Both Gentiana saxosa and G. serotina are protandrous. When flowers open, pollen is presented extrorsely around the closed stigma for one to six days. As the stigma opens, the stamens curve toward the corolla lobes. The length of the female phase, and therefore reproductive flower life, is determined by pollination, although in both species the corolla may remain fresh for longer than one month. Fresh female-phase flowers close at night and fail to reopen on the day following pollination. After five days in the female phase, flowers reacted less quickly to pollination and seed production was reduced; flowers pollinated on their tenth day of stigma presentation produced no seed although they appeared fresh. Senescence of unpollinated flowers differed between species: in G. saxosa the flowers remained open and gradually deteriorated, but in G. serotina the flowers eventually closed before full senescence. Pollination-induced flower senescence has been demonstrated for a number of other angiosperms, and the usual reactions to pollination are corolla abscission, color change, or wilting. In Gentiana, the closed corolla enfolds the large superior ovary and may serve to protect it from predators as well as prevent further pollinator visits. Pollination-induced flower senescence probably also minimizes flower maintenance costs by ensuring that the flower functions no longer than necessary. One correlate of this phenomenon in hermaphroditic flowers is protandry, which ensures pollen dispatch before flower closure. Floral senescence may be either time-dependent (endogenous) or exogenous (usually pollination-induced). However, there have been few detailed investigations of the factors that determine floral senescence, and hence, floral longevity (Primack, 1985). Most such studies have concentrated on corolla color changes or other physiological reactions that follow pollination and signal senescence (e.g., Arditti et al., 1973; Arditti, 1976; Gottsberger, 1971; Gori, 1983; Strauss & Arditti, 1984; Casper & La Pine, 1984; Halevy, 1984), and less attention has been afforded structural changes such as wilting, flower closure, and corolla abscission (Mayak & Halevy, 1980). Most of this research is concerned with the proximate determinants of floral longevity rather than the evolution of particular responses (but see Stead & Moore, 1979; Gori, 1983; Casper & La Pine, 1984; Devlin & Stephenson, 1984). The paucity of research on the evolutionary aspect of flower senescence is somewhat surprising, because pollination-induced senescence in particular may have important consequences for the pollination system and ultimately for the plant's overall reproductive strategy. For instance, in hermaphrodite flowers pollination-induced flower senescence will limit the duration of pollen and stigma presentation and so may influence or be influenced by the extent and nature of dichogamy (Lloyd & Webb, 1986). New Zealand species of Gentiana (in the southern group of Philipson, 1972) are protandrous (Thomson, 1881; Simpson & Webb, 1980; Webb, 1 984a). Their large, relatively simple flowers make them particularly suitable subjects for experimental studies of flower function. This paper describes the response of flowers of two species to pollination and reports the results of experiments to determine the functional duration of male and female phases in terms of pollen presentation and seed production. MATERIALS AND METHODS The two species of Gentiana selected were those that grew best in cultivation. Gentiana saxosa Forster f. grows naturally in coastal sites of southern South Island and Stewart Island, New Zealand; plants were collected from Curio Bay, Southland. Gentiana serotina Cockayne occurs in grassland in inland central South Island; plants were collected from Lake Lyndon, Canterbury. The plants were grown in clay pots in an insectproof cage in greenhouses at Lincoln, Canter1 We thank I. C. Brown and garden staff at Lincoln, Canterbury, for maintaining the plants in cultivation, J. Miles for assistance with photographs, and P. Brooke for drawing Figures 2 and 3. We are grateful to L. F. Delph-Lively, E. Edgar, and D. G. Lloyd for comments on a draft of the manuscript. 2 Botany Division, Department of Scientific and Industrial Research, Private Bag, Christchurch, New Zealand. ANN. MISSOURI BOT. GARD. 74: 51-57. 1987. This content downloaded from 157.55.39.144 on Mon, 25 Jul 2016 04:48:36 UTC All use subject to http://about.jstor.org/terms 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN [VOL. 74 TABLE 1. Results of controlled pollinations to determine self-compatibility in Gentiana saxosa and G.
Read full abstract