Sugar Receding in Aril Benefits the Recalcitrant Seeds of Litchi (Litchi chinensis) and Longan (Dimocarpus longan) to Cope with Dry Spells after Maturation

Abstract

Litchi and longan are tropic/subtropic fruits harboring recalcitrant seeds that are covered with sugar-rich arils as the edible part. A rapid sugar content decline, called sugar receding, in the arils after the fruit maturation largely reduces the fruit quality, however, the mechanism is unclear. Litchi and longan fruits grow and mature in a hot and rainy season with dry spells between rainfalls. Here, we found that the seed maturation was around 2 weeks earlier than the fruit maturation, after which the sugar contents of the arils significantly decreased, while the fruits could stay on the tree for 1–2 months with high seed vigor. A continuously low-level fruit drop after the seed maturation resulted in continuous seed germination and seedling establishment in around 2–3 months. Blocking assimilate supply by storage of detached fruits or on-tree girdling-plus-defoliation for 7 days, the aril sugar contents of the treated fruits significantly decreased, while the sugar levels and vigor of the seeds increased, implying that the arils are sugar sources to maintain or even enhance the seed vigor and suggesting that sugar receding in arils benefits the recalcitrant seeds of litchi and longan to cope with dry spells after maturation.

1. Introduction

Litchi (Litchi chinensis) and longan (Dimocarpus longan) are members of the Sapindaceae family, originally distributed in southern China and nearby regions [1,2,3,4]. As important evergreen fruit crops, litchis and longans are widely grown in the tropical and subtropical regions across the world. The fruits develop and mature in a hot rainy season from late spring through early autumn, and the edible part of the fruits is a fleshy aril around the seed. The aril accumulates sugars up to 15–25% of the fresh mass during the development, giving the sweet taste of the fruits [2,5]. Additionally, the arils are translucent and juicy with exotic flavor, contributing to the attractivity of the fruits to consumers around the world [6,7].

Sweetness is an important eating quality for litchi and longan fruits. However, a phenomenon called “sugar receding” of the fruits, known as a significant decline in sweetness or total soluble solids (TSS) content of the arils soon after fruit maturation, has been found to largely reduce the fruit quality [8,9, 10,11]. This trait is the main factor shortening the harvest period (about 1 week) and limiting the industry development of the fruits. For example, around a 15% decrease in the contents of TSS was found in the on-tree Feizixiao litchi fruit after a harvest delayed by 7 days [12]. The TSS contents decline from 20% to 14% in 10 days after fruit maturation of Cihezhong longan [13]. This quick and dramatic decline in sugar or TSS contents after fruit maturation has rarely been found in other horticultural fruits. In contrast, even a slight increase in TSS contents was found in the fruits of some cultivars of citrus and grape when being stored on-tree for around 2 months after maturation [14,15]. Although intensive research has been focused on this issue, the mechanism of “sugar receding” is largely unclear so far [8,9,11,12]. It had long been believed that the sugars in the arils were transported back to the branches of mother trees to support the new leaf flushing, which was the origin of the term “sugar receding” in the industry. However, the transport back hypothesis controverts to the general photosynthate partition law in plants [16]. So far, the studies on “sugar receding” mainly focused on the changes in sugar components, activities of sugar-metabolism-related enzymes, and their coding genes [8,9,10,11,12]. These studies help us to understand the sugar metabolism in the arils during the “sugar receding”, but they cannot explain the reason for or role of the phenomenon. Based on evolutionary rationales, Sobral [17] proposed that all traits are functional in plants. In this context, we expected that the sharp sugar decrease in the arils of litchi and longan fruits might be connected to some uncharacterized process of the fruits or seeds.

Litchi and longan seeds are typical recalcitrant seeds sensitive to water loss and vigor loss [18,19,20]. Recalcitrant seeds are much more likely to be dispersed in the wet season and can immediately germinate when there is adequate water in the environment [21,22,23]. Recalcitrance of seeds is regarded as a major adaptation for their survival in the environment, for rapid germination of recalcitrant seeds may minimize both the risk of seed drying and post-dispersal predation (the duration of exposure to predation). In the wet season, dry spells recur persistently in tropical rainforest weather. After germination, the seedlings will still face unpredictable environments; this could be good for seedling establishment, or the seedlings might die when a long drought period occurs [24]. Theoretically, multiple germinating times of seeds can enhance the probability of plant propagation in a relatively long period with unpredictable dry weather; the robust seedlings have rapid access to soil water, thereby minimizing the risks of mortality induced by dry spells in the wet season.

Litchi and longan fruits belong to a distinct fruit type with recalcitrant seeds enveloped by a specialized seed appendage called the “aril”, which is an outgrowth of the seed and is an accessory part of the seed [20]. As a composite structure between the pericarp and seed coat (Figure 1A), the aril differentiates from the funicle of the ovule soon after fertilization and develops as a partial or complete envelope for the seed [25]. Fruits with aril-covered recalcitrant seeds are very common in tropical and subtropical plant species, such as ackee and rambutan besides litchi and longan of the Sapindaceae family; mangosteen of the Guttiferae family; durian of the Bombacaceae family; annonas of the Annonaceae family, etc. [26]. In general, arils as well as the edible pericarp of fleshy fruits are regarded as attractants and rewards for animal seed dispersers [27,28]. However, as a tissue originating from the funiculus, other ecological roles (if there are any) of arils are paid little attention in terms of their responses to environmental conditions. Recently, some studies have shown the importance of arils in regulation of seed germination time and viability maintenance. Evrard [29] reported that the presence of the aril lowered the seed germination rate of Afzelia bipindensis, for the aril may contain growth inhibitors such as flavonoids, which are necessary to temporarily promote seed longevity. Our previous and other studies have reported that the pericarp and aril could help to retain seed moisture and favor viability maintenance of these seeds in the dry spells of the wet season [20,30]. In litchi and longan fruits, whether the sugar-rich arils and the “sugar receding” of arils can modulate the seed germination vigor during unpredictable dry weather is unknown.

The aril is a noteworthy feature of lichi and longan fruits due to its important economic value. Numerous studies have focused on the initiation and development of the aril, as well as the accumulation of sugar during fruit development in litchis and longans [2,5,26]. However, as a differentiated structure of seeds, the biological significance of the arils to the environmental adaptability are largely overlooked or unknown [20]. Here, we expected that arils of lichi and longan fruit might serve as a sugar-rich tissue beneficial for the maintenance of seed vigor on-tree or in the dropped fruits, which might enhance chances of seed germination and allow seedling establishment for a longer period and consequently spread the propagation risks in unpredictable dry weather.

To test the hypothesis, we used data from three-year experiments to investigate the relationship between seed germination/vigor and sugar contents in the arils/seeds throughout the development of litchi and longan fruits. We further applied two approaches to block the transport of photosynthate from the mother trees to the fruits/seeds to analyze the change in seed germination/vigor and sugar contents in the arils/seeds. To our best knowledge, this work is the first to link the fruit “sugar receding” with the germination vigor of seeds, significantly advancing our understanding of the “sugar receding” of lichi and longan fruits and also germination strategies of aril-covered recalcitrant seed fruits.

2. Materials and Methods

2.1. Plant Materials

Litchi (Litchi chinensis Sonn. cv. Heiye) and longan (Dimocarpus longan Lour. cv. Shixia) fruits at different days after anthesis (DAA) throughout the fruit development season, specifically at the stages after aril initiation till the complete fruit drop (end of the season), were collected in 2021, 2022, and 2023 from an orchard located in the Conghua district of Guangzhou City, Guangdong, China, at latitude 113°53′ N, longitude 23°53′ E. The data of temperatures and rainfalls of the location were obtained from the Guangzhou weather website: www.tqyb.com.cn (accessed on 1 January 1990–2020 for the average values; and 1 April 2021–2023 for the daily values). Three litchi or longan trees with the same flowering stage were randomly selected in the orchard, and fruits evenly distributed in the north, south, east, and west directions were labeled for the fruit collection of each tree. At certain developmental stages, fruits at similar stages were collected and transported to the lab in two hours. Only those fruits with uniform appearance without any damage were chosen for the following experiments. The weight of a whole fruit/aril/pericarp/seed, TSS contents, seed germination rates, seedling establishment rates, and the stem/root length of the seedlings at different DAA were immediately determined after arrival at the lab. Aril and seeds for the determination of soluble sugar and starch contents were stored at −80 °C until the assays.

2.2. Fruit Drop Rate Evaluation

To characterize the fruit drop pattern after aril initiation, fruit abscission rates were monitored from 52 and 61 DAA for the litchi and longan trees, respectively, until the complete fruit drop after fruit maturation, namely to 108 DAA for the litchi trees and 173 DAA for the longan trees. For this purpose, the tree selection and fruitlet labeling in the trees were as described above. According to the method by Li et al. [31], cumulative fruit drop rate was calculated by subtracting the number of remaining fruits from the initial number, dividing by the initial number. Relative fruit drop rate was calculated by subtracting the number of remaining fruits from the last recorded number, dividing by the last number. Fruit drop was surveyed weekly for the litchi trees and bi-weekly for the longan trees. The cumulative and relative fruit drop rates were expressed as percentages. Means and standard errors were calculated on a per-tree basis.

2.3. Fruit Storage Treatment after the Fruit Detachment from the Trees

Litchi and longan fruits, at 66 DAA and 89 DAA, respectively, were detached from the trees. To assess the change in the fruits after the removal from the trees, 5 lots of detached fruits (with 30 fruits each, serving as 5 biological replicates) were packed separately in plastic bags to avoid rapid water loss and stored in an artificial climate chamber at 25 ± 1 °C and 80% relative humidity. After 7 days of storage, the fruits without any disease from each replicate were chosen for the measurement of the TSS, sugar contents, and seed germination.

2.4. Girdling Plus Defoliation (GPD) Treatment on Fruit Clusters

To assess the effect of the blockage of assimilate transport from the trees to the fruits/seeds, girdling plus defoliation (GPD) treatments of litchi and longan fruit clusters were conducted at 66 DAA and 89 DAA, respectively, with non-GPD-treated fruits as the control, according to the method by Li et al. [31]. About 20 to 30 clusters evenly distributed in different parts of the trees as described above were chosen in each tree for the 3 treatments (with 7–10 clusters for each). A ring of bark (including the cambium) about 1 cm in width was removed from the branch base, followed by removing all leaves above the girdle (defoliation). The fruits on trees before the treatment (0 d) and 7 days after the treatments of GPD and the control were collected for the measurement of the TSS, sugar contents, and seed germination.

2.5. Fruit Weight

Thirty litchi/longan fruits were selected randomly, and the weights of the whole fruit, aril, pericarp, and seed were measured by an analytical balance.

2.6. Ripe Seed Definition, Seed Germination, and Seedling Establishment Rate

After removing the pericarp and aril, 5 lots of seeds (with 10 seeds each) were used as 5 biological replicates. The seeds were cleaned with tap water and sown in moist soil. The seeds were kept in the dark at 25 °C for 14 days for germination and growth. The seed germination was defined as when the radicle protruded around 2 mm from the seed coat, and the emergence of the first pair of true leaves from the stem above the soil was defined as the beginning of seedling establishment. Fourteen days after the sowing, the seed germination rates, the root length, and stem length in each replicate were manually measured.

Germination rate (%) = number of germinated seeds/number of test seeds × 100

Seedling establishment rate (%) = number of established seedlings/number of test seeds × 100

The seeds are regarded as ripe seeds when they present around 90% seed germination and 80% seedling establishment.

2.7. Determination of the Total Soluble Solid Contents

At least 25 fruits as 5 biological replicates (with 5 fruits each) were used for determination of the content of total soluble solid (TSS) in the arils. The juice of the arils from 5 fruits (for one replicate) was squeezed and filtered by gauze and TSS determined by a digital display refractometer (ATOGO, Tokyo, Japan). It should be noted that, for young fruits with a thin layer of aril, at least 5 fruits were required to obtain enough juice for the determination.

2.8. Determination of the Total Soluble Sugars

The total soluble sugars in the arils or seeds were determined by the anthrone–sulfuric acid colorimetry method as described by Buysse et al. [32]. Briefly, the samples were immediately frozen in liquid nitrogen and ground into powder. Then, a 0.03 g sample of the aril or seed powder was placed into a centrifugal tube, mixed with 1 mL distilled water, and heated in a boiling water bath for 10 min. The extracts were centrifuged at 8000× g for 10 min at room temperature. The supernatant was collected, and the residues were extracted again as described above. The supernatants of the two extractions were combined and 200 µL of the combined extract was added to 1.3 mL of the anthrone–sulfuric acid reagents from the Micro Plant Soluble Sugar Content Assay Kit (Solarbio, Beijing, China) and heated in a 95 °C water bath for 10 min. After cooling to room temperature, the absorbance of the reaction solution was determined at 620 nm according to the method by Zhang et al. [33] by using a microplate spectrophotometer (51119670DP/Multiskan Sky with Touch Screen, Thermo Scientific, Waltham, MA, USA). Various concentrations of glucose solutions (0.00625–0.2 mg/mL) were used to construct a standard curve and the soluble sugar contents in the extracts were calculated based on the standard curve. The data were collected from five biological replicates.

2.9. Determination of the Starch Contents

The seed samples were immediately frozen in liquid nitrogen and ground into powder. A 0.03 g sample of powder was placed into a centrifugal tube and 1 mL of 80% ethanol was added. After incubating in an 80 °C water bath for 30 min, the extraction was centrifuged at 3000× g at room temperature for 5 min. The supernatant, which contained the soluble sugars, was removed and the residue was extracted again as described above and the supernatant was removed. The residue that contained starch was added to 0.3 mL distilled water and hydrolyzed by heating in a boiling water bath for 15 min. The content of the sugar moieties after hydrolyzation was determined using the anthrone–sulfuric acid colorimetry method as described in Section 2.8. The starch content in the residue was calculated with a Micro Plant Starch Content Assay Kit (Solarbio, Beijing, China) according to the method by Zhang et al. [33]. The data were collected from five biological replicates.

2.10. Data Statistical Analysis

Each experiment was repeated in three years with similar results, and data from representative, individual experiments are presented. The data of the figures in this paper are expressed as mean ± standard error (SE) with 30 biological replicates for weight values and 5 biological replicates for other parameters. Statistical significance of the experimental data was analyzed by SPSS 22.0, using t-tests.

3. Results

3.1. The Growth Profiles of the Arils of Litchi and Longan Fruits in the Developmental Season with Fluctuating Weather

The flowers of Heiye litchi bloom at the beginning of April, and Shixia longan flowers bloom in the middle of April at the location. The whole fruit development from the flower bloom to complete fruit drop after maturation lasts around 110 and 170 days for the litchi and longan, respectively, which ends in the middle of summer for the litchi and in the beginning of autumn for the longan. The fruit development season is the hottest season of the year at the location with average maximum temperatures of 29.1 °C based on the statistics of the last 30 years. The season is also the most rainy season of the year, with two rainfall peaks at the end of June and the beginning of September (Figure 1B). When monitoring the weather from April to September (the fruit development season) in 2021, 2022, and 2023, we found that the temperatures were relatively stable, while the rainfall fluctuated and dry spells of more than half a month occurred (Figure 1C,D).

The mature litchi and longan fruits are composed of pericarp, aril, and seed (Figure 1A), among which the seed and pericarp developed and grew first, in agreement with the findings in previous studies [3,34]. As shown in Figure 1E–G, the aril was initiated at around 52 DAA (21 May) for the litchi and at 61 DAA (18 June) for the longan, when the seeds almost reached their maximum weights. Then, the weights of the whole fruit and the aril increased rapidly in parallel until they reached the maximum weights at 80 DAA for the litchi and 131 DAA for the longan, when the fruits became fully red for the litchis and light brown for the longans. Then, a slight weight drop was found for both the whole fruit and aril for the two fruits (Figure 1H,I).

3.2. The Relationship between Seed Vigor and Sugar/Starch Contents in the Arils and Seeds throughout the Development of Litchi and Longan Fruits

With the development of the fruits, sugars and starch accumulate in the arils and seeds. The contents of total soluble solid (TSS) in the arils increased to the maximum values at 80 DAA (19.52%) and 103 DAA (25.32%) for the litchi and longan fruits, respectively, which are generally regarded as mature fruit stages in the industry. Then, the TSS contents decreased to 12.52% at 101 DAA for the litchis and to 12.48% at 159 DAA for the longans. Similar patterns were recorded for the total soluble sugar contents in the arils for both the fruits. The total soluble sugar contents in the arils of litchi and longan decreased 22% and 54%, respectively, from the peak values at 3 weeks after fruit maturation, consistent with the “sugar receding” phenomena found for the two fruits (Figure 2A,E).

The contents of total soluble sugars in the seeds increased first and then decreased during the development, in agreement with the theory that sugars in seeds are accumulated first and then converted to starch. The total soluble sugar contents peaked at 73 DAA and 68 DAA for the litchi and longan seeds, respectively. A slight increase in the sugar content was found for both the seeds in the late development period after fruit maturation. Rapid starch accumulation in the seeds was found in the periods before the seeds reached the maximum weights at 59 DAA for the litchis and 75 DAA for the longans, and then the accumulation slowed down before fruit maturation. A slight decrease in starch contents occurred after 80 and 131 DAA for litchi and longan fruits, respectively (Figure 2B,F).

For the litchi fruit, the seeds showed a 4.5% germination rate at 52 DAA and no seedlings were established when the seed coat was still pale green and the seed was only partially covered by a very thin layer of aril. At 59 DAA, when the seeds turned brown and were completely covered by arils, the seed germination rate sharply increased to nearly 100% and the seedling establishment rate (seedling rate) was around 12%. The seed germination/seedling rates were almost 100% at 66 DAA, indicating that the seeds were mature at this stage. At 101 DAA, 4 weeks after fruit maturation, the seed germination/seedling rates were maintained at high levels (95–100%) (Figure 2C). To further monitor seed vigor, we recorded the stem/root lengths of the seedlings at 14 d after the seed sowing. The maximum lengths of the stems (9.7 cm) and roots (10.1 cm) were found for the seeds from the fruits at 73 DAA. Then, the lengths of the seedlings declined slightly, and the length was around 80% of the maximum length at 101 DAA, indicating that the seed vigor was maintained at high levels for almost a month (Figure 2D).

Similar patterns were documented for the longan fruits, whose seeds started to have germination capacity (2.5%) at 61 DAA when the seeds were white and not fully covered by a thin layer of aril. The seed germination rate and seedling rate reached around 80% when the seeds turned brown and were completely covered by arils at 75 DAA, and the seeds were regarded as mature at this stage (Figure 2G). The lengths of the stems and roots of the seedlings were 4.8 and 5.1 cm, respectively, at 103 DAA when the fruit matured. The seedling lengths continuously increased to the maximum at 131 DAA but significantly decreased at 159 DAA when the seeds still showed around 95% germination and seedling rates (Figure 2H).

Based on these data, we found that the seed germination rates reached the maximum prior to the onset of the TSS peaks and stayed at almost 100% even after the significant decrease in the TSS contents after fruit maturation. With further decrease in the aril TSS contents, only a slight decrease in seed vigor was found for the litchi at 3 weeks after fruit maturation; and increased seed vigor was observed even at 4 weeks after the longan fruit maturation. The results indicate that the “sugar receding” in the arils may be beneficial to maintain (for the litchi) or even enhance (for the longan) the seed vigor.

3.3. Arils Benefit Seed Vigor Maintenance/Enhancement after Photosynthate Blockage of the Fruits on-Tree

To further confirm the function of the arils in the maintenance/enhancement of seed vigor, we carried out girdling plus defoliation (GPD) of the fruit clusters, so as to block the transportation of photosynthates from the trees to the seeds/fruits. The GPD experiments were carried out at the same time as the above-described storage experiment, namely at 66 DAA for the litchis and 89 DAA for the longans (Figure 3A–C). After the 7 d on-tree experiment, the TSS contents of the arils in the non-GPD control fruits increased from 15.2% to 16.29% for the litchis and increased from 17.98% to 20.59% for the longans. Meanwhile, the contents of the GPD-treated fruits dropped from 15.2% to 13.36% for the litchis and from 17.98% to 14.74% for the longans. Similar patterns were recorded for total soluble sugar contents in the arils, with a 8.2% and 32.3% decrease in the litchis and longans, respectively, 7 d after the GPD treatment (Figure 3D,H). In contrast to the arils, significant increases in the contents of total sugars and starch were detected in litchi seeds 7 d after the GPD, and a slight increase in these contents was found for the longan seeds (Figure 3E,I). The results imply that after the blockage of photosynthates from the trees by GPD, the increase in sugar and starch contents in the seeds is due to the transport of the sugars from the aril to the seeds.

Seed germination rates of the litchi fruit were 100% before and after the GPD/non-GPD treatment. The seedling establishment rates increased from 76% to 94% 7 d after GPD, and the rates increased from 76% to 98% for the non-GPD control fruit (Figure 5F). Seed germination rates of the longan fruit were around 82% before the girdling, while they increased to 92% and 94% for the GPD and non-GPD treated fruits, respectively, after 7 d. The seedling establishment rate of the longan fruit increased from 68% to 80% after the GPD and from 68% to 82% after the non-GPD treatment (Figure 3J). The seedling lengths of the litchi and longan seeds increased significantly after the GPD and non-GPD treatments (Figure 3G,K). The results indicate that after the block of photosynthate supply, the aril becomes the source for the seeds to maintain and even enhance the seed germination and seedling establishment.

3.4. Fruit Drop Profiles of the Litchi and Longan Trees at the Developmental Stages after Aril Initiation under Fluctuating Weather

There is a continuous fruit drop throughout the developmental season of litchi and longan trees, in which more than 90% of the total fruit drop occurs in the early developmental period that is dominated by the growth of seeds and pericarp, and the fruit drop dramatically slows down afterward [15]. As we found that the seeds started to have germination capacity after aril initiation, we monitored the fruit drop patterns after the aril initiation to understand the potential function of the aril on seed vigor. Here, we referred to the fruit drop rate at the first time point (52/61 DAA) as 0% and determined the fruit drop rates afterward. We found low relative fruit drop rates (less than 5% each week) were recorded before fruit maturation (80 and 103 DAA for the litchi and longan, respectively). A relatively higher rate, 14.9% for the litchi at 80 DAA and 5.9% for the longan at 68 DAA, was found when a severe typhoon was recorded at the location in 2023. However, significantly higher relative (30.6%) and cumulative (48.5%) fruit drop rates were recorded for the overripe litchi fruits (94 DAA, 2 weeks after the fruit maturation) (Figure 4A). Relatively to the litchis, the rise of fruit drop rates of the longans was postponed to 4 weeks after fruit maturation (131 DAA; as overripe fruit), when relative (11.4%) and cumulative (27.3%) rates were recorded (Figure 4B). Complete fruit drop was found at around 108 DAA and 173 DAA for the litchis and longans, respectively.

As continuous fruit drop happened throughout the developmental process, dropped litchi and longan fruits with various sizes were easily found under the trees at around the fruit maturation time (Figure 4C,F). Seedlings at different developmental stages, such as those with 2–10 leaves, and even just-germinated seeds, were observed beneath the tree canopies, indicating that the continuous fruit drop results in continuous seed germination and seedling establishment throughout the development (Figure 4D,E,G,H). The results further suggest that the continuous fruit drop behavior (Figure 4A,B) and high seed vigor maintained for a long duration described above (Figure 2C,D,G,H) enhance the probability for seed germination and seedling establishment and consequently spread the risk over a relatively long period with unpredictable dry weather.

3.5. Arils Benefit Seed Vigor Enhancement after Fruit Detachment from Trees

As described above, fruit drop occurred throughout the whole development season. To investigate whether the sugar-rich aril may benefit the seed vigor and seedling establishment after fruit drop, we collected the fruits 2 weeks before fruit maturation (66 DAA for the litchis and 89 DAA for the longans) and stored them in a chamber at 20 °C. After a 7 d storage period, the TSS contents of the litchi arils dropped from 15.5% to 13.85%, and that of the longan arils declined from 18.13% to 16.31%. Decreases of around 8% and 16% in total soluble sugar contents of the arils were recorded for the litchis and longans, respectively (Figure 5A,E). In contrast to the arils, significant increases in total sugars were detected in both the seeds after the storage. The starch content in the longan seeds significantly increased after the storage, and the content increased slightly for the litchi seeds. The results imply that the increase in sugar and starch contents in the seeds is relevant to the sugar decrease in the arils (Figure 5B,F).

Seed germination rates of the litchi fruit were 100% before and after the storage, while the seedling establishment rate increased from 76% to 100%. Seed germination rates of the longan fruit increased from 84% to 90% after the storage, and the seedling rate increased from 66% to 78% (Figure 5C,G). Significant increases in the lengths of seedling stems/roots were found for both the litchi and longan seeds after the fruit storage (Figure 5D,H). The results imply that after removal from the trees with no further photosynthate distribution, sugars in the arils may be the supply for the seeds to maintain and even enhance the seed vigor and seedling establishment.

4. Discussion

It has long been found that a dramatic decrease in fruit TSS or sugar content in a short time after maturation occurs in litchi and longan fruits, which is called “sugar receding” in the industry. It is a special phenomenon, for it has rarely been found in other horticultural fruits. Although many previous studies have attempted to understand the phenomenon, the underlying mechanism is still debated [8,9,10,11,12]. Here, we showed that the “sugar receding” of lichi and longan fruits could play a role in seed vigor maintenance via sugar supply for the seeds, which could spread the seed germination risk in unpredictable dry weather. To our best knowledge, this work is the first to link the “sugar receding” with seed vigor, which would significantly advance our understanding of “sugar receding” of lichi and longan fruit in the industry and seed germination strategies of aril-covered recalcitrant seed fruits.

Litchi and longan seeds are typical recalcitrant seeds and have a very short life span after release [18,19,20]. Their trees grow in tropical and subtropical environments and their fruits develop and mature in April to September, the most hot and rainy season of the year (Figure 1B), like other plant species with recalcitrant seeds [35,36,37,38]. When the seeds of litchi and longan shed and germinate in a wet period, seedlings can be well established with adequate water in the environment. However, even in a tropical rainfall season, dry spells always occur (Figure 1B–E). When a relatively long dry period occurs after seed shedding or germination, the seeds or seedlings are bound to die [24]. Based on a previous study [20] and the present study, we propose that litchi and longan trees have evolved special traits including “sugar receding” to cope with the variable environment.

In the litchi and longan industry, the fruits are regarded as mature when the TSS content of the arils reaches the maximum value, which is when the aril is the sweetest, at around 75–85 DAA and 95–115 DAA for litchi and longan fruits, respectively [2,5]. In the present study, we found that seeds matured around 2 weeks earlier than the fruits (Figure 1F,G and Figure 2A,E). Meanwhile, the TSS of the arils reached the maximum values at 80 and 103 DAA for the litchi and longan fruits, respectively (Figure 2C,G). These data indicate that the maturation of the seeds is much earlier than that of fruits. Moreover, we found that after maturation the fruits could stay on the tree for a relatively long period with a high seed germination rate until the complete fruit drop (Figure 3A,B). This is not common for most plants, which generally shed the fruits once seeds have matured [39]. Combined with the fact that litchi and longan fruits shed continuously during fruit development, we suggest that the long duration of seed vigor maintenance on-tree enhances the probability of seed germination and seedling establishment. This special trait evolved by litchi and longan trees can spread the risk over a relatively long period with unpredictable dry weather, which is a good plant regeneration strategy.

For such a mission, maintenance of high seed vigor for a long time is a prerequisite, in particular when no photosynthate is assigned after fruit maturation. It has long been found that a phenomenon called “sugar receding” of fruits commonly occurs in litchis and longans after fruit maturation, leading to a dramatic decrease in the aril sugar contents in around a week [8,9,10,11,12]. To date, the mechanism of “sugar receding” of fruits is still unclear. In the present study, we found that after fruit maturation, the seed vigor was maintained at high levels along with the aril TSS decline (Figure 2). Based on these findings, we propose that the aril may supply sugars for the seeds to maintain high vigor after fruit maturation for a long time, which leads to the dramatic sugar decrease in the arils, i.e., the “sugar receding” of fruits.

Litchi and longan fruits develop and mature in a hot rainy summer season (Figure 1) in which typhoons, heavy rain, or storms often happen and result in increasing fruit drop due to the action of rain and wind (Figure 3) [29]. In addition, the activity of birds, bats, and monkeys also increases the fruit drop of litchis and longans in a natural ecosystem [29,40]. When fruit drop occurs in a dry period, the dropped litchi and longan seeds are bound to die; in a rainy period, the seeds can immediately germinate and establish seedlings in the soil with adequate water. Since litchi and longan fruits could stay on the tree for a long period after seed maturation, the continuous fruit drop may give the seeds multiple chances to take advantage of rainfall. Here, we found seedlings with divergent ages (with 2–10 leaves) under the litchi and longan trees, indicating that the seeds germinated and established seedlings at different times (Figure 3C–H). Accordingly, the traits of continuous fruit drop and high seed vigor maintenance on the tree of litchis and longans provide multiple germinating times during the season, spreading the risk over a relatively long period in unpredictable dry weather, to avoid the high risk of mortality of the seedlings when an excessive single-time fruit drop occasionally occurs during a long drought.

In the field, litchi and longan fruits mainly drop as whole fruits with intact pericarp and aril. Our previous study reported that the arils of litchis and longans helped to avoid quick water loss of seeds due to their high content of water and accordingly favored seed vigor maintenance after harvest [20]. In this study, we showed that the treatment of girdling plus defoliation (GPD) resulted in a significant decrease in TSS contents in the arils due to the blockage of sugar supply, while slightly increasing the sugar and starch contents in the seeds. The treatment did not affect seed germination rates, and even significantly enhanced seed vigor and seedling growth (Figure 5). Similar results were found for the stored fruits that were removed from the trees (Figure 4). This is not a general pattern for plants, as many studies reported that seed vigor was reduced when smaller proportions of assimilates are transported to seeds due to partial defoliation [41,42] or severe herbivory [43]. Here, by on-tree GPD or removal from the trees to cut down the supply of sugars from the mother trees, our results provide direct evidence that the sugar-rich arils benefit the seed germination of litchi and longan.

The arils of litchis and longans evolved from the funicle of the ovule after fertilization. So far, many studies have shown that, in addition to the rewards for seed dispersers [27,28], the arils have multiple roles, including protection of the seeds from water loss and regulation of seed germination time to enhance survival under unfavorable weather conditions [20,30]. Here, we suggest that, in litchi and longan fruits, arils evolved not just for providing means for seed dispersal like fleshy fruit but also for other physiological roles due to their water-rich [20] and sugar-rich features for the purpose of germination and seedling establishment.

5. Conclusions

The seeds of the litchi and longan mature around 2 weeks earlier than the fruit. The fruits can stay on the tree for 1–2 months with high seed vigor after fruit maturation. The sugar receding phenomenon in the aril, as represented by a rapid sugar content decline, helps maintain or even enhance the seed vigor. A long duration of seed vigor maintenance on the tree or after detachment benefiting from the sugar receding is a good plant regeneration strategy to cope with unpredictable dry spells during the fruit development season.