ww.sciencedirect.com

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4

Available online at w

ScienceDirect

http: / /www.elsevier .com/locate/biombioe

Yield of perennial herbaceous and woody biomasscrops over time across three locations

Gregg A. Johnson*, Donald L. Wyse, Craig C. Sheaffer

University of Minnesota, Department Agronomy and Plant Genetics, 1991 Upper Buford Circle, 411 Borlaug Hall,

United States

a r t i c l e i n f o

Article history:

Received 16 November 2012

Received in revised form

30 September 2013

Accepted 7 October 2013

Available online 28 October 2013

Keywords:

Biomass feedstock

Crop production

Biomass production

Bioenergy crops

Renewable energy

* Corresponding author. Tel.: þ1 507 837 561E-mail address: johns510@umn.edu (G.A.

0961-9534/$ e see front matter ª 2013 Elsevhttp://dx.doi.org/10.1016/j.biombioe.2013.10.

a b s t r a c t

The use of perennial biomass crops is expected to increase and will likely be part of a

diversified approach to cropping system design that focuses on multiple economic,

ecological, and environmental benefits. Field experiments were conducted from 2006 to

2011 at three locations in Minnesota to quantify biomass production across a diverse set of

perennial herbaceous and woody crops. Herbaceous crops were harvested annually in the

fall while the woody crops were harvested once following five years of growth. Willow

produced more total biomass than all other woody and herbaceous biomass crops across

all locations. However, miscanthus biomass yield was similar to ‘SX67’ willow at St. Paul

and Waseca, but was dependent on the cultivar of miscanthus. Prairie cordgrass cultivars

were among the highest and most consistent yielding herbaceous biomass crops across

locations. Miscanthus cultivars produced the highest annual dry matter yield of

35 Mg ha�1 yr�1 biomass, but only during the final year of the study. Other herbaceous

crops such as switchgrass performed well in certain locations and may offer flexibility in

cropping choice. This unique information on comparative biomass yield across a diversity

of perennial crops will inform the overall decision-making process in a way that reduces

risk and optimizes productivity in specific environments. This study shows that several

biomass crop species can be successfully grown as part of a diversified biomass cropping

enterprise.

ª 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Cellulosic biomass is being considered as a feedstock for

numerous bioindustrial applications [1]. Ethanol, heat, and

electrical power generation are the most common platforms

that use cellulosic biomass as a feedstock. Cellulosic crops

have been proposed as major feedstock to achieve the

79.5 hm3 of advanced biofuels goals by 2022 as designated by

the Renewable Fuels Standards. New techniques and pro-

cesses are being developed that expand the range of products

7.Johnson).ier Ltd. All rights reserved013

being produced from biomass feedstock. For example, there is

great interest in using plant biomass for the industrial pro-

duction of organic compounds as well as the synthesis of new

products that do not have a synthetic counterpart [1e3]. The

bio-refinery concept is proposed as a method of generating

even greater value by integrating the production of first and

second generation transportations fuels, power, and an array

of chemicals at one facility [4,5]. In this model, renewable

feedstock enters the biorefinery and is converted through an

array of processes into a mixture of products [6]. However,

.

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4268

feedstock type and quantity are key drivers in the success of

these concepts.

Most of the cellulosic feedstock for current bioindustrial

applications is derived from annual crops (e.g. corn (Zea mays

L.) stover) or perennial grasses (e.g. switchgrass (Panicum vir-

gatum L.)). It has been argued that the use of perennial feed-

stock is the preferred choice for bioindustrial feedstock

because of its economic, environmental and ecological bene-

fits [7,8]. Boehmel [9] evaluated the performance of six crop-

ping systems and found that the perennial cellulosic crops,

willow and miscanthus, had the best combined high biomass

and energy yields with high land and energy use efficiency,

nitrogen fertilizer use and environmentally benign production

methods compared to other annual crops. The benefits of

perennial biomass crops provided by increased ecosystem

services, such as enhanced wildlife habitat, nutrient seques-

tration, improved water quality and retention, and reduced

soil erosion, are well-documented [7,10e12].

The development of a robust bioproducts industry is a

recognized way to increase farmer profitability while

addressing environmental, ecological, and social issues. As

such, the use of perennial biomass crops is expected to in-

crease. However, there is a great need to understand how

biomass crops can be effectivelymanaged in a sustained long-

term approach [8]. Gonzalez-Hernandez et al. [13] suggest the

need to evaluate multiple species with adaption to specific

climate and edaphic zones because there is no one crop that

will provide all the necessary attributes necessary to meet the

demands of industry. Our objective was to quantify biomass

yield of perennial herbaceous and woody crops over time and

environment in Minnesota.

2. Materials and methods

Field studies were conducted at three University of Minnesota

Research and Outreach Centers from 2006 to 2011. The loca-

tions represent a range of soils and climates typical of

southern Minnesota. The three locations included St. Paul

(44�590 N, 93�10 W),Waseca (44�430 N, 93�060 W), and Lamberton

Table 1 e Soil, precipitation, and temperature characteristics dWaseca, and St. Paul locations.

St. Paul

Normal precipitation (cm)a 52.6 51

Precipitation (cm)

2007 47.8 64

2008 30.1 43

2009 28.7 27

2010 66.3 87

2011 64.4 46

Mean annual temp (�C) 7.8 6.6

Soil classification Waukegan silt loam

(fine-silty, mixed

mesic typic hapludoll)

Ni

mi

aq

Soil organic matter (%b) 4.5 5.1

Soil pH 6.3 5.8

a 30-year normal from 1971 to 2000.b Mass fraction.

(44�150 N, 95�190 W). Locations follow amoisture gradient from

western to eastern Minnesota with the eastern location (St.

Paul) having the greatest amount of precipitation and the

western location (Lamberton) with the least amount of pre-

cipitation during the growing season. There is also a soil

texture and soil drainage pattern across Minnesota whereby

the southern location (Waseca) is characterized by finer

textured, poorly drained soils compared to courser textured,

well drained soils of eastern and western Minnesota. Soil and

environmental information for each location is presented in

(Table 1). All field locations tested high for soil P and K. Ni-

trogen was broadcast applied as urea in each plot, except for

the polyculture treatment, at a rate of 112 kg ha�1 N in April of

each year and location. The previous crop was soybean across

all locations.

The experimental design was a randomized complete

block with three replications at each location. Block was

restrictedwhereby tall woody plantswere grouped so as not to

alter growth of shorter herbaceous species in nearby plots.

Plot size was 4.25 m � 6.25 m for the herbaceous crops and

5.00m� 9.25m for the woody plants. Treatments consisted of

perennial woody and herbaceous plants that are being

considered for biomass production. Plantmaterial used in this

study was purchased from commercial nurseries with the

exception of false indigo, sunflower, Jerusalem artichoke, stiff

goldenrod, and late goldenrod which were obtained locally

from University of Minnesota collections in St. Paul, MN. All

herbaceous crops were established from transplants with the

exception of the polyculture treatment in May 2006. Woody

crops were established from rooted or un-rooted cuttings.

Details regarding cultivars, sources, plant spacing, and plant

density are provided in Table 2.

Weed control consisted of a preemergence application

of mesatrione ([2-[4-(methylsulfonyl)-2-nitrobenzoyl]-1,3-

cyclohexanedione]) at 20.6 g ha�1 active ingredient in spring

2007 and 2008 and acetochlor (2-chloro-N-(ethoxymethyl)-N-

(2-ethyl-6-methylphenyl) acetamide) at 2.2 kg ha�1 active

ingredient in spring 2009e2011. Selective weed control

consisted of a postemergence application of quizalofop ((�)-2-

[4-[(6-chloro-2-quinoxalinyl)oxy]phenoxy]propanoic acid) at

uring the growing season (MayeSeptember) at Lamberton,

Waseca Lamberton

.9 42.3

.2 36.4

.2 32.7

.9 32.4

.9 47.5

.8 44.8

6.8

collet clay loam (fine-loamy,

xed, superactive, mesic

uic hapludolls)

Coland clay loam (fine-loamy,

mixed, superactive, mesic

cumulic endoaquolls)

4.5

6.5

Table 2 e Biomass crop selection, source, and planting density.

Biomass crop Scientific name Source Plant spacing Plants/location

m No.

Herbaceous

‘Bonilla’ big bluestema Andropogon gerardii Vitman cv. Bonilla Minnesota 0.30 648

‘Bison’ big bluestema Andropogon gerardii Vitman cv. Bison Minnesota 0.30 648

‘Sioux Blue’ indiangrassa Sorghastrum nutans (L.) Nash cv. Sioux Blue North Carolina 0.30 648

M. x giganteusb Miscanthus sinensis x Miscanthus sacchariflorus Maryland 0.91 150

‘Goliath’ miscanthusb Miscanthus sinensis Maryland 0.91 150

‘Aureomarginata’ prairie cordgrassb Spartina pectinata Bosc ex Link cv. Aureomarginata Maryland 0.30 648

‘Red River’ prairie cordgrassb Spartina pectinata Bosc ex Link cv. Red River North Dakota 0.30 648

‘Chiefton’ reed canarygrassa Phalaris arundinaceae L. cv. Chiefton Minnesota 0.30 648

‘Vantage’ reed canarygrassa Phalaris arundinaceae L. cv. Vantage Minnesota 0.30 648

‘Cloud Nine’ switchgrassa Panicum virgatum L. cv. Cloud Nine North Carolina 0.30 648

‘Northwind’ switchgrassa Panicum virgatum L. cv. Northwind Maryland 0.30 648

‘Sunburst’ switchgrassa Panicum virgatum L. cv. Sunburst Minnesota 0.30 648

Sunflowerb Helianthus tuberosus x H. annuus Minnesota 0.91 150

Jerusalem artichokeb Helianthus tuberosus L. SE MN ecotype 0.91 150

Stiff goldenrodc Oligoneuron rigidum (L.) Small var. rigidum EC MN ecotype 0.6 � 0.3 378

Giant goldenrodc Solidago gigantean Aiton SE MN ecotype 0.6 � 0.3 378

Polycultured Various spp. MN ecotypes

Woody species

SX 67 willowe Salix miyabeana New York 0.61 � 0.76 252

SV1 willowe Salix dasyclados New York 0.61 � 0.76 252

False indigo (A16)c Amorpha fruticosa (MN accession#16) Minnesota 1.22 84

False indigo (A19)c Amorpha fruticosa (MN accession#19) Minnesota 1.22 84

Hazelnutf Corylus x hybrid Wisconsin 1.22 84

‘Precocious’ hazelnutf Corylus x hybrid Michigan 1.22 84

Late lilacg Syringa villosa Minnesota 1.22 84

Common lilacg Syringa vulgaris Minnesota 1.22 84

a Transplanted as 30 cm potted plants.b Transplanted as rhizomes.c Transplanted as 15 cm potted plants.d Commercially formulatedmesic prairie mix comprising grasses and forbs including big bluestem, wild rye, switchgrass, indiangrass, and stiff

goldenrod (Prairie Restorations, Inc., Princeton, MN 55271).e Planted as 15 cm un-rooted cuttings.f Planted as 15 cm bare root plants.g Planted as 30 cm bare root plants.

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4 269

61 g ai ha�1 in forbs and dicamba (3,6-dichloro-2-

methoxybenzoic acid) at 28.0 g ha�1 active ingredient in

grass crops. Hand weeding was also performed as needed

throughout the duration of the study. Herbaceous biomass

yield was measured by harvesting all plant material within

two 0.5 m2 quadrats to a height of 10 cm in each plot. Har-

vesting occurred between September 21e28, September 29 e

October 21, October 5e19, September 28 e November 9, and

October 17e20 in 2007, 2008, 2009, 2010, and 2011, respectively.

Samples were weighted immediately after harvest, dried in a

forced-air oven at 45 �C for 5e7 days, and weighed again to

determine moisture content and biomass on a dry weight

basis. After samples were harvested, all remaining plant ma-

terial was removed from the entire plot to simulate harvest

from a typical biomass production system. Datawere adjusted

to reflect actual crop stand and dry matter (d m) reported as

Mg ha�1. Data were also adjusted to account for actual sample

area among plots with different plant spacing.

Woody biomass crops were planted at the same time as

herbaceous plants (2006), but harvested only once during

the study period, in 2010. Therefore, biomass data from

woody crops represents five years of growth, except for the

willow crop. Willow was managed using a short-rotation

coppice strategy, i.e. plants were coppiced fall 2006 fol-

lowed by undisturbed growth from 2007 until harvest in

2010. Therefore, willow biomass yields were measured four

years after coppicing, compared to after five years for all

other woody crops. This is slightly longer than a normal

production harvest cycle for willow, i.e. 3 years vegetative

growth, but within normal production harvest cycles for

other woody crops. Woody plant samples were harvested

from two 5.95 m2 areas in each plot in November 2010 at

Waseca, May 2011 at St. Paul and June 2011 in Lamberton.

Wood samples were immediately chipped, weighed, and

dried at 45 �C for 10e14 days to determine biomass on a dry

weight basis.

Data were analyzed using the MIXED procedure in SAS

(SAS Institute, Cary, NC) at a ¼ 0.10. Location and year were

considered fixed while block (nested within location) was

considered random. A significant year and location by year

interaction effect was noted in the model. Therefore, data

were analyzed by year and site. Datawere analyzed separately

for herbaceous and woody crops because of differences in

harvest cycles.

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4270

3. Results

3.1. Rainfall

Rainfall amount varied considerably among and across loca-

tions from 2007 to 2011 (Table 1). At St. Paul, rainfall was near

normal in 2007 while a rainfall deficit of 22.5 and 23.9 cm

occurred in 2008 and 2009, respectively, during the growing

season. Rainfall was above normal in 2010 and 2011 by 13.7

and 11.8 cm, respectively. At Waseca, rainfall was above

normal in 2007 and 2010 by 12.3 and 36.0 cm, respectively.

Conversely, precipitation was below normal in 2008, 2009, and

2011 with a rainfall deficit of 8.7, 24.0, 5.1 cm, respectively. At

Lamberton, rainfall was below normal from 2007 to 2009 by

5.9, 9.6, and 9.9 cm, respectively during the growing season.

Precipitation was above normal in 2010 by 5.2 and 2.5 cm,

respectively. Growing season precipitation tended to be lower

at Lamberton compared to other location across all years, with

the exception of 2009. Growing season precipitation tended to

be greater at Waseca compared to St. Paul and Lamberton in

2007, 2008, and 2010. Precipitation at Waseca was less than

Lamberton in 2009 and less than St. Paul in 2011.

3.2. Crop establishment and persistence

Herbaceous crop survival was >85% of the original stand in

2007 (first year after planting) across all siteswith the exception

of Miscanthus x giganteus (43e63% survival) and ‘Goliath’ mis-

canthus (33e58% survival) due to winter injury during the

winter of 2006 and 2007. Replanting of M. x giganteus and

‘Goliath’ miscanthus rhizomes significantly improved plant

stand by spring 2008. Although crop establishment was excel-

lent for ‘Northwind’ switchgrass, ‘Cloud 9’ switchgrass, late

Table 3 e Combined total biomass production of herbaceous anthe St. Paul, Waseca, and Lamberton locations.

Treatment St. Paul

‘Bison’ big bluestem 21.5 cdb

‘Bonilla’ big bluestem 21.1 cd

‘Sioux Blue’ indiangrass 24.6 cd

Miscanthus x giganteus 41.9 b

‘Goliath’ miscanthus 43.0 ab

‘Red River’ prairie cordgrass 40.3 b

‘Aureomarginata’ prairie cordgrass 41.3 b

‘Sunburst’ switchgrass 36.6 b

Jerusalem artichoke (wild) 26.1 c

Stiff goldenrod 19.8 cd

Polyculture 16.6 de

‘SX67’ willow 50.7 a

‘SV1’ willow 37.1 b

False indigo (A16) 6.6 f

False indigo (A29) 7.0 f

Hazelnut 6.3 f

‘Precocious’ hazelnut 8.6 ef

Lilac (late) 21.5 cd

Lilac (common) 22.2 cd

a Combined biomass dry matter production.b For each location, means within the same column followed by the sam

goldenrod, Sunflower, and Jerusalem artichoke in 2007, plant

density declined significantly in 2008 and 2009 resulting in very

low biomass yields. Therefore, these crops were removed from

the experiment in 2009 due to lack of productivity. Both culti-

vars of reed canarygrasswere removed from the experiment in

2009 because of aggressive growth that threatened to interfere

with growth and development of crops in surrounding plots.

This is in agreement with Jakubowski et al. [14] that found reed

canarygrass to be among the most vigorous and aggressive

plants in an upland agricultural environment.

Woody crop survival was>85% at St. Paul andWasecawith

the exception of ‘Precocious’ hazelnut at St. Paul (57% sur-

vival). At Lamberton, crop survival was 83% for ‘SX67’ willow,

75% for ‘SV1’ willow, 20% for both cultivars of false indigo, 13%

and 5% for hazelnut and ‘Precocious’ hazelnut, and 95% for

both cultivars of lilac. Consequently, false indigo and hazelnut

were removed from the experiment at Lamberton due to poor

establishment. Below-normal precipitation during the first

three years of the study likely resulted in poor establishment

of woody crop species at Lamberton.

3.3. Cumulative woody and herbaceous biomassproduction

Woody biomass crops are typically harvested on a 3e4 year

cycle whereas herbaceous biomass crops are harvested

annually. To compare biomass productivity between all crops

(woody and herbaceous), cumulative herbaceous biomass

yield was summed over a 4-year period (2007e2010) and

compared to the one-time woody biomass of woody biomass

crops in 2010.

‘SX67’ willow produced the most biomass compared to

other woody and herbaceous species across all locations

dwoody biomass crops over a 4-year period (2007e2010) at

Waseca Lamberton

Mg ha�1a

19.0 ghi 15.3 d

23.8 g 19.8 d

19.4 ghi 13.0 d

69.4 a 21.5 cd

21.0 gh 18.4 d

40.5 c 36.4 bc

49.8 b 47.5 b

35.5 cd 28.8 cd

32.5 cdef 25.2 cd

33.5 cde 26.6 cd

30.4 def 27.7 cd

72.0 a 66.0 a

49.2 b 55.4 a

10.4 j e

15.5 hij e

11.3 i e

13.2 hij e

27.0 efg 25.7 cd

24.4 fg 27.2 cd

e letter are not significantly different (a > 0.10).

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4 271

(Table 3). However, ‘SX67’ willow drymatter yieldwas variable

between locations and tended to be greatest at Waseca

(72 Mg ha�1) followed by Lamberton (66 Mg ha�1) and St. Paul

(50.7 Mg ha�1). At St. Paul, biomass production of ‘Goliath’

miscanthus and M. x giganteus was not different from ‘SX67’

willow at St. Paul and Waseca locations, respectively. At

Lamberton, there was no difference between ‘SX67’ and ‘SV1’

willow.

At St. Paul, M. x giganteus, prairie cordgrass, ‘Sunburst’

switchgrass, and ‘SV1’ willow biomass production was similar

to ‘Goliath’ miscanthus, but less than ‘SX67’ willow (Table 3).

At Waseca, ‘Aureomarginata’ prairie cordgrass and ‘SV1’ wil-

low produced >49 Mg ha�1 dry matter of biomass which was

more than all other species, with the exception of ‘SX67’ wil-

low and M. x giganteus. Both cultivars of prairie cordgrass

produced >36 Mg ha�1 dry matter at Lamberton. ‘Sunburst’

switchgrass produced more biomass than the polyculture at

St. Paul. Conversely, there was no difference in biomass pro-

ductivity between ‘Sunburst’ switchgrass and polyculture at

Waseca and Lamberton.

3.4. Herbaceous biomass production over time

3.4.1. St. Paul location‘Red River’ and ‘Aureomarginata’ cultivars of prairie cord-

grass produced more biomass than all other herbaceous

crops in 2007 (Table 4). In 2008, both cultivars of prairie

cordgrass, ‘Goliath’ miscanthus and ‘Sunburst’ switchgrass

produced the most biomass. ‘Sioux Blue’ indiangrass, M. x

giganteus, Jerusalem artichoke, and stiff goldenrod biomass

production was not different from ‘Goliath’ miscanthus.

‘Goliath’ miscanthus, ‘Sunburst’ switchgrass, and ‘Aur-

eomarginata’ prairie cordgrass were among the top biomass

producers in 2009. In 2010, both cultivars of miscanthus

produced more biomass than any other herbaceous crop;

however, M. x giganteus produced more biomass than

‘Goliath’ miscanthus. As in 2010, both cultivars of mis-

canthus produced more biomass than any other species in

2011. However, biomass production between both cultivars

was not different.

Table 4 e Productivity of herbaceous biomass crops at St. Paul

Biomass crops 2007 2008

‘Bison’ big bluestem 4.8 cb 5.4

‘Bonilla’ big bluestem 5.5 c 5.9

‘Sioux Blue’ indiangrass 6.1 c 8.5

Miscanthus x giganteus 2.5 d 7.8

‘Goliath’ Miscanthus 2.9 d 10.0

‘Red River’ prairie cordgrass 10.2 a 12.0

‘Aureomarginata’ prairie cordgrass 11.4 a 11.9

‘Sunburst’ switchgrass 8.5 b 10.9

Jerusalem artichoke (wild) 5.0 c 8.8

Stiff goldenrod 5.3 c 7.7

Polyculture 2.6 d 5.4

a Biomass dry matter production.b For each year, means within the same column followed by the same le

‘Sunburst’ switchgrass was among the greatest and most

consistent biomass producers with yields ranging between 8.1

and 10.9 Mg ha�1 dry matter across all five years of the study

and was not different from both cultivars of prairie cordgrass,

except in 2007. There was also no difference between M. x

giganteus and ‘Goliath’ miscanthus in 2007 and 2008. In 2009,

‘Goliath’ produced more biomass than M. x giganteus while in

2010 M. x giganteus produced more biomass than ‘Goliath’

miscanthus. By 2011, there was no difference in biomass yield

between M. x giganteus and ‘Goliath’ miscanthus with both

cultivars producing >23 Mg ha�1 dry matter. Miscanthus

biomass yield tended to be higher during the last two years of

the study and likely contributed to high miscanthus biomass

yields.

3.4.2. Waseca location‘Red River’ and ‘Aureomarginata’ prairie cordgrass produced

more biomass than any other herbaceous biomass crop in

2007 (Table 5). However, there was no difference between ‘Red

River’ prairie cordgrass, ‘Sunburst’ switchgrass, and stiff

goldenrod. This trend continued in 2008, except that therewas

no difference between ‘Red River’ prairie cordgrass and M. x

giganteus and Jerusalem artichoke. By 2009, M. x giganteus

produced more biomass than all other herbaceous species.

Both cultivars of prairie cordgrass, ‘Goliath’ miscanthus,

‘Sunburst’ switchgrass, Jerusalem artichoke, stiff goldenrod,

and polyculture all produced similar amounts of biomass but

were less than M. x giganteus. As in 2009, M. x giganteus pro-

duced the most biomass in 2010 and 2011. ‘Goliath’ mis-

canthus and ‘Aureomarginata’ produced >12 Mg ha�1 dry

matter but was less than M. x giganteus in 2010. In 2011,

biomass production was similar between ‘Goliath’ mis-

canthus and polyculture, but less than M. x giganteus.

At Waseca, M. x giganteus biomass yield was greater than

‘Goliath’ miscanthus across all years. Prairie cordgrass yield

was not different between cultivars from 2007 to 2009. How-

ever, ‘Aureomarginata’ prairie cordgrass produced more

biomass than ‘Red River’ in 2010 and 2011. The polyculture

crop produced 3.9 Mg ha�1 dry matter in 2007, but by 2011 the

polyculture crop produced >20 Mg ha�1 dry matter.

from 2007 to 2011.

2009 2010 2011

Mg ha�1a

d 6.1 bcd 5.2 e 7.4 bcd

cd 5.0 cd 4.8 e 8.9 bcd

bc 3.9 d 6.1 de 5.5 d

bcd 6.3 bcd 25.3 a 25.2 a

ab 11.3 a 18.8 b 23.1 a

a 7.4 bc 10.7 c 12.9 b

a 8.7 ab 9.4 cd 12.6 bc

a 8.1 abc 9.2 cd 9.3 bcd

bc 5.1 cd 7.3 cde 10.5 bcd

bcd 3.3 d 3.5 e 7.3 cd

def 3.7 d 4.8 e 8.3 bcd

tter are not significantly different (p ¼ 0.10).

Table 5 e Productivity of herbaceous biomass crops at Waseca from 2007 to 2011.

Species 2007 2008 2009 2010 2011

Mg ha�1b

‘Bison’ big bluestem 4.5 da 6.4 ef 4.9 de 3.1 g 7.2 ef

‘Bonilla’ big bluestem 4.9 d 8.4 def 5.4 cde 5.1 efg 7.9 ef

‘Sioux Blue’ indiangrass 6.6 c 5.9 f 2.6 d 4.3 fg 3.0 f

Miscanthus x giganteus 4.0 d 13.8 ab 28.1 a 23.4 a 35.3 a

‘Goliath’ Miscanthus 1.8 e 3.0 e 10.6 b 12.6 bc 16.3 bc

‘Red River’ prairie cordgrass 10.7 ab 14.2 ab 8.1 bcd 7.5 def 9.5 de

‘Aureomarginata’ prairie cordgrass 11.4 a 14.9 a 10.1 b 13.4 b 13.5 cd

‘Sunburst’ switchgrass 9.2 b 10.7 cd 8.3 bc 7.3 def 10.4 de

Jerusalem artichoke (wild) 5.1 cd 12.0 bc 6.7 bcd 8.8 cde 7.5 ef

Stiff goldenrod 9.1 b 8.6 de 8.0 bcd 7.8 def 5.7 ef

Polyculture 3.9 d 9.0 d 8.3 bc 9.2 cd 20.3 bc

a For each year, means within the same column followed by the same letter are not significantly different (p ¼ 0.10).b Biomass dry matter production.

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4272

3.4.3. Lamberton location‘Aureomarginata’ prairie cordgrass produced more biomass

than all other herbaceous crops in 2007 (Table 6). ‘Red River’

prairie cordgrass and stiff goldenrod produced>8 Mg ha�1 dry

matter, but was less than ‘Aureomarginata’ prairie cordgrass.

In 2008, both cultivars of prairie cordgrass, ‘Sunburst’

switchgrass, and the polyculture crop producedmore biomass

than all other crops, with the exception of Jerusalem artichoke

and stiff goldenrod. ‘Aureomarginata’ continued to produce

the greatest amount of biomass in 2009, but was not different

from M. x giganteus, ‘Goliath’ miscanthus, and ‘Red River’

prairie cordgrass. The polyculture crop produced almost

9 Mg ha�1 dry matter, but was less than ‘Aureomarginata’

prairie cordgrass. In 2010, ‘Aureomarginata’ prairie cordgrass

produced more biomass than all other species with the

exception of M. x giganteus. By 2011, M. x giganteus produced

more biomass than any other herbaceous crop. ‘Goliath’

miscanthus produced >20 Mg ha�1 dry matter more than any

other crop with the exception of M. x giganteus.

There was no difference in biomass production betweenM.

x giganteus and ‘Goliath’ miscanthus except in 2011 whereM. x

giganteus produced more biomass than ‘Goliath’. The prairie

Table 6 e Productivity of herbaceous biomass crops at Lamber

Species 2007 2008

‘Bison’ big bluestem 3.2 ea 2.5

‘Bonilla’ big bluestem 4.5 d 4.2

‘Sioux Blue’ indiangrass 6.8 c 2.5

Miscanthus x giganteus 1.5 f 4.2

‘Goliath’ Miscanthus 2.0 f 2.6

‘Red River’ prairie cordgrass 8.9 b 8.9

‘Aureomarginata’ prairie cordgrass 11.9 a 10.7

‘Sunburst’ switchgrass 6.5 c 8.8

Jerusalem artichoke (wild) 6.3 c 7.9

Stiff goldenrod 8.3 b 6.5

Polyculture 4.2 de 8.0

a For each year, means within the same column followed by the same leb Biomass dry matter production.

polyculture crop produced only 4.2Mg ha�1 drymatter in 2007,

but 6.7e8.9 Mg ha�1 dry matter in subsequent years. Switch-

grass biomass yield was more variable at Lamberton with

yields ranging between 6.0 and 9.8 Mg ha�1 dry matter across

years.

4. Discussion

Crop establishment was generally good across all locations.

Among the switchgrass cultivars, ‘Sunburst’ was the only

cultivar to successfully establish and produce a consistent

source of biomass throughout the duration of the study.

Sunflower and late goldenrod failed to provide adequate

biomass across all locations. At Lamberton, false indigo and

hazelnut did not establish and were removed from the

experiment.

The goal of this study was to assess variability in crop

biomass production across locations over the course of a

typical cropping cycle. For the perennial herbaceous crops,

‘Red River’ and ‘Aureomarginata’ prairie cordgrass were the

most consistent and among the highest yielding herbaceous

ton from 2007 to 2011.

2009 2010 2011

Mg ha�1b

d 4.6 ef 5.0 cd 5.8 def

cd 5.8 de 5.4 cd 8.8 cde

d 1.6 f 2.0 d 3.4 ef

cd 11.2 abc 10.0 ab 25.9 a

d 11.5 ab 6.8 bc 20.2 b

ab 11.1 abc 7.6 bc 10.7 cd

a 12.6 a 12.3 a 13.9 c

ab 7.5 cde 6.0 c 9.8 cd

b 4.4 ef 6.6 bc 5.6 def

bc 7.0 de 4.8 cd 3.0 f

ab 8.9 bcd 6.7 bc 7.3 def

tter are not significantly different (p ¼ 0.10).

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4 273

biomass crops across locations and years. However, there

were differences in prairie cordgrass productivity across lo-

cations and years that appear to be driven primarily by soil

water. For example, prairie cordgrass biomass production

tended to be greater at Waseca and St. Paul compared to

Lamberton, especially during years when precipitation was

below normal. Prairie cordgrass yield was also reduced from

2008 to 2009 at Waseca and Lamberton due to below normal

rainfall in 2008. However, biomass production quickly recov-

ered in 2010 during which time precipitation was significantly

above normal. Interestingly, the ‘Aureomarginata’ cultivar of

prairie cordgrass performed better at Lamberton compared to

the other locations indicating that this cultivar may perform

better in drier environments than the ‘Red River’ cultivar.

Both M. x giganteus and ‘Goliath’ miscanthus produced the

most biomass by the end of the study period compared to

other crops. Heaton et al. [16] examined peer-reviewed articles

and reported M. x giganteus yields averaged 22 Mg ha�1 dry

matter after three ormore years after planting. Moreover, they

noted that M. x giganteus showed a strong response to soil

water. In our trials, M. x giganteus averaged 19.5 Mg ha�1 dry

matter across locations after four years with the greatest yield

obtained from locations with higher annual rainfall amounts.

In 2011, five years after planting, yield of M. x giganteus aver-

aged 28.8 Mg ha�1 dry matter. Although M. x giganteus and

‘Goliath’ miscanthus produced high yields during in the final

year of the study, there were issues related to overwintering

success during the first few years. Miscanthus is reported to

have poor over-winter survival in temperate areas with cold

winters and little snow cover [17,18]. Clifton-Brown and

Lewandowski [18] noted thatM. x giganteus rhizome survival is

affected when soil temperatures fall below �3 �C at the 5 cm

soil depth, particularly during the first winter after the

establishment year. In Minnesota, soil temperature typically

falls below this point, especially during periods with little or

no snow cover. For example, soil temperature reached �10 �Cat 5 cm soil depth on January 13, 2007 at Waseca and snow

depthwas<1.3 cm. Low soil temperature and little snow cover

likely resulted in poor establishment of M. x giganteus and

‘Goliath’ miscanthus. However, new cultivars are being

developed that have shown promise in colder climates pri-

marily due to greater survivability under cold soil conditions.

A concern with miscanthus and prairie cordgrass is that

these crops are typically established from rhizomes making

planting time consuming and costly. Use of crops that can be

established from seed with existing farm equipment can

improve efficiency and reduce establishment cost. Native

grass crops that dominated tallgrass prairie environments,

such as big bluestem, switchgrass, and indiangrass, are ex-

amples of potentially high-yielding native crops that can be

established from seed. Both cultivars of big bluestem pro-

duced stable amounts of biomass across contrasting wet and

dry environments between 2007 and 2010 with the exception

of Waseca, which experienced a reduction in biomass from

2008 to 2010. Conversely, ‘Sioux Blue’ indiangrass tended to

decline in productivity over time across all sites, especially in

dryer environments like Lamberton. Switchgrass, another

native grass established from seed, was also among the

highest yielding biomass crops. Although yield never

approached that ofmiscanthus or prairie cordgrass, it was still

consistently around 8e12 Mg ha�1 dry matter throughout the

duration of the study, especially for ‘Sunburst’ switchgrass,

depending on location. These observations demonstrate the

need to evaluate crop options given time, labor, and biomass

yield potential as a function of environment.

With the exception of the polyculture treatment, all

biomass crops in this study were planted in monocultures,

fertilized with N, and sprayed for weed control. Polycultures

consisting of mixtures of grass species or mixtures of grass,

forbs, and legume species may offer an alternative to mono-

cultures that maintain biomass productivity while reducing

input costs [19,20]. Biomass yield of the native plant poly-

culture was, low relative to N fertilized monocultures, during

the first year after establishment ranging from 2.6 to

4.2 Mg ha�1 dry matter in (2007), which is not unusual [19].

However, the polyculture treatments showed an upward

trend in yield over years, except at the Lamberton. This was

especially evident atWasecawhere initial biomass production

was 3.9Mg ha�1 drymatter in 2007. By 2011, however, biomass

production was 20.3 Mg ha�1 dry matter, yielding better than

any other crop with the exception of M. x giganteus.

There is little information available relating the produc-

tivity of woody biomass crops compared to perennial herba-

ceous crops. This is likely due to inherent differences in

harvest cycles betweenwoody and herbaceous biomass crops.

For example, willow is typically harvested every 3e4 years

whereas herbaceous crops are harvested annually. The

approach used in this study was to compare cumulative yield

of herbaceous crops over four years with woody biomass

harvested once during the study period. Willow produced

more biomass than all other woody and herbaceous biomass

crops during the 4-year period. However, miscanthus biomass

yield was similar to ‘SX67’ willow at St. Paul and Waseca, but

was dependent on the cultivar of miscanthus. Labrecque and

Teodorescu [15] reported biomass yields of 46.6 t ha�1 dry

matter for ‘SV1’ willow and 37.7 t ha�1 dry matter for ‘SX67’

willow in Canada. However, there is no information on the

performance of these clones in Minnesota soils and climate,

especially given a 4-year rather than the normal 3-year har-

vest interval. Although willow produced more biomass than

most other crops in this study, it is important to take into

account the effect of harvest frequency on the economics of

these systems.

5. Conclusions

Our objective was to quantify biomass yield of perennial

herbaceous andwoody crops over time and environments as a

basis for the development of more diversified landscape

management strategies. This study shows that several

biomass crop species can be successfully grown as part of a

comprehensive farm management strategy aimed at

increasing crop diversification. However, there were differ-

ences in biomass production among the crops tested as a

function of environment. Prairie cordgrass consistently pro-

duced the most biomass across years and locations. However,

the greatest biomass production was realized in eastern

and south central Minnesota where annual rainfall is gener-

ally higher. Miscanthus was also among the top biomass

b i om a s s a n d b i o e n e r g y 5 8 ( 2 0 1 3 ) 2 6 7e2 7 4274

producers, but only during the later part of the study period.

This was due to poor establishment during the first years of

the study and may therefore limit its use in colder climates.

Considering cumulative biomass production over a 4-year

period, willow and miscanthus generally produced the most

biomass compared to other crops. Other considerations must

be taken into account when making decisions related what

biomass crop(s) to plant such as economics, equipment

availability, and labor requirements. These considerations

may prevent using crops like miscanthus, prairie cordgrass,

and willow that typically cost more to plant. However, we

found that other crops such as switchgrass, big bluestem,

indiangrass, and the polyculture plantings all performed well

depending on location and year and can be considered good

alternatives to miscanthus and prairie cordgrass. For

example, switchgrass consistently produced 8e12Mg ha�1 dry

matter throughout the duration of study. The polyculture

treatment also performed well towards the end of the study

period, producing up to 20 Mg ha�1 dry matter at Waseca in

2011.

r e f e r e n c e s

[1] Marshall AL, Alaimo PJ. Useful products from complexstarting materials: common chemicals from biomassfeedstocks. Chem-Eur J 2010;16:4970e80.

[2] Gallezot P. Direct routes from biomass to end-products. CatalToday 2011;167:31e6.

[3] Langeveld JW a, Dixon J, Jaworski JF. Developmentperspectives of the biobased economy: a review. Crop Sci2010;50:S-142e51.

[4] Bridgwater AV. Review of fast pyrolysis of biomass andproduct upgrading. Biomass Bioenergy 2011;38:1e27.

[5] Hatti-Kaul R. Biorefineries e a path to sustainability? Crop Sci2010;50:S-152e6.

[6] Ragauskas AJ, Williams CK, Davison BH, Britovsek G,Cairney J, Eckert CA, et al. The path forward for biofuels andbiomaterials. Science 2006;311(5760):484e9.

[7] Blanco-Canqui H. Energy crops and their implications on soiland environment. Agron J 2010;102:403e19.

[8] Dale VH, Kline KL, Wright LL, Perlack RD, Downing M,Graham RL. Interactions among bioenergy feedstock choices,landscape dynamics, and land use. Ecol Appl2011;21:1039e54.

[9] Boehmel C, Lewandowski I, Claupein W. Comparing annualand perennial energy cropping systems with differentmanagement intensities. Agr Syst 2008;96:224e36.

[10] Berges SA, Schulte Moore LA, Isenhart TM, Schultz RC. Birdspecies diversity in riparian buffers, row crop fields, andgrazed pastures within agriculturally dominatedwatersheds. Agr Syst 2010;79:97e110.

[11] Jordan N, Boody G, Broussard W, Glover J, Keeney D,McCown B, et al. Sustainable development of the agriculturalbio-economy. Science 2007;316(5831):1570e1.

[12] Kort J, Collins M, Ditsch D. A review of soil erosion potentialassociated with biomass crops. Biomass Bioenergy1998;14:351e9.

[13] Gonzalez-Hernandez JL, Sarath G, Stein JM, Owens V,Gedye K, Boe A. A multiple species approach to biomassproduction from native herbaceous perennial feedstocks.In Vitro Cell Dev Plant 2009;45:267e81.

[14] Jakubowski AR, Casler MD, Jackson RD. Has selection forimproved agronomic traits made reed canarygrass invasive?PLoS ONE 2011;6:e25757.

[15] Labrecque M, Teodorescu TI. Field performance and biomassproduction of 12 willow and poplar clones in short-rotationcoppice in southern Quebec (Canada). Biomass Bioenergy2005;29:1e9.

[16] Heaton E, Voigt T, Long SP. A quantitative review comparingthe yields of two candidate C4 perennial biomass crops inrelation to nitrogen, temperature and water. BiomassBioenergy 2004;27:21e30.

[17] Anderson E, Arundale R, Maughan M, Oladeinde A,Wycislo A, Voigt T. Growth and agronomy of Miscanthus xgiganteus for biomass production. Biofuels 2011;2:167e83.

[18] Clifton-Brown JC, Lewandowski I. Overwintering problems ofnewly established miscanthus plantations can be overcomeby identifying genotypes with improved rhizome coldtolerance. New Phytol 2012;148:287e94.

[19] Mangan ME, Sheaffer C, Wyse DL, Ehlke NJ, Reich PB. Nativeperennial grassland species for bioenergy: establishmentand biomass productivity. Agron J 2011:509e19.

[20] Mulkey VR, Owens VN, Lee DK. Management of warm-season grass mixtures for biomass production in SouthDakota USA. Bioresour Technol 2008;99:609e17.