1202 Ecology, 80(4), 1999, pp. 1202–1220  1999 by the Ecological Society of America MODELING INVESTMENTS IN SEEDS, CLONAL OFFSPRING, AND TRANSLOCATION IN A CLONAL PLANT SHEA N. GARDNER 1,3 AND MARC MANGEL 2,4 1 Center for Population Biology, University of California, Davis, California 95616 USA 2 Department of Environmental Studies, University of California, Santa Cruz, California 95064 USA Abstract. Clonal plants that can switch facultatively between sexual and asexual re- production may respond plastically to the environment. We constructed a dynamic state variable model to examine how the measure of fitness, ramet and genet mortality, and the assimilation rates of a parent and its clonal offspring influence behavioral investments in ramet growth, clonal offspring, seeds, and continued resource translocation to clonal off- spring after establishment. The model leads to predictions that ramet and genet mortality rates and/or the fitness payoff from producing seeds must be high for seed production to capture a proportion of reproductive investments. If seed production occurs as a result of high ramet or genet mortality rates, then results indicate that it is better to produce seeds early in the season, regardless of ramet size. In contrast, if seed production is favored as a result of its large contribution to fitness, then it is predicted to depend on ramet size more than on time. While the total amount of biomass directed to reproduction is predicted to increase with a ramet’s own productivity, the proportion of this biomass invested clonally or sexually depends on the resource environment encountered by that ramet’s clonal offspring; more productive surroundings favor investment in clonal offspring that forage locally, reduce the risk of genet mortality, and increase the expectation for future seed production by the genet. The model we present also suggests that a higher rate of translocation to support clonal offspring benefits a genet when the parent and offspring ramets have contrasting produc- tivities. In addition, the model also leads to the predictions that translocation is more advantageous when the currency of fitness selects for increases in ramet size more than ramet number and when the probability of mortality is correlated among ramets. Key words: clonal plants; clonal reproduction; dynamic model; integration; seeds; sexual repro- duction; state variable; trade-off; translocation. INTRODUCTION In this paper, we aim to explain how the measure of fitness, ramet productivity, and genet and ramet mor- tality affect investments by clonal plants in seeds and vegetative propagation. Plants capable of producing po- tentially independent clonal offspring, called ramets, can spread through the habitat (de Kroon and Hutchings 1995). By enlarging a genet, the production of ramets increases a genet’s capacity for making more seeds in the future (Watson 1984, Madga et al. 1993), as well as reduces the risk of genet mortality (Cook 1979, Eriksson and Jerling 1990). Producing seeds or spores provides a genet the chance to disperse a fraction of offspring to colonize new, perhaps richer or safer, habitats, and may be particularly advantageous if rates of ramet or genet mortality are high (Hartnett 1987). Consequently, clonal Manuscript received 7 August 1997; revised 12 June 1998; accepted 15 June 1998. 3 Current address: Natural Environment Research Council (NERC) Centre for Population Biology, Imperial College at Silwood Park, Ascot, Berkshire SL5 7PY, UK. 4 Address correspondence to this author: E-mail: msmangel@cats.ucsc.edu. plants face reproductive trade-offs (Sohn and Policansky 1977, Law et al. 1983, Westley 1993, Worley and Harder 1996) that are affected by selection on foraging, repro- duction, dispersal, and growth. We focus on the behavioral, rather than the genetic, aspects of reproduction. Therefore, when we refer to seeds or sexual reproduction, it may be interpreted more loosely as investment in any offspring that do not remain connected to a parent ramet and that may dis- perse and colonize distant areas. Although these off- spring are often outcrossed seeds, they may also be a result of agamospermy (Crawley 1997) or, in the case of aquatic plants, widely dispersing clonal fragments (Room 1990, Philbrick and Les 1996). In addition, al- though sometimes clonal offspring may disperse farther than seeds (Cain and Damman 1997), seeds are often capable of dispersing away from the parent genet (Er- iksson 1989, 1992). Since rates of seedling survival in established populations are so low (Harper 1977, Wik- berg et al. 1994, Nishitani and Kimura 1995, Jonsson et al. 1996), it may be that the greatest gains from producing seeds are realized from those few seeds that travel farther afield or that establish after major dis-