Discussion
Our investigations demonstrated that the species H. mutabilisshowed a mixed mating system, and its sexual reproduction depended on pollinators. The floral reward (nectar) of H. mutabilis was consumed by various visitors, including four bee species and one moth species. The pollination efficiency of the moth species was relatively lower than all bee species. The body length of five pollinator species increased the amount of transferred pollen, and larger bees tended to remove more pollen from anthers and deposit more pollen on stigmas, resulting in higher pollination efficiency. However, there was no correlation between the intertegular distance and pollination efficiency of visitors in H. mutabilis . Body length may be a better index of body size than intertegular distance.
The nectar ofH. mutabilis was not displayed at the base of the petals but at the base of the ovary. So the flower structure of H. mutabilis shielded visitors with shorter proboscis. The best match between the length of the proboscis and the circumference of the nectar secretion at the base of the flower is the M. pyrrhosticta , which suggests the M. pyrrhosticta should be the most important pollinator in its native habitat (Fig. 5A). However, the investigation results found that bees were still important pollinators of H. mutabilis . Although the proboscis length does not match the circumference of the basal nectar, they can suck in nectar from the gap between the front petals. Different proboscis lengths of bees could absorb different amounts of nectar, which can improve the chances of pollen removal and pollen deposition.
Although the moth speciesM. pyrrhosticta had the longest body length among the five visitor species, its pollination efficiency was relatively lower than all bees. The interpretation of the relationship between body size and pollination efficiency may depend on the influence of other morphological traits of M. pyrrhosticta . Unlike bees, the moth without hairiness may carry a few pollen grains. Moreover, the moth has a long proboscis so that they can get nectar easily without entering deep into the corolla of flowers. In this case, the moth may have little contact with floral sex organs. In terms of four bee species, the body length of X. appendiculata was 1.8× longer than that of honey bee A. mellifera (Fig. 5B). Moreover, the pollen removal and pollen deposition of X. appendiculata were 1.9× and 2.4× more than that of A. mellifera(Fig. 6A and B), and the pollination efficiency of X. appendiculata was 1.2× more than that ofA. mellifera (Fig. 6C). These results demonstrated that larger bees removed and deposited more pollen than smaller bees in a single visit, and larger bees were more efficient pollinators in H. mutabilis . The reason was that larger bees were more likely to come into contact with floral sex organs. This finding is partly consistent with a previous study, which indicated that larger bee species deposited more pollen grains onto stigmas than smaller bees in Brassica napus (Chang et al . 2023), but they did not consider the effect of body size on pollen removal and pollination efficiency. Besides, we found that the intertegular distance was not an effective index of body size, because there was no correlation between the intertegular distance and pollination efficiency of visitors in H. mutabilis . Only the moth species was excluded, the intertegular distance of four bee species was significantly and positively correlated to pollen removal and pollen deposition. However, there were always significant positive correlations between the body length of four or five pollinator species and pollen removal, pollen deposition, and pollination efficiency. We concluded that body length may be a better index of body size than intertegular distance.
The visitation rate of bees is another important behavioral parameter. Honey bee A. mellifera was the most frequent visitor to flowers of H. mutabilis . The visitation rate of A. mellifera was 5.1× more than that of X. appendiculata , even though the pollination efficiency of A. mellifera was relatively lower. The high visitation rate of bees to flowers may increase the pollinator’s interaction with the stigma or pollen transfer, so A. melliferawas also a very efficient pollinator in H. mutabilis . Although the pollination efficiency of X. appendiculata was higher than honey bees, they visited fewer flowers than honey bees. Therefore, it is difficult to comprehensively assess which insect species contributes more pollination service to flowers in H. mutabilis , and further study is needed to investigate the pollination service of various pollinators to flowers throughout the flowering period of plants.
The flowers of H. mutabilis had a generalist pollination system and attracted various pollinator species by providing plenty of nectar rewards. Although H. mutabilis was highly self-compatible, there was pollen limitation under natural conditions. The sexual reproduction of H. mutabilis depended on pollinators. The visitation rate and pollination efficiency of various pollinator species were greatly different. The species H. mutabilis with a mixed mating system may be an adaptation to variable pollinator visiting.