Variability of morphometric traits of seeds of different genotypes of Lycium spp

For the successful cultivation of neglected and underutilized species and the production of new valuable varieties, a comprehensive study of their plant morphology and anatomy is necessary [1–6]. It is especially important to study the biological characteristics of seeds, since they characterize the most successful varieties, especially those propagating by seed. The morphological characteristics of the seeds of different plant species serve as taxonomical markings and also in deducing phylogenetic reUDC 582.951.4:581.48 https://doi.org/10.21498/2518-1017.17.1.2021.228198


Introduction
For the successful cultivation of neglected and underutilized species and the production of new valuable varieties, a comprehensive study of their plant morphology and anatomy is necessary [1][2][3][4][5][6]. It is especially important to study the biological characteristics of seeds, since they characterize the most successful varieties, especially those propagating by seed.
The morphological characteristics of the seeds of different plant species serve as taxonomical markings and also in deducing phylogenetic re- The unique biochemical characteristics of Lycium are well documented. However, information about the morphological variability of Lycium seeds is insufficient. It is important to study the genetic variability of seeds for improving selected characteristics in the future.
The purpose of this study was to determine the variability of morphological characteristics of Lycium spp. seeds. The obtained results will help to select promising genotypes for further breeding work.

Collection of plant material
Plants growing in M. M. Hryshko National Botanical Garden of NAS of Ukraine (Kyiv) from seeds or cuttings obtained from China, France, Slovak Republic and other Botanical Gardens of Ukraine. The research was conducted during 2016-2019. The following genotypes of the three Lycium species were studied in this work: L. barbarum (var. LB01, LB02 and LB03); L. chinense (var. LC01, LC02, LC03, LC04, LC05, and cv. Amber Sweet, Big Lifeberry, Delikat, Q1, Sweet Lifeberry, Tybet); Lycium truncatum (var. LT01 and cv. Super Sweet, Korean Big, N1 Lifeberry, New Big, Princess Tao). The ripened fruits were harvested in the maturity stage (August).

Morphometric analysis
Immediately after the harvest, 30 fruits of each genotype were taken and 30 seeds were randomly selected. The following morphometric parameters were measured: seeds weight (50 seeds), in g; seeds length, in mm; seeds width, in mm. Seeds weight was measured by using a digital balance with a sensitivity of 0.01 g (PS6000/C/1). Linear dimensions of seeds as length and width were measured by using a digital calliper gauge with a sensitivity of 0.01 mm than shape index was calculated by using length/width ratio.

Statistical analysis
Basic statistical analyses -the minimal and maximal values of the traits, arithmetic means, and coefficient of variation (V %) were performed using PAST 2.17 (Norway, 2001). Results of the morphometric analysis were determined by mean ± standard deviation (SD) and statistical significance was estimated. Hierarchical cluster analyses of similarity between phenotypes were computed by the Bray-Curtis similarity index and were performed using PAST 2.17.

Results and discussions
For the first time since 2016 in Ukraine in the M. M. Hryshko National Botanical Garden (Department of Acclimatization of Fruit Plants) work on the collection of different species of Lycium L. has begun. Until this time, Lycium spp. was not studied in Ukraine at all. The collection consists of 45 genotypes (from seeds or cuttings) received from China, France, Slovakia and other botanical gardens in Ukraine, 9 of which were selected for cultivars.
There is limited information on morphometric parameters of Lycium seeds. Descriptions of species in the flora of countries indicate rough parameters of seed length and width indicators, which are usually 2-3 and 1.5-2.0 mm, or indicate only length. This does not reveal the morphometric parameters fully.
Cultivars and varieties of different species of Lycium plants varied in weight, shape, and size of seeds ( Fig. 1).
The seeds color of the currently studied Lycium ranges from pale yellow, grayish yellow, and light brown.
The minimum and maximum values for the seeds weight, seeds length, seed width and shape index of seeds in the twenty-one cultivars and varieties are shown in Table 1.
Kazbekovna et al. [27] established a seed width range in L. barbarum from 2.50 to 3.0 mm and in L. ruthenicum from 1.5 to 2.0 mm. According to Zhang et al. [28], the seeds width of L. ruthenicum was also determined from 1.5 to 2.0 mm. The seeds width of L. chinense was determined to be between 2.5 and 3.0 mm and L. shawii between 1.5 and 2.0 mm [29].
The most variable characteristics in the studied genotypes were seed weight and length,     Variety studying and variety science which are important parameters for selection. They determine the pulp content and number of seeds in the fruit and the ratio of these parameters to each other. The smaller the seed in weight and length, the greater the pulp content of the fruit. It is through variability that promising varieties with small weight and lengths can be selected, securing them afterwards, as in other cultivars, vegetatively.
Determination of the complex of relationships of morphological characteristics of seeds of cultivars and varieties of Lycium spp. showed a strong correlation between the main morphological features (Fig. 6).

Ñîðòîâèâ÷åííÿ òà ñîðòîçíàâñòâî
Thus, the correlation coefficient between seed length and width was 0.964 and is reliable at all accuracy levels. The strength of the relationship between the length and seed weight was characterized by the coefficient of 0.894, and between the width and seed weight by 0.934.
Cluster analysis is very widely used to assess the genetic diversity of many plant species [1,6,30,31].
The above data (Figure 2-5) confirms cluster analysis. In clustering, all studying parameters of seeds for 21 cultivars and varieties of the Lycium spp. were used, and the resulting clusters are shown in Figure 7.
Based on the data presented in Figure 7, we can say that cluster analysis divides the collection into two main clusters. The largest number of samples (15 cultivars and varieties) were included in Cluster I. Cluster II consisted of 5 cultivars and 1 varieties of L. truncatum, which were the most distant from all other samples of Cluster I, and differed from the others by the smallest morphometric characteristics of seeds. Cluster analysis demonstrates the integrated character of the variability in seeds of the studied species, cultivars and indicates possible ways to artificially improve the genetic material.
Varieties of Lycium with big seeds weight and size not be seen as a practical unusable. The seeds of different Lycium species contain rich biologically active substances, mainly fatty acids. Oil from these seeds can be used in the pharmaceutical and food industries [32][33][34].

Conclusions
Diagnostic signs by seed morphometry for differentiation of Lycium species were revealed. The analysis of coefficient of variation showed the difference of variability in morphometric characteristics between some Lycium cultivars and varieties. The use of cluster analysis allowed us to establish a clear limitation of L. truncatum on a complex of diagnostic characters. Seeds of L. truncatum differed from other plant species by the lowest morphometric indices. The most variable characteristics in the studied genotypes were seed weight and length, which are important parameters for selection. They determine the pulp content and number of seeds in the fruit and the ratio of these parameters to each other. The smaller the seed in weight and length, the more pulp the fruit contains. It is through variability that promising varieties with small weight and lengths can be selected, securing them afterwards, as in other cultivars, vegetatively.

Acknowledgments
The publication was prepared with the active participation of researchers in International Network AgroBioNet within the project ITMS 25110320104 «Innovation of Test Methods and Procedures for the Detection of Sources of Bioactive Substances for the Improvement of Health and Quality of Life». The authors are grateful to Visegrad Fund (52011113).