Partitioning the components of relative growth rate: how important is plant size variation?

Plant growth plays a key role in the functioning of the terrestrial biosphere, and there have been substantial efforts to understand why growth varies among species. To this end, a large number of experimental analyses have been undertaken; however, the emergent patterns between growth rate and its components are often contradictory. We believe that these conflicting results are a consequence of the way growth is measured. Growth is typically characterized by relative growth rate (RGR); however, RGR often declines as organisms get larger, making it difficult to compare species of different sizes. To overcome this problem, we advocate using nonlinear mixed-effects models so that RGR can be calculated at a standard size, and we present easily implemented methods for doing this. We then present new methods for analyzing the traditional components of RGR that explicitly allow for the fact that (log)(RGR) is the sum of its components. These methods provide an exact decomposition of the variance in (log)(RGR). Finally, we use simple analytical and simulation approaches to explore the effect of size variation on growth and its components and show that the relative importance of the components of RGR is influenced by the extent to which analyses standardize for plant size.