Initial notes on site and experimental design
Some metadata has been lost for this study. Below are
the best approximations of the methods taken from publications,
particularly Hewins et al (2017) listed below.
The field experiment was conducted at the Jornada Experimental
Range (JER), near the northern terminus of the Chihuahuan Desert
biome and approximately 40 km NE of Las Cruces, New Mexico, USA.
We worked in replicate (n = 3) 2.25 ha livestock exclosures
(hereafter blocks) established in 2004 (Li et al. 2007). The
center point for these blocks was located at 32.563°,
-106.760°(lat./long.), and the three blocks were spaced
approximately 1 km apart. Within each block we established
replicate (n = 3) plots (50 9 30 m) separated by at least 25 m.
Within each plot, we randomly designated a grass, a shrub, and a
bare ground patch (hereafter microsites) for litterbag placement.
Microsites were at least 10 m apart and at least 3 m downwind of
any large shrubs ([0.5 m tall) to avoid wake effects (Okin 2008).
Grass microsites were shrub-free and dominated by B. eriopoda or
S. flexuosus. Shrub microsites consisted of a single mature P.
glandulosa plant with no or minimal sub-canopy vegetation. Bare
ground microsites consisted of exposed soil lacking vegetation.
Microsites encompassed a minimum of 3 m2, with exact size
dependent on vegetation structure. Our field sites were located on
the 'sand sheet' geomorphic surface at the JER. Surface soil
texture in each block was classified as very sandy (95% sand, 5%
silt and clay; Li et al. 2007).
Published methods
Decomposition was quantified for leaflets of P. glandulosa, a
native, deciduous N2-fixing shrub that now dominates former
grasslands in the Chihuahuan Desert and southern Great Plains of
North America. Leaflets were collected prior to abscission on
19–20 November 2009 and immediately air dried at 30 °C for 5 days.
Litterbags were filled with 2 g of leaflets. Empty mesh bags
served as a control in each microsite to assess aggregate
formation in the absence of litter. Litterbags were deployed on 19
April 2010 and retrieved after 0, 1, 6, 12, 24, and 30 months.
Litterbags were arrayed along transects oriented perpendicular to
the prevailing erosive winds (originating from the southwesterly
direction 79% of the time; Li et al. 2007). Litterbags in grass
and shrub microsites were placed below canopies, where litter
typically accumulates. Litterbags were affixed to the soil surface
with steel sod staples and spaced at least 30 cm apart to minimize
effects of neighboring bags on surface soil movement. We deployed
162 P. glandulosa litterbags (N = 3 blocks x 3 plots/block x 3
microsites/plot x 6 collection dates/microsite) and 54 empty
litterbags (N = 3 blocks x 1 plot/block x 3 microsites/plot x 6
collection dates/microsite). Litterbags were stored at -20 °C
immediately following retrieval from the field.
Litterbag contents (litter + accumulated soil) were separated
using a 1 mm mesh sieve. Litter was then manually dusted using
small brushes to remove additional soil from leaflets. The brushed
litter was frozen at -80 degrees C for 48 h, lyophilized for 48 h,
weighed, and then ground to a fine powder using a ball mill (8000D
Mixer/Mill, Spex Certiprep, Metuchen, NJ, USA). Subsamples of
litter were combusted at 550 degrees C for 6 h to determine the
inorganic matter content (% ash). Mass loss and litter C and N
content (elemental analyzer; ECS 4010, Costech Analytical
Technologies, Valencia, CA, USA) are expressed on an ash-free
basis. The % ash was also used as a conservative index of soil
accumulation that accounts only for soil adhering to litter
surfaces after sieving and brushing (see Throop and Archer 2007).
A large proportion of soil that infiltrates litterbags covers or
mixes with litter, but does not adhere to litter surface. The mass
of these bulk soils entering or exiting litterbags is responsive
to wind and water transport processes and is thus likely highly
dynamic relative to that of the soil-litter films that form on
litter surfaces. Quantifying the magnitude and dynamics of this
bulk component of the soil-litter matrix was beyond the scope of
this study.
Correction factors are used when calculating mass remaining
(estimates of transport loss, litter water content, initial ash,
etc.,), so at times, especially at the 0 month collection, the %
mass remaining may be a few % greater or less than 100%. Once
decomposition begins in earnest these become negligible.
References
Hewins, Daniel B., Robert L. Sinsabaugh, Steven R. Archer, and
Heather L. Throop. "Soil–litter mixing and microbial activity
mediate decomposition and soil aggregate formation in a sandy
shrub-invaded Chihuahuan Desert grassland." Plant ecology
218, no. 4 (2017): 459-474.
https://doi.org/10.1007/s11258-017-0703-4
Throop, Heather L., and Steven R. Archer. "Interrelationships
among shrub encroachment, land management, and litter
decomposition in a semidesert grassland." Ecological
Applications 17, no. 6 (2007): 1809-1823.
