Evaluation of Herbaceous Biomass Crops in the Northern Great Plains: Results and Discussion: Part 2

Chemical Composition of Perennial Species

Chemical composition of perennial biomass species as influenced by environments is presented in Table 4. Many chemical components had a significant year X site interaction indicating that the chemical composition can be expected to vary with site and production year. The significant interaction was caused by changes in relative ranking of a component among years. For example, the ash concentration was highest at Glenfield Good in 1989, 1991, and 1992, but lowest in 1990. The data suggest that it is impossible to accurately predict the chemical composition of a perennial biomass species at a particular site and year by other than a "ballpark" figure.

Nitrogen fertilization had little effect on the chemical composition of perennial biomass species in the first year of production (Table 5), and had several significant effects at only Glenfield Good and Carrington irrigated in 1990. Significant differences by N level were detected in about half of the comparisons of chemical components within sites in 1991 and 1992, but a consistent trend for N level effects across sites was evident only for N concentration (data not shown). Where significant N effects were detected, ash, ADF, TNC, and cellulose concentrations generally decreased with increasing N level and NDF, ADL, N, and hemicellulose increased. Even where significant differences in chemical composition occurred, relative differences were small for all but N concentration, indicating that N fertilization has its greatest effect on biomass yield and only minor effects on the chemical composition. Nitrogen effects within 1991 and 1992 were discussed by Meyer et al. (1993).Cool-season grass species/mixtures at Prosper generally differed little in chemical composition when meaned across years and N levels, the "ballpark" estimate (Table 6, Fig. 3). This is not too surprising when the makeup of field plots was evaluated. The three cool-season mixtures generally had few legume components and/or were dominated by intermediate wheatgrass. This caused bromegrass and brome-alfalfa, and intermediate wheatgrass, intermediate/western wheatgrasses, and the CRP

mixture to be quite similar in chemical composition and act more like two treatments than five. Still, reed canarygrass, bromegrass, and brome-alfalfa generally had higher ash and N concentrations than the other species/mixtures, but NDF, ADF, hemicellulose, and cellulose concentrations were very similar. Crested wheatgrass had a higher TNC and a lower ash content than most other cool-season species/mixtures. Bromegrass and brome-alfalfa were slightly higher in ADL.

Switchgrass and big bluestem [only isolated comparisons, Meyer et al. (1993)], warm-season species, had significantly greater NDF and hemicellulose and less ash, N, and ADL concentrations than most cool-season species/mixtures (Fig. 3, Table 6). These species were the major cause of significant species differences in chemical composition where included in the treatments (especially at Prosper). These differences were expected and follow known differences between warm- and cool-season grasses.

Cool-season grass species/mixtures at sites other than Prosper also differed little in chemical composition when meaned across years and N levels (Table 6). Crested wheatgrass generally was the lowest species/mixtures in ash, cellulose, and ADF concentrations across sites. Bromegrass and brome- alfalfa generally were highest of the cool-season species/mixtures in ash and N concentrations.

Species differences in chemical composition within years and sites frequently were detected (Table 7) in contrast to when meaned across years (Table 6). These differences have been discussed previously by Meyer et al. (1990, 1992, 1993).

The yield of chemical components (biomass yield times concentration of each component) of perennial biomass species at six sites in 1989-92 is presented in Table 8. The absolute yields were very variable among sites and years. For example, yields of NDF ranged form 0.70 to 6.29 Mg ha^-1 across years and sites, or a 9-fold difference. Likewise, TNC ranged from 0.08 to 1.07 Mg ha^-1 and ADL ranged from 0.5 to 0.59 Mg ha^-1. Most chemical component yields were primarily dependent on biomass yields; although, the small differences detected in chemical composition among species/mixtures did at times contribute to yield differences (data not presented).

Corn Grain Yields

Corn for grain was included in these experiments as an environmental check. Grain yield ranged from 9.2 to 10.9 Mg ha^-1 at the Carrington irrigated site across the 3 test years (Table 9), about that expected at this site. Grain yields at the dryland sites were much more variable. For example, grain yields at Prosper ranged from 2.0 Mg ha^-1 in the 1988 dry year to 11.0 Mg ha^-1 on fallow in 1990. Leonard had a very similar range to that of Prosper, but grain yields were significantly lower in 1991 and 1992 at Leonard than at Prosper. Grain yields at Glenfield were at best disappointing. Grain did not set at Hettinger in 1988, 1990, and 1991 due to droughty conditions (Table 1). Grain yields were not obtained at Glenfield Good and Poor due to an undetected zinc deficiency in 1988 and 1989.

Nitrogen level and cropping system (fallow vs. wheat) management intensities rarely were significant (Table 9), in part due to the few degrees of freedom to test the cropping system effects, but also due to variable yields. Prosper in 1988 and Leonard in 1991 were the only exceptions where significance was detected.

Biomass Yield of Annuals

Biomass yield of annual herbaceous crops as affected by cropping system at two sites is presented in Table 10. Biomass yields averaged 1 to 1.2 Mg ha^-1 greater on fallow than on recrop wheat land at the two sites. Biomass yields were not significantly affected by the cropping system in 1991 and 1992 at both sites, but were greater on fallow than on recropped wheat land in 1990 at Leonard. Biomass yields were unaffected by the cropping system at Glenfield Poor in 1991, but were higher from recrop than from fallow at Glenfield Good in 1991 (Table 11). The reason for this last response is unclear.

Nitrogen levels above 50 kg N ha^-1, the lowest level used, generally did not affect the biomass yield of annual herbaceous crops at most sites or years (Table 12). Likewise, the N level X species interaction generally was nonsignificant; therefore, species results are presented at 50 kg N ha^-1.

Biomass yields meaned across years of annual species at 50 kg N ha^-1 as affected by site and cropping system is presented in Figure 4. Maximum biomass yield (16.9 Mg ha^-1) was obtained with forage sorghum at Carrington during 1988-91. Likewise, forage sorghum has been the highest yielding annual species on dryland averaging 16.2 Mg ha^-1 at Prosper on fallow during 1988-92 and the highest yielding species on recrop or fallow at Prosper, Leonard, and Glenfield Good. Sorghum X sudan has been only slightly lower yielding than forage sorghum at the high-moisture sites and higher yielding at Hettinger and Glenfield Poor, the droughty sites. This result follows what we anticipated when these experiments were planned. Foxtail millet generally has been the lowest yielding species at most sites, but may have a place where a short-season crop is needed.

Kochia has been the most surprising species. Kochia and sweetclover were initially included as a volunteering species, but it became clear that kochia has allelopathic compound(s), which are autotoxic to the kochia plant. Biomass yield of volunteering kochia was less than half that of annually seeded kochia (data not presented). Therefore, volunteering kochia was deleted and annually seeded kochia included instead. In addition, sweetclover was not volunteering adequately so this treatment was discontinued also.

Biomass yields of annually seeded kochia have been nearly equal to forage sorghum at Prosper and the highest yielding species at Hettinger (Fig. 4, Table 11). As a weedy species commonly found in the northern Great Plains, this species appears well adapted to droughty environments. However, it did not perform well at Leonard; although, biomass yields reported at many sites may have been reduced by inappropriate harvest dates. See the section on maturity effects latter in this report. In addition, poor stand establishment in 1991 resulted in either no observation or reduced yields from the low plant density. It is clear that additional work is needed on use of kochia as an annually seeded biomass species.

The highest biomass yield recorded was 23.3 Mg ha-1 for forage sorghum grown at Prosper on fallow in 1990 (Table 11). Species effects within years have been discussed previously (Meyer et al., 1989-93).