Altered rainfall patterns increase forb abundance and richness in native tallgrass prairie

Abstract

Models predict that precipitation variability will increase with climate change. We used a 15-year precipitation manipulation experiment to determine if altering the timing and amount of growing season rainfall will impact plant community structure in annually burned, native tallgrass prairie. The altered precipitation treatment maintained the same total growing season precipitation as the ambient precipitation treatment, but received a rainfall regime of fewer, larger rain events, and longer intervals between events each growing season. Although this change in precipitation regime significantly lowered mean soil water content, overall this plant community was remarkably resistant to altered precipitation with species composition relatively stable over time. However, we found significantly higher forb cover and richness and slightly lower grass cover on average with altered precipitation, but the forb responses were manifest only after a ten-year lag period. Thus, although community structure in this grassland is relatively resistant to this type of altered precipitation regime, forb abundance in native tallgrass prairie may increase in a future characterized by increased growing season precipitation variability.

Figure 1. Growing season precipitation parameters between ambient and altered treatments (fewer, larger rain events) from 1997 (pretreatment year) through 2012.

(a) total number of rain events excluding events less than 5 mm. (b) mean number of days between rain events or the mean length of dry periods in days. (c) mean size of rain events (mm). Differences between ambient and altered treatments were significant (P = < 0.0001) for all parameters. Error bars denote one standard error.

Figure 2

(a) Forb, grass, and total cover by treatment from 1997–2012. Total cover was higher on average under the altered (fewer, larger rain events) treatment, primarily due to an increase in forb cover rather than grass cover. However, this effect did not emerge until after ten years of altered precipitation. (b) Forb, grass, and total richness by treatment from 1997–2012. Total and grass richness were consistently higher on average in the altered treatment over the duration of the experiment. Higher altered forb richness occurred after a lag of ten years. Error bars denote one standard error.

Figure 3. Non-metric multidimensional scaling (NMDS) of plant community composition for all plots (a), between 1997 and 2012 and (b), 2007 only, the year altered forb cover increased significantly from ambient.

Each point is the temporal average of the NMDS coordinates for a given RaMP (plot) over the 15-year period. Error bars are one standard error. Ellipses are 95% confidence clouds based on standard error. Treatment is indicated by color. Ambient treatments are plots in which precipitation was added immediately after a natural rain event and altered treatments are plots with fewer, larger rain events. Annual treatment differences were only significant in 2007 and 2009 based on the PERMANOVA.