Productivity of natural grassland of the order Arrhenatheretalia depending on nitrogen fertilization level

Natural grasslands are major animal feed resources in the moutainous region of Serbia. Proper use of mineral fertilizers in these grasslands can lead to a multiple increase in productivity. The paper analyzes the effect of NPK fertilizers with different levels of nitrogen (unfertilized – A0; N60:P40:K40 – A1; N100:P40:K40 – A2; N140:P40:K40 – A3) on forage yield and botanical composition of natural grassland of the Arrhenatheretalia order on the slopes of Mount Kopaonik. With the increase of the amount of nitrogen applied in spring from 40 to 80 kg ha, forage and hay yields also increased, whereas further increase innitrogen quantity to 120 kg ha resulted in the absence of the increasing yield trend. In the second cut, there were lower forage and hay yields compared to the first cut. A total of 50 different plant species of 44 genera and 20 families were identified on the grassland. In both cuts, in all fertilized treatments, the grass percentage was significantly higher than the control. The percentage of legumes in the first cut was low, while it significantly decreased in the second crop with the increase in the level of fertilizer. The percentage of other plants was higher in the second cut, as compared to the first one, and it significantly decreased in both cuts with the increase in fertilizer rate.


Introduction
In hilly and mountainous regions of the Republic of Serbia, natural grasslands are the most widespread meadow-pasture communities (Lazarević et al., 2009).According to the last agricultural census (2012) (RZS 2013), natural meadows and pastures are present on an area of 713 242 ha.According to SGS (2015) in 2014, in the Republic of Serbia, meadows covered 381 654 ha, with a total dry matter production of 784 111 t and an average dry matter yield of 2.1 t ha -1 , and the surface area of pastures was 331 589 ha, with a total dry matter production of 561 527 t and an average dry matter yield of 1.7 t ha -1 The yield and quality of forage mainly depend on the floristic composition of the grassland, the soil, the quantity and distribution of precipitation, thermal conditions and light (Đukić et al., 2008).Under the same conditions, the judicious appropriate use of fertilizers can increase forage yield up to several times (up to 20 t ha .The production of livestock feeds in meadows and pastures is relatively low and unstable, and the main reason for this is the lack of cultural practices (Dubljević, 2007). -1 The aim of the research was to analyze the influence of NPK fertilizers with different quantities of nitrogen on production characteristics and floristic composition of the natural grassland of the Arrhenatheretalia order.
), while at the same time improving the quality of the forage (Stevanović et al. 2004;Nešić et al., 2004;Vučković et al., 2004).Nitrogen is one of the most important nutrients and the most common limiting factor for achieving high yields of natural grasslands (Vitousek and Howarth 1991;Frink et al., 1999;LeBauer and Treseder 2008;You et al., 2017).

Materials and methods
The experiment was designed in 2011, and analyses were carried out in 2016, on the natural meadow of the Arrhenatheretalia order in the village of Rakovac near Jošanička Banja on the slopes of Mount Kopaonik (Serbia) (43º23'35,89'N; 20º48'29,24'E , 990 m above sea level).The experiment was set up on a dolomite calcomelanosol soil (pH H 2 The experiment was set up in a completely randomized block design in three replications, with the plot size of 7.5m O 6.5), which was poor in readily available phosphorus, rich in potassium and moderately supplied with humus.The mean annual air temperature for the given area is 7.9°C, and the annual rainfall is 771 mm.In 2016, the mean annual temperature was 0.15°C lower, and annual rainfall was 249 mm higher than the long-term average.Monthly rainfall (Figure 1) showed large variations during the growing season (April-September).
2 (5x1.5m).The following fertilization treatments were used: A 0 -unfertilized -control, A 1 -N60:P40:K40; A 2 -N100:O40:K40; A 3 -N140:P40:K40 kg ha -1 .Fertilization involved introducing the total amount of phosphorus, potassium and most of the nitrogen immediately before the beginning of the growing season, and 20 kg ha -1 of nitrogen per treatment, except in the control, was left for additional fertilization, which was done after the first cut.In each following years, fertilization was carried out in the same way.Monthly rainfall (mm) Analyses were carried out for two cuts obtained in 2016.Just before mowing, the botanical composition of the meadow was examined.The identification of species was performed according to the keys of Josifović (1970) and Javorka and Csapody (1991).
Mowing was carried out at the stage of inflorescence appearance in the most abundant species, i.e. tall oat grass (Arrhenatherum elatius L.).Forage yield was determined by measuring the total plot biomass weight immediately after mowing.After the measurement, samples were taken for the analysis of the water content in the forage and botanical composition.The percentages, by weight, of grasses (fam.Poaceae), legumes (fam.Fabaceae) and other species in the forage were determined in the measured sample (1000 g).After drying the samples at 65°C, hay yield (t ha -1 ) was calculated.The results were analyzed by the method of analyzing the variance of a single-factorial trial (ANOVA) using the SPSS 4.5 software.The significance of differences in mean values of the treatments was tested by the LSD test.

Results and discussion
In the first cut in 2016, significantly higher forage and hay yields as compared to the control were achieved in treatment A 1 , in which 40 kg ha -1 N was used in the spring (Table 1).However, compared to treatments A 0 and A 1 , forage and hay yields were significantly higher in treatments A 2 and A 3 .Therefore, forage and hay yields increased with the increase in the amount of nitrogen applied in the spring from 40 to 80 kg ha -1 , but further increase in the amount of nitrogen to 120 kg ha -1 led to no increasing yield trend.As found by Dubljević (2007), fertilization of natural grasslands with nitrogen results in a significant increase in forage yield and raw proteins, enhances growth and tillering of grasses, increases grass cover density, extends the growing season and slows down plant aging.Vučković et al. (2004) and Alibegović -Grbić et al. (2004) also pointed out that nitrogen has a major impact on yield, nutritional value and floral composition of highland natural grasslands.As reported by Vučković et al. (2004), grass biomass yield was increased by 153% under treatment with 160 kg N ha -1 compared to control.Fertilizing grasslands with nitrogen increases carbon yield in the above-ground part, directly by entering the plant life processes (Xia and Wan 2008) and indirectly through faster mineralization of organic matter in the soil (Nowinski et al. 2008).Apart from the type and quantity of fertilizer, the yield of natural meadows is also significantly affected by the timing of fertilizer application (Lazarević et al. 2004).This is especially true for lawns at higher altitudes where regeneration possibilities after cutting are reduced.The same authors pointed out that fertilizing grasslands with nitrogen is not economically justified at altitudes above 1000m a.s.l.In all fertilized treatments in the first cut, the percentage of grasses was significantly higher than in the control (Table 2).The percentage of legumes was low and did not change significantly across fertilization treatments, while the percentage of other plant species was decreased with increasing fertilizer rate.Fertilization had a similar effect on the yield of grasses, legumes and other plants.Nešić et al. (2004), Stevanović et al. (2004), Stevens et al. (2004) and Vučković et al. (2004) showed that fertilization can significantly increase dry matter yield and quality of natural grasslands and change their floral composition.Fertilization of natural grasslands with nitrogen reduces plant diversity and the presence of legumes and C 4 plants, while increasing the presence of C 3 In this experiment, a total of 50 different plant species of 44 genera and 20 families (Table 3) were identified in the grassland.The most common was the Poaceae family, dominated by the species Arrhenatherum elatius, which forms a typical community Arrhenatherion elatiorius.
plants (Gough et al. 2000;Stevens et al. 2004).This phenomenon is the result of increased competition for light of the plants (Xia and Wan, 2008).In the second cut, forage and hay yields were lower than in the first one.The reason for this, among other things, was a smaller amount of precipitation in the second part of the growing season.At that time, only A 1 treatment showed significantly higher forage and hay yields than the control (Table 4).A significantly higher content of dry matter in forage was recorded only in treatment A 3 in relation to the control.The percentage of grasses in the second cut was significantly higher in treatment A 1 than in the control, and in treatments A 2 and A 3 than in the previous two (Table 5).The percentage of legumes was higher than in the first cut, and decreased with increasing fertilizer rate.The percentage of other plants was 84% higher in the second cut than in the first cut, and significantly decreased with increasing fertilizer rate.Grass yield was significantly higher in all treatments as compared to the control, while yield of legumes and weeds decreased significantly with increasing fertilization rate.

Conclusion
In the first cut, significantly higher forage and hay yields, as compared to the control, were reported in all fertilizing treatments and the highest yield was achieved in treatments A 2 and A 3 .In the second cut, significantly higher forage and hay yields were reported only in treatment A1, as compared to the control.Therefore, forage and hay yield increased with the increase in the amount of nitrogen used in the spring from 40 to 80 t ha -1 , but further increase in the nitrogen amount up to 120 t ha -1 A total of 50 different plant species of 44 genera and 20 families were identified in the grassland.In the first and second cut, in all fertilized treatments, the percentage of grasses was significantly higher than in the control.The percentage of legumes in the first cut was low, while in the second cut it decreased significantly with increasing fertilizer rate.The percentage of other plants was higher in the second cut than in the first one, and significantly decreased in both cuts with increasing fertilizer rate.resulted in the absence of the yield increasing trend.In the first cut, fertilization caused no change in the content of dry matter in the forage, while in the second cut a positive effect was achieved.In the second cut, forage and hay yields were significantly lower than in the first one.

Table 1 .
Effect of fertilization treatments (A 0 -A 3 ) on forage yield, hay yield and dry matter content in the forage in the first cut of natural grassland in 2016 * * nsValues followed by different small letters within columns are significantly different (P<0.05) according to the LSD test; * F test significant at P≤0.05; ns -non significant.

Table 2 .
Effect of fertilization treatments (A 0 -A 3 ) on botanical composition (% of legumes, grasses and other species) and hay yield of grasses, legumes and other plants in the first cut of natural grassland in 2016 * F test significant at P≤0.05; ns -non significant.

Table 3 .
Plant species in the grassland

Table 4 .
Effect of fertilization treatments (A 0 -A 3 ) on forage yield, hay yield and dry mater content in the forage in the second cut of natural grassland in 2016 * F test significant at P≤0.05.

Table 5 .
Effect of fertilization treatments (A 0 -A 3 ) on botanical composition (% of legumes, grasses and other species) and hay yield of grasses, legumes and other plants in the second cut of natural grassland in 2016