Introduction

In arid and semi-arid zones, climate change is marked by an increase in temperatures and extreme weather events, such as floods and longer, more severe droughts (Cavin et al. 2013). In these areas, water scarcity is one of the main limiting factors for forage production (Rad et al. 2021; Rostamza et al. 2011) and this leads to a decrease in growth, development, and thus dry matter yield of several forage species (Staniak 2016). Tunisia, like other Mediterranean countries, is particularly affected by irregular rainfall and limited water resources (Fathalli et al. 2023), and forage crops are subject to severe sequences of water deficit. According to (Deandreis et al. 2021) a study conducted in Tunisia, compared to the reference period of 1981–2010, a decrease of annual precipitation (between -6 and -9% and between − 9 and − 18% in 2050 and 2100, respectively) is expected under climate change.

Common forage crops in Tunisia include alfalfa, clover, ryegrass and sorghum, with oats predominating, accounting for 60% of total forage crops. Since rainfed fodder makes up 79% of the crop, its production is largely dependent on the meteorological circumstances of each year (D.G.P.A 2017), which are becoming increasingly restrictive during the summer season. As a result, Tunisia's fodder resources are becoming more vulnerable and suitable areas to fodder and pastoral plants are expected to decline by an average of 19% (Deandreis et al. 2021).

The need to introduce alternative fodder species that can adapt to challenging growth circumstances and use less water is the reason behind this. Vetiver (Chrysopogon zizanioides L.) Roberty; syn.: Vetiveria zizanioides (L.) Nash.), recently introduced in Tunisia (Ben Romdhane et al. 2022), is a promising fodder specie, considering its many advantages.

This species is a perennial C4 grass belonging to the Poaceae family, characterized by short rhizomes and massive, finely structured root system, allowing it to reach 3–4 m soil depth (Truong 2002). Due to its deep root system, vetiver can endure a wide range of environmental stresses, such as waterlogging (Rad et al. 2021), soil and water salinity (Zhou and Yu 2009), extreme temperatures and drought (Truong and Baker 1995; Akhzari et al. 2013). Vetiver grass can survive under severe aridity conditions, reaching less than 200 mm rainfall per year (Fraser 1993).

A recent study revealed the benefits and fodder quality of young vetiver leaves in terms of digestibility, crude protein, phosphorus, and other minerals (Lal et al. 2023).

However, optimum water supply and plant density are two crucial factors that affect the yield and quality of fodder (Amaducci et al. 2000; Zarei et al. 2017).

Among the benefits of optimizing plant density includes the increase in photosynthetic activity (Yang et al. 2014), plant height, number of tiller and forage yield (Tang et al. 2018). Optimal plant density is dependent on species, genotypes and environmental factors such as soil texture and water availability (Cusicanqui and Lauer 1999). According to Rossi et al. (2020) high density significantly increases forage yield of chia (Salvia hispanica L.). Likewise, an increase in sunlight interception and yield has been recorded with high canola densities in China (Wang et al. 2023). Indeed, excessive sorghum planting density reduced tillering, dry matter yield and protein content (Bahrani and Deghani 2004). However, increasing planting density under water-scarce conditions enhances competition between plants, leading to yield reduction (Berenguer and Faci 2001). For the maize, a reasonable increase in planting density combined with an appropriate irrigation can simultaneously achieve high yields and economic return, and water use efficiency (Zhang et al. 2022).

Due to its recent introduction in Tunisia in 2010 as a forage species (Ben Abdallah 2012) vetiver remains understudied and relatively unknown. Furthermore, vetiver is a perennial species with high tillering potential (Ben Romdhane et al. 2022). Thereby, identifying the optimal density in a water-scarce environment could improve forage yield in terms of both quantity and quality.

Hence, a field vetiver grass experiment was performed under different irrigation conditions and plant density in three consecutives crop seasons, to (1) examine the changes in the agronomical and quality vetiver forage components and (2) explore the role of these traits as potential indicator to explain the forage yield variability. The results of the current study will provide novel information and inspiration for the best plant density of Chrysopogon zizanioides under optimal and limited irrigation scenarios.

Materials and methods

Experimental site

A 3-year field study was conducted at the experimental station of Agriculture High School of Mateur, Tunisia (37°03′15″N; 9°37′11″E; Altitude 20 m) during three growing seasons: 2015 (GS-1), 2016 (GS-2) and 2017 (GS-3). Experiments were carried out on a silty-clay soil texture in which the average field capacity of the root is 32%. Prior to planting, soil samples from each plot were extracted and analyzed. The soil showed an average organic matter amount of 1.93% and a low total nitrogen content of 0.2% (Table 1).

Table 1 Soil physico-chemical properties of the experimental field
Full size table

The average monthly precipitation, evaporation, average minimal and maximal temperature obtained during the 3 growing seasons and 30 years average were collected from the National Institute of Meteorology of Tunisia and are presented in Fig. 1.

Fig. 1
figure 1

Monthly rainfall (A), evaporation (B), average minimum (C) and maximum temperature (D) during growing seasons 2015–2016–2017 compared to 30 years average in the experimental site

Full size image

Experimental design and field Management

Treatments were arranged in three-replicated split-plot factorial design with four irrigation regimes as: optimum (Ir75), moderate (Ir100), severe (Ir125), and very severe irrigation (Ir150) (when evaporation reached respectively 75, 100, 125, and 150 mm from the surface of evaporation pan class “A”), combined with three plant densities: 10,000 (D1), 20,000 (D2) and 40,000 plants ha−1 (D3). The essay was arranged into 36 plots of 16 m2 (4 × 4 m) spaced by 3 m. Plant densities were arranged as follows: 12 plots (1 × 1 m), 12 plots (1 × 0.5 m) and 12 plots (0.5 × 0.5 m) for D1, D2 and D3, respectively (Fig. 2).

Fig. 2
figure 2

Experimental design

Full size image

Before planting, plots were surface homogenized by a roller to ensure equal water distribution. The plants were obtained from 25 cm long vetiver cuttings grown for 2 months in a greenhouse on brown peat substrate. Vetiver field transplantation was carried out on March 5th, 2015. During initial plant growth stages, each plot was irrigated equally, and a manual weeding was carried, irrigation regimes were applied two months after planting. During the three growing seasons, a homogenization cut was released to equalize and keep plant height at approximatively 10 cm above the ground before starting water irrigation treatments. For each treatment, irrigations were started when the corresponding evaporation levels were reached. Surface irrigation was adopted as a method of bringing water to the plots. The total number of irrigations for each growing season (two cuts/GS) was: 15, 10, 7, and 5 irrigations for Ir75, Ir100, Ir125 and Ir150, respectively. Irrigation’s intervals were on average between 6–7 days (Ir75), 9–10 days (Ir100), 12–13 days (Ir125) and 16–17 days (Ir150). A water meter was used to measure the amount of water applied. Each plot received approximately 0.4 m3 of water in each irrigation. During each growing season, each water regime received a total amount of water: Ir75 (375 mm), Ir100 (250 mm), Ir125 (175 mm), and Ir150 (125 mm).

During the 3 years of the study, irrigation treatments were started and interrupted on approximately the same dates. Two cuttings per season were carried out, 7 weeks apart. Table 2 records the beginning dates of irrigation treatments and cutting for the three growing seasons.

Table 2 Date of starting irrigation regimes and cuttings during the three growing seasons: 2015 (GS-1), 2016 (GS-2) and 2017(GS-3)
Full size table

Assessed parameters

Agronomic traits

During the three growing seasons, cuts were realized 7 weeks of plant growth later. All growth and yield parameters were recorded the day of cut. Five plants were chosen in the two centrals rows of each plot, marked and considered to measure: plant height (H), numbers of tillers per plant (Nt), chlorophyll content (Ch), and L/S ratio. The leaves situated in the mid part of plants were used to estimate Ch, by Chlorophyll meter (SPAD-502 plus; Kononica Minolta sensing, INC. Japan). The leaves (L) and stems (S) of each plant were separated and weighed to calculate the leaf per stem ratio (L/S). In each plot, the dry matter yield (DMY) (kg m−2) was estimated using 1 m2 quadrates, then after recording fresh weight, total samples were dried in a forced air oven at 70 °C for 72 h.

Fodder quality traits

Quality analysis of vetiver leaves was accomplished in a certified animal and forage laboratory (Central Laboratory for Feed Analysis, Soukra, Tunisia). Dried leaves samples were used to analyze plant nutritive components. The samples were ground through a 0.2 mm sieve in a cyclone sample mill. Ashes (Ash) was determined after burning the samples in a muffle furnace at 550 °C for 10 h. This procedure allowed us to calculate the organic matter (OM) content. Crude fat (CFa) was determined using Soxhlet (ISO 6492:1999). The plant calcium content (Ca) was determined by SAAF method (ISO6869:2000), and the phosphorus concentration (P) was estimated by spectrophotometric method (ISO 6491:1998). The crude protein (CP), water soluble carbohydrates (WSC), neutral detergent fiber (NDF, on dry matter basis) and acid detergent fiber (ADF, on dry matter basis) were determined using a near-infrared reflectance spectroscopy (NIRS). Dry matter digestibility (DMD) was estimated according to the following equation: DMD = 88.9—(0.779 × ADF) (Macedo-Barragán et al. 2011).

Data analysis

Collected data were subjected to analysis of variance using the GLM procedure. The treatment means differences were compared using the Tukey test (α = 0.05). Pearson correlation coefficient was determined for all the measured components for the three cropping seasons. The stepwise analysis was released with protein yield (PY) as a dependent variable and the choice of predictive variables was carried out by an automatic procedure. All statistical analyses were released using R 4.0.5.

Results

As a perennial crop, vetiver growth started with increased temperature in March and reached a maximum during the summer season where the temperature reached 35.3 °C in August 2017 (Vs 31.8 °C during a 30-year average) (Fig. 1D). Besides, growth is completely stopped in winter when the minimum temperature reached 6.7 °C in January (Vs 5.6 °C during a 30-year average) (Fig. 1C). The maximum values of evaporation were recorded during July and August with an average for the three growing seasons of 236 mm and 229 mm, respectively (Fig. 1B). Those evaporation values are higher than the 30 years average data of July and August by 15.5 mm and 17.3 mm, respectively.

For long-term average annual rainfall (30 years: 1984–2014) was 585 mm, of which over 70% occurred between November and March. During the first (GS-1; 2015) and third (GS-3; 2017) growing seasons, precipitation was 5.6% and 22.5% below the 30 years average. Only in the GS-2, rainfall increased by 11.96% compared to the region average (Fig. 1A). Evaporation during GS-1, GS-2 and GS-3 was higher than thirty years average by 6.40%, 8.71% and 11.95%, respectively (Fig. 1B).

Growth parameters and dry matter yield

During the three growing seasons, the vetiver's growth and production were significantly affected by irrigation regimes (Ir) and planting density (D). Moreover, a significant double interaction (Ir × D) was observed for most measured parameters (Table 3).

Table 3 Effects of four irrigation regimes (Ir), plant density (D) during three growing seasons (GS) on vetiver height (H, cm), number of tillers per plant (Nt), ratio of leaves fresh weight per stem fresh weight (L/S ratio), dry matter yield (DMY, T ha−1) and leaf chlorophyll content estimated by SPAD (Ch)
Full size table

Growth of vetiver

Plant height (H) was significantly (P < 0.05) affected by the interaction between irrigation regime (Ir) and plant density (D). Overall, the H was improved by irrigation and reduced by increasing density, mainly under longer irrigation intervals (Ir125 and Ir150) (Fig. 3). The combination of low-density D1 (10,000 plants ha−1) and short irrigation intervals (Ir75 and Ir100) allowed a better growth ± 248 cm. Additionally, despite variations in density, no significant differences were observed between both of those two irrigation treatments (Ir75 and Ir100). In contrast, the highest growth decline (42.77%) was recorded under the combination of severe water regime (Ir150) and the highest plant density D3 (40,000 plants ha−1).

Fig. 3
figure 3

Boxplot illustrating the plant height variation (H) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities (10,000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)), during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

The number of tillers per plant (Nt) in vetiver was significantly influenced (P < 0.01) by irrigation levels, different plant densities, and growing seasons (GS), no interaction effect was observed between these three treatments (Table 3). The Nt decreased as a result of water shortage and high densities. Extremely severe irrigation regime (Ir150) contributed to a maximum drop in Nt (21.81%), compared to optimum (Ir75). In addition, the increase in plant density led to a gradual decrease in Nt. The highest density (D3; 40,000 plants ha−1) was among a maximum decrease of Nt (11.81%) compared to D1 (10,000 plants ha−1). On the other hand, tillers production increased linearly over the three GS, with a maximum of 21.10% during the last growing seasons for all treatments.

In our study, variation in irrigation regimes and planting density significantly (P < 0.01) impacted the L/S, no interaction was observed. Water scarcity reduces the L/S ratio, while increasing density improves leaf weight compared to stems (data not shown). Therefore, under Ir150 the L/S ratio decreased by 25.82% compared to the optimum irrigation regime (Ir75). Increasing density from D1 to D3 resulted in a linear increase of L/S (+ 16.80%).

Leaf chlorophyll content estimated by SPAD (Ch) was significantly (P < 0.01) affected by D (P < 0.01) and Ir (P < 0.05). The highest values of SPAD were obtained in the less dense plots D1 (50.72) versus D3 (47.94) and plots with an irrigation regime corresponding to Ir125 (Table 3).

Dry matter yield (DMY)

The analyzed data showed that both factors Ir and D were among significant (P < 0.01) effects on DMY (Table 3). During all GS, drought stress led to a significant decrease in DMY. On the other hand, increasing plant densities improve DMY. Average plant density (D2; 20,000 plants ha−1) and highest densities (D3; 40,000 plants ha−1) yielded better DMY under the four irrigation treatments. Shorter irrigation intervals (Ir75 and Ir100) yielded higher DMY compared to longer irrigation intervals such as Ir125 and Ir150 for all plant densities. No significant difference was observed between optimum (Ir75) and moderate (Ir100) (Fig. 4).

Fig. 4
figure 4

Boxplot illustrating the dry matter yield variation (DMY) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities (10,000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3), during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

In all water regimes, D1 had the least fluctuation amount in DMY. The decrease of DMY was only by 38.31% (− 4.10 T ha−1) for D1 from optimum (Ir75) to severe water irrigation regime (Ir150). While for the same water stress (Ir150), DMY decreased by 44.86% (− 8.3 T ha−1) and 50% (9.7 T ha−1) for D2 and D3, respectively. Under severe (Ir125) and very severe (Ir150) irrigation regimes, DMY was higher for D2 and D3 compared to D1 (Fig. 4).

Vetiver forage quality

Crude protein and protein yield

The analyzed data showed that crude protein content (CP), expressed by gram per kilogram of dry matter, was significantly (P < 0.01) affected by Ir, D, and GS (Table 4). Reduced irrigation frequency led to a progressive rise in CP. The maximum increase of CP (15.83%) was observed under Ir150 compared to Ir75. In addition, CP showed a negative trend with increased plant densities. Indeed, D3 (40 000 plants ha−1) decreased CP by 8.77% compared to D1. The CP content was consistent during the two first growing seasons, while a decrease in 2017 (GS-3), by 4.56% compared to GS-2 was registered (Table 4).

Table 4 Effects of four irrigation regimes (Ir), plant density (D) during three growing seasons (GS) on vetiver crude protein (CP, g kg−1) protein yield (PY, kg ha−1), fat (CFa, %), organic matter (OM, %), ashes (Ash, %), calcium (Ca, %), phosphorus (P, %), dry matter digestibility (DMD, g kg−1), water soluble carbohydrates (WSC, g kg−1), acid detergent fiber (ADF, g kg−1) and neutral detergent fiber (NDF, g kg−1)
Full size table

During three growing seasons, protein yield (PY), expressed by kg per hectare, was significantly (P < 0.01) affected by variations in water regimes and plant densities. Also, a highly significant (P < 0.01) interaction was observed between these two parameters (Table 4). During three years, vetiver produced more protein per hectare under the increased frequency of irrigation and plant densities. The highest PY was observed under optimum (Ir75) and moderate irrigation (Ir100). Average tested plant density D2 (20,000 plants ha−1) yielded the best under the four water regimes. Moreover, PY was, in general, consistent under D2 and D3. The interactions of D2 × Ir100 and D2 × Ir75 were among the highest PY with 1978.46 kg ha−1 and 1934.97 kg. ha−1, respectively. In contrast, the longer irrigation interval (Ir150) combined with the less density (D1) showed the lowest PY with 831.76 kg. ha−1. Both D2 and D3 allowed higher PY under severe drought conditions (Ir125 and Ir150) compared to D1. The results showed clearly that D2 with half the number of plants than D3 could be adapted for all irrigation regimes in Mediterranean areas (Fig. 5).

Fig. 5
figure 5

Boxplot illustrating the vetiver protein yield variation (PY) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities [10000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)], during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

Dry matter digestibility (DMD)

In this study, water shortage improves DMD and increasing density reduces DMD. The triple interaction between Ir × D × GS significantly (P < 0.01) impacted the dry matter digestibility (DMD) of vetiver (Table 4). The interaction between longer irrigation intervals Ir150 and the lowest and middle density (D1: 10,000; D2: 20,000 plants ha−1) led to a rise in DMD during the three growing seasons. The maximum DMD values of 623.37 and 627.7 g kg−1 g kg−1 were observed under both Ir150 × D1 and Ir150 × D2, respectively during GS-1 and GS-2. On the other hand, short irrigation intervals coupled with high density reduced DMD. The lowest DMD recorded values of 560.8 g kg−1 and 581.75 g kg−1 were registered under Ir75 × D3 and Ir100 × D3, respectively during GS-1 and GS-2. It’s important to notice that in GS-3, D1 (10,000 plants ha−1) showed highest DMD mainly under drought conditions (Ir125 and Ir150) (Fig. 6).

Fig. 6
figure 6

Boxplot illustrating the vetiver dry matter digestibility variation (DMD) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities (10,000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)) during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

Water soluble carbohydrates (WSC)

During three growing seasons, water soluble carbohydrates (WSC) were controlled by the interaction effect (P < 0.05) of Ir × D. WSC increased as the water deficit and density increased. While the maximum WSC (139.33 g kg−1) was found under combination of highest density D3 (40 000 plants ha−1) and longer irrigation interval (Ir150). The minimal WSC value of 119.93 g kg−1 was recorded in the less dense plots D1 (10,000 plants ha−1) with an optimal irrigation level (Ir75) (Fig. 7).

Fig. 7
figure 7

Boxplot illustrating the variation of water-soluble carbohydrates (WSC) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities (10,000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)), during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

Acid detergent fiber (ADF) and neutral detergent fiber (NDF)

In our study, ADF and NDF were significantly affected by plant density changes and irrigation regimes variations. ADF and NDF followed an incremental trend as plant densities increased from D1 (10,000 plants ha−1) to a high of D3 (40,000 plants ha−1) and decreased when water shortage increased. Likewise, a significant (P < 0.01) interaction between irrigation levels and plant densities was observed. The Highest ADF (295.07 g kg−1) and NDF (582.2 g kg−1) content were recorded when vetiver was planted at the highest density D3 (40,000 plants ha−1) and under optimum irrigation level Ir75. Meanwhile, minimum ADF and NDF values were observed for the longest irrigation interval (Ir150) and the lowest plant density (D1; 10,000 plants ha−1) showing a respective decrease of 19.95% and 6.13% compared to the maximum at Ir75 × D3 (Fig. 8).

Fig. 8
figure 8

Boxplot illustrating the variation of concentration of acid detergent fiber ADF (A) and neutral detergent fiber NDF (B) under the interaction of irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and plant densities [10000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)], during three growing seasons (GS-1, GS-2 and GS-3). Outliers exceeding 2.5 standard deviations are plotted as individual points

Full size image

Organic matter, ashes, nutrient and crude fat contents

The results showed that vetiver OM and Ash contents were significantly influenced (P < 0.01) by Ir, D, and GS, but no interaction effect between treatments was observed (Table 4). The Highest OM was recorded when the plants were irrigated by Ir75 and planted at the highest density (D3) and during the last growing season (GS-3). The plots that received the maximum dose of water Ir75 showed statistically higher organic matter contents by 85.76% against 83.49% for Ir150. In addition, less densely planted plots (D1; 10,000 plant ha−1) had less OM content with 84.05% than densely planted plots (D3; 40,000 plants ha−1) 85.09% (Table 4).

In vetiver, increased water shortage and reduced planting density improves ash content. The highest Ash occurred under the lowest irrigation level (Ir150) and D1 plant density with a respective increase of 15.95% and 6.97% compared to Ir75 and D3, respectively. On the other hand, the minimum Ash content was observed during the third year of cultivation (GS-3) (Table 4).

Irrigation regimes have a significant (P < 0.01) effect on Ca and P content of vetiver. Lower irrigation frequency levels led to a progressive decrease in Ca and P content. Increasing water shortage from Ir75 to Ir150 reduced calcium and phosphorus respectively by 21.27% and 17.24% (Table 4).

In contrast, the effect of planting density was only observed on calcium content. Under the maximum density D3 a lower calcium content was observed compared to D1 and D2.

In this study, crude fat content was significantly (P < 0.01) affected by irrigation and planting density. Increasing the water restriction from Ir150 to Ir75 improved CFa accumulation by 2.29%. Similarly, the most densely planted plots (D3; 40,000 plants ha−1) showed a CFa content 2.64% higher than the density (D1; 10,000 plants ha−1) (Table 4).

The predictive variables of protein yield: step-wise analysis

To assess the relationships between the final protein yield (PY) and all measured parameters including both agronomical and quality vetiver forage components, a stepwise multiple linear regression analysis was assessed. The PY was selected as the most important factor due to the fact that this parameter is the final output of both quantitative and qualitative aspects of vetiver forage for feeding animals.

Under irrigation regimes, different models were generated depending on the drought conditions induced by the irrigation intervals. PY under optimal irrigation (Ir75) was established by two agronomic components (DMY and Nt) and only CP as a quality component. While under a moderate irrigation regime (Ir100), PY was controlled by only DMY and CP (Table 5).

Table 5 Multiple linear regression analysis (stepwise) explaining protein yield (PY) variation within irrigation regimes and plant densities as a dependent’s variables, and all measured parameters as independent variables
Full size table

With higher water stress as for a severe irrigation regime (Ir125), PY was under the control of DMY, Nt (agronomic components), and CP, NDF (quality components). Besides and for the most severe irrigation regime (Ir150), PY was controlled by DMY, CP, Nt.

The results showed clearly that under sufficient growth water supply, PY was dependent mainly on DMY, CP, Ch, and NDF. While under drought conditions, and in addition to previous parameters the Nt seems to be a predominant determinant factor for PY. Considering the variation of plant density, PY was dependent on a set of common components such as DMY and CP. Under low density (D1), PY was also determined by Nt. While, P, CFa and Ch were additional components in the model explaining PY under the highest plant density (D3).

The PY under the interaction of Ir × D was determined by a model including both agronomic (DMY) and qualitative (CP, ADF and DMD) components (R2 = 0.99).

Relationship between dry matter yield (DMY) and protein yield (PY)

DMY and PY are the two main components for vetiver quantitatively and qualitatively aspects. Those two parameters showed a linear relationship despite levels of Ir and D (Fig. 9). The linear correlation between DMY and PY was not dependent on irrigation regimes between 15 T ha−1 of (DMY) and 1500 kg ha−1 of (PY). The optimum irrigation for DMY and PY was observed for a moderate irrigation regime (Ir100) showing a higher slope compared to the linear regression obtained under Ir75. Irrigating vetiver moderately during the summer season could save water and insure DMY and PY but only less than 20 T ha−1 and 2000 kg ha−1, respectively (Fig. 9A). Considering plant density, the maximum DMY and PY were reached under average plant density (D2; 20,000 plant ha−1); while the lowest yields were obtained under low plant density (D1; 10,000 plant ha−1). It’s important to notice that the linear regression showed no difference between the slope of D1 and D2 till DMY of 17.5 T ha−1 and PY of 2000 kg ha−1.

Fig. 9
figure 9

Linear regression between protein yield (PY) and dry matter yield (DMY) under (A) the irrigation regimes (Ir75, Ir100, Ir125, and Ir150) and (B) plant densities [10,000 (D1), 20,000 (D2), 40,000 plants ha−1 (D3)]

Full size image

The linear regression between DMY and PY, showed that the highest plant density (D3; 40,000 plant ha−1) has a lower slope compared to D2. Planting vetiver at 20,000 plants ha−1 (D2) was the best plant density to reach maximum dry matter yield (DMY) and protein production (PY) (Fig. 9B).

Discussion

In a general context of water scarcity, the choice of an appropriate planting density makes it possible to optimize the use of the available water in the soil, and to fully express the genetic yield potential. In this work, we studied the effect of different irrigation regimes combined with several planting densities on the yield and forage quality of vetiver grass, to gain a better understanding of this species, as an alternative forage under Tunisian conditions.

Vetiver growth and yield as affected by water regimes and planting density interaction

Regardless of density effect, and under the three growing seasons, the highest vetiver growth (240 cm) and yield (16.2 T ha−1) were recorded under Ir75 and slightly less under Ir100 regimes. By increasing the water stress, growth is reduced by 24 and 37% at Ir125 and Ir150, respectively. Similarly, water scarcity had a detrimental impact on yields that are directly correlated with plant growth factors (Gebremedhin et al. 2019), the Ir150 regime recording the lowest average yields (8.7 T ha−1). Comparable results were also noted, under drought conditions, for berseem clover (Trifolium alexandrinum) (Daneshnia et al. 2015) maize and sorghum (Berenguer and Faci 2001).

This expected result is among the main indicators of forage production, and yield is considered a reliable criterion to evaluate plant tolerance to biotic and abiotic stress. In this context, Staniak (2016) reported a strong reduction of fodder grass yielding us as Dactylis glomerata, Festuca pratensis, Festulolium braunii, Lolium multiflorum, under severe drought conditions.

During the three growing seasons and under increasing water shortage (from optimum (Ir75) to very severe irrigation (Ir150), DMY was reduced on average by 46% (Table 3). Moreover, DMY was related to plant density where D2 (20,000 plants ha−1) and D3 (40,000 plants ha−1) densities were among the highest DMY regardless of Ir. Those observations are in complete agreement with those of Ferreira and Teets (2017), who reported that increasing plant density under no water constraint increases DMY. However, the lowest density D1 (10,000 plants ha−1) showed the smallest fluctuation in DMY when moving from an optimum water regime (Ir75) to the most severe (Ir150). This implies that the root system makes better use of the amount of water available in the soil. This aspect, which is related to the volume of soil explored by the roots, should be studied in depth under other conditions. However, when root density is high and water supply limited, plant competition for water and minerals increases (Balazadeh et al. 2021), mainly for nitrogen uptake, which considerably limits plant growth and development (Yang et al. 2014). In our experiment, no nitrogen amendment was applied, which probably led to competition for soil reserves and would certainly have an impact on vetiver yield from one year to the next and particularly with the highest densities D3.

The DMY decrease was also associated with the L/S ratio reduction. Control of vetiver leaf emission is considered a drought stress tolerance mechanism. This adaptation has been described in several forage species such as turf grasses (Carraw 1996), forage sorghum (Jahanzad et al. 2013), forage Maize (Rad et al. 2021; Panada et al. 2004), and Pearl millet (Rostamza et al. 2011). In fact, reducing the number of leaves and leaf area, combined with the thickening of the cuticle, restricts transpiration (Staniak and Kocoń 2015).

Water scarcity limits vetiver leaves emission, reduces transpiration surface and thus improves plant water status (Borrajo et al. 2018). Under drought conditions, leaf senescence allows tolerant species to redistribute reserve substances (proteins, carbohydrates, and fats) towards other organs (Munné-Bosch and Alegre 2004) mainly to let reborn tissues of perineal species. In fact, re-irrigation after drought conditions or after dormancy, will boost the growth of species that already stored reserves in roots (Hofer et al. 2017). The fructans were identified as the main stored reserves in entire tiller bases of perennial forage species (Volaire et al. 1998). In vetiver, the root system plays a key role in the accumulation of reserve substances essential for the regeneration of the aerial parts (Lal et al. 2023).

The vetiver regeneration process after water restriction, was observed during the three growing seasons by a linear increase in the number of tillers (Nt). This perennial grass, known for its tillering potential, exceeds 37 tillers per plant (Ben Romdhane et al. 2022). In this study, the emission of new shoots after each irrigation was observed in all irrigation treatments to delay height growth and escape unfavorable conditions (Zarei et al. 2017). This perennial character of Vetiver guarantees a continuous and reliable production cycle, providing a significant advantage compared to the annual summer crops commonly cultivated in Tunisia and the Mediterranean basin, such as maize and sorghum.

The stepwise analysis confirms that DMY is the main component in vetiver production. Vetiver DMY was not affected by water irrigation treatments till moderate drought conditions (Ir100). Vetiver’s tolerance drought could be attributed to its ability to develop a root system that is dense (Machado et al. 2015), deep up to 3 m (Islam et al. 2016), and consists mainly of fine roots (0.5–1.0 mm in diameter) (Truong 2002). Such a developed root system confers to most grasses the aptitude to absorb large amounts of water (Fariaszewska et al. 2017). The Vetiver root system can capture and manage rainfall throughout the year, especially the increasingly frequent torrential summer and autumnal rains, resulting in the ability to decrease soil erosion on especially vulnerable slopes.

Vetiver forage quality as affected by water regimes and planting density interaction

Water availability for fodder crops influences forage yield but also their nutritional value. Results indicated that a lack of water had a deleterious impact on PY, OM, ADF, NDF, and P. While, drought causes an increase in CP, DMD, WSC, Ash, and CFa.

Crude protein (CP) is one of the main indicators of forage quality (Küchenmeister et al. 2013). Results revealed that CP tended to increase under low soil moisture as previously reported for a set of species (Staniak 2016). Numerous factors influenced CP such as soil type, plant growth stage, plant mineral and organic fertilization, and water availability as well as duration and intensity of drought stress (Farooq et al. 2009; Jahanzad et al. 2013). Similar results were obtained for sorghum (Sorghum bicolor), a 14.28% increase in protein content was recorded when moving from an optimal irrigation level to a very restrictive water regime (Janhanzad et al. 2013).

Pearl millet (Pennisetum americanum) showed the same rise (Rostamza et al. 2011). The findings also demonstrated that, as has already been reported in a number of plants including corn (Abhishek and Basavanneppa 2020) and sorghum (Gosavi and Bhagat 2009) increasing vetiver planting density lowered CP content as a result of an intra-specific competition for water and nitrogen.

The ADF (cellulose and lignin) and NDF (hemicellulose, cellulose, and lignin) are considered to be the two most important factors in the digestibility of fodder species (Staniak 2016). Low fiber content is associated with high forage quality (Staniak and Harasim 2018), and it’s impacted by numerous growth factors such as plant growth stage, temperature, light, nutrients, and water availability (Liu et al. 2018).

Overall, in our study irrigation treatments had a significant impact on vetiver ADF and NDF fractions. Longer water shortage was associated with a declining fiber fraction. Reduction in ADF and NDF content under drought conditions was reported in forage species such as Dactylis glomerata, Festuca pratensis, Festulolium braunii and Lolium multiflorum (Staniak 2016), and in legumes (Küchenmeister et al. 2013; Staniak and Harasim 2018; Liu et al. 2018). Sorghum and other forage crops have shown increased lignin content associated with adequate proper water supply (Amaducci et al. 2000). A study conducted on sorghum bicolor, by Qadir et al. (2019), indicated that NDF and ADF were positively correlated with shorter interval irrigation and build up as intervals of irrigation decrease. The decline of NDF was attributed to the suppression of the enzyme activities that produce fibers (Sheaffer et al. 1992).

Therefore, the vetiver hemicellulose fraction is strongly affected by water restriction as reported for other grass species (Fariaszewska et al. 2017). On the other hand, in our study, the interaction between irrigation levels and plant densities is more important, increasing the vetiver’s population under adequate water status improves ADF and NDF. Similar results were reported for maize (Widdicombe and Thelen 2002; Zhang et al. 2022) and Sorghum (Jahanzad et al. 2013).

In contrast with ADF and NDF, vetiver's dry matter digestibility (DMD) is enhanced by water constraint. Several studies reported a negative correlation between DMD and both ADF and NDF (Huang et al. 2020; Balazadeh et al. 2021). Drought stress reduces fiber fractions (ADF and NDF) in many forage species (Fariaszewska et al. 2017). The improvement of vetiver digestibility under severe water treatments (Ir125 and Ir150) could be attributed to the decrease in hemicellulose content (Jahanzad et al. 2013; Staniak and Harasim 2018). In this study increasing plant density caused a reduction in the digestibility of vetiver, this result was observed in other forage species, likewise perennial rays-grass (Lolium perenne) (Jafari et al. 2003). The relationship between irrigation levels and vetiver digestibility should be examined in future research, in order to improve our knowledge of this species, which is still largely unknown.

Water soluble carbohydrates (WSC) constitutes an important factor impacting animal forage digestibility (Fariaszewska et al. 2017). The increase of WSC is mostly correlated with forage photosynthesis efficiency. In fact, the vetiver photosynthesis process was not affected by the drought (Table 4) allowing the synthesis of carbohydrates. This photosynthesis efficiency is probably due to PSII resistance to protein denaturation under biotic stress (Manaa et al. 2021). The results showing that water shortage increases vetiver WSC concentrations were in complete agreement with other species as perennial forage legumes (Küchenmeister et al. 2013), soybean (Nakayama et al. 2007) and legumes-grass mixtures (Staniak and Harasim 2018). Thus, increasing sugar concentration may contribute to adjusting cell osmotic potential, and therefore maintaining soil water uptake under drought and salt stress conditions.

Equally plant density impacted WSC content. Less dense plots had less WSC than densely planted plots. Also, the interaction effects between density and irrigation levels were significant in this study. For vetiver, high density associated with longer irrigation intervals improves WSC. In the same study conducted by Jahanzad et al. (2013), WSC content in sorghum increased when plant densities increased and water scarcity increased, also the effect of interaction was registered.

In addition, crude fat (CFa) content tends to increase under water shortage only since severe water regime (Ir125). The same tendency was observed in Festulolium and Alfalfa under soil moisture varying from 40 to 70% (Staniak and Harasim 2018). Soil water availability promotes vetiver calcium and phosphorus accumulation in stems and especially in leaves, which is partly explained by aerial growth observed under Ir75 and Ir100. Besides, water shortage increased vetiver ash content as already observed in sorghum (Jahanzad et al. 2013) and in Pearl millet (Rostamza et al. 2011).

Subsequently, depending on the water regime used, selecting the right beginning density is crucial for boosting or maintaining production and quality in a perennial grass like vetiver. Managing the available water resources more effectively can be achieved by adjusting the stand's density in line to the water availability in the area where the vetiver will be introduced.

Finally, for a perennial grass such as Vetiver, the choice of starting density is a key factor in maximizing or maintaining production and quality, depending on the water regime applied. By modulating the density of the stand according to the availability of water in the environment in which the vetiver will be introduced, it is possible to enhance the use of the available water resources.

Conclusions

This study is the first report to investigate the potential of vetiver grass as an alternative source of fodder in an overall context of water scarcity in Tunisia. Vetiver has shown a great adaptive plasticity, and density seems to be a decisive factor in optimizing the use of available water resources.

Regardless of the water regime applied, the best DMY and PY were recorded in D2 (14.7 Tha−1, 1638 kg ha−1) and D3 (14.67 Tha−1, 1540 kg ha−1) under the three growing seasons. Moreover, our investigation indicated that the interaction between moderate irrigation regime (Ir100) and average plant density D2 (20,000 plants ha−1) seems to be the best compromise for maximizing PY (1978.46 kg ha−1). More specifically, D2 has allowed maximum optimization of the different amounts of applied water. Despite, the low production of D1, it was distinguished by the slightest fluctuation in term of DMY by 38.31% from an optimum water regime (Ir75) to the most restrictive one (IR150) compared to other two plant densities with 44 and 50% for D2 et D3, respectively.

In light of our preliminary results, it would be judicious to study the effect of different planting densities without supplemental irrigation with the intention of introducing the vetiver to marginal and rangelands under different pedoclimatic conditions. Additionally, extending the study period to autumn and spring seasons whether or not fertilized and use treated wastewater could be useful to more explore the vetiver potential.