Changes in water content and germination rate of Quercus acutissima seeds during desiccation
Q.acutissima is a dicotyledon; its seeds are nuts, the testa is thick and hard, and the endotesta is thin film-like, when fresh, tightly attached to the cotyledons(Fig. 1).
As desiccation time increased, Fig. 2 shows a general declining trend in the water content and germination rate of Q.acutissima seeds. Freshly harvested seeds with an initial WC of 38.8% were dried to a final WC of 14.8%, and the germination rate of fresh seeds was 98.67%. Before drying for 51 hours, the water content of seeds decreased rapidly, and subsequently, the rate of decrease slowed. When the water content reduced from 38.8% to 35.8%, the germination rate declined from 98.67% to 94.67% without any significant difference. However, the germination rate began to drop precipitously as the degree of dehydration increased after 51 hours of drying. With further desiccation to 26.8% WC (209h), the germination rate was 52.0%, and nearly half of the seeds lost vigor, indicating that the semi-lethal water content of seeds is approximately 26.8%. Seed germination rate had dropped to 0 while the WC had decreased to 14.8%. Analysis of the Pearson correlation revealed a highly significant positive correlation (P<0.01, R=0.982) between the germination rate and water content of Q.acutissima seeds, indicating that they were extremely sensitive to water loss and were typical recalcitrant seeds. The complete lethal water content of seeds is roughly 14.8%.
Changes in the water content of Quercus acutissima seed coat, embryo axis, and cotyledon during desiccation
This study divided seed samples into three parts: seed coat, embryo axis, and cotyledons, according to their special structures. As shown in Fig.3, the water content of each part was 33.21%, 51.59%, and 40.10%, respectively. According to the declined trend, we found that the seed coat is the first portion of the seed to lose water during desiccation. The WC of the seed coat reduced from 33.21% to 15.52 % after 7 hours of drying, whereas the embryo axis and cotyledon WC did not change significantly. Therefore, the germination rate remained high, and the seeds were not injured by desiccation. During the further drying, the WC of the seed coat declined to 10.53% after 51 hours and remained at roughly 10%. Meanwhile, the WC of the embryo axis and cotyledon began to decrease, and the seed germination rate substantially reduced from 94% to 84%. Seeds were progressively damaged upon desiccation. Pearson correlation analysis showed that the WC of cotyledons and embryo axis were significantly positively correlated with the germination rate (Table 2). After 51 hours of drying, the WC of cotyledon and embryo axis began to decrease rapidly, with embryo axis WC decreasing more rapidly than cotyledon. During 120-209 hours of drying, the WC of the embryo axis showed an inflection point of rapid decrease. A precipitous decrease in germination rate was from 70% to 52%. When the seeds were completely dead, the embryo axis and cotyledon water content reduced to 19.30% and 16.66%, respectively, after drying for 545 hours.
Ultrastructure of the seed coat
Q.acutissima seed coat (Fig. 4A) is composed mostly of the cuticle, palisade layer, and parenchyma, as revealed by a scanning electron microscope image. There is a thick waxy deposit on the surface of the cuticle (Fig. 4B). The Palisade layer is immediately close to the cuticle. It comprises a single palisade cell, closely packed and without gaps. These two layers form a structure that primarily impedes water, efficiently reducing water loss.
However, Q. acutissima has two water entry and exit sites in its seed coat. The micropyle is one of the entry and exit points for water in Q.acutissima seeds (Fig 4D, E). As observed in the micropyle cross-section (Fig. 4E), the top of the seed is composed of parenchyma, and there are numerous small pores from the endotesta through which water can enter and exit (Fig 4F). At the same time, the other parts consisted of closely arranged perianth remnant structures without loose pore tissue (Fig 4C). The second water entry and departure point occur at the scar of the Q.acutissima seed testa (Fig. 4G). Many vascular bundles where water can enter and exit are present in the scar (Fig. 4H). Moreover, the cross-section of the scar consists of loose parenchyma (Fig. 4P). It contains xylem vessels, which may transport water more efficiently than the dense, waxy covering of the seed coat. It is hypothesized that scar is the primary water entry and exit point for Q.acutissima seeds.
MRI analysis during seed desiccation
MRI is a non-destructive detection technique that can obtain a weighted image of the 1H density within the tissue, reflecting the distribution of 1H in the sample. The different brightness levels in the images indicated different moisture contents, and the red region represents the largest 1H density [21]. Fig5 demonstrates the MRI of a Q.acutissima seed cross-section during the drying process. It can be observed that fresh Q.acutissima seeds had a high WC, the red regions were the most distributed, and the testa was plainly visible. The red water signal was concentrated in the embryonic axis and the middle of the cotyledon, gradually decreasing from the embryonic axis to the cotyledon margins. Therefore, MRI demonstrated that the embryonic axis and the center of the cotyledons have higher WC and a significant difference between the embryonic axis and the cotyledon margins. The water signal of the testa disappeared after 7 hours of drying, while the water signal of the cotyledon margins began to fade progressively. We could observe the path of water migration in the MRI from 51-120 h of drying. The red signal of water gradually migrated toward the scar, and the area of the red water signal detected by the embryonic axis began to decrease gradually. Water was lost from the scar and the micropyle. From 209 to 358 h of drying, the water signal at the embryonic axis gradually weakened. It indicated that the embryonic axis was seriously damaged, and the red water signal in the cotyledons also gradually disappeared toward the scar. Due to the low WC, the water signal at the embryonic axis gradually disappeared between 358 and 545 hours of drying, and the water signal area in the cotyledons also decreased.
Division of T2 relaxation time and water status during desiccation
Since the duration of the transverse relaxation time reflects the water-binding forces and the degree of free hydrogen protons, the water status of a sample can be determined by comparing the peak positions of the T2 relaxation curve. The shorter the relaxation time, the more tightly the hydrogen nucleus is attached to the substance, the more difficult it is to remove the hydrogen nucleus, and the lower water freedom there is [22]. Fig6 demonstrates that during desiccation, the lateral relaxation time of water in Q.acutissima seeds ranges from 0.1 to 1000 ms. The relaxation curve has three distinct peaks, and the T2 relaxation times in the interval where each peak is placed, can be identified as T21, T22, and T23, respectively. The shortest relaxation time of T21 (0.1-1ms) corresponds to bound water, primarily hydrogen-bonded to macromolecules in the cell, and has weak mobility [23]. The intermediate relaxation time of T22 (1-10ms) is immobile water, which is mainly intercellular water with restricted but slightly higher mobile than bound water. T23 with the slowest relaxation time (10-1000ms) is free water, which exists between cells in a free status and flows freely in the cells. Free water is a suitable solvent that can participate in the material metabolism of cells [24].
For the T2 test, 3×10 seeds with the same shape and size were chosen to exclude individual seed differences' influence on the experimental outcomes. The results are depicted in Fig6; the curves of the three sets of samples basically overlap, demonstrating that under the condition that the quality of each group is consistent, individual differences across seeds have a negligible effect on the experimental outcomes.
The influence of desiccation on the water status of Quercus acutissima seeds
Figure 7 and Table 1 show the results of the T2 inversion curves of Q.acutissima seed drying. The relaxation range can be calculated beginning and ending peaks of each water status. The relaxation range of bound water was first unchanged and gradually decreased during desiccation. The relaxation range did not change considerably when the water content was reduced from 38.8% to 29.8%. It began to drop when the WC was less than 29.8%; as seed WC dropped to14.8%, the bound water relaxation range significantly reduced from 0.34ms to 0.17ms(P<0.05). Meanwhile, a decreasing trend was shown in the free water relaxation time. When seed WC declined to 14.8%, the relaxation range decreased from 29.33ms to 18.04ms, a reduction of 38.5% relative to the initial value. The relaxation range of immobile water was more complex than that of bound water and free water, exhibiting a decreasing, then increasing, and then decreasing trend. When seed WC dropped from 38.8 % to 35.8%, the relaxation range reduced dramatically from 6.15ms to 5.58ms. Bound water relaxation range increased significantly to 6.78ms when seed WC reduced from 35.8% to 29.8%. As desiccation progressed, the relaxation range decreased significantly to 4.6ms when the WC decreased to 14.8%, a fall of 24.7% from the initial. In general, the relaxation ranges of bound water, immobile water, and free water within the seeds were significantly reduced during drying indicating that the water freedom of the seeds was continuously reduced.
Changes in freedom degrees of each water status also result in the shift in the peak time that occurs during drying. Table 1 reveals that the peak time of bound water decreased with a slow trend. The initial peak time of 0.13ms significantly decreased to 0.10ms as the water content dropped to 17.8%. The peak time of immobile water has been decreasing, resulting in a leftward shift of the peak time. In the early drying stages, the peak time decreased slowly (38.8% - 29.8%) but began to fall drastically in the later stages (26.8% -14.8%). Overall, the peak time of free water shifted to the left. When the WC decreased to 23.8%, the peak time decreased significantly from 30.03ms to 26.75ms, a decrease of 39.93%. It indicated that during desiccation, to slow down the loss of water, the structure of the cells became tighter, and the binding ability of water and macromolecules in the cells increased.
Table 1. T2 inversion of Quercus acutissima seeds during desiccation
|
Water content(%)
|
T21
|
T22
|
T23
|
Relaxation range/ms
|
Peak time/ms
|
Relaxation range/ms
|
Peak time/ms
|
Relaxation range/ms
|
Peak time/ms
|
38.8
|
0.34±0.02a
|
0.13±0.02a
|
6.15±0.24abc
|
1.91±0.00a
|
534.60±0.48a
|
29.33±1.22a
|
35.8
|
0.35±0.00a
|
0.12±0.01ab
|
5.58±0.71cd
|
1.91±0.00a
|
487.57±20.36b
|
28.02±1.14ab
|
32.8
|
0.35±0.01a
|
0.11±0.01bc
|
6.43±0.27ab
|
1.87±0.08a
|
475.34±19.85bc
|
29.38±2.04a
|
29.8
|
0.34±0.00a
|
0.11±0.01bc
|
6.78±0.26a
|
1.83±0.00a
|
444.10±48.35cd
|
28.02±1.14ab
|
26.8
|
0.30±0.01b
|
0.11±0.01bc
|
6.65±0.50ab
|
1.59±0.11b
|
412.61±0.28de
|
28.02±1.14ab
|
23.8
|
0.23±0.00c
|
0.11±0.01bc
|
6.41±0.27ab
|
1.48±0.00c
|
403.11±0.28e
|
26.75±1.06b
|
20.8
|
0.19±0.00d
|
0.11±0.01bc
|
6.00±0.23bcd
|
1.29±0.00d
|
376.07±0.26e
|
24.39±0.99c
|
17.8
|
0.19±0.01d
|
0.11±0.01bc
|
5.45±0.22d
|
1.10±0.04e
|
290.81±11.79f
|
18.47±0.75d
|
14.8
|
0.17±0.02d
|
0.10±0.00d
|
4.63±0.31e
|
0.78±0.03f
|
247.58±17.18g
|
18.04±0.00d
|
Dynamic changes of peak area and its proportion during desiccation of Quercus acutissima seeds
Desiccation of recalcitrant seeds is a complex process consisting of a series of physical and biochemical changes. The peak areas S21, S22, and S23 corresponding to different relaxation times T21, T22, and T23 can be used to quantify the relative contents of bound water, immobile water, and free water in the seeds of Q.acutissima. Fig8 depicts the dynamic changes in the peak areas of the three water status during the desiccation of Q.acutissima seeds. Throughout the drying process, the total water content declined consistently. In contrast, the content of each water status grew or decreased, indicating that the various water status in the seed cells were mutually transformed.
The peak area of free water declined as water loss intensified due to partial conversion and drying loss. It was the first water lost during seed desiccation, indicating that the free water in seeds was more active and easily moved. The Pearson correlation analysis revealed a highly significant correlation between the peak area of free water and seed germination (P<0.01, R = 0.986). The peak area of immobile water showed a trend of first decreasing, then increasing, and then slowly decreasing. At the beginning of drying (35.8%-32.8%), as seed WC decreased, the cell structure began to contract, and some free water combined with macromolecular groups to transform into immobile water. Simultaneously, when the seeds dried and lost water, the degradation of nutrients and enzymes in the seeds caused a portion of the bonded water to move to immobile water, causing the content of immobile water to increase gradually and decrease slowly in the later stages [25]. When seed WC dropped from 38.8 % to 32.8 %, there was no significant difference change in the peak area of bound water. The peak area of bound water gradually decreases when the WC dropped below 29.8%. Seed germination also began to decline rapidly at this time, and the correlation between seed germination and the peak area of bound water was highly significant (P < 0.01, R = 0.981). Immobile and bound water was harder to lose than free water. Free water dropped by 80.74 % eventually, while immobile and bound water decreased by 31.04 % and 44.84 %, respectively.
Bound, immobile, and free water accounted for 10%, 35%, and 55% of the total water in the seeds, respectively, in fresh Q.acutissima seeds (Fig9), with the highest content of free water, which ensures its vigorous metabolic activity and makes the seeds easier to germinate. During drying, the peak proportion of free water showed a decreasing trend, the peak proportion of bound water showed a slowly increasing trend, and the peak proportion of immobile water changed more complicated, showing a decreasing and then increasing trend. The percentage of immobile water decreased from 38.8% to 35.8%, while the percentage showed a significant increase when the water content decreased from 35.8% to 14.8%. When seeds die, bound water, immobile water, and free water account for 13%, 61%, and 26% of total seed water, respectively. At this stage, the proportion of immobile water is at its peak, most free water is lost during drying, and the water distribution of seed is out of balance.
Table 2. Pearson correlation analysis of various indexes of Quercus acutissima seeds during desiccation
|
|
Water content(%)
|
Germination rate(%)
|
Testa water content(%)
|
Embryo axis water content(%)
|
Cotyledons Water content(%)
|
S21
|
S22
|
S23
|
Water content(%)
|
1
|
|
|
|
|
|
|
|
Germination rate(%)
|
.989**
|
1
|
|
|
|
|
|
|
Testa water content(%)
|
.703*
|
0.625
|
1
|
|
|
|
|
|
Embryo axis water content(%)
|
.982**
|
.987**
|
0.592
|
1
|
|
|
|
|
Cotyledons Water content(%)
|
.990**
|
.992**
|
0.601
|
.987**
|
1
|
|
|
|
S21
|
.978**
|
.986**
|
0.544
|
.989**
|
.995**
|
1
|
|
|
S22
|
.830**
|
.786*
|
0.559
|
.809**
|
.836**
|
.827**
|
1
|
|
S23
|
.993**
|
.981**
|
.750*
|
.967**
|
.971**
|
.955**
|
.774*
|
1
|
Relationship between total NMR signal amplitude and relative seed water content during desiccation of Quercus acutissima seed
LF-NMR is a convenient technique to monitor water changes during drying. It is a suitable non-destructive method to analyze water changes during seed drying based on the trend of the total amplitude of the NMR signal [26]. The relationship between the transverse relaxation signal and the water content of Q.acutissima seeds during drying is seen in Figure 10. Water weight during seed drying showed a significant linear relationship with the total NMR signal amplitude, and the linear equation was y=2292.41x+16817.76 (R2=0.99736). By testing the total signal amplitude of the NMR relaxation peak area, the water weight of Q.acutissima seeds can be quickly estimated using the linear connection between water weight and total signal amplitude.