What do seedlings die from and what are the implications for evolution of seed size?

Abstract

Plants face a fundamental trade-off between producing many small seeds, each with a low probability of successfully establishing as a seedling and producing fewer, better provisioned seeds with higher establishment probabilities (reviewed in Leishman et al. 2000, Westoby et al. 2002). In the absence of frequency-dependent interactions between seedlings, this trade-off is predicted to result in a single optimum seed mass in a given environment (Smith and Fretwell 1974). However, plant communities usually contain species whose seed sizes span 5-7 orders of magnitude (Leishman et al. 2000). Theory to explain the wide range of seed mass strategies coexisting in plant communities invokes frequency-dependent processes. Consider, for example, the game-theoretical seed-size/number trade-off model (developed by Geritz 1995 and extended to track population and evolutionary dynamics by Rees and Westoby 1997). Under this model, a large-seeded strategy can be invaded by a small-seeded strategy because small-seeded strategies have greater seed output and therefore reach a proportion of living-sites that are not reached by largerseeded strategies. However, small-seeded strategies can be invaded by large-seeded strategies because seedlings from large seeds are able to outcompete the relatively poorly-provisioned seedlings produced by small seeds. Thus neither strategy excludes the other and a range of strategies can co-exist. Two assumptions of this model are 1) that small seeded species produce more seeds from a given ground area occupied and in this sense are better colonisers than large-seeded species and 2) that large-seeded species win sites through superior competitive ability. There is strong evidence for the first: small-seeded species do produce more seeds for a given amount of effort (Greene and Johnson 1994, Jakobsson and Eriksson 2000, Aarssen and Jordan 2001, Henery and Westoby 2001). This higher fecundity should give small-seeded species a greater number of colonising opportunities (empirically demonstrated in some, but not all situations; evidence reviewed in Leishman 2001). Evidence for the second assumption is not as strong. A positive relationship has often been found between seed size and competitive ability across species (Eriksson 1997, Turnbull et al. 1999, Leishman 2001). However, there is debate as to whether the competitive hierarchies observed are strong enough to maintain a range of coexisting strategies (Coomes and Grubb 2003). Further, as Leishman (2001) pointed out, competition between seedlings might not be important in deciding species composition in most real vegetation. Almost all studies of seedling-seedling competition have been performed at artificially high densities achieved through seed supplementation or seedling planting. Our field observations have led us to suspect that natural populations of seedlings seldom occur at high enough densities for seedling-seedling competition to be a major cause of seedling mortality. Evidence that seedling survival is dependent on competition under natural conditions and at natural densities is sparse. Silvertown and Bullock (2003) reported both frequencyand density-dependent mortality among seedlings establishing in experimentally created gaps in calcareous grassland in Oxfordshire, England and concluded that the evidence supported the assumptions of the seed-size/ number trade-off model. Coomes et al. (2002) studied spatial patterning among annuals establishing on sand dunes in Britain. They found negative associations between two large-seeded species and smaller-seeded species at a small spatial scale (interpreted as evidence for local exclusion of small-seeded species by largeseeded species), but only in one year out of the three studied, when seedling densities were particularly high. These authors concluded that the frequency of competitive interactions between species was low enough that competition was unlikely to be a major determinant of community structure. Moles and Westoby (in press)