Final Report for SW00-015
Project Information
Grazing sheep on wheat stubble resulted in higher wheat stem sawfly mortality than tillage or burning. Weed populations were either lower or did not differ from tillage or burning. Grazing did not negatively impact soil compaction. In a second study, we found that sheep grazing alfalfa residue dramatically reduced weevil numbers compared to non-grazed plots. Although grazed alfalfa plots had less biomass than non-grazed plots at the end of grazing in mid-spring, by harvest time there was no difference in hay yields. These results indicate that grazing sheep on alfalfa and wheat residue can help control pest insects and weeds.
Objectives and methods described in this proposal are for the first three years of a long-term commitment by the project team to integrated sheep into a crop residue management system. Specific objectives were derived from reactions and concerns from a preliminary study conducted by Hatfield, Blodgett, Walkers, and Swartzs (Hatfield et al., 1999a), a producer/scientist planning meeting to outline objectives and determine responsibilities (appendix b), producer organization input (appendix d), and reviewer comments from previous submitted SARE proposals.
Wheat stubble grazing and wheat stem sawfly and weed control
Compare burning, grazing, and tilling wheat stubble fields in a multi-farm study on:
Objective 1. Over-wintering WSS larvae and emergence of adult populations.
Objective 2. Soil nutrient profiles, nutrient cycling, and soil compaction.
Objective 3. Total biomass including wheat stubble, cheatgrass, volunteer wheat, and mustard weed.
Objective 4. Crop yields and plant health
Objective 5. Develop an economic model to evaluate long-term cost-benefits of the various methods of wheat stubble management and their respective outcomes.
Objective 6. Develop and conduct large, multi-farm field demonstrations. Communicate results to producers, students, scientists, and public on advantages of implementing sustainable alternative grain residue management strategies.
Alfalfa residue grazing and weevil control
Objective The objective of this study was to determine the effect of winter through spring sheep grazing on spring re-growth characteristics of alfalfa and change in alfalfa weevil densities in southwestern Montana
Long-term goal
The team’s long-term goal is to develop an integrated crop/livestock production system that is economical and environmentally sustainable and provides benefits to both grazing livestock and grain producers. In addition, we look forward to developing a holistic sheep grazing program based on weed, insect, brush, and fire control. This will result in rural development with a new paradigm for production based on the “marriage” of food and fiber production with landscape manipulation.
Background
In the United States, 789 million tons of crop residues are produced annually in excess of that needed to prevent soil erosion. In many production settings crop residues are considered a hindrance to production and profitability as well as a habitat for insect pests. Burning, cutting, and baling, or tillage to reduce biomass and control disease and insect is typically practiced. All of these practices have negative environmental and/or economic consequences. Burning stubble releases particulate matter and a number of gases including carbon dioxide, methane, carbon monoxide, and oxides of nitrogen. Emissions from residue burning are in the fine particulate range and include carbon-based particles such as soot, ashes, dirt, soil dust, acid aerosols, and plant matter. Tillage is costly in terms of fuel, equipment, and labor. Sheep can graze stubble, converting it into food and fiber for human consumption (Hatfield et al., 1999b). No published information (AGRICOLA 1970-1998) is available that incorporates pest control, nutrient cycling, or economic evaluation of a sheep/wheat production system into a “whole farm” systems approach. The goal of our project is to evaluate and demonstrate sheep grazing on small grain stubble, to show grazing as an alternative to traditional stubble management practices, and to illustrate benefits for pest management, nutrient cycling, and weed control.
Wheat stem sawfly (Cephus cinctus Norton, WSS) is the most damaging insect or disease pest to Montana’s $1 billion/year grain industry with an economic impact estimated in excess of $30 million/year (Blodgett et al. 1996). Wheat stem sawfly is distributed throughout the Northern Great Plains. Originally a pest of spring wheat, the adult emergence period has gradually shifted earlier, making WSS a significant winter wheat pest (Morrill & Kushnak 1996). Adults emerge in early summer and females lay single eggs within an elongating wheat stem. Eggs hatch and larvae feed on parenchyma tissue within the stem. As wheat matures, the larva completes its feeding, and travels to the base of the stem where it cuts and plugs the stem behind it forming a stub. Larval cutting weakens the wheat stem increasing the likelihood of lodging, rendering the grain unharvestable. The WSS passes most of its life, egg through pupae, within a single wheat stem, protecting it from unfavorable environmental influences and control practices, making WSS management difficult to achieve. Insecticide application provides inconsistent results because a single application targets a non-feeding mobile insect that emerges over a 4-6 week period. Techniques such as tillage (Goosey, 1999) or burning provide some reduction in WSS population but have expenses associated with their implementation and cause other problems including soil erosion and negative impacts on water and air quality. Resistant strains of wheat have been developed (McNeal et al. 1971), however, yield reductions of 10% over that of susceptible wheat varieties occur and are therefore not readily used by producers. All of these practices are costly and do not offer effective control. Targeting the overwintering stage of the WSS for control through grazing offers advantages for both residue and insect pest management (Hatfield et al. 1999a).
Worldwide, more than 90% of the sheep’s diet is composed of roughage. Cereal crop residues are primarily structural carbohydrates that are unusable for non-ruminant (including human) foods and lack the energetic density to warrant processing or transport. Because of the symbiotic relationship between ruminant animals and their rumen microbial populations that digest these structural carbohydrates, ruminant grazing is the most energetically efficient method of utilizing and managing the vast energy potential of cereal crop residue. By taking advantage of underutilized small grain crop residues, the integration of sheep into grain production has the potential to be an “environmentally friendly” method of meeting the world’s increasing protein needs.
The American Sheep Industry organization speculates that there will be excellent opportunities for skilled sheep husbandry people to not graze sheep solely for meat and wool production, but to use sheep to manipulate landscapes. Examples include controlling noxious weeds, thinning brush and weeds in tree plantations, controlling brush in areas susceptible to fire, and “processing” crop residues, including insect pest control. Sheep grazing is recognized as a effective non-chemical method for noxious weed control in Montana and has resulted in economically beneficial opportunities to sheep producers for control of leafy spurge (S. Walker, personal communication). Although using sheep in this situation requires changing from traditional sheep management practices, opportunities will arise for rural development through sheep’s role as landscape manipulators. Typically, the largest constraint to entering the grazing sheep industry is the purchase of land. The results of this project (both the biological and economic components) coupled with previous research on sheep’s role in weed, brush, and fire control will allow new and existing entrepreneurs to transform high feed costs in the form of purchased grazing land into a potential economic return via sheep’s role in landscape manipulation.
Literature Review
Grazing wheat stubble – wheat stem sawfly and weed control
Much is known individually about crop and animal components of agriculture production. However, seldom are crop and animal agriculture combined into a mutually beneficial system. The potential use of crop residues (NRC, 1983) and non-traditional grazing systems that incorporate residue grazing is impressive (Blackburn, 1991). In view of the worldwide application of grazing crop residues (Owen and Kategile, 1984) and the potential for grazing systems that incorporate their use, this subject has received little attention from the research community.
Although there has been much research conducted on feeding straw to livestock, most of this work revolves around various methods of treating straw to improve digestibility (Lesoing et al., 1981). These methods may improve sheep weight gains (Higgins, 1981) but harvest, transport, and treatment of straw rely heavily on fossil fuels and will not benefit U.S. agriculture’s competitive position in the world market. Coombe (1981) concludes that the subject of stubble grazing merits more attention, especially as harvesting of crop residues becomes less economical because of rising fossil fuel costs. Hatfield et al. (1999a) demonstrated that sheep grazing of wheat stubble can reduce populations of WSS larva. In addition, Hatfield et al. (1998 and 1999b) concluded that grain stubble and other crop residues are viable and economical methods of maintaining ewes during the winter and extending the traditional seasonal supply of high quality market lambs.
Burning small grain stubble is considered by most producers to be an inexpensive, labor-efficient means of removing unwanted crop residue prior to tillage and seedbed preparation. However, Dormaar et al. (1979) clearly demonstrated an increase in nitrate levels and a decrease in cereal grain yields as a result of long-term burning. Many producers believe that burning is also a viable method of controlling pest. Biederbeck et al. (1980) reported that the heat from burning only penetrated 1.3 cm into the soil and thus had minimal effects on weed seed, unwanted insects, and disease.
Stevens et al. (1997) estimated the cost of burning rangelands at $4.64/acre. Although this work is not directly applicable to burning small grain stubble, it does demonstrate the potential for production costs associated with burning. In previous research, Hatfield et al. (1999b), small grain stubble for sheep was leased from private farming operations for $.03 to $.05/sheep daily. Reported stocking rates for sheep on stubble range from 168 (Thomas et al., 1990) to 808 (Mulholland et al., 1976) sheep days/acre. Calculated returns per acre range from $5.04 using the stocking rate reported by Thomas et al. (1990) and the $.03/sheep daily to $40.42 using the stocking rate reported by Mulholland et al. (1976) and the $.05/sheep daily. Thus, if a small grain producer considers the loss in yields, the increased cost of fertilizer, the cost of burning, the loss of revenue, and the resulting air quality conflicts with downwind urban areas, the negative impacts of burning are substantial.
Traditional agricultural production systems based on intensive use of inputs have contributed greatly to feeding the world’s growing population. However, this may not be the preferred strategy for the future, for economic as well as environmental reasons (Poincelot, 1986; Papendick et al., 1987). Integrated, low input systems optimize output per unit of input, rather than maximize output at all cost (Stinner and House, 1989). The proposed project takes advantage of the potential mutual benefit of sheep herbivory with WSS (Hatfield et al., 1999a) and weed control (Walker et al. 1992) as well as stubble reduction and improved organic matter cycling in grain crops (Fritz, 1995 a, b).
In the past 50 years, researchers, and to a lesser extent producers, have focused on maximizing single commodity outputs rather than optimizing production through integrated production systems. A major problem resulting from this mentality has typically been increased input costs. Our proposed production systems will maximize sustained conversion of forage to animal products, maintain ecosystems with high resilience, and make better use of crop residues, all of which serve to minimize energy costs. Integrating animal and crop production systems to provide mutual benefits increases the economics and sustainability of cropping systems. Benefits for the grain and sheep producer will accrue from this research by developing relationships that can be used in mixed agricultural systems that may be developed locally that include the biological, environmental, and economical aspects of crop residue management.
Alfalfa residue grazing and weevil control
Alfalfa, Medicago sativa (L.), is grown on approximately 10.6 million ha in the United States (Bailey 1994) with a 1998 estimated on-farm value of $5 billion (Radcliffe and Flanders 1998) and represents the foremost forage crop in many semiarid and temperate states (Allen et al. 1986b, Bailey 1994). Two biological stressors (insects and weeds) combined with poor field management are primarily responsible for reduced alfalfa production (Latheef et al. 1988). The alfalfa weevil, Hypera postica Gyllenhal, is economically the most damaging phytophagous pest of alfalfa in the United States (Blodgett et al. 2000).
Montana sheep producers often rely on fall regrowth of alfalfa as a source of fall and winter pasture. Fall regrowth also is utilized as overwintering habitat by the adult alfalfa weevil (Dively 1970, Dowdy et al. 1986), which hibernates in leaf litter or around plant crowns (Blodgett 1996). In the southern U.S., the majority of weevil eggs are oviposited in alfalfa regrowth during fall and winter months, making fields with copious fall regrowth more attractive (Berberet et al. 1980). However, in colder northern states, such as Montana, temperatures restrict weevil winter activity, and little to no oviposition occurs during winter months (Blodgett et al. 2000). Because alfalfa weevil adults aestivate during summer, emerging in fall, they are in a resting state when temperatures are low during winter.
Multiple tactics have been examined to manage alfalfa weevil populations and limit damage with varied results. Alfalfa weevil-tolerant cultivars currently available to producers often do not provide sufficient protection from weevil larval damage to justify their use (Blodgett et al. 2000). Biological agents have reduced weevil populations below economic thresholds in the eastern U. S. (Richardson et al. 1971); however, their impact has been marginal in the western U.S. (Van den Bosch 1972, Kingsley et al. 1993, Brewer et al. 1998, Radcliffe and Flanders 1998). Insecticides that target alfalfa weevil larvae are used on approximately 34% of the alfalfa acreage in the U.S. (Bailey 1994). However, insecticide use is costly and requires intensive field monitoring, by producers, to determine when a treatment is economically justifiable.
Cultural practices for weevil management include late fall (Dowdy et al. 1992) and early spring harvest (Essig and Michelbacher 1933, Harper et al. 1990), burning (Bennett and Luttrell 1965), early harvest with raking (Blodgett et al. 2000), and grazing (Dowdy et al. 1992). Late fall harvest as practiced by Dowdy et al. (1992) reduced weevil eggs by 55% in fall regrowth but did not reduce spring larval numbers compared to unharvested controls. Blodgett et al. (2000) reported early harvest followed by raking reduced alfalfa weevil larval numbers in post harvest stubble by 43% compared to early harvest alone. This tactic represents the only recommended non-chemical means of weevil management in Montana.
Dowdy et al. (1992) reported a 67% reduction in weevil eggs and 25% reduction in spring larval numbers in grazed compared to nongrazed plots in Oklahoma. In northern U.S. states, cold winter temperatures restrict early spring weevil activity and oviposition (Blodgett et al. 2000). Therefore, these researchers speculated that winter/spring grazing would have no impact on spring larval populations. Spring larval populations that damage first-cut alfalfa in Montana hatch from eggs oviposited that spring (Blodgett 1996). Early spring harvest, as reported by Essig and Michalebacher (1933), can remove the majority of weevil eggs and young larvae, thus reducing subsequent damage. However, early summer harvesting, before physiological plant maturity, typically has a negative impact on yield. Hard winter grazing of many grass and clover species, as reported by Brougham (1960), can favor regrowth by removing foliar cover thus allowing sunlight to penetrate the canopy and raising soil temperatures favorable for plant growth sooner than in non-grazed or mowed plots. There is no published literature defining grazing dormant alfalfa with sheep in Montana and the impacts this grazing has on spring regrowth characteristics and alfalfa weevil densities.
Cooperators
Research
Methods (and results) are partitioned into the five major segments of the study. All research was conducted on private farms and ranches in Montana.
Incorporating sheep into dryland grain production systems: I. Impact on over-wintering larva populations of Wheat stem sawfly, Cephus cinctus Norton, (Hymenoptera: Cephidae)
Experiments were conducted in conjunction with two other studies published in this series (Hatfield et al. 200_b; 200_c). Studies were conducted at eight sites on four farms located in Montana with high WSS infestations to evaluate the effects of various management strategies on over-winter WSS larval mortality. The experimental design was a complete randomized block design, replicated four times at each site. Individual plots were 9 x 12.3 m and were the experiment unit.
Sites 1 and 2 were located in Stillwater County, south central Montana ( ). Sites 3 through 8 were located in Toole and Pondera counties, north central Montana ( ). All sites were seeded to winter wheat on grain stubble fields resulting from 2000 (sites 1 through 4) and 2001 (sites 5 through 8) crops. Precipitation was lower and the number of frost-free days greater than average for both years at all sites (NASS, 2003). The soil type at all sites was a clay loam (NRCS, 1997).
Plots were sampled to determine WSS larval numbers prior to treatment imposition in the fall (September and October) and following completion of treatment in the spring (May), but before adult WSS emergence occurred. Three samples were taken from each plot. A sample consisted of removing all stubble material, including plant crowns, from a 0.46 m length of a single stubble row. Samples were labeled and returned to the laboratory where WSS cut stems were identified and dissected to determine if the stem contained a live WSS larva. Response variables were post treatment (sampled in spring) live larval numbers and percent WSS mortality larvae in each plot. Calculations for percent mortality were made from the beginning (fall) and ending (spring) data from each plot and were calculated as percent mortality = [(fall samples – spring samples)/fall samples] x 100.
Experiment 1
Treatments were fall tilled (GT), fall grazed (GF), spring grazed (GS), fall and spring grazed (GFS,) and an untreated control (GC). For the grazing treatments, five mature western white-faced ewes were randomly assigned to each grazed plot. Sheep were kept in plots with electro-net temporary fence (Premier Fence Systems, Washington, IA) powered by Intellishock 40B energizers (Premier Fence Systems, Washington, IA) and Dura-Start deep cycle batteries (Exide Corp., Reading, PA). The GF and GS plots were grazed for 24 hours, resulting in a stocking rate of 400 sheep days per ha (2.1 aum). The GFS plots were grazed for 24 hours in the fall and again for 24 hours in the spring, resulting in a stocking rate of 800 sheep days per ha (4.2 aum). Tillage was done with a chisel plow. Plots were tilled once in the fall to a depth of approximately 20 cm. Tillage occurred within 72 hours of fall grazing.
Experiment 2
In Experiment 2, the untreated control (BC), fall grazed (BF), and fall tilled (BT) treatments were subsets of GC, GF, and GT, respectively. These treatments, along with the burned treatment (BB), were imposed at six of the eight sites. Plots were burned with a propane brush burner in the fall, within 72 hours of the fall grazing treatment.
Experiment 3
In Experiment 3, the untreated control (TC) was a subset of GC. The other treatments were fall clipped (TCP), fall trampled (TF), spring trampled (TS), and fall and spring combined trampled (TFS). These treatments were imposed on two of the eight sites. For the clipped treatment, the entire plot was mowed to an average stubble height of 4.5 cm with a rotary mower. For TF, TS, and TFS treatments, sheep were muzzled to prevent grazing. Stocking rates for the trampling treatments were the same as for the grazed treatments.
Statistical Analysis
A randomized complete block design with four blocks per location was used for each experiment. Data were analyzed using the GLM procedure of SAS (1993) with plot as the experimental unit. The study was conducted over a two-year period at four locations per year. Because yearly variation was not an objective of our study, location and year were combined into one term defined as site. Therefore, in Experiment 1 (for example) variation associated with location (df=3), year (df=1) and the interaction (df=3) were pooled into the site variable with 7 df. The model included effects of site, treatment, and site by treatment interaction. Models also included the following contrast statements: Experiment 1: a) GC vs grazed, b) GT vs grazed, c) GF and GS vs GFS; Experiment 2: a) BB vs BC, b) BB vs BT, and c) BB vs BF; and Experiment 3: a) TC vs trampled, b) TCP vs trampled, c) TF vs TS, d) TF and TS vs TFS. Wheat stem sawfly larval numbers at the beginning of each experiment were tested in each model as a covariable and included when significant (P£0.10). Comparisons were considered significant when P was £0.10.
Incorporating sheep into dryland grain production systems: II Impact on changes in biomass and weed density
Experiments were conducted in conjunction with research presented in two other articles published in this series (Hatfield et al. 200_ a and c). In this paper, we present the effects of sheep grazing and other management strategies on change in aboveground biomass and weed density in fallow fields in the Northern Great Plains. The experiments were conducted at eight sites on four farms over two years. For each site year, the experimental design was a randomized complete block with four replicates. Individual plot size was 9 x 12.3 m.
Sites 1 and 2 were located in Stillwater County, south central Montana. These sites had been seeded to winter wheat the previous crop year. Sites 3 through 8 were located in Toole and Pondera counties, north central Montana. These sites were also seeded to winter wheat the previous crop year. Sites were established on grain stubble fields resulting from 2000 (sites 1 through 4) and 2001 (sites 5 through 8) crops. Precipitation was lower and number of frost-free days greater than average for both years at all sites (NASS, 2003). The soil mapping units at all sites were clay loams (NRCS, 1997).
Plots were sampled to determine total biomass and weed density prior to treatment imposition in the fall (September and October) and following completion of treatment imposition in the spring (May). For biomass, three 0.1 m² sub-samples, consisting of all live and dead aboveground vegetative matter, were taken from each plot. Sub-samples were composited, labeled, and returned to a laboratory where they were dried at 50º C for 48 h and weighed. The response variables were ending plant biomass and percent change in plant biomass calculated as (ending mean/beginning mean) x 100.
Weed density was determined only in the first year of the trial due to severe drought and an absence of weeds in the second year. Three sub-samples were taken per plot as previously described for biomass sampling. For the determination of weed density, all weeds within each of three 0.10 m² rings were counted. Response variables were ending weed density and percent change in density, calculated as (ending treatment mean/ beginning treatment mean) x 100.
Experiment 1
Treatments were fall tillage (GT), fall grazed (GF), spring grazed (GS), fall and spring grazed (GFS), and an untreated control (GC). For the grazing treatments, five mature western white-faced ewes were randomly assigned to each grazed plot. Sheep were kept in plots with electro-net temporary fence (Premier Fence Systems, Washington, IA) powered by Intellishock 40B energizers (Premier Fence Systems, Washington, IA) and Dura-Start deep cycle batteries (Exide Corp., Reading, PA). The fall grazed and spring grazed plots were grazed for 24 hours, resulting in a stocking rate of 400 sheep d/ha (2.1 aum). The GFS plots were grazed for 24 hours in the fall and again for 24 hours in the spring, resulting in a stocking rate of 800 sheep d/ha (4.2 aum). Tillage was done with a chisel plow. Plots were tilled once in the fall to a depth of approximately 20 cm, within 72 hours of fall grazing.
Experiment 2
For Experiment 2, fall grazed (BF), fall tillage (BT), and the untreated control (BC) were subsets of GF, GT, and GC, respectively. These treatments, along with the fall burned treatment (BB), were imposed at six of the eight sites. Plots were burned with a propane brush burner within 72 hours of the fall grazing treatment.
Experiment 3
For Experiment 3, the untreated control (TC) was a subset of GC. The other treatments were fall clipped (TCP), fall trampled (TF), spring trampled (TS), and fall and spring combined trampled (TFS). These treatments were imposed at two of the eight sites. For the clipped treatment, the entire plot was mowed to an average stubble height of 4.5 cm with a rotary mower. For TF, TS, and TFS treatments, sheep were muzzled to prevent grazing within the plot. The stocking rates for the trampling treatments were the same as for the grazed treatments.
Statistical Analysis
A randomized complete block design with four blocks/location was used for each experiment. Data were analyzed using the GLM procedure of SAS (1993) with plot as the experimental unit. The study was conducted over a two-year period at four locations per year. Because yearly variation was not an objective of our study, location and year were combined into one term defined as site. Therefore, in Experiment 1 (for example) variation associated with location (df=3), year (df=1) and the interaction (df=3) were pooled into the site variable with 7 df. The model included effects of site, treatment, and site by treatment interaction. Models also included the following contrast statements: Experiment 1 a) GC vs grazed, b) GT vs grazed, c) GF and GS vs GFS; Experiment 2 a) BB vs BC, b) BB vs BT, and c) BB vs BF; and Experiment 3 a) TC vs trampled, b) TCP vs trampled, c) TF vs TS, d) TF and TS vs TFS. The appropriate variable (i.e. biomass for biomass analysis and weed density for weed analysis) at the beginning of each experiment was tested in each model as a covariable and included when it was a significant (P£0.10) source of variation. Comparisons were considered significant when P was £0.10.
Incorporating sheep into dryland grain production systems: II Impact on changes in biomass and weed density
Experiments were conducted in conjunction with research presented in two other articles published in this series (Hatfield et al. 200_ a and c). In this paper, we present the effects of sheep grazing and other management strategies on change in aboveground biomass and weed density in fallow fields in the Northern Great Plains. The experiments were conducted at eight sites on four farms over two years. For each site year, the experimental design was a randomized complete block with four replicates. Individual plot size was 9 x 12.3 m.
Sites 1 and 2 were located in Stillwater County, south central Montana. These sites had been seeded to winter wheat the previous crop year. Sites 3 through 8 were located in Toole and Pondera counties, north central Montana. These sites were also seeded to winter wheat the previous crop year. Sites were established on grain stubble fields resulting from 2000 (sites 1 through 4) and 2001 (sites 5 through 8) crops. Precipitation was lower and number of frost-free days greater than average for both years at all sites (NASS, 2003). The soil mapping units at all sites were clay loams (NRCS, 1997).
Plots were sampled to determine total biomass and weed density prior to treatment imposition in the fall (September and October) and following completion of treatment imposition in the spring (May). For biomass, three 0.1 m² sub-samples, consisting of all live and dead aboveground vegetative matter, were taken from each plot. Sub-samples were composited, labeled and returned to a laboratory where they were dried at 50º C for 48 h and weighed. The response variables were ending plant biomass and percent change in plant biomass calculated as (ending mean/beginning mean) x 100.
Weed density was determined only in the first year of the trial due to severe drought and an absence of weeds in the second year. Three sub-samples were taken per plot as previously described for biomass sampling. For the determination of weed density, all weeds within each of three 0.10 m² rings were counted. Response variables were ending weed density and percent change in density, calculated as (ending treatment mean/ beginning treatment mean) x 100.
Experiment 1
Treatments were fall tillage (GT), fall grazed (GF), spring grazed (GS), fall and spring grazed (GFS), and an untreated control (GC). For the grazing treatments, five mature western white-faced ewes were randomly assigned to each grazed plot. Sheep were kept in plots with electro-net temporary fence (Premier Fence Systems, Washington, IA) powered by Intellishock 40B energizers (Premier Fence Systems, Washington, IA) and Dura-Start deep cycle batteries (Exide Corp., Reading, PA). The fall grazed and spring grazed plots were grazed for 24 hours, resulting in a stocking rate of 400 sheep d/ha (2.1 aum). The GFS plots were grazed for 24 hours in the fall and again for 24 hours in the spring, resulting in a stocking rate of 800 sheep d/ha (4.2 aum). Tillage was done with a chisel plow. Plots were tilled once in the fall to a depth of approximately 20 cm, within 72 hours of fall grazing.
Experiment 2
For Experiment 2, fall grazed (BF), fall tillage (BT), and the untreated control (BC) were subsets of GF, GT, and GC, respectively. These treatments, along with the fall burned treatment (BB), were imposed at six of the eight sites. Plots were burned with a propane brush burner within 72 hours of the fall grazing treatment.
Experiment 3
For Experiment 3, the untreated control (TC) was a subset of GC. The other treatments were fall clipped (TCP), fall trampled (TF), spring trampled (TS), and fall and spring combined trampled (TFS). These treatments were imposed at two of the eight sites. For the clipped treatment, the entire plot was mowed to an average stubble height of 4.5 cm with a rotary mower. For TF, TS, and TFS treatments, sheep were muzzled to prevent grazing within the plot. The stocking rates for the trampling treatments were the same as for the grazed treatments.
Statistical Analysis
A randomized complete block design with four blocks/location was used for each experiment. Data were analyzed using the GLM procedure of SAS (1993) with plot as the experimental unit. The study was conducted over a two-year period at four locations per year. Because yearly variation was not an objective of our study, location and year were combined into one term defined as site. Therefore, in Experiment 1 (for example) variation associated with location (df=3), year (df=1) and the interaction (df=3) were pooled into the site variable with 7 df. The model included effects of site, treatment, and site by treatment interaction. Models also included the following contrast statements: Experiment 1 a) GC vs grazed, b) GT vs grazed, c) GF and GS vs GFS; Experiment 2 a) BB vs BC, b) BB vs BT, and c) BB vs BF; and Experiment 3 a) TC vs trampled, b) TCP vs trampled, c) TF vs TS, d) TF and TS vs TFS. The appropriate variable (i.e. biomass for biomass analysis and weed density for weed analysis) at the beginning of each experiment was tested in each model as a covariable and included when it was a significant (P£0.10) source of variation. Comparisons were considered significant when P was £0.10.
Incorporating Sheep into Dryland Grain Production Systems: A Field Size Approach
This study was conducted at 6 sites across Montana from October 2002 to May 2003. At each site, treatments were grazed and non-grazed plots in cereal grain stubble fields. Three non-grazed plots (9.1 x 12.2 m) were randomly located within each experimental field. Three grazed (9.1 x 12.2 m) plots were established, in the same field, 6.1 m north of each non-grazed exclosure. Sheep stocking rates and field sizes ranged from 70-285 sheep d/ha and 11.4-113 ha, respectively. Non-grazed enclosures were fenced at sites 1 and 2 on 20 August, 2002, 11 d prior to introducing sheep; at sites 3 and 4 on 15 September, 2002, 76 and 61 d prior to introducing sheep, respectively; at sites 5 and 6 on 8 October, 2002, 38 and 48 d prior to introducing sheep or goats, respectively. Fences remained standing at each site throughout the grazing period. All sites were established on cereal stubble resulting from the 2002 harvest. Weed and soil bulk density data were collected at each site. Wheat stem sawflies were only found infesting fields at sites 5 and 6. Therefore, WSS results are only presented for these sites.
Wheat stem sawfly
Three samples were taken from each plot (grazed and non-grazed) in September, prior to imposition of treatments, and in the following May, after treatments were removed but prior to WSS emergence. A sample consisted of all stubble material, including plant crowns, from a 0.46 m length of stubble row. Samples were labeled and returned to the laboratory where WSS cut stems were identified and dissected to determine if they contained viable WSS larvae. The percent mortality (i.e., percent reduction) of overwintering WSS larvae in each plot was calculated from these data.
Biomass (Weeds and Stubble)
Three samples were taken from each plot in September, prior to imposition of treatments, and in May, after treatments were removed. Each sample consisted of all aboveground biomass within a 0.1 m2 quadrat. Biomass was separated into wheat stubble and weeds, which were bagged separately. Samples were returned to the laboratory, dried at 50º C for 48 h, and weighed to determine grams of dry matter.
Soil Bulk Density and Moisture
Three soil bulk density samples were taken from each plot in September, prior to imposition of treatments, and in May, after treatment imposition. A sample consisted of a 5.08 x 5.08 cm soil core extracted using an AMS compaction core sampler (AMS, American Falls, ID). Samples were returned to the laboratory at Montana State University, dried at 105º C for 48 h, and weighed to determine soil bulk density. Additionally, three soil moisture samples were taken per plot post-spring using the AMS compaction core sampler. Extracted soils were immediately placed in air-tight, plastic bags and returned to the laboratory. Samples (soils and bags) were weighed to determine wet weight, then dried at 105º C for 48 h and weighed again to determine dry weight. Percent soil moisture was calculated, on a wet weight basis, from these data.
Statistical Analyses
The experimental design was a randomized complete block with three replicates at each site. The model included the effects of site treatment and site by treatment interaction with pre-fall biomass included as a covariable. The GLM procedure of SAS (SAS Institute 2000) was used to compute least squared means for treatment effects.
Evaluation of Alfalfa Weevil (Coleoptera: Curculionidae) Densities and Regrowth Characteristics of Alfalfa Grazed by Sheep in Winter and Spring
Research was conducted during two study years, 2002 and 2003, 13 km northeast of Dillon in southwestern Montana. During each study year, six non-grazed plots (9.1 x 12.2 m) were randomly located within a 2 or 3-y-old (depending on the study year), 36 ha field of ‘Geneva’ alfalfa. Grazed (9.1 x 12.2 m) plots were established, in the same field, 6.1 m north of each non-grazed enclosure. Plots were located within a fenced 324-ha area designated solely to alfalfa and hay barley production. Sheep had free access to the entire 324-ha area during the grazing period, but plots were established within a 36-ha portion of this area to eliminate bias associated with establishing research plots on different alfalfa cultivars.
Non-grazed enclosures were fenced on 16 October 2001 and 30 September 2002, 95 and 128 d prior to introducing sheep, respectively. Fences remained standing throughout the grazing period. Rambouillet ewe lambs (2002: n = 1,600) and breeding ewes (2003: n = 1,200) were introduced to the experimental field on 19 January 2002 and 5 February 2003 and were removed on 3 May 2002 and 15 May 2003, respectively, resulting in stocking rates of 469 and 363 sheep days per ha, respectively. Pre-graze biomass samples were taken from each plot on 16 October 2001 and 30 September 2002. Post-graze biomass samples were taken from each plot on and 6 May 2002 and 28 May 2003. Biomass samples were collected by removing all plant material from three 0.11 m2 quadrats per plot, dried at 48° C for 72 h, and weighed to determine dry matter. Plant height and weevil count samples were taken weekly at 4 sampling dates during both study years: 2002 (date 1 = 5 June, date 2 = 12 June, date 3 = 19 June, and date 4 = 26 June) 2003 (date 1 = 5 June, date 2 = 12 June, date 3 = 18 June and date 4 = 25 June).
Spring re-growth characteristics
Mean stem height (cm) was determined by randomly selecting 10 stems from 5 randomly located 0.11 m2 quadrats per plot on each sample date. Stem damage was determined by cutting 100 stems from 10 random locations in each plot on each sampling date. Each stem was visually inspected for weevil damage and assigned a designation of ‘yes’ or ‘no’ depending on the presence or absence of weevil larval damage. The percentage of plants damaged by weevil larvae was calculated from these data. To determine yield, 3 (45.7 x 50.8 cm) quadrats were hand harvested from each plot using a Stihl HS 75 gas hedge trimmer (Stihl Inc., Virginia Beach, VA) by cutting and harvesting all aboveground biomass. Forage samples were dried at 48° C for 72 h to determine dry matter yield. Three stems per yield sample were collected at harvest and bagged separately for plant nutrient analyses (dry matter (%), crude protein (%), acid and neutral detergent fibers (%)) conducted at the Montana State University Oscar Thomas Nutrition Center. Crude protein (kg/ha), and acid and neutral detergent fibers (kg/ha) were calculated by multiplying yield with nutrient concentration. Samples were oven dried and ground to pass a 1.0 mm sieve using a Wiley Mill (Thomas Scientific, USA). Crude protein was calculated using the AOAC Leco combustion method 990.03 (AOAC International 1999) and acid and neutral detergent fibers were calculated using methods of Van Soest et al. (1991). Bloom stage (a visual indicator of plant maturity) was determined by assessing the phenological stage of 100 randomly selected stems per plot on 26 June 2002 and 25 June 2003.
Alfalfa weevil densities
Alfalfa weevil adult and larval densities were determined by collecting one sample, consisting of 20 (180°) sweeps with a 38 cm diameter sweep net, per plot per sampling date.
The experimental design was a randomized block with plot as the experimental unit. The GLM procedures of SAS (SAS Institute 2000) were used to compute least squared means to make within date comparisons of treatment stem height, percentage of stems damaged by weevil larvae, and alfalfa weevil larval populations. No data transformations were preformed prior to analysis. Least squared means were also calculated to analyze treatment dry matter, crude protein, acid and neutral detergent fibers, and for comparing treatment pre- and post-biomass and treatment yield and maturity.
Results and literature cited are partitioned into the five major segments of the study. All research was conducted on private farms and ranches in Montana.
Incorporating sheep into dryland grain production systems: I Impact on over-wintering larva populations of wheat stem sawfly, Cephus cinctus Norton, (Hymenoptera: Cephidae)
Experiment 1
Site by treatment interaction were detected (P=0.01; ) for post treatment WSS live larva numbers. Therefore, these results are presented as the simple effects of treatment within site. No site by treatment interaction was detected (P=0.37) for percent over-winter larval mortality, therefore, main effects of treatment across sites are presented.
Post treatment WSS live larval numbers were greater (P < 0.04) for GC than grazed (GF, GS, and GFS) at four sites and did not differ (P>0.27) at the remaining four sites. The sites where no differences were detected had relatively low WSS infestations. Larva mortality was greater (P=0.01) for grazed treatments than GC.
At three of the eight sites, post treatment larva numbers were greater (P<0.05) in GT (fall tilled) than grazed plots, but post treatment larval densities between GT and grazed plots did not differ (P>0.25) at the remaining five sites. Larval mortality was greater (P=0.01) for grazed treatments than GT.
At two of the eight sites, post treatment larva numbers were greater (P<0.05) for GF than GS. However, at the remaining six sites, larval numbers did not differ (P>0.24) between GF and GS. Larval mortality did not differ (P=0.75) between GF and GS.
Although post treatment larval densities did not differ (P>0.26) between GFS and the mean of GF and GS, mortality was greater (P=0.02) for GFS than the mean of GF and GS.
These results agree with those of Hatfield et al. (1999) who reported that compared to a non-grazed control, grazing wheat stubble with sheep in the fall and fall/spring reduced WSS larva numbers in the spring by 60 and 87%, respectively. Although Hatfield et al. (1999) reported that spring grazing and control plots did not differ in WSS larva numbers, spring grazing resulted in a 46% reduction in larval numbers in the spring compared to the non-grazed control.
Goosey (1999) speculated that for tillage to be an effective method of WSS control, soil needed to be removed from around the crown area of the wheat stubble to increase WSS desiccation due to freezing and drying. In field trials using tillage equipment similar to that used in our studies, only 35% of the plant crowns were soil free following tillage (Goosey, 1999). We used standard shallow tillage with sweeps at approximately 20 cm depth, and our results were similar to those of Goosey (1999).
Other studies comparing tillage and grazing on WSS overwintering larva numbers or mortality are not available. We speculate that compared to tillage, grazing the stubble either breaks the stubs and disturbs the frass plug the larva has deposited in the stub to protect itself from overwintering conditions or disrupts the soil around the crown of the plant, increasing larval exposure to greater extremes in environmental conditions.
Goosey (1999), Farstad et al. (1945), and Holmes (1954a) emphasized that increased exposure of stems is needed to increase WSS mortality. Church (1955) found that larvae collected in the early spring were physiologically resistant to moisture deficits, and were unaffected by soil moisture level. However, Holmes (1954a) reported that to reach 96% WSS mortality, spring wheat stubs must be exposed to environmental conditions between May 25 and June 4. At that time, larvae within the stem were between diapause and pupation and were susceptible to desiccation. In our study, larvae were exposed to cold temperatures, less than 0° C, as well as dry winter conditions. The fact that GF and GS did not differ in WSS mortality may indicate that, at the stocking rates used in our study, season of grazing may not be a factor in WSS mortality. Hatfield et al. (1999) also reported no difference in WSS mortality between fall and spring grazed plots. One possible explanation for the increased WSS mortality in the fall and spring grazed (GFS) plots compared to GF and GS is the higher stocking rate used in dual season grazing.
Both Mulholland et al. (1976) and Thomas et al. (1990) concluded that cereal stubble with green weedy material was an acceptable grazing resource for sheep when stocked at 330 and 420 sheep d/ha, respectively. The stocking rate for our study was 400 sheep d/ha for the GF and GS treatments, similar to that used by Thomas et al. (1990) and slightly greater than that used by Mulholland et al. (1976). Given the similar stocking rates, we conclude that the levels of WSS mortality noted on our study, was accomplished within the realm of reasonable stocking rates on grain stubble used for sheep production.
Experiment 2
Site by treatment interactions were not detected (P>0.25) for all of the variables measured in Experiment 2, therefore main effects of treatments across sites are presented. Post treatment larva numbers and mortality did not differ (P>0.63) for the comparisons of BB vs BC or BB vs BT. Post treatment WSS larva numbers were lower (P = 0.01) and overwintering larva mortality greater (P = 0.01) in BF than BB treatments.
Burning, tillage, and the untreated control did not differ in WSS mortality, indicating insufficient disruption of the overwintering WSS environment to cause larval desiccation or other environmental factors that increase mortality. Burning does not disrupt the soil surface, which may be key to enhance overwintering desiccation of WSS, and soil is not heated to temperatures adequate to kill larvae below the soil surface (Salt, 1947; Church, 1955). Salt (1946 and 1947) and Church (1955) concluded that WSS larvae can withstand high temperatures for long periods of time without desiccating. The lack of disruption of the soil around the plant crown in the burned plots may, in part, explain the lower mortality in burned than grazed plots.
Experiment 3
Site by treatment interactions were detected (P<0.02) for post treatment WSS larval numbers for Experiment 3, therefore, results are presented as the simple effects of treatment within site. Site by treatment interactions were not detected (P>0.90) for percent mortality. Therefore, main effects of treatments across sites are presented.
The mean of the combined trampling treatments had lower (P=0.01) post treatment larval numbers than TC or TCP at Site 4, but did not differ (P>0.19) at Site 8. Larva mortality was greater (P<0.01) for the combined trampling treatments than TC or TCP. Post treatment larva numbers and mortality did not differ (P>0.44) between TF and TS. Post treatment larva numbers were lower (P=0.06) for TFS than the mean of TF and TS at Site 4, but did not differ (P=0.17) at Site 8. Overwintering larval mortality was greater (P=0.01) for TFS than the mean of TF and TS.
When the graze treatments were compared within season to the trample treatments no differences were detected (P>0.36) except GS had a lower (P=0.01) post treatment larva number than TS at Site 4. Graze and trample treatments did not differ (P>0.25) in percent mortality.
Trampling disrupts the soil surface around the roots and causes damage to the stubs of the wheat plants. Anslie (1920) and Church (1955) found temperature and moisture to be the prime factors obstructing larval development. The hoof action of the sheep “churn” the soil near the roots where the larvae overwinter, loosening the soil so that it is not as effective for insulating the WSS larvae. Because trampling may be as important as, or more important than, consumption in reducing larva populations, other options of WSS control may be possible. For example, winter feeding of hay on fallow ground may result in sufficient hoof action to drastically reduce WSS larva numbers.
Conclusions
Sheep can be used as a tool in an integrated pest management system to reduce wheat stem sawfly numbers. Although more research is needed, particularly on large commercial scale grain operations, results are promising. To fully utilize sheep to control insect or weed infestations, it requires that grain and sheep producers both view the animal as a biological control agent rather than a grazing animal for the sole use of producing food and fiber.
Acknowledgements
This research was supported in part by Grant No.SW00-015 from USDA, CSREES, WR-SARE. We acknowledge excellent assistance from Brenda Robinson in upkeep of the experimental animals and assistance in data collection and analysis.
References
Ainslie, C.N., 1920. The Western Grass-Stem Sawfly. U.S.D.A. Bulletin No 841. Washington D.C.
Blodgett, S. L., Goosey, H.B., Waters, D., Tharp, C.I., Johnson, G. D., 1996. Wheat stem sawfly control on winter wheat. Arthropod management tests. 22, 331-332.
Blodgett, S.L., Johnson, G.J., Kemp, W.P., Sands, J.K., 1997a. Developing an Extension pest management program using the needs-assessment process. JOE 35, 1-5.
Blodgett, S., Johnson, G., Lenssen, A., Morrill, W., Weaver, D., Goosey, H., Sing, S., Waters, D., Berg, J., Bruckner, P., Lanning, S., Talbert, L., Brastrup, B., Crouch, W., Graham, P., Kerby, B., Maatta, J., Ranney, J., Roos, B., Wargo, J., Carlson, G., Eckhoff, J., Kushnak, G., Wichman, D., 1997b. Montana Crop Health Report. Special Edition: Wheat Stem Sawfly. 10(10):1-7. http://scarab.msu.montana.edu/mchr/1010_2.html. Accessed 10/29/01
Bowman, H., Cash, D., Bruckner, P., Berg, J., 1996.
Although far from widespread acceptance, our results clearly demonstrate the potential “win/win” scenario for sheep and crop producers from the incorporating of sheep into farming systems. With weed and pest control and fallow management representing the largest variable cost to farming operations, our results show great promise for using grazing sheep to control weeds and insect pests and potentially reducing fuel and chemical costs associated with fallow management. We do not expect or desire widespread acceptance by farmers at this time. Rather, more research is needed to refine our recommendations to ensure proper application and positive results. We are however, extremely optimistic that we can help reduce crop farming variable costs while offering opportunities to sheep producers, not solely as meat and fiber producers, but also as environmentally sound landscape managers.
Research Outcomes
Education and Outreach
Participation Summary:
MANUSCRIPTS IN PEER REVIEW
Hatfield, P. G., S. L. Blodgett, T. M. Spezzano, H. B. Goosey, A. W. Lenssen, and R. W. Kott. 200_ Incorporating sheep into dryland grain production systems: I Impact on overwintering larva populations of wheat stem sawfly, Cephus cintus Norton, (Hymenoptera: Cephidae).
Hatfield, P. G., A. W. Lenssen, T. M. Spezzano, S. L. Blodgett, H. B. Goosey, and R. W. Kott. 200_. Incorporating sheep into dryland grain production systems: II Impact on changes in biomass and weed frequency.
Hatfield, P. G., H. B. Goosey, T. M. Spezzano, S. L. Blodgett, A. W. Lenssen, and R. W. Kott. 200_. Incorporating sheep into dryland grain production systems: III Impact on changes in soil bulk density and soil nutrient profiles.
Goosey, H. B., P. G. Hatfield, A. W. Lenssen, S. L. Blodgett, R. W. Kott, and T. M. Spezzano. 200_. Incorporating sheep into dryland grain production systems: IV. A field size approach
PEER REVIEWED JOURNAL PUBLICATIONS
Goosey, H. B., P. G. Hatfield, S. L. Blodgett, and S. D. Cash. Evaluation of Alfalfa Weevil (Coleoptera: Curculionidae) Densities and Regrowth Characteristics of Alfalfa Grazed by Sheep in Winter and Spring. J. of Entomol. Sci. 39:598-610.
Griffith, D., P. G. Hatfield, and R. W. Kott. 2001. Enterprise budgeting for ewe flock operations. Sheep and Goat Res. J. 17:29.
PROCEEDINGS
Spezzano, T. M. H. B. Goosey, P. G. Hatfield, S. L. Blodgett, S. D. Cash, P. M. Denke, R. W. Kott, A. W. Lenssen, and C. B. Marlow. 2003. Managing Insect Pest Damage to Wheat and Alfalfa by Integrating Sheep into Crop Production. Proc. Montana Livestock Nutr. Conf. 52:
Goosey H. B., T. M. Spezzano, P. G. Hatfield, S. L. Blodgett, P. M. Denke, and R. W. Kott. 2002, Using sheep in grain production systems to reduce pesticide use: I. Control of Wheat Stem Sawfly infestation in wheat stubble. West. Sec. Amer. Soc. Anim. Sci. 53.197.
Spezzano, T. M., H. B. Goosey , P. G. Hatfield, S. L. Blodgett, P. M. Denke, and R. W. Kott. 2002, Using sheep in grain production systems to reduce pesticide use: II.Comparing stubble grazing with tillage and burning on weed and soil characteristics. West. Sec. Amer. Soc. Anim. Sci. 53.208.
POPULAR PRESS
Goosey, H., P. Hatfield, S. Blodgett, and D. Cash. “Sheep grazing to control alfalfa weevil” 2003. Montana Woolgrower bulletin, spring issue 2003 page 9.
Hatfield et al. “Using sheep for fallow management” 2004 Montana Woolgrower bulletin, fall issue 2004.
Flaherty, C. “MSU research uses sheep to take bite out of sawfly problem”. 1999.
Billings Gazette, Tri-State Livestock News, Fairfield, Wibaux, Poplar, Glendive
Big Sandy, Chinook, Superior, Conrad, The Prairie Star, Montana Wool Growers
The benefits of incorporating strategic sheep grazing into dryland grain production has attracted producer attention and received coverage in agriculture and popular press media. See the following site for an example article. http://www.montana.edu/commserv/csnews/nwview.php?article=1945
Two presentations were given at each of the following sites. One by Terri Spezzano (graduate student) and the other by Hayes Goosey (research associate). All of these presentations were focused on our research of incorporating sheep into farming systems: Cabin Fever production seminar Havre MT 35 in attendance, Golden Triangle, production seminar Ft. Benton MT, 53 in attendance, Cropping seminars in Chester, Shelby, Cutbank, and Choteau MT, 120 in attendance, MAGIE Great Falls MT 45 in attendance. Hayes has also gave numerous high school presentations
INVITED SPEAKER
Gobierno De Chile, Instituto De Investigaciones Agropecuarias, Kampenaike “Western U.S. Sheep Production Systems” 2001.
Montana Wool Growers Association 120th Annual Convention. “Montana State University Incorporating sheep into dry land grain production.” 2003.
Education and Outreach Outcomes
Areas needing additional study
Listed below are potential objective for ongoing and future research.
Compare strategically managed sheep grazing to chemical and mechanical fallow on: a) weed and volunteer grain biomass reduction, b) soil nutrient cycling, and compaction, and c) grain production and grain quality.
Compare profitability and cash flow of grazing vs traditional fallow systems.
Conduct "on campus" field demonstrations. Publish results and communicate to both the production and scientific communities the results and potential economic and biological advantages of incorporating sheep grazing of stubble and weeds into sustainable grain farming systems.
Develop two computer-based decision-support tools for grain and sheep producers. Programs will be capable of handling the inherent variations among different grain and sheep production systems to evaluate potential cash flow and profitability for a single unique enterprise.
Conduct commercial-scale field trials in major grain-producing regions of ND, SD, and MT to evaluate differences between non-grazed and grazed grain residue on: a) biomass (weed, residue, and volunteer grain) reduction, b) pest insect occurrence, and c) soil compaction. Also monitor animal body weight and condition changes during field trials.
On commercial grain farms in three different grain-producing regions of Montana we will compare: 1) a no-input control, 2) fall and spring wheat stubble grazed by sheep to NRCS maximum allowable biomass removal, 3) Winter feeding of harvested feeds on wheat stubble, and 4) a combination of treatments 2 and 3 on a) overwintering WSS larva mortality, b)weed and volunteer grain biomass reduction, c) soil bulk density, and d) change in grazing sheep body weight and body condition.
Determine the impact of weather conditions and stubble disruption on timing and magnitude of WSS larva mortality in: field conditions and in the laboratory.
Determine the impact of stubble grazing by sheep on WSS predatory insect populations and the combined impact of sheep grazing and predatory insect populations on WSS larva mortality. If synergistic impact of grazing and predation on WSS larva exists, define the mechanism of this interaction.