Skip to main content
Dryad logo

Divergent patterns and spatial heterogeneity of soil nutrients in a complex and dynamic savanna landscape


Zhou, Yong; Taylor, Robert; Boutton, Thomas (2021), Divergent patterns and spatial heterogeneity of soil nutrients in a complex and dynamic savanna landscape, Dryad, Dataset,


Many grassy ecosystems around the world are experiencing woody encroachment. These woody encroachers often cause nutrient enrichment in the plant-soil environment, potentially facilitating their growth and reproduction. However, studies of encroachment effects on nutrient distributions have been confined to a few major elements (e.g., N, and P) and limited in spatial extent. We analyzed 19 elements in dominant plants and in georeferenced soils across a subtropical savanna landscape experiencing woody encroachment to quantify their spatial patterns and elucidate drivers responsible for these patterns. We found divergent patterns of spatial heterogeneity of these elements in surface soils across this complex landscape. Nutrient accumulation underneath woody canopies and redistribution by woody plants occurred in a subset of elements (i.e., N, P, S, Ca, Cu, and Sr). Though some of these elements are not necessarily growth-limiting, they do occur in higher concentrations in woody compared to herbaceous plants. Distributions of the other elements were closely related to spatial variation in soil pH, clay content, and slope rather than to woody encroachment. Our nuanced spatial sampling approach and analysis reveal significant variation in nutrient distributions in response to woody encroachment, and illustrate the role of landscape patterns in mediating ecosystem processes. These changes in the concentrations and distributions of key essential nutrients broaden our understanding of the biogeochemical consequences of woody encroachment, and provide new insights regarding the significance of long-term vegetation dynamics in dryland ecosystems.


All plant and soil samples were collected from Texas A&M AgriLife La Copita Research Area (27˚40̍ N, 98˚12̍ W) in southern Texas, USA.

Briefly, on an upland portion of this area, a 160 m × 100 m landscape subdivided into 10 m × 10 m grid cells (i.e., 160 grid cells in total) was established in 2002. In July 2014, two sampling points were randomly selected within each grid cell, yielding a total of 320 points. At each sampling point, two adjacent soil cores (2.8 cm in diameter and 5 cm in length) were collected. One soil core was used to estimate fine root (< 2 mm) biomass by washing through sieves. The other core was air-dried and then passed through a 2 mm sieve to remove coarse organic matter prior to subsequent analyses.

In addition, fine roots and fully expanded new leaves were collected from 16 dominant plant species (woody species = 5, grasses = 6, and forbs = 5) during the peak of the growing season. Four replicates were obtained for each species. In addition, leaf litter and bulk fine roots were collected from grasslands and woody patches. Leaf litter was collected within a 25 cm × 25 cm frame. After removing leaf litter, soil cores (7 cm diameter × 5 cm length) were collected to obtain bulk fine roots. All plant samples were carefully washed with deionized water and then oven-dried at 65 °C for subsequent analyses.

Soil samples were pulverized by mortar and pestle while plant samples were pulverized using a Retsch MM400 mixer mill (Retsch GmbH, Haan, Germany) with a titanium grinding jar to avoid metal contamination. All plant and soil samples were analyzed for a suite of biologically essential macronutrients (N, P, K, Ca, Mg, and S), essential micronutrients (Fe, Mn, Cu, Zn, Ni), and nonessential micronutrients (Al, Ba, Co, Cr, Li, Sr, V, and Zr).

Briefly, for each sample, 0.1 g sample, 0.5 mL nitric acid (16M), 0.5 mL hydrochloric acid (12M), and 4 mL deionized water were placed within a disposable borosilicate glass tube with a Teflon cap, and digested in a single reaction chamber microwave system (ultraWAVE, Milestone Inc., Milan, Italy) (U.S. EPA 2007). Digested solutions were diluted with deionized water to 50 mL, transferred to a screw-cap vial, and then analyzed with an Avio 500 inductively coupled plasma optical emission spectrometer (PerkinElmer, Inc., Waltham, MA, USA) at the Texas A&M Trace Element Research Laboratory.

In addition, N concentrations for soil and plant samples were analyzed using an EA 4010 Costech Elemental Analyzer (Costech Analytical Technologies Inc., Valencia, CA, USA) at the Texas A&M Stable Isotopes for Biosphere Science Laboratory. Soil pH was determined on a 1: 2 (10 g soil: 20 ml, 0.01 mol/L CaCl2) mixture using a glass electrode. Soil texture was analyzed using the hydrometer method.


National Science Foundation, Award: DEB/DDIG1600790

U.S. Department of Agriculture, Award: 1003961