Nutrient Balances and Forage Productivity in Permanent Grasslands Under Different Fertilisation Regimes in Western Poland Conditions

Abstract

Effective nutrient management in grassland ecosystems is essential for maintaining soil nutrient balance and ensuring high forage productivity. A field experiment was conducted between 2022 and 2024 on a permanent dry meadow at the Experimental Station in Poznań-Strzeszyn, western Poland. The trial, established in autumn 2021, was carried out under production conditions on large plots (140 m² each). Plots were assigned to different fertilisation regimes, varying in both type and dosage. The treatments included an unfertilised control, three levels of annual mineral NPK fertilisation (NPK1, NPK2, NPK3), three levels of annually applied farmyard manure (FYM1, FYM2, FYM3), and three levels of mineral and organic fertilisers applied every two years (NPK1/FYM1, NPK2/FYM2, NPK3/FYM3). Throughout the study, botanical composition, annual dry matter yield (DMY), nitrogen (N), phosphorus (P), and potassium (K) content in the plant biomass were assessed. A simplified nutrient balance was calculated based on nutrient input from fertilisers and nutrient output with harvested yield. The average N balance across three years ranged from −12.17 kg N ha⁻¹ in control to +20.6 kg N ha⁻¹ in FYM3. For phosphorus, average balances ranged from −7.2 kg P ha⁻¹ in the control to +9.8 kg P ha⁻¹ in FYM3. In contrast, potassium balances were mostly negative: from −51.7 kg K ha⁻¹ in FYM1 to −7.4 kg K ha⁻¹ in NPK1. The most balanced nutrient budgets were observed under alternate NPK/FYM fertilisation, with moderate surpluses of N and P and a smaller K deficit compared to FYM applied alone. In contrast, inorganic and organic fertilisation applied separately resulted in greater nutrient surpluses or a pronounced potassium deficit. This study emphasises the importance of balanced nutrient management in permanent meadows, showing that moderate fertilisation strategies, such as alternating FYM and mineral NPK, can maintain productivity, and reduce environmental impacts. These findings provide a practical basis for developing sustainable grassland management practices under variable climatic conditions.

1. Introduction

Permanent grasslands are a key component of the European agricultural landscape, covering nearly 38% of agricultural land [1]. Globally, they make up around 67% of all arable land [2]. These areas are used for forage production and are covered with grasses or other fodder crops that are not included in crop rotation for at least five years [3]. Beyond their productive role, permanent grasslands also contribute significantly to environmental protection by reducing soil erosion, supporting biodiversity, and providing a range of ecosystem services [4,5,6,7].

Fertilisation is one of the key factors determining yield and forage quality [8]. In practice, inorganic fertilisers and various forms of organic fertilisers, including manure and slurry, are used on grasslands [9,10]. Both types of fertilisation contribute to increased biomass production from grassland [8]. Organic fertilisation, especially with manure, also promotes the development of leguminous plants such as clover (Trifolium spp.) [11,12,13] improves feed quality and has a positive effect on protein content and fibre digestibility [10,11,14,15]. However, these effects are not always clear. Some studies have not found significant differences in feed quality [16].

Organic fertilisers provide not only major nutrients but also micronutrients, thereby supporting growth and the nutritional value of feed. In addition, the use of organic fertilisers has a positive effect on soil fertility, increases humus content, improves nutrient availability, and mitigates acidification [17]. It can also increase the humic substance content and improve the quality of organic matter [18].

Mineral fertilisers in contrast, offer faster nutrient release and more precise dosing, which allows for closer alignment with plant needs. However, long-term reliance on mineral fertilisers can degrade soil structure and reduce biological activity [19]. Nitrogen-based mineral fertilisers may acidify the soil (pH < 4.5) and negatively impact soil organic carbon accumulation [17]. Their use has also been linked to reduced species diversity, especially among dicotyledonous and leguminous plants.

To address these challenges, integrated fertilisation strategies that combine mineral and organic sources are increasingly adopted. Such approaches help to optimize yield and forage quality while maintaining or even improving soil fertility [20]. They also support soil microbial activity, improve microbial community structure, increase enzymatic processes, and facilitate more efficient nutrient cycling [21].

Nevertheless, excessive fertiliser application can result in economic losses and environmental harm [22]. This makes monitoring fertiliser efficiency, particularly through nutrient balance calculations using simplified net methods, an important management tool. These assessments help fine-tune fertilisation practices and limit nutrient losses or overaccumulation in the soil [23,24]. Nitrogen balance is especially critical, as surpluses can lead to greenhouse gas emissions and nitrate leaching [25], while deficits may impair soil fertility over time.

This study aimed to evaluate how various levels and types of manure and mineral fertilisers affect the nitrogen (N), phosphorus (P), and potassium (K) balance and the productivity of permanent grasslands on mineral soils. The working hypothesis was that alternating between organic and mineral fertilisation would improve nutrient efficiency and lead to more balanced nutrient budgets than relying on either source alone.

2. Materials and Methods

2.1. Experimental Site

The experiment was set up on a three-cut production meadow located at the Experimental Station in Poznań-Strzeszyn (52°28′29″ N, 16°50′59″ E), in western Poland. The area belongs to the Institute of Technology and Life Sciences—PIB (Figure 1).

A field-scale experiment was established in autumn 2021 on a permanent meadow situated on mineral soil. According to the World Soil Classification System [26], this soil is classified as Phaeozem.

2.2. Fertilisation Treatments

Ten experimental plots, each measuring 10 × 14 m (140 m²), were designated as follows: Control, NPK1, NPK2, NPK3, FYM1, FYM2, FYM3, NPK1/FYM1, NPK2/FYM2, and NPK3/FYM3. Nine plots received annual fertilisation using either mineral fertilisers (NPK1–3), farmyard manure (FYM1–3), or alternating applications of mineral fertiliser and manure (NPK/FYM combinations). Each treatment involved one of three fixed rates of nitrogen (N), phosphorus (P), and potassium (K), which remained constant throughout the experiment. The Control plot received no fertilisation. The experimental setup is shown in Figure 2.

The fertilisation levels used were adopted on the basis of previous work on the fertilisation of permanent grassland [27,28], taking into account the need for recommendations on low-input production methods. In addition, when determining manure doses, the nitrogen, phosphorus, and potassium content of the manure was taken into account, adopting the following equivalents for the use of these components from Barszczewski and Szatyłowicz [29] as follows: nitrogen—0.5, phosphorus—0.9, and potassium—0.7.

The macronutrient application rates used in each treatment are shown in

Table 1. Fertilisation treatments and nutrient application rates.

Nutrient Application Rates (kg ha−1)
Macronutrient	Control	Mineral Fertilisation			Farmyard Manure			Alternate Fertilisation
		NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
Nitrogen	0	30	45	60	32	46	62	30/32	45	60/62
Phosphorus	0	10	15	20	11	16.5	22	10	15	20
Potassium	0	27	41	54	32	48	64	27/32	41/48	54/64

.

On plots fertilised only with mineral fertilisers (NPK1, NPK2, NPK3), fertilisation was applied annually in 2022, 2023, and 2024. In spring, the full annual dose of phosphorus (P) and one-third of the nitrogen (N) and potassium (K) doses were applied. The remaining two-thirds of N and K were applied after the first and second harvest, respectively. The mineral fertilisers used included ammonium nitrate (34% N), potassium salt (60% K₂O), and granulated triple superphosphate (46% P₂O₅). For treatments fertilised with organic matter (FYM1, FYM2, FYM3), farmyard manure was applied each autumn (2021, 2022, and 2023) in three dosage levels corresponding to those of the mineral fertiliser treatments. Farmyard cattle manure with an average nutrient content of 3.0% N, 0.70% P, and 3.18% K (based on dry matter) and an average dry matter content of 34% was used for both organic and alternating treatments. The solid manure applied in the experiment was obtained from dairy cows kept in a livestock housing facility. Prior to application, the manure was aged in a heap for approximately six months. Chemical analyses of this manure were carried out before application to the plots. For this purpose, manure samples were taken from the heap in triplicate. On the alternately fertilised plots (NPK1/FYM1, NPK2/FYM2, NPK3/FYM3), manure was applied in the autumn of 2021 and 2023, while mineral fertilisers were used in 2022. The mineral fertilisers were applied at the same time and according to the same application scheme as in the NPK treatments.

2.3. Soil and Plant Sampling

In autumn 2021, prior to the application of manure, soil samples were collected from the experimental plots using an Egner stick at two depths: 0–10 cm and 10–20 cm. In each plot, 13 punctures (primary samples) were taken, separately for both layers. From these primary samples, two bulk samples were prepared for each plot, giving a total of 20 samples. The samples were analysed to determine soil pH as well as phosphorus (P), potassium (K), magnesium (Mg), and calcium (Ca) content. The characteristics of soil conditions at each test site are shown in

Table 2. Characteristics of soil conditions on individual experimental plots.

Soil Reaction/ Soil Layer	Control	Mineral Fertilisation			Farmyard Manure			Alternate Fertilisation
		NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
pH
0–10 cm	4.9	5.2	5.2	5.1	4.9	4.9	4.9	4.8	4.6	4.7
10–20 cm	4.5	5.2	5.2	4.6	4.8	4.4	4.5	4.6	4.6	4.7
Nitrogen
g 100−1 soil
0–10 cm	0.12	0.12	0.13	0.14	0.15	0.14	0.15	0.14	0.17	0.18
10–20 cm	0.08	0.07	0.09	0.08	0.14	0.11	0.11	0.10	0.12	0.12
Phosphorus
g 100−1 soil
0–10 cm	0.01	0.01	0.01	0.01	0.02	0.01	0.01	0.01	0.01	0.00
10–20 cm	0.01	0.01	0.01	0.01	0.02	0.01	0.01	0.01	0.01	0.00
Potassium
g 100−1 soil
0–10 cm	0.01	0.01	0.02	0.02	0.02	0.02	0.01	0.02	0.02	0.02
10–20 cm	0.01	0.01	0.01	0.01	0.02	0.01	0.02	0.02	0.02	0.01
Magnesium
g 100−1 soil
0–10 cm	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
10–20 cm	0.00	0.01	0.01	0.01	0.00	0.00	0.00	0.00	0.01	0.00
Calcium
g 100−1 soil
0–10 cm	0.03	0.06	0.06	0.04	0.07	0.06	0.04	0.04	0.01	0.07
10–20 cm	0.01	0.03	0.03	0.03	0.03	0.03	0.02	0.03	0.00	0.04

Treatments: Control—0 kg N ha⁻¹, 0 kg P ha⁻¹, 0 kg K ha⁻¹; NPK1—30 kg N ha⁻¹, 10 kg P ha⁻¹, 27 kg K ha⁻¹; NPK2—45 kg N ha⁻¹, 15 kg P ha⁻¹, 41 kg K ha⁻¹; NPK3—60 kg N ha⁻¹, 20 kg P ha⁻¹, 54 kg K ha⁻¹; FYM1—32 kg N ha⁻¹, 11 kg P ha⁻¹, 32 kg K ha⁻¹; FYM2—46 kg N ha⁻¹, 16.5 kg P ha⁻¹, 48 kg K ha⁻¹; FYM3—62 kg N ha⁻¹, 22 kg P ha⁻¹, 64 kg K ha⁻¹; NPK1/FYM1—30/32 kg N ha⁻¹, 10/11 kg P ha⁻¹, 27/32 kg K ha⁻¹; NPK2/FYM2—45/46 kg N ha⁻¹, 15/16.5 kg P ha⁻¹, 41/48 kg K ha⁻¹; NPK3/FYM3—60/62 kg N ha⁻¹, 20/22 kg P ha⁻¹, 54/64 kg K ha⁻¹. NPK1, NPK2, NPK3—mineral fertilisation, FYM1, FYM2, FYM3—farmyard manure, NPK1/FYM1, NPK2/FYM2, NPK3/FYM3—alternate fertilisation.

.

Each year, prior to the first harvest (in May), the botanical composition of the sward was assessed on all plots using the Klapp estimation method [30]. In each plot, a square with an area of 25 m² (5 m × 5 m) was randomly designated. In each square, all species were recorded in the order of grasses, legumes, and others (herbs and weeds). According to the adopted method, the percentage share of individual species in each group was estimated, with an accuracy of 1%, starting from the most numerous species. For all plots, the functional group structure was also determined, expressed as the percentage share of grasses, legumes, herbs, and weeds, along with the number of species in each group.

During the period (2022–2024), plant material samples were collected from each treatment plot during successive harvests to determine dry matter yield (DMY) and analyse chemical composition. Plant green mass samples were collected from each plot in four replicates. For this purpose, four equal areas were designated on each plot. On each of these areas, a random location for plant material sampling was selected using a 1 m² (1 m × 1 m) frame. The plants were cut manually with scissors to a height of 5 cm. The collected biomass samples were weighed and combined into an averaged sample for each treatment (10 samples in total). They were then left to dry to obtain air-dry weight.

2.4. Dry Matter Yield and Chemical Analyses

Dry matter yield (DMY) was calculated based on the fresh weight of plant material harvested from an area of 1 m². Dry matter content was determined by drying 100 g of plant material in a laboratory oven at 105 °C for 4 h. The results were expressed in tonnes of dry matter per hectare. The remaining plant material was used for chemical analysis. After drying and grinding, nitrogen (N), phosphorus (P), and potassium (K) content were determined for all samples.

Chemical analyses were conducted at the ITP–PIB laboratory. Soil pH was measured using the potentiometric method in a 1 mol dm⁻³ KCl solution [31]. Nitrogen (N) content was determined using the Kjeldahl method. Total phosphorus (P) was analysed colorimetrically using ammonium molybdate and sodium metabisulfite. Potassium (K) content was measured using flame emission spectrometry. Magnesium (Mg) and calcium (Ca) contents were determined by atomic absorption spectrometry (S series AA spectrometer, Thermo Fisher Scientific, Waltham, MA, USA).

2.5. Nutrient Balance Calculation

Based on the results obtained, simplified nutrient balances were calculated. Annual balances of nitrogen, phosphorus, and potassium were determined for each treatment by calculating the difference between inputs and outputs.

On the input side, nutrient loads from fertilisers and atmospheric deposition were included. The average nutrient input from precipitation was adopted from previous research on permanent grasslands in Poland conducted by Barszczewski and Burs [32] and amounted to 17.3 kg N ha⁻¹, 0.4 kg P ha⁻¹, and 4.4 kg K ha⁻¹ per year.

The nitrogen balance also accounted for biological nitrogen input from soil microorganisms and symbiotic nitrogen fixation by leguminous species. The microbial contribution to nitrogen input was assumed to be 10 kg N ha⁻¹, based on data from Pietrzak [33]. Nitrogen fixation by legumes was estimated according to their percentage share in the sward composition. A 1% share of leguminous species was assumed to correspond to 3 kg of fertiliser-equivalent nitrogen, in line with national agricultural recommendations [34], which include practices for protecting water resources against nitrate pollution from agricultural sources. On the output side, nutrient removal was calculated based on the nutrient content of the harvested plant biomass.

2.6. Meteorological Data

Meteorological data, including temperature and precipitation, were recorded throughout the study to evaluate their influence on nutrient uptake, sward yield, and nutrient balances, key objectives of this experiment. The data were obtained from the nearest weather station located in Poznań and sourced from a publicly accessible online database maintained by the Institute of Meteorology and Water Management—National Research Institute [35]. The dataset covered the growing season in Polish conditions, defined as the period from April to September, for the years 2022–2024. These data were used to calculate average monthly temperatures and total precipitation for each month of the growing season.

Based on the obtained data, the Sielianinov hydrothermal coefficient (HTC) was also calculated [36]. This index was used to characterise the pluviothermal conditions during the study period in more detail. The HTC takes into account both temperature and precipitation and it is flexible enough to be applied on a monthly or 10-day basis. It is sensitive to dry conditions, making it a useful tool for monitoring agricultural drought. The HTC was calculated using the following formula:

HTC = (P 10)/Σt

(1)

P—the total monthly rainfall (mm), t—the monthly total of average daily air temperatures > 0 °C.

The following classes were used to assess pluviothermal conditions: extremely wet HTC > 3.0; very wet 2.5 < HTC ≤ 3.0; wet 2.0 < HTC ≤ 2.5; fairly wet 1.6 < HTC ≤ 2.0; optimal 1.3 < HTC ≤ 1.6; fairly dry 1.0 < HTC ≤ 1.3; dry 0.7 < HTC ≤ 1.0; very dry 0.4 < HTC ≤ 0.7; extremely dry HTC ≤ 0.4.

2.7. Statistical Analysis

Due to the field-based nature of the study, the experiment did not include replications within a single year. Therefore, the statistical model treated fertilisation and year (as a blocking factor) as sources of variation. The effect of fertilisation treatment was assessed using two-way ANOVA, followed by Tukey’s post hoc test to compare means at a significance level of α = 0.05. All statistical analyses were performed using Statgraphics 18 software.

3. Results

3.1. Meteorological Conditions

The meteorological conditions in April–September in 2022–2024 are presented in

Table 3. Average monthly temperatures and precipitation in 2022–2024 at the meteorological station in Poznań [32].

Month	Temperature [°C]	Precipitation [mm]	Temperature [°C]	Precipitation [mm]	Temperature [°C]	Precipitation [mm]
	2022		2023		2024
April	13	36	13.2	37	16.5	50
May	20.8	23	19.4	30	23.4	52
June	26.4	63	25.7	39	24.4	93
July	26.4	23	26.8	62	26.6	75
August	27.8	53	24.6	156	27.2	39
September	18.5	32	25.0	3	22.9	56

.

In the first year of the study (2022), average temperatures during the growing season ranged from 13.0 °C in April to 27.8 °C in August. In 2023, temperatures varied between 13.2 °C in April and 26.8 °C in July, while in 2024 they ranged from 16.4 °C in April to 27.2 °C in August. The smallest temperature variation occurred in 2024. That year, the start of the growing season was noticeably warmer than in the previous years, but temperatures remained relatively stable throughout the following months. The warmest final month of the growing season was recorded in 2023, with an average temperature of 25.0 °C in September, while the coolest was in 2022, with 18.5 °C. Total precipitation across the study period amounted to 229 mm in 2022 (the driest year), 326 mm in 2023, and 365 mm in 2024 (the wettest year). The highest variation in monthly rainfall occurred in 2023, with a record 156 mm in August followed by just 3 mm in September. In the other years, the highest monthly rainfall was recorded in June. The driest months were May and July 2022 (23 mm each) and August 2024 (39 mm). Interestingly, August 2023, despite exceptionally high rainfall, was also the coldest August of the three-year period.

Table 4. Monthly values of the HTC for selected months in the years 2022–2024.

Year	Month
	April	May	June	July	August	September
2022	1.5	0.5	1.1	0.4	0.8	0.8
2023	1.5	0.7	0.7	1.0	2.5	0.1
2024	0.9	0.7	1.3	0.9	0.5	0.8

Extremely wet HTC > 3.0; very wet 2.5 < HTC ≤ 3.0; wet 2.0 < HTC ≤ 2.5; fairly wet 1.6 < HTC ≤ 2.0; optimal 1.3 < HTC ≤ 1.6; fairly dry 1.0 < HTC ≤ 1.3; dry 0.7 < HTC ≤ 1.0; very dry 0.4 < HTC ≤ 0.7; extremely dry HTC ≤ 0.4.

shows the values of the HTC.

Throughout the experiment, HTC ranged from 0.1 to 1.5, indicating mostly very dry or dry conditions, with occasional months classified as optimal. The only exception was August 2023, which, due to total rainfall of 156 mm, was classified as wet. The beginning of the growing season was optimal in both 2022 and 2023, while in 2024 it was already dry. May conditions were consistently very dry across all three years. June was particularly dry in 2023, with an HTC of 0.7. The greatest variation between years was observed during the final three months of the growing season. July was very dry in 2022, and dry in both 2023 and 2024. August showed the sharpest contrast: it was very dry in 2024 but wet in 2023. September 2022 and 2024 were classified as dry, while 2023 experienced extremely dry conditions. When considering the growing season, 2024 was the driest year on average, and 2023 was the wettest. However, taking the entire study period into account, all three years can generally be classified as dry.

3.2. Species Richness and Botanical Composition of the Sward

The number of plant species in the botanical composition of the meadow sward across individual treatments over the three-year study period is presented in Figure 3. Overall, the average number of species recorded per treatment showed little variation and did not exhibit any clear relationship with the type or dose of fertilisation. On the control plots, the number of species ranged from 15 to 17. Similar patterns were observed on fertilised plots, with only slight differences: 14–19 species under mineral fertilisation, 14–21 under manure fertilisation, and 14–20 under alternating fertilisation. No distinct year-to-year trends were identified. In the first year, species counts ranged from 14 to 19 across all treatments. In the second year, the range increased slightly to 16–21 species, while in the third year it narrowed to 15–18. These findings suggest that species richness remained relatively stable regardless of the fertilisation strategy applied.

The proportions of different plant groups in the meadow sward across treatments and study years are shown in Figure 4. Grasses consistently formed the dominant group in all treatments and played a key role in determining yield. Their average share in the sward was 83% in the first year, 81% in the second, and 80% in the third year. The most common species within this group included Arrhenatherum elatius (L.) P. Beauv. ex J. Presl & C. Presl, Poa pratensis L., Festuca rubra L., Holcus lanatus L., Dactylis glomerata L., Alopecurus pratensis L., and Festuca pratensis Huds.

The second largest functional group comprised broad-leaved dicotyledonous species, mainly classified as herbs and weeds. This group included, among others, Plantago lanceolata L., Taraxacum officinale F.H. Wigg., Achillea millefolium L., and Rumex acetosa L. Their average share in the sward was 16% in 2022 and 18% in both 2023 and 2024. A slightly higher proportion of these species was observed on plots receiving higher doses of mineral fertilisers (NPK2 and NPK3).

Legumes (Fabaceae) made up the smallest group, contributing only 1–3% to the sward composition. This group consisted mainly of Vicia cracca L., Vicia sepium L., Lotus corniculatus L., and occasionally Vicia angustifolia L. The average proportion of legumes was 1% in the first year, increasing slightly to 2% in the second and third years. A modest increase in Fabaceae presence (by 1–2%) was noted on all plots fertilised with farmyard manure (FYM1, FYM2, FYM3), as well as on those receiving alternating fertilisation. An exception was the NPK3/FYM3 treatment, where legumes were absent throughout the entire study period. On plots fertilised solely with mineral fertilisers, the share of legumes remained low and stable, with a decline observed in the third year on the NPK1 plot.

3.3. Meadow Sward Yield

The DMY of the meadow sward across individual treatments over the three-year study period are presented in

Table 5. Effect of analysed factors on DMY [Mg ha⁻¹].

Factor	Control	NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
Count	3	3	3	3	3	3	3	3	3	3
LS Mean	2.58	3.29	4.26	4.47	3.87	4.52	4.29	3.89	4.19	4.72
Homogeneous Groups	a	ab	bc	bc	bc	bc	bc	bc	bc	c

Means within columns followed by the same letter are not significantly different from each other, according to Tukey’s test at p < 0.05. Treatments: Control—0 kg N ha⁻¹, 0 kg P ha⁻¹, 0 kg K ha⁻¹; NPK1—30 kg N ha⁻¹, 10 kg P ha⁻¹, 27 kg K ha⁻¹; NPK2—45 kg N ha⁻¹, 15 kg P ha⁻¹, 41 kg K ha⁻¹; NPK3—60 kg N ha⁻¹, 20 kg P ha⁻¹, 54 kg K ha⁻¹; FYM1—32 kg N ha⁻¹, 11 kg P ha⁻¹, 32 kg K ha⁻¹; FYM2—46 kg N ha⁻¹, 16.5 kg P ha⁻¹, 48 kg K ha⁻¹; FYM3—62 kg N ha⁻¹, 22 kg P ha⁻¹, 64 kg K ha⁻¹; NPK1/FYM1—30/32 kg N ha⁻¹, 10/11 kg P ha⁻¹, 27/32 kg K ha⁻¹; NPK2/FYM2—45/46 kg N ha⁻¹, 15/16.5 kg P ha⁻¹, 41/48 kg K ha⁻¹; NPK3/FYM3—60/62 kg N ha⁻¹, 20/22 kg P ha⁻¹, 54/64 kg K ha⁻¹. NPK1, NPK2, NPK3—mineral fertilisation, FYM1, FYM2, FYM3—farmyard manure, NPK1/FYM1, NPK2/FYM2, NPK3/FYM3—alternate fertilisation.

and Figure 5. In the first year of the study (2022), the average DMY was 4.31 Mg ha⁻¹. The highest yields (4.52 Mg ha⁻¹ on average) were obtained on plots fertilised solely with mineral fertilisers (NPK1, NPK2, NPK3). Slightly lower yields were recorded on plots fertilised with farmyard manure applied in autumn 2021 (FYM1, FYM2, FYM3), where the average yield was 4.49 Mg ha⁻¹. For plots receiving alternating fertilisation (NPK1/FYM1, NPK2/FYM2, NPK3/FYM3), the average DMY amounted to 4.42 Mg ha⁻¹, which was only 2.3% lower than for mineral-fertilised plots and 1.6% lower than for those fertilised with manure.

The highest average DMY across the entire study were recorded in the second year (2023), reaching 4.40 Mg ha⁻¹. This was 2% higher than in 2022 and 32% higher than in 2024. A major contribution to this result came from the third cut. In August 2023, heavy rainfall occurred following the second cut in July, which significantly improved regrowth and boosted the final yield, as shown in Figure 5. In that year, the highest average DMY (4.96 Mg ha⁻¹) were achieved on plots with alternating fertilisation, followed by plots fertilised with manure (4.81 Mg ha⁻¹), while the lowest DMY (4.09 Mg ha⁻¹) were observed on plots fertilised with mineral fertilisers. It is worth noting that plots with alternate fertilisation in 2023 were fertilised with mineral fertilisers only, yet the high DMY may have been influenced by the residual effect of manure applied two years earlier (autumn 2021).

The lowest yields were recorded in the final year of the study (2024), averaging 3.33 Mg ha⁻¹. Average yields for this year were as follows: 3.42 Mg ha⁻¹ on plots with mineral fertilisation, 3.39 Mg ha⁻¹ on manure fertilised plots, and 3.44 Mg ha⁻¹ on plots receiving alternating fertilisation. These results were, on average, 1.52 Mg ha⁻¹ lower than in the previous year.

When considering the full three-year period, the lowest average DMY (2.58 Mg ha⁻¹) was recorded on the unfertilised control plot. This plot yielded 2.80 Mg ha⁻¹ in 2022, 2.39 Mg ha⁻¹ in 2023, and 2.56 Mg ha⁻¹ in 2024. The control plot significantly underperformed compared to the fertilised plots, with the exception of the NPK1 treatment, where the difference was not statistically significant (Table 5). The three-year average DMY for the fertilised plots were: 4.01 Mg ha⁻¹ for mineral fertilisation, 4.23 Mg ha⁻¹ for manure fertilisation, and 4.27 Mg ha⁻¹ for alternating fertilisation. No significant differences were observed among most of the fertilised treatments, except between NPK1 and NPK3/FYM3.

3.4. Nutrient Balance

3.4.1. Nitrogen Balance

Nitrogen balance values varied depending on the type of fertilisation and the year of the study (Figure 6). In the unfertilised treatment (Control), a negative nitrogen balance was recorded throughout the study period. The values ranged from −14.7 kg N·ha⁻¹ in 2022 to −8.9 kg N·ha⁻¹ in 2024.

In treatments fertilised only with mineral fertilisers, the nitrogen balance was mostly positive, except for NPK2 in 2022, where a negative value of −10.6 kg N·ha⁻¹ was recorded. In the first two years, nitrogen balances were relatively low and similar between treatments. The highest nitrogen balances were observed in the third year (2024), especially in the NPK3 treatment, where the balance reached +32.8 kg N·ha⁻¹.

In treatments fertilised each year with manure only, nitrogen balance values varied depending on the year and fertiliser dose. In the FYM1 treatment, the nitrogen balance was slightly negative in the first two years (−6.8 kg N·ha⁻¹ in 2022 and −4.1 kg N·ha⁻¹ in 2023) and became positive in 2024 (+5.5 kg N·ha⁻¹). The FYM2 treatment showed greater variability: the balance was strongly negative in 2023 (−17.4 kg N·ha⁻¹) and positive in 2024 (+18.5 kg N·ha⁻¹). The highest nitrogen balance values, especially in 2024 (+42.5 kg N·ha⁻¹), were recorded in the FYM3 treatment.

A similar pattern was observed in the alternating fertilisation treatments. In 2022, the balances were positive, reaching up to +12.1 kg N·ha⁻¹ in NPK3/FYM3. In 2023, all alternating treatments showed negative balances, with the lowest value of −26.8 kg N·ha⁻¹ also in NPK3/FYM3. In 2024, nitrogen balances were positive again, reaching up to +29.1 kg N·ha⁻¹ in the NPK2/FYM2 treatment.

Regardless of fertiliser type and dose, the year of the study had a clear effect on nitrogen balance values. In 2022, balances ranged from −14.7 kg N·ha⁻¹ (Control) to +19.9 kg N·ha⁻¹ (FYM3), and the differences between fertilised treatments were relatively small. In 2023, the results varied widely, from −26.8 kg N·ha⁻¹ (NPK3/FYM3) to +5.4 kg N·ha⁻¹ (NPK3). Negative balances were recorded in all manure and alternating treatments, as well as in the unfertilised control. In 2024, all fertilised treatments showed a positive nitrogen balance, ranging from +5.5 kg N·ha⁻¹ (FYM1) to +42.5 kg N·ha⁻¹ (FYM3), which was the highest value in the entire study period. The analysis of variance did not reveal a statistically significant effect of the treatments on nitrogen balance (p = 0.0513) (

Table 6. Effect of analysed factors on nitrogen balance.

Factor	Control	NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
Count	3	3	3	3	3	3	3	3	3	3
LS Mean	−12.2	6.3	5.2	16.3	−1.8	2.9	20.6	3.4	10.9	4.7

Since the differences between means were only marginally significant (p = 0.0513, ANOVA), no Tukey’s post hoc test was conducted.

). Although the differences were not statistically significant, a clear trend was evident. The most favourable nitrogen balance was recorded in the FYM3 treatment (mean = 20.6), while the lowest was observed in the Control (mean = −12.2).

3.4.2. Phosphorus Balances

Phosphorus balances were positive in most fertilisation treatments and showed little change between years (Figure 7). Negative balances occurred only in the unfertilised treatment (Control), where there were steady phosphorus losses throughout the study period, ranging from −6.6 kg P ha⁻¹ in 2022 to −8.3 kg P ha⁻¹ in 2024, indicating that soil phosphorus reserves were being depleted.

In mineral fertilisation, phosphorus balances were stable and increased with higher doses: from about +0.6 kg P ha⁻¹ and −0.6 kg P ha⁻¹ (NPK1) to just over 9.0 kg P ha⁻¹ (NPK3) in the last two years. The highest value was recorded in 2024 at the NPK3 site (9.8 kg P ha⁻¹). Positive phosphorus balances were also recorded in the first year, ranging from 2.1 to 11 kg P ha⁻¹.

Treatments with manure also showed positive phosphorus balances, especially at the highest dose (FYM3), reaching +11.2 kg P ha⁻¹ in 2022 and +10.2 kg P ha⁻¹ in 2024, although these values varied more between years compared to mineral fertilisation.

Alternating treatments (NPK/FYM) gave similar results to the organic and mineral fertilisers. Phosphorus balances increased with the dose, from 1 to 3 kg P ha⁻¹ at the lowest dose, up to over 9 kg P ha⁻¹ at the highest dose.

Like with nitrogen balance, phosphorus balance values changed depending on the year. In the first year, phosphorus balances ranged from −6.6 kg P ha⁻¹ (Control) to +11.0 kg P ha⁻¹ (NPK3) and +11.2 kg P ha⁻¹ (FYM3). The second year showed slightly better values, from −6.7 kg P ha⁻¹ (Control) to +9.1 kg P ha⁻¹ (NPK3). In the last year, phosphorus balances ranged from −8.3 kg P ha⁻¹ (Control) to +10.2 kg P ha⁻¹ (FYM3).

As observed for nitrogen, phosphorus balances also varied depending on the year of the study. In 2022, balances ranged from −6.6 kg P ha⁻¹ in the control to +11.0 kg P ha⁻¹ in NPK3 and +11.2 kg P ha⁻¹ in FYM3. In 2023, the values were slightly more favourable, spanning from −6.7 kg P ha⁻¹ (control) to +9.1 kg P ha⁻¹ (NPK3). The final year (2024) showed the widest range, from −8.3 kg P ha⁻¹ in the control to +10.2 kg P ha⁻¹ in FYM3.

The applied fertiliser regimes had a significant effect on the phosphorus balance. The Control recorded the lowest and significantly negative P balance, indicating the strongest negative impact on phosphorus content. All fertilised treatments resulted in significantly higher phosphorus balances compared to the Control (

Table 7. Effect of analysed factors on phosphorus balance.

Factor	Control	NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
Count	3	3	3	3	3	3	3	3	3	3
LS Mean	−7.2	0.7	4.7	10.0	0.4	4.3	9.8	2.1	4.8	7.3
Homogeneous Groups	a	bc	cd	e	b	bcd	e	bc	cd	de

Means within columns followed by the same letter are not significantly different from each other, according to Tukey’s test at p < 0.05. Treatments: Control—0 kg N ha⁻¹, 0 kg P ha⁻¹, 0 kg K ha⁻¹; NPK1—30 kg N ha⁻¹, 10 kg P ha⁻¹, 27 kg K ha⁻¹; NPK2—45 kg N ha⁻¹, 15 kg P ha⁻¹, 41 kg K ha⁻¹; NPK3—60 kg N ha⁻¹, 20 kg P ha⁻¹, 54 kg K ha⁻¹; FYM1—32 kg N ha⁻¹, 11 kg P ha⁻¹, 32 kg K ha⁻¹; FYM2—46 kg N ha⁻¹, 16.5 kg P ha⁻¹, 48 kg K ha⁻¹; FYM3—62 kg N ha⁻¹, 22 kg P ha⁻¹, 64 kg K ha⁻¹; NPK1/FYM1—30/32 kg N ha⁻¹, 10/11 kg P ha⁻¹, 27/32 kg K ha⁻¹; NPK2/FYM2—45/46 kg N ha⁻¹, 15/16.5 kg P ha⁻¹, 41/48 kg K ha⁻¹; NPK3/FYM3—60/62 kg N ha⁻¹, 20/22 kg P ha⁻¹, 54/64 kg K ha⁻¹. NPK1, NPK2, NPK3—mineral fertilisation, FYM1, FYM2, FYM3—farmyard manure, NPK1/FYM1, NPK2/FYM2, NPK3/FYM3—alternate fertilisation.

). The highest, statistically similar positive P balances were observed in the treatments with the highest fertiliser doses: NPK3 and FYM3. The NPK3/FYM3 treatment did not differ significantly from either the FYM3 and NPK3 group or the group receiving FYM2, NPK2, and NPK2/FYM2 application rates.

3.4.3. Potassium Balances

Potassium balances are shown in Figure 8. The data show that, for most fertilisation treatments, the potassium balance was negative, and the values varied greatly both between treatments and between years of the study. Positive potassium balances were recorded only in 2024, and only in treatments fertilised with higher doses (NPK2, NPK3, NPK2/FYM2, NPK3/FYM3). Treatments fertilised with both mineral fertilisers and manure had varying balance values, but generally, the negative potassium balance was greater in treatments with manure.

In the first year of the study (2022), potassium balances ranged from −8.4 kg K ha⁻¹ (NPK1) to −58.6 kg K ha⁻¹ (FYM1). Less favourable balances were found in treatments fertilised with manure and in alternating treatments. In the second year (2023), the range was even wider, from −11.0 kg K ha⁻¹ (NPK1) to −78.6 kg K ha⁻¹ (FYM2). This was also the year with the highest average dry matter yield, which suggests that potassium fertilisation was not enough to meet the plants’ needs, especially when manure was used.

In 2024, potassium balances were more favourable than in the previous two years, ranging from +7.9 kg K ha⁻¹ (NPK2/FYM2) to −22.7 kg K ha⁻¹ (FYM1). This was also the year with the lowest sward yield throughout the study, which resulted in lower potassium uptake and, consequently, more positive balance values.

As shown in

Table 8. Effect of analysed factors on potassium balance.

Factor	Control	NPK1	NPK2	NPK3	FYM1	FYM2	FYM3	NPK1/FYM1	NPK2/FYM2	NPK3/FYM3
Count	3	3	3	3	3	3	3	3	3	3
LS Mean	−26.4	−7.4	−10.6	−13.6	−51.7	−45.0	−22.8	−22.5	−21.2	−26.4
Homogeneous Groups	abc	c	bc	bc	a	ab	abc	abc	abc	abc

Means within columns followed by the same letter are not significantly different from each other, according to Tukey’s test at p < 0.05. Treatments: Control—0 kg N ha⁻¹, 0 kg P ha⁻¹, 0 kg K ha⁻¹; NPK1—30 kg N ha⁻¹, 10 kg P ha⁻¹, 27 kg K ha⁻¹; NPK2—45 kg N ha⁻¹, 15 kg P ha⁻¹, 41 kg K ha⁻¹; NPK3—60 kg N ha⁻¹, 20 kg P ha⁻¹, 54 kg K ha⁻¹; FYM1—32 kg N ha⁻¹, 11 kg P ha⁻¹, 32 kg K ha⁻¹; FYM2—46 kg N ha⁻¹, 16.5 kg P ha⁻¹, 48 kg K ha⁻¹; FYM3—62 kg N ha⁻¹, 22 kg P ha⁻¹, 64 kg K ha⁻¹; NPK1/FYM1—30/32 kg N ha⁻¹, 10/11 kg P ha⁻¹, 27/32 kg K ha⁻¹; NPK2/FYM2—45/46 kg N ha⁻¹, 15/16.5 kg P ha⁻¹, 41/48 kg K ha⁻¹; NPK3/FYM3—60/62 kg N ha⁻¹, 20/22 kg P ha⁻¹, 54/64 kg K ha⁻¹. NPK1, NPK2, NPK3—mineral fertilisation, FYM1, FYM2, FYM3—farmyard manure, NPK1/FYM1, NPK2/FYM2, NPK3/FYM3—alternate fertilisation.

, the most advantageous treatments were those grouped under the homogeneous letters “bc” and “c”, corresponding to NPK1, NPK2, and NPK3. Across all years, most treatments exhibited a negative potassium balance, indicating that the applied fertiliser doses did not fully meet the potassium requirements of the plants. Less negative balances were observed only in the year with the lowest yields, suggesting reduced potassium uptake due to lower biomass production.

4. Discussion

4.1. Effect of Fertiliser Type and Dose on the Botanical Composition of the Sward

The use of fertilisers often leads to a decline in plant species richness and changes in vegetation structure and functional composition [17,37,38]. This is important because species composition, both the number of species and their share in the sward, determines the productivity and practical value of permanent grassland [39].

As shown by a meta-analysis conducted by Shi et al. [8], mineral fertilisation, although it increases biomass production, caused an average reduction in species richness of 18%. Organic fertilisers, by improving soil structure, increasing sorption capacity, and supporting the soil microbiome, can promote the development of specific functional groups, including legumes. For example, Stybnarova et al. [10] reported that fertilisation with cattle manure and slurry, combined with moderate or intensive grassland use, increased the share of legumes in the sward.

Shi et al. [8] also report that both nitrogen and phosphorus fertilisation contributed to a decline in species richness. In the case of nitrogen, this is due to stronger, more competitive species outcompeting weaker ones [40,41]. In contrast, in our study, the number of species remained relatively stable throughout the observation period. No clear decline in species richness was observed in the mineral-fertilised treatments, nor was there a significant improvement in botanical composition due to organic fertilisation.

Apart from the fertiliser type, the applied dose also plays an important role. On intensively fertilised grasslands, increased light competition eliminates smaller and less competitive plants [42]. A meta-analysis by Francksen et al. [43] showed a negative correlation between nitrogen input and the number of species, estimating a loss of about 1.5 species per 100 kg N·ha⁻¹·yr⁻¹. Legumes are particularly sensitive to high nitrogen levels [44]. Kacorzyk et al. [45] observed that applying sheep manure at nitrogen rates below 69 kg N·ha⁻¹ increased the share of legumes, while higher rates (69–103 kg N·ha⁻¹) favoured grasses at their expense. In addition, some studies suggest that mineral phosphorus fertilisation can also significantly reduce species richness [46].

However, the impact of high mineral and organic fertiliser doses on species diversity may vary. For instance, Shi et al. [8] did not find a significant effect of organic fertiliser rate on total species number. This could be because nutrients in organic fertilisers become available only after the decomposition of organic matter, which delays their uptake and reduces competitive pressure.

In our study, the applied fertiliser dose also had no effect on the botanical composition of the sward. Overall, the lack of a clear effect in our experiment may be due to the moderate fertiliser doses used (up to 60 kg N·ha⁻¹ and 20 kg P·ha⁻¹) and the relatively short duration of the study (three years). Habitat-specific conditions should also be taken into account. The response of the sward to fertilisation depends not only on the fertiliser type and rate, but also on soil and climatic conditions, as well as the initial composition of the plant community, which may explain why our findings differ in some respects from studies conducted under more intensive or long-term fertilisation regimes.

4.2. Effect of Fertiliser Type and Dose on DMY

The DMY obtained on the control plots (i.e., without fertilisation) averaged 2.58 Mg·ha⁻¹, reflecting the natural production potential of dry meadows under temperate climate conditions. The application of different fertilisation types had a significant impact on yield compared to the control. On average, DMY increased by 55.3% with mineral fertilisation (NPK), by 63.9% with organic fertilisation (FYM), and by 65.45% with alternating NPK and FYM fertilisation (NPK/FYM). These results are broadly consistent with findings from previous studies. For instance, Shi et al. [8] reported that organic fertilisers increased biomass production more than mineral ones (on average by 56% ± 5% vs. 42% ± 2%), while Stybnarova et al. [10] found that fertilisation with manure and slurry fertilisation significantly increased DMY, by 51.9% and 56%, respectively, compared to the control (4.81 t·ha⁻¹). Our observed increases, particularly in the combined NPK/FYM treatments, slightly exceed these reported values, suggesting a potential synergistic effect under our experimental conditions.

In our study, the greatest biomass increase was observed in the treatment with the highest alternating NPK and FYM fertilisation rate (NPK3/FYM3), where yield rose by more than 83% compared to the control. This supports literature reports suggesting that the combined application of mineral and organic fertilisers produces a synergistic effect: it improves soil fertility, which in turn enables higher yields than when each type of fertiliser is applied separately [20,21,47].

Organic fertilisers supply numerous macro- and micronutrients. By gradually releasing nutrients, they increase soil humus content, improve soil structure, and support microbial development [48], contributing to sustained productivity over time. Mineral fertilisers, in contrast, provide nutrients more rapidly, supporting intensive plant growth and development [49]. Combining both fertiliser types helps reduce nutrient losses typical of mineral fertilisers, supplies carbon as an energy source for the soil microbiome, and compensates for the slower nutrient release from organic fertilisers [50], which can explain the higher yields in NPK/FYM treatments.

In our study, significant differences in yield were also observed between years, primarily due to variations in weather conditions. The lowest average DMY (3.33 Mg·ha⁻¹) was recorded in the final year of the study (2024). In contrast, the highest average DMY across all treatments occurred in the second year (2023), reaching 4.40 Mg·ha⁻¹. This was 2% higher than in 2022 and as much as 32% higher than in 2024. The high DMY in 2023 was largely due to the abundant third cut. In August, during the regrowth period following the second cut (in July), heavy rainfall created highly favourable conditions for plant regrowth, leading to a substantial increase in total annual yield.

The smallest differences between years and harvests were observed in the Control and NPK1 treatments. Analysis of the first harvest showed the maximum yield in 2022, likely due to optimal early-season conditions and a delayed harvest date (15 June). In contrast, the first harvest in 2023 occurred much earlier (22 May), a 24-day difference, contributing to lower relative yields. In 2024, the first cut was similar to 2023 (13 May), while lower yields were likely due to less favourable early-season weather. Differences in the second harvests were smaller between treatments and years. The third harvest in 2023, performed the latest (2 October), reached the highest yields, whereas in 2022 and 2024 the final cut took place at the beginning of September (Figure 4).

4.3. Effect of Fertiliser Type and Dose on Nitrogen, Phosphorus, and Potassium Balances

Nutrient balances calculated for different fertilisation types and doses are a useful tool for evaluating whether nutrient inputs match plant requirements under specific environmental conditions (soil, precipitation, temperature), which ultimately determine yield levels [51]. Small balance values, whether positive (+) or negative (−), indicate that the fertilisation applied is well aligned with actual crop needs and sustainable nutrient management.

The analysis of nitrogen, phosphorus, and potassium balances confirmed that both the type and dose of fertilisation significantly influenced nutrient surpluses or deficits. The control treatment showed consistent nutrient shortages in all years, with average negative balances of −12.2 kg N·ha⁻¹, −7.2 kg P·ha⁻¹, and −26.4 kg K·ha⁻¹. These results indicate ongoing soil nutrient depletion in the absence of external inputs.

In mineral fertilisation treatments, nitrogen balances were positive at all doses, especially at 30 kg N·ha⁻¹ (average +6.3 kg N·ha⁻¹). The highest N dose (60 kg N·ha⁻¹) led to a nitrogen surplus of +16.3 kg N·ha⁻¹, suggesting excess nitrogen relative to plant demand. Phosphorus balances also increased with dose, ranging from +0.7 kg P·ha⁻¹ at the lowest rate to +10.0 kg P·ha⁻¹ at the highest. Potassium balances, however, remained negative at all NPK levels (from −7.4 to −13.6 kg K·ha⁻¹), likely due to high plant uptake or leaching losses, a pattern observed in other studies of mineral fertilisation on light-textured soils [52].

Manure fertilisation also led to positive nitrogen and phosphorus balances, with the highest values being +20.6 kg N·ha⁻¹ and +9.8 kg P·ha⁻¹. Nevertheless, potassium balances were strongly negative across all FYM treatments (from −51.7 to −22.8 kg K·ha⁻¹), reflecting both the relatively low K content of FYM and the high mobility of potassium in soils, especially under leaching-prone conditions. Kayser and Isselstein [49] noted that potassium in grassland systems is poorly retained in the soil and is easily lost due to uptake by plants and export with harvested biomass, even under moderate fertilisation.

Interestingly, our results differ from Barszczewski and Ducka [53], who reported negative nitrogen and potassium balances and near-zero phosphorus balances at comparable fertiliser rates (60 kg N·ha⁻¹) in a permanent meadow on degraded black soil. This highlights how soil type and initial nutrient status can strongly influence nutrient balance outcomes.

The alternating application of NPK and FYM (NPK/FYM) had a positive effect on nutrient balances for all three elements. For nitrogen, it improved the balance compared to FYM alone, without causing the large surpluses typical of continuous NPK application. For phosphorus, it helped reduce the risk of accumulation observed under exclusive NPK fertilisation. Regarding potassium, alternating fertilisation reduced the deficit typical of FYM treatments and improved potassium availability to plants.

4.4. Impact of Weather Conditions on Nitrogen, Phosphorus, and Potassium Balances

The results, regardless of the type of fertilisation used, confirm that weather conditions in each year had a strong influence on nitrogen, phosphorus and potassium balances. In 2022, moderate spring temperatures combined with rainfall promoted efficient nutrient uptake by the plants. As a result, nitrogen and phosphorus balances were positive across all fertilisation variants. Despite the favourable conditions, potassium balances remained clearly negative, especially in the FYM treatments (e.g., FYM1: −58.6 kg K·ha⁻¹), indicating that manure alone may not fully satisfy the potassium demands of the plants.

In 2023, although temperatures were close to the long-term average, there was a large variation in rainfall, with water shortages in spring and heavy rains in August (156 mm). This contributed to the highest yield of the entire study period (average 4.40 Mg·ha⁻¹), which in term strongly influenced the nutrient balances. Negative nitrogen balances were recorded in most FYM and NPK/FYM variants. Only the highest NPK treatment showed a positive balance (+5.4 kg N·ha⁻¹). Phosphorus balances were also lower than in the previous year but remained positive in the NPK/FYM2 variant (+2.6 kg P·ha⁻¹). Potassium balances dropped to their lowest levels during the study, for example FYM2: −78.6 kg K·ha⁻¹, likely due to high leaching losses caused by intense rainfall in the second half of the growing season.

In 2024, spring was unusually hot (April: 16.5 °C, May: 23.4 °C) and rainfall was highly variable. Due to reduced yields (average 3.33 Mg·ha⁻¹), all fertilised variants showed positive nitrogen balances, with the highest in FYM3 (+42.5 kg N·ha⁻¹) and NPK3 (+32.8 kg N·ha⁻¹). These results suggest that plant nitrogen uptake was limited under heat stress and lower biomass production. Similarly, high phosphorus balances in the intensively fertilised variants (FYM3: +10.2 kg P·ha⁻¹, NPK3/FYM3: +9.1 kg P·ha⁻¹) also point to reduced nutrient uptake. Potassium balances improved slightly, especially in NPK3/FYM3 (+3.6 kg K·ha⁻¹) and NPK3 (+5.9 kg K·ha⁻¹), but remained negative in the FYM-only treatments, highlighting persistent K deficiency despite organic inputs.

4.5. Assessment of Nutrient Balances in the Context of Environmental Guidelines

For environmental reasons, nutrient balances should ideally be balanced. This is generally true in closed natural ecosystems where no plant biomass is harvested. However, under field conditions, perfectly balanced nutrient levels are rarely achieved, as confirmed by both the literature and our own research. As noted by Fotyma et al. [54], achieving completely neutral nutrient balances in agricultural practice is difficult, especially while maintaining profitable production. Some surplus, particularly nitrogen in intensive fertilisation, is inevitable, but its magnitude should be managed to reduce environmental pollution risks [55].

OECD experts consider nutrient balances between 20 and 30 kg ha⁻¹ per year to be within safe limits for fertiliser management and water protection. According to Toczyński [55], optimal nitrogen balance values in Polish conditions should range from 38.6 to 47.2 kg N ha⁻¹, phosphorus between −1.0 and +4.0 kg P ha⁻¹, and potassium between 8.7 and 13.7 kg K ha⁻¹.

In most of the fertilisation treatments, nitrogen balances were within safe or optimal ranges. Even the highest nitrogen balance (+42.5 kg N ha⁻¹), observed with the highest manure dose (FYM3), did not exceed Toczyński’s upper recommended limit [56].

A three-level water quality risk scale was also developed for phosphorus: balances below 20 kg P ha⁻¹ are considered safe; 20–30 kg P ha⁻¹ represent medium risk; and above 30 kg P ha⁻¹ indicate a high risk of phosphorus accumulation in soil and leaching into surface waters. Most phosphorus balances in our study fell into the low or moderate risk categories (below 20 kg P ha⁻¹). None exceeded the 30 kg P ha⁻¹ threshold, which marks the risk limit for accumulation and leaching. Moreover, many variants, especially in 2023 and 2024 under moderate fertilisation, had phosphorus balances close to the recommended range for Poland (−1.0 to +4.0 kg P ha⁻¹) [55]. This suggests that the applied phosphorus doses fully met plant nutritional needs without excessive accumulation [57].

The largest deviations from reference values were seen with potassium. In most variants and years, potassium balances were clearly negative, especially in manure-fertilised treatments, where a balance of −73.8 kg K ha⁻¹ was recorded in 2023. Only a few cases, mainly in 2024, showed a positive potassium balance. Considering the optimal potassium balance range in Poland is +8.7 to +13.7 kg K ha⁻¹ [56], these deficits likely resulted from both high plant uptake and losses through leaching, especially during wet seasons such as 2023. Such large negative potassium balances can reduce soil potassium reserves, lower yields in subsequent years, and degrade yield quality. Excessive potassium depletion can lead to soil degradation through damage to clay mineral structure, which is rich in potassium [58,59].

While this study provides valuable insights into nitrogen, phosphorus, and potassium balances depending on fertilisation type in permanent meadows, it has certain limitations. The experiment was conducted on mineral soils, limiting direct application of results to other soil types (e.g., organic soils) or climates. Also, balance calculations used a simplified method that did not account for all nutrient loss pathways, such as losses during fertiliser application. Another limitation is the relatively short three-year study period (2022–2024), which, although covering variable weather (Table 3 and Table 4), cannot fully reflect long-term effects of fertilisation on permanent grasslands seen in longer-term experiments, e.g., the 120-year study by Kidd et al. [17]. In addition, the absence of experimental replicates reduces the statistical robustness of the results and should be considered when interpreting the findings. Therefore, future research should extend observation periods and, if possible, include replicated plots to better capture changes in nutrient balances over the long term and refine practical fertilisation recommendations for permanent grasslands

The results of this study highlight several important considerations for fertiliser management in permanent meadows. Monitoring nutrient balances can help adjust fertiliser doses to better match plant requirements, improving nutrient use efficiency and maintaining yield stability. Careful nitrogen management is essential to prevent environmental pollution while ensuring adequate plant nutrition, particularly under variable weather conditions. Phosphorus fertilisation appears adequate in the studied systems, but persistent negative potassium balances indicate the need for supplementary potassium inputs to avoid soil depletion and maintain long-term productivity. Implementing combined fertilisation strategies, such as alternating FYM and mineral NPK, offers a promising approach to balance plant nutrition, support environmental protection, and enhance sustainable grassland management.

5. Conclusions

The type and dose of fertilisation clearly influenced both sward yield and nutrient balances. The highest positive nitrogen and phosphorus balances were observed in the intensively fertilised organic treatment (FYM3) and in the alternate fertilisation treatment (NPK3/FYM3). In contrast, potassium balances were mostly negative, particularly in manure-fertilised plots.

Weather conditions strongly affected N, P, and K balances. In years with favourable rainfall, nutrient uptake was more efficient, resulting in improved and more balanced nutrient budgets. Conversely, periods of water shortage reduced nutrient availability, leading to greater imbalances despite consistent fertilisation practices.

Nitrogen balances remained below levels considered environmentally risky. However, in most cases, they were also below the recommended range for Polish conditions, indicating limited nitrogen uptake under stressful weather.

Phosphorus balances were positive but remained within safe environmental limits, reflecting efficient use of this nutrient.

Potassium balances, especially in FYM-fertilised treatments, were negative. Persistent negative potassium balances may gradually deplete soil reserves unless supplemented by additional mineral potassium inputs.

Among the fertilisation treatments studied, alternate fertilisation—combining FYM and mineral NPK, emerged as a promising approach for improving nutrient use efficiency, increasing dry matter yield, and maintaining more balanced nutrient budgets.

The study emphasises the importance of balanced nutrient management in permanent meadows, showing that moderate fertilisation strategies, such as alternating FYM and mineral NPK, can maintain productivity while reducing environmental impacts. Future research should focus on long-term monitoring to better understand nutrient dynamics and optimize fertilisation practices for sustainable grassland management.