The Chronic Addition of Nitrogen Gradient Experiment (ChANGE): the effect of a gradient of nitrogen addition on productivity in a semi-arid shortgrass steppe and a mesic tallgrass prairie for the first five years of fertilization (2014-2018)
 
Study Sites:
This experiment is conducted in two grassland ecosystems: a semi-arid shortgrass steppe and a mesic tallgrass prairie.
The semi-arid shortgrass steppe site is located within the Central Plains Experimental Range in northeastern Colorado, USA (40° 50’ N, 104° 45’ W). Elevation at the study site ranges from 1,600 to 1,700 m, and mean annual precipitation is 366 mm (Griffin-Nolan et al. 2018). The study site is located in a pasture that has been ungrazed by cattle since 2000, with minimal topographic variation and relatively homogeneous vegetation (Griffin-Nolan et al. 2018). Perennial and annual forb species make up the bulk of diversity at the site.
The mesic tallgrass prairie site is located within the Konza Prairie Biological Station in the Flint Hills of northeastern Kansas, USA (39° 05’ N, 96° 35’ W)(Watershed R1B). This site has never been plowed. Elevation at Konza ranges from 320 m to 444 m and the mean annual precipitation is 835 mm (Collins et al. 2021). The study site is located in an ungrazed, upland site within a watershed that has been annually burned since 2001 (Collins et al. 2021).
 
Experimental Design:
This experiment was established in 2013. It is a randomized complete block design, in which 8 different levels of nitrogen (0, 2.5, 5, 7.5, 10, 15, 20, and 30 g m-2) were randomly assigned to one of eight 5 x 5 m plots within each of six blocks (8 treatments * 6 blocks = 48 total plots per site). Blocks were arrayed on the landscape so that there was a minimum 10 m distance between each block. Plots within each block were arrayed in two rows of four plots, with each plot separated by a 1 m buffer. Nitrogen in the form of time-release urea was evenly applied by hand to each 5 x 5 m plot during late May or early June.
 
Plant Diversity and Community Composition:
At each study site, plant community composition was surveyed within a 1 x 1 m permanent sampling area (located in a corner of the designated 2.5 x 2.5 m plot, with a 0.5 buffer from the edge of the plot) at the beginning (late May or early June) and end (mid- to late-August) of the growing season in each year of the study. Cover of each species was visually estimated as percentage aerial cover to the nearest 1% for species under 50% abundance and to the nearest 5% for species above 50% abundance for each species rooted within the sampling area. When early and late measures of cover were estimated for a species, the maximum cover estimate was used. Cover estimates of species within each plot were used to calculate species richness and species evenness (Evar; Smith and Wilson 1996) using the community_structure() function within R’s (version 4.4.1; R Core Team 2024) Codyn package, (version 2.0.5; Hallett et al. 2016, 2018).

Aboveground Net Primary Productivity Measurements:
ANPP was estimated at peak biomass (late August/early September) each year by clipping all vegetation at ground level within two 20 x 50 cm quadrats, located within the same subplot as the plant composition measurements. Quadrats were moved yearly to prevent resampling over time. Biomass was dried for 48 hours at 60 ºC, and weighed to the nearest 0.01 g. Quadrat values that fall outside the 99th percentile for weight were excluded as outliers. Mean biomass (excluding previous year’s dead) across the two quadrats in each plot was multiplied by 10 to estimate ANPP (g m-2). Only forbaceous and grass biomass is used to calculate ANPP, since woody biomass is rare, and not representative of the site. The effect of nitrogen addition on productivity is calculated as the absolute differences in ANPP between control plots and fertilized plots within a block.

Quality assurance:
All outlier values have been assessed for accuracy.

Citations:
- Hallett L, Avolio M, Carroll I, et al (2018) codyn: Community Dynamics Metrics R package (version 2)
- Hallett LM, Jones SK, MacDonald AAM, et al (2016) Codyn: An R package of community dynamics metrics. Methods Ecol Evol 7:1146–1151 (https://doi.org/10.1111/2041-210X.12569).
- R Core Team (2024). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. (https://www.R-project.org/).
- Smith B, Wilson JB (1996) A Consumer’s Guide to Evenness Indices. Oikos 76:70 (https://doi.org/10.2307/3545749).