@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Land and Food Systems, Faculty of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Nolan, Daryl Guy"@en ; dcterms:issued "2010-08-22T17:15:56Z"@en, "1989"@en ; vivo:relatedDegree "Master of Science - MSc"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """The problematic reinfestation of chemically-treated sites by diffuse and spotted knapweed {Centaurea diffusa and C. maculosa) is thought to occur from dormant seeds in the soil. This study confirmed that reserves of dormant seeds are present in the soil of infested sites, although greater numbers of seeds were recovered from senesced plants. Knapweed plants produce both non-dormant and dormant seeds (germination polymorphism), the relative proportions of which vary between individual plants within a site, as well as between bulk samples collected from different sites. Two types of dormant seeds were identified. Dormancy of some seeds was broken by exposure to red light ('light-sensitive seeds'). Light-sensitivity was evident at 10, 15, 20, 25, and 30°C. Germination in light-sensitive seeds was shown to be mediated by phytochrome. A lesser number of dormant seeds failed to respond to red light ('light-insensitive seeds'). Dry after-ripening released dormancy in both light-sensitive and light-insensitive seeds. However, no apparent loss of dormancy from after-ripening occurred when the relative humidity was too low or too high. At the highest relative humidity level tested (90.7%), dormancy was induced in some seeds while other seeds died. Dormancy was also induced when imbibed seeds were incubated in darkness at 25, 30, 35, and 40°C for 5 days. Dormancy induction was greatly enhanced by incubating submerged seeds in de-oxygenated water (anaerobiosis). However, some seeds died when incubated anaerobically for 5 days. Dormancy was broken in a small percentage of dormant seeds by incubation in a 10 mM solution of potassium nitrate or potassium nitrite; 100 mM potassium nitrite killed most seeds. Gibberellic acid was a much stronger germination stimulant. Some dormant seeds germinated at 25 °C if they were previously chilled at 3°C. To compare laboratory findings with field germination behaviour, seeds from two samples of each species were buried to a depth of about 3 cm in mesh packets during November, April and August near Salmon Arm, B.C. Seeds exhibiting higher levels of germination in darkness in vitro also germinated to higher levels in situ when burial occurred in November. However, burial in April and August led to lower germination levels in situ. Light sensitivity was still prominent following 17 months of burial. Most of the decline in viable seed numbers during burial were attributable to in situ germination. Theoretical discussions of the source of germination polymorphism in knapweed seeds, the importance of light to field germination and seedling mortality, and a potential strategy for controlling these weeds are presented."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/27605?expand=metadata"@en ; skos:note "SEED GERMINATION CHARACTERISTICS OF •CENTAUREA DIFFUSA AND C. MACULOSA By DARYL GUY NOLAN B.Sc, The University of British Columbia, 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Plant Science) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA November 1989 ©Daryl Guy Nolan, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date October m DE-6 (2/88) i i ABSTRACT The problematic reinfestation of chemically-treated sites by diffuse and spotted knapweed {Centaurea diffusa and C . maculosa) is thought to occur from dormant seeds in the soil. This study confirmed that reserves of dormant seeds are present in the soil of infested sites, although greater numbers of seeds were recovered from senesced plants. Knapweed plants produce both non-dormant and dormant seeds (germination polymorphism), the relative proportions of which vary between individual plants within a site, as well as between bulk samples collected from different sites. Two types of dormant seeds were identified. Dormancy of some seeds was broken by exposure to red light ('light-sensitive seeds'). Light-sensitivity was evident at 10, 15, 20, 25, and 30 ° C . Germination in light-sensitive seeds was shown to be mediated by phytochrome. A lesser number of dormant seeds failed to respond to red light ('light-insensitive seeds'). D r y after-ripening released dormancy in both light-sensitive and light-insensitive seeds. However, no apparent loss of dormancy from after-ripening occurred when the relative humidity was too low or too high. A t the highest relative humidity level tested (90.7%), dormancy was induced in some seeds while other seeds died. Dormancy was also induced when imbibed seeds were incubated in darkness at 25, 30, 35, and 40 ° C for 5 days. Dormancy induction was greatly enhanced by incubating submerged seeds in de-oxygenated water (anaerobiosis). However, some seeds died when incubated anaerobically for 5 days. Dormancy was broken in a small percentage of dormant seeds by incubation in a 10 m M solution of potassium nitrate or potassium nitrite; 100 m M potassium nitrite killed most seeds. Gibberellic acid was a much stronger germination stimulant. Some dormant seeds germinated at 25 ° C if they were previously chilled at 3 ° C . To compare laboratory findings with field germination behaviour, seeds from two samples of each species were buried to a depth of about 3 cm in mesh packets during November, Apr i l and August near Salmon A r m , B. C . Seeds exhibiting higher levels of germination in darkness in vitro also germinated to higher levels in situ when burial occurred i i i in November. However, burial in Apr i l and August led to lower germination levels in situ. Light sensitivity was still prominent following 17 months of burial. Most of the decline in viable seed numbers during burial were attributable to in situ germination. Theoretical discussions of the source of germination polymorphism in knapweed seeds, the importance of light to field germination and seedling mortality, and a potential strategy for controlling these weeds are presented. iv T A B L E O F C O N T E N T S A B S T R A C T ii L I S T O F T A B L E S viii L I S T O F F I G U R E S xi A C K N O W L E D G M E N T xiii 1.0 I N T R O D U C T I O N 1 2.0 L I T E R A T U R E R E V I E W O F K N A P W E E D B I O L O G Y A N D C O N T R O L 2.1 D e t r i m e n t a l C o n s e q u e n c e s of K n a p w e e d Invas ion 2.1.1 Forage Production 2 2.1.2 Rangeland Management 3 2.1.3 Livestock Injury 3 2.1.4 Miscellaneous Effects 3 2.2 G e o g r a p h i c D i s t r i b u t i o n 2.2.1 Present Distribution 4 2.2.2 Potential Distribution in Canada 4 2.3 C o n t r o l 2.3.1 Chemical 5 2.3.2 Biological Control 6 2.3.3 Cultural 2.3.3.1 Exclusion 6 2.3.3.2 Range seeding 7 2.3.3.3 Burning 7 2.3.3.4 Cultivation 7 2.3.3.5 Mowing or grazing 8 2.3.3.6 Fertilization 8 2.3.3.7 Irrigation 9 2.4 K n a p w e e d B i o l o g y 2.4.1 Taxonomy 9 2.4.2 Favoured Habitat 11 2.4.3 Life Cycle 2.4.3.1 Life span 12 2.4.3.2 Germination 14 2.4.3.3 Seedlings 16 2.4.3.4 Rosettes 16 2.4.3.5 Bolted plants 20 2.4.3.6 Flower production 22 2.4.3.7 Seed production 22 2.4.3.8 Seed dispersal 25 2.4.3.9 Seed bank 28 2.5 K n a p w e e d S e e d P h y s i o l o g y 28 V 3.0 EFFECT OF LIGHT ON KNAPWEED SEED GERMINATION 3.1 Background 3.1.1 Light Sensitive Germination in the Asteraceae 30 3.1.2 Properties of Phytochrome 30 3.1.3 Phytochrome Mediation of Field Germination 34 3.1.4 Objectives 36 3.2 Materials and Methods 3.2.1 Seed Collection and Storage 37 3.2.2 Incubation Conditions 37 3.2.3 Light Sources 39 3.2.4 Germination Behaviour of Seeds From Different Sites and Clutches ..39 3.2.5 Reversibility of R and FR Effects on Germination 41 3.2.6 Effect of R Duration on Germination 42 3.2.7 Statistical Procedures 42 3.3 Results 3.3.1 Seeds from Different Sites and Clutches 42 3.3.2 Effect of Sequential R and FR Light Exposures 46 3.3.3 Effect of Duration of R 46 3.4 Discussion 48 4.0 EFFECT OF LIGHT QUALITY DURING SEED MATURATION 4.1 Background 54 4.2 Materials and Methods 57 4.3 Results and Discussion 58 5.0 EFFECT OF AFTER-RIPENING ON GERMINATION BEHAVIOUR 5.1 Background 61 5.2 Materials and Methods 5.2.1 Effect of Aging on Germination Behaviour 61 5.2.2 Effect of Relative Humidity on After-Ripening 63 5.3 Results 5.3.1 Effect of Aging on Germination Behaviour 5.3.1.1 Level of ND seeds 63 5.3.1.2 Level of LI seeds 65 5.3.1.3 Level of IR seeds 71 5.3.2 Effect of Relative Humidity on After-Ripening 5.3.2.1 ND seed levels 71 5.3.2.2 IR seed levels 79 5.3.2.3 LI seed levels 79 5.4 Discussion 5.4.1 Conformity of Results with Previous Reports 79 5.4.2 Proposed Model of Dormancy Transition in Dry Knapweed Seeds 85 5.4.3 After-Ripening: A Source of Germination Polymorphism in Diffuse and Spotted Knapweed? 86 5.4.4 Regulation of Field Germination 88 5.4.5 After-Ripening: Considerations in Seed Germination Behaviour Studies 88 v i 6.0 E F F E C T O F T E M P E R A T U R E O N I M B I B E D K N A P W E E D S E E D S 91 6.1 C o n s t a n t T e m p e r a t u r e 6.1.1 Background 92 6.1.2 Materials and Methods 92 6.1.3 Results 6.1.3.1 Dark germination 93 6.1.3.2 Light sensitivity 93 6.1.3.3 Seed viability 93 6.1.4 Discussion 96 6.2 E f fec t o f T e m p e r a t u r e D u r i n g D a r k Incubat ion o n S u b s e q u e n t G e r m i n a t i o n B e h a v i o u r at 25 ° C 98 6.2.1 Materials and Methods 98 6.2.2 Results 6.2.2.1 Dark germination 99 6.2.2.2 Germination following exposure to 2 min R 102 6.2.2.3 Germination following 1 d R exposure 103 6.2.2.4 Discussion 103 6.3 E f fec t o f C h i l l i n g D u r a t i o n o n S e e d G e r m i n a t i o n 105 6.3.1 Materials and Methods 106 6.3.2 Results 6.3.2.1 Transfer to 20 ° C .....106 6.3.2.2 Transfer to 25 ° C 108 6.3.3 Discussion 108 7.0 E F F E C T O F A N A E R O B I O S I S O N S E C O N D A R Y D O R M A N C Y I N D U C T I O N 7.1 B a c k g r o u n d 113 7.2 M a t e r i a l s a n d M e t h o d s 113 7.3 R e s u l t s 7.3.1 Dark Germination 114 7.3.2 Seed Viability 114 7.3.3 Germination Following 2 min R 119 7.3.4 Germination Following l d R 119 7.4 D i s c u s s i o n 124 8.0 E F F E C T O F N I T R A T E A N D N I T R I T E O N G E R M I N A T I O N 8.1 B a c k g r o u n d 126 8.2 M a t e r i a l s a n d M e t h o d s 8.2.1 Effect of Nitrate on Germination 8.2.1.1 Effect of nitrate concentration and temperature on dark germination 126 8.2.1.2 Effect of light and nitrate 127 8.2.2 Effect of Nitrite 8.2.2.1 Prel iminary screening of effect on dark germination 127 8.2.2.2 Effect of nitrite and light 127 8.3 R e s u l t s 8.3.1 Nitrate 8.3.1.1 Dark germination 128 8.3.1.2 Effect of nitrate and R at 20 ° C 128 8.3.2 Nitrite 8.3.2.1 Dark germination 133 8.3.2.2 Nitrite and R treatment 133 8.4 D i s c u s s i o n 133 v i i 9.0 EFFECT OF GIBBERELLIC ACID ON KNAPWEED SEED GERMINATION 9.1 Background 139 9.2 Materials and Methods 9.2.1 Germination Dose Response to Exogenous G A g 139 9.2.2 G A g , Light , and Seed Coat Excision 139 9.3 Results 9.3.1 Germination Dose Response 141 9.3.2 G A g , Light , and Seed Coat Excision 141 9.4 Discussion 144 10.0 SEED PERSISTENCE IN THE SOIL AND ON SENESCED PLANTS 10.1 Background 145 10.2 Numbers of Soil-borne Knapweed Seeds 10.2.1 Materials and Methods 146 10.2.2 Results 147 10.3 Numbers of Seeds Retained on Senescent Plants 10.3.1 Materials and Methods 150 10.3.2 Results 151 10.3.3 Discussion 153 10.4 Effect of Burial on the Germination Characteristics of Knapweed Seeds 10.4.1 Materials and Methods 155 10.4.2 Results and Discussion 156 11.0 A THEORETICAL EXPLANATION FOR KNAPWEED DISTRIBUTION AND ITS POSSIBLE MANAGERIAL IMPLICATIONS 170 12.0 CONCLUSIONS 179 13.0 BIBLIOGRAPHY 181 14.0 APPENDIX 201 v i i i LIST OF TABLES Table Page 1 Knapweed Distribution in Western North Amer ica 4 2 Seed Collection Sites (British Columbia) 38 3 Germination of Diffuse Knapweed Seeds Collected F r o m Different Sites 43 4 Germination of Spotted Knapweed Seeds Collected F r o m Different Sites 44 5 Comparative Germination of Different Knapweed Seed Clutches Within Sites 45 6 Reversibility of R and F R Effects on Knapweed Seed Germination 47 7 M e a n Germination of Seeds Matured in Capitula Surrounded by Various L ight Filters 59 8 Effect of After-ripening at 25 ° C on Percentage of N D Seeds 64 9 Effect of After-ripening at 3 ° C on Percentage of N D Seeds 66 10 Effect of After-ripening at -20 ° C on Percentage of N D Seeds 67 11 Effect of After-ripening at 25 ° C on Percentage of L I Seeds 68 12 Effect of After-ripening at 3 ° C on L I Seed Percentages 69 13 Effect of After-ripening at -20 ° C on L I Seed Percentages 70 14 Effect of After-ripening at 25 ° C on Percentage of IR Seeds 72 15 Effect of After-ripening at 3 ° C on Percentages of IR Seeds 73 16 Effect of After-ripening at -20 ° C on Percentages of IR Seeds 74 17 Effect of R H Dur ing a 30 d After-ripening Period on Dark Germination of Diffuse Knapweed 75 18 Effect of R H Dur ing a 30 d After-ripening Period on Dark Germination of Spotted Knapweed 76 19 Effect of R H Dur ing a 30 d After-ripening Period on the Percentages of Non-viable Diffuse Knapweed Seeds 77 20 Effect of R H Dur ing a 30 d After-ripening Period on the Percentages of Non-viable Spotted Knapweed Seeds 78 21 Effect of R H Dur ing a 30 d After-ripening Period on Germination of Diffuse Knapweed Following a 2 min R Treatment < 80 22 Effect of R H Dur ing a 30 d After-ripening Period on Germination of Spotted Knapweed Following a 2 min R Treatment 81 ix 23 Effect of RH During a 30 d After-ripening Period on Germination of Diffuse Knapweed Following a 1 d R Treatment 82 24 Effect of RH During a 30 d After-ripening Period on Germination of Spotted Knapweed Following a 1 d R Treatment 83 25 Effect of Incubation Temperature on Diffuse and Spotted Knapweed Seed Viability 97 26 Effect of Chilling at 3 °C on Dark Germination of Diffuse Knapweed at 20 °C 107 27 Effect of Chilling at 3 °C on Dark Germination of Spotted Knapweed at 20 °C 109 28 Effect of Chilling at 3 °C on Dark Germination of Diffuse Knapweed at 25 °C 110 29 Effect of Chilling at 3 °C on Dark Germination of Spotted Knapweed at 25 °C I l l 30 Effect of Anaerobiosis on Diffuse Knapweed (D7) Dark Germination 115 31 Effect of Anaerobiosis on Spotted Knapweed (S10) Dark Germination 116 32 Effect of Anaerobiosis on Viability of Diffuse Knapweed (D7) Seeds Used in the Dark Germination Experiment 117 33 Effect of Anaerobiosis on Viability of Spotted Knapweed (S10) Seeds Used in the Dark Germination Experiment 118 34 Effect of Anaerobiosis on Germination of Diffuse Knapweed (D7) Seeds Following a 2 Min R Exposure 120 35 Effect of Anaerobiosis on Germination of Spotted Knapweed (S10) Seeds Following a 2 Min R Exposure 121 36 Effect of Anaerobiosis on Germination of Diffuse Knapweed (D7) Seeds Following a 1 Day R Exposure 122 37 Effect of Anaerobiosis on Germination of Spotted Knapweed (S10) Seeds Following a 1 Day R Exposure 123 38 Effect of Nitrate and R on Diffuse Knapweed Germination 131 39 Effect of Nitrate and R on Spotted Knapweed Germination 132 40 Effect of Nitrite on the Germination of Diffuse and Spotted Knapweed Seeds in Darkness at 20 °C 134 41 Effect of Nitrite and R on Diffuse Knapweed Germination 135 X 42 Effect of Nitrite and R on Spotted Knapweed Germination 136 43 Effect of Light and G A 3 on Diffuse Knapweed (D5) Germination 142 44 Effect of Light and G A 3 on Spotted Knapweed (Si) Germination 143 45 Seed Numbers in the Soil Profile on 3-30-1987 149 46 Seed Retention on Senesced Plants 152 47 Dark Germination of Exhumed Diffuse Knapweed Seeds (Falkland 1985) 158 48 Dark Germination of Exhumed Diffuse Knapweed Seeds (Winfield 1984) 159 49 Dark Germination of Exhumed Spotted Knapweed Seeds (Westwold 1985) 160 50 Dark Germination of Exhumed Spotted Knapweed Seeds (Falkland 1984) 161 51 Germination of Exhumed Diffuse Knapweed Seeds (Falkland 1985) Following 2 Min R 162 52 Germination of Exhumed Diffuse Knapweed Seeds (Winfield 1984) Following 2 Min R 163 53 Germination of Exhumed Spotted Knapweed Seeds (Westwold 1985) Following 2 Min R 164 54 Germination of Exhumed Spotted Knapweed Seeds (Falkland 1984) Following 2 Min R 165 55 Effect of R Duration on the Germination of Diffuse and Spotted Knapweed 201 56 Effect of Incubation Temperature on Diffuse Knapweed Germination 202 57 Effect of Incubation Temperature on Spotted Knapweed Germination 203 58 Effect of Temperature During a 5 d Dark Incubation Period on the Subsequent Germination Behaviour of Diffuse and Spotted Knapweed Seeds 204 59 Effect of Nitrate on Germination of Diffuse Knapweed Seeds in Darkness 205 60 Effect of Nitrate on Germination of Spotted Knapweed Seeds in Darkness 206 61 Effect of G A 3 on Dark Germination of Diffuse Knapweed 207 62 Effect of GAg on Dark Germination of Spotted Knapweed 208 xi LIST O F F I G U R E S Figure Page 1 Diffuse knapweed {top) and spotted knapweed (bottom) capitula 10 2 Diffuse knapweed plants observed in 1989 in the Canadian Pacific Railway yard in Nanaimo, B. C 13 3 Spotted knapweed perennation through shoot production from the crown 15 4 A dense clump of spotted knapweed seedlings 17 5 Diffuse knapweed (top) and spotted knapweed (bottom) rosettes 18 6 Spotted knapweed seedling 19 7 Diffuse knapweed rosettes 21 8 Achenes of diffuse knapweed (left) and spotted knapweed (right) 23 9 Senesced diffuse knapweed plants 26 10 Spotted knapweed plants with open capitula and perennation from the crowns 27 11 Spectral distribution of red, far-red, and green light sources 40 12 Effect of the duration of R exposure on the germination of diffuse and spotted knapweed seeds 49 13 Effect of incubation temperature on the germination of three samples of diffuse knapweed seeds incubated in darknes or previously exposed to 2 min R at 25 °C....94 14 Effect of incubation temperature on the germination of three samples of spotted knapweed seeds incubated in darkness or previously exposed to 2 min R at 25 °C....95 15 Effect of temperature during a 5 d dark incubation period on the subsequent germination behaviour of diffuse knapweed seeds incubated for an additional 5 d at 25 °C 100 16 Effect of temperature during a 5 d dark incubation period on the subsequent germination behaviour of spotted knapweed seeds incubated for an additional 5 d at 25 °C 101 17 Effect of potassium nitrate on the dark germination of diffuse knapweed seeds at different temperatures 129 18 Effect of potassium nitrate on the dark germination of spotted knapweed seeds at different temperatures 130 19 Dose-response of GAg on the dark germination of diffuse and spotted knapweed 140 20 Soil core collection 148 x i i 21 Soil core partitioning 148 22 Seed burial plot showing an exhumed pot containing soil and ready for transport back to the laboratory 157 23 Seed bank study site near Salmon Arm, B. C 174 x i i i A C K N O W L E D G E M E N T My highest thanks goes to Dr. M. K. Upadhyaya for his guidance, patience, and perserverance over the years. Sincere thanks also go to the remaining members of my committee, Dr. M. D. Pitt, Dr. W. Vidaver, and Dr. V. C. Runeckles, for their assistance. I thank Dr. Jolliffe for the use of the walk-in incubator and to Dr. G. W. Eaton for advice on statistics. I am grateful for the financial support provided by the British Columbia Agricultural Services Coordinating Committee and the British Columbia Cattlemens'* Association. I am especially grateful for the technical and emotional support provided by my wife Donna, and the financial and moral support provided by our parents. 1 1.0 I N T R O D U C T I O N Diffuse and spotted knapweed (Asteraceae: Centaurea diffusa Lam. and C. maculosa Lam.) are herbaceous introduced weeds which reduce rangeland productivity in western North America. Low palatability, prolific seed production, and drought tolerance enables knapweed to displace desirable forage species in grazed grasslands. Although several herbicides control knapweed, the relatively low monetary return from rangeland makes chemical control a financially untenable long-term strategy because of the problem of site reinvasion (Cranston et al. 1983). Biological control provides the promise of lasting control through the introduction of several agents that cumulatively stress knapweed enough to reduce populations to a level where forage loss is no longer a concern. However, the biological control programme has not yet reduced stands of these weeds. Less attention has been given to developing cultural methods of stressing knapweed that could augment biological and chemical control. Studies examining cultural means of controlling these weeds are often trial and error type studies that fail to target a specific weakness in the knapweed life cycle. If the critical factor(s) which delineate whether habitats are, or are not, suitable for knapweed development could be identified, alternative methods of stressing these weeds could be devised and tested in a systematic manner. The greatest mortality of knapweed populations occurs prior to rosette establishment (Roze 1981; Myers and Berube 1983). From the limited information available it is not possible to conclude which environmental factor is the most important cause of this mortality. In many species, timing of emergence is a critical determinant of seedling survival. Light, temperature, and nitrogenous ions have been identified as the key factors regulating seed germination in field situations (Roberts 1972). Consequently, the primary objective of this study was to better understand the influence of these environmental factors on diffuse and spotted knapweed seed germination. 2 2.0 L I T E R A T U R E R E V I E W O F K N A P W E E D B I O L O G Y A N D C O N T R O L 2.1. D e t r i m e n t a l C o n s e q u e n c e s of K n a p w e e d I n v a s i o n 2.1.1 Forage Production The knapweeds are the most serious rangeland weeds in British Columbia (Hoyles 1979). The primary negative impact of knapweed invasion is displacement of desirable forage species. Forage production of knapweed infested rangeland in British Columbia was only 15% of comparable areas in excellent condition (Watson and Renney 1974; Harris and Cranston 1979). Diffuse knapweed in the Crimea region of the U.S.S.R. constituted 65 to 73.6% of dry matter in long-fallow pasture and 40 to 65% of dry matter in alfalfa fields (Popova 1960). The displacement of desirable range species by knapweed is a serious problem because knapweed is so unpalatable to cattle that plants are generally consumed only in situations where succulent bolted plants are present in overgrazed sites (Watson 1972). Sesquiterpene lactones, toxic deterrents to herbivores which impart a bitter flavour to the foliage of many Asteraceae (Heywood et al. 1977), have been isolated from diffuse knapweed (Muir and Majak 1983). However, sheep, goats, and cattle in Montana eat substantial amounts of the succulent bolted stems of spotted knapweed (Kelsey and Mihalovich 1987). In addition, the coarse, spiny nature of the plants deters animals from utilizing underlying forage species (Popova 1960). The replacement cost of forage displaced by knapweed on 40,000 ha of rangeland in British Columbia was estimated to be $320,000 per annum; the corresponding loss in beef output value at the farm gate was $1.5 million per year (Cranston et al. 1983). If knapweed were to infest all invasion-susceptible rangeland in British Columbia, losses in beef production could approach $41 million annually (Cranston et al. 1983). The economic impact of these weeds would increase dramatically if knapweed spread to all susceptible rangelands in western Canada. Knapweed could displace 2.5 million tonnes of dry forage annually in this 10 million ha area (Harris and Myers 1984). 3 2.1.2 Rangeland Management Management of British Columbian rangelands is complicated by the fact that most current knapweed infestations encompass areas utilized for spring and fall grazing; land that that is in short supply in parts of the B. C. Interior (Muir 1986). As herd size is often dictated by the availability of spring and fall grazing areas, further invasion of these sites could force cattlemen to reduce stock numbers (Muir 1986). 2.1.3 Livestock Injury Spiny diffuse knapweed capitula can inflict physical injury to the mouths and digestive tracts of livestock (Popova 1960; Watson and Renney 1974). In addition, anecdotal reports suggest that ingestion of large quantities of knapweed causes toxic symptoms in horses and sheep (Popova 1960; Higgins and Schirman 1977; Maddox 1979). Reports have implicated other species of Centaurea, namely yellow star-thistle (C. solstitialis) and Russian knapweed (C. repens) in cases of stock poisoning (Fowler 1965; Larson 1970; Young et al. 1970a, 1970b; Perdomo and De Freitas 1978; Cordy 1978; Gard et al. 1979). In these cases, a disorder called nigropallidal encephalomalacia caused afflicted animals to starve due to development of brain lesions in areas of the brain responsible for mouth muscle control (Fowler 1965; Larson 1970; Young et al. 1970a, 1970b). Sesquiterpene lactones are believed to be the toxic principle in these plants (Cordy 1978; Stevens 1982; Stevens and Merrill 1985). However, cattle observed grazing bolted knapweed plants near Kamloops, B. C. exhibited no adverse effects (Watson 1972) and cattle, sheep, and goats in Montana were apparently unaffected by ingestion of spotted knapweed, whether eaten fresh in the field or in silage or hay (Kelsey and Mihalovich 1987). 2.1.4 Miscellaneous Effects Further detrimental effects attributed to knapweed infestation include: reduced market value of property (Maddox 1979; Cranston et al. 1983), flavouring of dairy milk 4 (Maddox 1979), reduced desirability of recreational areas (Watson and Renney 1974), and increased fence repair costs [caused by tumbling plants of diffuse knapweed] (Maddox 1979). 2.2 Geographic Distribution 2.2.1 Present Distribution Extensive infestations of diffuse and spotted knapweed occur in the dry intermountain regions of western North America. Knapweed infests an estimated 83,000 ha of rangeland in British Columbia (Hamlen and Hansen 1984) and a further 1.4 million ha in the states of Washington, Oregon, Idaho, and Montana (Maddox 1977, 1979). Diffuse knapweed was the more widespread species in all jurisdictions but Montana (Table 1). Table 2.1 Knapweed Distribution in Western North America Location Diffuse Spotted Reference British Columbia 69,000 ha 13,400 ha Hamlem and Hansen 1984 Washington/Idaho 331,600 ha 32,000 ha Maddox 1979 Oregon 300,000 ha Maddox 1979 Montana 600 ha 800,000 ha Maddox 1979; Story 1984 2.2.2 Potential Distribution in Canada Harris and Cranston (1979) estimated the potential range of diffuse knapweed in western Canada by comparing the soil types, and mean monthly precipitation and temperature levels of this area with infested areas in Europe. Similarly, Chicoine (1984) predicted the potential range of spotted knapweed in Montana by estimating the total land area in the state with edaphic and climatic conditions (soil type, elevation, frost-free days, mean July temperature maximum, and potential evapotranspiration) similar to infested areas of the state. 5 Climatic and soil conditions similar to those found in the native Eurasian distribution of these weeds occur in 8.4 to 10.7 million ha of grasslands in Western Canada; British Columbia contains 1.1 million ha of this total, with the remainder in Alberta and Saskatchewan (Harris and Cranston 1979). The actual risk of knapweed invasion over much of this area is not known because the factors responsible for site susceptibility to knapweed invasion are poorly understood. 2.3. Control 2.3.1 Chemical Herbicides are currently the only reliable means of eliminating stands of knapweed from most rangeland areas. Chemical control of spotted knapweed in Montana increased grass production 30 to 75% (Baker 1980). Chicoine (1984) reported 200 to 380% greater perennial grass production in plots treated with picloram (4-amino-3,5,6-trichloro-2-pyridine carboxylic acid) 4 years earlier to control spotted knapweed. Dicamba (3,6-dichloro-2-methoxybenzoic acid) and 2,4-D ([2,4-dichlorophenoxy] acetic acid) provide non-residual control of knapweed (Popova 1960; Furrer and Fertig 1965; Renney and Hughes 1969; Cranston 1985). However, knapweed stands typically re-establish from fall germinating seeds in sites treated with non-residual herbicides (Renney and Hughes 1969). Consequently, picloram is the only herbicide widely used to control knapweed because its residual action can provide three to four years of control (Cranston 1985). Even with picloram, the necessity for periodic retreatment limits use of this method of control to containment programmes and spot treatment of small infestations. The costs incurred in applying picloram are recovered only if forage production is restored to the average level of knapweed-free rangeland for 17 years (Cranston et al. 1983). Unfortunately, even under ideal soil and climatic conditions, a single application of picloram is only able to prevent reinvasion for seven years (Cranston et al. 1983). In addition, picloram adversely 6 affects the growth of young grass seedlings and thus complicates reseeding operations (Scifres and Halifax 1972; Hubbard 1975). 2.3.2 Biological Control The establishment of a self-perpetuating complex of biological control agents is a promising means of overcoming the inherent monetary limitations associated with controlling these weeds. Theoretically, the introduction of 6 to 8 agents should stress the weeds enough to reduce populations below an economically-damaging threshold (Harris and Cranston 1979; Cranston 1985). Two seed-reducing flies, Urophora affinis Frauenfeld and Urophora quadrifasciata Meigan have successfully established over much of knapweed's distribution and have reduced seed production up to 95% in some areas (Harris and Cranston 1979; Cranston 1985; Maddox 1979). However, even a 99.9% reduction in seed production may not cause population declines (Schirman 1981). One major weakness of Urophora fly attack is that large numbers of seed heads escape attack during the seven-day interval between the non-overlapping first and second generations of this agent (Roze 1981). In addition, diffuse knapweed produces large numbers of seeds late in the season following the first generation of the flies; the second generation is less effective at reducing seed numbers (Roze 1981). Several other agents have been introduced, but their eventual impact on knapweed populations is unknown. Detailed accounts of the biological control programme against the knapweeds are available elsewhere (Harris and Myers 1984; Schroeder 1984; Muir 1986). 2.3.3 Cultural 2.3.3.1 Exclusion Preventing seed introduction into uninfested areas by excluding knapweed contaminated hay or vehicles is the most effective means of preventing knapweed invasion 7 (Cranston 1985). Eradication of knapweed introductions can prevent the rapid escalation of control costs which occur if populations are allowed to spread (Hoyles 1979). 2.3.3.2 Range seeding Range seeding with a suitable perennial forage species reduced the susceptibility of disturbed or overgrazed areas to invasion (Watson and Renney 1974; Cranston 1985) and increased forage production (Hubbard 1975). Seeding desirable grass species, such as crested wheatgrass (Agropyron cristatum) and Russian wild rye (Elymus junceus), prevented diffuse knapweed invasion by increasing knapweed seedling mortality, presumably, through soil moisture depletion (Berube and Myers 1982). However, the success of reseeding appeared dependent on soil moisture availability. Although crested wheatgrass suppressed diffuse knapweed invasion in Cache Creek plots, the more mesic Pritchard plots were invaded (Berube and Myers 1982). Furthermore, the resistance of different grasses to knapweed invasion differed. Bluebunch wheatgrass (Agropyron spicatum) was the most susceptible to invasion of 6 grasses studied (Agriculture Canada 1979). Bawtree and Cranston (1984) reported that sodgrasses compete with knapweed better than bunchgrasses. 2.3.3.3 Burning Burning led to the almost complete disappearance of diffuse knapweed within two years in the Crimea (Popova 1960). The positive effect of burning was manifested gradually, and seemed to be associated with a subsequent stimulation of grass growth rather than direct knapweed mortality. However, the fire hazard posed to forested areas precludes the use of this technique in British Columbia (Watson and Renney 1974). 2.3.3.4 Cultivation Knapweed populations are susceptible to cultivation, but annual cultivation or deep ploughing is necessary to prevent reinfestation from soil-borne seeds (Popova 1960). A single cultivation in early May reduced diffuse knapweed mature plant and rosette numbers; a second cultivation in early August eliminated all remaining plants (Watson 1972). However, 8 cultivation produces favourable conditions for knapweed invasion (Strang et al. 1979). Disturbance of the soil to a depth of 10 cm in early May significantly increased diffuse knapweed seedling numbers relative to undisturbed controls by unearthing viable seeds in the seed bank (Watson 1972). Consequently, other measures are needed to prevent reinvasion. Regardless, cultivation is not feasible on much of the land infested by these weeds (Harris and Cranston 1979). 2.3.3.5 Mowing or grazing Popova (1960) reported that mowing increased diffuse knapweed populations in the Crimea. Conversely, mowing decreased seed producing diffuse and spotted knapweed plant numbers in British Columbia (Watson 1972; Watson and Renney 1974). These contradictory results may reflect the different methods of population measurement employed. Watson determined seed producing plant numbers shortly (one or two months) after mowing, whereas Popova counted bolted plants one year later. Mowing was also reported to reduce germination of diffuse and spotted knapweed seeds (Watson 1972; Watson and Renney 1974). Unfortunately, viability does not appear to have been determined in this study. Therefore, the results may reflect either reduced seed viability or, alternatively, lowered germinability resulting from greater dormancy in the progeny of mowed plants. Spotted knapweed was nearly eliminated in some Montana pastures by sheep grazing in spring and early summer, and again in August; sheep were subsequently removed to prevent overgrazing of associated grasses (Cox 1983). 2.3.3.6 Fertilization Although the competitive ability of desirable forage species improved with fertilization, applications of fertilizer to existing stands of knapweed increased the percentage cover of the weed (Watson 1972). Popova (1960) reported that fertilization was ineffective at 9 reducing diffuse knapweed populations in the Crimea. Application of ammonium sulphate fertilizer increased spotted knapweed production in Idaho (Wattenbarger et al. 1979). In Washington, fertilization using 16-20-0 NPK decreased spotted knapweed populations in the first year while increasing forage production (Sheley and Roche 1982). In areas where diffuse knapweed is highly stressed due to lack of moisture, nitrogen fertilization might increase stress by elevating moisture depletion by other species (Berube and Myers 1982). 2.3.3.7 Irrigation Knapweed does not persist in irrigated alfalfa (Harris and Cranston 1979). The mechanism by which irrigation suppresses knapweed is unknown. However, irrigation is not a practical control option over much of the infested rangeland in British Columbia. 2.4. Knapweed Biology 2.4.1 Taxonomy Centaurea species are members of the Cardueae tribe of the Asteraceae (Compositae) family. Members of this genus are discerned primarily on the basis of distinctive phyllary morphology. Diffuse knapweed phyllaries bear an apical spine, while phyllaries of spotted knapweed are pectinate with a dark brown or black margin (Figure 1). Although diffuse and spotted knapweed seedlings and rosettes are difficult to distinguish, the persistent nature of senesced aerial portions of these weeds can be utilized to identify the species present in a site. The terms diffuse and spotted refer to the characteristic decurrent branching habit, and phyllary colouration of these species, respectively. Detailed descriptions of the distinguishing taxonomic characteristics of diffuse and spotted knapweed are available elsewhere (Moore 1972; Moore and Frankton 1974; Watson and Renney 1974). Possible confusion regarding the taxonomic classification of the species referred to as Centaurea maculosa in North America is reviewed by Schroeder (1984). 10 Figure 1. Diffuse knapweed (top) and spotted knapweed (bottom) capitula. Note the variable flower colour and characteristic phyllary morphology. 11 2.4.2 Favoured Habitat Knapweed is found in various habitats, both in British Columbia (Watson 1972) and in the U.S.S.R. (Popova 1960). In the Crimea, only forested areas are free of the weed (Popova 1960). Stands of diffuse and spotted knapweed are common throughout the semi-arid intermountain areas of British Columbia. Spotted knapweed is more prevalent in cooler, more mesic areas within this region (Watson 1972), while areas with periods of summer drought are favoured by diffuse knapweed (Harris and Cranston 1979). Sites infested with diffuse knapweed in N.E. Washington were in a 40 to 50 cm precipitation zone, while sites infested with spotted knapweed in northern Idaho were in a 64 to 76 cm precipitation zone (Schirman 1981). Although areas experiencing summer precipitation deficits are prone to invasion, plants found in irrigated or high rainfall sites can be particularly robust (Harris and Cranston 1979). Irrigation increases plant size and flower production (Watson and Renney 1974). Diffuse knapweed's competitive advantage over other species may be restricted to a specific set of moisture conditions as knapweed density was lower within gullies and depressions near Kamloops, B. C , whereas, recruitment from sown seed was greater within gullies at a drier site near Cache Creek, B. C. (Berube and Myers 1982). Watson (1972) found that knapweed plant densities were not correlated with either chemical or physical soil properties. However, Brunisols and Brown Soil regions appear especially susceptible to invasion by diffuse knapweed, while spotted knapweed is associated with Dark Brown soils (Harris and Cranston 1979). Knapweed densities were correlated with subjective ratings of the degree of soil disturbance (Watson 1972). This supported Atkinson and Brink's (1953) conclusion that all sites in the dry southern interior of British Columbia with a disturbed A horizon appear susceptible to invasion. Disturbance of soil or vegetation cover through overgrazing, drought, or cultivation predisposes sites to invasion. Consequently, knapweed is common on roadsides, railway and hydro right-of-ways, and overgrazed rangelands or pastures (Popova 1960; Watson and Renney 1974; Cranston 1985). Conversely, intensive cultivation, irrigation (probably through stimulated growth of competing 12 species), and healthy perennial vegetation cover inhibit knapweed establishment (Watson and Renney 1974; Cranston 1985). Marsden-Jones and Turrill (1954) reported that several perennial Centaurea species found in Britain (C. jacea, C. nemoralis, C. nigra, and C. scabiosa) became increasingly less frequent the further north the location, where they were generally restricted to disturbed habitats such as road and railway verges. These British knapweed species are also notably intolerant of shaded conditions (Marsden-Jones and Turrill 1954). Knapweed can be found in areas of high precipitation where edaphic conditions or vegetation disturbance favour its survival. For example, mature diffuse knapweed plants were observed in the gravel roadbed of Westminster Highway, Richmond, B.C. from 1987 to 1989 (Nolan, personal observation). The combination of periodic mowing and gravel substrate likely allowed the plants to establish in this high rainfall area. Knapweed may have been introduced into this area by rail traffic as the plants are adjacent to a Canadian National Railway crossing. Rail traffic may also have introduced diffuse knapweed to Vancouver Island as a stand was observed in 1989 in the Canadian Pacific Railway yard in Nanaimo, B. C. (Figure 2). Other knapweed infestations have been associated with rail traffic (Yule 1987). Established populations of knapweed persist as the dominant species even if the area is not, subsequently, disturbed by grazing (Harris and Cranston 1979; Hoyles 1979). Overgrazing does not always precede diffuse knapweed invasion into grassland (Myers and Berube 1983). Disturbance of vegetation by wildlife (e.g. coyote dens and gopher holes) allow knapweed to establish in excellent condition rangeland (Hoyles 1979). Once established in disturbed areas, diffuse knapweed is often capable of spreading to neighbouring \"virgin land\" (Popova 1960). 2.4.3 Life cycle 2.4.3.1 Life span Depending upon environmental conditions, diffuse and spotted knapweed plants can exhibit monocarpic or polycarpic behaviour, and annual, biennial, or perennial life spans. Figure 2. Diffuse knapweed plants observed in 1989 in the Canadian Pacific Railway yard in Nanaimo, B. C. 14 Examination of secondary xylem rings in spotted knapweed taproots in Montana revealed plants as old as 9 years at some sites (Boggs and Story 1987). Diffuse knapweed plants have persisted in the rosette stage for 4 years (Agriculture Canada 1980). However, knapweed exhibits the formalized separation of vegetative and reproductive phases of the life cycle typical of \"biennial\" species (sensu Grime 1979). Spotted knapweed exhibits a propensity for polycarpic behaviour because of its ability to produce shoots from the crowns of flowering plants (Watson and Renney 1974) [Figure 3]. British knapweed species (C. jacea, C. nemoralis, C. nigra, and C. scabiosa) also exhibit this half-rosette hemicryptophyte behaviour where flowering stems die back and shoots arise from buds overwintering on the crown near the soil surface (Marsden-Jones and Turrill 1954). However, this vegetative shoot production does not enable knapweed populations to spread in the manner of rhizomatous weeds. Vegetative shoots produce larger plants; population expansion occurs entirely from seed production. 2.4.3.2 Germination Emergence of diffuse and spotted knapweed exhibits a bimodal distribution as seed germination occurs mainly during periods of favourable soil moisture in the fall and spring (Watson and Renney 1974; Schirman 1981), although lesser numbers of seedlings emerge throughout the growing season (Roze 1981). This is a common characteristic of ruderal species (Roberts and Feast 1970). Optimum germination levels are attained by seeds on the soil surface (Watson 1972; Watson and Renney 1974; Spears et al. 1980). Diffuse knapweed seeds do not emerge when buried deeper than 3 cm (Popova 1960; Watson 1972; Spears et al. 1980); spotted knapweed seed buried 5 cm deep failed to emerge (Spears et al. 1980), while Watson (1972) reported less than 10% emergence from this depth. Roze (1981) reported that smaller percentages of diffuse and spotted knapweed seeds germinated when the rate of sowing was increased, and that germination was possibly inhibited by shading of established rosettes. Figure 3. Spotted knapweed perennation through shoot production from the crown (Salmon Arm, B. C , March 30, 1987) 16 2.4.3.3 Seedlings Spotted knapweed seedling densities seldom exceeded 10,000 seedlings/m in Montana, while annual seed production in a mature stand ranged between 60,000 to 86,000 9 seeds/m (Chicoine and Fay 1984). In localized areas, knapweed seedlings form a continuous cover over the soil (Figure 4). Substantial diffuse knapweed mortality occurs prior to rosette establishment (Roze 1981; Myers and Berube 1983). Knapweed populations are believed to be regulated in part by density-dependent mortality of small seedlings (Roze 1981). Roze found 73% and 84% mortality in small seedlings of diffuse and spotted knapweed, respectively. Survival of large seedlings and rosettes of spotted knapweed was not regulated by intraspecific competition, whereas similar mortality in diffuse knapweed was density dependent (Roze 1981). Seedlings can persist in an arrested state during periods of low soil moisture (Roze 1981). In Spokane, Washington seedling leaves desiccated during August but the seedlings regrew in the fall provided they had emerged before June (Schirman 1981). Large seedlings of diffuse and spotted knapweed successfully over-winter (Roze 1981). However, dry conditions following seedling emergence can cause high mortality (Watson and Renney 1974; Schirman 1981). Survival through periods of summer drought appears related to germination timing, as delayed seeding reduced seedling numbers and flowering plant percentages the subsequent year (Schirman 1981). 2.4.3.4 Rosettes Seedlings develop into extremely drought resistant rosettes (Figure 5) [Berube and Myers 1982]. Muir (1986) reported that diffuse knapweed produces an elongate taproot while spotted knapweed roots are more fibrous. However, spotted knapweed also produces deep-reaching taproots. For example, spotted knapweed rosettes only 2 cm in diameter were found with taproots over 10 cm long (Figure 6). Taproots enable grassland forbs to access soil moisture reserves beyond the reach of typically shallow-rooted grasses (Grime 1979). Figure 4. A dense clump of spotted knapweed seedlings. (Salmon Arm, B. C, March 30, 1987) Figure 5. Diffuse knapweed (top) and spotted knapweed (bottom) rosettes. 19 Figure 6. Spotted knapweed seedling. (Salmon Arm, B. C, March 30, -1987) 20 Diffuse knapweed rosettes are the most competitive stage in the life cycle largely due to soil moisture depletion by their extensive root systems (Roze 1981). The production of allelopathic compounds was once proposed to be the mechanism of knapweed's aggressive competitive ability (Fletcher and Renney 1963). However, although sesquiterpene lactones have been isolated from diffuse knapweed (Muir and Majak 1983), field and pot studies have provided no evidence that allelopathic effects contribute to the invasiveness of these weeds (Agriculture Canada 1986; Muir et al. 1987). Knapweed rosettes are often among the first green plants to be seen in infested areas in the spring (Figure 7). Spotted knapweed rosettes can retain chlorophyllous tissue over the winter in Salmon Arm [removal of melting snow in early spring revealed green rosettes] (Nolan personal observation). Other Centaurea found in Britain also exhibit this \"winter-green\" characteristic (Marsden-Jones and Turrill 1954). 2.4.3.5 Bolted plants The transition from the vegetative phase to the reproductive phase in knapweed is marked by rapid stem elongation (bolting). Carbohydrates accumulated in taproots during the rosette stage, enable biennial plants to rapidly produce large flowering structures (Grime 1979). Rosettes usually bolt the May following overwintering. Vernalization is not essential for plants to bolt (Watson 1972; Roze 1981). Schirman (1981) noted that a few of the earliest emerging plants bolt in the year of seeding. The transition from rosette to bolted plant appears to be associated with rosette size (Roze 1981), and can be induced artificially with exogenous gibberellic acid (Upadhyaya 1986). Bolted spotted knapweed plants perennate by offshoot production from the crown (Watson 1972; Watson and Renney 1974; Nolan personal observation). However, although previously described as vegetatively produced rosettes (Watson 1972; Watson and Renney 1974), these shoots reflect perennation as they arise from the crowns of senesced plants and do not develop independent root systems (Nolan personal observation). Consequently, Figure 7. Diffuse knapweed rosettes. (Vernon, B. C, March 30, 1987) 22 multiple-stemmed spotted knapweed plants are common. Diffuse knapweed usually produces a single compound and spreading stem. However, removal of the bolted stems (prior to senescence) can stimulate multiple shoot production (Nolan personal observation). Diffuse and spotted knapweed can attain densities of 500 and 400 plants per square meter, respectively (Watson and Renney 1974). However, Schirman (1981) found flowering stem densities of diffuse and spotted knapweed averaged only 24 and 44 flower stems/m in Washington and Idaho. 2.4.3.6 Flower production Schirman (1981) found that plants were more likely to flower if seeding took place earlier in the previous growing season. From 70 to 95% of plants sown in March and April flowered, while all plants sown in June or July remained vegetative (Schirman 1981). The asteraceous capitulum is a racemose inflorescence in which flowers develop sequentially in a centripedal fashion (Burtt 1977). Capitula production is indeterminate, beginning in July or August and continuing until environmental conditions become unfavourable (Watson and Renney 1974; Roze 1981). Spotted knapweed generally ceases flowering by mid-August, while diffuse knapweed continues to initiate seed heads until cold kills the plants in October [provided adequate moisture is present] (Roze 1981). Asynchronous flowering ensures maturation of some seeds under adverse environmental conditions, such as drought (Roze 1981). 2.4.3.7 Seed production As \"plants within the same family tend to share the same basic seed structure\" (Corner 1976), a description of seed structure reported in other Asteraceae follows. Knapweed 'seeds' (Figure 8) are actually fruits: achenes. Some workers employ the alternate term cypselas for asteraceous fruits (Marsden-Jones and Turrill 1954; Popova 1960; Marks and Prince 1981, 1982). A distinction between achenes and cypselas is that the former arise 23 Figure 8. Achenes of diffuse knapweed (left) and spotted knapweed (right). Colour differences apparent in the figure are not characteristic of differences between the species rather seed colour in both species can range from light to dark. The black bar is 1mm wide. 24 from a monocarpellate ovary, while the latter arise from a polycarpellate ovary in which all but one ovule aborts during development (Martin 1978). Asteraceous achenes are dry, sclerenchymatic, indehiscent bicarpellate fruits arising from an inferior ovary, in which the seed has a single point of attachment to the pericarp (Esau 1967). In Lactuca, the embryo is enclosed by a fruit coat consisting of an outer layer of maternal tissue (pericarp), and an inner semi-permeable membranous layer (endosperm) [Borthwick and Robbins 1928; Atwater 1980]. Karawya et al. (1974) have described the macro and micromorphology of Centaurea calcitrapa. The point of attachment of the achene to the receptacle in Centaurea is marked by a characteristic oblique-basal scar (Moore and Frankton 1974). This basal scar in the Asteraceae has also been called callus, separation tissue, podocarp, or carpopodium (see Haque and Godward 1984). The carpopodium consists of specialized thick-walled tissue developing in the abscission zone between the seed and receptacle whose purpose is to facilitate detachment of the seed from capitulum (Haque and Godward 1984). In addition, the carpopodium and micropyle region is the primary route for water entry into the seed of many asteraceous species (Sheldon 1974). Pappus and seed morphology tend to improve the contact of this region with the soil and thus improve seed water uptake (Sheldon 1974). Spotted knapweed seeds are approximately 60% heavier than diffuse knapweed seeds and bear a pappus lacking on most diffuse knapweed seeds (Watson 1972; Roze 1981). Seeds begin to mature by mid-August, provided that insect pollination occurs (Watson and Renney 1974). Diffuse and spotted knapweed plants produced, on average, 13 and 30 seeds per capitula, respectively, prior to the introduction of biological control agents (Schirman 1981). Diffuse and spotted knapweed plants produced up to 900 and 400 seeds per plant under typical rangeland conditions, and up to 18,000 and 25,000 seeds per plant under irrigated conditions, respectively, prior to the introduction of biological control agents (Watson 1972; Watson and Renney 1974). Seed production by diffuse knapweed in N.E. Washington, 25 and by spotted knapweed in northern Idaho varied during 1973-1976 from 112,000 to 481,000 seeds/m2, and 113,000 to 296,000 seeds/m2, respectively (Schirman 1981). Although Urophora gall flies caused an 80% reduction of seed production, Roze (1981) found that diffuse and spotted knapweed plants, in sites with peak populations of the flies, 9 dispersed 3200 and 2000 seeds/m , respectively. Schirman (1981) estimated that only 0.1 percent of potential seed production was needed to maintain populations. 2.4.3.8 Seed dispersal Seed dispersal in knapweed occurs from late summer through the following growing season. Seed dispersal in knapweed is passive. Movement of stems by wind or passing animals propel seeds from the capitula (Watson and Renney 1974), a mechanism called jacitation (Van Der Pijl 1982). In addition, decay of the stem base of diffuse knapweed plants releases the senescent aerial portion of plants, whereafter, the globular shape of the plant facilitates a rolling, tumbleweed-like movement by the wind (Ridley 1930). Long-distance dispersal occurs when plants are caught in vehicle undercarriages, and when seeds become caught up in animal fur, or in vehicles (Watson and Renney 1974; Cranston 1985). The rapid spread of knapweed in British Columbia is believed to have been aided by the movement of infested hay from Washington state (Muir 1986). Diffuse knapweed capitula have small distal openings through which seeds pass one at a time (Watson and Renney 1974). Consequently, many diffuse knapweed seeds overwinter within capitula on the fibrous senesced plants (Figure 9). Over 80% of diffuse knapweed seeds overwintering on plants in Washington were viable the following April (Schirman 1981). Conversely, seeds of spotted knapweed are loosely held in the capitulum because phyllaries open widely when dry (Figure 10). Spreading of pappus bristles on drying pushes seeds from the capitula of Centaurea species (Harper et al. 1970). However, hygroscopic opening and closing of the phyllaries in response to drying and wetting of the capitulum can influence the timing of seed dispersal (Nolan personal observation). 26 Figure 9 . Senesced diffuse knapweed plants. (Chase, B. C, January 1, 1989) 27 Figure 10. Spotted knapweed plants with open capitula and perennation from the crowns. (Salmon Arm, B. C , March 30, 1987) 28 Furthermore, peripherally situated seeds appear to have a greater propensity for retention in the capitulum (Nolan personal observation) as they are often firmly held between the involucre and the receptacle, as reported in Senecio jacobaea and Picris echioides (Burtt 1977). 2.4.3.9 Seed bank Myers and Berube (1983) reported that the density of diffuse knapweed seeds in the top 3 cm of soil at a site near Kamloops, British Columbia was about 1,000 times the density of flowering knapweed plants. However, the proportion of viable seeds in this total was not reported. Numbers of viable spotted knapweed seeds in the soil of an infested site in Montana were as high as 1,000 seeds/m (Chicoine 1984; Chicoine and Fay 1984). Various cultural practices (harrowing, rolling, burning, mowing) did not hasten declines in the seed bank relative to plots where seed production was prevented with herbicide treatment (Chicoine 1984; Chicoine and Fay 1984). 2.5 K n a p w e e d S e e d P h y s i o l o g y Although some form of seed dormancy is prerequisite to the formation of a seed bank, germination regulation in knapweed is poorly understood. Watson (1972) found germination of both species occurred over the temperature range of 7 to 34 °C. Both species germinate to high levels (80 to 100%) under optimum conditions (Popova 1960; Watson 1972; Watson and Renney 1974; Schirman 1981). Light did not increase germination levels above those found in darkness for either species, although continuous white light treatments of 700 ft-c reduced germination of both species by 17% (Maguire and Overland 1959; Watson 1972; Watson and Renney 1974). Only 15% of spotted knapweed seeds buried at a depth of 2.5 cm germinated following 12.5 months burial in two sites in Montana (Chicoine 1984; Chicoine and Fay 1984). Ungerminated seeds exhibited 90% viability (Chicoine 1984; Chicoine and Fay 1984). Consequently, spotted knapweed seeds were reported to lack any form of physiological dormancy, and the failure of the seeds to germinate during burial was attributed to improper 29 temperature, light, or moisture conditions (Chicoine 1984). Although other workers have also considered the lack of light to be a factor responsible for enforced dormancy, light-sensitive germination (phytochrome-mediated dormancy) should be considered a true physiological form of innate dormancy equivalent to dormancy resulting from immature embryos, internal inhibitors, and impermeable seed coats. Roze (1981) found 24% and 3% of seeds collected from spotted and diffuse knapweed sites, respectively, were dormant (germination was checked every three days during imbibition at room temperature [23 °C] for 12 d). Watson and Renney (1974) reported that germination levels of diffuse and spotted knapweed increased from 40 to 68%, and 20 to 80%, respectively, following 25 days of dry storage at room temperature. These reports suggested that knapweed seeds possess a primary (innate) dormancy that is lost through dry-after-ripening. 30 3.0 EFFECT OF LIGHT ON KNAPWEED SEED GERMINATION 3.1 Background 3.1.1 Light Sensitive Germination in the Asteraceae Light plays an important regulatory role in the germination of many weedy members of the Asteraceae (e.g. Wesson and Wareing 1967, 1969a, b; Gorski 1975; Gorski et al. 1977, 1978; Bostock 1978; Silvertown 1980; Grime et al. 1981). Other genera within the same tribe as knapweed (Cardueae) exhibited greater levels of germination in light [e.g. Arctium lappa, (Gorski et al. 1978), A. minus (Maguire and Overland 1959; Grime et al. 1981), Cirsium arvense (Bostock 1978; Grime et al. 1981), C. acaulon, C. vulgare (Grime et al. 1981), and C. palustre (Stoutjesdijk 1972; Gorski et al. 1978; Grime et al. 1981; Pons 1983)]. The germination of several species of Centaurea was stimulated by white light (C. repens, C. solstitialis, Maguire and Overland 1959; C. nigra, C. scabiosa Grime et al. 1981), or inhibited by light rich in far-red wavelengths (C. cyanus Gorski et al. 1977; C. kotschyana, C. oxylepis, C. rhenana Gorski et al. 1978; and C. nigra Silvertown 1980). This information strongly suggested that light might play a role in the germination regulation of diffuse and spotted knapweed seeds. 3.1.2 Properties of Phytochrome The germination of light-sensitive seeds is controlled by the signal-transducing photoreceptor phytochrome. An overview of properties of phytochrome relevant to this study follows. Detailed reviews of the photophysiology and photochemistry of phytochrome are available elsewhere (Briggs and Rice 1972; Mitrakos and Shropshire 1972; Smith 1975; Kendrick and Frankland 1976; Kendrick and Smith 1976; Quail 1976; Satter and Galston 1976; Smith and Kendrick 1976; Pratt 1982). Phytochrome exists in two interconvertible forms, a red light (R) absorbing form designated P r , and a far-red (FR) light absorbing form designated P f r When P r absorbs R (630-680 nm) it transforms to the Pf r isomer. Absorption of FR (730-750 nm) converts Pf r to 31 P r . However, small amounts of Pfr are formed under monochromatic FR because the pigment absorbs radiation to various degrees over the whole visible spectrum (Black 1969). Differences in the quantum yields of these two conversions, and the extinction co-efficients of P r and Pf r, result in a greater energy efficiency of the P r to Pf r photoconversion (Bewley and Black 1982). Pf r is considered to be the active form of phytochrome as its presence in seeds initiates physiological processes that culminate in germination. Germination is likely controlled by a threshold response to the quantity of Pfr present (Bewley and Black 1982). Consequently, R stimulates the germination of most light-sensitive seeds, while FR either inhibits germination, or reverses the effect of a previous R exposure. Flint and McAlister (1935, 1937) first reported that R stimulated, and FR inhibited seed germination. Borthwick et al. (1952, 1954) first reported the reversibility of R and FR effects on germination and proposed the existence of the pigment phytochrome. Differences in the amount of light required for germination by different species may reflect quantitative differences in phytochrome (Bewley and Black 1982). For example, there are differences in the percentage of Pf r required to initiate germination in two cultivars of lettuce. May Queen lettuce seeds require much less than the 50% conversion of Pf r to P r required by Grand Rapids seeds; Grand Rapids seeds require exposure to R to germinate, while May Queen seeds germinate in darkness (Smith 1973). In addition, environmental stresses imposed on the seed embryo can influence the quantity of Pfr required for germination (Black 1969). For example, light-mediated dormancy in Grand Rapids lettuce was manifested only at supra-optimal temperatures (Borthwick et al. 1954), and in Progress lettuce, light-sensitivity was only apparent when seeds were treated with coumarin (Nutile 1945). Photoconversion of phytochrome in imbibed seeds is relatively unaffected by differences in water content (Hsiao and Vidaver 1973; Berrie et al. 1974; Loercher 1974; 32 Duke 1978) and temperature (Ikuma and Thimann 1964; Isikawa and Fugii 1961; Taylorson and Hendricks 1973b; Vidaver and Hsiao 1975). The germination response elicited by photoconversion of phytochrome is largely dependent on the spectral quality of light; the intensity and duration of light exposure have less effect (Cumming 1963). Other processes influencing the amount and state of the two phytochrome isomers include thermal reversion of Pf r to P r , synthesis of Pf r, and destruction of P r . However, discussion will be limited to thermal reversion, as there is a lack of strong evidence for the latter two processes in seeds (Frankland and Taylorson 1983). Pf r is thermodynamically unstable and undergoes dark reversion to the thermodynamically stable P r isomer in hydrated seeds (Bewley and Black 1982). Thermal dark reversion may explain high temperature effects on light-sensitive seed germination (Borthwick et al. 1954). Under such conditions, seeds may fail to germinate under prolonged exposure to low fluence rate R because \"the rate of non-photochemical reversion of Pf r to P r will be greater than the rate of photochemical conversion of P r to Pf r\" (Frankland 1981). However, a thermolabile factor has been implicated in Lactuca sativa germination regulation in addition to the phytochrome system (Takeba and Matsubara 1976). Some reversion of P r to Pf r also occurs in darkness (Smith 1973) and as a result some seeds will germinate in darkness unless intermittent or continuous FR exposures are given to continually remove the Pf r. For example, seeds of the lettuce variety May Queen germinate well in darkness but continuous FR prevents germination (Boisard 1969). The transformation of one phytochrome isomer to the other involves a series of intermediate structures (Kendrick and Spruit 1977) and is a pure photoreaction, independent of temperature and oxygen (Ikuma and Thimann 1964). Conversion of some intermediates is prevented in dehydrated tissue, consequently, full sensitivity to light is usually only apparent when seeds are, at least, partially imbibed (Bewley and Black 1982). Dry seeds (approximately 6% water content) typically exhibit little response to light, and maximum 33 sensitivity to R light is attained between 13 and 22% moisture content (Berrie et al. 1974; Hsiao and Vidaver 1971). Pfr was postulated to interact with a \"reaction partner\" X; increases in light responsivity during imbibition, and declines in sensitivity following prolonged incubation in darkness have been interpreted as changes in the quantity or receptivity of X (Karssen 1970; Taylorson and Hendricks 1973b; Duke et al. 1977). Factor X is believed to be a membrane (Kahn 1960; Taylorson and Hendrick 1973b; Duke et al. 1977). Phytochrome conversion from P r to Pf r may represent a shift from a peripheral associate of a membrane to a transmembrane position; this transformation may alter transmembrane transport by producing aqueous pores (Smith 1977). Evidence indicates that phytochrome is associated with plasma membranes. For example, phytochrome potentiates rapid changes in membrane behaviour (Tanada 1968). Processes initiated by the presence of Pf r, and culminating in germination, are respiration dependent as anaerobic conditions imposed following establishment of high Pf r levels inhibit germination (Ikuma and Thimann 1964). Phytochrome involvement in germination regulation can be demonstrated if repeatable red/far-red photoreversibility of germination occurs [i.e. red light stimulates germination but subsequent exposure to far-red light negates this effect] (Borthwick et al. 1954; Black 1969; Popay and Roberts 1970a; Toole 1973). Reversibility is lost gradually as the interval between R and FR exposures increases because the physiological processes initiated by Pf r commit the seed to germination (Borthwick et al. 1954). The gradual nature of the \"escape\" from reversibility reflects variability among individuals in the time Pfr action is needed for germination (Frankland 1981). Seeds requiring relatively long periods of Pfr action often fail to germinate in response to a single short duration R treatment because the quantity of Pfr declines as dark reversion to Pr occurs (Frankland 1981). 34 3.1.3 Phytochrome Mediation of Field Germination The energy reserves of germinating seeds are eventually depleted if photosynthetically-active radiation (PAR) is intercepted by overlying plants. Under conditions of limited PAR availability, mortality can be reduced by delaying germination until light conditions are favourable for seedling survival. Some species would not survive to maturity if light-sensitive germination did not prevent germination under vegetation cover (Sagar and Harper 1960). Similarly, the survival of small buried seeds is enhanced if germination is delayed until the seeds are on, or near, the soil surface as deeply buried seeds would exhaust their food reserves before emerging (Popay and Roberts 1970b). Such selective pressure has lead to the development of a light-sensitive germination mechanism that enables plants to cope with plant and soil layers overlying their seeds: phytochrome-mediated germination. Light-sensitive germination is less evident in domesticated species (Gorski et al. 1978), probably as a result of years of selection for strains giving immediate germination (Salisbury 1961). The phytochrome pigment system not only detects the presence of light, it also discerns light quality. Phytochrome-mediated germination enables seeds to detect an overlying plant canopy; a situation where light is present but low PAR may limit seedling survival (Bewley and Black 1982). Consequently, light-sensitive germination is a mechanism favouring germination in safe sites (Angevine and Chabot 1979; Silvertown 1980, 1981; Grime 1981; Marks and Prince 1982). Phytochrome has been implicated in the germination regulation of many species (see Smith 1975; Gorski et al. 1978; Karssen 1980/8la). This form of germination regulation is especially important for species whose seedlings are not competitive with established vegetation (Grime 1979, 1981; Gross and Werner 1982). Germination flushes following the disturbance of soil (Sauer and Struik 1964; Wesson and Wareing 1969a) or vegetation cover are often manifestations of this light-sensitive germination regulating mechanism. The seasonal distribution of germination in some species 35 may result from the interaction of the phytochrome pigment system with the incident light reaching the seed, and other environmental factors, especially temperature. The bichromatic ratio of photon fluence rates at wavelengths of 660 nm (R) and 730 nm (FR) is used to characterize light quality pertinent to phytochrome conversion (Monteith 1976; Smith and Holmes 1977). Sunlight has a red to far-red ratio (R:FR) of approximately 1.2 (Taylorson and Borthwick 1969; Frankland 1981). Consequently, as photoconversion favours Pf r formation, light-sensitive seeds exposed to unfiltered sunlight germinate (if other conditions are favorable). However, sunlight filtered through a plant canopy has an inhibitory effect on germination because chlorophyll alters spectral quality by attenuating R wavelengths more strongly than FR (Taylorson and Borthwick 1969). A single leaf or a dense leaf canopy can reduce R:FR to 0.18 and 0.1, respectively (Taylorson and Borthwick 1969; Frankland 1981). Such FR-rich light environments inhibit germination as phytochrome is largely converted to P r Chlorophyllous leaf tissue is such an effective FR filter it has been used to demonstrate reversibility in Lactuca sativa (Black 1969) and Rumex obtusifolius (Taylorson and Borthwick 1969) seed germination. Consequently, phytochrome is an excellent light detector allowing seeds to sense shading by neighbouring plants. Inhibition of germination by chlorophyllous vegetation is a well documented phenomenon (see Sagar and Harper 1960; Van Der Veen 1970; Stoutjesdijk 1972; Smith 1973; Grime and Jarvis 1974; King 1975). For example, germination of the asteraceous weeds Anthemis cotula, Carduus nutans, Cirsium arvense, C. vulgare, Senecio jaobaea and Silybum marianum was inhibited by pasture cover (Phung and Popay 1981). Similarly, fewer seedlings of Senecio vulgaris emerged when vegetation was left undisturbed (Popay and Roberts 1970b), while leaf-canopy-filtered light inhibited germination and induced a light requirement in seeds of Bidens pilosa (Fenner 1980a, b). Wesson and Wareing (1967) found that 20 of 23 species (6 Asteraceae) present in exhumed soil required light for germination. Species of the Cardueae tribe (e.g. Carduus) \\ 36 form persistent light-sensitive seed banks in the soil (Roberts and Chancellor 1979). Exclusion of light was the major factor preventing the germination of buried Senecio vulgaris seeds (Popay and Roberts 1970a). The seeds of many species acquire a light requirement during burial in the soil (Wesson and Wareing 1969b; Holm and Miller 1972). For example, Cirsium palustre seeds acquired a light requirement when buried (Pons 1984). Overlying soil also affects light quality incident upon shallowly buried seeds. Frankland (1981) found that light transmission through 1 mm of soil reduced the R:FR from approximately 1.2 to 0.6. The change in the R:FR is dependent upon the physical properties (i.e. soil type, water content) of the soil (Frankland 1981). The quantity of light energy passing through soil also drops sharply with depth (Frankland 1981). Consequently, the phytochrome system enables buried seeds to lie dormant until exposed to sunlight. This mechanism minimizes mortality that would arise if seeds germinated at soil depths from which food reserve depletion would precede seedling emergence (Roberts and Totterdell 1981), and allows the formation of the persistent soil-borne seed banks which complicate weed control (Frankland 1981). Seed banks disperse germination over time, thereby minimizing the danger of catastrophic population mortality should adverse environmental conditions, or control measures implemented by man, occur following seedling emergence. 3.1.4 Objectives The primary objectives of the studies in this section were 1) to confirm the existence of seed dormancy in diffuse and spotted knapweed, 2) to determine whether or not knapweed seed germination is light sensitive and mediated by phytochrome, and 3) to determine whether seeds collected from different wild populations and individual plants within a site exhibit different germination characteristics. 37 3.2 M a t e r i a l s a n d M e t h o d s The general methodology described hereafter also applies to subsequent chapters unless otherwise stated. 3.2.1 Seed Collection and Storage Seeds were collected during August and September of 1985 from populations growing in the interior of British Columbia (Table 2) by beating plants against the inside of a large pail. Seed collection was much more productive under dry conditions, especially in the case of spotted knapweed, as periods of high relative humidity (i.e. early morning dew or rainfall) caused hygroscopic closure of senesced capitula. Most plant debris was removed in the field by screening and winnowing the seeds before placement in sealed containers. Bulk samples, consisting of seeds pooled from a large number of plants, as well as samples from up to 10 individual plants within a site were collected. Seeds collected from individual plants within a site were stored in paper envelopes placed in a single sealed container. All seeds were stored at -20 °C within 24 h of collection. 3.2.2 Incubation Conditions Seeds were incubated in 9-cm petri dishes lined with a Whatman No. 1 filter disc moistened with 5 ml distilled water. Unless otherwise stated, each treatment consisted of 3 or 4 replicates of 50 seeds each. Germination (radicle protrusion) was recorded after 5 days of incubation in darkness at 25 °C. Preliminary studies indicated that germination commenced before 24 h and reached a maximum within 3 days of incubation at this temperature (data not shown). In initial runs, petri dishes were placed in metal tins sealed with aluminum foil (to exclude light) and a tight fitting plastic lid during incubation. However, a strong inhibition of germination in tins with rusted inner surfaces necessitated a switch to plastic containers. Therefore, at least one run of each experiment was incubated in sealed 5 litre Frig-O-Seal plastic food savers lined with a paper towel moistened with 40 ml distilled water. Where 38 Table 2. Seed Collection Sites (British Columbia) Species and Date of collection collection Site site code d/m/yr location Habitat Diffuse knapweed D l 30/8/85 Lytton Roadside D2 1/9/85 Falkland Disturbed area D3 2/9/85 Chase Pasture D4 1/9/85 Sunnybrae Roadside D5 2/9/84 Vernon Field D6 2/9/85 Vernon Roadside D7 2/9/85 Kelowna Disturbed area D8 2/9/85 Kelowna Gravel pit Spotted knapweed SI 14/8/85 Westwold Pasture S2 16/8/85 Canoe Disturbed area S3 16/8/85 Canoe Roadside S4 17/8/85 Salmon Arm Disturbed area S5 18/8/85 Salmon Arm Gravel pit S6 18/8/85 Chase Roadside S7 18/8/85 Kamloops Roadside S8 19/8/85 Enderby Gravel pit S9 19/8/85 Grindrod Roadside S10 19/8/85 Savona Roadside 3 9 prolonged light treatments were given, petri dishes were sealed with parafilm strips to prevent water evaporation from the dishes during placement in the light treatment box. To exclude light during dark incubation, containers were placed in cardboard file boxes covered with opaque cloth, or in specially constructed bags consisting of a layer of aluminum foil sandwiched between two layers of 8 mil black polyethylene. Experiments were conducted in a light-tight walk-in growth chamber. 3.2.3 Light Sources Addition of water and any subsequent manipulations were made under a dim green safelight (Westinghouse 20 W cool-white fluorescent tube, F20T12/CW, wrapped with 4 layers of Roscolux No. 90 dark yellow-green celluloid filter. Red light (R) and far-red light (FR) exposures were initiated after 8 h of imbibition and, unless otherwise stated, were of 2 and 10 min duration, respectively. R was obtained by filtering light from 5 cool-white fluorescent tubes (40W, General Electric, F40CW) through a 3 mm thick red filter (Rohm and Haas Plexiglas No. 2423). FR was produced by filtering light from 8 incandescent bulbs (100 W, Westinghouse) through single layers of red Plexiglas and Roscolux No. 95 medium blue green celluloid filters. The spectral distribution and irradiance of green, R and FR sources (Figure 11) was determined using an International Light IL700 radiometer and IL785A photomultiplier. Irradiances for R and FR sources were determined at seed level, and 5 cm from the fluorescent tube for the green safelight. The R:FR ratios of R and FR sources were 3.88 and 0.04, respectively. 3.2.4 Germination Behaviour of Seeds From Different Sites and Clutches Experiments were designed to determine whether different lots of diffuse and spotted knapweed seeds exhibit distinctive physiological responses (i.e. germination polymorphism sensu Williams and Harper 1965; Cavers and Harper 1966). Germination in darkness, and following exposure to R, was determined in factorial experiments for seeds collected from 40 16-i CN U C o D T I I 1 1 1 1 1 1 1 300 350 400 450 500 550 600 650 700 750 800 850 900 Wavelength [nm] Figure 11. Spectral distribution of red, far-red, and green light sources. _ g Irradiance values of R, FR, and green light need to be multiplied by factors of 10\"., 10' , 10\" respectively. 41 different sites and individual plants within a site. Employing the terminology used by Silvertown (1984), seeds from an individual mother plant will be called a \"clutch.\" Viable seeds that failed to germinate in darkness at 25 °C were considered primary (innately) dormant. Seed viability was determined at the end of the experiment by pouring off excess water, adding 2 ml of 1 mM GAg (Sigma), incubating under room lighting and temperature for 5 days, and then determining the percentage germination. The ability of GAg to stimulate germination is detailed in chapter 9. The viability of seeds that failed to germinate in response to GAg was tested by removing the distal end of the seed coat and cotyledons (less than one-fifth of total seed length) with a scalpel. As excised seeds exhibited the atypical germination behaviour described by Ikuma and Thimann (1963b) for lettuce (i.e. protrusion of the cotyledons preceded that of the radicle), seeds were considered to have germinated following hypocotyl elongation and cotyledon expansion. In preliminary studies, comparable seed viability values were obtained using this method and the tetrazolium staining method described by R. J. Moore (1972). Data analyses utilized germination values corrected for viability so that site-specific differences in germination were not confounded with viability differences among samples. However, viability differences between seeds utilized from different sites and clutches were not large. Sample viability averaged 98% and ranged from 90 to 100%. 3.2.5 Reversibility of R and FR Effects on Germination In order to test the hypothesis that knapweed seed germination is mediated by phytochrome, seeds were exposed to sequential R (2 min) and FR (10 min) treatments initiated after 8 h of imbibition in darkness at 25 °C. Control seeds were not exposed to light. Germination was recorded 5 days after the addition of water. 42 3.2.6 Effect of R Duration on Germination The objective of this experiment was to determine whether the failure of some primary dormant seeds to germinate following R treatments was because of inadequate exposure durations, or the possession of a deeper form of dormancy. Seeds were exposed to 2 min, 12 h, 1 d, 3 d, and 5 d of R. Seeds from individual plants with low germination in darkness and following exposure to R (plant 1 at site D8, and plant 5 at site S2) were chosen in order to determine if the failure to respond to R resulted from an inadequate duration of exposure. A shortage of seeds restricted the number of seeds per replicate to 29 and 37 (instead of 50) for diffuse and spotted knapweed, respectively. Parafilm-sealed petri dishes containing seeds were exposed to R following 8 h of dark incubation. Following the indicated R exposures, dishes were returned to darkness. Germination counts for all treatments were made upon the completion of the 5 d of R treatment. 3.2.7 Statistical Procedures All experiments (completely randomized design) were repeated at least two times with similar results. Results were analyzed by the analysis of variance and means separated by Fisher's protected LSD test (p<0.05). Percentage values were used in the ANOVA as tests indicated homogeneity of variance. 3.3 Results 3.3.1 Seeds from different sites and clutches Diffuse and spotted knapweed seeds collected from different sites (Tables 3 and 4 ) and from different clutches within selected sites (Table 5) exhibited variable germination behaviour; characterized by significant differences in germination in darkness and following exposure to R. R significantly stimulated germination of diffuse and spotted seeds from all sites (Tables 3 and 4 ) , and from all plants but one spotted knapweed individual (Table 5 ) . In all cases, dormancy was evident because some viable individuals in the seed samples collected 43 Table 3. Germination of Diffuse Knapweed Seeds Collected From Different Sites Germination ( %)a -Site code Dark Redb -D2 6 70 D3 9 65 D5 16 57 D6 23 78 D l 31 83 D7 34 75 D4 36 77 Analysis of variance Source DF Mean square F Site 6 717.6 15.86 ** Light 1 34900.1 771.37 ** Site X light 6 164.2 3.63 ** Error 42 45.2 aValues are the means of 4 replicates of 50 seeds. ^Seeds were exposed to 2 min R after 8 h of imbibition at 25 °C. 44 Table 4. Germination of Spotted Knapweed Seeds Collected From Different Sites Germination (%)a Site code Dark Redb S7 4 35 S3 8 28 SI 11 53 S2 12 56 S8 13 48 S4 14 36 S5 15 47 S9 16 62 S6 34 78 S10 35 83 Analysis of variance Source DF Mean square F Site 9 1561.8 37.37 ** Light 1 26875.4 643.04 ** Site X light 9 210.9 5.05 ** Error 60 41.8 Values are the means of 4 replicates of 50 seeds. 'Seeds were exposed to 2 min R after 8 h of imbibition at 25 °C. 45 Table 5. Comparative Germination of Different Knapweed Seed Clutches Within Sites Site code L a l b 2 3 4 5 6 7 8 9 10 (LSDQ05)<= Diffuse knapweed Dl D R 8 75 8 81 19 91 23 91 28 89 31 95 41 94 52 97 53 97 59 97 (15) D2 D R 1 90 5 93 6 83 6 87 7 65 7 98 14 96 14 99 16 94 16 96 (7) D3 D R 1 71 7 93 8 74 8 74 12 92 15 92 20 88 21 88 23 93 36 100 (9) D4 D R 17 77 20 88 24 95 34 96 39 99 56 100 71 100 87 99 95 100 96 100 (7) D5 D R 3 19 14 57 16 75 18 66 27 96 28 98 33 74 36 90 41 91 78 98 (12) D6 D R 13 92 18 85 19 87 20 91 24 80 29 86 29 93 31 93 57 97 60 100 (9) D7 D R 7 34 25 89 42 89 42 94 42 94 43 91 58 93 61 96 64 97 75 91 (14) D8 D R 4 29 6 82 7 51 10 60 16 80 22 75 25 69 28 85 29 86 38 83 (9) Spotted knapweed SI D R 1 35 1 19 2 16 3 38 3 47 6 75 8 52 8 57 13 50 - (8) , S2 D R 0 44 0 27 1 66 2 25 3 19 4 43 7 87 8 43 9 68 10 56 (10) S4 D R 7 38 8 14 11 62 12 39 14 57 16 51 17 43 '20 57 -- (10) * Light treatment: D = dark, R = 2 minutes R. b Plant number; arranged in order of increasing dark germination Jevels. c Fisher's protected LSD; ANOVA, revealed significant (P<0.05) plant, light and plant X light effects fit all sites. 46 from different sites germinated following R treatment but failed to germinate in darkness, while others did not respond to R treatment (Tables 3 and 4). This variable germination behaviour was also apparent within a clutch (Table 5). In most cases, clutches contained non-dormant (ND) and dormant seeds (consisting of a mixture of 2 minute R-sensitive (RS) primary dormant, and R-insensitive (RI) primary dormant seeds. This indicated that the polymorphic germination behaviour of bulk samples was not solely the result of mixing seeds from plants of different genotypes. The relative proportion of each seed type varied among diffuse and spotted knapweed clutches. Some diffuse knapweed plants produced only 1% ND seeds, while others bore over 90% ND seeds (Table 4). The range of ND seed levels among clutches also differed at different locations. For example, ND seed levels ranged from 1 to 16% at Falkland (D2) compared to 17 to 96% at the Sunnybrae site (D4). Similar differences in germination levels following R were also evident, although the range of the differences within samples from a single site were less than the range of ND seed levels. All spotted knapweed plants produced fewer than 25% non-dormant seeds; some plants produced no ND seeds (Table 5). Spotted knapweed seeds collected from different sites or clutches also exhibited differences in germination levels in darkness and following a 2 min R treatment. 3.3.2 Effect of Sequential R and FR Light Exposures Diffuse and spotted knapweed seeds exposed to sequential R and FR treatments exhibited the classic R/FR reversibility behaviour indicative of phytochrome-mediated germination (Table 6). Exposure to FR negated the germination stimulating effect of a previous R exposure. However, FR did not inhibit the germination of ND seeds. 3.3.3 Effect of Duration of R The results demonstrated that some dormant diffuse and spotted knapweed seeds fail to germinate in response to a R stimulus. These dormant seeds were classified as light-insensitive (LI). Extending the duration of R treatment from 2 minutes to 1 day increased 47 Table 6. Reversibility of R and FR Effects on Knapweed Seed Germination Germination (%) Treatment Diffuse knapweed Spotted knapweed D 5 a D10 SI S10 Dark 23 60 11 48 FR 24 59 15 45 R 81 95 69 83 R-FR 19 49 15 50 R-FR-R 79 91 77 85 R-FR-R-FR 17 59 13 51 L S D(0.05) 13 a Seed collection site code 48 diffuse knapweed germination by 38%, and spotted knapweed germination by 48% (Figure 12). However, increasing R duration further stimulated the germination of few of the remaining dormant seeds. Approximately, 45% of diffuse knapweed and 35% of spotted knapweed seeds in these samples were LI. 3.4 Discussion Three distinct types of germination behaviour were exhibited. All three dormancy categories were evident within individual diffuse and spotted knapweed seed clutches. The key criterion of germination polymorphism, that differences in innate dormancj' exist such that germination of seeds is discontinuously distributed (Popay and Roberts 1970b; Roberts 1972), was exhibited by both diffuse and spotted knapweed in this study. The majority of primary dormant seeds exhibited light-sensitive germination mediated by the phytochrome pigment system (LS seeds). In addition, a lesser number of dormant seeds were insensitive to R light (LI seeds). The remaining seeds exhibited no innate physiological barrier to germination at 25 °C (ND seeds). The LS dormancy evident in knapweed seeds could be considered a form of 'relative' dormancy (rather than true dormancy) because germination was restricted to specific environmental conditions of light and temperature (Karssen 1980/8la). On the other hand, LI dormant knapweed seeds must experience an additional unidentified environmental stimuli in addition to light before germination can occur. Phytochrome mediation is either absent in these seeds or an additional block to germination must be removed before light-sensitivity can be manifested. LI seeds may possess a form of 'true' dormancy, where seeds are unable to germinate regardless of light and temperature conditions (Karssen 1980/8 la). However, further studies employing a wider range of environmental conditions would be necessary to confirm this. Others (Salisbury 1964; Popay and Roberts 1970b) consider such differences in dormancy quantitative in nature and normally distributed. Consequently, under a given set of conditions, certain seeds are capable of germination while others require a further environmental stimulus in order to germinate (Popay and Roberts 1970b). Furthermore, the proportion of dormant diffuse and spotted knapweed seeds varied among 49 7 0 - i Duration of R [days] Figure 12. Effect of the duration of R exposure on the germination of diffuse and spotted knapweed seeds. Values represent the mean of two experiments. See the Appendix, Table 55 for means and S. E. M. 50 clutches and bulk samples. This variability can be considered a characteristic of polymorphic germination behaviour as it reflects the fact that differences in the percentage germination of samples is indicative of differences in the physiological response of seeds to the test conditions. Data presented by Marsden-Jones and Turrill (1954) for the germination behaviour of Centaurea nigra suggests (viability data was not presented) that germination polymorphism, similar in nature to that noted in this study for diffuse and spotted knapweed, is characteristic of other members of the genus as well. In that study, substantial differences in C. nigra germination were evident after 2 days of incubation at 25 °C (probably reflecting dark germination) were apparent among seeds collected from bulk samples (germination ranged from 14 to 47%) and from individual plants grown in Kew (germination ranged from 2 to 53%) [Marsden-Jones and Turrill 1954]. These differences persisted through several subsequent observations over a period of 28 days (during which time seeds were presumably exposed to white light). The seed clutch exhibiting the lowest germination at day 2 (2%) also attained the lowest germination level at day 28 (57%), while seeds with the highest germination initially (53%) also attained the highest final germination level (89%) on day 28. However, this study neither determined whether the differences were significant statistically, or whether ungerminated seeds were dormant or merely non-viable. The results obtained here for diffuse and spotted knapweed seeds demonstrated that variability in germination behaviour among samples from different sites or clutches exist and clearly reflect germination polymorphism present at the time of dispersal/harvest and not differences arising due to variable seed viability or pre-experimental handling of the seeds. Within-clutch variation has been identified as a source of germination polymorphism in a number of species: Rumex crispus (Cavers and Harper 1966; Maun and Cavers 1971a, b); Phleum arenarium (Ernst 1981); and Plantago coronopus (Dowling 1933; Schat 1981). Such germination polymorphism necessitates cautious interpretation of experimental results 51 when bulk seed samples are used as biologically important variability in germination behaviour is obscured (Salisbury 1965; Cavers and Harper 1966). Previous failures to detect a light requirement in knapweed seeds (Watson 1972; Watson and Renney 1974) could conceivably have occurred if seeds used were collected from plants producing no light-sensitive progeny. Polymorphisms for seed size, morphology, and germination within individual capitula are common in the Asteraceae (Harper 1965). Polymorphic germination behaviour improves the survival of weedy species by distributing germination temporally (Popay and Roberts 1970b; Bewley and Black 1982). Polymorphism also ensures the persistence of reserves of ungerminated seeds in the soil (Grime 1979). For example, depth of dormancy is positively associated with seed longevity in the soil (Taylorson 1970; Bostock 1978). The more deeply dormant LI knapweed seeds may aid the formation of soil-borne seed banks. Consequently, populations are not extirpated should control measures, or unfavourable environmental conditions eliminate all vegetative individuals. Environmental conditions following dispersal would determine whether seeds germinate or remain in a dormant or quiescent state. The ND component of diffuse and spotted knapweed seed production is capable of prompt germination under favourable water oxygen and temperature conditions. This is a favourable strategy when subsequent environmental conditions permit completion of the life cycle. Early germination and establishment has been correlated with increased fecundity in the asteraceous weed Lactuca serriola (Marks and Prince 1981). However, where competition for light from existing established plants limits seedling survival, or where winter mortality is severe, ND seed production and immediate germination is a less desirable strategy. The production of dormant (LS and LI) seeds by diffuse and spotted knapweed ensures that all progeny do not germinate immediately under such conditions. Although LS seeds are capable of immediate germination 5 2 in sites exposed to unfiltered sunlight, changes in light quality caused by overlying plants would minimize the P r to Pf r conversion needed for germination of these seeds. Light-sensitive seeds falling beneath chlorophyllous plant canopies would remain dormant until the canopy was removed or senesced. Consequently, germination of LS seeds would be delayed until the overlying (chlorophyllous) cover was eliminated by drought, freezing, grazing, or other disturbance. Depending on the relative sequence of climatic conditions, seeds could germinate in the fall (if senescence of competing plants was followed by later rains) or in spring prior to regrowth of a plant canopy (if dormancy was not induced over the winter). Germination in the autumn is thought to be the best strategy for seedling establishment in grasslands as competition at this time would be minimal due to the senescence or reduced growth of competing species that occurs during periods of summer drought (Chancellor 1982). Field experiments are necessary to establish the role of light in field germination as changing environmental conditions and seed light-sensitivity can affect germination in unexpected ways (Taylorson 1972; Baskin and Baskin 1980; Karssen 1980/8 lb; Pons 1983). Leaf-canopy inhibition of germination is thought to be a factor in the disappearance of Amaranthus patulus plants during succession following site disturbance (Washitani and Saeki 1984). Phung and Popay (1981) found that pasture cover inhibited the germination of several common asteraceous weeds and, therefore, may prevent invasion by these weeds. Dense swards of Agrostis stolonifera strongly inhibited germination of Lactuca serriola; when Agrostis plants were shortly cropped, Lactuca germination was comparable to bare soil (Marks and Prince 1981). Range management practices could conceivably modify knapweed germination behaviour. Scheduling grazing so that plant cover was not removed during periods favourable to knapweed germination would inhibit the germination of LS seeds. However, this practice per se might not have a substantial impact on knapweed populations because knapweed seed 53 germination is polymorphic; some ND seeds would still germinate. Furthermore, unless plant cover could be established before conditions became favourable for knapweed germination in the spring, germination of the remaining LS seeds would only be delayed until spring. While increasing the time knapweed seeds spend in the dormant condition would probably increase mortality and decrease seeding vigour (see Chicoine 1984), it is unlikely to have a significant impact on knapweed seedling populations in light of the prolific seed production of these species. However, germination is also contingent upon the temperatures experienced by the dormant seed. If seeds were dispersed, buried and moistened earlier in the season when soil temperatures were relatively high, thermodormancy induction (see section 6.2.2) could prevent germination in the spring. Non-dormant seeds buried at the same time would be expected to germinate completely in response to rising soil temperatures in the spring, if low temperature or dry soil conditions prevented their germination in the fall. Seeds dispersed and buried earlier in the season, and subsequently experiencing moisture and temperature conditions conducive to germination, would germinate in situ. However, if the same seeds experienced high moisture and soil temperature conditions (in excess of 20 °C), a number of them would be induced into thermodormancy (see section 6.2). A number of these seeds, determined by the depth of dormancy induced and the relative stimulatory strength of stratification and rising temperatures in the spring, could germinate in the spring. Light-sensitive seeds entering the seed-bank in late summer or fall should remain dormant until either exposed to light by soil disturbance, or in the case of the least dormant of them, stimulated by rising temperatures in the spring. However, the probability of them germinating in this latter case would be influenced by the relative influence of conditions conducive to thermodormancy induction (see 6.0) and dormancy loss through after-ripening (see 5.0) experienced during burial. 54 4.0 E F F E C T O F L I G H T Q U A L I T Y D U R I N G S E E D M A T U R A T I O N 4.1 B a c k g r o u n d The factors responsible for polymorphic germination behaviour are poorly understood. Both genetic and environmental influences appear to contribute to germination polymorphism. Polymorphism is generally associated with the maternal parent in the Asteraceae (Harper 1965). However, Globerson et al. (1974) concluded that, although factors determining seed dormancy in Lactuca sativa (Grand Rapids) were inherited, such traits were not maternal in origin. Eenink (1981) reported that dormancy in Lactuca sativa was associated with a single gene. However, caution is warranted prior to attributing germination polymorphism to genetic factors on the basis of comparative germination studies as the influence of environmental factors during seed maturation and after-ripening can markedly affect seed germination behaviour (see Baskin and Baskin 1973b). Seed morphology and germination behaviour differences in the Asteraceae are associated with position and relative timing of development within the capitulum. Seeds arising from ray florets in Bidens bipinnata (Dakshini and Aggarwal 1974), Bidens pilosa (Forsyth and Brown 1982), Grindelia squarrosa (McDonough 1975), Heterotheca subaxillaris (Baskin and Baskin 1976a; Awang and Monaco 1978), and Heterotheca grandiflora (Flint and Palmblad 1978) exhibited more dormancy than seeds arising from disc florets. Similarly, Xanthium pennsylvanicum produces two-seeded burs containing an upper, deeply dormant seed and a lower, shallowly dormant seed (Esashi and Leopold 1968). In these instances, morphological differences often distinguished the two seed morphs. Colour was the only obvious difference within clutches of diffuse and spotted knapweed seeds, however, germination behaviour did not appear to be consistently related to colour. Association of morphological and positional characteristics with germination behaviour has been reported in other families as well. Germination polymorphism within clutches of Chenopodium album seeds was associated with differences in seed coat colour and surface 55 texture (Williams and Harper 1965). Rumex seeds from the upper half of panicles were heavier and more dormant than seeds from the lower half of the panicle (Cavers and Harper 1966). Environmental conditions prevalent during seed maturation also affect germination behaviour (Roller 1962b). Soil fertilization improved germination of Lactuca sativa seeds (Thompson 1937). Dormancy is also heightened when soil moisture levels are high. Dormancy in Arenaria patula seeds was higher in wetter years (Baskin and Baskin 1975) and Avena fatua seeds were more dormant when plants were grown under high soil moisture conditions (Sexsmith 1969). Daylength affects the germination of many species. Seeds of Lactuca scariola (Gutterman et al. 1975), Chenopodium album (Karssen 1970), Chenopodium polyspermum (Pourrat and Jacques 1975), and Portulaca oleraceae (Gutterman 1974) maturing under short day conditions germinated to higher levels than seeds matured under long days. In many cases, day length affected such seed coat characteristics as thickness, colour, and water permeability (Jacques 1957, 1968; Cumming 1959; Karssen 1970; Pourrat and Jacques 1975; Gutterman 1978). Seed germination characteristics can be influenced by the light quality experienced by the mother plant during seed maturation. Arabidopsis thaliana plants grown under light rich in FR wavelengths produce more light-requiring dormant seeds than plants reared in a FR poor light environment (McCullough and Shropshire 1970; Hayes and Klein 1974). Similarly, seeds of Cucumis prophetarum, and C. sativus contained higher proportions of Pf r, and greater dark germination, when harvested fruits were exposed to R rather than FR (Gutterman and Porath 1975). However, this effect was lost if seeds were dried prior to conducting germination tests. Evidence suggested that light quality effects on germination of Arabidopsis thaliana seeds were localized to the floral stalk region and were not translocatable from the rest of the plant (Hayes and Klein 1974). Isolated floral buds irradiated with with fluorescent light exhibited about 95% dark germination while seeds from the rest of the plant (irradiated 56 with incandescent light) failed to germinate in darkness. Cauline leaves did not appear to be involved in the mediation of the effect. The effect of light was evident up to one day prior to seed maturation when seed moisture was 50% on a dry weight basis (Hayes and Klein 1974). Kendrick and Spruit (1977) proposed that the spectral quality of light falling on a developing embryo, and possibly the rate of dark reversion during seed dehydration, affects the amount of Pf r in the mature seed. In turn, the spectral composition of light falling on a developing embryo may be influenced by the tissues investing the seed, which in turn may influence the germination behaviour of the mature seed (Cresswell and Grime 1981). Cresswell and Grime (1981) found that species retaining chlorophyll in tissue surrounding seeds during seed maturation produced seeds with high levels of dormancy, while species losing chlorophyll prior to maturation exhibited minimal dormancy. The authors hypothesized that seed phytochrome can be arrested in either the Pf r or P r form, depending upon the R/FR of light intercepted during the critical point in the maturation process where seed moisture content drops below the level permitting phytochrome transformations. Similarly, intra-capsular germination polymorphism could arise if chlorophyll loss in the investing maternal tissue, moisture loss from seeds, or both, varied within developing inflorescences (Cresswell and Grime 1981). The concept does not conflict with reports that dormancy expression appears to develop during the latter stages of seed maturation in Lactuca sativa (Gutterman 1973; Globerson 1981) and Arabidopsis thaliana (Hayes and Klein 1974). In a similar fashion, germination polymorphism in knapweed may arise from variability in the light quality experienced by developing seeds. Chlorophyll loss from phyllaries, while seeds are sufficiently hydrated to permit phytochrome photoconversion, could enable conversion of phytochrome to Pf r and, therefore, production of non-dormant seeds. If chlorophyll was retained throughout seed maturation, attenuation of red wavelengths would favour phytochrome conversion to P and dormant seed production. 57 Floret and capitula development in knapweed occurs asynchronously. Consequently, seeds could mature under different light qualities if phyllary chlorophyll content varied over time. Such a scenario is quite possible because water stress, which undoubtedly changes with time in the typical knapweed habitat, influences chlorophyll retention in plant tissue (Alberte and Thornber 1977). Water stress effects were proposed to be the mechanism by which long photoperiods produce higher germination levels than short photoperiods in Lactuca sativa (Roller 1962b). Rnapweed phyllaries appear particularly prone to chlorophyll loss as totally senesced capitula are often borne by otherwise chlorophyllous plants (Nolan, personal observation). The hypothesis that higher levels of phytochrome-mediated primary dormant spotted knapweed seeds are produced when seeds are exposed to light rich in FR wavelengths during maturation, was tested in the following experiment. 4.2 M a t e r i a l s a n d M e t h o d s On 14 August 1985, capitula of spotted knapweed plants growing near Westwold, B. C. were enveloped with filters of either Roscolux medium blue green filter #95 combined with Roscolene orange #818 (FR filter), Roscolene orange #818 alone (R filter), aluminum foil, mesh, or clear plastic. R and FR filters provided contrasting R- and FR-rich light environments. Aluminum-foil, nylon mesh and clear plastic filter treatments acted as controls in case factors other than light (e.g. temperature or relative humidity) were important in dormancy determination. Capitula with moist, recently wilted corollas were selected to ensure pollination had occurred, while minimizing pre-treatment seed maturation processes. Phyllaries were chlorophyllous at this time. Filter types were assigned in a completely randomized manner to capitula as they were located in the stand. Two weeks later capitula were cut from the parent plants with the filters in place. At this time, capitula in the mesh and clear treatments were observed to be senesced. Three days later filters were removed and the seeds were extracted from the capitula under green 58 light. Seeds from the different capitula in each treatment were pooled and stored at -20 C for 3 months prior to experimentation while other experiments were being conducted. Germination in darkness and following 2 min R treatments was examined in separate experiments. Each treatment utilized 5 replicates of 30 seeds. Data were analyzed by the analysis of variance and treatments compared with orthogonal contrasts. The experiment was conducted once. Seeds maturing in capitula covered with far-red or aluminum foil filters (as phytochrome is formed in the P r form, exclusion of light would prevent photoconversion) were expected to exhibit greater levels of dormancy than seeds maturing beneath mesh, red, or clear filters. The mesh and clear plastic filter treatments were expected to produce seeds of similar germination behaviour if light quality was the key factor determining dormancy expression. However, similar levels of dormancy in all treatments would result if chlorophyll was retained throughout seed maturation in all treatments. In this event, the P r form of phytochrome would be expected to predominate because of R attenuation by chlorophyll, and, in the dark treatment, by the exclusion of light. 4.3 Results and Discussion Placement of filters over the capitula during seed maturation significantly affected seed germination in darkness and following exposure to 2 min R (Table 7). However, treatment effects were not consistent with the hypothesis that the light quality incident during seed maturation determines seed dormancy characteristics through its effect on the phytochrome pigment system. Dormancy, as measured by the percentages of seeds failing to germinate in darkness and following R treatment, was lowest in the treatments expected to yield the highest levels of dormant seeds (i.e. FR and aluminum foil). Treatments expected to favour phytochrome photoconversion to Pf r (R, clear, mesh) yielded more dormant seeds. The mesh filter treatment produced the most dormant seeds in terms of germination in both darkness and following 2 min R. 59 Table 7. Mean Germination of Seeds Matured in Capitula Surrounded by Various Light Filters Germination (%) Filter type Dark Reda Dark (foil) 71 95 Far-red 76 98 Red 65 97 Clear 47 90 Mesh 26 80 Analysis of variance and orthogonal contrasts of filter data Source D.F. Mean Square Significance'3 Germination in darkness Filters 4 2099 =M= D,FR vs R,C,M 1 4571 ** D vs FR 1 71 NS R vs C,M 1 2617 * * C vs M 1 1136 * * Error 20 63 Germination following 2 : min R Filters 4 280 * * D,FR vs R,C,M 1 378 ** D vs FR 1 17 NS R vs C,M 1 477 * * C vs M 1 250 * Error 20 35 a Seeds were exposed to 2 min R following 8 h imbibition at 25 °C. k NS = non-significant; * = significant (p<0.05); ** = significant (p<0.01) 60 The significant differences in both dark and R germination levels between the mesh and clear plastic treatments indicated that light quality was not the sole factor determining dormancy expression in spotted knapweed. Temperature and RH levels were likely influenced by filter placement over the capitula and may have influenced dormancy expression. In retrospect, aluminum foil and FR filters were sealed more carefully than the other treatments to exclude unfiltered sunlight. This lack of uniformity would have created differences in RH, and perhaps temperature, between capitula in different treatments. Gutterman (1974) found that light quality during Portulaca oleraceae seed maturation affected germinability. Germination in continuous white light at sub-optimal temperature was generally higher when mother plants were exposed to FR. However, Gutterman concluded that light quality does not affect germination behaviour through the phytochrome system sensu stricto as dark germination was unaffected. Seed dormancy is influenced by ambient temperature during seed formation in Lactuca sativa (Harrington and Thompson 1952; Roller 1962b; Eenink 1977) and Rosa spp. (Von Abrams and Hand 1956). However, temperature and RH were not monitored in this study examining the effects of light quality on knapweed seed dormancy. Consequently, although it is possible to conclude that environmental conditions during seed maturation affect the germination behaviour of spotted knapweed, the factor(s) responsible can not be identified on the basis of this experiment. Although the results did not support the hypothesis that seed dormancy is the result of light quality interactions with seed phytochrome, the confounding of light,temperature, and RH effects could mask the effects of light quality. More stringently controlled experiments are necessary to clarify the role light quality incident upon developing seeds has on knapweed germination behaviour. 61 5.0 E F F E C T OF AFTER-RIPENING ON GERMINATION BEHAVIOUR 5.1 Background Grime et al. (1981) reported that storage increased germination in 8 of 12 asteraceous species that initially exhibited less than 50% germination. This phenomenon, termed after-ripening, is characterized by a progressive loss of dormancy resulting from changes in seed physiology occurring during aging of low water content seeds. Although the physiological basis of the process is unknown, metabolic involvement is suggested by the positive relationship between the rate of after-ripening and temperature or oxygen levels (Roberts and Smith 1977; Bewley and Black 1982). After-ripening may occur under field conditions when seeds are in a dry state for prolonged periods (Bewley and Black 1982). Temperature, ambient oxygen level, and seed moisture content influence after-ripening (Bewley and Black 1982). Because field temperature and seed moisture content are more variable than oxygen levels, and because the moisture content of mature seeds is influenced by the ambient relative humidity, the experiments in this section examined the role of relative humidity and temperature on dry after-ripening in diffuse and spotted knapweed. Experiments were conducted in order to determine : 1) if dry after-ripening leads to a progressive loss of seed dormancy in knapweed, 2) if the effect of after-ripening on seed germination behaviour is similar in diffuse and spotted knapweed, and 3) if relative humidity and temperature influence the after-ripening process. 5.2 Materials and Methods 5.2.1 Effect of Aging on Germination Behaviour Seeds were selected from diffuse and spotted knapweed bulk collections exhibiting high levels of primary dormancy: samples D3 and S7 (see Table 3). Dry seeds placed in petri dishes were sealed in metal coffee tins and stored in darkness at -20, 3, and 25 °C. Following 0, 30, 60, and 120 d of after-ripening, 5.0 ml water was added. In some experiments a 62 shortage of seeds necessitated dropping the 30 d after-ripening treatment. Germination in darkness and following 2 min or 1 d R treatments was determined after 5 d at 25 °C. Characteristics of germination behaviour examined were the level of ND seeds (ND = % germination in darkness); the level of \"light-insensitive\" (LI) seeds (LI = 100 - % germination following Id R treatment); and the level of seeds \"insensitive to 2 min R treatment\" (IR seeds; IR = 100 - % germination following 2 min R) was determined in order to detect subtler changes in light sensitivity during after-ripening. Viability was routinely determined for all samples. It should be noted that changes in methodology of the after-ripening experiments may reduced experimental error. The effects of the different after-ripening durations were were probably confounded with chance variability in incubation conditions arising from the sequential nature of the germination tests; this was especially noticeable in experiments examining IR seed percentages during storage at 3 and -20 °C. Although there was no significant effect of after-ripening duration on IR percentages at these temperatures, quite large differences in final germination values were evident among different after-ripening durations. This source of error could be eliminated if conditions arresting after-ripening were determined first (e.g. storage at -20 °C for knapweed). Then the experiment could be designed so that the initiation of after-ripening was staggered (by removing seeds from cold storage at different times) in such a manner that seeds in all treatments were imbibed at the same time for germination determinations. Separate experiments were conducted for each temperature (i.e. 25, 3, -20 °C) and light (i.e. dark, 2 min R and 1 d R) treatment combination. Each experiment was conducted at least twice with similar results. Germination values were adjusted for differences in viability. 63 5.2.2 Effect of Relative Humidity on After-ripening Saturated solutions of potassium acetate, magnesium chloride, manganese chloride, sodium chloride, and potassium nitrate were used to produce relative humidity (RH) levels of 22.0, 32.8, 53.9, 75.6, and 90.7%, respectively, at 25 °C (Hall 1957). Replicate samples of seeds placed in open plastic vials were sealed in plastic containers (Frig-O-Seal 2.25 L) containing 200 mL of salt solution. In addition, seeds in petri dishes were stored at ambient RH in metal tins at -20 and 25 °C to simulate conditions used for seed storage in the freezer and in the previous after-ripening experiment (see section 5.2.1). These additional treatments were not included in the ANOVA. Following 30 d storage in darkness at 25 °C, seeds were transferred from vials to petri dishes; water was then added, and germination in darkness and following 2 min or 1 d R treatments (initiated 8 h after addition of water) was determined 5 d later. Seed viability determinations revealed a partial loss of viability following storage at 90% RH for 30 d. Consequently, an analysis of variance was done on seed viability data. Hollow seeds were not included in the non-viable seed totals to eliminate any error arising from variable numbers of such seeds in the treatments. 5.3 Results 5.3.1 Effect of Aging on Germination Behaviour 5.3.1.1 Level of ND seeds The duration of after-ripening at 25 °C had a significant (p<0.01) effect on ND seed levels for diffuse and spotted knapweed (Table 8). There was a significant positive, linear relationship between duration of after-ripening and ND seed percentages for both species. These results indicated that after-ripening at 25 °C caused a progressive loss of dormancy in the seed samples of diffuse and spotted knapweed used in this experiment. Species and species X time interaction effects were not significant. Therefore, the number of ND seeds, as 64 Table 8. Effect of After-ripening at 25 °C on Percentage of ND Seeds Mean germination (%) Time (d) Species 0 30 60 120 Diffusea 9.3 27.3 43.3 80.7 Spotted13 3.3 30.0 38.7 76.0 Analysis of variance Sources DF Mean square F Time 3 5448.61 173.97 ** Species 1 60.17 1.91 NS Time X species 3 23.28 0.74 NS Error 16 31.50 a y = 9.07 + 0.59x, r 2 = 0.96 ** b y = 6.53 + 0.58x, r 2 = 0.96 ** 65 well as the rate of dormancy loss attributable to after-ripening, were the same in the diffuse and spotted knapweed samples used in this experiment. There was no significant change in ND seed levels in diffuse or spotted knapweed seeds during 120 d of dry storage at either 3 or -20 °C (Tables 9 and 10). This indicated that the dormancy loss associated with dry after-ripening was temperature dependent, and arrested or retarded by low temperatures. 5.3.1.2 Level of LI seeds The duration of after-ripening at 25 °C had a significant (p<0.01) effect on levels of LI seeds in diffuse and spotted knapweed (Table 11). A significant negative, curvilinear relationship existed between the duration of after-ripening and LI seed numbers in both species. The decline in LI seed numbers indicated that after-ripening was responsible for dormancy loss in this class of seeds. The significant species effect reflected the lesser number of LI seeds in the diffuse knapweed sample. The significant species X time interaction effects did not necessarily indicate that spotted knapweed LI seeds became less dormant more rapidly because declines in the numbers of LI seeds were essentially completed by the first observation period (30 d). Therefore, the significant interaction effect arises primarily from the 15% difference in initial levels of LI seeds in the two samples. Further examination of LI seed germination behaviour in the 0 to 30 d interval of after-ripening would be required to clarify the comparative rates of dormancy loss in LI diffuse and spotted knapweed seeds. There was no significant (p<0.01) change in the percentage of LI diffuse or spotted knapweed seeds during 120 d of dry storage at either 3 or -20 °C (Tables 12 and 13). Again, the significant species effect reflects the initial differences in LI seed numbers in the two samples. 66 Table 9. Effect of After-ripening at 3 °C on Percentage of ND Seeds Mean germination (%) Time (d) Species 0 30 60 120 Diffuse 3.3 2.0 2.0 1.3 Spotted 2.7 4.7 4.7 1.3 Analysis of variance Sources DF Mean square F Time 3 5.50 1.22 NS Species 1 8.17 1.81 NS Time X species 3 4.61 1.02 NS Error 16 4.50 67 Table 10. Effect of After-ripening at -20 °C on Percentage of ND Seeds Mean germination (%) Time (d) Species 0 30 60 120 Diffuse 4.7 1.3 2.7 2.7 Spotted 5.3 2.7 2.7 4.7 Analysis of variance Sources DF Mean square Time 3 10.22 1.61 NS Species 1 6.00 0.95 NS Time X species 3 1.11 0.17 NS Error 16 6.33 68 Table 11. Effect of After-ripening at 25 °C on Percentages of LI Seeds Mean germination (%) Time (d) Species 0 30 60 120 Diffusea 14.7 0.7 2.0 1.3 Spotted13 30.0 6.0 2.7 3.3 Analysis of variance Sources DF Mean square F Time 3 581.50 59.14 ** Species 1 204.17 20.76 ** Time X species 3 65.94 6.71 ** Error 16 9.83 a y = 13.32 - 0.37x + 0.002x2, r 2 = 0.89 ** y = 28.51 - 0.75x + 0.005x2, r 2 = 0.87 ** 69 Table 12. Effect of After-ripening at 3 °C on LI Seed Percentages Mean germination (%) Time (d) Species 0 60 120 Diffuse 13.3 12.7 12.7 Spotted 30.0 15.3 25.3 Analysis of variance Sources D F Mean square F Time 2 90.89 1.67 NS Species 1 512.00 9.40 ** Time X species 2 78.00 1.43 NS Error 12 54.44 70 Table 13. Effect of After-ripening at -20 °C on LI Seed Percentages Mean germination (%) Time (d) Species 0 60 120 Diffuse 11.3 9.3 6.0 Spotted 30.0 31.3 41.3 Analysis of variance Sources DF Mean square Time 2 20.22 0.61 NS Species 1 2888.00 87.81 ** Time X species 2 116.67 3.55 NS Error 12 32.89 71 5.3.1.3 Level of IR seeds After-ripening at 25 °C significantly (p<0.01) affected the number of IR seeds(i.e those failing to germinate in response to a 2 min R exposure) in both diffuse and spotted knapweed (Table 14). There was a significant (p<0.01) negative linear relationship between the duration of after-ripening and IR seed percentage. The spotted knapweed sample had significantly greater numbers of IR seeds. However, the time X species effect was not significant. Therefore, rates of decline in IR numbers were similar for the diffuse and spotted knapweed samples used in this study. No significant changes in the percentage of IR seeds occurred when seeds were stored at 3 or -20 °C (Tables 15 and 16). 5.3.2 Effect of Relative Humidity on After-ripening 5.3.2.1. ND seed levels After-ripening in diffuse and spotted knapweed was a RH-dependent process. Dark germination (ND seed numbers) of diffuse and spotted knapweed seeds following 30 d at 25 °C was significantly (p<0.01) affected by the RH of the air surrounding seeds during the after-ripening period (Tables 17 and 18). Both species exhibited the greatest increase in dark germination (relative to \"unafter-ripened\" seeds stored at -20 °C) in the 32.8% RH treatment: 86% versus 39% and 47% versus 13% in diffuse and spotted knapweed, respectively (Tables 17 and 18). Dark germination of diffuse knapweed also increased significantly from 39% to 59% when seeds were after-ripened at 22.0% RH, but this increase was significantly less than the increase attained at 32.8% RH. However, in the second run of this experiment (data not shown) there was no significant increase in dark germination at 22.0% RH. Although dark germination was always lowest in the 90.7% RH treatment, the difference was only significant from the control in diffuse knapweed seeds (Table 17). Germination was not significantly different from the control in the remaining treatments. 72 Table 14. Effect of After-ripening at 25 °C on Percentages of IR Seed Mean germination (%) Time (d) Species 0 30 60 120 Diffusea 60.0 44.0 37.3 3.3 Spottedb 76.0 47.3 46.0 17.3 Analysis of variance Sources DF Mean square F Time 3 3381.94 63.02 ** Species 1 661.50 12.33 ** Time X species 3 48.61 0.91 NS Error 16 53.67 a y = 60.4 - 0.51x, r 2 = 0.86 ** b y = 70.3 . 0.50x, r 2 = 0.84 ** Table 15. Effect of After-ripening at 3 °C on Percentages of IR seeds 73 Mean germination (%) Time (d) Species 0 30 60 120 Diffuse 64.7 61.3 74.0 70.0 Spotted 76.7 77.3 68.7 77.3 Analysis of variance Sources DF Mean square F Time 3 19.72 0.23 NS Species 1 337.50 3.88 NS Time X species 3 128.61 1.48 NS Error 16 87.00 74 Table 16. Effect of After-ripening at -20 °C on Percentages of IR Seeds Mean germination (%) Time (d) Species 0 30 60 120 Diffuse 66.0 72.7 71.3 69.3 Spotted 78.0 86.0 75.3 77.3 Analysis of variance Sources DF Mean square F Time 3 64.67 1.64 NS Species 1 522.67 13.29 ** Time X species 3 26.67 0.68 NS Error 16 39.33 75 Table 17. Effect of RH During a 30 d After-ripening Period on Dark Germination of Diffuse Knapweed RH (%) Controla 22.0 32.8 53.9 75.6 90.7 Germination (%) 39.0 58.0 86.0 43.3 34.0 13.3 L S D 0.01 11.4 Analysis of variance Sources DF Mean square F RH 5 1790.79 85.17 * * Error 12 21.03 aControl treatments were stored at -20 C. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 80 _+ 2 % (S.E) dark germination. 76 Table 18. Effect of RH During a 30 d After-ripening Period on Dark Germination of Spotted Knapweed RH (%) Control21 2.0 32.8 53.9 75.6 90.7 Germination (%) 12.5 15.3 46.9 16.5 11.0 9.2 L S D 0 Q 1 14.9 Analysis of variance Sources DF Mean square F RH 5 598.10 16.80 ** Error 12 35.59 aControl seeds were stored at -20 °C. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 35 _+ 3 % (S.E.) dark germination. 77 Table 19. Effect of RH During a 30 d After-ripening Period on the Percentage of Non-viable Diffuse Knapweed Seeds RH (%) Control 22.0 32.8 53.9 75.6 90.7 Non-viable seeds (%) 0 0 0 0 0 6.1 Analysis of variance Sources DF Mean square F RH 4 22.32 9.30 ** Error 10 2.40 78 Table 20. Effect of RH During a 30 d After-ripening Period on the Percentage of Non-viable Spotted Knapweed Seeds RH (%) Control 22.0 32.8 53.9 75.6 90.7 Non-viable seeds (%) 0 0 0 0 1.4 22.3 L S D 0 < 0 1 7.2 Analysis of variance Sources DF Mean square F RH Error 5 12 242.75 8.28 29.31 ** 79 Some of the decline in dark germination (6%) was attributable to the significant loss of seed viability in the 90.7% RH treatment (Table 19). An even greater loss of viability occurred in spotted knapweed seeds at the same RH (Table 20). No significant losses of viability occurred in any other treatment. Seeds in the 90.7% RH treatment were also the only ones to develop a high incidence of fungal growth on the seed coat during imbibition in water. In most cases, seeds supporting fungal growth germinated in a normal manner. 5.3.2.2 IR seed levels RH during the 30 d after-ripening period significantly (p<0.01) affected the percentage of both diffuse and spotted knapweed seeds responding to a 2 min R treatment (Tables 21 and 22). Again, the greatest loss of dormancy occurred in the 32.8 % RH treatment in both diffuse (91 versus 99%) and spotted (30 versus 64%) knapweed. However, this increase was only significant (p<0.01) in spotted knapweed. Germination of seeds stored at 90.7 RH was lower than control seeds stored at -20 °C in both diffuse (60 versus 91%) and spotted knapweed (23 versus 30%), but was not significant (p<0.01) in spotted knapweed. Diffuse and spotted knapweed viability declined 9 and 32% in the 90.7% RH treatment. 5.3.2.3. LI seed levels Although RH had a significant effect on LI germination levels, the only significant difference between the control and treatments was the suppression of germination in the 90.7% RH treatment in both species (Tables 23 and 24). However, seed viability declines in these treatments could account for the differences: 8% of the 12% decline in diffuse knapweed and all of the 37% decline in spotted knapweed germination. f 5.4 Discussion 5.4.1 Conformity of Results With Previous Reports The experiments in this chapter demonstrated that the germination behaviour of diffuse and spotted knapweed seeds was influenced by after-ripening. The dormancy loss associated with dry after-ripening in diffuse and spotted knapweed seeds was temperature 80 Table 21. Effect of RH During a 30 d After-ripening Period on Germination of Diffuse Knapweed Following a 2 Min R Treatment RH (%) Controla 22.0 32.8 53.9 75.6 90.7 Germination (%)a 90.8 93.2 98.7 98.6 97.2 60.4 L S D 0 .01 10.2 Non-viable (%)b 0.7 1.3 0 0 0.7 8.7 Analysis of variance Sources DF Mean square F RH 4 772.40 57.93 * * Error 10 13.33 a Control treatments were stored at -20 °C. b Non-viable seed data is included in the table for comparison purposes only; this data was not included in the ANOVA. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 99 +_ 1 % (S.E.) germination. 81 Table 22. Effect of RH During a 30 d After-ripening Period on Germination of Spotted Knapweed Following a 2 Min R Treatment RH (%) Controla 22.0 32.8 53.9 75.6 90.7 Germination (%) 29.9 47.3 63.7 54.7 40.1 23.3 L S D 0.01 20.5 Non-viable (%)b 1.3 0.7 0 0.7 2.0 32.0 Analysis of variance Sources DF Mean square F RH 4 822.67 13.30 ** Error 10 61.87 aControl seeds were stored at -20 °C. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 74 _+ 9% (S.E.) germination. Non-viable seed percentages are included for comparison purposes only and were not included in the ANOVA. 82 Table 23. Effect of RH During a 30 d After-ripening Period on Germination of Diffuse Knapweed Following a 1 d R Treatment RH (%) Controla 22.0 32.8 53.9 75.6 90.7 Germination (%) 97.2 94.7 100 8.0 98.0 85.1 L S D o . o i b - 3.9 Non-viable (%) 0 0.7 0 0 1.3 8.3 Analysis of variance Sources DF Mean square F RH 5 87.17 35.90 ** Error 12 2.43 aControl seeds were stored at -20 °C. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 97 _+ 3 % (S.E.) germination. ^Non-viable seeds are included for comparison purposes only and were not included in the ANOVA. 83 Table 24. Effect of RH During a 30 d After-ripening Period on Germination of Spotted Knapweed Following a 1 d R Treatment RH (%) Controla 22.0 32.8 53.9 75.6 90.7 Germination (%) 72.0 77.6 86.4 91.3 72.6 34.3 L S D 0 .01 25.6 Non-viable(%)b 0.3 2.0 2.7 0.6 2.7 42.0 Analysis of variance Sources DF Mean square F RH 5 1208.20 11.43 Error 12 105.71 aControl seeds were stored at -20 °C. Seeds after-ripened under ambient RH in tins (i.e. no control of RH) exhibited 65 _+ 3 % (S.E.) germination. bNon-viable seed percentages are included for comparison purposes only and were not included in the ANOVA. 84 dependent, being arrested or retarded by low temperatures as reported for other species (e.g. Stokes 1965; Roberts and Smith 1977). Germination of diffuse and spotted knapweed in darkness and following 2 min or 1 d R treatments became progressively less restricted by dormancy as seeds after-ripened at 25 °C. The changes in germination behaviour observed in this study were similar in nature to those previously reported for Lactuca sativa; germination became less restricted by a requirement for light as the duration of after-ripening at 18 or 23 °C increased (Suzuki et al. 1980; Suzuki 1981; Bewley and Black 1982). The decline in LI and IR seed numbers in diffuse and spotted knapweed is similar to the increased light sensitivity that accompanied dormancy loss in Lactuca sativa (Suzuki 1980 et al.; Suzuki 1981). Similarly, seeds of Lactuca sativa cv. Grand Rapids required a period of after-ripening before germination in response to light, temperature, and growth regulator type compounds became evident (Globerson et al. 1973). Rates of after-ripening apparent in ND and IR seed number determinations were similar for the diffuse and spotted knapweed samples used in these experiments. However, the RH of the air surrounding the seeds during storage affected the rate of after-ripening. Supra- or sub-optimal RH levels arrested or slowed dormancy loss and, at 90.7% RH, increased dormancy levels evident in darkness or following 2 min R exposures. These findings concur with previous reports that adequate moisture content favours after-ripening but further such increases may deter the process due to secondary dormancy induction and viability loss (Toole 1950; Quail and Carter 1969; Baskin and Baskin 1979a; Bewley and Black 1982). Although the'optimum RH range for after-ripening in Draba verna differs from that of knapweed, the same qualitative trends are apparent: after-ripening is inhibited by low RH, and exposure to supra-optimal RH stimulates rotting of seeds (Baskin and Baskin 1979a). The viability loss in knapweed seeds stored under high RH conditions is common in many species, especially at high temperatures, whereas, low RH prolongs seed viability 85 (Toole 1950). Fully imbibed seeds do not experience such a pronounced viability loss because the disruption of enzymes and membranes common in partially hydrated seeds (Villiers 1972) are counteracted by more active cellular repair mechanisms (Villiers and Edgcumbe 1975). Further studies utilizing greater storage durations and closer increments of RH are required to determine 1) the optimum RH for after-ripening and 2) whether after-ripening is arrested, or merely slowed as RH deviates from the optimum RH level. Further investigation of possible interactions of light, RH, and temperature during after-ripening could clarify the ecological significance of this phenomenon. 5.4.2 Proposed Model of Dormancy Transition in Dry Knapweed Seeds The increased percentage of ND seeds in diffuse and spotted knapweed samples stored dry at 25 °C clearly indicated that some dormant individuals in the sample became non-dormant. Although recruitment directly from the LI fraction could have accounted for some of the increase, declines in the number of LI seeds were essentially complete following 30 d of after-ripening (Table 11) while ND seed numbers continued to increase linearly for 120 d (Table 8). This indicated that ND seeds were recruited from the LS pool. This was consistent with the finding that IR seed numbers also declined linearly with time (Table 14). As the IR class consisted of both LI seeds and some LS seeds, and as declines in IR numbers continued beyond 30 days, some LS seeds clearly became less dormant as they after-ripened. Based on these findings, I propose that ND, LI, and LS seeds reflect different phases of germination behaviour exhibited by individual after-ripening knapweed seeds. The proposed sequence of dormancy transition in knapweed seeds is: LI, to LS, to ND (i.e highest to lowest relative dormancy; although a direct transition from LI to ND can not be ruled out). This is not to say that all seeds are initially LI, merely that, whatever the dormancy type upon maturation, the process of after-ripening can lead to a change in dormancy expression. These dormancy classes are probably not unique in terms of morphological or physiological characteristics. Instead, ND, LS, and LI seeds probably reflect the arbitrary experimental 86 criteria imposed upon a population of individuals whose differences in all likelihood form a continuum, a normal distribution from least dormant to most dormant. The physiological processes induced in seeds by after-ripening enable seeds to germinate under conditions of light and temperature previously conducive to dormancy expression. This model conforms to the accepted concept that after-ripening results in a gradual widening of conditions under which germination occurs (Vegis 1963; Bewley and Black 1982). For example, more Senecio vulgaris seeds germinated at sub- and supra-optimal temperatures following 10 weeks dry storage at 35 ° C (Popay and Roberts 1970b), while Suzuki (1981) determined that Lactuca sativa seeds underwent a physiological transition from dormant to non-dormant and finally to a deteriorating state as the seeds age. The diffuse and spotted knapweed seeds used in this study underwent changes in behaviour consistent with this dormant to non-dormant transition. 5.4.3 After-ripening: A Source of Germination Polymorphism in Diffuse and Spotted Knapweed? After-ripening has been proposed to be a mechanism responsible for the seasonal variations in germination behaviour evident in seeds collected before dispersal. For example, in the Asteraceae, variable germination behaviour in Senecio vulgaris seeds collected in different months m a y arise from climatic influences on the seeds during ripening (Popay and Roberts 1970b). Seeds produced in early spring exhibited 70% dormancy in light, whereas, seeds produced in the summer exhibited little dormancy; possibly because higher temperatures accelerated dormancy loss through after-ripening (Popay and Roberts 1970b). Simi lar ly , Harrington and Thompson (1952) found a positive correlation between Lactuca sativa germination at 26 ° C and the average mean temperature experienced in the 30 day period prior to seed harvest, while Thompson (1937) reported decreasing dormancy levels in Lactuca sativa seeds as the date of harvest progressed. Comparable results have been reported in other families as well. A positive linear relationship between germination in light 87 or darkness and the temperature experienced during seed maturation was noted in the grass Dactylis glomerata (Probert et al. 1985). Similarly, seeds of Silene dioica (Thompson 1975), Hyacinthoides non-scripta (Thompson and Cox 1978), and Milium effusum (Thompson 1980) exhibited less dormancy when collected from plants in warmer regions. Stellaria media seeds exhibited greater dormancy when matured at lower temperatures (Van Der Vegte 1978). Germination polymorphism of this type is strong evidence for a strong interaction of the environment with seed genotype (Thompson 1981). Germination polymorphism in diffuse and spotted knapweed could arise, at least in part, through after-ripening. Diffuse knapweed seeds are especially prone to after-ripening in the field because of their prolonged retention within the capitulum following maturation. Variable levels of seed dormancy associated with different sites could reflect environmental differences during the period of pre-harvest after-ripening. Furthermore, as knapweed seed development is asynchronous, both within a single capitulum, and between different capitula on a single plant, the duration and rate of after-ripening could vary among progeny. For example, seeds borne by a diffuse knapweed plant could conceivably differ in age by as much as 3 months by the end of the growing season. If seeds are initially dormant, the greater duration of after-ripening experienced by seeds in earlier maturing capitula would lead to more dormancy loss and, therefore, a higher proportion of ND seeds than later maturing seeds. Consequently, germination polymorphism may reflect spatial or temporal environmental heterogeneity in the sites, plants, or capitula from which the seeds are collected, or differences in the timing of collection or seed dispersal relative to the date the after-ripening process began. However, factors other than timing of seed development clearly interact with the after-ripening process in some species. For example, Forsyth and Brown (1982) found that disc achenes of Bidens pilosa after-ripened more rapidly than ray achenes (from approximately 50 to 100% germination in 14 days versus an increase of 20 to 40%, respectively). 88 5.4.4 Regulation of Field Germination Plants inhabiting a specific niche tend to exhibit similar germination behaviour (Angevine and Chabot 1979). The after-ripening requirement for germination is prevalent among winter annuals of shallow or sandy soil, where it is thought to prevent premature germination in these dry habitats (Ratcliffe 1961; Newman 1963; Baskin and Baskin 1972a, b, 1973a, 1976b, 1979b, 1986; Grime et al. 1981). Seeds of these winter annuals are shed dormant in spring and require a period of after-ripening in the summer before they will germinate in the autumn. For example, field germination of Senecio vulgaris coincided with periods of high rainfall following warm dry periods (Popay and Roberts 1970b). Dormancy loss through after-ripening was proposed to be the mechanism by which this phenomenon occurred. The requirement for after-ripening in LI diffuse and spotted knapweed seeds is an additional means of distributing germination temporally. Further field experiments are necessary to determine whether the after-ripening requirement restricts germination of LI seeds to the spring or merely delays germination to later in the autumn. 5.4.5 After-ripening: Considerations in Seed Germination Behaviour Studies An important point emphasized by this section is that seed dormancy characteristics can change during storage! The possibility of misleading conclusions arising through inattentiveness to storage of seeds has been raised elsewhere (see Cavers 1974). However, researchers still overlook this fact. For example, in a recent paper, the effect of provenance on the germination characteristics of Parthenium hysterophorus (Asteraceae) was examined (Pandey and Dubey 1988). Unfortunately, the germination characteristics exhibited by the seeds used in this study may not accurately reflect that of the field populations at the time of dispersal because the seeds were stored at room temperature for 5 months prior to examination. After-ripening during this time could have dramatically changed germination behaviour. To their credit, Pandey and Dubey (1988) described the temperature conditions 89 under which the seeds were stored; other papers have omitted this information. Changes associated with after-ripening contribute to the variable light sensitivities of different Lactuca sativa seed stocks (Evenari and Neuman 1952). Storage conditions also affected the viability and germination characteristics of Rumex crispus (Cavers 1974). Another study, which concluded that light quality affects germination behaviour directly through its effect on the phytochrome pigment system during seed maturation (McCullough and Shropshire 1970), used seeds reared under different light regimes and then stored for 30 to 40 days at 25 °C and a RH of 40% or greater. The conclusions drawn in this report are suspect because of the combination of delay in examining the seeds, and storage of seeds under conditions conducive to after-ripening. If seed moisture content differed in seeds reared under the different light regimes employed in this study, the differences in dark germination could have arisen due to post-harvest after-ripening (the authors even reported that light sensitivity changed with time due to after-ripening). In addition, the effect could arise from differences in seed ripening prior to collection, resulting from differences in growth chamber conditions, rather than light quality effects on the seed phytochrome per se. Obviously, carefully controlled experiments are needed to separate the effects of light quality and after-ripening on seed maturation. The usefulness of the experiment examining the effect of light quality during seed maturation on spotted knapweed germination characteristics in the previous section of this thesis (see section 4.0), was flawed by this inability to discern light quality and after-ripening effects. Similarly, the outcome of the experiment in this section on the time course of changes in germination behaviour during after-ripening was dependent upon the ambient RH during storage. Fortunately, the experiment was serendipitously conducted at a RH favouring after-ripening (as the control treatment where seeds were after-ripened in tins with no RH manipulation produced levels of dark germination comparable to that attained in the 32.8 % RH treatment). This experiment could quite easily have produced results leading to the interpretation that after-ripening had 90 no effect on knapweed seed germination behaviour had the ambient RH level been either sub-or supra-optimal. Clearly, researchers wanting to determine if after-ripening affects a species' germination must be aware of the influence of RH on after-ripening lest they unknowingly work under RH conditions unfavourable to the process. The fact that polymorphic germination behaviour can be generated in different samples of seeds stored at room temperature through after-ripening is another important consideration. For example, if differences in seed moisture content exist at the time of collection, and seeds are stored in sealed containers in which the volume of air is relatively small in relation to seed volume, moisture content differences will persist during storage and potentially affect the rate of after-ripening. Consequently, the greater the delay in examining seeds, the greater the uncertainty that observed differences are not artifacts of seed handling, unless seeds are stored under conditions arresting after-ripening. The possible implications of unquantified after-ripening effects in the seed germination literature are immense. For example, most studies of Lactuca sativa germination behaviour have used stored seeds (Globerson 1981). In cases where proper storage of seeds was questionable, and seeds exhibited a lack of dormancy, germination characteristics may need to be re-examined to ensure that the results accurately reflect the true behaviour of freshly mature seeds. Storage of seeds at room temperature may have contributed to the inability of previous studies (Watson 1972; Watson and Renney 1974) to detect light sensitivity in diffuse and spotted knapweed. 91 6.0 EFFECT OF TEMPERATURE ON IMBIBED KNAPWEED SEEDS The level of dormancy exhibited by a sample of imbibed seeds is generally temperature-dependent. Germination is limited to specific temperature ranges, and, in many cases, only occurs after dormancy is overcome by exposure to low temperature (stratification), or specific diurnal fluctuations in temperature. Temperature-dependent differences in lettuce seed germination are not believed to be associated with phototransformation of phytochrome alone as this process is unaffected by temperature from 0 to 50 °C (Vidaver and Hsiao 1972). However, the effects of light and temperature are often additive or synergistic (Roberts and Totterdell 1981). Partial denaturation of the phytochrome pigment system and reversion of Pf r to P r are proposed causes of germination inhibition in Lactuca sativa resulting from incubation at 35 °C (Ikuma and Thimann 1964). Lipid organization in cellular membranes (believed essential for P^r action) is disrupted as temperature increases above 32 °C (Hendricks and Taylorson 1978) and high temperature is believed to favour rapid reversion of Pf r to P r (Borthwick et al. 1954). A seed's response to temperature can restrict germination to a particular season and is thus important in temperate climates where temperature is the most variable environmental factor (Roller 1964). Wide temperature ranges for germination indicate that temperature control of germination is probably not an important factor in the species' ecology unless soil moisture availability interacts with temperature (Roller 1964). Experiments in this section examined the effect of temperature on diffuse and spotted knapweed seed germination behaviour in darkness and following exposure to R. 92 6.1 C o n s t a n t T e m p e r a t u r e 6.1.1 B a c k g r o u n d Diffuse and spotted knapweed germination behaviour has been compared over a wide temperature range (Watson 1972; Watson and Renney 1974). However, the seeds used in those studies did not exhibit light sensitivity and they were treated with sodium hypochlorite, a known germination stimulant (Hsiao and Quick 1985). The following experiments re-examined the influence of temperature on knapweed seed germination in darkness and following potentiation by exposure to R. 6.1.2. M a t e r i a l s a n d M e t h o d s The effect of temperature on dark germination was examined by incubating seeds at 3, 10, 15, 20, 25, 30, 35, 40, 45 and 50 °C for 5 days. Sensitivity to a 2 min R exposure was examined at the same time as dark germination. However, these seeds were all first incubated at 25 °C for 8 h, exposed to R at 25 °C, and then incubated for 5 d in the same incubators used for determination of dark germination. The period of incubation at 25 °C assured similar levels of hydration and Pf r in the seeds prior to transfer to the various temperature treatments. Following germination counts, the viability of ungerminated seeds was determined in all treatments. Seeds from 3 bulk collections of each species (D2, D4, D6, S4, S8, S10, see Table 2) were used to compare the temperature responses of different samples of diffuse and spotted knapweed. Due to limitations of time and incubator availability, the temperature of each incubator was fixed and all replicates within a run were in the same incubator. Consequently, valid statistical analysis is not possible as the assumption of independence of experimental errors is not met, therefore, only means and standard errors are reported. 93 6.1.3 Results 6.1.3.1 Dark germination No major qualitative differences in dark germination were evident among different samples of a species, or between diffuse and spotted knapweed (Figures 13 and 14). Quantitative differences in dark germination were evident among different samples of both species, especially in the 15 to 25 °C temperature range. In both species, maximum germination was generally attained at 20 °C, although in one spotted knapweed seed sample (S10) germination at 15 °C slightly exceeded that at 20 °C. As the incubation temperature deviated from these values, germination progressively declined. Seeds which germinated in the 10 to 35 °C temperature range appeared normal. The few seedlings present at 40 °C were dead when examined and had produced less than 1 cm of radicle growth. No seeds germinated at 3, 45 and 50 °C in darkness. 6.1.3.2 Light sensitivity Germination levels attained following exposure to R were qualitatively similar to the results obtained in darkness in both species (Figures 13 and 14). However, maximum germination occurred more often at 15 °C. In the temperature range 10 to 30 °C, germination in R treatments was markedly greater than that obtained in darkness. A small number of spotted knapweed seeds germinated at 3 °C in all samples following exposure to 2 min R. However, the three samples of diffuse knapweed failed to germinate at 3 °C in one run (Figure 13), while the same seeds attained an average of < 10% germination in a second run (data not shown). 6.1.3.3 Seed viability Seed viability determinations revealed that some of the depression of germination at supra-optimal temperatures was a result of seed death. Viability began to decline after incubation at 35 °C and no viable seeds remained in any sample following incubation at 94 0 5 10 15 20 25 30 35 40 45 50 Temperature [°C] Figure 13. Effect of Incubation Temperature on Germination of Three Samples of Diffuse Knapweed Seeds Incubated in Darkness and Previously Exposed to 2 Min R at 25 °C. Values indicated are the means of 3 replicates of 50 seeds. See Appendix, Table 56 for means and standard errors. 95 Temperature [°C] Figure 14. Effect of Incubation Temperature on Germination of Three Samples of Spotted Knapweed Seeds Incubated in Darkness and Exposed to 2 Min R at 25 °C. Values indicated are the means of three replicates of 50 seeds. See Appendix, Table 57 for means and standard errors. 96 50 °C (Table 25). A greater proportion of spotted knapweed seeds (average of 40%) died at 40 °C compared to samples of diffuse knapweed (average of 8%). 6.1.4 D i s c u s s i o n The characteristic bell-shaped distribution of dark germination versus incubation temperature for diffuse and spotted knapweed germination is similar to previous reports (Watson 1972; Watson and Renney 1974). However, germination levels at 25 °C were more markedly depressed relative to levels attained at 20 °C in this study than had been reported in those previous studies. In addition, no evidence was found that spotted knapweed was able to germinate in darkness at lower temperatures than diffuse knapweed. The influence of temperature on knapweed germination is qualitatively similar to that exhibited by the asteraceous weed Senecio vulgaris (Popay and Roberts 1970a). Likewise, the cultivated annual flowers Ageratum houstonianum (Thompson and Cox 1979a) and Callistephus chinensis (Thompson and Cox 1979b) have seeds that germinate over the range of 6 to 35 °C, and 3.5 to 35 °C, with optima of 23, and 21 °C, respectively. Conversely, Lactuca sativa (Georghiou and Thanos 1983) germination was stimulated at low temperature although the upper temperature cut-off point was also 35 °C (Borthwick et al. 1954; Ikuma 1964). The viability loss which occurred at high temperatures in this experiment is consistent with previous reports that high moisture content seeds are prone to viability loss when incubated at high temperatures (Toole 1950). Data presented by Takeba and Matsubara (1976) indicated that imbibed Lactuca sativa seeds were killed after approximately 40 and 10 h at 45 and 50 °C, respectively. Light stimulated diffuse and spotted knapweed germination substantially relative to levels attained in darkness at temperatures above 3 °C and below 35 °C. Although Watson (1972) did not find any evidence of light sensitivity in knapweed seeds, he noted that 97 Table 25. Effect of Incubation Temperature on Diffuse and Spotted Knapweed Seed Viability Viability (%) a Diffuse knapweed Spotted knapweed Temperature D2 D4 D6 S4 S8 S10 3 ° C 100 100 99.3 100 100 100 10 °C 100 100 100 100 100 99.3 15 °C 100 100 99.3 99.3 100 100 20 °C 100 100 100 100 100 98.6 25 °C 100 100 100 100 100 100 30 °C 100 98.7 99.3 99.3 99.3 99.3 35 °C 98.7 94.7 96.7 91.3 84.0 78.5 40 °C 98.0 82.0 95.2 64.0 62.7 50.0 45 °C 32.0 0 0.7 0 4.7 0 50 °C 0 0 0 0 0 0 aExpressed as a percentage of filled seeds. 9 8 germination increased within the 7 to 30 C temperature range when germination counts were made 2 and 10 days after sowing. This result could conceivably have been a consequence of light-sensitivity, rather than a requirement for increased incubation time, as seeds were exposed to diffuse white light for brief periods as counts were made. Senecio vulgaris germination also exhibited light sensitivity over the whole temperature range at which germination occurred (Hilton 1983). However, knapweed and Senecio germination behaviour is unlike the widely utilized Grand Rapids lettuce seed system, where seeds require light for germination only at supra-optimal temperatures (Berrie 1966). Other cultivated asteraceous species, Ageratum houstonianum (Thompson and Cox 1979a) and Callistephus chinensis (Thompson and Cox 1979b) also exhibit this temperature-dependent light-sensitivity. Although the consistent light-sensitivity exhibited by diffuse and spotted knapweed seeds is not unique, it does make them potentially valuable systems for the study of phytochrome-mediated germination. In addition, knapweed seeds are easier to manipulate than lettuce seeds because of their larger size and rounder shape. 6.2 E f fec t o f T e m p e r a t u r e D u r i n g D a r k Incubat ion o n S u b s e q u e n t G e r m i n a t i o n B e h a v i o u r at 25 °C. The effect of exposure to different temperatures during a period of dark incubation was examined to determine the effect of thermo-inhibition (sensu Vidaver and Hsiao 1975) on subsequent diffuse and spotted knapweed seed germination behaviour. 6.2.1 M a t e r i a l s a n d M e t h o d s Imbibed seeds (samples D4 and S6) were incubated in darkness at 3, 10, 15, 20, 25, 30, 35, and 40 °C for 5 d. Changes in dark germination behaviour were determined by transferring seeds to 25 °C for a further 5d of incubation. Seeds in the dark germination determination remained in light-tight containers throughout the experiment until counts were conducted. Germination was compared to a control incubated at 25 °C in darkness for 10 days. Light-sensitivity was assessed by exposing separate replicates of seeds pre-incubated at 99 3 to 40 °C as above to 2 min R and 1 d R treatments. Consequently, petri dishes containing the seeds were removed from the light-tight containers during the period of light exposure. Germination was compared to controls incubated at 25 °C and exposed to 2 min and 1 d R after 8 h incubation. Differences in seed temperature were present (differences in dish temperatures were obvious during handling) during light exposure as insufficient time passed for seeds to equilibrate to room temperature (25 °C). Due to limitations of time and incubator availability, the temperature of each incubator was fixed and all replicates within a run were in the same incubator. Consequent^, valid statistical analysis was not possible and, therefore, only means and standard errors are reported. 6.2.2 Results 6.2.2.1 Dark germination The temperature experienced during 5 d of dark incubation affected the overall cumulative germination of diffuse and spotted knapweed seeds following transfer to a further 5 d incubation at 25 °C (Figures 15 and 16). Dark germination of both species was lower than controls incubated at 25 °C when seeds were first incubated at 10, 30, 35, and 40 °C. Although some of the decrease in germination at 40 °C was attributable to a partial loss of viability (see section 6.1.3.3), declines at 30 and 35 °C clearly indicated that secondary dormancy was induced in some ND seeds. Dark germination was substantially greater than controls following incubation at 3 °C. Therefore, either the period of chilling, or the temperature shift which occurred when seeds were transferred to 25 °C, stimulated germination of dormant diffuse and spotted knapweed seeds. Seeds treated at 20 °C exhibited higher germination than the dark control in all cases. Similarly, seeds incubated at 15 °C in some runs germinated to higher levels than the 25 °C dark control (Figures 15 and 16), while in other runs germination of dark controls was greatest (data not shown). However, the germination levels in these treatments was expected 100 Temperature [°C] Figure 15. Effect of temperature during a 5 d dark incubation period on the subsequent germination behaviour of diffuse knapweed seeds incubated for an additional 5 d at 25 °C. Horizontal lines indicate germination of controls incubated at 25 °C in darkness or exposed to 2 min and 1 d R following 8 h of imbibition. Values indicated are the means of three replicates of 50 seeds. See Appendix, Table 58 for means and standard errors. 101 100-1 90-80-70-g eoH C O \"o 50-| c 0) O 40 30-20-10-l 10 15 20 25 30 Temperature [°C] 35 40 Figure 16. Effect of temperature during a 5 d dark incubation period on the subsequent germination behaviour of spotted knapweed seeds incubated for an additional 5 d at 25 °C. Horizontal lines indicate germination of controls incubated at 25 °C in darkness or exposed to 2 min and 1 d R following 8 h of imbibition. Values indicated are the means of three replicates of 50 seeds. See Appendix, Table 58 for means and standard errors. 102 to be higher than controls because seed germination at 15 and 20 °C exceeds that at 25 °C (see section 6.1). Consequently, seed germination completed during the 5 d incubation at 15 or 20 °C would raise the final germination percentage above that of the control seeds incubated at 25 °C for 10 d. As pointed out by Karssen (1980/81a), \"studies of temperature effects on dormancy induction are complicated by the fact that temperature both influences the dormancy status of the seeds and determines the result in the germination test\". 6.2.2.2. Germination following exposure to 2 min R Qualitative changes in germination behaviour, similar to those evident in darkness, occurred in seeds treated with 2 min R. Again both dormancy induction and release were apparent. In all runs, germination of both species was lower than the control when seeds were first held in darkness at 25, 30, 35, and 40 °C (Figures 15 and 16). Clearly, light sensitivity declined following periods of dark incubation at these temperatures. At 40 °C the loss of light sensitivity was associated with increasing loss of seed viability (data not shown), however, the magnitude of the decline in germination greatly exceeded seed death. Incubation at supra-optimal temperatures therefore caused changes in ND and/or 2 min R-sensitive (SR) seeds which resulted in a change in their behaviour to IR. Seeds previously incubated at 3 °C germinated to higher levels following exposure to 2 min R. However, unlike results obtained in darkness (see section 6.2.2.1), no apparent induction of dormancy occurred at 10 °C. Instead, diffuse knapweed germination exceeded the 25 °C following incubation at 10 °C (and 15 °C). with near complete germination occurring in seeds treated at 3 °C. However, spotted knapweed germination was only stimulated by treatment at 3 °C. Either the chilling treatment, or the temperature shift which occurred when seeds were transferred to 25 °C, increased the effectiveness of a 2 min R exposure in seeds treated at 3 °C. In other words, seeds that initially expressed IR behaviour became less dormant and responded to 2 minute R (became SR). The R light treatment was sufficiently promotive of germination that the induced dormancy evident in darkness at 10 °C was not 103 reflected in the final germination percentages of seeds exposed to 2 min R. Presumably, those seeds made secondarily dormant during incubation in darkness at 10 °C, had merely acquired a light requirement. 6.2.2.3 Germination following 1 d R exposure Similar qualitative behaviour to that exhibited in darkness and following 2 min R were evident when seeds were exposed to 1 d R (Figures 15 and 16). Exposure to low temperature reduced the number of diffuse and spotted knapweed individuals expressing LI behaviour as germination levels following treatment at 3, 10, and 15 °C exceeded that of a control incubated at 25 °C and exposed to R after 8 h. Seed sensitivity to a 1 d R light treatment declined following incubation at 25 to 40 °C in one run (Figures 15 and 16). However, in two other runs, germination levels were similar to controls (data not shown). Consequently, it is difficult to conclude whether supra-optimal temperatures increase LI seed percentages, or if the 1 d R treatments are sufficient to overcome any dormancy induced in the 5 day treatment period. Further experiments utilizing longer treatment periods are necessary to clarify this point. 6.2.2.4 Discussion The temperature experienced during a 5d period of dark incubation affected the germination behaviour of diffuse and spotted knapweed seeds subsequently transferred to 25 °C. Incubation at temperatures supra-optimal for germination increased seed dormancy. The increased dormancy was apparent as lowered germination levels in darkness as well as following exposure to R. The failure of ND seeds to germinate upon return to favourable temperatures following high temperature treatments is one form of a phenomenon called thermodormancy. Some workers distinguish a further form of thermodormancy termed skotodormancy. Occurring in light-sensitive seeds of many species (Karssen 1967; Taylorson and Hendricks 104 1973b; Speer et al. 1974; Kivilaan 1975; Vidaver and Hsiao 1975), this form of dormancy is characterized by progressive loss of light sensitivity following prolonged incubation in darkness, as well as declines in responsiveness to temperature, gibberellic acid and other inductive agents (Ikuma and Thimann 1960; Taylorson and Hendricks 1973a; Vidaver and Hsiao 1974; Bewley 1980; Khan 1980/81; Georghiou and Thanos 1983; Hsiao et al. 1984). Although skotodormancj' is considered different from innate dormancy by some (Bewley 1980), the \"fundamental biochemical or physiological distinctions\" between innate (primary) dormancy and such forms of induced (secondary) dormancy are poorly understood (Bewley and Black 1982). Thermodormancy is common among members of the Asteraceae. For example supra-optimal temperatures induced seed dormancy in several Lactuca sativa cultivars (Vidaver and Hsiao 1974; Blaauw-Jansen and Blaauw 1975; Thompson et al. 1979; Bewley 1980; Kristie et al. 1981; Georghiou and Thanos 1983; Hsiao et al. 1984; Hsiao and Vidaver 1989) and in Ambrosia artemisiifolia (Willemsen 1975a). Thermodormancy has also been reported in the seeds of Cirsium palustre (Pons 1984): a weed in the same tribe (Cardueae) as Centaurea. In addition to inducing dormancy, temperature also affected dormancy release in diffuse and spotted knapweed. Incubation at 3 °C stimulated substantial increases in germination in darkness, and after exposure to R, following transfer to 25 °C. Germination of other asteraceous species is also enhanced by stratification: Lactuca sativa (Ikuma and Thimann 1964; Roth-Bejerano et al. 1966; Van Der Woude and Toole 1980), Senecio vulgaris (Popay and Roberts 1970a, Cirsium arvense (Kumar and Irving 1971; Bostock 1978), Artemisia vulgaris (Bostock 1978), Chrysanthemum segetum (Vincent and Roberts 1979), Achillea millefolium (Kannangara and Field 1985). Species from temperate regions often have a cold stratification requirement which prevents germination in the autumn of seed dispersal (Totterdell and Roberts 1979; Vincent and Roberts 1979). Stratification, or chilling, occurs when seeds are exposed to temperatures 105 slightly above freezing in the presence of adequate oxygen and moisture; temperatures in the range of 1 to 10 °C are most effective (Stokes 1965; Vincent and Roberts 1977). Cyclical changes in seed germination behaviour are often associated with seasonal variations in temperature (Karssen 1980/8la). The germination characteristics exhibited by diffuse and spotted knapweed seeds in this section suggest that knapweed seeds could become more or less dormant during burial depending on soil temperature. For example, knapweed dormancj' would be expected to increase when soil temperature exceeds the optimum for germination. Conversely, germination may be stimulated by large fluctuations in soil temperature similar to that experienced in vitro by seeds chilled at 3 °C then transferred to 25 °C in this study. Dormancy in Ambrosia artemisiifolia (Bazzaz 1970; Willemsen 1975b; Baskin and Baskin 1977), and Cirsium palustre (Pons 1984) is broken following a period of stratification in the winter. In some species, the stimulatory action of stratification is only evident if seeds are subsequently exposed to other stimulants such as light, nitrate, or temperature fluctuations (Vincent and Roberts 1977). The results in this section also demonstrate the importance of routine viability determinations in experiments examining germination behaviour. For example, without knowledge of the effect of temperature on seed viability, one might wrongly attribute lower germination levels solely to the induction of dormancy without recognizing the contribution of seed death. Viability determinations are essential to distinguish the confounded effects that induced dormancy and seed death have on germination values. 6.3 Effect of chilling duration on seed germination In many species, a single temperature shift accounts for the stimulatory effect of stratification (Isikawa and Fujii 1961; Bewley and Black 1982). For example, lettuce seeds require only a few hours of chilling for maximum germination stimulation (Van Der Woude and Toole 1980). For example, germination of Lepidium seeds was improved by a single temperature shift immediately prior to, or following, R treatment (Toole et al. 1955). 106 Germination of Rumex species is also stimulated by a single temperature alternation (Totterdell and Roberts 1979). The following experiment examined whether the dormancy loss associated with incubation at 3 °C (see section 6.2). was dependent upon the duration of the chilling period. 6.3.1 M a t e r i a l s a n d M e t h o d s Seeds from D4 and S4 were incubated in darkness at 3 °C for 0, 15, 30, 60, 120, and 240 h prior to transfer to 20 °C for a further 5 d incubation in darkness. The rationale for choosing 20 °C instead of 25 °C for subsequent incubation was that dormancy levels in knapweed was lowest at this temperature (see section 6.1). Consequently, any elevation in germination noted at this temperature would result from germination stimulation of seeds expressing dormancy at the optimal incubation temperature, not merely seeds expressing dormancy enhanced by supra-optimal temperature. However, after the first run of the experiment it was apparent that the stimulation of dark germination by chilling at 3 °C was not nearly as great when seeds were transferred to 20 °C as it had been when seeds were transferred to 25 °C. Consequent!}7, additional experiments examining seed transfer to 25 °C were conducted. Although experiments examining diffuse and spotted knapweed seed germination were conducted concurrently, they were handled as separate experiments. 6.3.2 R e s u l t s 6.3.2.1 Transfer to 20 °C Although germination was slightly higher in chilled diffuse knapweed seeds than in. unchilled controls, no significant differences in germination levels were detectable among treatments in this experiment (Table 26). However, in another run of this experiment, germination was significantly increased by chilling, although no differences among chilling durations were detected (data not shown). Similarly, in spotted knapweed, the average germination levels found across all durations of chilling did not significantly differ from the 107 Table 26. Effect of Chilling at 3 °C on Dark Germination of Diffuse Knapweed at 20 °C Chilling duration (h) Germination (%) 0 33 15 44 30 46 60 41 148 48 240 38 Analysis of variance and orthogonal contrasts Source Sum of squares D.F. Mean square Signif. Treatments 444.7 5 88.9 NS 0 v s l 5 - 240 250.0 1 250.0 NS 15 vs 30 - 240 1.7 1 1.7 NS 30 vs 60 - 240 32.1 1 32.1 NS 60 vs 148 & 240 10.9 1 10.9 NS 148 vs 240 150.0 1 150.0 NS Error 1501.3 12 125.1 108 unchilled control (Table 27). However, germination in the 15 h treatment was significantly greater (p<0.01) than the average germination at all other chilling durations. 6.3.2.2 Transfer to 25 °C Dark germination of diffuse knapweed increased from 8 to 31% following 15 h of chilling, and increased further to 48% after 30 h of treatment when seeds were transferred to 25 °C (Table 28). These differences between the unchilled control and chilled samples, and 15 h and 30 h treatments were significant. No further significant increase in germination was detectable as chilling duration increased above 30 h. Dark germination of spotted knapweed rose from 18 to 30% following 15 h of chilling at 3 °C, and increased further to 45% following 30 h of treatment (Table 29). Differences between the unchilled control and chilled samples, and 15 h and 30 h treatments were significant. The 148 h treatment resulted in a significantly higher level of germination than the 240 h treatment. 6.3.3 Discussion Although the germination of chilled knapweed seeds was generally higher than unchilled control seeds at 20 °C, the effect was not statistically significant, while chilled knapweed seed germination was significantly greater than the unchilled controls following incubation at 25 °C. Interesting^, the average level of germination of chilled seeds was similar in the separate 20 °C and 25 °C experiments: 43% versus 47% and 40% versus 47% in diffuse and spotted knapweed, respectively (Tables 26 through 29). So the statistical significance of the chilling effect in the 25 °C experiment appeared to be largely a consequence of the markedly lower germination of the 25 °C controls comparative to the 20 °C controls. Other workers have examined the effect of chilling on seed germination. Lactuca sativa seeds chilled at 1 °C and transferred to 25 °C exhibited a gradual increase in germination as the duration of stratification increased from 0 to 4 days (Ikuma and Thimann 1964). Berrie (1966) also reported greater germination in lettuce as the duration of an 109 Table 27. Effect of Chilling at 3 °C on Dark Germination of Spotted Knapweed at 20 °C Chilling duration (h) Germination (%) 0 36 15 55 30 35 60 29 148 44 240 37 Anatysis of variance and orthogonal contrasts Source Sum of squares D.F. Mean square Signif. Treatments 1173.3 5 234.7 * * 0 vs 15 - 240 40.0 1 40.0 NS 15 vs 30 - 240 806.7 1 806.7 * * 30 vs 60 - 240 4.0 1 4.0 NS 60 vs 148 & 240 242.0 1 242.0 * 148 vs 240 80.7 1 80.7 NS Error 322.7 12 26.9 110 Table 28. Effect of Chilling at 3 °C on Dark Germination of Diffuse Knapweed at 25 °C Analysis of variance and orthogonal contrasts Source Sum of squares D.F. Mean square Signif. Treatments 5018.0 5 1003.6 ** 0 vs 15 - 240 3763.6 1 3763.6 ** 15 vs 30 - 240 976.1 1 976.1 ** 30 vs 60 - 240 32.1 1 32.1 NS 60 vs 148 & 240 5.6 1 5.6 NS 148 vs 240 240.7 1 240.7 NS Error 1048.0 12 87.3 I l l Table 29. Effect of Chilling at 3 °C on Dark Germination of Spotted Knapweed at 25 °C Chilling duration (h) Germination (%) 0 18 15 30 30 45 60 49 148 67 240 45 Analysis of variance and orthogonal contrasts Source Sum of squares D.F. Mean square Signif. Treatments 4271.3 5 854.3 ** 0 v s l 5 - 240 2131.6 1 2131.6 ** 15 vs 30 - 240 1109.4 1 1109.4 ** 30 vs 60 - 240 186.8 1 186.8 NS 60vsl48& 240 176.3 1 176.3 NS 148 vs 240 726.0 1 726.0 ** Error 530.7 12 44.2 ( 112 interjected chilling treatment was increased from 4 to 24 hours, but when the duration of chilling exceeded 24 hours, the treatment stimulated less germination. Increasing germination of Verbascum blattaria seeds occurred as the duration of stratification at 5 °C increased (Kivilaan 1975). If the germination stimulation noted in diffuse and spotted knapweed seeds following chilling occurs in the field as well as in vitro, a portion of the dormant seeds in the seedbank would be expected to germinate as a result of rising soil temperatures in the spring. However, the stimulating effect of stratification in vitro was not strong enough to cause complete germination of the samples. Consequently, a substantial number of seeds would likely lie dormant in the soil until a stronger stimulus (such as light) was received. 113 7.0 E F F E C T O F A N A E R O B I O S I S O N S E C O N D A R Y D O R M A N C Y I N D U C T I O N 7.1 B a c k g r o u n d Results in the preceding chapter indicated that thermodormancy was induced in diffuse and spotted knapweed following prolonged incubation in darkness at temperatures supra-optimal for germination. The induction of thermodormancy was reported to be an aerobic process in Rumex crispus (Le Deunff 1973), Lactuca sativa (Vidaver and Hsiao 1975; Karssen 1980/8la), Sisymbrium officinale and Chenopodium bonus-henricus (Karssen 1980/81a). Conversely, anaerobic conditions induced dormancy in some cases (Vegis 1964; Holm 1972), including Lactuca sativa (Ikuma and Thimann 1964). In this chapter, experiments were conducted to examine the effect of anaerobic conditions on secondary dormancy induction in diffuse and spotted knapweed seeds. 7.2 M a t e r i a l s a n d M e t h o d s Seeds of both species were imbibed anaerobically or aerobically in darkness at 25 °C for 8 h or 120 h. Separate experiments conducted concurrently examined germination in darkness, following a 2 min R exposure, and following a 1 day R treatment. Aerobic incubation was done in petri dishes as described earlier. Anaerobic conditions were attained by incubating each replicate of 50 seeds in a 25 mL Erlenmeyer flasks filled with autoclaved water previously cooled to 25 °C in a closed system. Flasks were sealed firmly with a rubber stopper to exclude air bubbles, and a rubber band held the stopper in place. After anaerobiosis, water was strained off and seeds were transferred to petri dishes where 5 ml distilled water was added. Seeds were exposed to the green safelight for approximately 20 minutes during seed transfer. Light treatments were initiated immediately after all transfers were completed. Experiments were conducted twice using a completely randomized design. Means were separated using orthogonal contrasts. 114 7.3 Results 7.3.1 Dark Germination Dark germination was significantly lower following anaerobiosis in both diffuse (Table 30) and spotted knapweed (Table 31). Diffuse knapweed germination declined from 34 to 19%, and that of spotted knapweed from 43 to 9%, following 8 h of anaerobiosis. This indicated that the anaerobiosis treatment used in this experiment induced secondary dormancy in ND seeds. Increasing the duration of anaerobiosis from 8 h to 120 h did not significantly decrease dark germination further in spotted knapweed. The 9% decrease in diffuse knapweed germination was significant at the 5% level, however, no significant difference was evident in a second run of the experiment (data not shown). Germination in the 8 h and 120 h aerobic treatments did not differ significantly. This was expected as the aerobic incubation treatment was identical to the standard incubation conditions utilized following most experiments in this thesis (i.e. 5 days incubation in darkness at 25 °C). 7.3.2 Seed Viability Seed viability in the seed samples used in the dark germination experiment decreased significantly (p<0.05) in both species following anaerobiosis (Tables 32 and 33). Diffuse knapweed viability declined from 99 to 96%, and that of spotted knapweed from 100 to 96%, following 8 h of anaerobiosis. A greater decline in viability occurred following 120 h of anaerobiosis: from 100 to 92%, and from 100 to 83% in diffuse and spotted knapweed, respectively. However, this difference was only significant (p<0.05) in the spotted knapweed seed sample. The declines in viability noted could not account for the much larger declines in dark germination (section 7.3.1). Viability loss resulting from anaerobiosis was much greater in the first run of this experiment. Diffuse knapweed viability declined from 100% (aerobic control) to 86% and 69% following 8 and 120 h of anaerobiosis, respectively; similarly, 115 Table 30. Effect of Anaerobiosis on Diffuse Knapweed (D7) Dark Germination Treatment duration Aerobic incubation a Anaerobic incubation 8 h 34_+2 19_+4 120 h 34 + 3 10 + 2 Summary of analysis of variance and partitioning of SS for orthogonal contrasts Source Sum of squares D.F. Mean square Signif.b Treatments 1252 3 417 * * AE vs A N C 1121 1 1121 * * 8 h AE vs 120 h AE 0 1 0 NS 8 h AN vs 120 h AN 131 1 131 * Error 189 8 24 Mean germination (%) +_ standard error b * * _ p