Forage yield and quality of simple and complex grass-legumes mixtures under Mediterranean conditions

R E G U L A R A R T I C L E *Corresponding author: Teresa Carita, National Institute for Agricultural and Veterinary Research (INIAV, I.P.). Phone: 00351268637759, Mobile: 351 964442687, E-mail: teresa.carita@iniav.pt Received: 14 February 2016; Revised: 19 March 2016; Accepted: 22 March 2016; Published Online: 29 April 2016 Carita, et al.: Forage yield and quality of grass-legumes mixtures 502 Emir. J. Food Agric ● Vol 28 ● Issue 7 ● 2016 by Mediterranean climate (warm to hot dry summers and mild to cool wet winter; rainfall in this area has a strongly seasonal pattern). Two winter cereals (Triticale (X Triticosecale wittmack), Oat (Avena sativa L.)) and five legumes (Grasspea (Lathyrus sativus L.), Red vetchling (Lathyrus cicera L.). Hairy vetch (Vicia villosa Roth), Common vetch (Vicia sativa L.), Field pea (Pisum sativum L.)) in pure stands as well as is mixtures were evaluated (Table 1). INIAV-Portugal cultivars were chosen to this study. A randomized complete block design with three replications was used in the experiment. Each plot consisted of 8 rows with 5 m length. The space between rows was 25 cm. Sowing was done by hand on November 14th, 2013. Dry matter yield was evaluated by two sequential cuts throughout the growing season. The plant samples were taken from a randomly selected 0,438 m2 (1,75 m x 0,25m) area of each plot. The first harvest time was based on (i) the occurrence of flowering (at 10% of plants) in legumes pure stands and (ii) at appearance of the first node (Zadoks stage 31) (Zadoks et al., 1974) for cereal pure stands and for all mixtures. Plots were clipped for a second time, in the end of April at cereal grain watery ripe GS71 according to Zadoks’s scale Zadoks et al., 1974). Plant sub-samples were taken from each plot, dried in a forced-air oven at 50oC and weighted to obtain dry matter yield. The crude protein content was measured by Kjeldahl method (ISO 5983-1: 2005) and the conventional factor of N x 6.25 was used. The NDF content (neutral detergent fiber) was measured according to ISO 16472.2005 with lower modifications. The in vitro digestibility was estimated according to the determination in vitro of the dry matter digestibility of samples that was performed by the two stage pepsin-cellulase enzymatic method (Jones and Hayward, 1975). Data were analyzed using SPSS version 17.0 software (SPSS Inc., Chicago, IL). The differences between means were separated by Tukey multiple range test (P ≤0.05). RESULTS AND DISCUSSION The annual precipitation for this agricultural year was 775 mm, and the mean monthly temperature varied between 8.9°C in December and 24.9°C in July. After sowing date, a very dry period of 1.5 months (rainfall below 15 mm per month) was reported. Therefore, emergence was low and in every plot the distribution of plants was irregular. Significant differences were observed among treatments in dry matter yield, crude protein content, neutral detergent fiber content and digestibility, indicating considerable genetic and agronomic variability for these parameters. Dry matter yield The dry matter yield recorded at two different growth stages of pure stands and forage mixtures is presented in Table 2. Average dry matter (DM) content increased with delaying cutting date from first to second cutting date. DM varied from 581 kg ha-1 to 3327 kg ha-1. From Table 2 it can be seen that, at cut I, the highest dry matter was obtained by the common vetch (T10) that was not significantly different from T8 compared to almost all the other treatments. For for both cuts and unlike another studies (Ansar et al., 2010; Alami et al., 2015), no significant differences were recorded between binary and ternary mixtures. This occurrence is not in agreement with the conclusions of Papadopoulos et al. (2012), who reported increased forage yields when mixture complexity increased. Likewise, Picasso et al. (2011) asserted that choosing a single well-adapted species for maximum productivity could not be the best alternative over the long term, and that high levels of species diversity should be Table 1: Species, mixtures, varieties, and seed density used in the experiment Treatments Seed density T1 Triticale (Fronteira) 160 kg ha−1 T2 Oat (Sta Eulália) 130 kg ha−1 T3 Grasspea A 160 kg ha−1 T4 Grasspea B 80 kg ha−1 T5 Red vetchling A (Grão da Gramicha) 130 kg ha−1 T6 Red vetchling B (Grão da Gramicha) 65 kg ha−1 T7 Hairy Vetch A (Amoreiras) 30 kg ha−1 T8 Hairy Vetch B (Amoreiras) 15 kg ha−1 T9 Field pea A (Pixel) 160 kg ha−1 T10 Common vetch (Graveza) 80 kg ha−1 T11 Triticale+Grasspea 60 kg ha−1+80 kg ha−1 T12 Triticale+Red vetchling 60 kg ha−1+65 kg ha−1 T13 Triticale+Hairy vetch 60 kg ha−1+20 kg ha−1 T14 Triticale+Field pea 60 kg ha−1+80 kg ha−1 T15 Triticale+Common vetch 60 kg ha−1+50 kg ha−1 T16 Oat+Grasspea 45 kg ha−1+80 kg ha−1 T17 Oat+Red vetchling 45 kg ha−1+65 kg ha−2 T18 Oat+Hairy vetch 45 kg ha−1+20 kg ha−3 T19 Oat+Field pea 45 kg ha−1+80 kg ha−4 T20 Oat+Common vetch 45 kg ha−1+50 kg ha−4 T21 Triticale+Grasspea+Hairy vetch 60 kg ha−1+60 kg ha−1+10 kg ha−1 T22 Triticale+Grasspea+Field pea 60 kg ha−1+60 kg ha−1+60 kg ha−1 T23 Triticale+Red vetchling+Hairy vetch 60 kg ha−1+45 kg ha−1+10 kg ha−1 T24 Triticale+Red vetchling+Field pea 60 kg ha−1+45 kg ha−1+60 kg ha−1 T25 Oat+Grasspea+Hairy vetch 45 kg ha−1+60 kg ha−1+10 kg ha−1 T26 Oat+Red vetchling+Hairy vetch 45 kg ha−1+45 kg ha−1+10 kg ha−2 Carita, et al.: Forage yield and quality of grass-legumes mixtures Emir. J. Food Agric ● Vol 28 ● Issue 7 ● 2016 503 included in the design of productive and ecologically sound agricultural systems. At the present study, in pure stand, there were significant differences among some treatments and between cutting time (Cut I and Cut II). The highest dry matter yield was obtained in oat (Cut II (cut at the end of spring without winter use) 6430 kg ha-1), which has typically a highest-yielding forage, and in common vetch (Cut I 2014 kg ha-1). Common vetch and hairy vetch are late flowering species, when compared with the other legumes. So, the total accumulated dry matters until occurrence of flowering (Cut I) were bigger. On the other hand the lowest dry matter yield was observed in red vetchling (Cut I – 133 kg ha-1) and grasspea (Cut II – 1266 kg ha-1). Crude protein content Forage protein content is often considered a good index of quality. The crude protein content of different treatments at different growth stages are shown in Tables 3 and 4. The average crude protein contents declined by increasing crops maturity (Mean Cut I: 16.7%; Mean Cut II: 13.0%). Crude protein content was always lower for triticale and oat pure stand and for binary mixture cereal + field pea (T1, T2 and T14) compared with Lathyrus pure stand treatments (T3, T4 and T5). One of main qualities of these forages crops (grasspea and red vetchling) consist of its high protein content (Polignano, 2007). The results of this study, show higher protein content than observed by Foster et al. (2014) and similar than Van Saun (2016) results. This is because, there were selected for this study various forage legumes varieties which generally produce high quality forage. Digestibility Concerning to the forage quality, digestible energy (digestibility) is the most common limiting factor. However, there are times when protein and minerals are the nutrients that limit animal performance, especially in grazing situations when supplementation is impractical. (Ball et al, 2001). Maturity stage at harvest is the most important factor that determines forage quality of a given species, including digestibility. According to Ball et al (2001), digestibility declines by 1⁄3 to 1⁄2 percentage units per day until it reaches a level below 50% (it may happen after the blooming period). In the first cut, digestibility values ranged in this study from 61.8% (T8-Hairy vetch) to 84.7% (T2-Oat) (Table 5). The beginning of flowering for vetch species (date of first cut) happened late in relation to all the other species in this study. Hence, this is related to the lowest values of their digestibility, which are significantly lower than those obtained from other treatments. As it is known, digestibility varies greatly. This parameter decreases as the plant matures. So, as expected, the value for this characteristic is lower in the second cut (average variation between cuts: 62% (Cut II) – 78% (Cut I). The lowest value in the second cut was obtained in T1-Triticale (49.3%) and the highest in the T9field pea (70.7%). Neutral detergent fiber (NDF) content Legumes usually have less neutral fibre content (NDF) than grasses which promote a higher intake of the former family by animals. Neutral Detergent Fibre content is important because they reflect the amount of forage consumed by the animal. The higher the NDF percent, the lower the dry matter intake. Favourable average NDF were recorded in Table 2: Dry matter yield (DM) of 26 treatments in both cutting times Treatments DM (kg ha‐1) Treatments DM (kg ha‐1) Cut I Cut II Cut I Cut II T1 658bc 3520abc T14 661bc 3276abc T2 434bc 6430a T15 409bc 4849abc T3 594bc 1266c T16 517bc 5366ab T4 351bc 1374c T17 571bc 3760abc T5 133c 1332c T18 530bc 4617abc T6 223c 2055bc T19 574bc 2721abc T7 447bc 1949bc T20 501bc 3904abc T8 1153ab T21 374bc 4777abc T9 918bc 2795abc T22 650bc 3045abc T1


INTRODUCTION
During the past three decades, the livestock sector is growing due to the global expansion in production and consumption of animal products (FAO, 2016).The cattle sector production costs are important, mainly as a result of feed costs, which represent more than 50 percent of the total cost (Sulas et al., 2012).This makes necessary to find solutions to provide high quality animal feed using sustainable production methods.The dominating aspect of Mediterranean climate on agriculture is the scarcity in rainfall and the erratic distribution along the year, showing a coincidence of drought and high temperatures during the summer season.Furthermore, it is also expected that in winter, temperature increase in the Mediterranean region, +1 to +2°C based in the majority of models (Dumont et al, 2011).It is believed these climate changes will affect plant growth and development along with crop yield, Increased inter-annual variability of rainfall and temperature may be another significant aspect of climate change, and this is of high ecological relevance.For this region a slight reduction of precipitation ranging between 0 an 10% in winter and 10 and 20% in summer is predicted (Leliévre et al., 2010).Thus, herbage fodder mixtures suitable to grow in these conditions and with high nutritional value in late winter/early spring (time of greatest need of fodder) should be found.Forage nutritional value depends on many factors and it has a direct effect on forage value, animal performance, and, ultimately on farm profits (Ball et al., 2001).The most important factors affecting the forage nutritional value are forage species and growth stage in the moment of harvest.Among the factors affecting forage quality, the objective of this study was to select species and fodder mixtures for high dry matter content and for high quality potential (crude protein, digestibility and NDF).

MATERIALS AND METHODS
The experiment was conducted during 2013/2014 at the National Institute for Agrarian and Veterinarian Research (INIAV) experimental station in Elvas, Portugal (38°53′ N, 7°08′ W, 220 m above sea level), under vertic-calcaricchromic Cambisol soil.Portugal is mainly characterized by Mediterranean climate (warm to hot dry summers and mild to cool wet winter; rainfall in this area has a strongly seasonal pattern).Two winter cereals (Triticale (X Triticosecale wittmack), Oat (Avena sativa L.)) and five legumes (Grasspea (Lathyrus sativus L.), Red vetchling (Lathyrus cicera L.).Hairy vetch (Vicia villosa Roth), Common vetch (Vicia sativa L.), Field pea (Pisum sativum L.)) in pure stands as well as is mixtures were evaluated (Table 1).INIAV-Portugal cultivars were chosen to this study.
A randomized complete block design with three replications was used in the experiment.Each plot consisted of 8 rows with 5 m length.The space between rows was 25 cm.Sowing was done by hand on November 14 th , 2013.Dry matter yield was evaluated by two sequential cuts throughout the growing season.The plant samples were taken from a randomly selected 0,438 m 2 (1,75 m x 0,25m) area of each plot.The first harvest time was based on (i) the occurrence of flowering (at 10% of plants) in legumes pure stands and (ii) at appearance of the first node (Zadoks stage 31) (Zadoks et al., 1974) for cereal pure stands and for all mixtures.Plots were clipped for a second time, in the end of April at cereal grain watery ripe -GS71 according to Zadoks's scale - Zadoks et al., 1974).Plant sub-samples were taken from each plot, dried in a forced-air oven at 50 o C and weighted to obtain dry matter yield.The crude protein content was measured by Kjeldahl method (ISO 5983-1: 2005) and the conventional factor of N x 6.25 was used.The NDF content (neutral detergent fiber) was measured according to ISO 16472.2005with lower modifications.The in vitro digestibility was estimated according to the determination in vitro of the dry matter digestibility of samples that was performed by the two stage pepsin-cellulase enzymatic method (Jones and Hayward, 1975).Data were analyzed using SPSS version 17.0 software (SPSS Inc., Chicago, IL).The differences between means were separated by Tukey multiple range test (P ≤0.05).

RESULTS AND DISCUSSION
The annual precipitation for this agricultural year was 775 mm, and the mean monthly temperature varied between 8.9°C in December and 24.9°C in July.After sowing date, a very dry period of 1.5 months (rainfall below 15 mm per month) was reported.Therefore, emergence was low and in every plot the distribution of plants was irregular.
Significant differences were observed among treatments in dry matter yield, crude protein content, neutral detergent fiber content and digestibility, indicating considerable genetic and agronomic variability for these parameters.

Dry matter yield
The dry matter yield recorded at two different growth stages of pure stands and forage mixtures is presented in Table 2.
Average dry matter (DM) content increased with delaying cutting date from first to second cutting date.DM varied from 581 kg ha -1 to 3327 kg ha -1 .From Table 2 it can be seen that, at cut I, the highest dry matter was obtained by the common vetch (T10) that was not significantly different from T8 compared to almost all the other treatments.For for both cuts and unlike another studies (Ansar et al., 2010;Alami et al., 2015), no significant differences were recorded between binary and ternary mixtures.This occurrence is not in agreement with the conclusions of Papadopoulos et al. (2012), who reported increased forage yields when mixture complexity increased.Likewise, Picasso et al. (2011) asserted that choosing a single well-adapted species for maximum productivity could not be the best alternative over the long term, and that high levels of species diversity should be

T26 Oat+Red vetchling+Hairy vetch
included in the design of productive and ecologically sound agricultural systems.At the present study, in pure stand, there were significant differences among some treatments and between cutting time (Cut I and Cut II).The highest dry matter yield was obtained in oat (Cut II (cut at the end of spring without winter use) -6430 kg ha -1 ), which has typically a highest-yielding forage, and in common vetch (Cut I -2014 kg ha -1 ).Common vetch and hairy vetch are late flowering species, when compared with the other legumes.So, the total accumulated dry matters until occurrence of flowering (Cut I) were bigger.On the other hand the lowest dry matter yield was observed in red vetchling (Cut I -133 kg ha -1 ) and grasspea (Cut II -1266 kg ha -1 ).

Crude protein content
Forage protein content is often considered a good index of quality.The crude protein content of different treatments at different growth stages are shown in Tables 3 and 4. The average crude protein contents declined by increasing crops maturity (Mean Cut I: 16.7%; Mean Cut II: 13.0%).
Crude protein content was always lower for triticale and oat pure stand and for binary mixture cereal + field pea (T1, T2 and T14) compared with Lathyrus pure stand treatments (T3, T4 and T5).One of main qualities of these forages crops (grasspea and red vetchling) consist of its high protein content (Polignano, 2007).The results of this study, show higher protein content than observed by Foster et al. (2014) and similar than Van Saun ( 2016) results.This is because, there were selected for this study various forage legumes varieties which generally produce high quality forage.

Digestibility
Concerning to the forage quality, digestible energy (digestibility) is the most common limiting factor.However, there are times when protein and minerals are the nutrients that limit animal performance, especially in grazing situations when supplementation is impractical.(Ball et al, 2001).Maturity stage at harvest is the most important factor that determines forage quality of a given species, including digestibility.According to Ball et al (2001), digestibility declines by 1⁄3 to 1⁄2 percentage units per day until it reaches a level below 50% (it may happen after the blooming period).In the first cut, digestibility values ranged in this study from 61.8% (T8-Hairy vetch) to 84.7% (T2-Oat) (Table 5).The beginning of flowering for vetch species (date of first cut) happened late in relation to all the other species in this study.Hence, this is related to the lowest values of their digestibility, which are significantly lower than those obtained from other treatments.
As it is known, digestibility varies greatly.This parameter decreases as the plant matures.So, as expected, the value for this characteristic is lower in the second cut (average variation between cuts: 62% (Cut II) -78% (Cut I).The lowest value in the second cut was obtained in T1-Triticale (49.3%) and the highest in the T9-field pea (70.7%).

Neutral detergent fiber (NDF) content
Legumes usually have less neutral fibre content (NDF) than grasses which promote a higher intake of the former family by animals.Neutral Detergent Fibre content is important because they reflect the amount of forage consumed by the animal.The higher the NDF percent, the lower the dry matter intake.Favourable average NDF were recorded in  treatments with grasspea (T3, T4) and red vetchling (T5, T6) on pure stand in both cutting times (Fig. 1).On the other hand, the highest average NDF content (62.63%) was determined in triticale harvested at cereal grain watery ripe -GS71 of Zadoks's scale (Zadoks et al., 1974).These results were comparable to values reported by Lithourgidis et al. (2006) and Kocer and Albayrak (2012).Neutral Detergent Fibre contents of all treatments increased by delaying harvesting; this was as expected because of the maturation progress of each species.

CONCLUSIONS
This study showed that for combined positive productivity and quality it is essential include legumes on fodder mixtures, especially those with higher contribution to increase the protein content and forage digestibility, as well as to decrease NDF content, such as, the genus Lathyrus.It has also been shown that increasing the complexity of the mixture had slightly different effects on the forage production and forage quality.

Fig 1 .
Fig 1. Neutral detergent fibre values of the 26 treatments in cut I (left) and in Cut II (right) cutting times (means followed by the same letter(s) are not significantly different at P<0.05 according to Tukey analysis).

Table 2 : Dry matter yield (DM) of 26 treatments in both cutting
Means followed by the same letter are not significantly different at P<0.05

Table 4 : Crude protein (CP) content of 26 treatments in second
Means followed by the same letter (s) are not significantly different at P<0.05

Table 5 : Digestibility (Digest) content of 26 treatments in two cutting times Treatments Digest (%)
Means followed by the same letter (s) are not significantly different at P<0.05