Effect of four irrigation doses and four varieties on agronomical characteristics and yield of cotton cultivation in central Greece

In recent years, reduced irrigation has been applied for the cultivation of cotton. While this strategy remains desirable, it needs to be evaluated as it affects the yields and growth of cotton. During 2015 and 2016, two similar cotton experiments (Gossypium hirsutum L.) were performed in Greece, in the area of Karditsa and especially in the location of Palamas. An experiment was established designed according to split-plot design, with main plots four levels of irrigation (IRR. 2, IRR. 4, IRR. 6, IRR. 8) and subplots four varieties of cotton (Dp 419, Campo, Andromeda, Lider) for two growing periods. The results indicated that agronomical characteristics were affected by irrigation dose, while LAI (Leaf Area Index) was affected by irrigation levels and year. LAI higher values were noticed in RR8 level for all varieties. Second year values of LAI were higher than the first year. Irrigation levels affected the number of closed bolls. Closed bolls were ranged from 12 (RR2 with variety Dp 419) to 144.67 (RR8 in same variety). The yield was significantly affected by irrigation levels and variety; the highest value was observed in second year by the Lider variety (3,110 kg ha-1). A strong positive correlation was mentioned between yield and total fresh weight (r=0.72, p=0.001). On a pooled basis, all varieties responded positively to the largest amount of irrigation (RR8).


INTRODUCTION
In cotton (Gossypium hirsutum L.) irrigation is an important factor during the cultivation and is necessary in all stages, from sowing to the splitting of the bolls. Irrigation season, frequency and quantity of water in each irrigation greatly affect the prematurity, height and quality of production and depend on many factors such as the mechanical composition of the soil, the variety, the early maturity of the plantation and the fertilization.
Water is a critical resource for summer crops in Greece, which usually has hot, dry summers and cool and wet winters that can vary from year to year. So, water conservation is becoming more and more important especially in recent years where the impact of climate change is becoming more intense and the periods of drought are longer. Groundwater is depleted in many areas by over-pumping [1].
The interest for the production with reduced irrigation in the cotton cultivation in Greece is increasing through years. To improve the overall management of irrigation water for the rational use of available water resources, various different irrigation technologies and strategies are discussed [20]. Increased irrigation costs due to declining water availability in the last decade, motivated growers to reduce irrigation water, but this reduction on the amount of water for irrigation, even today, raises doubts that will affect fiber yields and quality [4]. Producers with irrigation potential can irrigate to minimize part of the deficit trend [17]. DeLaune et al. [6] note that crops such as cotton (Gossypium hirsutum L.) or sorghum (Sorghum bicolor L.) can be included in a strategy to reduce irrigated water. Moreover, Karamanos et al. [10] emphasize that in Greece there is a need to consider different cultivation practices for soil conservation, since cotton as of great economic importance is cultivated in most cases as monoculture.
For the application of a deficient irrigation program it is necessary to check the yields of the crops before application either during certain stages of development or throughout the season [11]. Also important is the texture of the soil and how much water it can hold. Kirda [12] reports that fine-textured soils have the ability to retain more water. An agricultural practice that is proposed, in addition to reducing the amount of irrigation, in a range of crops, drip irrigation has been proposed as a means of irrigation [2].
Well, management practices that lead to efficient water use are imperative for the sustainability of irrigated cotton production. The objectives of this study were to evaluate the yield potential and accuracy of four commercial varieties in Greece with four levels of irrigation. Most studies focus on crop yields under reduced irrigation. It is certainly the central point of interest for the producer and in this study the performance is presented. But it is important to consider from the total yield weight what part the shoot, leaves and fruiting bodies occupy. Thus, in this study, the individual elements of the above ground part of the plant are investigated in detail. Also, for the evaluation of the growth of the plant, the development of the leaf surface is noted using the LAI index and the number of closed bolls.  (Table 1) and the subplots different varieties of cotton (Dp 419, Campo, Andromeda, Lider) ( Table 2). The soil properties in the experimental field are presented in Table 3.

Location and Experimental Design
Moreover, the total experimental area was 1,920 m 2 (4 x 120m 2 ). The mean temperature and precipitation during the experimental periods for both experiments are shown in Figure 1.

Measurements and Methods
In order to perform the measurements, οne hundred bolls were randomly selected per plot. All measurements concerned plant agronomic characteristics. The leaf area index (LAI) was determined using the SunScan devices (Delta-T Devices Ltd). Additionally, there were measured the fresh weight of leaves (g plant -1 ), shoots (g plant -1 ), upper parts (g plant -1 ), Darawsheh, et al.
fruiting bodies (g plant -1 ) and the total weight of them (kg ha -1 ) in 129 DAS. Moreover, there were estimated the dry matter of leaves (g plant -1 ), shoots (g plant -1 ), upper parts (g plant -1 ), fruiting sites (g plant -1 ) and the total weight of them (kg ha -1 ) in 129 DAS. Yield was estimated 150 DAS. For all parts studied, dry weights were determined after drying for 48 hours at 70 o C.
Another agronomic characteristic, that was measured, was the number of closed bolls per 10m. Finally, the total yield (kg ha -1 ) as well as the ratios between the total dry weight with the total fresh weight, the yield with the total fresh weight and the yield with the total dry weight (%) were determined.

Statistical Analysis
Analysis of variance was carried out on data using the STATISTICA (Stat Soft, 2011) logistic package as a split-plot design. For the significance of differences between treatments estimation, Tukey's test in significant 5% level (p=0.05) was used.

RESULTS
In the Table 4 Table 5). The varieties did not have a statistically significant difference between them in comparison to Table 4.
Concerning the dry matter of fruiting sites per plant in the 2015 the IRR.2 had not statistically significant difference with the IRR.4 and the IRR.6 had not statistically significant difference with the IRR.8 in all varieties (Table 6). But in the 2016 there were no statistically significant differences between     (Table 8). It was observed that as the increased the irrigation the yield is also increased. There was no statistically significant difference between the varieties (Table 8).
Additional, in the total dry weight/total fresh weight there was no statistically significant difference between the treatments in the 2015 (  (Table 9).

DISCUSSION
Ünlü et al. [21] said that dry matter yields increased as water use increased. Furthermore, Dadgale et al. [3] noted that the dry matter production increased as the frequency of irrigation increased. This was also observed in our study where leaf dry matter and stem dry matter increased as irrigation regimes increased and as the frequency increased, although lower values were shown in IRR.2 and IRR.4 treatments. A similar course was followed by fresh weight where from minimum irrigation regime to over-irrigation it increased. The dry matter of the upper parts consists of the leaves and shoots, which means that as the dry and fresh weight of the leaves and shoots increases with irrigation, so does the total weight.
Regarding fruiting sites, Reddell et al. [18] reported that early flowering is water sensitive. Also, Lashin et al. [13] observed that there was an increase in the number of flowers and bolls per plant and therefore in yield, as a result of the water stress that appeared in the stage before flowering. Thus, as the number of flowers increases, this has resulted in an increase in the fresh and dry weight of the fruiting sites. In our study, as the level of irrigation increased the dry and the fresh weight increased. The fresh and dry weight of upper parts had positive correlation with the LAI (r= 0.59, p=0.001) and with the number of closed bolls (r= 0.43, p= 0.001) as shown in Table 10. This means that increased irrigation results in increased LAI as well as higher number of closed bolls and therefore due to the positive correlation that exists there was an increase in both fresh and dry weight of upper parts.
The ratio between total dry weight and total fresh weight seems to be directly affected by irrigation. The highest values were observed mainly at the lowest irrigation doses. In addition, according to Table 10, there was a strongly positive correlation between this ratio and the ratio between yield and total fresh weight (r=0.72, p=0.001). In terms of the ratio between yield and total dry and total fresh weight, only the latter was affected Darawsheh, et al. by irrigation, while there was a positive correlation between them (Table 10) (r=0.57, p=0.001). Their highest values are presented in the irrigation regimes with the lowest quantities of irrigation water. These ratios inform us whether irrigation affects biomass production at the expense of yield. Therefore, it is observed that the higher the amount of irrigation is, as well as the higher its frequency is, it favors the crop production not necessarily yield.
In general, irrigation and genotype affect the evolution of the leaf area index [16]. Zhang et al. [22] stated that LAI had appeared reduced in deficit irrigation while it was high as irrigation increased. According to our results, the leaf area index LAI, was affected significantly by irrigation treatment. This is explained by the fact that irrigation leads to an increase in biomass production [5]. In addition, an increase in the irrigation dose led to an increase in LAI, up to 100 days after plant emergence, after which fruit production is promoted [9].
Moreover, the number of closed bolls per 10 m, was affected only by irrigation treatments. More specifically, as irrigation doses increased, the number of closed bolls also increased. Similar results were presented in a study by [15], in which it was reported that irrigation treatment negatively affected the number of closed bolls per plants. Mahadevappa et al. [14] said that the number of bolls per plant increased as such the irrigation increased.
Regarding yield, according to the results of the present study, it was affected by irrigation, variety as well as by year. The yield was positively affected by the irrigation treatment. In terms of variety, the highest yield value was observed in the Lider, in the IRR. 6 treatment. Onder et al. [15] reported that the yield of cotton seed was correlated with the number and weight of green bolls per plant. In addition, the number of closed bolls was strongly correlated with yield [7]. In contrast, Kang et al. [8] reported that the higher irrigation regime did not lead to an increase in yield. On the other hand Shinde et al. [19] observed that the higher dry matter production increased number of bolls per plant and as a result there was an increased in seed yield.
In terms of varieties, Lider recorded the highest yield values in the IRR. 6 treatment, presenting at the same time high values in all agronomic characteristics and low number of closed bolls. On the other hand, the lowest value of yield presented in Dp 419, although it had the fewest closed bolls.

CONCLUSION
The high production potential of cotton in relation to irrigation water remains a major issue today that can establish the future of the crop. By evaluating the agronomic characteristics, we can conclude that different irrigation regime had effect on them (LAI, yield, closed bolls) in cotton varieties. More specifically, higher irrigation regime treatment, the highest yields were recorded in all varieties in both years, while at the same level the highest values of closed bolls were recorded. According to varieties, no differences were presented in either of the two experimental years. Varieties Andromeda and Lider presented the best agronomic characteristics. A basic knowledge is structured while further research is needed on quality fiber and bolls.  Correlation coefficients are significant at the 0.05 probability level; ('ns': not statistically significant; *: statistically significant for a significance level of p <0.05; **statistically significant for a significance level of p <0.01; ***: statistically significant for a significance level of p <0.001