Size distributions and dispersions along a 485-year chronosequence for sand dune vegetation

Using a sand dune chronosequence that spans 485 years of primary succession, we collected nearest-neighbor vegetation data to test two predictions associated with the traditional "size-advantage" hypothesis for plant competitive ability: (1) the relative representation of larger species should increase in later stages of succession; and (2) resident species that are near neighbors should, over successional time, become more similar in plant body size and/or seed size than expected by random assembly. The first prediction was supported over the time period between mid to later succession, but the second prediction was not; that is, there was no temporal pattern across the chronosequence indicating that either larger resident species, or larger seeded resident species, increasingly exclude smaller ones from local neighborhoods over time. Rather, neighboring species were generally more different from each other in seed sizes than expected by random assembly. As larger species accumulate over time, some relatively small species are lost from later stages of succession, but species size distributions nevertheless remain strongly right-skewed-even in late succession-and species of disparate sizes are just as likely as in early succession to coexist as immediate neighbors. This local-scale coexistence of disparate sized neighbors might be accounted for-as in traditional interpretations-in terms of species differences in "physical-space-niches" (e.g., involving different rooting depths), combined with possible facilitation effects. We propose, however, that this coexistence may also occur because competitive ability involves more than just a size advantage, with traits associated with survival (tolerance of intense competition) and fecundity (offspring production despite intense competition) being at least equally important.