Ronald K. Salemme, Jennifer M. Fraterrigo, 2021.




Stiltgrass invaded plots had reduced fire intensity compared to uninvaded plots (contrary to other studies).

Pre-fire abundance, and post-fire persistence and resprouting of naturally established tree seedlings were lower in stiltgrass invaded plots.

Lengthening fire return intervals at grass-invaded sites may help to increase post-fire tree seedling persistence.

Non-native grass invasion negatively affects long-term resilience of temperate forests.

In this study, the authors examined how prescribed fire and the invasive, shade-tolerant grass, Microstegium vimineum (stiltgrass) affect regeneration of naturally established tree seedlings and saplings. A globally important issue, exotic grass invasions have been shown to alter fire regimes and vegetation communities, resulting in a positive-feedback loop known as the grass-fire cycle. With prescribed fire use and widespread grass invasions increasing in temperate forests, there is a pressing need regarding how non-native grass invasions affect tree persistence post-fire.

Three objectives of this study were to, 1) quantify differences in fire intensity (and its drivers) at sites invaded and uninvaded by stiltgrass, 2) determine how fire intensity and invasion affect tree regeneration with respect to tree size, and, 3) evaluate long-term effects of fire and invasion on stand development (e.g., forest resilience) using a forest growth and yield simulation model.

The study took place on the Shawnee National Forest (SNF, southern Illinois, USA), where a large area has been invaded by stiltgrass (first reported in 1986), and prescribed fire is commonly used as a forest management tool. Forty plots were established across six prescribed fire units, all which were previously burned within 2-6 years. Half of the plots were invaded by stiltgrass (≥ 70% plot surface covered by stiltgrass). Prescribed fire was applied to all plots, four in spring and two during fall.

Prior to burning, surface litter and woody fuel biomasses, and moisture levels of fuels and soils were measured. Tree seedlings (≤1 m tall), saplings (1-2 m tall), and overstory trees (DBH ≥ 12.7 cm) were tagged, and species, height, and diameter were recorded. Post-fire, woody fuels were remeasured to determine consumption, and tagged trees were relocated and their status (survived/resprouted or not) recorded. Fire intensity was measured as maximum fire temperature and fire residence time over 60° C (above which plant cell death can occur) using thermocouples at the soil surface; flame lengths were measured via scorch on flame-retardant soaked string.

Tree seedling abundance pre-fire, and persistence levels post-fire, were higher in areas not invaded by stiltgrass (Microstegium vimineum, left side of image) compared to invaded areas (right side). (Photo: Ronald K. Salemme)

Statistical models were fit to pre- and post-fire fuel loads, fuel consumption, flame height, fire temperature residence time to evaluate relationships between invasion status, fuel conditions, and fire intensity. Fixed effects in these models were soil and litter moisture, surface litter biomass, fuel loading, air temperature, wind speed, and relative humidity (RH). Burn-day weather conditions were obtained from fire weather online archives. To evaluate long-term effects on stand development, post-fire tree regeneration was simulated per decade from 2016-2066 using USDA Forest Service Forest Vegetation Simulator (FVS) and the Fire and Fuels Extension (FFE), parameterized by field measurements.

Fire intensity

Contrary to findings from other studies, fire intensity, flame length, percentage of area burned, and fire residence time were all lower in invaded versus uninvaded plots. Rather, flame length was negatively associated with soil moisture and RH, and positively associated with air temperature. Percentage of area burned was negatively related to stiltgrass biomass. Fall burns had marginally longer fire residence times above 60° C, and marginally lower percentage of area burned. Stiltgrass was still green during these fires and did not burn even with direct fire application by drip torch.


Fig. 2. a) Maximum fire temperature, b) flame length, c) residence time over 60 °C, and d) percentage of area burned in uninvaded (control) and Microstegium-invaded plots. Values are least squares means and standard errors. Data are averaged across 2015 to 2017 prescribed fires. Symbols indicate statistical significance (‘**’ for P < 0.05 and ‘*’ for P < 0.10). Chart reprinted with permission from
Forest Ecology and Management.


Tree regeneration/   persistence

Pre-fire tree seedling density was lower in invaded plots than uninvaded plots, and was negatively correlated with stiltgrass biomass. Average seedling size was greater in invaded than uninvaded plots (16.5% larger stem diameter and 21.9% greater height). Post-fire, invaded plots had lower seedling persistence than uninvaded plots, despite lower fire intensity and being comprised of larger individuals. There were no significant differences in sapling survival by invasion status, maximum fire temperature, or diameter. There was a direct negative effect of stiltgrass biomass and fire intensity on seedling persistence, and a direct positive effect of seedling diameter.

Long-term forest resilience

FVS model simulations showed stiltgrass invasion significantly reduced post-fire seedling resprouting for both low- and moderate-intensity simulated fires, resulting in a 62.7% reduction in resprout density compared to uninvaded plots.


Findings of this study suggest that stiltgrass invasion affects post-fire seedling persistence directly through its presence and indirectly through effects on micro-site fire intensity, fuel continuity, and resource competition. Modeling showed that grass invasion had large, negative effects on seedling resprouting, regardless of fire intensity effects. Though grass invasion can have varying effects on fire intensity, it consistently alters forest dynamics by reducing the resilience of tree regeneration to fire. The authors suggest that based on understanding of resprouting and tree size associations, lengthening the time between fire applications in grass invaded forests may be needed to allow juvenile trees to reach adequate size to survive burning.


The fate of individual tree seedlings in stiltgrass-invaded forest sites (pictured) and not invaded, were tracked after prescribed burns. (Photo: Ronald K. Salemme)