Evaluation of Herbaceous Biomass Crops in the Northern Great Plains: Results and Discussion: Part 4

Biomass Cropping on Normally Fallowed Land

Legume biomass yields on normally fallowed land in 1990 ranged from 2.1 to 3.9 Mg ha^-1 for 1- and 2- cut alfalfa and 2.7 Mg ha^-1 for 1-cut sweetclover (Table 16). The poor alfalfa yield was due to severe alfalfa leafhopper infestations on both growths, the first time in 20 years an economic infestation occurred. Biomass yields would normally have been nearly double with the precipitation received without the leafhopper infestation. Sweetclover biomass yields were about 1 Mg ha^-1 less than expected, possibility due to weed competition.

Barley grain yields meaned across biomass treatments and tillage increased from 2438 to 2967 kg ha^-1 with application of 75 kg N ha^-1. Additional N additions did not increase grain yield.

Barley grain yield (meaned across N levels) was the highest following no-till wheat-barley rotation, primarily due to superior yield of the unfertilized treatment (Table 16). Apparently, enough N was carried over from the previous crop for near maximum grain yield. Grain yields of the 1-cut no-till and green-manured, tilled sweetclover treatments were similar to the no-till wheat-barley rotation and greater than the conventional-tilled wheat-barley rotation. The low grain yield following no-till green-manured sweetclover may have been due to a tie up of N in decomposition of the residue that was incorporated with spring tillage. These responses are similar to data reported by Meyer (1987) and Badaruddin and Meyer (1989).

Barley grain yields following 1- or 2-cut alfalfa were less than the wheat-barley rotation conventionally tilled and about equal to fallow (Table 16). It is unclear why the fallow treatment yielded less than the wheat- barley rotation unless moisture accumulation over the winter was a major factor or moisture storage on fallow was negligible due to tillage for weed control.

Reduced tillage results were erratic. No-till wheat-barley tended to yield greater than conventional-tilled, and no-till 1-cut sweetclover was slightly greater yielding than conventional-tilled. However, no-tilled 1- or 2-cut alfalfa was lower yielding than like treatments tilled.

Nonfertilized wheat grain yield was highest following the 1-cut sweetclover no-till treatment (Table 17). Most other treatments were quite similar.

Wheat grain yields (meaned across cropping systems) increased about 300 kg ha^-1 with application of 75 kg N ha^-1 (Table 17). Highest grain yields at the medium fertilization level occurred on wheat following wheat treatment. This is contrary to most previous results (Meyer, 1987; Badaruddin and Meyer, 1989) from legume cropping systems in North Dakota where normally fertilized wheat on wheat grain yield was nearly equal to nonfertilized grain yields following legumes either hayed or green-manured. One possible explanation for these results might be more water use by the legumes than wheat in 1990. An additional unit of N generally decreased or only slightly increased grain yields compared with 75 kg N ha^-1 except on 1-cut alfalfa with fall incorporation. The reason for this high grain yield is unclear.

These 1-year data suggest that a leguminous biomass crop (especially sweetclover) could be removed without reducing subsequent crop productivity substantially. However, additional work is needed on stand establishment of the leguminous biomass species. Inadequate stands to perform this experiment were obtained in 2 out of 3 years, partially due to two very dry establishment seasons. However, no-till establishment of legumes may allow consistent stand establishment in all but the driest environments.

Economic Analysis of Biomass Cropping

The first phase to evaluate the economic feasibility of biomass cropping was to develop model farms and enterprise budgets for three areas (Johnson et al., 1990). A linear programming model was used to analyze the impact of producing herbaceous energy crops on the model farms. Returns over variable costs were determined for all herbaceous biomass crops and conventional crops (Johnson et al., 1993a). The following is the abstract of the final report; see Johnson et al. (1993b) for more details.

The economics of producing herbaceous biomass crops were evaluated for three regions of North Dakota. Typical farms were modeled for eastern (Cass County), central (Foster County), and western (Adams County) North Dakota. At a $35-per-ton price, biomass crops were included in profit-maximum farm plans in all three areas. The increase in net income through introduction of biomass crops was substantial (up to $20,000) in Cass County. Kochia was the biomass crop included in the farm plan in all regions. When kochia is excluded, sudan/sorghum (Adams), forage sorghum (Cass), and sorghum X sudan (Foster) were the most profitable biomass crops. Including biomass crops changed the labor distribution but did not necessarily eliminate high labor-use periods. At lower biomass prices, production would be eliminated first on the Cass County farm followed by the Foster and Adams Counties model farms.

Maturity Effects on Biomass Yield of Kochia

Kochia in 1989 was harvested in August to prevent seed production so it would not cause a weed problem. But, results in 1989 suggested that the early harvest reduced the biomass yield. Therefore, our objective of these experiments were to determine the influence of maturity on biomass yields of kochia.

Biomass yields dropped sharply in 1990 when harvested 19 days after a killing frost (Table 18). Ash and ADL concentrations increased and N concentration tended to decrease with maturity. Cellulose, hemicellulose, NDF, and ADF concentrations changed little with maturity from mid-August through October.

This experiment was seeded at Prosper and Leonard in 1991, but the kochia seed source in 1991 did not establish adequate stands. It is unclear why the seed was poor, since we used the same methods of obtaining seed as previous years.

A native stand from a sparse 1991 stand was sampled about every 10 days starting July 10 in 1992. Biomass yields increased from 29 July (stem elongation) to 17 September (maturing seed) (Table 19). Biomass yield dropped 6.6 Mg ha^-1 following a -2.2oC frost on 22 September and a -5.5oC frost on 28 September due to seed and leaf loss. Likewise, biomass yield dropped 3.4 Mg ha^-1 following the frost in a late spring-seeded experiment (Table 20). Kochia must be harvested for biomass prior to frost to prevent sizable yield losses!

Biomass yield of kochia in the native stand (Table 19) increased 4.8 Mg ha^-1 from early to mid September, much of it due to seed production. If kochia is used as a biomass species, seed drop must be prevented in order to keep it from infesting subsequent crops. Therefore, optimum harvest time will be 2 to 3 weeks prior to normal first frost.

Fibrous component (NDF, ADF, ADL, hemicellulose, and cellulose) concentrations increased and N, TNC, and ash decreased as kochia matured in both 1992 experiments (Tables 19 and 20). These data are in contrast to the 1990 (Table 18) data where fibrous components changed little with maturity. The reason for differences in chemical composition among the years is unclear, but 1992 data is more what would be anticipated from a maturing crop.

Effects of Biomass Cropping on Subsequent Crop Yields

Our objective was to determine the "cost" of biomass cropping on normally fallowed lands, i.e., not having fallow land for subsequent cropping. Fallow and wheat treatments were included with the annual experiments at Prosper and Leonard in 1991. 'Grandin' spring wheat was seeded during April 1992 across all species. During May, it became obvious that this experiment failed due to atrazine carryover. Unknown to us, the previous research specialist had hand-sprayed green and yellow foxtail with atrazine a second time, which increased atrazine carryover. The carryover was enough to kill the emerging wheat plants for about 15 cm over the old sorghum and corn rows. As a result, the experiment was abandoned at both sites.

Maturity Effects on Stand Maintenance of Perennial Grasses

The objective of these preliminary experiments was to evaluate the effect of delayed harvest typical of biomass cropping on yield and stand maintenance in herbaceous perennial biomass species. Biomass yields of bromegrass, crested wheatgrass, and intermediate wheatgrass were unaffected by harvest date at Fargo, ND, in 1992 (Table 21). Increasing N level from 0 to 150 kg ha^-1 increased biomass yields 130% as a mean of species in the 5th year of fertilization. No lodging occurred in this experiment and in a similar experiment at Carrington, ND, in 1992 (data not presented). Likewise, no differences in stands could be detected by harvest date; although, ground cover was slightly greater in the nonfertilized compared with highly fertilized plots.

An old bromegrass sod fertilized with six levels of N since 1954 was harvested for biomass determination at two dates during 1990-92. Average biomass yields have been slightly higher when harvested 3 to 4 weeks after anthesis than harvesting at anthesis (Table 22). Biomass yields increased with N level up to 74 to 149 kg ha^-1, but decreased with increasing N level when harvested at anthesis. This decrease is due to a small stand loss in about 1 out of 4 years. However, little stand deterioration has occurred in plots fertilized with 74 or 149 kg N ha^-1, indicating that stand deterioration should not be a problem at economically productive levels of N fertilization.

Stand maintenance was evaluated following 4 years of biomass cropping at the six sites. Little deterioration of stand occurred at any site (data not presented)! Stands actually improved at the two Glenfield sites for all rhizomatous species. Less lodging occurred at Leonard than at Carrington irrigated and Prosper sites. As a result, no stand deterioration was observed at Leonard. Extensive lodging of cool-season species/mixtures (reed canarygrass excepted) occurred each year at Prosper and Carrington irrigated. If harvest of these species had been delayed into August or September, we are convinced that significant stand loss would have occurred. But, with the harvest date selected, little deterioration of stand occurred. Even plots showing slight stand deterioration at 200 kg N ha^-1 had better ground cover than the long-term fertilization experiment with bromegrass harvested at anthesis (Table 22).