ISSN 0373-580 X Bol. Soc. Argent. Bot. 36 (3-4): 279 - 282. 2001

SELF- AND CROSS-POLLEN TUBE GROWTH RATES IN LONGIFLORA ()

LUIS M. PRIMO and LEONARDO GALETTO1

Summary:We measured in vivo pollen tube growth rate of Nicotiana longiflora to evaluate the effects of the pollen source. Pollen readily germinated following outcross-, self-, and mixed pollinations. On average, pollen tubes grew at a constant rate (1.8 mm tv1). Two' days after pollination, pollen tubes had reached the ovules. Differences in the rate of pollen tube growth may reflect the ability of self- and outcross-pollen grains to effect fertilization. It is reasonable to expect that the fitness cost of autogamy can promote selection on traits which will eventually reduce the possibility of self-fertilizations (e.g., a slower growth rate In self-pollen tubes). Nevertheless, N. longiflora showed that self-pollen tubes were faster than outcross- pollen tubes. Available published data indicate that pollen tube competition does not appear to be a general mechanism utilized by for avoiding self-fertilization. Key Words: Nicotiana longiflora, pollen performance, pollen source, pollen tube growth.

Resumen: Tasa de crecimiento de los tubos polínicos provenientes de autopolinizaciones y polinizaciones cruzadas en Nicotiana longiflora (Solanaceae). Con el objetivo de evaluar los efectos de la fuente de polen, medimos la tasa de crecimiento in vivo de los tubos polínicos en Nicotiana longiflora. El polen germina rápidamente luego de realizar autopolinizaciones, polinizaciones cruzadas o con polen mixto. En promedio, los tubos polínicos crecen a una tasa constante (1,8 mm ir1). Los tubos polínicos alcanzan los óvulos dos días después de realizada la polinización. Las diferencias en la tasa de crecimiento dé¬ los tubos polínicos pueden reflejar la capacidad de polen autógamo o xenógarho para efectuar la fertili¬ zación. Es razonable esperar que los costos de supervivencia provocados por la autogamia, puedan promover la selección de caracteres de la planta que eventualmente reduzcan la posibilidad de fertilizaciones por polen autógamo (por ejemplo, una tasa de crecimiento más lenta en los tubos polínicos autógamos). Sin embargo, en N. longiflora los tubos polínicos autógamos fueron más rápidos que los xenógamos. Los datos publicados hasta el momento indican que la competencia de tubos polínicos no sería un mecanismo general utilizado por las plantas para evitar las fertilizaciones autógamas. Palabras Clave: Nicotiana longiflora, desempeño del polen, fuente de polen, crecimiento del tubo polínico.

INTRODUCTION origin may be one factor that can affect pollen tube performance inthe style. Whenever self-pollen grains In many flowering plants the initial phase of re¬ arrive at the stigma not only do they germinate suc¬ production(i.e., frompollination to fertilization) de¬ cessfully, but also their pollen tubes can grow down pends on the pollination and on the regulation of the style at the same rate than outcross-pollen tubes pollen-pistilinteractions. These interactions involve do. In some species, self- and outcross-pollen tubes recognition signals that may affect pollen tube may differintheir growthrate (e.g. Cruzan, 1989, 1990; growth through the style, and directional guidance Hessing, 1989; Aizenetal., 1990).If self-pollen tubes ofpollen tubes toward the ovules (e.g.,Newbigin et can grow down the style in N. longiflora, there is a al., 1993; Bell, 1995; Herrero & Hormaza, 1996; possibility that geitonogamous pollen deposited Nettancourt, 1997). Thepistilprovidesnot only sup¬ withinthe plant willachieve fertilizationin competi¬ port but it can also constrain pollen.tubes growth tion with outcross-pollen. (Herrero & Hormaza, 1996). Therefore, fertilization Nicotiana longiflora Cav. is a perennial herb of success and subsequent fruit development may be 0.5-1.5 mheight, with a withish tubular corolla of 9-, influencedby different factors andmechanisms that 12 cm length. This species provides an ideal system operate from pollination to'seed maturation. Pollen to study pollen tube growth and to measure puta¬ tive differences between self- and outcross-pollen 1 1nstituto Multidisciplinario de Biología Vegetal (Universidad tubes because of its long style (ca. 100 mm). This Nacional de Córdoba-CONICET), Casilla de Correo 495, South American species is mainly distributed in 5000, Córdoba, Argentina, e-mail: [email protected].

279 Bol. Soc. Argent. Bot. 36 (3-4)' 2001

northern Argentina andParaguay, andit is frequently rate. The different treatments applied to each inflo¬ found under full sun or partial shade sites as open rescence were: (i) self-pollination, 30 flowers were fields or roadsides (Goodspeed, 1954). Self-compat¬ pollinated with pollen of the same flower, (ii) out- ibility was mentionedby Pandey (1979) for cultivated cross-pollination, 30 flowers were pollinated with plants of N. longiflora, a species trait confirmed by pollen of plants located at > 20 m distance, and (iii) Primo (1997) for natural populations from Córdoba, mixedpollination,30 flowers werepollinatedwithpol¬ Argentina. len from one anther ofthe same flower and from one In this study, we examined the possible effects anther of a plant located at > 20 m distance. Mixed- of the pollen source on the pollen tube growth rate pollination treatment was done because the possi¬ in N. longiflora. If pollen tube growth depends on bility that pollen tube growth rate may be different thepollen source, we expect differencesbetween self- compared to the other two treatments. Pollinations and outcross-pollen. were done simultaneously to all flowers within the first five hours after opening. Then, 2-3 flowers of each treatment were cut from the plants and gyno- MATERIAL AND METHODS ecia were placed and stored in ethanol 70 % at the following successive times: 4, 6, 12, 14, 18,22,26,28, Plant material 36, 40, 43,and46hafterpollinations wereperformed. Datawereobtainedfrom30differentindividuals from Squashes were done according to Dafni (1992) with anaturalpopulationofN. longiflora locatedinCórdoba some modifications: gynoecia were softened with (31°25’ S, 64° 10’ W,Dept. Capital,Prov. Córdoba, Ar¬ sodiumhydroxide 0.8Nfor 3-4h, washedtwice with gentina), where it occupies an area of ca. 4 ha. Voucher distilledwater, stainedwith'anilineblue for 15-30min, specimens were deposited at Museo Botánico de and squashed under a cover-slip to spread the sty- Córdoba(CORD,Galetto&Primo 655-657). lar tissue. Pollen tubes were finally observed under an epifluorescence microscope (Zeiss Axiophot Pollen tube growth equipped withUV filters). A singleparameter, the dis¬ Flower life span was determined by tagging two tance that pollen tubes had grown into the style (i.e., recently open buds in each plant arid following their the greatest distance up to which most pollen tubes development.Timeintervals forpollentubegrowthwere hadreached down the style; this pollen tube front dis¬ chosen from preliminary observations on the pollen tance is evident for its remarkable fluorescence), was tube growthrate fromadditional flowers andalsobased measuredwithadigitalcaliper. onthe flower Ufe span. This species possess ameanof 2256 ± 700 ovules per ovary and a mean of 880500 ± Data analysis 188641pollengrainsper flower (Fernández & Galetto, All distributions were tested for randomness of unpublished data). Because differences in the amount nominal data (Rims test), homogeneity of variances ofpollen grains deposited onto stigmas can influence (Levene test), and departures from normality the pollen tube growth rate (e.g. Cruzan, 1986; (Kolmogorov-Smimov test for goodness of .fit). A Niesenbaum& Schueller, 1997),hand-pollinations were regression analysis was performed to evaluate pol¬ performedusingtwonew anthers frombagged flowers len tube growth rates for each treatment. Compari- in order to standardize pollen load sizes. We touched son of regression lines of pollen tube growth rates one side of the stigma with one anther and the other betweenpollination treatments were done by analy- half-size with the other anther. Considering that one sis of covariance (ANCOVA). We specifically ask if anther has ca. 175000pollen grains, we were sure that there are differences among groups for the pollen finalpollenloaddepositedonstigmas largely exceeded tube growth rate (dependent variable) when differ- the total number of ovules per ovary. In addition, enees among groups in the time after pollination(in- xenògamouspollinations were carriedoutusingpollen dependent variable) are considered (Sokal & Rohlf, of individuals growing at moderate distances (i.e, >20 1995). A suitable significance level was calculated m) from the maternal plants because of.the potential by using Bonferroni’s approximation (a/m), where influences ofoutbreedingdepressiononthepollentube a is the normally acceptableprobability andm is the growthrate (Waser, 1993 and citations therein). number ofcomparisonsmade (Fry, 1996). The statis- Bagged inflorescences on with recently open ticalprogrampackage SPSS (1992) wasusedfor these flowers were used to compare pollen tube growth analyses.

280 L. M. Primo & L. Galetto, Pollen tube growth rates in Nicotiana longiflora

RESULTS TREATMENT ' Flower traits Flower opening always occurs in the afternoon. f 1 * Flowers last four days. After this period, flowers’ |, 60 s present some wilting of their corollas. Five stamens are located at the upper portion of the corolla tube and have short filaments. Anthers dehiscence takes - place during the first night. The exerted stigma is re¬ X ceptive from the moment of flower opening and R * throughout the flower lifetime. Stigma-anther dis¬ a 30 i, tance usually exceeds +4 mm (range 2-6.mm). The hours after pollination recéptive stigmatic zone includes the entire surface Fig. 1. In vivo pollen tube growth rate under different pollina¬ of the stigma. tion treatments in Nicotiana longiflora. Regression equations are statistically significant (P<0.0001) where x is time (hours) Pollen tube growth andy is pollen tube length (mm): self-pollination, y = 0.66 + \.19x = = Pollenreadily germinated followingoutcross-,self, (r2 0.90); mixed-pollination, y -5.27 + 1.86x (r2 = 0.96); outcross-pollination, y = -7.54 + 1.87x (r2 = 0.87). Each point and mixed-pollinations. On average, two days after correspond to a flower data. pollination the pollen tubes had reached the ovules, wfiichare located at about 9.0-100mm fromthe stigma an increasing floral age. An “age” effect canbe dis- (average growth rate of pollen tubes =1.8 mm h1). charged in N. longiflora because pollinations were With the ANCOVA, we tested whether the meanpol- done as soon as flowers opened. On the other hand, len tube growthrate is different for the three pollina- competitive environment effects were not experimen- tion treatments. The intercepts and the slopes were tally considered here. significantly different(Table 1). Throughmultiple tests, Differences inthe rate ofpollentube growthmay separatepairs oflines were compareduntilit was clear reflect the ability of self- and outcross-pollen grains whichlines were different. Self-Pollentubes were faster to effect fertilization. Considering that seed germi¬ than outcross-pollen tubes (Fig. 1). nation is significantly reduced by autogamy in N. longiflora (unpublished data), it is reasonable to expect that the fitness cost of geitonogamy can Table 1. ANCOVA table for the comparison of regression pro¬ lines of pollen tube growth rates between pollination mote selection on plant traits which will eventually treatments (see Material & Methods). reduce the possibility of self-fertilizations (e.g., a slower growthrate in self-pollen tubes, the presence Source of Variation df . SS MS F P of an Pollination Treatment 2 621.1 310.4 4.66 0.01 incompatibility system). Nevertheless, N. Error 86 5736.2 66.7 longiflora is a self-compatible species (Primo & Galetto, unpublished data) andhere we showed that self-pollen tubes were faster than outcross-pollen DISCUSSION tubes. Recent investigations on several unrelated species showed that a slower growth of seíf-pollen Many factors may differentially influencepollen tubes outcross-pollen tubes was found in some of performance (e.g.Fenster & Sork, 1988; Stephenson them but not in others (Johnston, 1993; Snow & et al., 1992a; Niesenbaum & Schueller, 1997). The Spira, 1993 and references therein). Thus, when the competitive environment (e.g., the number of grains few available data up to now are considered, pollen deposited onto a stigma or the number of pollen .tube competition does not appear to be a general tubes per pistil) and the flower age at the moment of mechanism for avoiding self-fertilization. This con- pollination are factors witha clear ecological signifi- trasting pattern of pollen tube growth variation canee. Stephenson et al. (1992b) reported a style- among species implies strong selection promoting age effect on the performance of self- and outcross- pollenperformance. Anadaptive explanationofpol- pollen tubes in Campanula rapunculoides lenperformance variation in angiosperms wouldlie (Campanulaceae). They foundareductioninthe abil- in understanding how pollen tube growth affects ity of some plants to prevent self-fertilization with plant reproductive success.

281 Bol. Soc. Argent. Bot. 36 (3-4) 2001

ACKNOWLEDGEMENTS JOHNSTON, M. O. 1993. Tests of two hypotheses con¬ cerning pollen competition in a self- compatible, long- styled species (Lobelia cardinalis: Lobeliaceae). Amer. The authors thank M. Aizen, R. Aguilar, G. J. Bot. 80: 1400-1406. Bemardello, andC. Torres for theiruseful comments NETTANCOURT, D. de 1997. Incompatibility in an- on early versions of this manuscript, and “Consejo giosperms. Sex. Plant Reprod. 10: 185-199. Nacional de Investigaciones Científicas y Técnicas” NEWBIGIN, E„ M. A. ANDERSON & A. E. CLARKE. (CONICET),“Consejo deInvestigaciones Científicas 1993. Gametophytic self-incompatibility systems. Plant y Técnicas de la Provincia de Córdoba”, “Agencia Cell 5: 1315-1324. Nacional de Promoción Científica y Tecnológica”, NIESENBAUM, R. A. & S. K. SCHUELLER. 1997. Effects “Secretaría de Ciencia y Técnica de la Universidad of pollen competitive environment on pollen perfor¬ mance in Mirabilis jalapa (Nyctaginaceae). Sex. Plant Nacional de Córdoba”, and “Agencia Córdoba Reprod. 10: 101-106. Ciencia” for financial support. LG is a member of PANDEY, K. K. 1979. The genus Nicotiana: evolution of CONICET. incompatibility in flowering plants. In: HAWKES, J. G., R.N. LESTER & A. D. SKELDING (eds.), The Biology and of the Solanaceae, pp. 421-433. Aca¬ BIBLIOGRAPHY demic Press, Linnean Society Symposium, London. PRIMO, L. M. 1997. El sistema reproductivo de Nicotiana longiflora Cav. (Solanaceae): calidad de frutos y semillas AIZEN, M. A., K. B. SEARCY & D. L. MULCAHY. 1990. y su relación con los polinizadores y estrés hídrico. Tesis Among- and within-flower comparisons of pollen tube de Grado, F.C.E.F.N., U.N.C. growth following self- and cross-pollinations in Dianthus chinensis (Caryophyllaceae). Amer. J. Bot. 77: 671-676. SNOW, A. A. & T. P. SPIRA. 1993. Individual variation in the vigor of self pollen and selfed progeny in Hibiscus BELL, P. R. 1995. Incompatibility in flowering plants: Ad¬ moscheutos (Malvaceae). Amer. J. Bot. 80: 160-164. aptation of an ancient response. Plant Cell 7: 5-16. SOKAL, R. R. & F. J. ROHLF. 1995. Biometry. W. H. Free¬ CRUZAN, M. B. 1986. Pollen tube distribution in Nicoti- man, San Francisco, California. ana glauca: Evidence for density dependent growth. Amer. J. Bot. 73: 902-907. SPSS Inc. 1992. SPSS for Windows: base system user’s guide, Release 5.0. SPSS Inc., Chicago. CRUZAN, M. B. 1989. Pollen tube attrition‘in Erythronium grandiflorum. Amer J. Bot. 76: 562-570. STEPHENSON, A. G„ T. LAU, M. QUESADA & J. A. WINSOR. 1992a. Factors that affect pollen perfor¬ CRUZAN, M. B. 1990. Pollen-pollen and pollen-style in¬ mance. In: WYATT, R. (ed.), Ecology and evolution of teractions during pollen tube growth in Erythronium plant reproduction, pp. 119-136. Chapman & Hall, New grandiflorum (Liliaceae). Amer J. Bot. 77: 116-122. York. DAFNI, A. 1992. Pollination Ecology. A practical approach. STEPHENSON, A. G., J. A. WINSOR, T. E. RICHARDSON, Oxford University Press, Oxford. A. SINGH & T. KAO. 1992b. Effects of style age on FENSTER, C. B. & V. L, SORK. 1-988. Effect of cross¬ the performance of self and cross pollen in Campanula ing distance and male parent on in vivo pollen tube rapunculoides. In: OTTAVIANO, E., D. L. MULCAHY, growth in Chamaecrista fasciculata. Amer. J. Bot. M. SARI GORLA & G. BERGAMINI MULCAHY (eds.), 75: 1898-1903. Angiosperm pollen and ovules, pp.117-121. Springer- Verlag, New York. FRY, J. C. 1996. Biological data analysis. A practical ap¬ proach. Oxford University Press, Oxford. WASER, N. 1993. Population structure, optimal outbreed¬ ing, and assortative mating in angiosperms. In: GOODSPEED, T. H. 1954. The genus Nicotiana. Chron. THORNHILL, N. W. (ed.), The natural history of in- Bot. 16: 1-536: breeding and outbreeding, pp. 173-199. The Univer- HERRERO, M. & J. I. HORMAZA. 1996. Pistil strategies * sity of Chicago Press, Chicago. controlling pollen .tube growth. Sex. Plant Reprod. 9: 343-347. HESSING, M. B. 1989. Differential pollen tube success in Recibido el 11 de Abril de 2001, aceptado el 10 de Julio de Geranium caespitosum. Bot. Gaz. 150: 404-410. 2001'.

282