Clonal reproduction within Cyperaceae: Allocation, translocation, and response to nutrient availability

Background/Question/Methods

Globally, wetlands are dominated by the presence of graminoids which tend to invest heavily in clonal growth, i.e., the production of genetically-identical new individuals (ramets) through rhizome or stolon development. Understanding patterns in ramet production is essential to predicting architecture of entire genets, which, in turn, informs trends in plant growth rates and competitive ability. While substantial work exists on genet architectural configurations over long growth periods, quantification of parental investment in daughter ramets has remained surprisingly unexplored. We hypothesized that production of each new ramet requires a constant absolute amount of resources and therefore larger plants (e.g., under higher nutrients) should allocate a smaller percentage of parent biomass per daughter ramet, although they may make more daughter ramets.  To evaluate this hypothesis, two common wetland sedges—Schoenoplectus tabernaemontani and Eleocharis elliptica—were grown in a greenhouse under high and low nitrogen levels. After production of three to twelve ramets, genets were harvested immediately after the terminal ramet breached the soil surface, isolating parental investment into daughter ramets. Genets were then divided into individual ramets, and further broken down into component tissue types. 

Results/Conclusions

Logistic regressions of parent size at the time of new daughter production imply that a threshold size may need to be reached before a daughter ramet can be generated.  While nutrient addition increased the overall growth of genets and the rate at which daughter ramet production occurred, the percent of parent biomass allocated to daughter ramets remained constant across parent size between N treatments for both species. Because percent allocation to daughters remained unchanged relative to parent size, daughter ramets varied considerably in mass, refuting our hypothesis that daughter ramet cost is constant. Translocation, the movement of resources within a genet, also appears to have a significant effect in some cases; incorporating the biomass of up to five ramets preceding the terminal ramet into regression models of new ramet production increased fit for one of the two species (S.validus: 1 parent allocation- R²=0.42, 5 parents allocation- R²= 0.53; E. elliptica: 1 parent allocation- R²=0.64, 5 parents allocation- R²= 0.41), suggesting that resource sharing within genets differs among species. Overall, these findings indicate that the cost of clonal reproduction within sedges is highly variable, yet predictable as a proportion of parent size. Even so, nuances between individual species prevent broad generalization.