008.7
F-Value	13.63	30.89	35.1	51.07	300.86	822.26
Potassium (P) kg ha− 1
0	5.25	4.85b	21.12	20.22	23537b	24874b
50	5.34	5.28a	20.79	20	26314a	28011a
LSD	0.29	0.36	1.12	1.3	732.82	637.97
F-Value	0.44	12.49	0.29	0.21	67.86	134.06
Bacteria (B)
B-	5.3	5.11	20.77	19.83	22343b	23428b
B+	5.29	5.02	21.15	20.5	27508a	29457a
LSD	0.29	0.36	1.12	1.3	732.82	0.03
F-Value	0.01	0.24	0.57	1.04	200.95	282.19
Interaction (F value)
I×P	3.1	4.76*	0.51	3.03	ns	ns
I×B	1.88	0.28	5.29**	0.62	5.61**	9.08***
P×B	0.81	1.72	0.01	1.28	8.15**	8.09*
I×P×B	2.07	0.12	2.79	0.66	4.88**	2.89*

The difference between the values shown with different letters in the same columns is significant at P ≤ 0.05 level. ns: p ≥ 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001

Hemicellulose (%)

The hemicellulose content varied across irrigation treatments and between the two years of study (Table 5). In 2023, the highest hemicellulose content was recorded under the R3-R6 irrigation level (25.11%), followed by control (23.66%) and the rainfed (22.19%) treatments. The lowest value was observed under the VT-R3 treatment (15.00%). In 2024, R3-R6 again produced the highest hemicellulose content (25.12%), while VE-VT recorded the lowest (15.75%). Notably, hemicellulose content decreased in 2024 for most treatments compared to 2023, particularly in the control and rainfed plots. In 2023, plots receiving no potassium showed slightly higher hemicellulose content (21.12%) compared to those receiving 50 kg ha⁻¹ (20.79%). A similar trend was observed in 2024, with hemicellulose content of 20.22% and 20.00% for 0 and 50 kg ha⁻¹, respectively. This suggest that the rate of potassium has a minimal influence on hemicellulose levels. Bacterial inoculation was statistically non-significant in both years. However, marginally increased hemicellulose content in both years compared to uninoculated control (B-). In 2023, B+ plots recorded 21.15% compared to 20.77% in B-, while in 2024, the values were 20.50% and 19.83% for B + and B-, respectively. These results indicate a slight but consistent enhancement of hemicellulose accumulation with bacterial inoculation. A significant interaction between irrigation and bacterial inoculation (I×B) was observed in 2023 (Fig. 1), indicating that the effect of bacterial treatments on hemicellulose content was influenced by the irrigation regime. All other interaction effects (I×P, P×B, and I×P×B) were statistically non-significant in both years.

Biological yield (kg ha^− 1)

Biological yield (BY) was significantly affected by irrigation regime, potassium application, and bacterial inoculation in both 2023 and 2024 (Table 5). Control irrigation produced the highest yields (34583 and 36528 kg ha⁻¹), followed by VE-VT irrigation. Rainfed treatment resulted in the lowest yields (14052 and 15424 kg ha⁻¹). Potassium application (50 kg ha⁻¹) increased BY to 26,314 and 28,011 kg ha⁻¹, compared with 23,537 and 24,874 kg ha⁻¹ without potassium in 2023 and 2024, respectively. Bacterial inoculation also significantly enhanced yield, increasing BY from 22,343 to 27,508 kg ha⁻¹ in 2023 and from 23,428 to 29,457 kg ha⁻¹ in 2024.

Significant interactions were observed between irrigation and bacterial inoculation (I×B) in both years, suggesting that the response to bacteria depended on the irrigation regime (Fig. 1). Similarly, potassium and bacteria (P×B) interactions were significant in both years (Fig. 2). Biological yield increased progressively with the combined application of potassium and bacterial inoculation in both years. In 2023, yield ranged from 21,475 kg ha⁻¹ in the non-fertilized, non-inoculated control (K0 × B-) to 29,417 kg ha⁻¹ under 50 kg K ha⁻¹ with bacteria (K50 × B+). A similar response was recorded in 2024, with values spanning 22,369 to 31,536 kg ha⁻¹, again peaking in the K50 × B+ treatment. Across both seasons, treatments receiving a single input produced intermediate biomass levels. Likewise, the three-way interaction (I×P×B) was also significant in both years (Table 7). In 2023, the highest biological yield (41208 kg ha⁻¹) was obtained under full irrigation combined with 50 kg K ha⁻¹ and bacterial inoculation, while the lowest yield (10861 kg ha⁻¹) occurred under rainfed conditions without potassium and bacteria. A similar pattern was observed in 2024, with yields ranging from 43,347 kg ha⁻¹ under fully irrigated K50 × B + to 12,306 kg ha⁻¹ under rainfed, non-treated conditions. Overall, the combined application of potassium and bacteria consistently maximized yield, particularly under adequate water supply (Table 7).

Table 7.

Interactive effects of seasonal deficit irrigation, potassium fertilization and bacterial inoculation on crop responses during the 2023 and 2024 growing seasons

Treatments			NDF	CP	IWUE	WUE		Biological yield
Irrigation	Potassium	Bacteria	2023	2023	2023	2023	2024	2023	2024
Control	KO	B-	79.83ef	2.03j	65.34b-d	53.12c-e	60.74c-e	31653bc	31792d
	K50	B+	83.57d	2.57f-h	60.8c-e	56.45c	70.85b	33639b	37083b
	KO	B-	81.77de	2.27ij	58.7d-f	53.42 cd	64.75c	31833bc	33889c
	K50	B+	86.87bc	2.97d	73.20b	69.16a	82.82a	41208a	43347a
VT-R3	KO	B-	67.03j	2.23 h-j	47.51f-h	42.92gh	52.90 g-i	23056f	23,847 h
	K50	B+	68.67ij	2.87de	44.91gh	45.94 fg	54.75f-h	24681f	24681f-h
	KO	B-	76.33gh	2.63e-g	50.75e-g	50.83de	60.82c-e	27306e	27417e
	K50	B+	83.10d	3.57c	46.75gh	55.69c	69.97b	29,917 cd	31542d
R3-R6	KO	B-	83.40d	2.03j	39.12hi	34.27j	48.65i	19,306 g	20889i
	K50	B+	88.57b	2.37 g-i	43.20gh	42.01gh	57.93d-f	23667f	24875f-h
	KO	B-	87.00bc	2.03j	44.72gh	40.48hi	56.22e-g	22806f	24139gh
	K50	B+	91.73a	2.67ef	31.40i	43.84gh	60.68c-e	24694f	26056e-g
VE-VT	KO	B-	67.00j	2.30 h-j	71.04bc	48.69ef	53.17 g-i	22500f	23,014 h
	K50	B+	76.13gh	2.47f-i	88.51a	63.00b	74.04b	29111de	32,042 cd
	KO	B-	70.53i	2.40f-i	63.67b-d	51.03de	61.13 cd	23583f	26458ef
	K50	B+	76.00gh	3.90b	94.19a	72.16a	83.37a	33347b	36083b
Rainfed	KO	B-	73.87 h	2.97d	-	27.86k	50.08hi	10861i	12306k
	K50	B+	78.67 fg	4.33a	-	36.62ij	62.97c	14,278 h	15472j
	KO	B-	76.13gh	2.90a	-	33.73j	53.98f-h	13153hi	13264k
	K50	B+	84.40 cd	3.97b	-	45.95 fg	84.06a	17,917 g	20653i

The difference between the values shown with different letters in the same columns is significant at P ≤ 0.05 level.

Biological yield varied significantly with irrigation regime and Bacillus inoculation in both years. In 2023 and 2024, the highest yields were obtained under control irrigation with bacterial inoculation, while the lowest yields occurred under rainfed conditions without bacteria. Bacterial application was most effective under control and VE-VT irrigation, whereas its benefit was limited under rainfed conditions. Overall, Bacillus inoculation increased biological yield across irrigation regimes, but water stress consistently constrained biomass production (Fig. 1).

Water use efficiency (kg ha⁻¹ mm^− 1)

Water use efficiency (WUE) was significantly affected by irrigation regime, potassium application, and bacterial inoculation in both 2023 and 2024 (Table 6). In 2023, the highest WUE was recorded under VE-VT irrigation (58.72 kg ha⁻¹ mm⁻¹), followed by the control, whereas in 2024 the control treatment showed the maximum WUE (69.79 kg ha⁻¹ mm⁻¹), with VE-VT also performing well. Rainfed and R3-R6 treatments produced the lowest WUE in both years. Potassium application (50 kg ha⁻¹) significantly increased WUE compared with no potassium, and bacterial inoculation markedly enhanced WUE, increasing values from 43.14 to 53.57 in 2023 and from 55.63 to 70.75 in 2024. The irrigation × bacteria interaction was highly significant in both seasons, indicating that bacterial effects on WUE depended on irrigation regime (Fig. 1).

Table 6.

Water Use Efficiency (WUE) and Irrigation Water Use Efficiency (IWUE) as influenced by irrigation regimes, potassium application and bacterial inoculation

Treatments	Examined traits
	WUE (kg ha− 1.mm)		IWUE (kg ha− 1.mm)
Irrigation Levels (IL)	2023	2024	2023	2024
Control	58.03a	69.79a	64.52b	53.21a
VT-R3	48.84b	59.60c	47.47c	37.33b
R3-R6	40.15c	55.86d	39.68d	39.67b
VE-VT	58.72a	67.92a	79.35a	52.34a
Rainfed	36.04d	62.77b	-	-
LSD	2.27	2.37	5.60	4.37
F-Value	155.84	80.37	128.51	87.86
Potassium (P) kg ha− 1
0	45.57b	59.21b	58.32	45.71
50	51.14a	67.17a	57.19	45.57
LSD	1.44	1.50	3.96	3.09
F-Value	64.15	198.98	0.32	0.02
Bacteria (B)
B-	43.14b	55.63b	54.37b	44.45
B+	53.57a	70.75a	61.14a	46.82
LSD	1.44	1.50	3.96	3.09
F-Value	214.42	303.44	11.14	1.65
Interaction (F value)
I×P	ns	ns	ns	ns
I×B	7.92***	10.86**	9.01**	8.61**
P×B	7.99*	11.17**	ns	9.67**
I×P×B	3.80*	4.47**	4.08*	ns

The difference between the values shown with different letters is significant at P ≤ 0.05 level. ns: p ≥ 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001

For the irrigation × bacteria interaction, the highest WUE in 2023 was observed under VE-VT irrigation with bacterial inoculation, while the lowest WUE occurred under rainfed conditions without bacteria. A similar pattern was noted in 2024, with maximum WUE again under VE-VT with bacteria, followed by the control with bacterial application, whereas the lowest values were recorded under R3-R6 and rainfed treatments without bacteria. Overall, bacterial inoculation most effectively improved WUE under VE-VT and control irrigation, while its effect was limited under R3-R6 and rainfed conditions (Fig. 1). The potassium × bacteria interaction was significant in both years (Fig. 2). In 2023, WUE was highest under 50 kg K ha⁻¹ with bacterial inoculation (57.36 kg ha⁻¹ mm⁻¹) and lowest in the untreated control (41.37 kg ha⁻¹ mm⁻¹). A similar response occurred in 2024, with maximum WUE (76.18 kg ha⁻¹ mm⁻¹) in K50 × B + and minimum WUE (53.11 kg ha⁻¹ mm⁻¹) in K0 × B-, confirming the consistent synergistic effect of potassium and bacterial inoculation on water use efficiency.

Irrigation water use efficiency (kg ha⁻¹ mm^− 1)

Seasonal deficit irrigation significantly affected IWUE in both 2023 and 2024 (Table 6). The highest IWUE values were recorded under VE-VT irrigation (79.35 and 52.34 kg ha⁻¹ mm⁻¹) and the lowest values occurred under R3-R6 irrigation in both years. Control irrigation showed intermediate IWUE. Potassium application had no significant effect on IWUE. Bacterial inoculation increased IWUE, from 54.37 to 61.14 kg ha⁻¹ mm⁻¹ in 2023 and from 52.0 to 53.2 kg ha⁻¹ mm⁻¹ in 2024; however, the increase was only statistically significant in 2023.

A highly significant irrigation × bacteria interaction was observed in both years (Fig. 1). In 2023 and 2024, the highest IWUE occurred under VE-VT irrigation with bacterial inoculation, while R3-R6 irrigation consistently resulted in the lowest IWUE, particularly with or without bacteria. Overall, bacterial inoculation improved IWUE primarily under VE-VT and full irrigation, whereas its effect was limited under R3-R6 and rainfed conditions. The three way interaction (irrigation × potassium × bacteria) was significant in 2023 but not in 2024 (Table 7). No significant irrigation × potassium interaction was detected in either year, while a potassium × bacteria interaction was significant only in 2024 (Fig. 2).

Discussion

Seasonal deficit irrigation, potassium fertilization, and bacterial inoculation significantly influenced acid detergent fiber (ADF) and neutral detergent fiber (NDF). Early irrigation (VE-VT) consistently reduced fiber accumulation, while late irrigation (R3-R6) promoted higher ADF and NDF levels due to increased lignification during the reproductive stage. These results imply that deficit irrigation during the early vegetative stage (VE-VT) is preferable for maintaining lower ADF content. In contrast, late irrigation (R3-R6) increases fiber accumulation in the forage. Adequate or excessive irrigation promotes vigorous vegetative growth, particularly stem and leaf elongation. This leads to greater deposition of structural carbohydrates, such as cellulose, hemicellulose, and lignin, which are the primary components of ADF and NDF [42]. The application of potassium, particularly at a rate of 50 kg ha⁻¹, enhanced NDF while reducing ADF in some cases, likely due to improved nutrient translocation and cellulose biosynthesis, which support structural development [43]. Interestingly, bacterial inoculation significantly increased both ADF and NDF, suggesting that these microbial treatments may enhance nutrient uptake or activate metabolic pathways favoring lignocellulose formation. Although bacterial inoculation increased forage yield, it was also associated with higher ADF and NDF content, which can negatively affect digestibility and intake. The elevated ADF and NDF concentrations observed under late-season deficit irrigation have direct consequences for forage utilization. Increased ADF reflects greater lignification, which reduces cell wall digestibility, whereas higher NDF limits voluntary dry matter intake by livestock. Consequently, irrigation strategies that increase fiber accumulation during reproductive stages may compromise forage feeding value, despite maintaining or increasing biomass production. These findings emphasize the importance of irrigation timing in managing the trade-off between forage yield and nutritional quality. During periods of forage shortage, the deficit is usually compensated for with cereal straw, which has low nutritional value. Maize leaves and husks have moderate nutritional value and are highly palatable. Therefore, the plant residues (maize stover) remaining after grain harvest are used in animal feed. Using maize stover as a low cost source of animal feed is of considerable economic importance. Significant interaction effects, particularly irrigation × bacteria and potassium × bacteria, indicate that microbial stimulation of fiber traits is modulated by water and nutrient availability. While increased NDF can limit voluntary intake and elevated ADF reduces digestibility [44], these results underscore the need to optimize irrigation timing and inoculant selection to balance yield with forage quality. For ADF, the significant interaction between irrigation and bacteria indicates that bacterial inoculation increased lignified fiber most significantly under R3-R6 irrigation. This suggests that bacterial stimulation of lignin deposition is amplified during reproductive stages, when water availability supports secondary cell wall formation. In contrast, the interaction between irrigation and bacteria for NDF suggests that bacterial inoculation mainly increased total cell wall components under VE-VT and VT-R3 irrigation. This indicates enhanced structural carbohydrate accumulation during active vegetative growth rather than increased lignification.

Although bacterial inoculation increased forage yield, it was also associated with higher ADF and NDF contents. This indicates a clear trade-off between biomass production and forage quality. These results suggest that bacterial inoculation is most suitable when combined with early-season deficit irrigation (VE-VT), as yield benefits can be achieved with smaller penalties to digestibility [44]. Overall, the interaction among water availability, potassium nutrition, and bacterial inoculation indicates that fiber accumulation is driven by growth stage specific responses where water and nutrients regulate whether microbial stimulation favors vegetative biomass expansion or lignified structural development.

The crude protein content of maize fodder was significantly affected by irrigation, potassium fertilization, and bacterial inoculation. The increase in crude protein values under rainfed and deficit irrigation conditions may be due to the crops ability to combat stress. During drought, plants accumulate free amino acids and stress-related proteins to protect cells. This leads to increased protein synthesis as a defense mechanism [45]. Potassium fertilization increased protein content, possibly due to its role in supporting nitrogen assimilation and enzyme activity [46, 47]. Bacterial inoculants further increased protein content by improving nitrogen provision and nutrient uptake [48, 49]. Under deficit irrigation, the combined effects of potassium nutrition and bacterial inoculation likely improved nitrogen assimilation efficiency. This enabled sustained protein synthesis despite reduced water availability. These findings underscore the importance of integrating water and nutrient management with microbial inoculation to optimize forage protein quality. The ash content of maize fodder was primarily influenced by seasonal deficit irrigation. Early-season water deficits, from emergence to tasseling, significantly increased ash levels, likely due to concentration effects or enhanced mineral uptake under mild stress [50]. In contrast, prolonged or late season water stress (R3-R6) reduced ash content. This decline reflects limitations on root growth and transpiration-driven nutrient flow induced by stress. It indicates that severe or prolonged water stress restricts mineral acquisition rather than concentrating nutrients in plant tissues. This suggests that extended drought limits root activity and mineral translocation [51, 52]. Potassium fertilization and bacterial inoculation had a minimal impact on ash levels, which is consistent with previous reports that these inputs did not significantly alter total mineral accumulation in forage crops [53]. The lack of significant treatment interactions further underscores that ash content is largely governed by soil moisture availability during the early growth stages rather than by nutrient or microbial inputs.

The hemicellulose content varied significantly with seasonal deficit irrigation, increasing during late stage irrigation (R3-R6). This increase is likely due to enhanced secondary cell wall biosynthesis during reproductive growth [54, 55]. During the reproductive stage, the allocation of biomass shifts toward stem thickening and tissue maturation, increasing the deposition of hemicellulose as part of the reinforced cell wall architecture [56]. Lower values under VE-VT and VT-R3 deficit irrigation suggest reduced structural carbohydrate synthesis under early water stress [57]. Early-season water stress likely suppresses cell expansion and delays secondary cell wall formation. This reflects a stress-adaptive strategy that prioritizes survival and metabolic maintenance over structural biomass accumulation. Potassium application had minimal influence, which is consistent with reports that hemicellulose accumulation depends more on water than on nutrients [47, 53]. Bacterial inoculation slightly increased hemicellulose content, possibly by promoting root growth and nutrient uptake under stress conditions [48, 49, 58]. This response suggests that bacterial inoculation may improve nutrient assimilation efficiency and carbon allocation to structural polysaccharides when water is available for continued growth. Overall, the irrigation-dependent response of hemicellulose indicates that water availability dominantly regulates cell wall carbohydrate accumulation. Bacterial effects primarily become evident under conditions that support active growth rather than severe stress.

The biological yield of maize increased significantly with irrigation, potassium fertilization, and bacterial inoculation. The maximum biological yield was recorded under full irrigation, reflecting optimal conditions for photosynthesis and biomass accumulation [59]. Meanwhile, yield reduction under rainfed conditions indicates limitations on growth and assimilate transport due to drought [45]. With an adequate water supply, biomass partitioning favors vegetative tissues, such as stems and leaves. This results in greater accumulation of harvestable stover and a higher biological yield. A water deficit reduces leaf expansion, photosynthetic capacity, and carbon fixation, ultimately constraining biomass accumulation and biological yield [51].

Potassium may have contributed to the improvement in yield by supporting stomatal function, enzyme activity, and sugar translocation. Increased potassium availability improves assimilate transport and carbon allocation efficiency, which supports sustained biomass production, even under moderate water limitations [60, 61]. Bacterial inoculants have been reported to promote biomass production by enhancing nutrient uptake, improving root architecture, and increasing plant resilience to environmental stresses [48, 49, 58, 62]. The positive effects of Bacillus inoculation on biomass production and yield can be attributed to several physiological and biochemical mechanisms. Bacillus species are known to enhance root growth and root surface area through the production of phytohormones such as indole-3-acetic acid (IAA) and cytokinin, which improves nutrient acquisition and supports vegetative growth [48, 58]. Significant two-way interactions highlight the synergistic potential of combining microbial and agronomic inputs under water-limited or nutrient-deficient conditions.

Deficit irrigation during the early growth stages (VE-VT), during which no water was applied from emergence to tasseling, resulted in the highest IWUE in 2023. Water limitation during these early stages likely induced adaptive physiological responses, such as improved stomatal control and reduced non-productive water loss. This allowed the plants to maintain biomass production with lower irrigation input. These results are consistent with previous studies that emphasized the critical role of early vegetative water management in optimizing biomass accumulation and yield under limited water conditions [63, 64]. Although full irrigation produced similar WUE values, VE-VT was more efficient in terms of irrigation-specific productivity. This confirms that strategic irrigation scheduling enhances resource use. This indicates that, under VE-VT deficit irrigation, assimilates were preferentially allocated to harvestable biomass rather than excessive vegetative growth. This improves the yield per unit of applied irrigation water. Potassium application significantly improved water use efficiency (WUE) but had no effect on irrigation water use efficiency (IWUE), which aligns with reports that potassium enhances drought tolerance and stomatal regulation, especially under limited water conditions [46, 53]. Improvements in osmotic regulation and photosynthetic efficiency mediated by potassium enhance carbon assimilation per unit of water transpired. This increases water use efficiency (WUE) without directly affecting irrigation-derived efficiency. These findings suggest that potassium is more beneficial in improving water productivity under rainfed or partially irrigated systems. Bacterial inoculation significantly increased WUE and IWUE in 2023, likely due to improved root function, nutrient uptake, and stress mitigation mechanisms, as documented in the literature [48, 65]. Although IWUE improvements in 2024 were not statistically significant, the trend persisted, indicating year-dependent variability in the effects of microbes. Significant I×B interactions in both years imply a synergistic effect between early irrigation and microbial inoculation. Furthermore, the P×B interaction for WUE suggests that integrated nutrient-microbe strategies could optimize water use under specific conditions [66]. The beneficial effects of microbial inoculation on WUE and IWUE were most pronounced under early-stage deficit irrigation, where physiological efficiency rather than the absolute amount of water supplied determined productivity. Similarly, the potassium-bacteria interaction for WUE suggests that a combination of nutrients and microbes can improve water use efficiency under conditions of moderate water limitation. Improvements in water use efficiency and irrigation water use efficiency following Bacillus inoculation may be associated with enhanced regulation of plant water status under deficit irrigation. Bacillus spp. have been reported to influence root hydraulic conductivity and improve soil root interactions, facilitating more effective water uptake at lower soil moisture levels [67–69]. Additionally, plant growth promoting bacteria maintain cellular function under drought conditions by accumulating compatible solutes (e.g., proline, glycine betaine, trehalose), production of heat shock proteins, and synthesis of extracellular polymeric substances. These substances improve membrane stability, protect enzymes, and enhance microenvironmental water retention [70–73].

Conclusion

This study clearly demonstrated that the timing of irrigation, potassium fertilization, and bacterial inoculation significantly influence the quality of the forage, the biological yield, and the water use efficiency of maize. Early-season deficit irrigation from crop emergence to the tasseling stage (VE-VT) most effectively enhanced crude protein content and water productivity. This irrigation strategy is preferable and recommended under water-limited conditions to improve forage quality and efficiency. In contrast, late-season irrigation (R3-R6) increased fiber fractions but reduced protein content and yield. Higher fiber accumulation may negatively affect forage digestibility and intake, representing a quality and quantity trade-off. Bacterial inoculation improved plant performance consistently across all parameters, including crude protein content, yield, and water use efficiency. Potassium application contributed to a higher yield and improved nutrient uptake; however, its effects on ash and hemicellulose content were minimal. The significant interaction effects, particularly between irrigation and bacterial inoculation, underscore the importance of integrated management strategies for optimizing maize forage production. Integrated management helps balance yield improvement with forage quality, rather than only optimizing production.

Recommendation

Deficit irrigation should be applied during the vegetative crop growth stage, particularly before the appearance of tassel (VE-VT), to enhance protein content and water use efficiency. Deficit irrigation should be avoid during the R3-R6 stages to prevent yield and quality losses. If sufficient water is available, full irrigation should be applied to maximize crop potential; however, under limited water conditions, irrigation should be strategically scheduled based on critical growth stages to ensure optimal plant performance, especially under Mediterranean climatic conditions. The use of plant growth-promoting bacteria such as Bacillus is strongly recommended to improve nutrient uptake, biomass and efficiency particularly under moderate water availability and adequate baseline soil fertility conditions. Applying potassium at 50 kg ha⁻¹ can support protein synthesis and overall plant development, depending on initial soil potassium status. An integrated approach combining timely irrigation, moderate potassium fertilization, and bacterial inoculation offers the most effective strategy for optimizing maize fodder production.

Limitation

This study was conducted under semi-arid conditions. Extrapolation to other environments should be made with caution, and multi-location studies are needed for validation. Moreover, bacterial inoculation and potassium application increased forage yield, they also raised ADF and NDF contents, potentially reducing forage quality. This highlights the need to balance yield gains with digestibility and intake considerations.

Authors’ contributions

H.N. İ.A. and U.Ş. designed the study; C.T. and H.N conducted the research and wrote the first draft; H.N. C.T. İ.A. and U.Ş. revised the paper; H.N. and C.T contributed to the discussion of the paper.

Funding

This research did not receive any fund.

Data availability

Data that supports the findings are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Not appropriate.

Consent for publication

Not appropriate.

Competing interests

The authors declare no competing interests.