Description and Location: In February of 1984 the north bank of an abandoned silage pit north of Headquarters (grid C-15) was selected for the root window site. (The old excavation required only a minor amount of digging to put in the windows.) Four windows, approximately 50 cm wide by 50 cm deep, were installed using the following procedure: First, the walls of the old silage pit were cut into so that a clean, 50 cm x 50 cm vertical face of soil was exposed. A triangle was used to cut the wall of soil so that the exposed surface was indented by a 10% angle from top to bottom (Figure 1). Plexiglas was carefully installed so that about 0.5 cm of space remained open between the Plexiglas and the soil. Soil that had been removed from this surface, dried and sieved was then used to back-fill this open space. Six inches (ca 15 cm) of Styrofoam insulation was placed against the windows when not in use, and the Styrofoam was covered with burlap.

In spring of 1984 this site was burned, and two of the four windows were destroyed by fire. The remaining two windows were monitored for the next two years as a non-LTER data set, (Hayes, D.C. and T.R. Seastedt. Root dynamics of tallgrass prairie in wet and dry years. Canadian Journal of Botany (1987)).

In spring of 1986 six additional windows were installed, and the site was again burned. Litter was returned to the soil surface about four of the eight windows, and data was collected on root lengths during the 1986-growing season from May 1 to October 1. In November 1989 the two remaining original windows were replaced due to insect damage to wooden frames.

Variable Measured: Data available from this procedure include the actual root lengths and the estimates of new lengths and decomposed lengths for each quadrat. Data from this study are available through 1989, at which time study was terminated. The original data (the mylar sheets) is stored for each window for each year. The pre-LTER sheets have been archived, but the pre-LTER encoded data is not available.

Methods: Lengths are obtained by placing a fitted mylar sheet over the window and tracing all roots. Each sheet is divided into a 50 cm wide by 40 cm deep grid, and lengths are recorded for each 10 cm x 10 cm quadrat. Production is calculated during each two-week interval for each quadrate by measuring the lengths of roots present at time t+1 that are not present at time t. Likewise, decomposition is estimated by the roots presence at time t that are not present at time t+1. Quality control is obtained by having the investigator trace the window over two consecutive days. Errors created by failing to draw all roots at time t appear as a false production estimate. False disappearance (decomposition) results by failing to draw all roots at time t + 1. It is highly unlikely that errors will result from drawing roots that are not there, thus, subtracting the estimate of false production from the bi-weekly production estimate and the false decomposition from the disappearance data is suggested to produce realistic data.

Root tracing is best conducted during the cool of the day, before the sun is high enough to cause flare on the Plexiglas. Insets and poison have distracted from the enjoyment of this exercise.

For additional metadata information see: http://lter.konza.ksu.edu/sites/default/files/DC.pdf

For additional methods information see: http://lter.konza.ksu.edu/sites/default/files/MM.pdf