Recovery of working grasslands following a megafire in the southern mixed-grass prairie

Abstract

While fire is a necessary ecological driver for grassland systems, Great Plains grasslands have undergone extensive land use change following European settlement (conversion, fragmentation, fire suppression, intensive grazing, etc.). Recent studies have documented the benefits of re-introducing fire to grasslands, but work has largely focused on small-scale, low-intensity fire, often at a pasture scale (i.e., prescribed fire). Over the last 30–40 years, the size and frequency of wildfires in the Great Plains has increased due to long-term fire suppression, woody encroachment, and climate change, resulting in megafires (wildfire >40,000 ha). While there is a wealth of information regarding effects of prescribed fire on Great Plains grasslands, knowledge of how large megafire events affect modern, fragmented working grasslands (i.e., grazed grasslands) is lacking and needed in the face of increasing megafire activity. To assess grassland response and recovery following a 2017 megafire (~254,000 ha), we compared vegetation characteristics pre- (2014–2015) and post-fire (2018–2019) in the mixed-grass prairie of Kansas, USA. We examined linkages between vegetation characteristics and a metric of burn severity (differenced normalized burn ratio [dNBR]) and evaluated megafire effects on limiting reproductive habitat for a declining grassland species, the lesser prairie-chicken ( Tympanuchus pallidicinctus ). One-year post-fire, we documented increased bare ground (+59%) and decreased visual obstruction (−39%), litter depth (−31%), and forb cover (−35%). Decreased visual obstruction and increased bare ground led to an 81% decrease of lesser prairie-chicken nest habitat in the first-year post-fire, but grassland structure, functional group cover, and available nest habitat largely recovered 2.5 years post-fire. Recovery to pre-fire conditions 2.5 years post-fire was primarily due to elevated growing season precipitation received in the years following the fire (>700 mm/year). Percent cover of grass (β = 0.38) and bare ground (β = −0.36) exhibited the strongest relationships with burn severity pre-fire, but overall grassland structure and functional group cover were not strongly influenced dNBR burn severity post-fire. While our results suggested that recovering grasslands were more homogenous due to the large size of megafire, working grasslands in the mixed-grass prairie appeared largely resilient to the effects of megafire.