Evaluation of Herbaceous Biomass Crops in the Northern Great Plains: Results and Discussion: Part 1

Precipitation

Growing-season precipitation as departure from the long-term average is presented in Table 1.. Precipitation ranged from 112 to 243 mm below the long-term average during 1988, which made it nearly impossible to establish small-seeded perennials even with light supplemental irrigation.

Growing-season precipitation at Prosper increased each year of the experiment with 1989 below normal, 1990 near normal, and 1991-92 above normal (Table 1.). Distribution was poor during most years. Normal and above-normal years were associated with excessive precipitation during late May and June.

Hettinger was the most droughty environment as expected (Table 1.), but 2 of 4 years were very dry even by its standard. Leonard was much below normal in 1988 and 1989, above normal in 1990 and 1991, and near normal in 1992. Glenfield was extremely dry in 1992, near normal in 1991 and 1989 due to late-season precipitation, and above normal in 1990 due to excessive precipitation in June. Carrington was very dry in 1992, and above normal during 1989-91. Distribution however was extremely erratic in 1989 and 1990.

Perennial Biomass Yields

Biomass yields meaned across years and N levels of several perennial species/mixtures at six North Dakota sites is presented in Figure 2. The greatest average biomass yield (8.66 Mg ha^-1) was produced at the Leonard site primarily due to the exceptionally high yields produced with two cuts in 1990 (Table 2). This was somewhat surprising since the site was chosen as a marginal site due to its extreme vulnerability to wind erosion. Carrington irrigated and Prosper followed closely behind Leonard (Figure 2).

Biomass yields at Carrington were disappointing for an irrigated site. Stands were excellent, but irrigation may not have been as timely as needed for maximum yield. Perennials at Glenfield Good averaged 3.75 Mg ha^-1, 0.92 Mg ha^-1 greater than the Poor site. Hettinger, as expected, had the lowest average biomass yields of the six sites due to the poor environmental conditions (Table 1.) and poorest stands of all sites.

Nitrogen fertilization treatment had little effect on biomass yields during 1989 and 1990, but had significant effects at most sites in 1991 and 1992 (Table 3). Applications of 50 kg N ha^-1 produced near maximum yields at Hettinger in 1991 and Glenfield Good and Poor in 1991 and 1992, 100 kg N ha^-1 was required at Leonard and Carrington irrigated in 1991 and Prosper in 1992, and 200 kg N ha^-1 at Leonard and Carrington in 1992. Apparently, some N is being lost with the flood irrigation at Carrington since biomass yields at 200 kg N ha-1 were less than 100 kg N ha^-1 at Prosper and Leonard. These results follow closely the present recommendation by North Dakota State University to apply 45 kg N ha^-1 to established cool- season grasses in low rainfall areas and 80 to 110 kg N ha^-1 in higher rainfall areas.

Response to N fertilization was similar among species/mixtures as indicated by the lack of a species X N level interaction at most sites in 1991 and 1992 (Table 3). A lack of a N response to N fertilization for switchgrass was the primary cause of the significant interaction in 1992 at Prosper and may be associated with 1992 being the first year a N fertilization response was detected. Alternatively, switchgrass is a warm-season grass which may have a different N-use pattern than the cool-season grasses used at the other sites. However, biomass yields averaged over N level and years are reduced in 1991 and 1992 by including the nonfertilized treatment; therefore, reported biomass yields for the species/mixtures are conservative and would increase if the optimum fertilization was used to calculate species and site effects.

Switchgrass, where adequate stands developed, produced the greatest biomass yields of species tested (Figure 2, Table 2). Switchgrass averaged 11.3 Mg ha^-1 at Leonard and 9.0 Mg ha^-1 at Prosper during 1990-92. The single-plot observation at Glenfield Poor site in 1991 indicated that switchgrass was competitive in biomass yields with the other species. The major disadvantage of switchgrass was the poor establishment compared with the cool-season species/mixtures. Early observations indicated that switchgrass usually emerged, but plants did not establish. When it became obvious that we had establishment problems with switchgrass, a late summer seeding of switchgrass was seeded to initiate an add-on experiment. However, the 20 to 25-cm-tall, 2 to 3-leaf switchgrass plants did not overwinter. It is clear that additional work on switchgrass establishment is warranted.

Bromegrass and brome-alfalfa mixture generally were the lowest yielding biomass species/mixture at all sites (Figure 2, Table 2). Little alfalfa established in the brome-alfalfa treatment due to the fall seeding so it is no surprise that the biomass yields of these treatments were very similar at most sites. Intermediate wheatgrass, intermediate-western wheatgrasses, and the CRP mixture generally were similar in biomass yields across sites, the intermediate-western mixture at Leonard a possible exception (Figure 2). Again, the CRP mixture had little legume component and was dominated by intermediate wheatgrass. Crested wheatgrass was included as a drought-tolerant species and was expected to perform well in the droughty environments, but there was little evidence to support this.

Biomass yields of bromegrass at Prosper during the four years of this experiment (Table 2) have been very similar to well-fertilized forage yields of bromegrass harvested at anthesis growth stage at Fargo over the past 35 years (Meyer et al., 1992). This suggests that the Prosper data should be near that expected over many years.