The transition to flowering is a crucial moment in a plant’s life cycle of which the mechanism

A plant needs to flower at a suitable time of the year in order to produce a large number of viable seeds. Given its importance, it is not surprising that flower initiation is controlled by many environmental and endogenous factors. Physiological research in different plant species has identified numerous of these factors, leading to a multifactorial model of the control of flowering (Bernier, 1988). More recently, the combination of physiological and genetical research, especially in Arabidopsis (Koornneef et al., 1991) and pea (Weller et al., 1997), has identified several of the genes that play a role in this process. In Arabidopsis, the interactions among these genes have been further studied by including molecular approaches (Levy and Dean, 1998), leading to the identification of several pathways involved in the regulation of flowering time. A model for flowering initiation has been established consisting of a photoperiodic promotion pathway that promotes flowering under long-day (LD) light conditions, a vernalisation promotion pathway that promotes flowering at low temperatures and an autonomous promotion pathway that promotes flowering autonomously, independently of the promoting effects of the other two pathways. (MartinezZapater et al., 1994; Koornneef et al., 1998; Levy and Dean, 1998). This flowering model constitutes an appropriate framework for the analysis of flowering but is still far from complete and many questions remain. One way to refine the current model is to identify additional flowering time genes. The screening for late-flowering mutants in Arabidopsis has been quite exhaustive. Few early-flowering mutants were obtained initially because most screens were performed in early-flowering ecotypes under LD conditions, which enhance flowering and do not allow a wide window for selection. However, some screens in which early-flowering mutants were obtained have been described (Zagotta et al., 1992; Ahmad and Cashmore, 1996). In addition, several earlyflowering mutants were identified on the basis of other pleiotropic phenotypes. For instance, some phytochromedeficient mutants (Goto et al., 1991) and mutants that are hypersensitive to gibberellins (Jacobsen et al., 1996) or that overproduce cytokinins (Chaudhury et al., 1993) also flower early. Recently, another class of early-flowering mutants was obtained by overexpressing specific genes in transgenic plants (Cardon et al., 1997; Kania et al., 1997). How these earlyflowering mutants interact with the late-flowering mutants of the photoperiodic promotion or the autonomous promotion pathways is very poorly understood, however. Screens to obtain new early-flowering mutants can be done more efficiently in genetic backgrounds or under environmental conditions where Arabidopsis flowers late. For example, vernalisation-defective mutants have been obtained by mutagenising the late-flowering and vernalisationresponsive fca mutant (Chandler et al., 1996). Here, we describe a novel early-flowering mutant, early 4763 Development 126, 4763-4770 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 DEV0248