VARIABILITY OF AGRONOMIC TRAITS OF MAIZE HYBRIDS INFLUENCED BY THE ENVIRONMENTAL FACTORS VARIJABILNOST AGRONOMSKIH OSOBINA HIBRIDA KUKURUZA POD UTICAJEM FAKTORA SREDINE

In this study 36 maize hybrids of different FAO maturity groups were observed in three successive years (2011, 2012 and 2013), on 8 locations. The main objective of this experiment was to observe the GxE interaction concerning yield, grain moisture, grain yield per ear and test weight. The experiment was set up according to the RCBD. Based on the obtained results average estimates, CV and overall ranking of hybrids were calculated. ANOVA was applied in order to estimate the effect of factors: genotype, environment and interaction. Thus the significance of all these factors was observed. Results of this research indicate the importance and necessity of performing multilocation and multiyear trials with the aim of observation and understanding the intensity of GxE interaction, as well as its influence on the grain yield and it components.


INTRODUCTION
According to the sowing area of crops, maize is on the third place in the world, and on the first place in Serbia (Statistički godišnjak, 2016).That is due to the fact that maize has such a versatile usage from the unprocessed product for livestock feed, to that it is processed in many different industries such as food, pharmaceutical and more and more nowadays maize is row material for production of energy.The main goal in production of this crop is achieving high and stabile yields, and recently more and more, higher biomass.Grain yield beside the genetic potential is highly influenced by many factors such as: applied crop practices, soil fertility, and level of ground water, altitude, amount and distribution of precipitation, i.e. conditions of the environment.Filipović et al. (2015),emphasized that maize breeders are due to the global climatic changes, challenged to create highly adaptable genotypes, which are capable to produce high and stabile grain yields in different environments.Genotype stability in different environments is the consequence of its genetic structure, but there is a few information about genetic components that determine genotype stability, and how the selection and breeding has the influence on them (Lee et al., 2003).Factors that influence grain yield and thus the economic aspect of maize production are associated with polygene action, but are also under great influence of the environment.In the research of Pavlov and Crevar (2014), it was confirmed that beside the hybrid combination as the major factor, very important influence expressed environmental factors such as years and locations on the parameters of seed production.Therefore trials in maize breeding process that are focused on grain yield are performed on the larger number of locations and in several successive years.These experiments usually observe the relative success of genotype performances in different environments (Kandus et al., 2010).The aspect of the GxE interaction is very important in breeding programs and as well in the commercial introduction of new hybrids.Deitos et al. (2006) indicate that GxE interaction is important in breeding process because it influences the genetic gain, as well as the recommendation and choosing the varieties with high adaptability.Petrović et al. (2009) concluded that breeding for the targeted environment highly depends on the identification of the major sources of phenotypic variation in that region.In order to develop a variety or a hybrid which possess lower GxE interaction, for the dominant sources of variation, variety should have the balanced proportion between stabile and high yield (Boakyewaa, 2012).
The crucial impact on the maize grain yield has the amount and the distribution of precipitation.This crop responds very stressfully to the drought, especially in certain developmental phases.Filipović (2012) emphasized that high and stabile yields under our climatic conditions could be only achieved with the production under irrigation systems.In production years with pronounced drought, yield of certain hybrids could be lowered up to 80 %.
The aim of this research was to estimate GxE interaction for 36 maize hybrids of different FAO maturity groups on 8 locations and in three production years and to establish the intensity of these interactions.Based on these interactions genotypes that express highest stability and yielding potential in different production conditions and years should be selected.

MATERIAL AND METHOD
This research included 36 maize hybrids which were classified according to FAO maturity groups as followed: FAO 300: H1-H4; FAO 400: H5-H12; FAO 500: H13-H22; FAO 600: H23-H32; FAO 700: H33-H36.Trials were performed in the period 2011-2013, according to the RCBD in three replications on 8 locations in Serbia.Locations were selected according to the major regions in which this crop is most intensively produced (Sombor, Kikinda, Senta, Pančevo, Sremska Mitrovica, Svilajnac, Loznica i Šimanovci).Each hybrid was sown in four rows, out of which two inner rows were harvested, while outer rows were border rows.In each replication hybrids were sown according to different randomization so that GxG interaction could be excluded.The same sowing density was applied for all hybrids and the number of plants per ha was 62,643.Inter and intra row distance was adjusted to the mechanical sowing and harvesting.Following agronomic traits were observed: grain yield (JUS standard t/ha); grain moisture ( %); grain yield per ear (kg) and test weight (kg/hl).The three factorial ANOVA according to the fixed model (Steel and Torrie, 1960), was applied to analyze observed results.Comparisons between average estimates were performed by LSD test.The correlation between traits in this research was analyzed by Spearman's rank coefficient correlation (Zar, 1999).

RESULTS AND DISCUSSION
Grain yield: Significantly high variation of observed traits, especially maize grain yield was expected, considering the three very different production years in terms of the amount and distribution of precipitation during vegetation.
Significant differences were observed in amounts of precipitation, relative air humidity and average daily temperatures in certain critical developmental phases of maize, above all silking, pollination, and grain filling in the period July -August (Figure 1 and 2).The variation in grain yield of maize hybrids was significantly influenced by the difference in yielding potential, as well as the differences in FAO maturity group of hybrids.The highest average grain yield was achieved in 2011 (11.62 t/ha), while the lowest was in dry and stressful 2012 (6.90 t/ha).Average grain yields of observed hybrids in the first year of this trial ranged from 10.38 t/ha (H1) to 13.23 t/ha (H36).In the second year average grain yields ranged from 5.75 t/ha (H7) to7.87 t/ha (H3), while in 2013 average grain yields of hybrids ranged from 8.76 t/ha for hybrid H5 up to 12.01 t/ha for hybrid H36 (Table 1).Higher average grain yields were noticed in hybrids of longer vegetation in 2011 and 2013, while in 2012 the highest average grain yield had the hybrid H3 that belongs to the FAO 300.That could be explained by high temperatures and low air humidity during the period of pollination.Hybrids of shorter vegetation went through the silking and pollination phases, before temperature peaks occurred in July, therefore ear pollination was higher and the share of barren plants was lower.The highest average grain yield during three years and on 8 locations produced hybrid H36 (10.88 t/ha), while the lowest was observed in hybrid H14 (8.17 t/ha).Six hybrids had higher average grain yields compared to the trial average which were highly significant, and two hybrids had significantly higher average grain yields than the trial average.The highest coefficient of variation for grain yield in the observed years was established in the stressful 2012 (19.30%), while the lowest was in 2011 (9.93 %), which was also the most productive year.Kovačević et al. (2007) explained that highly significant differences between average estimates between years are expected and justified and that they point to the fact that meteorological factors during vegetation significantly influence yielding productivity, which confirms the significance of the GxE interaction, such was found in this research.That clearly points that the yielding variability is under the influence of the environment.
Grain moisture: Grain moisture is the relative indicator and it depends on the moment of harvest.The lowest grain moisture was observed in very dry 2012 (14.86 %), while between estimates in 2011(19.47 %) and 2013 (19.52 %) there were no statistical differences.In all three experimental years, harvest begun in mid September and ended by mid October.The order of harvesting certain location was under the influence of meteorological condition but it didn't differ significantly among years.Since the average daily and monthly air temperatures in 2012 were significantly higher than in 2011 and 2013 (Figure 1), hybrids accumulated sums of thermal units for specific developmental phases faster, including the phase of formation of black layer, i.e. the phase of physiological maturity.All this together with the warm and dry autumn of 2012, contributed the lower grain moisture at the moment of harvest.Videnović et al., (2011), also concluded that grain moisture is highly dependent on the factors such as genotype, the year and also sowing date.Average grain moisture of examined hybrids in the first year ranged from 14.55 % (H4) up to 22.91 % (H32).In the second year average grain moisture ranged from 12.64 % (H4) to 16.87 % (H35), while in the third year it ranged from 16.88 % for hybrid H19 to 23.07 % for hybrid H32 (Table 1).The hybrid from the FAO 300 H4 had the lowest average grain moisture (14.73 %), while the highest grain moisture was observed for the hybrid of FAO 600 H32 (20.67 %).The highest coefficient of variation in grain moisture was found in 2011 (13.07 %), while the lowest was in 2012 (7.51 %).
Grain yield per ear: Grain yield per ear is directly proportional to the grain yield, and as such it is very important indicator.The highest average grain yield per ear was observed in 2011 (0.171 kg), while such as average grain yield the lowest was in 2012 (0.089 kg).2).Average test weight was highest in 2012, and it was noticed in shorter season hybrids.This could be explained by the lower grain moisture at the moment of harvest, and because those hybrids avoided drought stress.On the other hand precipitation in September of 2013 was very high and with lower average daily temperatures, so the hybrids had higher grain moisture at harvested lower test weight.Similar variation in test weight in relation to the grain moisture at harvest in 2012 and 2013 was found in the research of Bowling (2014).Rankin ( 2009) also explains that it is essential to understand the relationship between grain moisture at harvest and test weight, because the dry grain is naturally smooth and slippery, so it provides better packaging.The highest test weight in this trial achieved H1 (75.07 kg/hl), while the lowest test weight had H31 (69.10 kg/hl).Highly significant differences between trial average and hybrids were found for 22 hybrids and for one significant difference.The highest coefficient of variation was found in 2012 (3.64 %), while the lowest was in 2011 (1.78 %).Test weight could be very interesting and important information to the farmers who are paying dry grain storage in big silos.The better the maize grain could be packed, the more kg could fit per hl, so the storage per ton is cheaper.
The correlation between observed traits: Spearman's rank coefficient correlation showed that in this trial there is significant positive correlation between grain yield and grain moisture (0.42*).Highly significant positive correlation was established between grain yield and grain yield per ear (0.91**).Moreover this correlation was very strong.On the other hand between grain yield and test weight correlation was negative and not significant (-0.33).Test weight was also in negative correlation with grain moisture and grain yield per ear, although these correlations were significant (-0.61**-0.41*,respectively).Highly significant positive correlation was found between grain moisture and grain yield per ear (0.61**).

CONCLUSION
Obtained results showed that the amount and distribution of precipitation, as well as average daily temperatures significantly influenced the variation of the observed traits and the ranking of hybrids.Analysis of variance showed that there were significant differences among genotypes, environments and GxE interactions.In the dry 2012 higher average grain yields produced hybrids of a shorter season, while in the years with average and above average precipitation higher yields were observed in hybrids with longer vegetation.On average for all three years hybrid H36 had the highest grain yield, and it was the most productive hybrid in 2011 and 2013.Similar trend was noticed for grain yield per ear for this hybrid.Grain moisture is a relative indicator and it depends on the speed of grain dry down of selected hybrids and the moment of harvest.Hybrids of longer vegetation had higher grain moisture at the moment of harvest than the shorter season hybrids.Average grain yields per ear were fully consent with the average grain yields of certain hybrid, which was confirmed by the Spearm's rank coefficient correlation.Test weight was in inverse proportion to the grain yield and moisture at harvest.Grain moisture is the most significant factor influencing this trait.Dry maize grain could be better packed, so the storage expenses could be lowered.Farmers lacking irrigation systems should be advised to sow hybrids of different length of vegetation, in order to achieve stabile yields.

Table 1 .
Average estimates and indicators of variability for grain yield (t/ha) and grain moisture (%) of 36 maize hybrids for all locations -significant at 0.05 and 0.01 probability level, respectively one year.This trait is in reverse proportion to the grain yield and grain moisture.Moreover grain moisture at the moment of harvest is the major reason for the variability of test weight.The highest test weight was noticed in dry 2012 (74.84 kg/hl), while the lowest test weight was in the year with the highest precipitation 2013 (70.47 kg/hl).Average test weight of examined hybrids in the first experimental year ranged from 69.05 kg/hl (H22) to 74.64 kg/hl (H13).In 2012 the lowest test weight had H36 (71.56 kg/hl) and H1 had the highest test weight (78.04 kg/hl), while in 2013 test weight ranged from 65.43 kg/hl (H31) to 74.14 kg/hl (H3) (Table