期刊
PLOS ONE
卷 10, 期 12, 页码 -出版社
PUBLIC LIBRARY SCIENCE
DOI: 10.1371/journal.pone.0144300
关键词
-
资金
- Andrew W. Mellon Foundation
- Alaska EPSCoR NSF award [EPS-0701898]
- state of Alaska
- Utah Agricultural Experiment Station
- Utah State University
Root biomass distributions have long been used to infer patterns of resource uptake. These patterns are used to understand plant growth, plant coexistence and water budgets. Root biomass, however, may be a poor indicator of resource uptake because large roots typically do not absorb water, fine roots do not absorb water from dry soils and roots of different species can be difficult to differentiate. In a sub-tropical savanna, Kruger Park, South Africa, we used a hydrologic tracer experiment to describe the abundance of active grass and tree roots across the soil profile. We then used this tracer data to parameterize a water movement model (Hydrus 1D). The model accounted for water availability and estimated grass and tree water uptake by depth over a growing season. Most root biomass was found in shallow soils (0-20 cm) and tracer data revealed that, within these shallow depths, half of active grass roots were in the top 12 cm while half of active tree roots were in the top 21 cm. However, because shallow soils provided roots with less water than deep soils (20-90 cm), the water movement model indicated that grass and tree water uptake was twice as deep as would be predicted from root biomass or tracer data alone: half of grass and tree water uptake occurred in the top 23 and 43 cm, respectively. Niche partitioning was also greater when estimated from water uptake rather than tracer uptake. Contrary to long-standing assumptions, shallow grass root distributions absorbed 32% less water than slightly deeper tree root distributions when grasses and trees were assumed to have equal water demands. Quantifying water uptake revealed deeper soil water uptake, greater niche partitioning and greater benefits of deep roots than would be estimated from root biomass or tracer uptake data alone.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据