Sexual reproduction in flowering plants depends on successful transfer of pollen that encloses male gametes from anthers to pistils. The pollen tube that carries male gametes must find its way to the ovule, which houses the female gametes, to achieve fertilization. While much attention has been paid to the pollination process in attempts to understand flower diversity and plant–pollinator interactions, a recent finding in arrowhead Sagittaria (Alismataceae) – a mechanism of reallocation of pollen tubes described by Wang et al. (2002) – highlights just how little we know about the fate of pollen in postpollination events. Nevertheless, it does appear that unusual, diverse growth pathways of pollen tubes may, potentially, provide a supplementary means of reproductive assurance under conditions of unpredictable pollination.
Postpollination events in Sagittaria
When a pollen grain lands on a compatible stigmatic surface, it germinates and extrudes a pollen tube. In general, the pollen tube elongates within the transmitting tissue in the style, eventually reaching the ovary, where it enters an ovule and penetrates the embryo sac, and then releases the sperm cells for fertilization (Cheung, 1995). Pollen tube growth has been used as a model system to study signal transduction in plants (Cheung & Wu, 2001; Herrero, 2001; Higashiyama et al., 2001; Wheeler et al., 2001). Once an ovule has been fertilized, the other pollen tubes usually stop growing towards it (Cheung & Wu, 2001). This pathway of pollen tube growth characterizes almost all observed plant species, but exceptions do exist. Some redundant pollen tubes reaching an ovary in the apocarpous gynoecium of Sagittaria potamogetifolia could grow through the base of the ovary and the receptacle tissue into adjacent unfertilized ovules (Fig. 1). Wang et al. (2002) harvested more than 10 achenes after applying approx. 30 pollen grains to one stigma of this aquatic monoecious herb in which one pistil has one ovule, confirming that pollen tubes enter neighbouring ovules when pollen is deposited on a single stigma. The first report of such intercarpellary growth of pollen tubes was in Illicium (Illiciaceae) by Williams et al. (1993), which was ignored by Wang et al. (2002). Both independent observations found that pollen tube growth between ovules in apocarpous angiosperms enhanced reproduction. This strategy of unusual pollen tube growth may provide supplementary reproductive assurance in plants that experience inadequate pollination.
Seed production is often limited by pollen because of the stochastic nature of pollinator services (Burd, 1994). Plants have developed diverse strategies to enhance pollination, such as prolongation of floral longevity, or increased attractiveness to pollinators by floral advertisements and rewards. Recent studies have demonstrated that diverse floral designs function principally to facilitate effective pollen transfer (Barrett, 2002). However, pollen is usually deposited unequally on the multiple stigmas of a gynoecium (Carr & Carr, 1961). For example, the pollen load ranged from 0 to 60 grains per stigma in apocarpous Liriodendron chinense (Magnoliaceae) in which some unpollinated pistils do not set seeds (Huang & Guo, 2002). The development of syncarpy with associated stigmas is a major innovation in angiosperms (Mulcahy, 1979; Endress, 1982), which may permit pollen tubes to cross between carpels and increase pollen competition during pollen tube growth in the pistils (Carr & Carr, 1961; Mulcahy, 1979; Endress, 1982; Williams et al., 1993; Armbruster et al., 2002). An early observation in Daucus carota (Apiaceae) showed that pollen tubes growing through either style could cross over and eventually fertilize either ovule (Borthwick, 1931). This phenomenon was observed in another genus, Lomatium (Apiaceae), and seems common in syncarpous flowers with separate styles (M. Schlessman, unpublished). The main advantage of fused carpels relates to offspring quality, increasing the intensity of pollen competition (Mulcahy, 1979; Endress, 1982; Armbruster et al., 2002). Pollen tube growth involving a long pathway has been observed in Dalechampia (Euphorbiaceae), in which species have expanded stigmatic surfaces. When pollen grains land on the lateral stigmatic surfaces, pollen tubes grow first to the stylar lip, bend 180 degrees, and then grow through the style to the ovules (Armbruster et al., 1995). Despite these potential advantages and the prevalence of syncarpy, reverse transitions from syncarpy to apocarpy do occur in angiosperms (Stebbins, 1974; Armbruster et al., 2002). Theoretical analyses suggest that the repeated evolution of fused carpels is influenced by pollination dynamics (Armbruster et al., 2002). In addition, a comparative analysis suggests that polycarpic plants seem no more likely to be pollen limited than monocarpic plants (Larson & Barrett, 2000).
It will be interesting to know how many angiosperms show such intercarpellary growth of pollen tubes, and how this phenomenon has evolved. Given that this pollen tube behaviour promotes fertility under conditions of unpredictable pollination, why do many plants seem not to employ this mechanism when pollen limits plant fecundity in various taxa? An alternative solution to the problem of inadequate pollination in plants is the shift to self-pollination (Baker, 1955; Wyatt, 1988; Schoen et al., 1996), which may have occurred in the genus Sagittaria, which is basically monoecious. The retention of functional stamens within perfect flowers of S. guyanensis, an andromonoecious species, could be selected for to allow self-fertilization in cases of inadequate pollination (Huang, 2003). This poses an interesting evolutionary question: how have the different strategies to diminish pollen limitation evolved in Sagittaria?
In cleistogamous flowers of the Malpighiaceae pollen tubes reach ovules by growing through the filament into the receptacle from the indehiscent anther (Anderson, 1980). Many chasmogamous flowers experiencing loss of pollinators in unfavorable habitats may adopt this means to achieve fertilization, although this has been little studied. For example, in the underwater pollen tubes of the insect-pollinated Ranalisma rostratum (Alismataceae) achieve fertilizations by growing from indehiscent anthers to reach stigma surfaces (Wang et al., 1993). Another unusual pattern of pollen tube growth was observed in the monoecious Callitriche (Callitrichaceae), an aquatic genus. In underwater conditions, pollen grains germinate in the indehiscent staminate flower. Pollen tubes grow down the filament, and through vegetative tissue across to the pistillate flower, and enter the ovary from the base (Philbrick, 1984).
These observations of unusual pollen tube growth in both dicotyledons and monocotyledons, although largely unexplored, suggest diverse means of achieving reproductive assurance during pollination. Pollen tubes growing through various different tissues to target ovules also provide natural cases to support the recent experimental observation that pollen tube growth is guided by a signal derived from the synergid cell of unfertilized ovules (Cheung & Wu, 2001; Higashiyama et al., 2001).
The author wishes to thank Mark Schlessman for providing unpublished data and, as well as Amots Dafni, for discussion; Spencer Barrett, Scott Armbruster and Lynda Delph for their valuable comments; Sarah Corbet for correcting English and providing helpful suggestions on an earlier draft of the manuscript; and three anonymous reviewers for their improvements to the manuscript. The research was supported by the National Science Foundation of China (Grant no. 30070054).