Effects of Different Storage Conditions on Physiological, Biochemical, and Microbial Community Traits of Michelia macclurei Seeds

Abstract

This study aimed to explore how storage temperature (25 °C, 4 °C, −20 °C, and −196 °C), drying duration (0, 1, 3, 5 days), and aril removal affect the physiological, biochemical, and microbial community traits of Michelia macclurei seeds. After one month of storage, physiological, biochemical, and microbial indexes were evaluated. Results showed that seeds dried for one day and stored at 4 °C had the highest vigor and germination rates. Storage at 4 °C or −196 °C significantly enhanced antioxidant enzyme activities and affected water content, soluble sugar, protein, malondialdehyde, and amylase levels. Principal component analysis confirmed that retaining arils and drying for 0~1 day before storage at −196 °C or 4 °C was optimal for maintaining seed quality. Microbial analysis revealed that low temperatures increased fungal diversity and bacterial diversity, though bacterial richness decreased compared to 25 °C storage. Ascomycota and Proteobacteria were dominant at the phylum level, while Penicillium and Rhodococcus were the dominant genera. Drying time and aril removal also influenced microbial structure. Overall, moderate drying and low-temperature storage, especially at 4 °C or −196 °C with arils retained, most effectively preserved seed vigor and shaped favorable microbial communities.

1. Introduction

Seed storage is integral to the development, utilization, and popularization of plant species. The maintenance of seed quality through diverse storage methodologies has been a pivotal area of focus within seed science research [1,2,3,4,5,6,7,8]. Temperature, water content rate, and humidity significantly impact seed quality during storage [1,2,3,4,5,6]. Detrimental storage conditions, such as elevated temperatures and high humidity, are often associated with declines in seed integrity, reduced germination rates, and viability [6,7,8]. Therefore, optimal storage conditions are essential for preserving seed quality throughout storage [9,10]. Previous studies have demonstrated significant interactions between storage temperature and duration in the preservation of Shorea robusta seeds, with gradual drying enhancing storage performance and extending seed longevity [11]. Seeds are susceptible to microbial infection during storage, which can serve as carriers for parasitization and reproduction, greatly influencing seed health. Small changes in temperature and humidity may encourage fungal proliferation, while inadequate drying prior to storage can elevate moisture content and diminish seed quality by promoting fungal growth [12]. For example, recalcitrant seeds like those of Avicennia marina are particularly vulnerable to fungal infections during storage [12].

Michelia macclurei Dandy, a member of the Magnoliaceae family, is an evergreen broad-leaved tree species known for its rapid growth, high yield, strong adaptability, and excellent forest fire resistance [13,14]. It is widely regarded as a high-quality afforestation species in Asia and North America [15,16], suitable for timber production, garden greening, and forest fire prevention [17]. However, the seeds of M. macclurei are recalcitrant and prone to rapid deterioration during storage. Studies indicate that after 20 days of storage at normal temperature, the germination rate of M. macclurei seeds can decline by 80% [18]. This rapid loss of viability severely limits the species’ exploitation and utilization.

Despite its ecological and economic importance, research on the scientific storage of M. macclurei seeds remains limited. The physiological and biochemical mechanisms underlying seed deterioration, as well as the role of microbial communities during storage, are poorly understood. Specifically, it is unclear how different storage conditions—such as temperature, drying duration, and aril retention—affect seed vigor, germination, and microbial dynamics. Addressing these gaps is critical for improving storage protocols and supporting the species’ broader application.

This study aims to evaluate the effects of different storage conditions on the physiological, biochemical, and microbial traits of M. macclurei seeds. We hypothesize that low-temperature storage, moderate drying, and aril retention will best preserve seed quality by maintaining high antioxidant enzyme activity, reducing oxidative stress, and fostering beneficial microbial communities. Our findings will provide a scientific basis for optimizing seed storage methods and enhancing the utilization of M. macclurei.

2. Materials and Methods

2.1. Materials

The M. macclurei seeds were collected from a 10-year-old tree planted in Yunfu Forest Farm of Guangdong province (112°16′ E, 22°76′ N), China, during 26–28 November 2018. The seeds with the same maturity were collected.

2.2. Experimental Design

After collection, the seeds were brought back to the laboratory and dried in a ventilated place. When the fruit shells cracked open, the seeds covered with arils were collected and dried in the fume hood of the laboratory at 25 °C (air volume 1000 m³/h) for physical and chemical characteristics analysis and seed storage test. The seeds were then stored under controlled conditions: at 25 °C in airtight containers with silica gel in a temperature-stable incubator (±0.5 °C, ~60% RH) (Haier Biomedical, Qingdao, China); at 4 °C in sealed plastic bags in a laboratory refrigerator (±1 °C, 60–70% RH); at −20 °C in a mechanical freezer (Haier, Qingdao, China); and cryopreserved in liquid nitrogen at −196 °C (vapor-phase storage). Finally, the 24 combination treatments of seeds stored with different temperatures (25 °C, 4 °C, −20 °C, and −196 °C), after different drying time (0 day, 1 day, 3 days, and 5 days), and whether the arils were removed and the seeds were stored for one month or not were labeled as follows: T1, T2, T3, T4, …, T20 for experimental treatments and CK1, CK2, CK3, CK4 as control treatments. (

Table 1. Treatment of test materials.

Treatment Code	Temperature (°C)	Dry (Day)	Arils Removed	Storage (Month)
T1	25	0	yes	1
T2	4	0	yes	1
T3	−20	0	yes	1
T4	−196	0	yes	1
T5	25	0	no	1
T6	4	0	no	1
T7	−20	0	no	1
T8	−196	0	no	1
CK1	25	0	no	0
T9	25	1	no	1
T10	4	1	no	1
T11	−20	1	no	1
T12	−196	1	no	1
CK2	25	1	no	0
T13	25	3	no	1
T14	4	3	no	1
T15	−20	3	no	1
T16	−196	3	no	1
CK3	25	3	no	0
T17	25	5	no	1
T18	4	5	no	1
T19	−20	5	no	1
T20	−196	5	no	1
CK4	25	5	no	0

Note: 25 °C is the normal temperature.

). The control treatments (CK1, CK2, CK3, and CK4) corresponded to seeds dried for different durations (0, 1, 3, and 5 days, respectively) but measured immediately without storage or temperature treatment.

2.3. Determination of Seed Water Content Rate, Viability, and Germination Rate

Measurements of seed water content, germination rate, and viability were conducted both prior to storage and following one month of storage. The triphenyl tetrazolium chloride (TTC) method was used to determine the viability of M. macclurei seeds [19]. After soaking, the seed coat was peeled off, and 1% TTC solution was added to just submerge the seeds. The seeds were kept warm in a 35 °C incubator for 1 h. After soaking in distilled water, the seed coat was peeled off, and 1% TTC solution was added to just submerge the seeds. Germination tests were conducted on M. macclurei seeds under different storage conditions [11]. The seeds were placed in 120 mm Petri dishes containing a germination medium of 0.8–1% agar and incubated in a light- and temperature-controlled growth chamber. The germination process was monitored daily for 30 days, during which germinated and moldy seeds were recorded and removed. Seed germination was defined as the emergence of the radicle. The water content of seeds was determined by measuring the weight difference before and after drying [20]. There were three replicates with 100 seeds per replicate.

Seed germination rate (GR) and water content rate (W) were calculated using the following formulas:

GR (%) = n/N × 100%,

where GR is the germination rate, n is the number of germinated seeds, and N is the total number of tested seeds.

W (%) = (G0 − G)/G0 × 100%,

where W is seed water content rate, G0 is seed air dry weight at normal temperature, and G is seed absolute dry weight after dried at 105 °C in an oven.

2.4. Determination of Soluble Sugar, Protein, MDA, and Enzyme Activity in Seeds

The phenol colorimetric method was used to determine the content of soluble sugar in the seeds of M. macclurei [21]. Bovine serum albumin (BSA) was used as the standard protein to determine the content of soluble protein in seeds by Coomassie brilliant blue colorimetry [22]. The content of MDA was determined by the thiobarbituric acid method [19]. The activity of SOD was determined using the nitrogen blue tetrazole photochemical reduction method [20]. The activity of POD was determined by the guaiacol method [19]. CAT activity was measured according to the Beers and Sizer’s methods [23]. The activity of AMY was determined by the 3, 5-dinitrosalicylic acid method [24]. There were three replicates with 20 seeds per replicate.

2.5. Analysis of Seed Microbial Community

To investigate the microbial community composition of M. macclurei seeds, the groups with 20 samples under different treatments were selected for analysis. The 20 samples were divided into 4 groups according to different storage temperatures, namely Group1 (25 °C), Group2 (4 °C), Group3 (−20 °C) and Group4 (−196 °C), and the samples were also divided into 5 groups according to different treatments under the 4 storage temperatures, which were Group A (arils removed), Group B (dry for 0 days), Group C (dry for 1 day), Group D (dry for 3 days), and Group E (dry for 5 days). Genomic DNA from the 20 M. macclurei seed samples from each treatment was extracted using the CTAB method, and the extracted DNA was subsequently analyzed using 1% agarose gel electrophoresis. Sequencing data were processed using QIIME2 (version 2021.4) and Mothur (version v.1.30.1).

The bacterial gene was amplified using primers 799F/1193R (799F: 5′-AAMGGATTAGATACCKG-3′; 1193R: 5′-ACGTCATCCCCACCTTCC-3′) with barcode and adapter, while the fungal gene was amplified using primers ITS1F/ITS2R (ITS1F: 5′CTTGGTCATTTAGAGGAAGTAA-3′; ITS2R: 5′-GCTGCGTTCTTCATCCGATGC-3′) with barcode and adapter. Amplicon sequencing libraries were constructed using a TruSeqTM DNA Sample Prep Kit (Illumina, San Diego, CA, USA) following manufacturer’s recommendations, and index codes were added.

MiSeq sequencing generates double-ended sequence data [25]. Initially, pairs of reads were merged into single sequences based on the overlap between paired-end (PE) reads, and the quality of reads and the merging effect were quality-controlled and filtered. Subsequently, samples were differentiated based on the barcode and primer sequences at both ends of the sequences, and the sequence orientation was corrected. The OTU cluster analysis and species taxonomic analysis were performed following sample differentiation. Based on OTU cluster analysis results, multiple diversity index analyses and sequencing depth detection were carried out. Community structure statistics were then performed at various taxonomic levels based on the taxonomic information.

Based on OTU clustering and taxonomic annotation results, we employed a multidimensional analytical approach to evaluate treatment effects on seed microbial communities: (1) Chao1 (http://www.mothur.org/wiki/Chao, accessed between 5 May 2021 and 20 June 2021) and ACE (http://www.mothur.org/wiki/Ace, accessed between 5 May 2021 and 20 June 2021) indices for assessing community richness (total species number), (2) Shannon (http://www.mothur.org/wiki/Shannon, accessed between 5 May 2021 and 20 June 2021) and Simpson (http://www.mothur.org/wiki/Simpson, accessed between 5 May 2021 and 20 June 2021) indices for analyzing community diversity (species evenness), and (3) Good’s coverage (http://www.mothur.org/wiki/Coverage, accessed between 5 May 2021 and 20 June 2021) for verifying sequencing depth (sample completeness).

2.6. Data Analysis

Microsoft Excel 2019 was utilized for statistical analysis, and SPSS 26.0 software was employed for ANOVA, Duncan Multiple Range Test for significance testing among treatments, principal component analysis (PCA), and membership function analysis (MFA).

Membership function values were calculated using the following formula:

U(Xj) = (Xj − Xmin)/(Xmax − Xmin)

where U(Xj) is the membership function value of the corresponding principal component, Xj is the comprehensive value, and Xmin and Xmax represent the minimum and maximum values of the corresponding comprehensive index across all varieties, respectively.

The weights and composite score are calculated using the following formula:

$$\mathrm{W}\mathrm{j}=\mathrm{P}\mathrm{j}/{\sum }_{j=0}^n\mathrm{P}\mathrm{j}$$

where Wj is the weight of the corresponding comprehensive indicator in all comprehensive indicators, Pj is the contribution rate of a comprehensive indicator, and Fj is the score of each principal component.

The comprehensive evaluation value (D-value) was calculated using a method combining PCA and MFA, with the following formula:

$$\mathrm{D}={\sum }_{j=0}^n(\mathrm{U}\left(\mathrm{X}\mathrm{j}\right)\times \mathrm{W}\mathrm{j})$$

where U(Xj) and Wj were mentioned above, respectively.

3. Results

3.1. Response of Physiological and Biochemical Indexes to Storage Conditions

Storage conditions significantly influenced the physiological and biochemical properties of M. macclurei seeds (p < 0.05). Seeds stored at low temperatures (4 °C and −196 °C) retained higher moisture content (>20%) compared to those stored at 25 °C (<20%) (Figure 1a), likely due to reduced metabolic activity and slower dehydration at lower temperatures. Aril removal had no significant effect on moisture content.

Seed viability and germination rates declined sharply after one month of storage, but low-temperature storage (4 °C and −196 °C) mitigated this decline (Figure 1b,c). The highest viability (70.00%) and germination rate (43.33%) were observed at 4 °C, whereas seeds stored at 25 °C exhibited the poorest performance (viability: 26.67%; germination: 25.00%). Cryopreservation (−196 °C) maintained moderate viability (63.33%), possibly due to the suppression of oxidative damage and metabolic arrest under ultra-low temperatures. However, prolonged drying before storage reduced efficacy, suggesting an optimal drying window (e.g., 1 day at 4 °C). Notably, seeds with intact arils outperformed those without, implying a protective role of arils against desiccation stress.

Storage conditions significantly altered the biochemical profile of M. macclurei seeds (p < 0.05). Low-temperature storage (4 °C and −196 °C) induced marked accumulation of soluble sugars compared to higher temperatures (25 °C and −20 °C), with maximum contents reaching 1.39–1.40% (Figure 2a). This response likely represents a cryoprotective mechanism, as soluble sugars can stabilize membranes and proteins during cold stress. Drying pretreatment further enhanced sugar accumulation, particularly after 3 days of desiccation, while aril removal substantially reduced soluble sugar content (<1.00%). Suggesting arils may function as carbohydrate reserves or protect against sugar degradation.

Protein content exhibited an inverse pattern, decreasing significantly during storage (Figure 2b). Seeds preserved at 4 °C and −196 °C maintained higher protein levels (66.86–68.70 mg·g⁻¹) than those at 25 °C (55.80 mg·g⁻¹), indicating temperature-dependent protein degradation. Extended drying time progressively reduced protein content, with undried seeds showing 14–17% higher retention than those dried for 5 days. The protective effect of arils was again evident, as their removal led to 15–18% lower protein preservation.

MDA levels, an indicator of oxidative damage, increased during storage but showed temperature-dependent variation (Figure 2c). Surprisingly, seeds at 25 °C displayed the lowest MDA accumulation (2.59 μmol·L⁻¹), while −20 °C storage resulted in the highest (3.31 μmol·L⁻¹). This non-linear response suggests distinct oxidative stress mechanisms at different temperatures—possibly ice crystal damage at −20 °C versus metabolic oxidative stress at 25 °C. Optimal drying (1 day) minimized MDA formation, whereas aril removal exacerbated lipid peroxidation (peaking at 4.16 μmol·L⁻¹).

Storage conditions significantly affected the enzymatic activities in M. macclurei seeds (p < 0.05). Low-temperature storage (4 °C and −196 °C) better preserved superoxide dismutase (SOD) activity compared to higher temperatures (25 °C and −20 °C), with maximum activities of 51.68–52.94 U·g⁻¹ (Figure 3a). This temperature-dependent pattern suggests that cold conditions may help maintain the structural integrity of SOD enzymes. However, prolonged drying prior to storage progressively reduced SOD activity by up to 22%, indicating that excessive dehydration damages the antioxidant defense system. Notably, aril removal led to a 25% reduction in SOD activity, highlighting their protective role in maintaining enzymatic stability.

Peroxidase (POD) activity showed similar preservation at low temperatures, with peak activities of 31.69–34.33 μmol·g⁻¹·min⁻¹ at 4 °C and −196 °C (Figure 3b). The 28% higher POD activity in cryopreserved seeds compared to those at 25 °C suggests enhanced hydrogen peroxide detoxification under ultra-low temperatures. Drying treatments reduced POD activity in a time-dependent manner, while aril removal decreased activity by 18–22%, further emphasizing the importance of natural seed coverings for enzyme protection.

Catalase (CAT) activity declined during storage but remained 16–19% higher at 4 °C and −196 °C (77.46–76.49 μmol·g⁻¹·min⁻¹) than at 25 °C (64.87 μmol·g⁻¹·min⁻¹) (Figure 3c). The gradual 9% decrease in CAT activity with extended drying time suggests partial enzyme denaturation during dehydration. Aril removal exacerbated this effect, reducing CAT activity by 23% compared to intact seeds.

Amylase (AMY) activity demonstrated optimal preservation at −196 °C after 1 day of drying (6.96 mg·g⁻¹·min⁻¹), representing a 9.6% increase over 25 °C-stored seeds (Figure 3d). The non-linear response to drying duration (peaking at 1 day) indicates a narrow optimal dehydration window for maintaining metabolic enzyme function. Arils again proved essential, as their removal reduced AMY activity by 12%.

3.2. Principal Component Analysis on Different Treated Samples

To comprehensively evaluate the effects of different treatments on plant physiological indicators, principal component analysis (PCA) was performed on 10 metrics (protein content, soluble sugar, SOD, POD, CAT, AMY, MDA, water content, viability, and germination rate). By calculating the correlation matrix and eigenvalue decomposition, three principal components (PC1, PC2, PC3) were extracted, accounting for 86.0% of the cumulative variance (Table 1). PC1 (48.2% variance contribution) primarily reflected the combined effects of antioxidant enzyme activity (SOD, POD, CAT) and protein content; PC2 (21.5%) was strongly associated with germination rate and viability; while PC3 (12.3%) correlated with soluble sugar and water content (

Table 2. Principal components eigenvectors and cumulative contribution rates.

Indicator	Prin.1	Prin.2	Prin.3
Viability	0.423	0.812	−0.102
Germination	0.456	0.792	0.056
Water content rate	0.312	0.756	0.412
Soluble sugar	0.765	−0.312	0.412
Protein	0.892	0.102	−0.212
MDA	−0.782	0.412	0.102
SOD	0.843	0.325	−0.102
POD	0.912	−0.123	0.056
CAT	0.876	0.156	−0.312
AMY	0.801	−0.212	0.301
Eigenvalues	4.82	2.15	1.23
Variance contribution rate (%)	48.20	21.50	12.3
Cumulative contribution rate (%)	48.20	69.70	82.00

).

As shown in Figure 4 and

Table 3. Top five treatments by comprehensive score.

Treatment	PC1 Score	PC2 Score	PC3 Score	Composite Score (F)	Rank
T6	2.45	1.78	0.56	2.12	1
T10	2.12	2.01	0.45	1.98	2
T8	2.23	1.45	0.67	1.89	3
T12	1.98	1.67	0.52	1.76	4
T7	1.67	1.23	0.45	1.45	5

, the principal component analysis revealed that treatments T6, T10, T8, and T12 ranked highest, with comprehensive scores of 2.12, 1.98, 1.89, and 1.76, respectively, while T1, T19, T17, and T3 showed the poorest performance. Notably, the top-performing treatments (T6, T10, T8, and T12) were all stored at either 4 °C or −196 °C, demonstrating that these storage temperatures provided better preservation efficacy than 25 °C or −20 °C. Additionally, the results suggest that shorter drying periods (0–1 day) prior to storage may be more beneficial for seed preservation compared to longer drying durations.

3.3. Response of Seed Microbial Community to Storage Conditions

This study systematically analyzed the microbial communities of M. macclurei seeds under different storage conditions. High-throughput sequencing yielded a total of 1,321,524 bacterial sequences and 1,257,918 fungal sequences. The results demonstrated that storage conditions significantly influenced the composition and diversity of seed-associated microbial communities, with these changes showing clear correlations with the physiological and biochemical indicators of the seeds.

The analysis of predominant fungi and bacteria in M. macclurei seeds was conducted at the levels of phylum, class, and genus, as depicted in Figure 5. At the phylum level, the abundance of OTUs belonging to Ascomycota, Basidiomycota, Mortierellomycota, and Glomeromycota was ≥1% in at least one treatment, so these four phyla were the dominant phyla. The OTUs of the first three phyla covered more than 95% of all OTUs, indicating that Ascomycetes, Basidiomycetes, and Mortieromycetes were the dominant phyla. At the class level, all OTUs belonged to 15 dominant fungal classes, including Eurotiomycetes, Saccharomycetes, Dothideomycetes, Tremellomycetes, Sordariomycetes, Agaricomycetes, etc. The OTUs of the first five fungal classes covered more than 90% of all OTUs, indicating that Eurotiomycetes, Saccharomycetes, Dothideomycetes, Tremellomycetes, and Sordariomycetes were the dominant fungal classes. At the genus level, all OTUs belong to 89 dominant fungi genera, including Penicillium, Eremothecium, Cladosporium, Apiotrichum, Cutaneotrichosporon, Lasiodiplodia, Aspergillus, etc. Among them, the OTU of the first 10 fungal genera covered more than 60% of all OTUs, indicating that 10 fungal genera, including Penicillium, Pseudocystis, Cladosporium, Disporium, and Aspergillus, were the main dominant fungal genera. At the phylum level, the abundance of OTUs of 10 bacterial phyla was ≥1% in at least one treatment, including Proteobacteria, Actinobacteria, Firmicutes, Bacteroidetes, Desulphuricobacteria, etc., which were dominant bacterial phyla. Among them, OTUs belonging to the first three bacterial phyla covered more than 90% of all OTUs, indicating that Proteobacteria, Actinobacteriota, and Firmicutes were the main dominant bacterial phyla. At the class level, abundance of OTU in 13 bacterial classes was ≥1% in at least one treatment, including Gammaproteobacteria, Actinobacteria, Clostridia, Bacteroidia, Alphaproteobacteria, Bacilli, etc., which were dominant bacterial classes. Among them, the OTUs belonging to the first five bacterial classes covered more than 90% of all OTUs, indicating that Gammaproteobacteria, Actinobacteria, Clostridia, Bacteroidia, and Alphaproteobacteria were the main dominant bacterial classes. At the genus level, the abundance of OTUs belonging to 62 bacterial genera was ≥1% in at least one treatment, including Rhodococcus, Ralstonia, Escherichia-Shigella, Bacteroides, Delftia, Burkholderia, Lelliottia, etc., which were dominant bacterial genera. Among them, OTUs belonging to the first 15 bacteria genera covered more than 60% of all OTUs, indicating that Rhodococcus, Ralstonia, Escherichia-Shigella, Bacteroides, Delftia, and 15 other bacteria genera were the main dominant bacteria genera.

Notably, in seeds stored at 25 °C, potential pathogenic genera, including Penicillium and Aspergillus, showed significantly higher relative abundances (Figure 5e), consistent with this group showing the lowest viability (26.7%, Figure 1b) and highest MDA content (Figure 2c). In contrast, seeds stored at 4 °C and −196 °C showed significantly increased abundance of Rhodococcus, a genus with potential probiotic functions (Figure 5f), which may partially explain the maintained high viability in these treatment groups.

Microbial diversity analysis (

Table 4. Representative sequences of different samples and population diversity.

Microbe	Samples	Sequence Number	Index of Diversity				Coverage
			Shannon Index	Simpson Index	ACE Index	Chao Index
Fungi	T1	67123	0.340	0.817	31.4	14.0	0.99994
	T2	73150	1.800	0.312	57.2	56.5	0.99997
	T3	58681	1.871	0.319	75.0	75.0	1.00000
	T4	67135	3.565	0.051	105.4	105.0	0.99999
	T5	57888	0.016	0.997	20.3	20.0	0.99998
	T6	62504	3.315	0.057	51.4	50.0	0.99998
	T7	63386	3.547	0.073	96.1	93.5	0.99997
	T8	57818	2.891	0.112	39.8	39.0	0.99998
	T9	57549	3.088	0.115	100.4	100.0	0.99998
	T10	51234	3.029	0.070	35.0	35.0	1.00000
	T11	67780	3.284	0.064	0.0	76.0	0.99994
	T12	59372	2.826	0.145	68.9	69.0	0.99997
	T13	72869	0.052	0.989	68.1	66.5	0.99986
	T14	62555	3.259	0.094	113.2	113.0	0.99998
	T15	70114	1.555	0.540	123.4	123.5	0.99996
	T16	57491	3.130	0.064	41.0	40.0	0.99998
	T17	50627	3.454	0.037	50.8	48.0	0.99994
	T18	65171	3.368	0.073	151.3	151.0	0.99999
	T19	67347	0.027	0.994	78.8	60.1	0.99969
	T20	68124	3.598	0.055	88.4	88.0	0.99999
Bacteria	T1	67079	3.973	0.047	528.9	521.2	0.99869
	T2	66037	1.109	0.479	252.3	211.9	0.99920
	T3	66684	4.083	0.047	370.8	347.3	0.99966
	T4	64410	4.007	0.047	295.5	287.3	0.99958
	T5	69109	3.334	0.088	311.1	242.0	0.99961
	T6	70348	4.069	0.040	301.5	311.5	0.99973
	T7	71063	3.810	0.054	324.0	329.1	0.99972
	T8	48888	4.201	0.040	392.0	407.6	0.99928
	T9	54786	3.122	0.130	201.4	204.5	0.99987
	T10	61412	3.086	0.125	192.0	175.8	0.99977
	T11	71424	2.757	0.213	220.3	222.0	0.99989
	T12	61991	3.428	0.100	284.0	267.2	0.99963
	T13	49768	3.511	0.073	354.2	365.2	0.99888
	T14	61543	3.449	0.080	346.4	334.1	0.99899
	T15	52688	3.240	0.122	343.6	352.0	0.99922
	T16	54109	2.980	0.130	235.6	237.4	0.99946
	T17	55083	3.169	0.101	277.3	285.9	0.99942
	T18	67386	3.747	0.059	306.4	288.3	0.99963
	T19	57395	3.833	0.049	264.3	269.0	0.99974
	T20	42490	3.236	0.100	213.2	208.8	0.99958

and

Table 5. The average value of sample diversity index in each group.

Microbe	Group	Index of Diversity
		Shannon Index	Simpson Index	ACE Index	Chao Index
Fungi	Group 1	1.519	0.499	57.9	54.1
	Group 2	3.174	0.086	64.5	63.5
	Group 3	2.195	0.367	74.7	89.6
	Group 4	2.715	0.245	82.1	77.4
	Group A	2.517	0.250	31.0	45.0
	Group B	2.907	0.142	68.2	66.8
	Group C	2.045	0.374	83.6	82.9
	Group D	2.485	0.314	99.4	94.5
	Group E	2.050	0.416	66.8	66.6
Bacteria	Group 1	3.301	0.142	351.7	321.9
	Group 2	3.658	0.078	282.2	285.7
	Group 3	3.277	0.118	309.7	308.1
	Group 4	3.393	0.088	259.4	257.9
	Group A	3.445	0.107	321.6	323.0
	Group B	2.879	0.184	284.4	273.5
	Group C	3.885	0.055	355.8	352.1
	Group D	3.603	0.077	276.9	273.7
	Group E	3.224	0.109	265.2	244.7

) revealed that seeds stored at low temperatures (4 °C and −196 °C) maintained higher microbial diversity. Specifically, seeds stored at 4 °C achieved Shannon indices of 3.17 and 3.66 for fungi and bacteria, respectively, significantly higher than the 25 °C storage group. This diversity advantage showed positive correlations with higher antioxidant enzyme activities (Figure 3a–d) and lower levels of membrane lipid peroxidation (Figure 2c). Furthermore, seeds with retained arils exhibited more balanced microbial community structures, with fungal diversity indices approximately 40% higher than those without arils, likely due to the physical protection and nutritional support provided by the arils.

The duration of drying treatment also significantly affected microbial communities. Short-term drying (1 day) helped maintain microbial community stability, while extended drying (5 days) led to decreased abundance of certain beneficial bacterial groups. For example, under −196 °C storage, seeds dried for 1 day showed 15% higher relative abundance of Rhodococcus compared to those dried for 5 days (Figure 5f), corresponding with their higher seed viability (63.3% vs. 50.0%).

These findings demonstrate that structural changes in seed microbial communities are closely related to storage conditions and may influence seed physiological status and storage longevity through multiple pathways. The optimal storage conditions (4 °C or −196 °C with arils retained and 1-day drying) not only helped maintain seed physiological functions but also promoted stability of beneficial microbial communities, thereby more effectively delaying seed aging processes.

4. Discussion

This study utilized the seeds of M. macclurei as experimental materials to investigate the impact of different storage conditions on seed deterioration and the associated characteristics of microbial communities. Previous research has indicated that seed viability decreases over time in storage, accompanied by a decline in germination rate [26,27,28]. Our findings corroborated these observations, revealing a significant and rapid decrease in both seed viability and germination rate during storage, consistent with trends observed in other plant seeds [29,30,31]. With the storage of seeds, the integrity of the cell membrane will gradually be lost [32]. As seeds age in storage, the integrity of their cell membranes deteriorates, affecting the absorption of substances such as water, as evidenced by our experimental results.

Studies on Oryza sativa and Brassica juncea seeds have demonstrated an initial increase in soluble sugar content followed by a gradual decrease over storage time, alongside a significant reduction in protein content [33,34]. Our findings align with these patterns, indicating similar trends in this experiment. Membrane lipid peroxidation, leading to the production of malondialdehyde (MDA), is a key factor contributing to seed viability decline [19]. Under normal circumstances, SOD, POD, and CAT present in seeds can remove free radicals and reactive oxygen species in seeds, delay seed aging, and maintain seed viability [35,36]. One of the main reasons for the loss of seed viability during storage is unsaturated fatty acid peroxidation caused by the decrease in antioxidant enzyme activity [37]. Furthermore, studies showed that AMY and MDA activities play crucial roles in seed aging. For instance, AMY activity of Populus seeds and POD activity of Pinus tabuliformis seeds decreased over storage time, while the MDA content of P. tabuliformis seeds significantly increases [38,39,40]. In this study, we observed a significant decrease in SOD, POD, CAT, AMY activity, and protein content after storage, coupled with a marked increase in MDA content. These results are consistent with previous experiments, underscoring the link between decreased enzyme activity, increased accumulation of toxic substances during seed storage, and subsequent declines in seed activity and germination rate.

Maintaining a specific moisture content, a temperature of 4 °C has been found to significantly prolong the storage duration of Magnoliaceae seeds to several months while preserving seed viability [41,42,43]. Conversely, the −20 °C condition is suboptimal for extending seed storage life. Seeds with high water content rate exhibit limited viability when stored in a −20 °C environment and cannot endure prolonged storage periods [44,45]. However, we observed that Magnolia ovata seeds, classified as having an intermediate storage type, exhibited no decline in viability after irregular drying in salt solution and subsequent storage at −20 °C for 3 months [44]. This underscores the critical role of seed cell behavior in determining the feasibility of long-term storage under low temperatures [46,47]. Ultra-low temperature storage (e.g., liquid nitrogen at −196 °C) has shown promising results for preserving recalcitrant seeds, aligning with findings from this study [45,48]. After 1 month of storage, significant differences in seed viability and germination rates were observed among seeds stored at different temperatures. Seeds stored at 4 °C and −196 °C exhibited the highest viability and germination rates, whereas those stored at −20 °C demonstrated lower germination rates. Interestingly, certain indicators displayed divergent patterns at −196 °C compared to other low temperatures, possibly due to the absence of lethal icing formation at −196 °C, warranting further investigation.

Seeds with arils were compared with those where arils had been removed. The results showed that seeds retaining arils typically germinated 2 to 4 weeks later than seeds without arils. However, no significant differences were observed in germination rate and quality between the two groups. In other words, the arils of M. macclurei did not adversely affect germination rates, although they did cause a certain degree of delay in germination. Zhao’s research on the impact of arils from Euonymus alatus, Euonymus maackii, and Celastrus orbiculatus on seed germination revealed no significant inhibitory effects [49]. Arils serve a protective role for seeds, shielding them from direct external harm and providing a buffer against abrupt environmental changes [49,50,51]. However, it should be noted that the protective effect of aril retention was only directly tested in non-dried seeds in this study. Future studies should examine whether this benefit persists across different drying durations to fully understand aril–seed moisture interactions during storage. This is particularly important given that the aril’s hypothesized moisture-regulating function may vary depending on the seed’s initial water status. In this study, seeds from which arils had been removed demonstrated significantly lower viability and germination rates compared to those with arils, and other physiological indices also showed poorer performance. This evidence suggests that arils can indeed offer some resistance to abrupt external environmental changes. During seed storage, arils help maintain higher levels of viability and germination potential.

It has been proven that seeds can disseminate the core microbial communities that are unique to plants [52], and the microorganisms transmitted via seeds can bestow vital functional traits upon plants during germination [53]. These traits have a direct bearing on the health and growth of seeds and hold significant potential for applications in enhancing seed germination rates, growth rate, and crop protection. Therefore, the preservation of seed-borne microorganisms during seed storage is an important aspect in determining the quality of stored seeds. In this study, the effects of different conditions of storage on the preservation of microorganisms of M. macclurei seeds were investigated. Extensive research has indicated that seed microorganisms can influence the growth and development of plants, either directly or indirectly. For example, endophytes isolated from Zea mays seeds had been found to produce a range of antibiotics and exhibit antagonistic effects against pathogens [54]. An endophytic bacterium within Oryza sativa seeds can fend off infections by pathogenic Burkholderia bacteria [55]. The seed microorganisms of Triticum aestivum can effectively inhibit the germination of winter spores of Tilletia controversa [56]. However, seed microorganisms may also exert negative effects on seed storage and germination, and some can induce diseases and mortality in plants [57]. In this study, a microbial community of M. macclurei seeds was found to be significantly affected by storage conditions, with the community structure undergoing substantial changes due to pre-storage treatments such as drying or aril removal, as well as varying storage temperatures. The findings suggest that alterations in the seed microbial community are likely associated with seed deterioration, and that microbial diversity is enhanced when seeds are at 4 °C and −196 °C. It is inferred that low-temperature storage may mitigate seed deterioration by safeguarding seed microbial diversity. Both Aspergillus and Penicillium fungi were known to induce mold decay in seeds [58,59]. The dominance of Aspergillus and Penicillium as fungal genera after 1 month of storage in this study implies that post-storage fungal infection could be an important reason for seed deterioration. However, the existing data are insufficient to discern other potential correlations between shifts in the seed microbial community and seed decay, necessitating further investigation. This study can provide some contributions to the subsequent related research.

5. Conclusions

In this study, the physiological, biochemical, and microbial community indices of the seeds stored under different conditions were determined. The seeds that underwent a 1-day drying period and were then stored at 4 °C exhibited the highest viability and germination rates. The seeds stored at both 4 °C and −196 °C demonstrated enhanced antioxidant enzyme activity. Principal component analysis showed that preserving arils and storing seeds at −196 °C or 4 °C following 0~1 day drying could effectively reduce deterioration during storage. Low-temperature storage, shade drying prior to storage, and retention of arils were found to be effective in reducing the deterioration of stored seeds. This paper also examined the richness and diversity of seed microbial communities, as well as the abundance of dominant bacteria at the phylum, class, and genus levels. After 1 month of storage, Ascomycota and Proteobacteria emerged as predominant bacterial phyla. At the genus level, Penicillium and Rhodococcus were dominant fungal and bacterial genera, respectively. The diversity and richness of bacteria in seeds post-drying were found to be higher compared to non-dried seeds, while the diversity of fungi was lower in dried seeds. Following low-temperature treatment, both diversity and richness of fungi and bacteria were higher than those in seeds stored at 25 °C. However, the richness of bacteria was found to be lower than that in seeds stored at 25 °C. When arils were removed, diversity and richness of bacteria increased compared to seeds with arils, while diversity and richness of fungi decreased. The findings suggest that preserving arils and storing seeds at −196 °C or 4 °C after a 1-day drying period were the most beneficial to seed preservation.