This study was conducted on the Jornada Experimental Range (JER) from
2007-2009. The study area is located on a level upland with flat
slopes (1-5%) that were fenced from cattle grazing since 1996. The
vegetation was dominantly desert grassland (Bouteloua eriopoda, black
grama) followed by honey mesquite (Prosopis glandulosa). Soils are
coarse-textured, well-drained, sandy loams soils (Typic Paleothids).
A randomized complete block design was employed to randomly assign one
treatment combination to each plot (2.5 × 2.5 m) within one of three
blocks. Precipitation (PPT) manipulations consisted of five levels of
PPT relative to ambient PPT: −80, −50%, control, +50, +80%, with 12
replicates per treatment. Reduction of 50% or 80% of rainfall was
achieved using different numbers of transparent acrylic V-shaped
\"shingles\" per shelter. The shingles routed incoming PPT
to outside gutters. For increased PPT treatments, a PVC-pipe
irrigation system with sprinklers was installed over the plots.
Irrigation occurred the day of the rain event if PPT was > 2mm
using rainfall collected off-site. Control plots received ambient PPT
and had neither shelters nor irrigation systems. In June 2009 a subset
of the plots were switched to a different precipitation treatment (see
“water_09_trt” variable in the data set) but the nitrogen treatments
were not altered.
Nitrogen fertilization treatments consisted of two levels: 1) liquid
ammonium nitrate fertilizer applied twice each growing season in mid-
and late July from 2006 to 2008 (treatment), and 2) application of the
same amount of water, equivalent to a 2mm rain event (controls). The
ammonium nitrate fertilizer consisted of 10 grams of ammonium nitrate
dissolved in water per meter per year. Treatments began in October
2006.
Volumetric soil water content was measured (VWC) to evaluate the
effectiveness of the water manipulation treatments. VWC measurements
were taken the 1st, 3rd and 5th days following any precipitation event
> 2 mm, and every 2 to 3 weeks during the period in between
rainfall events. Some VWC measurements were timed to coincide with
plant ecophysiological measurements. VWC was monitored using Decagon
ECH2O EC-5 and EC-20 moisture sensors (Decagon Devices, Pullman, WA)
at shallow (0-5 cm, EC-5) and deep (30-50 cm, EC-20) depths. Probes
were calibrated for soils at the study site following instructions
from the manufacturer. The is data set contains information on the
date of data collection, the Julian Day, block number, plot number,
2007-2008 precipitation treatment, 2009 precipitation treatment,
nitrogen treatment, depth of soil probe, sensor voltage, and
volumetric water content. Soil moisture monitoring was discontinued in
2009.
For further information and results, see:
Throop, H., L. G. Reichmann, O. Sala, and S. Archer. (2012), Response
of dominant grass and shrub species to water manipulation: an
ecophysical basis for shrub invasion in a Chihuahuan desert grassland.
Oecologia 169: 373-383. https://doi.org/10.1007/s00442-011-2217-4
Reichmann, L. G., O. E. Sala, and D. P. C. Peters. (2013), Water
controls on nitrogen transformations and stocks in an arid ecosystem.
Ecosphere 4(1):11. https://doi.org/10.1890/ES12-00263.1
