BACD (biodiversity and climate with drought): Manipulating Temperature, Rainfall, and Biodiversity
Experiment Id

Ecosystems are undergoing and will continue to experience multiple, simultaneous global changes, including the loss of biodiversity, warming, and increased frequency and intensity of droughts. It is largely unknown whether and how these simultaneous shifts will interactively alter ecosystem processes. To investigate this, the Biodiversity and Climate with Drought (BACD) experiment (e303) at Cedar Creek fully crosses three treatments: Temperature (ambient or warmed ~1.5° C above ambient), Rainfall (ambient or reduced by ~43%, which corresponds to a 1 in 100 year dry event), and Biodiversity (plots planted with 1, 4, or 16 species). 

Building on our earlier diversity-stability studies, we are developing and testing new theoretical predictions regarding how various components of stability (resistance, recovery, resilience, temporal stability) can be predicted from one another, how stability components vary among species in monoculture, and how they depend on plant diversity.

Experiment Design

from: J. M. Cowles, P. D. Wragg, A. J. Wright, J. S. Powers, D. Tilman, Shifting grassland plant community structure drives positive interactive effects of warming and diversity on aboveground net primary productivity. Global Change Biology 22, 741-749 (2016) 10.1111/gcb.13111

Nested within the BioDIV experiment (E120), the BACD (Biodiversity and Climate with Drought, pronounced “baked”) experiment (E303) has a split-plot design that fully crosses the BioDIV plant species richness treatment, which are applied to main plots (9 by 9 m), with warming and drought (rainout shelter) treatments, which are applied to 128 subplots (2 by 2 m, but see below for specific sizes of drought and warming treatment effects). Specifically, BACD is nested in 32 BioDIV plots, including 14 monocultures, 9 plots planted with four species, and 9 plots planted with 16 species. Each of these 32 main plots includes all four combinations of the warming and drought treatments, for a total of 128 subplots.

The BACD heating treatment is applied by 1200 W infrared heaters suspended 1.8 m above the ground. These lamps were previously used as the high level of warming in the BAC (Biodiversity and Climate) experiment (experiment 249 of the Cedar Creek Long-Term Ecological Research program), a predecessor to the BACD experiment that ran from 2009 to 2015 (Cowles et al. 2016). BACD differs from BAC in several ways. First, BAC had three levels of warming (none, low, high), rather than two (none or high). Second, BAC had no rainfall reduction treatment. Third, BACD subplots were shifted the West by 70 cm and towards the edges of the plots (either North or South) by 50 cm to avoid continuing to sample within the areas that were heavily impacted by BAC sampling. The 1200 W infrared heaters used in BACD warm the plots to a different extent during different months, at different heights aboveground, and at different soil depths. On average, throughout the growing season and from 25 cm aboveground to 30 cm belowground, the lamps warm the air and soil by about 1.5 ºC above the ambient temperature (Cowles et al. 2016). Beginning in 2016, warming has been applied each growing season from May through November.

The BACD drought treatment is applied by rainout shelters that are 2.5 meters by 2 meters in size and that remove 43% of rainfall, which corresponds to a 1 in 100 year drought at our study site. Specifically, according to monthly precipitation records from nearby Fort Snelling, which includes precipitation measurements from 1836 to 1993, mean annual precipitation is 682 mm per year, and precipitation during a 1 in 100 year drought (1st quantile) is only 389 mm per year, which corresponds to a 43% reduction in precipitation. Fixed, passive rainout shelters consist of slats of clear plastic, mounted on a PVC frame, tilted for water to drain into a gutter, with a tube that carries the water away from the plot. These shelters reduce rainfall by a percentage that corresponds to an extreme drought at the study site, without altering the precipitation regime. Beginning in 2017, shelters have been deployed each growing season from May through September.

Treatment Tables

Download treatment tables (zip file)