Published online 31 May 2007
Published in Crop Sci 47:929-934 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
CROP BREEDING & GENETICS
Genetic Variation for Biomass Production in Prairie Cordgrass and Switchgrass
Arvid Boe* and
D. K. Lee
Plant Science Dep., South Dakota State Univ., Brookings, SD 57007. This research was supported by the South Dakota Agricultural Experiment Station, the USDOE through contract DE-FC36-02G012028, A000 with Great Plains Institute for Sustainable Development, Minneapolis, MN, and USDOE's Biomass program through contract DE-A105-900R2194 with Oak Ridge National Lab. (ORNL). ORNL is managed by UT-Battelle, LLC, for the USDOE under contract DE-AC05-00OR22725. South Dakota Agric. Expt. Stn. Journal Series No. 3450
* Corresponding author (arvid.boe{at}sdstate.edu).
 |
ABSTRACT
|
|---|
Prairie cordgrass (Spartina pectinata Link.) is tall, rhizomatous, and native to marshes, drainage ways, and moist prairies in North America. Our objectives were to determine genetic variation among cordgrass populations for biomass production, to describe the distribution of biomass among phytomers and between leaf and stem components of cordgrass, to compare biomass production and composition of cordgrass to switchgrass (Panicum virgatum L.), and to determine heritability for biomass production in switchgrass. Seven populations of cordgrass and ‘Cave-In-Rock’, ‘Summer’, and ‘Sunburst’ switchgrass were harvested in October in 2001 through 2004. Mean biomass production across years ranged from 5.1 to 7. 9 Mg ha–1 among cordgrass populations. Yields of cordgrass (6.0 Mg ha–1) were similar to Cave-In-Rock (6.8 Mg ha–1) for the first two years. However production in the fourth year was greater for cordgrass (6.8 Mg ha–1) than Cave-In-Rock (2.0 Mg ha–1). Two cordgrass populations produced more biomass (9.3 Mg ha–1) than Summer and Sunburst (4.8 Mg ha–1) in the fourth year. Leaf comprised 70% of the biomass of cordgrass, and differences occurred among phytomers for leaf and internode traits. Cellulose and hemicellulose concentrations were similar for cordgrass and switchgrass, but cordgrass had higher levels of total N and ash. Narrow-sense heritability estimates for biomass production in Summer and Sunburst switchgrass were 0.6. Biomass production of native warm-season grasses intended for biofuel purposes in the northern Great Plains may be enhanced by selecting among populations of cordgrass and among families within cultivars of switchgrass.
Abbreviations: ADF, acid detergent fiber; ADL, acid detergent lignin; NDF, neutral detergent fiber; TN, total N.
|
INTRODUCTION
|
|---|
PRAIRIE CORDGRASS (Spartina pectinata Link.) is a tall, rhizomatous, warm-season species found in marshes, wet meadows, potholes, and drainage ways throughout Canada to 60° N lat and throughout the continental USA, with the exceptions of Louisiana to South Carolina in the Southeast, and California, Nevada, and Arizona in the West (Hitchcock, 1950; Mobberley, 1956; Stubbendieck et al., 1982). The genus Spartina has the most northerly distribution of any of the C4 perennial grasses (Potter et al., 1995). Prairie cordgrass is commonly associated with sedges (Carex sp.) and rushes (Juncus sp.). It is recognized for having tolerance to salinity and value for wetland revegetation, streambank stabilization, and wildlife habitat. Selected germplasms have been released by the USDA-NRCS from natural populations in Minnesota, North Dakota, and South Dakota (‘Red River Germplasm’) and Nebraska (‘Atkins Germplasm’) (N. Jensen, pers. comm., 2006).
Weaver (1954) described a prairie cordgrass–dominated community that occurred over hundreds of hectares along the Mississippi and Missouri Rivers and their tributaries in the tallgrass prairie from northern Minnesota to Texas. He concluded that its dominance in dense pure stands was due to its height and propagation by rhizomes. Prairie cordgrass is adapted to soils that are too wet and not sufficiently aerated for big bluestem (Andropogon gerardii Vitman) and switchgrass (Panicum virgatum L.), grows more rapidly than other tall grass prairie grasses, and is conspicuously taller than switchgrass and big bluestem where their distributions overlap (Weaver, 1954). Mobberley (1956) concluded that although prairie cordgrass primarily inhabits marshes, sloughs, and flood plains in the eastern USA and Canada, its primary habitat in the midwestern USA is open dry prairie and high ground along railroad rights-of-way.
Potter et al. (1995) reported that a clone of prairie cordgrass from New Haven, CT, averaged in excess of 10 Mg ha–1 yr–1 without fertilizer over seven successive years of harvest in late October in southeastern England (52° N lat). They concluded that prairie cordgrass and Spartina cynosuroides (L.) had potential for energy efficient production of biomass in eastern England. Madakadze et al. (1998) evaluated prairie cordgrass, switchgrass, big bluestem, indiangrass [Sorghastrum nutans L. (Nash)], and prairie sandreed [Calamovilfa longifolia (Hook) Scribn.] in southwestern Quebec and determined that prairie cordgrass and switchgrass had the greatest potential for biomass production in short-season areas of southern Canada.
Patterns of biomass distribution among leaf and stem components of phytomers have been described for genetically diverse switchgrass cultivars (e.g., Boe and Casler, 2005). However, similar information does not exist for prairie cordgrass. Description of the distribution of biomass among phytomeric subunits within a sward and determination of the relative contributions of leaf and stem fractions of each phytomer to biomass production can provide a framework for developing species-specific morphological approaches to breeding for biomass production.
Lee and Boe (2005) pointed out that stockpiling switchgrass biomass over winter in the northern Great Plains traps valuable moisture in the form of snow and provides cover for wildlife. They also indicated that loss in biomass due to weathering over winter could be partially compensated for by harvesting at a lower stubble height in the spring.
The primary objectives of this study were to determine variation for biomass production among seven populations of prairie cordgrass from eastern South Dakota and to describe the distribution of biomass among phytomers and leaf and stem components of cordgrass. Secondary objectives were to compare biomass production and composition of cordgrass to that of switchgrass and to estimate heritability of biomass production in two switchgrass cultivars adapted to the northern Great Plains.
|
MATERIALS AND METHODS
|
|---|
Seed was collected during January 2000 from seven populations of prairie cordgrass in eastern South Dakota. Five of the populations (PCG 1, 2, 3, 4, and 5) were located on a railroad embankment north of SD Hwy 50 between Vermillion (42° 56' N, 97° W) and Gayville, SD (42° 54' N, 97° 11' W), one (PCG 6) was located near Arlington, SD (44° 23' N, 97° 9' W), and one (PCG 7) was located near DeSmet, SD (44° 29' N, 97° 33' W).
Seedlings of the seven cordgrass populations and ‘Cave-In-Rock’, ‘Summer’, and ‘Sunburst’ switchgrass were transplanted to a field nursery near Aurora, SD (44° 19' N, 96° 42' W) during June 2000. The experiment was a randomized complete block design with two replications. Each replication consisted of single-row eight-plant plots of the seven cordgrass populations, and 30 half-sib families for each of Summer and Sunburst, with each family planted in single-row eight-plant plots. Additionally, three single-row eight-plant plots of Cave-In-Rock were included in each replication. The nursery was intended to provide an environment for among-family comparisons for Summer and Sunburst and among-population comparisons for cordgrass for biomass production in eastern South Dakota. Cave-In-Rock was included to provide a high-yield-potential check. Interrow spacing was 90 cm, and intrarow spacing was 35 cm. Soil type was a Brandt silty clay loam (fine-silty, mixed, superactive, frigid Calcic Hapludolls). Weed control was as needed by rototilling between rows during the springs of 2000 through 2004. No fertilizer was applied during the study. The previous crop was soybean (Glycine max L.). Plots were harvested with a sickle-bar mower at a stubble height of about 10 cm once annually during early October 2001 through 2004. Plot fresh weights were taken in the field at harvest. Moisture samples were dried at 60°C until constant weight for dry matter concentration determination. During the period of the experiment, the annual precipitation at Aurora ranged from 489 mm in 2003 to 642 mm in 2004, bracketing the long-term average of 582 mm.
Biomass data were analyzed by using the Linear Models General AOV procedure in Statistix 7 (Analytical Software, 2000). For comparison of cordgrass populations and switchgrass cultivars, analyses of variance were conducted on individual plot data for the cordgrass populations, means of the three plots in each replication for Cave-In-Rock, and means for the 30 families of Summer and Sunburst in each replication. The 30 families from each of Summer and Sunburst were chosen at random to determine within-cultivar among-family genetic variation for biomass production. Therefore, the grand means of the 30 families estimated their respective cultivar means. Entries (i.e., cordgrass populations and switchgrass cultivars) were main plots, and years were subplots. Summer and Sunburst were developed at South Dakota State University and are well adapted to the northern Great Plains. Cave-In-Rock was included because of its high biomass yield potential in eastern (Casler and Boe, 2003) and central (Lee and Boe, 2005) South Dakota. Entries and years were considered fixed and replications were considered random. Main and interaction effects were considered significant when p < 0.05 for the appropriate mean square ratio. As a result of a significant entry x year interaction, separate analyses of variance were conducted for each year.
Separate analyses of variance were conducted for Summer and Sunburst to provide estimates of variance components for the purpose of determining narrow-sense heritability estimates on a family mean basis for biomass production. The expression for phenotypic variance among family means and formula for calculating heritability from plot totals were from Nyquist (1991). Families, years, and replications were considered random. Asymmetric confidence intervals for heritability estimates were determined as described by Knapp et al. (1985).
Distribution of biomass among leaf and stem components of individual phytomers was determined from a random sample of 12 reproductive tillers severed with a rice knife near ground level from each of five populations located near Vermillion, SD, during January 2000. The number of phytomers tiller–1 was determined, and tillers were separated in to leaf blade, leaf sheath, internode, and inflorescence components for individual phytomers. The numbering system for phytomers used in this article designates the apical phytomer as phytomer 1 and subtending phytomers in sequence (Boe et al., 2000). Leaf and stem components were weighed on a balance with decigram accuracy. Analysis of variance (Analytical Software, 2000) was used to determine the importance of population, tillers within populations, and phytomer as sources of variation in blade, sheath, and internode lengths and weights. Population and phytomer were considered fixed. Main and interaction effects were considered significant when p < 0.05 for the appropriate mean square ratio. Fisher's protected least significant difference was used to separate means. If the F test for phytomer was significant, the linear and quadratic regression partitions for phytomer were tested. If the F test for phytomer x population was significant, variability among populations in linear regression coefficients for phytomer was tested.
To compare chemical composition of overwintered biomass of switchgrass and cordgrass, three replicates of 0.1-m2 plots of two populations of cordgrass (PCG 2 and 7) and three randomly chosen families of Sunburst switchgrass were hand-clipped at soil level during April 2006. Samples were dried at 60°C for 48 h in a forced-air oven and ground in a Wiley mill (Thomas-Wiley Mill Co., Philadelphia, PA) to pass a 1-mm screen. Concentration of neutral detergent fiber (NDF) and acid detergent fiber (ADF) were determined using an Ankom200 Fiber Analyzer (ANKOM Technology Corp., Fairport, NY) (Ankom Technology, 2003a, 2003b), acid detergent lignin (ADL) was determined with a Daisy Incubator II Digestor (ANKOM Technology Corp., Fairport, NY) (Ankom Technology, 2002), and total N (TN) was quantified using a Vario Max CNS elemental analyzer (Elementar Instrument, Mt. Laurel, NJ). Ash concentrations were determined using the methods described by Undersander et al. (1993). To estimate cellulose and hemicellulose yield, the concentration of cellulose and hemicellulose was calculated by the subtraction of ADL from ADF and ADF from NDF, respectively. Analysis of variance (Analytical Software, 2000) was used to determine effects of population on concentrations of cellulose, hemicellulose, ADL, TN, and ash. Fisher's protected least significant difference (p = 0.05) was used to separate means.
|
RESULTS AND DISCUSSION
|
|---|
Biomass Production of Cordgrass and Switchgrass
An entry x year interaction was found for biomass production (ANOVA not shown). For the first production year (2001), PCG 3 outproduced all entries other than PCG 1 and PCG 2, and PCG 1 and PCG 2 produced more biomass than Sunburst. In 2002, PCG 3 outproduced PCG 4, PCG 6, PCG 7, Summer, and Sunburst, and yields of PCG 3 and Cave-In-Rock were similar. In 2003, PCG 3 outproduced all entries other than PCG 2. In 2004, all entries other than Sunburst outproduced Cave-In-Rock, and PCG 2 and PCG 5 outproduced Summer and Sunburst (Table 1). The cordgrass populations were strongly rhizomatous and formed dense sods by the third production year. Stands of Cave-In-Rock were productive for the first 2 yr, but declined substantially, compared with the better adapted Summer and Sunburst, by the fourth production year (Table 1). A prairie cordgrass population from Wisconsin had similar biomass yields to Cave-In-Rock (425 g plant–1 in 1994 and 850 g plant–1 in 1995 averaged across species) in Quebec, but trials of at least 5 yr in duration would be needed to determine persistence characteristics of those populations (Madakadze et al., 1998). That experiment also found cordgrass consistently had higher tiller numbers than switchgrass throughout the growing season.
View this table:
[in this window]
[in a new window]
|
Table 1. Mean biomass production for seven populations of prairie cordgrass (Spartina pectinata Link.) and three cultivars of switchgrass (Panicum virgatum L.) harvested during early October 2001 through 2004 at Aurora, SD.
|
|
All entries showed yield variation across years (Table 1). However, the only conspicuous pattern was a substantial decline in biomass production for Cave-In-Rock between the second and fourth years of production.
The large differences for biomass production among seven prairie cordgrass populations from eastern South Dakota (Table 1) indicated that evaluation of a collection of natural populations in a common environment is an efficient way of identifying promising strains. The germplasm evaluated here may be immediately suitable for seed increase and release without artificial selection, or valuable source populations for methodical selection to improve biomass production, seed production, disease resistance, and/or other agronomic traits.
Phytomer Morphology of Cordgrass
The number of aerial phytomers per reproductive tiller was similar (grand mean = 8.0) for the five populations (PCG 1–5) in their natural habitats near Vermillion, SD. This was expected considering the proximity of the populations. Since tillers were not collected from PCG 6 and 7, we did not determine if number of phytomers per tiller varied between the natural populations from 
42° N lat (PCG 1–5) and 44° N lat (PCG 6 and 7), as has been reported for switchgrass (Boe and Casler, 2005). Madakadze et al. (1998) reported nine leaves per tiller for a prairie cordgrass population from Wisconsin and seven leaves per tiller for Cave-In-Rock when both were grown in Quebec.
Population x phytomer interactions were found for leaf blade weight and internode length. The linear phytomer x population mean square was also significant, indicating variation in linear responses of populations across phytomers for those two traits. Phytomer mean squares were significant for all traits other than internode weight. In every case, with the exception of internode weight, the linear and quadratic partitions for phytomer were significant (Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2. Mean squares for leaf and internode traits of the five uppermost leaf-bearing phytomers for reproductive tillers of five natural populations (PCG 1–5) of prairie cordgrass (Spartina pectinata Link.) harvested during January 2000 near Vermillion, SD.
|
|
Phytomer 5 had the longest leaf blades (91 cm) with phytomer 2 having the shortest (44 cm) for each of the five populations (data not shown). Weaver (1954) reported that the lengths of mature leaves of prairie cordgrass ranged from 75 to 150 cm. Madakadze et al. (1998) reported that the uppermost leaf on tillers of cordgrass in July in Quebec was 113 cm long, about twice as long as the uppermost leaf of switchgrass.
Leaf blades exhibited an acropetal decrease in weight, with blades from phytomer 6 among the heaviest and those from phytomers 2 and 3 the lightest (Table 3). A similar pattern of acropetal decrease in blade length and weight occurred in big bluestem (Boe et al., 2000). Since tillers were harvested during early winter, the inconsistent ranks for populations for blade length and weight for phytomers 4 and 5 may reflect differential weathering of these leaves among populations. Weaver (1954) noted that the bottom two to three leaves of prairie cordgrass were usually dead by midsummer.
View this table:
[in this window]
[in a new window]
|
Table 3. Weights of leaf blades and sheaths from phytomers of reproductive tillers of five natural populations (PCG 1–5) of prairie cordgrass (Spartina pectinata Link.) harvested during early January 2000 near Vermillion, SD.
|
|
Phytomer 2 ranked first (49 cm) and the middle phytomers (3 and 4) ranked last (37 cm) for sheath length for all five populations. Similarly, the length of the sheath from the second phytomer was longer than those from basipetal phytomers in big bluestem (Boe et al., 2000).
Sheaths of phytomer 6 were heavier than those of phytomers 2, 3, and 4 (Table 3). Similarly, sheaths from phytomer 2 were generally lighter than those from basipetal phytomers in big bluestem (Boe et al., 2000).
Although no difference was found among phytomers for internode weight, mean internode weights differed among populations (Table 2), ranging from 0.73 g for PCG 4 to 1.54 g for PCG 5. The lack of differences among phytomers for this trait contrasts to what was reported for big bluestem (Boe et al., 2000) and switchgrass (Boe and Casler, 2005).
The biomass distribution between leaf and stem fractions was quite different for prairie cordgrass than for switchgrass. For mature reproductive tillers of switchgrass, the weight of the internode component was about 1.5 times that of the leaf component (Boe and Casler, 2005). For the cordgrass populations in this study, the weight of the leaf (blade + sheath) component ranged from 1.3 to 3.1 times that of the internode component. In comparison, in southwestern Quebec, the leaf-to-stem ratio for cordgrass was 0.50, nearly twice that of Cave-In-Rock (0.28) (Madakadze et al., 1998). Prairie cordgrass is known for its long, coarse, and tough leaves used by indigenous Americans for thatching lodges and by pioneers for thatching roofs and covering haystacks and corncribs (Weaver, 1954).
Similarities in patterns of lengths and weights of leaf sheaths, leaf blades, and internodes were found for homologous phytomers in reproductive tillers of five populations of cordgrass in their natural habitats in southeastern South Dakota. These data provide evidence for innate morphological controls to biomass partitioning among phytomers in tillers of prairie cordgrass, as was reported previously for big bluestem (Boe et al., 2000).
Comparison of Chemical Composition of Cordgrass and Switchgrass
Cordgrass had higher concentrations of TN and ash and lower concentration of ADL than Sunburst switchgrass in overwintered biomass. Cellulose and hemicellulose concentrations were similar for the two species (Table 4). Cell wall concentrations were similar to those reported for cordgrass and switchgrass in Quebec (Madakadze et al., 1998). Total nitrogen for cordgrass was 15% lower and ash was correspondingly 15% higher than for cordgrass harvested in Quebec in October (Madakadze et al., 1998). Although we did not test both autumn and overwintered biomass in our trial, these differences are expected due to leaching of nitrogen during the overwintering period.
View this table:
[in this window]
[in a new window]
|
Table 4. Chemical composition in April 2006 of overwintered biomass of two populations (PCG 2 and 7) of cordgrass (Spartina pectinata Link.) and ‘Sunburst’ switchgrass (Panicum virgatum L.) at Aurora, SD.
|
|
Ice- and snowstorms occurred at Aurora during the winter of 2005–2006 and caused differential lodging between switchgrass and cordgrass based on visual observations. Switchgrass was well suited for stockpiling over winter because its resistance to lodging made it possible to be windrowed in the spring. Conversely, all of the cordgrass populations were completely lodged and would not have been able to be harvested with conventional hay-making equipment in spring 2006.
Genetic Variation among Switchgrass Families for Biomass
Differences occurred among years and among families within each switchgrass cultivar for biomass production in transplanted rows. Family x replication and family x year were also significant components of the phenotypic variation (ANOVA not shown), similar to what was found for northern upland and southern upland and lowland populations of switchgrass in spaced-plant nurseries in Oklahoma (Das et al., 2004). The range in family means was 2.9 Mg ha–1 for Summer and 3.6 Mg ha–1 for Sunburst. Estimates of heritability among family means were moderately high and similar for the two cultivars (Table 5).
View this table:
[in this window]
[in a new window]
|
Table 5. Ranges in family means and narrow-sense heritability (h2) estimates and their confidence intervals (CI) for biomass production for two cultivars of switchgrass (Panicum virgatum L.) harvested for biomass once annually during early October, 2001 through 2004 at Aurora, SD.
|
|
Hopkins et al. (1993) reported an among-family narrow-sense heritability of 0.22 for forage yield for polycross families from an upland population in seeded rows (76-cm row spacing) in Nebraska. Estimates of among family narrow-sense heritability for biomass on an individual spaced-plant (1.07-m spacing) basis in a lowland population composed of 33 families ranged from 0.52 (Godshalk et al., 1986) to 0.59 (Talbert et al., 1983). The heritability estimates obtained from the spaced-plant lowland population described by Talbert et al. (1983) were similar to what we found in two upland populations with similar numbers of families evaluated. However, the intrarow spacing between plants in this study was only 35 cm.
Large ranges in family means and moderately high heritability estimates for biomass production for Summer and Sunburst switchgrass in transplanted single-row plots indicated progress from selection can be expected. However, the correlation between biomass production in transplanted rows and seeded sward of switchgrass is not known, and this will affect the amount of gain realized. The greatest progress would likely result from direct selection for biomass using the between- and within-family selection procedure described by Vogel and Pedersen (1993).
|
CONCLUSIONS
|
|---|
Prairie cordgrass has potential for biomass production in the northern Great Plains. After 4 yr of a single annual harvest in early October, populations of cordgrass had formed dense and highly productive sods. High levels of persistence, sod formation, and productivity on a well-drained upland site in eastern South Dakota suggest that cordgrass may have wider environmental amplitude than switchgrass for biomass production in the northern Great Plains. Unlike prairie cordgrass, switchgrass would not be expected to be adapted to poorly drained and/or saline areas (Weaver, 1954). However, cordgrass because of its susceptibility to lodging was not as well suited as switchgrass for stockpiling over winter.
As perennial grasses become more important for lignocellulosic biomass production, it would be ill advised to plant large tracts of monocultures. Prairie cordgrass may be able to play a role in lignocellulosic biomass feedstock production systems by adding diversity, productivity, and environmental quality to the landscape.
|
NOTES
|
|---|
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.
Received for publication May 17, 2006.
|
REFERENCES
|
|---|
- Analytical Software. 2000. Statistix 7 user's manual. Analytical Software, Tallahassee, FL.
- Ankom Technology. 2002. Method for determining acid detergent lignin in the Daisy II Incubator. Ankom Technology Corp., Fairport, NY.
- Ankom Technology. 2003a. Method for determining neutral detergent fiber. Ankom Technology Corp., Fairport, NY.
- Ankom Technology. 2003b. Method for determining acid detergent fiber. Ankom Technology Corp., Fairport, NY.
- Boe, A., R. Bortnem, and K.D. Kephart. 2000. Quantitative description of the phytomers of big bluestem. Crop Sci. 40:737–741.[Abstract/Free Full Text]
- Boe, A., and M.D. Casler. 2005. Hierarchical analysis of switchgrass morphology. Crop Sci. 45:2465–2472.[Abstract/Free Full Text]
- Casler, M.D., and A. Boe. 2003. Cultivar x environment interactions in switchgrass. Crop Sci. 43:2226–2233.[Abstract/Free Full Text]
- Das, M.K., R.G. Fuentes, and C.M. Taliaferro. 2004. Genetic variability and trait relationships in switchgrass. Crop Sci. 44:443–448.[Abstract/Free Full Text]
- Godshalk, E.B., J.C. Burns, and D.H. Timothy. 1986. Selection for in vitro dry matter disappearance in switchgrass regrowth. Crop Sci. 26:943–947.[Abstract/Free Full Text]
- Hitchcock, A.S. 1950. Manual of the grasses of the United States. 2nd ed. Rev. A. Chase. USDA Misc. Publ. 200. U.S. Gov. Print. Office, Washington, DC.
- Hopkins, A.A., K.P. Vogel, and K.J. Moore. 1993. Predicted and realized gains from selection for in vitro dry matter digestibility and forage yield in switchgrass. Crop Sci. 33:253–258.[Abstract/Free Full Text]
- Knapp, S.J., W.W. Stroup, and W.M. Ross. 1985. Exact confidence intervals for heritability on a progeny mean basis. Crop Sci. 25:192–194.[Abstract/Free Full Text]
- Lee, D.K., and A. Boe. 2005. Biomass production of switchgrass in central South Dakota. Crop Sci. 45:2583–2590.[Abstract/Free Full Text]
- Madakadze, I.C., B.E. Coulman, A.R. McElroy, K.A. Stewart, and D.L. Smith. 1998. Evaluation of selected warm-season grasses for biomass production in areas with a short growing season. Bioresour. Technol. 65:1–12.[CrossRef][ISI]
- Mobberley, D.G. 1956. Taxonomy and distribution of the genus Spartina. Iowa St. Coll. J. Sci. 30:471–574.
- Nyquist, W.E. 1991. Estimation of heritability and prediction of selection response in plant populations. Crit. Rev. Plant Sci. 10:235–322.
- Potter, L., M.J. Bingham, M.G. Baker, and S.P. Long. 1995. The potential of two perennial C4 grasses and a perennial C4 sedge as ligno-cellulosic fuel crops in N.W. Europe: Crop establishment and yields in E. England. Ann. Bot. (Lond.) 76:513–520.[Abstract/Free Full Text]
- Stubbendieck, J., S.L. Hatch, and K.J. Kjar. 1982. North American range plants. 2nd ed. Univ. of Nebraska Press, Lincoln.
- Talbert, L.E., D.H. Timothy, J.C. Burns, J.O. Rawlings, and R.H. Moll. 1983. Estimates of genetic parameters in switchgrass. Crop Sci. 23:725–728.[Abstract/Free Full Text]
- Undersander, D., D.D. Mertens, and N. Thiex. 1993. Total ash in forages. Forage analysis procedures. National Forage Testing Association, Omaha, NE.
- Vogel, K.P., and J.F. Pedersen. 1993. Breeding systems for cross-pollinated perennial grasses. Plant Breed. Rev. 11:251–274.
- Weaver, J.E. 1954. North American prairie. Johnsen Publishing, Lincoln, NE.
TOP
ABSTRACT
INTRODUCTION
