Society for Range Management
Characterization of Diversity among 3 Squirreltail Taxa
Author(s): T. A. Jones, D. C. Nielson, J. T. Arredondo, M. G. Redinbaugh
Source: Journal of Range Management, Vol. 56, No. 5 (Sep., 2003), pp. 474-482
Published by: Allen Press and Society for Range Management
Stable URL: http://www.jstor.org/stable/4003839
Accessed: 14/12/2009 14:15
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=acg.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Allen Press and Society for Range Management are collaborating with JSTOR to digitize, preserve and extend
access to Journal of Range Management.
http://www.jstor.org
J. Range Manage.
56: 474-482 September 2003
Characterizationof diversity among 3 squirreltail taxa
T.A. JONES*, D.C. NIELSON, J.T. ARREDONDO, AND M.G. REDINBAUGH
T.A. Jones and D.C. Nielson, Research Geneticist and Entomologist, USDA-ARSForage and Range Research Laboratory, Utah State University,Logan,
Utah 84322-6300; former Research Associate, J.T. Arredondo,Departamentode IngenieriaAmbientaly Manejo de Recursos Naturales. InstitutoPotosino de
Investigacion Cientifica y Tecnologica (IPICYT)Rio Kennedy205, Col. Bellas Lomas San Luis Potosi, SLP, CP78210 Mexico; M.G. Redinbaugh,Research
Plant Physiologist, USDA-ARS,Corn and Soybean Research, Ohio Agriculturaland Research Development Center, Selby Hall, 680 Madison Ave., Wooster,
Ohio 44691-4096.
Abstract
Resumen
Squirreltail (Elymus elymoides, E. multisetus) is a complex of 5
taxa whose systematic interrelationships are uncertain. Our
objectives were to determine whether the 3 taxa studied here,
Elymus elymoides ssp. elymoides, E. elymoides ssp. brevifolius, and
E. multisetus, can be distinguished by several ecological and
physiological traits and whether geographical origin is correlated
with these traits across accessions within taxa. A multivariate
principal component analysis of materials collected in the 10 contiguous western states successfully distinguished taxa, but no pair
of the 3 taxa appeared to be more ecologically similar than any
other pair. Elymus elymoides ssp. elymoides, which prevails in the
semi-arid cold desert, was shortest and exhibited the lowest total
plant dry-matter, earliest phenology, and lowest seed mass.
Elymus elymoides ssp. brevifolius, which prevails in the Rocky
Mountains, exhibited slowest emergence, highest specific root
length, lowest nitrate reductase activity, and lowest root-to-shoot
ratio. Elymus multisetus, which is most common in areas with relatively warm springs, exhibited fastest emergence (particularly
from deep seeding), greatest root length, and greatest root-toshoot ratio. Elymus elymoides ssp. brevifolius accessions clustered
into 3 groups: late-maturing high-seed mass accessions originating in Colorado, New Mexico, and Arizona (Group A), earlymaturing low-seed mass accessions originating in Colorado and
Utah (Group B), and intermediate-maturing low-seed mass
accessions originating in the Snake River Plain of southern Idaho
(Group C). The ecologically distinct subspecies and groups within ssp. brevifolius are indicative of the highly ecotypic nature of
the squirreltails, suggesting that restoration practitioners should
match site with genetically and ecologically appropriate plant
material for these species.
"Squirreltail (Elymus elymoides, E. multisetus) es un complejo
de 5 taxas cuyas interrelaciones sistematicas son inciertas.
Nuestros objetivos fueron determinar si las 3 taxas estudiadas
aqui Elymus elymoides ssp. elymoides, E. elymoides ssp. brevifolius, y E. multisetus, pueden ser distinguidas por varias caracteristicas ecologicas y fisiologicas y si el origen geografico esta
correlacionado con estas caracterristicasa traves de las accesiones
dentro de la taxa. Un anailisismultivariado de componentes principales de los materiales colectados en los 10 estados contiguos
del oeste distinguio en forma exitosa la taxa, pero ningun par de
las tres taxas parecio ser mas ecologicamente similar que
cualquier otro par. Elymus elymoides ssp. elymoides, el cual
prevalece en el desierto frio semiarido, fue el mas corto y mostro
la menor materia seca por planta, la fisiologfa mas temprana y la
menor biomasa de semilla. Elymus elymoides ssp. brevifolius, que
prevalece en las Montauias Rocallosas presento la emergencia
mas lenta, la mayor longitud especifica de raiz, la mas baja
actividad de nitrato reductasa y la relacion mas baja de
tallo:raiz. Elymus multisetus, que es mas comuinen aireascon primaveras relativamente calientes, mostro la emergencia ma'srapida (particularmente en siembras profundas), la mayor longitud
de raiz y la mayor relacion tallo:raiz. Las accesiones de Elymus
elymoides ssp. brevifolius se agruparon en tres grupos: accesiones
de madurez tardia-alta biomasa de semilla originarias de
Colorado, New Mexico y Arizona (Grupo A), accesiones de
madurez temprana -baja biomasa de semilla originarias de
Colorado y Utah (Grupo B) y un grupo de accesiones de
madurez intermedia-baja biomasa de semilla originarias del
Snake River Plain del sudeste de Idaho (Group C). Las subespecies y grupos ecologicamente distintos dentro de ssp. brevifolius son indicativos de la naturaleza altamente ecotipica de los
"Squirreltails, sugiriendo que los que practican la restauracion
de pastizales deben aparejar el sitio con un el material vegetal
gen'tica y ecol6gicamente apropiado de estas especies.
Key Words: Elymus elymoides, genetic variation, plant autecology, Sitanion hystrix
The squirreltails (Elymus elymoides [Raf.] Swezey andE. multisetus [J.G. Smith] Burtt-Davy) are increasingly important grasses for restoring rangeland communities on disturbed sites in the
Intermountain Region of western North America. Systematists
generally recognize up to 5 squirreltailtaxa (Wilson 1963), which
we refer to in this paper as elymoides, brevifolius, californicus,
hordeoides, and multisetus. These taxa have been recognized in
This research was supported by the Utah Agricultural Experiment Sation, Utah
State University, Logan, Utah 84322-4810. Approved as journal paper no. 7386.
*Corresponding author (tomjones@cc.usu.edu).
Manuscript accepted 10 Oct. 02.
474
both Sitanion and in Elymus, with the latter being the currently
preferred classification (Barkworth et al. 1983). However, rank of
the taxa varies with taxonomic treatment.
There seems to be little dissension regarding the retention of 2
taxa in E. elymoides, namely ssp. elymoides and ssp. californicus.
Wilson (1963) mentioned the marked similarity of their karyotypes. Also, M.E. Barkworth and J.J.N. Campbell (pers. comm.
2000) noted that they appear to hybridize when sympatric.
Taxonomic treatments have varied in their handling of specimens keying to brevifolius and hordeoides in Wilson's (1963)
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
treatment.Brevifoliuswas treatedas a separatespecies by Wilson (1963) (S. longifolium J.G. Smith),reluctantlytreatedas a
separate species by Holmgren and
Holmgren(1977), treatedas a subspecies
while others do so only reluctantly
(Holmgrenand Holmgren 1977). Wilson
(1963) notedthatthe karyotypeof multisetus was the most distinctiveof the 5 squirreltail taxa. He reportedextensive introof E. elymoides (ssp. brevifolius) by M.E. gressive hybridizationbetween multisetus
Barkworth and J.J.N. Campbell (2000 and elymoidesin southeasternCalifornia
pers.comm.),andunrecognizedas a sepa- and southernNevada,thoughnot in other
rate taxon by Hitchcock (1950) and locales where these 2 taxa are sympatric.
Arnow(1993). Hordeoideshas been treat- HolmgrenandHolmgren(1977) suggested
ed as a separatespecies by Wilson (1963) thata reasonableapproachwouldbe amal(S. hordeoides Suksdorf)or as part of a gamation of all 5 taxa into a single
species including elymoides, californicus,
species.
and brevifolius by Holmgren and
Holmgren(1977) and by M.E. Barkworth
andJ.J.N.Campbell(pers.comm.).
Some systematists readily recognize
multisetusas worthy of specific rank (E.
multisetus [J.G. Smith] Burtt-Davy = S.
Wilson (1963) provides a reliable key
that easily separatesthe 5 taxa based on
floral appendages. We have used it to
identifyover 100 accessionsin ourcollection, which includesrepresentativesof all
taxaexcepthordeoides.Ourobjectivewas
jubatum J.G. Smith) (Hitchcock 1950,
Wilson 1963, Barkworth et al. 1983),
to determine how 3 taxa, multisetus, brevifolius, and elymoides, compare for traits
relatingto seedlingestablishment,leaf and
root morphology,and phenologybecause
such traitsmay be correlatedwith adaptation to particular range sites. We also
wished to determinethe degreeof separation or overlapbetweenthe taxa basedon
these traitsandto gatherany insightsfrom
the relationshipbetweenperformanceand
geographicaloriginof the accessions.
Materials and Methods
Data set 1
Twenty-one of the 27 accessions
(Tables 1, 2) in dataset 1 originatedfrom
USDA-NRCS Plant MaterialsCentersin
Aberdeen,Ida. and Bridger,Mont.,or the
Upper Colorado Environmental Plant
Centerin Meeker,Colo.. Two accessions
Table 1. Means for 1 californicus, 17 elymoides, 5 brevifolius,and 4 multisetus squirreltail accessions in data set 1 for 8 traits in a 25-day greenhouse trial.
Accession
Collection location
Days to
emergence
(no.)
Leaf length
(10 days Total plant
post-plant) dry-matter
(mm)
Root-toshoot
ratio
(mg plant-') (mg mg-')
Leaf
area
Specific
leaf area
Root
length
Specific
root
length
(mm2
(mm2
(mm
(mm
plant-')
mg')
plantl-)
mg ')
californicus
PI 531604
Toe Jam Creek,Elko Co., Nev.
6.0
73
10.46
0.389
174
31.3
914
324
elymoides
Acc:1107
Acc:1 108
Acc:1109
Acc: 1110
Acc:1111
Acc:1112
Acc:1113
Acc:1114
Acc:1 115
Acc:I 116
Acc:1117
Acc:1124
PI 619489
PI 619491
PI 619555
PI 619561
9019219
sec 32 T22N R22E, Lemhi Co., Ida.
sec 36 T4N R26E, Butte Co., Ida.
sec 11 T4NR26E,ButteCo.,Ida.
sec 26 T4N R26E, Butte Co., Ida.
sec 11 T4NR26E,ButteCo.,Ida.
sec 11 T4NR27E,ButteCo.,Ida.
1737 m, CusterCo., Ida.
MorseCreek,CusterCo.,Ida.
LeatonGulch, CusterCo., Ida.
BradburyFlat, CusterCo., Ida.
RoundValley, CusterCo., Ida.
Squaw Butte, HarneyCo., Ore.
Whitehall,Mont.
Big Piney, Wyo.
Warren,Mont.
Big Piney, Wyo.
btwn Worland& Ten Sleep, Wyo.
5.4
6.3
6.0
5.8
6.2
5.1
5.3
6.2
6.2
6.0
5.6
6.1
5.4
5.6
6.2
5.8
5.8
65
61
58
58
58
67
74
54
62
64
62
60
63
60
57
60
64
9.56
9.57
9.26
6.98
8.38
9.79
8.85
7.14
9.57
9.56
10.18
10.27
8.86
10.33
9.36
9.69
10.15
0.424
0.315
0.306
0.516
0.442
0.365
0.423
0.291
0.405
0.335
0.354
0.359
0.427
0.352
0.285
0.436
0.325
148
168
182
110
158
175
157
116
167
178
182
150
142
191
174
167
214
29.1
29.4
31.9
31.0
33.5
31.0
31.3
27.7
32.6
33.8
32.9
25.2
29.0
32.1
31.7
32.0
35.2
916
698
652
687
853
838
1057
641
717
668
814
960
830
711
629
821
667
317
316
324
289
337
330
410
410
270
275
322
368
323
273
298
306
284
5.8b#
62b
0.374b
163b
31.1a
5.7
6.4
7.0
70
71
56
12.44
9.35
10.50
0.358
0.280
0.267
197
157
154
27.6
28.0
23.8
1017
700
794
343
376
368
7.4
6.3
49
65
10.75
12.10
0.305
0.333
158
210
25.0
29.7
836
943
361
332
6.6 a
62 b
11.03 b
0.309 c
175 b
26.8 b
858 b
356 a
4.8
4.4
4.8
5.5
85
85
82
84
16.60
18.48
13.89
15.45
0.483
0.483
0.392
0.429
269
303
245
262
31.2
31.0
32.3
31.2
1434
1980
1520
1161
268
335
393
278
4.9 c
84 a
16.11 a
0.447 a
270 a
31.4 a
1524 a
319 a
mean
brevifolius
Acc: 1123
Acc:1130
PI 531605
9040187
9040189
btwn PrairieCity & Dixie Pass, Ore.
Savageton,Wyo.
btwn N. LaVetaPass
& Gardner, Colo.
Wet Mountains,CusterCo., Colo.
Buford, Rio Blanco Co., Colo.
mean
multisetus
Acc:1132
PI 531603
PI 531606
PI 619466
mean
ParadiseValley, Nev.
Lake Tahoe, Nev.
CentralFerry,Wash.
Redlands,Lassen Co., Cal.
9.26c
774b
321a
#taxameans followed by differentlettersin the same column are significantlydifferentat P < 0.05.
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
475
Table 2. Means for 1 californicus, 17 elymoides, 5 brevifolius, and 4 multisetus squirreltail acces-
sionsin dataset 1 for 5 traitsin fieldand greenhousetrials.
Accession
Headingdate
(days after 30/4/95)
Seed mass
(mg seed-)
Emergenceindex
(60 mm)
(20 mm)
Nitrate
reductaseactivity
(per mg protein)
californicus
PI 531604
29
4.14
6.36
3.64
7.66
elymoides
Acc: 1107
Acc: 1108
Acc: 1109
Acc:1110
Acc:1111
Acc: 1112
Acc:1113
Acc: 1114
Acc: 1115
Acc: 1116
Acc:1117
Acc: 1124
PI 619489
PI 619491
PI 619555
PI 619561
9019219
mean
15
27
15
31
19
24
31
17
33
33
36
36
31
29
33
27
29
27 b#
3.27
2.67
2.71
2.95
3.02
2.47
2.94
3.07
3.43
3.49
3.33
3.44
3.26
3.48
3.00
2.86
4.07
3.14 b
6.71
6.54
6.62
6.51
6.31
6.74
7.21
6.36
6.67
6.83
7.43
6.64
5.71
6.44
6.60
6.56
6.18
6.59 a
3.01
2.01
3.03
2.34
2.80
2.90
3.38
3.50
4.36
4.40
4.80
2.93
1.14
3.40
2.82
3.41
4.32
3.21 ab
7.83
7.34
7.01
5.38
7.17
6.52
6.03
7.50
8.48
9.46
9.13
6.68
7.34
8.48
7.83
8.64
8.64
7.62 a
brevifolius
Acc:1123
Acc: 1130
PI 531605
9040187
9040189
mean
47
29
46
59
45
45 a
3.94
3.98
6.83
6.32
5.51
5.32 a
5.67
5.92
5.05
5.62
5.46
5.54 b
3.73
3.94
3.40
3.43
2.94
3.49 b
3.91
6.85
4.57
5.54
5.87
5.35 b
multisetus
Acc:1132
PI 531603
PI 531606
PI 619466
42
50
47
33
5.29
6.33
4.89
4.69
6.58
6.10
7.30
7.25
4.03
4.24
4.25
3.63
9.62
8.64
8.15
8.15
mean
43a
5.30a
6.81 a
4.04a
8.64a
#taxameans followed by differentlettersin the same column are significantlydifferentat P < 0.05.
were obtainedfrom USDA-ARS,i.e., W6
22031 (David C. Ganskopp,Bums, Ore.)
and W6 20963 (James A. Young, Reno,
Nev.). The 4 remaining accessions (PI
531603 throughPI 532606),obtainedfrom
theNationalPlantGermplasm
System,were
originally from the collection of the late
D.R. Dewey of our laboratory.We determinedthe taxonof eachaccessionusingthe
dichotomous key provided by Wilson
(1963) thatdistinguishesthe taxabasedon
numberandbifurcationof awns.Seed samples of theseaccessionsaredepositedin the
National Plant GermplasmSystem at the
USDA-ARS Western Regional Plant
IntroductionStation in Pullman, Wash.
(http://www.ars-grin.gov/npgs).
Data were collectedfrom evaluationsin
the field and4 greenhousetrials.The field
evaluation was established on 2-3 Sept.
1993 at GreenvilleFarm,NorthLogan,Ut.
(41045'59" N, I I 1148"34"W) on a
Millville silt loam (coarse-silty,carbonatic, mesic Typic Rendolls; 2-4% slope).
476
Accessionswere establishedas field transplants on 0.5-m centers in a 7 x 7 plant
block. Each block was filled with 49
plants unless fewer were available.
Because the blocks were not replicated,
datacollectedfromthis trialwere only for
2 traits of high heritability,which were
headingdate (the numberof days after30
Apr. 1995 thatthe majorityof the plantsin
a plot had at least 1 spikeemergefromthe
boot) and seed mass (mg seed-' averaged
across 1994 and 1995harvests). Seed harvested fromthese blocks was used to conduct a seedling growthtrial in the greenhouse, 2 depth-of-seeding trials, and a
nitrate reductase activity assay in the
greenhouse.Becauseoutcrossingin squirreltail is negligible (Jensen et al. 1990),
seed harvestedfrom such a block may be
presumedto be geneticallyrepresentative
of the originalcollection.
The seedlinggrowthtrialwas plantedin
a greenhousemaintainedat 20 ? 5C on 8
Jan. 1996 in a soil consisting of 3 parts
Kidmanfine sandy loam (coarse-loamy,
mixed, mesic Calcic Haploxerolls):1 part
peat moss. Five seeds of an accession
were planted5 mm deep in 65 X 255-mm
cone-containers. Accessions were
assigned to 10 replicationsin a randomized complete block design. The third
seedlingto emerge was retained,days-toemergence was recorded, and the other
seedlings were removed.Waterwas supplied daily with a 32% Rorison nutrient
solution (Arredondoet al. 1998). Length
of the first leaf was measuredon day 10.
Plants were harvested on day 25 and
divided into shoots and roots. Leaf area
was determinedwitha leaf-areameter(LI3000, LI-COR, Lincoln, Neb.). Roots
were recoveredby gentle washingwith a
slow-movingstreamof water.Root length
was measured on fresh tissue with
'Branching' software (version 1.52,B)
(Berntson 1992). Shoots and roots were
then oven-dried at 70?C for 48 hours
before weighing. Specific leaf area was
calculated as leaf area/shoot mass, and
specific root lengthwas calculatedas root
length/rootmass.
To measure seedling emergence, 2
greenhousetrials were conductedat 20 ?
50
C at depths of 20 mm and 60 mm. Both
trials were plantedin above-groundbeds
filled with the greenhousesoil described
above. This was underlain by a gravel
layerfor drainage.The beds were 1.067-m
wide and furrows were 50.8-mm apart.
Fifty seeds of an accession were planted
per half-length of each furrow. Another
randomizedaccession was plantedin the
oppositehalf. Accessions were plantedin
a randomizedcompleteblock design with
6 replications.The 20-mmtrialwas planted on 9 to 10 Dec. 1996 and seedling
emergence was tallied daily. Emergence
index was calculatedas Y[(gn-9(n-1))/n],
where (gn-g(n-)) is the number of
seedlings germinatingsince the previous
day and n is the numberof days since the
germinationtrial began (Maguire 1962).
Therefore, large numbers indicate fast
emergence and small numbers indicate
slow emergence,the oppositedirectionof
the days-to-emergence variable in the
above-mentioned seedling growth trial.
The 60-mm trial was planted on 4 Feb.
1997 andconductedsimilarly.
High nitrate reductase activity may
facilitatehigherassimilationof soil nitrogen, therebypromotingseedlingestablishment. For the nitratereductaseassay, the
27 accessions were grown in a 20 ? 5?C
greenhouseas seedlingsin the samesoil as
the depth-of-seeding evaluations. Fresh
samplesof leaf tissue were harvested2 to
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
4 hours after sunrise,placed directlyinto
liquid N2, and stored at -80?C until use.
Frozentissue was powderedin liquid N2
andextractswere madeby grindingin 5 to
10 vol. of 50 mM MOPSbuffer(pH 7.5)
containing 10 mM MgC12,1mM EDTA,
0.1% TritonX-100, 2.5 mM DTT, 1 mM
PMSF, 10 pM leupeptin,and 10 pM chymostatin. After centrifugation of the
extractsat 12,000 g for 4 min, the supernatantwas transferredto a fresh tube and
used immediately to assay total
NADH:nitrate
reductaseactivityaccording
to Redinbaughet al. (1996). Assays were
performedin a 200-pl volume containing
50 mM MOPS, 10 mM K-phosphate
buffer(pH 7.5) containing13 mM KNO3,
500 pM NADH, and 10 pM FAD. Soluble
protein in the extracts was determined
using a dye-binding assay (Redinbaugh
andCampbell1985).
of the plantsin a plot had at least 1 spike
emerge from the boot), seed mass (the
mean mg seed-' across 1997/1998), and
plant height (mean across 1997/1998)
were measuredon these plots. Seed harvestedfromthese blocks was used to conduct a seedling growthtrial in the greenhouse as for dataset 1.
The seedlinggrowthtrialwas plantedin
a greenhousemaintainedat 20 ? 5?C on
12 Mar. 1999 in the same design and soil
as used in data set 1. Waterwas supplied
without nutrientsolution. Days-to-emergence of the third seedling was recorded
and its leaf length was measured17 days
after planting.The plants were harvested
14 Apr. (replications 1 to 3), 15 Apr.
(replications4 to 8), and 16 Apr. (replications 9 to 10) and dividedinto shoots and
roots. For data set 2, we used a Comair
root length scanner (Commonwealth
AircraftCorp.Ltd., Melbourne,Australia)
to measureroot lengthof fresh tissue that
had been stored in refrigeratedvials of
water.This greatlyexpeditedthe process
comparedto the techniqueused for data
set 1. After scanning, shoots and roots
were oven-dried at 70?C for 48 hours
before weighing.Specific root length was
calculatedas rootlength/rootmass.
ized to unit variancefor each traitbefore
multivariate analysis. Thirteen and 9
orthogonal principal components, the
numberof variablesin each dataset, were
generatedfor datasets 1 and2, respectively (Morrison1976). Because a large proportionof the total variancewas accounted for by the first 2 principalcomponents
for both data sets, only the first 2 were
retained. These 2 principal components
were each definedby a vector of scalars,
each representinga trait.These are termed
the principal component loadings for
Principal Component 1 and Principal
Component 2, respectively. Principal
Component1 loadingsarethe single set of
numbers,13 in the case of dataset 1 and9
in the case of dataset 2, thatbest circumscribes the data set. Likewise, the
Principal Component2 loadings are the
secondsingle set of numbersthatbest circumscribesthe variationremainingafter
Principal Component 1 variation is
Data set 2
removed.We rotatedprincipalcomponent
For the second data set, data were colloadings using an orthogonal varimax
lected froma field evaluationand a greenrotation
in PROC FACTOR (SAS
house seedling growth trial. Forty-seven
Institute, Cary, NC) to maximize the
accessions (Table 3) were establishedas
amountof informationincluded in the 2
field transplants on 3-4 Sept. 1996 at
retained
principal components (Kaiser
Evans Farm, Millville, Ut. (41?41'39"N
1958). Principalcomponentscores were
111?49'58"W)on a Nibley silty clay loam
calculatedfor each accession for each of
(fine, mixed mesic Aquic Argiustolls;0the 2 principalcomponentsby multiplying
3% slope). All 47 accessionswere collect- Statisticalanalysis
ed from their native sites in 1995. The
For each trait, differences between the vector of standardizedmeans by the
plantingdesign was similarto data set 1. meansof any 2 taxawere testedfor devia- vector of rotated principal component
Heading date (the mean numberof days tion from zero using a t-test at cc = 0.05. loadings. Accessions were plotted in 2
after 30 Apr. 1997/1998that the majority Values for all accessions were standard- dimensionsbasedon PrincipalComponent
1 and Principal Component 2 scores.
Because2 of the 13 traitsmeasuredin data
Table 3. Loadings for the first 2 principal components in data sets 1 (13 traits) and 2 (9 traits).
set 1 and 3 of the 9 traitsmeasuredin data
set 2 were taken from unreplicatedplots,
Principalcomponent
comparisonsweremadeonly betweentaxa
Trait
1
2
ratherthanbetweenindividualaccessions.
Data set 1
The individualaccessionsprovidedrepliDays to emergence
-0.563
-0.635
cation
for theirrespectivetaxa.
Leaf
length
Total plant dry-matter
Root-to-shootratio
Leaf area
Specific leaf area
Root length
Specific root length
Headingdate
Seed mass
Germinationindex (20 mm)
Germinationindex (60 mm)
Nitratereductaseactivity
0.818
0.968
0.437
0.903
0.055
0.924
-0.040
0.655
0.702
0.028
0.477
0.222
0.339
0.018
0.526
0.227
0.854
0.059
-0.517
-0.552
-0.593
0.791
0.163
0.749
Data set 2
Days to emergence
Leaf length
Total plantdry-matter
Root-to-shootratio
Root length
Specific root length
Headingdate
Plantheight
Seed mass
0.353
0.082
0.512
-0.754
0.441
0.302
0.912
0.905
0.902
-0.882
0.921
0.824
0.395
0.788
-0.654
0.008
0.001
0.291
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
Results and Discussion
Data set 1
In the 25-day greenhousetrial,multisetus accessionsemergedfaster,had longer
leaves at 10 days, and had greatertotal
plantdry-matterat harvestthanthe other2
taxa (Table 1). The fasterseedling development of multisetus may relate to its
prevalencein regionswithrelativelywarm
springslike the ColumbiaandSnakeRiver
Plains. Brevifolius emerged slowest,
which may be because of the cooler
springs and late freezes of the Rocky
Mountains,a regionwherethis taxonpredominates.Elymoideshad the least total
plant dry-matterof the 3 taxa, probably
477
because it is most common in the semiarid cold desert, a stressful environment
where conservative growth is probably
advantageous. Root-to-shoot ratio was
least for brevifoliusand.greatestfor elymoides. Evaporativedemand,necessitating a greaterroot-to-shootratio,is typically greater in the semi-arid cold desert,
whereelymoidespredominates,thanin the
Rocky Mountains,where brevifoliuspredominates. Brevifolius and elymoides were
similar for leaf area and root length and
lower than multisetus, which again may
relateto the latter'sprevalencein warmer
regions.Brevifoliushadlower specificleaf
area, i.e., thickerleaves, than the other 2
taxa, but no differencesamongtaxa were
observed for specific root length.
Likewise, Arredondoet al. (1998) found
that 4 brevifolius accessions had lower
specific leaf areathanthe multisetus,Sand
Hollow.Low specific leaf areais associated with high carboninvestmentand long
leaf life-span(Reich 1997), a characteristic of perennials relative to cheatgrass
(Arredondoet al. 1998). We are uncertain
as to whether brevifolius plants have
greater longevity in their environments
than do multisetus or elymoides plants.
especially at low (<10 mM) external
nitrateconcentrations(Johnsonet al. 1976,
Feil et al. 1993). In addition, barley
mutantswith low nitratereductaseactivity
grew slower in the cool early-seasonand
matured later than wild-type plants
(BlackwoodandHallam,1979). Measured
nitratereductaseactivityis the resultof a
complex interactionof nitrate, carbohydrateavailability,plant development,and
light. But plant materials with higher
nitratereductaseactivity, e.g., elymoides
and multisetus,that can more effectively
assimilate and compete for soil nitrogen
may grow or become established faster
than those with lower nitrate reductase
activity,e.g., brevifolius.Indeed,the slow
emergence of brevifolius from a 5-mm
planting depth relative to elymoides and
multisetuswas notedabove.Principalcomponentloadingsfor days to emergenceand
nitratereductaseactivity were negatively
correlatedfor both principalcomponents1
and2, but the relationshipwas strongerfor
principalcomponent2 (Table3).
Principal Components 1 and 2
explained 38 and 29% of the variation,
respectively,amongaccessionsfor the 13
traits.Loadingsfor PrincipalComponent1
indicate that accessions with higher
PrincipalComponent1 scores had generally more rapid emergence at a 5-mm
plantingdepth,greaterday-10 leaf length,
and greatertotal plantdry-matter,root-toshoot ratio, leaf area, and root length at
harvestin the greenhousetrial;laterheading dateandgreaterseed mass in the field;
and more rapidemergencefrom a 60-mm
planting depth (Table 3). Principal
Component 1 separatedmultisetus(high
scores) from elymoides (low scores).
Brevifolius was intermediate between
these 2, but closer to elymoides(Fig. 1).
Loadingsfor PrincipalComponent2 indicate that accessionswith higherPrincipal
Component2 scores had generally more
rapid emergence at a 5-mm planting
depth;greaterroot-to-shootratio, greater
specific leaf area (thinner leaves), and
lowerspecificrootlength(thickerroots)at
harvest in the greenhouse trial; earlier
heading date and lower seed mass in the
field; more rapid emergence from a 20mm planting depth; and higher nitrate
reductaseactivity.PrincipalComponent2
separated elymoides and multisetus (high
scores) from brevifolius (low scores).
Overlapamong the 3 taxa was absent in
the 2-dimensionalscatterplot, but beyond
that,accessionsdid not appearto be organizedin anygeographicalmanner.
Values for the lone californicusaccession, PI 531604, were withinthe rangeof
elymoidesaccessions for all traitsexcept
total plant dry-matterin the greenhouse
and seed mass in the field, where PI
531604 had slightlygreatervalues (Tables
1, 2). Accession PI 531604 was clearly
more similar to elymoidesthan to brevi-
In the field trial,elymoidesheadedearlier and produced seed with lower mass
than the other taxa (Table 2), suggesting
that it is adaptedto regions lacking the
summerprecipitationthat is found in the
Rocky Mountains, for example. In the
folius or multisetus, supporting a close taxgreenhouse depth-of-emergence trials,
multisetus emerged significantly faster
than brevifolius at both 20 and 60-mm
depths. The 25-day greenhousetrial also
demonstratedthe slow germinationrateof
01
x
x
brevifolius, based on emergence from a
x
.
W 0r
shallow 5-mm depth. Elymoides emergence was similarto multisetusand faster
U -2
0
than brevifolius at 20 mm, but lagged
j
E
behindwith brevifoliusat 60 mm. The low
0
seed mass of elymoides accessions may
0.
havebeen a disadvantageat the deeper60mm planting depth, while the high seed
.E -3
mass of brevifoliusmayhave compensated
for its inherentlyslow germinabilityat this
-4
depth. Correlations between seed mass
4
2
3
0
1
-1
-2
and emergence index were negative (r =
-0.50; P < 0.01) at 20 mm and positive (r
Principal component I score
= 0.34; P < 0.10) at 60 mm. Thus, largerseeded accessions emerged faster at 60
mm, while the reversewas trueat 20 mm.
Emergenceindices at the 2 depths were
uncorrelated
acrossaccessions(P > 0.10).
Nitratereductaseactivities were lower
for brevifoliusthanthe other2 taxa (Table
A
2). In agronomiccrops, high leaf nitrate
reductase activity was correlated with Fig. 1. Plot of the first 2 principal componentscores for 27 Elymuselymoidesand E. multiseincreased plant growth and grain yield,
tus squirreltailaccessionsin data set 1.
* ssp. californicus* ssp. elymoides
ssp. brevifolius x E. multisetus
478
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
Table 4 Means for 10 elymoides, 21 brevifolius, and 16 multisetus squirreltail accessions in data set 2 for 6 traits in a 34-day greenhouse trial.
Taxon/ Accession
Collectionlocation
Elevation
(m)
elymoides
T-el 173
T- 1174
T-1 175
T-1191
T-1 192
T-1193
T-1 198
T-1223
T-1224
PI 619553
mean
brevifolius
groupA
T-1180
T-1233
T-1238
T-1239
T-1242
T-1249
T-1260
T-1264
T-1265
T-1271
T-1272
T-1277
T-1299
(mean)
groupB
T-1228
T-1243
T-1308
(mean)
groupC
T-1202
T-1203
T-1204
T-1205
T-1206
(mean)
mean
multisetus
PI 619454
PI 619456
PI 619457
PI 619458
PI 619459
PI 619460
PI 619461
PI 619462
PI 619463
PI 619464
PI 619465
PI619467
PI 619564
T-1165
T-1183
T-1214
Leaf
length (17
Root
Days to
days post- Total plant Root-toSpecific root
length
plant)
emergence
dry-matter shootratio length
(no.)
(mm)
(mg plant-) (mg mg') (mmplant-) (mmmg-')
east of MountainHome, Ida.
west of MountainHome, Ida.
Ditto Creek,ElmoreCo., Ida.
CountyRd lO, Moffat Co., Colo.
Point of Rocks, SweetwaterCo., Wyo.
Superior,Wyo.
btwn Sage Junction& Laketown,Ut.
Fish Creek,Carey, Ida.
Atomic City, Bingham Co., Ida.
Shoshone, Ida.
1050
950
1000
2150
2200
2200
2200
1450
1500
1300
1633
5.7
4.4
4.6
5.0
4.7
5.9
5.0
4.3
4.5
4.8
4.9b#
38
38
42
41
47
37
42
49
49
38
42b
4.09
5.18
4.58
5.36
5.69
4.04
4.47
5.94
5.26
5.29
4.99c
northeastof Wagonmound,N.M.
northof Hermosa,Colo.
east of ChimneyRock, Colo.
northeastof Pagosa Springs,Colo.
northof Lake City, Colo.
west of Sargents,Colo.
south of Westcliffe, Colo.
northof Colmor,N.M.
northof Wagonmound,N.M.
southwestof Ocate, N.M.
La Cueva, N.M.
Tres Piedras,N.M.
northof Flagstaff,Ariz.
2000
2000
2050
2350
2450
2600
2450
1850
1900
2300
2200
2350
2150
(2204)
7.4
5.8
5.8
6.4
6.4
6.8
6.6
7.0
7.1
6.3
6.4
6.6
6.2
(6.5)
30
44
44
39
37
35
42
35
37
45
43
44
47
(40)
5.52
9.53
8.89
7.33
6.16
7.68
6.61
5.33
5.56
6.86
5.46
5.70
8.44
(6.85)
Colton, Ut.
northof Powderhorn,Colo.
northwestof Almont, Colo.
2150
2750
2500
(2467)
6.5
6.3
5.7
(6.2)
37
39
39
(38)
hwy 75 X 20, Blaine Co., Ida.
east of Fairfield,Ida.
east of Hill City, Ida.
west of Hill City, Ida.
east of Dixie, ElmoreCo., Ida.
1350
1500
1550
1600
1600
(1520)
2100
5.4
5.4
5.8
4.2
5.0
(5.2)
6.1 a
950
1000
800
1100
1000
1000
950
1100
1000
800
1000
800
1100
850
1000
1100
3.9
4.1
3.8
4.0
3.9
4.3
3.9
3.7
3.9
3.8
4.4
4.3
4.9
4.3
4.2
4.2
A-line canal, Gem Co., Ida.
CartwrightRd., Ada Co., Ida.
Little Land & Livestock, CanyonCo., Ida.
northwestof Mayfield, ElmoreCo., Ida.
Boise, Ida.
Bogus Basin Rd., Ada Co., Ida.
Barber,Ida.
southeastof Mayfield, ElmoreCo., Ida.
Ditto Creek,ElmoreCo., Ida.
Lower Hatley, CanyonCo., Ida.
Seaman'sGulch Rd., Ada Co., Ida.
Middle Hatley, Gem Co., Ida.
Ditto Creek,ElmoreCo., Ida.
King Hill, Ida.
MountainHome, Ida.
Black's Creek,Ada Co., Ida.
mean
950
4.1 c
#taxameansfollowedby differentlettersin thesamecolumnn
aresignificantly
different
atP < 0.05.
JOURNALOF RANGEMANAGEMENT
56(5) September2003
640
860
720
870
1090
730
840
870
770
780
817c
278
297
313
302
329
289
336
281
255
270
295b
0.84
0.61
0.50
0.85
0.88
0.65
0.76
0.74
0.82
0.81
0.92
0.91
0.76
(0.77)
710
1210
970
1160
1060
1080
1060
800
980
1050
850
1030
1500
(1035)
282
335
327
345
367
356
372
354
392
342
326
380
411
(353)
3.70
5.10
4.37
(4.39)
1.31
1.07
1.07
(1.15)
910
1070
980
(987)
433
405
434
(424)
42
52
43
44
45
(45)
41 b
4.71
6.28
6.55
5.92
5.92
(5.88)
6.27 b
1.26
0.84
1.17
1.26
1.18
(1.14)
0.91 b
720
1060
1010
1080
1080
(990)
1018 b
274
369
286
327
336
(319)
355 a
61
49
64
51
55
53
51
58
55
56
53
52
44
63
65
49
10.49
9.00
9.91
9.92
9.21
9.05
9.24
8.74
9.17
9.61
9.87
8.25
7.38
10.87
10.26
7.99
55a
9.31 a
1.28
1.27
1.01
1.16
1.39
1.68
1.27
1.09
1.35
1.20
1.27 a
1.14
1.17
1.22
1.31
1.20
1.16
1.16
1.39
1.34
0.99
1.23
1.07
1.04
1.25
0.99
1.04
1660
1560
1360
1470
1530
1510
1390
1360
1490
1280
1510
1160
1030
1610
1380
1200
297
321
250
262
305
311
280
267
284
268
278
272
273
267
270
294
1.17a
1406a
281 b
479
onomic relationshipbetween californicus traits.Loadingsfor PrincipalComponent1
and elymoides, as favored by Wilson indicatethataccessionswithhigh Principal
Component1 scores had generallygreater
(1963).
total plant dry-matter,lower root-to-shoot
ratio,and greaterroot length at harvestin
Data set 2
In the greenhouse trial, multisetus the greenhouse trial; and later heading
emerged faster, had longer leaves at 17 date,greaterheight,and greaterseed mass
days, and had greatertotal plant dry-mat- in the field (Table3). PrincipalComponent
ter at harvestthanthe other2 taxa (Table 1 separated brevifolius and multisetus
4). Brevifoliusemergedmost slowly from (high scores)from elymoides(low scores)
a 5-mm plantingdepthand elymoideshad (Fig. 2). Loadings for Principal
the least total plant dry-matter.Root-to- Component2 indicatethataccessionswith
shoot ratio was least for brevifolius and higherPrincipalComponent2 scores had
generallymorerapidemergenceat a 5-mm
greater for multisetus and elymoides.
Specific root lengthwas greaterfor brevi- plantingdepth;greaterleaf length on day
folius (thinroots),especiallygroupB (see 17 and greater total plant dry-matter,
below), than for multisetus or elymoides. greaterrootlength,andlower specificroot
Multisetus had the greatest root length length (thicker roots) at harvest in the
greenhousetrial. PrincipalComponent2
with brevifolius intermediate and elymoides least. In the field trial (Table 5), separatedmultisetus (high scores) from
elymoides had the earliest heading date, elymoides and brevifolius (low scores).
the lowest seed mass, and the shortest This separationresultedin plots that oriheight. Brevifolius and multisetus did not entedthe 3 taxain a similarmannerto data
set 1 (Fig. 1). Variationamongaccessions
differfromone anotherfor thesetraits.
The first 2 principal components was greaterfor brevifoliusthanthe other2
explained 41 and 40% of the variation, taxa;this was reflectedmostlyby Principal
respectively, among accessions for the 9 Component1 scores.
Together,PrincipalComponents1 and2
separated brevifolius accessions into 3
groups (Fig. 2). "GroupA" accessions
were late-maturingwith high-seed mass
and originatedfrom mediumto high elevations (1850 to 2600 m) in Colorado,
New Mexico, and Arizona. "GroupB"
accessions were early-maturing(as early
as elymoides) with low-seed mass and
originatedfrom high elevations (2150 to
2750 m) in ColoradoandUtah."GroupC"
accessions were intermediate-maturing
with low-seed mass and originatedfrom
low to mediumelevations (1350 to 1600
m) in southernIdaho.Amplifiedfragment
length polymorphisms (AFLP), a DNA
markertechnique,have verifiedthatthese
3 groupsaregeneticallydistinct(Larsonet
al. 2003).
In the greenhouse (Table 4), days to
emergence were generally lesser for
GroupC accessions and more similar to
elymoidesthanwere GroupA or B accessions. Totalplantdry-matterwas generally
greatestfor GroupA accessions,least for
GroupB accessions,and intermediatefor
GroupC accessions. Root-to-shootratio
Table 5. Means for 10 elymoides, 21 brevifolius, and 16 multisetus squirreltail accessions in data set 2 for 6 traits in a 2-year field trial.
Taxon/
Accession
Heading
date
Plant
height
Seed
mass
(days after4/30)
(m)
(mg seed-')
elymoides
T-1173
T- 1174
T-1175
T-1191
T-1192
T-1193
T-1198
T-1223
T-1224
PI 619553
21
21
22
18
20
15
20
26
24
23
0.28
0.32
0.31
0.28
0.27
0.38
0.48
0.42
0.42
0.38
2.21
2.37
2.16
3.29
2.83
2.36
2.91
3.01
2.73
3.05
mean
21 b
0.35 b
2.69 b
multisetus
PI 619454
PI 619456
PI 619457
PI 619458
PI 619459
PI 619460
PI 619461
PI 619462
PI 619463
PI 619464
PI 619465
PI 619467
PI 619564
T-1165
T-1 183
T-1214
44
42
42
42
41
42
41
45
45
44
42
42
42
33
41
41
0.53
0.58
0.58
0.68
0.53
0.58
0.64
0.65
0.57
0.63
0.61
0.56
0.66
0.53
0.62
0.60
4.88
5.50
4.72
5.83
5.05
5.01
5.18
5.07
4.72
4.87
5.14
4.62
4.87
4.15
5.39
5.49
mean
42 a
0.59 a
5.03 a
Taxon/
Accession
Heading
date
Plant
height
Seed
mass
(days after4/30)
brevifolius
groupA
T- 1180
60
T-1233
62
T-1238
47
T-1239
49
T-1242
42
T-1249
46
T-1260
51
T-1264
65
T-1265
60
T-1271
57
T-1272
57
T-1277
53
T-1299
63
(mean)
(55)
(m)
(mg seed-')
0.66
0.64
0.75
0.75
0.77
0.68
0.53
0.71
0.71
0.72
0.72
0.63
0.66
(0.68)
4.81
4.51
5.10
5.34
5.36
5.71
4.58
4.86
5.02
5.23
5.48
4.51
6.16
(5.13)
groupB
T-1228
T-1243
T-1245
(mean)
23
24
17
(21)
0.51
0.53
0.58
(0.54)
3.39
3.77
3.37
(3.51)
groupC
T-1202
T-1203
T-1204
T-1205
T-1206
(mean)
34
32
34
38
38
(35)
0.50
0.54
0.58
0.62
0.60
(0.57)
3.75
3.74
4.61
3.47
3.27
(3.77)
mean
45a
0.64a
4.57a
#taxa means followed by differentlettersin the same column are significantlydifferentat P < 0.05.
480
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
data set 2 included plant height. With 1
exception (root-to-shootratio) the 2 data
sets rankedthe taxa in the same orderfor
all 8 common traits. However, for 3 of
these traits (root-to-shoot ratio, root
length, and specific root length), significant differencesamongtaxonmeanswere
slightly different.For root-to-shootratio,
data set 1 grouped middle-rankingely-
0
N1
x
*1
c
S
-2
moides with lowest-ranking breviolius,
while data set 2 groupedhighest-ranking
-3
0.~.
-2
elymoides with middle-ranking multisetus.
0X
-1
Principal component I score
Forrootlength,dataset 1 groupedlowestranking elymoides with middle-ranking
brevifolius,while data set 2 groupedlowest-rankingelymoidesapartfrom middleranking brevifolius. For specific root
length, data set 1 grouped all 3 taxa
together,while data set 2 separatedhighest-ranking brevifolius apart from elymoides and multisetus.
* ssp. brevifolius(A) * ssp. elymoides
A ssp. brevifolius(B) x E. multisetus
z Ssp.brevifolius
(C)
Fig. 2. Plot of the first 2 principal componentscores for 47 Elymuselymoidesand E. multisetus squirreltailaccessionsin data set 2.
was generally lower for GroupA accessions thanGroupB or C accessions,which
were more similarto elymoidesandmultisetus. Specific root length was generally
greaterfor GroupB accessionsthanGroup
A or C accessions,which were moresimilar to elymoides and multisetus. In the
field trial(Table5), seed mass was generally greaterfor GroupA accessions,which
were more similar to multisetus, than
Group B or C accessions, which were
more similar to elymoides. Plant height
was generallygreaterfor GroupA accessions thanGroupB or C accessions,which
were more similarto multisetus,which in
turnwas greaterthanelymoides.
While the geographicalrange of Group
C is clearly separatefrom GroupsA and
B, GroupsA and B themselvesare overlappingand may occur as biotypesat the
same site. Accession T-1245, a group B
accession originating northwest of
Almont,Colo., was collected at the same
site as T-1308, which was not includedin
data set 2, but has been confirmed as a
GroupA accessionbasedon its AFLPprofile (Larson et al. 2003). Likewise, T1264, a Group A accession originating
north of Colmor, N.M., was collected
from the same site as T-1309, also not
includedin data set 2. Based on its early
maturity(unpublisheddata), however, T1309 is likely a GroupB accession.
Relativeto GroupA, GroupB is earlier
in maturity, is shorter, and has greater
root-to-shoot ratio, greater specific root
length (thinner roots), lower total plant
dry-matter,andlowerseed mass.We wonderwhethersome or all of these traitsmay
be associated with adaptationto higher
elevations. In fact, T-1243 (Powderhorn,
Colorado),a GroupB accession,was collected at 2,750 m, the highestelevationof
any accession. While our elevation data
(Table4) do not directlysupportthis speculation, we may have missed a trend by
failing to collect at higherelevationsthan
we did.
Conclusions
In both data sets, the 3 squirreltailtaxa
were easily separatedby the traitsexamined. Dataset 1 had 13 traitsanddataset 2
had 9 traits. The 8 traits in common to
both data sets were days to emergence
from5 mm;leaf length(at 10 days in data
set 1 and 17 days in dataset 2), totalplant
dry-matter,
root-to-shootratio,rootlength,
and specific root length at harvest;heading date;and seed mass. In additionto the
8 traits in common, data set 1 included
leaf area,specific leaf area,days to emergence from 20 mm, days to emergence
from 60 mm, and nitratereductaseactivity. In additionto the 8 traitsin common,
JOURNAL OF RANGE MANAGEMENT56(5) September 2003
Of all taxa, elymoideswas the earliest,
shortest, and had the lowest seed mass,
seedling dry-matter, and root length.
Brevifolius had the lowest root-to-shoot
ratio,specific leaf area,and nitratereductase activity, the highest specific root
length, and generally the slowest emergence from a normalplantingdepth.Low
specific leaf area is associatedwith high
carboninvestmentand greatleaf life-span
(Reich 1997). Multisetushad the fastest
emergencefrom a normalplantingdepth,
longest leaf length at 10 days (dataset 1)
or 17 days (data set 2) post-plant, and
greatestseedlingdry-matter,leaf area,and
rootlength.
Wilson (1963) stated that brevifolius
displays the greatestecological diversity
of the 5 squirreltailtaxa, being found in
desert to montane habitats from 600 to
3000 m elevationand with extremevariation in size. Variationamong accessions
was similarfor the 3 taxain dataset 1, but
brevifolius had the greatest variation in
data set 2, perhapsbecause these brevifolius accessionsrepresenteda widergeographical distribution than the other 2
taxa. Nevertheless, we found a greater
degreeof discontinuity,e.g., GroupsA, B,
and C, withinbrevifoliusthanwithinmultisetus or elymoides. In data set 2, acces-
sions thatplottednearthe convergenceof
the 3 taxa were from the environs of
southemIdaho,a regionwith amplerepresentationof all 3 taxa. Overall,GroupC
accessionswere moresimilarto elymoides
and multisetusthan were Group A or B
accessions(Fig. 2).
Five brevifoliusaccessionsin dataset 1
can be reinterpretedbased on findingsof
data set 2 and subsequentAFLP analysis
(Larsonet al. 2003). Threelate-maturing,
481
high seed-mass accessions, 9040189,
904187, and PI 531605, qualify as Group
A, according to AFLP data. While no
AFLP data have been collected on the
fourthaccession, Acc:1123, it appearsto
be a member of Group C based on its
Oregon origin, fast emergence (Table 1)
and low seed mass (Table2). It shouldbe
noted, however, that its heading date is
similarto the 3 GroupA accessions.The
fifth accession, Acc:1130 (Savageton,
Wyo.), is unique in this study. Based on
AFLP results, it belongs to "GroupD",
accessionsof which we have collectedon
the HighPlainsfromAlbertato Colorado.
Ourdata supportthe recognitionof elymoides, brevifolius, and multisetus taxa,
but they do not provide direction as to
whethereach meritsspecific or subspecific rank. From an ecological perspective,
these data do not supportthe suggestion
that any one pair of elymoides, multisetus,
and brevifolius is more similar than any
otherpair.Furthermore,
these resultsprovide guidance as to which specific commerciallyavailableplant materialsshould
be plantedin which locations.For example, the 3 distinct groups of brevifolius
accessions,as well as the differenttaxa,are
ecologicallydistinctandare likely adapted
to regionsin whichtheirparticulargroupis
found in nature.These results reflect the
highlyecotypicnatureof the squirreltails.
482
LiteratureCited
Jensen, K.B., Y.F. Zhang, and D.R. Dewey.
1990. Mode of pollination of perennial
speciesof the Triticeaein relationto genomiArnow, L.A. 1993. GramineaeA.L. Juss. p.
cally defined genera. Can. J. Plant Sci.
775-833. In: S.L. Welsh, N.D. Atwood, S.
70:215-225.
Goodrich,and L.C. Higgins (eds.) A Utah Johnson, C.B., W.J. Whittington, and G.C.
flora. 2nd ed. BrighamYoung Univ., Provo,
Blackwood.1976. Nitratereductaseas a posUt.
sible predictive test of crop yield. Nature
Arredondo, J.T., T.A. Jones, and D.A.
262:133-134.
Johnson. 1998. Seedling growth of Kaiser, H.F. 1958. The varimaxcriterionfor
analytic rotation in factor analysis.
Intermountainperennialand weedy annual
Psychometrika
23:187-200.
grasses.J. RangeManage.51:584-589.
Berntson,G.M. 1992. A computerprogramfor Larson, S.R., T.A. Jones, C.L. McCracken,
and K.B. Jensen. 2003. Amplifiedfragment
characterizingroot systems branchingpatlength polymorphism in E. elymoides, E.
tern.PlantSoil 140:145-149.
multisetus, and other Elymus taxa. Can. J.
Barkworth, M.E., D.R. Dewey, and R.J.
Bot. 81 (in press).
Atkins. 1983. New generic concepts in the
Triticeaeof the IntermountainRegion: Key Maguire, J.D. 1962. Speed of germinationaid in selection and evaluationof seedling
and comments. Great Basin Natur.
emergenceandvigor.CropSci. 2:176-177.
43:561-572.
Morrison, D.F. 1976. Multivariatestatistical
Blackwood, G. and R. Hallam. 1979. Nitrate
methods. 2nd ed. McGraw-Hill Book Co.,
reductase activity in wheat (TriticumaesNew York,N.Y.
tivumL.) II. The correlationwith yield. New Redinbaugh, M.G. and W.H.
Campbell.
Phytol.82:417-425.
1985. Adaptationof the dye-bindingprotein
Feil, B., R. Thiraporn, and P. Stamp. 1993.
assay to microtiterplates. Anal. Biochem.
In vitronitratereductaseactivityof laborato147:144-147.
ry-grownseedlings as an indirect selection Redinbaugh, M.G., S.W. Ritchie, K.W.
criterionfor maize.CropSci. 33:1280-1286.
Hendrix, C.T. MacKown, and W.H.
Hitchcock,A.S. 1950. Manualof the grassesof
Campbell. 1996. Expression of enzyme
the United States. 2nd ed. (rev. A. Chase).
activitiesand transcriptsrequiredfor nitrate
USDA Misc. Publ.No. 200. U.S. Govt.Print.
assimilationin maize (Zea mays L. A188 X
Office.Washington,D.C.
BMS) suspension cell cultures. Plant Sci.
Holmgren, A.H. and N.H. Holmgren. 1977.
119:169-181.
Poaceae.p. 175-464. In: A. Cronquist,A.H. Reich, P.B. 1997. From tropics to tundra:
Global convergence in plant functioning.
Holmgren,N.H. Holmgren,J.L. Reveal, and
Proc.Nat.Acad.Sci. USA 94:13730-13734.
P.K. Holmgren (eds.) Intermountainflora.
Wilson, F.D. 1963. Revision of Sitanion
vol. 6. ColumbiaUniv. Press,New York.
(Triticeae,Gramineae).Brittonia15:303-323.
JOURNAL OF RANGE MANAGEMENT56(5) September 2003