Effect of different conditions of storage on seed viability and seedling growth of six European wild fruit woody plants

Malus sylvestris(L.) Mill.,Pyrus communis(L.),Sorbus aucuparia(L.),Prunus avium(L.),Prunus padus(L.), andCornus sanguinea(L.) are related, co-occurring species producingorthodoxseeds. However, we observed differences in their response to storage conditions, such as storage at different seed moisture contents (5%, 8%, and 11%) and/or temperatures (− 3 °C, − 18 °C, and − 196 °C). Severe desiccation to ca. 5% of MC negatively affected seeds ofM. sylvestris. Seeds ofP. aviumwere sensitive to storage in LN or at − 18 °C.S. aucupariaseeds are best stored at − 3 °C, whereasC. sanguineaseeds tolerate desiccation and storage in LN. In general, species with deeper physiological dormancy (S. aucuparia,P. padus, andC. sanguinea) tended to be more tolerant to desiccation and low temperatures. For all species, storage conditions did not affect seedling growth. Wild fruit woody species face many threats such as genetic loss, population fragmentation, and alien species; thus, their genetic variability should be preserved. To examine the effect of storage conditions on seed viability and the initial growth of seedlings of six European wild fruit species: Malus sylvestris (L.) Mill., Pyrus communis (L.), Sorbus aucuparia (L.), Prunus avium (L.), Prunus padus (L.), and Cornus sanguinea (L.). Seeds were desiccated to three different levels of moisture content (ca. 5, 8, or 11%) and stored at three different temperatures (− 3°, − 18°, or − 196 °C; liquid nitrogen, LN) for up to 3 years. Germination and seedling emergence tests were performed as well as measurement of seedling growth. Desiccation of M. sylvestris seeds from 10.7 to 4.9% significantly lowered germination from 91 to 77% and seedling emergence from 88 to 74%. In P. avium, LN storage significantly inhibited seedling emergence, both in the laboratory and the greenhouse, but did not affect total seed germination. In P. communis, P. padus, and C. sanguinea, neither germination nor seedling emergence was affected by seed storage conditions. There were small or no differences in stem height and root collar diameter in the first year of seedling growth of stored seeds. Species with deeper physiological dormancy (S. aucuparia, P. padus, C. sanguinea) tended to be more tolerant of various storage conditions. Seeds of P. padus and C. sanguinea can be stored long term at − 18 °C or in LN at 5–8% MC without losing viability. M. sylvestris and P. avium seeds are sensitive to desiccation below 6% MC or low temperature of storage at − 18 °C or − 196 °C, respectively. We observed that storage conditions had significant influence on germination and seedling emergence but had no effect on seedling growth after the first growing season.


Introduction
Biodiversity is the multiplicity of life, including variation among genes, species, and functional traits, in an ecosystem and has a profound effect on the functioning of ecosystems (Cardinale et al. 2012). The benefits of biodiversity in forest ecosystems have been well recognized in recent years, indicating resistance to natural disturbances (drought, severe winds, etc.), greater ecosystem productivity, and higher resilience (Liang et al. 2016;Jactel et al. 2017;Isbell et al. 2015). However, Europe's forest biodiversity remains under pressure due to land-use changes, invasive species, pollution, habitat, and climate changes (EEA 2016). This results in declining species richness in European forests as well as in global forests (Hooper et al. 2005;Schulze et al. 2014;Forest Europe, 2015).
Some of the species that make the greatest contributions to broad-leaved forest ecosystem function are fruit-bearing trees and shrubs. They provide multiple ecosystem services as a source of nutrition, pollination, genetic variation for breeding programs, and even of valuable income for people living around them (Powell et al. 2013). Wild fruit trees are considered less economically important than timber species and thus have often been neglected in conservation programs. Although wild fruit species are usually not considered endangered by European countries, their genetic diversity is threatened (Dirlewanger et al. 2002;Larsen et al. 2006;Wünsch and Hormaza 2007;Urrestarazu et al. 2016). Wild fruit trees are usually pioneer or intermediate species with weak competitive ability (Stephan et al. 2003;Russell 2003). Even considering their usually broad natural distribution in Europe due to their rare occurrence and scattered distribution pattern, populations are disappearing at an alarming rate (Coart et al. 2003). Their high light and soil demands, as well as slow growth, often result in low natural regeneration restricted to forest ecotones (Stephan et al. 2003;Russell 2003). A number of wild species often hybridize with cultivated species, leading to loss of their true wild form and genetic swamping (Fellenberg et al. 2000;Bleeker et al. 2007). The plantation-based model of fruit collection, as well as the lack of a uniform plan for preserving genetic resources in most of these species in the EU, leads to a further decrease in these populations (Wolf et al. 1999). Nevertheless, a number of wild fruit tree populations registered in the EUFGIS database for forest genetic resources are underrepresented and restricted to few countries in comparison with timber species. Therefore, more efforts to conserve the genetic resources of those species must be made.
One very efficient method for conserving the genetic resources of temperate forest trees is seed banks (FAO 2014). All wild fruit tree species from temperate and boreal regions produce orthodox seeds that can be stored for long periods of time at a low seed moisture content, below 5%. Such desiccation tolerance allows the storage of seeds beyond average tree lifespan at a conventional temperature of − 20°C (Pritchard et al. 2014) and later use in reintroduction programs. Although the total number of temperate tree species stored in seed banks is rather high, interspecies genetic diversity is generally not well conserved (FAO 2015). In addition to conventional methods, cryopreservation in liquid nitrogen (LN) is used as a secure backup for plant genetic resources. Cryopreservation is usually utilized for storing recalcitrant seeds, but due to its simple methodology, it is used for many orthodox seeds as well . Effective seed cryopreservation has been proven for many temperate forest tree species, e.g., Alnus glutinosa L. (Chmielarz 2010), Corylus avellana L. , Malus sieversii M.Roem. (Kushnarenko et al. 2010), Fraxinus excelsior L. (Chmielarz 2009a), Prunus armeniaca L. (Jaganathan et al. 2015), Populus nigra L. (Michalak et al. 2015a), and Prunus padus L. (Popova et al. 2016).
In our study, we examined seed storage and seedling growth in 6 wild fruit species: European crab apple (Malus sylvestris L.), European pear (Pyrus communis L.), mountainash (Sorbus aucuparia L.), wild cherry (Prunus avium L.), bird cherry (Prunus padus L.), and common dogwood (Cornus sanguinea L.). All described species suffer from fragmentation and disappearance of their natural habitats as mentioned earlier. In the case of the endangered European crab apple, the hybridization between this species and the domesticated apple could lead to the extinction of its wild populations (Reim et al. 2012). On the other hand, populations of bird cherry have to compete for habitats with highly invasive species such as black cherry (P. serotina; Pairon et al. 2010). Nevertheless, apart from European crab apple and wild cherry, due to their usage in fruit production, information about seed storage behavior in the scientific literature is very limited and has not been properly examined. However, with the current climate and environmental changes, it is essential to assess the best possible storage conditions for effective conservation of these often-neglected species.
Seeds stored in gene banks must produce plants capable of reproduction in natural conditions, and high seedling quality is crucial for the initial success of any reforestation program (Grossnickle 2012). Often, seedling death is the result of environmental stress, animal grazing, disease, or insects. However, in many cases, survival can be significantly increased with good-quality planting stock (Campo et al. 2010). Seed storage conditions and practices have a strong influence on seed quality and therefore on seedlings. The effects of these practices on early seedling establishment need to be understood to produce high-quality seedlings that have a high probability of being "coupled" into forest ecosystems and reproducing (Grossnickle 2005). To our knowledge, studies examining seedling quality after storage are still rare. Recent publications assessing this subject confirm a possible effect on seedling quality both in annual plants (Marshall and Lewis 2004;Arguedas et al. 2018) and in woody species (Benamirouche et al. 2018).
The aim of this study was to examine the seed biology of 6 wild fruit tree species before and during storage under controlled conditions for 3 years. Our research questions were as follows: how do storage temperature and seed moisture content affect seed germination, both in controlled and field conditions? Do desiccation-tolerant seeds always tolerate storage at ultra-low temperatures (− 196°C) and does it affect subsequent seedling growth? What are the possibilities for ex situ conservation in nurseries, seed, and cryobanks of examined species? 2 Material and methods

Plant material
Mature seeds of the six tested species, European crab apple (Malus sylvestris L.), European pear (Pyrus communis L.), mountain-ash (Sorbus aucuparia L.), wild cherry (Prunus avium L.), bird cherry (P. padus L.), and common dogwood (Cornus sanguinea L.), were collected between May and September 2012 (Table 1) in natural stands in Poland (Wawrzyniak et al. 2020). Seed lots of each species were collected from individual trees (2-3), with the exception of European crab apple, where seeds from three adjacent individuals were mixed together. All experiments were conducted on whole seeds, with P. avium, P. padus, and C. sanguinea seeds consisting of endocarp. All tested species produce fleshy fruit; in order to obtain the seeds, collected fruits were pressed and seeds were extracted from the pulp, as described by Suszka et al. (1996). Then, seeds were air-dried in ambient conditions (20°C) until they reached approximately 11% moisture content (control). Seeds were stored at 3°C until the start of the experiment (1-2 months depending on species).

The moisture content of the seeds
Seed moisture content (MC) refers to the seed-containing embryo and seed coat. MC was calculated on a fresh weight basis (%) using the formula: where MC 1 is the moisture content, FW 1 is the initial fresh weight, and DW is the dry weight. The MC of the seeds was adjusted prior to seed storage with desiccation over activated silica gel to obtain two levels of MC: 5% and 8%. The adjustment of the MC of seeds was based on the FW of the seeds, according to the formula: where FW 2 is the desired weight and MC 2 is the desired moisture content. The moisture content of the seeds was achieved specified by drying in the oven at 103 ± 2°C for 17 h (4 replicates of 10 seeds). The exact seed MC levels achieved are shown in Table 2.

Seed storage
Seeds of each species were dry-stored at three levels of approximate MC (5, 8, and 11%; Table 2), in plastic air-sealed bags at temperatures of − 3°C, − 18°C, or cryopreserved at − 196°C for 12, 24, and 36 months. As a control, we used nonstored seeds at approximately 11% MC. Cryopreserved seeds in sealed polypropylene bags were directly immersed in liquid nitrogen (LN). After storage, seeds were thawed in a water bath at 40°C for 10 min. Subsequently, seeds were placed in a stratification substrate and underwent stratification.

Stratification and germination
After each storage interval and prior germination and seedling emergence tests, seeds of all the examined species underwent a species-specific stratification process, as presented in Table 3. The stratification substrate was composed of a moist mixture (v/v, 1:1) of quartz sand (fraction > 1 mm) with sieved peat of pH 3.5-4.5. Seeds were mixed with the substrate at a ratio of 1:3 and placed in loosely sealed plastic containers. The condition of the seeds and the substrate was checked every 2 weeks. Stratification continued until the appearance of the first germinating seeds (≤ 5%). Directly after stratification, germination and seedling emergence tests were conducted in a laboratory under controlled conditions. Tests were performed in a growth chamber at cyclically alternating temperatures of 3°/20°C with an interval of 16/8 h. Seeds with a radicle of ≥ 3 mm were counted as germinated. Germinated seeds were examined weekly, counted, and eliminated until the last seed germinated. After the test, non-germinated seeds were cut and evaluated as healthy and dormant or decayed. During the germination test, the total germination percentage (GP), as well as the mean germination time (MGT or T ), was calculated according to the formula given by Ranal and Santana (2006) based on Labouriau (1983). Germination time was given in weeks and counted from the beginning of cold stratification.
where T i is the time from the start of the experiment to the ith observation and N i is the number of seeds germinated in the ith time (the number of germinated seeds corresponding to the ith observation); k is the last observation. Seedling emergence (SE) tests were conducted in a similar way using a mixture of sand and peat. Stratified seeds were sown in plastic boxes and covered with a 1 cm substrate layer. Seedling emergence tests were conducted at the same temperatures as germination tests. The seed was regarded as emerged when cotyledons and epicotyl were observed above ground. After seedlings reached 2-3 cm in height, boxes were moved into the chamber with light and a temperature of 20°C. A sprouting index (SI) was calculated using the same formula as for calculating MGT, counted from the end of stratification. For germination and seedling emergence tests, 3 replicates of 30 seeds or 4 replicates of 50 seeds (in the case of mountainash and common dogwood) were used. For each test, separate seed samples were used.

Seedling emergence and further seedling growth in a greenhouse
In addition to the laboratory tests, a field emergence test was conducted. Stratified seeds were sown into plastic containers (FlexiFrame, BCC Sweden), with each seed in a separate cell, and placed in random blocks in the greenhouse. Seeds were placed into a mixture of peat, perlite (v/v, 2:1), and longlasting mineral fertilizer (Osmocote Extract 5-6 M 15-912 + 2MgO + TE, 2-4.5 mm). The test was performed on 3 replicates of 30 seeds or 4 replicates of 50 seeds (in the case of mountain-ash and common dogwood). One month after  sowing, seedlings were counted. In autumn, after the vegetation period, all 1-year-old seedlings were measured for their height (H) and root collar diameter (RCD). These tests were repeated each year during the 3-year seed storage period. The temperature of the air and ground in the greenhouse was recorded during the experiment (Table 7).

Statistical analysis
The fixed effects of storage conditions on germination and seedling emergence were evaluated separately using a generalized linear model (GZLM) with a binomial distribution with loglink function (Ranal and Santana 2006). For the germination rate and sprouting index comparison, a linear regression model was used. Germination means used in statistical analyses were counted from 3 to 4 replicates (for each 30-50 seeds), each obtained from separate germination tests. A three-way analysis of variance (ANOVA) with interaction between main effects (storage time, moisture content, and storage temperature) was used in the linear regression model (Table 9). For GZLM, special analysis was based on deviance (Carvalho et al. 2018). For multiple comparisons between treatments, we adjusted p values using Bonferroni-Holm method. For the linear regression models, we tested normality and equal variance assumptions of residuals by examining the plots of residuals. For growth analysis (height and root collar diameter measurements), a nonparametric Kruskal-Wallis test was performed (p < 0.05; from agricolae package). For all statistical analyses of the data, R statistical software (R Core Team 2020) was used.

Effects of storage conditions on seed germination
After the third year of storage, seeds' average germination capacity of all species was above 80%, regardless of the storage conditions used. The main effects (storage time, seed MC, and storage temperature) had visible impact on seed germination and seedling emergence process. The interaction effects on germination were recorded between storage time and seed MC in bird cherry and between storage time and temperature in common pear as well as wild cherry (Table 8). The MC of stored seeds mainly affected germination of crab apple and mountain-ash. Different storage temperatures had major effect on wild cherry seedling emergence. Although common dogwood seeds germinated and emerged at high percentage regardless of the storage conditions used, nonstored seeds germinated significantly better than stored seeds. The mean time to germinate and seedling emergence index in most cases were alike. The seedling emergence tests were more sensitive on influence of storage conditions and therefore better indicate seed quality after the storage in comparison with germination test.

European crab apple
The best results were for European crab apple seeds stored at − 18°C at 8.5% seed MC, where germination and seedling emergence both in the laboratory and in the greenhouse were high, at approximately 93%, 89%, and 83% and similar for each tested year of storage, respectively (Fig. 2). Germination after storage at − 3°C showed a slight but significant decrease. The germination percentage was significantly lower after 1 and 3 years of storage than that of nonstored and seeds that were stored for 2 years (Table 4). The mean germination time was the highest after 3 years of storage: 12.2 weeks after the start of stratification. Seedling emergence both in the laboratory and in the greenhouse significantly decreased in each tested year. The nonstored seed emergence was 90.7% in the laboratory and 91.5% in the greenhouse, in comparison with 76% and 78% emerged after the third year of storage, respectively. The mean time of seedling emergence varied after each year of storage, with the lowest after the first year (2.1 weeks) and the highest after the second year (2.6 weeks) of seed storage. The temperature used for storage did not alter the germination percentage, which was between 85 and 88% (Table 5). Seed desiccation to 4.9% significantly lowered germination, from 92% in the control to 77% (Table 6). However, the mean germination time was a week faster (11.4 weeks) in strongly desiccated seeds than in the control (12.3 weeks). A similar effect was recorded for seedling emergence, which was significantly lower in seeds desiccated to 4.9% both in the laboratory and in the greenhouse. Desiccated seeds of European crab apple (5.5%) had a significant decrease in seedling emergence after storage at each tested temperature. Seedling emergence of M. sylvestris seeds stored at − 196°C decreased after the first year of storage, regardless of seed MC.

European pear
In general, nonstored seeds of European pear (P. communis) germinated at a slightly but significantly higher percentage (99%) than stored seeds (96-97%; Table 4). After the first year of storage of seeds desiccated to 7.1% MC and stored at − 196°C, we recorded some decrease in all values of the conducted tests, but after the third year of storage, seeds maintained their germination ability at the same level as the control seeds (Fig. 3). The mean germination time was 16 weeks after the start of desiccation in each tested year. Seedling emergence was the highest in the nonstored control and after the third year of storage (95-97%). However, the mean time of seedling emergence differed significantly: after 1 year of storage, seedlings emerged almost 1 week faster (1.3 weeks) than after the other tested storage times (approximately 2 weeks). Seedling emergence in the greenhouse was significantly higher after the third year of storage (93%) than for nonstored seeds (83%). Seedling emergence after exposure in seeds stored in LN was significantly lower (90%) than the emergence for seeds stored at − 18 (97%) and − 3°C (96%; Table 5). Seedling emergence in the greenhouse was 83% after storage in LN, 92% after storage at − 18°C, and 90% after storage at − 3°C. Seed moisture content had little or no effect on seed germination and emergence ( Table 6). The highest germination percentage (99%) was recorded at 7.1% MC. Seedling emergence in the laboratory was significantly lower for seeds desiccated to 5.3% (91%) than for those in the control treatment (95%). There were no significant differences between seedling emergence percentages in the greenhouse; at all tested MC levels, the percentage was 87-89%. Seeds of P. communis desiccated to ca. 5% were more vulnerable to storage at − 196°C. Higher MC during storage in LN resulted in better germination percentage and seedling emergence in greenhouse.

Mountain-ash
The seeds of mountain-ash (S. aucuparia) had significantly lower germination and seedling emergence in the laboratory after storage in all tested conditions in comparison with nonstored seeds. The germination percentage after the first and second year was 10% lower than in the control seeds Index means in the same row marked with the same letter are not significantly different at p < 0.05. Means without letter marks are not significantly different at p > 0.05 (85%; Table 4). The mean germination time was the lowest after the second year of storage (20.3 weeks). Seedling emergence in the laboratory also significantly decreased after storage. Seedling emergence in a greenhouse was significantly lower after the first and second years of storage. The highest germination percentage was recorded in seeds stored at − 3°C (80%) in comparison with 77% at both lower temperatures (Table 5). Similar results were recorded for seedling emergence in the laboratory, where emergence after storage at − 3°C was the highest (72%). The temperature had no effect on either the mean germination time (20 weeks) or the sprouting index (1.8 weeks). Desiccated seeds of mountain-ash (5.5% MC) had a significantly higher germination percentage and seedling emergence in the greenhouse compared with the control (11.1% MC; Table 6). Seedling emergence in the laboratory was recorded at 67% for the control and 74% for desiccated seeds. In seeds of mountain-ash, all tested temperatures and MCs after the third year of storage resulted in high viability of seeds in comparison with the nonstored control. Only desiccated seeds (8.4 and 5.5% MC) stored in LN had a significantly lower germination percentage after the third year. The most consistent results were recorded after 3 years of storage for seeds with 8.4% MC at − 3°C (Fig. 4).

Wild cherry
After 3 years of storage of wild cherry seeds, its germination rate remained high (89%) at temperatures of − 18°C and − 3°C . Storage in LN significantly decreased seedling emergence both in the laboratory and in the greenhouse; however, germination after storage in LN remained high. We obtained the best results at a storing temperature of − 18°C in seeds with 11.2% MC (Fig. 5). The seeds germinated at a similar level in the control and after all tested storage years (Table 4). Seedling emergence in the laboratory in all the tested conditions was significantly lower after storage (73-81%) than in the control (89%). The mean time of seedling emergence was 1 week faster in fresh nonstored seeds, 2.7 weeks in comparison with 4.3 weeks after the third year of storage. After the third year of storage, seedling emergence in the greenhouse was 89% in comparison with the control result, 86% of germinated seeds. The storage temperature significantly affected the percentage and the rate of germination and seedling emergence. After storage in LN, the germination percentage decreased from 92% at − 3°C and 91% at − 18°C to 82%, and the mean germination time was significantly longer, increasing from 15.4 to 16.6 weeks (Table 5). Seedling Index means in the same row marked with the same letter are not significantly different at p < 0.05. Means without letter marks are not significantly different at p > 0.05 emergence in the laboratory after storage at traditional temperatures was 89% at − 3°C and 85% at − 18°C, while seedling emergence for seeds stored in LN was 66%. Seedling emergence in the greenhouse after LN storage was almost 20% lower than that of seeds stored at the other tested temperatures. The mean seedling emergence time was higher after LN storage. Desiccation of P. avium seeds to 5.5% slightly decreased germination from 90% at 11.2% MC to 86% (Table 6). Seeds stored at higher MC levels germinated faster, from 15.9 weeks at 5.5% MC to 15.7 weeks at 11.1% MC. Seedling emergence in the laboratory was the lowest for desiccated seeds, at 77%. In the greenhouse, seedlings emerged at 82% for all tested MC levels.

Bird cherry
Storage conditions had little or no effect on bird cherry (P. padus) seed germination and seedling emergence (Fig.  6). Germination was similar among nonstored seeds and all storage times, varying from 88 to 91% (Table 4). The mean germination time varied at each year from 23.9 weeks after 2 years of storage to 24.7 weeks after the third year. Seedling emergence in the greenhouse showed no significant differences among all storage years except for the second year, where it was slightly lower (84%) in comparison with the first and third years of storage (89-88%). Seeds germinated in all the tested temperatures with a mean germination time of approximately 24 weeks after the start of the cold stratification phase. Desiccation of seeds did not result in changes in the germination percentage (89-90%; Table 6) or the mean germination time (24 weeks).

Common dogwood
Seeds of common dogwood (C. sanguinea) were characterized by similar durability, in which the only factor that slightly affected germination and seedling emergence was storage time. Nonstored seeds of common dogwood germinated at 97% in comparison with 93-94% for stored seeds (Table 4). The mean germination time varied from 26.4 weeks after the first year of storage to 27.6 weeks after the third year. Seedling emergence in the laboratory was significantly higher in nonstored seeds (97%) and decreased in storage seeds (87-92%). Additionally, the mean time of emergence varied each year, resulting in the fastest, 4.4 weeks in the seeds stored for 2 years, to the slowest, 5.4 weeks in nonstored seeds. The results of seedling emergence in the greenhouse were much lower than those in the laboratory, resulting in emergence between SI, weeks 4.9 ± 0.07ab 5.0 ± 0.07b 4.9 ± 0.08a SE in greenhouse, % 63.9 ± 1.28 64.9 ± 1.31 64.7 ± 1.26 Index means in the same row marked with the same letter are not significantly different at p < 0.05. Means without letter marks are not significantly different at p > 0.05 germination percentage or seedling emergence both in the laboratory and in the greenhouse (Table 5). Seeds germinated at a high level, 94-95%, at all the tested temperatures, and the mean germination time was 27 weeks from the beginning of stratification. Seedling emergence in the laboratory was between 91 and 92%, much higher than in the greenhouse, where it was 63-66% (Fig. 7). Similar results were obtained for different seed MC levels. At all MC levels, between 94 and 95% of seeds germinated with a mean time of 27 weeks (Table 6).

Height and root collar diameter of seedlings
The height of 1-year-old seedlings from seeds stored in controlled conditions was comparable in all tested conditions. All tested species' seedling height and diameter were the lowest in nonstored seeds and the highest after the third year of storage ( Fig. 1a, d). In the case of seedlings of European crab apple, seeds stored at controlled temperatures did not differ among the tested temperatures, and seedling height varied from 34 to 36 cm (Fig. 1b, e). The root collar diameter in seedlings from seeds stored at − 3°C was 3.92 mm and was significantly lower than that from seeds stored at other storage temperatures. The shortest seedlings (33 cm) were obtained from seeds stored at 8.5% MC, while seedlings grown from seeds with 5 and 10.7% MC were statistically significantly taller (Fig. 1c, f). A similar result was obtained for European pear, in which the shortest seedlings grew from seeds stored at 7.1% MC (34 cm) in comparison with 35 and 36 cm for the heights of seedlings grown from seeds stored at 5.3 and 11.9% MC, respectively (Fig. 1c). The thickest root collar diameter was 4.28 mm in seedlings grown from seeds at 7.1% MC. The tallest seedlings were obtained from mountain-ash seeds stored at − 3°C, at 53 cm, and they were 2 cm higher than seedlings grown from seeds stored at lower temperatures (Fig.  1b). The root collar diameter in mountain-ash was similar at all the tested temperatures and varied from 5.29 to 5.34 mm. Seedlings grown from desiccated seeds (5.5%) were 53 cm in height in comparison with 51 and 52 cm heights of seedlings grown from 8.4 and 11.1% of MC, respectively. Seedlings of wild cherry grown from cryostored seeds grew 2 cm less (34 cm) than seedlings stored at higher temperatures (Fig. 1b). The thickest root collar diameter was obtained in seedlings from seeds stored at − 18°C. The height and root collar diameter were similar for all tested MC levels. The moisture content of the stored bird cherry seeds had no effect on the height of seedlings, which was 29 cm for all tested MC levels of seeds. The root collar diameter (4.36 mm) was slightly larger in seedlings grown from seeds with 8.3% MC. In common dogwood, seedlings grown from cryostored seeds had the smallest root collar diameter of 3.56 mm, in comparison with 3.66 and 3.64 mm in seedlings from seeds stored at − 3 or − 18°C (Fig. 1e). The shortest seedling, at 49 cm, was grown from seeds stored at 11% MC, 4 cm shorter than seedlings grown from seeds with greater MC. The root collar diameter for seedlings grown from seeds stored at 8% MC was 3.51 mm, in comparison with 3.7 and 3.65 mm for seedlings grown from seeds stored at 5.5% and 11% MCs, respectively.

Discussion
Seeds of the investigated species responded variously to storage conditions. Although they belong to the same storage behavior category (orthodox or uncertain in case of C. sanguinea according to the SID database) and originate from the same ecological niche, their MC and temperature during storage had different effects on germination and seedling emergence. Probert et al. (2009) showed that seed longevity is affected primarily by the environment and endosperm presence. Non-endospermic seeds from dry, hot environments tend to have a longer life span in storage. Oil content and seed mass, according to those authors, have no effect on seed longevity. However, the oligo-/disaccharide ratio in seeds plays an important role in seed longevity (Lin and Huang 1994). In our research, seeds were rather small, with 1000-seed weight of 7-32 g in Malus sylvestris, Pyrus communis, and Sorbus aucuparia, 52-84 g in Cornus sanguinea and Prunus padus, and 184 g in P. avium. The large size of P. avium seeds could have a negative effect during its freezing in LN. Also differences in seed coat thickness can influence seed longevity, as seeds with less permeable seed coats deteriorate slower (Debeaujon et al. 2000). Besides, Nguyen et al. (2012) show that dormancy in Arabidopsis mutants is negatively correlated with seed longevity. In our study, we observed that species with more complex dormancy (C. sanguinea, P. padus, and S. aucuparia; Table 6) tended to be more tolerant to both desiccation and storage temperature.
Storage of Malus sylvestris seeds at 5.5% of MC in our experiments resulted in a decrease in all performed seed viability tests. Storage of seeds below 5% of MC was considered to increase seed longevity by other authors (Perez-Garcia et al. 2009;Ellis 1995;Ellis et al. 1996). However, an experiment by Hong et al. (2005) shows that ultra-dry storage can be beneficial in ambient conditions but not always at subzero temperatures. This was also confirmed by experiments of Haiying et al. (2002) on tomato seeds (Solanum lycopersicum) stored at 2.3% of MC for 13 years at 20°C and of Ellis et al. (1986), in which seed desiccation of Sesamum indicum from 5 to 2% of MC increased storage time 40-fold. Walters (2007) indicated that, below a critical moisture content, longevity does not increase with further drying. A similar conclusion was reported by Michalak et al. (2015b), as storage of European crab apple desiccated to 4% of MC also resulted in decreased germination and seedling emergence. We noted that seed storage of M. sylvestris at MC 5.5% in LN resulted in a significant decrease in seedling emergence. Despite the decline in seedling emergence, desiccated seeds of European crab apple tend to release dormancy a week earlier than the control. In general, desiccation plays a crucial role in the after-ripening process, which prepares seeds to germinate. As noted by Angelovici et al. (2010), gene expression and metabolic processes during seed desiccation strongly resemble those observed during germination. This suggests that one of the functions of the desiccation process is preparation for germination. Seed storage conditions of Pyrus communis affected neither germination nor seedling emergence. However, despite being desiccation-tolerant (to MC 5.3%), seeds stored at low MC in LN germinated at a lower rate. There is little information about storing Sorbus aucuparia seeds in the literature. Gordon et al. (1992) recommend storage of Sorbus spp. at 6-8% of MC, at 1-4°C for up to 8 years. As reported by Granstrom (1987), mountain-ash seeds survive and remain viable for up to 5 years in the soil bank. In our experiment, nonstored seeds germinated faster and better than seeds stored for 1-2 years. This might be due to the delay in seed sowing caused by a lack of space in the growth chambers, as the proportion of decayed seeds after M . s y lv e s tr is P . c o m m u n is S . a u c u p a r ia P . a v iu m P . p a d u s C . s a n g u in e a M . s y lv e s tr is P . c o m m u n is S . a u c u p a r ia P . a v iu m P . p a d u s C . s a n g u in e a M . s y lv e s tr is P . c o m m u n is S . a u c u p a r ia P . a v iu m P . p a d u s C . s a n g u in e a M . s y lv e s tr is P . c o m m u n is S . a u c u p a r ia P . a v iu m P . p a d u s C . s a n g u in e a RDC, mm RDC, mm RDC, mm  Fig. 2 Germination and seedling emergence in laboratory and greenhouse of seeds of Malus sylvestris after 1, 2, and 3 years of storage in controlled conditions. One-way ANOVA; columns marked with the same letter are not significantly different at p < 0.05, by Tukey's test. Means without letter marks are not significantly different at p < 0.05. Means ± SE stratification never exceeded 4.5%. Therefore, to hinder germination, nonstored seeds were placed at − 1°C for 2 weeks prior to sowing. According to Afroze and O'Reilly (2015), as well as our observations (Fig. 8), freezing can result in higher germination. For confirmation, additional seeds of mountain-ash were examined, with a 2-week treatment at − 1°C after the end of stratification. Seeds subjected to freezing germinated approximately 10 times better than the control. The lower temperature may be beneficial in breaking seed-coat-induced dormancy (Raspé et al. 2000). Perez-Garcia et al. (2009) found that, in some species of the family Brassicaceae, long-term storage increased germination due to dormancy loss. However, after the third year of storage, both the rate and total germination showed no difference from nonstored seeds. Information on the storage behavior of Cornus sanguinea seeds is scarce in the literature. The Seed Information Database (Kew Royal Botanical Gardens 2019) states that its storage behavior is uncertain. Seed survival in the transient seed bank was less than a year, which may indicate that seeds are short-lived under ambient conditions (Thompson et al. 1997). In our study, seed desiccation to 5.5% of MC did not affect germination, which reached 95% (vs. 94% in the control). The seeds remained viable at all the tested storage temperatures for 3 years; therefore, seeds of this species should be classified as orthodox. Our study shows for the first time the feasibility of cryopreservation of common dogwood seeds in the safe range of 5.5-11.0% of MC. The seedling emergence test in the greenhouse showed much lower results than in its laboratory counterpart. This is most likely due to the late sowing date (second half of April), in which high soil temperatures could induce secondary dormancy (Benech-Arnold et al. 2000).
Seeds of Prunus padus germinate even after desiccation to 3.5% according to Popova et al. (2016). At the tested temperatures, the seeds remained viable; therefore, seeds of bird cherry can be successfully stored in conventional conditions defined by gene bank standards (FAO 2014). In the literature, for medium-term storage of this species, temperatures of 3-5°C with MC below 12% were recommended (Gordon and Rowe 1982), whereas for long-term storage, a temperature of − 18°C at ca. 5% of MC is recommended (Kew Royal Botanical Gardens 2019). Our results indicate that long-term storage of cryopreserved seeds of bird cherry is possible at all the tested MC levels (seedling emergence 76-79%). Popova et al. (2016) report that the safe MC for cryogenic storage of bird cherry seeds is 3.5-15.0%, which is a wider range than in other "oily" orthodox seeds. On the other hand, seeds of wild cherry in our study tolerated desiccation to MC 5.5% and germinated at 86%, compared with 90% in the control (MC 11%). In general, wild cherry seeds are desiccation-tolerant to 1.6% of MC (Chmielarz, 2009;Royal Botanic Gardens Kew, 2019;Jensen and Eriksen 2001). However, partial drying of freshly collected seeds can result in a significant decrease in germination after immediate stratification and sowing (Michalak et al. 2015c). Further storage of dry seeds in tightly closed vials restores germination capacity to higher values (Suszka 1964). In our research, storage of seeds at ultralow temperatures resulted in a significant decrease in seed viability, which was detected in the seedling emergence test (although in the germination test the results were comparable), so wild cherry seed can be classified as desiccation-tolerant and LN-sensitive. Cryopreservation considerably hampers seed emergence (66% in the laboratory and 70% in a greenhouse) but is feasible at a narrow range of seed MC (Chmielarz, 2009). The cited study suggests that low tolerance to subzero temperatures may result from the high content (45%) and composition of storage lipids in the seeds of wild cherry. Damages can occur during changes in the phase transitions of the membranes from liquid to Pyrus communis Fig. 6 Germination and seedling emergence in laboratory and greenhouse of seeds of Prunus padus after 1, 2, and 3 years of storage in controlled conditions. One-way ANOVA; columns marked with the same letter are not significantly different at p < 0.05, by Tukey's test. Means without letter marks are not significantly different at p < 0.05. Means ± SE crystalline (Stanwood 1985). A similar composition of seed storage lipids is also observed in Hippophaë rhamnoides, which-like wild cherry-germinates but does not develop into viable seedlings after storage in LN (Wawrzyniak unpublished). Hor et al. (2005) report that there is a negative correlation between lipid content of seeds and unfrozen water, which affects their ability to regenerate after exposure to LN. Out of the investigated 6 species, Prunus avium, Pyrus communis, and M. sylvestris seeds do not fully tolerate storage in LN after severe desiccation to MC ca. 5%. However, in contrast to Prunus avium and Pyrus communis, whose seeds tolerate desiccation to MC 5%, M. sylvestris seeds are negatively affected by severe desiccation to MC ca. 5%. S. aucuparia, Prunus padus, and C. sanguinea seeds tolerate storage in LN at MC ca. 5%.
Although there are many publications describing the effect of seed storage on germination, few studies have addressed the further growth of seedlings. For example, seedlings of Coffea arabica grown from seeds stored for 9 months at 10-11% of MC and 10-15°C were not suitable for nursery planting. Additionally, seedlings grown from seeds stored at 18% of MC and 20°C were 50% shorter than from those stored at higher MC levels (da Rosa et al. 2011). Mucha et al. (2015) showed that seed storage temperature influences the root anatomy of black poplar (Populus nigra) and consequently may alter nutrient absorption. Good-quality seedlings with greater height and root collar diameter confer a higher chance of survival in natural conditions (Grossnickle 2012). Therefore, we conducted field research after the first year of the growing season in the greenhouse. In all the tested species, seedling emergence in the greenhouse was slightly lower than in laboratory conditions but still high, regardless of storage conditions. The height and the root collar diameter differ significantly between the tested years but probably due to differences in temperatures during the growing season (Table 7) rather than seed storage time. Differences in morphological attributes at different moisture levels and temperatures were small and did not affect seedling quality despite their statistical significance. However, this is not always the case. Arguedas et al. (2018) found that seedlings of corn (Zea mays) grown from cryostored seeds had delayed growth, although they did not differ phenotypically. The measurements in our study were made only at the end of the growing season, so we cannot conclude whether any differences were visible at the initial phase of growth. However, even if there were any differences initially, the seedlings from all the tested variants evened out by the end of the season.
Basing on our results, we can recommend storing seeds of Malus sylvestris at − 18°C and ca. 8%. For seeds of Prunus avium, the safest temperature for storage is − 3°C. Desiccation to ca. 8% before storage at low temperature can be recommended for seeds of Sorbus aucuparia. In the case of Pyrus communis, severe desiccation should be avoided if storage in LN is planned. Prunus padus and Cornus sanguinea can be safely stored at low seed MC and temperature, as they proved to be the most tolerant. All the tested seed storage methods can be successfully used in nursery production, as no or minimal morphological differences occurred in seedling growth.
In general, species with deeper physiological dormancy (S. aucuparia, P. padus, C. sanguinea) tended to be more tolerant to desiccation and low temperatures. However, our study confirms that severe desiccation can damage some seeds even if classified before as orthodox species. Therefore, more accurate experiments on the long-term storage of seeds of woody species should be conducted, especially for species subject to rapid environmental changes.   Fig. 3 Germination and seedling emergence in laboratory and greenhouse of seeds of Pyrus communis after 1, 2, and 3 years of storage in controlled conditions. One-way ANOVA; columns marked with the same letter are not significantly different at p < 0.05, by Tukey's test.
Means without letter marks are not significantly different at p < 0.05. Means ± SE  Fig. 7 Germination and seedling emergence in laboratory and greenhouse of seeds of Cornus sanguinea after 1, 2, and 3 years of storage in controlled conditions. p < 0.005, Tukey test. One-way ANOVA; columns marked with the same letter are not significantly different at p < 0.05, by Tukey's test. Means without letter marks are not significantly different at p < 0.05.

Means ± SE
Page 17 of 20 58 Annals of Forest Science (2020) 77: 58 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Fig. 8 Germination rate of seeds of Sorbus aucuparia after stratification in 3°C and no additional storage (control) or with additional freezing in 2week storage in − 3°C (freeze). Error bars represent standard error for mean. Additional test was made for 3 replicates per 30 seeds, after 1 year storage in − 3°C