DOT Project Number:  90-00-LRTF-715

Fiscal Year:  1997

Award:  $25,000

Principal Investigator:  Troy Siefert, Iowa State University; Dr. Thomas Rosburg, Department of Biology, Drake University, thomas.rosburg@drake.edu

Summary Report:

THE EFFECTS OF FIRE AND MOWING ON THE COMMUNITY COMPOSITION AND DEVELOPMENT OF RECONSTRUCTED PRAIRIE. YEAR 1 PROGRESS REPORT: EVALUATION OF A RECONSTRUCTED PRAIRIE COMMUNITY AND INITIAL MOWING RESULTS

This report utilized 1996 and 1997 inventories to evaluate the progress of the reconstruction and to address the effect of the first mowing on the mow plots, which occurred in early summer of 1997, the second year of the reconstruction. The mow plots were mown once between June 5 and June 25 with either a rotary mower (7 plots) or disc mower (8 plots). The rotary mower is typically attached behind the tractor such that some vegetation is driven down before mowing, while the disc mower cuts vegetation to the side of the tractor. Rotary mowers also tend to chop and shred plant material more than disc mowers. The impact of driving down vegetation and differences in the structure of the litter and stubble after mowing could potentially impact prairie community composition and structure. Personnel from the Iowa Department of Transportation did the mowing with equipment typically used by IRVM managers. Plots were mown by driving through them in a direction perpendicular to the road. Mowing height was approximately 10 cm.

Net productivity and litter was measured on each site at the time of the 1997 inventory in two subsamples totaling 0.5 m2. The subsamples were located adjacent to the control plots and outside the 3x5 m treatment area so that community composition would not be affected by sampling.

Evaluation of Prairie Quality
To evaluate the progress of the planting, the absolute percentage frequency of species observed in the 60 25x25 quadrats sampled in the original belt transects established in 1996, was compared with the absolute percentage frequency of species observed in the 40 25x25 quadrats sampled in the control and mowing treatment plots in 1997 (20 in each). Both evaluation criteria (RASS and PSSP) exhibited general increases from 1996 to 1997. Increases in mean RASS for both mesic (44.8% to 47.2%) and dry-mesic (37.5% to 39.8%) were small and statistically insignificant. RASS ranged from a low of 13.6% on site 713S to a high of 72.6% on 767S (both occurred in 1996; the range of data was not as great in 1997). Increases in mean PSSP for both mesic (23.0% to 29.2%) and dry-mesic (18.9% to 25.2%) were statistically significant (p<0.05). PSSP ranged from a low of 10.0% on site 598N to a high of 40% at site 610S.

The average PSSP of both seed mixes increased about 6 percentage points, which is equivalent to about 4 additional species. Since there were no significant changes in the average RASS, the primary effect of the second growing season on the reconstruction was an increase in the richness of prairie species. The mesic sites generally exhibited higher quality than the dry-mesic sites since their average RASS and PSSP were higher.

Population Patterns
After two growing seasons, the seeded species with the greatest abundance included big bluestem, little bluestem, side-oats grama, black-eyed susan, grey-headed coneflower, and frost aster (mean absolute frequency > 10%, Table 3). Two forb species that are not native to central Iowa – common wormwood (Artemisia absinthium) and false aster (Boltonia decurrens) – were among the most abundant forbs on the plots. Common wormwood was likely included accidentally instead of prairie sage (Artemisia ludoviciana) and was relatively abundant on both mesic and dry sites. Common wormwood is not native to North America, but is naturalized in states north and west of Iowa. False aster was purposefully substituted for its congener Boltonia asteroides, which is native to Iowa. It mostly occurred on the mesic sites. False aster is a federally protected species naturally restricted to a small area of southern Illinois and eastern Missouri.

The 1996 inventory indicated that the herbicide application used to prepare the seedbed resulted in a successful burn down of smooth brome (Bromus inermis) and Kentucky bluegrass (Poa pratensis), however it appeared that these species made a comeback in 1997. Smooth brome, Kentucky bluegrass and quackgrass (Agropyron repens) increased significantly on the mesic plots to levels that may represent their pre-reconstruction abundance. Smooth brome also rebounded significantly on the dry-mesic sites, whereas a slight increase in Kentucky bluegrass was not significant. Apparently the lower soil moisture of these plots was an important additional stress on bluegrass. Tall fescue (Festuca arundinacea) remained abundant (50% frequency) and unchanged between 1996 and 1997 on the mesic plots, but a 75% decrease was recorded in the dry-mesic plots. This nearly significant decrease may have been caused by a delayed effect of a herbicide and dry soil interaction, or it may have resulted from an interaction between dry soils and increasing competition from prairie grasses.

All four main prairie grasses increased in frequency from 1996 to 1997 (big bluestem on both mesic and dry-mesic sites, little bluetem on mesic sites, and Indiangrass and side-oats grama on dry-mesic sites). Among the prairie forbs, most species remained unchanged from 1996 to 1997. Exceptions were false aster and smooth aster (Aster laevis), which both exhibited significant increases. A troublesome non-native species, birds-foot trefoil (Lotus corniculatus), exhibited a significant increase of about 25% on the dry-mesic plots, where its mean frequency was about 44%.

Biomass, Litter and Effect of Mowing
The 1997 inventories of the control and mow plots were analyzed to assess for mowing effects. Net productivity and litter were also measured to provide ancillary information to the burn and mow study. As expected, the mesic plots had significantly higher productivity than the dry-mesic plots for the control plots. The mean biomass of dry-mesic control plots (340 g/m2) was 61% of the mean biomass on mesic control plots (557 g/m2) (two sample t-test, p=0.045). However, the lack of a similar significant difference in the mow plots suggests that mowing impacted (decreased) the net productivity on the mesic sites more than the dry-mesic sites. While both the mesic and dry-mesic sites exhibited a higher percentage litter on the mow plots (32% and 42% respectively) than the control plots (24% and 31% respectively), the absolute amount of litter was only different on the dry-mesic sites (higher on mow plots). Thus mowing does effectively increase the amount of mulch present, but more so on dry sites. On mesic sites, the amount of litter was about the same on control and mow plots; rather it was the decrease in total biomass created by mowing that caused the percentage of litter to increase. Most likely the percentage litter was higher on the dry-mesic sites than on mesic sites because dry microclimates would reduce decomposition. Comparison of the effect of the types of mowing on percentage litter indicates that disc mowing only slightly increased the percentage litter (by 21%), whereas the increase in percentage litter on plots with rotary mowing was appreciably greater than control plots (51%). On the other hand, net productivity on the plots with disc mowing was more greatly reduced (65% of the control) than on the plots with rotary mowing (76% of the control). If a goal is to create litter (increase mulch), a rotary mower is more effective; if the goal is to reduce growth (lower competition), a disc mower works better.
A thorough analysis of the effects of mowing on species abundance could not be done because there was no baseline (i.e., pretreatment) data for the plots (the plots didn’t exist in 1996). Nevertheless, assuming the mow and control plots were originally fairly homogenous, mowing effects (or the lack thereof) can be ascertained by comparing the 1997 species densities on the mow and control plots. One early summer mowing during the second year did not have prominent effects on the abundance of prairie grasses or forbs. There was a suggestion that mowing tended to increase the density of big bluestem, little bluestem, side-oats grama, and frost aster. More conspicuous mowing effects were observed among non-prairie species. Smooth brome, tall fescue and horseweed had much lower densities on mow plots than control plots, while forb seedling and foxtail densities were considerably higher on mow plots than controls. This suggests that summer mowing in the early stages of the prairie reconstruction could have benefits in the reduction of non-native cool-season grasses and the enhancement of germination of seeded species.