HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Fonseca, C.E.L. Articles by Viands, D.R. Search for Related Content PubMed Articles by Fonseca, C.E.L. Articles by Viands, D.R. Agricola Articles by Fonseca, C.E.L. Articles by Viands, D.R.

CROP BREEDING, GENETICS & CYTOLOGY

Near Infrared Reflectance Spectroscopy Prediction and Heritability of Neutral Detergent-Soluble Fiber in Alfalfa

^a EMBRAPA-Cerrados, Rodovia BR 020 Km 18, Planaltina-DF, Brazil 73301-970
^b Dep. of Plant Breeding, 523 Bradfield Hall, Cornell Univ., Ithaca, NY 14853-1902 USA
^c Dep. of Animal Science, 325 Morrison Hall, Cornell Univ., Ithaca, NY 14853-4801 USA

drv3{at}cornell.edu

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Increasing neutral detergent-soluble fiber (NDSF) concentration has been considered a strategy for breeding alfalfa (Medicago sativa L.) with improved quality. Pectin is the main component of NDSF and is highly digestible, but the assay for NDSF is time-consuming. The objectives were (i) to determine the potential use of near infrared reflectance spectroscopy (NIRS) to predict NDSF concentration in alfalfa and (ii) to estimate NDSF heritability and gain from selection. Herbage was sampled five times across 2 yr from two populations of half-sib (HS) progeny tested at Ithaca, NY. Calibration equations were developed by multiple linear regression of laboratory values of NDSF on NIRS spectra. Samples from one harvest were assayed twice to compare extraction methods. Means of NDSF from 90- and 80%-ethanol extractions were 213.1 ± 7.21 and 179.5 ± 5.89 g kg^-1, respectively. Assay standard deviations were low for both ethanol concentrations. Because a better calibration equation was obtained with the 80%-ethanol assay, all remaining samples were assayed by that technique. Standard errors of calibration were small, ranging from 4.71 to 7.54 g kg^-1. Coefficients of multiple determination (R²) ranged from 0.72 to 0.97. Most heritability estimates for NDSF were significantly greater than zero, ranging from -0.03 ± 0.20 to 0.56 ± 0.17 and averaging 0.41. Expected gain from selection of parents based on HS progeny ranged from 4.6 to 19.0 g kg^-1 and averaged 10.3 g kg^-1. Selecting parents with higher NDSF concentrations predicted by NIRS evaluations of HS progenies may improve forage quality of alfalfa.

Abbreviations: EIR, ethanol-insoluble residue • EIRCP, ethanol-insoluble residue crude protein • EIROM, ethanol-insoluble residue organic matter • HS, half-sib • NDF, neutral detergent-insoluble fiber • NDFCP, neutral detergent-insoluble fiber crude protein • NDSF, neutral detergent-soluble fiber • NIRS, near infrared reflectance spectroscopy • SEC, standard error of calibration • SEL, standard error of laboratory determinations • SEP, standard error of prediction • and SEP(C), standard error of prediction corrected for bias

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
ALFALFA is the dominant forage legume in dairy cattle rations in the northern USA. The factors associated with its widespread use are high yield and high forage quality. Cultivars with high leaf/stem ratio, high crude protein, and low fiber concentrations are the results of 20 yr of research in breeding alfalfa for higher forage quality and consequently higher digestibility. Increasing digestibility by increasing pectin concentration has been considered a new strategy for breeding alfalfa with superior quality characteristics (Hatfield, 1996; Van Soest, 1995; Viands, 1995a, 1995b). Pectin is the predominant component of NDSF in alfalfa (Hall et al., 1997; Van Soest et al., 1991). Neutral detergent-soluble fiber refers to the portion of the forage soluble in neutral detergent solution and resistant to mammalian enzymes, and contains the pectin, ß-glucan, and fructan fiber-fractions (Van Soest et al., 1991). A high NDSF concentration in alfalfa forage is desirable because: (i) alfalfa pectin is rapidly and completely fermented by rumen microorganisms, providing a better utilization of the high protein levels in current cultivars; (ii) pectin fermentation does not decrease the ruminal pH, often associated with animal acidosis; and (iii) pectin is not subject to metabolic turnover in the plant, as are sucrose and starch (Hatfield and Weimer, 1995; Hatfield, 1996; Van Soest et al., 1991). Alfalfa forage with higher NDSF would have higher total fiber that is not digestible by mammalian enzymes. However, NDSF is completely fermented by rumen microbes such that selecting for higher NDSF would not be expected to increase the proportion of indigestible fiber in the forage.

No information is available on estimating NDSF in alfalfa for breeding purposes. Large-scale screening by conventional methods is complicated, expensive, and time-consuming. Due to these constraints, the extent of genetic variation and possible gains from selection for NDSF concentration in alfalfa has not been investigated. A simple method that accounts for changes in mass between extracted residues has recently been developed for estimating NDSF concentration of feeds (Hall et al., 1997). Although simpler, its usefulness in a breeding program is still limited because the complete procedure involves several laboratory assays for each sample. Thus, the development of alfalfa populations with improved NDSF concentration is still dependent on faster and easier assay techniques. A potential alternative would be the assay method developed by Hall et al. (1997) as a reference procedure to develop NIRS equations for predicting NDSF in alfalfa.

The successful use of NIRS in forage quality studies was first demonstrated by Norris et al. (1976). Since then, accurate equations have been developed for quality components such as acid detergent fiber, neutral detergent fiber, acid detergent lignin, crude protein, and in vitro dry matter digestibility (Barton and Windham, 1988; Marten et al., 1984, 1988). A NIRS equation for predicting pectin from six cool-season legumes and two cool- and four warm-season grasses resulted in good accuracy, but the authors concluded that better calibration equations may be obtained by using individual species or groups of similar forage species (Fairbrother and Brink, 1990). To our knowledge, no information is available for using NIRS to predict NDSF concentration in alfalfa. The screening of many forage samples in a breeding program would be facilitated by NIRS.

Our objectives were (i) to determine the extent that NIRS predicts NDSF concentration in alfalfa and (ii) to estimate NDSF heritability and expected gains from HS family tests of two alfalfa populations.

	Materials and methods

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Samples
Two genetically broad-based alfalfa populations, NY9505 and NY9515, from the forage breeding program at Cornell University were used. Both populations are a mixture of germplasm of moderate fall dormancy (rating of 4) and were bred primarily for multiple disease resistance. Population NY9505 was also developed from selection for higher hemicellulose concentration.

A random sample consisting of 75 HS families from each population was used to set up two independent HS family tests (Fonseca, 1998). Each experiment was laid out in a three-replicate randomized complete block design with about 25 plants per HS row in each replicate. In June 1995, 6-wk-old seedlings were transplanted from the greenhouse to a field in Ithaca, NY with a mechanical transplanter. Plant spacing was 20 cm within and 90 cm between rows. The soil type was Williamson silt loam (coarse-silty, mixed, mesic Typic Fragiochrept) with 2 to 6% slope. The field was limed and fertilized as recommended for alfalfa for the upper northeastern USA, and insects and weeds were controlled when necessary. In August 1995, the field was overseeded with common timothy (Phleum pratense L.) in order to simulate competition in a sward. Half-sibs were sampled at 6-wk intervals, three times in 1996 beginning in the first week of June (Harvests 1, 2, and 3). In 1996, the prediction equations for Harvests 2 and 3 had higher R² values than did Harvest 1; thus only Harvests 2 and 3 were sampled in 1997.

Each sample weighed about 400 g fresh and was composed of randomly selected stems harvested with manual clippers at a stubble height of 5 cm from 20 or more plants in each HS row. All samples were dried for 72 h in a forced-air oven at 55°C, sequentially ground through a 2-mm screen in a Wiley mill (A.H. Thomas Co., Philadelphia, PA), through a 1-mm screen in a Udy cyclone mill (Udy Corp., Boulder, CO), and stored in 0.2 L plastic bags for laboratory analyses (Whirl-Pak by NASCO, Fort Atkinson, WI).¹

Spectra from 1100 to 2498 nm for 2430 samples, 486 from each harvest, were collected on a Pacific Scientific 6350 scanning monochromator (Pacific Scientific, Silver Springs, MD).

Neutral Detergent-Soluble Fiber Laboratory Analysis
From each harvest, 56 to 60 samples (12–13% of the total) were randomly selected, assayed for NDSF, and used to calibrate the NIRS. To begin the NDSF study, samples from Harvest 2 in 1996 were selected because they exhibited the best calibration equations for NDF, ADF, CP, acid detergent lignin, and in vitro dry matter true digestibility (Fonseca, 1998). Neutral detergent-soluble fiber analyses were performed according to procedures described by Hall et al. (1997) with a few modifications. Neutral detergent-soluble fiber was calculated as the difference between ethanol-insoluble residue organic matter (EIROM) and neutral detergent-insoluble fiber (NDF) after correction for ethanol-insoluble residue crude protein (EIRCP) and neutral detergent-insoluble fiber crude protein (NDFCP) using the formula: . All values were expressed as g kg^-1 of original sample dry matter. In the method described by Hall et al. (1997), NDSF was corrected for ethanol-insoluble residue starch. Because the starch concentration in ethanol-insoluble residue (EIR) of alfalfa is very small, ranging from 6 to 17 g kg^-1 of sample dry matter (Hall et al., 1997), no attempt was made to correct NDSF for starch.

Ethanol-insoluble residue was prepared by stirring 0.5 g of sample in 100 mL of 90% (v/v) ethanol on a magnetic stir plate for 4 h. Each EIR was filtered through a coarse porosity Gooch crucible, washed twice with 90% ethanol and then twice with acetone, vacuum-filtered until dry, oven-dried at 105°C for 16 h, and weighed while hot. Ethanol-insoluble residue organic matter was calculated as the difference in EIR weight before and after ashing at 520°C for 5 h. Ethanol-insoluble residue for N analysis was prepared by stirring 1.0 g of sample in 200 mL of 90% ethanol and then it was filtered through Whatman 54 filter paper (Fisher Scientific, Pittsburgh, PA), washed twice with 90% ethanol and then twice with acetone, vacuum filtered until dry, and then oven dried at 55°C for 72 h. Neutral detergent-insoluble fiber analysis was performed by using sodium sulfite during refluxing time and heat-resistant -amylase (ANKOM Tech. Corp., Fairport, NY) according to Van Soest et al. (1991) as modified by Komarek et al. (1994). Neutral detergent-insoluble residue for N analysis was prepared by digesting 1.5 g of sample according to Van Soest et al. (1991) using sodium sulfite and heat-resistant amylase during refluxing and then oven drying at 105°C for 16 h. Kjeldahl N concentration was determined in a Tecator Kjeltec Auto 1030 digestion–distillation system (Tecator AB, Höganäs, Sweden). Crude protein in ethanol and in NDF residues was calculated as N times 6.25.

For Harvest 2 in 1996, all EIR determinations are means of two replicates. In an attempt to improve the calibration results obtained with 90%-ethanol extraction, the above procedure and the same samples from Harvest 2 in 1996 were used for new EIR and EIRCP determinations by replacing 90% with 80% ethanol. Eighty percent ethanol is the most commonly used concentration for extraction and precipitation of carbohydrates (Association of Official Analytical Chemists, 1990). Because NDF and NDFCP determinations are independent from EIR determinations, the same estimates for both fractions were used to calculate the NDSF from both ethanol concentrations.

Analysis of variance comparing NDSF composition of 90- and 80%-ethanol extractions was performed as a completely randomized design by using ethanol concentrations and samples as the main factors. A simple correlation coefficient was computed for NDSF concentration between the ethanol concentrations. Calibration statistics were used to compare the equations developed for each ethanol concentration.

Calibration
Neutral detergent-soluble fiber equations were developed for each harvest and year using the Infrasoft International software program (ISI, Port Matilda, PA) CAL, version 1.5, 1985. The recommended statistic for comparing two calibration equations when external validation samples are not available is the standard error of prediction (SEP) of an internal validation set (Windham et al., 1989). For each harvest, every third sample in the calibration file was used for equation validation tests. Selection of equation terms was based either on all possible combinations of four wavelengths or, when more than four terms were selected, on a combination of modified stepwise regression and all possible combinations of four wavelengths. The best calibration equations were chosen by the optimum combination of the following statistics: (i) the difference between standard deviations (SD) of calibration and validation samples <20%; (ii) calibration equations with small standard error of calibration (SEC), large R², and F-test > 10.0 on each selected wavelength; and (iii) validation samples with the lowest standard error of prediction corrected for bias [SEP(C)] and no more than 1.33 larger than the SEC, small bias, large R², and slope closer to 1.0 (Windham et al., 1989). Final equations were derived by using all calibration samples, omitting calibration and/or validation outliers.

Heritability and Expected Gain
Near infrared reflectance spectroscopy equations for individual harvests were used to predict NDSF concentrations of all samples from both field experiments. Narrow-sense heritability on a HS progeny-means basis was computed for NDSF concentration, within and across harvests and years, according to Nguyen and Sleper (1983). Expected gains from direct selection for NDSF were computed by the formula , where k is the standardized selection differential = 1.76 for 10% selection pressure, h² is the narrow sense heritability estimate for NDSF, and S is the phenotypic standard deviation (Hallauer and Miranda, 1981; Nguyen and Sleper, 1983).

For HS analysis of variance, harvests were considered fixed effects, and replicates, years, and families were random effects, in a cross-classified model. All statistical analyses were performed using JMP — Statistics Made Visual software, version 3.2.1 (SAS Institute, 1994).

	Results and discussion

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Comparison of Ethanol Extractions
Means of the various EIR fractions used to calculate NDSF and means of NDSF were significantly different (P < 0.001) for samples, ethanol concentrations, and their interaction. Extraction with 90%-ethanol concentration resulted in greater EIR, EIR ash, EIROM, EIRCP, and NDSF than extraction with 80%-ethanol concentration (Table 1) . These differences in extractions are similar to those previously reported for alfalfa (Hall et al., 1997), and are explained by greater polysaccharide extraction by lower alcohol concentrations of aqueous ethanol solutions (Gaillard, 1958; Smith, 1973). It is still unclear whether the carbohydrates extracted by 80% ethanol, but not by 90% ethanol, are pectic carbohydrates. Alfalfa pectin is formed by long regions of contiguous galacturonic acid (polygalacturonan) interrupted by regions of rhamnose (rhamnogalacturonan) that contain a high degree of intramolecular substitution and are the major sites of arabinose and galactose branching (Hatfield, 1993). These polysaccharides are not exclusive components of pectins. They are also reported as components of the primary cell wall (Åman, 1993; Carpita and Gibeaut, 1993). Therefore, traces of arabinose, rhamnose, and xylose, and a significant amount of glucan were found in 80%-ethanol extracts of 90% ethanol-EIR from alfalfa leaves and stems (Hall et al., 1997), and could have had their origin either as pectin sidechains or as separate polysaccharides.

A significant sample x ethanol concentration interaction indicated that variability existed among samples for polysaccharide solubility. The 80%-ethanol concentration extracted less polysaccharide than did the 90% concentration in 4 of the 56 samples. Hall et al. (1997) reported a similar interaction, but their sample set included both legumes and grasses. The high standard deviation of NDSF (Table 1) laboratory determinations for both ethanol concentrations indicated that sufficient variation existed for NIRS calibrations.

Near Infrared Reflectance Spectroscopy Calibration of Neutral Detergent-Soluble Fiber
Neutral detergent-soluble fiber SEC from 90%-ethanol extraction was more than twice that for 80%-ethanol extraction (Table 2) , indicating a better fit of the laboratory data with the NDSF 80%-ethanol equation. The SEC values for crude protein, NDF, acid detergent-insoluble fiber, and in vitro dry matter digestibility were lower than or equal to those reported elsewhere (Marten et al., 1984; Windham et al., 1989). The standard error of calibration of NDSF from 80%-ethanol extraction was also lower than that reported for pectin from a wide range of forage species (Fairbrother and Brink, 1990). The standard error of prediction (SEP) for NDSF was 1.4-fold higher than the SEC for NDSF from 90%-ethanol extraction and about equal to the SEC for NDSF from 80%-ethanol extraction. The mean bias for NIRS prediction of NDSF was very small in 90% ethanol (2.7 g kg^-1) and negligible in 80% ethanol (-0.10 g kg^-1), reflecting standard errors of prediction corrected for biases [SEP(C)] similar to SEP for both ethanol concentrations.

Near Infrared Reflectance Spectroscopy Prediction Equations of Neutral Detergent-Soluble Fiber
The best NDSF prediction equations were obtained when laboratory data from individual harvests were used to calibrate the NIRS (Table 3) . The SEP(C) for combined harvests in 1996 was higher than the SEC, while in 1997 the SEP(C) was more than 1.3-fold higher than the SEC; however, the SEP(C) for each individual harvest was about equal to the SEC. The SEC values for predicting NDSF from the five individual harvests ranged from 4.71 to 7.54 g kg^-1. Squared coefficients of multiple determination (R²) of individual harvest NDSF equations ranged from 0.72 to 0.97 (Table 3) and were highly associated with the standard deviations of the laboratory data sets . The lower R² for Harvest 1 in 1996 and for Harvests 2 and 3 in 1997 may be explained by the narrower range in NDSF. Data sets with small ranges have generally been associated with lower R² (Windham et al., 1989). Factors that possibly affected the range of NDSF at individual harvests are: (i) plant maturity—since plants were at early bud stage at Harvest 1 in 1996, at late bud stage at Harvests 2 and 3 in 1997, and at early flower stage at Harvests 2 and 3 in 1996—and (ii) seasonal climate variations, such as colder than normal conditions before Harvest 1 in 1996 and colder conditions with irregular rainfall distribution between Harvests 1 and 2 in 1997 (Northeast Regional Climate Center, 1998).

The variance component estimate for HS families across years from NY9515 was significant (0.205 ± 0.062) and was greater than those for HS family x environment interactions (from 0.008 ± 0.003 to 0.156 ± 0.063). Across harvests within each year, estimates of variance components for NY9515 HS means were significant (0.186 ± 0.057 and 0.170 ± 0.052 for 1996 and 1997, respectively). Also, they were greater than those of HS family x harvest interactions (0.057 ± 0.055 and 0.098 ± 0.035 for 1996 and 1997, respectively). These results indicate that genotype x environment interaction effects were small relative to genetic effects. Jung et al. (1997) reported similar results in which small magnitudes of genotype x environment interaction effects were obtained for other NIRS-predicted quality traits from the USA alfalfa core collection evaluation. Buxton and Casler (1993) also reported genotype x environment interaction effects as generally absent or small. In contrast, the variance component for NY9505 HS families in 1996 was greater than zero (0.195 ± 0.064) but of the same magnitude as for HS family x harvest (0.181 ± 0.056).

Heritability estimates for NDSF across harvests or years were moderate, ranging from 0.35 to 0.54 for NY9515 and from -0.03 to 0.52 for NY9505 (Table 4) . Seven of the eight heritability estimates were significantly greater than zero for NY9515, while five were significant for NY9505. Heritability estimates across harvests or years were higher than for individual harvests of NY9515. For NY9505, only the heritability estimate across harvests for 1996 was higher than for individual harvests. Heritability estimates were more consistent and larger for NY9515. The difference was caused by the larger HS family variance component estimates in population NY9515.

	Conclusions

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Although both ethanol concentrations were consistent for NDSF laboratory determinations, 80%-ethanol concentration was preferable to 90% for NIRS calibrations due to the better precision of the equations. The use of NIRS to predict NDSF concentration in alfalfa samples resulted in equations with very low SEC and SEP(C) and moderate to high R² values, suggesting it as an acceptable screening tool to quickly evaluate large numbers of forage samples. However, a sample of selected genotypes should be verified by laboratory reference methods due to potential bias of NIRS predictions, as reported by Buxton and Mertens (1991). Calibration samples randomly selected from a population with a broader range of NDSF concentrations may improve the calibrations. Additional studies on other genetically diverse populations and at different locations may result in an expanded range of NDSF concentrations. More genetic variability is expected among individual plants than among HS families. Also, the range in NDSF values was greater when forage was harvested at the early flower stage rather than the bud stage. Thus, harvesting on a fixed 6-wk interval is not recommended. Significant genetic variability existed among HS families in both populations for NDSF predicted by NIRS equations. In view of the obtained heritability estimates and expected gains, selection of parental genotypes with high NDSF concentrations on the basis of HS family evaluation and NIRS equations is likely to be effective.

	ACKNOWLEDGMENTS

 
Special thanks to Dr. James B. Robertson for his advice, discussions, and assistance. We thank Dr. Peter Schofield and Dr. Mike Van Amburgh for the use of their laboratory equipment and assistance. We also thank Pedro Arcuri and Dr. Francisco Juarez Lagunes for assistance on laboratory procedures.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Part of a dissertation submitted by the senior author in partial fulfillment of the requirements for a Ph.D. degree at Cornell University. Joint contribution of EMBRAPA-Cerrados and Dep. of Plant Breeding, Cornell Univ.

¹ Use of specific products does not constitute an endorsement or recommendation by Cornell University and does not imply approval to the exclusion of other suitable ones.

Received for publication July 1, 1998.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 

• man P. Composition and structure of cell wall polysaccharides in forages. In: Jung et al H.G., ed. Forage cell wall structure and digestibility. Madison, WI: ASA, CSSA, and SSSA, 1993:183-199.
• Association of Official Analytical Chemists. Official method of analysis, 15th ed Arlington, VA: AOAC, 1990.
• Barton F.E., II, Windham W.R. Determination of acid-detergent fiber and crude protein in forages by near-infrared reflectance spectroscopy: Collaborative study. J. Assoc. Off. Anal. Chem. 1988;71:1162-1167.[Medline]
• Buxton D.R., Casler M.D. Environmental and genetic effects on cell wall composition and digestibility. In: Jung et al H.G., ed. Forage cell wall structure and digestibility. Madison, WI: ASA, CSSA, and SSSA, 1993:685-714.
• Buxton D.R., Mertens D.R. Errors in forage-quality data predicted by near infrared reflectance spectroscopy. Crop Sci. 1991;31:212-218.[Abstract/Free Full Text]
• Carpita N.C., Gibeaut D.M. Structural models of primary cell walls in flowering plants: Consistency of molecular structure with the physical properties of the walls during growth. Plant J. 1993;3:1-30.[ISI][Medline]
• Fairbrother T.E., Brink G.E. Determination of cell wall carbohydrates in forages by near infrared reflectance spectroscopy. Anim. Feed Sci. Technol. 1990;28:293-302.
• Fonseca, C.E.L. 1998. Breeding for neutral detergent-soluble fiber and associations among forage quality traits, vigor, and disease resistance in alfalfa. Ph.D. diss. Cornell Univ., Ithaca, NY. Diss. abstr. 59–06B:2494.
• Gaillard B.D.E. A detailed summative analysis of the crude fibre and nitrogen-free extractive fractions of roughages. J. Sci. Food Agric. 1958;9:170-177.
• Hall M.B., Lewis B.A., Van Soest P.J., Chase L.E. A simple method for estimation of neutral detergent-soluble fibre. J. Sci. Food Agric. 1997;74:441-449.
• Hallauer A.R., Miranda J.B. Quantitative genetics in maize breeding. Ames: Iowa State Univ. Press, 1981.
• Hatfield R.D. Cell wall polysaccharide interactions and degradability. In: Jung et al H.G., ed. Forage cell wall structure and digestibility. Madison, WI: ASA, CSSA, and SSSA, 1993:285-313.
• Hatfield, R.D. 1996. Improving alfalfa utilization: Biochemical strategies for improving alfalfa utilization in dairy production. p. 15–22. in Proceedings of the Dairy Forage Research Center Conference with Dairy and Forage Industries, Madison, WI. 1996. Agricultural Research Service, USDA.
• Hatfield R.D., Weimer P.J. Degradation characteristics of isolated and in situ cell wall lucerne pectic polysaccharides by mixed ruminal microbes. J. Sci. Food Agric. 1995;69:185-196.[ISI]
• Jung H.G., Sheaffer C.C., Barnes D.K., Halgerson J.L. Forage quality variation in the U.S. alfalfa core collection. Crop Sci. 1997;37:1361-1366.[Abstract/Free Full Text]
• Komarek A.R., Robertson J.B., Van Soest P.J. A comparison of methods for determining ADF using the filter bag technique versus conventional filtration. J. Dairy Sci. 1994;77:114.
• Marten G.C., Brink G.E., Buxton D.R., Halgerson J.L., Hornstein J.S. Near infrared reflectance spectroscopy analysis of forage quality in four legume species. Crop Sci. 1984;24:1179-1182.[Abstract/Free Full Text]
• Marten G.C., Halgerson J.L., Sleper D.A. Near infrared reflectance spectroscopy evaluation of ruminal fermentation and cellulose digestion of diverse forages. Crop Sci. 1988;28:163-167.[Abstract/Free Full Text]
• Nguyen H.T., Sleper D.A. Theory and application of half-sib matings in forage grass breeding. Theor. Appl. Genet. 1983;64:187-196.[ISI]
• Norris K.H., Barnes R.F., Moore J.E., Shenk J.S. Predicting forage quality by infrared reflectance spectroscopy. J. Anim. Sci. 1976;43:889-897.[Abstract/Free Full Text]
• Northeast Regional Climate Center. Monthly climate highlights. [Online] Cornell University, Ithaca, NY. Available at http://www.snow.cit.cornell.edu/climate/ithaca1998.
• SAS Institute. JMP statistics and graphics guide. Version 3. Cary, NC: SAS Inst, 1994.
• Smith D. The nonstructural carbohydrates. In: Butler G.W., Bailey R.W., eds. Chemistry and biochemistry of herbage. London: Academic Press, 1973:105.
• Van Soest, P.J. 1995. What constitutes alfalfa quality: new considerations. p. 1–15. In 25th National Alfalfa Symposium, Liverpool, NY. 27–28 Feb. 1995. Certified Alfalfa Seed Council, Davis, CA.
• Van Soest P.J., Robertson J.B., Lewis B.A. Methods for dietary fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal nutrition. J. Dairy Sci. 1991;74:3583-3597.[Abstract]
• Viands, D.R. 1995a. What breeding objectives really will improve forage quality of alfalfa? p. 24–28. In 25th National Alfalfa Symposium, Liverpool, NY. 27–28 Feb. 1995. Certified Alfalfa Seed Council, Davis, CA.
• Viands, D.R. 1995b. Breeding for higher forage quality. p. 13–17. In 11th Eastern Forage Improv. Conf., Ottawa, Ontario, Canada. 6–8 July 1995. USDA, Beltsville, MD.
• Windham R.F., Mertens D.R., Barton F.E., II. Protocol for NIRS calibration: Sample selection and equation development and validation. In: Marten et al G.C., ed. Near infrared reflectance spectroscopy (NIRS): Analysis of forage quality. Washington, DC: USDA ARS, 1989:96-103 Handb. 643..

This article has been cited by other articles:

				I. Y. Tecle, D. R. Viands, J. L. Hansen, and A. N. Pell Response from Selection for Pectin Concentration and Indirect Response in Digestibility of Alfalfa Crop Sci., March 27, 2006; 46(3): 1081 - 1087. [Abstract] [Full Text] [PDF]

				C.E.L. Fonseca, D.R. Viands, J.L. Hansen, and A.N. Pell Associations among Forage Quality Traits, Vigor, and Disease Resistance in Alfalfa Crop Sci., September 1, 1999; 39(5): 1271 - 1276. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (4) Citing Articles via Google Scholar Google Scholar Articles by Fonseca, C.E.L. Articles by Viands, D.R. Search for Related Content PubMed Articles by Fonseca, C.E.L. Articles by Viands, D.R. Agricola Articles by Fonseca, C.E.L. Articles by Viands, D.R.