Rey. Biol. Trop., 44(3)/45(1 ): 79-86, 1996-1 997
Germination of Tabebuia heterophylla seeds (Bignoniaceae) from a wet and dry forest of Puerto Rico
Roberto A. Cordero1and Brenda Molano-Flores.2 1 DepartmentofBiology, Universityof Puerto Rico, RioPiedras, Puerto Rico00931-3360, U.S. \. 2 DepartmentofBiological Sciences, Division ofBotany, University ofIowa, Iowa City,Iowa 52242, U.SA
(Rec. 14-I1-1996. Rey. 2J-V-1996. Accept.8-V III-1996)
Abstract: Seedgermination response of thePuerto Rican wá anddry forest populations of Tabebuia heterophyllatreo::s was testedusing a gradient of osmotic potentials from O to -1.5 MPa. Morphological comparisons were also made from adult specimens. Dryforest treo::s showed smaller leaves, fiuits,and seeds, andgreater specific leafweigjrt. Dry forestfiuits produced smallerseeds th an wá forest fiuitswhen similar small fiuitswere compared. Germination percentage was stronglyreduced as osmotic potential decreased, and was significantin both seedorigins. This reductioogreat was er 00 seeds fromthe wá forest population. Days for germination showedthe same responsebut thiswas less evident inthe seedsfrom the dryforesto Osmotic potential lower than -0.9 MPa. compláely arrested seed germination in both populations. No sigJlificant difference in final germination percentage anddays to germinate exist at higll osmotic potentials in both seedpopulations. This physiological capacity along with the morphological modifications indicates why this species can maintain viable populations in the contrasting environmental conditionsofthe dryand moist forest habitats.
Key words: osmotic stress, phenotypic yariation, PEG, seed size, seed germination, Tabebuia.
Because morpho-physiological variations of populations (Schlichting 1986), but not many tropical plants have been related to the of the studies include tropical tree forros existence of water gradients, in which plants (Sternmerman 1983, Abrams et al. 1992). At resist or avoid moisture stress (Medina 1983), the seed level, sensitivity of seed germination we expected that sorne morphological to water stress has been reported among and differences within species existed in response within temperate tree species (Vuillemin to stressing environments. Traits such as the 1982, Falusi et al.1983). The importance of presence of sclerophyllous leaves, deep root germination time for the establishment and systems, deciduousness, microphylly and growth of seedlings in a semideciduous compound leaves, greater belowground tropical forest is known (Augspurger 1979, biomass, and tissue osmotic adjustment Garwood 1986), and underscores how seed (Sobrado and Cuenca 1979, Medina 1983) are leve! processes related to water stress are of common adaptations to water stress, but they primary ecological interest. Vegetative have been mostly treated at the interespecific variability among Tabebuía species of West leve!. At the intraspecific level, phenotypic lndies has been noted, and is often associated plasticity has been demonstrated among with edaphic specialization (Gentry 1990). 80 REVISTA DE BIOLOGIA TROPICAL
Tabebuia heterophylla (De). Britton site during Juiy 1991. The length and width of (Bignoniac.eae) is no exception (Longwood the largest leaflet, specific leaf weight (SLW, 197 1, Gentry 1990). In Puerto Rico, glcm2), fruit length, number of developed vegetative variations of T. heterophy//a seem seeds per fruit, and seed and seed plus wing to be correlated with the amount of width were measured. Data on fruit length, precipitation, which defines a strong north- number of seed per fruit, and seed weight south gradient across the island We studied (calculated from the weight of 25 seeds) were the seed germination of two T. heterophy/la taken from collected fruits of at least 18 trees populations when subjected to a gradient of during late May 1992. water potentials. We also present some morphological traits related to the two Germination experiment: A pooled sample contrasting subtropical forests where this tree of seeds was obtained from available dry fruits species occurs. T. heterophy/la is a shrub or arbitrarily collected in June 1991 from 16 secondary tree up to 20 m in height. Itflowers trees at each forest. Small and damaged seeds mainly in early spring and sporadically were eliminated from the samples. Seeds were through the year (Little and Wadsworth 1989). treated with solutions of polyethylene-glycol The natural range of T. heterophylla includes (pEG, molar mass 6000, Sigma Co.), with Hispaniola, Puerto Rico, Virgin Islands osmotic potentialof -0.1, -0.3, -0.5, -0.7, -0.9, through the Lesser Antilles to Grenada and -1.2 and -1.5 MPa, at 23 C, following the Barbados. In Puerto Rico, it grows from sea procedures and conditions previously level to approximately 1000 m (Longwood considered (Cordero and Di Stéfano 1991, and 1971). references there in). Twenty-five seeds were put on Petri dishes with 10-15 mI of PEG MATERIALAND METHODS solutions or distilled water (e.g. control), and maintained on atable under laboratory Seeds were collected from El Verde and conditions (25 C ambient temperature and a Guánica. El Verde in the Luquillo 50% relative humidity). Treatments were Experimental Forest (northeastem Puerto replicated four times. Seed germination was Rico, 65°49'W, 18°20'N) has a mean annual recorded every day during 28 days. PEG precipitation of 3920 mm. The forest is in the solutions were renewed every eight days, as subtropical wet forest life zone (Ewel and well as the control. Number of days for initial Whitmore 1973), and the soil water content is germination was recorded, and the final maintained during most of the year. Guánica germination percentages were calculated. Commonwealth Forest and Biosphere Reserve Viability of seeds (tetrazolium test) collected in the southwestem part of Puerto Rico during May 1995 was not different to the seed (65°52'W, 17°52'N) presents a subtropical dry germination in destilled water under the same forest life zone (Ewel and Whitmore 1973). conditions as aboye, but a little higher than Soil moisture deficits can occur 10 months a the 1991 seeds. year, and a strong dry season commonly exists from January to May (Lugo et al. 1978). Statistical analyses: Morphological traits Rainfall varies among 600 to 1100 mm were compared using the non parametric Rank annually (Ewel and Whitmore 1973), but there Sum Test of the Statistix Program, due to non are two well-defined rainfall peaks in April homogeneity of variances in some of the and September (Lugo, el al. 1978). Long compared characters. The regression term precipitation patterns show strong coefficients between fruit length and seed unpredictability in annual precipitation and its number were compared for the two distribution in Guánica (Medina and Cuevas popuiations using a modified Student "t" test. 1990). Final germination percentages were Morphological measurements: The plant normalized with the arcsine transformation for material was collected from ten trees at each the statistical analysis. Two ways ANOVA R.A. Cordero & B. Molano-Flores: Gerrnination of Tabebuia heterophylla seeds (Bignoniaceae) 81
with seed souree (populations) and treatments trees. Strong differences in fruit and seed (osmotie solutions) as factors was performed variables were obtained among populations. El as suggested by Scott � al. (1984). Verde fruits were larger than Guánica fruits. Baektransformed values are reported. The Seed size at El Verde was twice larger than same analysis was performed for the days to Guánica. initial germination, eliminating those Also, significant differenees were found on treatments where no germination occurred. seed width and seed plus wing's width (Table Mean differences between treatments were 1). Guánica seeds were four times lighter than eompared by using the Tukey test. Statistical El Verde seeds (Rank Sum test, P TABLE 1 Mean and standar error (SE) o[ some morphologicaltraits measured on adult T. hc:terophylla treesfrom a wet[orest (El Verde) and a dry[orest (Guánica) in PuertoRico. Wetforest Dryforest ,r Mean SE ,r Mean SE pi Leafletlenglh (an) 14.14 0.58 30 3.45 0.15 38 ** Leafletwidth (an) 5.69 0.192 30 1.51 0.075 38 ** 2 ** Leafletarea (an ) 55.52 4.56 30 5.51 0.45 26 2 * Specificleafweigbt Froit lenglh1991 (an) 19.17 0.79 27 9.26 0.38 30 .** Froit lenglh1992 (an) 15.35 0.67 32 9.33 0.3 i 30 ** Seedsper hit 1991 68.2 9.29 10 74.53 3.56 30 NS2 2 Seedspec ftuit1992 79.4 5.9 32 71.6 2.63 30 NS Seed+ wingl(an) 2.73 0.05 40 1.85 0.04 40 ** Seedwidth (an) 1.21 0.02 40 0.633 0.07 40 ** Seedweigbt 1992 (g) 0.0174 3.7E-4 36 4.0E-3 4.3E-4 30 *- pl: indicates significantdiffe rence between means using SumRank Test �: no significant. P > 0.05. (*) P < 0.01. and (**) P< 0.001. nJ: samplesize. 82 REVISTA DE BIOLOGIA TROPICAL TABLE 2 Number of days for initiation of germination of T. heterophylla seeds under several osmoticwater potentials treatments. Seeds came from a wet forest (El Verde) and a the dryforest (Guánica) ofPuertoRico. Four replications per treatment. Statisticalanalysis explained in the texto No germinationoccurs in sornetreatments. Wetforest Dryforest . . Treatment Mean s.E. Mean s.E. Control 5.50a 0.86 2.25a 0.25 -0.1 MPa 5.50ab 0.96 3.75ab 0.63 -0.3 4.50ab 0.75 4.75bc 0.48 -0.5 7.00ab 0.91 5.00bc 0.41 -0.7 11.0b 1.22 7.00c 0.41 -0.9 15 6.00bc 0.71 -1.2 15 -1.5 *SE, mean standard error. Differentletter indicates significant differences between rneans of the Tukeytest at 0.05 leve/.. Germination experiment: A two-way percentageas osmotic potentialdecreased (two ANOVA for the days to germination showed way ANOVA, treatment effect, p "'"' WETFOREST DRYFOREST '#.'-' 75 MPa 75 �.... 0.0 MPa 60 60 -0.1 -0.1 0.0 I -0.3 45 45 -0.5 � -0.7 -0.3 -0.9 � -0.5 � 30 JO fi-; j � 15 -0.7 15 t;;¡ O 4) O U O 10 20 JO O 10 20 30 DAYS AFTER SOWING Flg. 2. Cumulative gennination curves fo r two populations of Tabebuia heterophylla seeds under a gradient of osmotic solutions ofPEG. Error bars indicatestIte greater error standard offour replications. Treatmentsat -1.2 and-1.5 MPa completely inhibitedgennin ation. An interaction effecf (P Several factors probably reduce soil water ,------, 80 O WETFOREST potentials at Guánica. This dry forest presents shallow soils with (requent limestone • DRYFOREST outcropping, where rain water rapidly percolates, runs off. or evaporates (Lugo et al. 60 1978). Then, biophysical properties of soil, high temperature and rapid use of available soil water by plants could favor the rnaintenance of reduced low soil water potentials, especially during dry to wet season transition, when deciduous species are resprouting and increasing the water absorption again. We consider that the higher germination rate of srnall seeds may confer advantages in both greater imbibition and seedling survival under low water potentials, o helping to maintain the population under reduced moisture -1.6 -1.2 -0.8 -0.4 0.0 availability. In sorne species, seedling survival is also greater for plants from small OSMOTIC POTENTIAL (MPa.) seeds under short-term drought (Wulff 1986, Hendrix and Trapp 1992). Under that Fig. 3. Relationship bttwem osmotic potential of situation, we hypothesize that the irregular germination treatment andtbe final germination p=tage and unpredictable rainfall pattern during the of T. heterophylla seros. Two populations were studied: Guánica dry forest (filled circ\es) and El Verde wet forest dry to wet season transition in the dry forest of (opencirc\es). Guánica (Medina and Cuevas 1990), produce a preconditioning effect of the germination responseof seeds. Seed Germination: The arrested germination Enhanced and faster germination after produced as the osmotic potential decreases is preconditioning treatrnents has been obtained well k:nownfrom cultivatedplants (Bodsworth for a tropical tree (Cordero and Di Stéfano and Bewley 1981), and in tropical tree species 1991) and others cultivated plants (Bodsworth (Cordero and Di Stéfano 1991). This kind of and Bewley 1981). Also, this preconditioning response might be selected in areas with effect could explain the absence of differences sporadic rains and great variation of quantity in germination and seedling survival to of rain (Gutterman 1980/81). Our results watering regimes found in other studies could also be due to the physics of water (Blain and Kellman 1991). This rnechanism imbibition as affected by seed size, but the would be irrelevant in wet forests where less if general appearance of seeds indicated that any soil rnoisture deficít occurs. In ecological they were well imbibed, which suggested that terms, it is predicted that preconditioned seeds radicle growth was the arrested process, and will have a more homogeneous germination not the seed imbibition. The most relevant and emergence through the dry to wet season result from this study líes on the differences in transltlon. Thus, the differences in response among dry and wet forest Tabebuia physiological phenotypic plasticity of both tree seeds. This confers an ecological seed populations and their environmental1y irnportance to the ability of seeds to germinate correlated morphological traits in the natural under moisture stresses, as the one adduced to populations are in concordance with the the habitat distribution oftwo Quercus species ability of T. heterophylla to survive and (Vuillernin1982), and to sorne populations of rnaintain viable populations in these two Pinus (Falusi et al. 1983). contrasting subtropical forests. Plant R.A. Cordero & B. Molano-Flores: Gennination of Tabebuia heterophylla seeds (Bignoniaceae) 8S demography, seed size variation among Ecology Division, UPR and tbe facilities of populations and its maternal effects and the tbe Dept. ofBiology. UPR. extent to which these variations will affect seedling traits in longer periods would REFERENCES produce different results (Hendrix and Trapp 1992). Abrams, M.O., B.O. KIoeppel & M.E. Kubiske. 1992. Ecopbysiological and morphological responses to shade and drought in two contrasting ecotypes of Prunus RESUMEN serotina. TreéPhysio!. 10:343-355. Ang:¡purger, C. K 1979. Irregular rain cues and the El efecto de la tensión hídrica sobre la germination and survival of a Panamanian shrub germinación de las semillas de Tabebuia (Hybanthus pruniifolius). Oecologiá44:53-59. heterophy/la (Bignoniaceae), procedentes de Blain, D. & M Kellman. 1991. Tbe effelt of water supply un bosque seco y otro muy húmedo en Puerto on tree seed germination and seedling survival in a Rico, se estudió utilizando un gradiente de tropical seasonal forest.in Veraauz, Méxioo. J. Trql. potenciales osmóticos (O a -1.5 Mpa). Eoo!. 7:69-83 Algunas comparaciones moñológicas fueron Bodsworth, S. & J.D. Bewley. 1981. Osmotic priming of hechas de individuos adultos de ambas seedsof aopspecies with polydhylcne glyool as a mesn poblaciones. Los árboles del bosque seco of enhancingearly and syndJronous germination at0001 temperatures.Can. J. Bot. 59:672-676. presentaron hojas, frutos. y semillas más pequeñas y el peso foliar específico fue mayor. Bradford, KJ. & T. C. Hsiao. 1982. Physiological responses Los frutosdel bosque seco contenían mayor to mocIa-atewater stress . p:227-265. In: Lange, 0.1... � cantidad de semillas más pequeñas los !l. (eds.)Physiologicsl Plant Eoology n Waterrelations que and carbon assimilation. Encyclopedia of Plant frutos de tamaño similar de árboles del Physiology, New Series, Vol 12B. Springer-Verlag. bosque muy húmedo. El porcentaje de Berlin. germinación se redujo drásticamente conforme Cordero;R.A & J.F. Di Stéfano. 1991. Efec:to del estrés el . potencial osmótico fue más negativo en osmótioo sobre la germinaciÓll de las semillas de ambas poblaciones, aunque fue más acentuado Tecoma stans (1...) H.B. K (Bignoniacese). Rev. Biol. en las semillas del bosque húmedo. El Trql. 39:107-110. número de días para el inicio de la Ewel, J. J. & J.1... Whitmore. 1973. Tbe eoologicsl life zones germinación. presentó la misma respuesta, ofPuerto Rioo andthe Virgin Islands. U.S. For. Secv. pero ésta fue menos marcada en las semillas Res. Papo ITF-18. Institute of Trqlical Forestry, Rio Piedras,Puerto Rioo. 72 p. de árbolesdel bosque seco . . La germinación fue inhIbida completamente a potenciales Falusi, M, R. Calamasi & A Tocci. 1983. Sensitivity of osmóticos menores que -0.9 MPa, Y no fue seed germination onseedlingroot_ growth to moisture diferente entre los potenciales osmóticos. más stress in four provcnances of Pinus halapensis Mill. Silvae Genetica32:4-9. altos, en ambas poblacionesrespetivamente. Estos efectos sobre la germinación junto con GarwOOd, N.C. 1986. Cónstraints on the timing of seed las modificaciones moñológico-ambientales germination ina tropical forest. pp:347-355. IN: Estrada, A & T. H. Fleming. (eds.) Frugivores andseed parecen capacitar a esta especie a mantener dispersa!. Dr. W. Junk Publisbers.Dordrecht. poblaciones viables en estos ambientes tan contrastantes. Gartty, AH. 1990. Evolutionary pattems in Neotropical Bignoniacese. Me.m. New York Bot. Garden 55:118- 129. ACKNOWLEDGMENTS Guttennan, Y. 1980\81. Intlucnces on seed germinability: phcnotypic maternal effeás during seed maturation. Criticisms and English corrections on :final Israel J. Bot. 29:105-117 manuscripts bythefaculty and friends at UPR are great1y appreciated, especiallyA Montana Harper, J. L. 1977. Population biology of plants. Academic, and S. Ward This study was partially NewYork. 892 p. supported by RIMI NSF grant RII-8903827, a Hcndrix, S.D. & E. J. Trapp. 1992. Population demography summer stipent to RAe from the Terrestrial of Pastinaca sativa (Apiacese): Effeás of seedsize QP 86 REVISTA DE BIOLOGIA TROPICAL emergence. survival. and recruitment. Amer. J. Bot. structure, nineral nutrient content, and drought 79:365-375. resistancein tropical rain fo rests ofthe Upper Río Negro Region. Biotropica22:51-64 Little. E.L. lr. & F.H. Wadsworth. 1989. Common trees of Puerto Rico andthe Virgin Islands. 2nd. ed. Agric. Melzack, R.N. & D. Watts. 1982. Variation in seed weight, Handbook No. 249. U.S. Dept. Agric., Forest Service. gerrnination and seedling vigor in the yew (Taxus WashingtonD. C. 556p. baccata L) in England. l. ofBiogeogr.9:55 Longwood, F.R. 1971. Present and potwtial commercial Schlichting, C .D. 1986. The evolution of phenotypic timbersofthe Caribbean.Agric. 2nd ed. Handbook No. plasticity in plants. Ann. Rev. Eco\. & Syst. 17:667- 207. US Dept. Agric., For. Servo Washington D.C. 693. 167p. Lugo, AE., lA González, B. CintrÓll & K Dugger.1978. Scott, S. J., RA Jones & WA WiIliams. 1984. Review of Structure, productivity, and transpiration of a data analysis mt'thods fo r seed gerrnination. Crop Sci. subtropical dry forest in Puerto Rico. Biotropica 24:1192- 1199. 10:278-291. Sobrado, MA & O. Cuenca. 1979. Aspectos del uso del Lugo, AE. & P.O. Murphy. 1986. Nutrient dynamics of a aguade especiesdeciduas y siempreverdes en1m bosque PuertoRican subtropicaldry fo resto J. Trop. Eco\.2:55- seco tropical de Venezuela. Acta Científica. Ven. 72. 30:302-308. Medina,E.V. 1983. Adaptations oftropicaltrees to moisture Stemmermann, L. 1983. EcologicaI studies of Hawaiian stress. p. X-X IN: F.B. Golley(ed.) Ecosystems ofthe Metrosideros in a successional context. Pacific Science world: tropical rain for es!., struc.ture, and fimction. 37:361-373. Elsevier,Amsterdam, Holland .. Vuillemin, J. 1982. Ecophysiologie compareé du Medina,E.V. & E. Cuevas.1990. Propiedadesfot osintéticas développment initial de Quercuspubescens WilI. á º' y eficiencia de uso de agua de plantas leñosas del ilex L. JI. Oerrnination á croissance radiculaire en bosque deciduo de Guánica: Consideraciones generales condition de stress hidrique. Ecología Mediterranea 8: y resuhados preliminares. Acta Científica 4(1-3):25- 147-151. 36. Wulff, R.D. 1986. Seed size variation in Desmodium Medina, E., V. Garcia & E. Cuevas.1990. Sclerophylly and paniculatum, l. Factor affecting seed sire. l. Eco!. oligotrophyc environments: relationship báween leaf 74:87-97. PAPER REPRODUCED FROMA CAMERA-READY ORIGINAL PREPARED BY AUTHORS.