INTRODUCTIONThe timing of the transition from vegetative growth to flowering is of paramount importance in agriculture, horticulture, and plant breeding because flowering is the first step of sexual reproduction. Studies to understand how this transition is controlled have occupied countless physiologists during the past half century and have produced an almost unmanageably large amount of information (Bernier et al., 1981a; Halevy, 1985 Halevy, -1989Bernier, 1988;Kinet, 1993).A majority of plants use environmental cues to regulate the transition to flowering because all individuals of a species must flower synchronously for successful outcrossing and because all species must complete their sexual reproduction under favorable externa1 conditions. Any environmental variables exhibiting regular seasonal changes are potential factors that control the transition to flowering. The major factors are photoperiod, temperature, and water availability. Plants that do not require a particular photoperiod or temperature to flower, i.e., the so-called "autonomous-flowering" plants, are usually sensitive to irradiance. The environmental factors are perceived by different parts of the plant. Photoperiod and irradiance are perceived mainly by mature leaves in intact plants. Temperature is perceived by all plant parts, although low temperature (vernalization) is often perceived mainly by the shoot apex. Water availability is perceived by the root system.There are strong interactions between these different factors, so that each factor can change the threshold value for the effectiveness of the others. Plants, as opportunists, will thus make use of a different critical factor in different environments. Melilotus officinalis, for example, is a biennial with a vernalization requirement in temperate zones and an annual long-day (LD) plant with no cold requirement in arctic regions. In photoperiodic species, such as the short-day (SD) plant Pharbitis nil and the LD plant Silene armeria, flowering in unfavorable photoperiods can be caused by changing temperature, irradiance, or nutrition or by removing the roots. Similarly, in some late-flowering mutants of Arabidopsis, vernalization and an increase in the proportion of far-red light in the light source can substitute for one another in promoting the transition to flowering (Martínez-Zapater and Somerville, 1990; Bagnall, 1992). Clearly, there are alternate pathways to flowering in most, ifTo whom correspondence should be addressed. not all, plants. Because the different flowering-promoting factors are perceived by different parts of the plant, this implies that these parts interact and that the fate of the apical meristem-remaining vegetative or becoming reproductive-is controlled by an array of long-distance signals from the entire plant.The ability of subsets of plant parts to control flowering is also underscored by the fact that some plants may flower almost normally after complete defoliation (Hyoscyamus niger, red Perilla, Chenopodium amaranticolor) or derooting (Perilla, Loli...

In order to test whether an increased export of carbohydrates by leaves and starch mobilization are critical for floral transition in Arabidopsis thaliana, the Columbia ecotype as well as its starchless mutant pgm and starch-in-excess mutant sex1 were investigated. Induction of flowering was achieved by exposure of plants to either one long day (LD) or one displaced short day (DSD). The following conclusions were drawn: (i) Both the pgm and sex1 mutants have a late-flowering phenotype in days shorter than 16 h. (ii) When inductive treatments cause a large, percentage of induced plants, there is always a large, early and transient increase in carbohydrate export from leaves. By contrast, when an inductive treatment results in only a low percentage of induced plants (pgm plants exposed to one DSD), the export of carbohydrates from leaves is not increased, supporting the idea that phloem carbohydrates have a critical function in floral transition. (iii) Starch mobilization is not required to obtain an increased carbohydrate export when induction is by one LD (extended period of photosynthesis), but is absolutely essential when induction is by one DSD (period of photosynthesis unaffected). (iv) Floral induction apparently increases the capability of the leaf phloem-loading system.

Understanding the complete picture of floral transition is still impaired by the fact that physiological studies mainly concern plant species whose genetics is poorly known, and vice versa. Arabidopsis thaliana has been successfully used to unravel signalling pathways by genetic and molecular approaches, but analyses are still required to determine the physiological signals involved in the control of floral transition. In this work, the putative role of cytokinins was investigated using vegetative plants of Arabidopsis (Columbia) induced to flower synchronously by a single 22 h long day. Cytokinins were analysed in leaf extracts, leaf phloem exudate and in the shoot apical meristem at different times during floral transition. It was found that, in both the leaf tissues and leaf exudate, isopentenyladenine forms of cytokinins increased from 16 h after the start of the long day. At 30 h, the shoot apical meristem of induced plants contained more isopentenyladenine and zeatin than vegetative controls. These cytokinin increases correlate well with the early events of floral transition.