Assessing morphological congruence in Dinobryon species and their stomatocysts, including a newly established Dinobryon pediforme–stomatocyst connection

The chrysophyte genus Dinobryon Ehrenberg consists of 44 taxa, which occur in freshwaters, rarely marine waters, mostly in temperate regions of the world. The taxa of Dinobryon produce characteristic solitary or dendroid colonies and resting stages called stomatocysts. Only 20 Dinobryon taxa have information on produced stomatocysts and only four stomatocysts are reliably linked with vegetative stages using modern identification standards employing scanning electron microscopy (SEM) analyses. In this study, an encysted material of Dinobryon pediforme (Lemmermann) Steinecke was collected in two lakes in contrasting regions of Poland. Light microscopy (LM) and scanning electron microscopy (SEM) analyses revealed that Dinobryon pediforme produces stomatocyst #61, Piątek J. that is described here as new morphotype following the International Statospore Working Group (ISWG) guidelines. This raises to five the number of reliable links between vegetative stages of Dinobryon species and corresponding stomatocysts. Phenotypic similarities between Dinobryon species and their stomatocysts, analysed for five reliably established links, showed no relationships in size and shape between loricas and stomatocysts belonging to the same species. The morphological characters of loricas and stomatocysts mapped onto the phylogenetic tree of the five Dinobryon species revealed only little congruence between their morphology and phylogenetic relationships.


Materials and methods
The encysted material of Dinobryon pediforme was collected by the second author of this study in two contrasting regions of Poland: from Suchar II lake in the Wigry National Park (north-eastern Poland) on 2 nd July 2008 and from Wielki Staw lake in the Karkonosze National Park (south-western Poland) on 29 th July 2008. Samples were taken from both lakes using a no. 25 plankton net, either from shore near a hiking trail in Suchar II lake or from shore near the outflow of Biały Potok brook in Wielki Staw lake.
Water temperature, pH and conductivity of Suchar II lake and Wielki Staw lake were measured immediately after material collection with an Elmetron CPC-401 pH/conductivity meter. Other water parameters, including total and carbonate hardness, concentration of oxygen and chloride ions were analysed by titrimetric methods using an Aquamerck Compact Laboratory (Merck) and a Chloride test (Merck).
The examination and identification of Dinobryon and stomatocyst specimens were made using standard light and phase contrast microscopy (LM) and scanning electron microscopy (SEM). LM studies were conducted from slide preparations mounted in water, using a Nikon Eclipse 80i light microscope. Micrographs were taken with a Nikon DS-Fi1 camera. For SEM analyses, two methods of material preparation were used. First, fresh material was pipetted onto clean cover glasses, air-dried and affixed to aluminium stubs with a double-sided transparent tape. The stubs were then sputter-coated with carbon using a Cressington sputter-coater and viewed with a Hitachi S-4700 scanning electron microscope. However, in such prepared materials the membrane chamber stuck hard to the stomatocysts, so they were covered, making SEM study difficult. Therefore, for detailed identification of stomatocysts (collar, pore morphology and ornamentation) it was necessary to get rid of the membrane chamber to observe naked specimens. For this purpose material was prepared in the laboratory, where samples were placed in a glasses scintillation vials, covered with 10% HCl and allowed to stand for 24 h. They were then boiled for 15 min, rinsed several times with distilled water, pour in glass vials and covered with 95% alcohol. Finally, such prepared materials were also sputter-coated with carbon and viewed with the microscope mentioned above.
The morphological features of vegetative colonies of Dinobryon pediforme found in Suchar II lake and Wielki Staw lake follows the terminology given by Kristiansen  Lorica length, lorica width (at the mouth), and stomatocyst diameter in Dinobryon pediforme from Suchar II lake and Wielki Staw lake were compared using appropriate statistical tests after testing the assumptions of normality (Shapiro-Wilk test) and equality of variances (F test) in Past, ver. 4.01. Lorica length (N = 40) was subjected to the t test, lorica width (N = 40) to the Welch test, and stomatocyst diameter (N = 20) to the Mann-Whitney test 25 . P values below the threshold 0.05 were treated as statistically significant.
Dinobryon pediforme lorica and stomatocyst newly analysed here, as well as four other Dinobryon taxa for which stomatocysts can be reliably assigned (D. cylindricum, D. divergens, D. sertularia, D. sociale var. americanum, for the sizes and morphological characteristics see Tables 1 and 2), were compared by hierarchical clustering in Past, ver. 3.18, and presented as classification dendrograms (Fig. 3a,b). Clustering was based on the Jaccard's coefficient (binary similarity index) and the UPGMA method 25 . The occurrence of quantitative morphological characteristics in the classification analysis was marked as "yes", their absence was marked as "no" (Table 2). Characteristics marked with an asterisk (*) were excluded from the cluster analysis to avoid occurrence common for all five taxa and duplicate values. Both dendrograms were next elaborated using CorelDRAW, ver. 9 (Corel Corporation). In order to analyse the level of congruence between morphology and the phylogenetic relationships of the five Dinobryon taxa under study, a schematic phylogenetic tree (Fig. 3c) was generated in MorphoJ, ver. 1.06d 26 , and next elaborated using CorelDRAW X4, ver. 9 (Corel Corporation). The tree was based on the evolutionary tree of chrysophytes constructed by maximum likelihood of 18 S rDNA published by Bock et al. 27 where phylogenetic relationships of all the five Dinobryon taxa under the present study can be found.

Results
field sampling, and chrysophyte and stomatocyst morphologies. The encysted specimens of Dinobryon pediforme originated from natural populations in two freshwater lakes: Suchar II lake and Wielki Staw lake. Suchar II lake (54°05′N, 23°01′E) is a dystrophic lowland lake situated in the Wigry National Park in north-eastern Poland. The lake is surrounded with Picea abies forest. Its area is 2.6 ha, maximum depth is 10 m 28 . Wielki Staw lake (50°45′N, 15°42′E) is a subalpine lake situated in the Karkonosze Mountains, in the Karkonosze National Park in south-western Poland. It lies at 1225 m a.s.l. and is the largest water body in these mountains, with the area of 8.3 ha and maximum depth 24.4 m 29 . The occurrence of Isoëtes lacustris on the bottom of Wielki Staw lake indicates its low trophy and allows to classify it to lobelia lakes 30 . During the sampling the following physicochemical characters were measured: temperature 21.4 °C and 18.3 °C, pH 5.5 and 5.8, conductivity 17.0 μS cm −1 and 11.0 μS cm −1 , oxygen concentration 7.1 mg l −1 and 7.2 mg l −1 , total and carbonate hardness below 1.0 mmol l −1 , chloride concentration ca. 6 mg l −1 Cl-in Suchar II lake and Wielki Staw lake, respectively.
The abundance of the loricas and stomatocysts differed between both populations. In the sample from Suchar II lake, loricas of Dinobryon pediforme occurred massively (Fig. 1a-d), while its stomatocysts occurred rarely, with www.nature.com/scientificreports www.nature.com/scientificreports/ only single specimens observed in LM and SEM preparations ( Fig. 2a-b). The reverse situation was observed in the sample from Wielki Staw lake: loricas of Dinobryon pediforme were rare ( Fig. 1e-f), but stomatocysts occurred in mass ( Fig. 2c-f). In both lakes, mostly empty loricas of Dinobryon pediforme without a vegetative cell inside were observed. The stomatocysts occurred separately, but usually within the encystment chambers; however, the stomatocysts included within chambers in the stage of leaving the chrysophyte lorica were also observed ( Fig. 2a), confirming their link to Dinobryon pediforme. The stomatocyst of Dinobryon pediforme was not described before by modern standard methods and is therefore assigned to new morphotype.
SEM morphology: Loricas in the anterior part cylindrical, straight or more or less wavy; mouth more or less extended; the posterior asymmetrical with 1-2 lateral processes, short with truncate cone (Fig. 1a) Table 2. Qualitative characteristics of the five Dinobryon taxa and corresponding stomatocysts analysed in this study. Occurrence of the characteristics was marked as "yes", their absence as "no". Characteristics marked with an asterisk (*) were excluded from the cluster analysis to avoid occurrence common for all five taxa and duplicate values. www.nature.com/scientificreports www.nature.com/scientificreports/  Table 2; (b) Similarities in the shape of stomatocyst between the five Dinobryon taxa having stomatocysts assigned to their vegetative stages on the basis of their qualitative characteristics computed using the Jaccard's coefficient and the UPGMA clustering. Codes of the characteristics are presented in Table 2; (c) Mapping lorica and stomatocyst morphologies of the five Dinobryon taxa having stomatocysts assigned to their vegetative stages on their phylogenetic relationships. The tree is based on the phylogeny of Bock et al. 27 . Sizes of the depicted loricas and stomatocysts are means of the sizes included in Table 1. www.nature.com/scientificreports www.nature.com/scientificreports/ (29.3-)32.0-46.7 μm long and 5.0-8.3, 6.7-9.2, 8.0-11.5 μm wide (posterior, middle and anterior parts, respectively).
Comments: The sizes of loricas within and between the populations of Dinobryon pediforme from Suchar II lake and Wielki Staw lake were roughly similar (length 32.0-49.0 μm and width 5.0-10.0 μm, length 27.1-40.4 μm and width 5.5-11.5 μm, respectively). Statistical differences were observed in lorica length (t = 10.91, p = 2.34 10 −17 ), but not in lorical width (t = 0.06, p = 0.96), between the two populations. Loricas were smooth and striated even within the same colony in both populations; sometimes, from one parental lorica two morphologically different daughter loricas were emerged: one smooth and other clearly striated (Fig. 1d).
SEM morphology: Stomatocysts smooth and spherical, 9.0-11.2 μm in diameter, sometimes located within an encystment chambers, which stuck to the cyst wall during the drying SEM preparations (Fig. 2b,e). Collar conical, sometimes with flat planar annulus. Collar in the basal part 1.2-2.9 μm and in the apical part 1.7-2.2 μm in diameter, and 0.2-0.5 μm high. Pore 0.4-0.9 μm in diameter, surrounded by a swollen pseudoannulus or rarely by a planar pseudoannulus.
Comments: The sizes of stomatocysts, and their collars and pores in two studied populations were roughly similar. The cyst diameter was 9.5-11.4 μm in Suchar II lake and 9.0-10.5 μm in Wielki Staw lake, but differed statistically (U = 32, p = 5.54 10 −6 ). The collar of stomatocysts had the same height in both populations, but in the population from Wielki Staw lake the collar had an irregular edge. Another difference was observed in the size of basal diameter of the collar, which was smaller in the population from Wielki Staw lake (1.2-2.0 μm) than in the population from Suchar II lake (2.1-2.9 μm). Finally, some differences were also observed in the size of the pore diameter, which was smaller in specimens from Wielki Staw lake (0.4-0.7 μm) than from Suchar II lake (0.7-0.9 μm).
The shape of vegetative stages of Dinobryon pediforme was most similar to the stages of D. divergens. Both taxa produce dendroid or scattered colonies composed of cylindrical loricas with short, sharpened and oblique ends. The other taxon, D. cylindricum, producing scattered colonies of cylindrical loricas with a short, oblique, but not sharpened ends, is similar to these two taxa. The lorica posterior part in D. pediforme can posses additional 1-2 lateral appendages, whereas D. divergens has a wavy middle part of lorica and D. cylindricum has a distended one. In contrast, the colonies of D. sertularia and D. sociale var. americanum are dendroid and composed of conical loricas with a straight posterior part (Fig. 3a,c, Table 2).
No relationships in size between loricas and stomatocysts belonging to the same taxa were observed. Four of five taxa under study with different sizes of lorica produce stomatocysts which are similar in size. Dinobryon sertularia, whose loricas are rather short, produces clearly bigger stomatocysts. In three taxa (D. divergens, D. pediforme and D. sociale var. americanum), their stomatocyst diameter is similar to the lorica width, except for that of D. sertularia which produces cysts much bigger than the lorica width ( Table 1).
The collar diameter is bigger that its height in both Dinobryon pediforme and D. sertularia. Both species produce stomatocysts with a short collar, compared to the other three taxa. The collar diameter and the collar height overlap in the remaining three taxa ( Table 1).
The shape of stomatocysts is not connected with the shape of vegetative stages which produce them, as taxa most similar to each other with respect to shape of loricas are not similar to each other with respect to the shape of stomatocysts. Dinobryon sertularia forms a common clade either with D. sociale var. americanum in the tree showing the shape of loricas or with Dinobryon divergens in the tree showing the shape of stomatocysts (Fig. 3a-b, Table 2).
The stomatocyts of Dinobryon divergens and D. sertularia, mostly similar with regard to their shape, are spherical or slightly oval (in D. divergens), smooth and have a cylindrical collar. The stomatocysts of D. pediforme and D. cylindricum are also very similar to each other. They are spherical and smooth, but have conical collars, Scientific RepoRtS | (2020) 10:9779 | https://doi.org/10.1038/s41598-020-65997-9 www.nature.com/scientificreports www.nature.com/scientificreports/ sometimes with an irregular edge. Dinobryon cylindricum produces stomatocysts, which are sphaerical or obovate, smooth and have conical collars with a hooked apex. In contrast, stomatocysts of D. sociale var. americanum, forming a separate lineage in the tree showing the cyst morphologies, are covered with scabrae and verrucae, mainly in the posterior hemisphere, and have obconical collars (Fig. 3b, Table 2).
Mapping morphological data on loricas and stomatocysts of the five analysed Dinobryon taxa on the phylogenetic tree shows that their sizes are not connected with phylogeny, whereas their shapes only little reflect natural evolutionary relationships of the taxa. Dinobryon sociale var. americanum, D. sertularia and D. cylindricum, being most closely related, produce different types of colonies, loricas and stomatocysts (including their collar and ornamentation). Dinobryon divergens, which forms a separate phylogenetic lineage, possesses wavy middle parts of loricas, which is a clear difference between this taxon and the remaining four taxa (Fig. 3c, Tables 1 and 2).

Discussion
Dinobryon pediforme is a freshwater species, occurring in subarctic regions and humic acid waters 34 . It is recorded from Europe and North America, and it was recently used as a bioindicator species tied with strongly acid, often humic conditions 35 . In this study, the encystment of Dinobryon pediforme was observed in two natural populations in Poland. It is worth to note that the encystment process of chrysophytes in natural populations is rarely observed 5 and usually only a small fraction of living chrysophyte cells undergoes the stomatocyst development 13,14 . Previously, the stomatocyst production by Dinobryon pediforme was documented only few times by line drawings from observations made in light microscope. The illustrative documentation was, however, weak and contradictory. Krieger 23 (repeated by Starmach 36,37 ), Nygaard 38 and Juriš 39 depicted globoid stomatocysts with a pore lacking collar. Stomatocysts were located in special encystment chambers and were in stage of leaving the lorica. The pore of stomatocysts was oriented towards the upper part of encystment chamber. By contrast, Skuja 40 and Ermolaev and Safonova 41 depicted globoid stomatocysts with a well developed collar (and therefore invisible pore). Stomatocysts were located in special encystment chambers and were in stage of leaving the loricas, but the collar (with invisible pore) of stomatocysts was oriented towards the lower part of the chamber. The morphology of stomatocysts of Dinobryon pediforme observed in Poland were more similar to those depicted by Krieger 23 , Nygaard 38 and Juriš 39 than to those illustrated by Skuja 40 and Ermolaev and Safonova 41 . The collar and pore were not visible in LM, but the low collar and pore were visible in SEM. The stomatocyst produced by Dinobryon pediforme cannot be assigned to any of the stomatocysts described by modern SEM-based methods and is therefore described in this study as new morphotype.
With this study, the number of reliably established Dinobryon-stomatocyst links raised to five and include americanum) are common cosmopolitan species, recorded frequently from different localities worldwide. Dinobryon pediforme has been recorded so far only from Europe and North America. The known geographical distribution of these Dinobryon species generally corresponds to the known distribution of their stomatocysts. Species of Dinobryon were recorded more frequently and from more continents because they were longer studied and by more researchers compared to stomatocysts. A good example might be the case of D. sertularia, which was recorded many times from all continents except Antarctica, while its stomatocyst -stomatocyst 48, Piątek J. 2012 21 was recorded only from Africa.
A common view is that stomatocysts are species-specific 10,16 . The same holds true for the analysed Dinobryon taxa -all of them have specific stomatocysts. An interesting case is Dinobryon cylindricum. Its cyst is assigned to stomatocyst 41, Duff & Smol 1989 7,42 . Two other cysts, stomatocysts 78, Duff & Smol 1991 31 and 233, Zeeb et al. 44 are similar morphologically and could be produced by closely related Dinobryon species 7 . It is also not excluded that they all are produced by Dinobryon cylindricum but represent three developmental phases: immature stomatocyst, slightly immature stomatocyst and mature stomatocyst 14,45 .
No relationships in size between loricas and stomatocysts belonging to the same taxa analysed in the present study may result from different amounts of elements which are used by individual taxa to build stomatocysts. So far, we do not have much data on environmental conditions causing the encystment process 5 , as many chrysophyte taxa are usually not willing to grow in cultures and finding the encysting specimens in natural and laboratory conditions is problematic 22 . The fact that the diameter of stomatocysts in analysed Dinobryon species was similar or bigger (in D. sertularia) than the lorica width resulted from the encystment process which in all Dinobryon species occurs within the special chamber at the mouth of the lorica 1 . This character may be, however, genetically fixed.
Similarly, the shape of stomatocysts of the five Dinobryon taxa analysed in the present study is not connected with the shape of vegetative stages which produce them. This means that the morphological evolution of loricas goes independently from that of stomatocysts. However, some genes can be linked and, consequently, responsible for common occurrence of some morphological traits, which can be observed in both plant and animals 46 .
The genomic and phylogenetic studies on the genus Dinobryon are still at the initial level 20,22,27,[47][48][49][50] . The comprehensive studies of Bock et al. 27 , whose 18S rDNA phylogeny was a reference for the tree reconstructed in the present study, analysed relationships between all five taxa being under the present study and additional two taxa, D. sociale and D. cf. sociale. Only little congruence between morphology of the five Dinobryon taxa and their phylogeny revealed in the present study suggests that morphological characters can be the effect of homoplasy or DNA regions used for reconstructing phylogenetic relationships so far are not sufficient. Such lack of congruence