1. Introduction
Climate exerts strong control over vegetation structure and function. Its influences extend from global biome distributions to rates of ecosystem processes at individual sites (Whittaker 1975; Lloyd and Taylor 1994; Mudrick et al. 1994). One way climate affects ecosystems is through periodic events such as severe storms or droughts that damage or kill plants. Such disturbances shape species composition, biodiversity, and landscape patterns (Frelich 2002). Broadscale spatial variations in disturbance regimes may influence geographic patterns of vegetation (Parker et al. 2001; Bond and Keeley 2005).
Here, we investigate how climate, specifically moisture, influences geographic patterns of wildfire in the eastern United States. Fire is an influential disturbance agent and major natural hazard that presents complex management challenges. Many factors, such as vegetation and land use, affect fire occurrence (Pyne 1982; Alaback et al. 2003), but climate is of fundamental importance in defining the conditions that permit fire. Thus climate helps shape geographic patterns of fire (Krawchuk et al. 2009). A long-recognized (Sauer 1952) and recently articulated (Meyn et al. 2007; Parisien and Moritz 2009) generalization about climate and fire is that locations with intermediate moisture provide the best fire “habitats.” These include, for example, tropical savannas and dry temperate forests. More extreme environments burn less often: for example, arid regions with sparse fuels and wet environments with moist fuels.
Beyond this generalization, a fuller understanding of how climate shapes fire regimes requires a consideration of climatic variability (Schoennagel et al. 2004; Meyn et al. 2007). Even a climate with intermediate moisture conditions is not constantly prone to fire but has alternating wet and dry periods. A number of studies have demonstrated the importance of anomalously dry periods for fire occurrence (e.g., Swetnam 1993; Westerling et al. 2006). A better understanding is needed about how the degree of climatic variability influences the overall susceptibility of an environment to fire.
Climatic variability exerts more influence than the climatic mean on some ecological processes (Jentsch and Beierkuhnlein 2008). Consequently, changes in climatic variability may affect ecosystems more severely than shifts in the mean. In mesic grasslands, for example, a variable precipitation regime with long rain-free intervals has been shown to cause greater vegetation stress and less ecosystem productivity than one with more consistent rainfall, even with identical mean annual precipitation (Fay et al. 2003; Heisler-White et al. 2009). Likewise, we expect that a variable precipitation regime would generate more days per year that favor burning, hence a more fire-prone environment, than a climate with more consistent precipitation and short rain-free intervals. If this is true, spatial differences in precipitation regime should contribute to geographic patterns of fire.
Previous research on spatial patterns of wildfire in the central Appalachian Mountains prompted the work reported here. Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007) examined wildfire data for federally managed lands that spanned three physiographic provinces with differing climates: First, the high eastern edge of the Appalachian Plateau physiographic province, which occupies the western edge of the study area, receives abundant orographic precipitation. Second, the Ridge and Valley province to the east, in the rain shadow of the plateau, has a drier climate. Farther east, the Blue Ridge has wetter conditions, similar to the plateau; it receives moisture from the Atlantic Ocean and also impedes some of that moisture from penetrating west to the Ridge and Valley.
The fire cycle (i.e., number of years needed to burn an area equivalent to the entire landscape) varied by two orders of magnitude among the three provinces, from 10 845 years on the Appalachian Plateau to 1274 years in the Ridge and Valley and 284 years on the Blue Ridge (Lafon and Grissino-Mayer 2007). All these are quite long, much longer than in the past when fires were not suppressed (e.g., Aldrich et al. 2010). Nonetheless, fire still occurs frequently enough in the Blue Ridge to play an important role in ecosystem functioning and vegetation patterns. Fire has a minimal influence on the Appalachian Plateau, however, despite annual precipitation totals that are about the same as those on the Blue Ridge. Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007) suggested that a regime of less frequent but heavier precipitation events leads to longer dry spells and more fire in the Blue Ridge than the Appalachian Plateau, with a less variable regime of frequent, light precipitation events that keep fuels moist for much of the time.
In the current paper, we follow up on that study to determine whether, indeed, the Blue Ridge has greater precipitation variability than the other physiographic provinces. We also extend the study to the entire eastern United States, which encompasses a wider range of climates than the Appalachian Mountains. We explore the relationship of fire to both the average annual moisture condition and the variability of daily precipitation amount.
2. Methods
2.1. Precipitation regimes in the Appalachian Mountains
We obtained the 1971–2000 mean annual precipitation data (NCDC 2011a) for 42 climate stations in the central Appalachian Mountains (Figure 1; Table 1). The stations occupy the study area of Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007). We also acquired daily precipitation amounts (NCDC 2005; NCDC 2009) for the period 1971–2000 and calculated the coefficient of variability (CV) for daily precipitation (Zar 1999) at each of the 42 stations. The CV expresses variability relative to the mean (Zar 1999) and therefore enables comparisons of precipitation between stations with different mean daily precipitation amounts. It is regularly used to characterize daily precipitation variability (e.g., Douville et al. 2001; Goswami et al. 2006).
The central Appalachian Mountains. The heavy lines portray state boundaries, and the thin lines indicate physiographic boundaries (based on Fenneman 1938; Rodgers 1970; Bailey 1999). The points represent the climate stations used to analyze precipitation regimes for the central Appalachian NF and NP lands used by Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007) in their analysis of central Appalachian fire regimes.
Citation: Earth Interactions 16, 11; 10.1175/2012EI000442.1
Climate stations in the central Appalachian Mountains.
We averaged the annual precipitation and CV data for the stations within each physiographic province. Kruskal–Wallis analysis (Zar 1999) was used to test whether the climatic variables differed among the provinces. We then performed pairwise comparisons between the provinces (Dunn 1964; Zar 1999).
2.2. Fire–climate relationships across the eastern United States
We identified 34 national forests (NFs) and national parks (NPs) that span the range of moisture conditions found across the humid continental and humid subtropical zones of the eastern United States (Table 2). These federal lands represent the major forest types and physiographic regions of the temperate forest region in the eastern United States: from subtropical pine (Pinus)–dominated forests to the southern boreal forests and from dry oak (Quercus) forests to mesophytic and northern hardwoods forests. For each federal land, we obtained wildfire occurrence data archived in the National Interagency Fire Management Integrated Database (NIFMID; USDA Forest Service 1998), which contained fire occurrence data (e.g., date, cause) for 1970–2009 and reliable area-burned data for 1986–2009. These records permitted us to calculate the mean annual density of wildfires for 1970–2009 and the mean annual area burned for 1986–2009. We expressed annual fire density and area burned on the basis of a 400 000-ha area (i.e., fires per 400 000 ha yr−1; area burned per 400 000 ha yr−1). The 400 000-ha area approximates the 1 000 000-acre basis conventionally used in fire density calculations in the United States (e.g., Schroeder and Buck 1970). These calculations required data on the land area of each national forest/park; we gathered these data from the U.S. Forest Service website (http://www.fs.fed.us/) and the NP index (NPS 2009).
Eastern U.S. federal lands incorporated in the study.
Both measures of wildfire activity—fire density and area burned—are commonly used in research on fire regimes (e.g., Mitchener and Parker 2005; Lafon and Grissino-Mayer 2007). Area burned is particularly important because it reflects the extent of burning within a landscape. It is influenced by suppression efforts, however, which could obscure the climatic and ecological conditions that favor burning. Suppression does not affect fire density. Fire density therefore reflects the climatic and ecological conditions that lead to ignition (Gralewicz et al. 2011), although it does not reveal how those factors influence the spread of fire over a landscape. Using both measures offers more reliable insights about fire–climate relationships than using either one alone.
We assigned a climate station to represent the general conditions associated with each federal land. For each station, we first characterized the average annual moisture availability by calculating its annual moisture balance [precipitation P − potential evapotranspiration (PET)]. We used Thornthwaite’s (Thornthwaite 1948) method to estimate PET from the 1971–2000 monthly temperature averages (NCDC 2011a). We then conducted correlation analyses (Zar 1999) to test whether fire activity (mean annual fire density and mean annual area burned) was related to annual moisture balance.
Second, we calculated the CV of daily precipitation for 1971–2000 at each of the stations, using data from the National Climatic Data Center (NCDC 2005; NCDC 2009). We correlated fire activity with this variable to look for an influence of precipitation variability on fire activity. Finally, as the two climate variables (annual moisture balance and CV of daily precipitation) were correlated with each other (r = −0.44, P < 0.05), we calculated partial correlation coefficients (Zar 1999) to examine the relationship of fire activity with each of the two climate variables independently.
3. Results
3.1. Precipitation regimes in the Appalachian Mountains
Mean annual precipitation for the central Appalachian climate stations varied spatially according to the generally recognized pattern in the region. Significantly less precipitation fell over the Ridge and Valley than over the other two provinces (Table 3). Mean precipitation for stations in the Ridge and Valley ranged between 835 and 1150 mm yr−1 (Table 1). The stations in the Appalachian Plateau generally had wetter climates, annually receiving 1028–1515 mm of precipitation. Precipitation totals in the Blue Ridge were similar to those on the Appalachian Plateau (Table 3), with station totals averaging 987–1394 mm yr−1.
Mean values of climatic variables for the weather stations in each physiographic province of the central Appalachian Mountains. Within a row, values followed by different letters differ significantly according to Kruskal–Wallis and post hoc pairwise tests.
Daily precipitation variability showed a different pattern, wherein the Appalachian Plateau had less variability than the more flammable provinces to the east (Table 3). Although precipitation variability did not differ significantly between the moderately fire-prone Ridge and Valley and the more fire-prone Blue Ridge, every station in the Blue Ridge but one (Pedlar Dam; Table 1) had a higher CV than the stations in the Ridge and Valley. The failure to obtain a statistically significant difference in precipitation variability between the Ridge and Valley and the Blue Ridge probably reflects the small sample size (n = 6) for the Blue Ridge.
Spatial variations in fire cycle correspond with the pattern of daily precipitation variability: fire cycle declined from 10 845 years on the Appalachian Plateau to 1274 years in the Ridge and Valley and 284 years in the Blue Ridge (Lafon and Grissino-Mayer 2007), paralleling the increase in CV of precipitation from west to east. The fire cycle did not vary in the same manner as total annual precipitation, however. The two wet environments (the Appalachian Plateau and the Blue Ridge) had contrasting fire regimes—one with little fire and the other with much fire—while the driest environment (the Ridge and Valley) had an intermediate level of burning compared to the wetter provinces. In the following section, we examine how fire activity relates to annual moisture level and daily precipitation variability across the entire eastern United States.
3.2. Fire–climate relationships across the eastern United States
Fire density tended to decline from the driest to wettest environments (Figure 2a), but the pattern was weak. Both high and low fire density could be found across the entire climatic moisture gradient, and a statistically significant correlation did not emerge (Table 4, relationship A). Area burned displayed a more pronounced decline (Figure 2b), with drier areas witnessing a greater areal extent of burning than moister environments, resulting in a negative correlation between area burned and moisture balance (Table 4, relationship B).
Relationship of fire activity to (a),(b) mean annual moisture balance and (c),(d) daily precipitation variability for NFs and NPs in the eastern United States. The graphs in (a),(c) portray fire activity in terms of mean annual fire density, while those in (b),(d) depict fire activity in terms of mean annual area burned. Each point on the scatterplots represents an NF or NP.
Citation: Earth Interactions 16, 11; 10.1175/2012EI000442.1
Correlations of fire activity to climate variables for national forests and parks across the eastern United States.
Fire activity showed stronger relationships with daily precipitation variability. Both fire density and area burned increased along the precipitation variability gradient (Figures 2c,d), and both showed positive correlations with CV of daily precipitation (Table 4, relationships C and D). In general, then, environments with high precipitation variability witnessed more fires and a greater areal extent of burning than did locations with relatively little precipitation variability.
Analyzing partial correlations revealed no significant relationship between fire activity and P − PET (Table 4, relationships E and F). According to this analysis, any tendency toward greater fire activity in the driest parts of the eastern United States was mostly a consequence of the greater precipitation variability in dry sites compared to wet sites, not a result of the average moisture levels themselves. Locations with higher precipitation variability had more burning than those with less variability, as indicated by positive partial correlations between fire activity and the CV of daily precipitation (Table 4, relationships G and H).
The interregional variations in fire activity across the eastern United States generally corresponded to their climatic differences (Figure 2). The fire-prone Southeast showed high precipitation variability. Most of the points on the right half of the precipitation variability axis (Figures 2c,d) represent locations in the Southeast. The southeastern locations were also relatively dry, in terms of annual moisture balance, compared to the Northeast and parts of Appalachia (Figures 2a,b). Delta National Forest, Mississippi, deviated in flammability from the rest of the Southeast. Its fire activity, in terms of both density and area burned, was among the lowest of all sites in the entire eastern United States, despite its climatic resemblance to the rest of the Southeast (Table 2).
The Great Lakes region had the driest climatic conditions of all the regions in terms of annual moisture balance (Figure 2; Table 2). It showed only a moderate level of burning, however, which matched its moderate precipitation variability. Ottawa National Forest, Michigan, witnessed an unusually low level of burning compared to the rest of the Great Lakes sites.
The wettest climates and least variable precipitation regimes occurred in the northeastern United States, which also had little fire activity (Figure 2; Table 2). Acadia National Park, Maine, was the only northeastern site with a moderate level of fire activity, at least in terms of fire density. It had one of the wettest climates in the entire eastern United States but exhibited relatively high precipitation variability, considerably higher than the other northeastern sites.
The Appalachian national forests and parks showed a wide range of climatic conditions and fire activity (Figure 2; Table 2), despite their geographic proximity. Fire activity was quite high along the eastern edge of Appalachia: in the Blue Ridge Mountains of Shenandoah National Park, Virginia, and Pisgah National Forest, North Carolina. Precipitation variability was quite high at these two locations. On the other hand, Monongahela National Forest, West Virginia, had little fire activity. It is situated primarily on the Appalachian Plateau, where the moisture regime resembled that of the Northeast.
4. Discussion
The spatial pattern of fire activity across the central Appalachian physiographic provinces parallels that of daily precipitation variability, as suggested by Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007). Fire activity does not correspond as well to the spatial pattern of total annual precipitation. On the basis of total annual precipitation, in fact, we would expect the Ridge and Valley to show the most fire activity and the Blue Ridge to exhibit less, perhaps about the same level as the Appalachian Plateau. Precipitation variability appears to offer a better explanation than total precipitation for the spatial differences in fire regime among the physiographic provinces of the Appalachian Mountains.
Clearly, the Appalachian Plateau is not too wet to burn on the basis of total annual precipitation, given that it receives about the same amount as the Blue Ridge. Instead, the plateau likely remains too wet to burn for much of the year because frequent light precipitation events persistently rewet the fuels. Konrad (Konrad 1994) showed that light precipitation events (defined as events with <1.27 cm of rain) occur more frequently on the western side of the Appalachian region than on the eastern side, where heavy events (>5.08 cm of rain) are more common. Despite similar annual averages, then, the regime of precipitation delivery varies from east to west as moisture-bearing winds and precipitation-generating mechanisms interact with the terrain. The eastern slopes of the Blue Ridge, in particular, favor convection of moist air imported from the Atlantic Ocean (Konrad 1994; Phillips 2001), leading occasionally to heavy rainfall in an otherwise relatively dry—and flammable—environment.
The patterns of fire and precipitation regime across three physiographic provinces do not suffice to establish a general trend. The broader fire–climate pattern for the entire temperate forest region of the eastern United States, however, evokes a general pattern consistent with that suggested by the central Appalachian fire and climate data considered in this study and in Lafon and Grissino-Mayer (Lafon and Grissino-Mayer 2007). Both annual moisture balance and daily precipitation variability appear to influence fire activity, but daily variability has the dominant effect, as indicated by the correlation analyses.
Some of the federal lands in our study, in fact, display high fire activity despite relatively wet climates, apparently because of their variable precipitation regime. The DeSoto National Forest, Mississippi, occupies one of the wettest locations in the Southeast, but its precipitation variability and fire density exceed those of all other study sites. Likewise, Acadia National Park, Maine, has a wet climate, but its precipitation variability and fire density are high compared to the other sites in the Northeast. Haines et al. (Haines et al. 1983) found that in the northeastern United States the probability of fire rises with the passage of more consecutive rain-free days. A variable precipitation regime should promote fire in a humid climate because it comprises episodic heavy rainfall events separated by multiple rainless days that permit fuels to dry and burn.
The southeastern United States, especially Florida and the Gulf Coastal Plain, is notable for its flammability. Despite its humid climate, the Southeast has a long history of frequent burning by humans and lightning (Pyne 1982; Lafon 2010) and currently has more wildfires annually than any other region of the United States (Andreu and Hermansen-Báez 2008). The abundant rainfall and high temperatures in the Southeast combine to generate a fire-prone environment. First, plant productivity is high under these moisture and temperature conditions, resulting in heavy fuel loads (Pyne 1982). Second, the fuels dry quickly because the warm environment induces high potential evapotranspiration. Third, despite abundant precipitation, the Southeast generally has less rain days annually than the other regions in our study (NCDC 2011b), giving the fuels longer to dry and burn. The high annual precipitation totals are possible because of the intense rainfall events that occur in the Southeast (Frederick et al. 1977). During summer, clockwise flow around the Bermuda high advects warm, moist air masses from the Gulf of Mexico into the Southeast. These unstable air masses produce convection that leads to afternoon thunderstorms (Soule 1998). Infrequent but heavy rain from tropical cyclones can contribute significant precipitation during the late summer and fall (Knight and Davis 2009).
Nonclimatic factors obviously affect fire activity too. Differences in terrain, vegetation, and human accessibility probably account for much of the scatter shown in our graphs. The Delta National Forest, Mississippi, for example, lags the rest of the Southeast in fire activity because it occupies frequently flooded bottomland hardwood forests. Further, in large and/or mountainous national forests or parks, our “broad brush” approach using a single climate station cannot represent the range of fire climates that exist across the entire landscape. Despite such limitations, our study reveals important patterns, particularly with respect to the consequences of precipitation variability for fire activity. Fire, like plant productivity (e.g., Fay et al. 2003; Heisler-White et al. 2009) and perhaps other ecological processes, responds to differences in precipitation variability as strongly as to differences in total moisture supply.
5. Conclusions and implications
In the eastern United States, precipitation variability influences flammability more strongly than does the total annual moisture supply. While the driest parts of the eastern United States tend to have more fire than the wettest parts, this pattern appears mainly to reflect the greater precipitation variability in dry sites, rather than an actual influence of average moisture levels. Our results suggest that the regime of precipitation delivery must be considered if we are to understand fire–climate relationships, at least within the range of environmental conditions found in the eastern United States. The association between fire and precipitation variability may apply to forested areas beyond the eastern United States. Preliminary analyses that we have conducted for the western United States (C. W. Lafon and S. M. Quiring 2011, unpublished manuscript) resemble those reported here and suggest the generality of the relationships.
Changes in precipitation variability could produce strong ecological effects (Jentsch and Beierkuhnlein 2008), but they have received less attention than changes in climatic averages. Our results suggest that a shift toward less frequent but more intense rainfall events could promote greater fire activity, regardless of whether the average moisture level changes. The observation record indicates that precipitation has increased in the eastern and southeastern United States over the last century (Karl and Knight 1998). Most of the increase in the southeastern United States can be attributed to greater intensities of very heavy and extreme precipitation events (top 10% of events). The observations also show that the duration of prolonged periods of warm season dry episodes increased in the eastern United States during the last 40 years (Groisman and Knight 2008). Apparently, then, eastern U.S. climates are trending toward greater variability in precipitation delivery, with prolonged dry periods punctuated by extreme precipitation events. The trend varies within the region (Keim 1997), but for the eastern United States as a whole it may portend increasing wildfire occurrences.
The fire–climate relationships that we have considered here likely influence biogeographic distributions. Several fire-dependent species—for example, the mountain golden heather (Hudsonia montana Nutt.; cf. USFWS 2011), eastern turkeybeard [Xerophyllum asphodeloides (L.) Nutt.; cf. Flora of North America Editorial Committee 1993], and Table Mountain pine (Pinus pungens Lamb.; cf. Della-Bianca 1990)—occupy eastern parts of the Appalachian Mountains but not the less flammable western portions. Similarly, the pine (Pinus L.) woodlands and savannas that historically dominated the coastal plains of the Southeast developed under a regime of frequent burning that restricted the survival of competing hardwood trees. Komarek (Komarek 1968) suggested that the irregular occurrence of convective storms interspersed with long, hot rainless periods enabled the frequent fires that sustained the pine stands. He noted that the deciduous forests to the north and interior occupy a region with more regular precipitation, which reduces opportunities for extensive burning.
References
Alaback, P., T. T. Veblen, C. Whitlock, A. Lara, T. Kitzberger, and R. Villalba, 2003: Climatic and human influences on fire regimes in temperate forest ecosystems in North and South America. How Landscapes Change: Human Disturbance and Ecosystem Fragmentation in the Americas, G. A. Bradshaw and P. A. Marquet, Eds., Springer-Verlag, 49–87.
Aldrich, A. R., C. W. Lafon, H. D. Grissino-Mayer, G. G. DeWeese, and J. A. Hoss, 2010: Three centuries of fire in montane pine-oak stands on a temperate forest landscape. Appl. Veg. Sci., 13, 36–46.
Andreu, A., and L. A. Hermansen-Báez, 2008: Fire in the South 2: The southern wildfire risk assessment. Southern Group of State Foresters Rep., 30 pp.
Bailey, C. M., 1999: Physiographic map of Virginia. College of William and Mary, 1 pp.
Bond, W. J., and J. E. Keeley, 2005: Fire as a global ‘herbivore’: The ecology and evolution of flammable ecosystems. Trends Ecol. Evol., 20, 387–394.
Della-Bianca, L., 1990: Pinus pungens Lamb.: Table Mountain pine. Silvics of North America, R. M. Burns and R. M. Honkala, Eds., U.S. Department of Agriculture Handbook 654, 604–612.
Douville, H., F. Chauvin, and H. Broqua, 2001: Influence of soil moisture on the Asian and African monsoons. Part I: Mean monsoon and daily precipitation. J. Climate, 14, 2381–2403.
Dunn, O. J., 1964: Multiple comparisons using rank sums. Technometrics, 6, 241–252.
Fay, P. A., J. D. Carlisle, A. K. Knapp, J. M. Blair, and S. L. Collins, 2003: Productivity responses to altered rainfall patterns in a C4-dominated grassland. Oecologia, 137, 245–251.
Fenneman, N. M., 1938: Physiography of Eastern United States. McGraw-Hill, 714 pp.
Flora of North America Editorial Committee, 2002: Magnoliophyta: Liliidae: Liliales and Orchidales. Vol. 26, Flora of North America North of Mexico, Oxford University Press, 752 pp.
Frederick, R. H., V. A. Myers, and E. P. Auciello, 1977: Five- to 60-minute precipitation frequency for the eastern and central United States. U.S. Department of Commerce National Weather Service Tech. Memo. NWS HYDRO-35, 37 pp.
Frelich, L. E., 2002: Forest Dynamics and Disturbance Regimes: Studies from Temperate Evergreen-Deciduous Forests. Cambridge University Press, 266 pp.
Gralewicz, N. J., T. A. Nelson, and M. A. Wulder, 2012: Spatial and temporal patterns of wildfire ignitions in Canada from 1980 to 2006. Int. J. Wildland Fire, 21, 230–242.
Goswami, B. N., V. Venugopal, D. Sengupta, M. S. Madhusoodanan, and P. K. Xavier, 2006: Increasing trend of extreme rain events over India in a warming environment. Science, 314, 1442–1445.
Groisman, P. Ya., and R. W. Knight, 2008: Prolonged dry episodes over the conterminous United States: New tendencies emerging during the last 40 years. J. Climate, 21, 1850–1862.
Haines, D. A., W. A. Main, J. S. Frost, and A. J. Simard, 1983: Fire-danger rating and wildfire occurrence in the northeastern United States. For. Sci., 29, 679–696.
Heisler-White, J. L., J. M. Blair, E. F. Kelly, K. Harmoney, and A. K. Knapp, 2009: Contingent productivity responses to more extreme rainfall regimes across a grassland biome. Global Change Biol., 15, 2894–2904.
Jentsch, A., and C. Beierkuhnlein, 2008: Research frontiers in climate change: effects of extreme meteorological events on ecosystems. C. R. Geosci., 340, 621–628.
Karl, T. R., and R. W. Knight, 1998: Secular trends of precipitation amount, frequency, and intensity in the United States. Bull. Amer. Meteor. Soc., 79, 231–241.
Keim, B. D., 1997: Preliminary analysis of the temporal patterns of heavy rainfall across the southeastern United States. Prof. Geogr., 49, 94–104.
Knight, D. B., and R. E. Davis, 2009: Contribution of tropical cyclones to extreme rainfall events in the southeastern United States. J. Geophys. Res., 114, D23102, doi:10.1029/2009JD012511.
Komarek, E. V., Sr., 1968: Lightning and lightning fires as ecological forces. Proc. Annual Tall Timbers Fire Ecology Conf., Tallahassee, FL, Tall Timbers Research Station, 169–197.
Konrad, C. E., II, 1994: Moisture trajectories associated with heavy rainfall in the Appalachian region of the United States. Phys. Geogr., 15, 227–248.
Krawchuk, M. A., M. A. Moritz, M.-A. Parisien, J. Van Dorn, and K. Hayhoe, 2009: Global pyrogeography: The current and future distribution of wildfire. PLoS One, 4, e5102, doi:10.1371/journal.pone.0005102.
Lafon, C. W., 2010: Fire in the American South: Vegetation impacts, history, and climatic relations. Geogr. Compass, 4, 919–944.
Lafon, C. W., and H. D. Grissino-Mayer, 2007: Spatial patterns of fire occurrence in the central Appalachian Mountains and implications for wildland fire management. Phys. Geogr., 28, 1–20.
Lloyd, J., and J. A. Taylor, 1994: On the temperature dependence of soil respiration. Funct. Ecol., 8, 315–323.
Meyn, A., P. S. White, C. Buhk, and A. Jentsch, 2007: Environmental drivers of large, infrequent wildfires: The emerging conceptual model. Prog. Phys. Geogr., 31, 287–312.
Mitchener, L. J., and A. J. Parker, 2005: Climate, lightning, and wildfire in the national forests of the southeastern United States, 1989–1998. Phys. Geogr., 26, 147–162.
Mudrick, D. A., M. Hoosein, R. R. Hicks Jr., and E. C. Townsend, 1994: Decomposition of leaf litter in an Appalachian forest: Effects of leaf species, aspect, slope position and time. For. Ecol. Manage., 68, 231–250.
NCDC, 2005: Data documentation for data set 3210 (DSI-3210): Summary of the day—First order. National Oceanic and Atmospheric Administration, 19 pp.
NCDC, 2009: Data documentation for data set 3200 (DSI-3200): Surface land daily cooperative summary of the day. National Oceanic and Atmospheric Administration, 19 pp.
NCDC, cited 2011a: U.S. climate normals. National Oceanic and Atmospheric Administration. [Available online at http://hurricane.ncdc.noaa.gov/cgi-bin/climatenormals/climatenormals.pl?directive=prod_select2&prodtype=CLIM81&subrnum=.]
NCDC, cited 2011b: Climate maps of the United States. National Oceanic and Atmospheric Administration. [Available online at http://hurricane.ncdc.noaa.gov/cgi-bin/climaps/climaps.pl?directive=welcome&subrnum=.]
NPS, 2009: The national parks: Index 2009–2011. U.S. Department of the Interior National Park Service, 128 pp.
Parisien, M., and M. A. Moritz, 2009: Environmental controls on the distribution of wildfire at multiple spatial scales. Ecol. Monogr., 79, 127–154.
Parker, A. J., K. C. Parker, and D. H. McCay, 2001: Disturbance-mediated variation in stand structure between varieties of Pinus clausa (sand pine). Ann. Assoc. Amer. Geogr., 91, 28–47.
Phillips, S. E., 2001: Climatological lightning characteristics of the southern Rocky and Appalachian Mountain chains, a comparison of two distinct mountain effects. M.S. thesis, Texas A&M University Department of Atmospheric Sciences, 142 pp.
Pyne, S. J., 1982: Fire in America: A Cultural History of Wildland and Rural Fire. Princeton University Press, 654 pp.
Rodgers, J., 1970: The Tectonics of the Appalachians. Wiley Interscience, 271 pp.
Sauer, C. O., 1952: Agricultural Origins and Dispersals. Bowman Memorial Lectures, American Geographical Society, 110 pp.
Schoennagel, T., T. T. Veblen, and W. H. Romme, 2004: The interaction of fire, fuels, and climate across Rocky Mountain forests. Bioscience, 54, 661–676.
Schroeder, M. J., and C. C. Buck, 1970: Fire weather: A guide for application of meteorological information to forest fire control operations. U.S. Department of Agriculture Forest Service Agriculture Handbook 360, 229 pp.
Soule, P. T., 1998: Some spatial aspects of southeastern United States climatology. J. Geog., 97, 142–150.
Swetnam, T. W., 1993: Fire history and climate change in giant sequoia groves. Science, 262, 885–889.
Thornthwaite, C. W., 1948: An approach toward a rational classification of climate. Geogr. Rev., 38, 55–94.
USDA Forest Service, 1998: National Interagency Fire Management Integrated Database (NIFMID): Technical guide. U.S. Department of Agriculture Forest Service Fire and Aviation Management Rep., 104 pp.
USFWS, 2011: Mountain golden heather: Hudsonia montana. U.S. Fish and Wildlife Service Rep. 2 pp.
Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam, 2006: Warming and earlier spring increase western U.S. forest wildfire activity. Science, 313, 940–943.
Whittaker, R. H., 1975: Communities and Ecosystems. 2nd ed. Macmillan, 387 pp.
Zar, J. H., 1999: Biostatistical Analysis. 4th ed. Pearson, 929 pp.
