Plant Ecology

, Volume 212, Issue 9, pp 1513–1525

Understory response to varying fire frequencies after 20 years of prescribed burning in an upland oak forest

Authors

  • Jesse A. Burton
    • National Park Service
    • Department of Natural Resource Ecology and ManagementOklahoma State University
  • Samuel D. Fuhlendorf
    • Department of Natural Resource Ecology and ManagementOklahoma State University
  • David M. LeslieJr.
    • U.S. Geological Survey, Oklahoma Cooperative Fish and Wildlife Research Unit, Department of Natural Resource Ecology and ManagementOklahoma State University
Article

DOI: 10.1007/s11258-011-9926-y

Cite this article as:
Burton, J.A., Hallgren, S.W., Fuhlendorf, S.D. et al. Plant Ecol (2011) 212: 1513. doi:10.1007/s11258-011-9926-y

Abstract

Ecosystems in the eastern United States that were shaped by fire over thousands of years of anthropogenic burning recently have been subjected to fire suppression resulting in significant changes in vegetation composition and structure and encroachment by invasive species. Renewed interest in use of fire to manage such ecosystems will require knowledge of effects of fire regime on vegetation. We studied the effects of one aspect of the fire regime, fire frequency, on biomass, cover and diversity of understory vegetation in upland oak forests prescribe-burned for 20 years at different frequencies ranging from zero to five fires per decade. Overstory canopy closure ranged from 88 to 96% and was not affected by fire frequency indicating high tolerance of large trees for even the most frequent burning. Understory species richness and cover was dominated by woody reproduction followed in descending order by forbs, C3 graminoids, C4 grasses, and legumes. Woody plant understory cover did not change with fire frequency and increased 30% from one to three years after a burn. Both forbs and C3 graminoids showed a linear increase in species richness and cover as fire frequency increased. In contrast, C4 grasses and legumes did not show a response to fire frequency. The reduction of litter by fire may have encouraged regeneration of herbaceous plants and helped explain the positive response of forbs and C3 graminoids to increasing fire frequency. Our results showed that herbaceous biomass, cover, and diversity can be managed with long-term prescribed fire under the closed canopy of upland oak forests.

Keywords

Fire frequencyFoliar coverHerbaceous plantsSpecies richnessUnderstory diversity

Introduction

Vegetation of eastern North America was shaped by human fire for thousands of years before the arrival of European settlers (Pyne et al. 1996; Delcourt 2004). As settlement of the continent progressed, the use of fire to manage vegetation was curtailed to protect property. A century of fire suppression has dramatically changed forest and savannah ecosystems by fostering increased tree density, changes in species composition, and reduction of ground-level vegetation in forest understories (Chapman et al. 2006; DeSantis et al. 2010). Increasing recognition of these changes has resulted in strong interest in the use of prescribed fire to restore biological diversity of ecosystems, reduce invasive species, and decrease hazardous fuel buildup (Pyne et al. 1996). But knowledge of how to use prescribed fire is lacking in many regions, and research is needed to learn how to use it effectively. Because redevelopment of fire practices in southeastern forests has been relatively recent, much of the research concerning effects of prescribed fire on forest vegetation has been relatively short-term (Taft 2003; Hutchinson et al. 2005). In addition, much of the research has concerned fire use for improving regeneration of commercially important forest trees especially oak (Quercus sp.) (Brose and Van Lear 1998) rather than ecosystem restoration.

Use of prescribed fire to restore ecosystems and manage fuels is a major concern worldwide. Where there is little information about historical fire regimes, research is needed to determine the role of fire and effective methods of fire management. Among the most important characteristics of a fire regime are frequency, intensity, and season of fire (Pyne et al. 1996). Fire regime can determine vegetation composition and diversity by its effects on regeneration. Although fire stimulates sprouting and creates a favorable seedbed, too frequent burning may not allow enough time for replenishing root system reserves needed for sprouting or renewing seed banks (Odion and Tyler 2002; Russell-Smith et al. 2002). Frequent fire has been found to favor grasses capable of rapid seedling establishment and vegetative growth over shrubs and trees that were slower to regenerate after fire (Vila et al. 2001; Haidinger and Keeley 1993). Life-history traits determine how plant species respond to fire frequency and intensity (Gill 1981).

The low intensity fires characteristic of many forests of southeastern North America rarely kill overstory trees, and it has been determined that overstory thinning may be necessary for restoration of herbaceous vegetation (Hutchinson et al. 2005). Stand thinning was found to enhance the effects of prescribed fire when the goal was to establish an earlier successional stage in pine-hardwood and pine stands (Masters et al. 1993; Phillips and Waldrop 2008). Harvesting or thinning of overstory trees can be very costly (Laughlin et al. 2008) and may conflict with societal values such as preservation of old-growth forests. The benefit of overstory thinning for production of grasses and forbs was lost unless the stands were burned every 3–4 years (Masters et al. 1993, 2006).

We studied the effects of prescribed burning on understory herbaceous vegetation in old-growth xeric oak forests of Oklahoma in south-central North America at the western limit of eastern deciduous forests. Fire frequency in these forests prior to European settlement was approximately 2.5 fires per decade (Clark et al. 2007; DeSantis et al. 2010); more recently fire frequency increased in some areas but has been completely suppressed in others. This study was conducted in an area where prescribed fire had been used for 20 years at frequencies from zero to five fires per decade. A recent study at this location showed prescribed fires did not affect overstory trees even after 20 years of burning at the highest frequency (Burton et al. 2010). In contrast, density and species diversity of mid-story woody saplings and shrubs were reduced by fire frequencies greater than two per decade. In this article we explore the relationships between fire frequency and diversity in the herb layer of the forest. We predict that biomass and diversity of understory herbaceous species increase with increasing fire frequency. The main objective was to determine whether differences in fire frequency affected biomass and species richness of understory herbaceous plants. Another important question was whether the different functional groups of understory plants—C4 grasses, C3 graminoids, forbs, legumes, and woody plants—responded differently to fire frequency. The expected outcome of this research was greater knowledge of fire frequency effects on understory vegetation for development of improved prescribed fire treatments.

Methods

Study area

This study was conducted at the Okmulgee Game Management Area (OGMA; 35°37′N, 96°03′W; Fig. 1) managed by Oklahoma Department of Wildlife Conservation (ODWC). The OGMA is primarily forestland that has remained relatively undisturbed due to poor timber value and shallow rocky soils (Stahle and Chaney 1994). We decided to limit the study to the Hector–Endsaw complex soil because it represented approximately 75% of the OGMA and was typical of upland oak forests of the region. This soil type was characterized as well-drained, non-arable, shallow stony fine sandy loam with bedrock at a depth of approximately 30 cm on hill or mountain topography of 5–30% slopes (Sparwasser et al. 1968). The forests were dominated by post oak (Quercus stellata Wangenh.) with subdominants of blackjack oak (Q. marilandica Münchh.), black oak (Q. velutina Lam.), black hickory (Carya texana Buckley), and winged elm (Ulmus alata Michx.) (Burton et al. 2010). Mean annual temperature is 16.1°C, and mean monthly temperatures are in the range of 33.9°C in July to −3.9°C in January. The area has a humid subtropical climate receiving approximately 111 cm of precipitation annually, with a range of 54.5–156.2 cm (Oklahoma Climatological Survey 2005). Precipitation is well distributed throughout the year and May has the highest monthly precipitation.
https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig1_HTML.gif
Fig. 1

Location of the Okmulgee Game Management Area (OGMA) and surrounding land resource areas. Modified from USDA, NRCS (2006). OGMA is in a forest peninsula at the western edge of the central forest region, surrounded by the lower lying mesic Cherokee Prairies. Numbers in burn treatment units correspond to fire frequency (fires per decade)

In 1989, the OGMA was divided into prescribed burning units of 100–600 ha and fire frequencies from zero to five per decade (Fig. 1). Although there were no records of fire occurrences in the OGMA prior to the prescribed fire program, a recent study within a single stand of 100 ha at the OGMA found fire scar evidence of 2.5 fires per decade for the period before Euro-American settlement between 1750 and 1900 and five fires per decade for the period after Euro-American settlement between 1900 and 1989 (DeSantis et al. 2010). It was assumed that vegetation was similar across all units prior to the prescribed burning program. All prescribed fires were low-intensity dormant season burns conducted in February and March (Table 1) when relative humidity was between 30 and 50%, temperature <27°C, and winds <25 kph.
Table 1

Year and month of prescribed fires by units. FPD fires per decade, YSLF years since last fire

Year

Burn unit

1

2

3

4

6

7

10

13

2008

   

Mar

Mar

   

2007

Feb

Feb

      

2006

        

2005

  

Feb

Feb

 

Feb

Mar

 

2004

 

Mar

 

Mar

    

2003

Mar

     

Mar

 

2002

   

Feb

Feb

   

2001

Mar

  

Feb

  

Mar

 

2000

Feb

Feb

Mar

     

1999

Feb

       

1998

Feb

       

1997

Feb

 

Feb

 

Feb

   

1996

     

Feb

Feb

 

1995

Feb

       

1994

 

Mar

 

Mar

  

Mar

 

1993

Mar

   

Mar

   

1992

 

Feb

 

Feb

  

Feb

 

1991

  

Feb

     

1990

        

1989

Feb

       

1988

        

Fires

10

5

4

7

4

2

6

0

FPD

5.0

2.5

2.0

3.5

2.0

1.0

3.0

0.0

YSLF

1.4

1.4

3.4

0.4

0.4

3.4

3.4

20+

Field sampling

Twenty 10-m × 10-m sample plots were located at random within each of eight prescribed burning units, hereafter called treatments, and measured during June to September, 2008. Plots were measured for diameter at breast height (1.4 m, dbh) of all trees >5 cm dbh, slope, and aspect. Four 1-m × 1-m sub-plots were nested within the four corners of the sample plots. At each sub-plot, we visually estimated percent cover utilizing a customized Braun-Blanquet cover scale (Kent and Coker 1992) for the following: exposed soil, rock, leaf litter, vascular plant functional group, and vascular plant species. Litter depth was measured at four points in the sub-plots. Overstory canopy cover was measured at each sub-plot using a forest canopy spherical densiometer. All plants were identified to species level with the exception of a few taxonomic groups that were classified to genus level; therefore, measurements of richness and diversity indices were conservative. Nomenclature for all plant species followed the PLANTS database (USDA, NRCS 2008).

In September of the same year, biomass samples were collected from three treatments: 0, 2.5, and 5 burns per decade. One year had elapsed since the last fire for both burned units. We collected biomass from five 200-m transects in each treatment. Transects were randomly located and consisted of five 0.5-m × 0.5-m quadrats spaced 50 m apart. All plant matter was collected to mineral soil from herbaceous plants and woody plants <1.4 m tall. Living plant matter was classified by functional group, and dead plant matter was classified as litter. Samples were dried for 24 h at 70°C and weighed.

Data analysis

Basal area (m2 ha−1) was calculated from the dbh measurements. Foliar cover was analyzed by species and sum of species by plant functional group. Species cover data were square-root transformed by plot prior to statistical analysis. Species cover data were used to calculate richness, diversity, and evenness indices. To examine diversity of the understory, Shannon’s indices were used to calculate species diversity (H = −∑Pi ln Pi) and equitability (J = H/ln S) from species cover values in each treatment, where Pi was the relative cover of species i and S was the species richness per treatment unit (Begon et al. 2006). The means of response variables including ground cover, foliar cover, species richness, and diversity indices were calculated for each of the eight treatments based on 20 samples per burn treatment.

Simple linear regression analysis was used to determine the significance of the relationship between individual response variables including ground cover, foliar cover, species richness, and diversity indices and both explanatory variables fire frequency (N = 8) and time since last fire (N = 7, excluded no-fire treatment). Analyses were done for all plant species together and separately for each species and plant functional group. Analyses were done separately for each explanatory variable. Analysis of variance was used to determine the significance of effects of fire frequency on biomass production of plant functional groups. Significance was determined with P ≤ 0.05 for all statistical tests.

Results

Forest canopy structure and ground cover

Fire frequency did not affect either total basal area which ranged from 22.6 to 26.7 m2 ha−1 or tree density which ranged from 940 to 1240 stems ha−1 among treatments. Overstory canopy cover ranged from 88.5 to 95.7% and likewise, was not affected by fire frequency. Regression analyses showed no relation between fire frequency and percent bare soil, litter cover, and litter depth (Fig. 2). In contrast, time since the last fire had strong effects on ground cover; bare soil declined nearly tenfold, litter cover increased over 50%, and litter depth increased twofold, as time since fire increased from 1 year to 3 years.
https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig2_HTML.gif
Fig. 2

Effects of fire frequency and years since last fire on bare soil, litter cover, and depth. Solid lines indicate significant linear relationships based on regression analysis

Understory functional group cover and biomass

The effects of variation in fire frequency on the structure and functional group composition of the understory vegetation was visually very striking (Fig. 3). Total cover was greatest for woody plants followed in decreasing order by forbs, C3 graminoids, C4 grasses, and legumes (Fig. 4). Total woody plant cover ranged from 68 to 104% among burning treatments and showed no relation to fire frequency. Individual species achieving the highest cover averaged across treatments were fragrant sumac (Rhus aromatica Aiton, 15%) and winged elm (13%). Post oak (12%), blackjack oak (8%), and black hickory (6%) were close behind. The only plant functional group that increased cover with time since last fire was woody plants; this group increased more than 30% during the 3 years after a burn.
https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig3_HTML.jpg
Fig. 3

Photographs of four treatment units: a zero fires per decade, b one fire per decade, c two and one half fires per decade, d five fires per decade. Canopy cover, basal area, and density of trees greater than 5 cm dbh were similar for all units; however, understory forb and C3 graminoid cover greatly increased with fire frequency

https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig4_HTML.gif
Fig. 4

Effects of fire frequency and years since last fire on understory cover by functional group. Solid lines indicate significant liner relations based on regression analysis

Forb cover ranged from 20 to 70% among treatments (Fig. 4) and increased threefold in response to increased fire frequency from zero to five fires per decade. Forb species with the highest cover across all treatments were elmleaf goldenrod (Solidago ulmifolia Muhl. ex Willd., 9%) and Parlin’s pussytoes (Antennaria parlinii Fernald, 9%).

C3 graminoid cover ranged from 24 to 57% among burning treatments (Fig. 4) and increased more than twofold with increased fire frequency. Individual C3 graminoids with the highest cover averaged across all treatments were Indian woodoats (Chasmanthium latifolium (Michx.) Yates, 8%), slimleaf panicgrass (Dichanthelium linearifolium (Scribn. ex Nash) Gould, 8%), and poverty oatgrass (Danthonia spicata (L.) P. Beauv. ex Roem. & Schult., 5%).

C4 grass cover ranged from 4 to 48% among burning treatments and was not affected by fire frequency (Fig. 4). C4 grasses with the greatest cover averaged across all treatments were muhly grass (Muhlenbergia spp. Schreb., 6%) and big bluestem (Andropogon gerardii Vitman, 5%).

Legume cover ranged from 5 to 23% among burn treatments and was not related to fire frequency (Fig. 4). Legumes with the highest cover averaged across treatments were creeping lespedeza (Lespedeza repens (L.) W. Bartram, 4%) and slender lespedeza (Lespedeza virginica (L.) Britton, 3%).

Seven understory herbaceous plant species showed a response to fire frequency and all were positive (Fig. 5). One woody species decreased cover with time since last fire and two herbaceous plants showed increasing cover as time progressed since last fire (Fig. 6).
https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig5_HTML.gif
Fig. 5

Effects of fire frequency on cover of selected understory species. Solid lines indicate significant liner relations based on regression analysis

https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig6_HTML.gif
Fig. 6

Effects of years since last fire on cover of selected understory species. Solid lines indicate significant liner relations based on regression analysis

Biomass of the understory averaged across treatments was greatest for woody plants (454 kg ha−1) followed in decreasing order by C3 and C4 graminoids (279 kg ha−1), forbs (76 kg ha−1) and legumes (3 kg ha−1). Only combined C3 and C4 graminoid biomass showed an effect of fire frequency, increasing from 130 kg ha−1 at zero fires per decade to 511 kg ha−1 at five fires per decade.

Understory functional group species richness and diversity

Over 170 taxa were identified. Mean plot species richness was greatest for woody plants followed in decreasing order by forbs, C3 graminoids, C4 grasses, and legumes (Table 2). Only forbs and C3 graminoids showed a plot richness response to fire frequency and it was positive. None of the functional groups showed a plot species richness response to time since last fire. Treatment richness was greatest for forbs followed in decreasing order by woody plants, C3 graminoids, C4 grasses, and legumes. None of the functional groups showed a response of treatment richness to fire frequency and time since last fire. Both total plot richness and total treatment richness showed no response to fire frequency and time since last fire.
Table 2

Plot richness, treatment richness, Shannon’s diversity index (H), and Shannon’s equitability index (J) by functional group and all species taken together and results of linear regression analysis for the relation with fires per decade (FPD) and years since last fire (YSLF)

Functional group

Plot richness

Treatment richness

Diversity (H)

Equitability (J)

\( \overline{X} \)

P value

\( \overline{X} \)

P value

\( \overline{X} \)

P value

\( \overline{X} \)

P value

FPD

YSLF

FPD

YSLF

FPD

YSLF

FPD

YSLF

Woody plants

5.93

0.968

0.368

21.75

0.403

0.134

2.20

0.478

0.700

0.72

0.086

0.228

Forbs

4.45

0.029

0.856

29.88

0.157

0.270

2.18

0.470

0.630

0.64

0.949

0.910

C3 graminoids

3.68

0.014

0.300

12.25

0.594

0.589

1.83

0.989

0.458

0.73

0.849

0.167

C4 grasses

1.66

0.227

0.134

7.00

0.391

0.467

1.22

0.804

0.530

0.76

0.496

0.321

Legumes

1.40

0.131

0.336

8.00

0.058

0.329

1.62

0.215

0.556

0.78

0.598

0.536

Total

17.11

0.066

0.409

78.88

0.100

0.589

3.30

0.018

0.422

0.76

0.128

0.038

Bold number indicates significant regression

Total species richness showed no response to fire frequency and time since last fire (Fig. 7). Total plant species diversity as measured by Shannon’s diversity index (H) responded positively to fire frequency but showed no response to time since last fire. However, species equitability (J) was highest immediately following the prescribed fire, had a negative relation to time since last fire, and no relation to fire frequency.
https://static-content.springer.com/image/art%3A10.1007%2Fs11258-011-9926-y/MediaObjects/11258_2011_9926_Fig7_HTML.gif
Fig. 7

Effects of fire frequency and years since last fire on species richness and diversity indices. Solid lines indicate significant liner relations based on regression analysis

Discussion

Results of this study demonstrated beneficial effects of low-intensity dormant season fire for understory herbaceous vegetation in upland oak forests. The strength of these findings was enhanced because they came from operational prescribed burning conducted on large forest tracts over a 20-year period. That these effects were strong under a full canopy (88.5–95.7% closed) suggested that low light was likely not the only factor limiting understory herbaceous plants. This was a very important finding, as it contrasted with a common conclusion of research in oak forests that a closed canopy must be thinned to get a significant response of herbaceous plants to burning (Hutchinson et al. 2005; Harrington and Kathol 2009; Franklin et al. 2003).

The dominance of woody plant cover in the understory most likely was due to the sprouting capability of the dominant species, which insured rapid recovery after fire. Oaks are well known for prolific sprouting after fire (Johnson 1992, 1993; Van Lear and Watt 1992). Although we found woody cover did not change across the range of zero to five fires per decade, a study in oak savannas and woodlands in Minnesota, USA found substantial reductions in woody cover when fire frequencies increased to between five and eight fires per decade (Peterson et al. 2007). Winter burns in longleaf pine (Pinus palustris Mill.) stands did not reduce the density of woody understory (Waldrop et al. 1992); only annual summer burns were sufficient to reduce woody cover. On the other hand, the number of years between winter fires determined the size of the woody understory consistent with our finding that woody cover increased with time since last fire. That the understory of the unburned stand after 20 years without fire had as much woody cover as the burned stands suggested woody cover will accumulate over time even without fire stimulated sprouting. On the other hand, because sprouting stimulated by burning did not result in higher woody cover, growth of the woody understory may have been limited by site resources such as light, nutrients or moisture.

Forbs often have the largest number of species in oak forests (Hutchinson et al. 2005; Harrington and Kathol 2009) and can show a positive response to burning (Peterson et al. 2007; Peterson and Reich 2008). This group included many species such as Parlin’s pussytoes, wild petunia (Ruellia spp. L.), and elmleaf goldenrod that tolerate or thrive after fire and grow well in savannas or closed forests (Bader 2001; Tyrl et al. 2002; Holzmueller et al. 2009). In prairie grasslands in the USA, the strong positive response of forbs to increasing fire frequency reflects the tendency for these species to benefit from reduction of the matrix grass species (Collins and Gibson 1990). In the understory of the closed forest, there was no reduction in matrix species indicating factors such as increased nutrient cycling may play a greater role than competitive interactions. If prescribed fire increased niche diversity across the large units in our study because the intensity of fire and its effects were patchy, forbs may be showing the benefit from this effect, because they had the greatest overall mean treatment richness. Forbs may have a higher capacity to make use of resources made available by fire even under a closed canopy while grasses require a more open canopy to realize the same benefits (Phillips and Waldrop 2008).

In contrast to a tallgrass prairie dominated by C4 warm season grasses (Collins and Gibson 1990), the forest in our study had higher abundance of C3 cool season grasses, sedges, and rushes in the understory with greater cover and species richness. The C3 grasses were dominated by species adapted to forest habitats including Indian woodoats, poverty oatgrass, and several species of rosette grasses (Dichanthelium spp. (Hitchc. & Chase) Gould) (Tyrl et al. 2002). One of them, poverty oatgrass, showed a strong positive cover response to increasing fire frequency. Research results concerning effects of fire on C3 graminoids are meager and highly specific; they showed fire can increase density but the effect depended on the season of the fire (Sparks et al. 1998; Taft 2003).

Lack of a benefit from fire among the C4 grasses contrasts sharply with studies in prairies where they show strong increases after fire apparently due to increased energy and nutrient availability. But the dominant C4 grass, muhly grass, is known to thrive under a closed forest canopy like those in our study (Tyrl et al. 2002). Although the next two important C4 species, big bluestem and little bluestem (Schizachyrium scoparium (Michx.) Nash), are known to predominately inhabit prairie rather than forest habitats, they both are important in the nearby oak-pine savanna ecosystem where they benefit from fire (Masters et al. 1993). Big bluestem showed a strong increase in cover over three growing seasons after fire indicating further research should explore the interaction of the effects of fire frequency and time since last fire. Lower cover and species richness of C4 grasses in the forest understory compared to C3 graminoids was consistent with their high light energy requirement which may reduce the advantage from their high water-use and nitrogen-use efficiency (Pearcy and Ehleringer 1984) when they compete for resources in low-light environments. The low light under a closed canopy may limit the capacity for C4 grasses to benefit from indirect effects of fire, such as removal of litter and competition and release of nutrients (Harrington and Kathol 2009).

Litter consumption by repeated burning could benefit herbaceous vegetation by creating a more favorable environment for germination and establishment of plants. Reduced litter may increase germination success by increasing mineral soil surface temperatures and reducing physical barriers to seed deposition and seedling emergence (Sydes and Grime 1981; Facelli and Pickett 1991). In xeric longleaf pine woodlands, herbaceous vegetation was more severely impacted by litter deposition and subsequent forest floor development than by overstory canopy cover and midstory tree density (Hiers et al. 2007). Although we found time since last fire was the major determinant of litter cover and depth, the number of times per decade that fire reduced the litter was the major determiner of cover and species richness of forbs and C3 graminoids.

The positive responses of forb and C3 graminoid cover and richness and of total understory cover and diversity to increased fire frequency in our study may have resulted, in part, from the significant reduction in density of saplings and shrubs >1.4 m tall at fire frequencies between two and five fires per decade reported elsewhere (Burton et al. 2010). A previous study in longleaf pine woodlands in Florida, USA found a reduction in litter production from saplings and shrubs due to their removal by prescribed burning benefitted understory herbaceous plants (Hiers et al. 2007). Reduction in below-ground competition for moisture and nutrients may explain the positive herbaceous response. However, the positive herbaceous response did not result from the reduction in total cover because there was no change in cover measured at 1.4 m with a densitometer.

We found that total plant diversity, forb and C3 graminoid cover and richness and combined C3 and C4 graminoid biomass significantly increased with an increase in fire frequency from zero to five fires per decade in upland oak forests. These forests have experienced fire frequencies higher than five per decade during the past 250 years according to fire scar analysis (DeSantis et al. 2010) leading to the question of what would be the plant response to more frequent burning. A study in longleaf pine stands in South Carolina and northeast Florida, USA showed additional fires beyond five per decade could continue to increase understory vegetation diversity (Glitzenstein et al. 2003). On the other hand, a study in oak savannas of Minnesota, USA found that increased fire frequency could reduce diversity, as more frequent fires began to remove species, particularly woody species (Peterson et al. 2007; Peterson and Reich 2008). That none of the understory species at OGMA had a negative response to increased fire frequency suggests resilience to a regime of biennial burning. Further study is needed to ascertain effects of prescribed fire frequencies greater than five per decade on vegetation composition and biomass.

Our results showed that herbaceous biomass, cover, and diversity can be managed with long-term prescribed fire under the closed canopy of upland oak forests. Fire at frequencies of less than two fires per decade led to declines in herbaceous vegetation most likely due to increased woody shrub and sapling diversity and density (Burton et al. 2010). More frequent fire, up to five fires per decade, benefited diversity and biomass of some herbaceous species and disadvantaged woody shrubs and saplings (Burton et al. 2010). This information provides the land manager the knowledge to use prescribed fire to create a diverse landscape with patches of vegetation in different conditions depending on the fire frequency. Managers of wildlife management areas will benefit from this knowledge, as their goal is to create a landscape with diverse habitats suitable to a wide range of wildlife species.

Acknowledgments

We thank Michael W. Palmer for assistance with study design, analyses, and specimen identification, Mark S. Gregory for help with GIS and GPS methods and Bruce H. and Louise D. Burton, Ryan D. DeSantis, Ryan J. Williams, Amber D. Breland, Glen M. Hensley, and Stephen L. Winter for assistance with field sampling. Funding was provided by the Federal Aid, Pittman-Robertson Wildlife Restoration Act under Project W-160-R of the Oklahoma Department of Wildlife Conservation and Oklahoma State University. The project was administered through the Oklahoma Cooperative Fish and Wildlife Research Unit (Oklahoma Department of Wildlife Conservation, Oklahoma State University, United States Geological Survey, United States Fish and Wildlife Service, and Wildlife Management Institute cooperating). The project received partial funding from the Oklahoma Agricultural Experiment Station.

Copyright information

© Springer Science+Business Media B.V. (outside the USA)  2011