Fire Ecology Volume 12, Issue 2, 2016 Stambaugh et al.: Scale Dependence of Fire Regimes doi: 10.4996/fireecology.1202065 Page 65 RESEARCH ARTICLE SCALE DEPENDENCE OF OAK WOODLAND HISTORICAL FIRE INTERVALS: CONTRASTING THE BARRENS OF TENNESSEE AND CROSS TIMBERS OF OKLAHOMA, USA Michael C. Stambaugh1*, Richard P. Guyette1, Joseph M. Marschall1, and Daniel C. Dey2 1 Department of Forestry, University of Missouri, 203 ABNR Building, Columbia, Missouri 65211, USA 2 USDA Forest Service, Northern Research Station, University of Missouri, 202 ABNR Building, Columbia, Missouri 65211, USA *Corresponding author: Tel.: +1-573-882-8841; e-mail: [email protected] ABSTRACT RESUMEN Characterization of scale dependence La caracterización de la dependencia de la es- of fire intervals could inform interpre- cala de los intervalos de fuego podría informar tations of fire history and improve fire sobre interpretaciones de la historia del fuego y prescriptions that aim to mimic histor- mejorar prescripciones que apunten a imitar ical fire regime conditions. We quan- condiciones históricas del régimen de incen- tified the temporal variability in fire dios. Nosotros cuantificamos la variabilidad regimes and described the spatial de- temporal en los regímenes de fuego y describi- pendence of fire intervals through the mos la dependencia espacial de intervalos de analysis of multi-century fire scar re- fuego a través del análisis de archivos de cica- cords (8 study sites, 332 trees, 843 fire trices de fuego por muchas centurias (8 sitios scars) derived from two historically de estudio, 332 árboles, 843 cicatrices de fue- post oak (Quercus stellata Wangenh.) go) derivados de dos paisajes históricos de ar- woodland landscapes. Despite large bustales de roble de los postes (Quercus stella- differences in fire environment condi- ta Wangenh). A pesar de las grandes diferen- tions, study sites (~1 km2) burned fre- cias en las condiciones ambientales de fuego, quently (mean fire interval [MFI] ≤10 los sitios de estudio (~1 km2) se quemaron fre- yr) before Euro-American settlement cuentemente (intervalo promedio de fuego (pre-EAS), with sites in Tennessee [MFI] ≤ 10 años) antes del asentamiento eu- showing higher overall fire frequency ro-americano (pre-EAS), mostrando una fre- than sites in Oklahoma, USA. Pre- cuencia general de fuegos en los sitios de Ten- EAS MFIs decreased exponentially nessee más alta que en los de Oklahoma, with increasing spatial extent from in- EEUU. Los MFIs en el pre-EAS decreció en dividual trees (~1 m2) to landscapes forma exponencial con el grado de aumento es- (~100 km2). The relationship between pacial desde árboles individuales (~1 m2) hasta MFI and spatial extent may help to paisajes (~100 km2). La relación entre en MFI explain how historical observations of y la extensión espacial puede ayudar a explicar annual burning could be recorded in como observaciones históricas de quemas woodlands, when experimental stud- anuales podrían ser recopiladas en arbustales, ies suggest that this is too frequent for cuando estudios experimentales sugieren que Fire Ecology Volume 12, Issue 2, 2016 Stambaugh et al.: Scale Dependence of Fire Regimes doi: 10.4996/fireecology.1202065 Page 66 tree recruitment. Further investiga- esto es muy frecuente para el reclutamiento de tions of spatial dependence of fire in- árboles. Otras investigaciones de la dependen- tervals would improve our ability to cia espacial de los intervalos de fuego podrían relate historical and experimental fire mejorar nuestra habilidad para relacionar datos data to present day fire prescriptions, históricos y experimentales de fuego a prescrip- and vice versa. ciones actuales de fuego y vice versa. Keywords: management, oak woodland, pyrodiversity, restoration, spatial extent, succession Citation: Stambaugh, M.C., R.P. Guyette, J.M. Marschall, and D.C. Dey. 2016. Scale depen- dence of oak woodland historical fire intervals: contrasting The Barrens of Tennessee and Cross Timbers of Oklahoma, USA. Fire Ecology 12(2): 65–84. doi: 10.4996/fireecology.1202065 INTRODUCTION form us about their relative importance as re- lated to oak woodland development. Nineteenth- to twentieth-century trends of Oak woodlands are highly variable forest oak (Quercus spp. L.) woodland communities communities with open canopies ranging from in the eastern US overwhelmingly show transi- 30 % to 100 % closure; sparse midstories; and tion to more closed-canopy conditions and a dense ground flora rich in forbs, grasses, and fire-intolerant tree species (Dyer 2001, Nowac- sedges (Nelson 2005). Historically, wood- ki and Abrams 2008, Hanberry et al. 2014). lands existed in ecoregions throughout the Prior to Euro-American settlement (pre-EAS) eastern deciduous forest (Braun 1950, Bailey and logging effects, fire was considered the 1997), including being embedded components primary disturbance that maintained oak wood- of glades, barrens, and oak-pine ecosystems, land communities. Interest in woodland resto- accompanying shortleaf pine (Pinus echinata ration and management is increasing, not only Mill.), longleaf pine (P. palustris Mill.), red because of historical prevalence and modern pine (P. resinosa Aiton), and pitch pine (P. rarity, but also due to the multiple ecological rigida Mill.). Mature canopy heights may benefits resulting from fire restoration com- range from 6 m to 27 m depending on site con- pared to fire suppression. Recently, research ditions. Nelson (2005) identified 18 different has focused on how fire treatments may be in- oak woodland communities in Missouri, USA, corporated into silvicultural systems (Ryan et and vascular plant species richness can exceed al. 2013, Brose 2014, Dey and Kabrick 2015). 200 species ha-1. Diverse species composi- Detailed and spatially explicit information is tions and open canopy structures of woodlands needed about the fire ecology of woodlands, are attributed to repeated and relatively fre- including how species and ecosystem function quent fires (Olmstead 1857, Swallow 1859), respond to specific fire regime conditions, and although effects of other interacting distur- how or if tree recruitment can be sustained bances and drought also promote canopy through repeated, long-term burning. Specifi- openness (McEwan et al. 2011). Oak wood- cally, understanding the scale dependencies of lands with these characteristics occurred fire intervals would improve the ability to throughout North America pre-EAS despite crosswalk between sources that characterize being relatively rare today (Hanberry et al. woodlands such as historical data, experimen- 2014). During the twentieth century to the tal data, and present day fire monitoring. Addi- present, rates of forest transitions from open- tionally, quantifying the variability in fire inter- to closed-canopy conditions have varied by re- vals across spatial and temporal scales may in- gion and species assemblages (Guyette et al. Fire Ecology Volume 12, Issue 2, 2016 Stambaugh et al.: Scale Dependence of Fire Regimes doi: 10.4996/fireecology.1202065 Page 67 2003, DeSantis et al. 2011, Cocking et al. forest management objectives (i.e., fire pre- 2012, Stambaugh et al. 2014a). In long-un- vention, wood fiber production, and human burned areas with minimal ground flora diver- health promotion; Weldon 1996). sity, the seedbank often evidences relict wood- Historical fire regimes of eastern US oak land conditions through its diversity and the woodlands vary at spatial scales from land- site requirements of seedbank species scapes to regions (Guyette et al. 2003, 2006; (Hutchinson et al. 2005, Waldrop et al. 2008, Stambaugh et al. 2014b). Through time, his- Kinkead et al. 2013). torical fire intervals within eastern oak wood- For these reasons and others, fire is in- lands can vary by an order of magnitude with- creasingly considered in silvicultural systems in the spatial extent of study sites (e.g., 1 km2). and ecological restoration (Albrecht and Mc- When managing for woodland conditions with Carthy 2006, Dey and Schweitzer 2014, fire, it is not clear whether maintenance of Kabrick et al. 2014). Benefits of fire treat- woodland structures should focus on temporal ments and woodland conditions are diverse or spatial variability in fire (or some combina- and arguably critical to sustaining the oak eco- tion). Understanding the historical distribu- system, including enhancing oak regeneration tion of woodlands, how fire regimes varied (Arthur et al. 2012, Brose et al. 2013); increas- across landscapes, and how vegetation re- ing understory plant species cover and rich- sponded to fire regime departures would bene- ness (Hutchinson et al. 2005, Ratajczak et al. fit understanding their successional pathways 2012, McCord et al. 2014); and improving di- and designing management systems. Fire dis- versity of native insects (Wood et al. 2011), turbance properties are often simplified by dis- birds (Reidy et al. 2014), and mammals regarding variability in metrics and how they (McShea et al. 2007, Starbuck et al. 2015). At change with scale (Falk et al. 2007). To char- larger scales, other benefits of woodlands may acterize the temporal variability in fire regimes be realized, such as improved wildlife diversi- and to describe the spatial dependence of fire ty through increasing early-successional habi- intervals, we analyzed eight multi-century fire tat (Thompson and DeGraff 2001), decreased scar records from two historically post oak hazardous fuel loads, and increased climate (Quercus stellata Wangenh.)-dominated