Unique Growth Paths of Heterospecific Pollen Tubes Result in Late

Plant Biology ISSN 1435-8603

RESEARCH PAPER Unique growth paths of heterospeciﬁc pollen tubes result in late entry into ovules in the gynoecium of Sagittaria (Alismataceae) N. Lyu, W. Du & X.-F. Wang College of Life Sciences, Wuhan University, Wuhan, China

Keywords ABSTRACT heterospecific pollen; interspecific pollination; pollen tube growth; pollen tube guidance; pollen-pistil interactions; Sagittaria. • Pollen-pistil interactions are a fundamental process in the reproductive biology of angiosperms and play a particularly important role in maintaining incipient species Correspondence that exist in sympatry. However, the majority of previous studies have focused on spe- X.-F. Wang, College of Life Sciences, Wuhan cies with syncarpous gynoecia (fused carpels) and not those with apocarpous gynoecia University, Wuhan, 430072, China. (unfused carpels). E-mail: [email protected] • In the present study, we investigated the growth of conspecific pollen tubes compared to heterospecific pollen tubes in Sagittaria species, which have apocarpous gynoecia. We Editor conducted controlled pollinations between S. pygmaea and S. trifolia and observed the G. Thiel growth of conspecific and heterospecific pollen tubes under a fluorescence microscope. • Heterospecific and conspecific pollen tubes arrived at locules within the ovaries near Received: 26 August 2015; Accepted: simultaneously. However, conspecific pollen tubes entered into the ovules directly, 24 September 2016 whereas heterospecific tubes passed through the carpel base and adjacent receptacle tissue, to ultimately fertilize other unfertilized ovules. This longer route taken by doi:10.1111/plb.12508 heterospecific pollen tubes therefore caused a delay in the time required to enter into the ovules. Furthermore, heterospecific pollen tubes displayed similar growth patterns at early and peak pollination. The growth pattern of heterospecific pollen tubes at late pollination was similar to that of conspecific pollen tubes at peak pollination. • Heterospecific and conspecific pollen tubes took different routes to fertilize ovules. A delayed entry of heterospecific pollen into ovules may be a novel mechanism of conspecific pollen advantage (CPA) for apocarpous species.

2010) and pistil-length mismatch (Lee et al. 2008). Environ- INTRODUCTION ment-dependent selection on pollen performance (Chapman Varying floral characters, such as corolla color and petal or pol- et al. 2005) may also play a role in pollen-pistil interactions. len scent (Adler & Irwin 2006), production of flower nectar These phenomena have been observed among taxa of Nicotiana (Mitchell & Waser 1992), as well as the structure of pollen (Lee et al. 2008), Iris (Emms et al. 1996), Mimulus (Fishman grain and stigma (Sannier et al. 2009) can determine outcomes et al. 2008), Helianthus (Rieseberg et al. 1995), Senecio (Chap- for plant pollination owing to their influence on complex man et al. 2005), Betula (Williams et al. 1999), Brassica (Hauser plant-pollinator interactions. For individuals of different spe- et al. 1997), Piriqueta (Wang & Cruzan 1998), Costus (Yost & cies that exist adjacent to one another, pollen exchange is pos- Kay 2009), and Chamerion angustifolium (Husband et al. 2002). sible via wind, and for those that exist in sympatry or Pollen-pistil interactions are more well characterized in syncar- parapatry, pollen exchange is possible via shared pollinators pous than apocarpous plants. (Arroyo & Dafni 1993; Brown & Mitchell 2001). Sagittaria is a genus of Alismataceae with conduplicate and Pollen-pistil interactions that lead mating barriers and prevent apocarpous carpels (Kaul 1976). There is incomplete fusion hybrid fertilizations are particularly important for the persistence of the mature carpel, which has an opening at the base due of incipient species that exist in sympatry. Hybrid fertilizations to the carpel margins being unfused (Huang 2014). In at reduce the fitness of one or both taxa as those ovules and sperm least four species of Sagittaria, pollen tubes from a carpel in are not used for conspecific matings (Levin 1975; Endress 1982; the gynoecium can either enter directly into an individual Armbruster et al. 2002; Husband et al. 2002; Ramsey et al. ovule or travel through the carpel basal opening and adjacent 2003). The pollen-pistil interactions of different taxonomic receptacle tissue to nearby unfertilized ovaries (Wang et al. groups employ varying strategies to maintain species integrity, 2002, 2006). The growth pattern of pollen tubes during such as inhibiting or retarding pollen germination on heterospecific pollination of these plants was unclear. There- heterospecific stigma (Hiscock & Dickinson 1993; Hodnett et al. fore, we used controlled pollination to examine the growth 2005), or unequal rates of pollen tube growth in the pistil (Car- of heterospecific pollen tubes in Sagittaria pygmaea gynoecia. ney et al. 1996; Klips 1999; Lankinene & Skogsmyr 2001). In We found that heterospecific and conspecific pollen tubes addition, several recent studies have reported mechanical mech- follow a different growth path in this species. This difference anisms of stigma closure and reopening (Sritongchuay et al. in pollen tube growth may represent a novel pollen-pistil

Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands 1 Alien pollen tube growth in Sagittaria species Lyu, Du & Wang interaction in apocarpous species, which may extend our recounted carpels in which the pollen tubes entered into the understanding of these interactions and their importance in ovules and passed into the surface layer of the receptacle tissue plant evolution. simultaneously (Fig. 1, Type VI). One-way analysis of variance (ANOVA), t-test, and statistical graphics were performed using the STATISTICA software package for Windows (single user MATERIAL AND METHODS version 6.0) StatSoft Inc. (2002). Plant materials and study site The two monoecious species S. pygmaea and Sagittaria trifolia RESULTS are both widely distributed in marshes, wetlands, and rice fields Pollen tube growth route after conspecific and heterospecific in the south of China. Their flowers have three white petals in pollination in S. pygmaea a spiral arrangement on the scapes and the flowering period is from May to November (Chen 1989). In total, 60 plants of the Conspecific pollen tubes grew through the recipient styles and two species were collected in 2013 from a rice field in Lichuan, reached the vicinity of the ovary at approximately 1 h AP Hubei Province, China (30°390 N, 109°180 E), and transplanted (Fig. 2A). At roughly 1.5 h AP, one tube began to turn toward into an outdoor pond at Wuhan University. the ovule and smoothly penetrated the micropyle. By approximately 2 h AP, a large proportion of the pollen tubes (44.74%) had entered the ovules. Redundant pollen tubes passed into Conspecific and heterospecific pollination and across the receptacle tissue to enter into other unfertilized ThefemaleflowersofbothS. pygmaea and S. trifolia were ovaries (Figs 2B and C, and 3A). bagged before flowering. The following day, flowers were polli- Heterospecific pollen tubes also reached the vicinity of the nated by applying pollen from either S. pygmaea or S. trifolia ovary at 1 h AP (Fig. 2D). Most heterospecific pollen tubes tan- directly onto the stigmata when the flowers opened naturally at gled together and displayed a crooked growth mode without any 0730 h which is when maximum pollen and stigma activity abnormal growth patterns. However, heterospecific growth tubes occurs in the two species (L Na, unpublished results) (herein were not seen to turn toward the ovules until 2.5 h AP (Fig. 3B). referred to as peak pollination). Pollinated flowers were then By 2.5 h AP, 5.13% of ovaries had received pollen tubes and harvested at 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, and 74.23% of carpels contained pollen tubes at the receptacle (Type 6.0 h after pollination (AP). We collected 3–6 flowers pollinated by each species at each sampling time and fixed them in formalin–acetic acid–alcohol (FAA). In addition, using a stereomicroscope, pollen grains from S. pygmaea were applied to a single carpel in the gynoecium of a pistillate flower from S. pygmaea, while the adjacent carpels were pollinated by S. trifolia. Pollinated flowers were collected 2 h AP and again fixed in FAA. Controlled pollination using pollen from S. trifolia on the gynoecia of S. pygmaea was also conducted at approximately 0930 h, meaning 2 h after the bagged flowers opened (herein referred to as late pollination). Similarly, we pollinated the gynoecia of S. pygmaea with S. trifolia at approximately 0630 h, meaning 1 h before peak pollination when the flowers were just opened (herein referred to as early pollination). Flowers were collected 2 h after early and late pollination and fixed in FAA.

Observation of pollen tube growth in pistils Sampled flowers were rinsed with water and placed in NaOH (5%, w/v) at room temperature until most of the tissues had become translucent. The flowers were again rinsed in water and then stained with aniline blue (0.1% in 0.03 M K3PO4) for 4 h (Wang et al. 2002). Carpels were randomly picked using tweezers and placed on slides for observation under a fluorescence microscope (Olympus BX-43, Tokyo, Japan). Fig. 1. Sagittaria pygmaea carpels classified into six types (I–VI) following Statistics and data analysis conspecific or heterospecific pollination, according to pollen tube front-end position. (I) Carpels with ungerminated pollen on the stigma. (II) Pollen ger- As shown in Fig. 1, the carpels were classified into six types mination and pollen tube beginning to emerge on the stigma. (III) Pollen (I–VI), according to pollen tube front-end position, with the tube navigating the style channel. (IV) Pollen tube arriving in the vicinity of exception of Type I that had carpels with ungerminated pollen. the ovary. (V) Pollen tube growing toward the ovule. (VI) Pollen tube passed At each time point from 0.5 to 6 h AP, we counted the carpels through the basal opening of the carpel toward the surface tissue layer of of each type using a fluorescence microscope. Furthermore, we the receptacle (arrow).

AB C

DE F

GH I

Fig. 2. Pollen tube growth in Sagittaria pygmaea carpels with conspecific (A, B and C) and heterospecific (D, E and F) pollen, and in Sagittaria trifolia carpels with heterospecific pollen (G, H and I) observed under a fluorescence microscope. (A) Pollen tubes of S. pygmaea pollen at 1 h after pollination (AP) on an S. pygmaea carpel. (B) Pollen tubes arriving at the base of the carpel, with one of them entering into the micropyle at 2 h AP. (C) One of the redundant pollen tubes passed into and across the receptacle, and entered a neighboring unfertilized ovary at 4 h AP. (D) Pollen tubes of S. trifolia 1 h after heterospecific pollination. (E) Pollen tubes of S. trifolia arriving at the base of the S. pygmaea carpel, but not travelling toward the heterospecific ovule, at 2 h AP. (F) Pollen tubes of S. trifolia travelling across the receptacle toward each other, with more than one pollen tube entering into the heterospecific ovary, at 4 h AP. (G) Pollen tubes of S. pygmaea on a S. trifolia carpel, at 1 h AP. (H) Pollen tubes of S. pygmaea arriving at the base of the S. trifolia carpel, but not entering into the ovules, at 2 h AP. (I) Pollen tubes of S. pygmaea travelling across the receptacle toward each other, with more than one pollen tube entering into the heterospecific ovary, at 4 h AP. Po, pollen; PT, pollen tube; St, stigma; Ov, ovule; Bars, 50 lm.

A 100 (I) (II) (III) (IV) (V) (VI) 80

40 Percentage (%) 20

0.5 h1.0 h1.5 h2.0 h2.5 h 3.0 h 3.5 h4.0 h4.5 h5.0 h5.5 h6.0 h Time after pollination B 100 (I) (II) (III) (IV) (V) (VI)

60 Fig. 3. (A) Percentage of Sagittaria pygmaea carpels with pollen tubes growing towards conspeciﬁc ovules at 40 different time points after pollination. (B) Percentage of

Percentage (%) S. pygmaea carpels with pollen tubes from Sagittaria 20 trifolia growing towards the ovules at different time points after pollination. I, ungerminated; II, stigma; III, 0 0.5 h1.0 h1.5 h2.0 h2.5 h 3.0 h3.5 h4.0 h4.5 h5.0 h5.5 h6.0 h style; IV, vicinity of the ovary; V, toward the ovule; VI, Time after pollination receptacle.

VI;Figs2EandF,and3B).Beforethistimepoint,mostofthe heterospecific pollen tubes passed the basal opening of the carpel and went straight into the surface tissue layer of the receptacle (Fig. 2E). This was also observed in outcrosses with S. trifolia as the maternal parent (Fig. 2G, H and I). Furthermore, 2 h AP following simultaneous deposition of conspecific and heterospecific pollen onto the stigma of S. pygmaea, conspecific pollen tubes entered into the ovule, while heterospecific pollen tubes went straight into the receptacle (Fig. 4).

Time course of heterospecific pollen tube growth toward the ovary in S. pygmaea A total of 3135 conspecific- and 3464 heterospecific-pollinated carpels were observed. As illustrated in Fig. 3A and B, the pollen tubes of S. trifolia reached the vicinity of the ovary of S. pygmaea at the same time (1 h AP) as conspecific pollen tubes. At 1.5 h AP, 18.62% of the carpels subject to conspecific pollination contained a pollen tube turning toward the ovules, and 13.52% of the carpels showed a redundant pollen tube navigating into the receptacle tissue. Very few heterospecific pollen tubes grew towards the ovules, or even exhibited a turning ten- dency, until 2.5 h AP. A significantly lower percentage of carpels subjected to heterospecific pollination contained pollen tubes that turned toward the ovary at 2.5 h AP than did those subjected to conspecific pollination (t = 7.184, d.f. = 12, P < 0.01). At 5.5 h AP, more than 80% of the ovules contained pollen tubes that reached the vicinity of the ovules. At that time, there was no difference in the proportion of carpels with conspecific Fig. 4. The growth path of pollen tubes of S. trifolia and S. pygmaea at the pollen tubes and those with heterospecific pollen tubes reaching carpels of S. pygmaea at 2 h AP. Heterospecific pollen tubes arriving at the = = > the vicinity of the ovules (t 0.37, d.f. 14, P 0.05). How- base of the carpel, but not entering into the ovules (left), and conspecific ever, heterospecific pollen tubes arrived in the vicinity of the pollen tubes already entering into the ovules (right). Po, pollen; PT, pollen ovules more than 1 h later than did conspecific pollen tubes. tube; St, stigma; Ov, ovule; Bars, 50 lm.

Table 1. One-way was used to compare the average percentages of Sagittaria pygmaea carpels that had pollen tubes derived from Sagittaria trifolia and S. pygmaea growing toward the ovules 2 h after peak, early, or late pollination. I, ungerminated; II, stigma; III, style; IV, vicinity of the ovary; V, toward the ovule; VI, receptacle.

S. pygmaea (♀) 9 S. pygmaea (♀) 9 S. pygmaea (♀) 9 S. pygmaea (♀) 9 The parameters S. trifolia (♂) (P) S. trifolia (♂) (E) S. trifolia (♂) (L) S. pygmaea (♂) (L)

I(%) 11.33 5.74 11.14 14.70 17.15 17.08 2.86 10.93 II(%) 4.05 3.29 0.71 1.69 0.44 1.07 0.15 1.92 III(%) 7.05 6.42 3.36 3.67 1.49 2.24 0.15 1.32 IV(%) 27.96 21.17 2.58 4.60 54.86 15.23 1.43 5.39 V(%) 0.55 1.13a 1.76 3.23a 22.45 12.73b 47.82 14.76c VI(%) 50.06 27.03 80.45 13.92 3.61 6.59 47.59 14.81

P = peak pollination (at the normal time); E = early pollination; L = late pollination. Different superscripts differ signiﬁcantly at P < 0.05.

Fig. 5. (A) Pollen tubes of S. trifolia entering into the ovules of the S. pygmaea, at 2 h after late pollination. (B) Pollen tubes of S. pygmaea arriving at the base of the S. pygmaea carpel, but not entering into the ovules, at 2 h after early pollination. Po, pollen; PT, pollen tube; St, stigma; Ov, ovule; Bars, 50 lm.

Heterospecific pollen tube growth in the carpels of S. pygmaea DISCUSSION at different pollination times The pollen–pistil interaction involves a complex series of cellu- After heterospecific peak pollination, the majority of carpels lar and molecular processes associated with active discrimina- had a pollen tube that grew through the basal opening of tion and rejection of ‘incompatible’ pollen at interspecific and the carpels and entered into the surface tissue layer of the intraspecific levels. The pollen-pistil interaction is a fundamen- receptacle. And after heterospecific early pollination, only a tal process in the reproductive biology of angiosperms and has much smaller proportion of carpels contained pollen tubes been the focus of intense research for many decades (Hiscock that reached the vicinity of the ovules and then entered into and Allen, 2008), especially in interspecies hybridization. the micropyles (Table 1; P = 0.407). Furthermore, two hours Heterospecific pollen may not germinate on the stigma, or the after either heterospecific early or peak pollination, the pro- growth of pollen tubes following germination may be prema- portion of carpels with a pollen tube reaching the vicinity of turely terminated, such as is seen for interspecies hybridization the ovules was significantly lower than that 2 h after late between Kunzea pomifera and K. ericoides (Myrtaceae) (Page pollination (Table 1; P = 0.00, Fig. 5A). In addition, in et al. 2010). In our study, pollen from S. trifolia germinated on instances where the maternal species was S. trifolia,we the stigma of S. pygmaea, and heterospecific pollen tubes observed that only 1.73% of carpels contained pollen tubes passed through the basal opening of the carpel and the surface that entered into the receptacle rather than into the ovules. tissue layer of the receptacle to finally enter into nearby ovaries. Regardless of the maternal species in heterospecific These different growth patterns for heterospecific pollen tubes hybridization, after late pollination, most of examined car- were observed for the first time in apocarpous Sagittaria plants, pels contained pollen tubes that turned straight to the ovary which have a unique gynoecium structure. and ovule rather than entering the surface tissue layer of the In both conspecific mating and heterospecific hybridization, receptacle. pollen tube growth through the tissues of the gynoecium is

Plant Biology © 2016 German Botanical Society and The Royal Botanical Society of the Netherlands 5 Alien pollen tube growth in Sagittaria species Lyu, Du & Wang influenced by developmental states of the carpel and pollen. sporophytic tissues (Shimizu and Okada, 2000; Shimizu, 2002; For example, Kandasamy et al. (1994)found that in very young Higashiyama et al. 2006; Palanivelu and Preuss, 2006; Kikuchi buds of Arabidopsis or Brassica that had undergone interspecies et al. 2007). We could hypothesize that pollen tubes from the hybridization, the growth pattern was the same for both two tested species may use the same signal systems in the heterospecific and conspecific pollen tubes. These authors also gynoecia but deploy of the gynoecia signal(s) may be sequen- found that Brassica pollen tubes could grow regardless of the tial. The early-stage signal may be sensed by both conspecific maturity of Arabidopsis pistils, whereas Arabidopsis pollen and heterospecific pollen tubes, whereas the late-stage signal could only germinate and grow pollen tubes on immature may be sensed by heterospecific pollen tubes until the attrac- Brassica pistils. In our study, most S. pygmaea carpels that had tant concentration reaches a threshold level. heterospecific pollen grains deposited on the gynoecia 1 h ear- In conclusion, we observed a new pattern of heterospecific lier than peak pollination had a pollen tube going straight pollen tube growth that occurs in apocarpous species following through the basal opening of the carpels and entering into the an inharmonious pollen-pistil interaction, which may represent surface tissue layer of the receptacle. This pattern was indistin- a novel mechanism of conspecific pollen advantage (CPA). guishable from that of heterospecific pollen tubes deposited at However, we should not overlook the fact that our present peak pollination and conspecific pollen tubes deposited at early results and those obtained by Higashiyama (2006) for Torenia pollination (Fig. 5B), almost all of the carpels with the pollen were gathered following single-parent pollination. In Narcissus tubes going straightly to the surface tissue layer of the recepta- triandrus, ovules cease to develop in response to a ‘long- cle. However, following late pollination, the growth pattern distance’ stimulus provided by conspecific pollen tubes (Sage exhibited by heterospecific pollen tubes was similar to that et al. 1999); i.e., pollen tubes regulate the development of shown after conspecific peak pollination. ovules. Pollen tube performance may also be affected by male- Pollen tubes are guided to the ovule micropyle by signals male interaction in the gynoecium (Bookman 1984; Weller & originating in the cells and the extracellular matrix of the pistil Ornduff 1991; Walsh & Charlesworth 1992). Future studies (Lord 2003). Cross fertilization between Torenia fournieri and should examine pollen tube growth over time following mixed- twelve related species displayed five stages of abnormal pollen parent pollination of gynoecia. tube growth, which correspond to the multiple pollen tube guidance signals currently known in Arabidopsis (Johnson and ACKNOWLEDGEMENTS Preuss, 2002; Kikuchi et al. 2007). In this study, aberrant pollen tube growth was not observed in the pollen tube channel from We thank Wen-Long Fu, Xiao-Wen Wang, and Shuai Wang the stigma to the base of the carpel. Unequal rates of for their help in sample collection and technical matters and heterospecific pollen tube growth in the vicinity of the ovary Lan-Jie Huang for his constructive opinions on the paper. may be the result of interaction, or lack thereof, with species- Financial support was provided by the National Natural specific guidance signal(s) from female gametophytic or Science Foundation of China (NSFC 30970194, 31270282).

45 – REFERENCES asymmetric hybridization in Louisiana Irises. Evolu- divergent Sorghum species. Crop Science, , 1403 tion, 50, 2201–2206. 1409. Adler L.S., Irwin R.E. (2006) Comparison of pollen Endress P.K. (1982) Syncarpy and alternative modes of Huang L.J. (2014) The Structure and development of transfer dynamics by multiple floral vistors: experi- escaping disadvantages of apocarpy in primitive incompletely closed carpels in an apcarpous species, ments with pollen and fluorescent dye. Annals of angiosperms. Taxon, 31,48–52. Sagittaria trifolia (Alismataceae). America Journal of Botany, 97,141–150. Fishman L., Aagaard J., Tuthill J.C. (2008) Toward the Botany, 101, 1229–1234. Armbruster W.S., Debevec E.M., Willson M.F. (2002) evolutionary genomics of gametophytic divergence: Husband B.C., Schenske D.W., Burton T.L., Goodwillie Evolution of syncarpy in angiosperms: theoretical patterns of transmission ratio distortion in Mon- G. (2002) Pollen competition as a unilateral repro- and phylogenetic analysis of the effect of carpel keyflower (Mimulus) hybrids reveal a complex ductive barrier between sympatric diploid and tetra- fusion on offspring quantity. Journal of Evolutionary genetic basis for conspecific pollen precedence. Evo- ploid Chamerion angustifolium. Proceedings of the Biology, 15, 657–672. lution, 62, 2958–2970. Royal Society of London Series B, 269, 2567–2571. Arroyo J., Dafni A. (1993) Interspecific pollen transfer Hauser T.P., Jørgensen R.B., Østergard H. (1997) Pref- Johnson M.A., Preuss D. (2002) Plotting a course: mul- among co-accurring heteromorphic and homomor- erential exclusion of hybrids in mixed populations tiple signals guide pollen tubes to their targets. phic species. Israel Journal of Botany, 41, 225–232. between oilseed rape (Brassica napus) and weedy B. Developmental Cell, 2, 273–281. Bookman S.S. (1984) Evidence of selective fruit pro- campestris (Brassicaceae). American Journal of Bot- Kandasamy M.K., Nasrallah J.B., Nasrallah M.E. (1994) duction in Asclepias. Evolution, 29, 467–474. any, 84, 756–762. Pollen-pistil interactions and developmental regula- Brown B.J., Mitchell R.J. (2001) Competition for polli- Higashiyama T., Inatsugi R., Sakamoto S., et al. (2006) tion of pollen tube growth in Arabidopsis. Develop- nation: effects of pollen of an invasive plant on seed Species preferentiality of the pollen tube attractant ment, 120, 3405–3418. set of a native congener. Oecologia, 129,43–49. derived from the synergid cell of Torenia fournieri. Kaul R.B. (1976) Conduplicate and specialized carpels Carney S.E., Hodges S.A., Arnold M.L. (1996) Plant Physiology, 142, 481–491. in the Alismatales. American Journal of Botany, 63, Effects of differential pollen-tube growth on Hiscock S.J., Dickinson H.G. (1993) Unilateral 175–182. hybridization in the Louisiana irises. Evolution, incompatibility within the Brassicaceae: further Kikuchi S., Kino H., Tanaka H., Tsujimoto H. (2007) 50, 1871–1878. evidence for the involvement of the self-incompat- Pollen tube growth in cross combinations between Chapman M.A., Forbes D.G., Abbott R.J. (2005) Pollen ibility (S)-locus. Theoretical and Applied Genetics, Torenia fournieri and fourteen related species. Breed- competition among two species of Senecio (Aster- 86, 744–753. ing Science, 57, 117–122. aceae) that form a hybrid zone on Mt. Etna, Sicily. Hiscock S.J., Allen A.M. (2008) Diverse cell signalling Klips R.A. (1999) Pollen competition as a reproductive American Journal of Botany, 92, 730–735. pathways regulate pollen-stigma interactions: the isolating mechanism between two sympatric Hibiscus Chen J.K. (1989) systematic and evolutionary studies on search for consensus. New Phytologist, 179, species (Malvaceae). American Journal of Botany, 86, Sagittaria from China. Wuhan University Press, 286–317. 269–272. Wuhan, China. Hodnett G.L., Burson B.L., Rooney W.L., Dillon S.L., Lankinene A., Skogsmyr I. (2001) Evolution of pistil Emms S.K., Hodges S.A., Arnold M.L. (1996) Pollen Price H.J. (2005) Pollen-pistil interactions result in length as a choice mechanism for pollen quality. tube competition, siring success, and consistent reproductive isolation between Sorghum bicolor and Oikos, 92,81–90.

Lee C.B., Page L.E., Mcclure B.A., Holtsford T.P. (Asteraceae). American Journal of Botany, 82, Wang J., Cruzan M.B. (1998) Interspecific mating in (2008) Post-pollination hybridization barriers in 515–519. the Piriqueta caroliniana (Turneraceae) complex: Nicotiana section Aletae. Sexual Plant Reproduction, Sage T.L., Strumas F., Cole W.W., Barrett S.C.H. effects of pollen load size and composition. Ameri- 21, 183–195. (1999) Differential ovule development following can Journal of Botany, 85, 1172–1179. Levin D.A. (1975) Minority cytotype exclusion in local self- and cross-pollination: the basis of self-sterility Wang X.F., Tao Y.B., Lu Y.T. (2002) Pollen tubes enter plant populations. Taxon, 24,35–43. in Narcissus triandrus (Amaryllidaceae). American Neighbouring ovules by way of receptacle tissue Lord E.M. (2003) Adhesion and guidance in compati- Journal of Botany, 86, 855–870. resulting in increased fruit-set in Sagittaria potamo- ble pollination. Journal of Experimental Botany, 54, Sannier J., Baker W.J., Anstett M.-C., Nadot S. (2009) getifolia Merr. Annals of Botany, 89, 791–796. 47–54. A comparative analysis of pollinator type and pollen Wang X.F., Tan Y.Y., Chen J.H., Lu Y.T. (2006) Pollen Mitchell J.R., Waser N.M. (1992) Adaptive significance ornamentation in the Araceae and the Arecaceae, tube reallocation in two preanthesis cleistogamous of Ipomopsis aggregata nector production: pollina- two unrelated families of the monocots. BMC species, Ranalisma arostratum and Sagittaria guya- tion success of single flowers. Ecology, 73, 633–638. Research Notes, 2, 145–155. nensis ssp. lappula (Alismataceae). Aquatic Botany, Page T., Moore G.M., Will J., Halloran G.M. (2010) Shimizu K.K. (2002) Ecology meets molecular genetics 85, 233–240. Breeding behaviour of Kunzea pomifera (Myrtaceae): in Arabidopsis. Population Ecology, 44, 221–233. Weller S.G., Ornduff R. (1991) Pollen tube growth self-incompatibility, intraspecific and interspecific Shimizu K.K., Okada K. (2000) Attractive and repul- and inbreeding depression in Amsinckia grandi- cross-compatibility. Sexual Plant Reproduction, 23, sive interactions between female and male gameto- flora (Boraginaceae). American Journal of Botany, 239–253. phytes in Arabidopsis pollen tube guidance. 78, 801–804. Palanivelu R., Preuss D. (2006) Distinct short-range Development, 127, 4511–4518. Williams J.H., Friendman W.E., Arnold M.L. (1999) ovule signals attract or repel Arabidopsis thaliana Sritongchuay T., Bumrungsri S., Meesawat U., Mzaer Developmental selection within the angiosperm pollen tubes in vitro. BMC Plant Biology, 6,7. S.J. (2010) Stigma closure and re-opening in Oroxy- style: using gamete DNA to visualize interspecific Ramsey J., Bradshaw H.D., Schemske D.W. (2003) lum indicum (Bignoniaceae): causes and conse- pollen competition. Proceedings of the National Components of reproductive isolation between the quences. American Journal of Botany, 97, 136–143. Academy of Sciences, 96, 9201–9206. monkerflowers Mimulus leswisii and M. cardinalis StatSoft, Inc. (2002) Statistica (data analysis software Yost J.M., Kay K.M. (2009) The evolution of postpolli- (Phrymaceae). Evolution, 57, 1520–1534. system), version 6. StatSoft, Inc., Tulsa, OK, USA. nation reproductive isolation in Costus. Sexual Plant Rieseberg L.H., Desrochers A.M., Youn S.J. (1995) Walsh N.E., Charlesworth D. (1992) Evolutionary Reproduction, 22, 247–255. Interspecific pollen competition as a reproductive interpretation of differences in pollen tube growth barrier between sympatric species of Helianthus rates. Quarterly Review of Biology, 67,19–37.