Plant growth is an essential ecological process, integrating across scales from physiology to community dynamics. Predicting the growth of plants is essential to understand a wide range of ecological issues, including competition, plant-herbivore interactions and ecosystem functioning.
A challenge in modeling plant growth is that growth rates almost universally decrease with increasing size, for a variety of reasons. Traditional analyses of growth are hampered by the need to remain within the structures of linear models, which handle this slowing poorly. We demonstrate the implementation of a variety of non-linear models that are more appropriate for modeling plant growth than are the traditional, linear, models.
Ecological inference is frequently based on growth rates, rather than model parameters. Traditional calculations of absolute and relative growth rates assume that they are invariant with respect to time or biomass, which is almost never valid. We advocate and demonstrate the calculation of function-derived growth rates, which highlight the time- and biomass-varying nature of growth. We further show how uncertainty in estimated parameter values can be propagated to express uncertainty in absolute and relative growth rates.
The use of non-linear models and function-derived growth rates can facilitate testing novel hypotheses in population and community ecology. Even so, we acknowledge that fitting non-linear models can be tricky. To foster the spread of these methods, we make many recommendations for ecologists to follow when their hypotheses lead them into the subject of plant growth.