Photo: Overgrazed sagebrush habitat, southwest Montana. © 2018 Delena Norris-Tull
Effectiveness of herbicides in rangelands
Summaries of the research and commentary by Dr. Delena Norris-Tull, Professor Emerita of Science Education, University of Montana Western, October 2020.
In an examination of research on the use of herbicides in rangelands, Sheley, et al., 2011, concluded that control of undesirable vegetation is rarely cost-effective, due to the low revenue value of rangeland. The use of herbicides, fire, and other typical practices often result in “rapid reinvasion.” Sheley, et al., recommend frequent use of grazing, “or infrequent use of more expensive strategies that provide longer-term control, such as restoration.”
Prescribed burns have had mixed results, sometimes resulting in increased infestations. Herbicides are effective “for preventing small… infestations from producing seeds and spreading. They are also effective for controlling weeds during restoration projects…. Controlling rangeland invasive plants rarely increases forage production enough to offset the herbicide costs…. Invasive annual grasses [in particular, cheatgrass] often proliferate after herbicides kill associated invasive forbs…. Large-scale herbicide treatments… often kill associated native forbs and shrubs,” which are less likely to recover than are the invasive species.
Sheley, et al., 2011, point out that rangeland herbicides are selective herbicides, typically killing either grasses or forbs, but not both. “Many native grasses increase following herbicide control of invasive forbs… Similarly, native forbs often increase after herbicides kill invasive grasses.” However, “herbicides are sometimes used to control invasive grasses even though herbicide-sensitive native grasses are present. Likewise, herbicides are used to control invasive forbs growing with native forbs and shrubs.” There are several cases wherein the native species were little affected and were able to recover. But there are other cases where herbicides caused extensive damage to native species. Some herbicides can have serious long-term damaging effects on native species. Sheley, et al., concluded that herbicides are expensive and often have only short-term effects on noxious weed control in rangelands.
Sheley, et al., 2011, describe research that suggests that atrazine has played a role in “global declines in amphibian populations.” They also found that “chemicals that target dicots can decrease plant community diversity, thereby reducing the food available for some wildlife species… For example, 2,4-D applied to western Colorado rangeland to favor grasses over forbs and shrubs reduced densities of northern pocket gophers… and least chipmunks…, while increasing densities of montane voles.”
Sheley, et al., 2011, found that research on current methods of invasive plant species control suggests that herbicides, tillage, and grazing can create increased disturbances, that increase invasions. They found that little research has been conducted on herbicide impacts on abiotic resources and native species.
Sheley, et al., found that “the impact of… herbicides on [soil, water, and air] resources is dependent on type of herbicide used, application rate, and soil characteristics, among other factors…. Glyphosate tightly adheres to soil,” thus reducing the risk of the chemical leaching into groundwater or affecting soil biota. But “dicamba and picloram are highly mobile in the soil… High application rates, high rainfall following application, or direct application… to water bodies can pose a significant threat to water resources.” But little research has been conducted on the impacts of herbicides on water and soil resources and wildlife. And few studies have compared impacts of herbicides on native versus invasive plant species. “Herbicides such as 2,4-D, clopyralid, or picloram…[used to control broadleaf weeds] can greatly decrease native forb density and cover… There is evidence suggesting that herbicide effects on native forbs are long-lasting and can drive a local decline in species richness… Desirable rangeland grasses have shown varying degrees of susceptibility to imazapic, a herbicide used to control invasive annual grasses, with evidence suggesting grasses within the Hordeae tribe may be more tolerant to imazapic than other grass species.” They concluded that, “The effect of herbicide on desirable vegetation remains difficult to predict.”
Reduction of invasive annual grasses near Cape Town, South Africa
In a study on the effectiveness of treatments used to reduce invasive grass species expansion within a South African wildflower reserve, Musil, et al., 2005, compared various treatments within 10,000 square meter parallel transects: within the middle transect, hand-removal of invasive grasses in the fall (the grass was then spread in an adjacent undisturbed transect); grass brush cut (mowed) to a height of about 5cm in alternate transects (with the grass spread in adjacent uncut transects). Following these two treatments, a 20-meter-wide fire-break was brush cut around the reserve’s perimeter. Then all vegetation in the reserve, except within two mowed transects, was given either intense or light burning. Forty-eight 100 square meter plots were randomized within the transects. This resulted in four primary treatments: hand-removal plus light burning; hand-removal plus intense burning; brush cut plus light burning; brush cut plus intense burning. These primary treatments were then divided into two secondary treatments: application of pre-emergent herbicide (Snapshot: trifluralin and isoxaban); or no herbicide.
The reserve is broadly categorized as “evergreen fire-prone vegetation.” In the two year study, within which the first year was significantly drier than the second year, abundances were measured of native and alien grasses, and native herbaceous and woody species. Species densities were calculated from counts of the average number of individuals present in sub-plots. In the case of mature perennial grasses, abundances were calculated from the average percentage covers of each species.
Musil, et al., 2005, results:
The obligate seeding native forbs had significantly higher densities (5 times greater) in the wetter second year, and their densities were significantly lower (1.9 times less) in plots that received pre-emergent herbicide, compared with the plots with no herbicide. This group’s densities were significantly altered in the mechanically manipulated plots. “General trends of increasing forb densities with increasing intensity of disturbance in the sequence mowed < lightly burnt < intensely burnt < hand-cleared were evident, both in those plots where the pre-emergent herbicide was added and in those where the herbicide was excluded, particularly in the wetter second observation year.”
For the obligate seeding, mainly invasive, annual grasses: total densities did not differ between drier and wetter years. Their densities differed significantly in the differently mechanically manipulated plots, “with measured densities averaging between 4.9 and 5.5 times higher in the lightly burnt, intensely burnt, and hand-cleared plots than in the mowed plots. Also, densities of this functional group were significantly… reduced in those plots where the pre-emergent herbicide was added.” There were significant interactions between some treatments. “The first interaction indicated that reductions in annual grass densities following addition of the pre-emergent herbicide… were larger in the lightly burnt plots (6.4-fold decrease) and mowed plots (4.0-fold decrease) than in the intensely burnt plots (2.7-fold decrease) and hand-cleared plots (1.4-fold decrease). The second interaction specified greater herbicide-induced reductions in total densities of annual grasses… in the drier first observation year (6.6-fold decrease) than the wetter second (1.7-fold decrease).”
For the sprouting native geophytes (plants with underground bulbs, corms, rhizomes, or buds), “total densities were not significantly (P ≥ 0.05) affected by the addition of the pre-emergent herbicide… However, densities of this functional group were significantly (P ≤0.001) higher (3.4 times greater) in the wetter second year than the drier first. Also, they were significantly (P ≤ 0.05) altered in the differently mechanically manipulated plots with a significant interaction (P ≤ 0.05) evident between mechanical treatment and observation year... This interaction indicated larger increases in geophyte densities between the drier first and wetter second observation year… in the mowed plots (10.5-fold increase) than in the hand-cleared plots (7.9-fold increase), intensely burnt plots (3.9-fold increase) and lightly burnt plots (1.6-fold increase).”
For the sprouting perennial grasses, “The relative densities of the native and alien perennial grasses in this functional group were not significantly… affected by the higher annual precipitation in the second observation year…, by any of the mechanical manipulations applied…, or by the addition of the pre-emergent herbicide... Total covers of perennial grasses were significantly… lower (2.3 times smaller) in the wetter second than in drier first observation year… but were not significantly… affected by the addition of the pre-emergent herbicide... However, covers of this functional group did differ significantly … between the differently mechanically manipulated plots, the measured amounts in the lightly burnt plots being consistently higher (up to 3.8 times more) than those in the mowed, hand-cleared and intensely burnt plots.”
The team also compared the costs of the different treatment methods. For managing invasive annual grasses, intense burning of uncut vegetation was the cheapest strategy. The limitation of intense burning was that it inhibited recruitment of native geophytes. But the slightly more expensive light burning did stimulate the growth of geophytes. “Control of invasive annual grasses by burning may be effective only if all adult plants and their vegetative propagating organs are destroyed and their seed banks eliminated.”
Mowing, slightly less expensive than light burning, “elicited responses of the smallest magnitude from the four designated functional groups” in the first year. But mowing resulted in increased growth of native forbs and geophytes in the second year. In contrast, “covers of creeping perennial grasses in the mowed plots were reduced in the second observation year… The cover of perennial grasses in this study was negatively correlated with the densities of annual forbs and geophytes, but not with those of the annual grasses.”
Use of the herbicide was the most expensive but most effective method for controlling invasive annual grasses in the first year, and in the lightly burned plots. It was less effective in the second year. “This treatment’s disadvantage was that it … inhibited recruitment of native species of forbs, though it did not detrimentally influence springtime densities of perennial grasses or indigenous geophytes. This was expected as such herbicides kill germinating seeds rather than established sprouting plants. Another short-term disadvantage of herbicide application is that inhibition of annual grass populations does not necessarily result in greatly decreased seed production.” The potential emergence of rapid herbicide resistance may make such treatments more costly over time.
Conclusions by Musil, et al., 2005: The most credible strategy for managing invasive annual grasses in this location appears to be mowing “prior to grass seed maturation, and removal of the cut grass biomass for use as fodder in restricted feed lots.” But in sites with heavy, long-term invasive annual grass infestations, “an initial light intensity autumn burn to remove excessive grass litter, and to promote flowering, seed production and smoke-stimulated germination of native taxa, followed by the addition of a pre-emergent herbicide…, may assist in partly recovering native floral diversity.”
Comparison of treatments for quail grassland habitat improvement
Madison, et al., 2001, conducted a field experiment to compare the effectiveness of a variety of treatments in reducing the density of tall fescue populations to potentially encourage the reestablishment of northern bobwhite populations in Kentucky grasslands. Tall fescue was imported to the US as a forage grass. It is now considered invasive. Tall fescue provides poor habitat for bobwhite feeding, nesting, and roosting. Other grasses and forbs provide better feeding habitat. Disturbance of a field densely populated with tall fescue can improve habitat for bobwhite. The study was conducted “on 4 Kentucky Department of Fish and Wildlife Resources …Wildlife Management Areas (WMA) representing different physiographic regions and bobwhite densities.” Each site was divided into 16 0.1-ha plots. Each plot was randomly assigned one of eight treatments: fall burning, fall disking (with a harrow), spring burning, spring disking, spring herbicide application, summer burning, summer disking, and a control. They implemented the treatments in fall 1991 and spring and summer 1992. Herbicide plots were mowed, allowed to regrow for two weeks, and then sprayed with RoundUp.
Madison, et al., 2001, conducted observations of the plot using 1.0 square meter quadrats placed randomly in each plot. They estimated “percent of vegetation canopy coverage, bare ground, and canopy coverage by individual plant species.” They visually estimated the first two characteristics. And they calculated percentage of canopy cover of individual plants by species, which included both native and non-native grass and forb species. They calculated the percentage of plants in each plot that represented preferred bobwhite foods. And they estimated plant height. They “sampled the vegetation once each season, one year prior to treatments” (four sample periods in 1990-1991). And they “measured the vegetation once each season for 2 years post-treatment, beginning in fall 1991 and ending summer 1994” (12 sampling periods). They collected seed in winter 1993, and air-dried it and weighed it.
Madison, et al., 2001, results:
“Tall fescue coverage did not differ significantly among [Wildlife Management Areas]…, and there was no WMA by treatment interaction through time.” This enabled them to pool data across WMAs. They “compared fescue coverage among treatments at 1 year pre-, 1 month post-, 1 year post-, and 2 years post-treatment.” And they compared fescue coverage within each treatment plot across time.
“Fall disking was the best treatment for potentially improving bobwhite winter feeding habitat during winter 1993. Canopy coverage of preferred bobwhite foods in summer 1992 was 45.9 ± 13.0%... Availability of bare ground also contributes to the quality of winter feeding habitat... Fall-disked plots provided 48.5 ± 8.2% bare ground during winter 1993.” Bare ground is necessary to enable bobwhite to forage for seeds, but bare ground > 60% increases their vulnerability to predation. The optimal bare ground coverage is estimated to be 30-60%.
“Spring-disked and herbicide-treated plots may have provided good to moderately good winter feeding habitat during winter 1993. Spring- disked plots averaged 45.1 ± 11.6% food plant canopy coverage during summer 1992 and produced greater seed biomass…between summer 1992 and winter 1993 than control plots; however, percentage of bare ground was >60%,” at the upper limit for the safety of bobwhite populations.
“Herbicide-treated plots averaged 59.2 ± 11.8% food plant canopy coverage in summer 1992 and produced greater seed biomass… from summer 1992 to winter 1993 than all other treatment plots, but percentage of bare ground was >80%.”
“Herbicide-treated plots during summer 1993 were the only treatment plots that satisfied most bobwhite nesting habitat requirements. Grass composition was 48.9 ± 13.0% and vegetation height was 46.9 ± 8.2 cm on herbicide-treated plots during summer 1993.” Optimal bobwhite nesting habitat is estimated to contain “vegetation cover comprised of at least 40-60% grasses approximately 40-60cm tall,” with 30-60% bare ground.
“The most effective treatment to reduce tall fescue canopy coverage was the spring herbicide application. Fescue canopy coverage 1 month post- (x=1 7 ± 1.7%), 1 year post- (x=9.8 ± 7.9%), and 2 years post- (x=32.5 ± 6.5%) treatment on herbicide treated plots was less than control plots (x=82.4 - 92.1%) and pre-treatment conditions (89.6 ± 7.8%).” However, the rapid decline in effectiveness of the herbicide treatment over the two years post-treatment would likely mean that additional herbicide treatments would be needed, which increases the risk for the development of weed species developing herbicide-resistance.
“Disking and prescribed burning did not reduce tall fescue canopy coverage for > 1 year. Fescue canopy coverage was reduced by 41 to 81% on burned plots and by 75 to 88% on disked plots 1 month post-treatment… By 1 year post-treatment, only fall (x = 46.9 ± 9.9%)-, spring (x=25.9 ± 10.9%)-, and summer (x=56.5 ± 10.4%)-disked plots had fescue canopy coverage less than control plots (x=92.1-96.9%) and pre-treatment conditions (x=83.9-95.4%).” By two years after treatment, fescue canopy coverage on disked and burned plots did not differ from control plots and pre-treatment. “Herbicide-treated plots were the only ones that provided good bobwhite feeding habitat in winter 1994. Canopy coverage of preferred bobwhite food plants during summer 1993 averaged 42.4 + 11.5% and percentage of bare ground in winter 1994 averaged 60.6 + 9.2% on herbicide-treated plots.”
“No treatment plots satisfied more than 2 bobwhite nesting habitat requirements [by the completion of the study] in summer 1994.” Schroder, 1985, upon whose research the optimal habitat requirements were based, stated that “low values of any one variable may be partially offset by higher values of the remaining variables.”
Madison, et al., 2001, concluded that, “The effect of these treatments on bobwhite habitat should be viewed cautiously. Bobwhite habitat requirements are well documented…, but implementing these treatments on a small scale only provides potential improvement in bobwhite habitat quality.” The fact that no treatment provided satisfactory results two years after treatment emphasizes the challenges of bobwhite habitat restoration. By two years after treatment, any improvements gained by disking or burning were gone, and benefits of herbicide treatment were rapidly diminishing. Madison, et al., 2001 compared treatments that were used alone. Recommendations for invasive species control advocate using multiple treatments and multiple herbicides sequentially. Refer to the description of Musil, et al., 2005, above.
References:
Links to additional information on herbicides and other pesticides:
Effectiveness of herbicides in rangelands
Summaries of the research and commentary by Dr. Delena Norris-Tull, Professor Emerita of Science Education, University of Montana Western, October 2020.
In an examination of research on the use of herbicides in rangelands, Sheley, et al., 2011, concluded that control of undesirable vegetation is rarely cost-effective, due to the low revenue value of rangeland. The use of herbicides, fire, and other typical practices often result in “rapid reinvasion.” Sheley, et al., recommend frequent use of grazing, “or infrequent use of more expensive strategies that provide longer-term control, such as restoration.”
Prescribed burns have had mixed results, sometimes resulting in increased infestations. Herbicides are effective “for preventing small… infestations from producing seeds and spreading. They are also effective for controlling weeds during restoration projects…. Controlling rangeland invasive plants rarely increases forage production enough to offset the herbicide costs…. Invasive annual grasses [in particular, cheatgrass] often proliferate after herbicides kill associated invasive forbs…. Large-scale herbicide treatments… often kill associated native forbs and shrubs,” which are less likely to recover than are the invasive species.
Sheley, et al., 2011, point out that rangeland herbicides are selective herbicides, typically killing either grasses or forbs, but not both. “Many native grasses increase following herbicide control of invasive forbs… Similarly, native forbs often increase after herbicides kill invasive grasses.” However, “herbicides are sometimes used to control invasive grasses even though herbicide-sensitive native grasses are present. Likewise, herbicides are used to control invasive forbs growing with native forbs and shrubs.” There are several cases wherein the native species were little affected and were able to recover. But there are other cases where herbicides caused extensive damage to native species. Some herbicides can have serious long-term damaging effects on native species. Sheley, et al., concluded that herbicides are expensive and often have only short-term effects on noxious weed control in rangelands.
Sheley, et al., 2011, describe research that suggests that atrazine has played a role in “global declines in amphibian populations.” They also found that “chemicals that target dicots can decrease plant community diversity, thereby reducing the food available for some wildlife species… For example, 2,4-D applied to western Colorado rangeland to favor grasses over forbs and shrubs reduced densities of northern pocket gophers… and least chipmunks…, while increasing densities of montane voles.”
Sheley, et al., 2011, found that research on current methods of invasive plant species control suggests that herbicides, tillage, and grazing can create increased disturbances, that increase invasions. They found that little research has been conducted on herbicide impacts on abiotic resources and native species.
Sheley, et al., found that “the impact of… herbicides on [soil, water, and air] resources is dependent on type of herbicide used, application rate, and soil characteristics, among other factors…. Glyphosate tightly adheres to soil,” thus reducing the risk of the chemical leaching into groundwater or affecting soil biota. But “dicamba and picloram are highly mobile in the soil… High application rates, high rainfall following application, or direct application… to water bodies can pose a significant threat to water resources.” But little research has been conducted on the impacts of herbicides on water and soil resources and wildlife. And few studies have compared impacts of herbicides on native versus invasive plant species. “Herbicides such as 2,4-D, clopyralid, or picloram…[used to control broadleaf weeds] can greatly decrease native forb density and cover… There is evidence suggesting that herbicide effects on native forbs are long-lasting and can drive a local decline in species richness… Desirable rangeland grasses have shown varying degrees of susceptibility to imazapic, a herbicide used to control invasive annual grasses, with evidence suggesting grasses within the Hordeae tribe may be more tolerant to imazapic than other grass species.” They concluded that, “The effect of herbicide on desirable vegetation remains difficult to predict.”
Reduction of invasive annual grasses near Cape Town, South Africa
In a study on the effectiveness of treatments used to reduce invasive grass species expansion within a South African wildflower reserve, Musil, et al., 2005, compared various treatments within 10,000 square meter parallel transects: within the middle transect, hand-removal of invasive grasses in the fall (the grass was then spread in an adjacent undisturbed transect); grass brush cut (mowed) to a height of about 5cm in alternate transects (with the grass spread in adjacent uncut transects). Following these two treatments, a 20-meter-wide fire-break was brush cut around the reserve’s perimeter. Then all vegetation in the reserve, except within two mowed transects, was given either intense or light burning. Forty-eight 100 square meter plots were randomized within the transects. This resulted in four primary treatments: hand-removal plus light burning; hand-removal plus intense burning; brush cut plus light burning; brush cut plus intense burning. These primary treatments were then divided into two secondary treatments: application of pre-emergent herbicide (Snapshot: trifluralin and isoxaban); or no herbicide.
The reserve is broadly categorized as “evergreen fire-prone vegetation.” In the two year study, within which the first year was significantly drier than the second year, abundances were measured of native and alien grasses, and native herbaceous and woody species. Species densities were calculated from counts of the average number of individuals present in sub-plots. In the case of mature perennial grasses, abundances were calculated from the average percentage covers of each species.
Musil, et al., 2005, results:
The obligate seeding native forbs had significantly higher densities (5 times greater) in the wetter second year, and their densities were significantly lower (1.9 times less) in plots that received pre-emergent herbicide, compared with the plots with no herbicide. This group’s densities were significantly altered in the mechanically manipulated plots. “General trends of increasing forb densities with increasing intensity of disturbance in the sequence mowed < lightly burnt < intensely burnt < hand-cleared were evident, both in those plots where the pre-emergent herbicide was added and in those where the herbicide was excluded, particularly in the wetter second observation year.”
For the obligate seeding, mainly invasive, annual grasses: total densities did not differ between drier and wetter years. Their densities differed significantly in the differently mechanically manipulated plots, “with measured densities averaging between 4.9 and 5.5 times higher in the lightly burnt, intensely burnt, and hand-cleared plots than in the mowed plots. Also, densities of this functional group were significantly… reduced in those plots where the pre-emergent herbicide was added.” There were significant interactions between some treatments. “The first interaction indicated that reductions in annual grass densities following addition of the pre-emergent herbicide… were larger in the lightly burnt plots (6.4-fold decrease) and mowed plots (4.0-fold decrease) than in the intensely burnt plots (2.7-fold decrease) and hand-cleared plots (1.4-fold decrease). The second interaction specified greater herbicide-induced reductions in total densities of annual grasses… in the drier first observation year (6.6-fold decrease) than the wetter second (1.7-fold decrease).”
For the sprouting native geophytes (plants with underground bulbs, corms, rhizomes, or buds), “total densities were not significantly (P ≥ 0.05) affected by the addition of the pre-emergent herbicide… However, densities of this functional group were significantly (P ≤0.001) higher (3.4 times greater) in the wetter second year than the drier first. Also, they were significantly (P ≤ 0.05) altered in the differently mechanically manipulated plots with a significant interaction (P ≤ 0.05) evident between mechanical treatment and observation year... This interaction indicated larger increases in geophyte densities between the drier first and wetter second observation year… in the mowed plots (10.5-fold increase) than in the hand-cleared plots (7.9-fold increase), intensely burnt plots (3.9-fold increase) and lightly burnt plots (1.6-fold increase).”
For the sprouting perennial grasses, “The relative densities of the native and alien perennial grasses in this functional group were not significantly… affected by the higher annual precipitation in the second observation year…, by any of the mechanical manipulations applied…, or by the addition of the pre-emergent herbicide... Total covers of perennial grasses were significantly… lower (2.3 times smaller) in the wetter second than in drier first observation year… but were not significantly… affected by the addition of the pre-emergent herbicide... However, covers of this functional group did differ significantly … between the differently mechanically manipulated plots, the measured amounts in the lightly burnt plots being consistently higher (up to 3.8 times more) than those in the mowed, hand-cleared and intensely burnt plots.”
The team also compared the costs of the different treatment methods. For managing invasive annual grasses, intense burning of uncut vegetation was the cheapest strategy. The limitation of intense burning was that it inhibited recruitment of native geophytes. But the slightly more expensive light burning did stimulate the growth of geophytes. “Control of invasive annual grasses by burning may be effective only if all adult plants and their vegetative propagating organs are destroyed and their seed banks eliminated.”
Mowing, slightly less expensive than light burning, “elicited responses of the smallest magnitude from the four designated functional groups” in the first year. But mowing resulted in increased growth of native forbs and geophytes in the second year. In contrast, “covers of creeping perennial grasses in the mowed plots were reduced in the second observation year… The cover of perennial grasses in this study was negatively correlated with the densities of annual forbs and geophytes, but not with those of the annual grasses.”
Use of the herbicide was the most expensive but most effective method for controlling invasive annual grasses in the first year, and in the lightly burned plots. It was less effective in the second year. “This treatment’s disadvantage was that it … inhibited recruitment of native species of forbs, though it did not detrimentally influence springtime densities of perennial grasses or indigenous geophytes. This was expected as such herbicides kill germinating seeds rather than established sprouting plants. Another short-term disadvantage of herbicide application is that inhibition of annual grass populations does not necessarily result in greatly decreased seed production.” The potential emergence of rapid herbicide resistance may make such treatments more costly over time.
Conclusions by Musil, et al., 2005: The most credible strategy for managing invasive annual grasses in this location appears to be mowing “prior to grass seed maturation, and removal of the cut grass biomass for use as fodder in restricted feed lots.” But in sites with heavy, long-term invasive annual grass infestations, “an initial light intensity autumn burn to remove excessive grass litter, and to promote flowering, seed production and smoke-stimulated germination of native taxa, followed by the addition of a pre-emergent herbicide…, may assist in partly recovering native floral diversity.”
Comparison of treatments for quail grassland habitat improvement
Madison, et al., 2001, conducted a field experiment to compare the effectiveness of a variety of treatments in reducing the density of tall fescue populations to potentially encourage the reestablishment of northern bobwhite populations in Kentucky grasslands. Tall fescue was imported to the US as a forage grass. It is now considered invasive. Tall fescue provides poor habitat for bobwhite feeding, nesting, and roosting. Other grasses and forbs provide better feeding habitat. Disturbance of a field densely populated with tall fescue can improve habitat for bobwhite. The study was conducted “on 4 Kentucky Department of Fish and Wildlife Resources …Wildlife Management Areas (WMA) representing different physiographic regions and bobwhite densities.” Each site was divided into 16 0.1-ha plots. Each plot was randomly assigned one of eight treatments: fall burning, fall disking (with a harrow), spring burning, spring disking, spring herbicide application, summer burning, summer disking, and a control. They implemented the treatments in fall 1991 and spring and summer 1992. Herbicide plots were mowed, allowed to regrow for two weeks, and then sprayed with RoundUp.
Madison, et al., 2001, conducted observations of the plot using 1.0 square meter quadrats placed randomly in each plot. They estimated “percent of vegetation canopy coverage, bare ground, and canopy coverage by individual plant species.” They visually estimated the first two characteristics. And they calculated percentage of canopy cover of individual plants by species, which included both native and non-native grass and forb species. They calculated the percentage of plants in each plot that represented preferred bobwhite foods. And they estimated plant height. They “sampled the vegetation once each season, one year prior to treatments” (four sample periods in 1990-1991). And they “measured the vegetation once each season for 2 years post-treatment, beginning in fall 1991 and ending summer 1994” (12 sampling periods). They collected seed in winter 1993, and air-dried it and weighed it.
Madison, et al., 2001, results:
“Tall fescue coverage did not differ significantly among [Wildlife Management Areas]…, and there was no WMA by treatment interaction through time.” This enabled them to pool data across WMAs. They “compared fescue coverage among treatments at 1 year pre-, 1 month post-, 1 year post-, and 2 years post-treatment.” And they compared fescue coverage within each treatment plot across time.
“Fall disking was the best treatment for potentially improving bobwhite winter feeding habitat during winter 1993. Canopy coverage of preferred bobwhite foods in summer 1992 was 45.9 ± 13.0%... Availability of bare ground also contributes to the quality of winter feeding habitat... Fall-disked plots provided 48.5 ± 8.2% bare ground during winter 1993.” Bare ground is necessary to enable bobwhite to forage for seeds, but bare ground > 60% increases their vulnerability to predation. The optimal bare ground coverage is estimated to be 30-60%.
“Spring-disked and herbicide-treated plots may have provided good to moderately good winter feeding habitat during winter 1993. Spring- disked plots averaged 45.1 ± 11.6% food plant canopy coverage during summer 1992 and produced greater seed biomass…between summer 1992 and winter 1993 than control plots; however, percentage of bare ground was >60%,” at the upper limit for the safety of bobwhite populations.
“Herbicide-treated plots averaged 59.2 ± 11.8% food plant canopy coverage in summer 1992 and produced greater seed biomass… from summer 1992 to winter 1993 than all other treatment plots, but percentage of bare ground was >80%.”
“Herbicide-treated plots during summer 1993 were the only treatment plots that satisfied most bobwhite nesting habitat requirements. Grass composition was 48.9 ± 13.0% and vegetation height was 46.9 ± 8.2 cm on herbicide-treated plots during summer 1993.” Optimal bobwhite nesting habitat is estimated to contain “vegetation cover comprised of at least 40-60% grasses approximately 40-60cm tall,” with 30-60% bare ground.
“The most effective treatment to reduce tall fescue canopy coverage was the spring herbicide application. Fescue canopy coverage 1 month post- (x=1 7 ± 1.7%), 1 year post- (x=9.8 ± 7.9%), and 2 years post- (x=32.5 ± 6.5%) treatment on herbicide treated plots was less than control plots (x=82.4 - 92.1%) and pre-treatment conditions (89.6 ± 7.8%).” However, the rapid decline in effectiveness of the herbicide treatment over the two years post-treatment would likely mean that additional herbicide treatments would be needed, which increases the risk for the development of weed species developing herbicide-resistance.
“Disking and prescribed burning did not reduce tall fescue canopy coverage for > 1 year. Fescue canopy coverage was reduced by 41 to 81% on burned plots and by 75 to 88% on disked plots 1 month post-treatment… By 1 year post-treatment, only fall (x = 46.9 ± 9.9%)-, spring (x=25.9 ± 10.9%)-, and summer (x=56.5 ± 10.4%)-disked plots had fescue canopy coverage less than control plots (x=92.1-96.9%) and pre-treatment conditions (x=83.9-95.4%).” By two years after treatment, fescue canopy coverage on disked and burned plots did not differ from control plots and pre-treatment. “Herbicide-treated plots were the only ones that provided good bobwhite feeding habitat in winter 1994. Canopy coverage of preferred bobwhite food plants during summer 1993 averaged 42.4 + 11.5% and percentage of bare ground in winter 1994 averaged 60.6 + 9.2% on herbicide-treated plots.”
“No treatment plots satisfied more than 2 bobwhite nesting habitat requirements [by the completion of the study] in summer 1994.” Schroder, 1985, upon whose research the optimal habitat requirements were based, stated that “low values of any one variable may be partially offset by higher values of the remaining variables.”
Madison, et al., 2001, concluded that, “The effect of these treatments on bobwhite habitat should be viewed cautiously. Bobwhite habitat requirements are well documented…, but implementing these treatments on a small scale only provides potential improvement in bobwhite habitat quality.” The fact that no treatment provided satisfactory results two years after treatment emphasizes the challenges of bobwhite habitat restoration. By two years after treatment, any improvements gained by disking or burning were gone, and benefits of herbicide treatment were rapidly diminishing. Madison, et al., 2001 compared treatments that were used alone. Recommendations for invasive species control advocate using multiple treatments and multiple herbicides sequentially. Refer to the description of Musil, et al., 2005, above.
References:
- Madison, L.A., Barnes, T.G., & Sole, J.D. (Summer, 2001). Effectiveness of fire, disking, and herbicide to renovate tall Fescue fields to bobwhite habitat. Wildlife Society Bulletin (1973-2006), 29(2):706-712.
- Musil, C.F., Milton, S.J., & Davis, G.W. (July/August 2005). The threat of alien invasive grasses to lowland Cape floral diversity: An empirical appraisal of the effectiveness of practical control strategies. South African Journal of Science, 101: 337-344.
- Sheley, R.L., James, J.J., Rinella, M. J., Blumenthal, D., & DiTomaso, J.M. (2011). Invasive plant management on anticipated conservation benefits: A scientific assessment. In D.D. Briske (Ed.) Conservation benefits of rangeland practices: Assessment, recommendation, and knowledge gaps. (pp. 293-336). USDA Natural Resources Conservation Service.
Links to additional information on herbicides and other pesticides: