Belowground plant competition: uncoupling root response strategies of peas

Abstract

Belowground plant competition has been shown to induce varying responses, from increases to decreases in root biomass allocation or in directional root placement. Such inconsistencies could result from the fact that root allocation and directional growth were seldom studied together, even though they might represent different strategies. Moreover, variations in belowground responses might be due to different size hierarchies between plants, but this hypothesis has not been studied previously. In a greenhouse rhizobox experiment, we examined the way both root allocation and directional root placement of Pisum sativum are affected by the size and density of Festuca glauca neighbours, and by nutrient distribution. We found that root allocation of P. sativum increased with the density and size of F. glauca. By contrast, directional root placement was unaffected by neighbour size and increased either towards or away from neighbours when nutrients were patchily or uniformly distributed, respectively. These results demonstrate that directional root placement under competition is contingent on the distribution of soil resources. Interestingly, our results suggest that root allocation and directional placement might be uncoupled strategies that simultaneously provide stress tolerance and spatial responsiveness to neighbours, thus highlighting the importance of measuring both when studying belowground plant competition.

Keywords: Pisum sativum; belowground competition; nutrient heterogeneity; plant–plant interactions; root allocation; root placement.

Figure 1.

Schematic illustration of the experiment exploring the ability of Pisum sativum plants to integrate information on nutrient distribution, neighbour size and neighbour density in their spatial belowground responses. The target P. sativum plant (in green) was planted in the rhizobox centre. Festuca glauca was used as a neighbour plant (in grey) with either small (a,b,e,f) or big (c,d,g,h) sizes, and single (a,c,e,g) or double (b,d,f,h) densities. Soil nutrient distribution was either patchy (a–d, brown column) or uniform (e–h, yellow background). Single neighbours were located on the same rhizobox section as the nutrient patch (a,c). Included are pictures from the experiment of rhizoboxes in the patchy soil nutrient distribution with a single small (i) or double big (l) neighbours, and in the uniform soil nutrient distribution with double small (j) or a single big (k) neighbour.

Figure 2.

Responses (means ± s.e.) of P. sativum to nutrient distribution (uniform versus patchy) as well as the size (small versus big) and density (single versus double) of neighbouring F. glauca, in root biomass (a), shoot biomass (b), root allocation (root to shoot biomass) (c), and root length (d). Different lowercase letters indicate statistically significant pairwise comparisons within the uniform or patchy nutrient distribution treatments (with the Benjamini–Hochberg false discovery rate correction).

Figure 3.

Standardized major axis (SMA) relationships between root and shoot biomass of P. sativum, according to the different treatment combinations of neighbour size and density (see table 2). Also shown are slope values for each treatment combination and their probability of equaling 1 (ns, p > 0.05; *, p < 0.05).

Figure 4.

Responses (means ± SE) of P. sativum to nutrient distribution (uniform versus patchy), as well as the size (small versus big) and density (single versus double) of neighbouring F. glauca, in directional root placement in the rhizobox relative to its centre. The nutrient patch and single neighbour were located in section B of the rhizobox. Directional root placement was estimated per plant as the relative change in root length measured between sections A and B of the rhizobox: [(B − A) / (A + B) × 100]. Negative and positive values indicate greater root placement in section A and B, respectively, while zero values indicate even root placement in both sections. Different lowercase letters indicate statistically significant pair-wise comparisons within the uniform or patchy nutrient distribution treatments (with the Benjamini–Hochberg false discovery rate correction).