Estimating dynamic allocation to clonal plant growth using static measurements

Background/Question/Methods

Clonal plants dominate many habitats such as grasslands and wetlands, and the majority of plants on earth exhibit at least some potential for clonal growth.  However, there is a large amount of variation among clonal plants in the amount of resources translocated to clonal growth and the degree of clonal integration.  Like many plant traits, clonal integration can potentially exert strong influences on population and community dynamics, but our advancement in understanding the influence of this trait is hindered by difficulties in measurement.  The three most common techniques for measuring clonal integration (isotope tracer studies, severing clonal connections, and growing connected ramets in heterogeneous resource patches) all pose nontrivial methodological problems.  Here we introduce two new methods to estimate translocation using simple static plant measurements.  The first of these measures uses logistic regression to describe the minimum amount of translocation required for clonal reproduction, while the second uses a mathematical model of source-sink dynamics for nutrients and/or photosynthate to estimate ontogenetically-determined allocation rates.  These measures of clonality were validated through simulations, and applied to empirical data to show differences in clonality among 10 wetland plant taxa common in the Laurentian Great Lakes region.

Results/Conclusions

The first method we developed was successful at predicting the minimum-required resource translocation needed for clonal reproduction in simulations.  Application of this method to empirical data also showed wide variation among common wetland taxa in the minimum requirement for successful clonal reproduction (p < 0.01).  Interestingly, this minimum requirement generally increased concomitantly with species' maximum size, meaning larger species required more carbon and/or nutrients for clonal reproduction.  As a result, when the minimum requirement for reproduction was expressed as a proportion of maximum ramet biomass, this proportion was constant between 5-15% for all taxa, regardless of size.  Our second method, which estimated ontogenetically-related source-sink dynamics across clonal connections, was successful at predicting age-related shifts in translocation in a simulation study.  However, when applied to empirical data, there was a large amount of variance associated with model-derived estimates, resulting in low-precision estimates.  Because the parameters estimated by this method contain more detailed information about translocation, we believe more empirical data are needed to derive accurate parameter estimates of translocation using this method.