1 Introduction

The efficiency of animal production in grazing areas is directly related to soil fertility, stocking rate, grazing management, water availability, and meteorological conditions of temperature and incident solar radiation that allow forages to express their productive potential and nutritional quality. Technologies that positively influence the forage production and utilization associated with proper management can increase productivity per unit area.

Increasing productivity and resource efficiency through sustainable intensification of grazing systems impose new challenges on crops and livestock. In temperate zones, the growth and development of tropical forages are irregular throughout the year, and they are influenced by several environmental factors (Esmaili and Salehi 2012; Newman et al. 2007). In order to reduce the damage caused by the low and inconsistent rainfall in specific periods, irrigation can be used to work around the effects of water deficit on forages. The irrigation prolongs the vegetative cycle and increases the forage daily production and stocking rate (Teixeira et al. 2013; Sanches et al. 2015). Still, in many grazing areas, the nutritional value of forage is insufficient to meet the animals’ requirements (Brown et al. 2012). The overseeding of black oat and ryegrass pasture on tropical pastures (Aruana Guineagrass) benefits grazing systems by supporting high stocking rates during the winter (Schmitz et al. 2020). It optimizes grazing areas, resulting in regular animal production throughout the year.

Ruminants digest temper legumes more rapidly than grasses. Thus, intake and production are typically greater than forage grasses. The high fiber digestibility enhances the efficiency of energy and protein utilization by ruminants compared to other forages (Christensen et al. 2015). It is likely that legume pastures under intensive grazing management could optimize ruminant productivity per unit area. Finding legumes adapted to subtropical production systems is complex since there is great variability in species, both perennial and annual. Furthermore, they are sensitive to physical and chemical soil conditions. So, clay content and organic matter are crucial in choosing a legume. Perennial species such as white clover and bird’s-foot trefoil show higher forage yield in spring. Bird vetch, an annual plant, presents slow establishment, showing sensitivity to competition in the forage canopy and trampling.

The legumes also reduce greenhouse gas effects and increase the nutritional value and efficiency of forage conversion into animal protein (Lüscher et al. 2014). In this sense, we hypothesized that over 3 years, adding irrigation and/or legumes to productive systems based on African Bermuda grass pasture, overseeded in the winter with black oat and ryegrass, would increase the productive performance of beef cattle (Fig. 1).

Fig. 1
figure 1

Use of irrigation and legumes in tropical pastures for beef steers. Photograph by Wagner Paris, UTFPR-DV.

Full size image

2 Materials and methods

2.1 Location, area, and fertility management

The experiment was carried out in the county of Dois Vizinhos, State of Paraná-Brazil, in an experimental area of the Beef Cattle Teaching and Research Unit, which belongs to the Universidade Tecnológica Federal do Paraná, located 25°44′ S and 53°04′ W, elv. 520 m a.s.l.

The soil is characterized as a Dystrophic Red Nitosol with a clay texture (Bhering et al. 2007). According to the Köppen classification, the climate is Cfa, humid subtropical. The climatological data of the evaluation period were recorded and obtained through an automatic surface weather station from UTFPR-DV, located about 100 m from the experimental area (Fig. 2).

Fig. 2
figure 2

Accumulated rainfall, average maximum and minimum temperatures, and water depth applied over the experimental months.

Full size image

Before the overseeding of temperate species, in each year, soil samples were collected at 20 cm depth. The acidity correction and base fertilization were performed as recommended by CQFS (2004), the amount of 200 kg/ha of a fertilizer, formulated with 5:20:10 NPK. The average soil analysis of 3 years were organic matter = 42.55 g dm−3; P = 5.25 mg dm−3; K = 0.39 cmol dm−3; Ca = 4.5 cmol dm−3; Mg = 1.9 cmol/dm3; bases = 7.12 cmol dm−3; Al3+ = 2.85%; and pH(CaCl) = 4.8. The nitrogen topdressing fertilization ((NH2)2CO) was carried out using 300 kg/ha of N that was split into three applications of 100 kg/ha per each of the three seasons.

2.2 Experimental design and treatments

The experiment was completely randomized in a 2 × 2 factorial arrangement. The treatments were irrigated pastures with or without legumes and non-irrigated pastures with or without legumes, with three replicates (paddocks). The experimental area had approximately 3.6 ha, distributed in 12 paddocks with an average size of 3000 m2. Each paddock was subdivided into four sub-paddocks. The winter comprised the months from July to September; the spring, from October to December; and the summer, from January to April, for all the 3 years. The period between May and June was used for germination and growth of the temperate species.

The research was conducted from 2016 to 2019. The grazing began for each year on 07/16/2016, 07/14/2017, and 08/07/2018, totaling 246, 265, and 210 grazing days, respectively. The delay in the last year was due to the adverse climatic conditions in the overseeding. The experimental area was an African Bermuda grass (Cynodon nlemfuensis Vanderyst) pasture since 2014. Every year, the overseeding of black oat and ryegrass pasture was carried out in May. The pasture was mowed using a brushcutter set to cut at 10 cm in May in order to carry out this study. In 2016, 55 kg/ha of ryegrass (Lolium multiflorum “Fepagro São Gabriel”) was used, and in 2017 and 2018, we used 30 kg/ha (Lolium multiflorum “BRS Integração”). The black oat (Avena strigosa Schreb “IAPAR 61”) was seeded at a 60 kg/ha density in 3 years.

The forage legumes used in the mixture were common vetch (Vicia sativa “Amethyst”), white clover (Trifolium repens “Zapican”), and bird’s-foot trefoil (Lotus corniculatus “São Gabriel”) at sowing densities of 30, 15, and 5 kg/ha, in the years 2016, 2017, and 2018, respectively. The vetch was sown in rows, and the clover and bird’s-foot trefoil were broadcast seeded before sowing the oat.

2.3 Animal and pasture evaluations

We used 72 castrated steers as testers (24 per year of evaluation — two testers per paddock) at 8 months old. The animals were predominantly Angus × Nellore crossbred (203 ± 14 kg of initial live weight). The tester animals were adapted to the pasture management for 15 days before the evaluations, and all steers had free access to mineral salt in troughs and water tanks with an automatic float valve throughout the experimental period.

For 3 years, the animals remained in the paddocks throughout the experimental period, alternating between the sub-paddocks according to the management of pre- and post-grazing heights. The animal performance was monitored at the beginning and end of each season by individual weighing of the tester animals, right after 14-h fasting of solids and liquids. Thus, the average daily gain (ADG; kg/day) and live weight gain per hectare (LWG; kg/ha/day) were obtained. The stocking rate was obtained by adding the average live weight of the testers on the days they were in each season, adding to the weight and permanence of the put-and-take animals.

The irrigation system was activated when the soil water potential reached −10 kPa, with 50% moisture in the field capacity. The amount of water to be applied (Fig. 2) was determined based on the water retention curve in the soil, and the readings of the matrix potential were obtained by tensiometers with a digital vacuometer. Eight tensiometers were placed, four in the irrigated area and four in the non-irrigated area.

The pasture was managed under rotational stocking with put-and-take grazing management (Mott and Lucas 1952). The steers shifted between paddocks whenever the average of the three replicates of the treatment had a 95% light interception (LI). The LI was measured using the Canopy Analysis System-SUNSCAN (Delta-T, Cambridge, England). The luminous intensity was measured between 11:00 a.m. and 13:00 p.m. at ground level (10 measurements per paddock). We measured the pasture height with a ruler in these exact spots. The put-and-take animals were kept in a similar grazing area. They were used to adjust the height of the forage mass (about 50% entrance height) whenever necessary.

In the winter and spring/summer, the average pasture height to allow the entry of the animals was 30 and 40 cm, 33 and 38 cm, and 30 and 35 cm, in 2016, 2017, and 2018, respectively. The grazing period had 4 days per paddock and 12 days for rest in the winter. There was an average of 5 days for grazing in spring and summer and 10 days for rest.

Pasture evaluations were made pre- and post-grazing, moments before the entry and right after the exit of the animals, respectively. The forage mass was directly determined by three cuts close to the ground (0.25 m2), made randomly on the sub-paddocks in pasture representative conditions. Then, aliquots were taken from these samples for botanical separation (grass and legumes) to estimate the proportion of each species in the yield and forage mass.

Forage mass represented the total dry matter of forage per hectare above ground level, expressed in kg DM/ha. The accumulation rate was calculated by the difference in the forage mass between pre- and post-grazing, divided by the interval between grazing periods. The yield of forage and forage legume was obtained by adding the initial forage mass to the sum of the forage accumulations of the rest intervals.

The forage allowance (kg of DM/100 kg LW) was calculated as described by Neves et al. (2009). The stocking rate was obtained by adding the average live weight of the testers on the days they were in each season, adding to the weight and permanence of the regulator animals in the paddocks.

2.4 Chemical analysis

For the chemical analysis of the pasture, the samples were obtained by the grazing simulation technique (Moore and Sollenberger 1997). The sampled material was dried in a forced-air oven at 55 °C for 72 h. After that, the samples were ground in a Wiley mill to pass a 1-mm sieve, and sent to the food analysis laboratory of the UTFPR-DV. We analyzed the contents for dry matter, ash, organic matter, and crude protein (AOAC 2006), and insoluble fiber in neutral detergent (NDF) and insoluble fiber in acid detergent (ADF) by the methodology of Van Soest et al. (1991) adapted for polyester bags with autoclave digestion (Senger et al. 2008). The in vitro dry matter digestibility (IVDMD) was performed according to Tilley and Terry (1963) using polyester bags with treatment of neutral detergent solution in an autoclave (Goering and Van Soest 1970; Senger et al. 2008). The Technal TE-150 incubator was used for the simulation of the ruminal environment.

2.5 Statistical analysis

The statistical analyses were performed within each season, i.e., the seasons were not compared, following the mathematical model:

$$Y_{ijk}=\mu+\iota_i+\lambda_j+\gamma_k+p_l+{(\iota^\ast\lambda)}_{ij}+{(\iota^\ast\gamma)}_{ik}+{(\lambda^\ast\gamma)}_{jk}+{(\iota^\ast\lambda^\ast\gamma)}_{ijk}+\varepsilon_{ijk}$$

where Yijk is the observation concerning the i-th irrigation (ιi) and legume use (λj) in the j-th experimental year (γk) and their interactions. The i-th grazing period (pl) as a random effect was included when necessary.

The data were subjected to analysis of variance through the Glimmix procedure (SAS Institute Inc 2013) using the generalized linear mixed model methodology. The choice of which distribution would best fit the data was made using the corrected Akaike information criterion (AICc). The F test was used (P = 0.05) to check whether the treatment effects and interactions were significant. In case of interactions described in the mathematical model, the means were compared through Student’s t-test (P = 0.05) using SAS University Edition.

3 Results

3.1 Chemical compounds and biomass responses

During the winter, the irrigated pastures showed lower ADF content (Table 1). Still, they presented lower IVDMD during the summer than the non-irrigated pastures (Table 1). The other nutritional variables were not influenced by irrigation or the presence of legumes.

Table 1 Nutritional composition of the grazing simulation in the pre-grazing on irrigated African Bermuda grass mixed with legumes, overseeded in the winter with oat and ryegrass. Leg, with legume; Nleg, without legume; Irri, with irrigation; NIrri, without irrigation; SEM, standard error of the mean; *P value for irrigation effect — Irri, legume effect — Leg and irrigation vs legume interaction — Irr*Leg; P = 0.05 considered statistically significant by F test or Student’s t-test.
Full size table

The forage legume in the pre-grazing biomass reached 11% and 7% in the winters of the first and second years. The percentage of forage legume biomass was 7% and 5% for treatments without irrigation, respectively. However, its percentage was below 2% in the third year because spring production did not develop due to competition with the already established pasture.

The pre-grazing biomass (Table 2) did not differ between treatments within the seasons. Only the post-grazing forage mass in the winter was higher for the irrigated system (Table 2). However, there was an interaction between irrigation management and the experimental years for pre- and post-grazing mass in the winter (Table 3). These variables were higher in the irrigated system in the second year of evaluation (Table 3). Legumes provided a higher accumulation rate in the summer without significantly influencing the other seasons. Also, lack of irrigation decreased forage yield in winter (Table 2). In winter and spring, legume yield was significantly lower in the irrigated area, but it did not affect the total forage yield.

Table 2 Forage mass from grazing systems based on irrigated African Bermuda grass mixed with legumes, overseeded in the winter with oat and ryegrass. Leg, with legume; Nleg, without legume; Irri, with irrigation; NIrri, without irrigation; SEM, standard error of the mean; *P value for irrigation effect — Irri, legume effect — Leg and irrigation vs legume interaction — Irr*Leg; P = 0.05 considered statistically significant by F test or Student’s t-test.
Full size table
Table 3 Effect of irrigation or legumes presence on pre- and post-grazing, live weight gain (LWG), and stocking rate (SR) in the year of evaluation of grazing systems managed on irrigated African Bermuda grass mixed with legumes, overseeded in the winter with oat and ryegrass. SEM, standard error of the mean; P = 0.05 considered statistically significant for the irrigation or legume effect on the experimental years.
Full size table

3.2 Animal gain and live weight yield per area

Seasonal variations related to animal performance were restricted to spring and summer (Table 4). The irrigation provided a higher LWG and stocking rate in the spring. In the summer, legumes and irrigation resulted in a higher ADG than irrigated grass without legumes. Also, the stocking rate and LWG were higher for the grass–legume mixture regardless of irrigation in the summer (Tables 2 and 4).

Table 4 Animal performance from grazing systems based on irrigated African Bermuda grass mixed with legumes, overseeded in the winter with oat and ryegrass. Leg, with legume; Nleg, without legume; Irri, with irrigation; NIrri, without irrigation; SEM, standard error of the mean; *P value for irrigation effect — Irri, legume effect — Leg and irrigation vs legume interaction — Irr*Leg; P = 0.05 considered statistically significant by F test or Student’s t-test.
Full size table

Significant interactions between the irrigation and years (Table 3) were observed. The spring of the first year showed a lower LWG for the irrigated system. However, the irrigation was positive in the spring of the second year, resulting in a higher stocking rate (+940 kg LW/ha) and consequently higher LWG (+ 53%). This result was also observed in the summer of the third year, with a higher stocking rate and weight gain in the irrigated system. Simultaneously, the legumes also provided a higher stocking rate (+822.9 kg LW/ha) in the summer of the third year (Table 3).

The ADG and annual LWG proved sensitive to the interaction between the treatments and years (Table 4). The lowest ADG and LWG were recorded in the areas without irrigation or legumes in the second year. The ADG was lower in the third year for the irrigated system (Table 4) without legumes. It is also noteworthy that the increase in LWG with legumes and irrigation was 28% higher than treatments without legumes and irrigation in the third year (Table 5). The highest annual LWG and total weight gain, summing the 3 years, was observed in the system with forage legumes regardless of irrigation (Table 5).

Table 5 Daily and annual live weight gain of beef cattle on irrigated African Bermuda grass mixed with legumes, overseeded in winter with oat and ryegrass. Leg, with legume; Nleg, without legume; Irri, with irrigation; NIrri, without irrigation; SEM, standard error of the mean; *P value for irrigation effect — Irri, legume effect — Leg and irrigation vs legume interaction — Irr*Leg; P = 0.05 considered statistically significant by F test or Student’s t-test.
Full size table

4 Discussion

The ADG and LWG above 0.75 kg/ha/day and 1459 kg/ha/year, respectively, show that it is possible to intensively produce beef cattle exclusively on pastures of African Bermuda grass overseeded with irrigated annual winter grasses and legumes. Irrigation increases forage production, allowing the maintenance of stocking rate in periods of water deficit (Table 2). Using legumes increases animal productivity with a significant intensive and sustainable response in the long term (Table 5).

4.1 Biomass yield and nutrient content

The lower ADF of the irrigated system in the winter (Table 1) resulted from the forage growth stimulated by irrigation, increasing the contribution of vegetative material (Pequeno et al. 2015). C3 species have lower cell wall contents than C4 species. Besides that, water restriction and other environmental factors are necessary conditions for low forage production in grazing systems (Lemaire and Chapman 1996). Therefore, the non-irrigated pastures also present lower contents of soluble nutrients due to less translocation, photosynthesis, and nutrient absorption by the roots (McCree and Fernandez 1989).

The chemical composition did not differ in the summer and spring, mainly due to the forage masses of pre- and post-grazing and forage allowance (Table 2), demonstrating similar pasture management between treatments. Garay et al. (2004) used Bermuda grass in different stocking rates and reported that the pasture’s nutritional value is not related to the gain, being more associated with the mass and availability of forage. The pasture showed lower in vitro digestibility in the irrigated areas during the summer (Table 1), demonstrating that the forages available to the animals tend to display a significant amount of support structures and less digestibility due to greater growth by irrigation. Pequeno et al. (2015) also found that irrigation during rainy seasons decreases in vitro digestibility. Pedreira et al. (1999) observed that lower CP and IVDOM in a cultivar of Florakirk Bermuda grass were generally associated with longer grazing cycles and, consequently, with older regrowth. The authors also reported that grazing management treatments that cause more regrowth directed to the stem could decrease digestibility.

The highest forage masses for post-grazing in the irrigated system in winter (Table 2) reflect the greater biomass in the winter of the second year (Table 3). These are consequences of the adequate water supply. With the maintenance of adequate moisture levels in the soil, there is a higher production of biomass and an increase in the residual mass. The beginning of the second experimental year was marked by a drop in temperature and low rainfall (Fig. 2), changing the structure of the forage sward. According to Fioreze et al. (2020), water stress acts in several ways on plants, affecting photosynthesis, water and solutes translocation, and plant growth. Under drought, there may be a decrease in the rate of new leaves and tiller initiation, which causes a decrease in the total mass of available forage. Some buds may just remain dormant, leading to fewer tillers in a given time frame (Lemaire and Chapman 1996). Despite the differences observed in irrigation for forage mass and ADF in winter, the animal production in this period was similar between treatments (Tables 2 and 4).

The percentage of forage legumes in winter and summer was close to 6% but considering only the first year, it was at most 11% in dry matter (Table 2). In winter and spring, the yield of forage legumes was lower in irrigated areas, mainly due to competition in the canopy with grasses, which have higher growth efficiency (Khatiwada et al. 2020). Competition was the main reason for the change of annual legume species between the years of evaluation, aiming for a greater contribution of legumes in the mix. However, even with a low percentage in the mix, animal production had positive effects over 2 years, with averages of 291 and 260 kg LWG/ha/year. These values were higher for forage legumes and irrigation together than for the non-use of any techniques for the second and third years, respectively (Table 5).

Temperate forage legumes are more productive during winter–spring growth (Scheffer-Basso et al. 2002). Their contribution to the system is usually long term, mainly for annual forage legumes, due to the degradation of dead material and its incorporation as organic matter, improving soil fertility and favoring grass growth (Barcellos et al. 2008; Carvalho and Pires 2008). Such a trend was observed for ADG, LWG, and stocking rate in the summer (Tables 2 and 4). It verifies the residual effect of the legumes, as described in the results from the third year (Table 5).

4.2 Pasture responses on animal performance

Here we show for the first time, considering long-term results (3 years), the benefits of using forage legumes even with a low sward participation in the pasture. This fact was verified by the higher LWG achieved in systems with legumes regardless of irrigation. The 20% stocking rate increase in legume treatments in the summer of the third year (Table 5) shows that the residual effect from the previous 2 years was not observed in the short-term (Schmitz et al. 2020). However, it proves efficient for a consolidated system with proper management, optimizing the potential of forage species. The management of the forage allowanced was essential for the results since high stocking rates can cause a decrease in individual gains (Garay et al. 2004).

Legumes can benefit the system by increasing N, which improves biomass degradation, and improving the C:N ratio (Sartor et al. 2011; Assmann et al. 2014). Including appropriate legumes in a mixed pasture will result in increased N availability, great nutritive value, reduced emission of greenhouse gas effects, increased persistence, great cattle performance, and increased biodiversity, leading to high farm profitability (Khatiwada et al. 2020).

The irrigation effect was fast and significant in the spring, which resulted in a high stocking rate and LWG, regardless of the legume presence (Table 2). Another consequence of the great irrigation in September and October of the second year (Fig. 2) was high productive indicators (Table 3).

The irrigation results may vary. In the winter, e.g., the similarity between pastures and climatic conditions did not present any significant variations in animal performance. The LWG was higher without the irrigation (Table 3) in the spring of the first year, a consequence of adequate rainfall. It led to more intense regrowth of the Bermuda grass as there was less competition with the temperate grasses. The irrigated ryegrass in winter had a longer vegetative cycle (Volesky et al. 2003). It affected the transition to spring and probably released allelopathic compounds that inhibited the growth of the Bermuda grass (McCarty et al. 2010). Such responses regarding the use of irrigation and legumes are important, as they highlight that those legumes associated with the overseeding of temperate forages benefit perennial species. The lower LWG in treatments without legumes demonstrates that their presence improves the system’s sustainability over the years when it is appropriately managed (Peprah et al. 2018).

Irrigation in regions that present variable climatic conditions throughout the year is a safe technology for forage production and, consequently, the maintenance of planned animal production, but the soil type must be considered (Vogeler et al. 2019). The occurrence of meteorological phenomena that favors droughts is worrisome. However, the water demand, supply, and storage must be carefully evaluated for investment decisions (Monjardino et al. 2015). Thus, it is evident that irrigation is a tool that should be used in critical periods, with significant results in specific periods of animal production. It facilitates the farmer’s decision-making with the lowest risk, as they will be able to achieve the planned goal regardless of climatic conditions.

5 Conclusion

The use of legumes or irrigation does not change the nutritional values of African Bermuda grass pasture mixed with oat and ryegrass, allowing for excellent animal performance throughout the year.

Irrigation increases the pasture’s stocking rate and, consequently, the live weight gain per hectare in times of water deficit, as observed in the spring and winter of the second year or the summer of the third year of our study.

Here we show that forage legumes, in low percentage, in the grazing system with proper management can contribute to increasing animal production per area in the long term. The presence of legumes is greater without irrigation.