| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
North Dakota State Univ., Dickinson Res. Ext. Cent., 1133 State Ave., Dickinson, ND 58601
* Corresponding author (pcarr{at}ndsuext.nodak.edu)
Received for publication January 5, 2005.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: CP, crude protein • DM, dry matter
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Integrated crop–livestock systems are not common in the region. Krall and Schuman (1996) estimated that <10% of agricultural land is dedicated to integrated systems. The two researchers concluded that the inability or unwillingness to integrate crop and livestock enterprises prevents benefits in environmental quality, economic diversity, and pest management from occurring. For example, wheat yield enhancements and weed pressure reductions occurred when forages were incorporated into rotations with wheat and other grain crops in the Canadian prairie region (Entz et al., 1995). Additional benefits and synergies between crop and livestock enterprises can occur when grazed pasture is rotated with wheat or other grain crops (Entz et al., 2002). A review by Martin (1996) suggested that grassy weed invasion was reduced by rotating grazed legume forages with wheat compared with cropping systems that did not include pasture.
Integrated crop–livestock systems have been developed in response to the environmental degradation and poor economic returns that occur when growing wheat in regions outside of the Great Plains. For example, wheat is rotated with clover (Trifolium spp.) and medic (Medicago spp.) pastures in Australian ley farming in an integrated crop–livestock approach that provides flexibility and diversity to dryland agriculturists (Carter et al., 1982). The annual clovers and medics regenerate from the soil seed bank and provide forage during the pasture phase for sheep (Ovis aries L.) and cattle (Bos taurus L.) as well as fix N biologically for subsequent use during the wheat phase. Ley farming comprised over 20 million ha in the wheat–sheep zone of southern Australia by the mid-1980s, as the benefits of this system compared with traditional wheat production systems became evident. These benefits included more profitable wheat production (Boyce et al., 1991), wheat pest suppression (Loomis and Connor, 1992), reduced fertilizer inputs along with improved air and water quality (Grierson et al., 1991), production of high quality forage (Mann, 1991), and soil conservation (Cocks et al., 1980).
Efforts to export ley farming to regions outside of Australia generally have been unsuccessful. The inability to identify legume species and cultivars that are adapted to local environments explains the failure of adopting ley farming, in some instances (Springborg, 1986; Halse, 1993). Koala (1982) concluded that cultivars of several Medicago spp. and subterranean clover (Trifolium subterranean L.) used in Australia were less adapted to growing conditions in Montana than black medic (Medicago lupulina L.), a naturalized legume species that occurs throughout the northern Great Plains. Subsequent studies suggested that the suitability of black medic for pasture may be limited in the region (Walsh et al., 2001; Carr et al., 2005). However, other legumes showed promise as regenerating pasture species in these two studies.
Important characteristics of legume species used in Australian ley farming include the ability to regenerate from the soil seed bank following the wheat phase and to produce adequate amounts of forage to support grazing livestock (Puckridge and French, 1983). A minimum population of 200 medic seedlings m–2 that regenerated from the soil seed bank was considered to be necessary to maintain a productive legume pasture (Walsh et al., 2001). Research in Wyoming and North Dakota identified legumes capable of regenerating more than 200 seedlings m–2 and producing forage dry matter (DM) in amounts exceeding 7 Mg ha–1 (Walsh et al., 2001; Carr et al., 2005). However, the ability of legumes to reseed naturally when growth was interrupted by a wheat phase was not considered in either study. Four of 16 annual legume cultivars regenerated from the soil seed bank following spring wheat in a field experiment in Montana (Koala, 1982), but forage DM production was not determined.
The ability of legumes to regenerate from the soil seed bank and produce forage following wheat in a pasture–wheat system has not been demonstrated in the northern Great Plains. Our objectives were to determine: (i) the impact that legume forages have on grain yield in a legume–wheat sequence and (ii) the ability of legume forages to regenerate from the soil seed bank and produce forage following spring wheat.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Herbicides were used to control legume growth before seeding spring wheat the third production year following the 2-yr legume phase in each experiment. In 2001, a pre-emergent spring application of glyphosate [N-(phosphonomethyl)glycine] at 0.64 kg a.i. ha–1 was used in the experiment established in 1999. Clopyralid (3,6-dichloro-2-pyridinecarboxylic acid) at 1 kg a.i. ha–1 and 2,4-D amine [(2,4-dichlorophenoxy)acetic acid] at 0.56 kg a.i. ha–1 were applied in the fall in 2001 in the experiment established in 2000. Clopyralid and 2,4-D amine in the fall in 2002 followed with glyphosate in the spring before seeding in 2003 were applied at rates used in the other two experiments in the experiment that was established in 2001.
Hard red spring wheat was sown at 296 live kernels m–2 on 26 Apr. 2001, 17 May 2002, and 19 May 2003 during the wheat phase in the third year of each experiment using a commercial, low-disturbance grain drill. Wheat seedlings were counted in a 0.5-m2 area at two random locations in each plot at 14 to 21 d after seeding. No in-crop applications of herbicides occurred during the wheat phase in the experiment established in 2001, but bromoxynil (3,5-dibromo-4-hydroxybenzonitrile) at 0.28 kg a.i. ha–1 in 2002 and thifensulfuron {3-[[[[(4-methoxy-6-methyl-1,3,5-triazin-2-yl)amino]carbonyl]amino]sulfonyl]-2-thiophenecarboxylic acid} at 0.02 kg a.i. ha–1 along with fenaxaprop-P {[2R]-2-[4-[(6-chloro-2-benzoxazolyl)oxy]phenoxy] propanoic acid} plus mefenpyr [1-(2,4-dichlorophenyl)-4,5-dihydro-5-methyl-1H-pyrazole-3,5-dicarboxylic acid] at 0.09 kg a.i. ha–1 in 2003 were applied during the wheat phase in the other two experiments.
Wheat grain was harvested at maturity (Zadoks Growth Stage 92; Zadoks et al., 1974) from a 26-m2 area in the center of each 56-m2 plot in 2001 and from an 11.5-m2 area in the center of each 25-m2 plot in 2002 and 2003 using a small-plot combine (Kincade Equip., Haven, KS).1 Grain test weight and 100-kernel weight were determined from subsamples. Crude protein (CP) concentration was determined for subsamples of grain harvested in 2001 and 2003 by near infrared spectroscopy (Infratec grain analyzer, UAS Service Corp., Hawley, MN).1
An early- to mid-May application of imazethapyr {2-[4,5-dihydro-4-methyl-4-(1-methylethyl)-5-oxo-1H-imidazol-2-yl]-5-ethyl-3-pyridinecarboxylic acid} at 0.03 kg a.i. ha–1 along with some hand weeding was used to control weeds following the wheat phase in the fourth year of the field experiments. The ability of legume species to regenerate from the soil seed bank was determined using the procedure described by Carr et al. (2005). Briefly, legume seedlings with opened cotyledons (BBCH Growth Stage 09; Lancashire et al., 1991) were counted in two 0.25-m2 areas in each plot following spring warm up (mid- to late April). Seedlings were counted approximately 15 and 30 d later, but only those with opened cotyledons having less than two trifoliolate leaves to avoid recounting seedlings. The total number of legume seedlings was determined by adding together the number of seedlings that were counted on all three dates.
Forage yields were determined by harvesting plants at early flowering (BBCH Growth Stage 60–61) in a 3.7-m2 area to a 6-cm height with a forage plot harvester (Swift Machine & Welding Ltd., Swift Current, SK).1 A 400-g subsample from the harvested material was dried from 3 to 8 d at 50°C until a constant weight to determine moisture content. Forage yield was reported on a DM basis.
Legume treatments were in three randomized complete blocks. The three experiments were analyzed separately because some treatments were not included in all experiments and because management differed across years and experiments. The ANOVA procedure from SAS (SAS Inst., 1985) was used in the analyses with legume treatments considered fixed and blocks considered random effects. The Ryan–Einot–Gariel–Welsch multiple range test was used to reduce the likelihood of making a Type II error when the F test indicated that a significant difference existed between treatment means at the P < 0.05 level.
A weather service station within 1 km of the field experiments indicated that overwinter precipitation (September through March) was near the 30-yr average of 141 mm before seeding spring wheat in both 2001 and 2003 (Fig. 1a) . Overwinter precipitation was 86 and 136% of the 30-yr average in 2002 and 2004, respectively. Growing season (April through August) precipitation was 134% of the 30-yr average of 284 mm in 2001, 126% in 2002, 77% in 2003, and 56% in 2004. Overall, greater-than-average amounts of precipitation occurred in 2001 and 2002 while lower-than-average amounts occurred in 2003 and 2004. Particularly dry conditions occurred during July in 2003 and during June and August in 2004 (Fig. 1b). Average air temperatures during the growing season were within 1°C of the 30-yr average of 15°C each year.
|
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Wheat yield was equal or superior following yellow-flowered sweetclover compared with the other legume treatments in 2003, except for fenugreek (Trigonella foenum-graecum L.; Table 1). Plots originally seeded with fenugreek were fallowed in 2002 because natural reseeding did not occur. Wheat yield averaged 2535 kg ha–1 following the fallowed fenugreek and was similar to yield in plots of six other annual legumes {Austrian winter pea, black lentil (Lens culinaris Medik.), chickling vetch (Lathyrus sativus L.), rigid medic [Medicago rigidula (L.) Allioni], roughpea (Lathyrus hirsutus L.), and woolypod vetch [Vicia villosa ssp. dasycarpa (Ten.) Cav.]} that also were fallowed because few legume plants regenerated from the soil seed bank in 2002. Soil water depletion is greater following yellow-flowered sweetclover compared with fallow (Badaruddin and Meyer, 1989), so less stored soil water probably was available in sweetclover plots. This may have depressed wheat yield since conditions during July in 2003 were dry (Fig. 1b).
The CP concentration of wheat grain averaged 135 g kg–1 in 2001 and 159 g kg–1 in 2003. Differences in grain CP concentration were not detected following legume treatments in either year. In contrast, test weight of wheat grain was comparable or heavier from plots established previously with yellow-flowered sweetclover compared with those from other legume treatments in 2001 and 2002 (Table 1). Grain test weight was lighter in 2003 following yellow-flowered sweetclover than in fallowed plots that originally were seeded with arrowleaf clover (Trifolium vesiculosum Savi), Austrian winter pea, fenugreek, rigid medic, and roughpea. Similarly, light grain test weights resulted when wheat was grown in birdsfoot trefoil plots compared with plots originally seeded with these five legume species in 2003 and with Austrian winter pea in 2001. No differences in grain test weight were detected when wheat was grown following birdsfoot trefoil and other legume treatments in 2002.
Differences in kernel weight were not detected when wheat was grown following legume treatments in 2001 and 2002 (Table 1). Heavier kernels were produced in 2003 when wheat followed fenugreek and other legume treatments that were fallowed in 2002 compared with kernel weights produced in plots where forage legumes reseeded naturally or persisted the previous year. Large kernel size probably was favored in 2003 in plots that were fallowed the previous year since grain fill occurred during July in 2003 when only 24 mm of precipitation was received (Fig. 1b).
Forage Legume Regeneration and Forage Yield following Wheat
Over 200 seedlings m–2 regenerated from the soil seed bank for eight legume species in at least 1 of the 3 yr following the wheat phase (Table 2). A minimum of 200 seedlings m–2 is considered necessary to maintain productive legume pasture in Australian ley farming (Walsh et al., 2001). Regenerating seedling numbers were equal or greater for black medic compared with other legume species and ranged from around 550 to 4375 m–2, depending on the year. Comparable numbers of seedlings regenerated from the soil seed bank for yellow-flowered sweetclover in 2003. Fewer seedlings occurred for yellow-flowered sweetclover than black medic in 2004, but over 3000 seedlings m–2 occurred for both species in that year. Only around 170 yellow-flowered sweetclover seedlings m–2 occurred in 2002. Fewer seedlings occurred in birdsfoot trefoil than black medic plots in all 3 yr, but seedling numbers always exceeded 200 seedlings m–2 for both species. More than 200 seedlings m–2 also occurred for alsike clover (Trifolium hybridum L.), kura clover (Trifolium ambiguum Bieb.), red clover (Trifolium pratense L.), and white-flowered sweetclover (Melilotus alba Medik.) in 2 yr and in 1 yr for white clover (Trifolium repens L.).
|
Total forage DM production exceeded 3 Mg ha–1 for birdsfoot trefoil following the wheat phase in 2002 and 2003. Forage DM production was less for yellow-flowered sweetclover than birdsfoot trefoil in 2002. However, differences were not detected between the two legume species for forage DM yield in 2003. Likewise, white-flowered sweetclover produced comparable amounts of forage DM compared with birdsfoot trefoil in 2003. White-flowered sweetclover was not included with birdsfoot trefoil and yellow-flowered sweetclover in the experiment that was harvested for forage DM in 2002. Dry conditions in 2004 (Fig. 1b) prevented the three legume species from producing more than 0.5 Mg ha–1 of forage DM.
Forage DM production was 2.5 Mg ha–1 for black medic in 2002. However, black medic produced <500 kg ha–1 of forage DM in both 2003 and 2004. Likewise, alsike clover produced only 500 kg ha–1 of forage DM in 2003 and even less in 2004. Forage production was <500 kg ha–1 both years that kura clover was grown. The inability of the two clover species and black medic to produce adequate amounts of forage DM to support grazing suggests poor potential as regenerating pasture species in a legume–wheat rotation even though the three species reseeded naturally following the wheat phase (Table 2).
Red clover produced <2 Mg ha–1 of forage DM following wheat in 2003. The amount of forage produced in 2004 was so limited that yield was not determined because forage could not be harvested mechanically. The red clover treatment was not included in the experiment harvested in 2002. Red clover regenerated from the soil seed bank following wheat in both years that the legume treatment was included in field experiments (Table 2). The clover species also had been identified as a potential pasture crop in previous research (Carr et al., 2005). However, red clover probably is not adapted as a regenerating pasture species in legume–wheat rotations because of limited forage yield potential following the cereal phase.
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Results of this investigation indicate that both white- and yellow-flowered sweetclover can regenerate from the soil seed bank following wheat in a legume–wheat rotation. The biennial growth habit of sweetclover may require a 2-yr legume phase when rotated with wheat since forage DM and seed production could be limited in the first year. However, observations by Stoa (1941) indicated that a 2-yr sweetclover stand can produce enough seed to support regeneration from the soil seed bank for several years. Results of the present study suggest that forage DM in the first year may be sufficient to support grazing by livestock in some years although careful management is needed to prevent bloat and other grazing concerns (Helm and Meyer, 1993). The potential of biennial sweetclover in a pasture–wheat rotation should be investigated more thoroughly. Annual sweetclover exists, and work also is needed to determine if annual sweetclover is better suited to ley farming than biennial sweetclover.
Balansa clover, berseem clover (Trifolium alexandrinum L.), black medic, burr medic (Medicago polymorpha L.), crimson clover (T. incarnatum L.), and Persian clover (T. resupinatum L.) were identified as regenerating legume species in previously reported research (Carr et al., 2005). Only black medic regenerated from the soil seed bank following wheat in the present study, and <1 Mg DM ha–1 was produced by black medic in 2 of 3 yr. These six legume species seem to have limited potential as regenerating legume species in a pasture–wheat rotation in western North Dakota and similar climatic regions.
The ability of legume species to maintain or enhance wheat yield in a pasture–wheat rotation compared with fallow–wheat or more diverse crop rotations must be demonstrated before ley farming will generate much interest among wheat producers in the northern Great Plains. Spring wheat yields never were elevated and sometimes were depressed following birdsfoot trefoil in our study. Yield depression may have occurred because soil water reserves were depleted during the 2-yr legume phase proceeding the wheat phase, as expected. Forage legumes depleted soil water reserves compared with fallow in rotations with wheat in eastern North Dakota (Badaruddin and Meyer, 1989), particularly when legume growth was unrestricted. Legume growth was not controlled until the fall of the second year in our study because we believed that terminating growth earlier could reduce seed production and subsequently the soil seed bank. Work is needed to determine when legume species must be terminated so soil water recharge can occur while also allowing adequate seed production to maintain the soil seed bank. Herbicide treatments sometimes failed to kill established birdsfoot trefoil, kura clover, and red clover plants before seeding wheat in this study. Wheat growth and grain yield were suppressed in plots where persisting legume plants occurred. Alfalfa was controlled by applications of clopyralid, glyphosate, and other herbicides before seeding wheat in a subhumid region (Bullied et al., 1999), but these same treatments did not control alfalfa consistently in western North Dakota (Carr, unpublished data, 2004). The efficacy of many herbicides declines when hot and dry conditions develop, as commonly occurs in late summer in much of the northern Great Plains. Research is needed to identify the herbicides and the timing of application that result in consistent termination of birdsfoot trefoil and other persisting legumes in western North Dakota and similar semiarid regions.
A 2-yr legume phase preceded the wheat phase in our study so that species with an ability to regenerate from the soil seed bank could be identified. We hypothesize that the soil seed bank created by the initial 2-yr phase could be maintained by subsequently rotating legume pasture on an annual basis with wheat in alternate years. Birdsfoot trefoil and sweetclover are tap rooted (McGraw and Nelson, 2003), and limiting stand persistence to a single growing season could reduce root development and soil water depletion compared with soil water removal when 2-yr legume stands are maintained. However, the ability of a single year of birdsfoot trefoil and possibly annual sweetclover to produce adequate seed to maintain the soil seed bank and forage to support grazing has not been determined. Additional research is necessary to determine if soil water is conserved and adequate forage DM and seed production is maintained when birdsfoot trefoil and annual sweetclover pastures are limited to a single growing season.
Crop–fallow is being replaced by rotations that include various crops in addition to wheat across the northern Great Plains. The increased complexity of cropping systems suggests that rotating legume pasture with wheat in alternate years may not be attractive to commercial producers interested in expanding grain and seed crop choices. Research is needed to determine if birdsfoot trefoil and other legume species can regenerate from the soil seed bank following other cereal crops besides wheat and particularly broadleaf crops. Specifically, work is needed to determine the maximum interval between successive pasture phases that still allows regeneration of forage legume seedlings from the soil seed bank in diverse rotations.
An important determinant in the success at regenerating legume pasture in ley farming is the amount of crop residue that remains on the soil surface following a wheat crop. Research in Australia indicates that density of regenerated medic seedlings was inversely related to surface residue cover and that crop residue in excess of 4 Mg ha–1 creates a surface mulch that may inhibit natural reseeding by medic seedlings (Reeves, 1987). Wheat residue production sometimes exceeded 4 Mg ha–1 following selected legume treatments in 2001 and 2003 in our study, based on crop residue/grain ratios for spring wheat reported by Engel et al. (2003). A possible strategy to reduce crop residue on the soil surface is to allow ruminants to graze crop residue following wheat harvest. Grazing crop aftermath would not only result in a decrease in wheat stubble through consumption by livestock but would also bury some residue through the hoof action of animals. Shallow burial of legume seeds produced during the previous pasture phase might also occur. Research is needed to determine if grazing crop aftermath following wheat harvest improves regeneration of legume pasture from the soil seed bank in no-tillage systems and if postharvest grazing contributes to soil compaction.
The grazing of legume pasture is an important component contributing to the economic and environmental sustainability of ley farming. Birdsfoot trefoil is nonbloating and produces high quality forage. However, no cropping system investigation has incorporated the grazing of birdsfoot trefoil or other legume species in the northern Great Plains. Unfortunately, many of the benefits attributed to ley farming occur because legume species are grazed and not hayed or plowed under as green manures. The grazing of legume species during the pasture phase should be included in future studies of ley farming to learn if the benefits resulting from this farming method in Australia can be transferred to the northern Great Plains region of North America.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Spring wheat generally cannot be grown profitably in western North Dakota without government subsidy payments. A majority of the wheat presently is grown in a continuous monoculture, and this production system requires relatively high inputs of N fertilizer. Incorporating grazed legume pasture into rotations with wheat by ley farming should reduce or even eliminate the need for N fertilizer during the cereal grain phase. Wheat production could become more sustainable economically by ley farming compared with continuous wheat monoculture if grain yield was maintained but N fertilizer inputs were reduced.
Crop–fallow once dominated wheat production methods in the northern Great Plains, and preceding spring wheat with fallow still is a popular practice. Bare soil is susceptible to both water and wind erosion, and sometimes even untilled soils can erode because of the limited amount of crop residue that remains after grain harvest. Incorporating legume forages into rotations with wheat provides vegetative cover for the soil surface during the pasture phase, thereby reducing or eliminating soil erosion and enhancing the environmental sustainability of wheat production systems in the region.
The growing popularity of reduced- and no-till systems along with the development of diverse rotations are important in sustaining crop production viability in western North Dakota and throughout the northern Great Plains. Many emerging strategies and technologies for cropping systems rely heavily on fossil fuels to power planting and harvesting equipment. Costs of fossil fuels continually increase, and less expensive alternative fuels or decreases in consumption are needed to counter the price increases. Grazing regenerating pasture is an alternative to harvesting forages mechanically that should reduce fossil fuel use directly by eliminating mechanized planting and harvesting operations. Incorporating legume pasture into rotations with wheat also may reduce fertilizer N needs in the cereal crop, thereby reducing natural gas consumption that occurs when manufacturing most N fertilizers.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
1 Mention of a proprietary product name is for identification purposes only and does not imply endorsement or warranty to the exclusion of other products. 
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
