Data from: Heterogeneity promotes resilience in restored prairie: implications for the ‘environmental heterogeneity hypothesis’
Data files
Jan 08, 2024 version files 323.05 KB
Abstract
Enhancing resilience in formerly degraded ecosystems is an important goal of restoration ecology. However, evidence for the recovery of resilience and its underlying mechanisms requires long-term experiments and comparison to reference ecosystems. We used data from an experimental prairie restoration that featured long-term soil heterogeneity manipulations and data from comparable remnant (reference) prairie to (1) quantify the recovery of ecosystem functioning (i.e., productivity) relative to remnant prairie, (2) compare resilience of restored and remnant prairies to a natural drought, and (3) test whether soil heterogeneity enhances resilience of restored prairie. We compared sensitivity and legacy effects between prairie types (remnant and restored) and among four prairie sites that included two remnant prairie sites and prairie restored under homogeneous and heterogeneous soil conditions. We measured sensitivity and resilience as the proportional change in aboveground net primary productivity (ANPP) during and following drought (sensitivity and legacy effects, respectively) relative to average ANPP based on four pre-drought years (2014-2017). In non-drought years, total ANPP was similar between remnant and restored prairie, but remnant prairie had higher grass productivity and lower forb productivity compared to restored prairie. These ANPP patterns generally persisted during drought. Sensitivity of total ANPP to drought was similar between restored and remnant prairie, but grasses in restored prairie were more sensitive to drought. Post-drought legacy effects were more positive in restored prairie, and we attributed this to the more positive and less variable legacy response of forb ANPP in restored prairie, especially in the heterogeneous soil treatment. Our results suggest that productivity recovers in restored prairie and exhibits similar sensitivity to drought as remnant prairie. Furthermore, imparting heterogeneity promotes forb productivity and enhances prairie resilience to drought.
README: Remnant and Restored Prairie ANPP 2014-2019
https://doi.org/10.5061/dryad.fqz612k0d
Remnant and Restored Prairie ANPP 2014-2019
Heterogeneity promotes resilience in restored prairie: implications for the ‘environmental heterogeneity hypothesis’
Ashley A. Wojciechowski, John M. Blair, Scott L. Collins, and Sara G. Baer
Ecological Applications
DATA:
Data are provided in one file (Remnant and Restored Prairie ANPP 2014-2019.csv) corresponding to annual aboveground net primary productivity (ANPP) data of two remnant prairie experiments and one restored prairie experiment located at the Konza Prairie Long Term Ecological Research site in Manhattan, KS. Datafile includes grass (= GRASS in file name), forb (= FORB in file name) and total (=TOTAL in file name) ANPP from 2013 to 2019.
We used a subset of whole plots in the restoration prairie experiment representing the most homogeneous (RESHOM) and most heterogeneous (RESHET) soil treatments for this study. Each whole-plot heterogeneity treatment was assigned to a 6 m x 8 m plot in each of four blocks according to a randomized complete block design (Baer et al. 1999). The RESHET plots contained a 2 x 3 factorial combination of soil depth and nutrient availability assigned to vertical and horizontal strips according to a split block design. Soil depth (two levels: shallow or deep) was assigned to two of four alternating 2 mx 6 m strips (Fig. 1).
Prior to the initial planting, shallow soil depth was achieved by excavating the soil to a depth of 20-25 cm, burying pieces of rough-cut limestone slabs in the maximum heterogeneity treatments to mimic the uplands of the Flint Hills, and returning the topsoil to the plots. All plots were excavated to control for disturbance prior to burying the limestone slabs (Baer et al. 1999). Soil nutrient availability (three levels: reduced-N, ambient-N, and enriched-N) was assigned to one of three 2 m x 8 m strips (perpendicular to depth treatments). Reduced-N strips were created by incorporating 5.5 kg dry sawdust/m2 prior to planting in 1998 and have received 84.2 g C/m2 (~200 g sucrose/m2) three times during the growing season since 2003 to maintain microbial demand and immobilization of N (Baer et al.2003). Enriched-N treatments have received 5 g N/m2 in June of every year since 1998 by hand broadcasting ammonium-nitrate pellets(Baer et al. 2003). The deep ambient-N strips never received experimental manipulations (Fig. 1). RESHOM plots receive no soil manipulations.
The remnant prairie experiments used in this study have historically been referred to as the “irrigation transect” and the “belowground plot” experiments at KPBS, referred to here as REMIRT and REMBGP, respectively. The irrigation transect experiment has manipulated water availability by irrigation since 1991 (described in detail by Knapp et al. [2001]). We used ANPP data from six non-irrigated control plots(10 m2)for comparison with the restoration experiment (Fig. 1). REMBGP has manipulated fire frequency, mowing, and nutrient availability since 1986 (described in detail by Carson and Zeglin [2018]). We used ANPP data from four annually burned, unmowed, and unfertilized control plots (12.5 m2) for additional comparison with the restoration experiment.
TAB: DATA
DETAILS: These data pertain to Figure 1 in Wojciechowski et al. 2024 (Ecological Applications) Grass, forb, and total aboveground net primary productivity (ANPP) were determined at the same time period (2014-2019) using the same methods in remnant and restored prairie experiments. In all experiments, plant material was harvested from 0.1 m2 quadrats at peak biomass. Biomass was sorted into grasses, forbs, and litter produced in the sampling year. All biomass was dried at 60oC and weighed to estimate ANPP (Briggs and Knapp 1991). In the restoration experiment, we harvested biomass from twelve quadrats in each plot (n = 4 per heterogeneity treatment). In REMIRT, ANPP was estimated from six quadrats in each of 6 control plots. In REMBGP, ANPP was estimated from two quadrats in each of 4 control plots. Biomass in each sample was multiplied by 10 to report ANPP values as g/m2/yr.
COLUMN DESCRIPTIONS
COL. LABEL DESCRIPTION UNIT
A PRAIRIE TYPE Remnant or Restored N/A
B LOCATION REMIRT, REMBGP, RESHOM, RESHET N/A
C PLOT Plot number N/A
D SUBPLOT Subplot number or letter N/A
E YEAR Year of collection N/A
F GRASS Grass ANPP (g/m2/yr)
G FORB Forb ANPP (g/m2/yr)
H TOTAL Total ANPP (g/m2/yr)
I COMMENTS
Methods
These data pertain to Figure 1 in Wojciechowski et al. (Ecological Applications).
Grass, forb, and total aboveground net primary productivity (ANPP) were determined for the same time period (2014-2019) using the same methods in remnant and restored prairie experiments. In all experiments, plant material was harvested from 0.1 m2 quadrats at peak biomass. Biomass was sorted into grasses, forbs, and litter produced in the sampling year. All biomass was dried at 60oC and weighed to estimate ANPP (Briggs and Knapp 1991). In the restoration experiment, we harvested biomass from twelve quadrats in each plot (n = 4 per heterogeneity treatment). In REMIRT, ANPP was estimated from six quadrats in each of 6 control plots. In REMBGP, ANPP was estimated from two quadrats in each of 4 control plots. Biomass in each sample was multiplied by 10 to report ANPP values as g/m2/yr.