**1. Introduction**

Research on reproductive characteristics and reproductive dynamics effectively identifies the course that endangers the plant population and the underlying mechanism [1]. Moreover, it clarifies the reproductive biological characteristics and the effects of external factors on the most critical aspect of plant reproduction, as well as the mechanism that endangers plant species [2].

At the stage of pollination, research on the reproductive system of plants is helpful for understanding their life history and causes of endangerment [3]. Recent studies have shown that corolla shape, diameter, color, odor, and other factors affect pollinator variety, visiting frequency, and visual and olfactory reactions, which reduces the possibility of hybridization and adaptability of

future generations. This effect leads to a decline in the number of species [4,5]. Pollen viability and stigma receptivity differ among plant species [6]. Highly dynamic pollen and stigma are conducive to pollination [7]. Conversely, presentation of pollen viability and stigma receptivity at different times influences pollination efficiency and limits the fruiting rate [8]. Moreover, the decline in pollination quality also causes various negative effects, including reproductive failure because of pollen restrictions, decline in genetic diversity, and depression from inbreeding [9]. Studies on plant species based on different population sizes and genetic variation demonstrate that inbreeding significantly affects seed rate, seed germination, survival, and resistance to stress [10].

Seeds are an important part of the life history of endangered plants [11]. Ecological research on the seeds of endangered plants reveals better approaches to breeding seeds and seedlings, which are of great significance for population expansion and preventing population extinction [12]. Some endangered plants have less natural production and low seed quality. This phenomenon fundamentally reduces the seed germination rate and restricts population growth [13]. Other endangered plants have special fruit morphologies or structures that are not conducive to the spread of seeds or germination, which also endangers these plants [14]. The seeds of some endangered plants have distinctive features, such as congenital dormancy or substances that inhibit germination, which leads to low sprouting rates and the scarcity of seedlings under natural forests [15]. Climate change and inefficient population regeneration in modern habitats can also be attributed to the limited expansion of endangered seeds and seedlings [16]. Some studies on *Phoebe bournei* (Hemsley) Yen C. Yang J.W. and *Abies chensiensis* Tieghem showed that animal damage to seeds and seedlings contribute to their scarcity in natural populations [17]. Moreover, the short life of seeds of endangered plants, such as *Cinnamomum micranthum* (Hay.) Hay and *Cathaya argyrophylla* Chun & Kuang, also restricts their expansion [18].

*Sonneratia* × *hainanensis* W.C. Ko, E.Y. Chen and W.Y. Chen is an aiphyllium from the genus *Sonneratia* (*Sonneratiaceae*) [19]. This species grows on beaches near the low water line or on low tide beaches along inland rivers with low tidewater salinity [20]. The adult tree of *Sonneratia* × *hainanensis* is only eight and they are currently only found at Dongge and Touyuan in the Qinglan protection zone in Wenchang City, Hainan Province [21]. China's State Forestry Administration lists this species as a wild plant species with an extremely small population [22]. Studies on *S.* × *hainanensis* show that it has an outcross reproductive system with a long flowering phase [23]. It has flowers that can bloom almost year-round, although some *Sonneratia* species have overlapping flowering phases [24]. The outcross reproductive system, interspecific crossing sympatry, and overlapping flowering phases of *Sonneratia* provide chances for interspecific hybridization [25]. Among these species, *S.* × *hainanensis* is a diploid (22 chromosomes) hybrid, possibly between *Sonneratia alba* Smith and *Sonneratia ovata* Backer [26]. The *S.* × *hainanensis* population has a low genetic diversity, but various groups have significant genetic differentiation [27]. This phenomenon is possibly closely related to its reproductive system. However, few studies have investigated the reasons for restricted *S.* × *hainanensis* population in terms of the effects of reproductive ecology and environment on seed germination.

We conducted field fixed-point observations and laboratory experiments to study the reproductive system and seed germination of *S.* × *hainanensis*, as well as their relationship with the environment to determine the reasons for its limited natural regeneration.

#### **2. Materials and Methods**

#### *2.1. An Overview of the Sample Plot*

The sample plot was located at Qinglangang Nature Reserve of Hainan Island. *Sonneratia* plant communities in the region grow on the beach near the low water line or low tide beach along the inland river with low tidewater salinity. The area is located at 19◦37 36" N and 110◦49 53" E. Mangrove plants present abundant varieties in these communities, including 24 species of true mangroves and 20 species of semi-mangroves. *Sonneratia caseolaris* (Linn.) Engler and *Bruguiera sexangula* (Loureiro) Poiret are the dominant mangrove species. In China, *Sonneratia* includes six species, namely, *S. caseolaris*, *S. alba*, *Sonneratia apetala* Buchanan-Hamilton, *S. ovata*, *Sonneratia* × *gulngai* N. C. Duke & B.R. Jackes, and *S.* × *hainanensis*. *S. apetala* is an introduced exotic plant with a developed population, whereas, *S.* × *hainanensis* is a seriously endangered plant species. The geographical location is shown in Figure 1.

**Figure 1.** Geographic distribution of the fixed plot.

## *2.2. Study Object*

We selected 8 trees of different *S.* × *hainanensis* strains from the bottomland of an enclosed sea harbor in Paigang Village, Dongge Town, Wenchang City, Hainan Province for the study. All trees were at least 100 years old, with strong trunks and branches. They were about 8 m high, with a diameter at breast height of 100 cm, and a crown breadth of 15 m × 15 m. The fruit type of *S.* × *hainanensis* was spherical berry with a diameter of 50–80 mm. The seeds were sickle-shaped, V-shaped, or irregular, with an average length of 10 mm and the outer seed coat was brown. There were 2 cotyledons, elliptical or oblong, 2–3 cm long, and 1–3cm wide.

#### *2.3. Determination of the Reproductive System*

#### 2.3.1. Estimating the Outcrossing Index (OCI)

Inflorescence diameter, flower size, and flowering behavior were measured and the assay of the reproductive system was specifically conducted according to the standards proposed by Dafni [28].

#### 2.3.2. Pollen to Ovule (P/O) ratio

Ten buds just coming into bloom with undehisced anthers were randomly selected and fixed in formalin-acetic acid-alcohol (FAA). The number of single pollen grains was counted according to the method by Cruden, and the number of ovules was measured by paraffin section. The pollen to ovule (P/O) ratio of each flower was calculated by dividing the number of pollen grains by the number of ovules [29].

#### *2.4. Determination of the Reproductive System*

#### 2.4.1. Changes in Stigma Receptivity and Pollen Vitality with Time

A total of 50 flowers were selected at the start and end of the period of high stigma receptivity. All stamens were carefully removed using tweezers before the blooming and isolated in a bag to prevent self-pollination. Fresh pollen grains were collected from other flowers. Flowers within 1 day to 5 days after pollination were isolated in a bag. Ten replicates were carried out during each flowering period.

A total of 50 flowers were selected at the start and end periods of pollen vitality. All stamens were carefully removed using tweezers before blooming. At the flowering day, the stigmas were pollinated with pollen from other flowers on the day of anther dehiscence and at 1 day to 5 days after dehiscence. Then, the stigmas were subjected to bag isolation. Ten repetitions were carried out for each pollen sample. Pollen germination rate and stigma receptivity were calculated according to the fruit setting rate during the preliminary stage. An ANOVA was used to compare the differences between pollen germination rate and stigma receptivity at different periods.

#### 2.4.2. Artificial Pollination and Bagging Experiments

During the bud stage, one strain and 280 buds of *S.* × *hainanensis* were selected for seven pollination treatments: A, natural control (natural hybrid); B, bagging with castration without pollination (apomixia); C, net isolation with castration (anemophily pollination); D, no bagging; E, no bagging with castration (bagging before flowering and pollen from the same flower was given after flowering); F, artificial geitonogamy (mutual pollination of No. 1 and No. 10, No. 2 and No. 9, and so on for 10 buds); and G, artificial xenogamy (pollen source was from another strain of *S.* × *hainanensis* 100 m away from the experiment strain). The E, F, and G treatments involved artificial pollination, whereas A, B, C, and D were performed for comparison. The fruit dropping time, number of fruits, fruiting rate, and seed setting rate of single fruits were determined.

#### *2.5. Seed Germination and Seedling Survival Experiment*
