Agronomy Journal 95:520-525 (2003)
© 2003 American Society of Agronomy
PRODUCTION PAPERS
Intercropping Irrigated Corn with Annual Legumes for Fall Forage in the High Plains
Craig M. Alford*,
James M. Krall and
Stephen D. Miller
Dep. of Plant Sci., P.O. Box 3354, Laramie, WY 82071
* Corresponding author (cmalford{at}uwyo.edu)
Received for publication February 15, 2001.
 |
ABSTRACT
|
|---|
Many farmers in the central High Plains graze corn (Zea mays L.) aftermath and are looking at ways to improve the quality and amount of this fall pasture resource. However, no information is available on intercropping annual legumes with irrigated corn in the region. Our objective was to determine the most appropriate pasture legume species that could be used for an irrigated corn–legume system. To accomplish this, field experiments were conducted under irrigation at four sites in Wyoming. Eight legume species planted with corn were compared with monoculture corn under weed-free and weedy conditions. Under weed-free conditions, corn grain yields were reduced by the presence of legume in some treatments while others were comparable to the check yields. Black medic (Medicago lupulina L.) did not reduce corn yields, but barrel medic (M. truncatula Gaertn.) and sphere medic (M. sphaerocarpus L.) reduced corn yields by 17%. Corn yields were reduced 62% by the presence of weeds regardless of legume species. Legumes did not suppress weed growth. Barrel medic produced the most forage in July; however, there was little difference among species by November. In November, when corn stalks would be grazed, black medic and yellow sweetclover (Melitotus officinalis Lam.) produced the highest quality forage. These results indicate that a mostly weed-free field is required to maximize corn and legume production. Of the legume species evaluated, black medic appears to offer the greatest potential for intercropping with irrigated corn in the central High Plains.
Abbreviations: ADF, acid detergent fiber • CP, crude protein • NDF, neutral detergent fiber • RFV, relative feed value • TREC, University of Wyoming Research and Extension Center at Torrington
|
INTRODUCTION
|
|---|
MANY OF THE FARMERS within the irrigated regions of the central High Plains graze corn aftermath. Intercropping with a forage legume could improve the quality and amount of this fall pasture resource. A study in Minnesota found that farmers generally perceived companion-cropping systems to be highly desirable due to increased forage production, reduced soil erosion, and weed suppression (Simmons et al., 1992). Recently, annual medics (Medicago spp.) have been evaluated in the midwestern USA as summer annual forages and as intercrops with small grains and corn (Zhu et al., 1996; Moynihan et al., 1996; De Haan, 1995; Zhu and Sheaffer, 1997). There was little reduction in grain yield when annual medics were intercropped with corn (De Haan, 1995). Additionally, annual medics intercropped with barley (Hordeum vulgare L.) in Minnesota reduced weed biomass 65% (Moynihan et al., 1996). Forage yields were increased and soil erosion reduced 90% when alfalfa (Medicago sativa L.) was intercropped with sorghum [Sorghum bicolor (L.) Moench] compared with a monoculture of sorghum (Hardin, 1996). These findings under rainfed conditions led us to believe that legumes intercropped with irrigated corn offer potential to improve the productivity and quality of fall grazing aftermath.
Annual medics are comparable to alfalfa in dry matter yield and forage quality (Zhu et al., 1996). When grown in monoculture in Minnesota, annual medic yields ranged from 0.5 to 5.7 Mg ha-1, depending on harvest time and species, and had crude protein (CP) levels equal to or higher than those of alfalfa. In the same study, snail medic [Medicago scutellata (L.) Mill.] had the highest acid detergent fiber (ADF) and neutral detergent fiber (NDF) while black medic and ‘Harbinger’ strand medic (Medicago littoralis Rohde ex Loisel.) had the highest CP concentrations. There was also evidence that forage quality can be quite variable across locations. However, there is little data available for annual legume production or forage quality in an intercropped irrigated corn system.
The objectives of this study were to (i) evaluate the effects of eight intercropped legume species on irrigated corn grain yield, weed populations, and weed growth; (ii) determine the effect of weeds and corn on legume production; and (iii) compare midseason forage quality of the eight legume species to that after corn grain harvest.
|
MATERIALS AND METHODS
|
|---|
Field experiments were conducted at four locations in Goshen County, WY, during 1996. Two experiments were at the University of Wyoming Research and Extension Center at Torrington (TREC), elevation 1249 m, on a Valentine sandy loam (mixed mesic, Ustic, Torripsamment) soil. Two experiments were established on producer fields, one at Lingle, WY, elevation 1271 m, on a Hargrave loamy sand (fine-loamy sandy, mixed, calcareous, mesic, Ustic Torrifluvents) and the other at Huntley, WY, elevation 1289 m, on a Mitchell sandy clay loam (coarse-silty, mixed, calcareous, mesic, Ustic Torriorthents). The environmental conditions for each site are described in Fig. 1
. Soils were sampled at each location and received the appropriate amounts of N and P for a yield goal of 11.3 Mg ha-1 corn grain. No adjustments were made for any potential N from the legumes. Legumes, weeds, and corn were seeded on 2, 10, 15, and 16 May at TREC 1, Lingle, Huntley, and TREC 2, respectively. All study sites were overhead sprinkler–irrigated with the exception of Lingle, which was furrow-irrigated. Irrigation commenced at the sprinkler sites immediately after planting and at the Lingle location after cultivation and ditching (4 July), approximately 7 wk after planting. Irrigation was applied every 7 to 10 d through the growing season, totaling 600 mm at TREC 1 and 2, 500 mm at Huntley, and 400 mm at Lingle.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 1. Temperature and precipitation data for year of study and 30-yr average at the three study sites.
|
|
The legume treatments were seven Medicago spp., including six annual and one perennial species and one biennial Melilotus sp. The legumes (‘George’ black medic, ‘Wrangler’ alfalfa, ‘Yellow Blossom’ yellow sweetclover, gama medic (Medicago rugosa Desr. cv. Paraponto), ‘Sava’ snail medic, ‘Caliph’ barrel medic, burr medic (Medicago polymorpha L. cv. Santiago), and ‘Orion’ sphere medic) were seeded at 172 pure live seeds m-2. Weedy and weed-free treatments were established with each medic species. Each treatment was replicated four times in a split-plot (weedy or weed free) arrangement with a randomized complete block design. Weed management strategy was the whole plot and legume species the subplots. Plot size was 3 by 6 m, with each plot containing four rows of corn. Weeds were seeded in the weedy treatments to ensure adequate weed populations. The weeds seeded were redroot pigweed (Amaranthus retroflexus L.), common lambsquarters (Chenopodium album L.), kochia [Kochia scoparia (L) Schrad.], longspine sandbur [Cenchrus longispinus (Hack.) Fern.], and yellow foxtail [Setaria glauca (L.) Beauv.]. The seeds from these weeds were mixed in appropriate proportions to give a combined weed seeding rate of 250 pure live seeds m-2. Weeds that were already present in the plots were also allowed to grow. Legume and weed seed were broadcast over the plots before corn planting and incorporated (<2 cm) using a Lely vertical tine tiller. A weedy, medic-free plot plus a weed- and medic-free plot were included in each replicate as basis for comparison.
Following incorporation, ‘Pioneer 3751 IR’ (imidazoline resistant) corn was planted in 76-cm rows at the rate of 120 000 seeds ha-1, except at the TREC 2 location, which was seeded at 79 000 seeds ha-1, a more normal seeding rate for the area, using a John Deere Maxemerge vacuum planter (John Deere, Moline, IL). Weed-free treatments received one application of imazethapyr {(±)-2-[4,5-dihydro-4-methyl-4(methylethyl)-5-oxo-1H-imidazol-2-yl]-5-ethyl-3-pyridine carboxylic acid} at 70 g a.i. ha-1, pendimethalin [N-(1-ethylpropyl)-3,4-dimethyl-2,6-dinitrobenzenamine] at 898 g a.i. ha-1, and nonionic surfactant at 0.25% (v/v) to control weeds.
Legume and weed counts were made approximately one month after planting at all locations. Four counts were made at random locations in each plot, using a 0.25-m2 quadrant. Corn population counts were made on a 1.5-m section of row at two random locations at the time of legume and weed counts. A legume and weed biomass sample was obtained by clipping at ground level at midseason (July), and a second legume biomass sample was obtained at the time of corn harvest (November) using a 0.3- by 1-m section over the corn row at two random locations in each plot. The samples were collected over the row to remain consistent across locations because the Lingle site had no legumes between the rows because of ditching for furrow irrigation. The weed and legume samples were dried for 48 h at 60°C and dry weight determined.
The weed-free forage samples from each replicate at the two Torrington sites and the Huntley location were combined and a subsample taken. This sample was then ground through a Wiley Mill and passed through a 2-mm sieve for forage quality analysis. The forage samples were analyzed for CP according to the method described by AOAC (AOAC, 1980). Neutral detergent fiber and ADF were analyzed using methods described by Goering and Van Soest (1970). These values were then used to determine the relative feed value (RFV) for each species. The RFV was calculated using an equation endorsed by the American Forage and Grassland Council (Linn and Martin, 1999; Undersander et al., 1985; Casler, 1990). Legumes from the weedy samples were not included in the forage quality analysis because they were of poor quality, and in many cases, there was insufficient sample to do a quality analysis.
Corn grain yields were determined by harvesting a 5.5-m section of row in each plot using a small-plot combine. Samples were weighed and yield expressed as megagrams per hectare at 15% moisture. Corn yield and weed populations were adjusted according to legume population using covariant analysis with the SAS (SAS Inst., 1988) statistical program. The adjustment for legume population was made because the populations were not uniform across species and it was desirable to give all species equal weight (Table 1).
Legume production results for the Lingle location are not reported because ditching destroyed much of the initial legume population and resulted in variable yield data. The remaining data were transformed using a square root transformation to eliminate problems with nonconstant variance in the data.
|
RESULTS AND DISCUSSION
|
|---|
Corn Production
Weed management significantly (P < 0.0001) influenced corn yield. Corn grain yields were reduced 43 to 69%, depending on legume species, when weeds were not removed. In a corn–legume intercropping system in Minnesota, weeds reduced corn yields 57% when they were not controlled (De Haan et al., 1997). Due to the large difference in yields between the weedy and weed-free treatments, the data were analyzed separately by weed and weed-free plots.
Within the weed-free treatments, there was a significant (P < 0.0001) location effect on corn grain yield (Table 2); however, there was no species x location interaction (Table 2), so results were combined across locations and are presented by medic species (Fig. 2)
. In weed-free treatments, corn grain yields were significantly (P < 0.05) influenced by legume species (Fig. 2). Only plots intercropped with black medic had yields that were not different (P 
0.05) from the weed- and legume-free check. De Haan et al. (1997) found that black medic did not reduce corn yield compared with the medic-free check when grown in fertility levels corresponding to those in this study. Black medic was found to be the least competitive medic species when it was intercropped with barley (Moynihan et al., 1996). Plots intercropped with the remaining legume species yielded 17 to 35% less than the weed- and legume-free check. Barrel medic provided the lowest corn yields; however, yields were not significantly lower than those of sphere medic, alfalfa, burr medic, and snail medic. In general, corn yields were lowest at Lingle. This is probably related to water stress caused by late initiation of furrow irrigation (4 July) and early termination of irrigation in mid-September. The TREC 1 site produced the highest corn yield and was planted 8 to 14 d earlier than the other three locations. Corn yields averaged 12.89, 11.88, 9.84, and 7.67 Mg/ha at TREC 1, TREC 2, Huntley, and Lingle, respectively. In the weedy treatments, there were no significant yield differences among legume species (P = 0.31).
View this table:
[in this window]
[in a new window]
|
Table 2. P values from ANOVA for corn grain yield from weedy and weed-free plots combined (C), weedy plots (W), and weed-free plots (WF) in Wyoming.
|
|
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2. Influence of legume species on weedy and weed-free irrigated corn grain yield averaged over four locations.
|
|
As with the weed-free treatments, in the weedy plots, there were differences in corn yield between locations (P = 0.001). Across locations, weed biomass was not significantly (P = 0.20) reduced by any legume species at the established density (data not shown). None of the legume species at the established densities significantly (P 0.05) reduced weed populations even though there were large differences in weed populations between locations (Fig. 3)
. Under weedy conditions, yields were highest at the furrow-irrigated Lingle location, which had been ditched. This ditching operation reduced both weed and legume populations. These data indicate that weed competition was excessive and was the main factor that reduced corn grain yields. In contrast to our study, weed biomass was reduced when medics were intercropped with corn in Minnesota (De Haan et al., 1997). However, the medic populations in the Minnesota study were twice as high as those used in this study. A successful weed control system using intercropped annual legumes must be able to suppress weeds under a wide range of weed population pressures. The annual legumes did not suppress weed growth at any of the weed populations present in this study. Therefore, weed pressure probably needs to be removed from the system for this intercropping approach to be successful.
Legume Production
Legume dry matter yield was significantly (P < 0.0001) reduced by the presence of weeds in the treatment (Table 3). Because weeds had a similar impact on corn grain yield, only weed-free production will be discussed. At each harvest time, there were significant differences in forage production between species and locations. Therefore, data for each location were analyzed by harvest date and location. At corn harvest, legume production was 57% higher in the weed-free treatments compared with the weedy treatments (data not shown) when averaged across locations. Results for three locations are presented in Fig. 4
. While there were significant (P < 0.05) differences in early-season (July) dry matter production between species, these differences disappeared by corn harvest (November). Black medic produced more forage at corn harvest than at the midseason harvest date at all three location while yellow sweetclover yielded more forage at harvest than midseason at TREC 1 and Huntley. Huntley was the only location where there were significant differences between the legume species for forage production at the end of the season. When grown in monoculture in Minnesota, snail medic, barrel medic, and burr medic produced the greatest yield, and black medic was one of the lower-yielding species (Zhu et al., 1996; Zhu and Sheaffer, 1997). Overall, medic production was less when medics were intercropped compared with being grown in monoculture, and species production trends were consistent between the two production systems (Zhu et al., 1996; Zhu and Sheaffer, 1997).
View this table:
[in this window]
[in a new window]
|
Table 3. P values from ANOVA for medic dry matter forage yield for two harvest periods from weed and weed-free plots combined (C), weedy plots (W), and weed-free plots (WF) in Wyoming.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4. Forage production of legumes in weed-free treatments harvested in July and November at three locations.
|
|
Forage Quality
Crude protein and RFV of annual legumes were 40 to 45% higher in July compared with November (Tables 4 and 5). The forage quality was most likely better in July than November because the forage was green and actively growing while it was mostly dead and desiccated by November. Because of significant differences between harvest dates, the data were analyzed separately for each harvest date. In July, yellow sweetclover and gama medic had RFVs of 153 and 149, respectively. Forage with an RFV value >151 is considered prime (Linn and Martin, 1999). None of the other species tested had RFV high enough to be considered premium. By July, snail medic and barrel medic had started to dry down and senesce leaves, which is reflected in the low RFV for these species. Crude protein and RFV for each of the species dropped considerably from July to November, as would be expected with maturity. Black medic and yellow sweetclover were the only species that had higher RFV than that of typical corn stalks alone. Usually corn stalks have a RFV of around 80 (Linn and Martin, 1999). These data indicate that Black medic is the most desirable species, with respect to forage quality, late in the season when it would be grazed. Yellow sweetclover was a more desirable species during the growing season. However, based on RFV, there was no significant difference between black medic and yellow sweetclover at either sampling time.
|
CONCLUSIONS
|
|---|
An irrigated corn–legume intercropping system might be a viable option for producers in the Rocky Mountain region. For this system to succeed, it will require a legume species that is not very competitive with corn so that corn yields can be maximized. Further, a mostly weed-free field will be required to maximize corn and legume production because the legume species in this study were not able to adequately suppress weed growth. Of the species evaluated, black medic appears to offer the greatest potential for intercropping with corn in the central High Plains. Black medic did not significantly reduce corn yields compared with the medic and weed-free check, whereas all other species caused significant yield reductions. Black medic also produced acceptable amounts of high quality forage late in the year. Black medic production of more forage later in the year makes this species suited to a production system utilizing fall grazing of corn stalks. It is suited to this system because it does not grow aggressively early in the year when it could reduce corn yield.
|
ACKNOWLEDGMENTS
|
|---|
We acknowledge the Western Region Sustainable Agricultural Research and Education Program (WSARE) for funding this project. Project no. 97-042.
|
NOTES
|
|---|
Contribution no. 1679, Wyoming Agric. Exp. Stn.
|
REFERENCES
|
|---|
- AOAC. 1980. Official methods of analysis. 12th ed. AOAC, Washington, DC.
- Casler, M.D. 1990. Cultivar and cultivar x environment effects on relative feed value of temperate grasses. Crop Sci. 30:722–728.[Abstract/Free Full Text]
- De Haan, R.L. 1995. Evaluation and development of annual Medicago species for integration into corn and small grain systems. Ph.D. thesis, Univ. of Minnesota, St. Paul.
- De Haan, R.L., C.C. Sheaffer, and D.K. Barnes. 1997. Effect of annual medic smother plants on weed control and yield in corn. Agron. J. 89:813–821.[Abstract/Free Full Text]
- Goering, H.K., and P.J. Van Soest. 1970. Forage fiber analysis (apparatus, regents, procedures, and some applications). Agric. Handb. 379. USDA-ARS, Washington, DC.
- Hardin, B. 1996. Inter-cropping, for more forage and less erosion. Agric. Res. 44(3):15.
- Linn, J.G., and N.P. Martin. 1999. Forage quality tests and interpretations. Bull. FO-02367. Univ. of Minnesota Coop. Ext. Serv., St. Paul.
- Moynihan, J.M., S.R. Simmons, and C.C. Sheaffer. 1996. Intercropping annual medic with conventional height and semidwarf barley grown for grain. Agron. J. 88:823–828.[Abstract/Free Full Text]
- SAS Institute. 1988. SAS/STAT user's guide. Release 6.03 ed. SAS Inst., Cary, NC.
- Simmons, S.R., N.P. Martin, C.C. Sheafer, D.D. Stuthman, E.L. Schiefelbein, and T. Haugen. 1992. Companion crop forage establishment: Producer practices and perceptions. J. Prod. Agric. 5:67–72.
- Undersander, D., W.T. Howard, and T. Smith. 1985. Making forage analysis work for you in balancing livestock rations and marketing hay. Bull. A 3325. Univ. of Wisconsin Coop. Ext. Serv., Madison.
- Zhu, Y., and C.C. Sheaffer. 1997. Growth analysis of spring and summer seeded annual Medicago spp. Crop Sci. 37:1514–1519.[Abstract/Free Full Text]
- Zhu, Y., C.C. Sheaffer, and D.K. Barnes. 1996. Forage yield and quality of six annual Medicago species in the north-central USA. Agron. J. 88:955–960.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
D. N. Baributsa, E. F. Foster, K. D. Thelen, A. N. Kravchenko, D. R. Mutch, and M. Ngouajio
Corn and Cover Crop Response to Corn Density in an Interseeding System
Agron. J.,
June 16, 2008;
100(4):
981 - 987.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
