1. Grassland plants, especially their root systems, are dynamic and can buffer changes resulting from exposure to multiple stressors; however, the interactive stressor effects on root dynamics and associated aboveground growth are poorly understood.

2. Here, we examine the effects of changed precipitation and air temperature, and defoliation intensity on root length dynamics and aboveground biomass using the third year data from a multifactor experiment conducted across three northern temperate grasslands.

3. We found that root length was more sensitive to the changes in environmental and management conditions than root mass, demonstrating the importance of root length as an indicator of rapid root system changes. Across all sites, warming, altered precipitation, and defoliation intensity interacted to affect root length while aboveground biomass was only affected by defoliation intensity, indicating that the root system was more responsive than aboveground biomass when climatic conditions change. Overall, drought reduced root length, particularly under low defoliation intensity, as well as in combination with warming and heavy defoliation, highlighting the risk of additive effects of such environmental stresses. Across unclipped plots, aboveground biomass was positively associated with total root length, the latter of which further interacted with precipitation, to affect aboveground biomass. Compared to defoliated communities, non-defoliated plant communities exhibited a greater ability to maintain aboveground biomass under drought conditions via increases in root system efficiency (the amount of aboveground biomass produced per unit of root length invested).

4. Our results highlight the rapid change of root length in the face of interactive stressors. We postulate that the degree of stability in aboveground biomass is driven by the altered root system dynamics or species turnover. Future studies are warranted that more directly assess how root length responses under climate change impact other important plant traits in grasslands.

In May 2007, one minirhizotron tube (5 cm inner diameter, 110 cm long) was installed in the center of each plot at a 45° angle from the soil surface. Tubes were made of clear cellulose acetate butyrate plastic and reached a depth of 40 cm. Image acquisition began in May 2009 (2 years after installation) to allow fine roots to colonize the disturbed soil around the tubes. High-resolution images were acquired using a Bartz BTC 100X video microscope (Bartz Technology Corporation, Santa Barbara, California, USA) on May 11, June 2, June 22, July 15, August 10, and September 2 throughout the growing season. Images were categorized into two depth classes (0-20 and 20-40 cm) and digitally analyzed using WinRHIZO TRON software (Regent Instruments, Quebec, Canada). For each individual root observed, we noted the date of root appearance (i.e., birth) and disappearance (i.e., “death”) and measured root length and width (i.e., diameter) for each session the root was present. Roots were categorized into fourteen diameter classes (<0.05, 0.05-0.1, 0.1-0.2, 0.2-0.3, 0.3-0.4, 0.4-0.5, 0.5-0.6, 0.6-0.7, 0.7-0.8, 0.8-0.9, 0.9-1.0, 1.0-1.5, 1.5-2.0, >2.0 mm) and measured after enlarging the images by a factor of 10.