RGR predicts short-term competitive dominance in a common garden experiment
Our nutrient addition treatment successfully created productive conditions with high nutrient and low light availability, while our sucrose addition treatment successfully created unproductive conditions with limited nutrient and high light availability. Sucrose addition reduced the amount of mineral nitrogen available to plants, in the second year of our common garden experiment, from an average of 2.3 ± 0.3 g N m-2 following nutrient addition to 0.9 ± 0.3 g N m-2 following sucrose addition. It also reduced biomass production in monocultures from 745 ± 39 g m-2(mean ± s.e.m.) following nutrient addition to 274 ± 25 g m-2 following sucrose addition (F1,48= 102.34, P < 0.001) and increased understory light availability measured just before the harvest from 13 ± 3% following nutrient addition to 65 ± 5% following sucrose addition (F1,48 = 54.25, P < 0.001) (Table S2).
After two years of treatment, the four parameters of the logistic growth curves used to calculate daily RGR of five perennial grass species growing in monoculture varied across species and nutrient treatments (Fig. S2A; Table S3). As a result, the rankings for species’ growth rates changed with both the growing season and nutrient treatment (Fig. S2B). For example, relatively high RGR early in the season was observed for H. lanatus under productive conditions, while A. pratensis had the highest early RGR under unproductive conditions.
We found that early season (day t =53) relative differences in species growth rates in monoculture were positively associated with relative differences in species biomass at harvest (day 171) in pairwise (Fig. 1A, Fig. S3A) and five-species mixtures (Fig. 1B, Fig. S3B) under both productive and unproductive conditions. The percentage of variance explained was maximum between day t = 53 and t = 112. This positive association was observed up to day t=133. Relative differences in species’ growth rates became smaller as the season progressed until they became negatively associated with differences in species biomass (from day t =135 in the year – 14.05.2008) (Fig. S3). The percentage of variance in species biomass at harvest explained by relative differences in species’ growth rates during the early stage of the growing season was approximately 50% under both productive and unproductive conditions for the pairwise mixtures (Fig. S3A) and approximately 60% under the productive condition and 50% under the unproductive condition for the five species mixtures (Fig. S3B).