ABSTRACT
Disturbance regimes, including the historical timing of disturbance, are important components of natural ecosystems and greatly influence ecosystem structure and functioning. Consequently, disturbance timing can be an important component of biodiversity management. We evaluated the effect of prescribed fires ignited during the warm and cool seasons (summer and spring, respectively) on the plant community of a calcareous grassland in northern Mississippi (USA). We found that fire season influenced plant community composition by having differential impacts on species with different life history traits. Differences among species were primarily driven by the dichotomy between cool-season (C3) and warm-season (C4) plants, independent of species native status. Spring burns reduced the cover of cool-season C3 graminoids, but had the opposite effect on C4 graminoids, which likely benefited from increases in resource availability due to the reduction of C3 species. However, summer burns decreased the abundance of C4 graminoids, as summer burns were ignited during the active growing and reproductive period for the C4 species. We found the same patterns for the number of inflorescences of the most abundant C3 and C4 graminoids. Summer burns also increased overall species diversity and the abundance of native C3 graminoids, forbs, and vines, resulting in significant differences in plant community composition between spring- and summer-burned areas. Programs that aim to restore native grassland communities in the short-term using prescribed fire should consider the life history traits of target plants (including invasive species) to determine the best time for prescribed fire implementation.
Programs that aim to restore native grassland communities using prescribed fire should consider the life history traits of target plants to determine when the best time is for prescribed burning to be implemented.
Species that share life history traits but have different biogeographic origins respond similarly to the timing of disturbance.
Prescribed burns ignited within the warm season may be an effective management tool to promote C3 graminoids, forbs and vines, suppress C4 grass species, and increase overall plant species diversity in fire-dependent ecosystems like grasslands.
Managers should combine fire with other management actions (e.g., mechanical removal, chemical control, and alteration of the physical environment) when both undesirable C3 and C4 invasive plants are present within the system.
Fire is widely recognized as one of the most important components of disturbance regimes and acts as a keystone process in multiple ecosystems worldwide (Murphy and Bowman 2012, Dantas et al. 2016). Fire plays an especially important role in grassland ecosystems, where it promotes biodiversity, suppresses woody encroachment, and, in some instances, minimizes negative ecological impacts of biological invasions (Leach and Givnish 1996, D’Antonio 2000, DiTomaso et al. 2006, Alstad et al. 2016).
Fire has been widely implemented as a management tool to manage biological invasions and restore native ecosystems (Knapp et al. 2009). However, one of the challenges managers currently face is deciding when the best time is for implementing prescribed fire. There is evidence that both warm- and cool-season burns can reduce the dominance of invasive plant species in grasslands (e.g., Parsons and Stohlgren 1989, Grace 1998, Meyer and Schiffman 1999, Emery and Gross 2005, Cox and Allen 2008, MacDonald et al. 2007, Howe 2011). Because grassland management outcomes differ among ecosystem types and invader traits, case studies of varying ecosystem types are needed to evaluate which fire season yields the best outcome for reducing impacts of biological invasions. The effectiveness of different burn seasons for invasive species control and restoration of native vegetation has also been attributed to changes in the fuel load and environmental conditions that affect fire intensity and severity (Meyer and Schiffman 1999, Kerns et al. 2006, Knapp et al. 2007, Gorgone-Barbosa et al. 2015, Reemts et al. 2019). Cool-season burns are usually conducted under higher fuel moisture conditions than warm-season burns, which reduces fire severity and fuel consumption (Kauffman and Martin 1989, Knapp et al. 2005). Moreover, the ecological effects of prescribed fire season on the overall plant community composition have been rarely evaluated (e.g., Owens et al. 2002, Céspedes et al. 2014, Tsafrir et al. 2018, Dickson et al. 2019), and there are mixed results about the effectiveness of fire timing in modifying plant community composition.
The phenological or life-history stage of plants at the time of prescribed burns also affects the response of individual species and, consequently, plant community composition, to burning (Kruger and Bigalke 1984, Copeland et al. 2002, Knapp et al. 2009). For example, due to their photosynthetic pathways, the active growing periods of C3 (cool season) and C4 (warm season) plants are spring/ late fall and summer/early fall, respectively (Edwards and Still 2008). Previous research suggests that warm-season prescribed burns that simulate the timing of lightning fires have more negative effects on warm-season C4 grasses than cool-season burns (Howe 1995, Simmons et al. 2007, Howe 2011). The suppression of dominant C4 grasses by warm-season fires can then lead to an increase in the abundance and richness of subdominant species (Copeland et al. 2002). In contrast, spring burns favor C4 grasses (Steuter 1987, Engle and Bidwell 2001, Towne and Craine 2014, Dickson 2019, Dickson et al. 2019) and tend to reduce the survival and growth of cool-season C3 plants (Whisenant and Uresk 1990). C3 plants are generally favored when burning at other times of the year (Engle and Bidwell 2001, Towne and Craine 2014, Dickson 2019, Dickson et al. 2019). Because prescribed fire timing may impact the survival and reproduction of species with variable life history traits, managers may need to consider which species are a priority for restoration when prioritizing prescribed burn plans (Meyer and Schiffman 1999, Valkó et al. 2014). Unfortunately, this could be especially challenging when both C3 and C4 plants are present within the system, due to their varied responses to fire timing.
Fire plays a pivotal role in the maintenance of remnant prairies in the Black Belt region, which extends across calcareous bedrocks of central Alabama and northern Mississippi, USA (Peacock and Schauwecker 2003, Barone 2005). At least 140,000 hectares of prairies surrounded by forest were present in this region in the 1830s. However, since the end of the 19th century, the area comprised by prairie has been dramatically reduced due to agriculture, development, erosion, and incursions of Juniperus virginiana (Hill 2004, Barone 2005). Currently, less than 1% of the historic area of Black Belt prairies remains in isolated scattered patches, making this a highly endangered ecosystem (Barone and Hill 2007). These prairies persist in roadside relicts and areas managed for early successional plant communities (Hill 2004, Barone 2005), which are prone to invasion by exotic species (Lázaro-Lobo and Ervin 2019). Because of the limited area of these ecosystems that remain, there is an urgent need for understanding the effectiveness of different management approaches at maintaining, or enhancing, the diversity of plant species representative of historic prairie communities. Restoration goals for these prairies also include the reduction of invasive plants with varying life history traits in order to provide more resources for extant native plant species to expand.
In this study, we used a field experiment to test the impact of prescribed fire timing on plant communities invaded by both C3 and C4 exotic plants in a remnant prairie of the Black Belt region. We evaluated whether burn season can be used as an invasive species management tool when both C3 and C4 invasive plants are abundant in the community. We categorized each plant species based on its life history traits (growth form and photosynthetic pathway) and biogeographic origin (native status). Based on previous work in grassland ecosystems, we hypothesized that burn season would affect plant community composition by having differential impacts on species with different life history traits. In this sense, we predicted that spring burns would suppress cool-season C3 species, whereas summer burns would have detrimental effects on warm-season C4 species, due to their respective active growing and reproductive periods. We also predicted that exotic and native species sharing life history traits would have similar responses to burn season, because ecological requirements and phenological patterns would be more important than biogeographic origin to determine species responses to fire timing. Lastly, we predicted that overall plant species diversity would be promoted by warm-season (summer) burns, because dominant warm-season plants would be controlled by fire disturbance during that season.
Materials and Methods
Study Area
The study was conducted in the Black Prairie Wildlife Management Area (33°34′ N, 88°58′ W; Figure 1), which covers an area of ~2,400 ha and hosts some of the remnant prairies in the Black Belt region. This prairie has historically been burned by managers on an arbitrary schedule during the warm and cool seasons (summer and spring/ fall, respectively), depending on the climatic conditions of the respective year and personnel availability. Long-term records of fire history were unavailable, but each part of the prairie where the present study was conducted has been burned with a return interval of 1–3 years during the past decade.
A. Study area. The shaded area corresponds to the counties within Mississippi (MS) and Alabama (AL), USA that comprised the Black Belt region during the 1830s (Barone 2005). The black dot corresponds to the Black Prairie Wildlife Management Area (33°34’ N, 88°58’ W) where this study took place. B. Block (90x70 m) formed by three randomly assigned plots. C. Plot (70x30 m). Each plot corresponded to a different treatment (spring burn, summer burn, and unburned). D. 1-m2 quadrat. Each plot was sampled using ten 1-m2 quadrats separated by at least five meters.
Experimental Design and Data Collection
We used a randomized complete block experimental design to conduct our study. Each block had different fire history, with varying fire frequency and burning seasons during the last ten years. We established eight 90 × 70 m blocks and randomly assigned three 30 × 70 m plots within each of them. All plots were placed perpendicular to a large trail to avoid disproportionate path edge effects on the experiment. Mowed fire lanes (5-m wide) were established around each plot to avoid the spread of fire to nearby vegetation and other plots. Each plot corresponded to a different treatment (spring burn, summer burn, and unburned; Figure 1). Spring burns were conducted in March and April 2018 and summer burns were conducted at the end of July 2018. Ambient temperatures during spring and summer burns ranged from 13–20 and 26–33°C, respectively.
We surveyed vegetation at different time periods (mid-late May, early September, and mid-November) during the year after treatment application (2019) to account for phenological variation of the plant community throughout the year. At each sampling time, we recorded the identity and percent cover of the plant species rooted within 1-m2 quadrats. In each plot, we placed ten 1-m2 quadrats separated by at least five meters and distributed along a 50-m longitudinal transect placed in the middle of the plot. Most of the quadrats were densely vegetated and reached percent cover values higher than 100% due to overlapping vegetation. We also counted the number of inflorescences within the 1-m2 quadrats for the two most abundant species in the study area, Schedonorus arundinaceus (an exotic C3 grass) and Sorghum halepense (an exotic C4 grass).
Statistical Analyses
All statistical analyses were carried out using R software (version 3.6.1; R Core Team 2020). We calculated the mean cover of each plant species and the mean number of S. arundinaceus and S. halepense inflorescences across quadrats within the same plot for the analyses.
The effect of burn treatment (spring burn, summer burn, and unburned) on plant community composition was evaluated with ordination methods, including non-metric multidimensional scaling (NMDS), permutational multivariate analysis of variance (PERMANOVA), and distance-based tests of homogeneity of multivariate dispersions (PERMDISP beta dispersion). We used the vegan package to carry out ordination methods (Oksanen et al. 2019). We used NMDS to visualize the response of the plant community composition to burn treatments. We adjusted the data using relativizations because NMDS is based on ranked distances and has no built-in relativization (McCune et al. 2002). We evaluated the appropriate number of ordination axes to balance the interpretability of the ordination versus reductions in stress (McCune et al. 2002). We selected three axes to conduct the NDMS analyses and used Bray-Curtis measure of dissimilarity, which is a robust index to evaluate distances between samples in ordination space (Faith et al. 1987). We then conducted permutational multivariate analysis of variance (PERMANOVA) with 1,000 permutations and the Bray-Curtis distance matrix to evaluate whether there were statistically significant differences in plant community composition among the different burn treatments. PERMANOVA is a non-parametric method based on permutation tests that analyzes the variance using distance matrices, which makes it appropriate to analyze multivariate data (Anderson 2001, Oksanen et al. 2019). Lastly, we used a dissimilarity-based multivariate analogue of Levene’s test to examine homogeneity of multivariate dispersions (beta dispersion) among burn treatments, which calculates the variance (internal spread) of each burn treatment and compares it among different burn treatments in ordination space (Anderson 2006, Anderson et al. 2006). We used Bray-Curtis dissimilarity to calculate distances between samples in ordination space and conducted 1,000 permutation-based tests to assess statistical differences among burn treatments.
The effect of burn treatment on percent cover of species grouped by their life history traits (growth form and photosynthetic pathway) and biogeographic origin (native status), as well as on diversity metrics, was evaluated with generalized linear mixed models (GLMMs) with block as a random factor. We categorized plant species based on their native status (native vs. exotic) and their growth forms (graminoids or grass-like flowering plants belonging to the families Poaceae, Cyperaceae, and Juncaceae; forbs and vines; and trees and shrubs; based on USDA [2020]). Graminoids were also categorized based on their photosynthetic pathway (cool-season C3 vs. warm-season C4 plants; all other species were C3). We calculated three measures of diversity based on the number of species and mean cover values obtained in the surveys (species richness, exponential of the Shannon-Wiener index, and inverse Simpson index; Oksanen et al. 2019). Species richness refers to the number of species present, whereas the other two indices consider both the number of species present and their relative abundances; all three indices provide estimates of the effective number of species in a community (Jost 2006, Gotelli and Chao 2013). The exponential of the Shannon-Wiener index is more influenced by richness than evenness, whereas the inverse Simpson index puts more weight on the evenness than on richness.
We used the “glmmTMB” package (Magnusson et al. 2017) using the appropriate data distribution for each response variable to build the GLMMs (Zuur et al. 2009). Percent plant cover was analyzed with beta distribution, which is widely used for cover proportions (Bolker 2008). Inflorescence number and richness were evaluated with negative binomial distribution, which controls for overdispersion. The other diversity indices (exponential of Shannon-Wiener index and inverse Simpson index) were analyzed with gaussian distributions because their model residuals met the assumptions of linear regressions (normality and homoscedasticity). When significant differences (p < 0.05) existed among burn treatments, we conducted post-hoc comparisons using estimated marginal means using the “emmeans” R package (Lenth et al. 2020) with Bonferroni’s adjustment method, which is appropriate for multiple comparison tests (Cabin and Mitchell 2000). We conducted GLMMs for each sampling time independently (May, September, and November), because prairie species composition (by biomass/percent cover) in our study area varies widely throughout the year, with different species dominating during different seasons.
Results
A total of 83 plant species with a mean total percent cover of 163% within each quadrat were recorded over the sampling period (Supplemental Material, Table S1). There were 63 forb/herb/vine species, 13 graminoid species (seven warm-season C4 and six cool-season C3), four shrub species, and three tree species. Forb/herb/vine, graminoid, and shrub/tree species had mean percent covers of 80, 66, and 13%, respectively. Fifteen species were exotic (two C4 graminoids, three C3 graminoids, nine forb/herb/vines, and one shrub) and their total cover averaged 53% across quadrats. The most common exotic species were the grasses Sorghum halepense and Schedonorus arundinaceus, followed by Daucus carota, Trifolium pratense, Medicago minima, Plantago lanceolata, and Bromus catharticus (all of which are C3 plants, expect for the C4 grass S. halepense; Supplemental Material, Table S1). The most common native species were Solidago canadensis, Andropogon virginicus, Rubus argutus, Desmanthus illinoensis, and Toxicodendron radicans (all of which are C3 plants, expect for the C4 grass A. virginicus; Supplemental Material, Table S1).
We found distinct plant community composition in plots with different burn treatments (Figure 2; PERMANOVA, p < 0.001). Across all sampling periods, spring and summer burned plots were largely separate from each other in ordination space, and unburned plots were somewhere in between. However, beta dispersion analyses indicated that there were not statistical differences (p > 0.05) in the homogeneity (internal spread) of plant community composition among burn treatments.
Non-metric multidimensional scaling (NMDS) ordination of sampling units (plots), based on Bray-Curtis dissimilarities of percent plant cover at each sampling time.
Burn treatments had significant effects on the percent cover of all growth form and photosynthetic pathway categories during the year following fire, and native and exotic species from the same photosynthetic pathway were similarly affected by burn treatments. Spring burns significantly decreased the percent cover of exotic C3 graminoids across all sampling periods relative to unburned and summer-burned plots (p ≤ 0.001; Figure 3; Supplemental Material, Table S2). Specifically, spring burns during the May, September, and November sampling periods had approximately 14, 9, and 8% less cover of exotic C3 graminoids than the other treatments (unburned and summerburns). Native C3 graminoids were much less abundant in the study area than exotic C3 graminoids, having an average percent cover of 1.5% across all sampling periods. However, as with exotic C3 graminoids, native C3 graminoids had lower cover following spring burns.
Mean response of plants categorized by their native status and growth forms to the different burn treatments at each sampling time. Graminoids were further divided into two categories based on their photosynthetic pathway (C3 vs. C4). Error bars represent standard errors (SE). Note variation in Y-axes among rows.
Burn treatments had a similar effect on native and exotic C4 graminoids across all sampling periods. Spring burns significantly increased the percent cover of native and exotic C4 graminoids across all sampling periods relative to unburned and summer-burned plots (p ≤ 0.001; Figure 3). Specifically, spring-burned plots during the May, September, and November sampling periods had approximately 15, 26, and 24% higher cover of native and exotic C4 graminoids than summer-burned plots, and approximately 10, 18, and 17% higher cover of native and exotic C4 graminoids than unburned plots. The percent cover of C3 graminoids was significantly different (p < 0.001) in different sampling periods in all treatments, and the same applied for C4 graminoids. Specifically, the mean percent cover of C3 graminoids across treatments was 14.2, 7.8, and 7.9% in May, September, and November, respectively, whereas the mean percent cover of C4 graminoids across treatments was 16, 28, and 25.5% in May, September, and November, respectively.
We also found differences in the number of inflorescences of the two most abundant exotic grass species between spring and summer burn treatments. Spring burns decreased the number of inflorescences of the invasive exotic C3 grass Schedonorus arundinaceus during its reproductive period (spring; p = 0.043; Supplemental Material, Table S2). On average, there were 1.7 fewer S. arundinaceus inflorescences in spring-burned plots than in summer-burned plots. However, summer burns decreased the number of inflorescences of the invasive exotic C4 grass Sorghum halepense during its reproductive period (summer; p = 0.044). There were 2.4 fewer S. halepense inflorescences after summer burns than after spring burns.
Summer-burned plots had 15% higher cover of native forbs and vines than spring-burned plots (p = 0.026) and 35% higher cover than unburned plots (p < 0.001) in November. There was no significant difference in the cover of native forbs and vines between unburned and spring-burned plots (p > 0.05). Similarly, the percent cover of exotic forbs and vines was ~4% higher in summer-burned plots relative to spring-burned and unburned plots in September (p ≤ 0.03). Furthermore, spring burns had 6 and 5% lower cover of native trees and shrubs than unburned and summer-burned plots during the September sampling period (p < 0.05; Figure 3). We only found one species of exotic tree/shrub (Ligustrum sinense), which was only present in one spring-burned plot.
The effective number of species (exponential Shannon and inverse Simpson diversity indexes) tended to be higher in summer-burned plots relative to spring-burned and unburned plots (Figure 4; Supplemental Material, Table S2). Specifically, the exponential Shannon index was 12.6 and 13.6% higher after summer burns than in spring-burned and unburned plots, respectively, during the November sampling period (p ≤ 0.015). Similarly, the inverse Simpson index was 18.5% higher after summer burns than spring burns during the September sampling period (p = 0.028), and 17.7 and 15.4% higher after summer burns than in spring-burned and unburned plots, respectively, in November (p ≤ 0.031). Also, we did not find significant differences in species richness among treatments.
Mean response of diversity metrics (species richness, exponential of Shannon-Wiener index, and inverse Simpson index) to the different burn treatments at each sampling time. The three diversity metrics are in units of species richness at the 1-m2 quadrat level. Error bars represent standard errors (SE).
Discussion
Our results suggest that, as in other grassland ecosystems, the timing of fire disturbance is an important factor affecting post-fire plant community composition during the year following fire. The response of plant species to fire season was influenced by their life history traits, as we initially hypothesized. We found that plants that shared similar life history traits responded similarly to the timing of disturbance, regardless of their biogeographic origin (i.e., native vs. exotic species). Our results also suggest that prescribed fires timed to occur during the active growing period of the most dominant species can promote the growth of subdominant species and, thus, increase species diversity. However, we found that burn season alone has limited use as an invasive species management tool when both C3 and C4 invasive plants are abundant in the community. Thus, managers should combine fire with other management actions (e.g., mechanical removal, chemical control, and alteration of the physical environment) when a system is invaded by species with different ecological requirements and phenological patterns.
In our study, life history traits of plant species, irrespective of their biogeographic origin, influenced plant species’ responses to fire season during the year following fire. Previous research suggests that phenological variation throughout the year (e.g., growing and reproductive period) could affect the response of individual plant species to timing of disturbance, due to species’ physiological processes (Meyer and Schiffman 1999, Knapp et al. 2009, Ruckman et al. 2012). Similarly, we found that the photosynthetic pathway (C3 and C4) of both native and exotic graminoids influenced their response to burn season. Spring burns, which occur during the growing period of C3 graminoids, significantly decreased the abundance and number of inflorescences of C3 graminoids during the year after burning. Whisenant and Uresk (1990) also found that spring burns decreased the standing crop of most of the C3 graminoids evaluated in their experiment. In contrast, spring burns increased the abundance and number of inflorescences of warm-season C4 graminoids during the year following fire. This could be explained by the deleterious effect of spring burns on C3 graminoids, which probably increased resource availability (e.g., light and nutrients) during the active growing period of C4 plant species (summer and early fall). Other studies conducted in grasslands also showed that spring fire favored C4 grasses in the short and long term (Steuter 1987, Engle and Bidwell 2001, MacDonald et al. 2007, Howe 2011, Towne and Craine 2014, Dickson 2019, Dickson et al. 2019). Similarly, as predicted, summer burns negatively affected the abundance and number of inflorescences of C4 graminoids, matching findings from previous research (Howe 1995, Simmons et al. 2007, Howe 2011). Furthermore, in our study, the abundance of exotic C3 graminoids was similar in summer-burned plots and unburned control plots, which suggests that summer burns have no effect on exotic C3 graminoids in the short term. In contrast, summer burns increased the abundance of native C3 graminoids relative to unburned plots, although the magnitude of the effect was very small. This finding suggests that summer burns could promote native C3 graminoids such as Carex spp., Elymus spp., and Sphenopholis spp. (which were rare in our study area) during the year following fire and fill an empty niche that would otherwise be occupied by the more abundant exotic C3 graminoids. The above-mentioned findings indicate that prescribed fires that are ignited during the active growing and reproductive period of C3 or C4 graminoid plant species generally decrease their abundance and population growth during the year following fire (Howe 1995, Simmons et al. 2007, Howe 2011, Reemts et al. 2019). Therefore, managers should consider species-specific responses to disturbance timing when managing and restoring prairie ecosystems. Prescribed fires should be conducted when undesirable species are most vulnerable and desirable species are less susceptible to the negative effects of fire (Meyer and Schiffman 1999, Valkó et al. 2014). However, management plans should also consider the effect of burn season on species diversity.
In our study, the abundance of both native and exotic forbs and vines (all C3 plants) during the year following fire was favored by summer burns. However, native forbs had a stronger positive response to summer burns than exotic forbs. Previous research conducted in central Texas and Kansas also suggests that summer burns benefit forbs in the short and long term (Towne and Kemp 2008, Reemts et al. 2019). On the other hand, Biondini et al. (1989) concluded that forb density was highest after spring and fall burns in South Dakota. This mixed evidence for fire timing effects on forbs could be explained by the inclusion of species with different life-history strategies and phenology within the same category or because forbs are primarily competing against C4 grasses in Texas and Kansas and C3 grasses in South Dakota. In our study, the reduction of dominant warm-season C4 graminoids following summer burns (especially Sorghum halepense [exotic] and Andropogon virginicus [native]) might have reduced competition during the year following fire thereby favoring higher abundance of forbs and increasing species diversity. The two diversity indices affected by fire season put different weight on richness and evenness, indicating that fire season affects both the number of species present and their relative abundances. Consequently, prescribed burns that are ignited in summer could stimulate the growth of subdominant species and allow the coexistence of a wider range of plant species (Copeland et al. 2002). Studies conducted in other regions of the United States also showed that summer burns increase species richness and/or diversity in the short and long term (e.g., Biondini et al. 1989, Copeland et al. 2002, Towne and Kemp 2008).
The implementation of seasonal burning as a restoration tool in Black Belt prairies would depend on the restoration goals. We found that prescribed fires ignited during the warm season (summer) in this ecosystem appear to be an effective management tool to suppress dominant C4 grass species and increase overall plant species diversity. However, our findings suggest that fire season alone cannot control invasive plants with different life history traits during the year following fire. As mentioned above, we recommend that managers combine fire with other management actions (e.g., mechanical removal, chemical control, and alteration of the physical environment) when both undesirable C3 and C4 plants are present within the system. Our findings also indicate that fires ignited at different seasons can shift plant community composition in this ecosystem during the year following fire. However, implementing spring and summer burns to decrease the abundance of C3 and C4 invasive plants would depend on the prevalence of such species.
There are multiple limitations of our study that could affect the development of management recommendations and the achievement of restoration goals. First, the vegetation was only sampled during the year after burning. Although burning in remnant Black Belt prairies is fairly frequent (sometimes every year), the short-term nature of this study could limit the scope of our findings. Further studies are needed to evaluate long-term responses of the plant community to burning season in this dynamic ecosystem. Moreover, burns could be implemented in multiple years to assess compound effects of burning on the plant community and invasive species, as well as the effect of interannual variability in weather conditions. Another limitation is that our experiment was conducted at a single location. Although our findings could be extrapolated to other remnant Black Belt prairies in the southeastern United States, because they share similar ecological, topographic, and geological characteristics (Peacock and Schauwecker 2003, Barone 2005), future studies could increase the spatial scale and establish experimental plots within multiple remnant Black Belt prairies.
Broom-sedge (Andropogon virginicus). a, spike; b, spikelet; c, glumes. Source: Whitney, W.D. 1911. The Century Dictionary: An Encyclopedic Lexicon of the English Language (New York, NY: The Century Co.), The Florida Center for Instructional Technology, College of Education, University of South Florida, fcit.usf.edu.
Johnson-grass (Sorghum halpense). Spikelets in a panicle. Source: Gaber, C.S. 1916. Fundamentals of Botany (Philadelphia, PA: P. Blakiston),The Florida Center for Instructional Technology, College of Education, University of South Florida, fcit.usf.edu.
Acknowledgments
The authors thank the managers at the Black Prairie Wildlife Management Area, specifically Chad Masley and John Gruchy for their help finding adequate areas to establish experimental plots and logistical support in conducting prescribed burns. This research was supported by USDA Forest Service grant 19-DG11083150-006 and Biology Student Research Award Spring 2018 of Mississippi State University.
Footnotes
Color version of this article is available through online subscription at: http://er.uwpress.org
Supplementary materials are freely available online at: http://uwpress.wisc.edu/journals/journals/er-supplementary.html
Author Contributions
A.L-L., M.A.L, and G.N.E conceived the research idea; A.L-L. collected data; A.L.L. and A.K.P. performed statistical analyses; A.L-L. wrote the initial draft of the manuscript; all authors discussed the results and commented on the manuscript.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.