Meristem Allocation: A New Classification Theory for Adaptive Strategies in Herbaceous Plants

Abstract

A major goal in studies of evolutionary ecology has been to define a classification of adaptive strategies that can be used for interpreting variation in fitness within species and patterns of diversity and species coexistence within and between habitats. An adaptive strategy is often associated with a particular pattern of allocation (e.g. of a limiting resource) that maximizes fitness in a given environment (Solbrig 1994). Different allocation patterns are assumed to be a consequence of tradeoffs between life history traits resulting from certain constraints (Stearns 1992). Hence, organisms may exhibit sets of traits that are predictably related to their ecology (Grime et al. 1988). Virtually all adaptive strategy theories based on allocation patterns in plants define strategies in terms of biomass allocation. A number of life history traits have been quantified in this way. The most common of these involves biomass allocation to reproductive versus vegetative functions (i.e. reproductive effort) (Hancock and Pritts 1987, Wilson and Thompson 1989, Bazzaz and Ackerly 1992). Biomass allocation to above versus below ground structures (Tilman 1988) and to growth, reproduction, structure, and defence (Taylor et al. 1990) have also been used to identify adaptive strategies in plants. Watson (1984) traces the roots of biomass allocation in life history strategies to the animal ecology literature (e.g. Cody 1966). Extending the use of biomass allocation to plant strategy theories seems to have arisen from convenience and established precedent in animal studies (Watson 1984). However, the use of biomass in allocation measures has a number of limitations. Problems have been most widely noted in studies of reproductive effort in plants. Partitioning biomass to reproductive and vegetative functions is difficult when support structures that are essential to, but not directly involved in reproduction, are considered. This is especially difficult when these structures have photosynthetic capability (a vegetative function) (Reekie and Bazzaz 1987a, Bazzaz and Ackerly 1992). Studies of biomass allocation also assume that there is one resource (i.e. carbon) that limits both growth and reproduction. However, different resources may limit growth in different environments and at different times in a plant's life (see Bazzaz and Ackerly 1992 for a review). A critical feature of life history strategy theories is that increased allocation to one function (e.g. reproduction) must tradeoff with a decreased allocation to another function (e.g. growth). To be of evolutionary significance, cost must be measured in terms of a true limiting resource (Watson 1984, Watson and Casper 1984, Reekie and Bazzaz 1987b, Bazzaz and Ackerly 1992). Another major criticism of biomass allocation is that it does not reflect internal constraints that may control plant growth and life history (Salomonson et al. 1994). Growth in grasses, for example, is constrained by tiller production (Law 1979). Any function that reduces tiller survival (e.g. flowering) decreases growth as only those tillers that survive can produce new tillers. Developmental choices of a plant can affect fitness through overall plant form (architecture). Biomass does not determine plant form (Smith 1984). In this paper we develop an alternative to biomassbased strategy theories based on allocation of above ground meristems in herbaceous flowering plants. Previous studies of meristems in ecology have focussed on several issues. Meristem demography (the births and deaths of modules within a plant) has been used to study growth patterns within a plant (Bazzaz and Harper 1977, Maillette 1982, 1985, 1987, 1990, Lehtila et al. 1994). Fitness may be defined at the level of the module (Tuomi and Vuorisalo 1989a, b). Modules are the units of reproduction, growth and survival and so the cycle of meristems producing new meristems may be used to estimate fitness in clonal plants (Fagerstr6m 1992, Wikberg 1995). Adaptive architecture has been studied via tradeoffs between growth (i.e. branching) and reproduction (Smith 1984, Watson 1984, Geber