Fall and Spring Forage Production and Quality of Winter Cereals Seeded at Three Fall Dates

Thomas C. Griggs, Assistant Professor, Department of Plants, Soils, and Biometeorology, Utah State University, Logan 84322-4820

Abstract

Irrigated perennial pasture growth is negligible during November through March in northern Utah. Winter cereals including barley (Hordeum vulgare L.), triticale (×Triticosecale Wittmack), and wheat (Triticum aestivum L.) may provide supplemental pasture during this period. This study evaluated fall and spring forage production and quality of cereal cultivars seeded on 22 or 27 August, 11 or 19 September, and 1 or 11 October in 2001-2002. Forage production declined by 53 to 69 kg/ha in fall, and by 42 to 85 kg/ha in spring, for each day that seeding was postponed beyond 22 or 27 August. Herbage mass for August seeding dates ranged from 1740 to 3730 kg/ha in November and from 1160 to 5810 kg/ha in April. Forage quality levels in November and April differed only slightly among cultivars and seeding dates, and ranged from 163 to 339 g/kg crude protein (CP), 691 to 948 g/kg neutral detergent fiber digestibility (NDFD), and 922 to 977 g/kg in vitro true dry matter digestibility (IVTD). When seeded in August, Stephens and Utah-100 wheat, and Forerunner and a TRICAL 102/2700 triticale blend, produced the most spring and total forage, averaging 6790, 6150, 6830, and 6180 kg total forage per ha, respectively, across years. Early-seeded winter cereals can extend pasture availability during mid-November to mid-April in the Intermountain West.

Productivity of irrigated perennial pastures is limited by low temperatures during November through March in northern Utah. One option for maintaining pasture availability and reducing hay-feeding costs during this period includes the use of cereal crops that grow under cooler conditions. Spring and winter cereals are widely grown for grain and harvested forage in rotation with alfalfa (Medicago sativa L.) in the Intermountain West. Species and cultivar rankings for forage production and quality have been inconsistent among environments and growth stages for spring and winter cereals harvested at early (2,14), late (9), or various (3,4,5,6,8,11) growth stages. In these studies, only one of which (9) was conducted in western North America, wheat has often been more productive than barley, while productivity of triticale has been more variable relative to wheat and barley. Spring forage production from winter triticale declined when seeded later than 25 September in Iowa (15). Mixing spring triticale with fall-seeded winter cereals increased fall forage production in Wisconsin, but decreased spring and total forage production relative to winter cereal monocultures (11). Greater fall production was related to internode elongation in spring triticale, as compared with vegetative growth of winter cereals during fall. This study focused on fall-seeded winter cereals because of their potential for earlier spring forage production than is provided by spring-seeded cereals (4,8). Few comparisons of forage production, quality, or seeding date responses of winter cereals have been made during fall through early spring in the Intermountain West. Our objective was to determine fall and spring forage production and quality of winter barley, triticale, and wheat cultivars seeded at three fall dates in northern Utah.

Experimental Treatments and Management

Experiments were conducted at North Logan, UT (41°46’N, 111°50’W, 1402 m elevation) on Millville silt loam soil (coarse-silty, carbonatic, mesic Typic Haploxerolls). Whole-plot treatments were three seeding dates and subplot treatments were nine cultivars in a split-plot design (Figs. 1 and 2) with four randomized complete blocks. Clean-tilled subplots 0.9 m by either 2.4 m (2001 seeding) or 3.4 m (2002 seeding) were seeded with a small-plot drill at a depth of 2 to 3 cm in five rows spaced at 15 cm. Early, middle, and late seeding dates were 22 August, 11 September, and 1 October in 2001, and 27 August, 19 September, and 11 October in 2002. Cultivars (Table 1) were ‘Kold’ and ‘Schuyler’ winter barley; ‘Forerunner’, ‘TRICAL 102’, ‘TRICAL 348’, and ‘RSI X349’ triticale; a commercial blend of 67% TRICAL 102 and 33% ‘TRICAL 2700’ triticale; and ‘Stephens’ and ‘Utah-100’ winter wheat. The role of TRICAL 2700, which is intermediate between spring and winter types, in the triticale blend was to increase fall productivity of the mixture relative to TRICAL 102 alone.


Fig. 1. Forage on 21 November 2002 in whole plots seeded 27 August (right), 19 September, and 11 October (left) 2002.		Fig. 2. Forage on 21 March 2003 in whole plots seeded 27 August (middle), 19 September, and 11 October (left) 2002.

Table 1. Cereal cultivars and seeding rates in 2001 and 2002.

Cultivar	Species and class	Spikelet awn character- istics	Fall seeding density and bulk seeding rate
			2001		2002
			seeds/ m²	kg/ha	seeds/ m²	kg/ha
Kold	6-row^x winter barley	Awned	416	120	355	103
Schuyler	6-row winter barley	Awned	370	120	355	115
Forerunner	Facultative triticale	Awnletted	216	120	384	198
TRICAL 102	Winter triticale	Awnletted	324	120	404	131
TRICAL 102/2700^y	Facultative triticale^y	Awned^y	312	120	356	137
TRICAL 348	Winter triticale	Awnletted	330	120	355	129
RSI X349	Winter triticale	Awned	469	120	356	91
Stephens	Soft white winter wheat	Awned	290	120	382	147
Utah-100	Hard red winter wheat	Awned	309	120	484	138
Mean			337	120	381	132

^x Six-row barley has three rows of fertile spikelets on each side of the rachis of the inflorescence.

^y Blend of TRICAL 102 and TRICAL 2700 (2:1). Characteristics are for TRICAL 2700, while seeding density and rate are for the blend.

Experiments were established each August in adjacent fields, following disking and harrowing of spring barley stubble in 2001 and summer fallow in 2002. Fall soil test levels were pH 8.0; 2.2 to 2.4% organic matter; and >14, 125, and 10 ppm P, K, and SO₄-S, respectively. Plots seeded in 2001 received 67 kg N per ha from ammonium sulfate on 7 September and 56 kg N per ha from ammonium nitrate on 12 April. Plots seeded in 2002 received 64 kg N per ha from ammonium sulfate on 11 September and 56 kg N per ha from urea on 21 March. Ammonium nitrate and urea were substituted for ammonium sulfate in spring because soil S levels were adequate. Seed densities (Table 1) were 216 to 469 seeds/m² in 2001 and were adjusted to a more uniform 355 to 484 seeds/m² in 2002. Previous research (9) suggests little to no variation in forage production across these ranges of seeding density. Plots were irrigated at 7- to 10-day intervals during September and October, but not in spring. Monthly mean maximum and minimum temperatures and total precipitation are shown in Table 2.

Table 2. Monthly mean air temperature and precipitation at Logan, UT during experimental periods in 2001-2003.

Seeding year	Air temperature mean (°C) or precipitation total (mm)^x	Aug	Sept	Oct	Nov	Dec	Jan	Feb	Mar	Apr
2001	Maximum temperature	31.9	28.0	19.5	9.9	-3.1	-1.7	-3.9	5.1	15.4
	Minimum temperature	10.2	7.0	-0.1	-2.8	-12.5	-11.9	-17.6	-6.6	1.3
	Precipitation	1	7	22	49	58	49	3	29	48
2002	Maximum temperature	30.0	24.6	15.4	7.0	4.2	5.9	3.0	11.1	11.1
	Minimum temperature	9.1	7.8	-0.7	-5.2	-5.4	-3.7	-7.4	-1.6	-1.1
	Precipitation	0	64	34	35	14	19	27	36	15

^x Monthly means and totals are of daily values. Data for August and April represent only those days between early seeding and spring sampling.

Forage Sampling and Analyses

Herbage mass (HM) was determined in four replications by hand-clipping within one 0.1-m² quadrat per subplot to 1 cm above soil surface and drying to constant mass at 55°C. Plots seeded in 2001 were sampled on 14 November 2001 and 29 April 2002, and plots seeded in 2002 (Figs. 1 and 2) were sampled on 18 November 2002 and 9 April 2003. Plots were trimmed to a 10- to 12-cm stubble height on 16 November 2001 and 16 December 2002 to simulate late-fall utilization. Total HM was the sum of initial fall growth and spring regrowth. In 2003 only, plots were resampled on 9 May to assess HM at a stage more relevant to mechanical harvesting than to pasture. Results from May 2003 sampling are reported separately from those for fall (November) and spring (April) sampling.

Herbage composition was determined in two replications with near-infrared reflectance spectroscopy (NIRS) of dried ground samples via scanning monochromator (Mod. 6500, FOSS NIRSystems, Inc., Silver Spring, MD), reference wet chemical analyses of 129 to 143 samples selected to represent the spectral distribution of the experimental material, and development of prediction equations with WinISI II chemometrics software (Infrasoft International, Port Matilda, PA) following procedures of Shenk and Westerhaus (16). Samples were ground in a cyclone mill to pass a 1-mm screen for NIRS and determination of dry matter (DM) and CP. Samples for other analyses were ground in the same mill to pass a 2-mm screen to approximate the particle size from a 1-mm screen of a shear mill (13), as is specified for NDF (12) and IVTD (7) procedures. Laboratory DM concentration was determined following overnight drying at 105°C. Nitrogen was determined (1) by a combustion procedure (LECO Corp., St. Joseph, MI) and multiplied by 6.25 to estimate CP concentration. Amylase-treated NDF (12) and IVTD (7) were determined with filter bags in batch vessels (Ankom Technology Corp., Fairport, NY). Samples for IVTD were incubated for 48 h at 39°C, then refluxed in neutral detergent solution. Digestibility of NDF was calculated from NDF and IVTD concentrations. Standard errors of cross-validation for NIRS prediction equations from modified partial least squares regression were 10, 22, and 21 g/kg for CP, NDF, and IVTD, respectively. Proportions of variation in CP, NDF, and IVTD concentrations in calibration samples accounted for by NIRS-predicted values were 0.96, 0.96, and 0.92, respectively.

Treatment, year, and year × treatment effects were determined by analysis of variance for a split plot in time (17). Orthogonal polynomial contrasts (17) and linear regression were also used to evaluate forage production responses to seeding date expressed as day of year. All analyses were performed with SYSTAT, ver. 11 (SYSTAT Software, Inc., Richmond, CA). Effects were considered significant at P ≤ 0.10 and subsequent references to effects or differences assume this level unless otherwise indicated. Following significant F tests, means were separated by Fisher’s LSD.

Forage Production and Seeding Date Responses

Growth stage (10) at fall and spring sampling was vegetative (Feekes stages 3 to 5), except for Forerunner triticale which was jointing (Feekes stage 7.5) in spring 2002. Significant effects on forage production included year and year × seeding date for fall, spring, and total HM; year × cultivar for spring and total HM; and year × seeding date × cultivar for fall and total HM. Treatment means for HM are therefore presented by year. Effects on forage quality included year for fall CP, spring CP, and fall NDF; and year × cultivar for fall NDF. Forage quality responses are therefore summarized across years, except for fall NDF.

Fall, spring, and total HM declined sharply with successive dates of seeding (Figs. 1 through 4), as has also been reported by Schwarte et al. (15). Forage production responses to seeding date had a linear component (P < 0.001) for fall, spring, and total growth in both years, and a quadratic component (P ≤ 0.001) only for fall and spring growth from plots seeded in 2002. Since the quadratic component accounted for only 2.1 and 4.4%, respectively, of seeding date responses in these two growth periods, linear regression was used to describe forage production responses to seeding dates for all growth periods (Table 3). For each day that seeding was postponed beyond 22 or 27 August, forage production declined by 53 to 69 kg/ha in fall, and by 42 to 85 kg/ha in spring.

Table 3. Results from linear regression of fall, spring, and total forage production on day of year of seeding winter cereals.

Forage production period		n^w	r^2 x	SEE^y (kg/ha)	Slope^z (kg/ha-day)
Fall	2001	27	0.92	263	-53
	2002	27	0.95	289	-69
	combined years	54	0.84	452	-58
Spring	2002	27	0.76	402	-42
	2003	27	0.87	617	-85
	combined years	54	0.42	1151	-54
Total	2001-2002	27	0.93	443	-95
	2002-2003	27	0.96	586	-154
	combined years	54	0.65	1492	-112

^w Number of herbage mass values used in regression, each of which was the mean of four replications.

^x Proportion of variation in forage production accounted for by date of seeding.

^y Standard error of the estimate from linear regression.

^z Regression coefficient showing reduction in forage production for each additional day that seeding is postponed beyond 22 August 2001 or 27 August 2002 (P < 0.001 in all cases).

Fig. 3. Fall, spring, and total herbage mass of barley, triticale, and wheat cultivars seeded 22 August, 11 September, and 1 October 2001 and sampled 14 November and 29 April.

Fig. 4. Fall, spring, and total herbage mass of barley, triticale, and wheat cultivars seeded 27 August, 19 September, and 11 October 2002 and sampled 18 November and 9 April.

Herbage mass of plots seeded in August varied among cultivars and years from 1740 to 3730 kg/ha in fall, and from 1160 to 5810 kg/ha in spring (Figs. 3 and 4). These results are consistent with findings of Maloney et al. (11) and Poysa (14). Forage production was more similar among years in fall than in spring. Spring HM in 2002 and 2003 comprised 48 and 59%, respectively, of total HM of plots seeded in August. Greater spring production in 2003 may have been due to higher temperatures in February and March than in 2002 (Table 2). Total HM of cultivars seeded in August 2001 and 2002 ranged from 3380 to 5510 kg/ha, and from 6180 to 9040 kg/ha, respectively.

Fall and spring sward height and HM following September and October seeding dates offered little to no extension of pasture carrying capacity, relative to perennial grass-legume mixtures. Across years, mean sward heights for August to October seeding dates were 6 to 14, 2 to 7, and 1 to 3 cm, respectively, in fall, and 19 to 23, 18 to 21, and 7 to 10 cm, respectively, in spring. Although this study focused on forage availability during November and April, plots were resampled in May 2003 for HM at a growth stage associated with mechanical harvesting. Mean HM at late jointing (Feekes stages 7.5 to 10.1) on 9 May was 10290, 9310, and 5000 kg/ha for August to October seeding dates, respectively, whereas HM on 9 April was 4430, 3180, and 600 kg/ha for the same seeding dates (no additional data shown).

Although broad statistical comparisons among cereal species was not an objective, spring and fall forage production of wheat cultivars was equal to or greater than that of barley and triticale cultivars for August seeding dates. Total HM of barley, triticale, and wheat cultivars averaged 4060, 4220, and 5370 kg/ha, respectively, for August 2001 seedings, and 6560, 7880, and 7570 kg/ha, respectively, for August 2002 seedings. Among years, rankings of August-seeded cultivars were inconsistent for fall HM, and were more consistent for spring and total HM (Figs. 3 and 4). In both years, August-seeded Stephens and Utah-100 wheat, and Forerunner and TRICAL 102/2700 triticale, exhibited greater spring and total HM than most other cultivars. Cultivars that were most productive in fall were not necessarily those with greatest spring and total forage production. A possible basis for cultivar selection might therefore be comparison of fall and spring forage production patterns relative to the season in which forage is most needed in a management system. Stevens wheat and Forerunner triticale are examples of cultivars with relatively even distribution of forage production across seasons.

TRICAL 102 and the TRICAL 102/2700 triticale blend often differed in fall and spring HM, but their differences were more consistent in spring, when HM was greater for the blend. Greater spring productivity for the TRICAL 102/2700 blend is in contrast to results from Wisconsin (11) and suggests that there was better winter survival and perhaps earlier spring growth of TRICAL 2700, relative to TRICAL 102. Despite inconsistencies in relative forage productivity of cultivars among years, fall and spring HM of most cultivars seeded in August was comparable with that of perennial pastures in the region during early spring and late summer.

Forage Quality and Seeding Date Responses

Forage quality levels were high in fall and spring (Tables 4 and 5), and are consistent with other reports (2,6,11) for vegetative growth stages of cool-season annuals. Forage quality responses to seeding date were less dramatic than those for forage production. Fall and spring CP levels increased, and NDF levels usually decreased, with later seeding dates. Digestibilities of DM and NDF did not differ among seeding dates in fall, but in spring they were greater for later seeding dates. Fall and spring NDF levels were consistently less for TRICAL 102, TRICAL 348, and RSI X349 triticale than for other cultivars seeded in August, but these differences and those for fall IVTD were small. Cultivar differences in forage quality have been reported by others (3,4,6,8,11) for spring cereals or later growth stages. While all sampling was at pre-heading growth stages, awns in late-harvested forage of some cultivars (Table 1) could negatively impact animal performance.

Table 4. Fall forage quality of cereals seeded 22 or 27 August and 11 or 19 September of 2001 and 2002, respectively. Plots seeded 1 or 11 October had insufficient growth for sampling. Two-year means, except for NDF, are for plots sampled 14 November 2001 and 18 November 2002.

Species and cultivar	CP (g/kg DM)		NDF^w (g/kg DM)				NDFD (g/kg NDF)		IVTD (g/kg DM)
	CP (g/kg DM)		2001		2002		NDFD (g/kg NDF)		IVTD (g/kg DM)
	Aug	Sept	Aug	Sept	Aug	Sept	Aug	Sept	Aug	Sept
Kold barley	186	218	353	329	311	274	825	823	945	948
Schuyler barley	170	212	318	305	289	273	795	761	941	934
Forerunner triticale	179	217	355	313	295	249	814	807	942	946
TRICAL 102 triticale	177	194	321	297	268	233	768	691	922	925
TRICAL 102/2700 triticale^x	185	235	377	318	289	247	770	831	931	947
TRICAL 348 triticale	196	228	317	313	263	232	795	805	931	947
RSI X349 triticale	194	220	321	306	275	–^y	805	782	937	933
Stephens wheat	188	217	351	296	300	245	840	822	941	944
Utah-100 wheat	163	185	372	297	282	228	835	762	944	939
Mean 2001	160	193	343	308	–	–	798	780	924	927
Mean 2002	203	235	–	–	286	248	812	794	951	954
P date	0.01		0.02		0.02		0.33		0.42
P, LSD^z cultivar	0.61, –		0.04, –		<0.01, 16		0.42, –		0.06, 9
P, LSD^z date × cultivar	0.70, –		0.04, 18		0.31, –		0.44, –		0.55, –

^w Data are by year due to year × cultivar interaction (P = 0.05).

^x Blend of TRICAL 102 and TRICAL 2700 (2:1).

^y Insufficient growth to permit sampling.

^z LSD shown only for significant F test and in absence of higher-level interaction.

Table 5. Spring forage quality of cereals seeded 22 or 27 August, 11 or 19 September, and 1 or 11 October of 2001 and 2002, respectively. Two-year means are for plots sampled 29 April 2002 and 9 April 2003.

Species and cultivar	CP (g/kg DM)			NDF (g/kg DM)			NDFD (g/kg NDF)			IVTD (g/kg DM)
Species and cultivar	Aug	Sept	Oct	Aug	Sept	Oct	Aug	Sept	Oct	Aug	Sept	Oct
Kold barley	223	254	322	371	350	334	842	893	919	938	957	968
Schuyler barley	225	235	339	352	338	306	843	893	925	943	963	977
Forerunner triticale	205	243	314	380	338	316	811	894	948	934	964	972
TRICAL 102 triticale	216	246	301	361	343	309	871	874	908	945	955	970
TRICAL 102/2700 triticale^x	185	243	302	385	352	324	825	861	933	929	953	973
TRICAL 348 triticale	204	259	303	362	330	311	840	910	930	938	963	977
RSI X349 triticale	219	248	297	360	347	323	846	888	920	942	963	965
Stephens wheat	212	233	306	374	372	344	858	866	919	937	947	970
Utah-100 wheat	209	249	278	369	354	346	868	885	908	949	967	971
Mean 2001	244	262	314	372	360	318	809	852	869	932	946	961
Mean 2002	177	229	300	364	334	329	881	917	978	947	973	982
P, LSD^y date	<0.01, 22			0.05, 14			0.10, –			0.07, 15
P, LSD^y cultivar	0.53, –			0.08, 12			0.30, –			0.25, –
P, LSD^y date × cultivar	0.66, –			0.78, –			0.08, 38			0.93, –

^x Blend of TRICAL 102 and TRICAL 2700 (2:1).

^y LSD shown only for significant F test and in absence of higher-level interaction.

Conclusions

Fall-seeded winter cereals can supply high-quality forage to extend pasture availability during November to mid-April in the Intermountain West, but production drops sharply for seeding dates after late August. Depending on seeding date, fall environmental conditions, and snow levels, livestock could begin grazing winter cereals as early as mid-November to early December. Assuming a sufficient regrowth period during March to early April under rotational stocking management, cereals could be regrazed in early to mid-April prior to availability of perennial pasture. One advantage of fall-seeded cereals for supplemental pasture is a relatively low irrigation water requirement that is limited to fall, as compared with a larger irrigation water requirement for summer production of perennial forage hay or silage for winter feeding. Cultivars varied in their relative fall and spring productivity, but pasture availability is often most limiting in early spring. Cultivars that produced the most spring and total forage following August seeding dates were Stephens and Utah-100 wheat, and Forerunner and TRICAL 102/2700 triticale, averaging 6790, 6150, 6830, and 6180 kg total forage per ha, respectively, across years.

Acknowledgments

Contributions of Robert Clawson, Darren Fillmore, Megan Guenter, Kristina Pack, and Randy Sessions are gratefully acknowledged. This research was supported by the Utah Agricultural Experiment Station.

Approved as Utah Agricultural Experiment Station journal paper no. 7780.

Literature Cited

1. AOAC. 1990. Official methods of analysis. 15th ed. Assoc. Off. Anal. Chem., Arlington, VA.

2. Brown, A. R., and Almodares, A. 1976. Quantity and quality of triticale forage compared to other small grains. Agron. J. 68:264-266.

3. Cherney, J. H., and Marten, G. C. 1982. Small grain crop forage potential: I. Biological and chemical determinants of quality, and yield. Crop Sci. 22:227-231.

4. Corrall, A. J., Heard, A. J., Fenlon, J. S., Terry, C. P., and Lewis, G. C. 1977. Whole crop forages: relationship between stage of growth, yield and forage quality in small-grain cereals and maize. Tech. Report No. 22. The Grassland Research Institute, Hurley, Berkshire, UK.

5. Edmisten, K. L., Green, J. T., Jr., Mueller, J. P., and Burns, J. C. 1998. Winter annual small grain forage potential. I. Dry matter yield in relation to morphological characteristics of four small grain species at six growth stages. Commun. Soil Sci. Plant Anal. 29:867-879.

6. Edmisten, K. L., Green, J. T., Jr., Mueller, J. P., and Burns, J. C. 1998. Winter annual small grain forage potential. II. Quantification of nutritive characteristics of four small grain species at six growth stages. Commun. Soil Sci. Plant Anal. 29:881-899.

7. Goering, H. K., and Van Soest, P. J. 1970. Forage fiber analyses. Agric. Handbook No. 379. USDA-ARS, Washington, DC.

8. Helsel, Z. R., and Thomas, J. W. 1987. Small grains for forage. J. Dairy Sci. 70:2330-2338.

9. Juskiw, P. E., Helm, J. H., and Salmon, D. F. 2000. Forage yield and quality for monocrops and mixtures of small grain cereals. Crop Sci. 40:138-147.

10. Large, E. C. 1954. Growth stages in cereals: Illustration of the Feekes scale. Plant Path. 3:128-129.

11. Maloney, T. S., Oplinger, E. S., and Albrecht, K. A. 1999. Small grains for fall and spring forage. J. Prod. Agric. 12:488-494.

12. Mertens, D. R. 2002. Gravimetric determination of amylase-treated neutral detergent fiber in feeds with refluxing in beakers or crucibles: collaborative study. J. AOAC Intl. 85:1217-1240.

13. Mertens, D. R. 2004. Using different grinder mills and screen apertures to prepare samples affects fiber analysis. Pages 61-63 in: U.S. Dairy Forage Res. Center 2003 Res. Rep. USDA-ARS, Madison, WI.

14. Poysa, V. W. 1985. Effect of forage harvest on grain yield and agronomic performance of winter triticale, wheat and rye. Can. J. Plant Sci. 65:879-888.

15. Schwarte, A. J., Gibson, L. R., Karlen, D. L., Liebman, M., and Jannink J. -L. 2005. Planting date effects on winter triticale dry matter and nitrogen accumulation. Agron. J. 97:1333-1341.

16. Shenk, J. S., and Westerhaus, M. O. 1991. Population structuring of near infrared spectra and modified partial least squares regression. Crop Sci. 31:1548-1555.

17. Steel, R. G. D., and Torrie, J. H. 1980. Principles and procedures of statistics. 2nd ed. McGraw-Hill, Inc., New York.