Reference values for fruit analyses at early vegetation and differences among cultivars in apple trees

Abstract Fertilizers are commonly applied to improve the yield and quality in orchards. Leaf analyses are usually used worldwide for conscious fertilization. However, there may not be linear relationships always between leaf and fruit nutrient contents. On the other hand, fruit quality problems are directly related to fruit nutrient contents. Therefore, fruit analyses assessed together with leaf analyses are more reliable in assessing nutritional status of orchards. Fruit analyses on the other hand are possible with the reference values to be established for fruits. In this study, 260 apple orchards were selected from Isparta province of Turkey in where apple is cultivated intensively. Fruit samples were taken at 6 different periods covering the timeframe from June drops to harvest and the reference values representing deficiency and excess limits for N, P, K, Ca, Mg, Fe, Mn, Zn and B in each period were determined. To determine the reference values for all sampling periods, the regression curves were created by using the reference values of each period. A decrease was observed in reference values of all nutrients with increasing number of days from the full bloom. The cultivar-induced differences were also put forth in assessment of fruit analyses. Cultivars were compared in 68 orchards composed of different cultivars grafted on seedling rootstocks and at full-yield in two districts with different soil characteristics. Cultivars were found to be significant for all nutrients, except for N, Fe, Cu and Zn and the greatest values were observed in Granny Smith.


Introduction
Beside basic nutrients, fruits are also rich in mineral salts mostly formed through combination of Ca, Mg, K, Na and Fe-like minerals with fruit acids. These mineral slats have various positive impacts on human health. Fruits and vegetables with low energy and high mineral and vitamin contents are important foods for nutrition and human health (Sezgin 2014).
Among the fruits, apple is grown in various parts of the world and consumed year-long in four seasons. World annual apple production is about 87.236.000 tons and Turkey with 3.619.000 tons production has the forth place worldwide (FAO 2019). Isparta province with about 550.000 tons apple production constitutes around 20% of apple production in Turkey (T € U _ IK. 2020).
Although varied based on cultivars and species, about 10-30% of annual fruit production of Turkey is spoiled and wasted throughout the supply chain from producers to consumers. When the losses experienced in fruits and vegetables were assessed through different stages, it was observed that 3-10% was lost during the storage ( € Ozdemir et al. 2003). Majority of storage losses are resulted from physiological disorders. Nutrients have significant impacts on fruit quality and physiological disorders. N, P, K, Ca and B are the most significant nutrients encountered in fruits (Fallahi et al. 2010). N-deficiency retards ripening, results in early fruit coloration, reduces storage capability and increases sensitivity to some storage diseases (Anonymous 2006). On the other hand, high N levels generally reduces color development (in red and yellow cultivars) and flesh firmness. Considering all the other factors as equal, a 0.1% increase in N level results in 5% decrease in fruit color. When the leaf N level drops below 2.2%, periodicity tendency increases especially in sensitive cultivars (Hoying, Fargione, and Iungerman 2004). P deficiency results in harder flesh firmness, but reduces sugar content and taste of fruits. Low P contents of fruit flesh (less than 0.011%) increase 'low temperature damage' during the storage (Bergmann 1992). Sufficient P fertilization reduces water core and increases fruit antioxidant capacity (Neilsen et al. 2008). In a previous study, the best tree growth was achieved at leaf P levels of over 0.22% and fruit P levels of between 100 and 200 ppm (Neilsen et al. 2008). K is an especially significant nutrient for fruits. Fully colored, quality and alluring fruits with high sugar contents are only possible with sufficient potassium supply. Fruit set may be normal in Kdeficient fruit trees, but fruits will be smaller than normal size, have dull colors and be tasteless because of insufficient acid levels and be thick skinned (Hoying, Fargione, and Iungerman 2004;Stiles 1994). For optimum yield and quality, K levels of apple leaves should be between 1.4 and 1.8% and fruit K levels should be between 0.12 and 0.15%. However, since there is an inverse relationship between fruit load and leaf K level, even the leaf K level of 1.3% was found to be sufficient at high-yield orchards (Hoying, Fargione, and Iungerman 2004;Bergmann 1992). High K levels in leaves and fruits negatively affect fruit quality and storage characteristics because of antagonistic effects between K and Ca-Mg (Bergmann 1992). Ca deficiency symptoms are more remarkable over the fruits than over the leaves. Ca deficiency is commonly manifest itself as abnormal peel tanning, darkening of lenticels toward to harvest and severe fruit cracking at harvest . However, 'bitter pit' is the most significant symptom of Ca deficiency. Bitter pit in apples is usually forms toward to harvest or during post-harvest storage. It is a physiological disorder creating malformations over the rind and manifests itself with brownblack pits (Meheriuk et al. 1982). Besides, core browning, jonathon spot, water core, low temperature and aging disorders are also among the physiological disorders experienced under Ca-deficient conditions . Although varied based on cultivars, especially lenticel browning, water core and cracking are commonly observed when the Ca content dropped below 40 ppm fresh weight (Perring 1984). Ca-K interaction is one of the most common processes in fruits. Fruit K:Ca ratio is a significant indicator for bitter pit and other physiological disorders to be experienced in autumn (Drahorad 1999). Dilmaghani et al. (2004) indicated that K:Ca ratio should be between 0.9 and 1.4 in leaves and between 19 and 46 in fruits. In Mg fertilization, it should be taken into consideration that resultant K:Mg ratio should not exceed 1.5 (Anonymous 2006). Excessive Mg treatments may result in bitter pits over the fruits grown in sandy soils with low buffering capacity . Malformations and corks can be experienced under B deficiency. Fruits are smaller than regular sizes, cracks are observed and early fruit drops are experienced under B-deficient conditions (Stiles 2004). Cork spot observed under B deficiency is similar to bitter pit observed under Ca deficiency (Meheriuk et al. 1982).
Foreknown fruit nutrient contents can be used to take some pre and/or post-harvest measures. Pre-harvest Ca fertilizations and potential physiological disorders during the storage can be estimated by using fruit nutrient contents. The earlier the fruit analysis is performed, the earlier the possibility of intervention in nutritional imbalances is possible. Uçgun et al. (2021a) determined that nutrient content of the fruits at the time of harvest could be monitored after the June drop since there is a correlation between nutrient contents of the fruits at the time of harvest and the contents of the fruits in the previous periods. But, as it was in plant tissues, fruit analyses also require reference values. In this study, reference values were obtained for long-interval fruit analyses in apple trees. The differences among cultivars in terms of these reference values were also put forth in this study.

Selection of apple orchards and collection of plant and soil samples
A total of 260 orchards were selected from different sections of Isparta province of Turkey with intensive apple cultivation. Sampled orchards were mostly single-culture and were larger than 20 decares in size. Coordinates of sampled orchards were taken with a GPS device and marked with strips. Fruit sampling from the selected orchards was initiated 42 days after full bloom and samples were taken on 56, 77, 98, 119 and 140th days (6 periods) in successive years. One sample was taken from each orchard in each sampling period. Soil samples were also taken from the selected orchards at 0-30 cm and 30-60 cm soil depths to find out the soil characteristics of the region.

Plant mineral analyses
Fruit samples were brought to laboratory and immediately washed through tap water, then they were washed through 0.1 N HCl and finally they were washed through deionized water and roughly dried out with drying papers. They were divided into small portions as whole fruits, placed in paper bags and dried at 65-70 C in a drying chamber until a constant weight (for about 96 hr). Dried samples were then ground and made ready for N, P, K, Ca, Mg, Fe, Mn, Zn and B analyses (Kacar and _ Inal 2010). Nitrogen content was determined through Kjeldahl wet digestion method, dry ashing method was carried out for P, K, Ca, Mg, Fe, Mn, Zn and B (Ryan, Estafan, and Rashid 2001). Readings were performed in an ICP-AES (Inductively Coupled Plasma Atomic Emission Spectrophotometer) device. NIST-brand reference apple leaf (1515) was used to check the accuracy of fruit analyses.

Statistical analyses
JMP software was used for statistical analyses. Data normality was checked through this software and extreme values were omitted. The remaining values were ordered in ascending manner and the values corresponding to 25% and 75% quartiles were considered as lower and upper reference values for each nutrient in each period. Then with these values, reference curves were created for entire sampling period of each nutrient. In other words, the regression curve passing through 25% quartile values were considered as deficiency limit and the curve passing through 75% quartile values was considered as excess limit (Aichner and Stimpfl 2002). Reliability of reference values was supported with correlation analyses. The correlations between nutrient contents of the last period and the earlier periods were investigated and resultant correlations were interpreted as possible traceability of fruit nutrients in early periods and possible set of reference values for each period. It was thought that cultivars may play a significant role in use of resultant reference values. Thus, differences were investigated among the cultivars in 68 orchards (28 orchards in E girdir and 40 orchards in Gelendost towns) with different soil characteristics. To prevent possible negative impacts while identifying the differences in rootstock and plant age of cultivars, the orchards grafted over seedling rootstocks and at full-yield were preferred. Sampled cultivars were composed of Starking Delicious (27%), Golden Delicious (34%), Granny Smith (13%) and Starkrimson Delicious (26%). All cultivars exist in both locations. Fruit samples collected 140 days after full bloom were used to determine the differences among cultivars. Significant means were compared with the use of LSD (Least Square Difference) test at P < 0.05 and P < 0.01 significance levels.

Results and discussion
Soil characteristics of experimental sites Soil samples were taken once from the selected orchards to determine soil characteristics of experimental sites. In general, about 50% of apple orchards in Isparta province have clay and clay-loam texture, all have unsaline and slightly alkaline soils with lime contents less than 25%. While surface soils had an organic matter content of below 60%, the value reached to 90% in deeper layers. Considering the micro element contents, Fe levels were below the sufficient levels at about 60% of soils, Cu levels were high and Mn levels were insufficient in all soils. Zinc and boron levels were sufficient at about 70% of surface soils and insufficient at 80% of sub-surface soils (Uçgun and Gezgin 2012).

References values of fruit samples
The 25 and 75% quartiles values of nutrients are provided in Table 1 and periodical changes of them are presented in Figures 1-5. The differences among the cultivars are provided in Table 2. The 25 and 75% quartiles varied respectively between 0.27-1.09% and 0.51-1.55% for N; between 0.06-0.19 and 0.08-0.22% for P; between 0.75-1.53% and 0.94-1.85% for K; 30-130 mg 100 g À1 and 50-190 mg 100 g À1 for Ca; between 37-110 mg 100 g À1 and 50-130 mg 100 g À1 for Mg; 8.96-19.20 ppm and 13.19-25.58 ppm for Fe; 3.24-7.52 ppm and 4.71-10.33 ppm for Cu; between 2.37-8.08 ppm and 3.88-13.24 ppm for Mn; between 1.66-8.67 ppm and 2.92-12.76 ppm for Zn; between 16.00-25.75 ppm and 22.92-35.17 ppm for B. While the greatest values for all nutrients were obtained from the first sampling period, the lowest values were obtained from the last period for N, P, K, Ca, Mg, Mn and Zn and during the last 3 periods for Fe, Cu and B. Such a case revealed that reference values decreased with the progress of sampling date for N, P, K, Ca, Mn and Zn and the values did not change from the 4th period and so on for Fe, Cu and B. In other words, besides sampling date, the date of full-bloom should also be known for reliable use of reference values.
In a study carried out by Uçgun et al. (2021a), regression curves were created for each nutrient to see the periodical changes in nutrient content of apple fruits. A periodical decrease was observed in fruit nutrient contents and such a decrease was faster throughout the earlier periods of vegetation. Correlations between fruit nutrient contents of the last period and the earlier periods were also investigated in the same study. Significant correlations were interpreted as possible traceability of fruit nutrients in early periods and possible set of reference values for each period. Except for N, Fe and Zn, positive results were obtained for the others. Uçgun and Gezgin (2017) searched whether leaf analysis could be used to determine the nutritional status of apples in the early growth period or not. Leaf samples were collected at six periods and correlations between the sixth period and the previous periods were examined. It was decided that leaf analysis could be carried out at any time from the beginning of vegetation for all elements except Fe and Cu. Apart from leaves, dormant shoots and flowers could be used to determine nutritional status of apple trees. Uçgun, Altındal, and Cansu (2018) obtained reference values for dormant shoot tissues. To check the reliability of shoot analyses, leaf samples were taken from the same orchards at 7 different periods beginning 14 days after full bloom and lasting until mid-vegetation season and correlations between nutrient contents of shoots and nutrient contents of leaves taken in all periods were investigated. Significant correlations were observed between N, P, K, Mg, Mn and B contents of shoots and leaves. Similarly, Uçgun et al. (2021b) performed a study to determine reference values for flower of apple trees. It was set forth that flower analysis could be used to determine the nutritional status of apples for P, Mg, Mn and B.

Cultivar-induced differences
As for cultivar-induced differences, the differences in N, Fe, Cu and Zn contents of cultivars were not found to be significant. The differences in other fruit nutrients were found to be significant  (5%) and the greatest values were obtained from Granny Smith cultivar. These differences between the cultivars should be taken into consideration while assessing fruit analyses. Granny Smith usually has the greatest number of days from the full-bloom to harvest (about 170-180 days). The other cultivars reach to harvest maturation at the same time (around 140-145 days). In other words, Granny Smith is different from the other cultivars in terms of harvest time. Fruit nutrient contents decreased with increasing number of days from the full-bloom. Such a case may then yield higher values for late-harvested cultivars. Besides, Granny Smith usually has lower yield levels than the other cultivars. The nutrients taken up by the roots are shared between the leaves and fruits and the nutrients going to fruits are shared among themselves. Low yield level of Granny Smith may therefore resulted in higher fruit nutrient contents than the others. Uçgun et al. (2016) carried out a study on leaves of the same cultivars and observed   similar results with the fruits in some nutrients and different results from the fruits in others. Significant differences were not observed in leaf Ca, Mn and Zn contents of the cultivars. However, significant differences were observed in fruit Ca and Mn contents of the cultivars and the values were higher in Granny Smith cultivar. P contents were similar both in leaves and fruits and the greatest values were again observed in Granny Smith cultivar. While the greatest fruit Mg and B values were obtained from Granny Smith, the case was totally opposite for leaves. Granny Smith was among the cultivars with low leaf N and K contents. As to conclude, high values observed in some nutrients in leaves did not support the nutrient contents in fruits. Uçgun et al. (2021b) conducted another study on flowers of the same cultivars and determined there were cultivar-induced differences for Ca, Mg, Fe, Mn, Zn and B in flower. The highest values were obtained in Granny Smith and Golden Delicious for Ca and Zn, in Granny Smith for Fe and Mn, in Spur Delicious for Mg, and in Spur Delicious and Starking Delicious for B.

Conclusions
In this study, reference values to be used in fruit analyses were determined for N, P, K, Ca, Mg, Fe, Cu, Mn, Zn and B contents of apples for the periods from June drops till harvest. For the relevant time frame, traceability of fruit nutrient contents, in other words the correlations between the nutrient contents at harvest and the earlier periods, indicated that fruit nutrient contents could be estimated 100 days ahead. Except for N, Fe and Zn, such a case was possible for the other nutrients. Possible differences among the cultivars are also significant issues for the use of fruit analyses to assess the nutritional status of apple orchards. Significant differences were  observed in some nutrients of the cultivars. Therefore, differences among the cultivars should be taken into consideration while using these reference values to assess the nutritional status of the orchards.