Illinois Fertilizer
Conference Proceedings
January 25-27, 1993
| Home | 1993 Index | Search |
|
|
| Home | 1993 Index | Search |
G. Bollero, D. Bullock, M. Cender, R. Farnsworth, T. Holsapple, T. Hortin1
Environmental and economic concerns have fueled interest in agriculture's economic competitiveness, soil erosion, N fertilizer efficiency, and nitrate leaching (Heichel and Barnes, 1984). We believe that these concerns will result in legislation, at both the federal and state levels, regulating N fertilizer and tillage use for summer feed grain crops in Illinois.
Winter cover crops will be probably required in this legislation. The cover crop will be either a legume which produces significant amounts of organic nitrogen via nitrogen fixation or a small grain which scavenges residual soil nitrates for the production of organic nitrogen and consequently reduces the amount of nitrates available for over-winter leaching.
Many winter cover crops can be used to reduce soil erosion and provide organic N. In the lower Corn Belt hairy vetch (Vicia villosa Roth.) is the preferred legume while rye (Secale cereale) is the preferred non-legume (Ebelhar et al. 1984, Frye et al. 1988). Most research indicates that hairy vetch contributes an equivalent of 60 to 90 pounds N per acre, but resulting yields are not usually economically competitive with corn receiving the normal 125 to 250 pounds fertilizer N per acre (Moschler et al. 1967, Mitchell and Teel 1977, Ebelhar et al. 1984). The reason for the yield reduction is explained as a combination of the large N demand of corn, the necessity of early corn planting dates, and the tendency of hairy vetch to deplete soil moisture. This previous research suggests that new legislation mandating cover crop use in order to reduce soil would have to allow supplemental fertilizer N use if Illinois corn farmers are to remain economically competitive.
Contrary to the published literature many Illinois farmers utilizing cover crops report that hairy vetch can, under the right conditions, produce in excess of 300 lb of N per acre (Holsapple, 1990). The difference may be an artifact of improperly managed cover crops in the published literature. This proposed research will help to resolve some of these questions. Rye has been shown to be inferior generally to hairy vetch in the lower Corn Belt (Ebelhar et al. 1984), but environmental conditions and corn planting date requirements often result in vetch not establishing well and/or not producing a substantial biomass. We speculate that under such conditions rye, which is easier to establish and has greater early season growth, may prove to be comparable to hairy vetch for the purposes described above.
Corn is the major feed grain crop in Illinois and virtually all of the winter cover crop / summer feed grain crop rotation work in this area has looked only at corn, but corn is probably not the best feed grain crop for such a system. Grain sorghum (Sorghum bicolor L.) has a shorter grain-fill period and thus can be planted later than corn without suffering a yield reduction (Olson et al. 1986). Grain sorghum also has a lower N requirement for optimum economic production than corn (80 pounds per acre vs 160 pounds per acre) (Olson et al. 1986) and thus may be a potential alternative summer feed grain. Especially on drought prone soil or if it is found that vetch must be allowed to grow relatively late into the spring in order to fix more organic N. Wagger (1987) reported that killing hairy vetch the first week of May instead of the third week of April resulted in an additional production of 1.78 tons dry matter per acre and 52 lbs N per acre. This is in agreement with Frye et al (1988) who noted that the majority of the vegetative growth of hairy vetch occurred in the three weeks prior to the onset of the reproductive stage.
The later planting date would result in more subsoil moisture depletion by the cover crop, but utilization of grain sorghum instead of corn would allow for a later planting date and thus permit a producer to wait longer for adequate soil moisture in the planting zone after the cover crop has been destroyed. Grain sorghum is also more drought tolerant than corn (Olson et al. 1986). Hargrove (1986) has demonstrated in Georgia a production system using winter legumes followed by grain sorghum is possible. Such a system has not been investigated in the central corn belt and farmers and scientists have little experience to predict the outcome of such a system.
Environment is a large factor in grain crop performance. When soils are deep and fertile and there is ample moisture, corn will produce more feed grain than grain sorghum. But if the soil is shallow, fertilizer inputs are reduced, or moisture is limiting, grain sorghum can yield as much or more feed grain than corn. For example, near Brownstown, Illinois the soils are relatively shallow, have low plant-available water holding capacity, and underlain with a dense acidic clay pan, and are thus very susceptible to drought. In the University of Illinois variety performance trials at Brownstown, grain sorghum often produces substantially more feed grain than corn. This relationship is seen commonly in much of the southern portions of Illinois. These areas also have longer growing seasons and have severe problems with soil erosion.
The question of tillage necessity in a cover crop system has not been answered. Some research has reported that the cover crop should be plowed under in order to enhance N mineralization (Varco 1986) while others report that it is equally effective if the winter cover crop is killed with a herbicide and then the following summer feed grain crop is no-till planted (Tripplet et al 1979; Flannery 1981). The no-till system is attractive because it reduces soil erosion, uses much less fuel, and requires smaller tractors, but it does mandate the use of herbicides. Farmers in Illinois have shown that the plow system can be successful with limited or no herbicides (personal conversation with T. Holsapple, 1989).
The objectives of this field experiment are to address the long term effects
of cover crops on production efficiency, N requirement, environmental consequences,
and economic competitiveness of production systems for a summer feed grain /
soybean rotation. The goal of this research is to identify and demonstrate economically
competitive and environmentally sound production systems for Illinois.
This five-year field experiment is a split-split-split-split plot in a randomized
complete block design with three replications per location and three locations.
The experiment is located at (from South to North) Tom Hortin's farm at Albion,
IL; Terry Holsapple's farm Toledo, IL; and Mark Cender's farm at Fisher, IL.
Whole plot treatments are tillage used the spring prior to feed grain planting:
1) no-till and 2) conventional. The split treatments are type of cover: 1) hairy
vetch, 2) rye, and 3) conventional i.e. no planted cover. The split-split plot
treatments are feed grain: 1) grain sorghum and 2) corn. The split-split-split
plot treatments are feed grain planting date: 1) May 1 and 2) May 20. The split-split-split-split
plot treatment are N fertilizer rate: 1) 0 lb N/ac, 2) 80 lb N/ac, 3) 1601b
N/ac, and 4) 2401b N/ac. Individual plot size are 20 feet wide (8, 30 inch rows)
by 65 feet long.
Soybean grain yield
Plots which were in corn and grain sorghum in 1991 were bulk planted to soybean in 1992. All plots were reconstructed and soybean yields were measured from each plot. Analysis of variance of soybean yield indicated significant differences due to the main effect of cover crop (Table 1) and the interactions of cover crop X date of planting (Table 2), and tillage X feed grain crop (Table 3). Soybean yield was not affected by any other main effect nor interaction.
In 1992 soybean yielded slightly more grain when in rotation with a rye cover crop or a hairy vetch cover crop than when in rotation with fallow (Table 1). There was no significant difference between the two winter cover crops. The advantage due to either crop as compared to fallow is not huge, but is large enough to be of economic interest. Assuming a mean increase of 1.85 bu/acre and bean price of $5.50/bu a gross advantage of $10.18/acre is realized. This contribution in combination with reduction in soil erosion with winter cover crops should be recognized when the economic efficiency of the cropping systems are compared.
While soybean grain yield did increase when in rotation with cover crops it was noted that the date of planting of the feed grain in 1991 interacted with that cover crop effect (Table 2). In particular soybean in rotation with rye and second date of planting for the feed grain yielded significantly less that soybean in rotation with rye and a first date of planting for the feed grain (i.e. 63.3 vs 60.4). None of the other differences in Table 2 are significant. This interaction may have been due to the time of killing the rye crop rather than the actual planting date of the feed grain. Cover crops were killed about a week prior to planting of the feed grain and thus the cover grew about an additional two weeks before it was killed if the feed grain was planted at the second date rather than the first date. This resulted in the rye crop being about 4 feet high and heading rather than 2 feet high.
Soybean yield in 1991 was also affected by the interaction of feed grain crop in 1991 and tillage used in 1991 (Table 3). In particular soybean yielded more when in rotation with conventionaltill grain sorghum than when in rotation with no-till grain sorghum (i.e. 64.5 vs 58.2). There was no difference in soybean yield when grown in rotation with conventional-till corn as compared to no-till corn (i.e. 59.8 vs 61.5). We believe this may have been due to excessive weed pressure. In 1991 we had much greater weed pressure in no-till grain sorghum than in conventional-grain sorghum. There was also a difference for conventional corn as compared to no-till corn, but the difference was not as great presumably, because corn is easier to establish in no-till systems, and also because corn is much taller than grain sorghum and is thus, much more competitive with weeds than is grain sorghum. In the 1992 soybean crop visual ratings indicated that weed pressure was larger for plots which had been in no-till grain sorghum in 1991 than for any other tillage X grain crop combination.
Feed Grain Yield
When locations (i.e. environments) were considered random and an alpha level of 0.05 was used, analysis of variance indicated a significant main effect of cover crop (C) X nitrogen fertilizer rate (N); a significant first order interactions for grain crop (G) X N and C X N, and significant second order interactions for tillage (T) X G X N, and T X C X N. No other main effects nor interactions were significant for grain yield in 1992.
Feed grain yield was significantly affected by the main effect of cover crop (Table 4). In this case the mean feed grain yield was significantly less when the feed grain was grown in rotation with rye than either fallow or hairy vetch. Note there is no significant difference between hairy vetch and fallow at the alpha =0.10 level. Hairy vetch was significantly different from fallow at approximately alpha= 0.13.
For the T X G X N means (Table 5) there was a significant quadratic effect of N on yield for all feed grain - tillage combinations. Thus, the interaction is due to the magnitude of the response rather than shape (i.e. linear vs quadratic) or direction (i.e. negative vs positive slope). From the means it can be seen that the differences in magnitude at the level of the second order interaction are not great. Tillage within a given feed grain crop did not dramatically effect the relationship between N fertilizer rate and grain yield. Rather the main difference appears to be at the G X N first order interaction level. This is supported by examination of the sums of squares for the G X N and T X G X N terms (13,889 vs 3,207). Corn yield: increased. with. increasing N rate up to about 180 lb N/acre while grain sorghum yield increased with increasing N rate up to only 60 lb N/acre. This is in agreement with last year's results from this project and the literature in general which shows that grain sorghum requires substantially less N for optimum grain yield than does corn.
For the T X C X N means (Table 6) there was a significant quadratic effect of N on yield for all cover crop - tillage combinations. Thus, just as was the case for the T X G X N the T X C X N interaction is due to the magnitude of the response rather than shape (i.e. linear vs quadratic) or direction (i.e. negative vs positive slope). From the means it can be seen that the differences in magnitude at the level of the second order interaction are mainly due to the response of yield to N rate of the fallow-conventional system. In other, words there appeared to have been a substantial increase in grain yield with N fertilizer applications in excess of 1801b N/acre. For all other cover crop - tillage combinations optimal yields were obtained at about 180 lb N/acre.
The overall relationship between winter cover crops and N rate is best viewed through the C X N interaction means (Table 7). Again in this case there was a significant quadratic relationship between N rate and grain yield. It appears that when all environments are averaged and considered random then hairy vetch and fallow systems both yield about the same, but that hairy vetch has a slightly lower economic optimum N rate. However, when compared to rye, either vetch or fallow systems are clearly better and this difference is not abated with increased N rate. To a certain extent the lower yields were due to poorer stands obtained with rye, but we also believe that rye is detrimental to grain sorghum and corn yield even when populations are constant. We suspect products released from the rye residue are detrimental to corn and grain sorghum yield.
So far in this discussion locations have not been separated, rather all discussion has regarded mean values over locations. That approach is appropriate in that locations are considered random in this study since we want to make inference to the entire state of Illinois rather than discuss individual farms. As this experiment develops we are beginning to see that location is interacting strongly with many of the main effects and first, second, and third order interactions in this study. The major reason for that is that cover crops in general, and hairy vetch in particular, does much better in southern Illinois (i.e. Tom Hortin's farm) than it does at either of the central Illinois locations. This results in much of the benefit seen at Hortin's farm being diluted. If for 1992 locations are considered fixed we can identify several highly significant interactions involving location. For example, there were tremendous differences in the biomass amount (Table 8) and N content of the biomass (Table 9) of cover crops at the three locations at each planting date of the grain crop. The additional biomass production and N content resulted in very different results in the response of corn and grain sorghum yields to these different cropping systems (Table 10). If this continues we will probably complete the project analyses by considering years random and locations fixed.
Other Measures
In addition to grain yield, we also measured plant height, plant nutrient content,
weed pressure, and various soil parameters. In deference to space these will
not be discussed in length, but they deserve brief attention. We have yet to
measure any differences in soil parameters such as bulk density, particle size,
or organic matter content, due to any treatment in this study. Perhaps this
is not surprising since the study is only in the second year and if such events
occur we probably should not expect them for several years at the earliest.
In regards to the other plant data we can summarize by saying that in general
they agree with the yield information. In general vetch seems to benefit corn
and grain sorghum growth and development while rye is detrimental to the same
events. Much, but not all, of that benefit from hairy vetch can be overcome
with additional N fertilizer. Much of the detrimental effect of aye can be overcome
by killing rye while it is still small (< 18").
In general this work shows that soybeans in rotations which involve either
hairy vetch or rye yield slightly more than soybeans in winter-fallow rotations.
It has also been shown that soybeans following no-till grain sorghum yield less
than soybeans following conventional-till grain sorghum. A similar tillage effect
was not seen corn. The soybean yield reduction may have been due to increased
weed pressure associated with the no-till grain sorghum. This work also shows
that location strongly affects cover crops. In particular, it is beginning to
appear that cover crops fit in well with existing corn/soybean and grain/soybean
rotations in far southern Illinois, but not so well in central Illinois. Conventional
wisdom suggests that hairy vetch must follow winter wheat in central Illinois,
because soybeans are harvested too late for proper vetch establishment. This
research clearly does not refute that argument.
Table 1: Effect of 1991 cover crop on soybean grain yield 1992
Table 2: Effect of 1991 feed grain planting date within cover crop on soybean grain yield in 1992
Table 3: Effect of 1991 tillage within feed grain crop on soybean grain yield in 1992
Table 4: Effect of winter cover crop on feed grain yield
Table 6: The effect of N fertilizer rate on feed grain yield within cover crop and tillage.
Table 7: The effect of N rate on feed grain yield within winter cover crops
1Graduate Student, and Asst. Professor of Agronomy; Farmer, Fisher, IL; Assoc Prof. Agric. Econ, Univ of IL; Farmer, Greenup, IL; and Farmer, Albion, IL respectively.
Ebelhar, S.A., W.W. Frye, and R.L. Blevins. 1984. Nitrogen from legume cover crops for notillage corn. Agron. J. 76:51-55.
Flannery, R.L. 1981. Conventional vs. no-tillage corn silage production. Better Crops 65:3-6 (Summer-Fall).
Frye, W.W., R.L. Blevins, M.S. Smith, S.J. Corak, and J.J. Varco. 1988. Role of annual legume cover crops in efficient use of water and nitrogen. p.. 129-154. In W.L. Hargrove (ed.) Cropping strategies for efficient use of water and nitrogen. ASA Spec. Publ. 51. ASA, CSSA, and SSSA, Madison, WI.
Hargrove, W.L. 1986. Winter legumes as a nitrogen source for no-till grain sorghum. Agron J. 78:70-74.
Heichel, G.H., and D.K. Barnes. 1984. Opportunities for meeting crop nitrogen needs from symbiotic nitrogen fixation. p.49-59. In D.A. Bezdicek et al. (ed.) Organic farming: Current technology and its role in a sustainable agriculture. Spec. Pub. 46. American Society of` Agronomy, Madison, WI.
Holsapple, Terry. 1990. Fertilizer freedom. The New Farm. January. pp. 29-30.
Moschler, W.W., G.M. Shear, D.L. Hallock, R.D. Sears, and G.D. Jones. 1967. Winter cover crops for sod-planted corn: Their selection and management. Agron. J. 59:547-551.
Mitchell, W.H., and M.R. Teel. 1977. Winter-annual cover crops for no-tillage corn production. Agron. J. 69:569-573.
Olson, R.A., W.R. Raun, Yang Shou Chun, and J. Skopp. 1986. Nitrogen management and interseeding effects on irrigated corn and sorghum and on soil strength. Agron. J. 78:856-862.
Triplett, G.B. Jr., F. Haghiri, and D.M. Van Doren, Jr. 1979. Plowing effect on corn yield response to N following alfalfa. Agron J. 71:801-803.
Varco, J.J. 1986. Tillage effects on transformation of legume and fertilizer nitrogen and crop recovery residue nitrogen. Ph.D. diss. Univ. of Kentucky, Lexington (Diss. Abst. 8705310).
Wagger, M. G. 1987. Timing effects of cover crop desiccation on decomposition
rates and subsequent nitrogen uptake by corn. p. 35-37. In J.F. Power (ed.)
The role of legumes in conservation tillage systems. Soil Conserv. Soc. AM.,
Ankeny, IA.
