Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Barriers to Sexual Reproduction in Polygonum viviparum: A Comparative Developmental Analysis of P. viviparum and P. bistortoides

Barriers to Sexual Reproduction in Polygonum viviparum: A Comparative Developmental Analysis of... Abstract Polygonum viviparum is widely distributed in arctic and alpine regions of the northern hemisphere. Fruit set has never been observed in North American populations and has been reported only very rarely in Europe. Although this species is extremely well studied, the impediments to successful fruit production are unknown. We investigated the sexual reproductive process in P. viviparum growing in the southern Colorado Rocky Mountains. For comparison, we also examined this process in the sympatric congener P. bistortoides, in which reproduction is exclusively sexual. Lack of viable fruit production in P. viviparum has no single developmental explanation; defects occur in each of the processes and structures associated with sexual reproduction studied, yet, these processes and structures also appear to function normally in at least some flowers or individuals. Development is abnormal in many ovules of P. viviparum, however, comparison with P. bistortoides shows that these abnormalities do not contribute to differences in seed production between the two species. The virtual absence of sexual reproduction in P. viviparum appears to be due largely to a low rate of fertilization and to embryo/fruit abortion. Key words: Asexual reproduction, embryo abortion, development, fruit abortion, pollen viability, Polygonaceae, Polygonum viviparum, Polygonum bistortoides, polyploidy, seed abortion, sexual reproduction. Received: 18 July 2001; Returned for revision: 12 September 2001; Accepted: 11 October 2001. INTRODUCTION The majority of flowering plants are capable of some form of asexual reproduction by means of a diverse range of structures and mechanisms. Clonal progeny may be produced by stolons, runners, rhizomes, tubers, buds on bulbs, corms and roots, layering of stems, and agamospermous seed (Grant, 1971). Despite the prevalence of asexual mechanisms of propagation, most clonal species also reproduce sexually (Stebbins, 1971; Silander, 1985; Richards, 1997) and exclusively asexual species are rare. Polygonum viviparum L. may be one of the few species that relies solely upon vegetative reproduction throughout much of its range (Fryxell, 1957; Callaghan, 1973; Engell, 1973; Petersen, 1981). Polygonum viviparum L. [= Bistorta vivipara (L.) S. Gray] (Polygonaceae) is widely distributed in both arctic and alpine regions of the northern hemisphere (Petersen, 1981; Lawet al., 1983; Callaghan and Emanuelsson, 1985; Bauert, 1993; Wookeyet al., 1994). Although most individuals of P. viviparum flower profusely, fruit set has never been observed in alpine populations of North America and has been reported only very rarely in Europe (Bauert, 1993). P. viviparum is a very common and important component of tundra plant communities and has been the subject of numerous ecological studies (Callaghan, 1973; Petersen, 1981; Bauert, 1993; Crawfordet al., 1993; Wookeyet al., 1994 and references cited therein), and most workers comment on the lack of seed set. The only published reports of sexual reproduction in P. viviparum include: Porsild and Porsild (1920) in Greenland; Söyrinki (1989) in Scandinavia; Bliss (1959) in the subalpine of the Rocky Mountains of Wyoming; Murray and Miller (1982) in interior Alaska; and Bauert (1993) in the central Swiss alps. Sexual reproduction is clearly a rare event in this species. In contrast to the rarity of fruit set by P. viviparum, asexual reproduction occurs readily via the production of bulbils, vegetative axillary buds borne within inflorescences (Troll, 1937; Engell, 1973; Diggle, 1997). After bulbils are dispersed, they germinate and grow under moist conditions (Callaghan, 1973; Engell, 1973, 1978; Petersen, 1981) and establish new, physiologically independent plants that are genetically identical to the parent. The impediments to successful fruit set in P. viviparum are unknown. Any of the many stages of development of male and female gametophytes, gametes, embryo and endosperm, or the processes of pollination and fertilization, might pose a barrier to sexual production of progeny in this species. Preliminary embryological studies of P. viviparum have been published (Edman, 1929;Schnarf, 1931; Engell, 1973); however, the causes of sterility were not identified. In order to explain the lack of fruit production, we investigated the sexual reproductive process of P. viviparum growing in the alpine tundra of the southern Colorado Rocky Mountains. Our study focused on ovule and female gametophyte development, pollen viability and early embryo/seed development. For comparison, we also examined these processes in the sympatric congener, P. bistortoides. P. bistortoides is very common at the study site, produces large numbers of seed and has no means of vegetative reproduction. MATERIALS AND METHODS Study site Study sites were located in the alpine tundra of Niwot Ridge (elevation 3750 m) in the Front Range of the eastern Colorado Rocky Mountains (40°03′N, 105°35′W). Niwot Ridge is managed cooperatively by the University of Colorado Mountain Research Station and the United States Forest Service as an ecological reserve and is the site of the Niwot Ridge Long‐Term Ecological Research Program. Polygonum viviparum and P. bistortoides were sampled haphazardly in a dry meadow community. Species description Polygonum viviparum is widely distributed in both arctic and alpine regions of the northern hemisphere (Petersen, 1981; Lawet al., 1983; Callaghan and Emanuelsson, 1985; Bauert, 1993; Wookeyet al., 1994). Within the study area, P. viviparum is common in all plant communities except in areas of very late melting snow (May and Weber, 1982). Polygonum bistortoides Pursh. (= Bistorta bistortoides (Pursh) Small), is sympatric with P. viviparum in the southern Rocky Mountains and is also widely distributed across all alpine communities, including snowbeds (Walkeret al., 1995). Both species are herbaceous perennials. The main axis of each is a monopodial unbranched rhizome that grows plagiotropically, with little internode elongation, below the soil surface. Rhizomes mature several preformed leaves and axillary inflorescences each year (Diggle, 1997; P.K.D., pers. obs.). P. bistortoides is larger than P. viviparum in most respects: rhizomes are longer and of greater diameter; leaves are longer, wider and more numerous; and inflorescences are taller. The reproductive characters of the two species are very similar. Inflorescences are condensed panicles bearing numerous white flowers. Flowers are small, approx. 3 mm in length. The perianth consists of five white tepals [Laubengayer (1937); Ronse Decraene and Akeroyd (1988), review interpretations of the perianth of the polygonaceous flower]. Stamen number is variable and the gynoecium consists of a three‐angled, uniovulate ovary with three styles and stigmas. The two species differ in one critical aspect of inflorescence structure (and consequently, life history): P. viviparum may bear bulbils in the proximal region of the inflorescence whereas P. bistortoides bears only flowers and reproduces exclusively by seed (Mooney, 1963). Both species flower in July and mature fruits of P. bistortoides are dispersed in August. P. viviparum is polyploid with 2n = 96, x = 12 (Löve and Löve, 1948, 1974, 1975; Wcislo, 1967; Engell, 1973; Löve, 1988). P. bistortoides is diploid with 2n = 24 (Mooney, 1963). Ovule and female gametophyte development All floral organs are initiated in the growing season prior to flower maturation. The preformed inflorescence‐bearing floral primordium undergoes dormancy below ground. Development resumes in the spring and the inflorescence begins to emerge above ground. Sampling of developing inflorescences began when the tips were first visible above the soil surface (first week in July, 1995) and continued until flowers reached anthesis. All inflorescences were fixed in formalin–acetic acid–alcohol (Berlyn and Miksche, 1976). Flowers and flower buds were sampled from basal positions of the short (two or three flowered) branches of the paniculate inflorescences. The gynoecium was dissected from mature flowers and large buds. Small buds were processed whole. Gynoecia and buds were dehydrated to 95% EtOH, embedded in JB4 methacrylate resin, serially sectioned on a Microm microtome at 4 µm, stained with toluidine blue (O’Brien and McCully, 1981) and observed with a Zeiss Axioskop. Pollen viability Mature inflorescences were collected, placed in sealed microcentrifuge tubes, and transported to the laboratory on ice. In 1992 (year 1), inflorescences from 150 individuals of P. viviparum were collected. In 1993 (year 2), inflorescences were collected from 20 P. viviparum and 47 P. bistortoides individuals. All flowers were examined within 12 h of collection. The FCR (fluorochromatic) test (Heslop‐Harrison and Heslop‐Harrison, 1970; Shivannaet al., 1991) was used to evaluate pollen viability. [See Thomsonet al. (1994) and Stoneet al. (1995) for discussions of the reliability of FCR as a measure of pollen viability.] The FCR test may give false negative results due to pollen drying (Heslop‐Harrisonet al., 1984; Shivana and Johri, 1985; Knoxet al., 1986; Thomsonet al., 1994). To prevent desiccation, flowers were kept in the humid atmosphere of the microfuge tubes until the test was performed. Three to five flowers per inflorescence were examined. All anthers were removed from a single flower and were placed on a glass slide. Pollen grains were teased from the anthers with fine forceps. Each slide was flooded with a solution of fluorecine diacetate (FDA; 0·02 g FDA/10 ml acetone in 2 ml of 10 % sucrose) for 5–10 min. Slides were examined with a Zeiss Axioskop equipped for epifluorescence. Brightfield was used to count all pollen grains present. The same slide was then examined under ultraviolet light. Stainability was calculated as the number of fluorescing grains divided by the total number of grains present. The stainability of all flowers from the same inflorescence was averaged, and this mean was used in subsequent statistical analyses. Pollen stainability was analysed with StatView for Macintosh. Stainability data were arcsine square root transformed prior to analysis. RESULTS Pre‐fertilization ovule development Ovule development is nearly identical in the two species, and the description that follows applies to both with exceptions as noted. Flowers of both species are uniovulate, and initiation of the ovary and nucellus had already occurred in all material collected. Inflorescence and floral primordia are preformed, and these events probably occur in the year preceding inflorescence maturation. In the youngest floral primordia sampled, the nucellus was already enclosed by the ovary wall (Fig. 1A, B). The large nucellus has an epidermal layer and one parietal layer. In ovules of both species, the epidermis of the nucellus is often elaborated into an apical ‘beak’ (e.g. Fig. 1C, D, F). The species differ in the number of megasporocytes per nucellus. In P. bistortoides, each nucellus develops one (rarely) to several megasporocytes (Fig. 1C). In P. viviparum there is typically only one megasporocyte per nucellus (Fig. 1D) although occasionally there are two. In both species, as megasporocyte differentiation occurs, the nucellus elongates, and the inner integument of the ovule is initiated (Fig. 1C, D). The megasporocyte (P. viviparum) or megasporocytes (P. bistortoides) undergo meiosis (Figs 1E and 2B) and each forms a linear tetrad (Fig. 2B). The chalazal‐most spore of each tetrad becomes the functional megaspore while the remaining three spores degenerate (Fig. 2C, D). In P. bistortoides there may be several functional megaspores per ovule, each the product of a separate meiotic event (Fig. 2C). A hypostase begins to differentiate at about the time of spore formation (Fig. 2C, D). At, or just prior to meiosis of the megasporocyte(s), the outer integument of the ovule is initiated (Figs 1E and 2B). As megagametophyte development begins, the integuments elongate and enclose the nucellus. The micropylar region of the inner integument expands, becoming several cell layers thick, and forms the micropyle. The outer integument remains two cell layers thick throughout development and does not contribute to the micropyle. One (P. viviparum) to several (P. bistortoides) gametophytes begin to develop within the nucellus. During this time, the gametophytes elongate only slightly. Nucellar elongation is primarily in the chalazal region; there are no further cell divisions in the parietal layers of the apical region of the nucellus, and the existing layer may be crushed. At the completion of mitosis to yield eight nuclei in seven cells, the female gametophyte elongates dramatically. The chalazal portion of the mature gametophyte is ‘wedge’ shaped, appearing very narrow in one dimension, and flattened in the perpendicular plane. No ovules of either species were observed to contain more than one mature gametophyte. The mature ovules of both species are orthotropic and crassinucellate. The polar nuclei of the female gametophyte usually fuse prior to fertilization so that the mature gametophyte comprises seven nuclei in seven cells (Fig. 3A–C). The two synergids are densely cytoplasmic and each has a distinct filiform apparatus. The egg cell is vacuolate and the large nucleus is located in the chalazal end of the cell (Figs 3A, C). The fusion nucleus, formed from the two polar nuclei, is located in the micropylar end of the gametophyte, either adjacent to the egg or near the wall of the central cell (Fig. 3B, C). The central cell is large and vacuolate. The antipodal cells are much smaller than the other cells of the gametophyte and typically persist beyond anthesis. The hypostase of mature ovules is pronounced (Fig. 3C) and the blue‐green staining with toluidine blue O suggests that the cell walls contain lignin. Development of abortive ovules Some ovules of both species fail to complete development. The relative frequency of such aborted development can be inferred from analysis of mature ovules. If those ovules that contain a mature female gametophyte and those containing a zygote or embryo are classified as ‘functional’, then 30 of 39 (76·9 %) mature P. bistortoides ovules were apparently functional, whereas nine (23·1 %) contained obviously malformed gametophytes (Table 1). Similarly, 31 of 42 (73·8 %) mature P. viviparum ovules appeared functional and 11 (26·2 %) were malformed. The frequency of functional ovules did not differ significantly between the two species (χ2 = 0·226, P > 0·1). (Clearly, the ovule becomes an immature seed upon fertilization, however, the term ‘ovule’ is retained here for simplicity.) Developmental analyses show that similar types of abnormalities occur in both species and that cessation of ovule development is not stage‐specific. Some aborted ovules of each species appear to be at the megasporocyte or megaspore stage, i.e. the nucellus is small (Fig. 4B) or has not yet been enclosed by the two integuments (Fig. 4A), yet there is no evidence of a megasporocyte or spore (e.g. compare Fig. 4A with Fig. 2A, C). Development of ovules can also cease at later stages (Fig. 4C–F). In some, the nucellus had begun to elongate but did not contain evidence of a female gametophyte. In other ovules, some, but not all, cells of a mature female gametophyte are present but the nucellus had not elongated to mature dimensions (Fig. 4C, D). In still others, the integuments and nucellus are indistinguishable from those of ovules containing a normal female gametophyte except that the nucellus is solid tissue and the central area is occupied by collapsed cells (Fig. 4E, F). It appears that a female gametophyte began to develop in these ovules and then collapsed. Post‐fertilization ovule development Although the frequency of apparently functional ovules did not differ between the species, a significantly greater proportion of gametophytes of P. bistortoides had been fertilized (Table 1). Sixteen of the 30 (53·3 %) apparently functional ovules of P. bistortoides contained zygotes or embryos compared with nine of 31 (29 %) ovules of P. viviparum (χ2 = 7·353, P < 0·01). Embryo development of the two species was not studied in detail. Ovaries that contained zygotes and embryos were collected from flowers at anthesis at the same time as those containing mature female gametophytes. Presumably, fertilization and the initiation of embryo development takes place very rapidly following anthesis. The embryos were in either the filamentous or globular stage and were surrounded by free nuclear endosperm. Development of one embryo of P. viviparum had progressed to the initiation of cotyledons and the endosperm had cellularized. Mature embryos were not observed in collected material of either species. Pollen viability Average pollen stainability for P. viviparum in year 1 was 20·6 % and ranged from 0– 75 %. In approx. half (54 %) of the individuals, less than 20 % of pollen grains stained positively with FCR, and stainability was greater than 70 % in only two individuals (Fig. 5A). In year 2, average stainability for P. viviparum pollen was 17·4 % and ranged from 0·66–43 %. Stainability was less than 20 % in the majority of individuals (65 %) (Fig. 5B). Stainability of P. viviparum pollen was not significantly different between years (Mann–Whitney U = 719; P = 0·5361). Average pollen stainability for P. bistortoides in year 2 was 56·6 % with a range of 13–83 % (Fig. 5C). Percent stainability differed significantly between the species in 1993 (t = 11·9, P < 0·0001). DISCUSSION Lack of viable seed production in P. viviparum has no single developmental explanation. Defects occur in each of the processes and structures associated with sexual reproduction studied, yet these processes and structures also appear to function normally in at least some flowers or individuals. Malformed ovules are common in P. viviparum; however, comparison with P. bistortoides shows that these abnormalities do not contribute to differences in seed production. The virtual absence of sexual reproduction in P. viviparum appears to be due largely to a low rate of fertilization and to embryo/fruit abortion. Pre‐fertilization development Beginning with the studies of Strasburger (1879), the origin and development of the female gametophyte in members of the Polygonaceae have become classic examples of ‘normal’ processes (Degeon, 1918). Development of the female gametophytes of P. bistortoides and of P. viviparum is similar and conforms to reports of other species of Polygonum and close relatives (summarized in Davis, 1966; also Stevens, 1912; Degeon, 1918; Schnarf, 1931; Mahony, 1935). Seemingly unusual features such as multiple megasporocytes and the conspicuous nucellar beak observed in both P. viviparum and P. bistortoides are common among other species of Polygonum but are not ubiquitous (Davis, 1966). Interestingly, developmental abnormalities of the ovule and female gametophyte are frequently observed among the species of Polygonum that have been studied and are not associated with low fruit production. There are no previous reports of ovule and female gametophyte development for Polygonum bistortoides. In contrast, there are least three published accounts of some aspects of reproductive development in P. viviparum, each motivated by the absence of seed set in this species. Edman (1929) and Schnarf (1931) found that irregularities in the development of female gametophytes were common and that degeneration could occur at any time following meiosis. Our results show that in addition pre‐meiotic failure can occur. Engell (1973) did not note abnormalities, but was unable to examine mature ovules and questioned whether P. viviparum was capable of sexual reproduction. While previous investigators have examined some aspects of female gametophyte development in P. viviparum in the context of low seed set, this study is the first to include an analysis of a ‘control’ taxon. Comparison of the development of P. viviparum ovules with that of P. bistortoides, a sympatric congener with plentiful fruit set, yields the unexpected conclusion that, although developmental irregularities are common in ovules of P. viviparum, they are probably not responsible for the extremely low fruit production. Malformed ovules occur with the same frequency in P. viviparum and P. bistortoides and ovule development appears to abort at the same range of stages in the two species. Clearly, the dramatic differences in fruit production between the two species are not attributable to abnormalities of pre‐fertilization ovule development in P. viviparum. While the frequency of apparently functional ovules is equivalent in the two species, pollen viability differs significantly. Average pollen viability (as assessed by the FCR reaction) is far less in P. viviparum than in P. bistortoides, and in both years a large proportion of flowers had no viable pollen. Engell (1978) also found a high frequency of small pollen grains in flowers of P. viviparum collected from the Faroe Islands and suggested that these were non‐functional. Thus, low pollen viability is also characteristic of northern European populations. Fertilization Pre‐fertilization ovule development of the two species is nearly identical and apparently functional mature female gametophytes occur with equal frequency, yet zygotes and embryos were observed in only 29 % of functional ovules of P. viviparum compared with 53 % in P. bistortoides. Assuming that these are the result of fertilization (i.e. not agamospermy), then the rate of fertilization is significantly lower in P. viviparum than in P. bistortoides. The difference in fertilization rate between the two species may be attributable to the large differences in pollen viability. If pollen is not available in excess, then the probability of fertilization occurring should be proportional to the frequency of viable pollen grains among potential donors. Fertilization of P. viviparum ovules also may be limited by pollinator service but pollen removal and deposition rates were not examined. A third potential explanation for differences in fertilization is self‐incompatibility. Because of the prevalence of asexual reproduction in P. viviparum, many potential pollen donors could be genetically identical to the recipient (e.g. Weis and Hermanutz, 1993; Tangmitcharoen and Owens, 1997; Negrón‐Ortiz, 1998). If the species is self‐incompatible, these pollinations would not lead to successful fertilization. Two facts argue against this explanation: (1) although the compatibility system of P. viviparum is unknown, this species is polyploid and polyploidy tends to disrupt self‐incompatibility (de Nettancourt, 1977; Richards, 1997; Miller and Venable, 2000); and (2) a genetic analysis of the study population found relatively high levels of genotypic diversity (Diggleet al., 1997), therefore, even if the species is self‐incompatible, compatible genotypes should be available within populations. Thus, low pollen viability or limited pollination are more likely explanations of the differences in fertilization between the two species. Although the frequency of ovules with zygotes or embryos is lower in P. viviparum than in P. bistortoides, the observed fertilization rate does not explain the apparent absence of fruit production. Embryo development was initiated in 29 % of ovules of P. viviparum and flower number per inflorescence at the study site ranged from 0 (some inflorescences bear only bulbils) to 65, with a mean of 16 (K. Stitt, unpubl. res.). If all fertilized ovules matured into seeds, there should have been a detectable number of fruit at the field site. Post‐fertilization development Although 29 % of P. viviparum ovules contained zygotes or embryos, we observed neither a mature embryo in sectioned material nor a mature fruit in the field; all flowers abscised from inflorescences within a few days of anthesis. Clearly, most fertilized ovules abort after initiation of embryo development. Embryo/seed abortion followed by abscission must occur rapidly because we did not observe any apparently malformed (dying) embryos or endosperms in sectioned material. Thus, in addition to a lower rate of fertilization compared with P. bistortoides, abortion of new sporophytes also must be extremely common in P. viviparum. Abortion of developing seeds and fruits is quite common among angiosperms (reviewed in Stephenson, 1981; Lee, 1988; Sedgeley and Griffen, 1989; Diggle, 1995). Failure of post‐fertilization development has been ascribed to a variety of factors including competition among seeds or fruits for nutrients (Stephenson, 1981; Lee, 1988; Diggle, 1995 and references therein; Shuraki and Sedgley, 1996), maternal selection among embryos of varying genetic quality (reviewed in Willson and Burley, 1983; Bawa and Webb, 1984; Guth and Weller, 1986; Briggset al., 1987; Marshall and Folsom, 1991; Akhalkatsiet al., 1999), late acting self‐incompatibility (Seavery and Bawa, 1986), and genetic load or other genetic abnormalities (Brink and Cooper, 1940, 1941, 1947; Cooper and Brink, 1940; Weins, 1984, 1987; Charlesworth and Charlesworth, 1987; Charlesworth, 1989; Andersson, 1993). Competition for nutrients and maternal selection among embryos implies that there are some ‘winners’, i.e. although abortion occurs some fruit should mature on each inflorescence or individual. This is clearly not the case for P. viviparum. Late acting self‐incompatibility also is not likely to occur in P. viviparum for the reasons discussed above for pre‐fertilization self‐incompatibility (see discussion under ‘fertilization’). Given that pollen development is affected so drastically in P. viviparum, and that fruit production is uniformly negligible among individuals, among years and among sites, genetic abnormalities are a likely cause of embryo abortion. Whereas P. bistortoides is diploid, P. viviparum is reported to be an octoploid (2n = 96, x = 12; Löve and Löve, 1948, 1974, 1975; Wcislo, 1967; Engell, 1973; Löve, 1988). Full or partial sterility is not uncommon among polyploids and is often attributed to irregular meiosis (e.g. Stebbins, 1971; Richards, 1997 and references therein). Meiotic aberrations would result in low pollen viability as observed in P. viviparum and could also explain embryo abortions. Even when gametes or gametophytes appear functional, they may have an unbalanced chromosome content and fertilization would produce zygotes that ultimately abort. The association between polyploidy and meiotic irregularity, however, is not well established (P. Soltis, pers. comm.) and many polyploids have normal chromosome pairing at meiosis (e.g. Grant, 1971; Lawet al., 1983). Further examination of meiosis in P. viviparum is warranted. Low pollen viability and embryo abortion also may be due to accumulation of mutations in these very long lived clonal organisms (Muller’s ratchet; Muller, 1964; Gabrielet al., 1993; Andersson and Hughes, 1996; Eckertet al., 1999). Many mutations have been identified in crop and model species that specifically affect meiosis (e.g. Kaul and Murthy, 1985). In addition, other mutations affect pollen viability (e.g. Kennell and Horner, 1985; Kaul, 1988; Benaventeet al., 1989; Chaudhury, 1993), ovule development (Reiser and Fischer, 1993; Christensenet al., 1998; Schneitz, 1999; Yang and Sundareasan, 2000) or embryo development (reviewed in Mayeret al., 1991; Uweret al., 1998; Albertet al., 1999; Heckelet al., 1999). Such mutations can be perpetuated by vegetative reproduction, and could accumulate over the lifespan of a genet, particularly if the mutations have no negative pleiotropic effects on survival (Eckertet al., 1999). In addition, embryo abortion could be the result of genetic load if the new sporophytes are products of self‐fertilization (Weins, 1984, 1987; Klekowski, 1988). The absence of any clear developmental correlate of embryo abortion is not unusual. There has been considerable developmental analysis of seed and fruit abortion, particularly in economically important species (reviewed in Brink and Cooper, 1940, 1941; Sedgeley and Griffen, 1989; Fernando and Cass, 1997). Most commonly, abortion is not stage‐specific and no visible evidence of the cause of abortion can be identified (e.g. Persica: Sedgeley, 1980; Pistacia: Shuraki and Sedgeley, 1996; Butomus: Fernando and Cass, 1997; Melilotus: Akhalkatsiet al., 1999; many tree crops including cherry, citrus, teak, mango and plum: Stephenson, 1981). In other cases, clear developmental abnormalities precede abortion. These include proliferation or death of the endosperm (Medicago: Cooperet al., 1937; Oxalis: Guth and Weller, 1986) proliferation of integuments (Asclepias: Moore, 1946), or pronounced development of the hypostase (Pisum: Briggset al., 1987). Although high rates of fruit abortion occur regularly in all of these species, the ultimate cause of abortion was rarely identified conclusively. Why is fruit production so low in P. viviparum? Although irregular ovule development is common in P. viviparum, it is no more frequent in this species than in the sympatric P. bistortoides. In fact, multiple developmental abnormalities are reported to be common throughout the genus Polygonum and may be normal for the genus (an odd situation for the ‘type’ genus for female gametophyte development!). Abortive ovule development does decrease maximum seed production in P. viviparum, but is not responsible for its absence. In contrast to the similarity of ovule development in the two species, the incidence of abnormal pollen is far greater in P. viviparum than in P. bistortoides and may be associated with differences in fertilization rates. Even low fertilization rates, however, cannot account for the rarity of fruit production. Abortion of young sporophytes must be the final phase of reproductive development that reduces fruit maturation to undetectable levels. Although fruit set in P. viviparum has not been observed in the study population, it is likely that it does occur occasionally. Genotypic diversity in the study populations is equivalent to that of clonal species known to have regular fruit/seed maturation (Diggleet al., 1997; see also Bauert, 1993, 1996 for European populations). The most likely explanation of the existence of multiple genotypes in this population is that viable seeds are produced, albeit rarely. Theoretical models suggest that a very low rate of seedling input into established populations is sufficient to maintain genetic diversity (Soane and Watkinson, 1979; Watkinson and Powell, 1993). The causes of low fruit production in P. viviparum are not absolute, and occasionally two normal gametes may unite to form a viable new sporophyte. ACKNOWLEDGEMENTS The authors thank Tara Doughty, Rachel Kaplan, Anne Klein, Amber Moody, Mingon Macias and Kristine Stitt for laboratory assistance, and Larry Hufford for comments on the manuscript. Support was provided by NSF DEB‐9357076, Apple Computers, Inc., the Niwot Ridge Long‐Term Ecological Research Program (NSF DEB‐9211776) and the University of Colorado Mountain Research Station (NSF BIR‐9115097). View largeDownload slide Fig. 1. Early ovule development in P. bistortoides (A, C and E) and P. viviparum (B, D and F). A, B, Nucellus surrounded by ovary wall. A megasporocyte is evident in B. C, D, Megasporocyte(s) entering prophase of meiosis I. The inner integument has been initiated. E, F, Megasporocyte(s) in meiosis, the outer integument has been initiated. B, Nucellar beak; ii, inner integument; M, megasporocyte; N, nucellus; O, ovary wall; oi, outer integument. Bars = 25 µm. View largeDownload slide Fig. 1. Early ovule development in P. bistortoides (A, C and E) and P. viviparum (B, D and F). A, B, Nucellus surrounded by ovary wall. A megasporocyte is evident in B. C, D, Megasporocyte(s) entering prophase of meiosis I. The inner integument has been initiated. E, F, Megasporocyte(s) in meiosis, the outer integument has been initiated. B, Nucellar beak; ii, inner integument; M, megasporocyte; N, nucellus; O, ovary wall; oi, outer integument. Bars = 25 µm. View largeDownload slide Fig. 2. Megaspore formation in P. bistortoides (A, C) and P. viviparum (B, D). A, End of meiosis II. The four spores that will form a tetrad are evident but the last walls have not yet formed. B, Linear tetrad. The chalazal‐most spore is on an adjacent section; its position is indicated by the arrow. C, D, The chalazal‐most spore (arrows) of each tetrad is functional, the remaining spores degenerate. A hypostase begins to differentiate at the base of the nucellus. Note multiple megaspores in P. bistortoides (C). ii, Inner integument; oi, outer integument; H, hypostase. Bars = 25 µm. View largeDownload slide Fig. 2. Megaspore formation in P. bistortoides (A, C) and P. viviparum (B, D). A, End of meiosis II. The four spores that will form a tetrad are evident but the last walls have not yet formed. B, Linear tetrad. The chalazal‐most spore is on an adjacent section; its position is indicated by the arrow. C, D, The chalazal‐most spore (arrows) of each tetrad is functional, the remaining spores degenerate. A hypostase begins to differentiate at the base of the nucellus. Note multiple megaspores in P. bistortoides (C). ii, Inner integument; oi, outer integument; H, hypostase. Bars = 25 µm. View largeDownload slide Fig. 3. Mature ovules of P. bistortoides (A and B) and P. viviparum (C). Part B shows the section adjacent to that in part A showing the position of the fusion nucleus (arrow) relative to the egg. C, Arrow indicates position of the fusion nucleus. E, Egg; ii, inner integuments; N, nucellus; oi, outer integuments; S, synergids. Bars = 50 µm. View largeDownload slide Fig. 3. Mature ovules of P. bistortoides (A and B) and P. viviparum (C). Part B shows the section adjacent to that in part A showing the position of the fusion nucleus (arrow) relative to the egg. C, Arrow indicates position of the fusion nucleus. E, Egg; ii, inner integuments; N, nucellus; oi, outer integuments; S, synergids. Bars = 50 µm. View largeDownload slide Fig. 4. Abnormal ovules of P. bistortoides (A, C and E) and P. viviparum (B, D and F). A and B, Ovule with both outer and inner integuments but no evidence of megasporocyte or megaspore formation. Compare with Fig. 2. Bars = 25 µm. C and D, Ovules appear nearly mature in anatomy yet have collapsed (A) or are incomplete (B) gametophytes. Part D shows a small gametophyte with synergids and an egg cell. Bars = 25 µm. E, F, Mature ovules with solid nucellus and collapsed gametophytes. The ovules are similar in size to those containing normal female gametophytes. The collapsed cells in the centre of the nucellus indicate that a gametophyte began, but did not complete, development. Bars = 50 µm. G, Gametophyte; ii, inner integument; N, nucellus; oi, outer integument. View largeDownload slide Fig. 4. Abnormal ovules of P. bistortoides (A, C and E) and P. viviparum (B, D and F). A and B, Ovule with both outer and inner integuments but no evidence of megasporocyte or megaspore formation. Compare with Fig. 2. Bars = 25 µm. C and D, Ovules appear nearly mature in anatomy yet have collapsed (A) or are incomplete (B) gametophytes. Part D shows a small gametophyte with synergids and an egg cell. Bars = 25 µm. E, F, Mature ovules with solid nucellus and collapsed gametophytes. The ovules are similar in size to those containing normal female gametophytes. The collapsed cells in the centre of the nucellus indicate that a gametophyte began, but did not complete, development. Bars = 50 µm. G, Gametophyte; ii, inner integument; N, nucellus; oi, outer integument. View largeDownload slide Fig. 5. Pollen viability as assessed by the FCR reaction. View largeDownload slide Fig. 5. Pollen viability as assessed by the FCR reaction. Table 1. Numbers and percentages of total (in parentheses) of mature ovules of differing classes. Ovule type  P. bistortoides  P. viviparum  Unfertilized  14 (35·9)  22 (52·4)  Fertilized (with zygote or embryo)  16 (41)   9 (21·4)  Total functional   30 (76·9)  31 (73·8)  Aborted   9 (23·1)  11 (26·2)  Total ovules  39  42   Ovule type  P. bistortoides  P. viviparum  Unfertilized  14 (35·9)  22 (52·4)  Fertilized (with zygote or embryo)  16 (41)   9 (21·4)  Total functional   30 (76·9)  31 (73·8)  Aborted   9 (23·1)  11 (26·2)  Total ovules  39  42   View Large References AkhalkatsiM, Pfauth M, Calvin CL. 1999. Structural aspects of ovule and seed abortion in Melilotus officinalis (Fabaceae). Protoplasma  208: 211–223. Google Scholar AlbertS, Despres B, Gilleminot J, Bechtold N, Pelletier G, Delseny M, Devic M. 1999. The EMB 506 gene encodes a novel ankyrin repeat containing protein that is essential for the normal development of Arabidopsis embryos. Plant Journal  17: 169–179. Google Scholar AnderssonDI, Hughes D. 1996. Muller’s ratchet decreases fitness of a DNA‐based microbe. Proceedings of the National Academy of Sciences of the USA  93: 906–907. Google Scholar AnderssonS. 1993. The potential for selective seed maturation in Achillea ptarmica (Asteraceae). Oikos  66: 36–42. Google Scholar BauertMR. 1993. Vivipary in Polygonum viviparum: an adaptation to cold climate? Nordic Journal of Botany  13: 473–480. Google Scholar BauertMR. 1996. Genetic diversity and ecotypic differentiation in arctic and alpine populations of Polygonum viviparum. Arctic and Alpine Research  28: 190–195. Google Scholar BawaKS, Webb CJ. 1984. Flower, fruit and seed abortion in tropical forest trees: implications of the evolution of paternal and maternal reproductive patterns. American Journal of Botany  71: 736–751. Google Scholar BenaventeRS, Skorupska H, Palmer RG, Shoemaker RC. 1989. Embryo sac development in the cv KS male‐sterile, female‐sterile line of soybean (Glycine max). American Journal of Botany  76: 1759–1768. Google Scholar BerlynGP, Miksche JP. 1976. Botanical microtechnique and cytochemistry. Ames: Iowa State University Press. Google Scholar BlissLC. 1959. Seed germination in arctic and alpine species. Arctic  11: 180–188. Google Scholar BriggsCL, Westoby M, Selkirk PM, Oldfirld RJ. 1987. Embryology of early abortion due to limited maternal resources in Pisum sativum L. Annals of Botany  59: 611–619. Google Scholar BrinkRA, Cooper DC. 1940. Double fertilization and development of seed in angiosperms. Botanical Gazette  102: 1–25. Google Scholar BrinkRA, Cooper DC. 1941. Incomplete seed failure as a result of somaplastic sterility. Genetics  26: 487–505. Google Scholar BrinkRA, Cooper DC. 1947. The endosperm in seed development. The Botanical Review  13: 423–541. Google Scholar CallaghanTV. 1973. A comparison of the growth of tundra plant species at several widely separated sites. Merlewood Research and Development Paper  53. Google Scholar CallaghanTV, Emanuelsson U. 1985. Population structure and processes of tundra plants and vegetation. In: White J, ed. The population structure of vegetation. Dordrecht: Dr. W. Junk, 399–439. Google Scholar CharlesworthD. 1989. A high mutation rate in a long lived perennial plant. Nature  340: 346–347. Google Scholar CharlesworthD, Chalesworth B. 1987. Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics  18: 237–268. Google Scholar ChaudhuryAM. 1993. Nuclear genes controlling male fertility. The Plant Cell  5: 1277–1283. Google Scholar ChristensenCA, Subramanian S, Drews GN. 1998. Identification of gametophytic mutations affecting female gametophyte development in Arabidopsis. Developmental Biology  292: 136–151. Google Scholar CooperDC, Brink RA. 1940. Partial self‐incompatibility and the collapse of fertile ovules as factors affecting seed formation in alfalfa. Journal of Agricultural Research  60: 453–472. Google Scholar CooperDC, Brink RA, Albrecht HR. 1937. Embryo mortality in relation to seed formation in Alfalfa (Medicago sativa). American Journal of Botany  24: 203–213. Google Scholar CrawfordRMM, Chapman HM, Abbott RJ, Balfour J. 1993. Potential impact of climatic warming on Arctic vegetation. Flora  188: 367–381. Google Scholar DavisGL. 1966. Systematic embryology of the Angiosperms. New York: Wiley. Google Scholar DegeonW. 1918. Morphology of Rumex crispis. Botanical Gazette  66: 393–422. Google Scholar DigglePK. 1995. Architectural effects and interpretation of patterns of fruit and seed development. Annual Review of Ecology and Systematics  26: 531–552. Google Scholar DigglePK. 1997. Extreme preformation in alpine Polygonum viviparum: an architectural and developmental analysis. American Journal of Botany  84: 154–169. Google Scholar DigglePK, Lower SS, Ranker TA. 1998. Clonal diversity in alpine populations of Polygonum viviparum. International Journal of Plant Sciences  159: 606–615. Google Scholar EckertCG, Dorken ME, Mitchell SA. 1999. Loss of sex in clonal populations of a flowering plant, Decodon verticillatus (Lythraceae). Evolution  53: 1079–1092. Google Scholar EdmanG. 1929. Zur Entwicklungsgeschichte der Gattung Oxyria Hill, nebst zytologischen, embryologischen und systematischen Bemerkungen über einige andere Polygonaceen. Acta Horti Bergiani  9: 165–291. Google Scholar EngellK. 1973. A preliminary morphological, cytological, and embryological investigation in Polygonum viviparum. Botanisk Tidsskrift  67: 305–316. Google Scholar EngellK. 1978. Morphology and cytology of Polygonum viviparum in Europe. I. The Faroe Islands. Botanisk Tidsskrift  72: 113–118. Google Scholar FernandoDD, Cass DD. 1997. Developmental assessment of sexual reproduction in Butomus umbellatus (Butomaceae): Female reproductive component. Annals of Botany  80: 457–467. Google Scholar FryxellPA. 1957. Mode of reproduction in higher plants. The Botanical Review  23: 135–233. Google Scholar GabrielW, Lynch M, Bruger R. 1993. Muller’s ratchet and mutational meltdowns. Evolution  47: 1744–1757. Google Scholar GrantV. 1971. Plant speciation. New York: Columbia University Press. Google Scholar GuthCJ, Weller SG. 1986. Pollination, fertilization and ovule abortion in Oxalis magnifica. American Journal of Botany  73: 246–253. Google Scholar HeckelT, Werner K, Sheridan WF, Duma C, Rogowsky PM. 1999. Novel phenotypes and developmental arrest in early embryo specific mutants of maize. Planta  210: 1–8. Google Scholar Heslop‐HarrisonJ, Heslop‐Harrison Y. 1970. Evaluation of pollen viability by enzymatically induced fluorescence: intracellular hydrolysis of fluorescin diacetate. Stain Technology  45: 115–120. Google Scholar Heslop‐HarrisonJ, Heslop‐Harrison Y, Shivana KR. 1984. The evaluation of pollen quality, and a further appraisal of the fluorochromatic (FCR) test procedure. Theoretical and Applied Genetics  67: 367–375. Google Scholar KaulMLH. 1988. Male sterility in higher plants. Berlin: Springer‐Verlag. Google Scholar KaulMLH, Murthy TGK. 1985. Mutant genes affecting higher plant meiosis. Theoretical and Applied Genetics  70: 449–466. Google Scholar KennellJC, Horner HT. 1985. Influence of the soybean male‐sterile gene (ms1) on the development of the female gametophyte. Canadian Journal of Genetics and Cytology  27: 200–209. Google Scholar KlekowskiEJJr. 1988. Mutation, developmental selection and plant evolution. New York: Columbia University Press. Google Scholar KnoxRB, Williams EG, Dumas C. 1986. Pollen, pistil, and reproductive function in crop plants. Plant Breeding Reviews  4: 9–79. Google Scholar LaubengayerRA. 1937. Studies in the anatomy of the polygonaceous flower. American Journal of Botany  24: 329–342. Google Scholar LawR, Cook RED, Manlove RJ. 1983. The ecology of flower and bulbil production in Polygonum viviparum. Nordic Journal of Botany  3: 559–565. Google Scholar LeeTD. 1988. Patterns of fruit and seed production. In: Lovett Doust J, Lovett Doust L, eds. Plant reproductive ecology: patterns and strategies. Oxford: Oxford University Press. Google Scholar LöveA. 1988. Chromosome number reports XCIX. Taxon  37: 396–399. Google Scholar LöveA, Löve D. 1948. Chromosome numbers in the northern plant species. Reykjavik: Ingolfsprent. Google Scholar LöveA, Löve D. 1974. Cytotaxonomical atlas of the Slovenian flora. Vaduz: Cramer. Google Scholar LöveA, Löve D. 1975. Cytotaxonomical atlas of the Arctic flora. Vaduz: Cramer. Google Scholar MahonyKL. 1935. Morphological and cytological studies on Fagopyrum esculentum L. American Journal of Botany  22: 460–475. Google Scholar MarshallDL, Folsom MW. 1991. Mate choice in plants: an anatomical to population perspective. Annual Review of Ecology and Systematics  22: 37–63. Google Scholar MayDE, Weber PJ. 1982. Spatial and temporal variation of the vegetation and its productivity on Niwot Ridge, Colorado. In: Halfpenny JC, ed. Ecological studies in the Colorado alpine: a Fetschrift for John W. Marr. Occasional Paper Number 37. University of Colorado, Institute of Arctic and Alpine Research, Boulder, Colorado. Google Scholar MayerU, Torres‐Ruiz RA, Berleth T, Miséra S, Jürgens G. 1991. Mutations affecting body organization in the Arabidopsis embryo. Nature  353: 402–407. Google Scholar MillerJS, Venable DL. 2000. Polyploidy and the evolution of gender dimorphism in plants. Science  289: 2335–2338 Google Scholar MooneyHA. 1963. Physiological ecology of coastal, subalpine, and alpine populations of Polygonum bistortoides. Ecology  44: 812–816. Google Scholar MooreRJ. 1946. Investigations on rubber‐bearing plants. III. Development of normal and aborting seeds in Asclepias syriaca L. Canadian Journal of Research  24: 55–65. Google Scholar MullerHJ. 1964. The relation of recombination to mutational advance. Mutation Research  1: 2–9. Google Scholar MurrayC, Miller PC. 1982. Phenological observations of major plant‐growth forms and species in montane and Eriophorum‐vaginatum tussock tundra in central Alaska. Holarctic Ecology  5: 109–116. Google Scholar Negrón‐OrtizV. 1998. Reproductive biology of a rare cactus, Opuntia spinosissima (Cactaceae), in the Florida Keys: why is seed set very low? Sexual Plant Reproduction  11: 208–212. Google Scholar de NettancourtD. 1977. Incompatibility in Angiosperms. Berlin: Springer‐Verlag. Google Scholar O‘BrienTP, McCully ME. 1981. The study of plant structure: principles and selected methods. Melbourne: Termatcarphi PTY. Google Scholar PetersenPM. 1981. Variation of the population structure of Polygonum viviparum L. in relation to certain environmental conditions. Meddelelser øm Gronland  4: 3–19. Google Scholar PorsildMP, Porsild AE. 1920. The flora of Disko Island and the adjacent coast of West Greenland from 66°–71° N lat. with remarks on phytogeography, ecology, flowering, fructification and hibernation. Meddelelser øm Gronland  58: 1–155. Google Scholar ReiserL, Fischer RL. 1993. The ovule and embryo sac. The Plant Cell  5: 1291–1301. Google Scholar RichardsAJ. 1997. Plant breeding systems. London: Chapman & Hall. Google Scholar RonseDecraeneL‐P, Akeroyd JR. 1988. Generic limits in Polygonum and related genera (Polygonaceae) on the basis of floral characters. Botanical Journal of the Linnean Society  98: 321–371. Google Scholar SchnarfK. 1931. Verglichende Embryology der Angiospermen. Berlin: Verlag von Gebrüder Bontrager. Google Scholar SchneitzK. 1999. The molecular and genetic control of ovule development. Current Opinion in Plant Biology  2: 13–17. Google Scholar SeaverySR, Bawa KS. 1986. Late‐acting self‐incompatibility in Angiosperms. The Botanical Review   52: 195–219. Google Scholar SedgeleyM. 1980. Anatomical investigation of abscised avocado flowers and fruitlets. Annals of Botany  46: 771–777. Google Scholar SedgeleyM, Griffen AR. 1989. Sexual reproduction in tree crops. London: Academic Press. Google Scholar ShivanaKR, Johri M. 1985. The angiosperm pollen: structure and function. New York: John Wiley and Sons. Google Scholar ShivannaKR, Linskins HF, Cresti M. 1991. Pollen viability and pollen vigour. Theoretical and Applied Genetics  81: 38–42. Google Scholar ShurakiYD, Sedgley M. 1996. Fruit development in Pistacia vera (Anacardiaceae) in relation to embryo abortion and abnormalities at maturity. Australian Journal of Botany  44: 35–45. Google Scholar SilanderJAJr. 1985. Microevolution in clonal plants. In: Jackson JBC, Buss LW, Cook RE, eds. Population biology and evolution of clonal organisms. New Haven:Yale University Press, 107–152. Google Scholar SoaneID, Watkinson AR. 1979. Clonal variation in populations of Ranunculus repens. New Phytologist  82: 557–573. Google Scholar SöyrinkiN. 1989. Fruit production and seedlings in Polygonum viviparum. Memoranda Societatis pro Fauna Flora et Fennica  65: 13–14. Google Scholar StebbinsGL. 1971. Chromosomal evolution in higher plants. Reading MA: Addison Wesley. Google Scholar StevensNE. 1912. The morphology of the seed of buckwheat. Botanical Gazette  53: 59–66. Google Scholar StoneJL, Thomson JD, Dent‐Acosta SJ. 1995. Assessment of pollen viability in hand‐pollination experiments: A review. American Journal of Botany  82: 1186–1197. Google Scholar StrasburgerE. 1879. Die Angiospermen und die Gymnospermen. Jena. Google Scholar StephensonAG. 1981. Flower and fruit abortion: proximate causes and ultimate functions. Annual Review of Ecology and Systematics  12: 253–279. Google Scholar TangmitcharoenS, Owens JN. 1997. Pollen viability and pollen‐tube growth following controlled pollination and their relation to low fruit production in teak (Tectona grandis Linn. f.). Annals of Botany  80: 401–410. Google Scholar ThomsonJD, Rigney LP, Karoly KM Thomson BA. 1994. Pollen viability, vigor, and competitive ability in Erythronium grandiflorum (Liliaceae). American Journal of Botany  81: 1257–1266. Google Scholar TrollW. 1937. Vergleichende Morphologie der Höheren Pflanzen I. Berlin: Gebrüder Borntraeger. Google Scholar UwerU, Willmitzer T, Altmann T. 1998. Inactivation of a glycyl‐tRNA synthase leads to an arrest in plant embryo development. Plant Cell  10: 1277–1294. Google Scholar WalkerMD, Ingersoll RC, Webber PJ. 1995. Effects of interannual climate variation on phenology and growth of two alpine forbs. Ecology  76: 1067–1083. Google Scholar WatkinsonAR, Powell JC. 1993. Seedling recruitment and the maintenance of clonal diversity in plant populations: a computer simulation of Ranunculus repens. Journal of Ecology  81: 707–717. Google Scholar WcisloH. 1967. Studies in Polygonum viviparum L. from the Tatra Mts. Rocznik Naukow. Dydaktyczny WSPw. Krakowie, Prace z botanika  28: 209–231. Google Scholar WeinsD. 1984. Ovule survivorship, brood size, life history, breeding systems, and repoductive success in plants. Oecologia  64: 47–53. Google Scholar WeinsD. 1987. Reproductive success, spontaneous embryo abortion, and genetic load in flowering plants. Oecologia  71: 501–509. Google Scholar WeisIM, Hermanutz LA. 1993. Pollination dynamics of arctic dwarf birch (Betula glandulosa; Betulaceae) and its role in the loss of seed production. American Journal of Botany  80: 1021–1027. Google Scholar WillsonMF, Burley N. 1983. Mate choice in plants: tactics, mechanisms, and consequences. Princeton, NJ: Princeton University Press. Google Scholar WookeyPA, Welker JM, Parsons AN, Press MC, Callaghan TV, Lee JA. 1994. Differential growth, allocation and photosynthetic responses of Polygonum viviparum to simulated environmental change at a high arctic polar semi‐desert. Oikos  70: 131–139. Google Scholar YangWC, Sundaresan V. 2000. Genetics of gametophyte biogenesis in Arabidopsis. Current Opinion in Plant Biology  3: 53–57. Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Annals of Botany Oxford University Press

Barriers to Sexual Reproduction in Polygonum viviparum: A Comparative Developmental Analysis of P. viviparum and P. bistortoides

Loading next page...
 
/lp/oxford-university-press/barriers-to-sexual-reproduction-in-polygonum-viviparum-a-comparative-28uBZh7I5g

References (94)

Publisher
Oxford University Press
ISSN
0305-7364
eISSN
1095-8290
DOI
10.1093/aob/mcf020
Publisher site
See Article on Publisher Site

Abstract

Abstract Polygonum viviparum is widely distributed in arctic and alpine regions of the northern hemisphere. Fruit set has never been observed in North American populations and has been reported only very rarely in Europe. Although this species is extremely well studied, the impediments to successful fruit production are unknown. We investigated the sexual reproductive process in P. viviparum growing in the southern Colorado Rocky Mountains. For comparison, we also examined this process in the sympatric congener P. bistortoides, in which reproduction is exclusively sexual. Lack of viable fruit production in P. viviparum has no single developmental explanation; defects occur in each of the processes and structures associated with sexual reproduction studied, yet, these processes and structures also appear to function normally in at least some flowers or individuals. Development is abnormal in many ovules of P. viviparum, however, comparison with P. bistortoides shows that these abnormalities do not contribute to differences in seed production between the two species. The virtual absence of sexual reproduction in P. viviparum appears to be due largely to a low rate of fertilization and to embryo/fruit abortion. Key words: Asexual reproduction, embryo abortion, development, fruit abortion, pollen viability, Polygonaceae, Polygonum viviparum, Polygonum bistortoides, polyploidy, seed abortion, sexual reproduction. Received: 18 July 2001; Returned for revision: 12 September 2001; Accepted: 11 October 2001. INTRODUCTION The majority of flowering plants are capable of some form of asexual reproduction by means of a diverse range of structures and mechanisms. Clonal progeny may be produced by stolons, runners, rhizomes, tubers, buds on bulbs, corms and roots, layering of stems, and agamospermous seed (Grant, 1971). Despite the prevalence of asexual mechanisms of propagation, most clonal species also reproduce sexually (Stebbins, 1971; Silander, 1985; Richards, 1997) and exclusively asexual species are rare. Polygonum viviparum L. may be one of the few species that relies solely upon vegetative reproduction throughout much of its range (Fryxell, 1957; Callaghan, 1973; Engell, 1973; Petersen, 1981). Polygonum viviparum L. [= Bistorta vivipara (L.) S. Gray] (Polygonaceae) is widely distributed in both arctic and alpine regions of the northern hemisphere (Petersen, 1981; Lawet al., 1983; Callaghan and Emanuelsson, 1985; Bauert, 1993; Wookeyet al., 1994). Although most individuals of P. viviparum flower profusely, fruit set has never been observed in alpine populations of North America and has been reported only very rarely in Europe (Bauert, 1993). P. viviparum is a very common and important component of tundra plant communities and has been the subject of numerous ecological studies (Callaghan, 1973; Petersen, 1981; Bauert, 1993; Crawfordet al., 1993; Wookeyet al., 1994 and references cited therein), and most workers comment on the lack of seed set. The only published reports of sexual reproduction in P. viviparum include: Porsild and Porsild (1920) in Greenland; Söyrinki (1989) in Scandinavia; Bliss (1959) in the subalpine of the Rocky Mountains of Wyoming; Murray and Miller (1982) in interior Alaska; and Bauert (1993) in the central Swiss alps. Sexual reproduction is clearly a rare event in this species. In contrast to the rarity of fruit set by P. viviparum, asexual reproduction occurs readily via the production of bulbils, vegetative axillary buds borne within inflorescences (Troll, 1937; Engell, 1973; Diggle, 1997). After bulbils are dispersed, they germinate and grow under moist conditions (Callaghan, 1973; Engell, 1973, 1978; Petersen, 1981) and establish new, physiologically independent plants that are genetically identical to the parent. The impediments to successful fruit set in P. viviparum are unknown. Any of the many stages of development of male and female gametophytes, gametes, embryo and endosperm, or the processes of pollination and fertilization, might pose a barrier to sexual production of progeny in this species. Preliminary embryological studies of P. viviparum have been published (Edman, 1929;Schnarf, 1931; Engell, 1973); however, the causes of sterility were not identified. In order to explain the lack of fruit production, we investigated the sexual reproductive process of P. viviparum growing in the alpine tundra of the southern Colorado Rocky Mountains. Our study focused on ovule and female gametophyte development, pollen viability and early embryo/seed development. For comparison, we also examined these processes in the sympatric congener, P. bistortoides. P. bistortoides is very common at the study site, produces large numbers of seed and has no means of vegetative reproduction. MATERIALS AND METHODS Study site Study sites were located in the alpine tundra of Niwot Ridge (elevation 3750 m) in the Front Range of the eastern Colorado Rocky Mountains (40°03′N, 105°35′W). Niwot Ridge is managed cooperatively by the University of Colorado Mountain Research Station and the United States Forest Service as an ecological reserve and is the site of the Niwot Ridge Long‐Term Ecological Research Program. Polygonum viviparum and P. bistortoides were sampled haphazardly in a dry meadow community. Species description Polygonum viviparum is widely distributed in both arctic and alpine regions of the northern hemisphere (Petersen, 1981; Lawet al., 1983; Callaghan and Emanuelsson, 1985; Bauert, 1993; Wookeyet al., 1994). Within the study area, P. viviparum is common in all plant communities except in areas of very late melting snow (May and Weber, 1982). Polygonum bistortoides Pursh. (= Bistorta bistortoides (Pursh) Small), is sympatric with P. viviparum in the southern Rocky Mountains and is also widely distributed across all alpine communities, including snowbeds (Walkeret al., 1995). Both species are herbaceous perennials. The main axis of each is a monopodial unbranched rhizome that grows plagiotropically, with little internode elongation, below the soil surface. Rhizomes mature several preformed leaves and axillary inflorescences each year (Diggle, 1997; P.K.D., pers. obs.). P. bistortoides is larger than P. viviparum in most respects: rhizomes are longer and of greater diameter; leaves are longer, wider and more numerous; and inflorescences are taller. The reproductive characters of the two species are very similar. Inflorescences are condensed panicles bearing numerous white flowers. Flowers are small, approx. 3 mm in length. The perianth consists of five white tepals [Laubengayer (1937); Ronse Decraene and Akeroyd (1988), review interpretations of the perianth of the polygonaceous flower]. Stamen number is variable and the gynoecium consists of a three‐angled, uniovulate ovary with three styles and stigmas. The two species differ in one critical aspect of inflorescence structure (and consequently, life history): P. viviparum may bear bulbils in the proximal region of the inflorescence whereas P. bistortoides bears only flowers and reproduces exclusively by seed (Mooney, 1963). Both species flower in July and mature fruits of P. bistortoides are dispersed in August. P. viviparum is polyploid with 2n = 96, x = 12 (Löve and Löve, 1948, 1974, 1975; Wcislo, 1967; Engell, 1973; Löve, 1988). P. bistortoides is diploid with 2n = 24 (Mooney, 1963). Ovule and female gametophyte development All floral organs are initiated in the growing season prior to flower maturation. The preformed inflorescence‐bearing floral primordium undergoes dormancy below ground. Development resumes in the spring and the inflorescence begins to emerge above ground. Sampling of developing inflorescences began when the tips were first visible above the soil surface (first week in July, 1995) and continued until flowers reached anthesis. All inflorescences were fixed in formalin–acetic acid–alcohol (Berlyn and Miksche, 1976). Flowers and flower buds were sampled from basal positions of the short (two or three flowered) branches of the paniculate inflorescences. The gynoecium was dissected from mature flowers and large buds. Small buds were processed whole. Gynoecia and buds were dehydrated to 95% EtOH, embedded in JB4 methacrylate resin, serially sectioned on a Microm microtome at 4 µm, stained with toluidine blue (O’Brien and McCully, 1981) and observed with a Zeiss Axioskop. Pollen viability Mature inflorescences were collected, placed in sealed microcentrifuge tubes, and transported to the laboratory on ice. In 1992 (year 1), inflorescences from 150 individuals of P. viviparum were collected. In 1993 (year 2), inflorescences were collected from 20 P. viviparum and 47 P. bistortoides individuals. All flowers were examined within 12 h of collection. The FCR (fluorochromatic) test (Heslop‐Harrison and Heslop‐Harrison, 1970; Shivannaet al., 1991) was used to evaluate pollen viability. [See Thomsonet al. (1994) and Stoneet al. (1995) for discussions of the reliability of FCR as a measure of pollen viability.] The FCR test may give false negative results due to pollen drying (Heslop‐Harrisonet al., 1984; Shivana and Johri, 1985; Knoxet al., 1986; Thomsonet al., 1994). To prevent desiccation, flowers were kept in the humid atmosphere of the microfuge tubes until the test was performed. Three to five flowers per inflorescence were examined. All anthers were removed from a single flower and were placed on a glass slide. Pollen grains were teased from the anthers with fine forceps. Each slide was flooded with a solution of fluorecine diacetate (FDA; 0·02 g FDA/10 ml acetone in 2 ml of 10 % sucrose) for 5–10 min. Slides were examined with a Zeiss Axioskop equipped for epifluorescence. Brightfield was used to count all pollen grains present. The same slide was then examined under ultraviolet light. Stainability was calculated as the number of fluorescing grains divided by the total number of grains present. The stainability of all flowers from the same inflorescence was averaged, and this mean was used in subsequent statistical analyses. Pollen stainability was analysed with StatView for Macintosh. Stainability data were arcsine square root transformed prior to analysis. RESULTS Pre‐fertilization ovule development Ovule development is nearly identical in the two species, and the description that follows applies to both with exceptions as noted. Flowers of both species are uniovulate, and initiation of the ovary and nucellus had already occurred in all material collected. Inflorescence and floral primordia are preformed, and these events probably occur in the year preceding inflorescence maturation. In the youngest floral primordia sampled, the nucellus was already enclosed by the ovary wall (Fig. 1A, B). The large nucellus has an epidermal layer and one parietal layer. In ovules of both species, the epidermis of the nucellus is often elaborated into an apical ‘beak’ (e.g. Fig. 1C, D, F). The species differ in the number of megasporocytes per nucellus. In P. bistortoides, each nucellus develops one (rarely) to several megasporocytes (Fig. 1C). In P. viviparum there is typically only one megasporocyte per nucellus (Fig. 1D) although occasionally there are two. In both species, as megasporocyte differentiation occurs, the nucellus elongates, and the inner integument of the ovule is initiated (Fig. 1C, D). The megasporocyte (P. viviparum) or megasporocytes (P. bistortoides) undergo meiosis (Figs 1E and 2B) and each forms a linear tetrad (Fig. 2B). The chalazal‐most spore of each tetrad becomes the functional megaspore while the remaining three spores degenerate (Fig. 2C, D). In P. bistortoides there may be several functional megaspores per ovule, each the product of a separate meiotic event (Fig. 2C). A hypostase begins to differentiate at about the time of spore formation (Fig. 2C, D). At, or just prior to meiosis of the megasporocyte(s), the outer integument of the ovule is initiated (Figs 1E and 2B). As megagametophyte development begins, the integuments elongate and enclose the nucellus. The micropylar region of the inner integument expands, becoming several cell layers thick, and forms the micropyle. The outer integument remains two cell layers thick throughout development and does not contribute to the micropyle. One (P. viviparum) to several (P. bistortoides) gametophytes begin to develop within the nucellus. During this time, the gametophytes elongate only slightly. Nucellar elongation is primarily in the chalazal region; there are no further cell divisions in the parietal layers of the apical region of the nucellus, and the existing layer may be crushed. At the completion of mitosis to yield eight nuclei in seven cells, the female gametophyte elongates dramatically. The chalazal portion of the mature gametophyte is ‘wedge’ shaped, appearing very narrow in one dimension, and flattened in the perpendicular plane. No ovules of either species were observed to contain more than one mature gametophyte. The mature ovules of both species are orthotropic and crassinucellate. The polar nuclei of the female gametophyte usually fuse prior to fertilization so that the mature gametophyte comprises seven nuclei in seven cells (Fig. 3A–C). The two synergids are densely cytoplasmic and each has a distinct filiform apparatus. The egg cell is vacuolate and the large nucleus is located in the chalazal end of the cell (Figs 3A, C). The fusion nucleus, formed from the two polar nuclei, is located in the micropylar end of the gametophyte, either adjacent to the egg or near the wall of the central cell (Fig. 3B, C). The central cell is large and vacuolate. The antipodal cells are much smaller than the other cells of the gametophyte and typically persist beyond anthesis. The hypostase of mature ovules is pronounced (Fig. 3C) and the blue‐green staining with toluidine blue O suggests that the cell walls contain lignin. Development of abortive ovules Some ovules of both species fail to complete development. The relative frequency of such aborted development can be inferred from analysis of mature ovules. If those ovules that contain a mature female gametophyte and those containing a zygote or embryo are classified as ‘functional’, then 30 of 39 (76·9 %) mature P. bistortoides ovules were apparently functional, whereas nine (23·1 %) contained obviously malformed gametophytes (Table 1). Similarly, 31 of 42 (73·8 %) mature P. viviparum ovules appeared functional and 11 (26·2 %) were malformed. The frequency of functional ovules did not differ significantly between the two species (χ2 = 0·226, P > 0·1). (Clearly, the ovule becomes an immature seed upon fertilization, however, the term ‘ovule’ is retained here for simplicity.) Developmental analyses show that similar types of abnormalities occur in both species and that cessation of ovule development is not stage‐specific. Some aborted ovules of each species appear to be at the megasporocyte or megaspore stage, i.e. the nucellus is small (Fig. 4B) or has not yet been enclosed by the two integuments (Fig. 4A), yet there is no evidence of a megasporocyte or spore (e.g. compare Fig. 4A with Fig. 2A, C). Development of ovules can also cease at later stages (Fig. 4C–F). In some, the nucellus had begun to elongate but did not contain evidence of a female gametophyte. In other ovules, some, but not all, cells of a mature female gametophyte are present but the nucellus had not elongated to mature dimensions (Fig. 4C, D). In still others, the integuments and nucellus are indistinguishable from those of ovules containing a normal female gametophyte except that the nucellus is solid tissue and the central area is occupied by collapsed cells (Fig. 4E, F). It appears that a female gametophyte began to develop in these ovules and then collapsed. Post‐fertilization ovule development Although the frequency of apparently functional ovules did not differ between the species, a significantly greater proportion of gametophytes of P. bistortoides had been fertilized (Table 1). Sixteen of the 30 (53·3 %) apparently functional ovules of P. bistortoides contained zygotes or embryos compared with nine of 31 (29 %) ovules of P. viviparum (χ2 = 7·353, P < 0·01). Embryo development of the two species was not studied in detail. Ovaries that contained zygotes and embryos were collected from flowers at anthesis at the same time as those containing mature female gametophytes. Presumably, fertilization and the initiation of embryo development takes place very rapidly following anthesis. The embryos were in either the filamentous or globular stage and were surrounded by free nuclear endosperm. Development of one embryo of P. viviparum had progressed to the initiation of cotyledons and the endosperm had cellularized. Mature embryos were not observed in collected material of either species. Pollen viability Average pollen stainability for P. viviparum in year 1 was 20·6 % and ranged from 0– 75 %. In approx. half (54 %) of the individuals, less than 20 % of pollen grains stained positively with FCR, and stainability was greater than 70 % in only two individuals (Fig. 5A). In year 2, average stainability for P. viviparum pollen was 17·4 % and ranged from 0·66–43 %. Stainability was less than 20 % in the majority of individuals (65 %) (Fig. 5B). Stainability of P. viviparum pollen was not significantly different between years (Mann–Whitney U = 719; P = 0·5361). Average pollen stainability for P. bistortoides in year 2 was 56·6 % with a range of 13–83 % (Fig. 5C). Percent stainability differed significantly between the species in 1993 (t = 11·9, P < 0·0001). DISCUSSION Lack of viable seed production in P. viviparum has no single developmental explanation. Defects occur in each of the processes and structures associated with sexual reproduction studied, yet these processes and structures also appear to function normally in at least some flowers or individuals. Malformed ovules are common in P. viviparum; however, comparison with P. bistortoides shows that these abnormalities do not contribute to differences in seed production. The virtual absence of sexual reproduction in P. viviparum appears to be due largely to a low rate of fertilization and to embryo/fruit abortion. Pre‐fertilization development Beginning with the studies of Strasburger (1879), the origin and development of the female gametophyte in members of the Polygonaceae have become classic examples of ‘normal’ processes (Degeon, 1918). Development of the female gametophytes of P. bistortoides and of P. viviparum is similar and conforms to reports of other species of Polygonum and close relatives (summarized in Davis, 1966; also Stevens, 1912; Degeon, 1918; Schnarf, 1931; Mahony, 1935). Seemingly unusual features such as multiple megasporocytes and the conspicuous nucellar beak observed in both P. viviparum and P. bistortoides are common among other species of Polygonum but are not ubiquitous (Davis, 1966). Interestingly, developmental abnormalities of the ovule and female gametophyte are frequently observed among the species of Polygonum that have been studied and are not associated with low fruit production. There are no previous reports of ovule and female gametophyte development for Polygonum bistortoides. In contrast, there are least three published accounts of some aspects of reproductive development in P. viviparum, each motivated by the absence of seed set in this species. Edman (1929) and Schnarf (1931) found that irregularities in the development of female gametophytes were common and that degeneration could occur at any time following meiosis. Our results show that in addition pre‐meiotic failure can occur. Engell (1973) did not note abnormalities, but was unable to examine mature ovules and questioned whether P. viviparum was capable of sexual reproduction. While previous investigators have examined some aspects of female gametophyte development in P. viviparum in the context of low seed set, this study is the first to include an analysis of a ‘control’ taxon. Comparison of the development of P. viviparum ovules with that of P. bistortoides, a sympatric congener with plentiful fruit set, yields the unexpected conclusion that, although developmental irregularities are common in ovules of P. viviparum, they are probably not responsible for the extremely low fruit production. Malformed ovules occur with the same frequency in P. viviparum and P. bistortoides and ovule development appears to abort at the same range of stages in the two species. Clearly, the dramatic differences in fruit production between the two species are not attributable to abnormalities of pre‐fertilization ovule development in P. viviparum. While the frequency of apparently functional ovules is equivalent in the two species, pollen viability differs significantly. Average pollen viability (as assessed by the FCR reaction) is far less in P. viviparum than in P. bistortoides, and in both years a large proportion of flowers had no viable pollen. Engell (1978) also found a high frequency of small pollen grains in flowers of P. viviparum collected from the Faroe Islands and suggested that these were non‐functional. Thus, low pollen viability is also characteristic of northern European populations. Fertilization Pre‐fertilization ovule development of the two species is nearly identical and apparently functional mature female gametophytes occur with equal frequency, yet zygotes and embryos were observed in only 29 % of functional ovules of P. viviparum compared with 53 % in P. bistortoides. Assuming that these are the result of fertilization (i.e. not agamospermy), then the rate of fertilization is significantly lower in P. viviparum than in P. bistortoides. The difference in fertilization rate between the two species may be attributable to the large differences in pollen viability. If pollen is not available in excess, then the probability of fertilization occurring should be proportional to the frequency of viable pollen grains among potential donors. Fertilization of P. viviparum ovules also may be limited by pollinator service but pollen removal and deposition rates were not examined. A third potential explanation for differences in fertilization is self‐incompatibility. Because of the prevalence of asexual reproduction in P. viviparum, many potential pollen donors could be genetically identical to the recipient (e.g. Weis and Hermanutz, 1993; Tangmitcharoen and Owens, 1997; Negrón‐Ortiz, 1998). If the species is self‐incompatible, these pollinations would not lead to successful fertilization. Two facts argue against this explanation: (1) although the compatibility system of P. viviparum is unknown, this species is polyploid and polyploidy tends to disrupt self‐incompatibility (de Nettancourt, 1977; Richards, 1997; Miller and Venable, 2000); and (2) a genetic analysis of the study population found relatively high levels of genotypic diversity (Diggleet al., 1997), therefore, even if the species is self‐incompatible, compatible genotypes should be available within populations. Thus, low pollen viability or limited pollination are more likely explanations of the differences in fertilization between the two species. Although the frequency of ovules with zygotes or embryos is lower in P. viviparum than in P. bistortoides, the observed fertilization rate does not explain the apparent absence of fruit production. Embryo development was initiated in 29 % of ovules of P. viviparum and flower number per inflorescence at the study site ranged from 0 (some inflorescences bear only bulbils) to 65, with a mean of 16 (K. Stitt, unpubl. res.). If all fertilized ovules matured into seeds, there should have been a detectable number of fruit at the field site. Post‐fertilization development Although 29 % of P. viviparum ovules contained zygotes or embryos, we observed neither a mature embryo in sectioned material nor a mature fruit in the field; all flowers abscised from inflorescences within a few days of anthesis. Clearly, most fertilized ovules abort after initiation of embryo development. Embryo/seed abortion followed by abscission must occur rapidly because we did not observe any apparently malformed (dying) embryos or endosperms in sectioned material. Thus, in addition to a lower rate of fertilization compared with P. bistortoides, abortion of new sporophytes also must be extremely common in P. viviparum. Abortion of developing seeds and fruits is quite common among angiosperms (reviewed in Stephenson, 1981; Lee, 1988; Sedgeley and Griffen, 1989; Diggle, 1995). Failure of post‐fertilization development has been ascribed to a variety of factors including competition among seeds or fruits for nutrients (Stephenson, 1981; Lee, 1988; Diggle, 1995 and references therein; Shuraki and Sedgley, 1996), maternal selection among embryos of varying genetic quality (reviewed in Willson and Burley, 1983; Bawa and Webb, 1984; Guth and Weller, 1986; Briggset al., 1987; Marshall and Folsom, 1991; Akhalkatsiet al., 1999), late acting self‐incompatibility (Seavery and Bawa, 1986), and genetic load or other genetic abnormalities (Brink and Cooper, 1940, 1941, 1947; Cooper and Brink, 1940; Weins, 1984, 1987; Charlesworth and Charlesworth, 1987; Charlesworth, 1989; Andersson, 1993). Competition for nutrients and maternal selection among embryos implies that there are some ‘winners’, i.e. although abortion occurs some fruit should mature on each inflorescence or individual. This is clearly not the case for P. viviparum. Late acting self‐incompatibility also is not likely to occur in P. viviparum for the reasons discussed above for pre‐fertilization self‐incompatibility (see discussion under ‘fertilization’). Given that pollen development is affected so drastically in P. viviparum, and that fruit production is uniformly negligible among individuals, among years and among sites, genetic abnormalities are a likely cause of embryo abortion. Whereas P. bistortoides is diploid, P. viviparum is reported to be an octoploid (2n = 96, x = 12; Löve and Löve, 1948, 1974, 1975; Wcislo, 1967; Engell, 1973; Löve, 1988). Full or partial sterility is not uncommon among polyploids and is often attributed to irregular meiosis (e.g. Stebbins, 1971; Richards, 1997 and references therein). Meiotic aberrations would result in low pollen viability as observed in P. viviparum and could also explain embryo abortions. Even when gametes or gametophytes appear functional, they may have an unbalanced chromosome content and fertilization would produce zygotes that ultimately abort. The association between polyploidy and meiotic irregularity, however, is not well established (P. Soltis, pers. comm.) and many polyploids have normal chromosome pairing at meiosis (e.g. Grant, 1971; Lawet al., 1983). Further examination of meiosis in P. viviparum is warranted. Low pollen viability and embryo abortion also may be due to accumulation of mutations in these very long lived clonal organisms (Muller’s ratchet; Muller, 1964; Gabrielet al., 1993; Andersson and Hughes, 1996; Eckertet al., 1999). Many mutations have been identified in crop and model species that specifically affect meiosis (e.g. Kaul and Murthy, 1985). In addition, other mutations affect pollen viability (e.g. Kennell and Horner, 1985; Kaul, 1988; Benaventeet al., 1989; Chaudhury, 1993), ovule development (Reiser and Fischer, 1993; Christensenet al., 1998; Schneitz, 1999; Yang and Sundareasan, 2000) or embryo development (reviewed in Mayeret al., 1991; Uweret al., 1998; Albertet al., 1999; Heckelet al., 1999). Such mutations can be perpetuated by vegetative reproduction, and could accumulate over the lifespan of a genet, particularly if the mutations have no negative pleiotropic effects on survival (Eckertet al., 1999). In addition, embryo abortion could be the result of genetic load if the new sporophytes are products of self‐fertilization (Weins, 1984, 1987; Klekowski, 1988). The absence of any clear developmental correlate of embryo abortion is not unusual. There has been considerable developmental analysis of seed and fruit abortion, particularly in economically important species (reviewed in Brink and Cooper, 1940, 1941; Sedgeley and Griffen, 1989; Fernando and Cass, 1997). Most commonly, abortion is not stage‐specific and no visible evidence of the cause of abortion can be identified (e.g. Persica: Sedgeley, 1980; Pistacia: Shuraki and Sedgeley, 1996; Butomus: Fernando and Cass, 1997; Melilotus: Akhalkatsiet al., 1999; many tree crops including cherry, citrus, teak, mango and plum: Stephenson, 1981). In other cases, clear developmental abnormalities precede abortion. These include proliferation or death of the endosperm (Medicago: Cooperet al., 1937; Oxalis: Guth and Weller, 1986) proliferation of integuments (Asclepias: Moore, 1946), or pronounced development of the hypostase (Pisum: Briggset al., 1987). Although high rates of fruit abortion occur regularly in all of these species, the ultimate cause of abortion was rarely identified conclusively. Why is fruit production so low in P. viviparum? Although irregular ovule development is common in P. viviparum, it is no more frequent in this species than in the sympatric P. bistortoides. In fact, multiple developmental abnormalities are reported to be common throughout the genus Polygonum and may be normal for the genus (an odd situation for the ‘type’ genus for female gametophyte development!). Abortive ovule development does decrease maximum seed production in P. viviparum, but is not responsible for its absence. In contrast to the similarity of ovule development in the two species, the incidence of abnormal pollen is far greater in P. viviparum than in P. bistortoides and may be associated with differences in fertilization rates. Even low fertilization rates, however, cannot account for the rarity of fruit production. Abortion of young sporophytes must be the final phase of reproductive development that reduces fruit maturation to undetectable levels. Although fruit set in P. viviparum has not been observed in the study population, it is likely that it does occur occasionally. Genotypic diversity in the study populations is equivalent to that of clonal species known to have regular fruit/seed maturation (Diggleet al., 1997; see also Bauert, 1993, 1996 for European populations). The most likely explanation of the existence of multiple genotypes in this population is that viable seeds are produced, albeit rarely. Theoretical models suggest that a very low rate of seedling input into established populations is sufficient to maintain genetic diversity (Soane and Watkinson, 1979; Watkinson and Powell, 1993). The causes of low fruit production in P. viviparum are not absolute, and occasionally two normal gametes may unite to form a viable new sporophyte. ACKNOWLEDGEMENTS The authors thank Tara Doughty, Rachel Kaplan, Anne Klein, Amber Moody, Mingon Macias and Kristine Stitt for laboratory assistance, and Larry Hufford for comments on the manuscript. Support was provided by NSF DEB‐9357076, Apple Computers, Inc., the Niwot Ridge Long‐Term Ecological Research Program (NSF DEB‐9211776) and the University of Colorado Mountain Research Station (NSF BIR‐9115097). View largeDownload slide Fig. 1. Early ovule development in P. bistortoides (A, C and E) and P. viviparum (B, D and F). A, B, Nucellus surrounded by ovary wall. A megasporocyte is evident in B. C, D, Megasporocyte(s) entering prophase of meiosis I. The inner integument has been initiated. E, F, Megasporocyte(s) in meiosis, the outer integument has been initiated. B, Nucellar beak; ii, inner integument; M, megasporocyte; N, nucellus; O, ovary wall; oi, outer integument. Bars = 25 µm. View largeDownload slide Fig. 1. Early ovule development in P. bistortoides (A, C and E) and P. viviparum (B, D and F). A, B, Nucellus surrounded by ovary wall. A megasporocyte is evident in B. C, D, Megasporocyte(s) entering prophase of meiosis I. The inner integument has been initiated. E, F, Megasporocyte(s) in meiosis, the outer integument has been initiated. B, Nucellar beak; ii, inner integument; M, megasporocyte; N, nucellus; O, ovary wall; oi, outer integument. Bars = 25 µm. View largeDownload slide Fig. 2. Megaspore formation in P. bistortoides (A, C) and P. viviparum (B, D). A, End of meiosis II. The four spores that will form a tetrad are evident but the last walls have not yet formed. B, Linear tetrad. The chalazal‐most spore is on an adjacent section; its position is indicated by the arrow. C, D, The chalazal‐most spore (arrows) of each tetrad is functional, the remaining spores degenerate. A hypostase begins to differentiate at the base of the nucellus. Note multiple megaspores in P. bistortoides (C). ii, Inner integument; oi, outer integument; H, hypostase. Bars = 25 µm. View largeDownload slide Fig. 2. Megaspore formation in P. bistortoides (A, C) and P. viviparum (B, D). A, End of meiosis II. The four spores that will form a tetrad are evident but the last walls have not yet formed. B, Linear tetrad. The chalazal‐most spore is on an adjacent section; its position is indicated by the arrow. C, D, The chalazal‐most spore (arrows) of each tetrad is functional, the remaining spores degenerate. A hypostase begins to differentiate at the base of the nucellus. Note multiple megaspores in P. bistortoides (C). ii, Inner integument; oi, outer integument; H, hypostase. Bars = 25 µm. View largeDownload slide Fig. 3. Mature ovules of P. bistortoides (A and B) and P. viviparum (C). Part B shows the section adjacent to that in part A showing the position of the fusion nucleus (arrow) relative to the egg. C, Arrow indicates position of the fusion nucleus. E, Egg; ii, inner integuments; N, nucellus; oi, outer integuments; S, synergids. Bars = 50 µm. View largeDownload slide Fig. 3. Mature ovules of P. bistortoides (A and B) and P. viviparum (C). Part B shows the section adjacent to that in part A showing the position of the fusion nucleus (arrow) relative to the egg. C, Arrow indicates position of the fusion nucleus. E, Egg; ii, inner integuments; N, nucellus; oi, outer integuments; S, synergids. Bars = 50 µm. View largeDownload slide Fig. 4. Abnormal ovules of P. bistortoides (A, C and E) and P. viviparum (B, D and F). A and B, Ovule with both outer and inner integuments but no evidence of megasporocyte or megaspore formation. Compare with Fig. 2. Bars = 25 µm. C and D, Ovules appear nearly mature in anatomy yet have collapsed (A) or are incomplete (B) gametophytes. Part D shows a small gametophyte with synergids and an egg cell. Bars = 25 µm. E, F, Mature ovules with solid nucellus and collapsed gametophytes. The ovules are similar in size to those containing normal female gametophytes. The collapsed cells in the centre of the nucellus indicate that a gametophyte began, but did not complete, development. Bars = 50 µm. G, Gametophyte; ii, inner integument; N, nucellus; oi, outer integument. View largeDownload slide Fig. 4. Abnormal ovules of P. bistortoides (A, C and E) and P. viviparum (B, D and F). A and B, Ovule with both outer and inner integuments but no evidence of megasporocyte or megaspore formation. Compare with Fig. 2. Bars = 25 µm. C and D, Ovules appear nearly mature in anatomy yet have collapsed (A) or are incomplete (B) gametophytes. Part D shows a small gametophyte with synergids and an egg cell. Bars = 25 µm. E, F, Mature ovules with solid nucellus and collapsed gametophytes. The ovules are similar in size to those containing normal female gametophytes. The collapsed cells in the centre of the nucellus indicate that a gametophyte began, but did not complete, development. Bars = 50 µm. G, Gametophyte; ii, inner integument; N, nucellus; oi, outer integument. View largeDownload slide Fig. 5. Pollen viability as assessed by the FCR reaction. View largeDownload slide Fig. 5. Pollen viability as assessed by the FCR reaction. Table 1. Numbers and percentages of total (in parentheses) of mature ovules of differing classes. Ovule type  P. bistortoides  P. viviparum  Unfertilized  14 (35·9)  22 (52·4)  Fertilized (with zygote or embryo)  16 (41)   9 (21·4)  Total functional   30 (76·9)  31 (73·8)  Aborted   9 (23·1)  11 (26·2)  Total ovules  39  42   Ovule type  P. bistortoides  P. viviparum  Unfertilized  14 (35·9)  22 (52·4)  Fertilized (with zygote or embryo)  16 (41)   9 (21·4)  Total functional   30 (76·9)  31 (73·8)  Aborted   9 (23·1)  11 (26·2)  Total ovules  39  42   View Large References AkhalkatsiM, Pfauth M, Calvin CL. 1999. Structural aspects of ovule and seed abortion in Melilotus officinalis (Fabaceae). Protoplasma  208: 211–223. Google Scholar AlbertS, Despres B, Gilleminot J, Bechtold N, Pelletier G, Delseny M, Devic M. 1999. The EMB 506 gene encodes a novel ankyrin repeat containing protein that is essential for the normal development of Arabidopsis embryos. Plant Journal  17: 169–179. Google Scholar AnderssonDI, Hughes D. 1996. Muller’s ratchet decreases fitness of a DNA‐based microbe. Proceedings of the National Academy of Sciences of the USA  93: 906–907. Google Scholar AnderssonS. 1993. The potential for selective seed maturation in Achillea ptarmica (Asteraceae). Oikos  66: 36–42. Google Scholar BauertMR. 1993. Vivipary in Polygonum viviparum: an adaptation to cold climate? Nordic Journal of Botany  13: 473–480. Google Scholar BauertMR. 1996. Genetic diversity and ecotypic differentiation in arctic and alpine populations of Polygonum viviparum. Arctic and Alpine Research  28: 190–195. Google Scholar BawaKS, Webb CJ. 1984. Flower, fruit and seed abortion in tropical forest trees: implications of the evolution of paternal and maternal reproductive patterns. American Journal of Botany  71: 736–751. Google Scholar BenaventeRS, Skorupska H, Palmer RG, Shoemaker RC. 1989. Embryo sac development in the cv KS male‐sterile, female‐sterile line of soybean (Glycine max). American Journal of Botany  76: 1759–1768. Google Scholar BerlynGP, Miksche JP. 1976. Botanical microtechnique and cytochemistry. Ames: Iowa State University Press. Google Scholar BlissLC. 1959. Seed germination in arctic and alpine species. Arctic  11: 180–188. Google Scholar BriggsCL, Westoby M, Selkirk PM, Oldfirld RJ. 1987. Embryology of early abortion due to limited maternal resources in Pisum sativum L. Annals of Botany  59: 611–619. Google Scholar BrinkRA, Cooper DC. 1940. Double fertilization and development of seed in angiosperms. Botanical Gazette  102: 1–25. Google Scholar BrinkRA, Cooper DC. 1941. Incomplete seed failure as a result of somaplastic sterility. Genetics  26: 487–505. Google Scholar BrinkRA, Cooper DC. 1947. The endosperm in seed development. The Botanical Review  13: 423–541. Google Scholar CallaghanTV. 1973. A comparison of the growth of tundra plant species at several widely separated sites. Merlewood Research and Development Paper  53. Google Scholar CallaghanTV, Emanuelsson U. 1985. Population structure and processes of tundra plants and vegetation. In: White J, ed. The population structure of vegetation. Dordrecht: Dr. W. Junk, 399–439. Google Scholar CharlesworthD. 1989. A high mutation rate in a long lived perennial plant. Nature  340: 346–347. Google Scholar CharlesworthD, Chalesworth B. 1987. Inbreeding depression and its evolutionary consequences. Annual Review of Ecology and Systematics  18: 237–268. Google Scholar ChaudhuryAM. 1993. Nuclear genes controlling male fertility. The Plant Cell  5: 1277–1283. Google Scholar ChristensenCA, Subramanian S, Drews GN. 1998. Identification of gametophytic mutations affecting female gametophyte development in Arabidopsis. Developmental Biology  292: 136–151. Google Scholar CooperDC, Brink RA. 1940. Partial self‐incompatibility and the collapse of fertile ovules as factors affecting seed formation in alfalfa. Journal of Agricultural Research  60: 453–472. Google Scholar CooperDC, Brink RA, Albrecht HR. 1937. Embryo mortality in relation to seed formation in Alfalfa (Medicago sativa). American Journal of Botany  24: 203–213. Google Scholar CrawfordRMM, Chapman HM, Abbott RJ, Balfour J. 1993. Potential impact of climatic warming on Arctic vegetation. Flora  188: 367–381. Google Scholar DavisGL. 1966. Systematic embryology of the Angiosperms. New York: Wiley. Google Scholar DegeonW. 1918. Morphology of Rumex crispis. Botanical Gazette  66: 393–422. Google Scholar DigglePK. 1995. Architectural effects and interpretation of patterns of fruit and seed development. Annual Review of Ecology and Systematics  26: 531–552. Google Scholar DigglePK. 1997. Extreme preformation in alpine Polygonum viviparum: an architectural and developmental analysis. American Journal of Botany  84: 154–169. Google Scholar DigglePK, Lower SS, Ranker TA. 1998. Clonal diversity in alpine populations of Polygonum viviparum. International Journal of Plant Sciences  159: 606–615. Google Scholar EckertCG, Dorken ME, Mitchell SA. 1999. Loss of sex in clonal populations of a flowering plant, Decodon verticillatus (Lythraceae). Evolution  53: 1079–1092. Google Scholar EdmanG. 1929. Zur Entwicklungsgeschichte der Gattung Oxyria Hill, nebst zytologischen, embryologischen und systematischen Bemerkungen über einige andere Polygonaceen. Acta Horti Bergiani  9: 165–291. Google Scholar EngellK. 1973. A preliminary morphological, cytological, and embryological investigation in Polygonum viviparum. Botanisk Tidsskrift  67: 305–316. Google Scholar EngellK. 1978. Morphology and cytology of Polygonum viviparum in Europe. I. The Faroe Islands. Botanisk Tidsskrift  72: 113–118. Google Scholar FernandoDD, Cass DD. 1997. Developmental assessment of sexual reproduction in Butomus umbellatus (Butomaceae): Female reproductive component. Annals of Botany  80: 457–467. Google Scholar FryxellPA. 1957. Mode of reproduction in higher plants. The Botanical Review  23: 135–233. Google Scholar GabrielW, Lynch M, Bruger R. 1993. Muller’s ratchet and mutational meltdowns. Evolution  47: 1744–1757. Google Scholar GrantV. 1971. Plant speciation. New York: Columbia University Press. Google Scholar GuthCJ, Weller SG. 1986. Pollination, fertilization and ovule abortion in Oxalis magnifica. American Journal of Botany  73: 246–253. Google Scholar HeckelT, Werner K, Sheridan WF, Duma C, Rogowsky PM. 1999. Novel phenotypes and developmental arrest in early embryo specific mutants of maize. Planta  210: 1–8. Google Scholar Heslop‐HarrisonJ, Heslop‐Harrison Y. 1970. Evaluation of pollen viability by enzymatically induced fluorescence: intracellular hydrolysis of fluorescin diacetate. Stain Technology  45: 115–120. Google Scholar Heslop‐HarrisonJ, Heslop‐Harrison Y, Shivana KR. 1984. The evaluation of pollen quality, and a further appraisal of the fluorochromatic (FCR) test procedure. Theoretical and Applied Genetics  67: 367–375. Google Scholar KaulMLH. 1988. Male sterility in higher plants. Berlin: Springer‐Verlag. Google Scholar KaulMLH, Murthy TGK. 1985. Mutant genes affecting higher plant meiosis. Theoretical and Applied Genetics  70: 449–466. Google Scholar KennellJC, Horner HT. 1985. Influence of the soybean male‐sterile gene (ms1) on the development of the female gametophyte. Canadian Journal of Genetics and Cytology  27: 200–209. Google Scholar KlekowskiEJJr. 1988. Mutation, developmental selection and plant evolution. New York: Columbia University Press. Google Scholar KnoxRB, Williams EG, Dumas C. 1986. Pollen, pistil, and reproductive function in crop plants. Plant Breeding Reviews  4: 9–79. Google Scholar LaubengayerRA. 1937. Studies in the anatomy of the polygonaceous flower. American Journal of Botany  24: 329–342. Google Scholar LawR, Cook RED, Manlove RJ. 1983. The ecology of flower and bulbil production in Polygonum viviparum. Nordic Journal of Botany  3: 559–565. Google Scholar LeeTD. 1988. Patterns of fruit and seed production. In: Lovett Doust J, Lovett Doust L, eds. Plant reproductive ecology: patterns and strategies. Oxford: Oxford University Press. Google Scholar LöveA. 1988. Chromosome number reports XCIX. Taxon  37: 396–399. Google Scholar LöveA, Löve D. 1948. Chromosome numbers in the northern plant species. Reykjavik: Ingolfsprent. Google Scholar LöveA, Löve D. 1974. Cytotaxonomical atlas of the Slovenian flora. Vaduz: Cramer. Google Scholar LöveA, Löve D. 1975. Cytotaxonomical atlas of the Arctic flora. Vaduz: Cramer. Google Scholar MahonyKL. 1935. Morphological and cytological studies on Fagopyrum esculentum L. American Journal of Botany  22: 460–475. Google Scholar MarshallDL, Folsom MW. 1991. Mate choice in plants: an anatomical to population perspective. Annual Review of Ecology and Systematics  22: 37–63. Google Scholar MayDE, Weber PJ. 1982. Spatial and temporal variation of the vegetation and its productivity on Niwot Ridge, Colorado. In: Halfpenny JC, ed. Ecological studies in the Colorado alpine: a Fetschrift for John W. Marr. Occasional Paper Number 37. University of Colorado, Institute of Arctic and Alpine Research, Boulder, Colorado. Google Scholar MayerU, Torres‐Ruiz RA, Berleth T, Miséra S, Jürgens G. 1991. Mutations affecting body organization in the Arabidopsis embryo. Nature  353: 402–407. Google Scholar MillerJS, Venable DL. 2000. Polyploidy and the evolution of gender dimorphism in plants. Science  289: 2335–2338 Google Scholar MooneyHA. 1963. Physiological ecology of coastal, subalpine, and alpine populations of Polygonum bistortoides. Ecology  44: 812–816. Google Scholar MooreRJ. 1946. Investigations on rubber‐bearing plants. III. Development of normal and aborting seeds in Asclepias syriaca L. Canadian Journal of Research  24: 55–65. Google Scholar MullerHJ. 1964. The relation of recombination to mutational advance. Mutation Research  1: 2–9. Google Scholar MurrayC, Miller PC. 1982. Phenological observations of major plant‐growth forms and species in montane and Eriophorum‐vaginatum tussock tundra in central Alaska. Holarctic Ecology  5: 109–116. Google Scholar Negrón‐OrtizV. 1998. Reproductive biology of a rare cactus, Opuntia spinosissima (Cactaceae), in the Florida Keys: why is seed set very low? Sexual Plant Reproduction  11: 208–212. Google Scholar de NettancourtD. 1977. Incompatibility in Angiosperms. Berlin: Springer‐Verlag. Google Scholar O‘BrienTP, McCully ME. 1981. The study of plant structure: principles and selected methods. Melbourne: Termatcarphi PTY. Google Scholar PetersenPM. 1981. Variation of the population structure of Polygonum viviparum L. in relation to certain environmental conditions. Meddelelser øm Gronland  4: 3–19. Google Scholar PorsildMP, Porsild AE. 1920. The flora of Disko Island and the adjacent coast of West Greenland from 66°–71° N lat. with remarks on phytogeography, ecology, flowering, fructification and hibernation. Meddelelser øm Gronland  58: 1–155. Google Scholar ReiserL, Fischer RL. 1993. The ovule and embryo sac. The Plant Cell  5: 1291–1301. Google Scholar RichardsAJ. 1997. Plant breeding systems. London: Chapman & Hall. Google Scholar RonseDecraeneL‐P, Akeroyd JR. 1988. Generic limits in Polygonum and related genera (Polygonaceae) on the basis of floral characters. Botanical Journal of the Linnean Society  98: 321–371. Google Scholar SchnarfK. 1931. Verglichende Embryology der Angiospermen. Berlin: Verlag von Gebrüder Bontrager. Google Scholar SchneitzK. 1999. The molecular and genetic control of ovule development. Current Opinion in Plant Biology  2: 13–17. Google Scholar SeaverySR, Bawa KS. 1986. Late‐acting self‐incompatibility in Angiosperms. The Botanical Review   52: 195–219. Google Scholar SedgeleyM. 1980. Anatomical investigation of abscised avocado flowers and fruitlets. Annals of Botany  46: 771–777. Google Scholar SedgeleyM, Griffen AR. 1989. Sexual reproduction in tree crops. London: Academic Press. Google Scholar ShivanaKR, Johri M. 1985. The angiosperm pollen: structure and function. New York: John Wiley and Sons. Google Scholar ShivannaKR, Linskins HF, Cresti M. 1991. Pollen viability and pollen vigour. Theoretical and Applied Genetics  81: 38–42. Google Scholar ShurakiYD, Sedgley M. 1996. Fruit development in Pistacia vera (Anacardiaceae) in relation to embryo abortion and abnormalities at maturity. Australian Journal of Botany  44: 35–45. Google Scholar SilanderJAJr. 1985. Microevolution in clonal plants. In: Jackson JBC, Buss LW, Cook RE, eds. Population biology and evolution of clonal organisms. New Haven:Yale University Press, 107–152. Google Scholar SoaneID, Watkinson AR. 1979. Clonal variation in populations of Ranunculus repens. New Phytologist  82: 557–573. Google Scholar SöyrinkiN. 1989. Fruit production and seedlings in Polygonum viviparum. Memoranda Societatis pro Fauna Flora et Fennica  65: 13–14. Google Scholar StebbinsGL. 1971. Chromosomal evolution in higher plants. Reading MA: Addison Wesley. Google Scholar StevensNE. 1912. The morphology of the seed of buckwheat. Botanical Gazette  53: 59–66. Google Scholar StoneJL, Thomson JD, Dent‐Acosta SJ. 1995. Assessment of pollen viability in hand‐pollination experiments: A review. American Journal of Botany  82: 1186–1197. Google Scholar StrasburgerE. 1879. Die Angiospermen und die Gymnospermen. Jena. Google Scholar StephensonAG. 1981. Flower and fruit abortion: proximate causes and ultimate functions. Annual Review of Ecology and Systematics  12: 253–279. Google Scholar TangmitcharoenS, Owens JN. 1997. Pollen viability and pollen‐tube growth following controlled pollination and their relation to low fruit production in teak (Tectona grandis Linn. f.). Annals of Botany  80: 401–410. Google Scholar ThomsonJD, Rigney LP, Karoly KM Thomson BA. 1994. Pollen viability, vigor, and competitive ability in Erythronium grandiflorum (Liliaceae). American Journal of Botany  81: 1257–1266. Google Scholar TrollW. 1937. Vergleichende Morphologie der Höheren Pflanzen I. Berlin: Gebrüder Borntraeger. Google Scholar UwerU, Willmitzer T, Altmann T. 1998. Inactivation of a glycyl‐tRNA synthase leads to an arrest in plant embryo development. Plant Cell  10: 1277–1294. Google Scholar WalkerMD, Ingersoll RC, Webber PJ. 1995. Effects of interannual climate variation on phenology and growth of two alpine forbs. Ecology  76: 1067–1083. Google Scholar WatkinsonAR, Powell JC. 1993. Seedling recruitment and the maintenance of clonal diversity in plant populations: a computer simulation of Ranunculus repens. Journal of Ecology  81: 707–717. Google Scholar WcisloH. 1967. Studies in Polygonum viviparum L. from the Tatra Mts. Rocznik Naukow. Dydaktyczny WSPw. Krakowie, Prace z botanika  28: 209–231. Google Scholar WeinsD. 1984. Ovule survivorship, brood size, life history, breeding systems, and repoductive success in plants. Oecologia  64: 47–53. Google Scholar WeinsD. 1987. Reproductive success, spontaneous embryo abortion, and genetic load in flowering plants. Oecologia  71: 501–509. Google Scholar WeisIM, Hermanutz LA. 1993. Pollination dynamics of arctic dwarf birch (Betula glandulosa; Betulaceae) and its role in the loss of seed production. American Journal of Botany  80: 1021–1027. Google Scholar WillsonMF, Burley N. 1983. Mate choice in plants: tactics, mechanisms, and consequences. Princeton, NJ: Princeton University Press. Google Scholar WookeyPA, Welker JM, Parsons AN, Press MC, Callaghan TV, Lee JA. 1994. Differential growth, allocation and photosynthetic responses of Polygonum viviparum to simulated environmental change at a high arctic polar semi‐desert. Oikos  70: 131–139. Google Scholar YangWC, Sundaresan V. 2000. Genetics of gametophyte biogenesis in Arabidopsis. Current Opinion in Plant Biology  3: 53–57. Google Scholar

Journal

Annals of BotanyOxford University Press

Published: Feb 1, 2002

There are no references for this article.