Invasive cogongrass (Imperata cylindrica) significantly elevates modeled surface fire spread and tree-torching potential across the southeastern United States, with behavioral shifts peaking in spring and at lower latitudes, according to a regional study published in Fire Ecology. Researchers evaluated fuel metrics and simulated fire behavior across 159 paired invaded and non-invaded plots spanning central Florida, north Florida, and southern Mississippi. Cogongrass established dense biomass that increased live herbaceous fuel loads by up to 1.96 metric tons per hectare during summer and increased dead fuel loads by a median of 1.88 metric tons per hectare across all sites. Custom BehavePlus simulations revealed that these fuel alterations accelerated predicted surface fire rate of spread by a median of 0.75 meters per minute and increased flame lengths by 0.16 meters, while raising crown fire transition probabilities. However, these effects diminished toward northern latitudes.

Invasive plants can alter fire regimes by modifying fuelbed quantity, continuity, and flammability. In the southeastern United States, cogongrass has long been recognized as an aggressive invader of pine savannas, roadsides, and commercial timberlands, producing standing fuel loads that frequently exceed native communities by more than one hundred percent. Previous studies linked these fuel accumulations to higher fireline intensity, elevated soil heating, and reduced native seedling emergence, but most work was restricted to localized sites and single seasons. Gaps remained regarding how cogongrass modifies fuel moisture, bed depth, and canopy ignition thresholds across environmental and seasonal gradients. To address these questions, researchers quantified how cogongrass presence and phenology alter fuel properties, developed customized fuel models, and modeled fire behavior across regional gradients.

To capture environmental variation, researchers sampled fuels across a southeastern latitudinal gradient stretching from 28.5 degrees North in central Florida to 31.6 degrees North in southern Mississippi, spanning longitudes from 89.8 to 81.9 degrees West. The study encompassed five public state and national forests alongside three private properties across three regions: central Florida, north Florida and southern Alabama, and southern Mississippi. Central Florida sampling included Withlacoochee State Forest and Ocala National Forest, providing 52 plots. The north Florida region encompassed 75 plots across Blackwater River State Forest, Escambia Experimental Forest, Jay Research Forest, and two private properties. Southern Mississippi sampling involved 32 plots in Bogue Chitto National Wildlife Refuge. Plots occupied forested areas dominated by upland sandhill ecosystems, alongside pine flatwoods and hardwood hammock communities.

The team established 159 plots measuring thirty by thirty meters, comprising 78 cogongrass-invaded plots and 81 non-invaded control plots. Invaded plots contained at least one cogongrass patch measuring a minimum of one by 0.5 meters and were located at least five meters from roadsides to minimize edge disturbance. Non-invaded plots were established at least forty-five meters away in areas with matching stand conditions and fire history. Fuel collections occurred between May 2024 and June 2025, timed to five-year Normalized Difference Vegetation Index trends to capture dominant phenological shifts. In addition to primary growing season sampling across all regions, researchers conducted seasonal sampling across a subset of north and central Florida plots, examining senesced fuels in January and February 2025 and transitional early green-up fuels in April 2025.

Field crews sampled fuels from three 0.25 by 0.25-meter quadrats along thirty-meter transects at each plot, measuring vegetation height, litter depth, and soil moisture with a ten-centimeter probe. Rooted herbaceous vegetation was clipped 2.54 centimeters above ground level, separated into live and dead components, and collected alongside surface litter while excluding woody debris. Samples were weighed, oven-dried at sixty degrees Celsius to constant mass, and evaluated for gravimetric moisture content. Overstory canopy metrics were documented using the point-centered quarter method to estimate tree density and trunk diameter at breast height. Shrub cover was measured along one by thirty-meter belt transects, enabling researchers to estimate herbaceous cover proportions and scale fuel bed depth based on differences in herbaceous height between invaded and non-invaded patches.

Field data parameterized custom surface and crown fire behavior simulations in BehavePlus software using the Rothermel models. The authors modified the standard SH4 shrub and litter fuel model using invasion-specific values for one-hour dead fuel load, live herbaceous load, fuel bed depth, and live and dead herbaceous moisture. Dead fuel moisture inputs were averaged across paired quartiles of dead herbaceous and litter moisture, with the dead fuel moisture of extinction set to sixty percent. Foliar moisture was held at one hundred percent, wind speed was set to the six-kilometer-per-hour annual average in Jay, Florida, with a 0.5 wind adjustment factor, and terrain was modeled as flat. Models incorporated longleaf pine savanna canopy bulk density and a 42.35 percent canopy base height ratio, yielding 1,458 summer simulations and 4,374 seasonal simulations.

Bayesian generalized linear mixed-effects models using Student-t error distributions demonstrated that cogongrass invasion substantially increased standing herbaceous fuel loads. In summer, invaded plots supported a median of 1.96 metric tons per hectare more live fuel than non-invaded plots, with a 95 percent credible interval of 1.62 to 2.30 metric tons per hectare and a probability of direction exceeding 0.99. Live fuel differences were smaller in spring at 1.26 metric tons per hectare and in winter at 0.96 metric tons per hectare. A credible interaction between invasion status and latitude (median estimate of -0.57 metric tons per hectare) indicated that non-invaded plots gained 0.35 metric tons per hectare per unit increase in latitude, whereas fuels in invaded plots decreased by 0.22 metric tons per hectare, diminishing fuel load differences at northern sites.

Dead standing fuel loads exhibited substantial increases in invaded plots across all conditions, with a median increase of 1.88 metric tons per hectare and a 95 percent credible interval of 1.53 to 2.25 metric tons per hectare. Seasonal dead fuel shifts showed a credible winter increase of 0.43 metric tons per hectare, while latitude exerted no directional effect. Conversely, cogongrass invasion had an uncertain effect on litter fuel load, with a median difference of -7.19 metric tons per hectare whose credible interval overlapped zero. Instead, season drove litter dynamics, with spring and winter yielding 1.40 and 1.16 metric tons per hectare more litter than summer. Invaded plots also featured taller vegetation across all seasons, with median height increases of 0.59 meters in summer and winter, and 0.43 meters in spring.

Fuel moisture and soil moisture patterns were driven by seasonal cycles rather than invasion status. Cogongrass invasion had minimal and uncertain effects on live, dead, and litter fuel moisture content. In contrast, live fuel moisture dropped credibly by a median of 56.15 percentage points in spring and 37.98 percentage points in winter compared to summer. Litter moisture was 10.30 percentage points wetter in winter and 8.89 percentage points drier in spring relative to summer, while spring soil moisture declined by 1.98 percentage points. Regression analyses revealed that modeled fire behavior in invaded plots was more sensitive to fuel moisture than in non-invaded plots, with live and dead fuel moisture suppressing spread rate with standardized coefficients of -0.64 and -0.70 under invasion, compared to -0.40 and -0.38 in non-invaded fuel complexes.

Simulations indicated that cogongrass invasion consistently elevated surface fire spread rates and flame lengths, driven primarily by elevated dead herbaceous and litter loads. In summer simulations, invasion increased predicted surface fire rate of spread by a median of 0.75 meters per minute and flame length by 0.16 meters, with credible intervals excluding zero. Seasonal modeling showed the greatest increases during spring green-up, where invasion raised spread rate by a median of 2.60 meters per minute and flame length by 0.80 meters over non-invaded stands. Winter increases were also substantial, adding 1.81 meters per minute in spread rate and 0.56 meters in flame length. However, surface fireline intensity showed no credible relationship with invasion status or season, as credible intervals overlapped zero across all scenarios.

Cogongrass invasion also heightened tree torching potential by increasing the crown fire transition ratio across all seasons. Invaded plots exhibited a median summer transition ratio increase of 0.03, with the largest increase occurring in spring at 0.29, followed by winter at 0.17 and summer at 0.07. Invaded plots displayed greater sensitivity to litter biomass, dead biomass, and fuel moisture deficits when initiating crown transitions. These findings hold critical significance for fire managers applying prescribed burns across southeastern pine savannas. Because regional prescribed burning predominantly occurs between late fall and early spring, cogongrass fuel accumulation and low seasonal moisture coincide with operational burn periods. By elevating spread rates, lengthening flames, and increasing torching risks, cogongrass invasion narrows safe operational burn windows and raises the danger of escaped fires.

Subject of Research: Predicting cogongrass invasion impacts on fire behavior and fuel loads across southeastern forest ecosystems

Article Title: Fueling the fire: predicting cogongrass (Imperata cylindrica) impacts on fire behavior across environmental and seasonal gradients

Article References: Ivory II, A. A., Wonkka, C. L., McGranahan, D. A., Akpejeluh, P., Kincy, S., Weidig, N. C., Roberts, C. P., Brock, K. M., Silva, C. A., & Donovan, V. M. (2026). Fueling the fire: predicting cogongrass (Imperata cylindrica) impacts on fire behavior across environmental and seasonal gradients. Fire Ecology, 22(1), Article 119. https://doi.org/10.1186/s42408-026-00574-5

Image Credits: AI Generated

DOI: 10.1186/s42408-026-00574-5

Keywords: cogongrass, fire ecology, invasive species, prescribed burning, wildfire behavior, fuel models, forest management, southeastern ecosystems, Fueling, fire, predicting, Imperata