Vaccinium corymbosum Berkeley study area in Jian City, Tonghua City, Jilin Province, China (126°18’ E, 41°12’ N), and the open ground and greenhouse were experimented in the same topography, landscape and at the same altitude of 108-1337 m to ensure that the soil condition, climate, and other environmental factors were consistent with the population area of about 500 m². The overall terrain of the study area is flat, and it belongs to the continental climate in the northern temperate zone, with a large diurnal temperature difference. There is a large temperature difference, with an average annual rainfall of 600-700 mm, an average annual temperature of 5.6°C ( Figure 1 ), an average annual frost-free period of 150 ± 10 d, and an average annual greenhouse temperature of 23°C and relative humidity of 56%. Experimental materials were selected from 100 plants of ‘Northland’ 5-years-old the open ground and greenhouse, respectively.

Data were organized using Excel 2018, and the data of corolla, filament, anther and stigma length, pollen number, pollen viability, visiting frequency, nectar secretion, and pollination experiment were statistically analyzed and compared differently using SPSS 19.0 statistical software. Differences in pollen grain number, pollen viability, visiting frequency, nectar secretion, and pollination experiments were compared using independent samples t-test (p < 0.05) in the two habitats, and differences in pollination experiments were compared using one-way ANOVA (p < 0.05). Experimental data are expressed as mean ± standard error. All graphs were produced using Origin 2021 software.

In the open ground and greenhouse, ‘Northland’ had a single white inverted bell-shaped flower with 10 stamens and 1 pistil. The single-flower opening process was divided into four periods based on the size of the corolla opening ( Figure 2 ), Phase I-corolla closed, Phase II-corolla slightly opened 1-3 mm (measured by corolla opening diameter), Phase III - corolla fully opened, 4-8 mm, and Phase IV-corolla with filament anthers detached together. The four periods is shown in differences in anther morphology, corolla morphology, anther color, and stigma color; the duration of single-flower opening was 13 ± 2 and 10 ± 2 d, with single-flower opening in the open ground being prolonged by about 3 d compared with the greenhouse; the corolla depth, filament, anther and stigma lengths within the greenhouse population were greater than those of the open ground population, with significant differences between filament and anther lengths within the two populations (p < 0.05) and no significant differences between corolla and stigma lengths (p > 0.05) ( Table 1 ).

The number of pollen grains in single flower of ‘Northland’ was 19000 ± 2550 and 24000 ± 1821 when the anthers were not dispersed, in the open ground and greenhouse. The number of pollen grains per flower remaining in the anthers at pre-flowering and peak flowering was highly significantly different (p < 0.01) in the open ground and greenhouse, and the number of pollen grains remaining in the anthers at the early flowering period was significantly different (p < 0.05) ( Figure 3 ), and the number of pollen grains remaining in the anthers at each period was 12,600 ± 597 grains, 13,500 ± 716 grains, 6560 ± 216 grains and 7830 ± 155 grains. The number of pollen grains in inflorescence remaining in the anthers was significantly different (p < 0.05) ( Figure 3 ), and the number of pollen grains remaining in the anthers at each period was 2680 ± 155 grains, 3900 ± 347 grains, 1560 ± 131 grains, 2470 ± 182 grains, 840 ± 54 grains and 1260 ± 176 grains. During flowering, the anthers were porous dehiscence, and the pollen grains inside the anthers were gradually released through the split anther top holes above the anthers, and the pollen dispersal of the anthers was in the form of gradual pollen presentation. Scanning electron microscopy showed that the anther top of ‘Northland’ formed two tubular structures connected with four dispensaries, and the epidermal cells of the anther were papillate. The stigma surface had a large number of papilla cells. The pollen surface was irregularly crinkled and corrugated, and the pollen grains were tetrads of complex pollen ( Figure 4 ).

In the open ground and greenhouse, ‘Northland’ had the highest pollen viability at the early flowering, 86 ± 2.9% and 75 ± 2.7% ( Figure 5 ); and there was a highly significant difference in pollen viability between open ground and greenhouse over the three time periods (p < 0.01). The stigma receptivity in the open ground and greenhouse was shown to be stronger in the pre-flowering period, strongest in the early flowering period, reaching a relative maximum, decreasing in the peak flowering period, and weakest at the end of the flowering period in all four periods. With in the open ground and greenhouse, the nectar accumulation of ‘Northland’ was the highest at the peak of flowering at 22 ± 2.5 μL and 18 ± 2.2 μL, respectively, with significant differences between nectar accumulation at the pre-flowering stage (p < 0.05), and highly significant differences between nectar accumulation at the early flowering stage and at the end of flowering (p < 0.01), and no difference existed during the peak period (p > 0.05).

The results of the study showed that nectar accumulation showed highly significant correlation (p < 0.01) with the visiting frequency under open ground and greenhouse at the early flowering, pollen grains number and visiting frequency showed significant correlation (p < 0.05) under greenhouse; nectar accumulation showed significant correlation (p < 0.05) with the visiting frequency under open ground and greenhouse at the peak flowering, nectar accumulation and pollen grains number under open ground (p < 0.05), under the open ground and greenhouse, nectar accumulation and pollen grains number showed significant correlation (p < 0.05), under the open ground and greenhouse, nectar accumulation and pollen grains number showed significant correlation (p < 0.01), under the open ground and greenhouse, nectar accumulation and pollen grains number showed highly significant correlation(p < 0.01) ( Table 4 ).

In the open ground and greenhouse, there were significant differences (p < 0.05) between the fruiting rates of natural pollination and different plants cross-pollination, significant differences (p < 0.05) between the same plant cross-pollination and different plants cross-pollination, the fruiting rates of different plants cross-pollination were significantly higher than that of the same plant cross-pollination, and there was a highly significant difference (p < 0.01) between non-emasculated bagging and different plants cross-pollination ( Table 5 ). Therefore, the type of breeding system was the same in the open ground and greenhouse, with predominantly outcrossing pollination, partial self-pollination, and the need for pollinators.

Fruit size and seed thousand-grain weight at different locations on the branches within different populations showed the same pattern of middle > top > base. In the open ground and greenhouse, there were no significant differences in fruit longitudinal diameter at the top and middle of branches (p > 0.05), significant differences in fruit longitudinal diameter at the base of branches (p < 0.05), and no significant differences in fruit transverse diameter at the top, middle and base of branches (p > 0.05). There was no significant difference in the thousand seed weight of fruit seeds at the top (p > 0.05), significant differences in the thousand seed weight of fruit seeds at the middle and base of branches (p < 0.05) in the open ground and greenhouse ( Table 6 ).

The study showed that the sugar content was 9.62 ± 0.14% in the open ground and 8.75 ± 0.22% in the greenhouse, which was significantly higher than that of the greenhouse population (p < 0.05); the acid content was 2.35 ± 0.18 g/kg in the open ground and 3.11 ± 0.14 g/kg in the greenhouse, which was significantly lower than that of the greenhouse(p < 0.05); and the VC content was 2.52 ± 0.21 mg/100g and 2.37 ± 0.17 mg/100g in open ground and greenhouse with no significant difference (p > 0.05) ( Table 7 ).

In this study, we found that in both open ground and greenhouse, the pollen dispersal of anthers in ‘Northland’ was a gradual mechanism, by contracting and squeezing the anthers, so that the pollen was released sequentially through the anther porous dehiscence and was removed by the pollinator one by one. This anther dehiscence can effectively avoid pollinators to take away too much pollen at a time, reduce pollen loss, so that pollinators can only get a small amount of pollen in a visiting, prolonging the pollen presentation time, improve pollination efficiency, and further promote the success of plant reproduction. Gradual pollen presentation is one of the typical floral characteristics of plants to improve staminate fitness. The interaction between the process of gradual pollen presentation and its pollinators enables accurate and efficient transfer of pollen to different plants stigmas, which not only improves staminate fitness, but also guarantees the success of different plants cross-pollination (Yang, 2016; De et al., 2015; He and Liu, 2023). Pollinator plant Bomareaacutifolia Snow A A et al. improves pollen transfer rate by gradual presentation of six stamens opening sequentially (Snow and Grove, 1995); in pollinator-deficient environments Gageanigra Zheng R Q et al. prolongs pollen dispersal time and reduces pollen wastage by gradual presentation (Snow, 1986; Zheng et al., 2016a). Pollen presentation is mainly divided into gradual pollen presentation and simultaneous pollen presentation (Thomson et al., 1994). In pollinator-scarce conditions, plants prefer simultaneous pollen presentation, while in pollinator-rich environments, plants strictly limit pollen output (gradual pollen presentation) so that only a small amount of pollen is available to each pollinator and more pollinators are involved in the pollination, thus increasing staminate fitness (Thomson et al., 1994, Thomson et al., 2000). Gradual pollen presentation not only improves the efficiency of pollen transfer, but also increases the plant’s ability to adapt to environmental changes to ensure the success of pistillate reproduction (Howell et al., 1993; Lu and Tan, 2007; He et al., 2024).

The floral characteristics include flower opening characteristics, pollen viability, stigma receptivity, nectar secretion, etc (Wang and Tan, 2011; Yang et al., 2019). The opening characteristics include flower structure, flower color, flower shape, etc. (Ji et al., 2023; He et al., 2010; Jing et al., 2023). Flowering patterns influence pollen dispersal patterns by attracting pollinators, which in turn affects the mating system of the plant (Tang and Han, 2007). It has been pointed out that the morphological characteristics of flower appearance of different plants are related to their reproductive and developmental characteristics as well as to the pollination of flowers (Ruan and Jiang, 2006). In this study, we showed that the single flowers of ‘Northland’ are inverted bell-shaped, which not only prevents rain from washing away the stamens and pistils and prolongs pollen viability and stigma pollination, but also reduces the negative effects of nectar dilution by rain and pollen removal by wind, and ensures the success of pollination. The flower characters of plants not only change according to the pollination status of pollinators, but are also closely related to environmental conditions (light, temperature, and humidity, etc.). Most plant populations often show more differences in floral characteristics across habitats (Ma, 2014; Tang, 2016; Huang et al., 2019). For example, the flower structures of Iris ruthenica show differences between under the forest and forest edge habitats (Wang et al., 2021). In this study, we measured each basic structure of the flower parts of ‘Northland’, and the corolla, anthers, filament length, and stigmas were longer in the greenhouse than in the open ground, but no significant differences existed, and the authors hypothesized that the greenhouse was more suitable for its growth due to its stable temperature and humidity factors.

The variability in the length of plant stigma receptivity and the duration of pollen viability, as well as the length of the overlap period, not only affects the pollination success rate of pollinators, but also influences the rate of self-crossing and outcrossing (Quan, 2011; Tang and Huang, 2012; Zheng et al., 2016b). Pollen viability is not only affected by genetic factors, but also by the external natural environment (Wei et al., 2006). Research showed that the floral characteristics of plants and their pollen viability affect the reproductive development of plants, and also indicated that the receptivity of the stigma is also strongly related to its reproductive development (Guan et al., 2009; Wan et al., 2011; Xiao, 2005). The pollen viability of ‘Northland’ differed significantly between the open ground and greenhouse, but the trend was the same, the pollen viability was highest at the early flowering and began to decline at the peak of flowering. Increasing pollen viability after flowering ensured that pollinators transferred the higher-viability pollen to the stigma after visiting to maximize the efficiency of staminate resource utilization, which was consistent with the results of the stigma-receptivity test. The stigma receptivity is strongest at the early flowering, and the stigma receptivity becomes weakened as the flowering time increases; pollen viability interacts with stigma receptivity at the early flowering. At the same time, ‘Northland’ secretes a large amount of nectar when it blooms, and when insects contact the stigma, pollination is completed and the chances of successful fertilization are guaranteed, which ensures that pollen adheres to the villi of pollinators at the peak of their viability, and promotes the pollination success rate of ‘Northland’.

Breeding systems are an expression of the interactions between the genetic mechanisms within the plant and the external environment, and their combination of pollinators and pollination behavior is a major factor in influencing the genetic composition and fitness of reproductive offspring, and plays an important role in determining the evolutionary pathways and characterizing variation in plants (He et al., 2023; Gran, 1881). The study of plant breeding systems helps to understand the responses of plants to environmental changes and is a necessary prerequisite for plant breeding research (Murúa et al., 2014; Wang et al., 2010). The visiting frequency of ‘Northland’ in the two habitats was correlated with the ambient temperature of the time period of the day, which was consistent with the results of the study on Iris ruthenica Wang Z Z et al. (Wang et al., 2021). The study showed in the open ground and greenhouse, the frequency of visiting increased as the temperature increased, and was highest when the daily temperature reached the highest. As the temperature decreased, the frequency of visiting also decreased, and the frequency of visiting by pollinators showed a tendency of increasing first to reach the highest and then decreasing gradually.

Plant breeding systems are an important part of plant life history and an important factor in plant evolution (Chen et al., 2023). The breeding system of plants is influenced by their own genetic factors and the external natural environment (He and Liu, 2003). Sexually reproducing plant breeding systems are categorized into self-crossing, outcrossing and various concurrent mating systems (Zhang et al., 2024). In nature, outcrossing is favorable to increase the genetic diversity of plants (Wang et al., 2024). According to the results of artificial pollination experiments, the type of breeding system of ‘Northland’ in the two habitats was partially self-crossing, outcrossing-dominated and required pollinators. Since the results of the different plants cross-pollination treatment were higher than the other treatment groups, the natural pollination and the same plant cross-pollination fruiting rates were lower than those different plants cross-pollination. Under natural conditions, the setting rate of open ground is higher than that of greenhouse, and under artificial pollination, the setting rate of greenhouse is higher than that of open ground, indicating that artificial pollination can be carried out in greenhouse to improve the setting rate.