Soil fertility effects on tree seedling performance are light‐dependent: evidence from a long‐term soil chronosequence

Soil chronosequences are a powerful tool for understanding how limitation of plant growth by nutrients and light changes throughout ecosystem development, but experimental tests of how availability of these resources interact to influence plant performance as ecosystem development proceeds are rare. We utilise the well‐characterised Franz Josef soil chrononosequence in New Zealand, a sequence of sites caused by a retreating glacier that spans ca 120 000 years and that includes all stages of ecosystem development from primary succession through to retrogression. Soil fertility is relatively low at either end of the sequence due to limitation of biological processes initially by N and ultimately by P whereas light availability is lowest at intermediate stages of the sequence dominated by tall forest. Growth and leaf traits of nine woody plant species, including those that occur widely along the chronosequence and those that are restricted to short portions of it, were quantified in a mesocosm experiment. Phytometers of these species were each grown in each of nine soils collected from throughout the chronosequence at either high (30%) or low (2%) light levels; these soil and light conditions represent the full variation observed along the sequence. Plant growth and biomass were greatest in soils from intermediate stages of the chronosequence and in high light. However, the stimulatory effects of soil fertility largely disappeared under shaded conditions that are characteristic of intermediate stages of ecosystem development. Our results demonstrate that long‐term changes in soil fertility and light availability that occur throughout ecosystem development had direct effects on plant species performance, but that there were stronger interactive effects of soils and light availability. Because light and soil resource availability shift predictably but have different trajectories throughout ecosystem development, our results help to understand variation in plant species performance and community assembly along complex environmental gradients.