Yield Formation and Stability of Maize Under Monoculture in Response to Biological Amendments, Weather Variability and Cultivar Maturity

Abstract

Contemporary agriculture faces the challenge of sustaining crop productivity amid increasing climatic pressures and simplified agronomic practices, such as monoculture. A field experiment conducted from 2022 to 2024 aimed to determine the effects of meteorological conditions and biological amendments on grain yield and yield structure in three maturity groups of continuous maize (Zea mays L.; FAO 200, 230 and 260). The split-plot experiment included applications of the biological amendments Neosol, Bactim Gleba and UGmax. Deteriorating agrometeorological conditions over the years studied led to a progressive decline in mean grain yield, reaching the lowest value in 2024 (5.06 Mg ha⁻¹). The cultivar belonging to the FAO 260 maturity group exhibited the highest yield potential. Application of all biological amendments resulted in a significant increase in grain yield and thousand-grain weight compared with the untreated control. The most effective treatment was UGmax which increased mean grain yield by approximately 14% and thousand-grain weight by 19% compared with the control. Path analysis revealed hierarchical relationships among components of ear structure and grain yield. The primary direct effect on yield increase was the number of kernels per ear, with thousand-grain weight also contributing significantly depending on maturity group. In later-maturing cultivars, kernel number per ear played the dominant role, whereas thousand-grain weight was more influential in earlier-maturing ones. The economic analysis demonstrated that all of the applied biological amendments generated a positive net profit, with the highest additional revenue obtained following the application of UGmax (160 USD·ha⁻¹). These results confirm that biostimulant application affected grain yield formation, and reduced yield losses under stress conditions.

1. Introduction

Maize (Zea mays L.) ranks among the most important crop species grown worldwide [1]. Its versatile uses make it a promising cereal [2,3]. Maize plays a key role in global food and energy security, a significance that has become even more pronounced in 2026 amid escalating climatic pressures, geopolitical instability, and environmental constraints. In many European regions practising intensive agriculture, maize has increased its share in cropping systems, with a growing tendency towards monoculture. This shift is driven by economic viability, farm specialisation, and simplified production organisation [4]. The results of numerous studies conducted under diverse soil and climatic conditions indicate that simplifications in the cultivation system, including long-term monoculture, may lead to reduced yield stability, increased susceptibility to abiotic stresses, and a gradual decline in agroecosystem productivity [5,6,7]. Moreover, the observed increased sensitivity of plants to abiotic stresses, including drought, may in the long term lead to yield reduction [8]. These trends underscore the need for sustainable agronomic practices that integrate biological and technical approaches to mitigate the drawbacks of monoculture, improve soil health, and enhance production system resilience to water deficits [9]. It has been demonstrated that the use of biological amendments can enhance plant productivity [10,11]. Moreover, their application may strengthen plant resistance to environmental stress, which in turn can improve both the quality and quantity of the yield [12]. Contemporary agriculture must maintain maize productivity under increasingly unpredictable weather patterns and simplified agronomic systems, including repeated cultivation on the same site—a practice not uncommon for maize [13]. The use of biological amendments therefore becomes essential, as they support plants during stress periods and help ensure satisfactory and relatively stable yields across successive years. It is also important to understand how maize cultivars of differing maturity respond to such interventions. Identifying these differences and the mechanisms underlying yield formation in altered environments can guide the development of cropping technologies that are both economically viable and environmentally sustainable. In maize research, information on grain yield alone is often insufficient. To achieve meaningful improvements in productivity, it is necessary to identify the specific traits contributing to yield and their interactions [14,15]. Path analysis is particularly valuable in this context, as it quantifies the direct and indirect effects of traits on yield under varying cropping conditions [16,17].

Because path analysis enables the identification of traits that genuinely determine grain yield [18,19] rather than merely correlate with it, this study applies Wright’s path coefficient method to address a significant gap in the existing literature. To date, research on biological amendments in maize has largely focused on yield magnitude, with limited attention given to the hierarchical structure of yield components and the causal relationships underlying productivity formation under monoculture conditions. The novelty of the present study lies in the integrated analytical framework combining classical analysis of variance (ANOVA) with Wright’s path analysis in a long-term field monoculture experiment. Such a methodological combination remains rarely applied in agronomic studies evaluating biological amendments. While ANOVA identifies statistically significant treatment effects, path analysis simultaneously clarifies biological cause–effect relationships among yield components. This dual approach allows not only quantification of productivity responses, but also explanation of the mechanisms driving those responses. Moreover, the study jointly evaluates three interacting determinants of maize productivity: interannual meteorological variability, cultivar maturity group (FAO 200, 230, 260), and biological amendment application. By integrating these factors within a single hierarchical model, the research provides a more comprehensive understanding of yield-formation mechanisms under monoculture stress. The theoretical contribution of this work therefore lies in identifying cultivar-specific yield-formation strategies and amendment-induced modifications in causal yield structure, which may support the design of climate-adaptive and resource-efficient cropping systems.

This research focused on the productive effects of biostimulant application and their influence on yield components, rather than on the dynamics of physicochemical soil parameter changes occurring over time. This study aimed to evaluate the effects of varying meteorological conditions and biological amendment applications on maize yield and yield-structure traits under monoculture conditions, and to determine hierarchical relationships among yield components using path analysis.

2. Materials and Methods

The field experiment was set up as a split-plot arrangement with three replicates and conducted from 2022 to 2024 at the Agricultural Experimental Farm in Zawady (52°03′ N, 22°33′ E), which belongs to the University of Siedlce. Each year, the experiment was established on the same site to achieve the effects of continuous monoculture cropping. The harvest area of each plot was 30 m². The experiment examined the following factors:

Factor A—Maize cultivars:

• A1—Early-maturing hybrid, FAO 200;
• A2—Medium–early-maturing hybrid, FAO 230;
• A3—Medium–late-maturing hybrid, FAO 260.

Factor B—Biological amendments:

• B1—Untreated control;
• B2—Neosol;
• B3—Bactim Gleba;
• B4—UGmax.

The soil was classified as a proper lessive soil (podzolic group) with a sandy granulometric composition, a very good rye complex, and a quality class IVa [20]. Prior to maize sowing, the soil nutrient status was analysed. The soil reaction was slightly acidic (pH ranged from 5.67 to 5.70). The soil showed low available phosphorus and medium available potassium and magnesium levels (

Table 1. Soil content of available nutrients before establishing the experiment.

Parameter	Unit	Value
pH in KCL	–	5.8
Phosphorus (P)	mg·kg−1	40.5
Potassium (K)	mg·kg−1	102.6
Magnesium (Mg)	mg·kg−1	41.4

).

Soil pH was determined using the potentiometric method. The content of the available forms of phosphorus and potassium was determined using the Egner–Riehm method (DL), while magnesium content was determined using the Schachtschabel method.

The physio-chemical soil parameters were determined prior to the establishment of the experiment. Detailed monitoring of the biological, chemical, and physical changes in the soil properties during the three-year study period was not included in the experimental design. The investigation focused primarily on the productive response of plants (yield and yield components) as an indicator of soil–plant system functioning. Therefore, interpretations concerning soil improvement refer to functional effects observed in plant responses rather than to directly measured changes in soil parameters.

Maize was grown in monoculture. Following harvest of the preceding crop, post-harvest tillage operations were performed, followed by winter ploughing to a depth of 25 cm. In early spring, agronomic practices were conducted to level the soil surface and promote warming. Mineral NPK was applied before planting. Potassium was supplied as potash salt at 250 kg ha⁻¹, equivalent to 124.5 kg K ha⁻¹. Nitrogen was applied at 100 kg N ha⁻¹, corresponding to 291 kg ha⁻¹ of ammonium nitrate (34.4% N). Fertilisers were incorporated into the soil using a combined cultivator. Each spring, pre-plant applications of biological amendments were made, and they were mixed into the soil to a depth of 5–10 cm. Bactim Gleba (Intermag sp. z o.o., Olkusz, Poland) is a bioproduct that accelerates revitalisation of ‘fatigued’ soils, i.e., those degraded to varying degrees by intensive cropping and unbalanced fertilisation. It contains Bacillus spp. bacteria isolated from natural soil environments, occurring in the form of spores, and a humic co-formulator supporting microbial development. It was applied as a soil spray at 1.5 L ha⁻¹. Granular Neosol (PRP Polska sp. z o.o., Warsaw, Poland) acts as an activator of soil microbial biomass, stimulates humic acid production, promotes humus formation through enhanced biological activity, and restores soil fertility. The active substance MIP SOIL in Neosol supplies minerals (iron, zinc, boron, and manganese compounds on a calcium and magnesium carbonate matrix) essential for soil microflora and for enzymes involved in organic matter transformations. It was applied at 150 kg granules ha⁻¹. UGmax (Agrowitan sp. z o.o., Skarszewy, Poland) contains microorganisms along with key macro- and microelements. It improves soil physical, chemical, and biological properties, increases organic matter content, and enhances soil structure. Its advanced formulation, including lactic acid bacteria (Lactobacillus) and an organic matter content of at least 40%, accelerates the natural decomposition of organic matter. UGmax was applied as a soil spray at 2 L ha⁻¹.

Maize was sown using a pneumatic precision drill with simultaneous row fertilisation using triple superphosphate (46% P₂O₅) at 71.8 kg ha⁻¹, equivalent to 33 kg P ha⁻¹. Seeds were placed in rows spaced 75 cm apart.

During the growing season, the experiment was conducted under real rainfall conditions and no additional irrigation was applied at the test sites.

Measurements were taken at full maturity (BBCH 89) on plants harvested from a 1 m² area within each plot, excluding border rows. The following biometric traits were assessed:

• Number of kernel rows per ear;
• Number of kernels per row;
• Number of kernels per ear;
• Thousand-grain weight (TGW, g).

Grain yield was determined from mechanical harvest of the plot area and recalculated to Mg ha⁻¹ at 14% grain moisture.

2.1. Meteorological Conditions

Data on the air temperature and atmospheric precipitation sums during the maize growing season were obtained from an automatic meteorological station located at the Agricultural Experimental Farm (AEF) in Zawady, belonging to the University of Siedlce, equipped with sensors and software manufactured by LAB-EL Laboratory Electronics sp. z.o.o, Reguły, Poland and regularly calibrated under laboratory conditions. Due to its local, institutional status, the station is not included in the Global Historical Climate Network Daily (GHCN-D) database and does not have an RSM identification number.

These data served as the basis for evaluating the thermal and moisture conditions in the years studied and their influence on maize growth and development. The Sielianinov hydrothermal coefficient (K) was calculated, and the hydrothermic conditions were classified according to Skowera and Puła [21]. Nine moisture classes were distinguished. Values of K ≤ 0.4 indicate extremely dry conditions. The range 0.4 < K ≤ 0.7 corresponds to very dry conditions, while 0.7 < K ≤ 1.0 indicates dry conditions. The interval 1.0 < K ≤ 1.3 denotes fairly dry conditions. Values of 1.3 < K ≤ 1.6 are considered optimal from the standpoint of moisture supply. The range 1.6 < K ≤ 2.0 characterises fairly moist conditions, 2.0 < K ≤ 2.5 moist conditions, 2.5 < K ≤ 3.0 very moist conditions, and K > 3.0 extremely moist conditions.

Mean air temperature during the study period (2022–2024) ranged from 8.2 °C in April to 21.0 °C in August (

Table 2. Meteorological conditions at the Zawady AEF station in 2022–2024.

Month/Year	April		May		June		July		August		September
	t	P	t	P	t	P	t	P	t	P	t	P
2022	5.2	31.5	13.6	31.1	19.9	26.5	19.3	95.7	21.0	39.3	11.7	64.9
2023	8.7	12.4	13.4	46.5	18.0	53.6	20.3	31.4	21.3	25.0	18.0	16.6
2024	10.8	23.2	16.6	5.2	19.0	65.4	21.6	59.7	20.7	55.5	18.1	25.5
Mean 2022–2024	8.2	22.4	14.5	27.6	19.0	48.5	20.4	62.3	21.0	39.9	15.9	35.7

t—Mean monthly air temperature, P—monthly total precipitation.

). The coolest month in the analysed half-year was April, while July and August were the warmest, favouring vigorous plant growth during the growing season. Interannual temperature variability was moderate, with 2024 showing the highest values in most months. Monthly precipitation totals exhibited substantial variability, both seasonally and between years. Over the 2022–2024 period, the highest average precipitation occurred in July (62.3 mm) and June (48.5 mm), consistent with the typical summer convective rainfall maximum. The lowest average totals were recorded in May (27.6 mm) and April (22.4 mm). Notably high precipitation was observed in July 2022 (95.7 mm), which may have caused temporary soil water excess. In contrast, precipitation in May 2024 was very low (5.2 mm), suggesting the potential for short-term dry conditions.

Moisture conditions during the maize growing season from 2022 to 2024 showed considerable variation, with a predominance of dry and very dry periods, particularly during the critical developmental stages of the plants. In 2022, favourable conditions occurred only in April (wet) and July (moist), whereas May (dry), June (very dry), and August (very dry) exhibited moisture deficits that likely restricted plant growth and grain filling (

Table 3. Values of Sielianinov hydrothermal coefficient (K).

Year/Month	2022	2023	2024
April	2.01 w.	0.48 v.d.	0.72 d.
May	0.74 d.	1.12 r.d.	0.10 ex.d.
June	0.44 v.d.	0.99 d.	1.15 r.d.
July	1.60 op.	0.50 v.d.	0.89 d.
August	0.60 v.d.	0.38 ex.d	0.86 d.
September	1.84 f.m.	0.31 ex.d	0.47 v.d.

Month: Extremely dry—ex.d., very dry—v.d., dry—d., rather dry—r.d., optimal—op., rather wet—r.w., wet—w., fairly moist—f.m.

). In 2023, dry and very dry conditions prevailed for most of the growing season, encompassing key phases of intensive growth, flowering, and grain filling. In 2024, moisture conditions were more variable, yet dry conditions still dominated during critical phases, especially in May (very dry) and during flowering and grain filling.

2.2. Statistical Analysis

Because the experiment was conducted on the same site across successive years, appropriate statistical methods were required to account for the repeated use of the location. The blocks and sub-blocks randomly selected in the first year remained fixed in subsequent years. In such cases, following standard principles for field experiments, the two-factor split-plot design was subjected to a combined (three-year) analysis according to a split-split-plot model, in which years were treated as main plots (factor A levels).

Analysis of variance for each trait was performed using the following mathematical model [22]:

y_ijlp = m + a_i + g_j + e_ij^/1/ + b_l + ab_il + e_ijl^/2/ + c_p + ac_ip + bc_lp + abc_ilp + e_ijlp^/3/,

where

y_ijlp is the value of the trait for the i-th level of factor A (year), l-th level of factor B (cultivar), and p-th level of factor C (biological amendment) in the j-th block (replicate). Other terms are defined as follows:

• m is the overall experimental mean;
• a_i, b_l, and c_p are the main effects of the respective factors;
• g_j is the effect of the j-th block;
• e_ij^/1/ is the whole-plot error;
• ab_il, ac_ip, and bc_lp are the two-factor interaction effects;
• e_ijl^/2/ is the sub-plot error;
• abc_ilp is the three-factor interaction effect;
• e_ijlp^/3/ is the sub-sub-plot error.

Means were compared and interactions evaluated using the Tukey’s Honestly Significant Difference (HSD) test [22].

In addition to analysis of variance, Wright’s path analysis was applied to determine the direct and indirect relationships between the yield components and final grain yield [14,23,24]. The analysis was based on a biologically justified hierarchical model that accounted for the sequential development of generative organs and the grain-filling process. The model adopted a sequential structure in which morphological traits (number of kernel rows and number of kernels per row) determine the number of kernels per ear, which in turn, together with thousand-grain weight (TGW), shapes the final yield.

Analysis of variance was conducted using Statistica 13.3 software (with the Zestaw Przyrodnika add-on). Direct and indirect path effects according to Wright were calculated using structural equation modelling (SEM) available in Statistica 13.3. All calculations were performed at a significance level of α = 0.05.

2.3. Economic Analysis

To assess the practical applicability of the examined technology, a simplified economic analysis was conducted using the gross margin method, which is commonly employed in agronomic research to evaluate the profitability of a single technological factor [25]. The analysis considered only the differences resulting from the application of soil conditioners, assuming identical costs for all other technological components. Such an approach makes it possible to determine the economic efficiency of the examined factor independently of production cost variability across farms. The economic assessment used the long-term average market price of maize grain, calculated on the basis of stock exchange quotations from the end of December in 2022, 2023, and 2024 [26]. Using the average grain price enabled an objective evaluation of the economic performance of the tested products by eliminating the influence of substantial year-to-year market fluctuations. The average purchase price of maize grain, converted to USD, amounted to 193 USD per Mg. The estimated purchase costs of the products, depending on the supplier, were as follows: UGmax (rate 2 L·ha⁻¹), approximately 24–29 USD·ha⁻¹; Bactim Gleba (rate 1.5 L·ha⁻¹), approximately 22–27 USD·ha⁻¹; and Neosol (rate 150 kg·ha⁻¹), approximately 44–53 USD·ha⁻¹ [27]. For data standardisation purposes, all costs and revenues were converted from PLN to USD according to the average exchange rate of the National Bank of Poland on the last working day of December for each of the analysed years (1 USD = 4.14 PLN) [28].

3. Results

Statistical analysis revealed significant effects of the meteorological conditions across the years studied, the genetic characteristics of the cultivars, and the applied biological amendments on maize grain yield.

Deteriorating agrometeorological conditions contributed to a progressive decline in plant productivity. The mean grain yield across all cultivars decreased by 2.25 Mg ha⁻¹ over the study period (from 7.31 to 5.06 Mg ha⁻¹) (Table 3). The primary factors limiting yield in the final year were precipitation deficits during the critical growth phase (May), combined with elevated mean air temperatures, which intensified abiotic stress on the plants.

The cultivar with the highest FAO number (260) displayed the greatest stability and yield potential, with a three-year mean grain yield significantly higher than those of the other cultivars. Cultivars with shorter growing seasons produced significantly lower yields, confirming that, under monoculture conditions, cultivars with greater biological potential make more effective use of available environmental resources.

Cultivar responses to the prevailing conditions varied across the years studied. In 2022, the longest-season cultivar (FAO 260) and cv. FAO 230 yielded similarly. In 2023, grain yield increased significantly with increasing length of the growing season. Under the drought conditions of 2024, cv. FAO 200 and FAO 230 produced similar yields.

All applied biological amendments (Neosol, Bactim Gleba, UGmax) resulted in a significant increase in grain yield compared with the untreated control. UGmax proved the most effective, increasing mean grain yield by approximately 14% compared with the control. The statistically significant differences among the products confirmed that the applied biological amendments improved plant performance under monoculture conditions.

The year × biological amendment interaction indicates that the efficacy of the amendments depended on the meteorological conditions during the growing season. Only in 2022 were significant differences observed in the yield-promoting effects of each product. In 2023 and 2024, yields obtained after the application of Neosol and Bactim Gleba were similar and significantly higher than the control. UGmax exhibited a stabilising effect on production, consistently producing the highest yields in every year (

Table 4. Grain yield of maize (Mg ha⁻¹) grown in monoculture by year, cultivar, and biological amendment.

	Year
Cultivar	2022	2023	2024	Mean
FAO 200	6.69 B	6.02 C	4.83 B	5.84 C
FAO 230	7.54 A	6.57 B	4.85 B	6.32 B
FAO 260	7.72 A	6.89 A	5.49 A	6.70 A
Mean	7.31 a	6.49 b	5.06 c
Biological amendment	2022	2023	2024
Control	6.78 D	6.13 C	4.76 C	5.90 D
Neosol	7.19 C	6.38 B	5.02 B	6.20 C
Bactim Gleba	7.44 B	6.58 B	5.06 B	6.39 B
UGmax	7.84 A	6.86 A	5.42 A	6.73 A

Means in rows followed by different lowercase letters (a, b, c) differ significantly at p ≤ 0.05 (for years). Means in columns followed by different uppercase letters (A, B, C, D) differ significantly at p ≤ 0.05 (for cultivars, biological amendments, year × cultivar interactions, and year × biological amendment interactions).

).

Analysis of the results demonstrated a significant effect of cultivar and year-to-year weather conditions on the thousand-grain weight (TGW) of maize (

Table 5. TGW of maize (g) grown in monoculture by year, cultivar, and biological amendment.

	Year
Cultivar	2022	2023	2024	Mean
FAO 200	266.06 a	249.19 b	241.54 c	252.26 C
FAO 230	297.22 a	286.82 b	274.44 c	286.16 B
FAO 260	306.63 a	309.67 ab	284.54 b	300.28 A
Mean	289.97 a	281.89 a	266.84 b
Biological amendment	2022	2023	2024
Control	268.2 a	255.3 b	242.0 c	255.16 D
Neosol	273.6 b	284.4 a	268.9 b	275.65 C
Bactim Gleba	299.8 a	289.9 a	264.4 b	284.70 B
UGmax	318.3 a	298.0 b	292.0 b	302.77 A
Biological amendment	Cultivar
	FAO 200	FAO 230	FAO 260
Control	237.16 B	260.84 C	267.48 D
Neosol	243.63 A	284.09 B	299.23 C
Bactim Gleba	257.82 A	288.29 B	307.99 B
UGmax	270.44 A	311.43 A	326.43 A

Means in rows followed by different lowercase letters (a, b, c) differ significantly at p ≤ 0.05 (for years, year × cultivar interactions, and year × biological amendment interactions). Means in columns followed by different uppercase letters (A, B, C, D) differ significantly at p ≤ 0.05 (for cultivars, biological amendments, and year × biological amendment interactions).

). The highest three-year mean TGW was recorded for cv. FAO 260 (300.28 g), a value significantly greater than those of cv. FAO 230 and FAO 200. The cultivar with the shortest growing season produced the lowest grain mass (252.26 g).

Weather conditions during the 2022 growing season favoured the accumulation of assimilates in the grain, resulting in the highest mean TGW (289.97 g). In subsequent years, a decline in this parameter was observed, although it was statistically significant only in 2024.

The application of biological amendments affected maize TGW. All of the tested products increased grain mass compared with the untreated control (255.16 g). The most beneficial effect was observed for UGmax, which increased mean TGW by approximately 19% compared with the control. Significant positive effects were also recorded for Bactim Gleba (284.70 g) and Neosol (275.65 g).

Analysis of the interaction between biological amendments and years showed that only in the control treatment did TGW decrease significantly from year to year (by 5% and 6%, respectively). The weather conditions prevailing in 2023 favoured the highest TGW in plants treated with Neosol and Bactim Gleba. UGmax contributed to maintaining TGW at a relatively stable level; after three years of monoculture (and under the least favourable climatic conditions), TGW did not decline markedly compared with 2023.

Cultivar responses to the applied biological amendments varied. The cultivar with the shortest growing season (FAO 200) produced grains of similar weight across all amendments, with values significantly higher than in the control. In cv. FAO 230, the TGW obtained with Bactim Gleba (288.29 g) and Neosol (284.09 g) was statistically similar, yet significantly lower than after the application of UGmax. In the longest-season cultivar, each amendment significantly differentiated the TGW, with values ranging from 299.23 g (Neosol) to 326.43 g (UGmax). In all maturity groups, the lowest TGW values were recorded in the control treatments.

Based on the presented data, both the year studied and the cultivars exerted a significant influence on the number of kernel rows per ear in maize. The overall mean number of kernel rows across the experiment ranged from 13.32 to 14.81. The most favourable year for ear structure development was 2022, when the highest mean number of kernel rows (14.64) was recorded. This value was statistically significantly higher than those achieved in the other years. In 2023 and 2024, a decline in the number of kernel rows was observed, although the difference between these two years was not statistically significant. This indicates that conditions in 2023 and 2024 were less conducive to ear development than those in 2022.

Analysis of the cultivar means revealed clear genetic differences. The later-maturing cultivars (FAO 230 and FAO 260) developed significantly higher numbers of kernel rows per ear (14.12 and 14.17, respectively). The difference between these two cultivars was not statistically significant (

Table 6. Number of kernel rows per ear in continuous maize by year and cultivar.

	Year
Cultivar	2022	2023	2024	Mean
FAO 200	14.49	13.58	13.32	13.79 B
FAO 230	14.81	13.83	13.72	14.12 A
FAO 260	14.63	13.92	13.97	14.17 A
Mean	14.64 a	13.77 b	13.67 b

Means in rows followed by different lowercase letters (a, b) differ significantly at p ≤ 0.05 (for years). Means in columns followed by different uppercase letters (A, B) differ significantly at p ≤ 0.05 (for cultivars).

).

The number of kernels per row was significantly influenced by all examined factors: the weather conditions during the years studied, the cultivars, and the biological amendments (

Table 7. Number of kernels per row of maize grown in monoculture by year, cultivar, and biological amendment.

	Year
Cultivar	2022	2023	2024	Mean
FAO 200	35.94	31.32	28.12	31.79 B
FAO 230	36.19	31.30	28.56	32.02 B
FAO 260	39.32	37.60	32.66	36.53 A
Mean	37.147 a	33.406 ab	29.78 b
Biological amendment	2022	2023	2024
Control	34.11 a	33.64 a	26.60 b	31.45 C
Neosol	37.29 a	33.38 b	30.02 c	33.56 B
Bactim Gleba	38.43 a	34.04 b	31.31 c	34.59 A
UGmax	38.77 a	32.56 b	31.20 b	34.17 AB
Biological amendment		Cultivar
	FAO 200	FAO 230	FAO 260
Control	30.30 B	30.02 B	34.03 B
Neosol	31.98 A	32.06 A	36.65 A
Bactim Gleba	33.93 A	32.54 A	37.31 A
UGmax	30.96 B	33.44 A	38.12 A

Means in rows followed by different lowercase letters (a, b, c) differ significantly at p ≤ 0.05 (for years, year × biological amendment interactions and cultivar × biological amendment interactions). Means in columns followed by different uppercase letters (A, B, C) differ significantly at p ≤ 0.05 (for cultivars and biological amendments).

). The highest mean number of kernels per row was recorded in 2022 (37.15), a value significantly greater than that observed in 2024 (29.78). The year 2023 represented an intermediate period and did not differ statistically from the other years. The later-maturing FAO 260 cultivar produced the highest number of kernels per row (mean 36.53), a statistically significant difference compared with cv. FAO 200 and FAO 230 (approximately 31–32).

All applied biological amendments had a beneficial effect on kernel number compared with the control (31.45). The most favourable results were obtained with Bactim Gleba (34.59) and UGmax (34.17). Neosol also significantly improved this trait compared with the control, although it was less effective than Bactim Gleba.

Statistical analysis revealed significant interactions between the biological amendments and year, as well as between the cultivars and biological amendments. In the untreated control, a drastic reduction in kernel number occurred in 2024 compared with the previous years. Products such as Bactim Gleba and Neosol helped maintain higher yield-structure parameters even in less-favourable years, although a downward trend was evident across all combinations. The analysis also showed that cultivar responses to biological amendments varied depending on maturity group. In the earliest cultivar (FAO 200), Neosol (31.98) and Bactim Gleba (33.93) significantly improved kernel number compared with the control. For the cultivars with longer growing seasons (FAO 230 and FAO 260), all tested amendments produced values significantly higher than the control. The highest absolute value was recorded in FAO 260 following application of UGmax (38.12). Across all maturity groups, the lowest kernel numbers per row were observed in the control.

The total number of kernels per ear showed trends similar to those observed for kernels per row (

Table 8. Number of kernels per ear of maize grown in monoculture by year, cultivar, and biological amendment.

	Year
Cultivar	2022	2023	2024	Mean
FAO 200	520.79	424.75	375.07	440.20 B
FAO 230	535.76	432.54	391.92	453.41 B
FAO 260	575.09	523.07	455.06	517.74 A
Mean	543.88 a	460.12 b	407.35 b
Biological amendment	2022	2023	2024
Control	494.33 a	452.50 b	359.47 c	435.43 C
Neosol	540.39 a	460.03 b	407.31 c	469.24 B
Bactim Gleba	568.29 a	476.07 b	436.89 c	493.75 A
UGmax	572.51 a	451.88 b	425.72 b	483.37 AB

Means in rows followed by different lowercase letters (a, b, c) differ significantly at p ≤ 0.05 (for years and year × biological amendment interactions). Means in columns followed by different uppercase letters (A, B, C) differ significantly at p ≤ 0.05 (for cultivars and biological amendments).

). The year 2022 was clearly the most favourable for kernel set. In 2023 and 2024, significant reductions were recorded—approximately 15% and 25%, respectively, compared with 2022. Cv. FAO 260 produced an average of 517.74 kernels per ear, a value significantly higher (by approximately 14–17%) than those of the other cultivars. Biological amendments significantly increased the number of kernels per ear. Bactim Gleba (493.75) was the most effective, while Neosol (469.24) also produced a significant improvement compared with the control but was less effective than Bactim Gleba. The untreated control had the lowest kernel number (435.43). Interaction analysis indicates that in the most challenging year, 2024, the use of biological amendments, particularly Bactim Gleba and UGmax, resulted in approximately 18–21% more kernels per ear compared with the control, suggesting their role in mitigating environmental stress in maize monoculture.

The Wright’s path analysis conducted for the period 2022–2024 revealed varying contributions of individual components to the formation of kernel number per ear and final kernel yield. Over the study period, the number of kernels per ear was determined by the number of kernel rows and the number of kernels per row (R² = 0.99). The number of kernels per row exerted the strongest direct effect (0.818), whereas the number of kernel rows had a significant but weaker direct influence (0.352). For kernel yield, the analysis showed that the number of kernels per ear was the primary determinant, displaying a strong direct effect (0.634). Thousand-grain weight (TGW), although significantly correlated with yield, had a stronger indirect rather than direct effect on the yield through kernel number (0.385 vs. 0.274). In 2022, a year characterised by a high coefficient of determination for yield (R² = 0.80), TGW played a dominant role. Its direct effect on yield (0.827) was considerably stronger than that of kernel number per ear (0.094), which acted mainly indirectly. In 2023, the structure of yield formation changed markedly. TGW almost entirely determined yield through a direct effect (0.975), whereas the influence of kernel number per ear was negative and operated solely through indirect mechanisms. This suggests that, in this season, factors affecting grain filling were of paramount importance. In 2024, the most balanced contribution from both components was observed. Both kernel number per ear (0.537) and TGW (0.314) exhibited statistically significant direct effects, with kernel number remaining the principal yield determinant (

Table 9. Path coefficients for maize yield components in 2022–2024.

Year	Outcome Variable	Causal Variable	Total Correlation	Direct Effect	Indirect Effect via Variable:
					Value	Variable Name
2022–2024	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.644 * 0.943 *	0.352 * 0.818 *	0.292 0.126	Number of kernels per row Number of rows
	Grain yield (R2 = 0.68)	TGW Number of kernels per ear	0.659 * 0.800 *	0.274 * 0.634 *	0.385 0.166	Number of kernels per row TGW
2022	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.483 * 0.870 *	0.491 * 0.875 *	−0.008 −0.04	Number of kernels per row Number of rows
	Grain yield (R2 = 0.80)	TGW Number of kernels per ear	0.894 * 0.686 *	0.827 * 0.094	0.067 0.593	Number of kernels per row TGW
2023	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.361 * 0.940 *	0.340 * 0.932 *	0.022 0.008	Number of kernels per row Number of rows
	Grain yield (R2 = 0.88)	TGW Number of kernels per ear	0.933 * 0.356 *	0.975 * −0.092	−0.042 0.488	Number of kernels per ear TGW
2024	Number of kernels per ear (R2 = 0.99)	Number of kernels per row	0.570 * 0.894 *	0.449 * 0.829 *	0.120 0.065	Number of kernels per row Number of rows
	Grain yield (R2 = 0.61)	TGW Number of kernels per ear	0.669 * 0.745 *	0.314 * 0.537 *	0.365 0.208	Number of kernels per ear TGW

*—significant relationships at p ≤ 0.05.

).

The application of different biological amendments modified both the strength and the nature of relationships among maize yield components (

Table 10. Path coefficients for maize yield components across biological amendments applied.

Biological Amendment	Outcome Variable	Causal Variable	Total Correlation	Direct Effect	Indirect Effect via Variable
					Value	Variable Name
Control	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.539 * 0.899 *	0.437 * 0.846 *	0.101 0.05	Number of kernels per row Number of rows
	Grain yield (R2 = 0.82)	TGW Number of kernels per ear	0.733 * 0.885 *	0.236 0.723 *	0.497 0.162	Number of kernels per ear TGW
Neosol	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.650 * 0.395 *	0.375 * 0.807 *	0.275 0.128	Number of kernels per row Number of rows
	Grain yield (R2 = 0.62)	TGW Number of kernels per ear	0.446 * 0.765 *	0.208 0.694 *	0.238 0.071	Number of kernels per ear TGW
Bactim gleba	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.707 * 0.946 *	0.360 0.787 *	0.346 0.159	Number of kernels per row Number of rows
	Grain yield (R2 = 0.68)	TGW Number of kernels per ear	0.715 * 0.793 *	0.325 0.570 *	0.390 0.222	Number of kernels per ear TGW
UGmax	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.692 * 0.969 *	0.282 0.830 *	0.410 0.139	Number of kernels per row Number of rows
	Grain yield (R2 = 0.63)	TGW Number of kernels per ear	0.661 * 0.772 *	0.257 0.599 *	0.403 0.173	Number of kernels per ear TGW

*—significant relationships at p ≤ 0.05.

). Combining correlation analysis with path analysis made it possible to determine how individual products altered the yield-forming structure relative to the control treatment. In the control, a very strong correlation was observed between the number of kernels per ear and the grain yield (r = 0.885). This association arose predominantly from the direct effect of that trait (0.723). Thousand-grain weight (TGW) was significantly correlated with yield (r = 0.733); however, path analysis indicated that its direct effect was not significant (0.236). The high correlation stemmed mainly from an indirect effect mediated through the number of kernels per ear (0.497). Biological amendments altered these relationships. Following Neosol application, the number of kernels per ear retained its dominant role as the main yield determinant (direct effect 0.694). In this treatment, TGW exhibited the lowest correlation with yield among all variants studied (r = 0.446), and its influence on yield was distributed between direct and indirect pathways. With Bactim Gleba, the highest coefficient of determination for yield was recorded (R² = 0.88), indicating that the adopted model explained the yield structure very precisely. Although the correlation between TGW and yield remained high and significant (r = 0.715), path analysis showed that the number of kernels per ear continued to act as a stable direct yield determinant (0.570), whereas TGW influenced yield largely indirectly by increasing the number of kernels per ear (indirect effect 0.390). As with the other amendments, UGmax application resulted in the number of kernels per ear strongly and directly shaping yield (0.599). In this variant, however, a clear indirect effect of TGW on final yield was evident (0.403), suggesting that the product enhances interdependence between kernel size and kernel number per ear. Analysis of the ear structural parameters revealed that, across all variants receiving biological amendments (Neosol, Bactim Gleba, UGmax), the primary direct determinant of total kernel number per ear was the number of kernels per row. These products strengthened the correlation between the number of rows and total kernel number compared with the control, indicating improved exploitation of the ear’s biological potential following soil biostimulation.

Path analysis revealed differences in the yield-forming strategies among maize cultivars differing in maturity group (

Table 11. Path coefficients for maize yield components according to cultivar earliness.

Cultivar	Outcome Variable	Causal Variable	Total Correlation	Direct Effect	Indirect Effect via Variable:
					Value	Variable Name
FAO 2000	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.724 * 0.949 *	0.353 * 0.781 *	0.371 * 0.168	Number of kernels per row Number of rows
	Grain yield (R2 = 0.76)	TGW Number of kernels per ear	0.785 * 0.765 *	0.513 * 0.466 *	0.271 0.300	Number of kernels per ear TGW
FAO 230	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.734 * 0.950 *	0.356 * 0.776 *	0.378 0.173	Number of kernels per row Number of rows
	Grain yield (R2 = 0.71)	TGW Number of kernels per ear	0.618 * 0.814 *	0.261 * 0.677 *	0.357 0.138	Number of kernels per ear TGW
FAO 260	Number of kernels per ear (R2 = 0.99)	Number of rows per ear Number of kernels per row	0.508 * 0.889 *	0.456 * 0.862 *	0.052 0.028	Number of kernels per row Number of rows
	Grain yield (R2 = 0.67)	TGW Number of kernels per ear	0.569 * 0.813 *	0.149 0.727 *	0.420 * 0.086	Number of kernels per ear TGW

*—significant relationships at p ≤ 0.05.

). In all FAO groups, the model for the variable ‘number of kernels per ear’ showed an almost perfect fit (R² = 0.99). In early-maturing cultivars (FAO 200), grain yield was determined to a nearly equal degree by thousand-grain weight (TGW, direct effect 0.513) and by the number of kernels per ear (0.466). This points to a balanced contribution from both yield components. In this maturity group, the number of kernels per ear depended predominantly on the number of kernels per row (direct effect 0.781), whereas the influence of the number of rows was exerted largely through indirect pathways. In the medium–early-maturing group (FAO 230), the relationships shifted towards ear structure. Here, the number of kernels per ear exerted a markedly stronger direct effect on yield (0.677) than did the TGW (0.261). Kernel weight played a secondary role, while the primary yield determinant was the number of kernels per row (direct effect 0.776). The most pronounced differences appeared in late-maturing cultivars (FAO 260). Path analysis indicated that, in this group, the sole key determinant of yield was the number of kernels per ear (very strong direct effect 0.727). It should be emphasised that TGW showed no significant direct influence in this FAO group (0.149); its correlation with yield (0.569) arose almost entirely from an indirect effect mediated through the number of kernels per ear (0.420). This indicates that, in late cultivars grown in monoculture, ear architecture, particularly the number of kernels per row, rather than individual kernel weight, governs yield success.

The economic analysis showed that all tested amendments generated a positive economic return (

Table 12. Economic efficiency of using biological amendments in maize cultivation (means for 2022–2024).

Biological Amendment	Yield Increase vs. Control (Mg·ha−1)	Additional Revenue (USD·ha−1)	Amendment Cost (USD·ha−1)	Net Economic Gain (USD·ha−1)	Profitability Index (BCR)	Break-Even Grain Price (USD·Mg−1)
Neosol	0.3	58	44–53	+14 to +5	1.31–1.09	146–176
Bactim Gleba	0.49	95	22–27	+73 to +68	4.3–3.5	45–55
UG Max	0.83	160	24–29	+136 to +131	6.5–5.5	29–35

). The highest additional revenue was obtained following the application of UGmax (160 USD·ha⁻¹), followed by Bactim Gleba (95 USD·ha⁻¹). UGmax also achieved the highest profitability ratio (5.50–6.60). In contrast, Neosol exhibited the lowest economic efficiency due to its relatively high cost in relation to the yield increase obtained, with a profitability ratio close to the break-even threshold (1.09–1.31).

4. Discussion

The United Nations Sustainable Development Goals highlight the crucial role of the agricultural sector in ensuring food security, protecting natural resources, and supporting adaptation to climate change. In countries with a strong reliance on crop production, such as Poland, technologies that stabilise yields under increasing environmental pressure, particularly drought and soil degradation in monoculture systems, are gaining strategic importance. Numerous long-term experiments have demonstrated that such systems lead to reduced yield stability and the deterioration of soil quality [29,30,31,32,33]. The study focused primarily on the productive response of plants (yield and yield components) as an indicator of soil–plant system functioning. The combined use of ANOVA and path analysis made it possible to assess the influence of the meteorological conditions, cultivar earliness, and the action of the biological amendments on the mechanism of yield formation through its individual components. ANOVA results obtained in this study showed that, under monoculture conditions, maize grain yield was strongly determined by meteorological conditions. Path coefficient analysis, in turn, enabled identification of the direct and indirect effects of individual traits on the final yield variable [16,34], providing deeper insight into the complex relationships between yield components and maize grain yield in monoculture.

Our findings demonstrated that the traits with the strongest direct influence on kernel yield were the number of kernels per ear and the thousand-grain weight (TGW). These results are consistent with observations by Huda et al. [35] and Matin et al. [16], who indicated that these parameters constitute the primary determinants of maize productivity. Path analysis further revealed that the relative importance of yield components varied depending on weather conditions. In the year with favourable weather (2022), TGW played the dominant role in shaping yield, while the effect of the number of kernels per ear was indirect. This indicates that kernel-filling efficiency was the decisive factor. Under water stress, the structure of relationships shifted: TGW exhibited a very strong direct effect on yield, whereas the direct effect of the number of kernels per ear was negative and insignificant. This suggests the presence of compensatory mechanisms, whereby plants partially offset reductions in kernel number by increasing individual kernel weight. At the same time, the application of biological amendments significantly mitigated the decline in monoculture productivity, which is particularly important from a sustainability perspective, as it contributes to yield stabilisation. The use of UGmax and Neosol led to a substantial reduction in yield losses in years characterised by unfavourable hydrothermal conditions. The observed response may be associated with soil-functioning improvement mechanisms described in the literature [3,21].

This is supported by numerous studies showing that organic or biological amendments can effectively mitigate the negative effects of monoculture [36,37,38,39]. UGmax produced the highest mean yield across the three-year cycle (6.73 Mg·ha⁻¹), exceeding the control by more than 14%, while Bactim Gleba and Neosol increased yields to 6.39 and 6.20 Mg·ha⁻¹, respectively. These production effects align with earlier research on UGmax in maize and potato, where yield increases and improved plant health were reported [40,41,42,43]. At the same time, Piotrowska et al. [44] demonstrated that this product increased soil microbiological activity and the content of plant-available nutrients. Novák et al. [45] noted that Neosol affected the physical properties of heavy soils as well as the physiological status of plants. The stabilisation of yield components observed in the present study is consistent with such findings. The plant response recorded in the present study may also be associated with mechanisms described in the literature, including those related to water management and soil aggregate structure [5,6,8].

From a physio-biological perspective, the plant responses observed here are consistent with studies describing the role of conditioners and organic amendments in shaping the rhizosphere microbiome and nutrient uptake efficiency [11,12,34].

ANOVA results further indicated that cv. FAO 260 achieved significantly higher mean yield (6.70 Mg·ha⁻¹) than FAO 230 (6.32 Mg·ha⁻¹) and FAO 200 (5.84 Mg·ha⁻¹). Cv. FAO 260 also exhibited the highest TGW (300.3 g), suggesting a greater capacity for assimilate accumulation even under declining water availability. This is consistent with findings by Berzsenyi et al. [30] and Księżak et al. [31], who reported that, in long-term monoculture, high-yielding cultivars utilise available resources more effectively than earlier-maturing forms.

Path analysis revealed clear differences in yield-forming strategies between maturity groups and biological amendment variants, confirming the usefulness of this method in agronomic research [14,24]. The differences observed between maturity groups correspond to the findings of Matin et al. [16] and Munawar et al. [46], who emphasised the variability of yield-formation mechanisms depending on genotype and environmental conditions. In cv. FAO 200, yield was shaped almost equally by TGW (direct effect 0.513) and the number of kernels per ear (0.466), indicating a balanced compensatory strategy. Similar mechanisms were reported by Rafiq et al. [15] and Soumya and Kamatar [47] who noted that early cultivars often compensate limitations in one component through another. In cv. FAO 230, the dominant role was played by the number of kernels per plant, with the primary determinant being the number of kernels per row. A similar pattern was described by Pavlov et al. [48] and Matin et al. [16], highlighting the importance of generative development in shaping yield in medium–early-maturing cultivars. The higher productivity of late-maturing cultivars, confirmed by ANOVA, resulted from the strong direct effect of the number of kernels per ear as the main yield-forming trait. This relationship aligns with the findings of Matin et al. [16] and Huda et al. [35], who emphasised that kernel number is the trait with the highest potential to determine yield in maize. This suggests that, under monoculture and water stress, cultivars which mature later in the season maintain productivity primarily by preserving kernel number, which may be crucial for yield stability under climate change.

Path analysis results for biological amendments are particularly noteworthy from a sustainability perspective. In the control, yield was mainly affected by the number of kernels per ear (direct effect 0.723), while TGW acted predominantly through indirect pathways (0.497), indicating limited kernel-filling capacity under declining soil fertility. After Bactim Gleba application, the number of kernels per ear remained the main and stable determinant of yield (0.570), with strong indirect effects of TGW (0.390).

For UGmax, the interdependence between grain number and grain weight was additionally strengthened (indirect effect of TGW = 0.403), which may indicate more favourable conditions for grain filling. This observation is consistent with the mechanisms described by Piotrowska et al. [44], Keteku et al. [49], and Odugbenro et al. [50], who reported links between biological amendments, soil structural properties, and nutrient uptake efficiency. From an agronomic perspective, this may imply the potential to stabilise yields without increasing fertilisation intensity, which is particularly relevant in the context of sustainable agriculture.

From a practical perspective, the profitability of using biological amendments is of key importance, as it depends on the balance between their cost and the resulting yield increase. Our results confirm that microbial-based products (UGmax and Bactim Gleba) are cost-effective solutions for mitigating the negative effects of maize monoculture. For Neosol, the break-even price is higher, indicating that its use is more sensitive to market volatility.

Detailed economic analyses that would account for the full production costs and the long-term effects of biological amendments on maize yield were beyond the scope of the present study. However, incorporating such analyses in future research is essential to fully assess the feasibility of implementing these solutions at the farm scale. The need for continued investigation in this direction is further justified by the inherent limitations of the current work. The most important of these include the relatively short, three-year duration of the experiment, which may not fully capture the long-term consequences of biostimulant use in maize monoculture. Moreover, this study focused on plant production parameters without direct monitoring of physicochemical and biological changes occurring in the soil. It should also be noted that the obtained results may have been influenced by specific local conditions and interannual meteorological variability.

5. Conclusions

Maize productivity under continuous monoculture was strongly determined by interannual meteorological variability. Increasing water deficit and heat stress during the generative development stage led to a progressive decline in yield, primarily through a reduction in the number of kernels per ear. This trait proved to be the key structural determinant of yield stability under abiotic stress.

The integration of ANOVA with Wright’s path analysis provided new insight into the biological hierarchy of yield formation in monoculture systems. The results demonstrated that maturity class determines the mechanism of productivity maintenance: early cultivars relied on a balanced contribution of kernel number and kernel weight, whereas medium- and late-maturing cultivars depended mainly on preserving kernel number, with thousand-kernel weight acting largely indirectly. These findings highlight the central role of reproductive structure preservation in sustaining yield under simplified cropping systems.

From a practical perspective, the targeted use of biological amendments combined with appropriate cultivar selection (in the present conditions particularly FAO 260) may improve yield stability in simplified cropping systems. Microbial-based products showed the most stable economic performance, with UGmax providing the highest profitability due to both yield increase and a favourable cost–benefit relation, whereas other amendments were more dependent on market conditions.

The study was limited by the relatively short three-year monoculture period and by the lack of direct measurements of the soil physical, chemical and biological properties. Therefore, long-term experiments integrating crop productivity with direct soil quality indicators are necessary to fully explain the mechanisms underlying yield stabilisation in monoculture maize systems.