Climate has a primary influence on the occurrence and rate of combustion in ecosystems with carbo... more Climate has a primary influence on the occurrence and rate of combustion in ecosystems with carbon-based fuels such as forests and grasslands. Society will be confronted with the effects of climate change on fire in future forests. There are, however, few quantitative appraisals of how climate will affect wildland fire in the United States. We demonstrated a method for estimating changes in fire probability based on future climate simulations of temperature and precipitation. The probability of a fire occurring in a particular climate was extracted from the Physical Chemistry Fire Frequency Model (PC2FM) and represented the rate of change in fire due to climate. Climate output data from two global climate models (GCMs) were applied to the PC2FM to estimate changes in fire probability. We calculated change in fire frequency and probabilities from the difference between current and future climates and mapped climate-forced percentage change in fire probability under each GCM for the nation at a 1.2 km 2 scale. Future fire probability estimates increased in cooler northern and high elevation regions but decreased slightly in some hotter and drier regions of the southwestern United States. Our approach's greatest strength may be reliance on only climate data and the simple principles of physical chemistry; many other nonclimatic factors that affect fire are often difficult to predict in the distant future.
A predictive model of fire frequency in North America using climate and human population density ... more A predictive model of fire frequency in North America using climate and human population density shows promise in estimating the importance of fire at broad temporal and spatial scales. Over 5 thousand fire scars from 120 sites and nine forest ecosystems were used to empirically derive and test a fire interval regression model. Three predictor variables were selected: a proxy
In this paper we describe a fire risk assessment of the Ozark Highlands. Fire risk is rated using... more In this paper we describe a fire risk assessment of the Ozark Highlands. Fire risk is rated using information on ignition potential and fuel hazard. Fuel loading, a component of the fire hazard module, is weakly predicted (r 2 = 0.19) by site-and landscape-level attributes. Fuel loading does not significantly differ between Ozark ecological landtypes. Drought and exposure are related to fuel moisture content. Drought is particularly important to the Ozark fire regime and fire risk as it is related to both ignitions and fuels.
Shortleaf pine community restoration requires an answer to the question, "What fi re frequency is... more Shortleaf pine community restoration requires an answer to the question, "What fi re frequency is appropriate for shortleaf pine regeneration and survival?" The answer to this question is one of the most critical to successful restoration through fi re management. We used three sources of information from Missouri to determine appropriate burning frequencies: a 400-yr historic shortleaf pine growth and fi re-scar database, fi re effects data from prescribed burning sites, and a vegetation dynamics prediction model that is widely used for characterizing fi re regimes. The historic shortleaf pine and fi re scar database provides actual past scenarios of regeneration dates, growth, survival, and associated fi re events. Shortleaf pine regeneration established most commonly during the 4 years following fi re events and generally decreased in abundance with years since fi re. Surviving seedlings were those that were not fi re scarred the year following establishment, and the mean number of years to a subsequent fi re was about 7 years. Fire effects data from prescribed burn sites revealed that both hardwood and pine regeneration showed substantially increased mortality after four consecutive dormant season burns, but oak and hickory species were more likely to survive frequent fi re. Mortality of advance regeneration was generally low in all hardwood species after one burn, while shortleaf pine seedlings had high mortality rates. The model showed 8 to 15 yr intervals are likely best for balancing both continual regeneration and recruitment. Model prediction runs for 500 years showed a signifi cantly decreased pine component in the absence of burning. Conversely, long-term frequent burning (1-to 3-yr intervals) resulted in abundant regeneration, but poor survival and ultimately decreased abundance in mid-and late-successional forests. In summary, all three sources support the effi cacy of frequent burning (1 to 4 yrs) in promoting pine regeneration, but survival and continued recruitment require longer fi re intervals (8 to 15 yrs). Fire management prescriptions that incorporate both frequent burning and longer intervals will likely provide for the most long-term regeneration and recruitment success.
We present the fire history of a 1-km 2 area that is a mosaic of oak forest, savanna, and fen on ... more We present the fire history of a 1-km 2 area that is a mosaic of oak forest, savanna, and fen on the White Ranch State Forest, Howell County, Missouri. We dated 135 fire scars on 35 cross-sections of post oak (Quercus stellata) trees and constructed a fire chronology dating from 1705 to 1997. Mean fire return intervals by periods were 3.7 years (1705 to 1830), 7.6 years (1831 to 1960), and 3.6 years (1961 to 1997). Fire frequency was positively correlated (r = 0.53, p <0.01) with the expansion and migration of the Osage Tribe from 1710 to 1830. Later, between 1831 and 1900, exponential increases in Euro-American population density were negatively correlated (r =-0.64, p <0.05) with fire frequency. Fires were more likely to have occurred in drought years before 1830 than after 1830. Droughts after 1830 were related to the percent of trees scarred during years with evidence of fire.
The purpose of this paper is to provide quantitative fire history information for a geographicall... more The purpose of this paper is to provide quantitative fire history information for a geographically unique region, the Loess Hills of northwest Missouri. We sampled 33 bur oak (Quercus macrocarpa Michx.), chinkapin oak (Q. muehlenbergii Engelm.), and black oak (Q. velutina Lam.) trees from the Brickyard Hill Conservation Area in northwest Missouri. The period of tree-ring record ranged in calendar years from 1671 to 2004 and fire-scar dates (n = 97) ranged from 1672 to 1980. Fire intervals for individual trees ranged from 1 to 87 years. The mean fire interval was 6.6 years for the pre-Euro-American settlement period (1672-1820), and 5.2 years for the entire record (1672-1980). A period of more frequent fire (mean fire interval = 1.6 for 1825 to 1850) coincided with Euro-American settlement of the area. The average percentage of trees scarred at the site was 16.8%, or about 1 in 7 trees sampled per fire. No significant relationship between fire years and drought conditions was found; however, events prior to 1820 may have been associated with wet to dry mode transitions.
Eos, Transactions American Geophysical Union, 2004
Understanding the Iong-term variability in climate has important societal implications, particula... more Understanding the Iong-term variability in climate has important societal implications, particularly for agricultural regions that are suppliers to global food markets such as the central United States. Paleoclimate information from agriculturally important regions of the central United States is limited in length and resolution. Tree rings provide one of the longest and highest-resolution paleoclimate records for North America; however, millenniumJength records are primarily from extreme environments (e.g., deserts, cliffs, high elevations) and do not represent climateplant responses in the cenl ral U.S. agricultural region. Recent research on the abundance and dendrochronology (tree-ring record) of modern and subfossil cnk (Quercus macrocarpa,Quercus bicolor) collected from streams in Missouri and Iowa (93"18'W40"22'N) may enable the development of one of the longest tree-ring records in the world, with initial data suggesting that the construction of a Holocenelength-plus proxy climate record is possible. The importance of constructing this record lies in its ability to provide information concerning past climate variability and timing of specific climatic events. This article reports on the American Long Oak Chronology (ALOC), a new research effort that aims to understand the past environmental conditions of the U.S. agricultural heartland from the construction of a Holocenelength tree-ring record. Previous contributions of millenniumlength tree-ring chronologies to geoscience have been extraordinaryOne of the most notable contributions of these data has been the development and use of high-precision radiocarbon calibration cu rves. In addition,tree-ring chronologies have provided a significant proportion of paleoecological information that includes air temperature, hydroclimate, fire, frost, and streamflow reconstructions, various scales of spatial and temporal variability in atmospheric circulation, climate change model calibration,and estimates of long-term agricu ltural productivity This list outlines only some of the potential uses of the ALOC.Additional uses might include studies of volcanology glaciology, and carbon budgets lGuyette et a1.,20021.
Historical evidence suggests that great wildfires burning in the Lake States and Canada can affec... more Historical evidence suggests that great wildfires burning in the Lake States and Canada can affect atmospheric conditions several hundred miles away (Smith 1950; Wexler 1950). Several ‘dark’ or ‘yellow’ days, as such events are commonly called, have been recorded, often with anecdotal or direct evidence pointing to wildfires as the source (Plummer 1912; Ludlum 1972). One such ‘dark day’ occurred across New England in 1780, a year in which people were technologically unable to confirm the source of such a phenomenon. Here we combine written accounts and fire scar evidence to document wildfire as the likely source of the infamous Dark Day of 1780.
The regeneration and dominance of northern red oak (Quercus rubra L.) has been associated with fi... more The regeneration and dominance of northern red oak (Quercus rubra L.) has been associated with fire throughout eastern North America. Red oak in central Ontario grows near the northern edge of its distribution in mixed hardwood-coniferous forests under mesic conditions where it competes with more shade-tolerant species. We hypothesized that the abundance of red oak in these stands was largely the result of anthropogenic burning and natural fires, which would favor the regeneration and recruitment of northern red oak over such shade-tolerant species as sugar maple (Acer saccharum Marsh.). Fire histories dating from the mid-1600s were constructed by dendrochronological methods from fire scars on stumps, trees, and natural remnants of red pine (Pinus resinosa Ait.), white pine (Pinus strobus L.), and red oak at six sites in south-central Ontario. Fire histories of the sites are characterized by abrupt changes in fire interval. As much or more variance in fire interval is found within s...
We used tree-ring dating and 14 C dating to document the temporal distribution and carbon storage... more We used tree-ring dating and 14 C dating to document the temporal distribution and carbon storage of oak (Quercus spp.) wood in trees recruited and buried by streams and floodplains in northern Missouri, USA. Frequency distributions indicated that oak wood has been accumulating in Midwest streams continually since at least the late Pleistocene, about 14,000 calibrated radiocarbon years before present (cal. BP). The median residence time of an oak bole in the study streams was 3,515 years (n = 200). More than 30% of sampled oak wood entered the floodplain sediments and stream waters within the last 1,000 years, though very few samples dated to the last 150 years. Temporal variability in the record of oak recruitment to streams suggests a potentially strong influence from shifts in climate and fluvial processes, although other possible influences are addressed. Recent human impacts on streams have altered the dynamics of oak input and sequestered carbon with unknown long-term consequences. The long duration of carbon storage (mean age = 1,960 years) in this waterlogged environment appears to be strongly limited by decreasing wood density resulting from reductions in cell wall thickness. Lack of evidence of biotic degradation may imply that wood loss is largely due to abiotic hydrolyses. These findings document a continuous and long-term form of carbon storage that is sensitive to changes in climate and anthropogenic alteration of fluvial processes.
Fire scar evidence in eastern North America is sparse and complex but shows promise in defining t... more Fire scar evidence in eastern North America is sparse and complex but shows promise in defining the dynamics of these fire regimes and their influence on ecosystems. We review fire scar data, methods, and limitations, and use this information to identify and examine the factors influencing fire regimes. Fire scar data from studies at more than 40 sites in Eastern North America document fire regimes in forests with oak. Fire frequency was highly variable in both time and space even at regional scales (less than 500,000 ha). Many sites burned frequently (2-to 3-year mean fire intervals) while nearby sites (less than 40 km distant) burned infrequently (mean fire intervals more than 20 years). The fire scar record shows that major factors controlling temporal differences in fire regimes are changes in human population density, culture, and annual drought. Spatial differences in fire regimes are influenced by regional temperature, human population density, and topographic resistance to the spread of fire. Severe fire years (more than 10 percent of trees scarred at the sites) were associated with strong regional droughts that covered most of the Eastern United States and southern Ontario, Canada. Major fire years in Eastern North America occurred about 3.6 times per century before suppression efforts in forests with an oak component. Fire regimes with numerous human ignitions were more influenced by droughts. We synthesize mean fire intervals during the pre-European settlement period using an empirically derived regression model. The model was developed using two variables to predict broad scale spatial differences in fire frequency based on fire interval data derived from dendrochronologically dated fire scarred trees. Sixty-three percent of the variance in mean fire intervals was explained by mean maximum temperature and 12 percent by mapped human population density and historical documentation. The model is used to map coarse scale fire intervals in forested regions of the Eastern United States.
A predictive equation for estimating fire frequency was developed from theories and data in physi... more A predictive equation for estimating fire frequency was developed from theories and data in physical chemistry, ecosystem ecology, and climatology. We refer to this equation as the Physical Chemistry Fire Frequency Model (PC2FM). The equation was calibrated and validated with North American fire data (170 sites) prior to widespread industrial influences (before $1850 CE) related to land use, fire suppression, and recent climate change to minimize non-climatic effects. We derived and validated the empirically based PC2FM for the purpose of estimating mean fire intervals (MFIs) from proxies of mean maximum temperature, precipitation, their interaction, and estimated reactant concentrations. Parameterization of the model uses reaction rate equations based on the concentration and physical chemistry of fuels and climate. The model was then calibrated and validated using centuries of empirical fire history data. An application of the PC2FM regression equation is presented and used to estimate historic MFI as controlled by climate. We discuss the effects of temperature, precipitation, and their interactions on fire frequency using the PC2FM concept and results. The exclusion of topographic, vegetation, and ignition variables from the PC2FM increased error at fine spatial scales, but allowed for the prediction of complex climate effects at broader temporal and spatial scales. The PC2FM equation is used to map coarse-scale historic fire frequency and assess climate impacts on landscapescale fire regimes.
The spatial structure of forest ecosystems is dominated by the horizontal and vertical distributi... more The spatial structure of forest ecosystems is dominated by the horizontal and vertical distribution of trees and their attributes across space. Canopy disturbance is a primary regulator of forest spatial structure. Although the importance of tree spatial pattern is widely acknowledged as it affects important ecosystem processes such as regeneration and recruitment into the overstory, quantitative reference spatial conditions to inform silvicultural systems are lacking. This is especially true for mixedwood forests, defined as those that contain hardwoods and softwoods in the canopy. We used data from a preexisting network of plots in a complex-stage mixedwood stand to investigate the influence of canopy disturbance on stand and neighborhood-scale spatial patterns. We reconstructed canopy disturbance history and linked detected stand-wide and gap-scale disturbance events to establishment and spatial patterns of shortleaf pine. The majority of shortleaf pine establishment coincided wi...
Climate has a primary influence on the occurrence and rate of combustion in ecosystems with carbo... more Climate has a primary influence on the occurrence and rate of combustion in ecosystems with carbon-based fuels such as forests and grasslands. Society will be confronted with the effects of climate change on fire in future forests. There are, however, few quantitative appraisals of how climate will affect wildland fire in the United States. We demonstrated a method for estimating changes in fire probability based on future climate simulations of temperature and precipitation. The probability of a fire occurring in a particular climate was extracted from the Physical Chemistry Fire Frequency Model (PC2FM) and represented the rate of change in fire due to climate. Climate output data from two global climate models (GCMs) were applied to the PC2FM to estimate changes in fire probability. We calculated change in fire frequency and probabilities from the difference between current and future climates and mapped climate-forced percentage change in fire probability under each GCM for the nation at a 1.2 km 2 scale. Future fire probability estimates increased in cooler northern and high elevation regions but decreased slightly in some hotter and drier regions of the southwestern United States. Our approach's greatest strength may be reliance on only climate data and the simple principles of physical chemistry; many other nonclimatic factors that affect fire are often difficult to predict in the distant future.
A predictive model of fire frequency in North America using climate and human population density ... more A predictive model of fire frequency in North America using climate and human population density shows promise in estimating the importance of fire at broad temporal and spatial scales. Over 5 thousand fire scars from 120 sites and nine forest ecosystems were used to empirically derive and test a fire interval regression model. Three predictor variables were selected: a proxy
In this paper we describe a fire risk assessment of the Ozark Highlands. Fire risk is rated using... more In this paper we describe a fire risk assessment of the Ozark Highlands. Fire risk is rated using information on ignition potential and fuel hazard. Fuel loading, a component of the fire hazard module, is weakly predicted (r 2 = 0.19) by site-and landscape-level attributes. Fuel loading does not significantly differ between Ozark ecological landtypes. Drought and exposure are related to fuel moisture content. Drought is particularly important to the Ozark fire regime and fire risk as it is related to both ignitions and fuels.
Shortleaf pine community restoration requires an answer to the question, "What fi re frequency is... more Shortleaf pine community restoration requires an answer to the question, "What fi re frequency is appropriate for shortleaf pine regeneration and survival?" The answer to this question is one of the most critical to successful restoration through fi re management. We used three sources of information from Missouri to determine appropriate burning frequencies: a 400-yr historic shortleaf pine growth and fi re-scar database, fi re effects data from prescribed burning sites, and a vegetation dynamics prediction model that is widely used for characterizing fi re regimes. The historic shortleaf pine and fi re scar database provides actual past scenarios of regeneration dates, growth, survival, and associated fi re events. Shortleaf pine regeneration established most commonly during the 4 years following fi re events and generally decreased in abundance with years since fi re. Surviving seedlings were those that were not fi re scarred the year following establishment, and the mean number of years to a subsequent fi re was about 7 years. Fire effects data from prescribed burn sites revealed that both hardwood and pine regeneration showed substantially increased mortality after four consecutive dormant season burns, but oak and hickory species were more likely to survive frequent fi re. Mortality of advance regeneration was generally low in all hardwood species after one burn, while shortleaf pine seedlings had high mortality rates. The model showed 8 to 15 yr intervals are likely best for balancing both continual regeneration and recruitment. Model prediction runs for 500 years showed a signifi cantly decreased pine component in the absence of burning. Conversely, long-term frequent burning (1-to 3-yr intervals) resulted in abundant regeneration, but poor survival and ultimately decreased abundance in mid-and late-successional forests. In summary, all three sources support the effi cacy of frequent burning (1 to 4 yrs) in promoting pine regeneration, but survival and continued recruitment require longer fi re intervals (8 to 15 yrs). Fire management prescriptions that incorporate both frequent burning and longer intervals will likely provide for the most long-term regeneration and recruitment success.
We present the fire history of a 1-km 2 area that is a mosaic of oak forest, savanna, and fen on ... more We present the fire history of a 1-km 2 area that is a mosaic of oak forest, savanna, and fen on the White Ranch State Forest, Howell County, Missouri. We dated 135 fire scars on 35 cross-sections of post oak (Quercus stellata) trees and constructed a fire chronology dating from 1705 to 1997. Mean fire return intervals by periods were 3.7 years (1705 to 1830), 7.6 years (1831 to 1960), and 3.6 years (1961 to 1997). Fire frequency was positively correlated (r = 0.53, p <0.01) with the expansion and migration of the Osage Tribe from 1710 to 1830. Later, between 1831 and 1900, exponential increases in Euro-American population density were negatively correlated (r =-0.64, p <0.05) with fire frequency. Fires were more likely to have occurred in drought years before 1830 than after 1830. Droughts after 1830 were related to the percent of trees scarred during years with evidence of fire.
The purpose of this paper is to provide quantitative fire history information for a geographicall... more The purpose of this paper is to provide quantitative fire history information for a geographically unique region, the Loess Hills of northwest Missouri. We sampled 33 bur oak (Quercus macrocarpa Michx.), chinkapin oak (Q. muehlenbergii Engelm.), and black oak (Q. velutina Lam.) trees from the Brickyard Hill Conservation Area in northwest Missouri. The period of tree-ring record ranged in calendar years from 1671 to 2004 and fire-scar dates (n = 97) ranged from 1672 to 1980. Fire intervals for individual trees ranged from 1 to 87 years. The mean fire interval was 6.6 years for the pre-Euro-American settlement period (1672-1820), and 5.2 years for the entire record (1672-1980). A period of more frequent fire (mean fire interval = 1.6 for 1825 to 1850) coincided with Euro-American settlement of the area. The average percentage of trees scarred at the site was 16.8%, or about 1 in 7 trees sampled per fire. No significant relationship between fire years and drought conditions was found; however, events prior to 1820 may have been associated with wet to dry mode transitions.
Eos, Transactions American Geophysical Union, 2004
Understanding the Iong-term variability in climate has important societal implications, particula... more Understanding the Iong-term variability in climate has important societal implications, particularly for agricultural regions that are suppliers to global food markets such as the central United States. Paleoclimate information from agriculturally important regions of the central United States is limited in length and resolution. Tree rings provide one of the longest and highest-resolution paleoclimate records for North America; however, millenniumJength records are primarily from extreme environments (e.g., deserts, cliffs, high elevations) and do not represent climateplant responses in the cenl ral U.S. agricultural region. Recent research on the abundance and dendrochronology (tree-ring record) of modern and subfossil cnk (Quercus macrocarpa,Quercus bicolor) collected from streams in Missouri and Iowa (93"18'W40"22'N) may enable the development of one of the longest tree-ring records in the world, with initial data suggesting that the construction of a Holocenelength-plus proxy climate record is possible. The importance of constructing this record lies in its ability to provide information concerning past climate variability and timing of specific climatic events. This article reports on the American Long Oak Chronology (ALOC), a new research effort that aims to understand the past environmental conditions of the U.S. agricultural heartland from the construction of a Holocenelength tree-ring record. Previous contributions of millenniumlength tree-ring chronologies to geoscience have been extraordinaryOne of the most notable contributions of these data has been the development and use of high-precision radiocarbon calibration cu rves. In addition,tree-ring chronologies have provided a significant proportion of paleoecological information that includes air temperature, hydroclimate, fire, frost, and streamflow reconstructions, various scales of spatial and temporal variability in atmospheric circulation, climate change model calibration,and estimates of long-term agricu ltural productivity This list outlines only some of the potential uses of the ALOC.Additional uses might include studies of volcanology glaciology, and carbon budgets lGuyette et a1.,20021.
Historical evidence suggests that great wildfires burning in the Lake States and Canada can affec... more Historical evidence suggests that great wildfires burning in the Lake States and Canada can affect atmospheric conditions several hundred miles away (Smith 1950; Wexler 1950). Several ‘dark’ or ‘yellow’ days, as such events are commonly called, have been recorded, often with anecdotal or direct evidence pointing to wildfires as the source (Plummer 1912; Ludlum 1972). One such ‘dark day’ occurred across New England in 1780, a year in which people were technologically unable to confirm the source of such a phenomenon. Here we combine written accounts and fire scar evidence to document wildfire as the likely source of the infamous Dark Day of 1780.
The regeneration and dominance of northern red oak (Quercus rubra L.) has been associated with fi... more The regeneration and dominance of northern red oak (Quercus rubra L.) has been associated with fire throughout eastern North America. Red oak in central Ontario grows near the northern edge of its distribution in mixed hardwood-coniferous forests under mesic conditions where it competes with more shade-tolerant species. We hypothesized that the abundance of red oak in these stands was largely the result of anthropogenic burning and natural fires, which would favor the regeneration and recruitment of northern red oak over such shade-tolerant species as sugar maple (Acer saccharum Marsh.). Fire histories dating from the mid-1600s were constructed by dendrochronological methods from fire scars on stumps, trees, and natural remnants of red pine (Pinus resinosa Ait.), white pine (Pinus strobus L.), and red oak at six sites in south-central Ontario. Fire histories of the sites are characterized by abrupt changes in fire interval. As much or more variance in fire interval is found within s...
We used tree-ring dating and 14 C dating to document the temporal distribution and carbon storage... more We used tree-ring dating and 14 C dating to document the temporal distribution and carbon storage of oak (Quercus spp.) wood in trees recruited and buried by streams and floodplains in northern Missouri, USA. Frequency distributions indicated that oak wood has been accumulating in Midwest streams continually since at least the late Pleistocene, about 14,000 calibrated radiocarbon years before present (cal. BP). The median residence time of an oak bole in the study streams was 3,515 years (n = 200). More than 30% of sampled oak wood entered the floodplain sediments and stream waters within the last 1,000 years, though very few samples dated to the last 150 years. Temporal variability in the record of oak recruitment to streams suggests a potentially strong influence from shifts in climate and fluvial processes, although other possible influences are addressed. Recent human impacts on streams have altered the dynamics of oak input and sequestered carbon with unknown long-term consequences. The long duration of carbon storage (mean age = 1,960 years) in this waterlogged environment appears to be strongly limited by decreasing wood density resulting from reductions in cell wall thickness. Lack of evidence of biotic degradation may imply that wood loss is largely due to abiotic hydrolyses. These findings document a continuous and long-term form of carbon storage that is sensitive to changes in climate and anthropogenic alteration of fluvial processes.
Fire scar evidence in eastern North America is sparse and complex but shows promise in defining t... more Fire scar evidence in eastern North America is sparse and complex but shows promise in defining the dynamics of these fire regimes and their influence on ecosystems. We review fire scar data, methods, and limitations, and use this information to identify and examine the factors influencing fire regimes. Fire scar data from studies at more than 40 sites in Eastern North America document fire regimes in forests with oak. Fire frequency was highly variable in both time and space even at regional scales (less than 500,000 ha). Many sites burned frequently (2-to 3-year mean fire intervals) while nearby sites (less than 40 km distant) burned infrequently (mean fire intervals more than 20 years). The fire scar record shows that major factors controlling temporal differences in fire regimes are changes in human population density, culture, and annual drought. Spatial differences in fire regimes are influenced by regional temperature, human population density, and topographic resistance to the spread of fire. Severe fire years (more than 10 percent of trees scarred at the sites) were associated with strong regional droughts that covered most of the Eastern United States and southern Ontario, Canada. Major fire years in Eastern North America occurred about 3.6 times per century before suppression efforts in forests with an oak component. Fire regimes with numerous human ignitions were more influenced by droughts. We synthesize mean fire intervals during the pre-European settlement period using an empirically derived regression model. The model was developed using two variables to predict broad scale spatial differences in fire frequency based on fire interval data derived from dendrochronologically dated fire scarred trees. Sixty-three percent of the variance in mean fire intervals was explained by mean maximum temperature and 12 percent by mapped human population density and historical documentation. The model is used to map coarse scale fire intervals in forested regions of the Eastern United States.
A predictive equation for estimating fire frequency was developed from theories and data in physi... more A predictive equation for estimating fire frequency was developed from theories and data in physical chemistry, ecosystem ecology, and climatology. We refer to this equation as the Physical Chemistry Fire Frequency Model (PC2FM). The equation was calibrated and validated with North American fire data (170 sites) prior to widespread industrial influences (before $1850 CE) related to land use, fire suppression, and recent climate change to minimize non-climatic effects. We derived and validated the empirically based PC2FM for the purpose of estimating mean fire intervals (MFIs) from proxies of mean maximum temperature, precipitation, their interaction, and estimated reactant concentrations. Parameterization of the model uses reaction rate equations based on the concentration and physical chemistry of fuels and climate. The model was then calibrated and validated using centuries of empirical fire history data. An application of the PC2FM regression equation is presented and used to estimate historic MFI as controlled by climate. We discuss the effects of temperature, precipitation, and their interactions on fire frequency using the PC2FM concept and results. The exclusion of topographic, vegetation, and ignition variables from the PC2FM increased error at fine spatial scales, but allowed for the prediction of complex climate effects at broader temporal and spatial scales. The PC2FM equation is used to map coarse-scale historic fire frequency and assess climate impacts on landscapescale fire regimes.
The spatial structure of forest ecosystems is dominated by the horizontal and vertical distributi... more The spatial structure of forest ecosystems is dominated by the horizontal and vertical distribution of trees and their attributes across space. Canopy disturbance is a primary regulator of forest spatial structure. Although the importance of tree spatial pattern is widely acknowledged as it affects important ecosystem processes such as regeneration and recruitment into the overstory, quantitative reference spatial conditions to inform silvicultural systems are lacking. This is especially true for mixedwood forests, defined as those that contain hardwoods and softwoods in the canopy. We used data from a preexisting network of plots in a complex-stage mixedwood stand to investigate the influence of canopy disturbance on stand and neighborhood-scale spatial patterns. We reconstructed canopy disturbance history and linked detected stand-wide and gap-scale disturbance events to establishment and spatial patterns of shortleaf pine. The majority of shortleaf pine establishment coincided wi...
Uploads
Papers by Daniel Dey