Research in the Klamath Basin
20 07 Annual R eport
Selection and Improvement of Teff Accessions for
Improved Forage Growth, Yield, and Nutritional
Quality, 2007
Richard J. Roseberg, Brian A Charlton, and Rachel A Shuck1
Introduction
As we described previously and elsewhere in this annual report, teff (Eragrostis
tef [Zucc.], Poaceae) is a warm season (C4) annual tropical grass that can produce good
quality forage during a short summer time frame, and thus has the potential to be a viable
crop choice when forage producers: a) desire a quick-growing, high quality mid-summer
annual forage; b) foresee less-than full season irrigation water supply; c) need an
emergency crop due to crop failure; or d) need a one-year forage rotation crop between
alfalfa stands. We began growing teff in a quasi-commercial field setting in 2003, and
began more rigorous testing of teff’s agronomic requirements in 2005 after
1
Associate Professor, Assistant Professor, and Research Technician, respectively, Klamath Basin Research
& Extension Center, Klamath Falls, OR.
Acknowledgments: We wish to thank Vicki Bradley at the USDA-ARS, WRPIS for providing the teff
accession seed, as well as helpful information about the accessions and the USDA germplasm system.
______________________________________________________________________________
Klamath Basin Research and Extension Center
Selection and Improvement of Teff Accessions for Improved
Forage Growth, Yield, and Nutritional Quality, 2007 Page 42
Research in the Klamath Basin
20 07 Annual R eport
unprecedented response from a popular press article about our early efforts indicated
widespread national interest in this “new” alternative forage crop (Zenk, 2005; Roseberg
et al., 2006).
Despite the promising results seen thus far, there are very few commercially
available brands of teff seed. Much of the teff available in commerce are common
landraces, not released varieties, and thus have varying degrees of uniformity and
performance. Despite the genetically narrow and uneven commercial seed supply, the
USDA germplasm system contains many accessions collected from teff’s native habitat
in east Africa. These accessions include a wide range of plant morphological types, seed
color, and other characteristics. Until recently, most interest in teff has focused on its
qualities relative to the seed used in human food (mainly enjera, a staple of Ethiopian
cuisine), but not its forage qualities. Because of this lack of germplasm evaluation and
varietal development, we decided to begin a multi-year evaluation of teff accessions to
evaluate the range of yield, quality, and morphological factors available, with the goal of
selecting improved cultivars for possible release as commercial varieties. 2007 was the
second year of this effort.
Objective
To evaluate the available accessions of teff with a goal of documenting forage
growth, yield, quality, and other parameters, and to begin selecting, and saving seed
from, superior types, with the eventual goal of releasing improved teff varieties to the
commercial market.
Procedures
Based on our results in 2006, we narrowed down the list of likely successful teff
accessions from the original 367 accessions to about 90. Based on seed availability, we
obtained and seeded 73 accessions using new seed obtained from the USDA-ARS,
Western Region Plant Introduction Station in Pullman, WA. In most cases we received
2.25 g, which was enough to plant two replicates; in a few cases we only had enough
seed to plant one replicate. Each individual plot was a single row, 15 feet long, seeded at
about ¼ inch depth, using a single-row V-belt planter on June 8. No fertilizer was applied
prior to or during planting. Four named teff seed brands that we also grew in the
concurrent variety trial were seeded in the same way as the numbered accessions to
provide a comparison with commercially available seed types.
On June 11, Solution 32 liquid fertilizer was applied through the irrigation water
at a rate of 100 lb/ac N. Weedmaster® herbicide (dicamba + 2,4-D amine, BASF) was
applied at 1.5 pint/ac on June 26, thus applying dicamba at 0.19 lb ai/ac plus 2,4-D at
0.54 lb ae/ac. No crop injury was apparent at any time after spraying.
A total of 12.65 inches of irrigation was applied on 16 dates between June 4 and
September 4. Rainfall amounts totaled 0.44 inch in June, 0.57 inch in July, 0.18 inch in
August, and 0.15 inch in September.
Observations on emergence and early seedling vigor were made in late June.
Observations on plant height, growth habit (upright or prostrate), leaf width,
______________________________________________________________________________
Klamath Basin Research and Extension Center
Selection and Improvement of Teff Accessions for Improved
Forage Growth, Yield, and Nutritional Quality, 2007 Page 43
Research in the Klamath Basin
20 07 Annual R eport
lateness/earliness of seedhead formation, and seedhead/leaf color were made on August 3
and September 24.
For each single-row plot, a one-foot section of the row was cut at two-inch height
on August 6 to provide an estimate of dry matter production. After drying and weighing,
samples were ground to 2-mm-sieve size in a Wiley Mill (Arthur H. Thomas Co.) and to
1-mm-sieve size in an Udy Mill (UDY Corporation) before being analyzed in a near
infrared spectrophotometer (NIRS) (NIRSystems, FOSS, NA, Minneapolis, MN) to
determine forage quality. Quality testing at KBREC is accomplished using the NIRS and
equations developed by the NIRS Consortium, Madison, WI (NIRS Consortium, 2007).
Calculated forage quality parameters included crude protein (CP), total digestible
nutrients (TDN), relative feed value (RFV) and relative forage quality (RFQ). We used
NIRS equations developed for other grasses due to the limited data available for teff.
After observations were made on September 24, all plants of one replicate of each
accession were cut by hand and placed into large seed bags for drying and storage. After
yield, quality, and morphology data was analyzed, selections were made for the 2008
season, and seed from those selections were individually threshed and cleaned using a
hammer mill, hand-held seed screens, and a Clipper table-top seed cleaner.
Results & Discussion
The NIRS quality analysis produced the following observations on the population:
CP ranged from 22.1 to 12.5%, a range similar to what was observed in 2006. Almost
half of the accessions had CP above 18.0%, and only five were less than 15.0%. RFV
ranged from 126.4 to 99.2, a similar range as observed in 2006. Four accessions had RFV
of greater than 120, and only about 1/3 of the accessions had an RFV below 110. Because
the forage yield estimates were based on only a one-foot section of a single row, their
absolute value is questionable, although relative comparisons between accessions are
______________________________________________________________________________
Klamath Basin Research and Extension Center
Selection and Improvement of Teff Accessions for Improved
Forage Growth, Yield, and Nutritional Quality, 2007 Page 44
Research in the Klamath Basin
20 07 Annual R eport
more valid. Estimated dry matter yield ranged from 5.7 ton/ac to 1.2 ton/ac, with about
2/3 of the accessions exceeding 3.0 ton/ac. The four commercial seed brands used as
controls (VAT1, Tiffany, Dessie, and Pharoah) had yields of 2.6 to 4.2 ton/ac using the
same methods. While these yield values were not used as an independent selection
screen, they were used to confirm whether the other screens seemed to accurately reflect
superior-looking plant types.
Growth characteristics such as plant height, stem/leaf type (fine, medium, or
wide), growth habit (upright or prostrate), seed head maturity (early or late), and
seedhead/leaf color, and seed color were measured or rated on a scale of one, two, or
three. As in 2006, we decided that the most desirable plant traits were tall (height), fine
(stem & leaf width), erect (growth habit), and late-maturing (delayed seedhead
formation). However, since most of the remaining selections were relatively tall, erect,
and later-maturing, we also placed strong emphasis on CP and RFV values, and to a
lesser extent the estimated yield. For example, if a selection had wide leaves, but
otherwise had favorable characteristics such as high CP and RFV, it was not rejected. In
general, the selections that were retained from 2007 to 2008 were among the upper half in
both protein and yield, and also had other favorable characteristics. We also kept a couple
selections that had more intense purple seedhead or leaf color (for possible ornamental
use), a couple with very white seed (for possible human food use), and three selections
with the lowest CP values so that we would retain selections with a wide range of CP for
analytical comparison. In the end, seed from 38 selections were threshed, cleaned and
retained for further selection and evaluation in 2008.
Conclusions
Because of the selections made in 2006, there was not as much variability in
growth characteristics in the 2007 selections. However, there was still a fairly wide
variation in forage quality, timing of maturity, and plant color among the selections in
2007. Some of them seemed to have growth and quality characteristics that were superior
to seed brands that are currently available on the commercial market. While there were
no accessions that were clearly superior in all factors measured, a careful evaluation of
the factors resulted in selection of a list that were at least “better than average” for most
(or all) parameters measured. We plan to re-select from this superior group of accessions
and re-plant the best candidates in 2008 using the saved, cleaned seed from the 2007 trial.
In addition to conducting another round of selection and seed collection based on
morphology and forage quality, we plan to seed a replicated forage yield trial using
standard forage yield plot methods to get a more accurate measurement of forage yield.
References
Roseberg, R.J., S. Norberg, J.E. Smith, B.A. Charlton, K.A. Rykbost, and C. Shock.
2006. Yield and quality of teff forage as a function of varying rates of applied
irrigation and nitrogen. In: Research in the Klamath Basin 2005 Annual Report.
OSU-AES Special Report 1069:119-136.
Zenk, Peg. 2005. Tons of teff. Hay & Forage Grower. February, p.4.
______________________________________________________________________________
Klamath Basin Research and Extension Center
Selection and Improvement of Teff Accessions for Improved
Forage Growth, Yield, and Nutritional Quality, 2007 Page 45
Research in the Klamath Basin
20 07 Annual R eport
NIRS Forage and Feed Testing Consortium, June 2007 grass hay calibration, file name:
gh50-2. Parameters used: DM, CP, ADF, dNDF48, NDF, Ca, P, K, Mg, Ash.
______________________________________________________________________________
Klamath Basin Research and Extension Center
Selection and Improvement of Teff Accessions for Improved
Forage Growth, Yield, and Nutritional Quality, 2007 Page 46