Introduction
Conservation of plant genetic diversity requires both in situ and ex situ approaches. In situ conservation, which works at the habitat level, is preferred, but protected landscapes remain threatened by several factors, such as fire, deforestation, diseases and habitat fragmentation. This vulnerability can lead to depletion of genetic diversity within populations, threatening their long-term stability survival (Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006; Cruz Neto et al., Reference Cruz Neto, Aguiar, Twyford, Neaves, Pennington and Lopes2014). Ex situ conservation can mitigate the loss of genetic diversity in fragmented habitats by maintaining reserves of germplasm in botanic gardens and seed banks. Technologies for ex situ conservation are readily available for species that produce seeds that survive drying to low relative humidity; once dried, these ‘orthodox’ seeds maintain viability when placed in freezer storage at −20°C (FAO, 2014). However, not all species produce seeds with high tolerance to desiccation.
Seeds that are sensitive to desiccation were classified by Roberts (Reference Roberts1973) as ‘recalcitrant’ because they cannot survive long-term storage or subzero temperatures. The basis for the desiccation tolerance/sensitivity in seeds is associated with the overlap between embryo development and germination metabolic programs, and this can lead to the appearance of a continuous range of seed physiologies (or tolerances to drying and temperature) rather than fixed categories (Berjak and Pammenter, Reference Berjak and Pammenter1994; Pammenter et al., Reference Pammenter, Naidoo, Berjak, Nicolás, Bradford, Côme and Pritchard2003; Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008).
Seeds from guaba (Inga vera) are highly recalcitrant and, like Castanospermum australe, represent a rare example of recalcitrant-seeded species within Fabaceae (Marques et al., Reference Marques2019). In this species, drying to water contents (WCs) below 35% (wet basis) is lethal (Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008), as is storage for 15 d or more at room temperature (Carvalho, Reference Carvalho1994; Bilia and Barbedo, Reference Bilia and Barbedo1997). The high metabolic activity might explain the requirement for water and short life span (Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008). Inga trees are used for food, firewood, land reclamation (Pritchard et al., Reference Pritchard, Haye, Wright and Steadman1995) and medicinal purposes (Stein et al., Reference Stein, Paiva, Vargas, Soares, Alves and Nogueira2010). Guaba is one of the most frequently planted tree species for ecological restoration of riparian forests (Caccere et al., Reference Caccere, Teixeira, Centeno, Figueiredo-Ribeiro and Braga2013; Cruz Neto et al., Reference Cruz Neto, Machado, Galetto and Lopes2015). Hence, the recalcitrance of the seeds has important economic and ecological implications.
Since lower, non-lethal water potentials slow metabolic activity (Vertucci, Reference Vertucci1989), we hypothesized that exposure of Inga seeds to osmotic media might extend shelf life (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006; Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006). Polyethylene glycol (PEG) (molecular weight = 8000) is an inert polymer that does not penetrate the cells and controls water movement between the seeds and the medium. This study aimed to find a water potential ×temperature combination that prolonged the survival of guaba seeds. In this paper, we compared the shelf life of these seeds using a commonly used practice of refrigerator storage and new methods that adjust the osmotic potential during storage in PEG solutions.
Materials and methods
Plant material
Mature fruits were collected in Ribeirao Vermelho, MG, Brazil, and taken to the Tree Seed Laboratory at the Federal University of Lavras for seed processing according to Faria et al. (Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006). Seeds were extracted from fruits manually and washed under the tap to remove sarcotesta and the extremely thin seed coat leaving the bare embryo. For the sake of simplicity and to avoid confusion with the polyembryonic condition often observed in this species, hereafter embryos will be referred to as seeds.
Water content determination
Seed WC was assessed by comparing the fresh and dry mass of seeds, with dry mass measured after heating seeds in an oven (105 ± 3°C) for 24 h according to Brasil (2009). The results are expressed as a percentage of the wet (i.e. fresh) mass basis.
Seed storage
Two methods of storage were tested: (1) cool storage – seeds stored in plastic bags in a cool chamber at 8°C; (2) osmotic storage – seeds stored in PEG solutions. For osmotic storage, seeds were placed in plastic trays and covered to a 0.5 cm depth with solutions of PEG in water to give water potentials of −1.6 and −2.4 MPa (following the protocol by Andreo et al., Reference Andreo, Nakagawa and Barbedo2006). The trays were covered with plastic film to decrease water evaporation. The solution was replaced to maintain a 0.5 cm level or when it was noticeably darkening. Trays were kept in an incubator at 10 ± 1°C in the dark, following the protocol by Andreo et al. (Reference Andreo, Nakagawa and Barbedo2006). Samples from both treatments were assessed for germination and WC almost monthly for a year or until germination percentage decreased to 0%.
Germination assay
Germination tests of fresh and stored seeds used 4 replications of 25 seeds per treatment, that were rolled in wet germitest paper (JProlab, Sao José dos Pinhas, Brazil) and kept in an incubator at 30 ± 1°C (Eletrolab – EL 212/4), under constant light, according to Andreo et al. (Reference Andreo, Nakagawa and Barbedo2006). In the case of seeds with more than one embryo, germination was considered completed when root protrusion was observed for at least one embryo. A normal seedling exhibited shoot and root, with normal aspect. The final counting of germination was done 14 d after seeds were sown on the paper (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006).
Optical microscopy
Embryonic axes were taken from each of the three sets of storage conditions about every 30 d and cut manually with a blade for microscopic analysis. The axes were stored in Karnovsky solution (Karnovsky, Reference Karnovsky1961) until the microscopic analysis procedures were performed. Lugol solution (Kraus and Arduin, Reference Kraus and Arduin1997) was used as an indicator test for the presence of starch (Ventrella and Almeida, Reference Ventrella and Almeida2013). The specimens were observed and photographed using an Olympus CX41 microscope coupled with a digital camera Belcam DIV-3000.
Data analysis
The experimental design was completely randomized with two factors: storage conditions (8°C in plastic bags and PEG solutions at −1.6 MPa and −2.4 MPa) and storage times (ranging from 0 to 330 d). Statistical analyses used R software (R Core Team, 2013) and were graphed using SigmaPlot. Data were analysed using general linear models with binomial error distributions. The significance of the treatments was observed using the χ 2 test, and averages were compared with LSD test at 95% of confidence. WC, germination and frequency of normal seedlings with storage time and storage condition were correlated using Spearman's correlation test.
Results
Water content during storage
Freshly harvested seeds were pale green, as described by Pritchard et al. (Reference Pritchard, Haye, Wright and Steadman1995), with 68% WC. WC did not change in seeds stored at 8°C in plastic bags (treatment 1) throughout the experiment, which ended on day 140, when seeds were dead (Fig. 1A). At 100 d, WC of seeds stored with no osmotic treatment (8°C in a plastic bag) and at −1.6 MPa PEG solution (10°C) were statistically similar (66 and 70%, respectively), the latter appearing to initially fluctuate. WC of seeds stored at −1.6 MPa declined to 62% near the 200th day of storage and thereafter remained unchanged until day 330 (Fig. 1A). Seeds stored in −2.4 MPa PEG had lower WC (53%) since the d 30, which did not change throughout the experiment, which ended on day 200 when all seeds were dead.
Fig. 1. Effect of storage time of guaba seeds on (A) WC (%, wet basis); (B) germination percentage (seed was considered germinated when a root protruded from at least one embryo); (C) total number of seedlings produced by 100 seeds; (D) number of normal seedlings produced by 100 seeds. Capital letters compare storage conditions within a given storage time and small letters compare storage times within storage condition, according to Fisher (LSD) with 5% of confidence.
Germination after storage
Fresh seeds were highly viable and usually, 100% germinated (radicle protrusion) within 2 d (data not shown). Germination percentages did not decline after 30 d of storage and were similar for the different storage conditions (93, 99 and 98% at 8°C, −1.6 and −2.4 MPa, respectively). Major changes in seed quality were visible in the 8°C stored seeds after 60 d: the seeds lost their greenish colour, smaller seeds turned black and bigger seeds turned pale pink; germination decreased to 14%. In contrast, germination of seeds stored in either PEG solutions was 99%, statistically equivalent to germination of fresh, non-stored seeds (Fig. 1B). Seeds stored at 8°C did not germinate after 90 d of storage, while those stored in PEG maintained 92% (−1.6 MPa) and 80% (−2.4 MPa). After 140 d of storage, germination of seeds stored in PEG at −2.4 MPa decreased to 41%, while germination of the −1.6 MPa treatment declined slightly to 90%, which was not statistically different from fresh seeds at the 0.05 level. After 200 d of storage, all seeds stored in PEG −2.4 MPa were dead, while 88% of those stored in −1.6 MPa germinated (not different statistically from fresh seeds at 0.05 level). The smaller seeds turned black and the bigger ones stayed pale green. At 260, 300 and 330 d of storage, 45, 24 and 0%, respectively, of the seeds stored in PEG −1.6 MPa germinated (Fig. 1B).
Guaba seeds are polyembryonic, and in our study 100 freshly harvested seeds yielded 225 seedlings (Fig 1C) of which 178 were normal, producing both roots and shoots (Fig. 1D). Both the total number and the number of normal seedlings produced from 100 seeds declined with storage time, with fastest decline observed in the 8°C (no osmotic) treatment (Fig. 1C, D). A decline in the number of seedlings occurred at similar rates in the −1.6 and −2.4 MPa treatments until about day 60 when the −1.6 MPa treatment supported higher incidence of polyembryony. By day 200, only one normal seedling was produced per seed (on average).
Correlations among germination metrics and storage conditions
Table 1 shows Spearman's correlation coefficients for response variables germination, total seedlings, normal seedlings, WC and storage conditions (8°C, −1.6 and −2.4 MPa in this order) and storage period (0, 30, 60, 90, 140, 200, 260, 300 and 330 d). Storage duration was the most significant factor affecting germination, total seedlings and normal seedlings (correlations of −63, −72 and −68%, respectively). WC positively correlated with germination, total and normal seedlings (33, 36 and 31%), meaning that higher germination performance occurs in seeds stored at higher WC.
Table 1. Spearman's correlation (level of significance) between storage condition, WC, storage time, germination, total seedlings and normal seedlings
Food reserves
Micrographs show that fresh seeds have high starch content (Fig. 2), characterizing such seeds as amylaceous. As the storage time progresses, the starch content decreases, consistent with previous observations of guaba seeds (Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006).
Fig. 2. Micrographs of guaba embryonic axes stained with Lugol's solution for starch. (A) axis from fresh seeds; (B) axis from seeds stored for 280 days in PEG −1.6 MPa. Arrows indicate starch granules. Bars = 10 μm.
Discussion
Longer shelf life of seeds requires reduced metabolic activity of both plant and fungal cells (Parisi et al., Reference Parisi, Biagi, Barbedo and Medina2013) to reduce the rate of deterioration (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006; Cardoso et al., Reference Cardoso, Binotti and Cardoso2012). Metabolic activity decreases with decreasing water potential and temperature (Walters et al., Reference Walters, Pammenter, Berjak and Crane2001) and greater storage potential can be achieved if these stresses do not damage seed cells. It is not possible to dry desiccation sensitive seed to very low levels, since this will damage or even kill them. Guaba seeds do not survive WCs below 35% (Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008). Hence, optimizing water supply using an osmotic medium, like solutions of PEG, is an interesting strategy because it may avoid big changes in WC of the seeds that could lead to deterioration, germination or death.
The WC of fresh guaba seeds was 68%, similar to that reported by Andreo et al. (Reference Andreo, Nakagawa and Barbedo2006) and Bonjovani and Barbedo (Reference Bonjovani and Barbedo2008) and also similar to seeds stored at −1.6 MPa (Fig. 1A). Neither fungal proliferation nor initiation of germination was observed during storage, although these are frequent problems in hydrated storage of recalcitrant seeds, according to Berjak and Pammenter (Reference Berjak and Pammenter2013). The high moisture levels used in our study remained slightly below the threshold for germination (Bradford, Reference Bradford, Kigel and Galili1995).
Lowering the moisture to 53% at −2.4 MPa did not appear to initially damage seeds (99% germinated after 60 d of storage), but did appear to cause more rapid deterioration than seeds stored at the higher water potential. Shifts in the longevity of seeds in hydrated storage have been previously noted (Ibrahim et al., Reference Ibrahim, Roberts and Murdoch1983; Walters et al, Reference Walters, Pammenter, Berjak and Crane2001), and were interpreted as hydration-induced changes in metabolism and structure of cells (Vertucci and Farrant, Reference Vertucci, Farrant, Kigel and Galili1995). The active metabolism is probably responsible for death due to shortage of food reserves (Fig. 2) or oxidative activity (Leprince et al., Reference Leprince, Buitink and Hoekstra1999; Roach et al., Reference Roach, Beckett, Minibayeva, Colville, Whitaker, Chen, Bailly and Kranner2010), altering the energy balance within cells (Pammenter and Berjak, Reference Pammenter and Berjak1994; Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006). At water potentials between 1.5 and −4.0 MPa, macromolecular surfaces are hydrated and water occupies spaces among the macromolecules (Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008) and contributes to the osmotic responsiveness (cell swelling and contractions). According to Marcos Filho (Reference Marcos Filho2005), water potentials that avoid stripping water from macromolecular surfaces will favour the conservation of seeds that are highly sensitive to desiccation.
Results showed that storage at −1.6 MPa and 10 ± 1°C extended survival times of guaba seeds compared to storage at lower or higher water potentials (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006; Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006). Successful storage of guaba seeds for 200 d (Fig. 1) is the longest storage period reported for this species, as far as we know. Other studies showed a faster decline in survival or did not test the duration that survival could be maintained: 50% of germination after 10 days of storage in glass containers covered with Parafilm at 5°C (Faria et al., Reference Faria, van Lammeren and Hilhorst2004); 94% after 30 days in PEG solution (−1.7 MPa) at 5°C (Faria et al., Reference Faria, Davide, Silva, Davide, Pereira, van Lammeren and Hilhorst2006); 75% after 90 days in PEG solution (−1.6 and −2.4 MPa) at 10°C (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006); 85% after 45 days in plastic bags at 8°C (Bonjovani and Barbedo, Reference Bonjovani and Barbedo2008); 100% of germination after 65 days in plastic bags at 7°C (Parisi et al., Reference Parisi, Biagi, Barbedo and Medina2013). Additional studies to elucidate the relationship between the mobilization of food reserves and viability lost during seed storage will contribute to the understanding of the interaction between metabolism and water potential as well as inform about supplementing reserves such as with nutrient-enriched media.
Methods that prolong the shelf life of guaba seeds is a significant accomplishment for seedling production in nurseries. Guaba is a key species for the restoration of deforested areas in Southeast Brazil, and sufficient seedlings at the right developmental stage are needed for planting during the wet season that extends from October to March. The phenology of the plant currently makes it difficult for restoration work because the fruits ripen in December through February (Andreo et al., Reference Andreo, Nakagawa and Barbedo2006; Stein and Fu, Reference Stein and Fu2007) and 4–6 months are needed to produced seedlings for planting (Davide and Faria, Reference Davide, Faria, Davide and Silva2008). Hence, seedlings are ready for planting in the dry season (April–August), when field survival tends to be much lower compared to seedlings planted during the rainy season. Maintaining seedlings longer in nurseries until the wet season increases the costs of production and decreases seedling quality by interfering with the proper balance between the root and shoot. Thus, extending the shelf life of guaba seeds to 6 months bridges the time to the wet season and optimizes seedling production in tree nurseries by sowing the seeds closer to the wet (planting) season.
Conclusion
Guaba seeds are highly recalcitrant and usually survive <2 months in storage. Here we show that storing seeds in a PEG solution at −1.6 MPa maintains germination percentage statistically equivalent to fresh seeds for up to 200 d in storage. We believe this water potential is low enough to prevent seeds from completing germination but high enough to prevent stress-induced metabolism that damages the cells. The extension of the shelf life, the longest reported for the species as far as we know, is useful for ecological restoration projects.
Financial support
A.C.J. and J.M.R.F. thank Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) for the research productivity fellowships (Processes 310976/2018-9 and 311556/2018-3, respectively). The other authors thank the following agencies for the scholarships: Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq (L.C.V.P.); Agricultural Greenhouse Gases Program under the Agriculture and Agri-Food Canada (R.C.M.); Fundação de Amparo à Pesquisa do Estado de Minas Gerais – Fapemig (C.R.Z.).
