Date Published: March 23, 2017
Publisher: Public Library of Science
Author(s): Tobias Ceulemans, Eva Hulsmans, Sigi Berwaers, Kasper Van Acker, Olivier Honnay, Harald Auge.
Anthropogenic activities have severely altered fluxes of nitrogen and phosphorus in ecosystems worldwide. In grasslands, subsequent negative effects are commonly attributed to competitive exclusion of plant species following increased above-ground biomass production. However, some studies have shown that this does not fully account for nutrient enrichment effects, questioning whether lowering competition by reducing grassland productivity through mowing or herbivory can mitigate the environmental impact of nutrient pollution. Furthermore, few studies so far discriminate between nitrogen and phosphorus pollution. We performed a full factorial experiment in greenhouse mesocosms combining nitrogen and phosphorus addition with two clipping regimes designed to relax above-ground competition. Next, we studied the survival and growth of seedlings of eight common European grassland species and found that five out of eight species showed higher survival under the clipping regime with the lowest above-ground competition. Phosphorus addition negatively affected seven plant species and nitrogen addition negatively affected four plant species. Importantly, the negative effects of nutrient addition and higher above-ground competition were independent of each other for all but one species. Our results suggest that at any given level of soil nutrients, relaxation of above-ground competition allows for higher seedling survival in grasslands. At the same time, even at low levels of above-ground competition, nutrient enrichment negatively affects survival as compared to nutrient-poor conditions. Therefore, although maintaining low above-ground competition appears essential for species’ recruitment, for instance through mowing or herbivory, these management efforts are likely to be insufficient and we conclude that environmental policies aimed to reduce both excess nitrogen and particularly phosphorus inputs are also necessary.
Anthropogenic activities have severely altered nutrient fluxes through excess rates of fertilizer application following agricultural intensification and elevated levels of atmospheric deposition, causing an estimated ten- to fiftyfold increase of nitrogen and phosphorus inputs in natural and semi-natural ecosystems worldwide [1–3]. Negative environmental effects of nutrient enrichment are commonly attributed to eutrophication, i.e. the increased biomass production of dominant plant species following the release from nutrient limited growth due to low nutrient availability [4, 5]. As this increase leads to less light in the understory, plant seedlings and small-statured and slow-growing plant species in the understory of the vegetation are outcompeted for light, resulting in plant species loss. Indeed, artificially supplying light to the understory of experimental grasslands experiencing increased above-ground competition, effectively prevented competitive exclusion of low-statured plant species and supported successful seedling recruitment . Also, a series of global grassland experiments showed that herbivory offsets plant species loss following nutrient enrichment by increasing ground-level light in otherwise more productive vegetation .
The experiment was conducted in a greenhouse in Leuven in 2014 (Belgium). The average temperature in the greenhouse was 18.3°C ± 1.8 SD, relative humidity of 69.0% ± 18.0 SD and 49.7 μmol m-2 s-1± 57.1 SD of light (all averaged over the natural day-night regime during the entire experiment). We performed a factorial experiment using grassland mesocosms. To construct the experimental mesocosms, we filled polymer containers of 42 cm by 38 cm and 10 cm height with a mixture of one part commercially available coarse Rhine sand and two parts of soil collected in a nutrient-poor grassland near the city of Leuven (Belgium). We treated the mesocosms a single time with four different nutrient solutions representing i) a control treatment, in which mesocosms were supplied with 500 mL of a solution of 13.9 g NaCl per liter of deionized water (C-treatment); ii) a nitrogen treatment, supplied with 500mL of a solution of 28.3 g NH4NO3+13.9 g NaCl per liter (N-treatment, equivalent of 309.9 kg N ha-1); iii) a phosphorus treatment with 500 mL of a solution of 16.4 mL concentrated H3PO4 per liter, buffered with 13 mL of 50% NaOH (P-treatment); and iv) a nitrogen + phosphorus treatment supplied with 500 mL of a nutrient solution, consisting of 250 mL of a solution of 32.8 mL concentrated H3PO4 per liter buffered with 13mL of 50% NaOH and 250 mL of a solution of 56.6 g NH4NO3 per liter (NP-treatment, same amounts of nitrogen and phosphorus as in the separate N- and P-treatments). To assess to what extent the nutrient treatments raised nutrient availability in the soil of the mesocosms, we randomly took three soil samples per mesocosm four weeks after the nutrient application. We quantified soil pH using a glass electrode in a 1:25 soil and deionized water mixture shaken for 30 minutes. Soil phosphorus availability was determined through Olsen-P extraction followed by colorimetric analysis using the molybdenum blue method . Soil nitrogen availability was determined by a 1 M KCl extraction of NH4+ and NO3- followed by a colorimetric analysis using a segmented flow analyzer (, Skalar, Breda, The Netherlands).
After planting, only twelve seedlings across different treatments died in the first week of the experiment and had to be replaced by new seedlings (3 B. media, 4 H. radicata and 5 S. pratensis). There was a significant difference in mean light availability at ground level in the vegetation of the mesocosms in the short-clipped treatment (71.5±16.8 μmol m-2 s-1) as opposed to the long-clipped treatment (35.7±7.8 μmol m-2 s-1). There were no significant differences in light availability among the different nutrient treatments (S1 Fig). The soil mixtures of the mesocosms contained 19.35±1.30 mg N kg-1 and 29.79±2.57 mg P kg-1 in the control treatment, 44.24±3.23 mg N kg-1 and 28.28±1.51 mg P kg-1 in the N-treatment, 18.56±1.05 mg N kg-1 and 66.44±4.27 mg P kg-1 in the P-treatment and 42.05±2.44 mg N kg-1 and 63.23±3.36 mg P kg-1 in the NP-treatment (S2 and S3 Figs). These levels are comparable to a gradient of soil nutrient levels in European grasslands (S4 and S5 Figs). Soil pH did not differ significantly between treatments (S6 Fig).
The most common hypothesis to predict environmental consequences of nutrient enrichment predicts competitive exclusion owing to an increase in above-ground biomass production (e.g. [6, 23]). Our results support this hypothesis by showing reduced seedling survival in the long-clipped treatment, designed to simulate higher above-ground competition, for all but the two grass species in the experiment. Furthermore, seedling growth rate was consistently higher in the short-clipped treatments. However, under the assumption that higher above-ground competition is uniquely responsible, the short-clipped treatment, designed to relax above-ground competition, should cancel out the negative impact of adding nutrients. This was not the case as all but one species (L. vulgare) showed a decline in seedling survival after nitrogen and/or phosphorus fertilization across both clipping treatments. The negative effects of nutrient addition on seedling survival of only one species was (partly) alleviated by lower levels of above-ground competition in the short-clipped treatments (H. radicata). However, lower above-ground competition did not fully alleviate its reduced seedling survival, indicating that unique negative effects of nutrient enrichment persist. These results suggest that higher levels of above-ground competition will negatively influence population recruitment, irrespective of nutrient enrichment. In other words, at any given level of soil nutrient availability, lower vegetation productivity should also allow for higher seedling survival. However, our results also indicate that fertilization negatively impacts species’ recruitment irrespective of the extant above-ground competition as all investigated species but H. radicata showed reduced seedling survival, despite lower levels of competition. Furthermore, only S. pratensis showed increased seedling growth rates under low competition, indicating nitrogen and phosphorus co-limited growth, which ultimately may alleviate the negative effects of nutrient enrichment on seedling survival.