RGR predicts short-term competitive dominance and exclusion in a field experiment
Our nitrogen and combined nitrogen and phosphorus addition treatments created productive conditions and reduced plant diversity while phosphorus addition alone did not significantly affect either productivity or diversity. In the third year of our field experiment, there was a marginally significant interaction between nitrogen and phosphorus addition on biomass production (F1,20=3.8, P=0.065) and plant species richness (F1,20 = 3.7, P = 0.069) (Table S4). Nitrogen addition increased biomass production from an average of 101 ± 11 g 0.25 m-2 (mean ± s.e.m.) in the control plots to 140 ± 11 g 0.25 m-2 and decreased species richness from 36 ± 2 species 0.25 m-2 to 22 ± 2 species 0.25 m-2. In contrast, the levels of biomass production (114 ± 11 g 0.25 m-2) and species richness (35 ± 2 species 0.25 m-2) under phosphorus addition were indistinguishable from those observed in the control plots. The combination of nitrogen and phosphorus addition had a large effect on productivity, which increased to 198 ± 11 g 0.25 m-2, while this treatment resulted in a smaller but still significant decrease in plant species richness than observed with just nitrogen treatment leading to 28 ± 2 species 0.25 m-2.
Similar to the results of our common garden experiment, rankings of species growth rates changed with both growing season and nutrient treatments (Fig. S4). We found that the relationship between early season relative differences in species growth rates and relative differences in species biomass varied with both the growing season and nutrient treatments (Fig. 2A, Fig. S5A). The percentage of variance explained was maximum at day 150 in the control (R2 = 0.29, F1,169 = 70.1, P <0.001), 146 with nitrogen addition (R2 = 0.35, F1,169= 89.3, P <0.001), 164 with phosphorus addition (R2 = 0.11, F1,151 = 18.0, P <0.001) and 146 with nitrogen and phosphorus addition (R2 = 0.26, F1,151 = 53.2, P <0.001). When significant, relationships were always positive (Fig. 2A, Fig. S5A). Results based on relative differences in species growth rates calculated between the two first measurements of biomass confirmed that early differences in growth rates predict competitive dominance at harvest except in the phosphorus addition treatment (Fig. S6).
We found that the relationship between early season species growth rate values and the likelihood of loss of a species varied with both the growing season and nutrient treatments (Fig. 2B, Fig. S5B). The percentage of variance explained was maximum at day 146 in the control (R2 = 0.05, F1,118 = 2.9, P = 0.11), 147 with nitrogen addition (R2 = 0.12, F1, 118 = 11.5, P = 0.003), 177 with phosphorus addition (R2 = 0.07, F1,112 = 3.6, P = 0.06) and 172 with nitrogen and phosphorus addition (R2 = 0.11, F1,118 = 11.6, P = 0.001). Short-term competitive exclusion could only be predicted by early differences in species growth rate under productive conditions (nitrogen and nitrogen & phosphorus addition) and, when significant, relationships were always negative (Fig. 2B, Fig. S5B). Under unproductive conditions (control and phosphorus addition), short-term competitive exclusion could not be predicted from early differences in growth rate.