Genotype–Environment Interaction in Shaping the Agronomic Performance of Silage Maize Varieties Cultivated in Organic Farming Systems

Abstract

Organic production systems impose strong environmental constraints on silage maize, yet the relative contributions of genotype, environment and their interaction (G × E) to key performance traits remain insufficiently resolved. This study evaluated six maize cultivars across 11 organically managed environments (location × year combinations) in Poland, assessing weed infestation, plant height, fresh matter yield, dry matter content and dry matter yield. Genotype × environment interaction was explicitly analyzed using AMMI-based models, and cultivar adaptability and stability were evaluated using complementary indices. Environmental effects consistently dominated all traits, explaining 78–91% of total variation, while G × E interactions, though smaller, were significant and altered cultivar rankings. Weed infestation ranged widely across environments, from below 10% to over 90%, and was almost entirely environment-driven. Yield-related traits followed a strong precipitation gradient, with Pawłowice and Śrem showing the highest biomass potential. SM Perseus produced the greatest dry matter yields (13.53 t·ha⁻¹), whereas SM Mieszko combined high dry matter content (37.73%) with outstanding stability. Mega-environment analysis identified distinct adaptive niches, confirming that no genotype performed consistently best across all conditions. These findings close a key knowledge gap regarding cultivar performance under organic management and demonstrate the necessity of multi-environment evaluation that integrates performance, stability and adaptability analyses to support site-specific cultivar recommendations that enhance biomass productivity and silage quality in ecologically managed maize systems.

1. Introduction

Maize (Zea mays L.) is one of the most widely cultivated cereal crops globally, serving as a fundamental component of food, feed and industrial systems. Under the temperate climatic conditions of Central Europe, including Poland, maize cultivation is dominated by hybrids belonging to earlier FAO maturity groups (approximately FAO 100–400), which are primarily grown for silage and grain production. These maturity classes are particularly suited to regions with a shorter growing season and higher interannual climatic variability. Worldwide demand continues to rise in response to population growth and expanding livestock sectors, with maize now representing the largest volume among traded cereals [1]. In temperate regions, maize is especially valued as a high-energy forage crop, and its suitability for silage production depends on achieving both high biomass productivity and favorable ensiling characteristics. The quality of maize silage—including its dry matter content, fermentation profile and aerobic stability—is strongly influenced by genotype-specific traits and environmental factors [2]. Against this background, increasing attention has been directed toward maize cultivation in organic farming systems, where environmental constraints and management limitations are more pronounced.

Increasing interest in sustainable agriculture and soil health has stimulated the expansion of organic maize cultivation. Long-term studies indicate that organic cropping systems can maintain comparable nutrient-use efficiency and yield potential when environmental conditions and management practices are well aligned [3]. However, compared with conventional systems, organic maize production is more vulnerable to nutrient fluctuations, climatic variability and particularly weed pressure, making hybrid selection a critical determinant of agronomic performance.

Cultivar differences in plant architecture, maturity, biomass allocation, physiological activity and senescence patterns substantially influence silage potential. Of particular importance is the stay-green trait, which is associated with prolonged photosynthetic activity during grain filling, improved nitrogen-use efficiency and enhanced biomass stability across environments. These relationships have been well documented in the work of Szulc et al. [4,5], who demonstrated that stay-green hybrids maintain functional leaf area longer and are often more resilient in variable conditions, including those characteristics of organic systems where nutrient buffering is reduced. These genotype-specific traits, however, do not act in isolation, but interact strongly with environmental conditions, giving rise to pronounced genotype × environment (G × E) interactions.

Multi-environment performance in maize is strongly shaped by genotype × environment interaction (G × E). Numerous studies using AMMI, GGE biplot and BLUP-based methodologies have shown that environment typically explains the largest portion of variation in maize yield, while G × E contributes significantly to cultivar ranking and stability [6,7,8,9,10]. Recent research highlights substantial phenotypic plasticity among maize hybrids, reflecting distinct reaction norms across environmental gradients [11]. Such variability underscores the necessity of multi-environment trials (METs) for identifying hybrids with broad or specific adaptation, particularly under organic conditions, where stress factors (nutrient scarcity, rainfall variability, weed interference) often interact.

Silage maize involves additional complexity because forage quality depends not only on biomass yield but also on dry matter content, structural composition, carbohydrate partitioning and harvest timing—all of which are strongly influenced by both genotype and environment [2,12]. Despite the extensive literature on G × E in maize, relatively few studies have focused on silage maize cultivated under certified organic systems, and even fewer have analyzed both agronomic traits and forage quality parameters (fresh matter yield, dry matter content, dry matter yield) across multiple contrasting environments.

Several studies have examined silage maize performance under organic or low-input management, focusing mainly on yield level and basic agronomic traits. For example, Cox and Cherney [13] demonstrated strong site- and year-dependent variability of maize biomass yield during the transition to organic farming systems, while Áldott-Sipos et al. [14] compared silage yield and performance of maize genotypes under organic and conventional conditions, highlighting differences in genotype response between production systems.

Despite these contributions, most available studies are limited to a small number of environments and do not explicitly address genotype × environment interactions using multi-environment, multivariate approaches. In particular, comprehensive AMMI-based evaluations of silage maize cultivated under certified organic conditions remain scarce, representing a key knowledge gap in cultivar assessment for organic forage systems.

Given the growing interest in sustainable forage systems and the need for improved hybrid recommendation for organic production, there is a clear gap in understanding how genotype, environment and G × E interactions jointly determine silage maize performance under organic management, especially in temperate regions subject to variable climatic stress.

By integrating multi-environment field trials with AMMI-based analysis under certified organic farming conditions, this study contributes to a deeper understanding of genotype × environment interactions in silage maize. The results provide a robust basis for identifying cultivars with broad or specific adaptation and support evidence-based selection of maize genotypes for organic silage production systems. Therefore, the aim of this study was to evaluate the effects of environment (location and year), genotype and genotype × environment interaction on weed infestation, plant height, fresh matter yield, dry matter content and dry matter yield of silage maize cultivated under certified organic farming conditions across four contrasting locations over three years. The findings provide new insights into cultivar performance and stability in organic systems, while contributing to the broader understanding of G × E in silage maize production.

2. Materials and Methods

2.1. Experimental Sites and Environmental Conditions

The study was conducted over three growing seasons (2022–2024) under certified organic farming conditions and involved silage maize (Zea mays L.). Field experiments were carried out at four experimental stations of the Research Centre for Cultivar Testing (COBORU): Krzyżewo, Pawłowice, Przecław and Śrem. These locations represent major maize-producing regions in Poland and differ in soil properties, climatic conditions, and cropping histories, thus providing a broad environmental gradient for cultivar assessment (

Table 1. Characteristics of Experimental Stations/Departments for Evaluation of Cultivar in Poland.

Location	GPS Coordinates	Altitude [m a.s.l.]	Voivodeship (Province)	Total Area [ha]	Experimental Crop Rotation [ha]	Main Soil Suitability Complexes	Dominant Soil Quality Classes
Krzyżewo (SDOO 1)	53.017 N, 22.767 E	135	Podlaskie	209.0	160.0	good wheat, very good rye, good rye, weak rye	II–VI
Pawłowice (SDOO)	50.467 N, 18.483 E	240	Silesian	156.0	152.3	faulty wheat, very good rye, good rye	IIIa–V
Przecław (SDOO)	50.183 N, 21.483 E	185	Subcarpathian	181.5	106.1	very good wheat, good wheat, faulty wheat, very good rye, good rye, weak rye	II–VI
Śrem (ZDOO 2)	52.083 N, 17.033 E	76	Greater Poland	389.0	104.3	very good rye, good rye	II–VI

¹ SDOO—Experimental Station for Cultivar name Evaluation; ² ZDOO—Experimental Department for Cultivar name Evaluation; Both are part of the Research Centre for Cultivar Testing (COBORU).

). The sowing density of maize varieties, regardless of the research location, was at the level of 8.88 pcs/m² (88.800 pcs per 1 ha).

Across the study period, all six cultivars were sown annually at all four locations, yielding an initial dataset of 12 site-years (4 locations × 3 years). Although the trial was initiated in Krzyżewo in 2023, this site-year was excluded from analyses. Severe wild boar damage caused substantial plant losses and prevented the maintenance of adequate and uniform stand density, making reliable data collection impossible. Following COBORU field experimentation protocols, the Krzyżewo 2023 trial was therefore discontinued and removed from the dataset.

Site-year specific agronomic characteristics—including soil pH, preceding crop, sowing and emergence dates, silking dates, and seasonal weather indicators (mean temperature and total precipitation)—are summarised in

Table 2. Summary of soil, cropping, and weather conditions of the experiments.

Year	Soil pH (KCl)	Preceding Crop	Sowing Date	Emergence Date	Silking Date	Total Precipitation (April–November) [mm]	Mean Air Temperature (April–November) [°C]
Krzyżewo
2022	6.1	WW	28 April	15 May	27 July	406	14.4
2023	–	–	–	–	–	–	–
2024	5.9	WT	6 May	17 May	13 July	263	17.2
Pawłowice
2022	6.6	WW	29 April	11 May	24 July	487	14.9
2023	6.4	FP	25 April	15 May	26 July	515	15.7
2024	6.6	WR	26 April	13 May	6 July	425	17.9
Przecław
2022	6.1	OAT	2 May	13 May	17 July	291	15.2
2023	6.1	WW	9 May	27 May	23 July	522	15.8
2024	7.0	WW	18 April	3 May	18 July	371	18.0
Śrem
2022	6.1	WW	2 May	10 May	13 July	325	16.8
2023	6.2	WW	2 May	16 May	16 July	314	17.1
2024	6.0	WW	30 April	7 May	6 July	516	18.3

WW—winter wheat; FP—field pea; WT—winter triticale; WR—winter rye; OAT—oats. Organic fertilization was applied each year in all locations.

. To further characterise climatic variability, detailed monthly precipitation totals and monthly mean air temperatures for all locations and years are provided in the Supplementary Material (Tables S1 and S2), based on data from weather stations situated within or near each experimental site.

2.2. Plant Material

Six maize (Zea mays L.) cultivars intended for silage production were evaluated: Farmonitz, Geoxx, SM Grot, SM Mieszko, SM Perseus, and SM Varsovia.

The cultivars differed in maturity class, biomass allocation patterns, plant architecture and stay-green behavior, representing the diversity of hybrids commonly used in temperate organic systems (

Table 3. Characteristics of the maize cultivars.

Cultivar	FAO Maturity Group	Hybrid Type	Stay-Green Expression 1
Farmoritz	260	SC (single-cross)	+++
Geoxx	240	SC (single-cross)	++
SM Grot	220–230	TC (three-way cross)	+
SM Mieszko	230	TC (three-way cross)	++
SM Perseus	250	TC (three-way cross)	++
SM Varsovia	250	TC (three-way cross)	+++

¹ Stay-green expression: + low, ++ medium, +++ high.

).

2.3. Experimental Design and Crop Management

Experiments were arranged in a randomized complete block design (RCBD) with four replications per location and year. Plot size was 16.5 m². Sowing density and agronomic management followed the COBORU methodology for silage maize evaluation.

All sites were managed strictly according to organic farming regulations. No mineral fertilizers, herbicides or chemical pesticides were applied. Organic fertilization was applied at all experimental sites in accordance with certified organic farming regulations. Fertilization consisted of farmyard manure applied at rates corresponding to local agronomic recommendations for silage maize under organic management. Application timing and rates were adjusted to site-specific conditions and crop rotation history.

Weed control was conducted in accordance with standard organic farming practices. Mechanical inter-row cultivation was applied at early vegetative stages of maize development, whereas manual weed removal within rows was performed as needed to limit competitive pressure. All quantitative weed assessments were conducted prior to the main weed control operations, corresponding approximately to the BBCH 59 growth stage. Subsequent weed control, when required, relied exclusively on manual interventions during later growth stages to avoid crop damage.

2.4. Weed Infestation Assessment

Weed infestation was assessed at the 6–9 leaf stage (BBCH 16–19). A visual estimation was performed on each plot to determine the percentage of soil surface covered by weeds, considering all weed species combined (total weed cover, %). To ensure consistency and minimize evaluator bias, a single trained evaluator conducted all assessments within each location across all study years. The entire plot was evaluated as a unit, without subsampling (into weed species).

2.5. Plant Height Measurement

Plant height was measured at tasseling (BBCH 55–59). In each plot, measurements were taken on 20 consecutive plants from the central rows, starting from the third plant in the row, by measuring the distance from the soil surface to the top of the tassel (male inflorescence). Plants from the central rows were selected to minimize border effects and ensure representative measurements of plant height. The mean plant height per plot was used for further analysis.

2.6. Biomass Yield and Dry Matter Determination

2.6.1. Fresh Matter Yield Determination

Fresh matter yield (FMY) was assessed at harvest at the appropriate silage stage (BBCH 85). The entire aboveground biomass from each 16.5 m² plot was cut and weighed with 0.1 kg accuracy. The mass was converted to t·ha⁻¹, accounting for the effective plot area.

2.6.2. Total Dry Matter Content Determination

Immediately after harvest, 2.5 kg of chopped whole-plant biomass was collected from each plot. Samples from the four replications within each cultivar × location × year combination were thoroughly mixed to create one composite sample of 7.5 kg. Pooling of samples was applied to obtain a single, representative estimate of dry matter content for each cultivar × location × year combination, consistent with standard procedures used in official silage maize cultivar evaluation. From this mixed material, two analytical subsamples of 700 g were taken. Drying was carried out in two steps for a total of 20 h: the first 6 h at ≤50 °C, followed by 14 h at 105 °C (±5 °C; reflecting between-batch oven calibration tolerance). After drying, the samples were left in a closed dryer for 2 h to equilibrate moisture content, and then weighed with an accuracy of 1 g. The dry matter content (DM) was expressed as the percentage of dry matter in the fresh biomass.

2.6.3. Dry Matter Yield Determination

Dry matter yield (DMY) was calculated as:

$$DMY=FMY\times {\displaystyle \frac{DM\%}{100}}$$

where: FMY—fresh matter yield (t·ha⁻¹), DM (%)—dry matter content.

2.7. Statistical Analysis

All statistical analyses were performed in R software (version 4.3.2) [15]. For each trait, statistical analyses were conducted separately within each environment, defined as a unique combination of year and location. Within each environment, a one-way analysis of variance (ANOVA) was performed with cultivar as the fixed effect, based on four plot-level replications. When the ANOVA indicated significant cultivar effects, mean comparisons were performed using Tukey’s honestly significant difference (HSD) test at α = 0.05. Results are presented as means ± standard error (SE), and significant differences among cultivars within a given environment are indicated by different letters (Tables S3–S12). The assumptions of the ANOVA model were evaluated by inspecting residual diagnostics, including normality of residuals and homogeneity of variances.

In addition, genotype × environment interaction patterns across all environments were explored using the additive main effects and multiplicative interaction (AMMI) model, based on cultivar means within environments.

Let $y_{ij}$ denote the mean value of a trait for the i-th cultivar ($i=1,\dots ,I$) in the j-th environment ($j=1,\dots ,J$). An environment was defined as a unique combination of location and year (e.g., “22Kr” = Krzyżewo 2022). Trait means were modelled using the Additive Main Effects and Multiplicative Interaction (AMMI) model with κ significant interaction principal components (IPCs), following the formulation of Gauch [7]. The number of significant interaction principal components (κ) was determined using F-tests for interaction principal components, as recommended by Gauch [7]:

$$y_{ij}=\mu +v_i+e_j+{\displaystyle \sum }_{k=1}^{\kappa }{\lambda }_k{\gamma }_{ik}{\delta }_{jk}+{\epsilon }_{ij},$$

where:

−

$\mu$ is the grand mean;

−

$v_i$ and $e_j$ are the main effects of cultivar and environment;

−

${\lambda }_k\ge 0$ is the eigenvalue of the k-th IPC;

−

${\gamma }_{ik}$ and ${\delta }_{jk}$ are cultivar and environment IPC scores;

−

${\epsilon }_{ij}$ is the residual error.

The number of significant IPCs (κ) was identified using F-tests for interaction principal components, as recommended by Gauch [7].

2.7.1. Stability Analysis

Cultivar stability was evaluated using the Weighted Average of Absolute Scores (WAAS) index, as proposed by Olivoto et al. [8]:

$$WAA{S}_i={\displaystyle \frac{{{\displaystyle \sum }}_{k=1}^{\kappa }\left|{\theta }_{k}IP{C}_{ik}\right|}{{{\displaystyle \sum }}_{k=1}^{\kappa }{\theta }_k}}$$

where

−

${\mathrm{IPC}}_{ik}$ is the score of the i-th cultivar on the k-th IPC;

−

${\theta }_k$ is the variance explained by the k-th IPC.

Cultivars were ranked based on AMMI-adjusted means and WAAS values, and both components were integrated into a Genotype Selection Index (GSI), combining performance and stability [8].

2.7.2. Adaptability Analysis

Adaptability across environments was assessed using the WTOP2 index:

$$WTOP{2}_i={\displaystyle \frac{n_i}{J}}$$

where $n_i$—number of environments in which the i-th cultivar ranked among the top wo, based on AMMIκ-derived fitted values.

In addition, mega-environments were identified by grouping environments according to their best-performing cultivar, following the classification approach of Yan et al. [6].

3. Results

3.1. Weed Infestation

3.1.1. Environmental Variation in Weed Pressure

Mean weed infestation levels across the 11 environments (22Kr, 22Pa, etc.) are presented in Table S3 (Supplementary Material). Across the 2022–2024 seasons, the lowest weed pressure was recorded in Przecław in 2022, where average infestation reached only 5–6%. In contrast, the highest values were observed in Śrem in 2024, with weed cover exceeding 90% in some cultivars. These pronounced differences corresponded closely with the weather conditions (Tables S1 and S2). In Śrem, the 2024 season was characterized by both the highest mean air temperatures (18.3 °C) and the greatest total precipitation (516 mm), creating conditions highly favorable for the emergence and proliferation of multiple weed species.

Conversely, the exceptionally low infestation levels in Przecław in 2022 coincided with reduced precipitation during June and July (21 mm and 85 mm, respectively), which likely suppressed weed competitiveness during their critical growth phase. These results emphasize the pivotal role of microclimatic variability in organic production systems, where the absence of herbicides makes weed pressure highly dependent on environmental fluctuations.

3.1.2. Genotype-Environment Interaction Structure (AMMI) for Weed Infestation

Weed pressure varied widely among the 11 environments, and this variability was strongly dominated by environmental effects. The AMMI ANOVA (

Table 4. Analysis of variance of the main effects and interactions for weed infestation.

Source of Variation	D.F.	S.S.	M.S.	F	Variability Explained [%]
Environment	10	30,670.1	3067.01	147.6538 ***	96.5
Cultivar	5	88.8	17.5	0.8547	0.3
Interactions	50	1038.6	20.77		3.3
PC1	14	792.9	56.64		76.3
PC2	12	196.3	16.36		18.9
Residuals	24	49.3	2.06		4.8

D.F. = degrees of freedom; S.S. = sum of squares; M.S. = mean square; IPCi = i-th interaction principal component; *** p < 0.001.

) showed that the environment accounted for 96.5% of the total sum of squares, whereas cultivar effects explained only 0.3%, and the G × E interaction 3.3%. Such an overwhelming environmental contribution confirms that weed infestation in organic maize is governed primarily by local meteorological and soil conditions rather than genetic differences.

The AMMI2 model (Table 4; Figure 1) revealed significant cultivar × environment interactions. The first two interaction principal components (IPC1 and IPC2) jointly accounted for 95.2% of the total G × E variation, with IPC1 explaining 76.3% and IPC2 an additional 18.9%. In the resulting biplot (Figure 1), environments 22Śr, 23Śr, and 24Śr, together with cultivars Farmonitz, Geoxx, and SM Varsovia, showed the strongest contributions to the interaction structure.

This pattern demonstrates that in Śrem, cultivar performance diverged more strongly than in the remaining locations—likely reflecting local soil characteristics (rye-complex soils of high agricultural quality) and the pronounced variability in precipitation recorded for this site (Table S1). Such environment-specific differentiation aligns with previously reported findings for multi-environment maize trials, where G × E interactions are particularly pronounced under heterogeneous climatic conditions.

3.1.3. Stability and Adaptability of Cultivars (WAAS, GSI and WTOP2) for Weed Infestation

Table 5. Mean weed infestation (%) and cultivar adaptability and stability assessed using WAAS, GSI, and WTOP2 indices.

i	Cultivar	Mean [%]	WAAS	GSI	WTOP2
1	Farmonitz	31.13 [1]	0.3884 [4]	5	0.45
2	Geoxx	31.81 [2]	0.5416 [5]	7	0.27
3	SM Grot	33.74 [5]	0.1622 [2]	7	0.54
4	SM Mieszko	34.41 [6]	0.2112 [3]	9	0.18
5	SM Perseus	33.64 [4]	0.0911 [1]	5	0
6	SM Varsovia	33.41 [3]	0.6033 [6]	9	0.54

Numbers in square brackets (e.g., [1], [2], etc.) indicate cultivar rankings within a given trait, based on AMMI-adjusted means. WTOP2 represents the proportion of environments in which a given cultivar ranked among the top two performers, based on AMMI-fitted values.

reports the AMMI-adjusted Cultivar means, values of the weighted average of absolute scores (WAAS), and the values of the genotype selection index (GSI). In the third column AMMI-adjusted Cultivar means are reported. The cultivar SM Perseus exhibited the highest stability of weed infestation across environments (WAAS = 0.091), whereas SM Varsovia showed the greatest instability (WAAS = 0.603). It should be emphasized that a low WAAS value does not necessarily reflect low infestation levels, but rather the predictability and consistency of cultivar performance across contrasting environments—a trait of major importance in organic farming, where opportunities to regulate weed pressure are inherently limited.

The GSI index further supported this pattern, identifying SM Perseus as the most desirable cultivar (GSI = 5) and SM Varsovia and SM Mieszko as the least favorable (GSI = 9).

In the WTOP2 analysis (Table 5), the highest adaptability was recorded for SM Grot and SM Varsovia. SM Grot ranked among the two least-infested cultivars every year in Pawłowice and Przecław (22Pa, 22Pr, 23Pa, 23Pr, 24Pa, 24Pr). This broad adaptability is particularly relevant in organic maize production, where reduced weed infestation directly translates into lower competition for water and nutrients, thereby improving the crop’s capacity to maintain biomass accumulation under non-chemical management.

3.1.4. Identification of Mega-Environments for Weed Infestation

The analysis of AMMI interaction means allowed the identification of four distinct mega-environments, as illustrated in Figure 2 of the manuscript. The first mega-environment comprised the sites 22Pa, 22Pr, 23Pa, 23Pr, 24Pa and 24 Pr where the cultivar SM Grot exhibited the greatest competitive ability against weeds and consequently the lowest infestation levels. The second group of environments, including 22Kr, 23Śr, and 24Kr, favored the cultivar Geoxx, which experienced the least weed pressure under these conditions. The third mega-environment, consisting of 22Śr proved particularly advantageous for SM Varsovia. The last mega-environment, consisting of 24Śr proved particularly advantageous for Farmonitz.

Beyond weed pressure, plant architectural traits such as height play an important role in biomass accumulation and competitive ability under organic conditions, and were therefore analyzed next.

3.2. Plant Height

3.2.1. Environmental Variation in Plant Height

Mean plant height values across the 11 environments are presented in Table S4. Plant height varied considerably among environments, with the environmental effect being dominant, similarly to the pattern observed for weed infestation. The tallest plants were recorded in Śrem in 2022, where mean height exceeded 280 cm for most cultivars. This was associated with favorable growing conditions, including adequate rainfall during May–July and elevated temperatures during grain filling (Table S2). In contrast, the shortest plants occurred in Przecław in 2024 (approximately 190–200 cm), coinciding with the lowest early-season precipitation recorded across all sites (Table S1). In organic maize production, plant height reflects the crop’s competitive ability against weeds and contributes directly to biomass accumulation for silage production.

3.2.2. Genotype-Environment Interaction Structure (AMMI) for Plant Height

The AMMI analysis (

Table 6. Analysis of variance of the main effects and interactions for plant height.

Source of Variation	D.F.	S.S.	M.S.	F	Variability Explained [%]
Environment	10	42.797	4279.7	26.9214 ***	78.4
Cultivar	5	3857	771.4	4.8526 **	7
Interactions	50	7948	159.0		14.6
PC1	14	3726	266.2		46.9
PC2	12	2438	203.1		30.7
PC3	10	1471	147.1		18.5
Residuals	24	314	22.4		3.9

D.F. = degrees of freedom; S.S. = sum of squares; M.S. = mean square; IPCi = i-th interaction principal component; *** p < 0.001; ** p < 0.01.

; Figure 3) revealed significant G × E interactions underlying variation in plant height. The first three interaction principal components were significant and jointly explained 96.1% of the interaction sum of squares, with IPC1, IPC2 and IPC3 accounting for 46.9%, 30.7% and 18.5%, respectively. The first two IPCs alone captured 77.6% of the interaction variance, justifying the use of a two-dimensional AMMI2 biplot (Figure 3), environments 22Śr and 23Śr were positioned furthest from the origin, demonstrating that cultivar height differed more strongly in these settings compared with other locations. Cultivars Farmonitz and Geoxx exhibited the largest interaction scores, reflecting increased sensitivity to environmental fluctuations.

3.2.3. Stability and Adaptability of Cultivars (WAAS, GSI, WTOP2) for Plant Height

The WAAS stability index (

Table 7. Mean plant height (cm) and cultivar adaptability and stability assessed using WAAS, GSI, and WTOP2 indices.

i	Cultivar	Mean [cm]	WAAS	GSI	WTOP2
1	Farmonitz	233.61 [6]	0.4506 [6]	12	0.09
2	Geoxx	248.82 [3]	0.3030 [2]	5	0.18
3	SM Grot	236.69 [5]	0.3378 [3]	8	0.18
4	SM Mieszko	246.19 [4]	0.3453 [4]	8	0.27
5	SM Perseus	253.57 [1]	0.3919 [5]	6	0.55
6	SM Varsovia	252.93 [2]	0.2776 [1]	3	0.73

Numbers in square brackets (e.g., [1], [2], etc.) indicate cultivar rankings within a given trait, based on AMMI-adjusted means. WTOP2 represents the proportion of environments in which a given cultivar ranked among the top two performers, based on AMMI-fitted values.

) indicated that SM Varsovia exhibited the highest stability in terms of plant height, whereas Farmonitz showed the lowest stability. The low WAAS value for SM Varsovia reflects high predictability of plant height across contrasting environments, which is particularly relevant under organic conditions where plant stature contributes to natural weed suppression. The GSI ranking further identified SM Varsovia as the most favorable cultivar, while Farmonitz ranked lowest, reflecting pronounced environmental sensitivity. The WTOP2 index highlighted the superior adaptability of SM Varsovia, which ranked among the two tallest cultivars in the highest proportion of environments, indicating efficient utilization of water and thermal resources under variable climatic conditions.

3.2.4. Identification of Mega-Environments for Plant Height

Based on AMMI-adjusted means (Figure 4), three mega-environments were identified. In Śrem (22Śr, 23Śr, 24Śr), SM Perseus and SM Mieszko achieved the greatest plant height, indicating superior adaptation to environments characterized by high temperatures and abundant rainfall. In Przecław (22Pr, 23Pr), Farmonitz performed best, whereas in Krzyżewo and Pawłowice (22Kr, 22Pa, 24Pa), Geoxx and SM Grot were consistently superior. The delineation of these mega-environments highlights the strength of G × E interactions and the necessity of environment-specific cultivar recommendations.

3.3. Fresh Matter Yield

3.3.1. Environmental Variation in Fresh Matter Yield

Mean fresh matter yields for the 11 environments are presented in Table S5. Yield showed substantial variation across environments, reflecting strong modulation by local weather and soil conditions. The highest yields were recorded in Pawłowice in 2022 and 2023 (39–46 t·ha⁻¹), coinciding with the highest seasonal precipitation totals (487–515 mm) and moderate temperatures that supported vigorous biomass accumulation (Tables S1 and S2). In contrast, the lowest yields were observed in Krzyżewo 2022 and Przecław 2024 (approximately 22–28 t·ha⁻¹), which experienced the lowest rainfall levels (263–291 mm) and early-season water deficits.

In organic maize systems, where nutrient and water availability rely solely on soil biological processes, fresh matter yield is a key indicator of cultivar suitability for silage. High biomass accumulation is directly linked to silage energy value and overall feed efficiency in dairy production.

3.3.2. Genotype-Environment Interaction Structure (AMMI) for Fresh Matter Yield

The AMMI analysis (

Table 8. Analysis of variance of the main effects and interactions for fresh matter yield.

Source of Variation	D.F.	S.S.	M.S.	F	Variability Explained [%]
Environment	10	368.682	36.868	45.6963 ***	87.1
Cultivar	5	14.569	2914	3.6116 **	3.4
Interactions	50	40.340	807		9.5
PC1	14	23.050	1646		57.1
PC2	12	6328	527		15.7
Residuals	24	10.962	457		27.2

D.F. = degrees of freedom; S.S. = sum of squares; M.S. = mean square; IPCi = i-th interaction principal component; *** p < 0.001; ** p < 0.01.

; Figure 5) indicated significant G × E interaction for fresh matter yield. The first interaction principal component (IPC1) explained 57.1% of the interaction sum of squares and was highly significant, whereas IPC2 accounted for an additional 15.7% but was not statistically significant. Together, the first two IPCs captured 72.8% of the interaction variance, providing an adequate basis for visualizing the interaction structure in a two-dimensional biplot.

The environments Pawłowice 2022, Pawłowice 2023, and Śrem 2022 contributed most strongly to the interaction, consistent with their high absolute yield levels. Among cultivars, SM Perseus and SM Varsovia exhibited the largest interaction scores, reflecting greater sensitivity to environmental variability.

The AMMI biplot (Figure 5) displayed a clear separation of environments along a precipitation gradient: high-rainfall environments (Pa, Śr) clustered at positive IPC1 scores, whereas drier environments (Kr, Pr, especially 2024) clustered at negative IPC1 values. This indicates that the primary driver of the G × E structure was variation in water availability, a key yield-limiting factor in organic biomass production.

3.3.3. Stability and Adaptability of Cultivars (WAAS, GSI, WTOP2) for Fresh Matter Yield

Table 9. Mean fresh matter yield (t·ha⁻¹) and cultivar adaptability and stability assessed using WAAS, GSI, and WTOP2 indices.

i	Cultivar	Mean [t·ha−1]	WAAS	GSI	WTOP2
1	Farmonitz	35.28 [4]	0.2835 [3]	7	0.27
2	Geoxx	34.93 [5]	0.1719 [1]	6	0
3	SM Grot	33.71 [6]	0.5460 [5]	11	0.18
4	SM Mieszko	35.54 [3]	0.1773 [2]	5	0.18
5	SM Perseus	38.40 [1]	0.4514 [4]	5	0.64
6	SM Varsovia	36.90 [2]	0.5474 [6]	8	0.73

Numbers in square brackets (e.g., [1], [2], etc.) indicate cultivar rankings within a given trait, based on AMMI-adjusted means. WTOP2 represents the proportion of environments in which a given cultivar ranked among the top two performers, based on AMMI-fitted values.

presents the WAAS stability scores and the GSI rankings for the evaluated cultivars for fresh matter yield. SM Mieszko showed the highest stability of fresh matter yield (lowest WAAS), whereas SM Varsovia and SM Grot were the least stable cultivars. Yield stability is particularly valuable in organic systems, where nutrient and water limitations occur more frequently than in conventional agriculture.

GSI rankings identified SM Perseusz and Farmonitz as the cultivars achieving the best combination of high yield and stability, whereas SM Varsovia ranked lowest.

The WTOP2 index highlighted SM Perseusz, Farmonitz, and Geoxx as the most adaptable cultivars. SM Perseus ranked among the top two yielding cultivars in 7 out of 11 environments, demonstrating broad adaptability, especially under wetter conditions.

3.3.4. Identification of Mega-Environments for Fresh Matter Yield

Based on AMMI-adjusted means (Figure 6), three mega-environments were identified for fresh matter yield. High-yielding environments (22Pa, 23Pa, 22Śr) favored SM Perseus and SM Mieszko. Moderately moist environments (23Śr, 24Pa, 24Śr) favored Geoxx and SM Grot, whereas dry environments (22Kr, 24Kr, 24Pr) were most favorable for Farmonitz. These patterns demonstrate distinct cultivar niches and highlight the role of water availability in determining optimal cultivar placement

3.4. Dry Matter Content

Dry matter content is a key quality parameter of silage maize, influencing fermentation efficiency, energy concentration, and aerobic stability. In the present study, dry matter content exhibited marked variability across environments and cultivars, reflecting the combined influence of weather patterns, soil conditions, and genotype-specific maturation dynamics.

3.4.1. Environmental Variation in Dry Matter Content

Mean dry matter contents for each environment and cultivar are presented in Supplementary Table S6. Across the 11 environments, dry matter content ranged from values typical of relatively immature silage (slightly above 30%) to levels exceeding 45%, indicating an advanced stage of whole-plant maturation. In 2022, the highest dry matter contents were recorded in Śrem, where several cultivars exceeded 45%, while Pawłowice and Krzyżewo generally showed lower values, around 30–35%. A similar pattern was observed in 2024, with Śrem and Przecław again reaching the upper range of dry matter content, whereas Krzyżewo and Pawłowice tended to remain closer to the lower or intermediate range. These gradients were consistent with the meteorological context: environments with higher temperatures and sufficient precipitation during the grain-filling period (e.g., Śrem 2022 and 2024) favored accelerated dry matter accumulation, while cooler or more water-limited sites resulted in slower progression towards silage maturity, as also reflected by the monthly rainfall and temperature profiles (Tables S1 and S2).

3.4.2. Genotype-Environment Interaction Structure (AMMI) for Dry Matter Content

The AMMI ANOVA for dry matter content (

Table 10. Analysis of variance of the main effects and interactions for dry matter content.

Source of Variation	D.F.	S.S.	M.S.	F	Variability Explained [%]
Environment	10	986.77	98.677	27.9941 ***	80.6
Cultivar	5	61.02	12.204	3.4622 **	5
Interactions	50	176.25	3.525		14.4
PC1	14	86.01	6.144		48.8
PC2	12	58.89	4.907		33.4
Residuals	24	31.35	1.306		17.8

D.F. = degrees of freedom; S.S. = sum of squares; M.S. = mean square; IPCi = i-th interaction principal component; *** p < 0.001; ** p < 0.01.

) confirmed that both environment and cultivar effects were statistically significant, with the environment accounting for 80.6% and cultivars for 5.0% of the total sum of squares. The cultivar × environment interaction explained 14.4% of the total variability, indicating a non-negligible, yet secondary, contribution of G × E to the overall variation in dry matter content. The first two interaction principal components (IPC1 and IPC2) jointly explained 82.2% of the interaction sum of squares, which justifies the use of a two-dimensional AMMI2 biplot to visualize the interaction structure.

In the AMMI2 biplot (Figure 7), environments 24Pa, 23Pr, and 22Śr were located furthest from the origin, which indicates a strong discriminatory capacity with respect to cultivar performance and a substantial contribution to the G × E interaction. Among cultivars, Farmonitz, SM Grot, and SM Varsovia showed the largest IPC scores, reflecting pronounced responsiveness of their dry matter content to environmental variation. In contrast, cultivars with IPC scores closer to zero displayed more consistent dry matter content across environments, suggesting a more stable maturation trajectory. The combination of relatively large interaction variance and the clear separation of several environments in the biplot underlines that dry matter content in organic silage maize is shaped not only by the main environmental gradient but also by specific genotype–environment combinations.

3.4.3. Stability and Adaptability of Cultivars (WAAS, GSI, WTOP2) for Dry Matter Content

WAAS and GSI results for dry matter content are summarized in

Table 11. Mean dry matter content (%) and cultivar adaptability and stability assessed using WAAS, GSI, and WTOP2 indices.

i	Cultivar	Mean [%]	WAAS	GSI	WTOP2
1	Farmonitz	35.52 [2]	0.5132 [6]	8	0.27
2	Geoxx	37.50 [3]	0.1246 [4]	7	0.45
3	SM Grot	37.40 [4]	0.3809 [2]	6	0.55
4	SM Mieszko	37.73 [1]	0.3252 [1]	2	0.45
5	SM Perseus	35.37 [6]	0.3101 [3]	9	0.09
6	SM Varsovia	36.23 [5]	0.4013 [5]	10	0.18

Numbers in square brackets (e.g., [1], [2], etc.) indicate cultivar rankings within a given trait, based on AMMI-adjusted means. WTOP2 represents the proportion of environments in which a given cultivar ranked among the top two performers, based on AMMI-fitted values.

. AMMI-adjusted means indicated that SM Mieszko had the highest average dry matter content (37.73%), followed by SM Grot and Geoxx, whereas SM Perseus showed the lowest mean value (35.37%). From a silage perspective, this places SM Mieszko in the most favorable range in terms of ensuring sufficiently high dry matter without reaching levels that would compromise compaction.

In terms of stability, SM Mieszko also exhibited the lowest WAAS value, confirming that its dry matter content was not only high, but also relatively consistent across environments. Conversely, SM Perseus and SM Varsovia showed higher WAAS scores, indicating greater fluctuation in dry matter content in response to environmental variation. The GSI rankings combined these two dimensions—mean performance and stability—and again identified SM Mieszko as the most desirable cultivar (lowest GSI), whereas SM Varsovia and SM Perseus received the highest GSI values, reflecting less favorable combinations of mean dry matter content and stability.

The WTOP2 index highlighted Geoxx, SM Grot, and SM Mieszko as the most adaptable cultivars in terms of dry matter content, as they most frequently ranked among the top two cultivars within individual environments. Geoxx reached the top group in environments such as 22Kr, 22Pa, 24Kr, 24Pa, and 24Pr, SM Grot in 22Kr, 22Pr, 22Śr, 23Pa, 23Pr, and 23Śr, while SM Mieszko achieved this status in 22Śr, 23Pa, 23Pr, 23Śr, and 24Śr. This pattern indicates that several cultivars can achieve desirable dry matter content under contrasting site conditions, which is important for organic producers seeking to balance harvest timing flexibility with silage quality.

3.4.4. Identification of Mega-Environments for Dry Matter Content

Based on the AMMI-adjusted cultivar × environment means (Figure 8) and the environment-wise comparisons reported in Supplementary Table S11, several mega-environments could be distinguished for dry matter content. The first group comprised environments 22Śr, 23Pa, 23Śr, and 24Śr, where SM Mieszko consistently achieved the highest or near-highest dry matter content, underscoring its particular suitability for conditions prevailing in Śrem and in Pawłowice in 2023. A second mega-environment included 22Kr and 24Kr, where Geoxx outperformed the remaining cultivars in terms of dry matter content, suggesting a good fit of this hybrid to the cooler and slightly less favorable conditions of Krzyżewo. A third group, represented by 22Pa and 24Pr, favored Farmonitz, which obtained the highest dry matter content in these environments.

Additional environments formed smaller clusters or single-environment niches in which specific cultivars showed locally superior performance, but without forming larger, clearly defined mega-environment complexes.

3.5. Dry Matter Yield

Dry matter yield integrates both fresh matter production and dry matter content and is therefore the most direct agronomic indicator of the amount of preservable energy harvested per unit area.

3.5.1. Environmental Variation in Dry Matter Yield

Dry matter yield showed pronounced variation across the 11 environments. The AMMI ANOVA (

Table 12. Analysis of variance of the main effects and interactions for dry matter yield.

Source of Variation	D.F.	S.S.	M.S.	F	Variability Explained [%]
Environment	10	65.019	6501.9	57.3059 ***	90.8
Cultivar	5	920	184	1.6214	1.3
Interactions	50	5673	113.5		7.9
PC1	14	2201	157.2		38.8
PC2	12	1409	117.4		24.8
Residuals	24	2064	86		36.4

D.F. = degrees of freedom; S.S. = sum of squares; M.S. = mean square; IPCi = i-th interaction principal component; *** p < 0.001.

) confirmed that the environment was by far the dominant source of variation, explaining 90.8% of the total sum of squares, while cultivar effects accounted for only 1.3%. The cultivar × environment interaction represented 7.9% of the total variation, and the first two IPCs together explained 63.6% of the interaction sum of squares. Mean dry matter yields for each environment and cultivar are provided in Supplementary Table S7. The highest yields were typically recorded in Pawłowice and Śrem during the 2022 and 2023 seasons, where relatively high precipitation and moderate temperatures supported vigorous vegetative growth and grain filling. In contrast, environments such as Krzyżewo 2024 and, to a lesser extent, Przecław 2024 showed reduced dry matter yields, which coincided with less favorable water supply or higher thermal stress during critical developmental stages (Tables S1 and S2). These patterns underline the central role of water balance and temperature regime in shaping biomass accumulation in organic maize.

3.5.2. Genotype-Environment Interaction Structure (AMMI) for Dry Matter Yield

In Table 12, the results of ANOVA from the AMMI2 model for dry matter yield are reported. The main effect of the environments was highly significant. The sum of squares for environments explained 90.8% of the total variation, while the sum of squares for cultivars explained 1.3% of the total variation and the sum of squares for interaction explained 7.9%. The first two interaction principal components (IPCs) jointly explained 63.6% of the whole effect it had on the variation in the dry matter yield.

The AMMI2 biplot for dry matter yield (Figure 9) illustrates the interaction structure between the six cultivars and 11 environments. Environments 24Kr and 23Pr were located furthest from the origin, indicating a strong ability to discriminate among cultivars and a major contribution to the G × E interaction. Among cultivars, Farmonitz, Geoxx, and SM Varsovia exhibited the largest IPC scores, reflecting greater sensitivity of their dry matter yield to environmental variation. In contrast, cultivars with IPC coordinates closer to zero demonstrated more stable yield responses across the tested sites.

3.5.3. Stability and Adaptability of Cultivars (WAAS, GSI, WTOP2) for Dry Matter Yield

Table 13. Mean dry matter yield (t·ha⁻¹) and cultivar adaptability and stability assessed using WAAS, GSI, and WTOP2 indices.

i	Cultivar	Mean [t·ha−1]	WAAS	GSI	WTOP2
1	Farmonitz	12.54 [6]	0.5730 [6]	12	0.18
2	Geoxx	13.12 [4]	0.3583 [4]	8	0.18
3	SM Grot	12.66 [5]	0.2152 [2]	7	0
4	SM Mieszko	13.44 [2]	0.0895 [1]	3	0.36
5	SM Perseus	13.53 [1]	0.2594 [3]	4	0.63
6	SM Varsovia	13.29 [3]	0.5122 [5]	8	0.63

Numbers in square brackets (e.g., [1], [2], etc.) indicate cultivar rankings within a given trait, based on AMMI-adjusted means. WTOP2 represents the proportion of environments in which a given cultivar ranked among the top two performers, based on AMMI-fitted values.

presents the AMMI-adjusted means, WAAS stability scores, and GSI rankings for dry matter yield. SM Perseus achieved the highest average dry matter yield (13.53 t·ha⁻¹), followed closely by SM Varsovia and SM Mieszko, whereas Farmonitz had the lowest mean yield (12.54 t·ha⁻¹). From a production standpoint, SM Perseus therefore combines an advantageous yield level with high relevance for organic silage systems.

3.5.4. Identification of Mega-Environments for Dry Matter Yield

The AMMI-based adaptive response patterns for dry matter yield (Figure 10) allowed the delineation of several mega-environments. The first mega-environment included environments 22Kr, 24Pr, and 24Śr, where SM Perseus displayed the highest dry matter yield, indicating a strong adaptation of this cultivar to conditions combining relatively high yield potential with variable water supply. A second mega-environment was formed by 22Pa, 23Pr, and 23Śr, where SM Varsovia achieved the highest yields, suggesting a particular fit of this cultivar to the more productive environments in Pawłowice and Śrem.

A third mega-environment comprised 22Pr and 22Pa, where Farmonitz was the best-performing cultivar, while a fourth mega-environment (22Śr, 24Kr) favored Geoxx. Finally, 24Pa constituted a separate mega-environment in which SM Mieszko was the highest-yielding cultivar. This configuration confirms that no single cultivar combines absolute yield superiority across all sites. Instead, distinct adaptation niches emerge, with SM Perseus and SM Varsovia dominating in many high-yielding environments, Farmonitz and Geoxx being better suited to a subset of specific conditions, and SM Mieszko combining high yield with outstanding stability.

4. Discussion

4.1. Environmental Effects as the Main Driver of Variation and the Role of G × E

Across the three-year multi-environment trial, environmental conditions associated with location and year emerged as the dominant source of variation in all measured traits, including weed infestation, plant height, dry matter content (DM), fresh matter yield (FMY) and dry matter yield (DMY). This pattern aligns with extensive multi-environment trials (METs) in maize, which consistently show that environment explains the majority of variation in yield-related traits, whereas genotype and G × E interaction contribute substantially, but to a lesser extent [9,10,16]. The magnitude of environmental effects observed here reflects marked differences among sites in soil type, fertility levels, temperature regimes and rainfall distribution, consistent with findings from large-scale silage maize analyses conducted in China [12], Europe [14] and Brazil [17].

Organic farming conditions further amplify these environmental contrasts because nutrient availability depends on mineralization dynamics rather than mineral fertilizers, and weed pressure is not mitigated by herbicides. As a result, environmental heterogeneity is more strongly expressed in organic systems, reinforcing the significance of G × E interactions for biomass production. Comparable findings were reported in studies assessing maize performance under organic or low-input conditions, where year-to-year and site-specific effects were much stronger than genotype main effects [18,19]. The present results therefore support the broader conclusion that G × E must be considered a fundamental component of cultivar evaluation when selecting hybrids for low-input and organic systems.

4.2. Weed Infestation as a Critical Constraint Under Organic Management

Among the environmental factors shaping organic maize performance, weed infestation emerged as one of the most critical constraints, warranting separate consideration. Weed infestation varied widely across environments, indicating its strong dependence on local conditions and management history. In our trial, weed cover ranged from very low levels in Przecław to exceptionally high values (>80%) in Śrem in 2024. Such variability is characteristic of organic maize, where the absence of herbicides intensifies weed competition and increases reliance on cultivar traits and mechanical management [13]. High weed pressure consistently reduced FMY and increased yield variability, confirming that weed-infested environments act as more stringent discriminators of genotypic performance. Similar conclusions were drawn by Mandale et al. [18], who observed strong genotype-dependent differences in biomass yield under organic conditions, largely driven by early-season weed competition.

Differences among cultivars in weed cover also indicated that weed suppression capacity is subject to G × E, consistent with studies showing that traits related to competitive ability (early vigor, leaf expansion, canopy architecture) differ across environments [10,20]. Furthermore, evaluations of phenotypic plasticity in maize [11] and recent G × E analyses [21] demonstrate that yield and competition-related traits commonly exhibit cross-over interactions. The present findings therefore highlight the need to include weed-infested environments in METs when selecting maize cultivars suitable for organic systems.

4.3. Plant Height and Architectural Adaptation Across Environments

Plant height exhibited clear differences between locations, with Pawłowice and Śrem supporting the tallest plants and Krzyżewo and Przecław producing shorter canopies. This reflects environmental constraints in heat accumulation, soil fertility and water availability. Height has long been considered an important driver of silage productivity due to its association with biomass accumulation and competitive ability [17]. Our findings are consistent with silage maize studies in Nepal and South America, where the most productive cultivars exceeded 2 m in height [17,22].

From the perspective of G × E, plant height demonstrated relatively stable cultivar rankings across environments. This is in line with AMMI and GGE studies that show developmental and architectural traits often have simpler interaction patterns than yield [10,16]. Given the importance of rapid canopy closure and shading ability in organic farming, stable height expression across years supports the use of height as an auxiliary selection criterion.

4.4. Fresh Matter Yield and Dry Matter Yield as Central Indicators of Biomass Productivity

While plant height reflects structural adaptation, fresh and dry matter yields provide direct measures of biomass productivity and agronomic value for silage production. Fresh matter yield varied markedly among locations, with the highest values recorded in Pawłowice and Śrem and the lowest in Krzyżewo. When integrated with DM content, these results correspond to dry matter yields of approximately 10–15 t·ha⁻¹, which is typical for silage maize in temperate climates grown under organic or reduced-input systems. These yields fall within ranges observed in other regions under comparable farming conditions, such as Brazil [17], Hungary [14] and southern Africa [19]. Although lower than yields achieved under intensive fertilization [12], the values are considered agronomically competitive under organic nutrient supply and weed pressure.

Environmental differentiation of DMY further supports the notion that biomass productivity in organic systems depends heavily on synchrony between environmental conditions, nutrient release and weed dynamics. This was similarly observed in studies analyzing genotype × environment × management (G × E × M) interactions in China [12] which showed that high-yielding genotypes often perform below potential when environmental constraints are not mitigated by external inputs.

4.5. Dry Matter Content and Implications for Silage Quality and Fermentation

Dry matter content was generally within the optimal range for whole-plant maize silage (30–40%), which ensures efficient fermentation, high energy concentration and good aerobic stability [23]. The highest DM values observed in Śrem and Przecław corresponded with warmer and drier late-season conditions, consistent with the environmental determinants of DM described by Zhao et al. [12]. Lower DM values observed in cooler or wetter conditions are typical of delayed maturity or reduced starch deposition.

Because DMY combines FMY and DM content, it is a more integrative measure of silage productivity. Our results fall within the ranges reported in comparable silage maize trials worldwide [17,22], confirming that DM content and maturity dynamics are critical for ensuring both high yield and adequate nutritive value. The observed genotypic differences in DM stability across environments highlight this trait’s relevance for cultivar selection in organic systems.

4.6. Cultivar Stability, G × E Interaction and Implications for Organic Maize Breeding

Integrating the observed responses across all traits highlights the importance of cultivar stability and adaptability when interpreting genotype × environment interactions in organic maize systems. Although environmental effects were dominant, meaningful genotypic differences were evident across traits. Some cultivars displayed broad adaptation, performing consistently across most environments, whereas others showed clear cross-over interactions driven by weed pressure or climatic variability. This pattern is characteristic of maize METs and highlights the complexity of selecting cultivars for environmentally heterogeneous organic systems [6,10,16].

Studies on maize phenotypic plasticity [11] and lodging resistance under varying environments [21] consistently show that high yield potential does not always coincide with stability. Our findings confirm that cultivar selection for organic systems must prioritize not only mean performance but also stability across contrasting environments.

The evaluated cultivars represent genotypes registered and commonly used under Polish organic farming conditions. While this may limit the direct transferability of cultivar-specific results to other regions, the multi-environment framework and AMMI-based analytical approach applied in this study are broadly applicable and can be readily extended to other genotype pools and agroecological zones.

Together, the results demonstrate that genotype–environment interaction is central to shaping agronomic performance in silage maize under organic farming. Multi-environment evaluations conducted directly within organic systems are therefore essential for identifying cultivars that combine competitive early growth, appropriate maturity, stable DM content and high DMY.

5. Conclusions

The results of this multi-environment study demonstrate that environmental conditions are the dominant factor shaping agronomic performance of silage maize under certified organic farming, while genotype × environment (G × E) interactions, although smaller in magnitude, significantly influence cultivar ranking and stability. Weed infestation, plant height, fresh matter yield, dry matter content, and dry matter yield were all strongly environment-dependent, highlighting the high sensitivity of organic maize production to climatic variability and local site conditions.

Despite the prevailing environmental effects, meaningful genotypic differences were identified across traits. Some cultivars exhibited broad adaptation and stable performance across contrasting environments, whereas others showed specific adaptation driven by precipitation patterns or weed pressure. The AMMI-based analysis and mega-environment classification proved effective tools for distinguishing these adaptation patterns and for identifying cultivars suited either to a wide range of organic environments or to specific production niches.

From a practical perspective, the findings emphasize that cultivar selection for organic silage maize should not be based solely on mean yield performance. Instead, stability across environments and predictable responses to variable climatic and weed conditions should be considered key selection criteria. Cultivars combining high dry matter yield with stable dry matter content and competitive early growth are particularly valuable for organic systems, where management options to mitigate stress factors are limited.

For plant breeders and cultivar testing programs, this study confirms the necessity of conducting multi-environment trials directly under organic farming conditions. Integrating yield performance with stability and adaptability analyses provides a robust framework for developing and recommending silage maize cultivars tailored to organic production systems. Overall, the present results support evidence-based, environment-specific cultivar recommendations and contribute to improving the resilience and productivity of organic silage maize farming.