Seed morphometric characteristics of European species of Elatine (Elatinaceae)

Elatine L. contains ca. 25 small, herbaceous, annual species distributed in ephemeral waters in both hemispheres. All species are amphibious and characterized by a high degree of morphological variability. The importance of seed morphology in Elatine taxonomy has been emphasized by many authors. The degree of seed curvature and seed coat reticulation have been traditionally considered very important in recognizing individual species of this genus. Seed morphometric characteristics of 10 Elatine species, including all European native taxa, are provided on the basis of material from two or three populations of each species. A total of 24–50 seeds were studied from each population, altogether 1,260 images were used for the morphometric study. In total, six parameters were measured from SEM pictures: object surface area, profile specific perimeter (object circuit), rectangle of the object (a) length, rectangle of the object (b) width, angle of the seed curvature, and number of pits in the seed coat counted in the middle row. Our study shows that the range of morphological variation of seeds in European species of Elatine is great, both between the species and the populations. Discrimination analysis showed that all six traits significantly differentiate the populations studied (λ = 0.001, p < 0.001), and the greatest contributions were “number of pits”, “rectangle_a”, and “the angle curvature”. Multidimensional scaling based on a correlation matrix of Mahalanobis distance of the six features studied revealed the greatest similarity between the three populations of E. alsinastrum, E. macropoda, and E. hexandra. Regarding interspecific differences, a Kruskal–Wallis tests showed that, in many cases, lack of statistically significant differences between species relative to the studied seed traits. If distinction of species is only based on seeds, especially if only a few seeds are evaluated, the following species pairs can be easily confused: E. alsinastrum and E. orthosperma, E. hexandra and E. macropoda, E. campylosperma and E. hydropiper, as well and E. gussonei and E. hungarica. We found no diversity in seed coat micromorphology within pits that could have potential taxonomic importance. An identification key and descriptions of species are provided on the basis of seeds traits.


INTRODUCTION
have been considered very importan tfor recognizing individual species (Cook, 1968;Uotila, 1974;Uotila, 2010;Tucker, 1986;Misfud, 2006;. There have been only a few studies addressing morphological variability of Elatine taxa (Mason, 1956;Molnár, Popiela & Lukács, 2013). Recently,  examined the level of phenotypic plasticity in Elatine. Analysis of morphological differences between aquatic and terrestrial forms of individual species clearly showed that vegetative traits are highly influenced by environmental factors and only seed traits are stable within species. According to Molnár et al. (2015), only seed morphology (aside from generative characteristics) is valuable for taxonomic purposes.
Consequently, we studied seed morphometric characteristics of 10 Elatine species, including all native European taxa, as a part of comprehensive surveys on taxonomy and phytogeography of this genus that have been conducted by a Hungarian-Polish research team since 2010. We assumed that advanced and methodically uniform seed characteristics are taxonomically important in this genus. Our aims were to (i) find statistical differences between Elatine species relative to seed morphological features, (ii) evaluate intra-and interspeciesseed variability, and then (iii) construct a guide to identifying species based onseed morphological features. Due to the small size of seeds the study was made by using SEM micrographs.

Plant material and cultivation
Plants studied were collected across Europe. In total, seeds were collected from all 10 Elatine species and from three populations each, with an exception of very rare E. brochonii and E. campylosperma, two populations, so altogether 28 populations were used for the study. The distance between the populations of each species ranged from approximately 10-2,000 km. For the localities of the original material, and the voucher specimens, see Table 1 and Fig. 1. The studied seeds were gathered directly from the field, or from cultivated plants grown from the original material; in some cases seeds from herbarium specimens were used. Culture was conducted at the Center for Molecular Biology at the University of Szczecin, Poland and/or a the Department of Botany at the University of Debrecen, Hungary. Plants were grown in climate-controlled culture chambers with 12 h/day light and 30,000 lux light intensity, temperatures: under light, 22 ± 2 • C, and under dark, 18 ± 2 • C.
Elatine hungarica, E. hydropiper and E. triandra are protected species in Hungary and were sampled with the permission of the Hortobágy National Park Directorate (Permission id.: 45-2/2000(Permission id.: 45-2/ , 250-2/2001. To determinate the variability and diagnostic features of seeds, 24-50 seeds obtained from several individuals from each population (Table 1) were used. A total of more than 1,500 scanning electron microscope (SEM) images of the seeds were obtained at ×200 magnification using an SEM (Zeiss Evo, Molecular Biology and Biotechnology Center, University of Szczecin, Szczecin, Poland); however, 1,260 images were used for the morphometric study, because all cracked seeds were excluded. In total, six parameters were measured ( Fig. 2): (A) object surface area; (B) profile specific perimeter (object circuit);

Data analysis
To distinguish the characteristics that have the greatest impact on population and species discrimination, multiple discriminant analysis was used. Wilks λ was used to measure the discriminatory power of the model (0-perfect discrimination; 1-no discrimination). Interpretation of discriminant functions was performed using canonical analysis. For visualization of the relationship between species and populations, Mahalanobis distance-based unweighted pair-group method using arithmetic averages to construct (UPGMA) trees was applied. Canonical values were shown using categorized scatterplots. The most discriminative traits were also independently tested by the non-parametric Kruskal-Wallis tests. All calculations were made in Statistica v. 12.5 software.

Variability of populations within species
Discrimination analysis showed that all six traits significantly differentiate the populations studied (λ = 0.001, p < 0.001). Of these, the greatest contributions were as follows: number of pits, rectangle a (length), and the angle of curvature (Table 2).
However, there were large ranges of variation for some traits, especially within the following populations: E. orthosperma from Finland, Fin1 (for acronyms see Table 1) (surface: SD 36131.9 and rectangle a: SD 78.9), E. hungarica from Slovakia (profile: SD  Table 4).
Multidimensional scaling based on a correlation matrix of Mahalanobis distance of the six features studied revealed the greatest similarity between the three populations of the following species: E. alsinastrum, E. macropoda, E. hexandra (Fig. 4).

Variability between species
Discriminant analysis showed that all variables could discriminate species (λ < 0.01). The greatest impact was from the following features: number of pits, the angle of curvature and rectangle a ( Table 5).
The Kruskal-Wallis tests showed, in many cases, lack of statistical significance between species relative to the studied seed traits (Table 6). Regarding the trait surface, only E. triandra seeds showed statistical significance compared with all species tested. Analysis of all characteristics showed the least amount of statistically significant differences between the following species pairs: E. alsinastrum and E. orthosperma, E. hexandra and E. macropoda, as well as E. gussonei and E. hydropiper (Table 6).
There was a large range of variation for the taxa studied regarding the following traits: seed size (traits: surface, profile, and rectangle a), especially within E. hungarica (SD 27183.7, SD 285.9, and SD 62.0, respectively); the angle of curvature, E. gussonei (SD 44.7); and number of pits, E. campylosperma (SD 7.2). The smallest variation was present in E. triandra (surface SD 14587.3, profile: SD 171.8, rectangle a: SD 54.6) and E. brochonii (rectangle b SD 20.6, the angle of curvature SD 14.7, pits: SD 1.3). The characteristics associated with size (surface, profile, rectangle a, rectangle b) revealed that the following species had the smallest seeds: E. brochonii and E. triandra, while the largest seeds in the studied species belonged to E. gussonei and E. hydropiper (Fig. 5, Table 7).
The classification matrix of the discriminant analysis showed that the level of classification varied from 86% (rectangle a, the angle of curvature, number of pits)  to 84% (surface, profile, rectangle a, rectangle b, the angle of curvature, number of pits). The highest values of classification were found for E. alsinastrum, E, brochonii, E. hydropiper, and E. orthosperma (all greater than 90%). The lowest values were found for E. campylosperma (57%, 55%) ( Table 8).
UPGMA clusters of Mahalanobis distance based on rectangle a, the angle of curvature, and number of pits yielded two groups: species with straight or nearly straight seeds, and species with curved and U-shaped seeds (Fig. 6). The greatest similarity was found  Table 1 between seeds of E. hexandra and E. macropoda, and E. campylosperma and E. hydropiper. The spatial distribution of observed characteristics of the analyzed species is depicted as a categorized scatterplot based on canonical analysis values (Fig. 7).

DISCUSSION
Our study shows that in Elatine tested seed variability is mainly associated with sizeconnected traits, especially surface, profile, rectangle b, and, to a lesser extent, rectangle a. This allowed us to draw the conclusion that to distinguish seeds of these species the most useful traits are the angle of curvature and number of pits, and to a lesser extent rectangle a (length). These findings confirm previous knowledge about the usefulness of these features in Elatine taxonomy (Misfud, 2006;Uotila, 2009a;Uotila, 2010;Molnár et al., 2015). Nevertheless, our study revealed that the range of variation of European Elatine morphological features is large, both between species and the populations of each species.
Regarding intraspecific variability, the traits studied were not statistically significantly different between studied populations of the following taxa: E. alsinastrum, E. macropoda, E. hexandra. Conversely, E. gussonei, E. campylosperma E. hungarica and E. hydropiper seeds showed statistically significant intrapopulation variability. The taxonomic status of the first three species is still being elucidated. Elatine gussonei, an enigmatic plant of the Mediterranean, was first described as a variety of E. hydropiper and was later classified as a separate species (Brullo et al., 1988;Misfud, 2006;Molnár, Popiela & Lukács, 2013). Elatine campylosperma was described by Seubert (1845) from Sardinia, and later greatly neglected by most researchers by synonymizing this species under   E. macropoda; at present, it is considered a separate species (Kalinka et al., 2015). Elatine hungarica was last collected in 1960 and rediscovered in Hungary in 1998 (Molnár et al., 1999); for years its taxonomic status was under discussion .
Our present study showed that regarding shape statistically only E. alsinastrum and E. orthosperma seeds are nearly straight and seeds of all other species are curved to varying degrees; the range of variation in some species is large in this respect, especially in E. gussonei, E. triandra, and E. hexandra. Figure 6 Morphological relationships of seeds among surveyed Elatine species displayed by Mahalanobis distance-based UPGMA cluster based on the following features: rectangle a, angle of curvature, and number of pits. For acronyms, see Table 1. If distinction of species is only based on seeds, it would be easy to confuse the following species pairs: E. alsinastrum and E. orthosperma, E. hexandra and E. macropoda, E. campylosperma and E. hydropiper, andE. gussonei and E. hungarica, especially if only a few seeds are evaluated. Previously, Misfud (2006), who worked on Malta and Mallorca populations, pointed out the importance of distinction based on greater seed curvature in    Misfud (2006) did not precisely describe the method of counting pits (especially in which row pits were counted), it is difficult to compare our results.  pointed out that seeds of E. hungarica are much  Table 1. more curved than those of E. gussonei, and especially of E. macropoda and E. orthosperma, but somewhat less curved than those of E. hydropiper. Our current study revealed that the range of variation for the feature the angle of curvature of E. hungarica seeds is similar to that of E. gussonei, and more curved seeds are found in E. hydropiper and E. campylosperma. These results are basically consistent with observations of , especially considering that more varied material was used in the current study. Our research confirms observations of Misfud (2006) and  concerning the evident semilunar membrane on the concave side of seeds . The membrane was present and clearly visible in all highly curved fresh seeds of the following species: E. gussonei, E. hydropiper, E. hungarica, and E. campylosperma. Regarding the seed testa, a very distinctive network-shape ornamentation pattern of E. triandra as visible 11). Elatine campylosperma seeds showed the most distinctive reticulation, and were characterized by a large number of narrow, rectangular pits (Fig. 10 (Figs. 10 and 11). Similar observations for some of these species were made by Misfud (2006), , Molnár, Popiela & Lukács (2013). We believe that the shape of pits may be an additional feature that helps distinguish seeds of individual species (Figs. 8 and 10). However, we found no diversity in seed coat micromorphology within pits (e.g., pores, strophioles) that could have potential taxonomic importance. Seed coats within pits were smooth with the exception of irregular strips, and the porosity of the seed coat is visible only in the inner layer of cracked seeds (Figs. 8Q-8R, 9I). Ornamentation pattern (pit shape) becomes distinct as the seeds dry up. The outer layer of the seed coat is very thin and easily destroyed ( Figs. 9A-9B; 9D-9F). Our research allowed us to construct a guide that can be useful to identify the studied taxa based on seed traits. We believe that this guide is important for better recognition of these rare and endangered species, and can be useful for elucidating the history of range formation of these taxa in the Holocene and their origin. Elatine subfossil finds were discovered in late-glacial and pre-boreal sediments in the last few centuries (Latałowa, 1992;Brinkkemper et al., 2008;Kowalewski et al., 2013). The ecological amplitude of this species provides robust clues for environmental reconstruction, which must have been a temporarily flooded fresh water area. ''...since this type of environment is strongly threatened on a worldwide scale, the presence of these species in the past may also provide interesting information for present nature development projects...'' (Brinkkemper et al., 2008).
Identification guide and descriptions for European species of Elatine based on seed morphology presented in Figs. 10 and 11. Note: the guide does not include exceptional values given in parentheses in the descriptions (min. outliers 1.5) 25%-75% (max. outliers 1.5).