American Journal of Botany 96(7): 1236–1244. 2009.

FLORAL ORGANOGENESIS AND THE DEVELOPMENTAL BASIS FOR POLLINATOR DECEPTION IN THE ASIATIC DAYFLOWER, COMMELINA COMMUNIS (COMMELINACEAE)1 CHRISTOPHER R. HARDY,2,3,4 LINDSEY L. SLOAT,2,5 AND ROBERT B. FADEN3 2James

C. Parks Herbarium, Department of Biology, Millersville University, P.O. Box 1002, Millersville, Pennsylvania 17551-0302 USA; and 3Department of Botany, MRC-166, National Museum of Natural History, Smithsonian Institution, P.O. Box 37012 Washington, D.C. 20013-7012 USA The upper half of flowers in Commelina communis deceptively lures potential pollinators with its showy petals and staminodes on the false promise of abundant pollen. This paper presents evidence that staminodization in the upper half is associated with a severe retardation of the entire upper floral hemisphere early in development. Possible consequences of this developmental retardation are seen also in the gynoecium, where the upper carpel of the three-carpellate ovary is underdeveloped and sterile at maturity. Only late in development do the upper petals and staminodes expand and acquire pigments necessary for their attractive function. We surmise that retardations of this severity are unlikely to be found for functionally fertile organs such as stamens and ovule-producing carpels, because key preparatory events preceding sporogenesis might otherwise be disrupted. Such differential growth about the floral apex resembles that known in some eudicots to be regulated by the TCP gene family; thus, future molecular developmental studies in Commelina may help to extend our understanding of the evolutionary genetics of floral monosymmetry to monocots. Key words: Commelina; Commelinaceae; congenital fusion of sepals; dayflower; floral development; floral monosymmetry; pollinator deception; staminode development; staminodization; TCP gene family.

Commelina L., a member of the monocotyledonous plant family Commelinaceae, is a cosmopolitan genus of some 170 species possessing notoriously delicate, ephemeral flowers that last only part of a day before deliquescing (hence their common name “dayflowers”). These rather complex flowers also are purported to have a generalist, yet deceptive, pollination biology involving flies and bees (Faden, 1992). Attraction and deceit of pollinators is thought to be served primarily by the upper half of Commelina flowers in which there are typically two showy petals and three showy, yet infertile or largely infertile staminodes (e.g., Fig. 1; see also Vogel, 1978; Dafni, 1984; Faden, 1992; Endress, 1994). Because pollen is the only reward for pollinators in Commelina flowers (no nectar is produced), the showy yellow antherodes (which sometimes have maroon centers) of the staminodes are thought to deceptively mimic large masses of pollen or anthers with large amounts of pollen to pollen-foraging insects (Faden, 1992, 2000). Once lured into a flower by these staminodes, the upper petals and sometimes the medial (also called the median) stamen, pollinators are thought to be rewarded primarily by pollen provided by the medial stamen whose anther is positioned in or near the upper half 1

Manuscript received 10 October 2008; revision accepted 23 February 2009.

This work was based on the undergraduate biology honors thesis of L.S., who gratefully acknowledges the support of the Millersville University Department of Biology and a Catherine Keever Grant, D. Dobbins for assistance with scanning electron microscopy and helpful advice, R. Byrne and A. Cuthbert for helpful critique, A. Faden for acquainting her with Commelina at the Smithsonian greenhouses, and D. Stevenson for sharing his large collection of pickled C. communis. M. Endress, A. Litt, an anonymous reviewer, and AJB editors provided very helpful critique of an earlier version of this manuscript. 4 Author for correspondence (e-mail: [email protected]) 5 Present address: Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona 85721-0088 USA doi:10.3732/ajb.0800344

of the flower via its curved filament, although this stamen is actually attached in the lower half of the flower. Then, while pollinators are foraging in the center or upper half of the Commelina flower, it is thought that two less conspicuously colored stamens in the bottom half of the flower deposit pollen bound for future pollinations onto the abdomen of the pollinator. The salient features of this theorized pollination syndrome for Commelina were largely verified experimentally recently by Ushimaru et al. (2007) for the Asiatic dayflower, Commelina communis L. (Fig. 1). Following the experimental, systematic removal of various subsets of the floral organs in different flowers, Ushimaru et al. quantified the pollinator responses to these alterations, as well as the consequent positive or negative impacts of these responses on the flower’s fitness. In so doing, they provided evidence that the two large blue upper petals (Fig. 1) of C. communis do, in fact, attract potential pollinators (namely, syrphid flies and Japanese honey bees). They also demonstrated that the bright nonrewarding staminodes (Fig. 1), as well as the medial stamen (which may have an anther similar in color to the antherodes and is positioned near them), serve both to attract and guide pollinators to landings that promote pollen–stigma transfer and pollination–pollen loading onto the pollinator. With the ecological context for staminodization and the more general upper/lower (dorsoventral) floral asymmetry in Commelina established, the primary objective of the present study was to elucidate the morphogenetic basis for this deceptive pollination biology, as well as to use what is learned about the timing of changes in the emergence and growth of floral organs to develop explicit hypotheses to guide future molecular developmental studies. Based on what is already known about the development of upper and lower floral asymmetry in certain core eudicot taxa (e.g., Hileman et al., 2003; Almeida and Galego, 2005; Kramer, 2007), one might hypothesize the action of a TCP-like transcription factor(s) early in floral development to suppress and thereby effect the staminodization of the upper

1236

July 2009]

HARDY ET AL.—DEVELOPMENT OF DECEPTION IN COMMELINA

1237

MATERIALS AND METHODS Commelina communis inflorescences of varying ages were collected for study from wild outdoor populations. A collection in the summer of 2005 at The Ephrata Bowman’s Club in Ephrata Township, Pennsylvania, USA, was vouchered as Sloat 3 in the James C. Parks Herbarium (MVSC, following Holmgren and Holmgren, 1998). A second collection was made during that same summer from a population located behind the Stayer Education Building, Millersville University (Sloat 2, MVSC). A large collection of FPA-preserved plants, made in the summer of 1999 in New York City, was donated by D. Wm. Stevenson of the New York Botanical Garden (NYBG; unused material is housed in the fluid collections at MVSC). Fresh collections were preserved in FPA (formalin : propionic acid : ethanol; 1 : 1 : 8, v/v) and then transferred to 70% ethanol for storage indefinitely. Specimens were dissected using a dissecting microscope (Olympus SZH10; Olylmpus, Tokyo, Japan). The mature flowers were removed from the cyme and the immature floral buds (often