Morphological and physiological responses of two savanna tree species to grass competiton

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT REQUEST OF AUTHOR.] Interactions between trees and grasses in savannas have been studies quite extensively, yet the mechanism by which trees and grasses coexist is still not completely clear. The strength of tree-grass interactions can potentially vary across environmental gradients and among tree species, but the ecophysiological explanations for this remain poorly understood. In particular, it remains unclear to what extent tree-grass rooting differences affect the strength of tree-grass competition. I conducted a greenhouse experiment in which I grew 20 seedlings of each of two common African savanna tree species (Acacia nigrescens and Colophospermum mopane) with and without grasses in large containers. To assess how water availability in upper vs. lower soil layers affects tree-grass interactions, I crossed the grass competition treatment with a watering treatment that provided irrigation at two different depths (surface irrigation vs. irrigation at 20 cm depth). I tied back the grasses to minimize effects of light limitation in containers with vs. without grass. I monitored tree seedling growth, soil moisture availability, photosynthesis rates and stomatal conductance over the course of 14 weeks. I simulated a watering regime of 650 mm y⁻¹ . Grass presence exerted clear effects on both tree species. I found significant negative effects on total seedling biomass, and leaf stomatal conductance and photosynthetic rates. In all cases I found significant species ₉ grass competition effects, suggesting that the extent of the grass effect varied by tree species. A. nigrescens, which was faster-growing and had higher photosynthetic rates than C. mopane, experienced proportionately greater declines in growth, photosynthesis and transpiration than C. mopane. The irrigation treatment had no impact on the tree response, either alone or in the presence of grass competition. I had assumed a priori that trees would be deeper-rooted than grasses, and benefit relative to grasses from deep irrigation, but an analysis of root profiles found that trees and and grasses did not differ in terms of median rooting depth. This result is somewhat consistent with field studies that find relatively little rooting separation in savanna trees and grasses in relatively mesic, sandy environments. I also found that the presence of grass in the containers had little to no effect of soil moisture availability, suggesting that the strong response of the tree seedlings to grass competition may not have been mediated by water limitation. Although I found evidence of a small effect of light limitation effects as a result of tree shading by grasses, the strong reduction in photosytnthesis rates in the presence of grasses is most likely the result of N limitation. Further research on the relative importance of water vs. N limitation in the tree-grass interaction would help resolve this question. My results are the first to show strong leaf-level physiological impacts of African savanna grasses on tree seedlings, and have important implications for our understanding of treegrass interactions in the savanna biome.