The Flowering Response of the Rice Plant to Photoperiod

The Flowering Response of the Rice Plant to Photoperiod A REVIEW OF THE LITERATURE FOURTH EDITION 1985 Los Banos, Laguna, Philippines Mail Address: P. O. Box 933, Manila, Philippines THE INTERNATIONAL RICE RESEARCH INSTITUTE First printing 1969 Partially revised 1972 Revised 1976 Revised 1985 The International Rice Research Institute (IRRI) was established in 1960 by the Ford and Rockfeller Foundations with the help and approval of the Government of the Philippines. Today IRRI is one of the 13 nonprofit international research and training centers supported by the Consultative Group for International Agricultural Research (CGIAR). The CGIAR is sponsored by the Food and Agriculture Organization (FAO) of the United Nations, the International Bank for Reconstruction and Development (World Bank), and the United Nations Development Programme (UNDP). The CGIAR consists of 50 donor countries, international and regional organizations, and private foundations. IRRI receives support, through the CGIAR, from a number of donors including: the Asian Development Bank, the European Economic Community, the Ford Foundation, the International Development Research Centre, the International Fund for Agricultural Development, the OPEC Special Fund, the Rockefeller Foundation, the United Nations Development Programme, the World Bank, and the international aid agencies of the following governments: Australia, Canada, China, Denmark, France, Federal Republic of Germany, India, Italy, Japan, Mexico, Netherlands, New Zealand, Norway, Philippines, Saudi Arabia, Spain, Sweden, Switzerland, United Kingdom, and United States. The responsibility for this publication rests with the International Rice Research Institute. Copyright @ International Rice Research Institute 1986 All rights reserved. Except for quotations of short passages for the purpose of criticism and review, no part of this publication may be reproduced, stored in retrieval systems, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without prior permission of IRRI. This permission will not be unreasonably withheld for use for noncommercial purposes. IRRI does not require payment for the noncommercial use of its published works, and hopes that this copyright declaration will not diminish the bona fide use of its research findings in agricultural research and development. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of IRRI concerning the legal status of any country, territory, city, or area, or of its authorities, or concerning the delimitation of its frontiers or boundaries. ISBN 971-104-151-0 CONTENTS Foreword Introduction 1 Rice as a Short-day Plant 1 Growth Phases 2 Basic Vegetative Phase 4 Photoperiod-Sensitive Phase 5 Photoinductive Cycles 7 Reception of the Photoperiodic Stimulus and Translocation 9 Light Intensity and Quality 9 Interruption of the Dark Period 11 Days from Photoinductive Treatment to Flowering 12 Biochemical Changes During Photoinduction 12 Effect of Temperature on the Flowering Response to Photoperiod 13 Measurements and Methods of Testing Photoperiod Sensitivity 14 Date-of-Planting Experiments 15 Ecology and Photoperiodism 17 Terminology Used in Describing Photoperiod Sensitivity 20 Inheritance of Vegetative Growth Duration 23 Problems in the Study of the Rice Plant? fs Photoperiodism 25 Summary 26 Appendix 28 Bibliography 38 Foreword This review, first published in 1969, has been an important reference in understanding the rice plant. It has had a small but continuing demand. Many new reports on the flowering response of the rice plant have been published since the first edition. More than 100 publications were included in the third edition; this edition includes another 103 publications. For ease of reading, numbers have been used to cite the references. This review was prepared with the cooperation of the IRRI Library Staff and the technical assistance of Mr. Romeo M. Visperas, and edited by Ms. Emerita P. Cervantes. M. S. Swaminathan Director General Introduction Photoperiod influences several aspects of plant growth. Some of its effects on rices have been reviewed by Best (24), Gwinner (111), Katayama (192), Morinaga (316), Sircar (439), and Wagenaar (534). This review is primarily concerned with the effect of photoperiod on the flowering of the rice plant. It includes more than 500 papers on the photoperiodism of rice, most of which are available at the International Rice Research Institute library. Several contributions in Japanese have been translated into English and also are available at the International Rice Research Institute library. A bibliography is given at the end of this review; not all papers listed were cited in this review but were nevertheless included as future references for interested workers. Rice as a short-day plant Rice is sensitive to photoperiod . long-day treatments can prevent or considerably delay its flowering. Rice cultivars exhibit a wide range of variation in their degree of sensitivity to photoperiod (87, 254, 319, 357, 531, 563). Figure 1 shows these variations, ranging from the very sensitive to the nearly insensitive. 1. Response curves of three representative types of rice cultivars. 2 The flowering response of the rice plant to photoperiod Most of the wild species of Oryza and many of the primitive cultivated rices ( O. sativa L. ) are photoperiod sensitive and may be classified as short-day plants. Most papers agree on such a classification, and therefore in this review, rice will be considered as a short-day plant. It also will be classified into photoperiod-sensitive and photoperiod-insensitive types, the latter showing a low response or a slight delay in flowering with an increase in photoperiod. The present tendency is to select photoperiod-insensitive cultivars so that most of the cultivated rices may eventually become photoperiod-insensitive ones. These improved, early maturing cultivars may fit into the multiple cropping system characteristic of progressive agriculture. There have been reports of cultivars whose flowering is delayed by short-day treatments and hence are considered long-day plants (1, 98, 99, 239, 254, 276, 277, 279, 283, 284, 287, 291, 303, 398,443,444, 488). Heenati, for instance, is often referred to in the literature as a long-day plant (1). Short photoperiods have delayed its flowering by 10 d, but this delay is relatively short and may be the result of nonphotoperiodic factors, such as low light intensity or relatively high temperature. The delay caused by short-day treatments ranged from 7 to 12 d in the Charnock and Panbira cultivars using an 8-h photoperiod (443), about 9 d in B. 76 (303), and 13 d in T. N. 32 and T. A. 64 (287). Many of the reported long-day and intermediate cultivars were found to be short-day cultivars in subsequent testing (522). The apparent long-day reaction of Heenati resulted from using photoperiods shorter than the optimum, which delayed flowering (34). Some rices may have been classified as long-day plants because inadequate facilities were used in testing the photoperiod reaction. The range of photoperiods used has been limited, usually involving only two treatments. In some instances, the classification was based on field reaction to different planting dates (98). Short-day-treated plants were often compared with plants grown under natural day lengths (291, 303, 304). The difference and changes in temperature and the photoperiods used have made it difficult to interpret the data intelligently. As will be discussed later, many photoperiod response curves show that photoperiods longer or shorter than the optimum delay the flowering of photoperiod-sensitive cultivars (34, 513). Photoperiod response differs markedly among rices; this also explains the diversity of the results reported on the photoperiodism of the rice plant (see Appendix). However, more than 400 cultivars have been critically tested at IRRI (l59, 160, 161, 162, 163, 164, 166, 167, 168, 169, 170), and not one so far has shown a long-day response. Growth phases The growth of the rice plant can be divided into three stages: 1) the vegetative growth phase, from germination to panicle initiation; 2) the reproductive phase, from panicle initiation to flowering; and 3) the ripening phase, from flowering to full development of grain. In the tropics, the reproductive phase is about 35 d while the ripening phase ranges from 30 to 35 d. Both phases are relatively constant, although low temperatures have been known to prolong them and high The flowering response of the rice plant to photoperiod 3 temperatures to shorten them. The ripening phase may be prolonged to as much as 60 d. However, it is the vegetative growth phase whose duration generally varies greatly and which largely determines the growth duration of a cultivar, especially in the tropics. The vegetative growth phase can be further divided into the basic vegetative phase (BVP) and the photoperiod-sensitive phase (PSP). The BVP refers to the juvenile growth stage of the plant, which is not affected by photoperiod. It is only after the BVP has been completed that the plant is able to show its response to the photoperiodic stimulus for flowering . this is the PSP of the plant. Figure 2 shows the growth phases and the typical response of a photoperiodsensitive rice and a photoperiod-insensitive rice. Based on the BVP and PSP, varietal response to photoperiod can be classified into four types as shown in Figure 3 (105, 526). 2. Growth phases and typical responses of a photoperiod-sensitive rice and a photoperiod-insensitive rice. BVP = basic vegetative phase, PSP = photoperiod-sensitive phase. 3. Four types of varietal response to photoperiod. BVP = basic vegetative phase, PSP = photoperiodsensitive phase. 4 The flowering response of the rice plant to photoperiod The BVP and PSP are two separable growth phases controlled by different genes. Although some tropical cultivars may be classified as the D type having both long BVP and long PSP, most were probably eliminated during domestication since they would have had an unusually long growth period and could be planted only within a narrow range of dates. Such cultivars were found in Bangladesh and are known as Rayadas (105). The four types shown in Figure 3 were classified under one temperature condition. Norin 20 (Type A) has a short BVP. When grown in the tropics, however, it has a much shorter BVP than when grown in the temperate areas (Fig. 1). In classifying cultivars based on BVP, most of those from the low latitudes were found to have long BVP? fs (531, 532). Basic vegetative phase At the early growth stages, the rice plant is photoperiod insensitive so that the photoinductive treatments are usually started when the plants are 10-63 d old (13, 90, 142, 175, 186, 213, 230, 232, 273, 304, 316, 401, 512, 531). Because of this insensitivity to photoperiod, the early growth stage has been termed the basic vegetative phase; it is also referred to as the juvenile growth stage of the insensitive phase of the plant. Suenaga recognized the BVP as early as 1936. He measured it by taking the duration of the vegetative growth phase at optimum day length. The BVP also has been measured by subtracting 35 d from the growth duration (sowing to flowering) of plants grown at the optimum photoperiod (526). This assumes that the period from panicle initiation to flowering is about 35 d. Anema (13) modified the determination of the BVP by subtracting 35 d and the minimum number of photoinductive cycles needed for panicle initiation from the heading date. The resulting BVP values are smaller but this complex method would mean determining the minimum number of photoinductive cycles needed for each cultivar. The range of BVP reported in the literature has varied from 10 to 85 d (105, 175, 266, 273, 326, 381, 383, 401, 407, 445, 512). In an F 2 population, BVP? fs of more than 100 d were reported (249), but a BVP of this length has not been found in conventional rice cultivars. It is possible that such characters are eliminated during cultivar selection. The appendix shows the range of the BVP of the cultivars tested at IRRI. The indica cultivars generally have longer BVP (583). Other workers have reported or measured BVP in terms of leaf number (93, 215, 340, 413, 551, 575). The minimum number of leaves can be less than five. The need for determining the BVP of a rice cultivar before using it as an experimental plant material is obvious but is frequently overlooked especially in the study of the inheritance of photoperiod sensitivity. Several experiments showed that short-day treatments of seedlings accelerated heading (393, 401, 437, 438, 445) or delayed it (16, 273, 284, 287, 296, 426, 443, 447, 551). The results indicate the possible effect of photoperiod while the plant is in its early growth stage and the possible existence of a very short BVP. On the other hand, long-day treatments of seedlings have been reported to induce earliness in flowering (418, 427). These varied and conflicting results may have been caused by nonspecific factors. A good example is seedling vigor, which is The flowering response of the rice plant to photoperiod 5 known to affect the flowering date, especially in the weakly photoperiod-sensitive cultivars. The degree of sensitivity of rice plants has been reported to increase with age (142, 190, 195, 202, 205, 347, 512). The increase in leaf area accompanying advancement in age does not explain this increase in sensitivity (413). An increase in sensitivity with age up to 28 d and then a decrease in sensitivity with older plants (35- to 42-d-old plants) has been reported (296). The delay probably resulted from the setback from delayed transplanting and not from plant age because the plants were already 63 d old when transplanted, with some already flowering. The optimum age of responsiveness is probably the result of growthlimiting factors, such as space and nutrients and delayed transplanting. Katayama (202) indicated that the BVP, or aging effect, probably resulted from small leaf area and (or) low metabolic activity and (or) lack of a specific metabolic pattern in young plants. The substance causing response to short-day conditions is produced in too small a quantity to affect morphogenesis at the growing point, but increases gradually with increasing age. Studying this aspect, Suge (460) found that the growth inhibitors in the plant were greatly reduced as the plant grew. However, it is not known whether these inhibitory substances are essentially involved in the sensitivity of the plant to photoperiod. In some instances, the apparent low sensitivity of the younger plants may be a matter of completing the BVP. If the photoinductive cycles were given before the BVP of the plants had been completed, the effective photoinductive cycles would be less and the resulting response of the plants would be smaller. The transition from the BVP to the PSP is not well known; it could be abrupt or it could involve a gradual buildup. Using several cultivars, Best (26) found that the insensitive phase (BVP) changed to the fully sensitive phase (PSP) within a week. The following are possible explanations for the existence of the BVP (26): 1) The first leaves formed are completely insensitive to photoperiod. 2) The first leaves formed have very low sensitivity that they do not reach an adequate level of induction to evoke floral initiation before the more sensitive leaves formed at higher nodes have reached this stage. ) The first leaves formed do not attain the induced stage before the (early) senescence of these leaves. 4) The total leaf area required before the plant can react by floral initiation to the inductive photoperiod is so large that it is reached only at a relatively late stage of plant development. 5) The growing point of the young plant is unable to react to the floral stimulus or the st imulus cannot reach the growing point. Photoperiod-sensitive phase The PSP or the eliminable phase (186) is the growth stage indicative of the rice plant's sensitivity to photoperiod. In photoperiod-sensitive cultivars, the PSP determines the rice plant? fs sensitivity. The PSP of photoperiod-insensitive cultivars ranges from 0 to 30 d while that of sensitive cultivars lasts from 31 d or longer. Under continually long photoperiods, 6 The flowering response of the rice plant to photoperiod some cultivars have been reported to remain vegetative even after 12 yr of growth (234). The PSP is usually determined by subtracting the minimum growth duration from the maximum growth duration of a cultivar (526). Because many cultivars remain vegetative for a long period if grown under long-day conditions, experiments are usually terminated after 200 d and the PSP of the cultivar is given the value of 200+. Besides measuring the PSP, there are many other ways . to be discussed later . of determining a cultivar? fs sensitivity to photoperiod. A rice cultivar? fs response to photoperiod may be measured by the length of the PSP, which in turn is determined by both the critical and optimum photoperiods of the cultivar. Because these two terms have been used interchangeably and in many ways, the following definitions will be adopted herein. Optimum photoperiod is the day length at which the duration from sowing to flowering is at a minimum (34). Critical photoperiod is the longest photoperiod at which the plant will flower or the photoperiod beyond which it cannot flower. Figure 1 shows that BPI-76 has an optimum photoperiod of 10 h and a critical photoperiod of 13 h. Tainan 3 has an optimum photoperiod of 12 h but no critical photoperiod because it flowered under all photoperiods. The critical photoperiod determines whether a cultivar will flower when planted at the usual time at a certain latitude, while the optimum photoperiod determines whether it will flower within a reasonable time if planted during a period with longer days than would normally occur during the growing season. With BPI-76, if the optimum photoperiod is 10 h and the delay under photoperiods longer than 10 h is great, one would expect the flowering of this cultivar to be greatly delayed when planted in the northern latitudes where the photoperiod during the growing season is about 14 h. If the critical photoperiod is 12 h, flowering will occur very late at high latitudes, and if flowering does occur, the crop will not mature in time because frost will kill it. A cultivar with a long optimum photoperiod or no critical photoperiod would have wider adaptability . it could be planted at any latitude and in any season, provided it is not too sensitive to temperature. Optimum photoperiod The optimum photoperiod differs with cultivars although many workers have observed it to be 8-10 h (39, 116, 135, 142, 311, 362, 371, 393, 512). Using intermediate photoperiods of less than and more than 10 h may reveal more important information. But this will require facilities in which a maximum of 15-min difference in photoperiods can be accurately obtained. There are also indications that the optimum photoperiod increases with increase in temperature Njoku (335) did not find any optimum photoperiod in the varieties he studied. The photoperiod he used was as short as 9 h, well below the range of natural day lengths. Cultivars with optimum photoperiods longer than 10 h have also been reported (26, 90, 320, 322, 362, 568). The less sensitivity to photoperiod, the longer is the (394). The flowering response of the rice plant to photoperiod 7 ptimum photoperiod (116, 311). However, others found no correlation between the optimum photoperiod and the photoperiod sensitivity of the many cultivars they tested (572). A photoperiod longer or shorter than the optimum has been shown to delay flowering, the delay depending upon the cultivar? fs sensitivity (311, 316, 319, 371, 393, 459, 5 13, 568). The term supraoptimum photoperiod has been used when the photoperiod is shorter than the optimum. Panicle initiation in plants receiving a photoperiod as low as 4 h has been reported (140). No flowering has resulted under a 2-h light period (140). Plants receiving 8-h light and varying dark periods from 16 to 64 h showed inhibited shoot apex conversion (219). This was ascribed to inadequacy of carbon compounds for synthesis of requisite quantity of flowering hormone. The turning point mentioned by Yu and Yao (568) is similar to the optimum photoperiod, but the photoperiod values they reported were larger because these were not the photoperiods at which growth is shortest but the photoperiods at which the first long-day effect is manifested. Critical photoperiod Scripchinsky (417), reviewing the literature on rice, indicated that the rice plants have a ? critical length of day for flowering.? h Later studies showed the presence of a critical photoperiod ranging from 12 to 14 h (175, 209, 244, 354, 478, 490, 500, 553). The critical photoperiods determined under controlled photoperiod rooms were almost the same as the day length from sunrise to sunset at 30 d before flowering under natural conditions (499). The lower the latitude of origin of a cultivar or strain, the shorter is its critical photoperiod (196, 356). The critical period is influenced by temperature (566) and lengthens as the plant becomes older (2 12). The PSP of a cultivar is probably a measure of the combined effect of photoperiod on its optimum photoperiod and critical photoperiod. The shorter the critical photoperiod, the longer is the PSP. Short optimum photoperiod is also associated with long PSP. Photoinductive cycles A photoperiodic cycle that induces the initiation of flowers on plants is called a photoinductive cycle. A 10-h photoperiod alternating with a 14-h dark period is one possible photoinductive cycle of a short-day rice cultivar. The minimum number of photoinductive cycles necessary to initiate the panicle primordium of a rice plant varies from 4 to 24. This required minimum number varies not only with cultivar, but also with the photoperiod being used (13, 21, 26, 142, 195, 292, 338, 344, 408, 449, 500, 527, 529). The number of photoinductive cycles necessary increases with photoperiod length (190, 195, 203, 204, 527). According to Katayama (190), the minimum number increases proportionally with the photoperiod used, although others (527) failed to obtain a proportional increase using a different cultivar. Katayama (190) found that the minimum number was lower in cultivars from higher latitudes than in those from lower latitudes. The flowering response of the rice plant to photoperiod Suge (463) showed that different numbers of photoinductive cycles produced different amounts of floral stimulus. He also found that Gibberellin A3 reduced the minimum number of photoinductive cycles necessary to induce flowering. However, gibberellin alone did not induce flowering under noninductive photoperiods. That a certain number of photoinductiv e cycles is required to induce flowering suggests that the stimulus produced by the treatment is cumulative and that flower induction occurs when the stimulus has reached a certain threshold level (205, 206, 208). Photoinductive cycles interrupted by noninductive cycles can negate to different degrees the effect of the photoinductive cycles (200, 206, 345). There are also indications that emergence of the panicle from the flag leaf sheath is a process separate from panicle initiation. For example, internode elongation, after the panicle has been initiated, proceeds more rapidly at shorter than at longer photoperiods (26, 37, 67, 135, 425, 451, 512, 529), and earliness is further induced if the treatment is prolonged until flowering (33, 438, 498). It is possible, however, that panicle initiation and exsertion are separate processes, but certainly the latter proceeds only after the panicle has been formed. The effect of photoperiod on exsertion may be on fuller development of the panicle, hence indirectly affecting elongation of the first internode or exsertion of the panicle. Plants subjected to insufficient photoinductive cycles sometimes form panicles but no emergence occurs (see Table 1) (92, 122, 344, 512, 526). A difference of two photoinductive cycles could make the difference between exsertion or nonexsertion of the panicle. Several workers, however, have reported that photoperiod has only a slight effect on culm elongation and panicle emergence (85, 116, 338, 473); but the cultivars used (85, 338, 473) were generally weakly photoperiodic because the differences between the control and the treated plants were relatively small (16 d at most). In another instance, the treatment was started at a later stage . 20 d before the standard heading time . at which time the plants had received sufficient photoperiodic stimulus for panicle initiation and emergence (1 16). In another experiment, long photoperiods had no effect on the terminal bud that had reached the stage of differentiation of secondary branch primordia (345). Reversals from a reproductive to a vegetative phase have been reported (54, 342). In some instances, however, the panicle is initiated and differentiated but Table 1. Response of 30-d-old BPI-76 seedlings given different numbers of 10-h photoinductive cycles. Days from sowing Days from sowing Cycles (no. ) to panicle to panicle initiation emergence 8 ** 10 47 ** 12 47 88 Continuous 46 66 *No panicle initiation 200 d after treament. **No panicle meregence 200 d after treament * The flowering response of the rice plant to photoperiod 9 does not emerge (526). The unexserted panicle ceases to grow, and instead the terminal growth is dominated by a shoot from a node below the panicle. Such a situation is not a true reversal of the growing point. In more recent histological studies, incomplete short-day treatment changed the bract pri mordium into a leaf primordium, a true reversal of some parts of the growing point (346). Reception of the photoperiodic stimulus and translocation The photoperiodic stimulus may be received by the leaves of the rice plant (24). The leaf sheaths can receive the stimulus as shown by removing the leaf blades and subjecting the plant to photoinductive treatments (26, 142, 481). More photoinductive cycles were needed to induce flowering when the leaf blades were removed (142). Defoliated plants responded to light interruption given during dark periods as well as the intact plants (142). In one cultivar, the culm received the photoperiodid stimulus (26). Evidently, the leaf most receptive to the stimulus is the youngest fully formed leaf (263). The first leaves, up to the sixth leaf, are either insensitive or have low sensitivity to photoperiod (26). It is difficult to study this aspect of leaf sensitivity because grafting experiments with the rice plant are difficult. Removing the leaves at regular intervals after the end of the photoinductive cycles showed that the floral stimulus moves gradually from the leaves to the terminal bud (142, 464). The translocation of the stimulus depends on temperature. It was also reported that the rate of translocation of the stimulus is the same regardless of the number of photoinductive cycles received by the plant (463). The question of stimulus movement from one tiller to another has also attracted the attention of several workers. When a plant was divided and half was kept under a 24-h photoperiod and the other half under an 8-h photoperiod, the half subjected to the short-day treatment flowered while that under long-day treatment remained vegetative (230, 232). The results indicate that the stimulus is not transmitted from one tiller to another. This finding has been substantiated by other workers using different cultivars and methods (263, 408, 521). Manuel and Velasco (263) concluded that the stimulus that induces flowering can be conserved in the stubble and later transferred to the ratoon but not to a neighboring tiller of the same age as the donor. Sasamura (413), however, reported that the floral stimulus goes from the main culm to its tillers. The irregularities observed in photoperiod-sensitive cultivars when planted during the off-season, for example, the high number of nonflowering tillers, have been attributed to the effect of the photoinductive cycles received by the plant and their nontranslocation to the succeeding tillers formed (521). Light intensity and quality The light intensities used to prevent or delay flowering varied from 1 to more than 200 lx. Incandescent, tungsten, as well as fluorescent bulbs have been used (69, 143, 310, 396, 484, 489, 503, 538, 565, 570, 577). The brighter the illumination, the stronger the retarding effect. 10 The flowering response of the rice plant to photoperiod Delay in flowering with light intensities varying from 10 to 100 lx and even at 1 lx (310, 484) has been reported (538, 565, 589). Extending the day length using light intensities of less than 200 lx during the first or last 3 h of the 12-h dark period did not prevent flowering (478). In another experiment, 2-h illumination at 15 lx before a 9-h dark period showed some inhibiting effect and 1-h illumination at 500 lx incandescent light before a 9-h dark period inhibited flowering (143). In correlating laboratory studies with field studies, the natural photoperiod used is usually based on the sunrise-to-sunset duration. Such measurements are unsatisfactory in assessing periods of effective light because very low light intensities have been known to effect photoperiod responses in some experiments. Civil twilight in the morning can generally delay flowering but civil twilight in the evening may or may not delay flowering (143, 196, 205, 502). Civil twilight ends when the light intensity is about 4 lx. Twilight, of course, varies with localities and within the year. The critical light that results in delayed flowering is around 5 lx and sometimes 10 lx, depending on variety and other factors (174). Twilight intensity also varies and may be higher in the morning than in the afternoon (Fig. 4). Katayama (196) attributes the greater effectiveness of the morning twilight to higher intensity. Cloudy weather affects twilight duration. Takimoto and Ikeda (478), however, concluded that the photoperiodically effective day length is equal to the astronomical day length (sunrise to sunset) because twilight (less than 200 lx) had little effect on photoperiodic induction in their experiment. Wormer (538) showed that low light intensities for 6 h (10-100 lx) given after a 12-h daylight can delay flowering. Farmers have complained that their rice plants did not flower regularly because of the electric lights installed along their fields (552). One incident has been reported in which the light from a flame of waste natural gas prevented normal 4. Change of light intensity during civil twilight (after Katayama [196]). The flowering response of the rice plant to photoperiod 11 flowering in rice. The effect of light was noticeable up to about 270 m from the flare (22). Although light from incandescent bulbs is generally used for photoperiod studies, other colors have been tried in rice. The blue-violet part of the spectrum has been shown to retard flowering (260) as has infrared light (323). The delay in flowering caused by green light is very slight, only 4-5 d later than natural day length (234). Green has, therefore, been used in light traps for the moth. Red light is the most effective in delaying flowering, while blue showed some effect only at high intensities and in the most photoperiod-sensitive cultivars (26, 146, 153, 503). The phytochrome pigment is generally regarded as the system that interacts with photoperiod or with different light qualities, such as red, far-red, and blue. Such pigment has been studied in rice coleoptile by Pjon and Furuya (378, 379). For panicle initiation, rice needs a high light intensity during the light period. The inhibition caused by low-intensity light during the light period can be overcome effectively by exposing the plant to high-intensity light immediately before or after the inductive dark period (140, 145). This phenomenon is similar to that reported in other short-day plants and is evidently a carbohydrate requirement. This requirement would explain why a 2-h light period followed by 22-h dark period did not induce flowering (140). Ikeda (145) reported, however, that plants growing in low-intensity light during the photoinductive period but briefly exposed to high-intensity light before the inductive dark period had floral induction, suggesting that light requirement for floral induction of rice is not entirely concerned with photosynthesis. In the flowering response of the rice cultivars to photoperiod, red light given during the dark period inhibited flowering (136, 146, 148, 411, 442). The effect of red light increased with intensity. Red light, as low as 10 ? EW/cm 2 given for 3 h or 290 ? EWc/cm 2 for 15 min in the middle of the dark period, inhibited flowering (146, 148, 149). Red light was most effective in inhibiting panicle initiation when given in the middle of the dark period (150). With red light, the period of exposure needed to inhibit floral development was shorter than with white light (146). The inhibiting effect of red light has also been shown in experiments involving red and far-red lights. Far-red after red nullifies the delaying effect of red light and promotes flowering (411). Far-red before a 9- or 10-h dark period promotes flowering and this effect can be reversed by red light (146, 149, 152). Far-red enhances flowering whereas blue retards flowering (185). Far-red after the critical dark period can shorten the critical dark period as well as reduce the minimum number of inductive cycles required (145). Interruption of the dark period Sensitive strains of rice respond to light interruption (26, 69, 218, 232, 260, 323, 449, 570, 577). Light given in the middle of the dark period delayed the flowering of the sensitive cultivar Shuan-chiang (570). The light intensity used was 1001x and the duration varied from a flash to as long as 15 min. The degree of delay was greater in the light interruption of a 12-h dark period (12 light and 12 dark) than of a 16-h dark period (8 light and 16 dark) (577). Interrupting the light period with darkness did not accelerate flowering. 12 The flowering response of the rice plant to photoperiod The earlier the interposition of the light during the dark period, the greater was the delay (449). The findings show that the flowering response of the plant is determined by the longest dark period. Days from photoinductive treatment to flowering The literature indicates that the number of days from panicle initiation to flowering is about 35. Many workers have reported that the difference among cultivars is small (7, 407, 511, 551). Others found that the number of days from panicle initiation to flowering ranges from 10 to 241 d (425). It seems obvious, however, that 10 d is too short for the full development of a panicle. Flowering may be delayed by long photoperiods after panicle initiation (176, 524). But if the plants are given photoinductive cycles beyond the minimum requirement, the subsequent photoperiods have very little effect on flowering and elongation (501, 524). Auxin application can nullify the delaying effect of long photoperiods (176). Under natural day length, the number of days from the first-bract differentiation stage to flowering varied from 27 to 46 d, depending upon the cultivar and time of sowing (14, 270). Reports vary on the number of days from the start of the photoinductive treatment to flowering. Misra (285) reported 37 d in 30-, 40-, 50-, 60-, and 70-d-old plants of the cultivar T. 36 using a 10-h photoperiod. Fuke (93) noted that the plants flowered about 28 d after treatment. The number of days from photoinductive treatment to flowering depends upon the photoperiod being used. Panicle initiation and flowering were earlier under the 10-h than under the 11- and 12-h photoperiods (527). Using 168 F 2 plants, those treated under the 10-h photoperiod took 30-47 d to flower, or a mean of 35. d (Li, unpublished data. For practical purposes, an estimate of 35 d should be workable. Thus, to obtain the BVP or the time of panicle initiation, 35 d can be subtracted from the minimum growth duration of the cultivar. In studying the effect of photoperiod on the flowering of the rice plant, the most fundamental consideration is panicle initiation because it marks the actual change from the vege tative to the reproductive phase. Instead of using this as a basis, however, most studies use the flowering date, which is only a projection of the variations of the date of panicle initiation. To a certain extent, several factors can affect the stage from panicle initiation to emergence. In some instances, panicle initiation can occur without the subsequent emergence. The panicle primordium is aborted and a vegetative shoot may dominate the growing tip (527). A methodological question might therefore arise regarding accuracy of the experiments based on flowering date. The practicality of the method, however, far outweighs the need for extreme accuracy. Biochemical changes during photoinduction Very little work has been done on the chemical changes occurring during photoinduction and panicle development in rice. An increase in the rate of respiration of rice shoot apices with each photoinductive cycle given to the eighth The flowering response of the rice plant to photoperiod 13 day, followed by a gradual decline in rate, has been reported (293). The peak of the respiration rate almost coincides with the minimum photoinductive cycles needed by the rice plant at 8 h of photoperiod. The results suggest that the photoperiodic mechanism in the flowering of rice involves a respiratory shift. This corroborates the findings of Elliot and Leopold (86) who used other plant species. The changes in carbohydrate and nitrogen content of rice plants subjected to short days were also studied by Misra and Mishra (299). Unfortunately, the difference in heading between treated and control plants was only 4 d. Khan and Misra (222) reported an increase in sugar and nitrogen content of the leaves when subjected to photoinductive cycles. Photoinduction increases the gibberellic acid activity, although the value is low (461). This immediate rice, visible after three photoinductive cycles, returns to a level lower than that of the original. The rice plant is difficult to use for studies on biochemical changes during reproduction. Perhaps it is best to leave this type of study to other short-day plants. Effect of temperature on the flowering response to photoperiod The flowering of the rice plant is mainly controlled by two ecological factors . day length and temperature . which are often interrelated. The plant may respond to temperature and photoperiod simultaneously, but the degree would vary according to the cultivar. Cultivars have been classified based on these two factors (248, 356, 530). Temperature affects both the photoperiod-sensitive and photoperiodinsensitive cultivars. Generally, high temperature accelerates and low temperature delays heading (5, 6, 90, 126, 186, 307, 339, 340, 370, 376, 409, 410, 439, 456, 531). Some reports, however, have shown that high temperature delays flowering (15, 18, 394). The acceleration of the photoperiod response by high temperature is an overall effect, but it does not indicate the specific effects on the different stages leading to flowering. The effect of temperature on the BVP, photoinductive period, panicle differentiation and development, and critical photoperiod has not been fully studied. Uekuri (506, 507) studied the effect of low temperature during the BVP and found a definite delay in attaining the PSP. The degree of extension of the BVP by low temperature varied with the cultivars used. The growing point of the shoot is the receptive organ for the low-temperature effect, not the leaf blades (506). Ahn (5) reported that high temperature reduced the BVP but had very little effect on the PSP. As early as 1931, Fuke had considered the effect of temperature during the photoinductive period. He used snow to lower the darkroom temperature, but the 5-10? †¹C decrease had little effect on heading. Temperatures above 20? †¹C to 29? †¹C accelerate panicle initiation (24, 341). Vergara and Lilis (524) showed that the vegetative primordium was converted to reproductive primordium at the same time or at the same morphological stage regardless of temperature (21-32? †¹C). 14 The flowering response of the rice plant to photoperiod Haniu et a1 (1 15) found similar results. These results contradict those reported by Noguchi and Kamata (341) and Best (24). Temperatures below 15? †¹C inhibited initiation and bud development (156). Floral induction, however, is possible at 15? †¹C (341) but not at 12 or 40oC (115). Because many test plants died in the growing process, 15? †¹ C is assumed to be near the lowest limit for rice growth (341). The optimum temperature reported for photoinduction is 30o C (1 15). The question still remains as to whether a critical temperature for photoinduction exists. The optimum temperature for photoinduction may vary depending upon the photoperiod being used. The optimum temperature tended to be higher under a longer photoperiod and vice versa (24, 364). Putting it another way, at a certain temperature each cultivar has its own optimum day length under which it flowers at the earliest date (459, 572). Detailed microscopic studies of the development of the panicle primordium have shown that high temperature accelerates panicle development (260). The critical temperature for young panicle differentiation has been reported to be 18oC (555). Best (24) has also shown that panicle development, especially in its later stages, is accelerated at high temperatures (35-37oC). On the other hand, low temperature markedly retards panicle primordium development, and, below 25oC, the panicle may not emerge completely from the flag leaf sheath (24). A night temperature of 24. 4oC was found more favorable than 29 and 35? C in accelerating the flowering of the Elon-elon cultivar (263). High night temperature accelerates flowering (220). This was attributed to increased production of florigen during the dark period. This may not be the case and dissecting plants after photoinductive treatments may reveal if it was an acceleration in panicle development and exsertion rather than in panicle initiation. Others have found that the acceleration in flowering with high temperature is the result of acceleration in panicle exsertion, which, in turn, is the result of shorter leafing interval (524). Obviously, caution should be taken in determining the time of panicle initiation by observing the heading date because the exact date of panicle initiation cannot be determined by this method. Measurements and methods of testing photoperiod sensitivity Most studies on the photoperiodism of the rice plant have been considered from two standpoints, namely, classification of the cultivar into photoperiod-sensitive and photoperiod-insensitive types and measurement of the degree of sensitivity. The classification may be relatively easy, but the measurement is rather complex (195). As a result, several methods of measuring photoperiod sensitivity have been developed. Studies on the measurement of photoperiod sensitivity are usually based on the reduction in the number of days as a result of short-day treatment (1 16, 195, 205, 327, 329, 357, 553, 574). Other methods were more specific; they measured the optimum photoperiod (40), critical photoperiod (351), or the gradient of the response curve (34, 192, 247) as the basis of sensitivity. Hara (116) was the first to measure photoperiod sensitivity using the formula: X The flowering response of the rice plant to photoperiod 15 = T . Y/Y X 100, where Y is the number of days required to head under standard conditions and T is the number of days required under an 8-h photoperiod. Several similar formulas have been used by other workers. The percentage or index obtained from such formulas, however, does not clearly define photoperiod sensitivity. The results usually apply only to the area where the rice was tested since the natural day length is usually used as the control. Chandraratna (37, 40) used second-degree polynomials to compute the minimum heading duration and optimum photoperiod; this method involved using at least three photoperiods. He showed that cultivars differ in both characters. Oka (352) and Katayama (192, 201) measured the critical photoperiod and the degree of sensitivity of several cultivars using different methods and formulas and came up with their preferred method of measurement. Both workers used the natural day length as a basis for computation and assumed that flowering occurs 30 d after photoinduction. Best (25) and Li (249), using a method similar to Chandraratna's (34, 37, 40), measured sensitivity based on response curves obtained by plotting the time from sowing to floral initiation on the ordinate and the photoperiod used on the abscissa. The method, however, requires a wide range of photoperiods. Li (249) also studied photoperiod sensitivity in terms of the BVP and the PSP. The BVP was obtained in plants grown under 10 h of light, and the PSP (which is a measure of sensitivity) by subtracting the growth duration under the 10-h photoperiod from that under the 16-h photoperiod. The PSP values obtained show the possible maximum range in growth duration as a result of extending the photoperiod. The photoperiodic characteristics of a rice plant have been described by Stewart (458) who used a different criterion based on 1) basic vegetative period in terms of degree-days (based on temperature accumulation), 2) photoinduction period in degree-days or degree-minutes (using accumulated night length), and 3) panicle development period in degree-days (based on temperature accumulation). Tests under field conditions were analyzed by this method and predictions were made on the response of the cultivar sown in different months. In Japan, the flowering response is evaluated using the floral stages (135, 463). The Japanese workers have used the scale of 0-7, based mainly on the length of the developing panicle. This destructive measurement is more accurate than the usual days from sowing to flowering or treatment to flowering. The choice of the most appropriate method of testing and describing the response to photoperiod depends upon the purpose of the experiment and the available facilities. From the physiological standpoint, however, controlled photoperiod and temperature are desired because of their advantages over natural photoperiods and temperatures. Date-of-planting experiments Day length changes rhythmically within a year and varies depending upon the latitude. The amount of change in day length during the rice cropping season differs from one latitude to another (Fig. 5). Even in locations at the same latitude the day length during the cropping season may differ because the planting dates 16 The flowering response of the rice plant to photoperiod 5. Day length changes during the cropping season at various locations in Asia. may differ greatly depending mostly on the rainfall pattern at each location. At northern latitudes (Sapporo, 43? †¹ N, and Konosu, 36? †¹ N) day ength increases and then decreases during the cropping season (Fig. 5). At lower latitudes (Taipei, 25? †¹N, and Los Banos, 14? †¹N) day length decreases during the main growing season. Near the equator (Bukit Merah, 5? †¹N) there is little change. These differences in day length during the growing season may account for the wide range of photoperiod response of rice cultivars. A rice cultivar that must have less than 12 h o daylight to flower will obviously flower too late at the northern latitudes because frost will set in before harvest. In the northern hemisphere, the longest days are in June and the shortest are in December. Taking these into account, the photoperiod response of the rice cultivars can be tested to a limited extent by planting the cultivars at a certain location at different dates. Maximum differences in growth duration can be obtained in the May and November plantings if temperatures are not too low for growth. If a rice? fs growth duration changes more than 30 d, agronomists usually consider it photoperiod sensitive or a seasonal cultivar. As Best (24) has pointed out, this criterion is not specific enough for research on photoperiodism, and caution should be taken in evaluating the data obtained. These phenological data, however, are important to breeders in selecting ecotypes. This method of testing sensitivity to photoperiod has been followed in Australia (245), Brazil (l03, 579), China (44, 356, 582), India (98, 99, 101, 214, 220, 295, 298, 423), Indonesia (467), Japan (533, 548), Korea (247, 466), Malaysia (74, 77, 244), Philippines (91, 512), Russia (452), Senegal (66), Sierra Leone (68, 536), Sri Lanka (112, 259, 402), Thailand (381), Trinidad (325), and United States of America (177, 180). The flowering response of the rice plant to photoperiod 17 These experiments strongly confirm the existence of wide cultivar differences in the effect of planting date on flowering date. Many of the results obtained from this type of testing, however, are not applicable to identical cultivars grown at different latitudes. A cultivar can be insensitive to day length in Malaysia but sensitive in Taiwan. Results of field tests at a certain latitude are, therefore, not always applicable at another latitude. Some published papers on the use of this testing method failed to mention latitude or the place where the tests were conducted. Under natural conditions very small differences in day length can affect the rice plant. In Malacca (Malaysia), the difference between the maximum and the minimum day lengths is only 14 min and yet the cultivar Siam 29 takes 329 d to flower when planted in January and only 161 d when planted in September (76). Another instance showing the sensitivity of the rice plant to small differences in day length was reported in a date-of-planting experiment in Malaysia (244). There was a difference of as much as 156 d in the growth duration of photoperiodsensitive cultivars when planted in the same month but in different years (Table 2). This presumably resulted from differences in weather during the critical periods. Cloudy weather early or late in the day shortens the twilight hour, thus shortening the day length. Toriyama et al (490) tested rice cultivars involving not only monthly planting but also sowing at different latitudes (Sri Lanka, Taiwan, and Japan). This gives a better idea of the photoperiodic response of the cultivars but involves much work and cooperation. Ecology and photoperiodism Rice can be grown over a wide range of environmental conditions, from the equator to about 53? N latitude, leading to the differentiation and establishment of various ecotypes and forms. The great diversity in photoperiod sensitivity from one latitude to another or within a latitude probably indicates that the rice cultivars predominantly cultivated in each area are those that have been selected on the basis of local adaptability (that is, adaptability to the temperature of the rice-growing season, day length, and duratio n of the growing season) to assure the full development of the plant and the best possible balance between vegetative and reproductive growth (423, 530, 532, 584, 585). Table 2. Growth duration (days from sowing to flowering) of photo. period-sensitive cultivars when planted in January 1962 and 1963 at several localities in Malaysia (244). Cultivar Locality Jan 1962 Jan 1963 Difference Engkatek Telok Chengai 136 292 156 Kota Bahru 146 243 97 Kuala Lumpur 134 97 37 Subang Bukit Merah 270 224 46 lntan 117 Kuala Lumpur 171 138 33 Kota Bahru 276 176 100 18 The flowering response of the rice plant to photoperiod A major problem in studying the ecology of the rice plant, especially in reference to photoperiodism, is that cultivars in farmers' fields keep changing. For example, Hara reported in 1930 that Japanese cultivars were more sensitive than the cultivars from mainland China and Taiwan. He concluded that the lower the latitude of the region of the native habitat, the less sensitive were the cultivars there. Wada (531), using 134 cultivars, showed contrasting results . the cultivars from the northern region of Japan had lower photoperiod sensitivity than those from the southern region. Recent papers, however, generally agree that among the photoperiod-sensitive cultivars, the lower the latitude ofdistribution, the higher the sensitivity (351, 352, 356, 531, 583). The cultivars in the tropics or lower latitudes are usually late maturing (long growth duration). Many studies show that the late cultivars are more sensitive to photoperiod than the early ones (116, 248, 357, 511, 563, 583). In the tropics, where rice can be grown any time of the year provided there is sufficient water, photoperiod sensitivity presents certain problems. During the off-season, when the day length during the early growth stage is increasing, the sensitive cultivars are uneconomical to use because they take a very long time to produce any grain. For wider adaptability, cultivars should have low photoperiod sensitivity (53, 70) and thus have little differences in growth duration when planted at different times of the year or at varying latitudes. Insensitive cultivars have been successfully grown at different latitudes where rice is used as a crop (45, 351, 352, 511, 532, 568. This indicates that it should not 6. Growth duration of IR8 planted in June or July at 12 sites in Asia. La Trinidad and Kanke are high-altitude areas (52). The flowering response of the rice plant to photoperiod 19 e difficult to introduce new photoperiod-insensitive cultivars to different ricegrowing areas or to culture them year-round in the tropics. The plant breeders, as the varieties coming out indicate, are developing more photoperiod-insensitive cultivars. Extensive testing in various rice-growing areas of the world has established the wide adaptability of photoperiod-insensitive cultivars. In general, the longer the BVP the less variation ingro wth duration and the stronger the PSP the greater the variation in growth duration (581). The wide adaptability and the stable growth duration of IR8, a photoperiod-insensitive cultivar, are indicated by the data furnished by cooperators in various parts of the world. IR8? fs growth duration varied within a range of 20 d at latitudes from 11o to 27oN except at high altitudes where low temperatures prevailed during part of the growing season (Fig. 6). A more illuminating example of the effect of temperature comes from monthly planting at Los Banos, Philippines, and at Joydebpur, Bangladesh (Fig. 7). A comparison between the monthly mean temperatures and mean photoperiods shows that the more variable heading pattern at Joydebpur is more closely associated with temperature rather than with the prevailing photoperiod. The effect of low temperature on the improved tropical cultivars becomes more obvious in photoperiod-insensitive cultivars. 7. Mean monthly temperatures and day length in relation to the growth duration of IR8 at Los Banos, Philippines, and Joydebpur, Bangladesh (52) 20 The flowering response of the rice plant to photoperiod Sensitivity to photoperiod of rice cultivars in the deep water areas is an important characteristic for survival (104, 520). The floating rice cultivars are highly photoperiod sensitive. They are planted early in the season when the soil can still be worked and without danger of submerging the young seedlings. Flowering occurs when the floodwater peaks or starts receding. If the cultivar flowers when the floodwater is still rising, it would mean the complete loss of the crop if the panicles are submerged. Elongation ability ceases after panicle emergence. Harvesting is usually done when the floodwaters have receded. The maturity of floating rice cultivars coincides with the receding of the annual floodwaters which may be 150-270 d after sowing. Such a long growth duration requires a photoperiod-sensitive cultivar. So far, there is no known tropical cultivar that has a long growth duration and is not sensitive to photoperiod. Photoperiod sensitivity may work as a safety mechanism when precise planting dates are not followed and environmental conditions such as water level cannot be effectively controlled. If the date of sowing or transplanting is delayed because of insufficient rainfall, a photoperiod-sensitive cultivar may still mature at its usual time (352, 382). Plants are not seriously damaged if left in the seedbed for prolonged periods because the growth duration of the main crop is sufficiently long for the plants to adjust. Thus, land preparation and transplanting can be staggered (382). Maturation of the crop at the same time. as with photoperiod-sensitive cultivars planted at different dates, may reduce rat and insect damage in any one field. Also, harvesting and drying are simplified. If the soil is not sufficiently fertile, a photoperiod-sensitive cultivar will continue its compelled vegetative growth until the short days come. This would give the plant enough time to reach a reasonable plant weight and accumulate enough carbohydrates before flowering (528). Thus, a photoperiod-sensitive cultivar generally may be more resistant to unfavorable conditions. Long-growthduration cultivars (essentially photoperiod sensitive) are least affected by strong soil reduction (549). Most upland rice cultivars have short growth duration and are photoperiodinsensitive (11, 12). However, in areas where the rainfall pattern is bimodal, as in northern Thailand, the cultivars are of medium growth duration and are photoperiod-sensitive . possibly another indication of the greater specific adaptability of long-growth-duration cultivars to adverse conditions. The sensitivity to photoperiod of wild species has also been studied in relation to their ecological distribution. Most of the wild rice materials tested were sensitive (191, 201, 205, 209, 353). They suggested that this sensitivity favors the wild rice plants and is perhaps essential to their survival. Terminology used in describing photoperiod ensitivity There is confusion in the terms used to describe the response of the rice plant to day length (515). Often, the terms used for growth duration are also used for response to photoperiod (see Table 3). As early as 1912, Kikkawa pointed out that The flowering response of the rice plant to photoperiod 21 Table 3. Some terms used in describing the growth duratio n and day length response of rice cultivars. Terms References Response to day lengths: date fixed vs period fixed season fixed vs period fixed season bound vs period bound timely fixed vs periodically fixed short-day plant vs long-day plant ensitive vs indifferent sensitive vs insensitive sensitive vs less sensitive short-day plant vs indifferent plant strongly photoperiodic vs weakly photoperiodic sensitive vs photosensitive vs photononsensitive day length sensitive vs day photoperiodic photoperiodic insensitive length nonsensitive early, medium, and late long-aged vs short-aged early flowering vs late flowering late maturing vs early maturing Season of planting: aman vs non-aman yala vs maha winter vs summer main-season vs off-season first crop vs second crop wet vs dry season aus, aman, boro, rabi, kharif Growth duration: 33 7 214, 511 308 1, 99, 336 3 68, 352, 353 21, 98, 449, 538 563 51 1 195, 352 339 574 91, 276, 277, 281 259 158 3, 230, 374 427 112 444 Malaysia, Indonesia, an d Thailand China Philippines Bangladesh, India it is meaningless to classify the rice cultivars of the world into such groups as early, medium, late, aus, or aman. He said, however, that this classification is useful in districts where the climates are similar. The use of the terms photoperiod-sensitive and photoperiod-nonsensitive in reporting the flowering response of a rice cultivar to changes in day length has been suggested (515). Weakly photoperiod-sensitive is sometimes used in place of photoperiod-nonsensitive because the existence of a completely photoperiod-nonsensitive cultivar is difficult to prove. Weakly photoperiod-sensitive is also used to describe cultivars whose flowering is delayed by as many as 70 d by long photoperiods. However, those types can be planted any month of the year in the tropics and can be expected to flower within the crop season. The terms short-day plant and long-day plant are not satisfactory because most rice cultivars today are short-day plants. Sensitive and insensitive, sensitive and indifferent, and sensitive and less sensitive are ambiguous terms. Because the response being described is a response to light period and not only to light, the terms photosensitive and photononsensitive are inappropriate. 22 The flowering response of the rice plant to photoperiod 8. Effect of four photoperiod treatments on the seeding-to-heading period of seven rice cultivars. Chang and Vergara (51, 52, 53) classified rice cultivars into four types using the length of the BVP and PSP as criteria (Fig. 8). Their classification was based on duration of plants grown in the greenhouse. Under this classification, the Japanese varieties, such as Fujisaka 5 and Norin 20 (Appendix), do not fall under any category because they have a short BVP and short PSP. Also, at least four photoperiods (10, 12, 14, and 16 h) are needed to classify the cultivars. A more practical grouping could be as follows (using also the length of the BVP and PSP). 1. Photoperiod nonsensitive . very short PSP (less than 30 d) and BVP varying from short to long. 2. Weakly photoperiod-sensitive . arked increase in growth duration when photoperiod is longer than 12 h; PSP may exceed 30 d, but flowering occurs under any long photoperiod. 3. Strongly photoperiod sensitive – sharp increase in growth duration with increase in photoperiod; no flowering beyond critical photoperiod; BVP usually short (not more than 40 d). Cultivars tested under only two photoperiods, such as 10 and 14 h, can also be classified according to these groupings (1 1). Agronomists and farmers would tend to use these groupings. The flowering response of the rice plant to photoperiod 23 Inheritance of vegetative growth duration The inheritance of the duration from seeding to heading in cultivated rices has been studied by many research workers, but the findings have resulted in diverse interpretations. Three categories of genetic postulates were generally offered: 1) monogenic or digenic control of heading date, with earliness dominant to lateness; 2) monogenic or digenic control of flowering date, with lateness being a dominant trait; and 3) multiple-factor inheritance in which the F 2 population showed a continuous and often unimodal distribution and in which the same population might produce a bimodal distribution when grown in a different season (44, 509). In experiments where photoperiod sensitivity was recognized, delayed flowering under a long photoperiod was generally inherited as a monogenic or digenic dominant trait (38, 242, 406, 424, 567). In several crosses involving distantly related parents, sensitivity to photoperiod appeared to be a recessive trait (242, 406). The continuous and transgressive segregation in several F 2 populations involving photoperiod-insensitive parents was ascribed to multiple genes, which indicated dominance of earliness (41, 95, 96, 97, 333, 389, 469, 554). However, in crosses among varieties in Yunnan Province in China, photoperiod sensitivity appeared to be a recessive trait in some F 1 hybrids (252). Some of the divergent interpretations just mentioned resulted partly from failure to recognize the composite nature of the vegetative growth period from seeding to panicle primordium initiation, partly from failure to control the interaction of the environmental factors (mainly photoperiod and air temperatures) and the different genes controlling the vegetative growth period, or from failure to relate the phenotypic expression with the revailing environment. Recent studies at IRRI (48, 161, 162, 163, 164, 165, 167, 168, 249) have demonstrated physiologically and genetically the feasibility of partitioning the vegetative growt