Switchgrass Production in Iowa: Economic Analysis, Soil Suitability, and Varietal Performance

 

 

 

E.C. Brummer, C.L. Burras, M.D. Duffy, and K.J. Moore

Iowa State University

 

 

under

Subcontract 90X-SY510V

 

 

 

Prepared August 2001

 

 

 

Research supported by

Office of Transportation Technologies

Activity No. EB 52 03 00 0

and

Office of Power Technologies

Activity No. EB 24 04 00 0

 

 

for

 

Bioenergy Feedstock Development Program

OAK RIDGE NATIONAL LABORATORY

Oak Ridge, Tennessee  37831-6422

 

 

managed by

 

UNIVERSITY OF TENNESSEE BATTELLE LLC

for the

U.S. DEPARTMENT OF ENERGY

under contract DE-AC05-00OR22725

TABLE OF CONTENTS

 

Conversion Factors.............................................................................................................................. 1

Executive Summary............................................................................................................................ 1

Project Personnel................................................................................................................................ 2

Principal Investigators................................................................................................................ 2

Technical Assistance................................................................................................................. 2

Introduction......................................................................................................................................... 2

Research Projects............................................................................................................................... 3

I.           Economics of Switchgrass Production............................................................................... 3

II.          Switchgrass Production in Relation to Soil Variability and Environmental Quality................... 8

Introduction..................................................................................................................... 8

II.1.      Fertility and Landscape Effects on Switchgrass Production and Quality....................... 8

Objective............................................................................................................... 8

Methods................................................................................................................ 8

Results and Discussion.......................................................................................... 9

II.2.      Hillslope Pedology and its Implications to Switchgrass Production in the Lake Rathbun Watershed, Iowa........................................................................................................... 13

Objectives............................................................................................................ 15

Materials and Methods.......................................................................................... 15

Intensive Project......................................................................................... 15

Extensive Project........................................................................................ 17

Results and Discussion......................................................................................... 18

Intensive Project......................................................................................... 18

Extensive Project........................................................................................ 25

Conclusion........................................................................................................... 32

III.         Biofuel Crop Germplasm Evaluation................................................................................. 33

III.1.     Switchgrass Germplasm Yield and Quality.............................................................. 33

Objective.............................................................................................................. 33

Methods.............................................................................................................. 33

Results and Discussion......................................................................................... 33

III.2.     Reed Canarygrass Breeding and Evaluation............................................................ 36

Biofuel Potential of Reed Canarygrass: A Literature Review...................................... 36

III.2.1.    Reed Canarygrass Variety And Harvest Management Evaluation................... 38

Objective............................................................................................... 38

Methods................................................................................................ 38

Results................................................................................................. 38

III.2.2.    Reed Canarygrass Germplasm Evaluation................................................... 42

Objective............................................................................................... 42

Methods................................................................................................ 42

Results and Discussion.......................................................................... 43

Acknowledgments............................................................................................................................. 48

Publications...................................................................................................................................... 48

Bibliography...................................................................................................................................... 48

Appendix I.          Detailed Establishment Year Cost Estimates For Seven Production Scenarios Described In Section I.1, And Expected Costs Of Reseeding Under Alternative Seeding Timings..................... 53

Appendix II.         Preliminary Budgets For Reed Canarygrass................................................................. 63

 

 

Conversion Factors

1 ton/acre (T/A) = 2.24 Mg/ha = 2400 kg/ha

1 Mg/ha = 1000 kg/ha = 0.45 tons/acre

1 g/m² = 10 kg/ha

1 g/kg = 0.1%

1 mg/kg = 1 ppm (part per million)

 

Executive Summary

Biofuel production in the Chariton Valley in southern Iowa would have desirable environmental effects by converting land usually planted to annual row crops into perennial grass cover.  Switchgrass, designated by DOE research as the most viable herbaceous biofuel crop, is native to Iowa and has been grown to a limited extent as a forage crop.  Its productivity as a biofuel needs to be assessed; the characteristics of a desirable biofuel crop differ from those of a forage, and agronomic practices will likely need to be altered.  Additionally, biofuel crops are targeted to the more erodible land in the region, land that varies considerably in soil characteristics, and hence, productive capacity.  Reed canarygrass could complement switchgrass, particularly in wet areas, and its ability to form a dense sod may improve erosion control in some instances.

 

Economic and agronomic analyses of biofuel crops–primarily switchgrass, secondarily reed canarygrass–are needed to determine the feasibility of growing these crops in southern Iowa.  In this report, we discuss preliminary research bearing on these issues.

 

The economic analysis of switchgrass production shows that yield and price are the determining factors for profitability.  With moderate yields (3 tons/acre) and price ($50 per ton), switchgrass could produce a significant positive impact for the regional economy.  Changing from a corn/soybean rotation to switchgrass will not make a substantial change in energy usage to produce the crop.

 

In field level trials, we have found switchgrass (cultivar ‘Cave-in-Rock’) yields to be relatively low when starting from long-term, poorly managed stands.  However, yields improved to nearly 4.3 Mg ha^-1 (about 2 tons/acre) after two years of fertilization with 112 kg N ha^-1 and weed control.  These yield levels are still low, but given that the stands in which the initial work was conducted were thin and poorly managed, we expect that yields can improve in well-managed stands.  The one caveat is that the inherent productivity of some highly erodible land is quite low, and high production in these areas, primarily sideslopes, may not be realistic.  Additionally, we found evidence of substantial erosion in some established switchgrass stands, a result that was unexpected.

 

Yields of various germplasm in small plot trials planted in 1997 ranged from 6.4 Mg ha^-1 in 1998 to 11.8 Mg ha^-1 in 1999 as the stands matured and filled in gaps.  The highest yielding variety in 1999 was ‘Alamo’, at 17 Mg ha^-1.  Alamo and several other lowland ecotypes produced the most biomass, higher than Cave-in-Rock, the normally recommended cultivar for southern Iowa.  These trials suggest that higher yields are possible under optimum management and with superior cultivars.  A cautionary note is that the lowland cultivars have not experienced a severe winter, and their winter hardiness may not be sufficient under those conditions.  In all cases, switchgrass quality appears adequate for a biofuel; variation among cultivars exists, suggesting that further improvements in quality are possible.

 

Preliminary evaluation of reed canarygrass suggests that two harvests, one in late spring and the other after frost, yield the most biomass.  Evaluation of a large collection of germplasm in Iowa and Wisconsin shows that higher yields are possible than those present in currently available cultivars.  Quality of reed canarygrass may be problematic:  ash, chlorine, and silica are higher than optimum.  Further analysis of quality is needed, especially because all data evaluated to date have been collected in central Iowa on soils quite different from those in southern Iowa.

 

All the field experiments discussed are continuing for at least another year.  More substantial discussion of the soil properties of fields and their relationship with biomass yield and quality will be completed over the next year.  In addition, new experiments to evaluate the best performing switchgrass cultivars in large strip trials, to test reed canarygrass side-by-side with switchgrass in large plots, and to determine field level yields and quality of reed canarygrass are underway.

 

Project Personnel

Principal Investigators

E. Charles Brummer       Project Coordinator; Biomass Crop Breeding

                                    brummer@iastate.edu       515-294-1415

 

C. Lee Burras                Soil Quality and Management

                                    lburras@iastate.edu          515-294-0559

 

Michael D. Duffy            Agricultural Economics

                                    mduffy@iastate.edu          515-294-6160

 

Kenneth J. Moore           Biomass Crop Production and Utilization

                                    kjmoore@iastate.edu        515-294-5482

 

Technical Assistance

Michael Barker              Biomass Crop Management, Evaluation, and Breeding, and Soil Characterization

 

Virginie Nanhou             Economic Analysis of Biofuel Production

 

Patricia Patrick              Biomass Quality Laboratory Analysis

 

Mark Smith                   Biomass Crop Small Plot Harvesting

 

John Sellers                  Large Field Plot Assistance

 

Introduction

Marginal soils, widespread throughout southern Iowa, are unsuited to annual row crop—corn and soybean—production.  Much of the landscape in southern Iowa is characterized by heavy, wet soils and significant slopes that allow substantial levels of erosion.  On-farm integration of biofuel crops with grain and forage crops and livestock may foster the long-term environmental and economic sustainability required for agricultural systems.

 

Switchgrass has been chosen as the model herbaceous biofuel crop, and its adaptation to Iowa is well known.  Profitable use of biomass crops requires sufficient understanding of agronomic aspects of their culture and economic realities of their production.  We intend to assess the productive potential of switchgrass across a range of soil types and landscapes, allowing us to more effectively pinpoint locations where it will perform well.

 

Reed canarygrass represents another potential biofuel crop, a cool-season grass alternative to switchgrass.  With its different growth pattern–it is most productive in spring and fall–and tolerance to both wet and droughty soils, reed canarygrass complements switchgrass in a diversified biofuel program.  Its strongly rhizomatous growth habit also make it appealing, particularly on soils on which switchgrass, a bunchgrass, does not form thick stands and erosion is a problem.

 

The research reported in this report is part of an ongoing project to understand the constraints to biomass production in southern Iowa and to develop production methods that will permit economically viable production of biofuel crops.  Although labeled a “final” report, most of the experiments discussed are continuing in the field for one to two more years.  Thus, only tentative conclusions are possible at this point.  Similarly, the economic analyses are necessarily preliminary and could change as production parameters developed in other phases of this program are implemented on-farm.

 

In the report, tables for each section follow immediately after the text for that section.  Figures are attached at the end of the document, after the appendices.

 

Research Projects

The research projects that will be discussed in this report are based on three objectives:

 

I.    Economic potential of switchgrass as an agronomic crop for bioenergy

1.   Document on-farm costs and resource commitments for switchgrass production

2.   Assess regional economic impacts of large-scale switchgrass production

3.   Quantification of energy consumption for switchgrass production

 

II.    Switchgrass production in relation to soil variability and environmental quality

1    Landscape and nitrogen effects on switchgrass production potential.

2.   Quantification of soil properties and their relation to switchgrass yield and quality, and assessment of the erosion potential in switchgrass fields

 

III.   Evaluate and develop switchgrass and reed canarygrass germplasm for bioenergy production and adaptation to Iowa

1.   Switchgrass cultivar evaluation for yield and biofuel quality

2.1. Evaluation of harvest management and varietal performance of reed canarygrass for biofuel

2.2. Evaluate diverse reed canarygrass germplasm and begin breeding new cultivars for bioenergy uses

 

 

I.      Economics of Switchgrass Production

The preparation of budgets for the costs of producing switchgrass has been completed.  This work has been prepared as an Iowa State University Extension Publication.  The publication is at the printers. 

 

The publication has the following outline:

 

What is switchgrass?

Description of the scenarios

General assumptions

      Assumptions on input costs

            Mchinery

            Seed

            Herbicides

            Fertilizers and lime

      Harvesting data

Summary of costs

Summary

 

The publication is entitled; Costs of Producing Switchgrass for Biomass in Southern Iowa, Iowa State University Extension Publication PM 1866.   There were 500 hard copies of the publication order.  In addition, the publication will be available electronically on the extension home page.

 

In addition to the extension publication, this work will be presented at the Fifth Annual Biomass Conference of the Americas.

 

Since the completion of the budgets reported in the extension publication we have learned more about the production of switchgrass.  To continue our work with switchgrass production costs we incorporated some of the changes into new budget estimations.  The primary changes that we examined were the impacts of increasing the seeding rates and changing the probability of needing to reseed. 

 

The extension budget estimations were based on using 6 pounds of pure live seed for the seeding rate.  In this new series of estimations we increased the seeding rate to 10 pounds pure live seed per acre.  The heavier seeding rate was more reflective of current production practices and it is consistent with what has been learned in the field.

 

The extension budget also assumed a 50% reseeding rate for spring seeded switchgrass and a 25% reseeding rate under a frost seeding system.  The heavier seeding rates and experience have shown the probability of reseeding varies.  Therefore, we also re-estimated the budgets using a 25, 15, 10 and 0% probability of reseeding. 

 

The new estimations were only for a frost-seeding regime.  The previous work showed that in all cases the frost seeding costs of production were lower than the spring seeding.  In addition, frost seeding regime was also selected because it has become the establishment technique of choice by producers in southern Iowa.  Therefore, we chose to concentrate further analysis on only the frost-seeding system.

 

Changing the seeding rate from 6 to 10 pounds made very little difference in the final costs per ton.  The estimated costs increased by 1% or less, depending on the yield.  Summary Tables 1 and 2 show the costs per ton for frost-seeding at 10 pounds per acre with alternative yield levels, alternative probabilities for reseeding, and alternative land charges.  Table 1 costs at $75 per acre and a 25% reseeding probability can be compared to Appendix 3 in the extension publication to obtain a comparison of the cost differences for 6 and 10 pound seeding rates.

 

Summary Table 1 shows that changing the probability of having to reseed causes little change in the costs of production.  At the lowest yield, 1.5 tons per acre on cropland, the cost per ton drops from $133.63 with a 25% probability of reseeding, to $130.34 per ton with no reseeding.  This is a change of only 2%.  The impact lessens the higher the yield.

 

Appendix I contains all the tables used to create Summary Tables 1and 2.  The appendix tables are for the establishment costs, the reseeding costs, and the various yield and reseeding probability scenarios. 

 

The analysis based on heavier seeding rates and alternative assumptions regarding the probability of reseeding do not change the basic conclusions from the initial work.  Yield per acre has the greatest impact on the costs per ton.  The second greatest impact is attributed to the land charge per acre.  With the highest yield, 6 tons per acre, the costs per ton vary from the low $50 range with a $75 per acre land charge to less than $45 per ton with a $25 per acre land charge. 

 

Examining alternative production techniques, reseeding rates, and other production aspects will not appreciably impact switchgrass costs of production.  The most important research must be on ways to increase yields.  This work has shown that the switchgrass at a 6 ton yield level can be cost competitive for biomass production.

 

We have completed work on estimating the costs of production for reed canarygrass.  These initial budgets will change as we learn more about production techniques and how to manage reed canarygrass.

 

The most significant reed canary production practices are the following:

 

·       Land preparation is usually done through no till drill following crops and killed sod.

·       The seed variety commonly used is Palaton, and seeding rate is 10 to 12 pounds pure live seed per acre.

·       Spring or late summer seeding, but late summer (August) seeding preferred.

·       No nitrogen application in the establishment year and two nitrogen applications during production years.

·       Two harvests per year, in large bales, weighing 1,100 pounds on average.

 

Summary Table 3 presents the estimated costs for establishing reed canarygrass following cropland and grassland.  We assumed a $50 per acre charge for grassland and a $75 per acre land charge for cropland.  We assumed that the stand would last for 11 years.  Further, we assumed there is no reseeding necessary.  Notice that there is no appreciable difference in the establishment cost estimates.  This is due to the assumptions used, especially regarding the herbicide choices.  These costs would change depending upon the production system chosen by the producer. The costs per ton range from a high of $79.62 per ton for the 3 ton yield on cropland ($75 per acre land charge) to a low of $45.17 per ton for the 6 ton yield on grassland ($50 per acre land charge).

 

Appendix II contains the tables used to create Summary Table 3.  The appendix tables are for the establishment costs and the estimated production costs for 3, 4, and 6 ton yield assumptions.

 

The costs of producing reed canarygrass follow a similar pattern to switchgrass in that yield is the most important variable in determining the costs per ton.  Land charges are the second most important variable.  However, as yield increases the effect of the land charge decreases.

        

Summary Table 1.  Summary of frost seeding on cropland, four levels of reseeding probability and two levels of land charge (seeding rate 10lbs/acre).
Scenario	Type of costs	Yield (ton/acre)	25% reseeding probability			15% reseeding probability			10% reseeding probability			0% reseeding probability
			$25	$50		$25	$50		$25	$50		$25	$50
Frost seeding on cropland	Yearly production cost	1.5	143.80	168.80		143.80	168.80		143.80	168.80		143.80	168.80
		3.0	183.90	208.90		183.90	208.90		183.90	208.90		183.90	208.90
		4.0	210.64	235.64		210.64	235.64		210.64	235.64		210.64	235.64
		6.0	264.11	289.11		264.11	289.11		264.11	289.11		264.11	289.11
	Total cost per acre	1.5	171.01	200.44		169.41	198.47		168.61	197.48		167.01	195.51
		3.0	211.11	240.55		209.51	238.57		208.71	237.59		207.11	235.62
		4.0	237.85	267.28		236.25	265.31		235.45	264.32		233.85	262.35
		6.0	291.32	320.76		289.72	318.78		288.92	317.80		287.32	315.83
	Total cost per ton	1.5	114.01	133.63		112.94	132.31		112.41	131.66		111.34	130.34
		3.0	70.37	80.18		69.84	79.52		69.57	79.20		69.04	78.54
		4.0	59.46	66.82		59.06	66.33		58.86	66.08		58.46	65.59
		6.0	48.55	53.46		48.29	53.13		48.15	52.97		47.89	52.64

 

 

 

Summary Table 2.  Summary of frost seeding on grassland, four levels of reseeding probability and two levels of land charge (seeding rate 10lbs/acre).
Scenario	Type of costs	Yield (ton/acre)	25% reseeding probability			15% reseeding probability			10% reseeding probability			0% reseeding probability
			$25	$50		$25	$50		$25	$50		$25	$50
Frost seeding on grassland	Yearly production cost	1.5	118.80	143.80		118.80	143.80		118.80	143.80		118.80	143.80
		3.0	158.90	183.90		158.90	183.90		158.90	183.90		158.90	183.90
		4.0	185.64	210.64		185.64	210.64		185.64	210.64		185.64	210.64
		6.0	239.11	264.11		239.11	264.11		239.11	264.11		239.11	264.11
	Total cost per acre	1.5	144.10	173.53		142.87	171.93		142.26	171.13		141.03	169.53
		3.0	184.20	213.63		182.98	212.04		182.36	211.24		181.14	209.64
		4.0	210.94	240.37		209.71	238.77		209.10	237.97		207.87	236.37
		6.0	264.41	293.85		263.19	292.25		262.57	291.45		261.35	289.85
	Total cost per ton	1.5	96.07	115.69		95.25	114.62		94.84	114.09		94.02	113.02
		3.0	61.40	71.21		60.99	70.68		60.79	70.41		60.38	69.88
		4.0	52.73	60.09		52.43	59.69		52.27	59.49		51.97	59.09
		6.0	44.07	48.97		43.86	48.71		43.76	48.57		43.56	48.31

 

 

Summary Table 3.  Summary for reed canarygrass production for two types of land (cropland, grassland) and three yield levels (3, 4 and 6 tons/acre).
Scenarios	Yield (ton /acre)	Prorated establishment cost ($)	Production cost per acre ($)	Production cost per ton ($)
Seeding on cropland	3.0	26.43	238.86	79.62
	4.0	26.43	258.28	64.57
	6.0	26.43	297.12	49.52
Seeding on grassland (1)(Burn down of grass and No till grass seed drill)	3.0	26.20	213.63	71.21
	4.0	26.20	233.05	58.26
	6.0	26.20	271.89	45.31
Seeding on grassland (2)(Plow and disk and grass seed drill)	3.0	25.33	212.76	70.92
	4.0	25.33	232.18	58.04
	6.0	25.33	271.02	45.17

 

II.     Switchgrass Production in Relation to Soil Variability and Environmental Quality

Introduction

The Chariton Valley in southern Iowa is well suited for agronomic crop production in many respects.  The average frost-free season and precipitation are nearly 170 days and 80 cm inches, respectively.  A well-developed farm culture is in place.  It consists of about 2500 farms, numerous agribusinesses and knowledgeable support organizations.  However, production is limited in parts of the region by soils that restrict the types of crops that can be profitably grown.  This limitation arises from the prevalence of soil consociations throughout the central Southern Iowa Drift Plain (Figures 1 and 2; see separate document “ISU 2000 Final Report Figures”) that are highly erosive, shallow to root restrictive zones and/or excessively wet.  Furthermore, dramatic differences among soils are common within a given field.  Consequently, development of a sustainable, profitable agronomic production scheme has been very difficult, especially over the last 40 years as the farmers have expanded machinery and field size.

 

The introduction of switchgrass (Panicum virgatum, L.) in CRP and as a biofuel has been widely supported because it was thought to thrive in an environmentally benign way across the soil-landscapes of the Chariton Valley while at the same time not competing with traditional farm crops.  The goal of this study was to document the reality of current switchgrass production practices vis-à-vis switchgrass yields and environmental benefits (or costs).  The specific objectives follow.

 

The areas within the Chariton Valley chosen for intensive plant and soil sampling are shown in Figures 3-5.  The predominant soil series within these fields is described in Table II.1.

 

II.1.   Fertility and Landscape Effects on Switchgrass Production and Quality

Objective

The objective of this experiment is to determine the effects of locations, years, harvest dates, landscape positions, and nitrogen levels on switchgrass yield and biomass quality traits. 

 

Methods

We began field experiments in 1998 using mature, established ‘Cave-In-Rock’ switchgrass fields at two southern Iowa locations:  near Derby in Lucas County and near Millerton in Wayne County.  The experimental design was a randomized complete block design with six replications at Derby and five replications at Millerton.  The replications are split across two fields in each location, which are owned and managed by the same farmer and which are adjacent to each other.  We have not observed a field effect within location; the two fields were merged.  One replication in Derby was dropped from data analysis because it behaved aberrantly, likely due to limestone dust from the adjacent road.  Thus, five replications at each location were used for analyses.  Each replication was 200’ wide and between 100’ and 400’ long, the variable length being necessary to allow incorporation of summit, backslope, and swale landscape positions within each plot.  This size plot was amenable to management by standard farm equipment.  Each replication included four randomly assigned plots, representing four nitrogen fertility treatments of 0, 56, 112, and 224 kg N ha^-1; each plot was 50’ wide and covered all three landscape positions.  In 1998 and 1999, plots were subsampled throughout the year for biomass yield and quality measurements using a 1 m2 quadrat.  In autumn 1998, 1999, and 2000, total plot biomass was harvested by mowing and baling the entire plot area.  Within each plot, soil samples of the ‘A’ horizons were taken at five points across the landscape.  Additionally, 30 1-m deep cores were taken across all plots.

 

These fields had a history of limited management prior to our use (they were enrolled in the Conservation Reserve Program [CRP] which only mandates a good ground cover be present) and had been in continuous switchgrass for at least five years.  The landscapes and soils are typical of the area with parent materials including Peorian loess, Yarmouth-Sangamon paleosol, Pre-Illinoisan till, or alluvium.  The total slope range across the research plots was 0 to 14%.  The soil types in the fields under investigation are shown in Table II.1.

 

Results and Discussion

Yield and plant height.  Biomass yield showed continued improvement in 2000 over the previous years (Table II.2).  The yield improvement demonstrated in these fields resulted from three years of nitrogen application and good management practices.  These fields were previously enrolled in the CRP and had received very limited management.  Thus, conversion of CRP switchgrass fields to biomass production will result in improved productivity, but several years may be needed to achieve maximum sustained production.  The yields seen in 2000 (averaging 6 Mg ha^-1, or nearly 3 T A^-1) make the economics of biomass production much more appealing than previous yield estimates had suggested.  Further gains in productivity may be possible.  The 2000 growing season was not ideal, with very low soil moisture during spring and autumn.  To an extent, the deep roots of switchgrass probably allowed the plants to avoid serious moisture stress, but a more consistent rainfall pattern during the growing season may have improved nitrogen use and growth.  The observed yields, while improving, are still relatively low, likely due to a combination of weather, site limitations (e.g., the fields consist of soils with severe B horizon limitations), and fertility and/or stand problems, and inappropriate switchgrass cultivars for southern Iowa.

 

The two locations (Lucas and Wayne) produced similar yields in 2000 (data not shown), although across all three years, Lucas slightly outyielded Wayne (Table II.2).  The important point is that two contrasting locations in the Chariton Valley, both of which started with less than optimal switchgrass stands, could be improved over the course of three years to produce similar, and acceptable, yields of biomass.  Given that some areas within the plots still have thin stands, further yield gains appear possible.  We will continue to monitor yield in these plots in 2001.

 

Nitrogen fertilization increased biomass both when averaged across the three years (Table II.2).  In 2000, the most striking response came with the addition of 56 kg ha^-1, with no difference between 56 and 112 kg ha^-1, or between 112 and 224 kg ha^-1.  The 224 kg ha^-1 level was higher than 56, however.  Across the three years, improvements in yield were realized by sequential increases of N from 0 to 56 kg ha^-1 and from 56 to 112 kg ha^-1.  Increasing nitrogen application above 112 kg ha^-1 did not result in further yield increases averaged across the three years or in 2000.  Thus, the recommended fertilization rate for switchgrass biomass production in this region of southern Iowa should be between 56 and 112 kg ha^-1.

 

Among landscape positions, summits had higher yields (based on subsampling) than the back and footslopes, not surprising given the better soil depth and quality at this location.  The end-of-year plot harvests were made across landscape positions and thus we don’t have this information on specific landscape points.  Except for subsample yields, differences among landscape positions were few, possibly because the size of the plots was not large enough (even though they were quite big) to represent striking differences in topography (see Tables II.5a,b in the 2000 Annual Report for more detail).

 

Plant height appears to be related to yield from 1998 to 2000 (Table II.2).  However, this relationship may not be completely accurate, as the measurements in 1998 and 1999 were made in August, about two months prior to harvest, but the 2000 data were collected at harvest time.  Heights did not differ in a meaningful manner between locations or among nitrogen treatments in 2000 (data not shown).

 

Cell wall components, nitrogen content, and ash.  Cell wall constituents differed among years (Table II.2), but the importance of these differences is not clear.  Harvest in 1999 occurred at the end of September, a month or more before the other years, and that could have caused lower cell wall content values because soluble material had not been leached as severely.  The most significant differences are that lignin (ADL) was lower and cellulose was higher in 2000 than in the other years.  This may be related to the yield improvement seen in 2000.  Otherwise, the differences among years followed no clear trend.  Ash values, determined as a byproduct of the cell wall digestion process, were about 5%. 

 

The two locations, Lucas and Wayne counties, were generally quite comparable for these traits, both averaged across years (Table II.2) and in 2000 (Table A.II.1).  Nitrogen in the plants, as determined using the Kjeldahl method, and ADL were slightly higher in Wayne, but this difference does not appear to be biologically important.  Among nitrogen fertilization levels, higher N rates generally led to higher concentrations of cell wall components (except hemicellulose).  No discernable trend was evident among N levels for nitrogen concentration or ash content.  The main conclusion from these data is that the cell wall content of switchgrass biomass does not appear to be altered greatly due to year, location, or fertility status, and those changes that are observed are not easily explained.  Certainly, increases in yield do not appear to have major effects on cell wall constituents.

 

Proximate, ultimate, and elemental analyses.  Proximate and ultimate analyses showed that differences occurred among years for all traits except sulfur (Table II.3), based on biomass samples collected at harvest time.  Like the cell wall results, the differences among years do not show any clear trend.  Ash was highest in 1999, nitrogen levels were highest in 2000, and BTU content was lowest in 2000; whether these results were related to environmental variation or to the higher yields obtained in 2000 is unknown.  Regardless, the differences are all relatively small, and probably would have little (if any) impact on using switchgrass as a biofuel.  Differences for these traits among N fertilization rates were similarly small.

 

Elemental analyses showed that the concentration of a number of elements differed between 1999 and 2000, but the differences are probably immaterial regarding biofuel quality (Table II.4).  Neither location nor N fertilization rate had a substantial impact on composition. However, chlorine varied by location, with Wayne having roughly the levels of Lucas, but both of these levels are within acceptable ranges for power plants.  The values obtained from proximate, ultimate, and elemental analyses are broadly congruent with those found previously for switchgrass by Miles (1996).

 

Note that the values of particular elements in Table II.4 vary between analyses because samples for the different analyses were prepared differently, being conducted on ashed samples, dry vegetation, or acid digested vegetation and because the different analysis types may result in loss or underestimation of particular elements.  However, in general, the values are comparable.

 

Large differences for most traits were observed among sampling dates (see Tables II.6a,b in the 2000 report for details).  Based on subsample yields (plot yields were not taken at multiple times), maximum dry matter yield appears to have accumulated by September (data not shown); thus, delaying harvest until frost serves only to lower the water content of the herbage.  Earlier harvests, if the material was acceptably dry, would expedite work in autumn when weather is unpredictable.  The leaf fraction of the harvested material declined through November.  This probably helps explain why nitrogen in the plant tissue declined throughout the year, reaching its low point by November, with little additional loss over winter.  Similarly, cellulose, lignin, ash, and digestibility fell as the plants matured.  Perhaps most interestingly, Cl, N, P, and S ions were substantially lower in March than November, which may be important for feedstock quality.

 

In general, overwintering material in the field results in slightly better biofuel, from an energy standpoint per unit dry weight, but the decline in yield during that time appears to more than offset the improved energy quality (see data in 2000 annual report).

 

Elemental analyses are presented in Table II.8 by location and by nitrogen level.  Only the September 1999 samples were analyzed due to limited samples from the 1998 growing season.  In general, neither location nor nitrogen treatment affected elemental composition of biomass, with the exception of Cl, P, and Ba.  Also, elemental values determined by ion chromatography corresponded very well with those determined by INAA and/or inductively coupled plasma emission spectometry (ICP).  Note that the values in Table II.8 vary between analyses because they were conducted on ashed samples, dry vegetation, or acid digested vegetation and because the different analysis types may result in loss or underestimation of particular elements.  However, in general, the values are comparable.

 

Table II.1.        Summary of soils information available from the Lucas and Wayne County soil surveys (Prill, 1960, and Lockridge, 1971, respectively).
		Field number* and estimated MU area (%)
Map unit	Series and great group classification	1	2	3	7
ClC2, CmC3	Clarinda, Vertic Argiaquoll			70	20
Gd	Grundy, Aquertic Argiudoll	100	60
Ha	Haig, Vertic Argiaquoll		10
Oa	Omitz-Gravity-Wabash, Cumulic Mollisolls		10
Sa	Shelby–Adair, Typic & Aquertic Argiudolls		20
SeB, SfC2	Seymour, Aquertic Argiudoll			15	80
ShD2	Shelby, Typic Argiudoll			15
*Field numbers 1 and 2 are in Lucas County, and 3 and 4 in Wayne County.

 

 

Table II.2.        Switchgrass yield, plant height, fiber content, nitrogen and ash for 1998, 1999, and 2000 in two southern Iowa locations and at four nitrogen fertilization rates.
	Yield	Height	NDF	ADF	ADL	Hemicellulose	Cellulose	N	Ash
	Mg/ha	cm	-----------------------------------------------g/kg-----------------------------------------------------
Year
1998	2.88	118	776.0	454.9	75.9	321.1	379.0	3.47	43.4
1999	3.90	145	710.7	414.1	70.7	296.6	343.4	5.48	56.1
2000	6.04	190	778.2	458.5	63.0	319.6	395.5	5.86	49.8
LSD (5%)	0.28	3	9.3	11.7	3.6	8.9	8.7	0.38	2.8
Location
Lucas	4.43	151	745.5	432.1	66.5	313.4	365.7	4.57	51.6
Wayne	4.12	151	764.4	452.9	73.3	311.5	379.6	5.30	47.9
LSD (5%)	0.23	ns	ns	9.5	2.9	ns	7.1	0.31	2.3
N Level
0	3.62	145	751.4	432.1	66.6	319.3	365.5	5.01	52.9
50	4.15	149	757.9	444.0	69.3	313.9	374.7	4.59	48.8
100	4.60	155	749.1	434.7	69.1	314.4	365.6	4.90	50.1
200	4.73	155	761.5	459.3	74.5	302.2	384.8	5.24	47.2
LSD (5%)	0.32	4	10.8	13.5	4.1	10.2	10.1	0.44	3.2
Grand mean	4.27	150.98	754.98	442.52	69.89	312.46	372.63	4.93	49.75
Harvest/sampling dates:  November 1998, September 1999, and October 2000.

 

Pedons were described using the procedures and nomenclature established by the Soil Survey Staff (1993).  A sample from each horizon described within each core was removed from the core, dried, ground, sieved, and stored in the same manner as the surface and hand core samples. 

 

Laboratory analysis.  Soil samples from horizons of a subset of the pedons were sent to the Iowa State University Soil Testing Laboratory for chemical analysis.  Analyses included pH, plant available phosphorus, plant available potassium, plant available zinc, percent organic matter, and total nitrate-nitrogen.  Additionally, the total carbon and nitrogen contents of pedon samples were determined by dry combustion using a LECO CHN-600 analyzer.

 

Bulk density, stable aggregate content, and particle size distribution were determined for selected soil samples using standard methods described in Soil Survey Staff (1996).

 

Yield.  The four study fields were harvested to determine total switchgrass yield in fall 1998, 1999, 2000 and 2001 although only the 1998 data is used herein.  Readers interested in greater year-by-year analysis of switchgrass yields are directed to Lemus (2000).  Harvesting methods were consistent with standard farm practices of the Lake Rathbun Watershed.

 

Statistical analysis.  Data were analyzed using the Statistical Analysis System (SAS) and the Microsoft Excel statistics package.  More sophisticated analyses were completed using SAS, more routine analyses using Excel. 

 

 

 

Extensive Project

Switchgrass yield was measured along 45 strips and 12 additional fields from throughout the Lake Rathbun Watershed following the 1999 growing season.  Strips were each about 1 ha in area and located in a larger field. The eight fields ranged from about 5 to 25 ha in area, which is typical for the Lake Rathbun Watershed.  Each strip and field was managed identically.  This included applying 160 kg ha^-1 N fertilizer prior to the growing season and use of recommended rates of atrazine and 2,4-D for weed control. 

 

Average yields for the strips and the fields were obtained by summing the weight of individual bales and then dividing this number by the total field area. 

 

Field and strip boundaries were determined using GPS having approximately 1-m accuracy.  These boundaries were then incorporated into GIS.  The GIS was then used in conjunction with the Iowa soil survey database in order to determine the area and selected attributes of each map unit.  Switchgrass yields-soil properties relationships were then examined using regression and stepwise analysis of variance. 

 

Results and Discussion

Intensive Project

Objective A.  Quantify selected pedon properties associated with shoulders, backslopes, and footslopes of 10-year old switchgrass fields from typical hillslope reaches in the Lake Rathbun Watershed.

 

The properties of pedons collected from summits, backslopes, and footslopes in fields of long-term switchgrass are surprisingly alike (Table II.6).  Few pedologically significantly differences are apparent although several statistically significant ones exist (Table II.7).  Summit pedons tend to be somewhat poorly drained while backslopes and footslope pedons are generally more poorly drained (Tables II.6 and II.7).  Epipedons and A-horizons average about 25 to 35 cm thick with the summit epipedons generally being the thickest.  The organic carbon content at each landscape position is around 2% with the footslope pedons having less carbon content than those on backslopes and summits.  The average common rooting depth is 50 to 70 cm with deeper rooting being more common in summit pedons.  Granular structure extends to the greatest depth (45 cm) in summit pedons.  Coarse fragment content becomes 3% on average at 73 cm in footslope pedons, which is more shallow by 20 cm than in backslope and summit pedons.  Mean stable aggregate content of the surface horizon ranges from 55 to 67%, with the lower mean being found in pedons from summits.  Pedons from all three landscape positions are consistently silty clay loam, silty clay, or clay textured throughout their sola (data not presented, see Molstad, 2000).  Clay content of the surface horizon and the B-horizon are around 27 to 29 and 44 to 46%, respectively.  The surface horizon C:N ratio is 10.  Solum pH ranges from around 5 to between 6.5 and 7.0 (Table II.6).