Viability assessment and estimation of the germination potential of charophyte oospores: testing for site and species specificity

Abstract

Anthropogenic disturbance over several decades has caused aquatic ecosystems to deteriorate, raising concerns about their condition and, consequently, prompting restorative activities. The recovery of submerged vegetation (notably charophytes) is a crucial requirement for the restoration of eutrophic aquatic systems. The success of this action often depends on the availability and germinability of aquatic plant diaspores. To characterize the potential of the diaspore bank for the restoration of charophyte stands, we investigated viability, germination potential and conditions for dormancy breaking of the oospores in three independent studies. The viability of the oospores was determined by means of the triphenyltetrazolium chloride (TTC) test. The germination potential of the diaspore banks from seven locations was estimated for potential irradiance dependency and the impact of factors influencing the dormancy status was assessed. For the latter aspect, the germination experiments were conducted with solitary oospores from plants and diaspore banks, pretreated under different temperature and desiccation regimens. The results showed species-specific differences in dormancy breakage and germination of 10 charophyte species. Five of the species tested, Chara filiformis Hertzsch, Lychnothamnus barbatus (Meyen) Leonh., Tolypella glomerata (Desvaux) Leonh., Nitella mucronata (A.Braun) Miq. in Hall and Nitella flexilis (L.) C. Agardh/Nitella opaca (C. Agardh ex Bruz.) C. Agardh, exhibited light-dependent germination. The irradiances where maximum germination was achieved correspond well to the species’ distribution within the vertical irradiance gradient in their habitats.

Keywords:

• Charophytes
• desiccation
• diaspore
• dormancy
• irradiance
• TTC test

Introduction

The deterioration of aquatic ecosystems resulting from anthropogenic disturbance has raised worldwide concerns and resulted in the release of national and international regulations, targeting the protection and restoration of water resources, e.g. the Water Framework Directive (European Union 2000) and the US Clean Water Act (Anonymous 1972). The primary intention of these regulations is to combat the deterioration of water quality by reducing the anthropogenic disturbance, such as pollution, habitat destruction and increased nutrient input, in order to restore the system’s ability to provide ecosystem services (e.g. drinking water, fishing and recreational use). As shown by numerous studies, the success of this action depends to a large degree on the macrophyte meadows, which play a crucial role in the regulation of aquatic food webs (e.g. Becker 2016; Blindow and van de Weyer 2016; Scheffer 1998). The sustainability of restoration measures also depends on the successful recovery of the submerged macrophyte belt in general, especially the recovery of charophyte meadows in temperate regions (e.g. Scheffer 1989). While the main factors in the decline of charophytes are well known (Becker 2016), further investigations are needed into the conditions of the aquatic environment required for their recovery.

The inoculation of vegetative thalli, as well as the seeding of diaspores, can support the reactivation of macrophyte meadows in cases where the internal restoration potential is too poor (Donabaum, Schagerl, and Dokulil 1999; Rodrigo et al. 2015). However, especially for rare species, the geographic provenance and adaptation to local conditions must be taken into account (Husband and Campbell 2004; Schaal and Leverich 2004). Therefore, activation of the internal diaspore bank should always be the first option considered, before considering transplantation with its potential negative consequences, such as disturbance of the source populations and contamination of the gene pool (Barett and Kohn 1991; Foster Huenneke 1991). Consequently, reliable estimations of the restoration potential of the diaspore bank, as well as the conditions required for its activation, are needed.

In particular, not only the number of diaspores, but also their viability (capacity to germinate) must be established. Viability can be determined by the “crush test”, i.e. compression of the individual oospores until they break. The resistance to crushing (indicated by a “pop”) and the presence of white starch grains are used to indicate viability (Bonis, Lepart, and Grillas 1995; Grillas et al. 1993; Matheson et al. 2005; Rodrigo, Alonso-Guillén, and Soulié-Märsche 2010; Sederias and Coleman 2007). As has been known for more than a century for terrestrial seeds, while the existence of the reserve carbohydrates is a prerequisite, it is not the only requirement for viability (Duvel 1904). Despite this knowledge, the methods in use for testing the viability of higher plant seeds have not been reported for oospores. Therefore, the applicability of seed-viability test methods for oospores was tested first in this study.

The second aspect of using diaspores for restoration is stimulation of the germination of viable diaspores. They allow the species to survive periods of unfavourable conditions (temporal dispersal) and enable rapid colonization after disturbances (Thompson 2000). The composition of the diaspore bank influences the species composition and diversity after disturbances (Jonsson 1993). If they are to be successful, the diaspores must be dormant until the conditions for growth and reproduction are favourable, especially in the case of non-seasonal disturbances. So, understanding the mechanisms that activate the diaspores (dormancy breaking) is a prerequisite for the use of diaspore banks for restoration. Little is known about dormancy control in charophytes. Several studies have found an irradiance effect on the germination of charophyte oospores (e.g. Forsberg 1966; Matheson et al. 2005; van den Berg et al. 1998), although there is also evidence for light-independent processes (De Winton, Casanova, and Clayton 2004; Shen 1966). Despite these varied results, several authors (e.g. Sokol and Stross 1992; Takatori and Imahori 1971) imply the presence of phytochrome systems in the oospores and their participation in the germination process.

Therefore, in addition to determination of oospore viability, the factors that influence dormancy breakage and induction of germination from natural diaspore banks are investigated in laboratory experiments, including the species-specific effects of temperature, irradiance, desiccation and nutrient conditions.

Material and methods

Viability

Four species from different localities were investigated in relation to viability (Table 1). Ripe oospores, taken directly from field-grown plants of Chara contraria A.Braun ex Kütz. (n = 120) and Chara globularis Thuill. (n = 134), were tested. The oospores of C. contraria (n = 120) were collected from Grosser Goitzschesee (51°37′37.90″ N, 12°20′47.26″ E) and the oospores from C. globularis from Kronwaldsee near Loitz (53°59′01.12″ N, 13°01′48.3″ E). The oospores of Chara vulgaris L. (pond near Marlow: 54°8′41.73″ N, 12°34′27.92″ E, n = 120) and Chara canescens Loisel. (n = 120, Lake Neusiedl, 47°51′06.07″ N, 16°48′44.33″ E, and German Baltic coast, 54°26′37″ N, 12°41′20″ E) were isolated from the seed sediment surface below monospecific stands. Before testing, the oospores were rinsed with distilled water and pretreated with hydrogen peroxide solution (30%) to ensure that they had no bacterial contamination on the outside oospore layer. After this procedure, the oospores were separated into individual sterile microtitre moulds, opened by pressing them gently with a glass rod and coated with 2,3,5-triphenyltetrazolium chloride (TTC) solution in different concentrations of 0.5%, 1% and 2%. The procedure of testing the oospores by means of TTC followed the International Rules for Seed Testing (International Seed Testing Association 2015), with the exception of modifying the concentration of TTC solution. The plates were stored in darkness at room temperature for at least 24 h (up to 96 h). Viability was evaluated by visual analysis (colour change) and carried out using a binocular SZX-16 microscope (Olympus, Japan). The reduction of TTC (colourless) to Formazan (red) only proceeds in the presence of active hydrogenases within the oospores. Three classes were distinguished: (1) totally red-coloured oospores, which were regarded as vital; (2) colourless oospores, regarded as non-vital; and (3) an intermediate group of partly red-coloured oospores. According to Lakon (1949), the third class had to be considered vital as well, as the colour change implies vital conditions. The differences between the classes with respect to TTC concentrations, as well as species, were analysed by the χ² test using the program SPSS (Norušis 2006). We used the standard p = 0.05 to determine statistical significance throughout this study.

Table 1. Triphenyltetrazolium chloride (TTC) viability test of Chara oospores.

TTC conc.	Colour	Chara contraria	Chara globularis	Chara canescens	Chara vulgaris
0.5%	Red coloured	40.5	81.2	52.5	36.8
	Partially red coloured	27.1	9.4	25	39.5
	Colourless	32.4	9.4	22.5	23.7
1.0%	Red coloured	52.6	60.0	45.0	30.4
	Partially red coloured	18.5	15.0	37.5	23.9
	Colourless	28.9	25.0	17.5	45.7
2.0%	Red coloured	63.1	43.9	52.5	48.7
	Partially red coloured	21.7	26.8	22.5	17.9
	Colourless	15.2	29.3	25.0	33.3

The numbers are the percentage of oospores tested vital (red coloured and partially red coloured) and non-vital (colourless) at 0.5%, 1.0% and 2.0% concentrations of TTC. The oospores of C. vulgaris (n = 120) and C. canescens (n = 120) were taken from the sediment samples; the oospores of C. contraria (n = 120) and C. globularis (n = 134) were taken directly from the field-grown plants.

Dormancy breakage

Investigating how to break the primary dormancy requires the removal of fresh, but fully ripened, oospores from the parental plants. For this experiment, field-grown individuals of Chara baltica Bruz. (Mallada, Spain; 36°43′12″ N, 4°25′13″ W), Chara canescens Loisel. (Lake Neusiedl and a Baltic coast lagoon; 54°47′09.0″ N, 25°20′05.4″ E and 54°21′46.15″ N, 12°48′20.94″ E), Chara hispida L. (Gummanz; 54°33′3.44″ N, 13°34′37.49″ E), Chara papillosa (Kütz.) Raam (Röblingen; 51°28′03.48″ N, 11°42′07.53″ E) and Lychnothamnus barbatus (Meyen) Leonh. (Lake Balsys; 54°47′09.0″ N, 25°20′05.4″ E), with well-developed oospores, were cultivated in the laboratory until the oospores were released and fell to the bottom of the vessels, where they were collected.

The experiment consisted of two steps: (1) the application of pretreatments to break the primary dormancy; and (2) treatments for the induction of germination. The pretreatment by desiccation was done by storing air-dried oospores at several treatment temperatures for a minimum of 2 weeks. The temperature pretreatment exposed the oospores to different temperatures in the presence or absence of microbially active sediment. A control group of oospores was transferred to the germination induction without any pretreatment. After pretreatment, the germination success was determined at 15°C and 30–40 μmol photons s⁻¹ m⁻². Three germination substrates were used: cultivation sediment (washed sea sand and phosphorus) (Wüstenberg, Pörs, and Ehwald 2011), cultivation sediment with the addition of organic plant material to increase nutrient supply, and pure agar as a nutrient-depleted, microbially inactive substrate. This experiment was carried out on the individual oospores. At least 10 (in the case of C. baltica from Spain) oospores were used for each combination of pretreatment and germination conditions. To test the influence of the different pretreatment/germination condition combinations, the Spearman’s rank correlation coefficient was calculated (Spearman 1904). The significance of the observed correlations was tested by means of Fisher’s exact test (Fisher 1922). All tests were computed using the program R (R Development Core Team 2011).

Seed bank germination

Seed bank material from seven locations was tested. These sites involved five lakes from North and Central Germany, one Lithuanian lake and one Baltic coastal lagoon: Fastensee (Schleswig-Holstein; 54°30′26.18″ N, 11°02′08.72″ E), Stoßdorfer See (Brandenburg, 51°50′15.77″ N, 13°48′57.15″ E), Obersee near Lanke (Brandenburg, 52°45′35.85″ N, 13°33′23.49″ E), Borkener See (Hessen, 51°1′39.25″ N, 9°15′55.65″ E), Grosser Goitzschesee (Sachsen-Anhalt, 51°37′37.90″ N, 12°20′47.26″ E), Lake Balsys (Lithuania, 54°47′09.0″ N, 25°20′05.4″ E) and Grabow lagoon (Mecklenburg-Vorpommern, 54°21′46.15″ N, 12°48′20.94″ E). All of these sites are habitats for several species of Chara L., Lamprothamnium J.Gr., Lychnothamnus (Rupr.) Leonh., Nitella Ag. and Tolypella (A.Braun) A.Braun. In the case of Nitella flexilis/opaca, no further differentiation could be made because of the lack of gametangia.

To test the irradiance dependency of germination, natural oospore-containing sediments, where dormancy status could be assumed to be uniform, were used. Except for Grabow, which had already been sampled in April 2013, all the seed bank samples were obtained between April and December 2014 using a corer (5.5 cm in diameter). Seven to 10 samples of the upper 3–5 cm of sediment were taken below the dense charophyte stands. The samples were stored at 5°C in darkness for at least 4 weeks before the start of the experiments. All the samples originating from one site were blended and 100 ml of each was transferred to at least 21 glass beakers (Duran 600 ml).

The beakers were filled with filtered (55 μm) habitat water, placed in a temperature-controlled water bath (15°C) and gently aerated. The highest type of irradiance was provided (Phillips TLD 36 W/950; light cycle 16 h:8 h) and then the beakers were filtered with different amounts of screening (fly screens): (1) 110–130 μmol photons s⁻¹ m⁻²; (2) 30–40 μmol photons s⁻¹ m⁻²; and (3) 15–20 μmol photons s⁻¹ m⁻². The irradiance was measured using a photosynthetically active radiation meter (LI-COR LI-250). The total duration of the test was 5 months, during which the water was replaced every 1–2 weeks. If no germination was detected after the first 3 weeks, the sediment was stirred to bring the oospores closer to the surface. After 5 months, the germlings that had established under the different irradiance conditions were counted. The germlings were determined following the key of van de Weyer and Schmidt (2011). Molecular analysis was applied to confirm doubtful cases, as described by Nowak, Schubert, and Schaible (2016). Every germling was examined to determine its origin (i.e. from oospores, bulbils, stem-tubers or root nodes).

The species-specific differences between light intensities were analysed by the Mann–Whitney U test (Mann and Whitney 1947) using SPSS (Green and Salkind 2010). We used p = 0.05 to determine the statistical significance throughout this study.

Results

Viability

In total, 67.6–90.6% of oospores from the fresh plant material (C. contraria and C. globularis) and 54.3–82.5% of oospores from the sediments (C. canescens and C. vulgaris) were classified as “vital” according to the TTC test criteria (Table 1). In contrast to this result, nearly 13% (38 oospores) of these vital oospores, as assessed by the TTC test, were classified as non-vital according to the crush test (data not shown). For C. globularis, significantly higher proportions of red-coloured oospores were detected for the lowest and highest TTC concentrations (Table 2). For C. contraria, a significantly higher proportion of colourless oospores was detected at the lowest TTC concentration. The fraction of partially red-coloured oospores of C. vulgaris was significantly higher at the highest TTC concentration (Table 2). Irrespective of these results, no general concentration dependency of the TTC method could be detected when analysing the whole data set.

Table 2. Results of the χ² test.

Species	Concentration (%)	Colour	χ^2	p (two-tailed)
Chara contraria	0.5	Red	2.52	0.112
		Partially red	0.184	0.668
		Colourless	5.789	0.016*
	1	Red	0.23	0.88
		Partially red	0.195	0.659
		Colourless	0.237	0.626
	2	Red	5.393	0.02*
		Partially red	0	0.997
		Colourless	1.069	0.301
Chara globularis	0.5	Red	10.968	0.001*
		Partially red	1.444	0.23
		Colourless	1.702	0.192
	1	Red	0.449	0.503
		Partially red	0.054	0.816
		Colourless	0.169	0.681
	2	Red	7.511	0.006*
		Partially red	1.946	0.163
		Colourless	2.927	0.087
Chara canescens	0.5	Red	0.073	0.787
		Partially red	1.102	0.294
		Colourless	1.57	0.21
	1	Red	0.593	0.441
		Partially red	0	0.988
		Colourless	0.467	0.494
	2	Red	0.232	0.63
		Partially red	1.024	0.311
		Colourless	0.249	0.618
Chara vulgaris	0.5	Red	1.111	0.292
		Partially red	2.894	0.089
		Colourless	1.868	0.172
	1	Red	0.038	0.845
		Partially red	0.36	0.548
		Colourless	1.894	0.169
	2	Red	1.628	0.202
		Partially red	4.971	0.026*
		Colourless	0.016	0.9

Standard p = 0.05 was used for significance (*).

The sediment-extracted oospores of C. canescens and C. vulgaris exhibited slightly lower viability than the oospores gathered directly from the parent plants (C. contraria and C. globularis) (Table 1).

Dormancy breakage

For dormancy breakage, either the temperature or the desiccation pretreatment was effective, with the combination of both being the most effective (Table 3). This effect of combined treatments was most pronounced for C. canescens and C. papillosa, both of which showed the highest germination rates after desiccation at a particular temperature. However, species-specific differences were also observed. While C. papillosa exhibited the highest germination rates after desiccation at 20°C (44–60%), for C. canescens (32–48%) and L. barbatus (66%) desiccation at 5°C was the most effective. For C. papillosa, the germination rate decreased after desiccation at 5°C (12%).

Table 3. Comparison of tested oospores and the combinations influencing dormancy, as well as the germination-inducing conditions and resulting percentage of germination.

Species; site	n (Oospores)	Germination (%)	T	M	D	Sediment
			Pretreatment	Pretreatment	Pretreatment	Germination process
			A	B	C	α
Chara canescens; Neusiedler Seewinkel, Austria	40	0	5	0	1	0
Chara canescens; Baltic Sea, Germany	40	0	15	1	1	0
Chara baltica; Mallada, Spain	10	0	−5	0	0	1
Chara baltica; Mallada, Spain	10	0	5	0	0	1
Chara baltica; Mallada, Spain	10	0	15	0	0	1
Chara hispida; Gummanz, Germany	30	0	–	0	0	1
Chara hispida; Gummanz, Germany	30	0	–	1	0	1
Chara hispida; Gummanz, Germany	30	0	−5	1	0	1
Chara hispida; Gummanz, Germany	30	0	−5	1	0	1
Chara hispida; Gummanz, Germany	15	0	5	0	0	1
Chara hispida; Gummanz, Germany	15	0	5	1	0	1
Chara papillosa; Röblingen, Germany	25	0	5	0	0	1
Chara canescens; Neusiedler Seewinkel, Austria	25	0	−5	1	0	1
Chara canescens; Baltic Sea, Germany	15	0	15	0	0	1
Chara papillosa; Röblingen, Germany	25	0	20	0	1	1
Chara canescens; Baltic Sea, Germany	15	0	15	0	1	1
Chara papillosa; Röblingen, Germany	25	12	5	0	0	2
Chara canescens; Baltic Sea, Germany	15	20	15	1	0	2
Chara canescens; Neusiedler Seewinkel, Austria	25	32	5	0	0	2
Chara canescens; Baltic Sea, Germany	15	33	15	1	1	2
Chara canescens; Baltic Sea, Germany	15	40	15	1	1	2
Chara papillosa; Röblingen, Germany	25	44	20	0	1	2
Chara canescens; Neusiedler Seewinkel, Austria	25	48	5	0	1	2
Chara papillosa; Röblingen, Germany	25	60	20	0	1	2

Pretreatments: T = temperature; M = microbial; D = desiccation; 0 = no microbial/not desiccated; 1 = microbial active phase/desiccation of oospores. Germination process: 0 = agar; 1 = washed sea sand and phosphorus; 2 = washed sea sand and phosphorus with organic addition.

The addition of organic material was a prerequisite for germination of all the Chara species tested. All set-ups without the addition of organic material to the cultivation sediment failed to produce germlings, regardless of the pretreatment conditions. In contrast to this result, L. barbatus germinated without the addition of organic material in the pretreatment or germination phase. For the Chara species tested, desiccation increased the percentage of germlings, but was not effective without the addition of organic material during germination (Table 3). The Spearman rank correlation revealed no effect of temperature on germination, when tested for the whole data set. However, when tested for all Chara data with the addition of organic material during the germination phase, the relation (rho = 0.171, p = 0.0258) was significant by Fisher’s exact test (p < 0.001).

The same was true for the combination of desiccation of oospores during pretreatment and the addition of organic material during germination. Tested for the whole data set, no influence of desiccation was detected. However, when restricted to all the Chara data with organic material during the germination phase, the increase in germination rate by desiccation was shown to be significant (p = 0.0011) by Fisher’s exact test.

Seed bank germination

Natural sediments, where the dormancy status could be assumed to be uniform, were used to test the irradiance dependency of the germination. Oospores of six Chara, two Nitella, one Lamprothamnium and one Nitellopsis species, as well as Lychnothamnus barbatus, germinated from the seven sediments tested (Table 4). The species composition, as well as the relative fraction of the respective species, depend on the site-specific oospore distribution and will not be discussed. Using the pooled homogenized sediment samples, the analysis concentrated on the effects on the germination success, as derived from the differences between the irradiance treatment groups.

Table 4. Total numbers of germinated oospores from natural sediments of seven locations under three light intensities.

Location	Replicates	Light irradiance	Total number of germlings	Lychnothamnus barbatus	Chara contraria	Nitellopsis obtusa	Chara globularis	Nitella flexilis/opaca	Nitella mucronata	Chara canescens	Chara filiformis	Tolypella glomerata	Lamprothamnium papulosum	Chara baltica/liljebladii
Lake Balsys	10	1	33	–	3.3 ± 2.5	–	0.7 ± 1.4	–	–	–	–	–	–	–
	10	2	30	2.6 ± 1.7	–	0.2 ± 0.6	–	–	–	–	–	–	–	–
	10	3	5	–	–	–	0.2 ± 0.4	–	–	–	–	–	–	–
Stoßdorfer See	7	1	142	–	–	2.4 ± 2	–	–	–	–	–	–	–	–
	7	2	11	–	–	0.9 ± 0.8	3.4 ± 4.7	–	–	–	–	–	–	–
	7	3	38	–	–	1 ± 1.4	3.4 ± 3	–	14.4 ± 6.6	–	–	–	–	–
Borkener See	10	1	89	–	9.3 ± 9.5	–	1.7 ± 1.4	–	–	8.6 ± 11.2	3.6 ± 3.1	–	–	–
	10	2	28	–	2.4 ± 3.9	–	0.1 ± 0.4	0.5 ± 1.1	–	16.7 ± 11.5	–	–	–	–
	10	3	36	–	4.5 ± 4.9	–	0.6 ± 1.4	1 ± 1.8	–	10.9 ± 11.7	–	3.1 ± 4.8	–	–
Obersee bei Lanke	7	1	13	–	0.9 ± 8.6	–	0.7 ± 1.4	–	–	–	–	–	–	–
	7	2	8	0.5 ± 1	1.3 ± 1.5	–	0.6 ± 0.7	–	–	–	–	–	–	–
	7	3	0	–	–	–	–	–	–	–	–	–	–	–
Großer Goitzschesee	10	1	164	–	11 ± 6.4	–	5.7 ± 4.8	–	–	–	–	–	–	–
	10	2	153	–	10.5 ± 7.5	–	2.5 ± 2	0.5 ± 1.5	–	–	–	–	–	–
	10	3	197	–	16.7 ± 8.6	–	1.1 ± 0.6	5 ± 5.4	–	–	–	–	–	–
Fastensee	10	1	77	–	–	–	–	–	–	0.1 ± 0.3	–	–	11 ± 9.9	–
	10	2	2	–	–	–	–	–	–	1.7 ± 1.9	–	–	0.3 ± 0.7	–
	10	3	0	–	–	–	–	–	–	0.9 ± 2	–	–	–	–
Grabow	7	1	145	–	–	–	–	–	–	0.4 ± 1	–	–	–	20.3 ± 23.7
	7	2	35	–	–	–	–	–	–	0.4 ± 0.7	–	–	–	4 ± 10.6
	7	3	78	–	–	–	–	–	–	–	–	–	–	10.7 ± 14
Generative (oospores) (%)		1	312	0	31	0	25	0	0	31	100	0	97.5	–
		2	239	63	16	0	16	0	0	47	0	0	2.5	–
		3	283	0	21	1	7	0	26	22	0	0	0	–
Vegetative (bulbils, stem-tuber or root nodes) (%)		1	100	0	5	25	34	0	0	0	0	0	0	–
		2	103	37	14	9	5	8	0	0	0	0	0	–
		3	167	0	13	65	13	92	74	0	0	100	0	–

Light intensity: 1 = 15–20 µmol photons s⁻¹ m⁻²; 2 = 30–40 µmol photons s⁻¹ m⁻²; 3 = 110–130 µmol photons s⁻¹ m⁻². Species-specific and site-specific mean values are listed with standard deviations, as well as the fractions of generative and vegetative reproduction.

The results indicated a separation into “irradiance-dependent” and “irradiance-independent” species, with respect to the dependence of germination on light intensity. Lychnothamnus barbatus, Nitella flexilis/opaca, Nitella mucronata, Tolypella glomerata, Chara filiformis and Lamprothamnium papulosum (Wallr.) J.Gr. showed irradiance-dependent germination, but differed with respect to the preferred intensity range. Tolypella glomerata and Nitella mucronata (p ≤ 0.001) germinated exclusively at the highest irradiance (Table 5). Nitella flexilis/opaca showed a significant preference for this range (p ≤ 0.017), but also germinated at lower irradiance. Lychnothamnus barbatus only germinated at medium irradiance (p ≤ 0.011) and failed to germinate at both lower and higher irradiances. Chara filiformis germinated exclusively at the lowest irradiance, with no germination observed at higher irradiances. Lamprothamnium papulosum showed a strong preference for the lowest irradiance (p ≤ 0.025), but two germlings were observed at the medium irradiance.

Table 5. Statistical analyses (Mann–Whitney U test).

Species	Groups of light irradiances
	1–2		2–3		1–3
	U	p (two-tailed)	U	p (two-tailed)	U	p (two-tailed)
Lychnothamnus barbatus	−2.673	0.008*	−2.536	0.011*	−0.014	0.989
Chara contraria	−0.629	0.529	−2.462	0.014*	−1.61	0.107
Nitellopsis obtusa	−0.332	0.74	−0.458	0.647	−0.183	0.855
Chara globularis	−1.434	0.152	−3.211	0.001*	−1.837	0.066
Nitella flexilis/opaca	−0.786	0.432	−2.376	0.017*	−2.874	0.004*
Nitella mucronata	−1.478	0.140	−3.351	0.001*	−3.343	0.001*
Chara canescens	−1.671	0.095	−1.556	0.12	−0.067	0.946
Chara filiformis	0.000	1	−2.357	0.018*	−2.46	0.014*
Tolypella glomerata	−0.971	0.331	−1.361	0.173	−0.333	0.739
Lamprothamnium papulosum	−1.296	0.195	−1.867	0.062	−2.241	0.025*
Chara baltica/liljebladii	−2.241	0.025*	−2.245	0.025*	−0.799	0.424

The U values and the significance level p are listed. Standard p = 0.05 was used for significance (*).

Chara baltica/Chara liljebladii Wallman, C. contraria and C. globularis germinated at all three irradiances tested (Table 4) without exhibiting a clear pattern. Significant differences were detected for the highest (p ≤ 0.025) and lowest (p ≤ 0.025) irradiance for C. baltica/liljebladii, tested against the medium irradiance. Significant differences between the germination rates at the medium and highest irradiance levels were detected for C. contraria (p ≤ 0.014) and C. globularis (p ≤ 0.001).

For C. contraria and C. globularis, germinating from sediments of five and four sites, respectively, site (population) dependency on the irradiance effects could be analysed. For C. contraria, site-specific differences in the irradiance dependency of germination could not be excluded. A preference for germination at low irradiance was very pronounced at Lake Balsys and, to a lesser extent, at Borkener See. This preference was not observed at the other two sites, as in the pooling of all data, the high number of germlings from Grosser Goitzschesee compared to the other sites masked all of the site-specific effects. This discrepancy resulted in a significant preference for the highest irradiance in the analysis of the pooled data, which was not supported by the results from any of the other sites. For C. globularis, the situation was somewhat different. For three out of the four sites, a preference for higher irradiance could be assumed. Only for Lake Balsys, where the germination rate was very low, was the pattern different. Consequently, the overall analysis resulted in a significant difference between the medium and the highest irradiance (p ≤ 0.001), whereas significance was not reached (p ≤ 0.066) for the difference between the lowest and the highest irradiance. However, given the low absolute number of germlings of C. globularis, no definite conclusion can be drawn about the site specificity of the irradiance preference for germination.

The same argument applies for C. canescens and Nitellopsis obtusa (Desv.) J.Gr., neither of which showed any clear irradiance dependency for germination (Table 4). Owing to insufficient numbers of germlings from different localities (C. canescens) and almost exclusively vegetative reproduction (N. obtusa), no site-specific analysis was performed for these two species.

Discussion

The results obtained here allow for better characterization of the germination potential of charophyte oospores. The first step is the estimation of the fraction of viable oospores, established conventionally by means of the crush test (Blindow et al. 2016; Bonis, Lepart, and Grillas 1995; Grillas et al. 1993; Steinhardt and Selig 2007, 2008, 2009, 2011), based mainly on the intactness of the oospore and the presence of starch. Since neither of these aspects is directly linked to the remaining physiological activity, a test of enzymic activity was adopted as a more meaningful alternative. The results showed that the TTC method, developed to detect the vitality of seeds (e.g. Cottrell 1947; Porter, Durell, and Romm 1947; Lakon 1939, 1942; Santos, Novembre, and Marcos-Filho 2007) is applicable not only to seeds of aquatic higher plants (Najas marina: L.Handley and Davy 2005; Zostera capricorni (Ascherson 1876): Conacher et al. 1994), but also to the much smaller oospores of charophytes. The discrepancy between the results of the TTC method and the crush test, the latter indicating a higher fraction of oospores as “intact”, requires further investigation before drawing definite conclusions about overestimation or underestimation of the colonization potential of diaspore banks, because neither approach allows for subsequent germination experiments. Overinterpretation of the present results should be avoided, since we just tested a more meaningful alternative test based on physiological parameters instead of physical integrity.

The TTC test determines the viability of the seed exclusively, as “still physiologically active” or “definitely dead”. The results can be interpreted as the maximum recolonization potential, but without reliable estimates in the terms of germlings. For reliable estimates of the germination potential of the diaspore banks, we need to know the conditions for the activation of the diaspore potential.

This activation consists of two processes: breaking dormancy and, subsequently, the induction of germination. The oospores of charophytes have been shown to gain a primary dormancy by leaving the parent plant (Forsberg 1965; Shen 1966). The loss of dormancy is both time dependent and influenced by external factors (Sederias and Coleman 2009). The fresh plant material thus exhibits a higher degree of dormancy than the oospores of the diaspore banks (Casanova and Brock 1996; Forsberg 1965; Shen 1966; Sokol and Stross 1986; Takatori and Imahori 1971). On the other hand, the viability decreases with time during deposition in the sediment, so breaking dormancy of the oospores collected from fresh plants is of interest for several reasons. Our results did not reveal a general temperature effect on the germination rates, which is in good agreement with the results of Proctor (1967) and Casanova and Brock (1996), confirming the species-specific differences with respect to temperature pretreatments. A different result was obtained for desiccation, where a clear and general positive effect was observed on germination rates after pretreatment. This desiccation effect is surprising to some extent because, except for the C. canescens oospores from Austria, all the samples were taken from permanent water bodies. Osmotic shock or desiccation as a trigger for germination has been reported previously for charophytes, but mainly for species growing in temporary water bodies, e.g. Tolypella salina R.Cor. and Lamprothamnium papulosum (Lambert 2012). On the other hand, desiccation was not an absolute prerequisite for germination in our investigations; rather, it just increased the germination rate. Therefore, it could be argued that this effect is just a part of a pioneer strategy, activating higher fractions of the diaspore potential after an episode of bed air exposure, when competition from other aquatic macrophytes can be assumed to be low.

The strongest effect observed was an absolute dependency of germination on organic material for all Chara species tested. No germination success was observed when providing inorganic substrate only, irrespective of the pretreatment conditions. These results include the pretreatment of oospores with a microbially active phase of organic sediment, leading to the conclusion that organic material is required for the germination process, not for breaking dormancy. This effect may be caused directly by the microbial interaction or by delimitation via the supply of trace compounds not present in the mineral medium. On the other hand, all of the species tested here were able to grow on the organic-free sediments used for the unsuccessful part of the germination experiments (Wüstenberg, Pörs, and Ehwald 2011 and also own observations). Therefore, limitation by trace compounds provided by the organic material and needed for growth can be excluded. However, some compounds released by the sediment microbiome (as the associated prokaryotic community; Bengtsson et al. 2017) are probably required for the activation of the germination process. A further explanation could be the difference between the requirement for a given compound for growth and for germination, the particular concentration being sufficient for vegetative growth but insufficient to satisfy the requirement for germination. Reasons for this difference could be an absence of specific ion transporters in the zygote cell membrane or simply the higher diffusion resistance of the oospore wall, compared with vegetative cells. The mechanism behind the dependency of germination on organic material needs to be investigated in more detail and is the subject of ongoing experiments.

Nitrogen compounds would be good candidates for investigation, because the physiological dormancy model of Hilhorst (1993) identified three main components of the activation process: (1) the influence of temperature in lowering the dormancy level and generating inactive phytochrome receptors; (2) the availability of nitrate to activate the phytochrome receptors; and (3) irradiance, responsible for the photo-transformation of the inactive phytochrome to the physiologically active P_fr form.

The third component listed by Hilhorst (1993), irradiance, was not tested in the dormancy approach. Constant illumination at medium irradiance was supplied for all combinations of pretreatment and germination conditions.

It must be taken into account that irradiance not only plays a role in the phytochrome system, activating the germination process, but also determines the growth and reproduction of aquatic macrophytes (Blindow 1992; Sabbatini et al. 1987; Schwarz, de Winton, and Hawes 2002; Wang, Liu, and Yu 2015; Wang, Yu, and Xiao 2008). The survival of germlings after exhausting the starch reservoir of the oospore therefore depends on irradiance, and the species-specific irradiance sensitivity of germination has been demonstrated in numerous studies on terrestrial (e.g. Lambers and Posthumus 1980; Karaguzel et al. 2004) and aquatic plants (Forsberg 1966; Matheson et al. 2005; van den Berg et al. 1998).

The results obtained in this study show some interesting similarities between the preference ranges for germination and the species distribution along the vertical irradiance gradient. For example, Nitella flexilis/opaca and Nitella mucronata colonize mostly shallow areas and they exhibited a significant preference for higher irradiance in the germination approach. They grow in deeper habitats only when the water is clear enough (Krause 1997; Schubert and Blindow 2004), which is consistent with our results. The fact that N. flexilis/opaca mainly reproduced vegetatively in our study can be explained by the observations of several authors (Asaeda, Rajapakse, and Sanderson 2007; Schwarz, Hawes, and Howard-Williams 1996; Wang, Liu, and Yu 2015; Wang, Yu, and Xiao 2008), all showing that both the production of and reproduction from the vegetative units are encouraged by light. Lychnothamnus barbatus also exhibited pronounced irradiance dependency of germination, preferring the intermediate irradiances. This species mainly inhabits calcareous, oligotrophic–mesotrophic lakes with high water transparency (Balevičius 2001; Casanova, García, and Feist 2003; Raabe et al. 2012; Sinkevičiené 2010). The stands of this tall plant usually do not reach the surface, with the dense meadows starting at about 0.6–1 m depth and only individuals being found in shallower water (own observations). Although both the vegetative and the generative germination were recorded here, there are no differences for this species between the two modes of reproduction with respect to irradiance.

In contrast to the above species, C. contraria and C. globularis produced seedlings under all irradiances, but differed with respect to their preference for generative or vegetative reproduction. The vegetative reproduction of C. contraria increased with increasing irradiance, whereas C. globularis showed maximum vegetative reproduction at low irradiance.

The results presented here add to the knowledge of the estimation of internal restoration potential. For both dormancy breaking and germination, species-specific differences must be taken into account. Especially in the case of irradiance dependency, it may be necessary to translocate diaspore-containing sediments to favourable depth (irradiance) horizons within the lake to maximize the germination rates. Pretreatment by desiccation may increase germination rates in general. The effect of temperature during desiccation is significant, but less important. Being species specific, no general pretreatment recommendation can be given. At least for testing the germination potential, the addition of organic material is mandatory for all Chara species tested to date. The mechanism of this effect still needs to be investigated.

These findings are not comprehensive. Successful colonization of a water body by macrophytes depends on a number of additional factors, including the activity and density of benthic invertebrates and fish (Barko, Gunnison, and Carpenter 1991; Fukuhara and Sakamoto 1987), seed size (Casanova and Brock 1990, 1996; Venable and Brown 1988), physical and chemical substrate characteristics (Barko, Gunnison, and Carpenter 1991; Barko and Smart 1986; De Winton, Clayton, and Champion 2000; Kalin and Smith 2007; Sederias and Coleman 2009), and many others. However, the findings presented here will help to improve the efforts of restoration measures in cases where the biotic and physicochemical circumstances have already been successfully optimized by technical measures.

Notes on contributors

Anja Holzhausen finished her PhD at the end of 2016. Since then, she has been employed as a Research Associate and project member at the University of Rostock. She has been conducting research for 4 years on the ecology, phenology and physiology of the Charophyceae of brackish and fresh waters. Contribution: general concept and research question, data analysis, interpretation of results and manuscript preparation.

Christian Porsche is a biologist specializing in aquatic ecology, working at the Institute for Biosciences, University of Rostock. He has focused on research fields including the ecophysiology of higher plants, ecosystem modelling, and design and statistical analysis of experiments for over 10 years. Contribution: data analysis, interpretation of results and manuscript preparation.

Hendrik Schubert is a marine biologist specializing in aquatic ecology, with a focus on the ecophysiology of autotrophs. He has been a Professor for Ecology at the University of Rostock for 15 years, investigating acclimation of organisms to the brackish conditions of the Baltic Sea and unravelling trophic interactions and the seasonality of coastal ecosystems. Contribution: general concept and research question, data analysis, interpretation of results and manuscript preparation.

Disclosure statement

We confirm that there are no known conflicts of interest associated with this publication.

Acknowledgements

The authors want to thank Maria Rodrigo, an anonymous reviewer and Carles Martín-Closas (invited editor) for their very helpful comments, which improved the manuscript substantially, as well as Mary Beilby for the English proofing. We would like to thank Petra Nowak (Rostock) and Claudia Lott (Rostock) for preparing genetic analyses, and Arne Schoor for laboratory and technical assistance.

One of the authors (AH) was supported by Landesgraduiertenförderung in Mecklenburg Vorpommern and Professorinnenprogramm II of the University of Rostock. The laboratory equipment was partly supported by the European Fund for Regional Development (EFRD).