Interpopulation variation in germination responses of Scotch thistle, Onopordum acanthium L., to various concentrations of GA 3, KNO 3, and NaHCO 3

Transcription

1 1156 Interpopulation variation in germination responses of Scotch thistle, Onopordum acanthium L., to various concentrations of GA 3, KNO 3, and NaHCO 3 Mirwais M. Qaderi and Paul B. Cavers Abstract: The effects of GA 3, KNO 3, and NaHCO 3 on germination percentages and rates were determined for cypselas (seeds) of four local populations of Scotch thistle, Onopordum acanthium L., collected during a 2-year period. Three (1996, stored cypselas) and four (1997, fresh cypselas) concentrations (0.025, 0.05, and 0.1% for both years, plus % for 1997) of the three compounds were used. Both freshly collected cypselas and those stored for 67 days at room temperature exhibited a range of germination responses from readily germinable to strongly dormant. GA 3,even at a very low concentration, caused most viable cypselas to germinate, but 5 6% of viable cypselas in two populations did not respond to any GA 3 treatment. In general, dry-stored cypselas germinated faster than fresh ones. Populations with more germinability (= proportion of cypselas capable of germinating) responded faster to GA 3 than those with less germinability. KNO 3 usually increased germination percentages but reduced the rate of germination. When treated with KNO 3, stored cypselas from the two populations with less germinability germinated more slowly than those of the other two. In general, germination percentage decreased with an increase in NaHCO 3 concentration. Overall, stored cypselas that were treated with NaHCO 3 germinated more slowly than those in the control. The large differences among four local populations in response to these chemicals suggest that the effects of germination stimulators and other chemicals should be observed on several populations per species. Key words: cypsela, germination promoters, germination inhibitors, local populations, Onopordum acanthium, Scotch thistle. Résumé : Les auteurs ont déterminé les effets du GA 3,duKNO 3 et du NaHCO 3 sur les taux et les pourcentages de germination des graines (cypsèles) de quatre populations locales de l Onopordum acanthium L., le chardon d Écosse, récoltées au cours de deux années. Trois (1996, cypsèles conservées) et quatre (1997, cypsèles fraîches) concentrations (0,025, 0,05 et 0,1 % pour les deux années, plus 0,0125 % pour 1997) des trois composés ont été utilisées. Les cypsèles fraîchement récoltées, aussi bien que celles conservées pendant 67 jours à la température de la pièce, ont montré une gamme de réactions de germination allant de la germination immédiate à la dormance profonde. La GA 3, même à de très faibles concentrations, provoque la germination de la plupart des cypsèles viable, mais 5 6 % des cypsèles viables des deux populations ne réagissent à aucun traitement avec la GA 3. En général, les cypsèles conservées à l état sec germent plus rapidement que les fraîches. Les populations comportant une germinabilité plus élevée (proportion des cypsèles capable de germer) réagissent plus rapidement à la GA 3 que celles ayant des germinabilités plus faibles. Le KNO 3 augmente généralement les pourcentages de germination, mais réduit le taux de germination. Lorsqu on traite des cypsèles conservées provenant des deux populations ayant des germinabilités moindres, elles germent plus lentement que celles des deux autres populations. En général, le pourcentage de germination diminue avec une augmentation de la concentration en NaHCO 3. Globalement, les cypsèles conservées qui ont été traitées avec le NAHCO 3 germent plus lentement que les cypsèles témoin. Les fortes différences observées entre les quatre populations locales, aux substances chimiques, suggèrent que les effets des stimulateurs de germination et autres substances chimiques devraient être observés sur plusieurs populations pour une espèce donnée. Mot clés : cypsèles, stimulateurs de germination, inhibiteurs de germination, populations locales, Onopordum acanthium, chardon d Écosse. [Traduit par la Rédaction] Qaderi and Cavers 1163 Introduction In many species, seed dormancy can be broken and germination enhanced by one or more chemicals, such as growth regulators, nitrogenous compounds, and oxidants (Bewley and Black 1994; Baskin and Baskin 1998). Gibberellic acid is particularly effective in breaking dormancy and promoting germination in seeds (Bewley and Black 1982), and it counteracts the action of abscisic acid (ABA) (Hilhorst and Karssen 1992). However, dormancy can be imposed by other Received February 23, M.M. Qaderi 1 and P.B. Cavers. Department of Plant Sciences, University of Western Ontario, London, ON N6A 5B7, Canada. 1 Author to whom all correspondence should be addressed. ( mqaderi@julian.uwo.ca). Can. J. Bot. 78: (2000)

2 Qaderi and Cavers 1157 chemicals. One of these is sodium bicarbonate, which is present in soils containing sodium and dissolved carbon dioxide (Kidd 1914; White 1997). Even though the effects of dormancy-breaking chemicals have been studied extensively in many species (Baskin and Baskin 1998), little is known about interpopulation variation in response to these chemicals. Pérez-García et al. (1995) studied seed germination patterns of four populations of the weedy mustard, Diplotaxis virgata (Cav.) DC. The application of gibberellic acid to seeds of D. virgata stimulated germination, especially in populations with stronger dormancy. In Onopordum nervosum Boiss., Pérez-García and Duran (1990) showed that gibberellic acid (GA 3 ) clearly stimulated cypsela (seed) germination in two populations, and the response increased with the concentration used. This increase was slightly greater in one population than in the other. Meier (1995) found significant differences in germination between two populations of Onopordum acanthium L. in response to 0.1%, but not 0.01%, GA 3. Nitrates, especially potassium nitrate (KNO 3 ), can replace the light requirement for germination in many positively photoblastic seeds (Evenari 1956; Toole et al. 1956; Roberts 1973). Potassium nitrate often has been used to stimulate germination of dormant seeds (Mayer and Poljakoff-Mayber 1989). Carbon dioxide occurs in the gaseous environment of soils and may affect seed germination. Carbon dioxide can enhance seed germination at low levels (2 5%) (Schönbeck and Egley 1981) or inhibit it at high levels (above 10%) (Corbineau and Côme 1995). The level of CO 2 in soils depends on biotic and abiotic factors (Egley 1986). One source of carbon dioxide in soils is sodium bicarbonate (NaHCO 3 ). Although NaHCO 3 is a common constituent of many soils, especially alkaline soils (Thompson 1957), its effect on seed germination has not been tested. We chose to work with Scotch thistle, O. acanthium, a species with a broad spectrum of dormancy. Onopordum acanthium is monocarpic and reproduces by cypselas (Alex and Switzer 1985). Some cypselas will germinate in the autumn shortly after dispersal from the mother plant, others in the following spring, but some remain dormant in the seed bank over several years (Qaderi 1998). Several authors have reported that dormancy can be broken in cypselas of O. acanthium by application of GA 3 (Scifres and McCarty 1969; Young and Evans 1972; Groves and Kaye 1989; Pérez-García 1993) or KNO 3 (Young and Evans 1972; Groves and Kaye 1989), but no study is known as to the effects of NaHCO 3 on this or any other species. The purpose of this study was to determine the effects of GA 3, KNO 3, and NaHCO 3 on the germination patterns of cypselas of four local populations of O. acanthium collected from London, Ontario. The following hypotheses were tested: (i) germination is promoted by GA 3 and KNO 3 and inhibited by NaHCO 3,(ii) rate of germination will parallel percentage germination, and (iii) responses of local cypsela populations will vary with concentration. Materials and methods Cypsela collection Bulk collections of cypselas of O. acanthium were taken from 100 randomly selected plants on September 4, 1996, and from 50 randomly selected plants on September 6, 1997, in each of four populations in or near London, Ontario (Table 1). The parent plants from the Q and Ola populations had germinated in situ while those from the ESW and GP populations had been transplanted from the original sites (at the rosette stage in the spring of 1996) to ESW, because they would have been destroyed in their original habitats. The cypselas were cleaned manually and stored at room temperature (20 25 C) until tested. Cypselas from the 1996 collection were stored for 67 days before testing while those from the 1997 collection were used fresh, after 5 days of storage. Aborted cypselas were discarded. Germination tests Three (0.025, 0.05, and 0.1% in 1996) and four (0.0125, 0.025, 0.05, and 0.1% in 1997) concentrations of GA 3, KNO 3, and NaHCO 3 were used. The choice of these concentrations was based on our preliminary experiments and on published results with these chemicals. For each chemical, 20 lots of 100 cypselas from each population in 1996 and 25 lots in 1997 (i.e., five replicates of each concentration) were each placed in a 9-cm-diameter glass Petri dish on one layer of blue germination filter paper (Anchor Paper Co., St. Paul, Minn.), initially moistened with 10 ml of one of the prepared solutions or distilled water (control). The cypselas were set to germinate in incubators at 25 C for 14 h of light and 10 C for 10 h of darkness. The temperature and light regimes were chosen to mimic soil surface conditions in late spring and early autumn in London, Ontario. The dishes containing GA 3 (Eastman Kodak Co., Rochester, N.Y.) were remoistened with the same concentrations of GA 3 (advice of Dr. R.I. Greyson since GA 3 is deactivated quickly under these conditions) and for NaHCO 3 with the same concentrations of NaHCO 3 as needed to maintain levels in the dishes. For KNO 3, distilled water was added as needed (about once per week) since the K + and NO 3 ions remained in solution and relatively little was taken up by the cypselas. Light was provided by cool white fluorescent tubes (PAR of µmol photons m 2 s 1 at the level of the Petri dishes, n = 30). Germinated cypselas (radicle of 2 mm or longer) were counted and removed daily. The dishes were arranged randomly in the incubator at the beginning of the experiment and replaced in different random patterns after each daily germination count. The experiments were terminated after a 5-day period with no germination (after 45 and 65 days in 1996 and 1997, respectively). Previous experiments have shown that a 5-day period with no germination is usually sufficient to ensure that little germination will occur thereafter. Firm, nongerminated cypselas were subjected to a viability test at the end of the experiments in 1996 by using a 1% (w/v) solution of tetrazolium chloride (2,3,5-triphenyl tetrazolium chloride, C 19 H 15 N 4 Cl, Sigma) (method of Delouche et al. 1962). In 1997, viability of nongerminated cypselas was tested by cutting them 1 mm from the cotyledonary end and setting them to germinate at 25:10 C. Statistical analyses Final germination percentages for each of the GA 3, KNO 3, and NaHCO 3 experiments were transformed to the arcsine square root of the percentage (Zar 1999) and analyzed by means of a balanced ANOVA for each year s data (McKenzie et al. 1995). Tukey s test was used to determine differences between treatments. Linear regression analyses were performed to determine the relationship between chemical concentration and germination percentage within each year (SAS Institute Inc. 1982; Sokal and Rohlf 1995). The coefficient of germination rate (CGR) was calculated for each replicate by dividing the final germination percentage (N) by the number of germinated seeds on the particular day on which a count was made (n i ) multiplied by the number of days from the start of the experiment (d i ) and summing for all days on which germination occurred (Alm et al. 1993):

3 1158 Can. J. Bot. Vol. 78, 2000 Table 1. Description of collection sites of Scotch thistle (Onopordum acanthium) populations in London, Canada. Population Latitude and longitude Soil texture Soil ph Habitat ESW (Environmental Sciences Western research station) N, W Silt loam 7.5 Arable field, noncultivated for 1 year a Q (Quarry) N, W Light gravelly loam m from edge of gravel pit, open area amongst deciduous trees, m from Thames River, never flooded Ola (Olalondo Road) N, W Gravelly loam m from edge of gravel pit, open area, m from Thames River, never flooded GP (Gibbons Park) N, W Sandy loam over gravel 7.6 On reconstructed area (1995) after Thames River erosion, 2 10 m from Thames River, frequently flooded and partially shaded b a Population originated in 1992 from gravel soil (ph 7.9) at woodland edge 20 m east of Thames River, in a frequently flooded habitat ( N, W). b Soil and gravel used for reconstruction came from Canada Building Materials quarries in north London ( N, W), 500 m from the Thames River. The GP population emerged from this substrate. [1] CGR = N nd i i All values of CGR are between 0 (no germination) and 1 (fastest germination rate). Then, a balanced ANOVA and a Tukey s test were applied (SAS Institute Inc. 1982; Sokal and Rohlf 1995). Results The effects of GA 3 Most fresh (1997) and stored (1996) cypselas germinated after GA 3 application, regardless of the concentration used (Table 2). From the ANOVA, populations (P < 0.001), concentrations (P < 0.001), and their interaction (P < 0.001) were significant. A significant relationship was found between GA 3 concentration and germination percentage in both 1996 (r 2 = 0.35, P < 0.001) and 1997 (r 2 = 0.51, P < 0.001). For all but one population year combination (ESW 1997), final germination percentages did not differ among the concentrations of GA 3, and in all GA 3 treatments % of the cypselas germinated. In contrast, germination of cypselas in control treatments was between 4 and 27% (1997) or between 42 and 89% (1996). Despite the similar response to all concentrations of GA 3 by all four populations, populations differed in final germination percentage. For example, germination in all treatments in the GP population was % versus only 89 93% in the Q population. Viability tests at the end of each experiment revealed that more than 97% of nongerminated cypselas were viable. The population with the least germination in the control treatment, ESW 1997, was the only population with significant differences in final germination percentages among the GA 3 treatments (Table 2). The % GA 3 treatment increased germination percentage from 4.4 to 85%, the 0.025% treatment increased it again to 95.2%, and the next highest concentration of GA 3 significantly increased it further to 99.6%. With regard to germination rates, the effects of population (P < 0.001), concentration (P < 0.001) and their interaction (P = 0.004, 1996; P = 0.002, 1997) were significant. A significant relationship was found between GA 3 concentration and germination rate in 1996 (r 2 = 0.32; P < 0.001), but not in 1997 (r 2 = 0.006; P = 0.215). For the stored cypselas (1996), germination was significantly faster in all but one GA 3 treatment (0.025% GA 3 GP) than in the control for the same population (Table 2). For the fresh cypselas (1997), germination rates in all GA 3 treatments (ESW population) or the 0.025% GA 3 treatment (Ola population) were slower than in the control treatment. In all other cases, there was no difference between the control and any GA 3 treatment nor between any GA 3 treatments (concentration) for the same population (Table 2). In general, Q cypselas germinated more rapidly (24.9 ± 0.7) and ESW cypselas more slowly (9.6 ± 0.8) than those from other populations (Table 2). Only for fresh control cypselas (1997) was this difference not significant. The effects of KNO 3 In general, application of KNO 3 increased germination percentages (Table 3). The effects of population (P < 0.001), concentration (P < 0.001) and their interaction (P < 0.001) were significant. A significant relationship was found between KNO 3 concentration and germination percentage in both 1996 (r 2 = 0.28; P < 0.001) and 1997 (r 2 = 0.68; P < 0.001). Unlike the pattern of response to GA 3, there was an increase in germination percentage in all populations (except GP in 1996) with increasing concentrations of KNO 3. The lowest concentration, %, affected neither percentage nor rate of germination (Table 3). The other concentrations of KNO 3 had a variety of effects on the different populations in the 2 years of the experiment. In the ESW population, with the lowest control germination percentage, the final germination percentage was higher in the 0.1% KNO 3 treatment than in the 0.05% KNO 3 treatment in both years and was higher in the 0.05% KNO 3 treatment than in the 0.025% treatment (1997) (Table 3). However, in the other three populations (1997), the germination percentage was higher in the 0.05% treatment than in the 0.025% treatment, but was not higher in the 0.1% treat-

4 Qaderi and Cavers 1159 Table 2. Effects of various concentrations of GA 3 on germination percentages and germination rates (means ± SE) of cypselas of Onopordum acanthium, collected from four populations (ESW, Q, Ola, and GP) in or near London, Ont. in Sept and Sept and incubated at 25 C for 14 h light and 10 C for 10 h dark. The 1996 cypselas had been stored dry for 67 days at room temperature when tests began, while the 1997 cypselas were freshly harvested. Final germination percentage Coefficients of germination rates b Year Concentration of GA 3 (%) ESW a Q Ola GP ESW a Q Ola GP (control) 41.8±4.3Cb 64.2±2.0Bb 53.6±7.6BCb 89.0±2.3Ab 10.6±1.0Bb 17.1±1.8Ab 10.4±1.3Bb 17.4±0.5Ab ±1.2Aa 91.2±2.0Aa 97.6±1.0Aa 98.0±0.6Aab 17.2±0.6Ba 23.1±1.1Aa 21.3±0.7ABa 21.6±0.7Aab ±2.1ABa 90.4±1.9Ba 96.2±1.6ABa 99.0±0.5Aa 18.2±0.5Ca 24.9±0.7Aa 20.5±1.1BCa 22.7±0.3ABa ±1.9Aa 90.2±1.7Aa 97.4±1.2Aa 97.4±0.8Aab 17.9±0.5Ba 22.7±0.4Aa 22.5±0.8Aa 21.9±0.3ABa (control) 4.4±0.8Bd 26.6±0.7Ab 17.6±1.8Ab 21.2±3.2Ab 16.9±1.5ABa 19.7±0.8Aa 16.5±0.9ABa 13.2±1.9Ba ±2.2Cc 88.8±2.0BCa 93.6±1.0Ba 100.0±0.0Aa 9.6±0.8Bb 15.5±0.5Aa 13.4±1.0ABab 12.9±0.8ABa ±1.0Bb 91.2±2.2Ba 93.8±1.3Ba 99.4±0.3Aa 11.1±0.3Bb 17.4±0.8Aa 11.6±0.4Bb 12.5±0.8Ba ±0.3Aa 93.4±0.5Ba 94.4±0.7Ba 99.2±0.6Aa 12.1±0.8Bb 16.5±1.7Aa 13.3±0.8ABab 14.1±0.4ABa ±0.6ABab 91.8±1.5Ca 93.8±1.4BCa 99.6±0.4Aa 10.3±0.6Cb 19.7±0.3Aa 13.1±0.2BCab 15.1±0.7Ba Note: Means followed by different upper-case letters within rows or by different lower-case letters within columns for each year are significantly different (P < 0.05) according to Tukey s honestly significant difference multiple range test (Tukey s HSD). a For details of the populations see Table 1. b For germination rates, all means have been multiplied by 100 to facilitate data interpretation. Table 3. Effects of various concentrations of KNO 3 on germination percentages and germination rates (means ± SE) of cypselas of Onopordum acanthium, collected from four populations (ESW, Q, Ola, GP). Final germination percentage Coefficients of germination rates Year Concentration of KNO 3 (%) ESW Q Ola GP ESW Q Ola GP (control) 41.8±4.3Bc 64.2±2.0Bb 53.6±7.6Bb 89.0±2.3Aa 10.6±1.0Ba 17.1±1.8Aa 10.4±1.3Ba 17.4±0.5Aa ±2.6Bbc 81.0±3.5Aab 57.4±7.7Bb 92.2±1.2Aa 7.5±0.4Ba 15.8±1.0Aab 9.1±1.3Ba 14.3±0.8Aab ±6.0Bb 82.6±2.2ABab 91.4±2.1Aa 93.0±1.6Aa 10.9±1.1Ba 14.2±1.1ABab 10.8±0.9Ba 16.0±0.7Aab ±2.2Aa 86.0±2.4Aa 86.0±3.2Aa 90.2±2.2Aa 9.7±0.3Aa 11.7±0.7Ab 10.1±0.5Aa 11.6±0.6Ab (control) 4.4±0.8Bc 26.6±0.7Ac 17.6±1.8Ac 21.2±3.2Ab 16.9±1.5ABa 19.7±0.8Aa 16.5±0.9ABa 13.2±1.9Ba ±1.7Cc 34.6±4.0ABbc 18.4±0.8Bc 36.4±4.7Ab 16.2±1.6Aa 13.9±1.3ABab 11.7±2.4ABab 8.7±1.1Bab ±1.1Bc 48.4±2.8Ab 39.6±3.6Ab 37.6±7.9Ab 9.9±1.5Ab 14.7±0.5Aab 10.4±1.5Ab 9.7±1.6Aab ±4.5Bb 75.2±3.9ABa 82.0±3.3ABa 89.6±2.5Aa 9.0±0.3ABb 13.4±0.4Ab 7.3±0.6Bb 6.6±0.3Bb ±2.4Aa 75.2±1.6Ba 68.8±3.4Ba 91.2±1.9Aa 7.4±0.5Ab 11.6±0.4Ab 7.2±0.3Ab 8.3±0.5Ab Note: Samples were collected in or near London, Ontario, in September 1996 and September 1997 and incubated at 25 C for 14 h light and 10 C for 10 h dark. The 1996 cypselas had been stored dry for 67 days at room temperature when tests began, while the 1997 cypselas were freshly harvested. For details of the populations see Table 1. Means followed by different upper-case letters within rows or by different lower-case letters within columns for each year are significantly different (P < 0.05) according to Tukey s honestly significant difference multiple range test (Tukey s HSD). For germination rates, all means have been multiplied by 100 to facilitate data interpretation. ment than in the 0.05% treatment (Table 3). Only the Ola population showed this pattern in The only population in which KNO 3 did not increase the final germination percentage in 1996 was GP, but germination of its control cypselas that year was 89% (Table 3). Viability of nongerminated cypselas was >98% in all treatments for all populations. With regard to germination rates, the effects of population (P < 0.001), concentration (P < 0.001), and their interaction (P = in 1996; P = in 1997) were significant. A significant relationship was found between KNO 3 concentration and germination rate in both 1996 (r 2 = 0.05, P = 0.024) and 1997 (r 2 = 0.39, P < 0.001). KNO 3 generally reduced germination rate, especially at the highest concentration. Only in the Ola and ESW populations in 1996 did the highest rate fail to slow germination (Table 3). The effects of NaHCO 3 In general, NaHCO 3 reduced final germination percentages in both the 1996 and 1997 tests (Table 4). The effects of population (P < 0.001), concentration (P < 0.001), and their interaction (P < in 1996; P = in 1997) were significant. A small but significant relationship was found between NaHCO 3 concentration and germination percentage in both 1996 (r 2 = 0.08, P = 0.007) and 1997 (r 2 = 0.09, P = 0.002). However, NaHCO 3 had no effect on the ESW population in either year nor on the Q population in Also, the % NaHCO 3 treatment (1997) did not affect the germination percentage or rate. Most other NaHCO 3 treatments reduced final germination percentages, but in all 1996 treatments at least one-third of the cypselas germinated. In all treatments, 99% of the nongerminated cypselas from all populations were viable.

5 1160 Can. J. Bot. Vol. 78, 2000 Table 4. Effects of various concentrations of NaHCO 3 on germination percentages and germination rates (means ± SE) of cypselas of Onopordum acanthium, collected from four populations (ESW, Q, Ola, and GP). Final germination percentage Coefficients of germination rates Year Concentration of NaHCO 3 (%) ESW Q Ola GP ESW Q Ola GP (control) 41.8±4.3Ca 64.2±2.0Ba 53.6±7.6BCa 89.0±2.3Aa 10.6±1.0Bb 17.1±1.8Ac 10.4±1.3Bb 17.4±0.5Aa ±0.9Ca 52.0±1.1Bab 25.6±0.7Cc 66.0±0.9Ab 17.5±0.5Ba 26.4±1.3Aa 15.8±1.0Ba 16.3±1.1Ba ±1.5Ba 51.0±1.3Bb 46.4±0.5Bab 71.4±0.5Ab 17.3±0.3ABa 21.8±0.3Ab 16.7±0.4Ba 19.0±0.3ABa ±1.1BCa 47.2±1.0Bb 33.2±1.2Cbc 62.2±0.9Ab 17.1±0.4Ba 23.2±0.6Aab 17.1±0.5Ba 18.6±1.2Ba (control) 4.4±0.8Ca 26.6±0.7Aab 17.6±1.8Ba 21.2±3.2ABa 16.9±1.5Aa 19.7±0.8Aa 16.5±0.9Aa 13.2±1.9Aa ±0.8Ca 32.2±1.0Aa 14.0±0.6Bab 17.2±1.7Bab 19.0±1.9Aa 18.0±1.0Aa 18.9±0.6Aa 13.5±1.3Aa ±0.7Ca 25.4±1.1Aab 11.2±1.2Bbc 11.6±0.8Bbc 16.4±1.6Aa 17.6±1.4Aa 16.0±1.4Aa 13.4±1.4Aa ±0.5Ca 24.4±1.3Aab 11.4±0.9Bbc 8.6±1.2Bc 20.0±0.5Aa 17.5±1.2Aa 18.4±1.1Aa 14.8±1.1Aa ±0.4Ca 20.2±0.7Ab 8.4±0.9Bc 9.0±1.6Bc 18.1±2.0ABa 19.6±0.7Aa 14.9±0.6ABa 12.7±0.6Ba Note: Samples were collected in or near London, Ontario, in September 1996 and September 1997 and incubated at 25 C for 14 h light and 10 C for 10 h dark. The 1996 cypselas had been stored dry for 67 days at room temperature when tests began, while the 1997 cypselas were freshly harvested. For details of the populations see Table 1. Means followed by different upper-case letters within rows or by different lower-case letters within columns for each year are significantly different (P < 0.05) according to Tukey s honestly significant difference multiple range test (Tukey s HSD). For germination rates, all means have been multiplied by 100 to facilitate data interpretation. In 1996, germination percentage was highest in the GP population and lowest in the ESW population. In 1997, it was highest in Q and lowest in ESW. With regard to germination rates, the effects of population (P < 0.001), concentration (P < 0.001), and their interaction (P < 0.001) were significant in However, in 1997, the effects of population (P < 0.001) but not concentration (P = 0.266) or the interaction between population and concentration (P = 0.518) were significant. A significant relationship was found between NaHCO 3 concentration and germination rate in 1996 (r 2 = 0.15, P < 0.001) but not in 1997 (r 2 =0,P = 0.933). There was no effect of NaHCO 3 on the rate of germination of fresh (1997) cypselas (Table 4). However, germination of stored (1996) cypselas of ESW, Q, and Ola was hastened by all three NaHCO 3 concentrations (Table 4). There was only one significant difference among populations in the rate of germination of fresh cypselas; under the 0.1% NaHCO 3 treatment, Q was faster than GP. In contrast, stored (1996) cypselas of the Q population germinated faster in all concentrations of NaHCO 3 and in the control than did cypselas of Ola; the stored cypselas of the Q population also germinated faster in all concentrations of NaHCO 3, except 0.05%, than did cypselas of ESW. Discussion Gibberellic acid and cypsela germination Several authors have reported that GA 3 can enhance the germination of cypselas of O. acanthium (Pérez-García 1993; Young and Evans 1972; Meier 1995). We found that GA 3 caused virtually all viable cypselas to germinate (Table 2), even at a much lower concentration (0.0125%) than the ones that were effective for other scientists (e.g., %, Pérez-García 1993; 0.1%, Meier 1995). In general, GA 3 increased the rate of germination of stored (1996) cypselas, but not of fresh (1997) ones (Table 2). Other experiments over several years have shown that freshly harvested cypselas have low germination percentages, which increase with dry storage (M.M. Qaderi and P.B. Cavers, unpublished data). Faster germination by stored cypselas was possibly due to rapid synthesis of GA 3 from precursors or its rapid transport from one part of a cypsela to another (Mayer and Poljakoff-Mayber 1989). Karssen and Laçka (1986) demonstrated that dry storage increased the sensitivity of Arabidopsis thaliana (L.) Heyn. seeds to 10 µm GA 4+7. Meier (1995) reported that 0.1%, but not 0.01%, GA 3 stimulated germination of O. acanthium cypselas, which does not agree with our results, even though the incubation temperature (25:10 C) was the same for both experiments. These discrepancies might be explained through differences in dates of collection. We used cypselas that were collected in early September while Meier collected her samples in early and late October. In another experiment, we found that cypselas collected later in the growing season had stronger dormancy (Qaderi and Cavers 2000). Meier s cypselas probably had stronger dormancy and hence did not respond as vigorously to GA 3 treatment. In her study, both populations (ESW and Q) had <65% germination following the application of 0.1% GA 3, while in our study, with the same concentration of GA 3, germination of all populations was >90%. Our findings contradict the results obtained by Scifres and McCarty (1969). None of the concentrations of GA 3 that they used (0, 8, 15, 31, 63, 125, or 250 mg/l) stimulated germination of nonstratified cypselas, but they were effective on stratified ones. In our study, GA 3 always had enhancing effects on nonstratified cypselas, whether they were stored or fresh. The differences may have resulted from the environmental conditions under which the cypselas matured (Nebraska, U.S.A., versus London, Canada), genetic differences, or differences in incubation temperature, as they used 30 C for 8 h light and 20 C for 16 h dark or constant temperatures in darkness. The comparatively small differences among populations in final germination percentage after GA 3 treatment (Table 2) could be of great importance in the long-term survival of populations. Populations such as Ola and Q with nearly 10% of their cypselas remaining dormant after exposure to high concentrations of GA 3 would probably have a longer-lived persistent seed bank than populations such as GP where al-

6 Qaderi and Cavers 1161 most all viable cypselas germinated in the higher GA 3 concentrations. Potassium nitrate and cypsela germination We obtained the same pattern of germination response to KNO 3 for O. acanthium as did Young and Evans (1972) for this species and Kolk (1962), Popay and Roberts (1970a), and Adkins and Adkins (1994) for other species. In general the higher the concentration of KNO 3, the higher was the germination percentage, but the rate of germination was slower for the KNO 3 -treated cypselas than for the control cypselas. Only one of our populations (ESW) had a higher germination percentage in 0.1% KNO 3 than in 0.05% (Table 3). Sodium bicarbonate and cypsela germination In general, the final germination percentage declined as the concentration of sodium bicarbonate increased (Table 4). The direct effect of NaHCO 3 in inhibiting seed germination was probably a result of the reduction in respiration caused by high CO 2 levels (Corbineau and Côme 1995). NaHCO 3 hastened germination in stored (1996) cypselas, except for GP, but not in fresh (1997) ones (Table 4). We have no explanation for this result, but we did observe that a few stored cypselas in each NaHCO 3 treatment germinated faster than any in the control. Ecological implications Cavers et al. (1995) found that some cypselas of Onopordum sp. that were extracted from the seed bank in Australia germinated immediately in a greenhouse when supplied with adequate moisture and suitable temperatures. Others germinated after several weeks in the greenhouse but only after the soil had been stirred. Still others required treatment with 0.01% GA 3 and some did not germinate at all but were apparently still dormant and viable at the end of the experiment. Our results suggest that cypselas of O. acanthium in London, Ont. have a range of germination responses, similar to those of Onopordum sp. examined by Cavers et al. (1995) in Australia. Our results also indicate that the seed banks in the soil will be different for different populations of O. acanthium, even if the populations mature in the same habitats. A concurrent 3-year study of seed banks of O. acanthium (Qaderi 1998 and unpublished data) has confirmed this supposition. An important factor for germination of seeds in soil is nitrate ion concentration (Steinbauer and Grigsby 1957). When the soil temperature is high and the moisture percentage is low the level of nitrate rises. For example, in England, values were near 0 mg/l in February but were as high as 90 mg/l in July (Popay and Roberts 1970b). This means that cypselas of O. acanthium in the soil will be exposed to different concentrations of nitrate ions at different times of the year. However, best germination of some species occurs when the soil has a moderate level of nitrate ions, a high percentage moisture and high day and night temperatures (Popay and Roberts 1970b). The levels of O 2 and CO 2 in the soil are also important for germination. In general, the greater the soil depth, the higher the CO 2 level (Popay and Roberts 1970a). Carbon dioxide in the soil can be dissolved and form compounds such as Na 2 CO 3 and NaHCO 3, which in turn may enhance or inhibit germination, depending on the species. NaHCO 3, which was used in this study, is found naturally in soils with alkaline ph levels and all of our plants were grown in alkaline soils (Table 1). The possession of a variety of germination rates is of great advantage for populations of O. acanthium. Cypselas that germinate quickly can establish a sizable seedling population in a few days under favourable conditions. Such seedlings often can survive subsequent environmental stresses whereas later-emerging seedlings at a younger phase of development will die. However, if under unexpectedly harsh conditions, all emerged seedlings die, there are still some cypselas emerging later that can replace the dead ones, fill the gaps, and perpetuate the species. This is one reason for the success of O. acanthium in many parts of the world. Variation among populations in response to chemicals Baskin and Baskin (1998) reported on hundreds of studies demonstrating that seeds from different populations within a species differ in germination response. In almost every case, they pinpointed clear-cut differences among the habitats of the different populations. In virtually all of the studies, the populations were separated by distances of many kilometres. We have not identified any other studies where the authors deliberately chose to compare the germination responses of seeds from a number of local populations, arising in similar habitats within a short distance of each other. The vast majority of studies of GA 3 or KNO 3 effects on germination have been conducted with single populations of a species; based upon our results, conclusions based on one population cannot characterize an entire species. Here, we show that even populations collected from very close proximity to each other, and from very similar habitats, can vary dramatically in their responses to these chemicals. The largest differences among populations were evident from the KNO 3 treatments. On the one extreme, fresh cypselas from the ESW population failed to respond to 0.025% KNO 3 (<6% germinated) but more than 74% germinated when exposed to 0.05% (Table 3). In contrast, nearly half of the Quarry population germinated in 0.025% KNO 3 and 75% in 0.05% versus 27% for the control. The same general response was found for the Ola and GP populations, but in both of these cases (and unlike Q and ESW) germination in 0.05% KNO 3 continued slowly from day 30 to the end of the experiment. In addition, the effects of KNO 3 on rate of germination varied among populations and among concentrations of KNO 3 within populations. Differences among populations were less for the GA 3 and NaHCO 3 treatments, but there were still several significant trends. For example, greater proportions of cypselas were triggered to germinate by GA 3 in populations with strong dormancy (ESW and Ola), in agreement with the findings of Meier (1995). Differences in final germination percentages among populations in response to GA 3 were fewer with stored cypselas than with fresh ones. GA 3 treatments revealed more differences among populations in the rate of germination than in final germination percentage. If we compare our results for final germination percentage with those of other scientists, the differences among different concentrations of GA 3 that they obtained

7 1162 Can. J. Bot. Vol. 78, 2000 can be explained (at least in part) by the shorter germination periods that they used. Differences among our populations became less over time because the germination percentages of populations with slower germination patterns eventually caught up to those of the others. Even seemingly small differences between populations can be of great importance. The 6 8% of viable cypselas from the Q and Ola populations that did not germinate in response to any GA 3 treatment could form a small but potentially long-lived pool of strongly dormant cypselas. The GP population had almost no cypselas in this category. NaHCO 3 treatments led to a significant decrease in final germination percentage of both fresh and stored cypselas in populations with greater germinability (GP, Q, and Ola), but not for ESW, a population with strong dormancy. This is a logical outcome, since NaHCO 3 would inhibit germination in nondormant but not dormant cypselas. Interpopulation variation could arise because of environmental or genetic factors or both. Cypselas from two populations studied (GP and ESW 1997) came from plants transplanted to the ESW research station in These plants grew there for more a year before producing cypselas. Differences in germination patterns in these two populations must have resulted from genetic differences and (or) differential sensitivity to environmental conditions between the populations. It is well known from studies of other species that seeds from different genotypes (parent plants) occupying the same habitat can have dissimilar dormancy patterns (Cavers and Harper 1966; Meyer and Allen 1999a, 1999b). Such dissimilarities could be the basis for eventual variation among local populations in germination response. In summary, our results clearly demonstrate that local populations can vary in response to chemical stimulators or inhibitors. Future studies of the effects of these chemicals should be done on several populations per species. Acknowledgements The authors would like to thank the Natural Sciences and Engineering Council of Canada for financial support through an operating grant to P.B.C. We are grateful to Dr. Robert Bailey for his statistical advice. We appreciate comments on the manuscript from Drs. Brenda Frick, Gordon Thomas, Phil Allen, and Frank Dennis, and an anonymous referee. References Adkins, S.W., and Adkins, A.L Effect of potassium nitrate and ethephon on fate of wild oat (Avena fatua) seed in soil. Weed Sci. 42: Alex, J.F., and Switzer, C.M Ontario weeds. Publication No Ontario Ministry of Agriculture and Food, Toronto. Alm, D.M., Stoller, E.W., and Wax, L.M An index model for predicting seed germination and emergence rates. Weed Technol. 7: Baskin, C.C., and Baskin, J.M Seeds: ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego, Calif. Bewley, J.D., and Black, M Physiology and biochemistry of seeds. 2. Viability, dormancy, and environmental control. Springer- Verlag, Berlin. Bewley, J.D., and Black, M Seeds: physiology of development and germination. 2nd ed. Plenum Press, New York. Cavers, P.B., and Harper, J.L Germination polymorphism in Rumex crispus and Rumex obtusifolius. J. Ecol. 54: Cavers, P.B., Groves, R.H., and Kaye, P.E Seed population dynamics of Onopordum over 1 year in southern New South Wales. J. Appl. Ecol. 32: Corbineau, F., and Côme, D Control of seed germination and dormancy by the gaseous environment. In Seed development and germination. Edited by J. Kigel and G. Gallili. Marcel Dekker, Inc., New York. pp Delouche, J.C., Still, T.W., Raspet, M.R., and Lienhard, M The tetrazolium test for seed viability. Tech. Bull. Miss. Agri. Exp. St. 51: Egley, G.H Stimulation of weed seed germination in soil. Rev. Weed Sci. 2: Evenari, M Light and seed dormancy. In Encyclopedia of plant physiology. Edited by W. Ruhland. Springer-Verlag, Berlin. pp Groves, R.H., and Kaye, P.E Germination and phenology of seven introduced thistle species in southern Australia. Aust. J. Bot. 37: Hilhorst, H.W.M., and Karssen, C.M Seed dormancy and germination: the role of abscisic acid and gibberellins and the importance of hormone mutants. Plant Growth Regul. 11: Karssen, C.M., and Laçka, E A revision of the hormone balance theory of seed dormancy: studies on gibberellin and/or abscisic acid-deficient mutants of Arabidopsis thaliana. In Plant growth substances Edited by M. Bopp. Springer-Verlag, Berlin. pp Kidd, F The controlling influences of carbon dioxide in the maturation, dormancy and germination of seeds. Part II. Proc. R. Soc. London Ser. B, 87: Kolk, H Viability and dormancy of dry stored weed seeds: studies in several species occurring frequently in Swedish grassland seed lots. Växtodling, Uppsala, Sweden. Mayer, A.M., and Poljakoff-Mayber, A The germination of seeds. 4th ed. Pergamon Press, Oxford. McKenzie, J., Schaefer, R.L., and Farber, E The student edition of Minitab for Windows. 2nd ed. Addison-Wesley, Reading, Mass. Meier, L.R Variation in seeds of Onopordum acanthium. M.Sc. thesis, Department of Plant Sciences, University of Western Ontario, London, Ont. Meyer, S.E., and Allen, P.S. 1999a. Ecological genetics of seed germination regulation in Bromus tectorum L. I. Phenotypic variance among and within populations. Oecologia, 120: Meyer, S.E., and Allen, P.S. 1999b. Ecological genetics of seed germination regulation in Bromus tectorum L. II. Reaction norms in response to a water stress gradient imposed during seed maturation. Oecologia, 120: Pérez-García, F Effect of the origin of the cypsela on germination of Onopordum acanthium L. (Asteraceae). Seed Sci. Technol. 21: Pérez-García, F., and Duran, J.M The effect of gibberellic acid on germination of Onopordum nervosum Boiss. seeds. Seed Sci. Technol. 18: Pérez-García, F., Iriondo, J.M., and Martinez-Laborde, J.B Germination behaviour in seeds of Diplotaxis erucoides and D. virgata. Weed Res. 35: Popay, A.I., and Roberts, E.H. 1970a. Factors involved in the dormancy and germination of Capsella bursa-pastoris (L.) Medick, and Senecio vulgaris L. J. Ecol. 58:

8 Qaderi and Cavers 1163 Popay, A.I., and Roberts, E.H. 1970b. Ecology of Capsella bursapastoris (L.) Medik and Senecio vulgaris L. in relation to germination behaviour. J. Ecol. 58: Qaderi, M.M Intraspecific variation in germination of Scotch thistle (Onopordum acanthium L.) cypselas. M.Sc. thesis, Department of Plant Sciences, University of Western Ontario, London, Ont. Qaderi, M.M., and Cavers, P.B Variation in germination response within Scotch thistle, Onopordum acanthium L., populations matured under greenhouse and field conditions. Écoscience, 7: Roberts, E.H Oxidative processes and the control of seed germination. In Seed ecology. Edited by W. Heydecker. Butterworths, London. pp SAS Institute Inc SAS user s guide: statistics. Cary, N.C. Schönbeck, M.W., and Egley, G.H Phase-sequence of redroot pigweed seed germination responses to ethylene and other stimuli. Plant Physiol. 68: Scifres, C.J., and McCarty, M.K Some factors affecting germination and seedling growth of Scotch thistle. Res. Bull. Nebr. Agr. Exp. St. 228: Sokal, R.R., and Rohlf, F.J Biometry: The principles and practice of statistics in biological research. 3rd ed. W.H. Freeman and Company, New York. Steinbauer, G.P., and Grigsby, B Interaction of temperature, light and moistening agent in the germination of weed seeds. Weeds, 5: Thompson, L.M Soils and soil fertility. 2nd ed. McGraw- Hill Book Company, Inc., New York. Toole, E.H., Hendricks, S.B., Borthwick, H.A., and Toole, V.K Physiology of seed germination. Annu. Rev. Plant Physiol. 7: White, R.E Principles and practice of soil science: the soil as a natural resource. 3rd ed. Blackwell Scientific Ltd, Cambridge. Young, J.A., and Evans, R.A Germination and persistence of achenes of Scotch thistle. Weed Sci. 20: Zar, J.H Biostatistical analysis. 4th ed. Prentice-Hall Inc. Upper Saddle River, N.J.