Effect of Location , Year and Variety on Winter Cereal Forage Yield and Quality in the Southern Plateau of the Spain

The objective of this research was to study the production and quality of forage at three different times of the year (April, June and July) of six winter cereals in the southern plateau of the Iberian Peninsula. The cereals studied were Triticale (xTriticosecale wittm) cv. “Tritano”, Oat (Avena sativa L.) cv. “Previsión” and cv. “Saia; Rye (Secale cereale L.) cv. “Gigantón”, Barley (Hordeum vulgare L.) cv. “Cameo” and cv. "Albacete”. The study was carried out in three different locations and over three successive years of harvesting. The three variables considered were location, year and cereal. The % dry matter (DM), % crude protein (CP), % acid detergent fiber (ADF), % neutral detergent fiber (NDF) and % ash content were determined for the three sampling periods and the quality was calculated in milk forage units (MFU/kg DM), the production in kg DM/ha, MFU/ha and kg CP/ha. In all three sampling periods the quality of the cereals was significantly influenced only by the year and by species. However, for production of dry matter (kg DM/ha), milk forage units (MFU/ha) and gross protein (kg CP/ha) all three variables were of significant influence as were their interactions. In the April sampling, the species which showed a significantly higher production (p≤0.05) was rye (1,693 kg DM/ha), which, along with its forage quality (16.56% CP, 0.886 MFU/kg DM) meant that the same occurred in MFU/ha and kg CP/ha. Significant differences between species were also found for the June sampling. The most productive cereal was again rye with 2,656 kg DM/ha, although its sharp fall in forage quality meant that barley cv Albacete (2,513 kg DM/ha) returned the highest production in forage units (1,934 vs. 1,951 MFU/ha) and barley cv. Cameo (2,413 kg DM/ha) in gross protein production (242 vs. 264 kg CP/ha). The significantly highest cereal production for July was barley cv. Albacete (4,923 kg DM/ha, 9.11% CP 0.722 MFU/kg DM). As a consequence of the results, we conclude that from the viewpoint of nutritional quality and production, rye is the most suitable for use in early spring in whatever year and location. However, barley cv. “Albacete” is the most appropriate for utilisation in later spring or early summer. (


INTRODUCTION
The southern plateau of the Iberian Peninsula in Spain lies at an average height of 500 meters above sea level, and has a typical semi-arid Mediterranean climate.Rainfall is low (300-500 mm/year) and of a highly irregular distribution over the year, with autumn and spring being the wettest periods.The overall annual temperature stands between 14 and 15°C while the majority of soils are from calcareous rock of sedimentary origin.
Sheep farming is common in the area (3.4×10 6 head according to MAPA 2004) as a subsidiary activity to agriculture.Flamant and Casu (1978), among others define it as a Mediterranean system based on complementary binomial cereal-sheep system.The main forage resources are natural pasture, shrublands, fallowlands and stubblelands.The climatic conditions that these resources are available for only short periods and are of low quality (Caballero, 2001).In general, the most critical period for food for these animals is spring and early summer since these agree in the majority of cases with the end of the gestation period and with suckling and so the flocks depend almost exclusively on trough foods.The lack of resources is further aggravated by crops being abandoned because of low harvests, which means that the flocks do not have access to these by-products.Different authors stated that one way of complementing the scarcity of natural pastures or the lack of sub-products would be to cultivate traditional cereals of the area under dry conditions, that can be grazed.These cereals could be barley, oats (Joy and Delgado, 1989;Francia et al., 2006), or rye (Gomez and Tarraga, 1991;Ciria et al., 1996).
Worldwide, there are a number of studies on the forage quality of cereals but mainly they deal with grain yield and occasionally with hay (Han and Kim, 1996), silo yield and quality (Kim et al., 2001aand b with rye, Shaoa et al., 2005vith oats and Zahiroddini et al., 2006with barley, Mahanta and Pachauri, 2005 with sorghum) in wetter conditions (Mengel, 2005) or under irrigation (Moret et al., 2007).Our study was focused on determining the potential of various winter cereals in different locations, in different years and at different periods: early spring, later spring and early summer.
Sowing was performed using a precision seeder in dry conditions in experimental units measuring 1.2×16 m.Doses were 150 kg/ha of seed with one fertilization of 20, 41 and 20 kg/ha of N, P and K, respectively.Dates of sowing and sampling are given in Table 1.Sampling was carried out in experimental units by measuring four times with a 25×25 cm metal square and cutting at a height of 2 cm.A 1 kg sub-sample was randomly taken for analysis from each of the sample.The samples were dried at 60°C for 72 h in a forced-air oven to determine the dry matter content.Samples were then ground, homogenized and passed through a 1 mm sieve.
The analyses performed were as follows: the ash content, following incineration at 550°C, crude protein content, using the Kjeldahl method (CP) (A.O.A.C. method 1990).Acid detergent fiber (ADF) and neutral detergent fiber (NDF), were determined by the methods of Goering and Van Soest (1970).Net energy was calculated (in milk feed units MFU) using the INRA method (1988) where the values for the digestibility coefficients were those given for similar phenological conditions in INRA (1988) and Option Méditerranéenes (1990).
Data were treated using the analysis of variance procedure (ANOVA) through the computer program SPSS + for Windows Realse 12.0 (2003).When significant differences were found, the Least Significant Difference (LSD) test was used for p≤0.05 to separate the means of the location, year and cereal variables (Steel and Torrie, 1980).For every sampling periods the considered fixed factors were: location, year and winter cereal.

RESULTS AND DISCUSSION
Precipitations and temperatures for the years under study are given in Table 2. Cumulattive precipitations over 25 years are very similar for Chinchilla and Bonete and higher (about 14 mm) than those of Aguas Nuevas.Accumulated precipitation during the first year is close to the mean in the three localities while those for the second and third years are 100 mm lower than the average.Mean temperatures over 25 years are very similar for the three locations (13.6, 13.5 and 13.6°C respectively).Average temperatures for the first year in the three locations are similar to those for the 25 year series, while for the second and third years they are higher.
Table 3 shows the significance of the variables, location, year and winter cereal and their interactions on the quality and production parameters.None of the three sampling periods showed significant differences between the three locations for the qualitative components (CP, ADF, NDF and MFU).Though, there were significant influence according to the location for productive parameters (kg DM/ha, MFU/ha and kg CP/ha).For this reason, the mean values for the three locations are shown in Table 4. Nevertheless, others authors have reported significant differences between locations for all components.For instance, Delogu et al. (2002) for varieties of triticale in Italy or Kim et al. (2005) for varieties of rye in South Korea.
In general, the values obtained for the qualitative components in the first two years were similar, and differed from those of the third year.
The differences in CP content (mean for six winter cereals) between the first and second years were not significant, although they were with respect to the third year for p≤0.05.Generally, the CP values obtained in the first and second years were higher than those for the third (Table 4).The study of the interactions among the location, year and cereal factors for CP content showed that the only significant result was the year×cereal interaction in the July sampling (Table 3), indicating that in the third year, the CP content was particularly low for the rye.The CP content in all cereals falls as the crop ripens, from April to July in whatever year, which is in line with what was reported by Tedla et al. (1992) for oats, Hadjipanayiotou et al. (1996) for oats and barley (Coffey et al., 2002) for rye and Delogu et al. (2002) and Zamora et al. (2002) for triticale.
In the April sampling, the cereal with highest mean CP content was rye with 16.56%.According by other studies, the results in the same vegetative stage are very variable (Options Mediterraneennes 1990; Ciria et al., 1996) to the extent that authors such as Kim et al. (2005) report values in the same study that range from 12.9 to 22.4%.The lowest values were returned by BarC and BarA, with 13.17 and 12.23% respectively.In the June sampling, OatP (11.67%) and OatS (11.42%) were the highest CP content.Rye and Trit fell appreciably in their CP content, and gave the lowest values, at 9.11 and 9.32%, respectively.For the July sampling, OatP presents the highest value in CP content, with 9.80%, followed by the two BarC and BarA varieties.
The lowest values for this sampling were found in rye, due mainly to the value of 5.16% recorded in the third year, which was associated with a very low grain production.The differences found between the oat varieties were significant in the July sampling.Differences between varieties of barley were significant for the April and June sampling.
The values for ADF and NDF in the first year were generally lower than those in the second and third (Table 4).This figure could partly be explained by a higher rainfall in spring for the first year.The interactions between the three factors varied according to the type of fiber considered and the sampling period (Table 3), with the greatest differences appearing in the ADF interactions in the July sampling, showing that the model response front annually climatic conditions are different for each cereal.
As with other forages, the chemical composition varies with the growth stage of plants (Firdous and Gilani, 2001;Kim et al., 2001b).These variations may also affect the concentration of micronutrients minerals (copper, iron, zinc, and manganese) according to Khan et al. (2006 and2007).The fiber content increased as the crops ripened, which is in agreement with what has been reported by authors such as Tedla et al. (1992) with oat, Hadjipanayiotou et al. (1996) with barley and oat, Delogu et al. (2002) and Zamora et al. (2002) with triticale.In the April sampling we found significant differences in the ADF and NDF contents of the different cereals.The lowest values of ADF and NDF content appeared in the two oat varieties.In this sampling, BarC presents the highest value of ADF with 26.25% while for NDF, it is rye, with 45.36%.In the June sampling rye returns the highest ADF values (32.61%) and the highest  NDF value is OatP, with almost 50%.In July the lowest fibre values were found in the two varieties of barley and the highest were in the two varieties of oat.No significant differences were found between the three locations for the values of the forage quality expressed as MFU/kg DM (Table 3).The effect of the variables and their interactions appeared as similar except for the June sampling, where no significant effect was found in the interaction between year×cereal and location×year×cereal.In the April sampling, values ranged from the 0.93 MFU/kg DM of oats and the 0.86 MFU/kg DM in the two varieties of barley.The net energy values decreased in the following sampling, which was consistent with what Zamora et al. (2002).In June it was the two varieties of barley that showed the highest values (BarC 0.79 MFU/kg DM and BarA 0.78 MFU/kg DM), while the cereal with the lowest value was rye, with 0.73 MFU/kg DM.Oat and Trit showed the lowest quality in the July sampling due to the very low proportion of grain returned by the cereals, especially in the second and third years.
The differences in the production of dry matter (kg DM/ha) were significant for the variables location, year and cereal.Likewise significant are all the interactions in the three samplings performed, with the exception of location× cereal and location×year×cereal in the June sampling.When considered individually or in terms of interactions, the effects of the different variables showed a similar pattern for the production expressed in MFU/ha and kg CP/ha (Table 3).
Dry matter productions in the April, June and July samplings in the CHMA and BONE locations were very similar (Table 5), and they were always higher than those of AGNU (p≤0.05).For all the samplings in all the locations, the first year was significantly more productive because of a greater abundance of rainfalls.The average accumulated rainfalls and temperatures in the second and third years were very similar and yet the productions of the second year were significantly higher than those of the third.The explanation may lie in the fact that the rainfall between February and June (late winter and spring) in the third year was much lower than in the second (117.2mm vs. 55.7 mm en AGNU, 115.1 mm vs. 81.9mm in CHMA and 108.6 mm vs. 77.5 mm in BONE).Dry matter productions differed significantly between cereal species in all the samples.In April, rye stood out as the most productive with an overall average of 1,693 kg DM/ha.The biggest production was 2,344 kg DM/ha for the first year in the BONE location, while the smallest was 952 kg DM/ha in AGNU in the third year.Other authors (Kim et al., 2005) reported rather higher values (near to 6,000 kg DM/ha) for rye.These differences may be accounted for by the varieties used and/or the type of soil since rainfall and temperature did not differ much from those in the study cited.The least productive species were OatP (1,094 kg DM/ha) and OatS (872 kg DM/ha), with a minimum of 717 kg DM/ha for OatS in the location of CHMA in the third year.
Significant differences also existed in dry matter production in the June sampling although these was not so marked as in the previous sampling.Rye production at 2,656 kg DM/ha was equal to that of BarA, with 2,513 kg DM/ha.OatP and OatS again returned the lowest values in June (1,770 and 1,936 kg DM/ha).Noro et al. (2003) found that rye was the most productive when they compared it to triticale and barley in a study covering the same period.In the July sampling, BarA production (4,923 kg DM/ha) was significantly higher than that of rye with 4,562 kg DM/ha.Average production for the other cereals was below 4,000 kg DM/ha.An observation of the growth dynamics of the crops led us to deduce that there are differences between the cereals.Rye had a high initial growth rate.Rye was the most productive cereal in the April sampling and then it leveled out.The barleys showed a more regular growth rate over the cycle, with BarA being the most productive in the last sampling.Unlike rye, OatP and OatS, showed a slow growth rate at the beginning and a more accelerated one at the end, although they never reached the productions of the other cereals in the climatic and soil conditions of this study.
In terms of MFU/ha (Table 6) the production increased significantly in each sampling since the increase in production in kg DM/ha always made up for the decrease in forage quality (MFU/kg DM).The only exceptions were the April production of rye in CHMA, which was similar to that of June (908 vs. 887) and the April production of OatS in BONE, which was higher than the June one (898 vs. 748).When expressing production in terms of kg CP/ha, rye presented lower values in the April sampling in the second and third years than in the June ones.The same phenomenon occurred in the location of AGNU in the first year.For all the other species, the amounts of kg CP/ha increased with time, with the greatest quantities coming in July and the lowest ones in April.This result was consistent with what was stated above about the dynamics of forage production and reflects that a sharp fall in the metabolic growth and development rate of the plant led to a lower proportion of protein content.The highest CP productions were observed in year 1; 663 and 575 kg/ha, in BarA and BarC, respectively (Table 7).
Given these results, we can conclude that in the conditions of our study location influenced on forage production but not on forage quality.The year was of influence because of the climatic conditions which occurred therein, with lower production and lower quality forage in years of lower rainfall.When comparing productions, rye was the most suitable cereal for early exploitation on account of its initial growth rate.For later exploitation, barley cv "Albacete" was the most appropriate.

Table 1 .
Field characteristics and culture practices at three locations and three years AGNU = Aguas Nuevas; CHIN = Chinchilla and BONE = Bonete.

Table 2 .
Mean air temperature and precipitation at three locations Source: Meteorological National Institute.Locations: AGNU = Aguas Nuevas; CHIN = Chinchilla and BONE = Bonete.

Table 3 .
Significance of main effects and their interactions in analysis of variant for quality and forage yield of winter cereals