Seed Mortality in Daucus Carota Populations: Latitudinal Effects1 Elizabeth

Daucus carota, a common herbaceous weed, grows over a wide latitudinal range in eastern North America. Viability and germination tests of mature seeds collected from 36° to 45°N were conducted to measure predispersal seed mortality. Viability and germination declined as latitude of the seed source decreased. Only 30-50% of the seeds from southern populations germinated owing to high embryo inviability and absence of embryos. Sixty to ninety percent of the seeds from northern populations germinated. Reciprocal planting of seeds in outdoor experimental plots at three latitudes and testing of seeds over two generations together showed that the environment in which seeds mature, rather than environmental preconditioning over generations or genetically-based differences among populations, explain this variation in germination ability. Within-latitude germination declined in experimental plots as population age of the seed source within latitudes increased. The data indicate that predispersal seed mortality can influence local population persistence and that seed mortality is an increasingly important factor in population regulation at the southern limit of the species' range. ALTHOUGH MORTALITY can occur at the seed stage in a plant life cycle, we do not yet know how strongly this mortality regulates the size of natural plant populations. The loss of a few hundred seeds may be inconsequential in view of the thousands entering the seed pool (Harper,1977).Conversely,seeddeathmayreduce the seed pool so dramatically that it renders post-germination mortality "trivial" (Hickman, 1979). To assess better the role that seed mortality plays in regulating natural plant populations, ecologists have recently begun to quantify seed mortality and identify its causes. In Minuartia uniflora, mortality at the seed stage far surpasses mortality at any other stage (Sharitz and McCormick, 1973). Predation (e.g., Janzen,1971; DeSteven,1983) and abortion (e.g., Stephenson, 1981; Aker, 1982; Lee and Bazzaz, 1982) can both act at the flower, fruit, or seed levels to lower seed set, and seed predation can severely reduce seedling recruitment in Haplopappus (Louda 1982, 1983). To provide further information about seed morl Received for publication 17 November 1983; revision accepted 14 July 1984. I thank C. Capps, E. Connor, S. Lowe, and E. Smith for help collecting data, North Carolina State Agricultural and Technical University, the Matthaei Botanical Gardens, and W. Canter for providing space for the outdoor experiments, the National Capitol Commission of Canada for permission to collect seeds, T. Meagher for use of his program for the G2 test, and C. Augspurger and D. DeSteven for helpful comments on an earlier draft of this manuscript. This study was supported by NSF Grant DEB 8012154 and by a UNC-G Research Council Grant. tality in natural populations, I here describe the variation in predispersal seed mortality in Daucus carota L. ssp. carota (Small,1978) and examine the contributions of latitude and population age to this variation. MATERIALS AND METHODS-The data come from four viability and germination experiments conducted on seed samples collected over the species' latitudinal range, from southern Canada to the Georgia piedmont, in eastern North America. Seed germination and viability studies are usually conducted to identify environmental conditions that induce germination in viable seeds (e.g., Ruhland, 1965; Kozlowski, 1972; Mayer and Poljako-Mayber, 1975; Harper, 1977; Baskin and Baskin 1982) or to identify the provenance that produces seeds best suited for agricultural or forestry purposes (see Barton,1967). In D. carota germination tests can also be used to assess mortality during fruit development. Fruit development continues whether or not the embryo dies during development, and the seed mortality agent does not remove fruits from the plant (Flemion and Henrickson,1949; Flemion and Olson, 1950; Robinson, 1954). Experiment 1: indoor germination testsDuring the summer of 1981 I sampled plants at four latitudes: Greensboro, NC-36°N, Charlottesville, VA-38°N, Ann Arbor, MI42°N, and Ottawa, Canada-45°N. (Fig.1). One population was sampled per latitude. The flowering season for D. carota spans approximately 2 months; in North Carolina flowering occurs 1175 Made available courtesy of Botanical Society of America: http://www.botany.org/ ***Reprinted with permission. No further reproduction is authorized without written permission from the Botanical Society of America. This version of the document is not the version of record. Figures and/or pictures may be missing from this format of the document.***