COS 104-5 - Fuel load manipulations influence prescribed fire intensity and severity in an experimentally restored Kentucky bluegrass savanna-woodland

Friday, August 16, 2019: 9:20 AM
L006, Kentucky International Convention Center
James Shaffer1, Scott Gleeson1, John M. Lhotka2 and John J. Cox2, (1)Biology, University of Kentucky, Lexington, KY, (2)Forestry and Natural Resources, University of Kentucky, Lexington, KY
Background/Question/Methods

Periodic ecological disturbances, such as fire, are necessary to maintain savanna systems. In Central Kentucky, the unique and imperiled Kentucky Bluegrass Savanna-Woodland (KBSW) is restricted in distribution and of conservation concern. Due to land use changes, few reference systems of KBSW remain, but historic descriptions suggest an open canopy structure. Climatic and edaphic conditions do not limit the growth of closed canopy forests in the region. Species traits of native hardwood trees show adaptations to periodic disturbances (e.g. fire, drought, and/or mammalian herbivory). To restore Midwestern oak savannas, native grass and tree species are planted in fields formerly used for agriculture, and fire is used to manage openness.

To understand the role of fire in the KBSW, an experiment with a native grass-forb mix and fourteen species of native hardwood tree seedlings was planted in March 2011. Experimental treatments include vegetation removal with mowing and herbicide. In March 2016, a dormant season prescribed fire was applied to one-third of the experiment. Estimates of fuel load were measured by weighing pre- and post-burn dry vegetation biomass in mowed and un-mowed plots. Fire temperatures were recorded, and fire severity was determined by measuring fire scarring of the planted tree seedlings post-burn.

Results/Conclusions

Fuel load of pre- to post-burn biomass was reduced by greater than half in both the mowed and un-mowed treatments. Fire intensity was greater in vegetated (i.e. un-mowed) plots where higher maximum temperatures were measured. However, residence times of maximum temperatures were longer in mowed plots due to smoldering of fuel. The average return time to ambient temperatures was similar in both mowed and un-mowed treatments. Fire severity was greater in vegetated plots as measured by taller fire scar heights, more seedlings showing 100% trunk scarring, and more individual trees being fully incinerated (e.g. a loss of biomass due to burning of the stem).

This research demonstrates that prescribed fires show a positive relation between available fuel load and fire intensity, resulting in tree seedlings experiencing a more severe fire when fuel is greater. However, native hardwood trees of the KBSW show the ability to re-sprout from stumps and therefore can persist even after severe fires. The occurrence of fire, both natural or from prescribed fires, would not eliminate these species from the overall pool. Fire may represent a heterogeneous disturbance that would have historically maintained an open habitat structure in Central Kentucky.