ROLE OF ENVIRONMENTAL FACTORS AND PLANT GROWTH REGULATORS ON THE FLORAL TRANSITION OF Exacum Styer Group by RANI KRISHNASAMY M.Sc (Horticulture), Tamil Nadu Agricultural University, India, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Plant Science) THE UNIVERSITY OF BRITISH COLUMBIA March, 2007 © Rani Krishnasmay, 2007 11 Abstract Exacum Styer Group exhibit many outstanding characteristics that make this crop suitable as a new flowering pot, bedding or cut flower crop. One requirement for a new crop's commercial success is a reliable system to predict or control flower initiation and development. Therefore, this study focuses on evaluating the role of environmental factors on the transition from vegetative to reproductive growth. Light treatments, based on a combination of shade nets and supplemental lights were designed to deliver between 1152 moles and 5211 moles of light over a 36 week period. Visual inspection of plants occurred daily with the dates of first visible bud and anthesis recorded from each individual within all populations (i.e., genetically related seedlings). Average number of days to visible bud and average number of days to anthesis for individual populations ranged from 170-218, and 237-280, respectively. Average number of days to anthesis for individual treatments were as follows: Treatment 1 (ambient light + 65% shade) 272 days; Treatment 2 (ambient light + 35% shade) 278 days; Treatment 3 (ambient light) 273 days; Treatment 4 (ambient light + supplemental light) 238 days. Analysis of variance results indicate significant effects (P<0.05) on days to visible bud and days to anthesis for population, treatment, and their interaction. These sources of variation are partly attributed to the high level of heterozygosity within these populations reflective of their interspecific background. Changes in meristem structure associated with floral initiation were characterized across populations of Exacum Styer Group. Meristem anatomy and morphology were characterized from tissue harvested weekly and were based on light (i.e., anatomy) and scanning electron (i.e., morphology) microscopy. Structural changes associated with Il l floral initiation were consistent across all individuals and populations and were successfully sub-divided into six relatively distinct stages. In Stage 1, the shoot apical meristem (SAM) is dome to conical shaped with a height to width ratio of 0.90 and without visible structures. In Stage 2, the apical dome becomes flattened as development progresses with the height to width ratio changing to 0.43 and remaining without visible structures. In Stage 3, five sepal primordia are initiated around the apex that resembles foliage leaves in both position and development. During Stage 4, the apex continues to flatten and broaden with a height to width ratio further reducing to 0.09, and with five petal primordia arising simultaneously alternate to the sepals. In Stage 5, initiation of the stamens occurs with their first appearance as short structures alternate to the petals. Finally, in Stage 6, the ovary is initiated. These observations on meristem transition are consistent with previous reports for many other angiosperm species and support these six stages of transition as fundamental in angiosperm development. Meristem width and height were used in allometric analyses and associated with environmental treatments (i.e., light unit sum, heat unit accumulation, time). Coefficients of determination obtained from simple allometric regression ranged from 0.468 (Population 02-65) to 0.980 (Population 02-62). To include additional sources of variation over that of the simple allometric regression (i.e., interactions between both the environment (i.e., treatments) and genotype (i.e., both within and among individual families)), an expanded power function model was developed. The best subset regression procedure was performed with the expanded allometric model to determine which independent variables were most appropriate to include in the multiple regression model. iv The models that displayed a low Mallow's Cp value, a low variance inflation value and a high R 2 were selected as the most useful. Direct comparisons between the two allometric models were based on the magnitude of the partial regression coefficients for each parameter in the best subset models. Using the best subset regression results, significant standard partial regression coefficients were observed for light sum (11 of 13 populations) and days (7 of 13 populations) indicating these two parameters significantly affected floral initiation. V TABLE OF CONTENTS Abstract ii Table of Contents v List of Tables vii List of Figures viii List of Acronyms ix Acknowledgement x Chapter 1: General Introduction and Background 1 1.1 Introduction 1 1.2 Research objectives 2 1.3 Literature review 3 1.4 References 16 Chapter 2: Effects of Cumulative Light and Temperature on Floral Initiation on Exacum Styer Group 25 2.1 Introduction..... 25 2.2 Materials and Methods 27 2.3 Results 30 2.4 Discussion 32 2.5 Conclusion 33 2.6 References 44 vi Chapter 3: Al lometr ic Analysis of the Shoot A p i c a l Meris tem's Transi t ion to Reproduction in Exacum Styer G r o u p as Affected by Cumulat ive L igh t 47 3.1 Introduction 47 3.2 Materials and Methods 48 3.3 Results 55 3.4 Discussion 60 3.5 Conclusion 63 3.6 References 69 Chapter 4: Conclusions and Recommendations for Fur ther work 73 Appendix 1: Population identification and parentage 74 Appendix 2: Effects of Plant Growth Regulator Treatments on Floral Initiation on Exacum Styer Group 75 V l l List of Tables Table 2.1 Summary of analysis of variance results for days to visible bud (G-VB) and days to anthesis (G-A) for Exacum Styer Group accessions following exposure to light treatments. 35 Table 2.2 Descriptive statistics of population performance for days to visible bud (G-VB) and anthesis (G-A) averaged across light treatments. 36 Table 2.3 Summary of analysis of variance results for node number at which the floral initiation occurred among Exacum Styer Group populations averaged across light treatments. 37 Table 2.4 Descriptive statistics of population effects on the node number at which floral initiation occurred averaged across light treatments. 38 Table 2.5 Descriptive statistics of light treatment effects on days to visible bud (G-VB) and anthesis (G-A) averaged across populations. 39 Table 2.6 Descriptive statistics of light treatment effects on node number at which floral initiation occurred averaged across populations. 40 Table 2.7 Statistics from multiple regression analyses where days to visible bud (G-VB) and days to anthesis (G-A) were dependent variables and light sum and heat sum were independent variables. 41 Table 3.1 Average values of cumulative moles of light at time of meristem transition, days to visible bud, and days to anthesis for light treatments averaged across populations. 64 Table 3.2 Parameters and statistics of simple allometric regressions using ln(y) = ln(oc) + (3ln(z) + ln(e), where (y) is meristem width and (z) is meristem height. 65 Table 3.3 Standard partial regression coefficients, Mallow Cp, variance inflation factor (VIF) and R 2 from the best subset regression tests for allometric relationships between meristem width and meristem height among Exacum Styer Group populations. 66 V l l l List of Figures Figure 2.1 Population performance for days to visible bud (G-VB) and anthesis (G-A) averaged across light treatments for selected populations 42 Figure 2.2 Light treatment effects on node number at which floral initiation occurred averaged across populations (error bars = SD). 43 Figures 3.1Thin section and SEM images of Exacum Styer Group apical meristems during the transition from vegetative to reproductive growth (Stages 1 and 2). Stage 1- domed to conical shaped meristem with leaf primordia, and numerous rudimentary axillary meristems visible (A-D); Stage 2- broadened meristem with the commencement of floral initiation with leaf primordia visible (E-F). 67 Figures 3.2 Thin section and SEM images of Exacum Styer Group apical meristems during the transition from vegetative to reproductive growth (Stages 3 through 6). Stage 3- Broadening of meristem with the initiation of five-sepal primordia around the apex(A-B); Stage 4- initiation of five petal primordia alternate to the sepals represents (C); Stage 5- stamen initiation appearing as short structures alternate to the petals(D-E); Stage 6- initiation of the ovary (F). 68 LIST OF ACRONYMS ix ABA Abscisic acid AL Ambient light AL+S Ambient light + supplemental light AL+SH65% Ambient light + shade net with 65% shade AL+SH35% Ambient light + shade net with 35% shade C° Degree Celsius Ca Carpel CCC Chlorocholinchloride DF Degrees of freedom DN Day neutral FAA Formalin-acetic acid-alcohol FM Floral meristem GA Gibberellic acid G-A Days from germination to anthesis G-VB Days from germination to visible bud H Heat units HS Sum of heat units / Heat sum LD Long day LP Leaf primodia LS Light sum MH Meristem height MS Mean Square MW Meristem width MYB Myeloblastosis binding protein PAR Photosynthetic active radiation Pe Petal PGR Plant growth regulators PPF Photosynthetic photon flux SAM Shoot apical meristem SD Standard deviation SD Short day Se Sepal SS Sum of square SEM Scanning electron microscope St Stamen wks Weeks VIF Variance inflation factor X A C K N O W L E D G E M E N T S I wish to thank my advisor Dr. Andrew Riseman and members of my advisory committee, Drs. Iain Taylor and Peter Jolliffe for their advice and constructive criticisms. Special thanks to Dr. Peter Jolliffe for his advice and help in completing the allometric analyses. I also wish to thank Dr. Mahesh Upadhyaya for his advice and encouragement that helped me to return to the program with determination and commitment. Finally I want to thank my family, for their continuous support and encouragement. 1 CHAPTER 1 General Introduction and Background 1.1 Introduction The genus Exacum L., is a member of Gentianaceae and is reported to contain approximately 65 species (Klackenberg, 1985). Of these, only Exacum affine L., also known as Persian violet, has been commercially bred for horticultural use. However, several additional species of Exacum have been identified as possessing horticultural potential. These species, native to Sri Lanka, have been collected and inter-crossed to produce multiple interspecific populations for use in a cultivar development breeding program. These populations exhibit many outstanding characteristics that make this crop suitable as a new flowering pot, bedding or cut flower crop. Notable characteristics include large flowers ranging in colour from violet to light blue, glossy green foliage in variable shapes and sizes, and several growth habits from upright to round. Through 12 sexual generations, these interspecific hybrids now form a cohesive group formally named Exacum Styer Group (Riseman et al., 2005). One requirement for a new crop's commercial success is a reliable production system that allows for the prediction or control of flower initiation and development. Unfortunately, this information has not been available for these Sri Lankan species or hybrids. 2 Therefore, this study focuses on evaluating the role of environmental factors (i.e., cumulative light, heat unit accumulation) and plant growth regulators on the transition from vegetative to reproductive growth by characterizing changes in meristem structure. 1.2 Research objectives The objectives of this investigation were: 1. To determine the affect of light quantity (i.e., integral) on the timing of floral initiation in Exacum Styer Group; 2. To evaluate the effects of selected plant growth regulators on floral initiation and development in Exacum Styer Group; 3. To document the extent of phenotypic variation among Exacum interspecific seedling populations for floral initiation and development across light treatments; 4. To characterize morphological changes in the apical meristem during the transition from vegetative to reproductive growth; 5. To analyze the relationship between meristem width and height during the transition from vegetative to reproductive growth. I believe these investigations will provide valuable data on the environmental stimuli that affect the initiation and development of flowers in Exacum Styer Group. This information is critical for both academic and industrial plant scientists working with this crop who wish to control the flowering cycle for either breeding efforts or commercial crop scheduling. 3 1.3 Literature Review Irradiation effects on floral initiation Flower initiation is primarily influenced by environmental factors that serve to communicate the time of year and or growth conditions favourable for sexual reproduction and seed maturation. These factors can include light intensity (i.e., Photosynthetically active radiation (PAR) or photosynthetic photon flux (PPF) between 400 - 700 nm wavelength of light), light integral (e.g., quantity of light delivered over a defined period of time; also known as daily light sum, daily light level, or daily PPF), photoperiod (i.e., day length), temperature, and/or vernalization (i.e., exposure to cold temperatures followed by warmer temperatures) (Yaron and Caroline, 1998). Large amounts of energy are typically required for flower initiation and development in many horticultural crops. This value is often achieved by supplementing ambient light with high intensity lighting. Many bedding plants and summer annuals can be forced to flower earlier than they otherwise would, when produced with supplemental lighting rather than with only ambient light. In many plants, daily light integral significantly affects floral initiation where greater total light sums promote earlier flowering. However, individual horticultural crops have specific light integral requirements for flowering and range from 5-10 mol nf" d" for African violet to 30-50 mol m" d" for rose and chrysanthemum (Hendricks et al.,1993). Using the mean daily light integral concept, floral initiation can be manipulated through the use of supplemental lighting (Hendricks et al, 1993). For example, growing small flowered petunia cultivars like 'Million Bells' and 'Carillion' with 13 mol m" d" daily light integral accelerated flowering by 15 days compared to 4 plants grown with a daily light integral of only 6.5 mol m/2 d"1 (Kaczperski et al., 1991). In addition, poinsettia (Euphorbia pulcherrima Willd. ex Klotzsch) production protocols suggest a light sum of 4.6 to 5.3 mol m"2 d"1 as this level shortened the production period, increased bract expansion, and promoted more uniform shoot development (Ludolph, 1994). Saintpaulia ionantha Wendl., (African violet) production guidelines recommend light sums between 4-8 mol m"2 d"1 to produce high quality plants (Ludolph, 1993). In addition, when light sums exceed 4.2 mol m"2 d"1, the timing of flower initiation and development were significantly enhanced. However, this response was highly dependent on cultivar (Ludolph, 1993). Flower development in Exacum affine is promoted by higher light integrals. For example, supplemental lighting of 183 pmol m"2 s"1 and a 16 hour photoperiod decreased time to floral initiation and produced more compact plants with improved branching (Holcomb and Craig, 1983). In other experiments with E. affine, daily light integral was positively correlated with days to anthesis where higher irradiance treatments promoted earlier flowering (Williams et al., 1983; Serek and Trolle, 2000; Tardiff and Dansereau, 1993). In addition, supplemental irradiance during the winter months significantly reduced the production period of Exacum cv Best Blue (Anderson and Abrandt, 1988) again supporting the positive effects of supplemental lighting on this crop. Photoperiod Photoperiodism, the response of plants to night length, regulates many responses in.plants including flower initiation and development, plant dormancy, seed germination, storage organ formation, and plant growth habit (Dole and Wilkins, 2005). Night-length 5 regulated flower initiation, referred to as photoperiodic induction, is considered independent from the effects of light integral. Phytochromes, referred to as a plant's internal clock, signal the change in the day length by changing between physiologically active and inactive forms with the ratio of these forms determining the plant response. Red light induces the formation of the active form while far-red light promotes the inactive form (Garner and Allard, 1920). The inactive form (Pr or phytochrome red) has light absorption maximum at 666 nm and following exposure, is converted into the active form, Pfr (phytochrome far-red). The active form returns to the inactive form by absorbing far-red light, approximately 730 nm, or by exposure to darkness. Cryptochromes, a subset of phytochromes, absorb both blue (400-500 nm) and ultraviolet light (UV) (100-380 nm) and function in phototropism (e.g., plants growing towards light), and stomata signalling. In Arabidopsis thaliana L., both far red light and blue light are known to promote flowering (Brown and Klein, 1971) with mutations in the gene encoding the blue light receptor CRYPTOCHROME (CRY) delaying flowering and reduceing the photoperiodic response (Guo et al, 1998) Phytochromes and cryptochrome act through a number of mechanisms including PHYTOCHROME-INTERACTING FACTOR 3 (PIF3), a basic helix-loop-helix transcription factor. Phytochrome B is activated by light and then binds to PIF3, resulting in the up regulation of both CIRCADIAN CLOCK ASSOCIATED! (CCA1) and LATE ELONGATED HYPOCOTYL (LHY), two proteins postulated to be constituents of the circadian oscillator (Martinez-Garcia et al, 2000). 6 In Arabidopsis, analysis of photoperiodic responses to flowering at the molecular level was reported by Koornneef et al., (1991). They identified CONSTANS (CO), FLOWERING LOCUS 7 (FT) and GIGANTEA (Gl) gene products as promoting flowering in response to long day photoperiods. The CONSTANS (CO) gene plays a key role in day length sensing and encodes a transcription factor that directly activates FLOWERING LOCUS T (FT). Mutants to CO flower later under long days, but were unaffected under short days. Over expression of CO caused earlier flowering under either long day or short day photoperiods (Putterill et al., 1995; Suarez-Lopez et al., 2001). Most photoperiodic plants can be grouped into three basic classes: short day (SD), long day (LD) and day neutral (DN). SD plants will flower when the dark period is longer than a critical species-specific critical length, and long day plants will flower when the dark period is shorter than a species-specific critical length. DN plants will flower regardless of the night length. For example, when Petunia hybrida Vilm. and Callistephus Cass, plants were grown under SD conditions, they were highly branched, compact, and with delayed flowering. However, under LD conditions, the growth pattern was reversed so that the plant habit was less branched and compact but flowered more quickly indicating floral induction occurred sooner (Kaczperski et al., 1991). In addition, photoperiodic effects can be modified by other factors including temperature. For example, experiments with Alstroemeria L., hybrids, a LD plant, indicate that flower initiation is promoted by a combination of long days (14-16 hrs) and low temperatures. Under these conditions, vegetative shoots initiated flowers earlier and had reduced inter-node lengths (Bakker et al., 1995; Baevre and Bakken, 1997). 7 Research on Eustoma Salisb, a closely related genus to Exacum and considered a quantitative long day plant (e.g., ability to flower under a wide range of day lengths but flowers more rapidly under long days), indicated that day-length extension, provided by incandescent lamps, was beneficial for decreasing the total number of nodes and number of days to flower (Greuber et al, 1985). Podivinsky (1993) supported this observation and concluded day length had a considerable effect on flower bud differentiation and that long day treatments accelerated the flowering process. In North America, production guidelines for E. affine recommend the use of high light intensities with long photoperiods. Under these conditions, high quality plants have been produced with fluorescent lamps (supplying 183 pmol m"2 s"1) and a 16 h photoperiod (Holcomb and Craig, 1983). However, in Exacum Styer Group, Anon (1994) reported that day length had little effect on floral initiation, floral development or overall growth but speculated that growth and flowering in these plants may be due to an interaction between irradiance and temperature. She based this thought on the observation that during the summer months (e.g., higher light levels and temperatures), plants produced greater number of buds, greater number of open flowers, increased plant height, and increased shoot fresh weight as compared to plants grown during the winter months (e.g., lower light levels and temperatures). Temperature Many plants are dependent on thermal energy (i.e., temperature) to control developmental rate as well as the transition from vegetative to reproductive phases (Landsberg, 1975). Temperature not only helps influence the time and rate of floral 8 development but also whether normal development occurs (Kinet et al, 1985). For example, compared to roses grown at optimum temperatures, roses grown below 12°C produced blind shoots while those.grown above 30°C resulted in abnormally small flowers (Moe, 1971). Pelargonium Xdomesticum Baily (regal geranium) require low temperatures for flower bud initiation (2-7 °C) and long days and high light integrals for flower development (Hackett et al., 1974). However, the temperature requirements may vary for each developmental stage of a plant. For example, in Saintpaulia ionanatha Wendl., (African violet), the greatest number of inflorescences were produced on plants grown at 18°C and a light intensity of 240 pmol m"2 s"1, while the greatest number of flowers per inflorescence were produced on plants grown at 24-27°C and 160 pmol m"2 s"1 (Hildrum and Kristoffersen, 1969). This report suggests that initiated flower buds had aborted during development when exposed to inappropriate temperatures and/or light levels. In Arabidopsis, wild type strains grown at ambient temperature flowered earlier at 23°C than at 16 °C (Blazquez et al., 2003). The flowering repression caused by low ambient temperature is independent of the repression produced by several photoreceptors (Halliday et al., 2003). In addition to average daily production temperature, the time period over which thermal energy accumulates is important. Accumulation of thermal energy over time is known as degree-days or heat unit accumulation, and has been used to determine planting dates, prediction of harvest dates, and selection of appropriate crop varieties. The accumulation of degree-days from a known starting point can help predict when a specific developmental stage will be reached. The date to begin accumulating degree-days, known as the biofix date (Arnold, 1959), varies among species. Biofix dates are 9 usually based on specific biological events such as planting dates or some other fixed event during crop production. Attempts have been made to construct prediction models based on accumulated degree-days for several flowering plants. With the construction of valid models, our understanding of plant development in response to temperature can lead to more reliable predictions of crop yield and may help in the design of crops that are better matched to their environment (Atkinson and Porter, 1996). However, prediction models constructed for ornamental plants (e.g., Lilium longiflorium Thunb, and Euphorbia pulcherrima) commonly express the development process as a function of time and not degree-days (Karlsson et al, 1988). Regardless, use of degree-days can help predict the stage of development reached following a specific time interval for some species. The relationship between developmental rate and daily average temperature is often linear within a range between the 'base' and 'optimum' temperatures for a specific crop (Arnold, 1959). This means that within this range, higher average daily temperatures will result in greater growth rates (e.g., shorter production period) and earlier flowering (Roberts and Summerfield, 1987). Therefore, for each crop, it is necessary to quantify the effects of temperature in terms of both absolute temperatures used during production as well as degree-day accumulation in order to better understand their influence on flower initiation and development. Hormones and Plant Growth Regulators (PGR) Plant hormones are an integral component in the control of plant growth and development. A plant hormone is generally described as an organic compound 10 synthesized in one part of the plant and translocated to another part, where in low concentrations (e.g., <1 mM and often <1 pM), it elicits a physiological response (Salisbury and Ross, 1992). The gibberellin class of plant hormones has long been thought to be involved in flower initiation, especially following exposure to long photoperiods or vernalizing conditions. In two LD plants, Lolium temulentum L., and Arabidopsis thaliana (L.,) Heynh., it has been shown that exogenously applied gibberellic acid (GA) promoted flower initiation (Mandel et al., 1992) while Arabidopsis mutants defective in GA biosynthesis (gal and gal-3) tended to flower later than wild type plants (Koornneef and Van der Veen, 1980). Molecular analysis of these mutants identified two putative GA-regulated genes that may be involved in floral initiation. The first gene, known as LEAFY, is thought to be involved in the conversion of a vegetative meristem to a reproductive meristem (Blazquez et al., 1997). The second gene belongs to the myeloblastosis binding protein (MYB) class of transcription factors (Blazquez et al., 1997) and is thought to mediate changes in gene expression, particularly of the LEAFY gene through the GA:Myb protein complex (Blazquez et al., 1998). In addition, a correlation has been identified between earlier flowering and increased GA concentrations within the shoot apex of these long day plants (Weigel and Nilsson, 1995). Physiologically, externally applied gibberellins can substitute for a long-day photoperiod or for a cold treatment, especially in species that grow as a rosette form when vegetative and 'bolt' (i.e., increased internode elongation) with floral initiation. This suggests that gibberellic acid may be a component of the flowering stimulus in 11 these plants because a known secondary effect of exogenous GA is to induce stem elongation (Taiz and Zeiger, 1998). In addition, declines in endogenous gibberellin concentrations were linked to flower abortion due to low light intensity or low temperature (Zieslin and Halevy, 1976) thereby supporting this hormone concentration: flower induction relationship. Several gibberellin biosynthesis inhibitors have been identified and include paclobutrazol, ancymidol, chlorocholinchloride (CCC) and abscisic acid (ABA). Of these, only ABA is an endogenous plant hormone that can inhibit flowering in both LD and SD plants (Bernier et al, 1981). However, ABA can also promote flowering in some SD plants (e.g., Plumbago indica L,) when grown under sub-optimal inductive conditions (Nitsch and Nitsch, 1967). In Impatiens balsamina L , plants grown under inductive photoperiod, ABA promoted flowering when applied externally by reducing the time required for flower initiation, increasing the number of floral buds initiated, and delaying the onset of reversion to a vegetative state upon plant transfer to non-inductive photoperiods (Saini and Sawhney, 1983). However, ABA failed to induce flowering when plants were grown under non-inductive conditions. The effect of CCC, and other anti-GA compounds on floral initiation has been widely studied. For example, the growth and flowering of Chrysanthemum cv. 'Local' was investigated by Godha et al, (2000) and found that CCC reduced plant height compared to the control, as expected, but also delayed flowering. Al Humaid (2001) reported that flowering time was increased (i.e., delayed flowering) as CCC concentration increased when applied to chrysanthemum. Additional effects of CCC 12 included more compact inflorescences with a higher number of florets per spike and increased stem thickness. Floral Initiation and Organogenesis in Exacum Styer Group No information is available on the phenology of flower development for any Exacum species. However, in other floricultural crops (e.g., chrysanthemum and carnation), detailed knowledge of phenological changes attributed to flower initiation have been instrumental in the development of successful production guidelines. However, in order for this information to be useful, it depends on accurate macro morphological characterization of both vegetative and reproductive meristems. Bernier et al., (1981) describe five major macro morphological features common to all plant species that take place during the meristematic transition from vegetative to reproductive growth. The first feature they describe to indicate floral initiation was stem elongation or 'bolting', manifested by an increase in internode length. This phenomenon occurs in plants that normally grow as rosette prior to flower initiation but also in some caulescent plants. Most often, bolting occurs before floral initiation can be detected in the meristem. In Exacum Styer Group, it has been observed that bolting and stem elongation occurs prior to the appearance of macroscopic flower buds and may indicate the beginning of the floral initiation process (Riseman, personal communication). The second major macro morphological feature associated with meristem transition is the breaking of apical dominance with the precocious growth of auxiliary meristems that normally develop into flower or inflorescence primordia. The third 13 feature described is associated with leaf growth and initiation. With the onset of reproduction, many species display an increase in the number of initiated leaf primordia. In addition, the size of the leaves along the reproductive axis changes from a large vegetative form to a smaller, appendage-like form, while becoming progressively simpler in shape. The final two morphological features that indicate the onset of reproduction are changes in meristem size/form, and changes in phyllotaxis. The vegetative meristem, commonly a flattened dome, often rises and broadens prior to visualization of the first floral primordia. The doming effect is attributed to changes in both cell division and cell elongation of the meristem. In addition, a higher order phyllotaxy, (i.e., more complex arrangement of leaves or appendages) is characteristic of the onset of reproductive growth. Satina and Blakeslee, (1941) described the changes in Datura stramonium L , that took place when a vegetative meristem transitioned to a reproductive shoot. Initially, the vegetative apex had the form of a low dome. With the onset of flowering, the apex became enlarged and more mounded. Simultaneous with enlargement, five sepal primordia were initiated around the apex. The apex then flattened and broadened while five petal primodia arose in alternate position to the sepals. Next, five stamen primodia were initiated in alternate position to the petals. Finally two crescentic carpel primordia were initiated and the remaining distal portion of the apex mounded to form the internal tissues of ovary. Wetmore et al, (1959) emphasized that the essential features of the transformation from vegetative to reproductive growth were the same regardless of 14 ultimate morphology of the reproductive structure. For example, they found similar developmental processes in the compact head of Xanthium pensylvanicum Walk, the branching inflorescence of Chenopodium album L , and the single flower of Papaver somniferum L. The development of a reproductive meristem can thus be divided into six distinct stages: (1) vegetative stage; (2) transition to reproductive stage; (3) sepal initiation; (4) petal initiation; (5) stamen initiation; and (6) carpel initiation. In Arabidopsis, floral meristems arise at the flanks of the inflorescence meristem at shoot apices. Floral meristem identity (FMI) genes are responsible for the transition, LEAFY (LFY), and APETALA 1 (API) (Weigel et al, 1992; Mandel et al, 1992). In Antirrhinum, these genes are FLORICAULA (FLO) and SQUAMOSA (SQUA) (Coen et al, 1990; Huijser et al, 1992). In Petunia hybrida gene UNSHAVEN (UNS) shares sequence similarity with the Arabidopsis thaliana flowering gerte SUPPRESSOR OF OVEREXPRESSION OF CONSTANS1, is expressed in vegetative tissues, and is down regulated upon floral initiation and the formation of floral meristems (Ferrario et al, 2004) Summary In order to develop a reliable production system for Exacum Styer Group, several critical pieces of information need to be obtained. These include knowledge of the environmental factors that affect the regulation of floral initiation and development and a detailed description of the morphological changes in meristem structure that occurs during the transition from vegetative to reproductive growth. With these two sources of information, precise identification of the meristem's transition to reproduction can be made and correlated to the production environments used. Therefore, this research 15 focuses on evaluating the role of cumulative light (PAR) and heat unit accumulation the transition from vegetative to reproductive growth by characterizing changes meristem structure. 16 1.4. References Al Humaid, A.I. 2001. Physiological responses of Gladiolus gandavensis cv. Rose Supreme to cycocel (CCC) application. Alexandria Journal of Agricultural of Research 46: 89-96. Anderson, H and Abrandt, Z. 1988. Kunstlyo til Exacum affine Best Blue. Gartner Tiden Development 32: 836-837. Anon, K.M. 1994. Environmental studies on the growth and flowering of interspecific hybrids of Exacum species (Gentianaceae) endemic to Sri Lanka. Master of Science Thesis. Department of Horticulture, The Pennsylvania State University. Arnold, C.Y. 1959. The determination and significance of the base temperature in a linear heat unit system. Proceedings of the American Society of Horticultural Science 74: 682-692. Atkinson, D and Porter, J.R. 1996. Temperature, plant development and crop yields. Trends in Plant Science 1: 119-124. Baevre, O.A and Bakken, A.K. 1997. Light intensive production of Alstroemeria under combinations of air and soil temperature. Scientia Horticulturae 68: 137-143. 17 Bakker, J.C, Bot, G.P.A., Challa, H., Braak, N.J and Van De. 1995. Greenhouse climate control. Wageningen Press. Wageningen. Bernier, G., Kinet, J and Sachs, R.M. 1981. The physiology of flowering. CRC Books, Boca Raton, Fla. Blazquez, M.A., Soowal, L.N., Lee, I and Weigel, D. 1997. LEAFY expression of flower initiation in Arabidopsis. Development 124: 3835-3844. Blazquez, M.A., Green, R., Nilsson, O., Sussman, M.R and Weigel, D. 1998. Gibberellins promote flowering of Arabidopsis by activating the leafy promoter. Plant Cell 10: 791-800. Blazquez, M. A., Ahn, J. H and Weigel, D. 2003. A thermosensory pathway controlling flowering time in Arabidopsis thaliana L. Nature Genetics 33: 168 -171 Brown, J.A.M and Klein, W.H. 1971. Photomorphogenesis in Arabidopsis thaliana (L) Heyhn. Plant Physiology 47: 393-399 Coen, E.S., Romero, J.M., Doyle, S., Elliott, R., Murphy, G and Carpenter, R. 1990. Floricaula: a homeotic gene required for flower development in Antirrhinum majus. Plant Cell 63:131-1322 18 nd Dole, J.M and Wilkins, H.F. 2005. Floriculture principles and species. 2 Edition. Printice Hall, New Jersey, USA Ferrario, S, Busscher, J , Franken, J , Gerats, T , Vandenbussche, M , Angenent, G.C and Immink, G.H.R. 2004. Ectopic Expression of the Petunia MADS Box Gene UNSHA VEN Accelerates Flowering and Confers Leaf-Like Characteristics to Floral Organs in a Dominant-Negative Manner. Plant Cell 16: 1490-1505 Garner, W.W and Allard, H.A. 1920. Effect of the relative length of day and night and other factors of the environment on growth and reproduction in plants. Journal of Agricultural Research 18: 553-606 Godha, S, Sharma, L.K and Kumar, A. 2000. Study on the influence of growth regulators on growth and flowering of chrysanthemum. Journal of Phytological Research 13: 175-178. Greuber, K . L , Corr, B.E and Wilkins, H.F. 1985. Eustoma grandiflorum. Pennsylvania Flower Growers Bulletin 361: 4-5, 9. Guo, H , Yang, W , Mockler, T.C and Lin, C. 1998. Antagonistic actions of Arabidopsis Cryptochromes and Phytochrome B in the regulation of flowering. Science 279: 1360-1363 19 Hackett, W., Kister, J and Tse, A. 1974. Flower induction of Pelargonium domesticum Bailey cv. Lavender Grand Slam with exposure to low temperature and low light intensity. HortScience 9: 63-65 Halliday, K. J., Salter, M. G., Thingnaes, E and Whitelam, G. C. 2003. Phytochrome control of flowering is temperature sensitive and correlates with expression of the floral integrator FT. Plant Journal 33: 875 -885. Hendricks, L., Ludolph, D and Moe, R. 1993. Supplemental lighting in greenhouse. Lehr-und Versuchsanstalt fur Gartenbau Hannover-Ahlem, Taspo-Praxis:15. Thalacker, Braunschweig. Hildrum, H and Kristoffersen, T. 1969. The effect of temperature and light intensity on flowering in Saintpaulia ionanatha. Acta Horticulturae 14: 249-255. Holcomb, E.J and Craig, R. 1983. Producing Exacum profitably. Greenhouse Grower 1: 18-57. Huijser, P., Klein, J., Lonnig, W-E., Meijer, H., Saedler, H and Sommer, H. 1992. Bracteomania, an inflorescence anomaly, is caused by the loss of function of the MADS-box gene squamosa in Antirrhinum majus. EMBO Journal 11: 1239-1249 20 Kaczperski, M.P., Carlson, W.H and Karlsson, M.G. 1991. Growth of Petunia xhybrida as a function of temperature and irradiance. Journal of the American Society for Horticultural Science 116: 232-237. Karlsson, M.G, Heins, R.D and Erwin, J.E. 1988. Quantifying temperature-controlled leaf unfolding rates in 'Nellie White' Easter lily. Journal of the American Society for Horticultural Science 113: 70-74. Kinet, J .M, Sachs, R.M and Bernier, G. 1985. The physiology of flowering: the development of flowers. CRC Books, Boca Raton, Fla. Klackenberg, J. 1985. A reevaluation of the genus Exacum (Gentianaceae) in Ceylon. Nordic Journal of Botany 3: 355-370. Koornneef, M and Van der Veen. 1980. Induction and analysis of gibberellin sensitive mutants of Arabidopsis thaliana (L.) Heynh. Theoretical and Applied Genetics 58: 257-263. Koornneef, M , Hanhart, C. J and Van Der Veen. 1991. A genetic and physiological analysis of late flowering mutants in Arabidopsis thaliana. Molecular General Genetics. 229:57-66 21 Landsberg, J.J. 1975. Temperature effects and plants response. Academic Press Inc, Orlando, Florida, U.S.A Ludolph, D. 1993. Effect of light intensity on growth and flowering of ornamental plants and its use in crop control. Verlag Ulrich E. Grauer, Stuttgart. Ludolph, D. 1994. Lighting of poinsettia? Poinsettia Growers Association 17: 112-115. Mandel, A. M , Brown, C.G., Savidge, B and Yanofsky, M.F. 1992. Molecular characterization of the Arabidopsis floral homeotic gene APETALA1. Nature 360: 273-277 Martfnez-Garcia, JF, Huq, E and Quail, P. 2000. Direct targeting of light signals to a promoter element-bound transcription factor. Science 288: 859-863 Moe, R. 1971. Propagation, growth and flowering of potted roses. Acta Horticulturae 31:25-30 Nitsch, C and Nitsch, J.P. 1967. The induction of flowering in vitro in stem segments of Plumbago indica (L). Planta 72: 371-384. 22 Podivinsky, E. 1993. Lisianthus, the story to date: a review of the current literature available on Lisianthus. Plant Pigment Group, Levin Research Center, New Zealand. Pp 1-10. Putterill, J., Robson, F., Lee, K., Simon, R and Coupland, G. 1995. The CONSTANS gene of Arabidopsis promotes flowering and encodes a protein showing similarities to zinc finger transcription factors. Cell 80: 847-857. Riseman, A., Sumanasinghe, V.A., Justice, D and Craig, R. 2005. New name 'Styer Group' proposed for interspecific hybrids of Exacum species native to Sri Lanka. HortScience 40: 1580-1583. Roberts, E and Summerfield, R. 1987. Measurement and prediction of flowering in annual crops. In: J.G. Atherton (ed.), Manipulation of flowering. Butterworth's. London. Saini, Y.R and Sawhney, S. 1983. Promotion of flowering by abscisic acid in Impations balsaminaL. Indian Journal of Plant Physiology 16: 326-329. Salisbury, F.B and Ross, CW. 1992. Plant Physiology. 4th Ed. Wadsworth Publications, Belmont, California. Satina, S and Blakeslee, A.F. 1941. Periclinal chimeras in Datura stramonium in relation to development of leaf and flower. American Journal of Botany 28: 862-71. 23 Serek, M and Trolle, L. 2000. Factors affecting quality and post-production life of Exacum affine. Scientia Horticulturae 86: 49-55. Suarez-Lopez, P, Wheatley, K , Robson, F , Onouchi, H , Valverde, F and Coupland, G. 2001. CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature 410: 1116 -1120. Taiz, L and Zeiger, E. Editors. 1998. Plant Physiology. 3rd ed. Andrew O Sinauer, Editor. Sinauer Associates Inc., Sunderland. Tardiff, L and Dansereau, B. 1993. Exacum cultivar trails. Cecil Del worth Bulletin 22: 14-15 Weigel, D and Nilsson, O. 1995. A developmental switch sufficient for flower initiation in diverse plants. Nature 377: 495-500. Weigel, D., Alvarez, J , Smyth, D.R, Yanofsky, M.F and Meyerowitz, E.M.1992. LEAFY controls floral meristem identity in Arabidopsis. Plant Cell 69: 843-859 Wetmore, R.H, Gifford, E.M and Green, M.C. 1959. Photoperiodism and related phenomena in plants and animals. R.B. Withrow, Editor. American Association of Advanced Sciences Publications. Washington DC. 24 Williams, S., Wolf, S and Holcomb, E. J. 1983. Growth of Exacum affine at three radiant energy levels. HortScience 18: 366-367. Yaron, Y.L and Caroline, D. 1998. The transition to flowering. Plant Cell 10: 1973-1990. Zieslin, N and Halevy, A.H. 1976. Components of axillary bud inhibition in rose plants. Botanical Gazette 137: 291-296 25 CHAPTER 2 Effects of Cumulative Light and Temperature on Floral Initiation On Exacum Styer Group 2.1 Introduction The genus Exacum is a member of the Gentianaceae with species distributed from the central zone of Africa, Madagascar, Arabian Peninsula, eastern India, Himalayas and Sri Lanka through Malaysia to the eastern tip of New Zealand (Klackenberg, 1985). Several species native to Sri Lanka have been collected and inter-crossed to produce multiple interspecific populations now known as E. Styer Group (Riseman et al., 2005). E. Styer Group exhibits many outstanding characteristics that make this crop suitable as a new flowering pot, bedding, or cut flower crop and include large flowers ranging in colour from violet to light blue, glossy green foliage in variable shapes and sizes, and several growth habits from upright to round. Flowering is controlled by variety of interrelated mechanisms. In many plants, the environment influences the production of a floral stimulus, historically referred to as 'florigen', which propagates from the site of reception (e.g., leaves) to the site of action (i.e., meristems). Meristems become committed to the production of flowers following the receipt of a sufficient amount of stimulus. Depending on the species, stimuli affecting flowering can be transmitted via light (e.g., daily light integral, photoperiod), temperature (e.g., degree-days, vernalization), or moisture sensing mechanisms. 26 Within a specific floral induction mechanism, variation is present for the quantity of stimuli sufficient for floral induction for individual species and cultivars. However, this response tends to be linear in nature where increasing levels of stimuli increases flower initiation and/or production. For example, Begonia cv Barkos plants displayed a linear decrease in flowering time and a linear increase in dry matter production as light sums increased from 3.4 to 13.6 mol m" d" (Yrket, 1998). For many cultivars of rose hybrids, flower bud initiation increases linearly over a range of light sums between 4 and 2 1 8 mol m" d" . For example, in cultivars 'Rosamini orange' and 'Pink', the addition of 2 - 1 2.8 mol m" d" supplemental light resulted in linear responses with a shortened production period, increased plant height, and increased dry matter (Zieslin and Tsujita, 1990) In addition to a single mechanism affecting flowering, two or more mechanisms may interact to affect flowering. For example, flower initiation and development in Gerbera jamesonii Adlam., (gerbera daisy) was dependent on both light sum and temperature with supplemental lighting shortening the time to anthesis (Erwin et al., 1991). In Pelargonium Xhortorum Bailey., (garden geranium), plant flowering was strongly dependent on both light sum and temperature where increased light sum from 2 1 3.2 to 6.5 mol m" d" and an optimum temperature of 21-23°C reduced the time from germination to visible bud from 87 to 67 days. However, temperatures below 21°C slowed the development (Langton and Runger, 1985; Kaczperski and Heins, 1995). This research investigated the affect of light sum and its interaction with heat on flowering of Exacum Styer Group. The specific objectives of this study were to: 27 1. determine if light integral (PAR) affects the timing of flower initiation of Exacum Styer Group; 2. evaluate the interaction between light integral (PAR) and heat unit accumulation on the timing of flower initiation. 2.2 Materials and Methods Plant material Thirteen genetically related seedling populations were selected for inclusion in this study (Appendix 1). Seeds were surface sown in April on Rediearth Commercial Mix (Sun Gro Horticulture, Bellevue, WA, USA), that was previously soaked with 1 mg L"1 Benlate™ fungicide (Dupont Canada Inc., Mississauga, Ontario), covered with a Humidome™ (Growtex, Burnaby, British Columbia) and placed into a growth room maintained at 24°C and a 16 h photoperiod provided by cool-white fluorescent lamps 2 1 (120 pmol m~ s" ). Germination occurred within 15-20 days. Approximately two months after germination, seedlings were individually transplanted into plug trays containing a potting mix composed of peat (70%), perlite (20%), pasteurized mineral soil (10%) and osmocote (18-6-12) (3 kg m"). The seedlings were placed on greenhouse benches, located in the Horticultural Greenhouse, University of British Columbia, and covered by shade cloth until acclimated (approx three weeks). After 2 months, seedlings were transplanted into individual 10 cm pots containing the same potting mixture as described above. Pots were placed on greenhouse benches with the aerial temperatures maintained between 20-22°C and the humidity maintained between 85-90%. Plants were watered as needed and fertilized (Plant -Prod 21-7-7 acid fertilizer, 28 Plant Products Co., Brampton, Ontario) biweekly with 200 ppm total nitrogen. Once a month, 15 g 100 L"1 EDTA Zn chelate (Plant Food Company Inc., Cranbury, NJ, USA) was added to the fertilizer solution to prevent the development of zinc stress. Electrical conductivity and pH were checked monthly and maintained between 1-2.25 mS cm"1 and 5-6, respectively. Standard pest control strategies using Safer's™ Natural Insecticide (Safer Ltd., Scarborough, Ontario), Ambush 50EC™ (Zeneca Agro., Calgary, Alberta) and Benlate (Dupont Canada Inc., Mississauga, Ontario), were applied as needed. Light treatments Experimental conditions were created using a combination of shade nets and supplemental lights so that light delivered over a 36 week period was as follows: (1) ambient light + 65% shade (SH65), 1152 moles; (2) ambient light + 35% shade (SH35), 2130 moles: (3) ambient light (AL), 3006 moles; (4) ambient light + supplemental light, 5211 moles. Supplemental light was provided by 430W high pressure sodium lamps (Sun Agro (Philips), Somerset, NJ, USA) positioned to deliver PAR light of 300 pmol 2 1 m" s" to the top of the plant canopy. Supplemental light was provided between 9:00 pm to 5:00 am. Light sum (PAR) was calculated using day length and lighting period multiplied by average light intensities from the day of germination through anthesis. Heat unit calculation Heat units were calculated using the following formula: Heat Units = ((Daily Maximum Temp + Daily Minimum Temp / 2) - Threshold Temp) (Arnold, 1959). The threshold temperature is known only for a few species, and is unknown for Exacum. 29 Therefore, I used a threshold temperature of 10°C because previous research has shown that this is a reasonable approximation for many species (Herms, 1998). The sum of the heat units (HS) from the day of germination through anthesis was used in data analyses. Morphological Data Collection Visual inspection of plants for macroscopic buds occurred daily. Date of first visible bud and date of anthesis were collected from each individual within all populations. Anthesis was determined at the time of petal reflex and full pigmentation. These data were then used in calculations for days from germination to visible bud (G-VB) and days from germination to anthesis (G-A). Statistical analyses A completely randomized block design was used with light treatments as blocks. Thirteen populations (i.e., specific combination of parents) each with 28 individual seedlings were included in each treatment. Descriptive statistics (means, SD), analysis of variance, and mean separation procedures were performed by SYSTAT (SPSS Inc, 1998). Significant differences (P<0.05) among treatments were tested by pair-wise multiple comparisons evaluated by Tukey's significant difference test. In addition to the descriptive statistics listed above, simple and multiple regression calculations were performed separately for each family using SigmaStat (SPSS Inc, Chicago, IL, version 2.03, 1997) software. All data were log transformed prior to regression analysis. The following equations were used in the multiple regression analyses: 30 Y = b + (b.,*LS) + (b2*HS) where Y = dependent variable (i.e., either days from germination to visible bud (G-VB) or days from germination to anthesis (G-A)); LS = light sum; HS = heat sum; b = constant; b\ and b2 = regression coefficients. 2.3 Results Population effects on flowering Analysis of variance results indicate significant effects (P<0.05) for days to visible bud and days to anthesis for population, treatment, and their interaction (Table 2.1). Average population performance in days to anthesis ranged from 237 days (population 02-174) to 280 days (population 02.-59) (Table 2.2). In addition to the 43 day span in average population flowering times, a high level of variation was observed within individual populations with the overall average standard deviation equalling 40.3 days indicating significant within population variation. Mean separation results identify populations 02-174, 02-39 and 02-59 as flowering significantly earlier from all other populations. Results from ANOVA for the node number at which floral initiation occurred indicated no significant differences between treatments and population interaction (Table 2.3). However, there was a significant population effect. Three out of thirteen populations (i.e., 02-174, 02-39 and 02-59) reacted to reproductive growth at node eight while all remaining populations averaged node nine for floral initiation (Table 2.4). 31 Treatment effects on flowering Light treatments significantly affected days to visible bud and days to anthesis where plants grown under the ambient + supplemental light treatment flowered significantly earlier than the other three treatments. This high light treatment produced plants that averaged 165 days and 238 days for day to visible bud and days to anthesis, respectively (Table 2.5). Plants grown under all other treatments took an additional 44 days or more to reach a comparable developmental stage. Treatment effects on the node number at which floral initiation occurred were significant. The ambient + supplemental treatment produced plants that flowered, on average, at node eight while plants from all other treatments transitioned to reproduction at node nine (Table 2.6). Multiple regression analyses When days to visible bud (G-VB) and days to anthesis (G-A) were individually regressed against light sum for individual populations, R 2 values ranged from 0.967 to 0.999 and 0.961 to 0.867, respectively (Tables 2.7). These high values indicate that regardless of population (i.e., genetic heterogeneity among populations), a strong negative relationship exists between light sum and days to visible bud/anthesis where increasing light sums significantly reduce days required. Also, significant P values (<0.001) indicate that light sum strongly contributed to accurately predicting G-VB or G-A. 32 2.4 Discussion Bernier (1988) reported that when environmental, hormonal and genetic signals accumulate in the appropriate concentrations and timing, flowering can be induced. In this study, I found that light sum significantly promotes floral initiation and development in Exacum Styer Group. For most populations, anthesis occurred earlier under the ambient + supplemental light treatment as compared to all other treatments. Similar results were reported with Begonia longifolia Blume, Saintpaulia sp. (African violets), Rosa chinensis var. minima (Sims) Voss (potted miniature roses), and Hydrangea, macrophylla (Thunb.) (Verberkt, 1995; Moe and Mortensen, 1992). When poinsettias (Euphorbia pulcherrima Willd.) and Rosa chinensis L. were grown under supplemental light, it promoted a shorter production period, expansion of bracts, and more uniform shoot development. (Bredmose, 1993; Ludolph, 1994). In previous research on Exacum Styer Group, many genotypes responded positively to higher light intensities with increased flower production and increased rate of development, but not earlier flowering (Anon, 1994). My results are in partial agreement with these findings in that populations grown under the highest light levels displayed the greatest growth rates while all other treatments produced lower but comparable growth rates. In addition, my populations grown under the supplemental light treatment initiated flowering at node eight. However, unlike Anon's results, the increased growth rate I observed translated into earlier flowering. 33 Several reports corroborate the positive effects of increased light sum on flower initiation and development in the related species, Exacum affine L (Williams et al., 1983; Serek and Trolle, 2000; Tardiff and Dansereau, 1993; Holcomb and Craig, 1983). My results were similar in that plants grown with higher light sums displayed faster flowering, with the most rapid flowering occurring with the highest irradiance treatment. However, as in E. affine where significant genotypic variation was observed for days to flower (Rubino, 1993), E. Styer Group also displayed a high level of genotypic variation for this trait. Within an individual treatment, I observed significant variation among populations for days to visible bud and time to anthesis. This is in agreement with the findings of Anon (1994) who reported significant genotypic variation (e.g., continuous flowering genotypes vs. periodic flowering genotypes) was present in E. Styer Group accessions where supplemental irradiance promoted flowering in some genotypes but not others. She also suggested that in genotypes not responsive to higher irradiance, other environmental or internal factors were influencing flower initiation. 2.5 Conclusion My findings indicate that a high level of variation was observed for days from germination to visible bud or anthesis within individual populations, this may be due to genetic heterogeneity among populations. Multiple regression analysis indicates light sum plays an important role in floral initiation of E. Styer Group genotypes and positively influences flower initiation and time to anthesis. However, other factors may 34 be involved that either interact with light sum or act independently to influence flower initiation. This effect may be indirect through node number. 35 Table 2.1. Summary of analysis of variance results for days to visible bud (G-VB) and days to anthesis (G-A) for E. Styer Group accessions following exposure to light treatments. Days from germination (G) to visible bud (VB) (G-VB) Source DF SS MS F P Population 12 182706 15225 7 <0.001 Treatment 3 365090 121696 59 <0.001 Population* Treatment 36 207918 5775 2 <0.001 Residual 752 1543692 2052 Days from germination (G) to anthesis (A) (G-A) Source DF SS MS F P Population 12 107024 8918 6 <0.001 Treatment 3 208009 69336 53 <0.001 Population * Treatment 36 123324 .3425 2 <0.001 Residual 752 974840 1296 36 Table 2.2. Descriptive statistics of population performance for days to visible bud (G-VB) and days to anthesis (G-A) averaged across light treatments. Populations count Days from germination to visible bud (G-VB) Meanz SD Days from germination to anthesis (G-A) Meanz SD 02-21 179 a b 35.3 251 b 34.4 02-39 218 e 47.0 273 d e t 36.0 02-59 218 e 47.0 280 f 33.8 02-60 214 d e 55.0 274 d e f 39.9 02-61 199 53.6 262 b c d 43.7 02-62 203 d e 56.1 268 d e t 41.8 02-64 201 c d 53.4 266 c d e 41.5 02-65 200 c d 70.8 263 b c d 55.6 02-77 214 d e 48.7 277 e t 36.9 02-125 186 b c 49.2 255 b c 41.8 02-141 200 c d 40.8 269 d e t 35.5 02-171 207 d e 55.1 2 6 8 c d e t 40.0 02-174 170 a 53.2 237 a 42.9 Within column entries with same letter superscript are not different by Tukey's significant difference test. 37 Table 2.3. Summary of analysis of variance results for node number at which the floral initiation occurred among E. Styer Group populations averaged across light treatments. Source DF SS MS F P Population 12 50.160 4.180 7.754 <0.001 Treatment 3 6.077 2.026 3.758 0.011 Population x Treatment 36 21.025 0.584 1.083 0.344 Residual 526 283.558 0.539 38 Table 2.4. Descriptive statistics of population effects on the node number at which floral initiation occurred averaged across light treatments. Population count Mean z of Node number SD 02-21 8.9b 0.59 02-39 8.6 e 0.62 02-59 8.7 e 0.63 02-60 9.1de 0.73 02-61 9.1cd 0.61 02-62 9.2 d e 0.60 02-64 9.3 c d 0.68 02-65 9.2 c d 0.68 02-77 9.5 d e 1.42 02-125 9.1 b c 0.67 02-141 9.2 c d 0.59 02-171 9.0 d e 0.72 02-174 8.4 a 0.67 z Within column entries with same significant difference test. etter superscript are not different by Tukey's 39 Table 2.5. Descriptive statistics of light treatment effects on days to visible bud (G-VB) and days to anthesis (G-A) averaged across populations. Treatments Populations count G-VB Mean z SD G-A Mean z SD AL+S 210 165 a 41.6 238 a 39.2 AL 200 215" 51.9 273 b 39.1 AL+SH35% 200 216 b 50.8 278 b 36.9 AL+SH65% 194 209 b 51.0 272 b 39.6 z Within column entries with same letter superscript are not different by Tukey's significant difference test. Table 2.6. Descriptive statistics of light treatment effects on node number at which floral initiation occurred averaged across populations. Treatments Mean z SD AL+S 8.9 a 0.93 AL 9.2 b 0.85 L+SH65% 9.2 b 0.91 AL+SH35% 9.1 b 0.88 z Within column entries with same letter superscript are not different by Tukey's significant difference test. 41 Table 2.7. Statistics from multiple regression analyses where days to visible bud (G-VB) and days to anthesis (G-A) were dependent variables and light sum and heat sum were independent variables. Populations G-VB R2 MS Residual G-A R2 MS Residual P 02-62 0.980 0.000286 0.906 0.000465 <0.001 02-64 0.982 0.000268 0.902 0.000514 <0.001 02-65 0.985 0.000470 0.961 0.000411 <0.001 02-60 0.972 0.000370 0.899 0.000460 <0.001 02-61 0.986 0.000212 0.928 0.000426 <0.001 02-141 0.992 0.0000781 0.936 0.000245 <0.001 02-77 , 0.978 0.000255 0.896 0.000418 <0.001 02-171 0.975 0.000350 0.891 0.000515 <0.001 02-174 0.993 0.000115 0.951 0.000301 <0.001 0-21 0.999 0.000007 0.895 0.000375 <0.001 02-125 0.989 0.000151 0.946 0.000295 <0.001 02-39 0.977 0.000248 ' 0.869 0.000501 <0.001 02-59 0.967 0.000298 0.867 0.000403 <0.001 42 Fig 2.1. Population means for days to visible bud (G-VB) and days to anthesis (G-A) averaged across light treatments for selected populations. 350 -300 -250 -in >. IS 200 -O c s 150 -s 100 • 50 -0 - _ AL+S A L AL+35% Populations • G-A II G-VB AL+65% 43 Fig 2.2. Light treatment effects on node number at which floral initiation occurred averaged across populations; bars = SD. 9.5 8.5 8 T T T X IIMHI piiii|ftiliii • H Bllillli WKm • H • 1 1 1 1 • H llilplllliili I M M M •IIMI iSliliiilil 1111111111111 AL+S AL AL+SH65% AL+SH35% T r e a t m e n t s 44 2.6 References Anon, K.M. 1994. Environmental studies on the growth and flowering of interspecifc hybrids of Exacum species (Gentianaceae) endemic to Sri Lanka. Master of Science Thesis. Department of Horticulture, The Pennsylvania State University. Arnold, C.Y. 1959. The determination and significance of the base temperature in a linear heat unit system. Proceedings of the American Society of Horticultural Science 74: 682-692. Bernier, G. 1988. The control of flower evocation and morphogenesis. Annual Reviews of Plant Physiology and Plant Molecular Biology 39: 175-219. Bredmose, N.B. 1993. Effect of year-round supplementary lighting on shoot development, flowering and quality of two-glass house rose cultivars. Scientia Horticulturae 54: 69-85. Erwin, J , Heins, R and Carlson, W. 1991. Pot Gerbera production. Minnesota Flower Growers Association Bulletin 40: 1-6. Herms, D.A. 1998. Ornamental plants - Annual Reports and Research Reviews. Special Circular 165-99. The Ohio State University, Columbus, OH, U.S.A. 45 Holcomb, E.J and Craig, R. 1983. Producing Exacum profitably. Greenhouse Grower 1: 18-57. Kaczperski, M. P and Heins, R. D. 1995. The effect of timing and duration of supplemental irradiance or flower initiation of plug grown Geraniums sp. HortScience 30: 760. Klackenberg, J. 1985. The genus Exacum (Gentianaceae). Opera Botanica. AiO Print Ltd. Copenhagen Pp 1-144. Langton, F. A and Runger, W. 1985. Hand book of flowering. A.H. Halvey, Editor. CRC Books, Boca Raton, Fla. Ludolph, D. 1994. Lighting of poinsettia? Poinsettia Growers Association Bulletin. 17:112-115. Moe, R and Mortensen, L.M. 1992. Thermomorphogenesis in pot plants. Acta Horticulturae 305: 19-25. Riseman, A , Sumanasinghe, V . A , Justice, D and Craig, R. 2005. New Name 'Styer Group' Proposed for Interspecific Hybrids of Exacum Species Native to Sri Lanka. HortScience 40: 1580-1583. 46 Rubino, B.D. 1993. Genotype and season interaction for time to flowering and flower and plant diameter Exacum affine Balf. HortScience 28: 211-212. Serek, M and Trolle, L. 2000. Factors affecting quality and post-production life of Exacum affine., Scientia Horticulturae 86: 49-55. Tardiff, L and Dansereau, B. 1993. Exacum cultivar trails. Cecil Delworth Bulletin 22: 14-15 Verberkt, H. 1995. Lighting strategies for potted plants. Aalsmeer, Research Station for Floriculture and Glasshouse Vegetables, Report 11. Williams, S, Wolf, S and Holcomb, E. J. 1983. Growth of Exacum. affine at three radiant energy levels. HortScience 18: 366-367. Yrket, G. 1998. More light advances flowering of Begonia. Vakblad voor de Bloemisterji 32: 52. Zieslin, N and Halevy, A.H. 1976. Components of axillary bud inhibition in rose plants. Botanical Gazette 137: 291-296 Zieslin, N and Tsujita, M. J. 1990. Response of miniature roses to supplementary illumination. 1. Effect of light intensity. Scientia Horticulturae 42: 113-121 47 CHAPTER 3 Allometric Analysis of the Shoot Apical Meristem's Transition to Reproduction in Exacum Styer Group as Affected by Cumulative Light 3.1 Introduction The genus Exacum is a member of Gentianaceae and is reported to contain approximately 65 species (Klackenberg, 1983). Of these, only Exacum affine L , also known as Persian violet, has been commercially bred for horticultural use. However, several additional species of Exacum have been identified as possessing horticultural potential. Several of these species, native to Sri Lanka, have been collected and inter-crossed to produce interspecific populations for use in a cultivar development breeding program. These populations now form a cohesive taxonomic group named Exacum Styer Group (Riseman et al, 2005) One requirement for a new crop's commercial success is a reliable system to predict or control flower initiation and development, usually governed by a complex set of physiological interactions between the plant and the environment (Lang, 1965; Evans, 1969; Bernier et al, 1981). Unfortunately, this information is not available for E. Styer Group. However, for the related species E. affine, flowering was positively affected by both warm temperatures and high total cumulative light (Holcomb and Craig, 1983). In order to develop a reliable production system for E. Styer Group, it is essential to know taxon-specific environmental effects on floral initiation and development. Typically, the transition from a vegetative to a reproductive state is distinguished by 48 characteristic morphological and anatomical changes in the apical meristem which alter its physical proportions (Gould, 1966). These changes in meristem form can be mathematically described through allometric analysis (Niklas, 1994), a procedure that allows for the evaluation of sources of morphometric variation. This research focused on evaluating the roles of cumulative light and heat unit accumulation on the transition from vegetative to reproductive growth by characterizing changes in meristem structure during floral organ initiation and development. It offers an opportunity to precisely identify the time of transition and the environmental conditions required for initiation. The specific objectives of this study were: (1) to characterize anatomical and morphological traits in the vegetative apical meristem of Exacum Styer Group; (2) to characterize anatomical and morphological changes associated with floral initiation and meristem transition to reproduction; (3) to identify distinct morphological markers associated with discrete developmental stages during transition to reproduction; and (4) to relate environmental conditions (i.e., cumulative light and heat units) to floral initiation and development. 3.2 Materials and Methods Plant material Genetically related seedling populations were selected for inclusion in this study (Appendix 1). Seeds were surface sown in April 2002 on Rediearth Commercial Mix (Sun Gro Horticulture, Bellevue, WA, USA) that was previously soaked with the fungicide Benlate™, (Dupont Canada Inc, Mississauga, Ontario) at a rate of 1 mg L"1, 49 covered with a Humidome™ (Growtex, Burnaby, British Columbia) and placed into a growth room set at 24°C. Germination occurred within 15 to 20 days. Approximately two months after germination, seedlings were transferred into plug trays containing a potting mix composed of peat (70%), perlite (20%), pasteurized mineral soil (10%) and osmocote (18-6-12) (3 kg m" ), and placed under shade cloth for acclimatization. After another 2 months, each seedling was transplanted to an individual 10 cm pot containing the same potting mixture. Plants were watered as needed. Fertilizer (Plant-Prod 21-7-7 Acid fertilizer, Plant Products Co, Brampton, Ontario), at a rate of 200 ppm, was applied biweekly. Once a month, 15 g 100 L"1 Zn-EDTA chelate (Plant Food Company Inc, Cranbury, NJ, USA) was added to the fertilizer to prevent the development of zinc stress. Electrical conductivity (EC) and pH of the medium were checked monthly. The EC was maintained between 1-2.25 mS cm"1 and pH between 5-6. Standard pest control strategies using Safer's™ Natural Insecticide (Safer Ltd, Scarborough, Ontario), Ambush 50 EC™ (Zeneca Agro, Calgary, Alberta) and Benlate™ (Dupont Canada Inc, Mississauga, Ontario), were applied as needed. Light treatments Experimental light treatments, using a combination of shade netting and supplemental lighting, were as follows: (1) ambient light (AL) + 65% shade (SH65); (2) AL + 35% shade (SH35); (3) ambient light; and (4) AL + supplemental light. Supplemental light was provided by 430W High Pressure Sodium lamps (Sun Agro (Philips), Somerset, NJ, USA) and positioned to deliver PAR light of 300 pmol m"2 s"1 to 50 the top of the plant canopy. Supplemental light was provided between 9:00 pm to 5:00 am. The treatments were designed to deliver, on average, the following total quantity of light (PAR) over 36 weeks: Treatment 1: 1152 mols, Treatment 2: 2130 mols, Treatment 3: 3006 mols, and Treatment 4: 5211 mols. Light sum (PAR) (LS) values used in the allometric calculations were summed from germination to anthesis and calculated using day length and lighting period multiplied by average light intensities. Heat unit calculation Heat units were calculated using the following formula: Heat Units = (Daily Maximum Temp + Daily Minimum Temp / 2) - Threshold Temp) (Arnold, 1959). The threshold temperature is known only for a few species, and is unknown for Exacum. Therefore, I used a threshold temperature of 10°C because previous research has shown that this is a reasonable approximation for many species (Herms, 1998). The sum of the heat units (HS) from the day of germination through anthesis was used in data analyses. Plastochron cycle effects The plastochron cycle is morphogenesis at the shoot apical meristem that repeats itself at regular time intervals (Erickson and Michelini, 1957). Therefore protocols to investigate morphogenesis at the shoot apex must be able to resolve morphogenesis within a plastochron, which, in apices producing relatively large leaf primordia, usually lasts between 1-3 days (Dale and Milthorpe, 1983). In my study, apical meristems were harvested and fixed for the purpose of recording changes in meristem structure during 51 floral initiation. Since the apical meristems were destroyed, the data on plastochron cycle were not recorded. Tissue preparation Apical meristems from individual seedlings, from each population and treatment combination, were collected weekly beginning eight weeks after germination. Leaves covering the mersitem were counted prior to observation to calculate node number in relation to floral initiation. Apical meristems were then fixed for either scanning electron microscopy (SEM) or histological analysis. Fixation for SEM included treatment with 3% glutaraldehyde in 0.01 M phosphate buffer for 8 h. Post fixation treatment included placing the meristems in a 0.01 M phosphate buffer (aqueous phosphoric acid (H3P04)) with 1% Os0 4 for 12 h. Samples were then rinsed with distilled water and dehydrated with 30% ethanol. Dehydration was completed in an ethanol series (i.e., from 50-100% in 10% increments) followed by three changes of 100% ethanol. The meristems were critically point dried using liquid C O 2 , mounted on aluminium stubs, coated with 30 nm of gold in a sputter coater, and observed using a Hitachi S4700 SEM (Biolmaging Facility, UBC). Images were recorded digitally. Apical meristems used for histological observations were prepared, according to the FAA fixation protocol of Meyerowitz (1987), embedded in paraffin, stained with Eosin yellow (Sigma, Oakville, Ontario, Canada) and sectioned into 8 pirn slices. Meristem width (MW) and meristem height (MH) were measured using a light microscope (Carl Zeiss, West Germany) calibrated with an ocular micrometer (American Optical Company, New York, NY, USA) and observed at 16X magnification. 52 MW was measured across samples at the meristem's maximum width while MH was measured from base (the point at which the meristem originates) to the tip of the meristem. MW and MH values are expressed in micrometers. Allometry Analysis and Calculation I decided to use allometric statistics to characterize changes in meristem structure during floral initiation because simple regression and multiple regression techniques may not adequately identify and describe the relationship between a meristem's width and height. Allometry can provide this information as it is used specifically to analyse changes in the proportions of various parts of an organism as a consequence of growth. Moreover, allometric regression allows a researcher to include additional parameters that may influence the observed changes in proportion. Allometric analysis involves the assessment of quantitative relationships between different measures of an organism, and is commonly accomplished using the general power function, y = azp where: y and z are two different measures of growth of an organism, a is a constant, and (3 is the allometric exponent (growth ratio) where y intercepts the vertical axis (Huxley, 1932; Niklas, 1994). In my research, y represents meristem width (MW, pm), z represents meristem height (MH, pm), a indicates the size of y when z equals 1.0 (pm), and (3 relates the proportionality between y and z. For an individual at some point in its life cycle, this is an exact relationship. However, for a population of non-identical plants, additional variation exists which can be accommodated for through variation in a, (3, and though the addition of a third parameter, e, which accounts for residual variation in y not associated with its allometry 53 with z- The time interval between germination and day of collection (i.e., meristem harvest day) was calculated in days, t. Meristem width and height data (i.e., y and z) were loge (In) transformed prior to analyses because log transformation of data tends to diminish the differences among large numbers and accentuate the differences among small numbers and is a simple way to explain proportionality. In the log transformed frame of reference, the general power function becomes: Eq. (1): ln(y) = ln(cx) + pTn(z) + ln(e) This model accounts for both environmental and developmental changes on y and therefore, are expressed through their influences on z and the parameters a, P and e. However, these influences can be more fully explained by expanding the loge In-transformed allometric power function to include three additional sources of variation, days (t), heat sum (HS), and light sum (LS), and their interactions with each other. The expanded allometric equation is as follows (Jolliffe et al,1988): Eq. (2): In (y) = In (a) + p 0 ln(z) + p, t ln(z) + p 2 LS ln(z) + p 3 HS ln(z) + p 4 tLS ln(z) + p 5 tHS ln(z) + p 6 HSLS ln(z) + p 7 tHSLS ln(z) + y, ln(t) + y2 ln(LS) + y3 ln(HS) + y4 ln(tLS) + y5 ln(tHS) + y6 ln(HSLS) + y7 ln(tHSLS) + ln(e) Given suitable data, values can be assigned to parameters a, Pk&yk in Eq.(2) using regression. Using this regression technique, a treatment effect on y can be 54 associated with adjustments in the allometric exponent reflecting changes in the proportionality between y and z (Pi-7), and whether they arise from direct effects or from a combination of treatment effects on y without involving changes in y:z proportionally (71-7). Allometric relationships were established using the best subset multiple regression procedure. Best subset regression is a method to determine which independent variables are the most appropriate to include in the multiple regression model (Daniel and Wood, 1971). In establishing allometric relationships, Eq. (1) was used while for the best subset multiple regressions, Eq. (2) was used. In addition, the variable subset with the lowest Mallow Cp combined with the lowest variance inflation factor (VIF) value was included in model construction. Mallow Cp is the measure of the difference of a fitted regression model from a true model, including random error. The average value of Cp is p+1, where p is number of parameters. VIF measures the multicollinearity and inflation of the variance for an independent variable due to redundant information in other independent variables. Data were analyzed using SigmaStat v. 2.03 (SPSS Inc, Chicago, IL.) software and used to predict the relationships between y and z variables. Linear regression (i.e., ln(y) = ln(a) + P ln(z) + ln(s)) was performed separately for each family. The best subset regression was performed to select the variables that best contributed to predicting the dependent variable, y. The selected subset was then tested with multiple linear regression analyses to obtain partial regression coefficients and standard partial regression coefficients. The partial regression coefficient relates how the dependent variable changes when one independent variable is increased by one unit of measurement while all other 55 independent variables are held constant. The purpose is to measure the amount of variation (R ) that can be attributed to one or a set of independent variables. 3.3 Results Morphological characterization Time-course sections show the progression of Exacum flower initiation and development (Figs. 3.1-3.2). This developmental process can be divided into six relatively distinct stages. In Stage 1, the floral meristem is dome to conical shaped with a height to width ratio of 0.90 (Fig. 3.1 A-B). At this stage, it produces leaf primordia as well as numerous rudimentary axillary meristems (Fig. 3.1 C-D). The apical dome becomes flattened as development progresses from Stage 1 to Stage 2 with the height to width ratio reducing to 0.43. During Stage 2, the apex gradually occupies proportionally less of the meristematic region as floral initiation commences and successive floral parts are formed (Fig. 3.1 E-F). With the onset of floral initiation, the apex becomes broadened with increased mitotic divisions throughout the apex, particularly in the central region (Fig. 3.2 A-B). Concurrent with this enlargement, five-sepal primordia are initiated around the apex indicating the start of Stage 3 (Fig. 3.2 A-B). As floral development progresses to Stage 4, the apex continues to flatten and broaden (e.g., height to width ratio -0.09) with five petal primordia arising simultaneously alternate to the sepals (Fig. 3.2 C-D). Stage 5 is characterized by stamen initiation with first appearance as short structures alternate to the petals (Fig. 3.2 D-E). Finally, Stage 6 is reached by the first appearance of the ovary primordium (Fig. 3.2 E-F). There was considerable variation in the timing of bud initiation both within and among the 56 populations. However, regardless of the timing of the transition to reproduction, the morphological characters described were consistent across all individuals. Light treatments There was significant variation among the populations evaluated for floral initiation with some populations unaffected by treatments while others significantly affected (Table 3.1). The earliest population (02-21) developed from vegetative to reproductive growth within 12 wks from germination regardless of treatment. In addition, transition typically occurred at node eight. In total, 18-20 weeks were required to reach the visible bud stage and a total 28-30 wks to reach anthesis. In contrast, population 02-39, while the second earliest population to initiate flower buds, was significantly affected by treatment. For example, the apical meristem of all individuals from the supplemental light treatment transitioned to a floral apex in 16 wks and at node eight. However, in all other treatments, individuals required 22 wks to transition with the floral apex typically at node ten. These differences in transition time were consistent through the dates of first visible bud and anthesis. All other populations required 18-22 wks to transition, irrespective of treatment, with transition occurring either at node nine or ten. Heat units In this experiment, temperature was the same on all light treatments. Results indicate heat units did not have a significant influence on the flowering of Exacum. 57 Allometric analyses Simple allometric relationships, as calculated by Eq. (1), varied among populations with the strength of the relationships reflected in the coefficients of determination values. These values are always positive and range between 0-1. Values close to 1 confirmed variation in the dependent variable was contributed to by the independent variables listed (Table 3.2). R 2 values ranged between 0.980 (population 02-62) and 0.468 (population 02-65) (Table 3.2) and values less than 1.00 indicate the presence of unexplained variation that was not accounted for by Eq. (1). This may be due to the genetic variation inherent in these populations. Because the accessions used were only genetically related and not genetically identical, total variability observed in the relationship between y and z is composed of interactions between both the environment (i.e., treatments) and genotype (i.e., both within and among individual populations). Therefore, additional components to the model were required as detailed in Eq. (2) that accounted for these additional sources of variation. The best subset regression procedure was performed with the expanded allometric model (Eq. (2)) to select the variables that best contributed to predicting the dependent variable, meristem width (y), and included total cumulative light (LS), heat unit (HS), days (t), and their combinations. This expanded model produced greater R2 values, ranging between 0.994 (population 02-62) and 0.848 (population 02-141) (Table 3.3), thereby accounting for additional variability beyond the ability of Eq. (1). For example, the best subset model for the population with the strongest allometric relationship (i.e., 02-65), displayed a low Cp value (1.200), a low VIF value (1.631), a high R 2 (0.945) and a probability of P<0.05 (Table 3.3). The low VIF value confirmed 58 that multicollinearity was not strong. Similarly, the low Cp value indicated that it is unlikely that a relevant variable was omitted during model construction. In contrast, the population with the weakest allometric relationship (i.e., 02-141) displayed a high Cp value (3.000) coupled with a high VIF value (1.728) and the lowest R 2 (0.848) value. The expanded allomeritc models clarified the roles of the experimental factors light sum (LS), heat unit (HS), and days (t) on meristem width (MW) and floral initiation. However, because these factors are in different scales of measurement, their influences are not comparable using the model coefficient values. However, direct comparisons can be made using the magnitude of the standard partial regression coefficients for each term in the best subset models. Using this method, the direct roles of light sum and days on MW can be clearly illustrated by reviewing the standard partial regression coefficients (Table 3.3). For example using the parameter light sum, 11 of 13 populations displayed a significant standard partial regression coefficient ranging from 0.078 to 1.169. Also, standard partial regression coefficients for parameter days (t) were significant in 7 of 13 populations (Table 3.3). When the simple linear regression models from Eq. (1) (Table 3.2) are compared to the models from Eq. (2) (Table 3.3), the increase in R values support the enhanced ability of Eq. (2) to better explain the variation observed in the dependent variable, meristem width and floral initiation. Significant differences were observed among populations for the parameters that describe the allometric relationship where some populations displayed the relationship via the 3 parameter while others displayed it via the y parameter. In the allometric analyses, the parameters Po, Pi.... P7 were the allometric coefficients and represented ln(z), t ln(z), LS ln(z), HS ln(z), tLS ln(z), tHS ln(z), HSLS ln(z), and tHSLS ln(z), 59 respectively. In addition to the general observation that most populations displayed allometric relationships through the direct effects of either light sum (ln(LS)) or days (ln(t)), 3 populations (02-60, 02-61 and 02-62) displayed significant effects via P5, P6, and P7 indicating there were significant interactions between these allometric exponents and treatment factors (Table 3.3). For example, in population 02-62, the allometric coefficient p7 (tHSLS ln(z)) illustrates that days (t), heat sum (HS), and light sum (LS) act through variable z with a standard regression coefficient of 0.151 (Table 3.3). However, in population 02-61, the standard regression coefficient values were not significant, thus indicating no allomertic relationships were present within those data. The terms containing yk (Eq. (2)) express the relationship between the experimental treatments and the variation in MW, which may arise in two ways. First, there may be direct effects of treatments on MW, independent of allometry, via the residual term as calculated in Eq. (1) (Table 3.3). However, another possibility is that the terms containing yk in Eq. (2) may reflect treatment effects on the allometric coefficient (a). Our data indicate that the yk parameters, specifically terms yi days (ln(t)) and j2 light sum (ln(LS)), had significant and large effects on the variable meristem width (ln(y)). Expression of treatment effects via the yk parameters indicate that they were not associated with the allometric exponent (Pk). For most populations, multiple regression analyses produced the highest standard regression coefficient values for light sum followed by days (Table 3.3). This indicates that these independent variables directly contributed to the variance values calculated for variable y, meristem width. However, this relationship was not consistent among all populations. In populations 02-141, 02-171 and 02-174, only light sum contributed to 60 all variance components of variable y while in the remaining populations, both light sum and days contributed. The strong effects of days and light sum expressed in all models support that these variables significantly affect a meristem's transition to reproduction. In addition, all models supported the significant effects for the parameters representing a, (3 and y indicating that the independent variables light sum and days can be used to predict the dependent variable, meristem width. 3.4 Discussion The shift from vegetative to reproductive growth is termed 'floral transition' and is commonly controlled by environmental signals such as temperature (Kinet et al, 1985), light intensity or cumulative light (Kinet et al, 1985), and/or photoperiod (Hammer and Bonner, 1938). In general, light intensity impacts the quality and rate of floral development while cumulative light impacts initiation (Kinet et al, 1985). The shoot apical meristems of all Exacum Styer Group accessions exposed to supplemental light (considered a cumulative light treatment since exposure occurred in the evenings) developed into floral apices significantly earlier than all other treatments, with an average initiation time of 12 weeks and occurring at node eight. This pattern and developmental rate were consistent through first visible bud and anthesis. Higher cumulative irradiance was also reported to promote earlier floral initiation in two additional Gentianaceae, E. affine L , (Holcomb and Craig, 1983) and Eustoma grandiflorum L , (Zaccai and Edri, 2002). However, interspecific Exacum grown under 100 pmol m"2 s"' supplemental lighting did not initiate flowers earlier than control plants but only had accelerated floral development (Anon, 1994). 61 The transition to reproduction is highly controlled and progresses through a set of well-defined morphological stages, from changes in meristem morphology and plant architecture through visible bud and anthesis. The time between observable changes in plant architecture (i.e., increased stem elongation or branching intensity) and changes in meristem morphology is termed 'floral evocation' (Evans, 1969; McDaniel et al., 1991). The first gross morphological change in Exacum Styer Group indicative of floral initiation is interrnode elongation or bolting (Anon, 1994). However, this occurs after the first observable change in meristem morphology. In Exacum, the development of a reproductive meristem is similar to many other angiosperm species previously evaluated and can also be divided into six distinct stages (Bernier et al., 1981; Bernier, 1988; Kinet et al., 1985). Despite significant variation in the time of transition in E. Styer Group individuals, the morphological characters observed were consistent across all accessions, populations, and treatments. This emphasizes the conclusions of Wetmore et al., (1959) that the essential features of.transformation from vegetative to reproductive growth were the same regardless of the species or ultimate morphology of the mature reproductive structure. Initial morphological/anatomical signs of transition from vegetative to reproductive phase typically begin with a broadening of the apex and an increase in mitotic divisions (Steeves and Sussex, 1989). These changes ultimately result in the consumption of the apical meristem by the production of floral meristems and the subsequent development of floral organs (Hicks and Sussex, 1971; McDaniel et al., 1985). Observations on E. Styer Group are consistent with these generalities in that the first sign of transition was a broadening of the meristem followed by complete 62 replacement of the vegetative meristem by developing floral organs as transition progressed from Stage 1 (i.e., vegetative meristem) to Stage 2 (i.e., sepal initiation). Plants commonly undergo significant changes in morphology, anatomy, and process as they mature due to many interacting factors. To measure these changes, various mathematical techniques are available with regression analyses commonly used to establish the form of functional relationships between variables (Niklas, 1994). In this research, I evaluated whether an allometric relationship existed between meristem morphology and environmental factors. The allometric equations developed for this study were able to explain a significant amount of the variation observed in meristem morphology. Use of a simple allometric equation displayed significant allometry between meristem width and meristem height. However, use of an expanded allometric equation explained more of the variation present by increased partitioning of the variance between direct treatment effects, allometry with z, treatment effects on allometric relationships, and residual variation. In addition, it was able to demonstrate that a linear relationship exists between the two features measured (i.e., meristem width and height). An allometric relationship between any two characters of a plant expressed across treatments or environments is interpreted as a type of plasticity (Weiner and Thomas, 1992). In this research, allometric analysis helped explain part of the phenotypic plasticity observed as being due to environmental factors. When the expanded allometric power function equation was used, 10 of 13 families showed that treatment effects significantly contributed to MW, either through direct treatment effects on allometry or through allometry with z. 63 3.5 Conclusion My findings indicate that changes in morphological characters during flower initiation and development were consistent across all populations in Exacum Styer Group. These populations required 18-20 weeks to reach visible bud stage and 28 to 30 weeks to reach anthesis. Allometric analysis with the expanded model produced high coefficient of determination (R2) values, low variance inflation factor values (VIF) and low Mallow Cp values for all populations. Direct comparisons were made using the magnitude of standard partial regression coefficients for each term. Standard partial regression coefficients were obtained for 7 out of 13 populations for the parameter days (ln (t)) (yi); 11 out of 13 for the parameter light sum (ln (LS)) (y2) and 2 out of 13 for the parameter heat sum (ln (HS)) (73). These indicate that the parameter light sum plays most significant role in the development process of flower initiation followed by days. In the expanded allometric model, values for Pi.7 terms were not significan for 9 out of 13 populations, and may be due to correlation with the parameter 'days' or 'light sum' . The majority of populations (9 out of 13) did not exhibit a significant relationship with ln(z) (Po) indicating that meristem height is not a major driver in changes in meristem width. Overall, this study indicates that there are direct affects of light sum and days on meristem width. 64 Table 3.1 Cumulative moles of light at time of meristem transition, average days to visible bud and average days to anthesis for light treatments across populations. Treatment N Cumulative moles of light at time of meristem transition (mols m" ) Initiation D - V B D-A Days to visible bud; Mean (SD)Z Days to anthesis Mean (SD)z AL+S 210 2049 4046 5072 165 (41.6)a 238 (39.2)a AL 200 1392 2984 3065 215 (51.9) b 273 (39.1)b AL+SH35% 200 1077 1920 1993 216 (50.8) b 278 (36.9) b AL+SH65% 194 724 1027 1173 209 (51.0) b 272 (39.6) b Within column entries with same letter superscript are not different by Tukey's Protected LSD at P<0.05. Table 3.2 Parameters and statistics of simple allometric regressions using ln(y) Pln(z) + ln(e), where y equals meristem width and z equals meristem height 65 = ln(a) + Population a P Std. Error a p Adjusted R 2 02-21 2.843t -0.455f 0.0556 0.0507 0.706 02-39 2.6161 -0.724f 0.0455 0.0530 0.849 02-59 2.850t -0.382f 0.0357 0.0315 0.816 02-60 3.095f -0.496f 0.0472 0.0408 0.816 02-61 2.994f -0.773f 0.0449 0.0440 0.903 02-62 2.658t -0.492| 0.0110 0.0123 0.980 02-64 2.769f . -0.488f 0.0528 0.0555 0.698 02-65 3.0341 -0.6191 0.1200 0.1130 0.468 02-77 2.6801 -0.4501 0.0121 0.0150 0.687 02-125 3.1471 -0.589f 0.0534 .0.0470 0.825 02-141 3.069f -0.466f 0.0458 0.0472 0.746 02-171 2.847f -0.68 I f 0.0351 0.0350 0.919 02-174 2.818f -0.519t 0.0365 0.0363 0.755 Table 3.3 Standard partial regression coefficients, Mallow Cp, variance inflation factor (VIF) and R2 from the best subset regression tests for allometric relationships between meristem width and meristem height among Exacum Styer Group populations. Independent Variables Parameters Populations and their standard partial regression coefficients 02-21 02-39 02-59 02-60 02-61 02-62 02-64 02-65 02-77 02-125 02-141 02-171 02-174 Intercept d -8.336 -10.903 -5.394 3.156 -13.360 -2.660 -3.460 -8.100 -7.250 -5.678 -1.513 0.332 -12.068 ln(z) Po -0.839 -0.140 -0.325 -0.546 tln(z) Pi LS ln(z) P2 HS ln(z). P3 tLS ln(z) P4 tHSln(z) P5 ^ 0.528 HSLS ln(z) P6 0.078NS tHSLS ln(z) P7 0.151 ln(t) Yi 0.774 0.728 0.864 0.888 1.169 0.821 0.892 ln(LS) Y2 0.265 0.327 0.164 0.339 -0.278 0.867 0.205 0.127 0.798 0.846 0.868 ln(HS) Y3 0.658 0.155 ln(tLS) Y4 ln(tHS) Y5 n(HSLS) Y6 ln(tHSLS) Y7 Mallow Cp 3.211 3.441 3.452 1.267 3.012 1.955 1.566 1.200 4.629 1.701 3.000 3.000 2.017 VIF 2.236 2.236 2.236 3.288 1.406 1.880 2.236 1.631 2.236 2.236 1.728 2.591 2.229 0.973 0.990 0.985 0.994 0.957 0.994 0.961 0.945 0.967 0.980 0.848 0.944 0.977 67 Figs. 3.1 Time section and SEM images of Exacum Styer Group apical meristems during the transition from vegetative to reproductive growth (Stages 1 and 2). Stage 1 - domed to conical shaped meristem with leaf primordia, and numerous rudimentary axillary meristems visible (A-D); Stage 2- broadened meristem with the commencement of floral initiation with leaf primordia visible (E-F). 1 bar = 0.1 pm SAM-Shoot apical meristem; LP-Leaf primodia; FA-Floral apex; Se-Sepal 6 8 Figs 3.2 Time section and SEM images of Exacum Styer Group apical meristems during the transition from vegetative to reproductive growth (Stages 3 through 6). Stage 3- Broadening of meristem with the initiation of five-sepal primordia around the apex (A-B); Stage 4- initiation of five petal primordia alternate to the sepals represents (C); Stage 5- stamen initiation appearing as short structures alternate to the petals (D-E); Stage 6- initiation of the ovary (F) lbar =0.1 pm. Se-Sepal; FM-Floral meristem; Pe-Petal; Ca- Carpel; St-Stamen 69 3.6 References Anon, K.M. 1994. Environmental studies on the growth and flowering of interspecifc hybrids of Exacum species (Gentianaceae) endemic to Sri Lanka. Master of Science Thesis. Department of Horticulture, The Pennsylvania State University. Pp 77-117. Arnold, C.Y. 1959. The determination and significance of the base temperature in a linear heat unit system. Proceedings of the American Society of Horticultural Science 74: 682-692. Bernier, G. 1988. The control of flower evocation and morphogenesis. Annual Reviews of Plant Physiology and Plant Molecular Biology 39:175-219. Bernier, G., Kinet, J and Sachs, R.M. 1981. The physiology of flowering. CRC Books, Boca Raton, Fla. Daniel, C and Wood, F.S. 1971. Fitting equations to data: computer analysis of multifactorial data for scientist and engineers. Wiley-Interscience, New York. Pp 342. Dale J.E and Milthorpe, F.L. 1983. General features of the production and growth of leaves. In: Dale JE, Milthorpe FL, eds. The growth and functioning of leaves. Cambridge: Cambridge University Press, 151-178. 70 Erickson, R. O and Michelini, F.J. 1957. The plastochron index. American Journal of Botany 44: 297-305. Evans, L.T. 1969. The induction of flowering: The nature of flower induction. Ithaca, New York, Cornell University Press. Gould, S.J. 1966. Allometry and size in ontogeny and phylogeny. Biological Reviews 41: 587-640. Hammer, K.C and Bonner, J. 1938. Photoperiodism in relation to hormones as factors in floral initiation. Botanical Gazette 100: 388-431 Herms, D.A. 1998. Ornamental plants - Annual Reports and Research Reviews. Special Circular 165-99. The Ohio State University, Columbus, OH, U.S.A. Hicks, G.S and Sussex, I.M. 1971. Organ regeneration in sterile culture after median bisection of the flower primordia of Nicotiana tabacum. Botanical Gazette 132: 350-363. Holcomb, E.J and Craig, R. 1983. Producing Exacum profitably. Greenhouse Grower 1: 18-57. 71 Huxley, J.S. 1932. Problems of relative growth. New York, L. MacVeagh, The Dial Press, Methuen, London. Jolliffe, P.A., Eaton, G.W and Potdar, M.V. 1988. Plant growth analysis: Allometry, Growth and interference in orchard grass and Timothy. Annals of Botany 62: 31-42 Kinet, J.M., Sachs, R.M and Bernier, G. 1985. The physiology of flowering: the development of flowers. CRC Books, Boca Raton, Fla. Klackenberg, J.1983. A reevaluation of the genus Exacum (Gentianaceae) in Ceylon. Nordic Journal of Botany 3: 355-370. Lang, A. 1965. Encyclopedia of plant physiology: Physiology of flower induction. Springer, Berlin. McDaniel, C.N., King, R.W and Evans, L.T. 1991. Floral determination and in vitro floral differentiation in isolated shoot apices of Lolium temulenum. L. Planta 185: 9-16. McDaniel, C.N., Singer, S.R., Dennin, D.A and Gebhardt, J.S. 1985. Floral determination: Stability, timing and root influence. M. Freeling, Editor. Proceedings of UCLA Symposium, Plant Genetics, Los Angeles, CA. 72 Meyerowitz. 1987. In situ hybridization to RNA in plant tissue. Plant Molecular Biology Reporter 5: 242-250. Niklas, K.J. 1994. Plant Allometry, the scaling of form and process. University of Chicago Press, Chicago. Riseman, A , Sumanasinghe, V . A , Justice, D and Craig, R. 2005. New Name 'Styer Group' Proposed for Interspecific Hybrids of Exacum Species Native to Sri Lanka. HortScience 40: 1580-1583. Steeves, T.A and Sussex, I. 1989. Patterns in plant development. Cambridge, New York. Cambridge University Press. Weiner, J and Thomas, S.C. 1992. Competition and allometry in three species of annual plants. Ecology 73: 648-656. Wetmore, R.H, Gifford, E.M and Green, M.C. 1959. Photoperiodism and related phenomena in plants and animals: Proceedings. R.B.Withrow, Editor. American Association for the Advancement of Sciences. Publications number 55. Washington. Zaccai, M and Edri, N. 2002. Floral transition in lisianthus (Eustoma grandiflorum). Scientia Horticulturae 95: 333-340. 73 CHAPTER 4 Conclusions and Recommendations for Further Work Flower initiation and development in Exacum Styer Group were characterized in an effort to identify the environmental conditions responsible for floral initiation. Significant variation in flowering time was present for both individuals and populations. This high level of variation was likely due to the level of heterozygosity present in these populations due to their interspecific pedigrees. Light sum was identified as the primary environmental factor that influences floral initiation. However, another yet unidentified factor acts through node number to influence flower initiation and either acts independently or interacts with light sum. Further research is required to more precisely identify all the factors that influence flower initiation in Exacum Styer Group. In this experiment, the effects of photoperiod on flower initiation were not studied. The most reasonable factor to be researched in the future is temperature and its interaction with light sum and may be photoperiod. A series of experiments that include treatments that allow a more accurate assessment of these factors are necessary. In addition, future research should try to reduce the level of genetic variation introduced into the experiments. This can be achieved by either developing homozygous lines via inbreeding or by using clonally propagated materials. 74 Appendix 1 Appendix 1. Population identification and parentage. For parentage designations, the pistillate parent is listed first followed by the staminate parent. R = Reciprocal Parents Population Identification Parentage 02-62 01-42-7x01-12-13 02-64 01-42-7x01-36-1 02-65 01-42-7 x 01-36-11 K 02-60 01-42-7 selfed 02-61 01-42-7 x 01-9-1 02-141 01-68-10x01-69-28 02-77 01-46-32 x 01-42-7 02-171 01-83-7x01-53-17 02-174 01-83-7 x 01-6-28 02-21 01-24-2x01-24-1 02-125 01-50-42x01-37-40 02-39 01-36-11 x 01-42-7 K 02-59 01-37-68x01-68-10 Appendix 2 75 Effects of Plant Growth Regulator Treatments on Floral Initiation on Exacum Styer Group Research Objectives This research investigated the effects of plant growth regulators (especially anti-gibberellins) on flowering of Exacum Styer Group genotypes. The specific objectives of this study were: 1. to evaluate the effects of abscisic acid on floral initiation and development; 2. to evaluate the effects of chlorocholinchloride on floral initiation and development; 3. to evaluate the effects of ancymidol on floral initiation and development. Materials and Methods Plant Material Thirteen genetically related seedling populations were selected for inclusion in this study (Appendix 1). Seeds were surface sown in April on Rediearth Commercial Mix (Sun Gro Horticulture, Bellevue, WA, USA), that was previously soaked with 1 mg L"1 Benlate™ fungicide (Dupont Canada Inc., Mississauga, Ontario), covered with a Humidome™ (Growtex, Burnaby, British Columbia) and placed into a growth room maintained at 24°C and a 16 h photoperiod provided by cool-white fluorescent lamps 2 1 (120 pmol rn s" ). Germination occurred within 15-20 days. Approximately two 76 months after germination, seedlings were individually transplanted into plug trays containing a potting mix composed of peat (70%), perlite (20%), pasteurized mineral soil (10%) and osmocote (18-6-12) (3 kg m"3). The seedlings were placed on greenhouse benches, located in the Horticultural Greenhouse, University of British Columbia, and covered by shade cloth until acclimated (approximately three weeks). After 2 months, seedlings were transplanted into individual 10 cm pots containing the same potting mixture as described above. Pots were placed on greenhouse benches with the aerial temperatures maintained between 20-22°C and the humidity maintained between 85-90%. Plants were watered as needed and fertilized (Plant -Prod 21-7-7 acid fertilizer, Plant Products Co, Brampton, Ontario) biweekly with 200 ppm total nitrogen. Once a month, 15 g 100 L"1 EDTA Zn chelate (Plant Food Company Inc, Cranbury, NJ, USA) was added to the fertilizer solution to prevent the development of zinc stress. Electrical conductivity and pH were checked monthly and maintained between 1-2.25 mS cm" and 5-6, respectively. Standard pest control strategies using Safer's™ Natural Insecticide (Safer Ltd, Scarborough, Ontario), Ambush 50EC™ (Zeneca>Agro, Calgary, Alberta) and Benlate (Dupont Canada Inc, Mississauga, Ontario), were applied as needed. Growth Regulator Treatments The following anti-GA compounds were evaluated: abscisic acid (ABA), chlorocholinchloride (CCC) (BASF Canada, Mississauga, Ontario), and ancymidol (A-REST™) (Plant products co, Brampton, Ontario). The concentration used for all 77 compounds was 10"4 M. Treatments were applied directly to the apical meristem by injecting 0.2 pi treatment solution with a 1 cc (1ml) syringe weekly. One cc syringe have markings showing tenths of a cc (0.10 is one-tenth of a cc; 1.0 is one cc). Control plants were treated as above but with sterilized distilled water. Results Population effects on flowering Significant effects (P<0.05) were observed for days to visible bud and days to anthesis for population, treatment, and their interaction (Appendix 2. Table 1). Average (i.e., across all treatments) population performance (i.e., days to anthesis) ranged from 263 days (population 02-39) to 285 days (population 02-65) (Appendix 2. Table 2). Treatment effects on flowering All anti- GA treated plants, regardless of the compound tested, delayed flowering by more than ten days compared to the control plants (Appendix 2. Table 3). The maximum delay was observed in the CCC treatment, where an additional 26 days were required to reach anthesis as compared to the control plants. Conclusions All anti-GA treatments significantly delayed floral initiation. However, no plant growth regulator treatment prevented floral initiation. Further research is required to, better understand the mode of action of these compounds and to better formulate strategies to prevent floral initiation. 78 Appendix 2 Table 1. Summary of analysis of variance for days to visible bud (G-VB) and days to anthesis (G-A) among Exacum Styer Group populations when treated with various anti-GA compounds. G-VB DF SS MS F P Populations 12 12886 1073 4 <0.001 Treatment 3 57289 19096 74 <0.001 Populations * Treatment 36 26012 722 2 <0.001 Residual 468 119501 255 G-A Populatons 12 9455 787 3 <0.001 Treatment 3 58233 19411 84 <0.001 Populations * Treatment 36 23784 660 2 <0.001 Residual 468 107467 229 / 79 Appendix 2 Table 2. Population mean and standard deviation (SD) for days to visible bud (G-VB) and anthesis (G-A) averaged across anti-GA treatments. Population G-VB Mean SD G-A Mean SD 02-21 187 18.5 268 19.8 02-39 184 22.4 263 24.5 02-59 190 19.0 266 24.3 02-60 188 21.9 268 21.1 02-61 196 17.6 272 15.7 02-62 196 15.1 270 18.4 02-64 193 16.7 270 13.7 02-65 199 22.5 285 16.7 02-77 187 12.8 273 15.2 02-125 211 18.1 272 12.2 02-141 211 22.0 268 16.5 02-171 218 19.9 264 17.6 02-174 180 15.4 269 17.6 80 Appendix 2 Table 3. Anti-GA treatment effects on days to visible bud (G-VB) and anthesis (G-A) averaged across populations. Treatments count G-VB Mean 2 SD G-A Mean z SD Control 130 189 b 16.1 255 b 17.2 ABA 130 213 b 14.3 277 b 14.6 CCC 130 215 a 14.0 278 a 13.9 Ancymidol 130 202 b 17.5 266 b 17.7 z Within column entries with same letter superscript are not different by Tukey's significant difference test.
UBC Theses and Dissertations
Role of environmental factors and plant growth regulators on the floral transition of Exacum Styer group Krishnasamy, Rani 2007
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
- 831-ubc_2007-0145a.pdf [ 4.44MB ]
- JSON: 831-1.0100686.json
- JSON-LD: 831-1.0100686-ld.json
- RDF/XML (Pretty): 831-1.0100686-rdf.xml
- RDF/JSON: 831-1.0100686-rdf.json
- Turtle: 831-1.0100686-turtle.txt
- N-Triples: 831-1.0100686-rdf-ntriples.txt
- Original Record: 831-1.0100686-source.json
- Full Text
Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url: