PRELIMINARY RESULTS REGARDING THE OPTIMAL SEED RATES OF SOME WINTER WHEAT VARIETIES

. Before to extended on large surfaces, the new wheat varieties need to be examinate to finding their requirements regarding technological components, like planting density. In accord with this demand, during the agricultural year 2021 – 2022 were tested at Oradea, in north-west of Romania, our news registered varieties Dacic and Biharia, beside of other 23 Romanian or foreign varieties of wheat, at seven seeding rates:200, 300, 400, 500, 600, 700 and 800 seeds/m 2 . Statistical processing of the yield and wheat features were dan by usual correlation (Pearson), Spearman rank correlations, the calculations of linear and quadratic equations of correlations between grain yield and seeding rate and calculation of linear, quadratic, exponential and logarithmic trend between the same characters, the trend of the most significant determination coefficient being utilised in graphical transpositions. Some genotypes (Voinic, Bogdana, Dacic, Cezara and Crisana) have their best ranks of yield at low densities while other (Gabrio, Consecvent, Anapurna and Abundent) at the greatest ones. Taking in consideration the average of ranks, the genotype Consecvent has the best average position, followed by Gabrio, Abundent, Bogdana and Voinic. These five genotypes have a good capability to adjust their yields components to vary densities of plants. At low densities (200, 300, 400 and 500 germinal seeds/m 2 ), the genotypes yield ranks are comparable but at high densities (more than six hundred seeds/m 2 ), the ranks are stronger effected by density. Every genotype has an optimum seed density, depending on its capacity to tolerate or no high density: Crisana-500 seed/m 2 , Biharia-600 seed/m 2 , Consecvent-600 to 700 seed/m 2 , Anapurna-800 seed/m 2 , etc. Some varieties, like Voinic, have a pour response to seeding rate, they yielded well even at small density, being able to compensate the reduced density by tillering capacity, number of grains/spikes, better test weight and bigger grain size.


INTRODUCTION
Bread wheat (Triticum aestivum L.) is one of the most important cereal crops in the world, in terms of area coverage and production (Behzad and Amani, 2020).It is a major source of nutrition for humans and livestock, its contribution of protein per year is as much as 60 million tons.
The creation of new varieties requires determination of their reaction to crop density which should be performed in some sites due to the microclimate specificity (Bokan and Malesevic, 2004).The new wheat varieties need to be examinate to determining their requirements regarding technological components, planting density being one of them.
To get maximum yield, it is necessary to use quality seeding and improved agronomic techniques such as optimum seeding rate, time of seeding, fertilizer application, irrigation, weeding, time of harvest, etc. (Behzad and Amani, 2020).Wheat yield and end-use quality depend upon the environment, genotype, and their interactions (Zecevic et al., 2014).Many studies indicated that environmental factors may have higher impact on the expression of yield components than genetic factors (Mandea et al., 2019).
Application of different seeding rates has significant effects on yield and yield components (Gebeyehu, 2020).Seeding rate above or below the optimum may reduce the yield significantly.It depends on the tillering capacity of the varieties, sowing date, soil conditions, cultivation intensity.In addition, seeding rate increases for varieties with larger seed size and low tillering capacity.
Optimization of seed rate with variety of wheat is essential to obtain optimum plant density for maximum grain yield and potential benefits (Zohaib et al., 2020).Optimum plant density is different for each variety and population density could be used to improve the wheat production (Casas et al., 2005).However, increase in seeding rate may only enhance production cost without any increase in grain yield (Behzad and Amani, 2020).
Optimum plant densities vary greatly between areas, climatic conditions, soil, sowing time and varieties.Since cultivars genetically differ for yield components, individual cultivars need to be tested at a wide range of seeding rates to determine their optimum seeding rate (Zecevic et al., 2014).Each new cultivar released from a breeding program should be tested at several seeding rates to fully understand the optimal seeding rate for maximum yield (Mehring et al., 2020).
Seeding rate is an important management practice affecting wheat yield (Lollato et al., 2017).Under optimal planting density conditions, a plant population can use water, mineral nutrients and sunlight most efficiently (Bokan and Malesevic, 2004).High plant density due to high seed rate don't always increase grain yield because of high interplant competition and availability of nitrogen and soil moisture (Zohaib et al., 2020).
In case of wheat, agronomic optimum plant density is depending on environment conditions, from 397 plants/m 2 for the low yield environment to 191 plants/m 2 for the high yield environment (Bastos et al., 2020).In Serbia (Zecevic et al., 2014), the optimal planting density of winter wheat should be about 650 germinating seeds/m 2 , what will produce enough good-quality spikes with adequate yield structure and quality.The planting density of the varieties Yugoslavia and Rana Niska should amount to 600 germinating kernels/m 2 .
Wheat productivity could be improved by ridge sowing with optimized seed rate for different varieties (Zohaib et al., 2020).Seeding rate for maximum grain yield can differ for diverse hard red spring wheat cultivars and is derived from a yield response curve to seeding rate (Mehring et al., 2020).
In Spain, in a Mediterranean climate, plant density is a factor of particular importance in wheat production systems because it can be controlled (Lloveras et al., 2004).Comparing six seeding rate (from 150 to 500 seeds/m 2 ), the plant densities giving the highest yield were at list 400 to 500 plants/m 2 for most of the varieties studied.In Belgium and northern France, the optimum spike densities for wheat are 475 to 500 spikes/m 2 and are normally obtained with a target density of 200 plants/m 2 at the end of winter.In north Mexico, the authors (Casas et al., 2005) recommended the use of 16 plants/m 2 for varieties Tepoca and Oasis and 24 plants/m 2 for another varieties.
In China (Yonggui et al., 2015), planting densities of up to 750 plants/m 2 are currently used to achieve high yield.It is possible because of introducing semi-dwarfing genes such as Rht-B1b and Rht-D1b in newer cultivars.However, high density reduces photosynthetic potential, provides a favourable environment for fungal growth and leaf disease development, and thus increases agronomic costs and causes yield loss.
Tillering potential is an important modulator of yield (Bastos et al., 2020).Based on this, a producer may select different seeding rates and genotypes with varying levels of tillering potential depending on given field environments.The seeding density for wheat should consider genotype tillering potential and location (Valerio et al., 2013).Genotypes with low tillering potential express higher effect on grain yield and ear weight, as a function of an increase in seeding rate.
The ability of wheat to produce tillers is characteristic when plant density is low (Casas et al., 2005).The varieties only differ in their ability to fill spaces, which could be an indicator of their capacity to subsist under competitive conditions by maintaining production.
The number of tillers increased consistently with increasing seeding rates (Gebeyehu, 2020).The highest average number of tillers per plant (6.0) were obtained at the seeding rate of 100 kg/ha while the lowest number of tillers per plant (2.2) was obtained at 150 kg/ha.Thickening plant density caused by changing plant height and stem thickness because of inter-specific competition to more absorption light.At higher seeding rates, competition among the plants started before the maximum tillering stage, which was manifested in a low increase in spike production.
Wheat varieties differ in their tillering capacity and therefore in their yield response to seeding rate (Lollato et al., 2017).Increasing plant population decreased the number of spring tillers sustained by the different varieties from more than eight tillers per plant to fewer than four tillers per plant.There were varietal differences in tillers per plant.
The spike number is greater than the planting density, being the results of productive wheat tillering (Bokan and Malesevic, 2004).The spikes obtained at higher planting densities had a smaller kernel number as well as lower thousand kernel weight values.
High tillering potential reduced agronomic optimal plant density mainly in high and low yield environments (Bastos et al., 2020).A compensation between heads per plant and kernels per head were the main factors contributing to yield with different tillering potential.
In a study regarding the effect of the variety and plant density on yield and other components of wheat (Casas et al., 2005), the number of tillers decreased drastically with increased plant density, plant height was affected too, the harvest index (HI) was not, while the number of spikes/m 2 was different for each variety.
Grain yield of bread wheat is a product of three yield components: the numbers of ears per unit area, the number of kernels per ear and individual kernel weight (Gebeyehu, 2020).Increasing sowing density from 200 up to 400 plants per m 2 , significantly increased yield and its components.At higher plant density, most kernels would fade at an early stage because of competition between growing kernels to absorb preserved matters and as the result, low kernels would produce.
Varieties have different abilities to compensate for low or high plant populations by modifying the number of tillers and consequently the number of spikes/m 2 , the number of kernels/spike or the grain weight (Lloveras et al., 2004).
Compensation between components is one of the main barriers to improve yield using this approach (Mandea et al., 2019).The yield showed the highest correlation with the number of grains/m2 and number of grains/spikes.A higher number of grains per spike is accompanied by a lower grain weight.Most correlation among the yield components were negative, illustrating the difficulty of combining in the same cultivar high values of more than one component.Cultivars showing positive or small negative deviation from the regression between negatively correlated yield components might be useful in breeding for reducing compensations between yield components.
The higher planting density increased lodging through increased leaf area index, plant height, centre of gravity and length of basal internodes as the important indicators for lodging resistance (Yonggui et al., 2015).Higher plant density could be used to select genotypes with lodging resistance in irrigated environments.
Crops compete for nutrients, water, and light (Yahuza, 2011).Yield-density relationship can be literally defined as a mathematical quantification of crop response to increase in plant population density.At low densities, the relationship between density and yield is typically linear function (y = a + bx).As competition begins as the density is increased, the relationship usually deviates from linearity such that the gradient declines until yield plateaus or subsequently declines.The situation where a decline in crop yield occurs at the highest densities as the density is increased further is parabolic.In these cases, yield can be quantified more appropriately by a quadratic function: y = a + bx + cx 2 .

MATERIAL AND METHODS
The experiment vas conducted during the year 2021-2022, in the experimental field of wheat breeding laboratory of the Agricultural Research and Development Station Lovrin.This experimental field vas placed near the town Oradea, in the North-west of Romania.
The soil is pre-latosol, a type with middle favourability for wheat: acid, low humus content, middle phosphorus supply, groundwater depth at 8 m. (table 1).
The autumn of year 2021 were normal et the beginning but unusually very worm and very humid, prolonged until the middle of December (table 2).The wintertime was mildly, the minimum temperature (-14.5 0 C) being registered for one single day (January 13).From January until July, the precipitation regime vas in deficit, associate with many days with temperature up to 30 0 C. By ensemble, the agricultural year 2021-2022 vas with + 1.2 0 C wormer than multiannual average and more drought, the deficit of precipitation being -223.2mm.Consequently, the atmospheric humidity vas decreased.
The wheat genotypes tested in this experiment consists in 25 Romanian and foreign cultivars and breeding lines, different regarding their precocity, height, type of ear, tillering capacity, resistance to different diseases, etc.The sowing date (11 October) vas optimal, the conditions for spring and development of the young plants being very favourable.The purpose of this experiment vas to determine the influence of plant density on grain yield and yield components and for this, we utilised seven seeding rates: 200, 300, 400, 500, 600, 700 and 800 seeds per square meter, the surface of plots being 5 m 2 .
To quantify the relation between the yield, yield components and genotypes, were utilised the usual correlation (Pearson method).To investigate the relation between two variables, the Spearman rank correlation test vas used, too.This correlation coefficient is appropriate to use especially when the data have violated parametric assumption, sample size is small and there is an outlier problem in data set (Temizhan et al., 2022).The formula for Spearman's rank correlation coefficient is: ]}, where: rsr = Spearman's rank correlation coefficient, with values between -1 and +1 di = difference between the two ranks of each observation n = number of observations.When rsr value is ≤ 0,1 correlation is negligible, between 0.10 and 0.39 is weak, between 0.40 and 0.69 is moderate, between 0.70 and 0.89 is strong and between 0.90 and 1.00 correlation is very strong.
In addition, for every genotype, were calculated the regression equations for response of grain yield to seeding rates: linear (y = a + bx) and quadratic (y = a + bx + cx 2 ).Also, linear, and non-linear regression analyses were used to test the grain yield response to variable wheat plant population.The best trend line model of interactions between characters (linear, exponential, power, logarithmic, polynomial order 2 and 3) were computed and then the most significant were presented in the figures.

RESULTS AND DISCUSSIONS
In table 3 are presented the genotypes yielding classifications for every seed rate (200, 300, 400, 500, 600, 700 and 700 seed/m 2 ).The 25 genotypes classifications are different from one seed rate to another, even the besides ones.Taking in consideration the yield average of all seven seed rates, the breeding line Consecvent has the best position between all 25 genotypes, followed by cultivar Gabrio and breeding lines Abundent and Bogdana.These genotypes probably have a good capacity to compensate the reduced number of plants by a good tillering capacity.The cultivars: Alex, Dumitra, Apache, Pitar, Falado and Cezara are placed in the latest positions of the genotype's top.
It can see that the averages of the yields are greatest from the smaller seed density to thickly: from 4942 kg/ha at 200 seed rates to 6170 kg/ha at 700 seed/m 2 but at 800 seed/m 2 , the yield average of the genotypes are smaller (5813 kg/ha).This is the first sign that, in our specific conditions of testing, the agronomic optimal plant density must be around 700 plants per square meter.
Because of the high number of dates and the absence of the repetitions, we were making option to analyze the rank of genotypes, for every seed rate (table 4) and to use Spearman rank correlation which gives the measure of the relation between variables.
According to this, it is more facile to evaluate the reaction of every genotype to plants densities and the effect of growing seeding rate on yield and yield components of wheat.Every genotype has a rank between all 25 genotypes, in every seven seeding rate.For example, the cultivar Biharia has the best rank at 600 seed/m 2 and the worthless at 800 seed/m 2 , the cultivar Crisana has the best rank at 200 seed/m 2 and the worthless at 800 seed/m 2 , etc.
If we take in the consideration the average of ranks, the genotype Consecvent has the best average position (with ranks average 2.3), followed by Gabrio, Abundent, Bogdana and Voinic.These five genotypes have their ranks in every seeding rate upper than the average of the respectively density.It can conclude that they have a good capability to adjust their yields components to vary densities of plants.
For a better understanding of the influence of seeding rate on the plant and spikes density, tillering capacity, disease attack (Septoria tritici), height, thousand kernel weight (TKW) and test weight (hectoliter mass), there were calculated the correlations between all these features (table 5).There were used the wheat features from all 175 plots (all genotypes in all seeding rates), the significant correlations values at 5% and 1% level of probability, positive or negative, being bolded.
The yield vas positive and significant influenced by seeding rate, plant, and spike densities and negative by tillering coefficient.The wheat yield correlate negative with TKW, when the yield increases, the weight of the grains decreases.The disease attack (Septoria tritici), which appear to the end of vegetative period, correlate significant positive only with test weight.The more evaluation symptoms of attack are intense, the test weight is greater.
The plants are higher when the density increase, because of competition for the sun light.Thousand kernel weight increase there where tillering, heigh and test weight (hectoliter mass) increase.These correlations have a general validity, exceeding the genotype factor.Between seeding rate, plant density and spike density exist a strong positive dependence, while with tillering ability the correlation is negative.In evidence mode, the coefficient of tillering decrease when the density increase.
Table 6 present the correlations of varieties ranks between the seeding rates (200, 300, 400, 500, 600, 700 and 800 seeds/m 2 ).For this time, we used the Spearman rank correlations because it will be appropriate especially when the data have violated parametric assumptions, sample size is small and there is an outlier problem in data set (Temizhan et al., 2022).The number of cases is 25 (the number of genotypes) and the Spearman rank correlation coefficient (rsr) has only positive significant values.
Its ca see that the correlation of genotypes ranks is strong between the seeding rate: 200, 300, 400 and 500 seed/m 2 (rsr ≥ 0.70).Also, Spearman correlation coefficient is strongly significant between 500 and 600 seeds/m 2 , 600 and 700 seeds/m 2 respectively 700 and 800 seeds/m 2 seeding rate.It can appreciate that, at low densities, the genotypes have yield ranks comparable but at high densities (more than six hundred seeds/m 2 ), the effect of density on genotypes yield average is stronger.The Spearman correlation coefficient is negligible or weak between seven hundred seed/m 2 and 200, 300,400 and 500 seed/m 2 .At seeding rate of eight hundred seeds/m 2 , the Spearman rank correlation is strongly significant only with seven hundred seeds/m 2 .The greater the difference in densities are bigger, the more genotypes' reactions are different.
When are calculate the Spearman correlation coefficient between the genotypes (table 7), results that a genotype correlate positive with someone and negative with another one.There are a group of genotypes that do not corelate with anyone: Ursita, Consecvent, Andrada and Codru.The variety Alex reacts different to seeding rate than Ciprian, the other one created at ARDI Lovrin.The reaction of the varieties Crisana and Dacic to seeding rate look to be similar, their genealogy being resemble.These varieties yielded more at middle seeding rate.Crisana correlate positive with Dacic, Pitar and Dumitra and negative with Dumbrava, Anapurna, Apache, Falado and Pibrac.On the other hand, the variety Biharia reacts to seeding rate in similar mode with Glosa and Otilia.
Between the varieties: Anapurna, Apache, Falado, Pibrac and Gabrio, the mark of correlations is positive, significant, or strong significant, their reaction being like the variation of seeding rate.
The response curves of the grain yields (y) with increasing seeding rates (x) for each variety are presented in table 8. Linear (y = a + bx) and quadratic (y = a + bx + cx 2 ) regression analyses were used to test the grain yield response to seeding rate.
Taking in account the average of all 25 varieties, the coefficient of determination is very significant for quadratic regression equation, having a great value (R 2 = 0.9815), and significant for linear regression equation (R 2 = 0.8666).That means that grains yield is very dependent by increasing of seeding rate.
There are three genotypes (Andrada, Cezara and Dumitra) which have not a significant value of determination coefficient, meaning that linear and quadratic equation are not adequate to describe the relation between seeding rate variation and yield response.These are exceptions because the yield responses of all others genotype to seeding rate increasing are significant for one or both equations.
Considering the regression to be linear, the grain yield increase with every one hundred seed/m 2 from 35,7 kg/ha (variety Dacic) and 52.8 kg/ha (variety Crisana) to 518.5 kg/ha (variety Falado) and 425 kg/ha (variety Anapurna).
At low densities, the relationship between density and yield is typically linear.(Yahuza et al., 2011).As competition begins as the density increased, the relationship usually deviates from linearity such that the gradient declines until yield plateaus or then declines.This situation, when yield decline at high density, correspond to a parabolic curve, typically to quadratic regression.In our experiment there are two exceptions: the varieties Anapurna and Apache, the only that have in equation the coefficient of x 2 with positive sign.That means that even at high density (800 seed/m 2 ), the trend of grains yield is ascendent.without a plateau, these two varieties tolerating well this density.Instead of, variety Biharia has the biggest coefficient of x 2 in regression equation (-119.29), the decline of its yield at eight hundred seeds/m 2 being abrupt.We calculated, for every variety and for all wheat plots, indifferent of variety, the trend curve for yield dependence to seeding rate.The equations calculated for trend lines were linear, exponential, quadratic, power and logarithmic, but the most significant coefficient of determination was for quadratic equation, in most cases.In general, the optimum seeding rate for the wheat is around of seven hundred  2), and for the range of varieties tested.

Figure 1. The medium response of wheat grain yield to the increase of seeding rate
There are some situations, like Crisana variety (fig.2), which has optimum seeding rate around of 500 seeds/m 2 , when realize around of 6500 kg/ha.The explication of this situations consists in the fact that this variety is tall, with greats spikes and flats leaves, so that the competition for light limit the number of plants.

Figure 2. The grain yield response of variety Crisana to the increase of seeding rate
Another variety, Biharia (fig.3), has the optimum seeding rate around of 600 seeds/m 2 , up to this density the grain yield declines abrupt, from more than 7000 kg/ha to less than 600 kg/ha.On the other hand, the variety Consecvent yielded similar at 700 seeds/m 2 , around of 7500 kg/ha, and after yield decline slowly (fig.4).
It is important to underline that the variety Anapurna has an increasing trend line of grain yield until eight hundred seeds/m 2 , probably more than this (fig.5).The reason of this situation may be its superior capacity to use the sunlight in photosynthetic processes.In opposition, the variety Voinic (fig.6) has a pour response to seeding rate, the grain yields rise slowly from the minimum yield (5960 kg/ha at 200 seeds/m 2 ) to the maximum (6600 kg/ha at 700 seeds/m 2 ).This variety yielded well even at small density, its capacity to compensate by tillering, number of grains/spike, test weight and grain size, being the probable explanation.Grain yield (q/ha) Seeding rate (seeds/m2) y = -0.0008xGrain yield (q/ha) Seeding rate (seeds/m 2 ) Figure 6.The grain yield response of variety Voinic to the increase of seeding rate

CONCLUSIONS
The genotypes: Voinic, Bogdana, Dacic, Cezara and Crisana have their best ranks of yield at low densities while the genotypes: Gabrio, Consecvent, Anapurna and Abundent at the greatest ones.
The genotypes: Consecvent, Gabrio, Abundent, Bogdana and Voinic have a good capability to adjust their yields components to vary densities of plants.
At low densities, the genotypes yield ranks are comparable but at high densities (more than six hundred seeds/m 2 ), the ranks are stronger effected by density.
Every genotype has a best seed density, depending on its capacity to tolerate well high density: Crisana at 500 seed/m 2 , Biharia and Consecvent at 600 seed/m 2 , Anapurna 800 seed/m 2 , etc.Some varieties (Voinic) have a pour response to seeding rate, they yielded well even at small density, being able to compensate by tillering, number of grains/spikes, test weight and grain size.Grain yield (q/ha) Seeding rate (seeds/m 2 )

Figure 3 .Figure 4 .Figure 5 .
Figure 3.The grain yield response of variety Biharia to the increase of seeding rate

Table 1 .
Soil characteristics from the experimental field in Oradea

Table 3 .
Genotypes yields classifications for every seeding rate

Table 4 .
Ranks of varieties for every seeding rate

Table 5 .
The correlation between features of winter wheat genotypes Seeding rate

Table 6 .
The correlations between the seeding rates of varieties ranks

Table 8 .
Yield response curves to seeding rate for tested varieties (fig.1).The grain yield increase with seeding rate until 700 seeds/m 2 , then stagnate and decrease at 800 seeds/m 2 .This result is valid in the case of the soil type from Oradea (table 1) and climatic conditions of the year 2022 (table