The Effect of Drought on Agronomic and Plant Physiological Characteristics of Cocksfoot (Dactylis glomerata L.) Cultivars

Abstract

Cocksfoot (Dactylis glomerata L.) is becoming increasingly important for grassland farming due to climate change, which alters precipitation and increases droughts. Although it is generally considered to be drought-tolerant, little is known about the differences between cultivars. This study aimed to investigate the effects of four different field capacity (FC) levels (80%, 60%, 40%, and rewetting to 80% after a period of 40% FC) on the yield, crude protein content, water consumption, water use efficiency (WUE), and drought susceptibility index of five European cocksfoot cultivars (cv). A pot experiment was conducted in a greenhouse subjected to the specified irrigation treatments over three growth periods. The results revealed significant differences in the cultivars’ responses to the irrigation treatments. Dry matter yield decreased under simulated drought conditions, while crude protein content and WUE increased. Prolana cv achieved the highest yield under drought conditions, Tandem cv had the highest WUE, and Laban cv exhibited the highest crude protein content. Rewetting to 80% FC in the last growth period resulted in similar dry matter and crude protein yields for all cultivars compared to full irrigation. These findings highlight the importance of selecting and breeding drought-tolerant cocksfoot cultivars to maintain high yields and quality in perennial grassland farming under future climate conditions.

1. Introduction

The growth and productivity of forage crops are influenced by various biotic and abiotic factors [1]—in particular, climate change-induced changes in precipitation distribution and the more frequent occurrence of droughts [2]. The latter poses an increasing challenge, especially in the southern and western parts of Europe [3,4], but now also in the Alpine regions despite their humid climate [5,6]. Drought is a complex natural phenomenon that alters the hydrological balance [7]. Agricultural drought reduces the growth of crops and can result in extreme reductions in yields and, thus, in revenues [7,8]. Especially in alpine regions and other disadvantaged areas, where ley farming and cultivation are limited, permanent grasslands are the primary land type and an essential resource for livestock farming [9]. However, the water demand of permanent grassland is considerably higher compared to arable crops, which makes grassland particularly affected by droughts. In recent years, significant progress has been made in understanding the mechanisms of plant responses to environmental stress, particularly drought [10]. Researchers have identified key physiological and molecular pathways that plants employ to cope with water deficit conditions, such as key genes for an efficient root architecture like those regulating early root growth and nutrient transport [11]; interactions with mycorrhizae [12] molecular mechanisms, which increase stress and drought tolerance [13,14,15]; and transcription factors enhancing water use efficiency [16] as well as drought and salt tolerance [17].

In contrast to most arable crops, permanent grassland mainly consists of a heterogeneous plant community with various species that can be assigned to specific functional groups [18]. Generally, it is a mixture of different grasses with varying amounts of dicotyledonous species, including legumes and non-leguminous forbs. Due to the interaction of the different functional groups, swards of permanent grasslands are stable under both ambient climate and stress conditions such as drought. Compared to monocultures, they are also more productive due to niche complementarity and positive interspecific interactions [19,20,21]. Grasses are generally the functional group with the greatest impact on yield and forage quality, with Lolium, Festuca, Dactylis, and Phleum being the most important genera in temperate regions [22,23,24]. Among the grasses, cocksfoot (Dactylis glomerata L.) belongs to the group of drought-tolerant species. It is generally of lower forage quality than perennial ryegrass (Lolium perenne), but provides high yields even in the summer months and is more tolerant to abiotic stress [25,26,27]. Cocksfoot is native to Eurasia and North Africa [28] and was introduced worldwide by European settlers in the 19th and 20th centuries [29]; its cultivation has become significant around the world [30,31]. Diploid, tetraploid, and some hexaploid populations of cocksfoot occur naturally [27,29]. While differences in the drought-yielding ability of different grass species have been described [32,33], very little literature is available on varietal differences within more drought-tolerant species, especially in Europe.

To address these existing knowledge gaps, we conducted an experiment with five cocksfoot cultivars of different European origins, which were exposed to different field capacity (FC) levels.

The aims of this study were to investigate the effects of three different irrigation level intensities (80%, 60%, and 40% FC) and the effect of rewetting to 80% FC after a period of strongly reduced irrigation at 40% FC, for two growth periods, on dry matter yield (DMY), crude protein (CP) content, water consumption (WC), and water use efficiency (WUE). We hypothesized that varying drought level intensities would significantly affect the yield and qualitative traits, and that these effects would vary significantly among the individual cultivars.

2. Materials and Methods

2.1. Study Site and Plant Material

A potting experiment was set up in the greenhouse at AREC Raumberg-Gumpenstein (47°29 N, 14°06 E), Austria, in 2020/2021. Five different cocksfoot cultivars of European origin were each exposed to four different irrigation levels, with all treatments being repeated six times in randomized order, giving a total number of 120 pots (Figure 1). Cultivar 1 was Amba (Denmark), cultivar 2 was Laban (Norway), cultivar 3 was RGT Lovely (France), cultivar 4 was Prolana (Czech Republic), and cultivar 5 was Tandem (Austria).

The plants were sown on 10 August 2020 in commercial standard soil for propagation, and on 25 August 2020, the plants were transplanted into seed trays filled with standard potting soil. Finally, on 24 September 2020, seven vital seedlings were each repotted in so-called Mitscherlich pots (21 cm high, 20 cm in diameter), each filled with 6500 g of quartz sand (diameter 0.3–1 mm). A covered water drain at the bottom of these containers prevents the roots from penetrating and the substrate from leaking into the collection container below. Following Hesse et al. [34], each pot received a starting fertilization of 0.315 g N, 0.7 g P, 1.1 g K, and 0.3 g Mg. To buffer the pH value in the substrate, each pot received 0.8 g CaCO₃ once at the start of the experiment.

The mixed substrate had a pH value of 7.2, and the bulk density was 1.47 g cm⁻³. The vessels were brought to a uniform weight (±2 g) using additional weights, thus facilitating gravimetric measurements to determine the water consumption. We then irrigated the plants three times a week to avoid water stress and to allow for rapid sprouting. On 29 September 2020, we treated all plants showing rust symptoms with a fungicide containing the chemical active ingredient tebuconazole (active ingredient content 250 gL⁻¹). On 30 September 2020, all pots received 1 g trace element fertilizer (0.5 mg B, 0.3 mg Cu, 0.9 mg Zn, 1.7 mg Mn, 0.2 mg Mo and 5.2 mg Fe) dissolved in 10 mL water. On 25 October 2020, we placed the pots in a darkened room where they were overwintered. The room was not heated, but the temperature was consistently above freezing. The plants remained in their winter quarters between 25 October 2020 and 8 March 2021 and were only watered twice during this time due to the low water consumption.

On 8 March 2021, we put the pots back into the greenhouse and watered them to ensure consistent growth without drought stress. Throughout the experiment, we recorded humidity, temperature, and solar radiation (Figure 2) with an all-in-one weather station (ATMOS 41, Meter Group, Pullman, WA, USA).

In order to achieve a uniform starting situation after the winter, the plants were cut to a height of 5 cm on 11 March 2021, and this day was set as the start of the experiment. We examined three growths during the trial (cutting dates were 13 May 2021, 9 June 2021, and 20 July 2021). At the beginning of each growth, the pots received 0.124 g N, 0.124 g P, and 0.09 g K as liquid fertilizer (this fertilization corresponds to 40 kg N, 40 kg P, and 29 kg K ha⁻¹). All pots were fully irrigated until 4 April 2021, and then the different irrigation levels were applied. A visualization of the experimental course is shown in Figure 3.

2.2. Drought Level Intensities

The water content was determined gravimetrically [35]; the potting substrate was dried at 80 °C to constant weight to determine the dry weight (DW) [36]. We then watered slowly until the substrate was saturated and water emerged from the bottom openings of the pots. To determine the FC [37], we stored the pots for three days, and then the final saturated weight (SW) was determined. We calculated the field capacity as FC = SW − DW [38], and irrigation levels were determined based on this calculation.

• Treatment 1(FC₈₀): irrigation to 80% FC;
• Treatment 2 (FC₆₀): slight lowering of irrigation to 60% FC;
• Treatment 3 (FC₄₀): strong reduction in irrigation to 40% FC;
• Treatment 4 (FC_40rw): Strong reduction in irrigation to 40% FC combined with rewetting after the second cut to 80% FC.

The water contents in the individual pots were determined gravimetrically. We weighted all 120 pots using a balance (SB16001 DeltaRange^®, Vienna, Austria) with an accuracy of ±1 g three to four times per week and then irrigated again to the specified weight. For this, we used a measuring cup (1000 mL) and dispensed the water manually across the entire soil surface.

2.3. Determination of Forage Yield and Quality Parameters

We defoliated the plants at a height of 5 cm and dried them for 24 h in a warm air-drying cabinet at 60 °C for further storage. After the third defoliation, the remaining stubble mass (0–5 cm) was also harvested and dried for 24 h. We milled (ZM 200, Retsch, Haan, Germany) the dried samples (grinding size 1 mm), then determined the residual moisture using near-infrared spectroscopy (SpectraStar™ XT Unity, KPM Analytics, Westborough, MA, USA) and the nitrogen content with an organic elemental analyzer (vario MAX cube, elementar, Langenselbold, Germany). The analyses were carried out according to the methods of Naumann et al. [39], including supplementary sheets (Methods 2.1.1 (1983); 31.3 (2004); 4.1.2 (2004)).

2.4. Determination of Water Consumption and Water Use Efficiency

To determine the water consumption (WC), the water quantities of the individual irrigations of each pot within the individual growths were summed. To determine the water use efficiency (WUE) as a function of total crop biomass to total water consumed [40], the dry matter yields of the individual growths were divided by the sum of the water consumption during the specific growth period. To determine the total dry matter over the entire duration of the experiment, the dry matter of the stubble mass was also included.

2.5. Drought Susceptibility Index (DSI)

A drought susceptibility index (DSI) was calculated according to Fischer and Maurer [41] to evaluate plants’ adaptability to drought. The value of DSI is based on the formula: $\mathrm{D}\mathrm{S}\mathrm{I}=(1-\frac{{\mathrm{Y}}_{\mathrm{d}}}{{\mathrm{Y}}_{\mathrm{n}}})/(1-\frac{{\mathrm{X}}_d}{{\mathrm{X}}_n}),$ where Y_d = the average DM yield of the specific cultivars under reduced irrigation; Y_n = the average DM yield of the specific cultivars in FC₈₀; X_d = the average DM yield of all cultivars under reduced irrigation; and X_n = the average DM yield of all cultivars in FC₈₀. This allows the cultivars to be classified into drought stress-susceptible (DSI > 1) and drought stress-resistant (DSI < 1) [42].

2.6. Statistics

We performed all the statistical analysis with SAS (Version 9.4; SAS Institute Inc., Cary, NC, USA). As the pot experiment followed a split-plot-design, we used the following model to analyze the effects of irrigation, cultivar, growth, and their interactions on the response variables DMY, CP content, WC, and WUE:

y_ijkht = µ_t + α_it + β_jt + γ_kt +(αβ)_ijt + (αγ)_ikt + (βγ)_jkt + (αβγ)_ijkt + b_ht + f_ih + e_ijhkt

y_ijhkt stands for the response variables of the i^th irrigation, the j^th cultivar, and the k^th growth in the h^th block; µ_t is a general intercept; α_it is the main effect of i^th irrigation; β_jt is the main effect of the j^th cultivar; γ_kt is the main effect of the k^th growth; (αβ)_ijt, (αγ)_ikt, (βγ)_jkt, and (αβγ)_ijkt are the interaction effects; f_hi is the main-plot error associated with the h^th block and i^th irrigation level, assumed to be random with zero mean and variance σ²; and e_ijhkt is the residual plot error associated with y_ijhkt. Because we repeatedly measured on the same experimental unit in time, we took care of the correlation of the longitudinal data. Subscript t represents this time-specific effect of the k^th growth associated with the parcels that are unique combinations of the h-th block, i-th irrigation, and j-th cultivar. Based on AIC, we used the compound symmetry model for the covariance structure for the response variable, water use. For all other response variables, we used the unstructured covariance model. A detailed description of the covariance models which we utilized can be found in the documentation of the GLIMMIX procedure (Version 9.4; SAS Institute Inc.).

We used the Keward–Roger approximation of degrees of freedom and the residual maximum likelihood method (REML) to estimate the covariance parameter. We calculated the adjusted means of the treatments with LSMEANS and used Tukey–Kramer tests to determine significant differences. We created the plots using R-studio (Version 1.1.463, R Foundation for Statistical Computing, Vienna, Austria).

3. Results

3.1. Dry Matter Yield and Drought Susceptibility Index

Our study overall showed significant influences of cultivar, irrigation treatment, growth, and the interaction of cultivar and irrigation (p < 0.05) on the DMY.

3.1.1. Drought Intensity Levels

The study indicated (

Table 1. Dry matter yield (g DM m⁻²) of cocksfoot cultivars as a function of irrigation treatments in different growths (arithmetic mean ± standard deviation).

1st Growth	FC80	FC60	FC40	Mean
Amba	288.5 ± 35.2 a e	278.4 ± 46.5 a ef	303.8 ± 28.6 a e	289.7 ± 36.8 FG
Laban	266.1 ± 39.6 a e	248.2 ± 57.0 a f	253.3 ± 15.5 a ef	254.6 ± 39.3 G
RGT Lovely	303.5 ± 46.1 a e	291.6 ± 51.0 ab ef	239.3 ± 42.2 b f	276.9 ± 52.3 FG
Prolana	287.3 ± 30.3 a e	332.0 ± 31.5 a e	303.9 ± 59.5 a e	308.8 ± 44.3 F
Tandem	300.7 ± 52.7 a e	262.3 ± 22.2 a f	252.0 ± 28.4 a ef	270.6 ± 40.8 FG
Mean	289.7 ± 40.8 A	280.43± 50.1 A	273.7 ± 44.4 A
2nd Growth	FC80	FC60	FC40	Mean
Amba	172.1 ± 4.7 a ef	159.3 ± 8.6 a ef	157.2 ± 15.9 a e	162.3 ± 12.2 FG
Laban	156.1 ± 16.3 a f	143.7 ± 23.6 a f	149.9 ± 12.2 a e	149.6 ± 17.7 G
RGT Lovely	197.6 ± 27.4 a e	182.1 ± 22.9 ab e	164.8 ± 12.4 b e	181.4 ± 24.7 F
Prolana	178.6 ± 23.9 a ef	187.2 ± 41.0 a e	167.3 ± 28.4 a e	178.3 ± 31.2 F
Tandem	180.1 ± 28.2 a ef	154.6 ± 16.9 a ef	156.4 ± 8.8 a e	162.3 ± 22.0 FG
Mean	178.3 ± 24.5 A	165.5 ± 28.7 AB	159.2 ± 17.0 B
3rd Growth	FC80	FC60	FC40	Mean
Amba	189.2 ± 25.8 a e	165.4 ± 11.8 ab e	158.4 ± 20.7 b e	171.9 ± 23.4 FG
Laban	166.1 ± 21.5 a e	160.6 ± 15.1 a e	146.0 ± 24.0 a e	156.0 ± 21.2 G
RGT Lovely	191.6 ± 31.4 a e	180.1 ± 27.6 a e	164.7 ± 17.8 a e	178.3 ± 27.1 F
Prolana	176.3 ± 15.7 a e	168.3 ± 18.7 a e	161.8 ± 9.8 a e	168.7 ± 15.5 FG
Tandem	185.5 ± 14.4 a e	168.7 ± 22.6 ab e	155.1 ± 12.8 b e	168.7 ± 20.6 FG
Mean	181.4 ± 23.1 A	168.7 ± 19.6 B	156.0 ± 17.8 B
Sum of growths (total yield)	FC80	FC60	FC40	Mean
Amba	649.9 ± 34.3 aef	603.1 ± 63.9 a ef	619.3 ± 60.0 a e	624.1 ± 54.8 FG
Laban	588.2 ± 65.5 a f	552.4 ± 85.7 af	549.2 ± 29.5 a e	563.3 ± 63.3 H
RGT Lovely	692.6 ± 102.0 a e	653.8 ± 83.0 ab ef	568.8 ± 57.5 b e	638.4 ± 94.25FG
Prolana	642.2 ± 50.8 a ef	687.5 ± 64.6 a e	632.9 ± 81.5 a e	654.2 ± 67.4 F
Tandem	666.4 ± 85.0 a ef	575.3 ± 57.6 b f	573.8 ± 30.2 b e	605.2 ± 73.2 GH
Mean	647.9 ± 75.2 A	614.4 ± 83.7 B	588.8 ± 75.2 B

Index a,b indicates the differences between treatments within a particular cultivar and growth. Index A,B indicates the differences between treatment means within a particular growth. Index e,f indicates the differences between cultivars within a particular treatment and growth. Index F–H indicates the differences between cultivar means within a particular growth.

) that the highest yield across all growths and cultivars was recorded in FC₈₀, followed by FC₆₀ and FC₄₀. Regardless of cultivar or treatment, the first growth produced the highest yield. With regard to the cultivars, Prolana achieved the highest yield on average across all growths and treatments, with this yield differing significantly only from Tandem and Laban. Laban recorded the lowest yield on average across all growths and treatments.

In the first growth, Prolana achieved the highest average yield, and Laban achieved the lowest average yield. It is noteworthy that Prolana tended to achieve the highest yield in FC₆₀, and the decrease in yield from FC₈₀ to FC₄₀ was only significant for RGT Lovely.

In the second growth, RGT Lovely and Prolana achieved the highest average yield, and Laban achieved the lowest average yield. While RGT Lovely achieved the highest yield in FC₈₀, Prolana again achieved the highest yield in FC₆₀. In this growth, the decline in yield from FC₈₀ to FC₄₀ was also only significant for RGT Lovely. In the third growth, RGT Lovely achieved the highest yield on average. Although the decline in yield from FC₈₀ to FC₄₀ was noticeable for all cultivars, it was only significant for the cultivars Tandem and Amba.

Regarding the treatments, it was evident that the cultivars RGT Lovely and Tandem showed promising results in all growths under FC₈₀ and were always markedly above the treatment average. In FC₆₀, Prolana achieved the highest yield, although this was only significantly higher compared to Laban. In FC₄₀, Prolana and Amba achieved higher yields than the treatment mean over the sum of all growths, with only Prolana exceeding the treatment mean in every single growth.

The DSI (Figure 4) also showed major differences between the cultivars. In both drought treatments, the cultivar Prolana showed the lowest values (<1). The negative value in FC₆₀ resulted from the higher DM yield in this treatment compared to FC₈₀. Furthermore, the cultivars Amba and Laban also showed low values in FC₄₀ (<1).

3.1.2. Rewetting

Rewetting after the second cut (FC_40rw) did not lead to significantly higher DMY for any of the cultivars in the subsequent growth compared to FC₈₀ (

Table 2. Dry matter yield in the 3rd growth (g DM m⁻²) of cocksfoot cultivars after rewetting. (arithmetic mean ± standard deviation).

Treatment	FC80	FC40	FC40rw	Mean
Amba	189.2 ± 25.8 a	158.4 ± 20.7 b	183.3 ± 13.1 a	178.3 ± 23.7 F
Laban	166.1 ± 21.5 a	146.0 ± 24.0 a	163.6 ± 7.6 a	159.2 ± 20.2 G
RGT Lovely	191.6 ± 31.4 a	164.7 ± 17.8 b	177.3 ± 13.7 ab	178.3 ± 23.8 F
Prolana	176.3 ± 15.7 a	161.8 ± 9.8 a	177.5 ± 11.5 a	171.9 ± 13.9 FG
Tandem	185.5 ± 14.4 a	155.1 ± 12.8 b	181.6 ± 20.5 a	175.1 ± 20.7 FG
Mean	181.4 ± 23.1 A	156.0 ± 17.8 B	178.3 ± 14.7 A

Index a,b indicates the differences between treatments within a particular cultivar. Index A,B indicates the differences between treatment means. Index F,G indicates the differences between cultivar means.

). Compared to FC₄₀ (without rewetting), an increased yield was observed for all cultivars, but this was only significant for Amba, Laban, and Tandem, as well as the average of all cultivars.

3.2. Crude Protein

Our study showed significant influences of cultivar; irrigation treatment growth; and interactions of cultivar × irrigation, cultivar × growth, and growth × irrigation (p < 0.05) on crude protein content.

3.2.1. Drought Intensity Levels

On average, the highest crude protein (CP) content across all growths and cultivars was recorded in FC₄₀, followed by FC₆₀ and FC₈₀. However, the difference between FC₄₀ and FC₈₀ was only significant in the first two growths (

Table 3. Crude protein content (g kg⁻¹) of cocksfoot cultivars as a function of irrigation treatments in different growths (arithmetic mean ± standard deviation).

1st Growth	FC80	FC60	FC40	Mean
Amba	151.0 ± 23.0 ab e	145.6 ± 14.5 b g	172.6 ± 16.0 a ef	156.4 ± 20.9 GH
Laban	168.4 ± 12.8 a e	180.3 ± 25.9 a ef	183.9 ± 23.4 a e	177.5 ± 21.3 F
RGT Lovely	144.9 ± 21.6 a e	151.9 ± 17.1 a fg	161.4 ± 14.0 a ef	152.7 ± 18.1 H
Prolana	150.9 ± 17.8 a e	129.8 ± 13.8 ag	149.5 ± 16.4 a f	143.4 ± 18.1 H
Tandem	155.6 ± 16.6 b e	183.7 ± 18.4 a e	180.5 ± 28.1 ab e	173.3 ± 24.1 FG
Mean	154.2 ± 19.1 B	158.3 ± 27.1 AB	169.6 ± 22.8 A
2nd Growth	FC80	FC60	FC40	Mean
Amba	153.2 ± 13.7 a e	144.2 ± 4.9 a ef	144.7 ± 10.9 a g	147.4 ± 10.7 G
Laban	152.6 ± 11.5 b e	159.4 ± 18.7 b e	179.6 ± 10.8 a e	163.8 ± 17.8 F
RGT Lovely	132.4 ± 14.0 b f	136.0 ± 13.6 b f	155.0 ± 12.4 a fg	141.1 ± 16.2 G
Prolana	141.3 ± 15.3 a ef	135.6 ± 4.2 a f	148.8 ± 8.8 a g	141.9 ± 11.3 G
Tandem	152.3 ± 10.0 b e	158.3 ± 12.9 ab e	170.2 ± 14.6 a ef	160.2 ± 14.1 F
Mean	146.3 ± 14.7 B	146.7 ± 15.5 B	159.6 ± 17.2 A
3rd Growth	FC80	FC60	FC40	Mean
Amba	127.3 ± 11.0 a fg	136.9 ± 14.8 a ef	134.2 ± 9.9 a e	132.8 ± 12.1 FG
Laban	144.4 ± 10.5 a e	139.8 ± 12.3 a e	139.0 ± 10.9 a e	141.0 ± 10.9 F
RGT Lovely	123.8 ± 7.6 a g	122.7 ± 12.6 a f	128.4 ± 5.9 a e	125.0 ± 9.0 G
Prolana	141.3 ± 10.1 a ef	128.9 ± 4.4 a ef	132.2 ± 5.9 a e	134.1 ± 8.7 FG
Tandem	133.6 ± 6.3 a efg	137.7 ± 12.2 a ef	138.5 ± 11.7 a e	136.6 ± 10.1 F
Mean	134.1 ± 11.8 A	133.2 ± 12.8 A	134.4 ± 9.5 A
Average of all growths	FC80	FC60	FC40	Mean
Amba	143.8 ± a ef	142.2 ± a f	150.5 ± a fg	145.5 ± 17.9 G
Laban	155.1 ± b e	159.8 ± ab e	167.5 ± a e	160.8 ± 22.7 F
RGT Lovely	133.7 ± b f	136.8 ± ab f	148.3 ± a g	139.6 ± 18.6 H
Prolana	144.5 ± a ef	131.4 ± b f	143.5 ± a g	139.8 ± 13.7 H
Tandem	147.2 ± b e	159.9 ± a e	163.1 ± a ef	156.7 ± 22.8 F
Mean	144.9 ± 17.4 B	146.0 ± 21.9 B	154.6 ± 22.8 A

Index a,b indicates the differences between treatments within a particular cultivar and growth. Index A,B indicates the differences between treatment means within a particular growth. Index e–g indicates the differences between cultivars within a particular treatment and growth. Index F–H indicates the differences between cultivar means within a particular growth.

). The CP content was highest in the first growth and decreased in all cultivars on average across the treatments until the third growth.

With regard to the cultivars, Laban showed the highest CP content in the mean of the treatments in each growth, while RGT Lovely and Prolana had the lowest CP contents in the mean of the treatments and growths. In the first growth, the CP contents of the cultivars were relatively close to each other in FC₈₀, whereas they differed significantly in FC₄₀ and especially in FC₆₀. This effect was also maintained in the second growth.

Looking at the individual growths, very high CP contents of over 180 g kg⁻¹ could be observed in the first growth for the cultivars Tandem and Laban in FC₆₀ and FC₄₀. These two cultivars also had the highest CP contents in the second growth, with Laban tending to be higher in FC₄₀ in particular. They also performed well in the third growth period, with Tandem tending to fall behind in FC₈₀.

In terms of crude protein yield, the treatments showed little influence on the cultivars, and the differences between the cultivars were also minor. A detailed analysis of the results is shown in Tables S1 and S2 in the Supplementary Materials.

3.2.2. Rewetting

The effects of rewetting on the CP content are shown in

Table 4. Crude protein content in the 3rd growth. (g kg⁻¹) of cocksfoot cultivars as a function of rewetting (arithmetic mean ± standard deviation).

Treatment	FC80	FC40	FC40rw	Mean
Amba	127.3 ± 11.0 a fg	134.2 ± 9.9 a e	124.5 ± 9.9 a ef	128.7 ± 10.5 G
Laban	144.4 ± 10.5 a e	139.0 ± 10.9 a e	139.9 ± 11.0 a e	141.1 ± 10.4 F
RGT Lovely	123.8 ± 7.6 a g	128.4 ± 5.9 a e	125.1 ± 13.0 a ef	125.8 ± 9.0 G
Prolana	141.3 ± 10.1 a ef	132.2 ± 5.9 ab e	127.2 ± 9.1 b ef	133.6 ± 10.0 FG
Tandem	133.6 ± 6.3 ab efg	138.5 ± 11.7 a e	121.8 ± 9.6 b f	131.3 ± 11.5 G
Mean	134.1 ± 11.8 A	134.5 ± 9.5 A	127.7 ± 11.8 B

Index a,b indicates the differences between treatments within a particular cultivar. Index e–g stands for the differences between cultivars within a particular treatment. Index A,B indicates the differences between treatment means. Index F,G indicates the differences between cultivar means.

. While the CP contents of the cultivars hardly differed between FC₈₀ and FC₄₀, they were significantly lower on average in FC_40rw, especially for the cultivars Prolana and Tandem. Laban achieved the highest CP content in all treatments and was also the only cultivar with a higher CP content in FC_40rw than in FC₄₀ without rewetting.

3.3. Water Consumption

Our study showed significant influences of irrigation treatment and growth (p < 0.05) on water consumption.

3.3.1. Drought Intensity Levels

The highest water consumption across all growths and cultivars was observed in FC₈₀, followed by FC₆₀ and FC₄₀ (

Table 5. Water consumption (L m⁻²) of cocksfoot cultivars in relation to irrigation treatments from 1st to 3rd growth (arithmetic mean ± standard deviation).

1st Growth	FC80	FC60	FC40	Mean
Amba	278.3 ± 15.9 a e	228.5 ± 53.2 b f	207.9 ± 11.0 b e	238.2 ± 43.2 FG
Laban	237.0 ± 31.4 a f	204.8 ± 30.3 b fg	172.3 ± 46.1 c f	204.7 ± 43.9 I
RGT Lovely	260.6 ± 41.2 a ef	237.1 ± 23.2 a f	183.9 ± 26.5 b ef	227.2 ± 44.2 GH
Prolana	270.8 ± 27.4 a ef	273.7 ± 24.1 a e	216.1 ± 15.4 b e	253.5 ± 34.7 F
Tandem	246.3 ± 37.5 aef	191.8 ± 43.8 b g	184.4 ± 41.9 b ef	207.5 ± 48.0 HI
Mean	258.6 ± 33.4 A	227.2 ± 44.7 B	192.9 ± 33.6 C
2nd Growth	FC80	FC60	FC40	Mean
Amba	125.6 ± 11.4 a e	105.9 ± 10.7 ab e	95.2 ± 4.6 b e	108.9 ± 15.7 FG
Laban	106.2 ± 14.3 a e	95.2 ± 13.6 a e	89.5 ± 12.0 a e	97.0 ± 14.4 FG
RGT Lovely	130.8 ± 12.1 a e	119.3 ± 3.9 ab e	96.2 ± 5.1 b e	115.4 ± 16.6 F
Prolana	125.6 ± 13.9 a e	120.6 ± 12.1 a e	99.0 ± 4.1 a e	115.1 ± 15.7 FG
Tandem	107.3 ± 9.9 a e	95.0 ± 13.4 a e	84.5 ± 9.0 a e	95.6 ± 14.1 G
Mean	119.1 ± 15.6 A	107.2 ± 15.5 A	92.9 ± 8.8 B
3rd Growth	FC80	FC60	FC40	Mean
Amba	208.7 ± 16.5 a e	181.8 ± 10.5 ab e	153.5 ± 8.1 b e	181.4 ± 25.9 FG
Laban	182.3 ± 17.6 a e	166.3 ± 19.5 ab e	147.4 ± 5.5 b e	165.4 ± 20.7 G
RGT Lovely	206.0 ± 9.7 a e	186.2 ± 9.4 a e	152.6 ± 7.6 b e	181.6 ± 24.2 FG
Prolana	212.4 ± 10.9 a e	195.8 ± 14.2 a e	160.4 ± 5.4 b e	189.5 ± 24.5 F
Tandem	189.1 ± 12.0 a e	165.1 ± 14.1 ab e	143.7 ± 13.0 b e	166.0 ± 22.7 G
Mean	199.7 ± 17.5 A	179.0 ± 17.7 B	151.5 ± 9.7 C
Sum of all growths	FC80	FC60	FC40	Mean
AMBA	612.6 ± 40.6 a e	516.1 ± 71.4 b efg	456.6 ± 21.2 b e	528. 5 ± 80.6 FG
LABAN	525.5 ± 53.2 a f	466.4 ± 49.9 ab fg	409.2 ± 61.3 b e	467.0 ± 71.1 H
RGT LOVELY	597.4 ± 61.4 a ef	542.6 ± 26.2 a ef	432.7 ± 36.9 b e	524.2 ± 81.7 G
PROLANA	608.8 ± 37.5 a e	590.1 ± 41.5 a e	475.5 ± 16.0 b e	558.2 ± 68.4 F
TANDEM	542.7 ± 48.5 a ef	451.9 ± 66.6 b g	412.6 ± 61.7 b e	469. 1 ± 79.1 H
Mean	577.4 ± 58.5 A	513.4 ± 71.6 B	437.3 ± 48.3 C

Index a,b indicates the differences between treatments within a particular cultivar and growth. Index A–C indicates the differences between treatment means within a particular growth. Index e–g indicates the differences between cultivars within a particular treatment and growth. Index F–H indicates the differences between cultivar means within a particular growth.

). WC peaked in the first growth and was second highest in the third growth. Looking at the individual cultivars, Prolana showed the highest water consumption over the sum of the growths, especially in growths 1 and 3, although the differences from the other cultivars were not consistently significant. In contrast, the cultivars Laban and Tandem consistently showed significantly lower water consumption in all growths.

In the first growth, the water consumption of all cultivars in FC₄₀ and FC₆₀ (except Prolana and RGT Lovely) was significantly lower than in FC₈₀. In the second growth, FC₆₀ and FC₈₀ did not differ from each other in any cultivar, and only the strong reduction in irrigation (FC₄₀) resulted in lower water consumption than full irrigation (FC₈₀) for Amba and Lovely. For water consumption, there were only significant differences in the irrigation treatments during the first growth.

Considering the average of all growths, it can be seen that for all cultivars, water consumption was higher in FC₈₀ than in FC₆₀ (significant for cultivars Amba and Tandem) and significantly higher in FC₈₀ than in FC_40. Looking at the average WC of the cultivars across all growths and treatments, the cultivar Prolana had the highest and the cultivars Laban and Tandem had the lowest WC.

3.3.2. Rewetting

In terms of rewetting (

Table 6. Water consumption (L m⁻²) in the 3rd growth of cocksfoot cultivars as a function of rewetting (arithmetic mean ± standard deviation).

Treatment	FC80	FC40	FC40rw	Mean
Amba	208.7 ± 16.5 a	153.5 ± 8.1 b	199.2 ± 9.3 a	187.1 ±27.2
Laban	182.3 ± 17.6 a	147.4 ± 5.5 b	202.1 ± 16.1 a	177.3 ± 26.8
RGT Lovely	206.0 ± 9.7 a	152.6 ± 7.6 b	202.2 ± 4.5 a	186.9 ± 26.0
Prolana	212.4 ± 10.9 a	160.4 ± 5.4 b	215.0 ± 11.0 a	196.0 ± 27.3
Tandem	189.1 ± 12.0 a	143.7 ± 13.0 b	196.2 ± 7.6 a	176.3 ± 26.1
Mean	199.7 ± 17.5 A	151.5 ± 9.7 B	203.0 ± 11.7 A

Index a,b stands for the differences between treatments within a particular cultivar. Index A,B indicates the differences between treatment means.

), all cultivars in FC_40rw showed significantly higher water use compared to FC₄₀, although no differences were found between cultivars, and there were also no differences between FC_40rw and FC₈₀.

3.4. Water Use Efficiency (WUE)

Our study showed significant influences of cultivar, irrigation treatment, and growth (p < 0.05) on water use efficiency.

3.4.1. Drought Intensity Levels

Concerning the water use efficiency, no significant differences were found between the cultivars within any treatment or growth (

Table 7. Water use efficiency (g DM L⁻¹ H2O) of cocksfoot cultivars in relation to irrigation treatments from 1st to 3rd growth (arithmetic mean ± standard deviation).

1st Growth	FC80	FC60	FC40	Mean
Amba	1.040 ± 0.14 b	1.253 ± 0.20 ab	1.459 ± 0.08 a	1.251 ± 0.23 F
Laban	1.124 ± 0.10 b	1.227 ± 0.31 ab	1.588 ± 0.53 a	1.313 ± 0.4 F
RGT Lovely	1.180 ± 0.21 a	1.225 ± 0.12 a	1.356 ± 0.49 a	1.254 ± 0.31 F
Prolana	1.069 ± 0.14 a	1.216 ± 0.10 a	1.411 ± 0.28 a	1.232 ± 0.23 F
Tandem	1.234 ± 0.22 a	1.379 ± 0.36 a	1.466 ± 0.26 a	1.36 ± 0.29 F
Mean	1.129 ± 0.17 A	1.26 ± 0.23 A	1.456 ± 0.35 B
2nd Growth	FC80	FC60	FC40	Mean
Amba	1.381 ± 0.15 a	1.513 ± 0.11 a	1.653 ± 0.18 a	1.516 ± 0.18 G
Laban	1.486 ± 0.21 a	1.525 ± 0.27 a	1.710 ± 0.35 a	1.574 ± 0.28 FG
RGT Lovely	1.508 ± 0.13 a	1.524 ± 0.16 a	1.717 ± 0.15 a	1.583 ± 0.17 FG
Prolana	1.426 ± 0.14 a	1.551 ± 0.27 a	1.686 ± 0.24 a	1.554 ± 0.24 FG
Tandem	1.698 ± 0.36 a	1.653 ± 0.28 a	1.873 ± 0.25 a	1.741 ± 0.3 F
Mean	1.5 ± 0.23 AB	1.553 ± 0.22 A	1.728 ± 0.24 B
3rd Growth	FC80	FC60	FC40	Mean
Amba	0.911 ± 0.15 a	0.910 ± 0.05 a	1.036 ± 0.16 a	0.9525 ± 0.13 FG
Laban	0.914 ± 0.11 a	0.981 ± 0.17 a	0.993 ± 0.18 a	0.9624 ± 0.15 FG
RGT Lovely	0.928 ± 0.13 a	0.965 ± 0.13 a	1.082 ± 0.13 a	0.9918 ± 0.14 FG
Prolana	0.833 ± 0.10 b	0.862 ± 0.10 ab	1.009 ± 0.07 a	0.9016 ± 0.12 G
Tandem	0.985 ± 0.11 a	1.025 ± 0.13 a	1.084 ± 0.11 a	1.031 ± 0.12 F
Mean	0.914 ± 0.12 A	0.949 ± 0.13 A	1.041 ± 0.13 B
1st–3rd growth	FC80	FC60	FC40	Mean
Amba	1.066 ± 0.101 b	1.176 ± 0.086 ab	1.356 ± 0.121 a	1.199 ± 0.16 G
Laban	1.122 ± 0.102 b	1.195 ± 0.225 ab	1.375 ± 0.267 a	1.23 ± 0.23 FG
RGT Lovely	1.161 ± 0.144 a	1.202 ± 0.112 a	1.330 ± 0.239 a	1.231 ± 0.18 FG
Prolana	1.056 ± 0.073 b	1.169 ± 0.123 ab	1.332 ± 0.177 a	1.186 ± 0.17 G
Tandem	1.235 ± 0.174 a	1.294 ± 0.206 a	1.409 ± 0.152 a	1.313 ± 0.18 F
Mean	1.128 ± 0.13 A	1.207 ± 0.16 A	1.36 ± 0.19 B

Index a,b indicates the differences between treatments within a particular cultivar and growth. Index A,B indicates the differences between treatment means within a particular growth. Index F,G indicates the differences between cultivar means within a particular growth.

). However, considering the average of the treatments, the cultivar Tandem always showed the highest WUE, although the differences from the other cultivars were not always significant. Across all growths, Tandem again showed the highest WUE, although the differences were only significant compared to the cultivars Prolana and Amba. In addition, the results indicated a continuous increase in water use efficiency (WUE) from FC₈₀ to FC₆₀ and further to FC₄₀ across all growths. With regard to the individual growths, the WUE was highest in the second growth and lowest in the third growth for all cultivars.

Furthermore, there were differences between treatments within some cultivars. In the first growth, Amba and Laban, and in the third growth, Prolana had significantly higher WUE values in FC₄₀ than FC₈₀. Considering the average of all growths, all cultivars showed the highest WUE in FC₄₀. Looking at the individual treatments, no difference could be found between the cultivars.

3.4.2. Rewetting

With regard to rewetting (

Table 8. Water use efficiency (g DM L⁻¹ H2O) in the 3rd growth of cocksfoot cultivars as a function of rewetting in the 3rd growth (arithmetic mean ± standard deviation).

Treatment	FC80	FC40	FC40rw	Mean
Amba	0.911 ± 0.137 a	1.036 ± 0.163 a	0.920 ± 0.036 a	0.956 ± 0.13 FG
Laban	0.914 ± 0.111 ab	0.993 ± 0.182 a	0.812 ± 0.043 b	0.906 ± 0.14 FG
RGT Lovely	0.928 ± 0.129 b	1.082 ± 0.128 a	0.877 ± 0.070 b	0.963 ± 0.014 FG
Prolana	0.833 ± 0.101 b	1.009 ± 0.075 a	0.826 ± 0.046 b	0.889 ± 0.11 G
Tandem	0.985 ± 0.108 ab	1.084 ± 0.113 a	0.925 ± 0.097 b	0.998 ± 0.12 F
Mean	0.914 ± 0.12 A	1.041 ± 0.13 B	0.872 ± 0.07 A

Index a,b stands for the differences between treatments within a particular cultivar. Index A,B indicates the differences between treatment means. Index F,G indicates the differences between cultivar means.

), there were no differences in WUE between FC₈₀ and FC_40rw in the mean of the cultivars; in FC_40, it was significantly higher compared to FC_40rw in all cultivars with the exception of Amba.

4. Discussion

4.1. Dry Matter Yield and Drought Susceptibility Index

4.1.1. Drought Intensity Levels

Our study demonstrates significant variability in the response of cocksfoot cultivars to drought treatments concerning dry matter yield across different growth periods. In the initial two growth periods, RGT Lovely showed a significantly higher yield under full irrigation (FC₈₀) compared to severe drought conditions (FC₄₀). However, this pattern was observed for Amba and Tandem only in the third growth period. When considering the entire year, a general trend of yield reduction under drier conditions was evident for all cultivars, consistent with previous findings in the literature for cocksfoot [1,43,44] and various perennial grasses [35,45]. This decrease in yield was particularly significant for the cultivars RGT Lovely and Tandem. These observations align with the results from Bristiel et al. [46], indicating that drought does not significantly affect all cultivars. The drought susceptibility index (DSI) further highlighted the differences in drought tolerance among the tested cultivars. Prolana emerged as the most drought-resistant cultivar in terms of yield, likely due to its adaptation to a continental climate, which inherently involves more variable and often lower precipitation levels. Similarly, Amba and Laban demonstrated resilience to drought stress. Conversely, Tandem and RGT Lovely, despite their high yields under optimal irrigation conditions (FC₈₀), were more susceptible to drought stress. This susceptibility against drought can be attributed to their origins in Alpine and Atlantic regions, which typically receive higher precipitation.

4.1.2. Rewetting

Rewetting during the third growth increased DMY for all cultivars compared to FC₄₀, although this was only significant for Amba and Tandem. However, no cultivar showed an increase in yield compared to FC₈₀. This contrasts with the results of Korup et al. [47], where cocksfoot and reed canary grass cultivars partially achieved higher yields after the drought compared to the control treatment.

4.2. Crude Protein Content

4.2.1. Drought Intensity Levels

Our findings reveal that the crude protein (CP) content generally increased under severe drought conditions (FC₄₀) compared to full irrigation (FC₈₀) across most cultivars, except for Prolana. This trend aligns with previous studies reporting similar increases in CP content under drought stress in cocksfoot as well as other forage grasses, such as meadow fescue (Festuca pratensis), festulolium, timothy (Phleum pratense), and Italian ryegrass (Lolium multiflorum) [35,48,49,50]. The increase in CP content under drought conditions can most likely be attributed to changes in the leaf-to-stem ratio [51,52] or the concentration effect due to reduced biomass production [53], which is a common response among forage species under water stress [48,54]. Also, the results of Küsters et al. [55] from studies with the cocksfoot cultivar Tandem showed an increased CP content in drought experiments.

Significant differences concerning CP content were observed between the cultivars in the individual growths and treatments. However, the cultivar Laban showed the highest CP content in all treatments and growths (except growth 1 FC₆₀). This suggests that Laban may possess inherent genetic traits that confer higher protein accumulation under varying water conditions. In contrast, Prolana consistently showed lower CP content under FC₆₀ compared to FC₈₀ and FC₄₀, although this difference was only significant when considering the entire year. While there were significant differences in protein content between cultivars and treatments, this was not the case for crude protein yields throughout the year, as the falling DMY and the rising CP contents neutralized each other.

4.2.2. Rewetting

Our study showed that rewetting had varying effects on CP content and yield among the cocksfoot cultivars. Generally, the CP content decreased in all cultivars except Laban when rewetting was applied compared to FC₄₀. Despite the decrease in CP content, the protein yield tended to increase across all cultivars, driven primarily by an increase in DMY.

The effects of rewetting on the CP content and yields were mixed when compared to the fully irrigated treatment (FC₈₀). Most cultivars exhibited similar CP contents and yields under rewetting and full irrigation, except for Tandem and Prolana. In the case of Prolana, the lower CP content observed under rewetting (FC40rw) can be explained by the higher biomass production. The dilution effect [53], by which a higher biomass yield leads to lower concentration of CP, likely contributed to this outcome.

4.3. Water Consumption and Water Use Efficiency

The investigated cultivars showed WUE values in the range of 1.1–1.4 g DM H₂O L⁻¹, which aligns with the results of Lelièvre et al. [56], who found WUE values between 1 and 2 g DM H₂O L⁻¹ for various cocksfoot cultivars. Our study found no significant differences in WUE among the cocksfoot cultivars under either full or limited irrigation. This aligns with the results of Norton et al. [57], who also found no significant differences in the WUE values of cocksfoot cultivars under drought conditions. However, when averaging WUE over the treatments across the second and third growth periods and calculating the overall average of all growths, significant differences emerged. Specifically, the cultivar Tandem exhibited a significantly higher WUE than Amba and Prolana.

In the first growth, significantly lower water use efficiency at FC₈₀ was observed compared to FC₄₀ for Amba. This can be attributed to the lower biomass production at FC₈₀ despite significantly higher water consumption. Similarly, the cultivar Laban exhibited significantly lower WUE in FC₈₀ compared to FC40, where the DM yield was higher at FC₈₀, but the water consumption was 37% lower.

In the second growth, despite partially significant differences in water consumption, there were no differences in WUE between treatments for any cultivar, as the DM yield increased with increasing water consumption. In the third growth, the cultivar Prolana exhibited a significantly higher WUE in FC₄₀ than in FC₈₀, which can be explained by the significantly lower water consumption and the only slightly reduced DM yield.

Considering all growths, it can be seen that the WUE of all cultivars is always higher in FC₄₀ than in FC₆₀ and higher in FC₆₀ than in FC₈₀. The results in the literature show varying trajectories here, while our results are consistent with those of [47], who also reported higher WUE for various grass species and cultivars under limited irrigation compared to full irrigation. In contrast, the results of Bahrani et al. [58] showed an opposite trend, as WUE decreased with lower irrigation. Considering the influence of rewetting, it can be found that the WUE in FC_40rw was significantly lower than in FC₄₀, and also tended to be lower than in FC₈₀.

5. Conclusions

The results of our study showed substantial effects of irrigation treatments on dry matter yield, forage quality parameters, and WUE. While the dry matter yield decreased under simulated drought, the crude protein content and WUE increased, confirming that irrigation level intensities significantly affected yield and qualitative traits, although the differences were not significant in all cases. Rewetting after the second cut resulted in similar dry matter and crude protein yields for all cultivars, as observed under full irrigation. The cultivars reacted differently in terms of dry matter yield and crude protein concentrations at different irrigation levels. In terms of WUE, no significant differences occurred among cultivars within a treatment, but there were significant differences on average across all treatments. The different responses of cocksfoot cultivars to drought emphasize the importance of breeding and using cocksfoot in perennial and permanent grassland. The ability of cocksfoot to utilize water efficiently and ensure high yields and good forage quality even under difficult growing conditions, such as drought, makes it a promising candidate for breeding to meet the challenges of climate change in grassland farming.