Full text of "The Great Basin naturalist"

See other formats

HARVARD UNIVERSITY

Library of the

Museum of

Comparative Zoology

, MCZ
LIBRARY

H E ^^^ 0 7 15

1995

, . HARVARD
UNIVZ:RSITy

GREAT BASIN

NMRAUST

VOLUME 55 N2 1 _ JANUARY 1995

BRIGHAM YOUNG UNIVERSITY

GREAT BASIN NATURALIST

Editor Assistant Editor

Richard VV. Baumann Nathan M. Smith

290 MLBM 190 MLBM

PO Box 20200 PO Box 26879

Brigham Young University Brigham Young University

Provo, UT 84602-0200 ' Provo, UT 84602-6879

801-378-5053 801-378-6688

FAX 801-378-3733 E-mail: NMS@HBLL1.BYU.EDU

Associate Editors

Michael A. Bowkrs Paul C. Marsh

Blandy Experimental Farm, University of Center for Environmental Studies, Arizona

Virginia, Box 175, Boyce, VA 22620 State University, Tempe, AZ 85287

J. R. Callahan Stanley D. Smith

Museum of Southwestern Biology, University of Department of Biology

New Mexico, Albuquerque, NM University of Nevada-Las Vegas

Mailing address: Box 3140, Hemet, CA 92546 Las Vegas, NV 89154-4004

Jeffrey J. Johanskn Paul T. Tueller

Department of Biology, John Carroll University Department of Environmental Resource Sciences

University Heights, OH 441 18 University of Nevada-Reno, 1000 Vallev Road

Reno, NV 89512
Boris C. Kondratieff

Department of Entomology, Colorado State Robert C. Whitmore

University, Fort Collins, CO 80523 Division of Forestry, Box 6125, West Virginia

University, Morgantown, WV 26506-6125

Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoolog>';
William Hess, Botany and Range Science; H. Duane Smith, Zoology. All are at Brigham Young University.
Ex Officio Editorial Board members include Steven L. Taylor College of Biolog>- and Agriculture; Stanley L.
Welsh, Director, Monte L. Bean Life Science Museum; Richard W Baumann, Editor Great Basin Naturalist.

The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University.
Unpublished manuscripts that further our biological understanding of the Great Basin and sunounding areas
in western North America are accepted for publication.

Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1995 are $25 for individual sub-
scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back
issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi-
ness should be directed to the Editor Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young
University, Provo, UT 84602-0200.

Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin Naturalist
through a continuing exchange of scholarly publications should contact the Exchange Librarian, 6385
HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889.

Editorial Production Staff

JoAnne Abel Technical Editor

Jan Spencer Assistant to the Editor

Ofticial publication date: 16 January 1995 1-95 750 12935

The Great Basin Naturalist

Published at Provo, Utah, by
Brigham Young University'

ISSN 0017-3614

Volume 55 31 January 1995 No. 1

LIFE HISTORIES OF STONEFLIES (PLECOPTERA) IN THE
RIO CONEJOS OF SOUTHERN COLORADO

R. Edward DeWaltl'2 and Kenneth W. Stewartl

Abstract. — Thirty-one stonefly species representing eight famihes were collected during the March 1987 to May
1990 study period. Genera represented by more than one species included Capnia. Utacapnia, Taenionema. Siiuallia,
Triznaka, Isogenoides, and Isoperla. Peak species richness was recorded on or near the summer solstice in 1988 and
1989. Climatic differences between years were reflected in nymphal development and emergence phenology of most
species. New or important corroborative life histoiy data are presented for 1 1 stonefly species of this assemblage. The
hyporheic nymphal development of most chloroperlid species limited the number of early instars sampled and our
capacity' to inteipret voltinism. Limited nymphal data suggested a univoltine-slow cycle for Phimiperia diversa (Prison).
Adults of Suwallia pallidula (Banks) and S\ uardi (Banks) were present for an extended summer period, but the bulk of
their respective emergence times was temporally separated. Isogenoides zionensis Hanson, Pteronarcella hadia (Hagen),
and Pteronarcijs califomica Newport were all showTi for the first time to have a 9-10-mo egg diapause, and all three
species have a semivoltine life cycle. Skwcda americana (Klapalek) and Isoperla ftdva Claassen were further confimied to
have univoltine-slow cycles. Univoltine-fast and univoltine-slow life cycles are reported for the first time in /. phalerata
and /. qiiinquepunctata, respectively. Regression analysis revealed that si.x of the eight abimdant species had extended
emergence patterns (slopes of <5%/d), while only two had synchronous patterns. Warmer spring and summer tempera-
tures in 1989 increased the slopes for five of the eight species studied, but did not change their synchrony designation.
Nine of 11 abundant species advanced their median emergence date in 1989 over 1988. This and the higher slope values
are consistent with a hurried nymphal development and narrower emergence period due to the warmer thermal regime
of 1989.

Key words: Plecoptera, life history, biodiversity, life cycle. Rocky Mountains.

Stoneflies (Plecoptera) are one of the integral understood (Sheldon and Jewett 1967, Stewart
and often dominant insect orders in stream and Stark 1988). Precise life histories are
ecosystems; therefore, they are important as known for <5% of the more than 575 North
biological indicators, as fish food, and as part American species, and knowledge of stonefly
of the energy and nutrient economy of streams life histories and ecology in southern Rocky
(Stewart and Stark 1988). Taxonomy of the Mountain streams is sparse. This has limited
North American fauna is now well known; our ability to increase understanding of eco-
however, information on their life histories, logical relationships between cohabiting stone-
local species richness, and ecology is still poorly fly species in this region.

'Department of Biological Sciences. University' of North Te.\as, Denton, TX 76203.

^Present address; Department of Zoology and Physiology, Louisiana State University, Baton Rouge, LA 70S0.3.

Great Basin Naturalist

[Volume 55

One objective of this study was to deter-
mine richness of the stonefly assemblage of
the Rio Conejos of southern Colorado, a large
drainage that has not been previously studied.
Second, we documented the important life
history events of its dominant species for
which sufficient individuals and observations
could be gathered by intensive monthly sam-
pling and by living streamside during spring
and summer.

Research was patterned after the classic
studies of Harper (1973a, 1973b) and Harper
and Hynes (1972), who studied a substantial
portion of the eastern Canadian fauna and
addressed critical aspects of life histories such
as egg development, diapause, and adult
behaviors that are often overlooked. H. B. N.
Hynes, in an address to the International Ple-
coptera Symposium (1992), emphasized the
need for more attention to these aspects to
support the eventual development of a para-
digm of life history evolution within the Ple-
coptera. We have also adopted the approaches
of Knight and Gaufin (1966), Harper and
Magnin (1969), Sheldon (1972), Barton (1980),
Ernst and Stewart (1985a, 1985b), and
Hassage and Stewart (1990) in comparatively
studying an assemblage of species. This report
is the first to address, on a large scale, such an
assemblage in a western North American
stream since the works of Knight and Gaufin
(1966), Sheldon (1972), and Stanford (1975).

Methods

Study Stream

The Rio Conejos is located in the southern
Rocky Mountains of south central Colorado.
The river flows east to west for 145 km from
its headwaters in the Rio Grande National
Forest of the San Juan range to the Rio Grande
32 km northeast of Antonito, CO. Three sam-
pling sites were established along the Rio
Conejos to ensure access to at least one of
them during the winter and to enhance collec-
tion of stonefly species that were not abundant
at all sites. These were located at elevations be-
tween 2400 and 2600 m above sea level. The
primary site (106° 15 'W longitude, 37°03'N
latitude) consisted of a 1-km stretch located 24
km west of Antonito, Conejos County, CO, off
Colorado highway 17. Sites two and three
were located 22.5 km west of Antonito, also on

highway 17, and 4 km north of Antonito at the
Colorado highway 285 bridge, respectively.

Stream temperatures varied from below
freezing during the winter months to near
20 °C in August. Ice cover was common from
December through March. Snowmelt began
in April, usually leading to peak flows in June.
Base flows were attained by late August and
continued through the winter. Water released
from Platoro Reservoir, 48 km upstream, aug-
mented river flow during summer low-flow
periods. Bottom substrates were characterized
by large boulders, cobble, gravel, and sand.
These were covered by a thin layer of silt in
quiet water. Important organic substrates
included the flooded coppices of willows and
cotton woods and their entrained leaf packs.
Willow {Salix spp.), cottonwoods and aspens
{Popidus spp.), and alder {Alnus sp.) con-
tributed to the riparian corridor.

Physical Conditions

Stream temperature was monitored at site
one from June through August 1988 using a
Ryan™ continuous recording thermograph.
High, low, and mean daily stream tempera-
tures were calculated from temperatures
recorded at 0400, 0800, 1200, 1600, 2000, and
2400 h. Water temperatures were not record-
ed during 1989 due to equipment failure.
However, summer air temperature highs and
lows and rainfall were recorded (1300 h daily,
mountain time) for both 1988 and 1989 at the
Conejos Peak U.S. Weather Service reporting
station at site one. Flow data for site two were
gathered from Petsch (1987-90).

Nymphal Growth

Nymphs were collected monthly (except
December due to poor weather conditions) at
all sites from March 1987 to May 1988.
Additional collections were made at irregular
intei-vals until March 1990. Samples were col-
lected by disturbing the substrate (mineral
and organic) upstream of a BioQuip rectangu-
lar dipnet until debris clogged the net. The
net was composed of a coarse, 1-mm mesh
first stage, modified by the addition of a coni-
cal second stage of 153-/xm mesh size. The lat-
ter collected even the smallest instars. A
plankton bucket was attached to the second
stage to facilitate sample removal. Contents of
the plankton bucket and the coarse stage con-
stituted a sampling unit and were stored in

1995]

Stonefly Life Histories

70% isopropyl alcohol. The number of sam-
pling units per month varied with the effort
necessaiy to secure approximately 50 nymphs
of all abundant species.

Nymphs were separated from sample
debris with the aid of 4-lOX magnification on
a stereo-dissecting microscope, sorted to
species when possible, and stored in 80%
ethanol until measurement. Head capsule
width (HCW, greatest distance across the
eyes) was measured with a calibrated ocular
micrometer fitted to a stereo-dissection micro-
scope. Nymphs from all sites for the 3-yr sam-
pling period were pooled by species and
month of collection to increase the number of
nymphs per month and to allow construction
of more robust growth histograms. Gender of
nymphs was assessed by a gap in the posterior
setal margin of the eighth sternum of females
(Stewart and Stark 1988) and by developing
external genitalia of females. Sex-specific kite
diagrams were constructed by placing male
and female nymphs into 0.1- or 0.2-mm size
classes. The frequency of these classes was
converted to a percentage of the total number
of nymphs (males + females + unsexed
nymphs) collected for that month. Polygons
were constructed for each month depicting
the relative proportion of all nymphs at that
size class.

Adult Emergence

Adults of winter- and early spring-emerg-
ing stoneflies were collected from bridge abut-
ments, from shoreline debris, and under the
cobble at streamside to provide a general
emergence period for each species. Adults
were also reared from preemergent nymphs.

A combination of sampling methods and
observational procedures was used during the
summers of 1988 and 1989 to evaluate emer-
gence, duration of adult presence, and behav-
ior of these species. Adult traps and methods
included a 2.25-m^ basal area BioQuip malaise
trap, two 0.25-m2 basal area floating emer-
gence traps, pitfall traps, sweepnetting of
streamside vegetation, exuviae collection, and
day and night transect walks. Pitfall traps were
emptied on alternate days, and the others
were emptied daily between 0900 and 1100 h.
All of these methods were used at site one;
sweepnetting was employed at site three on
several occasions.

The malaise trap was deployed among wil-
low and Cottonwood coppices, where its olive-
drab coloration mimicked the surrounding
vegetation. Flying, or crawling, adults inter-
cepted by the trap ascended the screening
into a dry apical collection chamber Addition-
ally, all adults on the trap mesh were collected
using an aspirator

Emergence traps were anchored over shal-
low riffles during the 1988 field season.
Natural diurnal changes in water level and
erratic discharges due to water release from
Platoro Reservoir rendered these ineffective
at times; therefore, their use was discontinued
in 1989.

Pitfall traps consisted of 28.3-cm2 modified
aluminum soda cans that were buried flush in
streamside substrates. A mixture of 70%
ethanol and ethylene glycol (the latter to
retard evaporation) was used as a preservative.
In 1988, 12 traps were installed 1 m from the
stream at 1-m intervals on an open beach with
nearby vegetation. This was expanded in 1989
to three transects, each consisting of 30 cans
set 1 m apart in transects 1 m, 5 m, and 8 m
from the initial shoreline. These traps moni-
tored not only adult presence of ground-tra-
versing, brachypterous stoneflies, but also
their potential to move laterally from the
stream.

Sweepnetting was conducted over a 15 x 2-
m willow and cotton wood riparian zone. The
entire area was methodically swept, working
from the base of each clump of vegetation up-
ward. Exuviae removal was the only method
used to assess emergence of Claassenia sabu-
losa (Banks) and was used for no other species.
In 1988 exuviae were removed daily from the
same 15 X 1-m area of cobble shoreline, and
the frequency of each sex was noted. In 1989
the removal area was expanded to 30 x 1 m of
shoreline area and up to 5 m into the water for
collecting exuviae from emergent substrates.

Year and sex-specific kite diagrams of adult
presence were produced for all abundant sum-
mer stoneflies by pooling all methods and
expressing daily catches as a percentage of the
total catch. Duration of emergence of Ptero-
narcijs californica Newport would be greatly
overestimated by including pitfall trap collec-
tions due to its synchronous emergence and
since pitfall traps were emptied on alternate
davs.

Great Basin Natur.\list

[Volume 55

Dates of first capture, 50% cumulative
catch, and last collection, plus total duration of
adult presence, were determined for the 11
most abundant species collected in the sum-
mers of 1988 and 1989. Emergence synchrony
was estimated using linear regression of the
cumulative percentage catch (all methods
pooled) versus days since first capture. Slopes
generated for each species were used as an
index of synchrony. Steeper slopes indicated a
more synchronous emergence. Slopes > 5%/d
were chosen to be indicative of synchronous
emergence since species with these slopes
emerged their entire population within a few
days and had steep, j-shaped, cumulative emer-
gence curves. Differences between slopes for
1988 and 1989 were tested using a modified t
test (Zar 1984). Common slopes were calculated
if no differences between years were noted.
This was a purely descriptive approach de-
signed to detect and compare patterns; there-
fore, it is not our aim to model emergence for
the purpose of prediction, but only to describe
patterns of emergence.

Since most adult collection methods em-
ployed in this stvidy collected adults of unknown
age, results reflected adult presence rather
than, in the strictest sense, emergence. No
attempt was made to discard old males and
females using any index of age. However, pat-
terns of adult presence should follow that of a
true emergence pattern, and since longevit\' of
most adults approached only 1 wk in the labo-
ratory, we believe these results to be useful.

Behavioral observations were made from
0800 to 1300 h and from 2000 to 2300 h for sev-
eral days during emergence of each species.
Observations made during intervening hours
produced little adult behavior. Timing of adult
activities, their relative distance from the
stream, and substrates on which activities took
place were monitored by walking the stream
margin, turning logs and rocks, and exposing
leaf-entrained bases of marginal vegetation.
Details of these observations have been narra-
tively described for each species in this paper.

Fecundity and Egg Incubation

Eggs of several species were incubated in
the laboratory to confirm proposed voltinism
based on growth histograms. Eggs were
placed into 1-cm-diameter dialysis tubing
bags and reared in a Frigid Units Living
Stream"', or they were stored in 100 x 15-

mm plastic petri dishes in an environmental
chamber. In both instances these were incu-
bated at approximate stream temperature and
light regime.

Fecundity was estimated from number of
egg batches deposited, number of eggs per
batch, and, for Skwala americana (Klapalek)
only, total number of eggs remaining in the
ovarioles. Females were housed at streamside
in screened, glass containers and provided
with moist cotton balls as a source of water.
Alternatively, some species were reared in
Denton and held under simulated streamside
conditions in large cotton-stoppered shell
vials.

Results

Physical Conditions

Mean daily stream temperatures in 1988
increased from near 10 °C in early June to
15 °C in mid- July (Fig. 1). The stream cooled
dramatically between 8 and 12 July. This coin-
cided with cool, damp weather conditions
(Fig. 2). Summer air temperature highs rarely
exceeded 30 °C in 1988, and rainfall occurred
at regular intervals throughout the summer
(Fig. 2). However, 1989 was marked by many
days above 30 °C with rainfall relegated to late
July and August (Fig. 2). The mean montlily dis-
charge of the Rio Conejos during 1987-1989
fluctuated predictably. Peak discharge
occurred typically in June but occurred in
Mav during the warm, windy spring of 1989
(Fig. 3).

Species Richness

More than 13,000 nymphs and adults were
studied over the 3-yr period. Among these

mean

low

high

£ 16

hi ^

;•.;■ \ r. A

f '/l

\N/^

a 12
CO

!\M/^-/i

fv^-A/y

it^f'-- ••

V ',

■--' ••■•/

7/14

Dates

Fig. 1. Daily mean, high, and low stream temperatures in
the Rio Conejos, summer 1988.

1995]

Stonefly Life Histories

5/27 6/6 6/16 6/26 7/6 7/16 7/26 8/5 8/15 8/25

Dates

C .

5/1 5/11 5/21 5/31 6/10 6/20 6/30 7/10 7/20 7/30 8/9

Dates

Fig. 2. Daih' high and low air temperatures and rainfall
for summer 1988 and 1989.

Fig. 3. Mean, minimum, and maximum monthly stream
discharge of the Rio Conejos during the study period.

were 31 species (Table 1) in eight families. The
Chloroperlidae, Perlodidae, and Capniidae
were the most speciose families with six,
seven, and seven species, respectively. Seven
genera were represented by more than one
species: Capnia, Utacapnia, Taenioneina, Suwal-
lia, Triznaka, Isogenoides, and Isoperla (Table 1).
Peak species richness occurred on or near
the summer solstice in both years (Fig. 4).
Pattern differences existed between years.

including an early waning and a more peaked
distribution of species richness in 1989.

Leuctridae

Paraleuctra vershina Gaufin and Ricker.

This was the only leuctrid found at our sites.
No nymphs were recovered from the stream,
indicating a probable hyporheic existence.
Adults were abundant in riparian vegetation
during June and July (Fig. 5). No variation in
adult presence parameters was noted for P.
vershina (Table 2). Emergence was classified
as extended in both years, although slopes of
these cumulative emergence curves were sig-
nificantly different over the 2 yr (Table 3).

Table 1. Stoneflies collected from the Rio Conejos,
Colorado, March 1987 through March 1990.

Euholognatha

Capniidae

Capnia coloradensis Claassen^

Capnia confitsa Claassen

Capnia vernalis (Newport)

Isocapnia crinita (Needham & Claassen)^

Utacapnia logana (Nebeker & Gaufin)*

Utacapnia poda (Nebeker & Caufin)!
Leuctridae

Paraleuctra vershina Gaufin & Ricker*
Nemouridae

Amphinemura banksi Baumann & Gaufin*

Prostoia hesemetsa (Ricker)*

Zapada frigida (Claassen)*
Taeniopterygidae

Taenionema pallidum (Banks)*

Taenionema pacificum (Banks)*

Doddsia occidentalis (Banks)*
Systellognatha
Chloroperlidae

Paraperia frontalis (Banks)*

Phimiperla diversa (Prison)*

SuualJia lincosa (Banks)*

Suuallia pallidula (Banks)*

Suuallia icardi Kondratieff & Kirchner*

Triznaka pintada (Ricker)*

Triznaka signata (Banks)*
Perlidae

Claassenia sabulosa (Banks)*

Hesperoperla pacifica (Banks)*
Perlodidae

Isogenoides zionensis Hanson*

Isogenoides prob. colubrimis (Hagen)*

Isoperla fill va Claassen

Isoperla monnona Banks*

Isoperla phalerata (Smith)*

Isoperla quinquepunctata (Banks)

Skwala americana (Klapalek)
Pteronarcyidae

Pteronarcella badia (Hagen)

Pteronarcijs californica Newport*

^New drainage and county records.

Great Basin Naturalist

[Volume 55

5/19 5/29 6/8 6/18 6/28 7/8 7/18 7/28 8/7 8/17

Dates

Fig. 4. Temporal species richness pattern of adult stoneflies collected daily fiom the Rio Conejos during the summers
of 1988 and 1989.

Chloroperlidae

Representatives from two subfamilies in-
habited the stream. The early- and mid-instar
nymphs of the Chloroperlinae genera could
not be reliably identified to genus. This neces-
sitated the illustration of a portion of the
nymphal growth of Plumiperla cliversa (Prison)
and Triznaka signata (Banks) as Chloroperlinae
spp. (Fig. 6). Growth of reliably identified mid-
to late-instar nymphs was illustrated separately.

Paraperla frontalis (Banks) (Paraperlinae).
Nymphs were collected infrequently among
marginal substrates during the colder months
of the year. All were pale, very thin, and had
eyes set far forward as described for mature
nymphs (Stewart and Stark 1988). These limit-
ed data are presented for the first year of the
presumed semivoltine growth pattern of this
large chloroperlid (Fig. 6). Less than 10 adults
were collected in early June during the 3-yr
study.

Plumiperla diversa (Chloroperlinae). No
adults were collected on which to base specif-
ic identity; however, nymphs of this genus are
distinctive, and only P. diversa has been col-
lected in this region (Baumann et al. 1977).
Nymphs were identifiable to genus by March.
Females were readily distinguished from
males at this time. Growth continued through
May when females attained a median HCW
9.6% larger than males. The limited nymphal

data suggested a univoltine-slow life cycle for
this species.

Suwallia pallidula (Banks) (Chloroper-
linae). Only 59 nymphs of Suwallia spp. were
collected from the Rio Conejos, even though
adults were abundant. Nymphs were hyporhe-
ic until immediately prior to emergence. This
habitat preference and our present inability to
distinguish congeners of Suwallia nymphs
precluded generation of meaningful his-
tograms and designation of voltinism for either
species. Adults of Suwallia wardi Kondratieff
& Kirchner were consistently larger than S.
pallidula. This trend followed in nymphs, too,
with proposed female nymphs of S. wardi in
June (peak emergence) being 22.0% larger

Paraleuctra vershina

M : F

25 % of total catch 36:58

9 M. kk 1988

20 30

MAY

JUNE

JULY

Fig. 5. Emergence of Paraleuctra vershina from the Rio
Conejos, 1988 and 1989. Polygons indicate daily relative
proportion of total catch.

1995]

Stonefly Life Histories

Table 2. Range of dates for adult presence parameters
(appears first) and 1989 from the Rio Conejos. Duration is
able for C. sabulosa, which emerged past our study period.

for 11 summer-emerging stonefly species collected in 1988
mean ± SD of the number of days. All parameters not avail-

Species

Date 1^'
captiue

Date
50% catch

Last date
capture

Duration (d)

P. vershina

2 June
1 June

12 June
17 June

5 Julv
7 July

35.0 ± 0.0

S. wardi

467
352

6 June
19 May

10 July
25 June

15 August
18 July

66.0 ±7.1

S. pallidida

276
162

30 June
IJuly

28 July
14 July

23 August
4 August

44.5 ± 4.8

T. signata

662
2697

9 June
2 June

28 June
19 June

23 August
12 July

59.0 ± 24.0

C. sabulosa

356
1195

19 July
16 July

—

I. fulva

19
61

9 June
9 June

22 June
18 June

7 July
28 June

24.5 + 7.8

I. phalerata

12
20

20 June
18 June

28 June
25 June

24 July
8 July

22.5 ± 3.5

I. quinqiiepunctata

9
12

24 June
19 June

14 Julv
5 July

27 Julv
15 July

30.0 ± 4.2

I. zionensis

200

8 June
10 June

19 June
17 June

28 June
24 June

15.5 ±4.9

P. badia

215
480

10 June
7 June

22 June
20 June

17 July
7 July

34.5 ± 3.5

P. califomica

55
21

6 June
4 June

8 June
5 June

12 June

13 June

6.0 ±2.8

than the July (peak emergence) females of S.
pallidula. Only two proposed male nymphs of
the latter were collected over the 3-yr period.

Adults of S. pallidula were collected in July
and August in both years (Fig. 7, Table 2).
Slopes from regression models were different
between years (t = -64.7, p < .0001), but
below the 5%/d criterion. We categorized this
species as an extended emerger (Table 3). The
median emergence date was advanced by 2
wk in 1989 over that of 1988 (Table 2). The
adult sex ratio over the two seasons was 13 6 :
415 9. Six field-collected and laboratory-
maintained females produced only one egg
batch (Table 4).

Suwallia wardi (Chloroperlinae). This was
the most abundant of the three Suwallia
species collected from the Rio Conejos. Adults
were first collected in late May or early June,
reached 50% cumulative catch by mid- July,
and disappeared from streamside by early

August (Table 2). It had the longest mean
duration of presence (66 d) for any stonefly
studied on the Rio Conejos (Table 2). Like its
congener, S. wardi' s 1989 date of median catch
was advanced by 2 wk over that of 1988 (Table
2, Fig. 7). Emergence of S. wardi was extend-
ed, and no significant slope differences were
noted between years (Table 3). No egg data
were collected for this species.

Riparian vegetation was used by this large,
yellow-green chloroperlid as a staging ground
for adult behaviors. Suwallia wardi was active
throughout the morning on sunny days and
again for 2-3 h before sunset if conditions
were warm and dry. During cool, rainy days
the low vegetation was devoid of S. wardi or
any other stonefly species.

Triznaka signata (Banks) (Chloroperlinae).
Identifiable, late-instar nymphs were collect-
ed during a 5-mo period in the spring and
summer. Nymphs of this univoltine-slow

Great Basin Naturalist

[Volume 55

Table 3. Synchrony and linear regression statistics lor
the years 1988 (appears first) and 1989. Slopes between
years were tested: * = significance .()5-.0I, ** = <.0()1
level or lower probability, and NT = not tested.

= 2 individuals

Species

Slope

fi2

Svnchronv

P. vershina
T. signata
S. piiUidttla
S'. wardi
C. sahiilosa
I. zionensis
P. badia

3.0
3.5*

1.5
3.0

2.5

3.5*

2.2
2.2

2.8
4.1

7.5
7.9

3.6
4.4*

P. califomica 13.3

18.9NT

.85
.90

.87
.91

.96
.90

.84

.85

.94
.99

.92
.95

.81
.92

.84
,97

.0001
.0001

.004
.103

extended
extended

synchronous
synchronous

extended
extended

synchronous
synchronous

species were largely full grown by April (Fig.
6) with some degree of sexual dimorphism
present at this time.

Adults first appeared in early June, reached
50% cumulative catch 2 wk later, and could no
longer be collected by late August (Fig. 7,
Table 2). Emergence was protandrous, but
slightly female-skewed sex ratios dominated in
both 1988 and 1989 (Fig. 7). Triznaka signata
displayed the greatest variation in last date of
capture and duration of presence of all stone-
flies in the river (Table 2). It advanced its 1989
median emergence date by 9 d over that of
1988. Regression slopes indicated an extended
emergence in both years (Table 3). Differences
between slopes for 1988 and 1989 were signif-
icant (t = -11.35, p < .0001).

Attempts during the entire study to obtain
eggs from laboratory-reared and -mated
females were unsuccessful. The mean number
of eggs from six females caught during ovipo-
sition flights was lower than any first batches
for other stoneflies studied (Table 4). Although
these females were held for a prolonged peri-
od of time, no additional egg batches were
laid.

Adults were never seen emerging in the
field, despite many hours of oliservation along
the shoreline, day and night, in habitats where

I — I = 50 % of monthly catch
n = 83

♦ * -- oviposition

\
emergence

Paraperia frontalis

> 1.2

. "4f4 ^^ ^

n = 453
Plumiperia diversa

n = 653
Triznaka signata

;^^

n = 694
Chloroperlinae spp.

Ill IV

V VI VII VIII IX X XI

Months

Fig. 6. Growth of Chloroperlidae nymphs collected
from the Rio Conejos, 1987-1990.

they were collected in abundance during the
day. Adults inhabited marginal vegetation,
where males were observed actively searching
willow stems and leaves for females. No drum-
ming was observed during the two summers
of intensive fieldwork. Large flights of adults
of both sexes took place just before dark, at
which time females were observed oviposit-
ing. Egg masses were dropped from up to 4-5
m above the stream.

Perlidae

Claassenia sabulosa. Although two perlids
were present in the Rio Conejos (Table 1),
only C. sabulosa was sufficiently abundant for
growth and emergence interpretation. Nymphs
of this species were found among larger rub-
ble of midstream. The life cycle was semivol-
tine and appeared to require 3 yr of nymphal
growth (Fig. 8). Recruitment occurred
throughout the fall with possibly some addi-
tional recruitment in March from overwinter-
ing eggs. Eggs containing eyespots were
recovered from the stream in October and
November. Sexual dimorphism in both size
and external genitalia occurred when nymphal
size reached 2.2 mm HCW. The size disparity
increased until the third year of growth when
little overlap between the sexes remained.

A protandrous emergence began in mid-
July in both years (Fig. 9, Table 2). Exuviae of
this species were abimdant throughout August,
possibly into September. Emergence ol C.
sabulosa was extended and slopes were signif-
icantly different between years (t = -10.7, /; <
.0001,' Table 3).

1995]

Stonefly Life Histories

= 2 ind. I = 5% of total catch

-^g^"-

Sex ratio
M:F

Triznaka signata ^^^ 263:446

1989 1181:1515

Suwallia wardi

,w^ ^|-g^

1988 119:348

1989 150:191

Suwallia pallidula

A 1988 13:253

1989 0:162

20 30 9 19 29 9 19 29 8 18 28

May June July August

Fig. 7. Emergence of Chloroperlidae from the Rio
Conejos, 1988 and 1989. Polygons indicate daily relative
proportion ot total catch.

Claassenia sahulosa produced the greatest
mean number of eggs of any stonefly species
studied (Table 4), with females producing up to
four batches. Longevity of seven females was
3.9 ± 1.9 d. Egg production lasted through
80% of the adult life. Several egg batches were
incubated, but none hatched within 6 mo of
obsei'vation.

Emergence occurred between 2000 and
2200 h. Nymphs crawled out of the water onto
emergent cobble and boulders to transform,
the entire molting process taking less than 5
min. Hardened and newly transformed males
ran over all emergent substrates, searched for
females in a circular pattern, and drummed
mostly on large mineral substrates. Pitfall trap
collections of 1989 caught a total of 115 male
adults in transect 1 and only 12 in transects 2
and 3. Only two females were collected in the
pitfall traps, presumably because of their less-
intensive and unidirectional movement pat-
tern. Therefore, excursions of great distance
away from the water's edge for either sex were
infrequent. Females were often found in the
morning under dry cobble with abdomens de-
void of eggs or with large egg masses suspend-
ed between the cerci. Several females were
observed at night running over the surface of
the water, but the cause of this behavior could
not be determined. No females were actually
observed ovipositing. Males were distinctly
cursorial, which fits with their brachypterous
morphology; however, females were never
observed flying, nor did they inhabit tall sub-
strates, even though they had full wings.

Perlodidae

Isogenoides zionensis Hanson (Perlodinae:
Perlodini). The large range in size of nymphs
from July samples (Fig. 10) could not be
accounted for by nymphs hatching from eggs
laid by June-mated females. June eggs reared
at simulated stream conditions hatched in
March and April, 9-10 mo after oviposition.
Therefore, at least some individuals of this
species have a semivoltine life cycle with eggs
diapausing over their first summer and winter.
Early-instar nymphs were missed in benthic
samples during their second spring, possibly
due to high water or their occurrence deep in
the substratum. Sexual dimorphism in size
and morphology was apparent by July of the
second year when nymphs approached 1.8
mm HCW (Fig. 10). This disparity increased
steadily throughout the rest of their growth.
Little overlap in size of the sexes existed by
May prior to emergence.

The adult presence parameters of /. zionen-
sis showed little variation over the 2 yr studied
(Table 2). Emergence was not protandrous,
but the sex ratio was heavily skewed towards
males (Fig. 11). This species was one of two
that emerged synchronously (Table 3). No dif-
ference in slope was found between years (t =
0.82, p > .2); therefore, a common slope of
7.6%/d was calculated.

Laborator\'-reared females put nearly 75%
of their total egg complement into a first batch
(Table 4). Only one of four females produced
additional batches.

Transformation of /. zionensis took place
from 2030 to about 2200 h. Nymphs crawled
away from the stream until they reached wil-
lows or other vegetation, then ascended < 1 m
vertically where they molted. Daylight activity-
began by 0700-0800 h at the base of small wil-
low coppices, where adults were often found
in emergent leafpacks. Adults ascended stream-
side willows as the sun rose. Drumming, mat-
ing, and egg batch formation took place from
these perches. Females crawled to the tops of
these willows and flew to the stream where
they fluttered on the water to release their
black egg masses. Most activity ceased by
1300-1400 h on days when air temperature
reached near 25 °C. On cloudy, cool days this
ascendance did not occur. Most adults could
then be found in the leaf-entrained bases of
riparian vegetation. Drumming on willow stems

Great Basin Naturalist

[Volume 55

Table 4. Mean eggs per hatch, nuinher of hatches, and mean total egg complement for nine species of stoneflies
occurring in the Rio Conejos, Colorado.

Eggs

/ hatch
n

Species

Total

S. pallidula

54.7 ± 26.6
6

—

54.7 ± 26.6
6

T. signata

42.2 ±17.4
6

—

42.2 ± 17.4
6

C. salmlosa

2166.0 ±774.0

902.0 ± 246.2
5

158.0 ±91.0
5

40.0

3188.0 ±613.0

I. zionemis

588.0 ± 86.0
4

327.0

185.0
1

—

843.2 + 141.4
4

I. fulva

231.5 ±7.8
2

—

231.5 ± 7.8
2

I. phalerata

703.0

—

703.0

S. amcricana

884.7 ± 267.3
6

—

884.7 ± 267.3^
6

P. badia

339.0 ± 86.0
30

58.4 ± 37.2
5

56.8 ± 39.7
4

—

351.0 ±101.0
30

P. calif ornica

393.0 ± 125.6
4

191.3 ±130.2
4

94.3 ± 49.5
4

69.8 ± 24.7
4

51.3 ±29.3
4

58.5 ± 23.3
2

57.0

845.3 ± 90.5
4

^Total fecundity includes those eggs remaining in o\aii(ilts

was observed at night, even when tempera-
tures approached 10 °C.

Isoperla fulva Claassen (Isoperlinae). We
collected this species in benthic samples only
occasionally, but enough individuals were
obtained to allow a tentative interpretation of
voltinism. Recruitment of nymphs was first
detected in August (Fig. 12). These measured
0.4-0.8 mm HCW and grew at a slow rate
throughout the fall until a winter decrease in
growth rate. Their size increased dramatically
after February, until emergence in June and
July. This species conformed to a univoltine-
slow growth pattern.

Adults were collected for the first time on 9
June in both years (Fig. 11, Table 2). Sex ratios
for the small number of 1988 adults were
approximately equal, but heavily skewed
towards males in 1989. Numbers of adults col-
lected in both years were too small to warrant
an analysis of synchrony.

Fecundity was difficult to assess since few
mature nymphs were available for rearing.
One egg batch from each of two field-oviposit-
ing females was collected (Table 4). Longevity
of three field-collected adult females was 5.7
± 0.58 d.

Isoperla phalerata (Smith) (Isoperlinae).
Although the number of nymphs collected was
small, no month supported more than one size
class (Fig. 12). Therefore, we have tentatively
proposed a univoltine-slow growth pattern for
this species. Adults were taken from mid- June
through mid- July (Table 2, Fig. 11). No assess-
ment of synchrony was made for 7. phalerata
due to low numbers of adults captured.
Females did not produce eggs in captivity. A
single egg batch from a field-collected individ-
ual contained 703 eggs. Four field-caught
females lived 11.3 ± 3.6 d past date of capture.

Isoperla quinquepunctata (Banks) (Isoper-
linae). This species was more common at site

1995]

Stonefly Life Histories

Claassenia sabulosa
4 % of total catch

I II III IV V VI VII Vlli IX X XI

Months

Fig. 8. Growth of Claassenia sahtilo.sa n\'mphs collected
from the Rio Conejos, 1987-1990.

three. The data suggested that /. qiiinquepunc-
tata had a univoltine-fast growth pattern.
Recruitment occuiTed in Januaiy and Februaiy
(Fig. 12), and growth was rapid from March
through May. Sexual dimoiphism in nymphal
size was not as evident in this species as in its
congeners. Emergence began in mid-June and
lasted through much of July (Table 2, Fig. 11).
No eggs were collected.

Skwala americana (Klapalek) (Perlodinae).
This species displayed a univoltine-slow
growth pattern and grew faster during sum-
mer and fall months than all other perlodids in
the Rio Conejos (Fig. 13). Nymphs were re-
cruited in June and increased their median
HCW from 0.4 mm to about 2.8 mm by
January. Growth was nearly completed by this
time. Sexual dimorphism was apparent as early
as August, and female nymphs reached a medi-
an HCW before emergence that was 21.4%
greater than males. Female nymphs in April
were found to contain fully sclerotized eggs in
their oviducts; hence, this species is fully
capable of mating and egg-laying immediately
upon emergence.

Emergence was in April and early May when
our sampling was still on a monthly basis;
therefore, no detailed analysis of emergence
phenology and synchrony can be offered.
Adults were collected mainly from emergent
logjam debris or under cobble at the stream
margin.

Egg batches collected in mid-April from
four laboratory-reared females hatched syn-
chronously after a mean of 61.0 ± 7.3 d. This
corroborates field collections of early-instar
nymphs in June. Only a single egg batch was
collected from each of six laboratory-reared
females (Table 1).

July

18
August

Fig. 9. Emergence of Claassenia sabulosa from the Rio
Conejos, 1988 and 1989. Polygons indicate daily relative
proportion of total catch.

Pteronarcyidae

Pteronarcella hadia (Hagen). This species
was found to have a semivoltine growth pat-
tern. Recruitment of nymphs began in March
and April from eggs laid the previous June
(Fig. 14). Many small nymphs were available
in benthic samples by mid-April when they
were at 0.2-0.4 mm HCW. This scenario was
coiToborated by laboratoiy incubation of several
egg batches that hatched in March and April
after a 9-10-mo diapause. Growth of nymphs
was rapid throughout their first spring. Size
differentiation among sexes was not apparent
until August, a full 14 mo after oviposition.
Median size of females just before emergence
the following May was 21% greater than that
of males.

Emergence began by early June, with slight
protandry and a preponderance of males being
collected (Fig. 15). Median emergence occurred
in the third week of June in both years (Table
2). Emergence was extended (Table 3) and
slopes were significantly different between
years (t = -2.2, p < .05).

Females generally laid only single egg
batches, but a small number produced up to
three egg batches (Table 4). Most females laid
their first egg batch within 24 h of mating and
often waited 2-d intei-vals before laying others.
Longevity of seven females under simulated
field conditions was 7.7 ± 4.2 d.

Pteronarcella hadia emerged just after dusk
and typically used willows, cottonwoods, and
stream margin sedges as transformation sites.
Males were observed actively searching the
willows and drumming for females at night.

Great Basin Naturalist

[Volume 55

E
E
S

Isogenoides zionensis
if

emergence — ' ~ oviposiiion

(-(• <p I — I = 1 5% of monthly catch

™ = 4 individuals
n = 421

I II III IV V VI VII VIII IX X XI

Months

Fig. 10. Growth of Isogenoides zionensis nymphs col-
lected from the Rio Conejos, 1987-1990.

even when air temperatures were near 10 °C.
Adults entered the leaf-choked bases of wil-
lows as the night progressed and were often
found the next morning in large mating aggre-
gations under these debris. These individuals
ascended the willows as the sun warmed the
air at streamside. Drumming, mate searching,
mating, and egg batching took place in mid-
moniing hours, while most activity' ceased by
1200 h when air temperatures reached
22-25 °C. Females oviposited by launching
themselves from the tips of tall riparian shrubs
toward the stream, where they would jettison
their white egg mass a few meters above the
surface of fast-flowing water.

Pteronarcys californica Newport. This
species is commonly known as the salmonfly.
Recruitment began in April (Fig. 16) after a
9-10-mo egg diapause. Nymphs grew to only
about 1 mm HCW through their first year.
Sexual differences in size and morphology
were apparent by June of their second year,
when they were nearly 1.5 mm HCW.
Nymphs grew for two more years, by the end
of which time preemergent females had
attained a 20% larger median HCW than
males. These data suggest a semivoltine life
history of 4-yr duration for this species.

Adults were first found on 6 June during
both years (Fig. 15, Table 2). Emergence was a
highly synchronous event (Table 3). Slopes
were not tested for significant differences due
to small sample size.

Most laboratory-reared females produced
five egg batches, but one individual produced
seven (Table 4). Egg production lasted
through 82% of the 15.0 ± 1.8 d (n = 4) aver-
age adult female life span.

■ = 1 iND. Perlodidae

X = 20 % of total catch

Isogenoides zionensis

1988

M: F
146:55

%W^ y

1989

53:18

Isoperla fulva

^M ■-»=

1988

11:9

43-«^

1989

53:7

^ r^ I

Isoperla phalerata

1988

4:8

1989

7:13

Isoperla quinquepunctata
_ _ 1988

3:6

1989

6:6

19
June

19
July

Fig. 11. Emergence of Perlodidae from the Rio
Conejos, 1988 and 1989. Polygons indicate daily relative
proportion of total catch.

Adults utilized marginal vegetation, much
as did ?. badia, as a staging ground for mating
and ovipositing. However, they tended to select
the taller cottonwoods and Engelmann spruce
rather than the shorter willows for their activi-
ties. Salmonflies oviposited by flying over the
stream and dropping their salmon-colored or
bluish egg masses (dimoiphism in egg color was
observed) from as high as 10 m. Ovipositing
adults were heavily fed upon by opportunistic
Eastern Robins {Turdus migratorius) and
Steller's Jays {Cyanocitta stelleri).

Discussion

Species Richness

The Rio Conejos displayed a great diversity
of Plecoptera. Twenty of the 31 species were
evenly distributed among the Capniidae, the
Chloroperlidae, and the Perlodidae. The only
North American family not represented was
the Peltoperlidae, which occurs transcontinen-
tally, but not in latitudes below the northern
Rocky Mountains (Baumann et al. 1977). Nearly
all species collected were adapted for a mon-
tane existence and were characteristic of
streams with high biotic integrity. Twenty-five
species were both new drainage and county
records (Baumann et al. 1977, Szczytko and
Stewart 1979, Nelson and Baumann 1989; Table
1), though all of them had been previously
reported from Colorado and neighboring New
Mexico. This demonstrates that we have yet to
adequately investigate the fine-scale diversity

1995]

Stonefly Life Histories

E
£1

Isoperlinae spp.

I = 30 % of monthly catch ^ i f

= 1 individual emergence

cf 9 „ -^ [

Isoperia fulva n = 105

> T

Isoperia phalerata
■^ n = 25

f Y Isoperia qumquepunctata
n = 229

III IV V VI VII VIII IX

Months

Fig. 12. Growth of Isoperia spp. n>iTiphs collected from
the Rio Conejos, 1987-1990.

and distribution of this order of aquatic insects
in at least some portions of the southern
Rocky Mountains.

Responses to Altered Thermal Regime

We became aware of substantial climatic
differences (Fig. 2) between the two summers
when adults were intensively studied. Though
no water temperatures were available for
1989, air temperatures (Fig. 2) and hydrologic
data (Fig. 3) suggested that the stream warmed
more quickly and attained peak summer highs
much earlier than in 1988. Consequently,
development of several species was hurried,
which narrowed the window of time adults
were present streamside. At the assemblage
level of organization, this trend is demonstrat-
ed by the species richness pattern of Figure 4.
The 1989 pattern was more peaked and great-
ly truncated over that of 1988. Species-level
responses can be demonstrated by inspection
of the flight diagrams for each species. Nine of
the 11 species presented in Table 2 show in-
creased median emergence dates. Additionally,
slopes produced by linear regression that
were different between years (Table 3) were
always higher in 1989. This result was consis-
tent with a hurried nymphal development and
shorter emergence period for each species.

Life History Parameters

Leuctridae

Paraleiictra vershina. Harper (1973b) reports
that most Leuctra ferruginea in an Ontario
stream are semivoltine, but that some univol-
tine individuals exist. Huryn and Wallace
(1987) propose a 2-yr life cycle for a composite
of Leuctra spp., most of which were probably

c?9

Skwala americana

^ = 30 % of monthly catch
■ =3 individuals
n = 227

Fig. 13. Growth of Skwala americana nymphs collected
from the Rio Conejos, 1987-1990.

L. ferruginea (Walker). Snellen and Stewart
(1979) record univoltine fast cycles for
Zealeuctra claasseni and Z. hitei in streams of
north Texas. Additionally, Ernst and Stewart
(1985a) report Leuctra tenuis as univoltine-fast
in an Ouachita Mountain stream.

Chloroperlidae

Most Chloroperlidae exhibit a univoltine-
slow or -fast growth pattern. Haploperla brevis
(Banks) is widespread from Oklahoma to
Quebec and west to Alberta, Canada. Ontario
(Harper and Magnin 1969), Quebec (Harper
et al. 1994), and Oklahoma (Ernst and Stewart
1985a) populations exhibited univoltine-fast
growth with a 2-5-mo diapause, while Alberta
populations were univoltine-slow (Barton
1980). European populations of Chloroperla
tripunctata (Scopoli) (Elliott 1988), Siphono-
perki torrentium (Pictet) (Elliott 1967), and S.
hunneisteri (Pictet) (Benedetto 1973) also
exhibited univoltine-slow growth. Species
with semivoltine growth include Sweltsa
onkos (Ricker) and possibly Utaperla gaspe-
siana Harper and Roy (Haiper 1973a, Harper
et al. 1994), S. mediana (Banks) (Cushman et
al. 1977), and S. lateralis (Banks) (Huiyn and
Wallace 1987).

Paraperla frontalis. Stanford and Gaufin
(1974) presented some evidence for semivol-
tine growth of this species. Emergence for this
species and for P. wilsoni Ricker occurs from
May through July (Stewart and Stark 1988).
Paraperlinae are radier robust chloroperlids that
tend to be hx^Dorheic for most of their nymphal
development. Their larger size, the more sta-
ble stream temperatures in the hyporheic
environment (Hendricks 1993), and the possi-
blv low availabilitv of some nutrients in the

Great Basin Naturalist

[Volume 55

Pteronarcella badia

oviposition

V VI VII VIII

Months

Fig. 14. Growth of Pteronarcella haclia inmphs collect-
ed from the Rio Conejos, 1987-1990.

hyporheic habitat (Stanford and Ward 1993)
may have contributed to a preponderance of
semivoltinism in this subfamily.

Plumiperla diversa. Stewart et al. (1990) re-
ported a univoltine-slow cycle for this species
on the North Slope of Alaska. Emergence
occurred from May through September, with
recruitment of nymphs from a direct hatch in
July. Growth occurred through the summer
months with most nymphs attaining maximimi
size before a winter quiescence. This assess-
ment compared well with our limited data.
Failure to collect adults was probably due to
our infrequent sampling during their pre-
sumed early May emergence.

Suwallia pallidula and Suwallia wardi. No
aspects of the life histories of either S. paUidiiki
or S. wardi have been reported. The latter was
recently described from a Colorado Front
Range springbrook (Kondratieff and Kirchner
1990). It was one of the most abundant chloro-
perlids in the Rio Conejos. This suggests that
its ecological tolerance is wide and that it may
soon be found in a variety of streams in the
southern Rocky Mountains.

Several explanations are possible for the
heavily female-skewed sex ratio (13 cj:425 ?)
of S. pallidula adults. The most probable is a
combination of limited use of emergence traps
coupled with an inaccessible microhabitat of
adult males, probably high in the vegetation.
Parthenogenesis may also be possible, but it is
exceedingly rare in stoneflies. HaqDcr (1973a)
reported that a few eggs of a perlid, Parag-
netina media (Walker), hatched without fertil-
ization. We did not attempt rearing of eggs
from virgin females to check for parthenogen-
esis in either Suwallia spp. These sex ratios
are a perplexing problem, compounded by the

Reronarcyidae

= 20% of total catch

M:F

9 /^f(s. Pteronarcys calif ornica

1988

23:32

<3 W""

M; -

1989

11:10

^ Pteronarcella badia

_ /v_:^»'^r>w„

1988

134:79

■•nr^r^

. nr^-j^^T^^ «.

1989

308:172

1 1 1 1 1

30 9 19 29 9

1
19

1
29

June July

Fig. 15. Emergence of Pteronarcella badia and
Pteronarcys californica from the Rio Conejos, 1988 and
1989. Polygons indicate daily relative proportion of total
catch.

fact that 0 6 : 657 ? of the closely related S.
lineosa were caught during concurrent sam-
pling on Massey Creek, a tributary of the Rio
Conejos.

Triznaka signata. Hassage and Stewart
(1990) studied the widely distributed T. signata
in the Rio Vallecitos of northern New Mexico.
They reported a univoltine-slow growth pat-
tern, with which we concur. No study of the
emergence of this species has previously been
published.

Perlidae

Claassenia sabulosa. Hassage and Stewart
(1990) and Barton (1980) report a merovoltine
(>2 yr) growth pattern for New Mexico and
Alberta populations of this species. No egg
batches from the Rio Conejos hatched in our
laboratory, but this Colorado population showed
some evidence of an extended hatch leading to
cohort splitting (Stewart and Stark 1988). Eggs
may undergo a temperature-dependent quies-
cence as occurs in Dinocras cephalotes (Curtis)
when fall temperatures decline to 8°C (Lille-
hammer et al. 1989). Presence of first-instar
nymphs in the fall, eyed eggs in October and
November, and more first-instar nymphs in
March supported this contention.

Life histories have been reported for at least
one species in every genus in the tribe Perlini,
to which C. sabulosa belongs. All growth pat-
terns involve 2-3 yr of development. Agnetina
flavescens (Walsh), from an Ozark stream,
exhibits a 2-yr life cycle, a short egg incuba-
tion period, and an extended emergence period

1995]

Stonefly Life Histories

§3

I — I = 20 % ol monthly catch emergence
■ = 1 individual

i* Pteronarcys califomica

^^ oviposition

VI VII Vlll IX X XI

Fig. 16. Growth of Pteronarcys califomica nymphs col-
lected from the Rio Conejos, 1987-1990.

(Ernst and Stewart 1985b). Agnetina capitata
(Pictet) was shown to have a 3-yr cycle, ex-
tended emergence, and a 40-80-d egg incuba-
tion period in Ontario (Harper 1973a). This
range of incubation coupled with a long emer-
gence promotes great differences in size of
nymphs that ultimately prevents the separa-
tion of cohorts and determination of voltinism.
This was also a problem for C. sabulosa in the
Rio Conejos.

Perlodidae

This family contains over 115 species (Stark
et al. 1986, Stewart and Stark 1988) in the
Nearctic. Although life histories of only 26
species are known, a clear trend toward uni-
voltine-slow cycles occurs among the subfami-
lies Isoperlinae and Perlodinae (Stewart and
Stark 1988). Growth and emergence had not
previously been studied for three of the seven
perlodids in the Rio Conejos. These include /.
zionensis, 1. quinquepunctata, and /. phalerata.

Isogenoides zionensis. Few detailed life his-
tory studies of the genus have been reported
(Stewart and Stark 1988). Barton (1980) supect-
ed semivoltinism for an Alberta population of
I. cohibrinus, since two size classes of nymphs
were collected in early May. Flannagan (1977)
reported great body length variation in May for
this species in another Alberta watershed but
concluded a univoltine-slow cycle. Hilsenhoff
and Billmeyer (1973) and Dosdall and Lehm-
kuhl (1979) proposed univoltine growth pat-
terns for the May-June-emerging /. frontalis
in Wisconsin and Saskatchewan streams,
respectively, based on samples taken a few
months of the year. Semivoltinism, as reported
for /. zionensis in the Rio Conejos, may also

occur in its congeners, but this will be con-
firmed only when detailed studies using small
mesh nets, frequent sampling, and egg rearing
have been conducted.

Isoperla spp. Of the three Isoperla whose
partial growth patterns are presented here,
only 7. fulva has been previously reported.
Hassage and Stewart (1990) reported a univol-
tine-slow cycle, with a June emergence in the
Rio Vallecitos of New Mexico. We concur with
the New Mexico study. Our results agree well
with reviews of Isoperla biology, summarized
for 12 Nearctic species through 1987 (Stewart
and Stark 1988). Ten species were univoltine-
slow, while only two were univoltine-fast.

In more recent literature Stewart et al. (1990)
reported univoltine-slow growth for /. petersoni
Needham & Christenson of Alaska. Additionally,
Harper et al. (1994) added as univoltine-slow
/. francesca Harper and 7. montana (Banks)
from Quebec populations. These and our Rio
Conejos work bring to 17 the Nearctic Isoperla
species known to exhibit univoltine-slow
cycles, while only three species appear to be
univoltine-fast. Isoperla grammatica (Poda)
and 7. difformis (Klapalek) (Malmqvist and
Sjostrom 1989) and 7. obscura (Zetterstedt)
studied by Ulfstrand (1968) are univoltine-
slow in the Palearctic.

Up to seven species of Isoperla commonly
occur in streams in North America (Stewart
and Stark 1988); conversely, in Scandinavia
rarely more than two species occur simultane-
ously (Malmqvist and Sjostrom 1989). Congen-
erics of aquatic insects often partition resources
along one or more resource gradients (Grant
and Mackay 1969). Though only small numbers
of adults were collected, a pattern of succes-
sive emergence of 7. fulva, I. quinquepunctata,
and 7. phalerata was clear in the Rio Conejos.
Fifty percent cumulative catch dates for 7.
fulva, I. phalerata, and 7. quinquepunctata were
22 June, 28 June, and 14 July, respectively, for
1988. These dates for 1989 were 18 June, 25
June, and 5 July. Temporal segregation
brought about by a gradual change in domi-
nance (lilies 1952) of these species may have
accounted for the pres-ent coexistence of
these stoneflies.

Skwala americana. Two other studies re-
ported univoltine-slow cycles with emergence
from February through April for this species in
northern New Mexico and central Colorado
(Short and Ward 1980, Hassage and Stewart

Great Basin Naturalist

[Volume 55

1990). Skwala curvata (Hanson) of" California
also exhibited a univoltine-slow cycle, with
emergence in April and May (Sheldon 1972).
Other Arcynopterygini with univoltine-slow
growth include Frisonia picticeps (Hanson) in
California (Sheldon 1972), Megarcys signata
(Hagen) in Utah (Cather and Gaufin 1975),
and Perhnodes aurea (Smith) in California and
Alberta (Radford and Hartland-Howe 1971,
Sheldon 1972).

Sheldon (1972) estimated average total
fecundity of S. curvata to be near 1780 eggs
for preemergent nymphs. This is much greater
than that proposed for S. americana from the
Rio Conejos. He used interocular width as an
index to predict fecundity. Conversion of inter-
ocular width to HCW likely involves a factor
of 2X, which would make S. curvata the larger
of the two stoneflies. This largely accounts for
differences in fecundity. Mutch and Pritchard

(1986) reported that S. americana (as S. paralle-
la) had a warm, stenothermal egg development.

Most species in this family have conserved
the life history traits that Lillehammer et al.
(1989) proposed as ancestral. These traits
include uni\'oltine-slow cycles, temperature-
dependent growth, and direct egg develop-
ment. Isoperla quinquepunctata and /. zionen-
sis have likely abandoned all of these except
temperature-dependent growth.

Pteronarcyidae

Pteronarcella badia. Gaufin et al. (1972)
reported that a 2-yr life cycle was possible for
this species in Utah; however, S. Perry et al.

(1987) and Stanford (1975) reported a univol-
tine life history in Montana. No eggs were
reared in either Montana study, and it is
apparent from their growth histograms that
early instars were missed entirely. Therefore,
semivoltine life history is most probable
throughout its range.

Nymphs of this species are more likely to
be found aggregated on filter paper leaf mod-
els than alone (Hassage et al. 1988). We have
also observed nymphs aggregating under mar-
gin cobble immediately before emergence.
Adults aggregate in leaf debris at the base of
willow and cottonwood coppices at the Rio
Conejos. This behavior may be attributable to
the transformation and nighttime refuge sites
being contagiously distributed. Hassage et al.

(1988) also postulated that aggregation in P.
badia lowers individual risk to predation.

Pteronarcys californica. The egg diapause
plus 38-mo nymphal life span places total life
span of this population at 4 yr. This is one of
the longest-lived aquatic insects known to
occur in the Nearctic. Additionally, this
species is perhaps our most synchronously
emerging stonefly.

Two- to 3-yr life cycles with a 9-10-mo egg
diapause occur in other Pteronarcys such as P.
dorsata (Barton 1980), P. proteus (Holdsworth
1941a, 1941b, W. Perry et al. 1987), and P
scotti in the southern Appalachian Mountains
(Folsom and Manuel 1983). However, Lechleit-
ner and Kondratieff (1983) detailed a 1-yr life
histoiy for P. dorsata in Virginia.

Multiple-year life histories are common
among larger species of the Pteronarcyidae
(Stewart and Stark 1988). Accompanying this
long nymphal growth, and perhaps contribut-
ing to it, is another life history trait, long egg
diapause. Univoltine growth patterns and
direct egg development are ancestral patterns,
while the semivoltine growth and diapause of
P. badia and P. californica are derived traits
(Lillehammer et al. 1989). Future studies of
egg incubation in lower latitudes of North
America will enable us to outline the range of
responses of which Pteronarcys and Pteronar-
cella are capable.

Unanswered Questions

Several largely unanswered questions per-
sist about the life histories of stoneflies in and
along the Rio Conejos. We have found that
nymphs of many chloroperlids are not avail-
able in surface sediments until just prior to
emergence. They must be hyporheic in their
habitat choice. Second, chloroperlids of the
present study did not readily produce eggs in
captivity, and those incubated never hatched.
We can still ask many questions about their
life histories. The answers would require a
detailed study of the hyporheic habitat of an
open-sediment stream like the Rio Conejos.
This study should concentrate only on the
chloroperlids, since they are generally abun-
dant and diverse. Such a study would still fit
within the comparative study approach of
Sheldon (1972), but the guild would involve
hyporheic chloroperlids.

To settle the dilemma of aberrant sex ratios
in this family, studies must concentrate on the
presence of male nymphs in the stream. In
this way the search for adult males whose

1995]

Stonefly Life Histories

whereabouts are unknown need not take
place, since both sexes of nymphs presumably
enjoy a similar microhabitat. If no male
nymphs are located, then incubation of eggs
from virgin females should be conducted to
confirm the possibility of parthenogenesis.

An exciting observation we made during
the study was that of basking in the sun of
nearly all adults of summer-emerging species.
Most displayed a remarkably consistent pat-
tern of ascendence of riparian vegetation be-
ginning at about 0800 h. Activity usually ceased
by 1300 h when air temperatures were hottest.
This ascendence culminated for females in
egg batching and oviposition flights, while
males used these riparian staging grounds for
mate searching, drumming, and mating.
Stoneflies should be investigated for potential
to benefit from basking, an unreported phe-
nomenon for Plecoptera.

Acknowledgments

We thank the Conejos Peak District of the
U.S. Forest Sei"vice for providing lodging and
laboratory space during the study. Special
thanks go to J. B. Moring for his help with
sample collection and D. Ziegler for providing
some fecundity data for P. badia. This project
was partially funded by faculty research funds
of UNT and a National Science Foundation
grant, BSR 8308422, to BCWS.

Literature Cited

Barton, D. R. 1980. Obsei-vations on the life histories and
biology of Ephemeroptera and Plecoptera in north-
eastern Alberta. Aquatic Insects 2: 97-111.

Baumann, R. W, a. R. G.wfin, a.nd R. E Surdick. 1977.
The stoneflies (Plecoptera) of the Rocky Mountains.
Memoirs of the American Entomological Society 31.

Benedetto, L. A. 1973. Growth of stonefly nymphs in
Swedish Lapland. Entomologisk Tidskrift 94: 15-19.

Gather, M. R., and A. R. Gaufin. 1975. Life history and
ecology of Megarctjs signata (Plecoptera: Perlodidae),
Mill Greek, Wasatch Mountains, Utah. Great Basin
Naturalist 35: 39^8.

GusHMAN, R. M., J. W. Elwood, and S. G. Hildebrand.
1977. Life history and production dynamics of A//o-
perla mediana and Diplectrona modesta in Walker
Branch, Tennessee. American Midland Naturalist
98: 354-364.

Dosdall, L., and D. M. Lehmkuhl. 1979. Stoneflies
(Plecoptera) of Saskatchewan. Quaestiones Entomo-
logicae 15: 3-116.

Elliott, J. M. 1967. The life histories and drifting of
Plecoptera and Ephemeroptera in a Dartmoor stream.
Journal of Animal Ecolog\' 36: 343-362.

. 1988. Interspecific and intraspecific variation in

egg hatching for British populations of the stoneflies

Siphonoperla torrenfium and Chloroperla tripunciata
(Plecoptera: Chloroperlidae). Freshwater Biology' 20:
11-18.

Ern.st M. R., and K. W Stewart 1985a. Growth and drift
of nine stonefly species (Plecoptera) in an Oklahoma
Ozark footliills stream, and conformation to regression
models. Annals of the Entomological Society of
America 78: 635-646.

. 1985b. Emergence patterns and an assessment of

collecting methods for adult stoneflies (Plecoptera)
in an Ozark foothills stream. Ganadian Journal of
Zoology 63: 2962-2968.

Flannagan, J. F 1977. Life cycles of some common aqua-
tic insects of the Athabasca River, Alberta. Alberta
Oil Sands Environmental Research Program, Report
11.20 pp.

FoLSOM, T. G., and K. L. Manuel. 1983. The life cycle of
Pteronarctjs scotti (Plecoptera: Pteronarcyidae) from
the southern Appalachians, U.S.A. Atjuatic Insects 5:
227-232.

Gaufin, A. R., W E. Ricker, M. Miner, R Milam, and
R. A. Hays. 1972. The stoneflies (Plecoptera) of
Montana. Transactions of the American Entomological
Society 98: 1-161.

Grant, P R., ,\nd R. J. Mackay. 1969. Ecological separa-
tion of systematically related stream insects. Gana-
dian Journal of Zoology 47: 691-694.

Harper, P. P. 1973a. Emergence, reproduction, and
growth of setipalpian Plecoptera in southern Ontario.
Oikos 24: 94-107.

. 1973b. Life histories of Nemouridae and Leuctri-

dae in southeiTi Ontario (Plecoptera). Hydrobiologia
41:309-356.

Harper, R P, and H. B. N. Hynes. 1972. Life histories of
Capniidae and Taeniopteiygidae in southeiTi Ontario
(Plecoptera). Archiv fiir Hydrobiologie, Supplement
40: 274-314.

Harper, R R, and E. Magnin. 1969. Gycles vitaux de
quelques Plecopteres des Laurentides (Insecta).
Ganadian Journal of Zoology 47: 483-494.

Harper, P R, M. Lauzon, and F. Harper. 1994. Life
cycles of sundry stoneflies (Plecoptera) from
Quebec. Review d'Entomologie Quebec 36: 28-41.

Hassage, R. L., and K. W. Stewart. 1990. Growth and
voltinism of five stonefly species in a New Me.xico
mountain stream. Southwestern Naturalist 35:
130-134.

Hassage, R. L., R. E. DeWalt, and K. W. Stewart 1988.
Aggregation of Pteronarcella badia nymphs and
effects of interaction with Claassenia sabulosa
(Plecoptera). Oikos ,53: 37-40.

Hendricks, S. R 1993. Microbial ecology of the hyporhe-
ic zone: a perspective integrating hydrology and
biology. Journal of the North American Benthologi-
cal Society 12: 70-78.

Hilsenhofe W. L., and S. J. Billmeyer. 1973. Perlo-
didae (Plecoptera) of Wisconsin. Great Lakes Ento-
mologist 6: 1-14.

HoLDSWORTH, R. P 1941a. The life histor>' and growtli of
Pteronarcys proteiis Newman (Pteronarcyidae:
Plecoptera). Annals of the Entomological Society of
America 34: 495-502.

. 1941b. Additional information and a correction con-
cerning the growth of Pteronarcys protean Newman
(Pteronarcyidae: Plecoptera). Annals of the Entomo-
logical Society of America 34: 714—715.

Great Basin Naturalist

[Volume 55

HURYN, A. D., AND J. B. Wallace. 1987. The exopterygote
insect community of a mountain stream in North
Carohna, USA: life histories, production, and func-
tional structure. Aquatic Insects 9: 229-251.

Hynes, H. B. N. 1992. Some thoughts on unanswered
questions about stoneflies. XI International Sympos-
ium on Plecoptera, Tomahawk, Wl.

IllIES, J. 1952. Die Plecoptcren luid das Monardsche
Prinzip. Berlin Limnologische Fluhstation Freuden-
thid 3: 53-69.

Knight, A. W, and A. R. Gaufin. 1966. Altitudinal distribu-
tion of stoneflies (Plecoptera) in a Rocky Mountain
drainage system. Journal of the Kansas Entomological
Society 39: 668-675.

Kondr.at[eff, B. C, and R. K Kirchner. 1991. New
Nearctic Chloroperlidae (Plecoptera). Journal of the
New York Entomological Society 99: 199-203.

Lechleitner, R. a., and B. C. Kondr.\tiefe 1983. The
life history of Pteronarcijs dorsata (Say) (Plecoptera:
Pteronarcvidae) in southwestern Virginia. Canadian
Journal Zoology 61: 1981-1985.

LiLLEHAMMER, A., J. E. BRITTAIN, S. J. SaLTVEIT, AND R S.
Nielsen. 1989. Egg development, nymphal growth
and life cycle strategies in Plecoptera. Holarctic
Ecology 12: 173-186.

Malmqvist, B., and R Sjostrom. 1989. The life cycle and
growth of Isoperla grammatica and /. diffonnis
(Plecoptera) in southernmost Sweden: intra- and
interspecific considerations. Hydrobiologia 175:
97-108.

Mutch, R. A., and G. Pritchard. 1986. Developmental
rates of eggs of some Canadian stoneflies (Plecoptera)
in relation to temperature. Journal of the North
American Benthological Society 5: 272-277.

Nelson, C. R., and R. W. Baumann. 1989. Systematics
and distribution of the winter stonefly genus Capnia
(Plecoptera: Capniidae) in North America. Great
Basin Naturalist 49: 289-366.

Perry, S. A., W. B. Perry, and J. A. Stanford. 1987.
Effects of thermal regime on size, growth rates and
emergence of two species of stoneflies (Plecoptera:
Taeniopterygidae, Pteronarcyidae) in the Flathead
River, Montana. American Midland Naturalist 117:
83-93.

Perry, W. B., E. F. Benfield, S. A. Perry, and J. R.
Webster. 1987. Energetics, growth, and production
of a leaf-shredding stonefly in an Appalachian
Mountain stream. Journal of the North American
Benthological Society 6: 12-25.

Petsch, H. E., Jr. 1987. Water resources data — Colorado
water year 1986. Vol. 1. Missouri River Basin,
Arkansas River Basin, and Rio Grande River Basin
water data report CO-86-1. United States Geological
Survey, Department of the Interior.

. 1988. Water resources data — Colorado water year

1987. Vol. 1. Missouri River Basin, Arkansas River
Basin, and Rio Grande River Basin water data report
CO-87-1. United States Geological Survey,
Department of the Interior.

. 1989. Water resources data — Colorado water year

1988. Vol. 1. Missouri River Basin, Arkansas River
Basin, and Rio Grande River Basin water data report
CO-88-1. United States Geological Survey,
Department of the Interior.

. 1990. Water resources data — Colorado water year

1989. Vol. 1. Missouri River Basin, Arkansas River
Basin, and Rio Grande River Basin water data report
CO-S9-1. United States Geological Survey,
Department of the Interior.

Radford, D. S., and R. Hartland-Rowe. 1971. The life
cycles of some stream insects (Ephemeroptera,
Plecoptera) in Alberta. (Canadian Entomologist 103:
609-617.

Sheldon, A. L. 1972. Comparative ecology o{ Arcijnoptenjx
and Diura (Plecoptera) in a California stream. Archiv
fiir Hydrobiologie 69: 521-546.

Sheldon, A. L., and S. Jewett. 1967. Stonefly emergence
in a Sierra Nevada stream. Pan-Pacific Entomologist
43: 1-8.

Short, R. A., and J. V. Ward. 1980. Life cycle and produc-
tion of Skwahi parallela (Frison) (Plecoptera: Perlo-
didae) in a Colorado montane stream. Hydrobiologia
69: 273-275.

Snellen, R. K., and K. W Stewart. 1979. The life cycle
and drumming behavior of Zealeiictra claasseni
(Frison) and Zealeuctra hitei Ricker and Ross
(Plecoptera: Leuctridae) in Te.xas, U.S.A. Aquatic
Insects 1: 65-89.

Stanford, J. A. 1975. Ecological studies of Plecoptera in
upper Flathead and Tobacco rivers, Montana. Un-
published doctoral dissertation. University of Utah,
Salt Lake City.

Stanford, J. A., and A. R. Gaufin. 1974. Hyporheic com-
munities of two Montana rivers. Science 185:
700-702.

Stanford, J. A., and J. V. Ward. 1993. An ecosystem per-
spective of alluvial rivers: connectivity and the
hyporheic corridor. Journal of the North American
Benthological Society 12: 49-60.

Stark, B. P, S. W. Szczytko, and R. W Bau.mann. 1986.
North American stoneflies (Plecoptera): systematics,
distribution, and taxonomic references. Great Basin
Naturalist 46: 383-397.

Stewart, K. W, and B. P Stark. 1988. Nymphs of North
American stonefly genera (Plecoptera). Entomological
Society of America, Thomas Say Foundation 12:
1-460.'

Stewart, K. W, R. L. Hassage, S. J. Holder, and M. W
OswooD. 1990. Life cycles of six stonefly species
(Plecoptera) in subarctic and arctic Alaska streams.
Annals of the Entomological Society of America 83:
207-214.

Szczytko, S. W, and K. W Stewart. 1979. The genus
Isoperla (Plecoptera) of western North America:
holomorpholog\' and systematics, and a new stonefly
genus Cascadoperla. Memoirs of the American
Entomological Society 31.

Ulfstrand, S. 1968. Life cycles of benthic insects in
Lapland streams (Ephemeroptera: Plecoptera:
Trichoptera: Diptera-Simuliidae). Oikos 19:
167-190.

Zar, J. H. 1984. Biostatistical analysis. 2nd edition. Prentice-
Hall, Inc., Englewood Cliffs, NJ. 718 pp.

Received 1 October 1993
Accepted 11 April 1994

POLLINATOR SHARING BY THREE SYMPATRIC MILKVETCHES,
INCLUDING THE ENDANGERED SPECIES ASTRAGALUS MONTH

S. M. Geerl-3, V J. Tepedino^-l, T. L. Griswold^, and W. R. Bowlinl

Abstfl\ct. — Insects visiting flowers of the endangered Heliotrope milkvetch. Astragalus montii, were compared with
those visiting two common sympatric congeners, A. kentrophyta and A. miser, on three sites on the Wasatch Plateau of
central Utah for 2 yr. We recorded 27+ species of bees, most of which were uncommon, visiting the three species. All
three species were primarily visited by native bees of the genera Osmia (15 species) and/or Bombus (4 species). Most
Osmia species visited the three species oi Astragalus indiscriminantly; binnblebees preferred A. miser and avoided A.
montii. Our hypodiesis that A. montii flowers would receive fewer total bee visits and be visited by fewer bee species
than their common congeners was rejected: A. montii was intermediate to the two common species in its attractiveness
to bees. Also rejected was our hypothesis that the greater similarity between A. montii and A. kentrophyta in flower size,
flower moiphology, and microhabitat would be associated with greater similarity of flower visitors than either had with
A. miser The data suggest that, rather than competing with each other for pollinators, the three species of Astragalus
facilitate each other's visitation rates.

Key words: Astragalus, milkvetch, endangered plant, reproduction, pollination, facilitation, bee diversity, conservation,
Fabaceae, Osmia.

Many insects such as dipterans and lepidop-
terans use flowers only as fuel stations (Elton
1966); they collect nectar and burn it as they
search for suitable spots to lay eggs. Such
insects may merely pass through areas where
flowers are sparse. Bees, in contrast, are central-
place foragers (Orians and Pearson 1979) that
must consistently reap profits in both nectar and
pollen, for they forage not simply to under-
wiite their own movements, but to provide food
to rear their progeny as well (Stephen et al.
1969). Because bees are under strong, selective
pressure to be profitable foragers, they are
attracted to dense patches of flowers (Heinrich
1976, 1979, Thomson 1982). Bumblebees, for
example, quickly recognize and exploit partic-
ularly rewarding flower patches (Heinrich
1976, 1979); other bees probably do so also.

Density-dependent foraging behavior by
bees has important implications for certain rare
plants. Rabinowitz (1981) distinguished seven
types of rarity in plants using the following
three criteria: (1) local abundance, (2) habitat
specificity (narrow or wide), and (3) geographic
range (large or small). Those species with both
narrow habitat specificity and small local pop-
ulations (regardless of geographic range) are

sparse and likely to attract foraging bees only
incidentally. We expect such species to be pol-
linator-vulneiable and, therefore, to be highly
self-compatible and perhaps primarily self-
pollinating (Karron 1987). It is less clear
whether plants in other categories of rarity,
especially endemics (Rabinowitz 1981, Kruck-
berg and Rabinowitz 1985), are also pollinator-
vulnerable. Endemics have narrow habitat
specificit)' but may be locally abundant.

One such endemic, the rare Heliotrope
milkvetch, Astragalus montii Welsh, is limited
to a few isolated populations in limestone
gravel outcrops on the Wasatch Plateau of
central Utah at about 3350 m. There it grows
with two common congeners, A. kentrophyta
var. tegetarius (S. Wats.) Dorn, hereafter A.
kentrophyta, and A. miser var. oblongifoliiis
(Rydb.) Cron., hereafter A. miser In all three
species, seed production requires, or is in-
creased by, pollinator visits to flowers (Geer and
Tepedino 1993). Information on the identity
and biology of these pollinators is important,
for A. jnontii occurs on rangelands that are
grazed by domestic livestock and sprayed with
insecticides to control grasshoppers. Successful
management of this rare species requires

^Department of Biolog\. Utah State University, Logan, UT 84322-5305.

2USDA, ARS Bee Biology and Systematics Laboratory, Utah State University, Logan, UT 84322-5310.
^Present address: Wallowa Whitman National Forest, Highway 82, Box 88401, Enterprise, OR 97828.
^Author to whom correspondence should be addressed.

Great Basin Naturalist

[Volume 55

knowledge of how such spraying may affect its
pollinators.

In this report we compared composition
and abundance of pollinator fauna of A. montii
with those of its two sympatric congeners.
Because there may be wide variation in a
species' pollinators between years and sites
(Tepedino and Stanton 1981, Herrera 1990,
Eckhart 1992), we censused pollinators of A.
montii and its congeners for 2 yr at three sites.
We hypothesized that A. montii would (1)
attract fewer individual pollinators, (2) have
lower pollinator species diversity than its two
common congeners, and (3) share more species
of flower visitors with A. kentrophyta than
with A. miser because similarity in plant and
flower size, flowering time, and microhabitat
is greater with the former than with the latter.

Species and Study Areas

All three species of Astragalus are small
perennial herbaceous legumes. A. montii is re-
stricted to three mountaintops on the Wasatch
Plateau in central Utah. Although Isely (1983)
proposed that A. montii be reduced in status
to a variety of A. limnocharis Barneby, it was
listed as endangered under the Endangered
Species Act in 1987 as A. montii and remains
so (Anonymous 1991). Therefore, we refer to
this taxon as A. montii.

A. kentrophijta and A. miser are widespread
species that occur with A. montii at three sites
on two of the mountains; the third mountain is
less accessible and was not included in the
study. A. kentrophyta is widespread and abun-
dant in the Rocky Mountains, mostly between
2280 and 3650 m. A. miser, one of the most
common species of Astragalus in the Rocky
Mountains, is locally abundant from sagebrush
foothills to the spruce-fir belt (Barneby 1989).
The three species co-occur at 3250 to 3350 m
in an Engelmann spruce {Picea engelmannii
Parry)/subalpine fir {Ahies lasiocarpa [Hook.]
Nutt.) community. A. montii and A. kentrophyta
are intenningled in limestone gravel outcrop-
pings where A. miser is found only occasional-
ly. A. miser is most abundant nearby where
soil is deeper and less rocky. A. montii and A.
miser occur at similar local densities on
Heliotrope Mountain (9.3 ± O.l/m^ and 12.6
± 8.3/m2; Geer unpublished data). There are
fewer A. kentrophyta (2.6 ± O.S/m^; Geer un-
published), but individuals cover more ground

than do those of its congeners. The three
species overlap in bloom time for about 3 wk
(Fig. 1).

Heliotrope milkvetch is a subacaulescent
plant 1-5 cm tall that arises from a branched
caudex. Flowers are deep purple with white
wingtips. There may be a dozen to a hundred
or more flowers (7.8 ± 1.5 mm long, N = 10;
Geer unpublished) per plant, two to eight per
raceme (Barneby 1989). It does not appear to
reproduce vegetatively (personal observation).
In 1989 and 1990 A. montii commenced flow-
ering with final snowmelt beginning as early
as June and continuing for about 4 wk until
mid-July (Fig. 1).

The common species A. kentrophyta started
to flower approximately 1 wk before A. montii
and continued to flower through early August.
It is prostrate, with stems that fork repeatedly
and closely to form low convex cushions cov-
ered with small blue-white to puiplish flowers
(6.6 ±1.2 mm long, N = 10; Geer unpublished),
only two per raceme (Barneby 1989).

The other common congener, A. miser, com-
menced flowering 1-2 wk after A. montii and
continued flowering until September. It is taller
(2-20 cm) than A. montii or A. kentrophyta.
Flowers are larger (11.4 ± 1.4 mm long, N =
11; Geer unpublished) and vaiy in number per
raceme (3-15; Barneby 1989) and in color;
flowers may be white, pink, or lavender.

All Astragalus species have papilionaceous
blossoms composed of a showy standard or
banner petal, a keel that protects the joined
stamens and pistil, and two wings that, along
with the keel, typically serve as a landing plat-
form (Kalin Arroyo 1981). To trip A. miser
flowers, bees land on the keel and force their
way under the banner (personal obsei-vation) as
they do for other species of Astragalus (Green
and Bohart 1975, Fliegri and van der Fiji 1979).
Visitors to A. montii or A. kentrophyta spread
the wing petals with their midlegs and take
nectar, or comb pollen from the anthers to
their abdominal pollen baskets with their
forelegs (personal observation). Stylar hairs
(termed a brush mechanism) aid in the collec-
tion of pollen by transporting it from the keel
outward (Kalin Arroyo 1981).

Sexual reproduction by A. miser and A.
kentrophyta requires insects to transfer pollen;
A. montii is capable of unassisted self-pollina-
tion (autogamy). However, fruits produced
autogamously by A. montii may be inferior in

1995]

Pollinators of Sympatric Milkvetches

A. kentrophyta

wind, and no precipitation). Initially, sight
identification of some taxa was attempted so as
to reduce impact on the poHinator community.
It soon became obvious that it was impossible
to identify Osmia and other individuals with-
out laboratory examination. Subsequently, all
flower visitors were collected whenever possi-
ble. Few insects other than bees visited the
flowers.

Diversity of bee visitors to each Astragalus

1 7(5 ^ 30 ^ ^ '^ To 20 species was calculated using Simpson's diver-

J""^ J"iy A"g"=t sity index, D = 1 - Z^j^j (Pi)2, where Pj = the

„. 1 Di J i f u • r proportion of individuals that belong to each

I'lg. 1. Blooming dates tor tliree co-occumng species ot f" '^ ^ ,

Astragalus at the SSH site. Solid line = 1989; dashed line Dee species (Southwood 1978). Simpson S
= 1990. index gives little weight to rare species and

more weight to common ones. Similarity of

,.,,., 1 , , . the bee fauna visiting Astragalus species was

qualitv to tliose produced by geitonogamous .. . i . /-. i ^ •' ■ i -^

^ ' 1 1 11- .• 1 estimated using Lzekanowski s similarity

or xenogamous hand pollinations, or open-pol- . j ^ ^,,7/ ii i \ i m • .^i

,. ^ J* ^ 1 ^ f 1. /M r index: C<. = NJ/(a+b + ...n), where N is the

Imated control treatments (there are fewer , r i 1 • i • i t •

J r -1. J J u r^ number oi plant species being compared, I is

seeds per truit and seeds are smaller; Geer ^i i ri • i i i i

J T- J- inno\ T-u 11 ..L • the number ot bee species shared by those

and lepedino 1993). Ihus all three species , , . , , , \

1111 r.. r . ,.■■...■ plant species, and a, b, etc., are the total num-

probably beneht from insect visitation. f pV . ... , ,

ber or bee species visiting each plant species

Methods (Southwood 1978). C^ is based on species

presence alone. We also calculated Cj, which

Insect visitors were collected for about 3 wk adjusts for the number of individuals per

in 1989 and for 2 wk in 1990 at the following species (Southwood 1978). The indices range

three sites, starting when A. montii was in peak ^om 0 (no similarity) to 1.0 (complete similari-

bloom: the head of Mill Stream on Ferron ^v)- They were calculated between pairs of

Mountain (HMS), south side of Heliotrope species and among all three species.
Mountain (SSH), and east end of Heliotrope Probable pollinators of the three Astragalus

Mountain (EEH). In 1990 collections from all species were ascertained by examining flower

three Astragalus species were made only at visitors and recording areas of their bodies on

the SSH site because only two insect collec- which pollen was found. Specimens were then

tors were available instead of four, as in 1989. relaxed and pollen was removed using an

We concentrated on the SSH site in 1990 to insect pin or by dabbing it with acid-fuchsin

make the number of collector hours there gel (Beattie 1971). The pollen was placed on a

equivalent to the 1989 effort. In 1990 visitors glass slide with acid-fuchsin gel, warmed until

to A. kentrophyta were collected at the SSH liquid, and a cover slip applied (modified from

and HMS sites, and visitors to A. miser were Faegri and Iverson 1964). One slide per leg or

collected at the SSH and EEH sites. Following two slides per abdomen were made for each

are approximate direct distances between sites: insect. All slides were viewed at lOOX magni-

HMS to SSH = 3.6 km, HMS to EEH = 2.4 fication and the pollen compared to a pollen

km, and EEH to SSH = 1.2 km. reference collection of species in bloom at the

Pollinators were collected with a standard study sites,
butterfly net and killed in cyanide jars. Cold

temperatures, strong winds, and frequent pre- Results

cipitation (snow and rain) prohibited pollina-
tors from flying during all but brief windows Bees were scarce at the study sites in both

of calm, sunny weather, so opportunistic collec- years (Table 1, Appendices I, II). Bee visitors

tion was necessary to ensure an adequate sam- per plant species ranged from about 0.5 to just

pie size. Collections were made from all three over 3 per hour, a small number considering

species contemporaneously, whenever weath- that many flowers of each species were being

er permitted (i.e., temperatures >13°C, little monitored. Bee numbers were higher in 1990;

Gkeat Basin Naturalist

[Volume 55

Tablf. 1. Number of person hours spent collecting and number ot bee individuals collected or observed visiting flow-
ers of Astragalus montii (Asmo), A. kentrophyta (Aske), and A. miser (Asmi) at three sites on the Wasatch Plateau in 1989
and 1990. SSH, EEH = south and east side Heliotrope Mountain, respectively; HMS = head of Mill Stream, Perron
Mountain.

SSH

EEH

HMS

Asmo

Aske

Asmi

Asmo

Aske

Asmi

Asmo

Aske

Asmi

1989

Hours

Individuals

Individuals/hour

1.2

1.1

1.0

0.8

0.5

0.6

0.7

1.1

Species

1990

Hours

—

Individuals

—

Individuals/liour

1.9

0.5

2.3

3.3

2.0

—

1.3

—

2.0

Species

—

when categorized by site and Astragalus species
visited, six of seven categories had more indi-
viduals per hour in 1990 than in 1989.

The initial hypothesis, that A. montii would
have fewer individual flower visitors than
would its common congeners, received little
support (Table 1, Appendices I, II). In 1989
there was little difference among species in
visitors per person hour at SSH. At EEH A.
montii flowers were visited more often than
the other species. Conversely, at HMS A.
montii flowers received the fewest visits. In
1990 comparisons of number of visitors among
all three Astragalus species could be made
only at the SSH site where A. montii had an
intermediate number of visitors per hour. At
EEH, A. montii again had more visits per hour
than A. kentrophyta, and at HMS it had fewer
visits per hour than A. miser

The prediction that species richness and
species diversity of bees visiting the three
Astragalus species would be lowest for A. montii
was also provisionally rejected. The number of
species captured on A. montii commonly ex-
ceeded those captured on the other species,
both when more hours were spent collecting
from A. montii than the other species (1989
SSH) and when collecting hours were equal
(1990 HMS; Table 1). Only once, when fewer
hours were spent collecting on A. tnontii than
on the other Astragalus species (1989 HMS),
was A. montii visited by the fewest species of
bees. When all sites were considered, total
number of species collected on A. montii in
1989 exceeded those captured on A. kentro-
phyta and equaled those captured on A. miser
(Table 2). In 1990 more species were caught

visiting A. montii than the other two species,
but this difference is probably because we col-
lected at three sites for A. montii but at only
two for each of the other two species.

Calculations using species diversity, D',
also failed to yield expected trends (Table 2).
In 1989 diversity of visitors to flowers of A.
montii was very similar to diversity recorded for
A. kentrophyta and A. miser Comparisons for
1990 are more tenuous because of the differ-
ences among species in number of sites sam-
pled. However, diversity of flower visitors was
highest for A. miser and similar for A. montii
and A. kentrophyta. Diversity in 1990 was
generally lower than in 1989, although num-
ber of individuals captured was greater.

The most frequent visitors to these Astra-
galus species in both 1989 and 1990 were
Osmia bees (Table 3). For the small-flowered
A. montii and A. kentrophyta, in both years
>70.0% of all visitors were Osmia bees. Only
for A. miser in 1990 did the percent Osmia
visitors drop below 50%. A. miser was more
frequently visited by bumblebees, especially
at SSH. The abundance of bumblebees caused
SSH to have the lowest percentage of Osmia
individuals recorded at any site in both years.
Even so, Osmia bees were always more than
60% of the total flower visitor fauna recorded
in any site -year.

Because of greater similarities in flower size,
color, and microclimate, we expected A. montii
and A. kentrophyta to have more visitors in
common than either did with A. miser This
was not true in either year. The three pairings
of Astragalus did not differ much in the num-
ber of bee species they shared, though results

1995]

Pollinators of Sympatric Milkvetches

Table 2. Number of individuals, number of species, and species diversity (D) of bees found visiting three species of
Astragalus at three sites on the Wasatch Plateau. In 1989 collections were made for each species at all three sites; in
1990 collections were made at all sites for A. montii, but at only two sites for the other two species. For comparative pur-
poses, collection data for the latter two species are shown in 1989 for all three sites and for only the two sites collected at
in 1990. D = Simpson's diversity index.

Astragalus

Individuals

.Species

species

3 sites

2 sites

3 sites

sites

3 sites

2 sites

1989

montii

—

0.87

kentrophyta

0.79

0.81

miser

0.88

0.87

1990

montii

113

—

0.62

kentroplujta

—

0.60

miser

—

0.79

'In 1989 onK- indiviuals that were collected were used in calculations, because uncaptured Otmia indi\iduals were not identifiable to species.

Table 3. Percent visitors that were Osmia bees to the flowers of three Astragalus species (abbreviations as in Table 1).
Data showai grouped by species across sites, and by site across species, for 2 yr. For comparative purposes, 1989 data are
shown in entirety' (3 sites or 3 species) or only for the 2 sites or 2 species sampled in 1990.

Asmo

Aske

Asmi

SSH

EEH

HMS

— Across sites - - -

- - Across species - -

1989

3 (sites/species)

88.9

71.8

73.7

62.3

88.1

76.5

2 (sites/species)

—

78.6

64.3

—

85.7

87.0

1990

93.8

74.2

47.5

62.6

87.5

95.0

varied somewhat with year and with index used
(Table 4). In 1989 the three pairings of Astraga-
lus species had about the same number of bee
species in common. In 1990 A. miser and A.
montii had about twice the number of species
in common as did the other pairings. Neither
coefficient of similarity, C^ or Cj, consistently
supported the hypothesis; in 1989, but not
1990, C^ and Cj were highest for the A. mon-
tii-A. kentrophyta comparison.

Many bees visiting Astragalus flowers car-
ried pollen on their bodies: 43% of the bees
captured, primarily females of the genus Osmia,
had been collecting pollen. Pollen loads com-
prised primarily Astragalus pollen (all means
>80%; Table 5). It is unknown whether loads
commonly contained more than one species of
Astragalus because pollen grains could not be
distinguished to species with the light micro-
scope.

Our observations of foraging bees suggest
some interspecific movement. In 1989 few
Osmia individuals flew between A. montii and
A. miser or A. kentrophyta; of 74 interplant

movements only two were interspecific. In
1990, 4 of 21 observed interplant movements
were between species. Interspecific visits
occurred most commonly where species grew
intermingled.

Discussion

Two hypotheses make predictions about the
abundance and diversity of visitors to the flow-
ers of rare plants. For entomophilous plants,
Levin and Anderson (1970), Straw (1972), and
Karron (1987) proposed that pollinators should
be more flower constant to abundant plant
species than to rare ones, that this differential
flower constancy would result in more suc-
cessful reproduction by "majority" species
than by "minority" species, and that over time
minority species would become extinct because
of dwindling recruitment or would evolve
some method of self-reproduction (Levin 1972).
A corollar)' of this hypothesis is that both the
number and diversity of visitors to the flowers
of rare plants should be lower than they are to
abundant ones.

Great Basin Naturalist

[Volume 5."

Table 4. Number of Ix-e species (S) collected on each A.str(i^alii.s species, and number of species shared (C) and simi-
larity' indices for each j^airin^ for each year. C^ = Czekanowski s similarity index for bee species presence-absence; Cj
= index weighted by individuals captured.

Astragalus
species pair

1989

1990

montii

miser

iiioiitii

kcntrophijta

kentrophyta

miser

All three species

0.50

0.55

0.45

0.35

0.34

0.43

0.27

13
12

0.56 0.37

0.40 0.35

0.32 0.53

0.28 0.30

In conticLst, the facilitation hypothesis (re-
viewed by Rathcke 1983) predicts that rare
species growing with attractive, more abun-
dant species may actually reproduce more
successfully because the latter draw many
more pollinating insects into the area than
would otherwise be present. If so, rare and
abiuidant sympatric species should have simi-
lar visitor diversity, and visitor abundances
should reflect respective frequencies of the
plants. This study indirectly assessed the
importance of facilitation and competition. A
direct assessment is difficult because (1) the
experiments necessar>' to distinguish between
alternatives cannot be conducted when the
"plant protagonist" is protected by the Endan-
gered Species Act; and (2) A. montii did not
occur in the absence of its congeners on our
study sites, so visitation rates of "facilitated"
and "unfacilitated" populations could not be
compared.

Our results supply consistent, though indi-
rect, support for the facilitation hypothesis.
Except for bumblebees, which foraged almost
exclusively from large-flowered A. miser, bees
did not discriminate against A. montii but
rather seemed to treat all three Astragalus
species as one taxa. First, A. montii did not
consistently attract fewer visitors per hour
than did the other species. Indeed, visitation
rates to A. montii were higher than to the
other species in three of six site-years (Table 1).
Second, neither species richness nor species
diversity of pollinators was consistently lower
for A. montii than for the other species (Table 2).
In fact, an ecjual or greater nimiber of species
visited A. montii than visited the others in
both years. And finally, bees were observed
moving between species on individual foraging

trips. Gross (1992) also reported that bees for-
aging on closely related legumes commonly
moved between species. Thus, there was no
detectable rare species disadvantage and no
evidence that endemics, at least those growing
in close proximity to abundant congeners, are
pollinator- vulnerable.

The shared microhabitat and similarities in
flower size and morphology of A. montii and
A. kentrophyta led us to expect that facilitation
would be more likely between these two species
and, therefore, that they would have more visi-
tors in common than either would with A. miser
For example, Thomson (1978, 1981, 1982) found
that, in two-species mixtures, the degree of
intermingling and the similarity in structure
and appearance of congeners' flowers deter-
mined the importance of competition and
mutualism. The more similar the flowers, the
more likely that visitation rates to rare species
would be bolstered by the presence of abun-
dant species and the more likely that visitors
would be shared. Our data supported this
expectation for 1989 but not for 1990 (Table 4).
In 1990 C^ for the A. montii-A. kentrophyta
comparison was intermediate to the other
comparisons; for C; it was lower than the other
comparisons. Thus, results for the similarity
analyses also tend to support the hypothesis
that most bees do not distinguish among these
Astragahis species when foraging, and that the
Astragalus species tend to facilitate each
other's visitation rates.

Only bumblebees seem uninfluenced by
Astragalus flowers in the aggregate. They
clearly preferred flowers of A. miser and
avoided those of the other Astragalus species.
Flowers of A. miser are large, probably more
rewarding, and provide a landing platfonii from

1995]

Pollinators of Sympatric Milkvetches

Table 5. Percent Astragalus pollen grains in pollen loads, and location of pollen loads carried by bees collected on
three Astragalus species at three sites on the Wasatch Plateau in 1989 and 1990.

Astragalus
species

montii

kentrophyta

Number
po'len lo

•of
ads

Mean %

Astragalus

pollen ( + SE)

Location of
pollen

Abdomen Legs

82 ±4

42 3

90 ± 1

19 —

95 ± 1

5 —

which large, energy-demanding bumblebees
can readily forage. Other large-flowered Astra-
galus species also attract numerous large bees
such as bumblebees {Bombiis spp.) and antho-
phorids (Green and Bohart 1975, Sugden 1985,
Karron 1987). In comparison, bumblebees
seemed unable to land on the small, weakly
supported A. montii flowers which are borne
above the foliage; they did occasionally exploit
the tiny A. kentrophyta blossoms while perched
on the foliage of that cushion plant.

Factors other than flower abundance can
influence the flight path of foraging bees.
Because bees are central-place foragers (Orians
and Pearson 1979), travel time and energy
expended between flower patches and nest
are also important. Thus, bees may patronize a
flower patch because of its proximity to their
nest, even though flowers are more abundant
elsewhere. For example, Osniia bees mated
and nested at the sheltered EEH site where
relatively few A. kentrophyta or A. miser plants
grew; the population of A. montii was small
but dense. Nevertheless, bees visited flowers
at least as frequently at EEH as at the other,
more flower-rich, sites (Tables 1, 2). Thus, suit-
ability of nesting habitat at EEH, rather than
Astragalus flower abundance, may best account
for the abundance of bees there. The effect of
wild bee nesting sites on seed production of
surrounding vegetation is poorly studied and
warrants additional attention.

Rigorous subalpine communities of the
Wasatch Plateau, with frequent high winds,
thunderstorms, and below-freezing tempera-
tures during the blooming season, support a
surprisingly rich bee fauna. In 2 yr we collected
27+ bee species foraging on Astragalus flow-
ers during 2-3 wk (Appendices I, II). These
bees are invaluable pollinators of native plants
both rare and common. Their welfare must also
be considered in management plans for rare

plants. Land managers must eliminate losses
of bees to insecticide applications made for
rangeland grasshoppers and minimize physical
damage to nest sites. The present insecticide-
free buffer zone (currently 4.8 km) around rare
plant populations should continue to be main-
tained. Areas where bees nest in soil should
also be protected from livestock trampling,
off-road vehicle use, and foot traffic (Sugden
1985). Such diversity, comparable to or greater
than that of other subalpine areas in North
America (Moldenke and Lincoln 1979), is to
be marveled at and preserved.

Acknowledgments

We are grateful to the many people who
assisted in this study. Etta Sechrest and Mike
Cram were reliable field and laboratory assis-
tants. John Healey, Don Riddle, and Bob
Thompson of the U.S. Forest Service and Lairy
England, U.S. Fish and Wildlife Service, helped
in a variety of ways, fi^om locating plant popu-
lations to putting a roof over our heads. The
manuscript was constructively reviewed by M.
Barkworth, K. Harper, and E. Sugden. This
study was funded as part of the APHIS Grass-
hopper IPM Project. It is Journal Paper #4436
from the Utah Agricultural Experiment Station.

Literature Cited

Anonymous. 1991. Endangered and threatened wildlife
and plants; .50 CFR 17.11 and 17.12, July 15, 1991.
Publication unit, U.S. Fish and Wildlife Service,
Washington, DC.

Barneby, R. C. 1989. Fabales. Pages 12-167 in A.
Cronquist, A. H. Holmgren, N. H. Holmgren, J. L.
Reveal, and P K. Holmgren, eds., Intennountain flora
3, Part B. New York Botanic Garden, Bronx, NY.

Beattie, a. J. 1971. A technique for the study of insect-
borne pollen. Pan-Pacific Entomologist 47: 82.

EcKHART, V M. 1992. Spatio-temporal variation in abun-
dance and variation in foraging behavior of the polli-
nators of gynodioecious Phacelia linearis (Hydro-
phyllaceae).' Oikos 64: 573-586.

Great Basin Naturalist

[Volume 55

Elton, C. S. 1966. The pattern of animal eoiiiniunities.
Methuen, London. 432 pp.

F.\EGRI, K., AND J. IvERSON. 1964. Textbook of pollen
analysis. Hafrier Co., NY. 237 pp.

Faegri, K., and L. VAN DER FiJL. 1979. The principles of
pollination ecology. 3rd revised edition. Perganion
Press, Oxford. 244 pp.

Geer, S. M., and V. J. Tepedino. 1993. Breeding .systems
of the rare Heliotrope milkvetch {Astragalus montii
Welsh: Fahaceae) and two conmion congeners. Pages
334-344 ill R. Sivinski and K. Lightfoot, eds., Proceed-
ings of the Southwestern Rare and Endangered Plant
Conference. New Mexico Forestry and Resources
Conservation Division, Santa Fe.

Green, T. W., and G. E. Bohart. 1975. The pollination
ecology of Astragalus cibarius and Astragalus uta-
hensis (Leguminosae). American Journal of Botany
62: 379-386.

Gross, C. L. 1992. Floral traits and pollinator constancy:
foraging by native bees among three sympatric
legumes. Australian Journal of Ecology 17: 67-74.

Heinrich, B. 1976. Foraging specializations of individual
bumblebees. Ecological Monographs 46: 105-128.

. 1979. "Majoring" and "minoring" by foraging

bumblebees, Bombus vagans: an experimental analy-
sis. Ecology 60: 245-255.

Herrera, C. 1990. Daily patterns of pollinator activity,
differential pollinating effectiveness, and floral re-
source availability, in a summer-flowering Mediter-
ranean shrub. Oikos 58: 277-288.

Isely, D. 1983. New combinations and two new varieties
in Astragalus, Orophaca, and Oxytropis (Legmnino-
sae). Systematic Botany 8: 422.

Kalin Arroyo, M. T. 1981. Breeding systems and pollina-
tion biology in Leguminosae. Pages 723-769 //; R. M.
Polhill and P H. Raven, eds.. Advances in legume
systematics. Part 2. Royal Botanic Gardens, Kew,
UK.

Karron, J. D. 1987. The pollination ecology of co-occur-
ring geographically restricted and widespread species
of Astragalus (Fabaceae). Biological Consei-vation 39:
179-193.

Kruckeberg, A. R., and D. Rabinowitz. 1985. Biological
aspects of endemism in higher plants. Annual Review
of Ecolog\' and Systematics. 16: 447-479.

Levin, D. A. 1972. Competition for pollinator sei-vice: a
stimulus for the evolution of autogamy. Evolution
26: 668-674.

Levin, D. A., and W. VV. Anderson. 1970. Competition
for pollinators between simultaneously flowering
species. American Naturalist 104; 455-467.

Moldenke, a. R., and P Lingoln. 1979. Pollination ecol-
ogy in montane Colorado: a community analysis.
Phytologia 42: 349-379.

Orians, G. H., and N. E. Pearson. 1979. On the theory of
central place foraging. Pages 155-177 in D. J. Horn,
G. R. Stairs, and R. D. Mitchell, eds., Analysis of
ecological systems. Ohio State University Press,
Columbus.

Rabinowitz, D. 1981. Seven fomis of rarity. Pages 20.5-218
in H. Synge, ed.. The biological aspects of rare plant
conservation. Wiley, New York, NY.

R.ATHGKE, B. 1983. Competition and facilitation among
plants for pollination. Pages 305-329 in L. A. Real,
ed., Pollination biology. Academic Press, New York,
NY

Southwood, T. R. E. 1978. Ecological methods. 2nd edi-
tion. Chapman and Hall, London. 524 pp.

Stephen, W. P, G. E. Bohart, and P F Torchio. 1969. The
biology and external moiphology of bees. Agricultural
Experiment Station, Oregon State University,
Coi"vallis. 140 pp.

Str.\w, R. M. 1972. A Markov model for pollinator con-
stancy and competition. American Naturalist 106:
597-620.

Sugden, E. A. 1985. Pollinators of Astragalus monoensis
Barneby (Fabaceae): new host records; potential
impact of sheep grazing. Great Basin Naturalist 45:
299-312.

Tepedino, V. J., and N. L. Stanton. 1981. Diversity and
competition in bee-plant communities on short-grass
prairie. Oikos 36: 35-44.

Thomson, J. D. 1978. Effect of stand composition of insect
visitation on two-species mixtures of Hieracium.
American Midland Naturalist 100: 431-440.

. 1981. Field measures of constancy in bumble-
bees. American Midland Naturalist 105: 377-380.

. 1982. Patterns of visitation by animal pollinators.

Oikos 39: 241-250.

Received 29 April 1993
Accepted 2 June 1994

Appendi.x I. Species of bees collected and obsei-ved visiting flowers of A. montii (Asmo), A. miser (Asmi), or A. kentro-
phyta (Aske) at three sites in 1989. Entries represent number of males/females collected. Obsei-vations are in parenthe-
ses. Site abbreviations as in Table 1.

SSH
1-21 June

EEH
14-25 June

HMS
1-J^22 June

Bee species

Asmo

Aske

Asmi

Asmo

Aske

Asmi

Asmo

Aske

Asmi

Andrenidae
Andrena transnigra Vier.
Andrena spp.

0/1

(1)

1995]

Pollinators of Sympatric Milkvetches

Appendix I. Continued.

SSH
1-21 June

EEH
14-25 June

HMS
14-22 June

Bee species

Asino

Aske

Asmi

Asnio Aske Asmi

Asnio Aske

Asmi

API DAE

Boinbiis hifarius Cr.
Boiiibiis Jlavifrons Cr.
Botnbtts hiintii Greene
Boinbiis nevadensis Cr

0/1

(0/2)
(0/2)
0/1

0/2

Halictidae
Evtjlaeus niger (Viereck)

Megachilidae
Anthidium temiiflorae Ckll.
Megachile spp.
Osmia cijanopoda Ckll.
Osinia htirdii White
Osinia longula Cr.
Osmia nigrifrons Cr.
Osinia atf. nigrifrons
Osinia paradisica Sanh.
Osinia penstcinonis Ckll.
Osinia pikei Ckll.
Osmia piisilla Cr
Osmia sladeni Sanh.
Osmia sladeni &/or alpestris
Osmia taniwri Sanh.
Osmia spp.

0/1

(2)1/0
(1/0)

0/1

0/3

0/1

(1/0)

(1) (1)1/2(1)

1/2(2)

1/0

0/1

0/4

1/3

(5/9)

2/0

(1/4) (0/3)

0/2

0/1
0/3

0/1

1/0

2/2

0/1

1/0

4/0

3/0

2/0

0/2

0/3

0/2

0/1

0/2

0/5

1/2

0/1

(8/9)

(1/1)

(1/3)

(1/1)

(0/1)

Appendix H. Species of bees collected and observed visiting flowers of A. montii (Asmo) at three sites and A. miser
(Asmi) and A. kentrophijta (Aske) at two sites each in 1990. Entries represent number of males/females collected.
Observations are in parentheses. Site abbreviations as in Table 1.

SSH
19 June-4 July

EEH
19-29 June

HMS
21-29 June

Visitor

Asmo Aske Asmi

Asmo Aske

Asmo Asmi

Andrenidae
Andrena nigrihirta (Ashm)
Andrena transnigra Vier.

0/1

Apidae
Apis mellifera L.
Bombus hifarius Cr
Bombus flavifrons Cr
Bombus hunt a Greene
Bombus nevadensis Cr.

0/1

0/1(5)
0/1

0/3(6)
0/3(5)

Megachilidae
Anthidium temiiflorae Ckll. 1/0(2)

Hoplitis fidgida Cr.
Megacile melanophaea Smith
Megachile perihirta Ckll. 1/0

2/0

2/0
1/0

0/1(3)
3/0

1/0

Great Basin Naturalist

[Volume 55

Appf.ndix II. ("ontinued.

SSH

EEH

HMS

19 J

une-4 July

19-

-29 June

21-

-29 June

Visitor

Asmo

Askc

Asmi

Asmo

Aske

Asmo

Asmi

Megachilidae (continued)

Osmia lon^ida Cr.

2/0

Osmia montana Cr.

1/0

Osmia afF. nig.n,frons

0/1

Osmia paradisica Sanh.

1/0

0/2

3/0

1/2

1/0

Osmia penstemonis Ckll.

0/1

Osmia pusilla Cr.

0/1

Osmia sculleni Ckll.

2/0

1/0

Osmia sladeni Sanh

19/13

4/0

1/0

8/16

7/8

3/6

1/21

Osmia suhawstralis Ckll.

4/0

1/0

Osmia tanneri Sanh.

9/2

1/0

9/1

1/0

0/1

FACTORS AFFECTING SELECTION OF WINTER FOOD AND
ROOSTING RESOURCES BY PORCUPINES IN UTAH

Dave Strickland -2, Jerran T. Flinders 1'3^ and Rex G. Gates ^

Abstract. — Ecological and phytochemical factors potentially affecting winter dietar)' discrimination by porcupines
{Erethizon dorsatiim) in tlie mountain brush zone of Utah were studied. Porcupines utilized gaml^el Oiik {Quercits gamhelii)
as their primary winter food and roosting resource. Big-tooth maple [Acer grandidentatum) was the most conmion tree
species in the study area but was rarely utilized by porcupines. Conifer species were used as a food and roosting
resource significantly less often than they occurred in the study area, despite themial advantages provided by their rela-
tively dense canopies. Oak feed trees were successfully separated from conifer feed trees by discriminant analysis 100%
of the time. Oak trees were correctly classified as feed and nonfeed trees 71% of the time. Gambel oak contained higher
amounts of crude protein, fiber, and tannins, but was lower in ether extract fractions and fatty acid content than conifers.
A layer of adipose tissue used as an energy reserve by porcupines may have relaxed energy intake demands sufficiendy
to permit them to concentrate on a diet of oak tissue, which is high in protein, rather than a high-fat conifer diet. A diet
relatively high in protein may have facilitated digestion of food material high in fiber. Temperature did not affect selec-
tion of tree species used for roosting. Rock and snow caves were utilized infrequently and the study population ranged
widely. Three of 15 study animals were eaten by predators.

Keij words: porcupine, Erethizon dorsatum, gambel oak, Quercus gambelii, dietary selection, mountain brush
predation.

'.one.

Porcupines {Erethizon dorsatum) roost and
feed in canopies of deciduous trees and shrubs
for extended periods during winter in much of
western North America (Oveson 1983, Craig
and Keller 1986, Sweitzer and Berger 1992).
Apparent localized interspecific and intra-
specific preferences for food and shelter
resources by porcupines imply that chemical
and/or physical advantages are available to
them. Further, since snow caves, rock dens,
and cover in canopies of coniferous tree
species likely offer increased thermal advan-
tages in the form of energy savings to porcu-
pines (Clarke and Brander 1973, Roze 1987,
1989), their dependence on a deciduous food
and roosting resource (which does not offer
those advantages) further strengthens the
implication that chemical and/or physical
selective advantages are realized by dietary
selection. Predator avoidance may also be an
important force in food and roost tree selec-
tion. The objective of this research was to
investigate physical, phytochemical, and eco-
logical agents involved in selection of gambel
oak by porcupines in south central Utah.

Study Area

The study was conducted in the mountain
brush zone near the mouth of Spanish Fork
Canyon in north central Utah. Elevations at
the study site range from 1650 to 2075 m. The
general exposure is northern, and terrain is
steep. Overstory woody vegetation is dominated
by gambel oak {Quercus gambelii) and big-
toodi maple {Acer grandidentatum). Aspen {Pop-
ulus tremidoides), chokecherry {Primus virgini-
ana), Douglas fir {Pseudostuga menziesii), white
fir {Abies concolor), and mountain maple {Acer
glabrum) are also represented in the woody
flora. The climate in Spanish Fork Canyon
during the winter of 1984-85 was not atypical.
Data from the Spanish Fork U.S. Climatological
Station, located approximately 5.5 km from
the study site, indicate that temperatures were
slightly colder and precipitation was slightly
higher than average (U.S. Climatological Data
for Utah 1984-85). Coyote {Canis latrans) and
mountain lion {Felis concolor) tracks were fre-
quently encountered in the study area. Private
access into the study area allowed observation

^Department of Botany and Range Science, Brigham Young University, Provo, UT 84602.
^Present address: USDA Forest Service, Pleasant Grove, UT 84062.
■'Address correspondence and reprint requests to this author

Great Basin Naturalist

[Volume 55

of a porcupine population relatively free from
human disturbance.

Methods

Fieldwork

We conducted fieldwork from late Decem-
ber 1984 through April 1985, at which time
the study population had shifted from a diet of
inner bark (phloem and cambium) of woody
vegetation to herbaceous vegetation. The study
area was systematically searched by researchers
on snowshoes. Study animals were captured
by hand, usually while they were still in tree
canopies. This was accomplished by grasping
distal guard hairs at the posterior end of the
tail between thumb and forefinger and pulling
the tail taut. The captured animal was then
secured by grasping the tail with the free hand
using a backward stroking motion to flatten
the quills. Fifteen porcupines, 10 females and 5
males, were instrumented with radio transmit-
ter collars (Telonics, Inc.). Animals were located
daily by triangulation, and visual sightings were
made on each animal approximately weekly.

Percent occurrence of woody species was
calculated from point-quarter measurements
using the feed/roost tree as the center point
(Cottam and Curtis 1956). Percent occurrence
of woody species vs. percent utilization of each
feed tree species was compared using chi-
square analysis to test whether feed tree selec-
tion was random. Diameter at breast height
(dbh), species, and distance from the feed tree
center point were recorded for the nearest
woody stem in each quadrant. Point-quarter
measurements were repeated using the near-
est neighbor nonfeed tree of the same species
as the center point. Tissues from feed and
nonfeed trees were collected to investigate
possible differences in chemical makeup.
Tissue samples from feed trees were collected
where fresh bark removal indicated the roost-
ing animal had foraged. Samples from nearest
neighbor nonfeed trees were taken from
branches at the same height and with a diame-
ter similar to those from corresponding feed
trees. Bark samples were frozen and analyzed
for dietary components. Results from those
analyses reasonably approximated values
reported for gambel oak (Smith 1957, Kufeld
et al. 1981, Welch 1989). Location, slope,
aspect, snow depth, and climatic conditions
were recorded at each feed tree site. High and

low temperature readings were taken dail>' at
an elevation of 1597 m, as well as from the
Spanish Fork climatological station.

Laboratory and Statistical Methods

Tissues from feed and nonfeed trees were
analyzed for protein and phosphorus using the
auto analyzer semiautomated method #12 for
feeds (Horwitz 1980). Calcium, magnesium,
potassium, and sodium content were deter-
mined by the atomic absorption method #2
for plants (Horwitz 1980). Sulphur content
was determined by a wet-ash process using
nitric and perchloric acid. Crude fiber was
determined by the acid detergent fiber and
lignin #21 method (Hoi-witz 1980). An evalua-
tion of crude fat was made using the direct
method (Hoi-witz 1980) on a Lab Con soxlet
extractor. A limited number of tissue samples
were analyzed on a Hewlett Packard model
5995 gas chromatograph/mass spectrometer
(GCMS) for fatty acids and terpenes. Tannin
content was measured by the radial diffusion
method (Hagerman 1987) with quebracho tan-
nin being the standard, and by astringency
(Gambliel et al. 1985). Soluble carbohydrates
were determined according to daSilveira
(1978). Urine samples of captive porcupines
on a strict diet of gambel oak were analyzed
for calcium and phosphorus content when lab-
oratory results indicated the Ca/P ratio in the
tissue of food materials was greater than ex-
pected. Eight oak tissue samples were chosen
at random and retested for calcium and phos-
phorus content according to Horwitz (1980)
on a Beckman DU-30 spectrophotometer

Differences between oak, white fir, and
Douglas fir feed and nonfeed trees were statis-
tically analyzed to help discern foraging pat-
terns used by instrumented porcupines.
Chemical and ecological factors were evaluat-
ed for between-species differences using two-
sample t tests, and for within-species differ-
ences with paired t tests (Minitab 1982).
Statistical results are reported at the p < .05
and p < .1 levels. Chi-square analysis was
used to determine if utilization of feed tree
species by porcupines differed from the ex-
pected. Discriminant analysis using backward
elimination and forward selection (SAS 1985)
was used to determine chemical and ecologi-
cal factors that best discriminate between tree
species, and between feed and nonfeed trees
of the same species.

1995]

Wintering Porcupines in Gambel Oak

Table 1. Mean values for factors tested for possible effects on porcupine herbivory.

Oak(l)

White fii

:• (2)

Douglas

fir (3)

N on feed

Feed

Nonfeed

Feed

Nonfeed

Feed

tree

**„ = 46

n =46

n =3

« =3

n = 7

n =7

Distance from conifer (m)

—

207

—

Distance to feed tree, same sp. (m

0 -

3582

—

5431

—

.377

Wind speed (mph)

—

5.53

—

3.7

—

9.71

Slope (%)

—

33.53

—

36.5

—

42.41

Elevation (m)

—

17792

—

19371-3

—

16802

Dbh (cm)

*13.2

16.52-3

*25.4

40.41

33.3

34.51

Crude fiber (%)

43.3

44.23

43.6

48.03

42.4

40.11-2

Protein {%)

4.9

5.02-3

4.0

4.21

4.0

3.91

Phosphorus {%•)

0.038

0.039

0.087

0.064

0.038

0.042

Ether extract fractions (%)

9.0

9.12-3

15.7

12.71-3

16.5

I8.9I-2

Water

41.0

39.52.3

46.1

49.71

53.4

50.71

Potassium (%)

0.39

0.393

0.36

0.31

0.16

0.191

Calcium (%)

2.7

2.73

2.8

2.7

1.7

1.71

Magnesium (%)

0.137

0.1422-3

0.083

0.0921-3

0.068

0.0651-2

Sodium (ppm)

51.1

54.0

53.7

60.0

71.6

58.0

Sulfur (%)

0.20

0.19

0.70

0.48

0.14

0..35

4.7

4.73

4.7

4.73

4.3

4.41-2

Tannins (radius in cm)

30.5

29.62.3

17.5

17.91-3

26.2

25.02-3

***Astringency (mg/g fw)

85.8

83.0

*48.7

66.3

104.3

95.8

Sodium salts (%)

2.9

3.0

3.3

2.9

3.4

Soluble carbohydrates {%)

16.02

16.20

17..33

16.,58

FA (GCMS count units)

827,905

399,239

—

2,609,969

—

1,2.59,531

Superscript values indicate differences between species at thep < .1 level or less. 1 = oak, 2 = white fir, 3 = Douglas fir.

*V'alues different between feed and nonfeed trees of the same species at or below p < I.

** Multiple locations in the same tree responsible for different n values used in calculations of cbemistn, and climatic data. Climatic data n vah

as reported in Table 3.

*** Not comparable across species boundaries.

n \alues for factors below dashed line not as reported for rest of column. Not statistically comparable due to smaller sample size.

i are the same

Results

Oak and white fir feed trees were larger than
nonfeed trees of the same species {p < .05,
Table 1). Herbivory by porcupines in decidu-
ous species occurred in the canopies of large
trees or in shrubs where branch diameters
were relatively small. In coniferous species
herbivory was also concentrated in the canopy
rather than on the tree bole. Only two instances
of chipping bark off the bole to expose the
inner bark were noted in our study, both on
deciduous tree species. There were no trends
correlating calendar date or temperature to
selection of feed tree species. Douglas fir feed
trees contained greater amounts of crude pro-
tein than Douglas fir nonfeed trees (p < .05).
Crude protein content of both conifer species
was less than that of oak trees (Douglas firp <
.05, white fir p < .1). Total tannins (as mea-
sured by radial diffusion) were higher in oak
than in conifers (Douglas fir p < .1, white fir p
< .05). Astringency (protein binding capacity)

was not comparable among species but was
greater for white fir feed trees than nonfeed
trees (p < .1). Ether extract fractions were
lower in oak than in conifers (p < .05) and
lower in white fir than Douglas fir (p < .1).

Tissue from Douglas fir contained less
crude fiber than tissue from oak and white fir
(p < .05), and Douglas fir feed trees contained
still less than nonfeed trees (p < .1). Water con-
tent was lower in oak tissue than in conifer tis-
sue (p < .05). Oak contained higher levels of
potassium and calcium than Douglas fir (p <
.05). White fir was also higher than Douglas fir
in calcium (p < .05). Magnesium levels for oak
were greater than for either conifer species (p
< .05). White fir and oak tissue had higher pH
values than tissues from Douglas fir (p < .05).
Oak feed trees were higher in sodium salts
than Douglas fir feed trees (p < .1). Calcium-
phosphorus ratios for feed trees were higher in
oak than in Douglas fir (p < .05). The calcium-
phosphorus ratio for oak is well above accept-
able limits for mineral absoiption by mammals

Great Basin Naturalist

[Volume 55

(Underwood 1966). High calcium-phosphorus
ratios have also been reported by Masslich
(1985) for aspen {Popidus tremuloides) tissue
utilized by beaver. After an independent test
of feed tree tissue confirmed the high ratio,
we tested the mineral content of feces and
urine from captive porcupines on an oak diet.
Calcium-phosphorus ratios from fecal material
were 10:1, while ratios from urine were
approximately 221:1.

Tissue samples from feed trees were ana-
lyzed by GCMS primarily as a check on ether
extract fractions. The small sample size did
not permit statistical analysis, but trends
showing lower fatty acid content in oak than in
conifers concurred with our observation of
lower ether extract fractions in oak. The
amount of fatty acids was lower in oak than in
either conifer species.

Discriminant analysis correctly classified
feed trees as either conifer or oak 100% of the
time (Table 2). Six factors were important con-
tributors to the model. Conifer feed trees had
higher amounts of phosphorus and a greater
ether extract fraction than oak feed trees.
Alternatively, oak feed trees were higher in
protein, calcium, tannins, and magnesium.
Although tannins entered into the model, they
were not a significant contributor These dif-
ferences between oak and conifer feed trees
generally are in agreement with differences in
Table 1. The classification of oak feed and non-
feed trees was less successful (71%, Table 2).
Oak feed trees were significantly higher in
sodium and fiber than nonfeed trees, while
nonfeed trees were higher in water content.

Porcupines used gambel oak as a food source
more often than it occurred in the study site
{p values listed in Table 3). Six of 15 animals
were found roosting and feeding exclusively in
oak, while 9 roosted and fed in conifer species
at least once. Snow depths and temperatures
were analyzed for the winter period before the
main snowmelt (judged to be 18 March).
Average snow depths at porcupine location
sites for that time period were 0.60 m.
Maximum snow depth was 1.20 m (median
0.65 m). Mean minimum temperature for the
night previous to locating study animals was
-10°C; the extreme low was -27°C. Mean
temperature for the night previous to locating
animals in rock or snow caves was -12 °C.
There was no statistical difference between
the minimum nightly temperature previous to

locating porcupines in station trees compared
to locating porcupines in rock or snow dens.

There were approximately 7.0 porcupines/
km^ in the study area. Radio-collared animals
were far ranging and did not utilize a single
den or station tree as a base from which to
launch foraging expeditions. Rather, they
roosted and fed in a single tree for one to sev-
eral days and then moved to another roost and
feeding tree. Death loss due to predation and
other causes left only 3 of 5 male and 6 of 10
female porcupines instrumented with radio
transmitting devices for the entire winter This
sample size made statistical analysis of home
ranges unreliable. Several animals spent the
winter in relatively small areas, but most had
relatively large, overlapping home ranges.
Male home range extremes were 6.8 and 47.5
ha. Extremes for females were 9.2 and 61.8 ha.
One female's home range overlapped those of
three males and at least four other females.
Movements of up to 400-500 m between relo-
cations of some of the larger, mature animals
were not uncommon. Some juvenile animals
had reduced home ranges and movements,
which generally agrees with observations by
Roze (1989). Mean distance from oak feed
trees to a potential conifer feed tree was sig-
nificantly less {p < .05) than the distance of an
average move by a porcupine from an oak feed
tree to any other feed tree (Table 1).

Three of 15 porcupines (20%) were eaten
by predators in a 4-mo period. Tracks in the
snow indicated that one porcupine was pur-
sued, worried, and killed by two coyotes. The
other two porcupines eaten by predators died
late in the season on south-facing slopes bare
of snow; neither the cause of death nor carni-
vore species could be positively determined.
Carcasses of two other porcupines that died
presumably of starvation and/or exposure dur-
ing the course of the study were not scav-
enged by coyotes.

Discussion

Chemical Factors

Dietary alternatives in the form of different
feed tree species, with significantly different
chemical makeup, were available to the study
population. In winter, vegetative oils have the
potential to be the most important source of
energy for porcupines. Data from ether extract
fi-actions derived fi-oni feed tree tissues indicate

1995]

Wintering Porcupines in Gambel Oak

Table 2. Standardized canonical discriminant function coefficients for factors that discriminated between oak and
conifer feed trees (100% correct classification), and between oak feed trees and oak nonfeed trees (71% correct classifi-
cation).

Oak ( + ) vs. conifer (-) feed trees n

= 56

Oak feed ( + ) vs. nonfeed (-)

trees

n = 46

Coefficient

Prob > b

Coefficient

Prob > b

Phosphorous -1.24

.00001

Water content -0.62

.006

Ether extract fractions -0.60

.0001

Sodium +0.61

.02

Protein +1.18

.0005

Fiber +0.59

.001

Calcium +0.39

.019

Tlmnins +0.29

.175

Magnesium +0.24

.006

that gambel oak, the major food source of our
animals, had lower values of ether extract fiac-
tions than tissues from conifers. Evaluation of
fatty acids by GCMS confirmed that fatty acid
content was higher in conifer tissue. Additional
research on known digestible fractions is
needed, but until data indicating otherwise
are presented, we will operate under the
premise that for porcupines conifers provide a
greater source of useable fats than do oaks.
Discriminant analysis was used to determine
if, when all variables were taken together,
there would be general support from this
analysis with the t test. Significant differences
found by these analyses comparing oak and
conifer feed trees were in agreement (Tables
1, 2). Phosphorus and the ether extract frac-
tion were higher in conifer feed trees com-
pared to oak feed trees, and protein, calcium,
tannins, and magnesium were higher in oak
feed trees. Discriminant analysis was less suc-
cessful in classifying feed and nonfeed trees
within oak (Table 2). An important reason for
this less successful classification was that the
cloning nature of oak was emphasized by the
point-quarter method. This method may have
resulted in selecting nonfeed trees from the
same clone as the feed tree. Future research
should involve delineating the boundary of the
clone and selecting a nonfeed tree from a
clone different from the feed-tree clone.

Conifer roost sites also offer greater thermal
advantages than deciduous roost sites (Clarke
and Brander 1973, Roze 1989). Despite multi-
ple options, porcupines depended heavily on
an oak diet low in fats and associated themial
advantages but higher in tannins. The advan-
tage of the oak diet may well be that it is high-
er in protein. High levels of crude fiber (e.g.,
cellulose) reduce the digestibility of crude
protein in monogastrics (Glover and Duthie
1958a, 1958b). Therefore, herbivores on a

high-fiber diet would be expected to maximize
the intake of crude protein to compensate for
a low digestibility rate. Implications of a diet
high in calcium and tannins are less clear, but
it is possible that porcupines may deal with
high levels of calcium in their food material by
concentrating calcium in the urine. Tannins
function as protein binding agents (Rhoades
and Gates 1976). It is now evident that some
insects can circumvent tannins through a
higher gut pH and the presence of surfactants
(Bernays 1981, Martin and Martin 1984,
Martin et al. 1985). However, pH values for
the mid-caecum (6.6), and the pyloric (1.8) and
esophageal (3.2) regions of the stomach of a
laboratory porcupine on a diet of oak were
consistent with gut pH for monogastrics of
comparable size (Hume 1982).

Oveson (1983) measured subcutaneous adi-
pose concentrations on the rump of porcu-
pines and reported a thickness of 15.1 mm (±
2.6 mm) in early winter. By late February and
early March fat reserves were virtually non-
existent. A similar phenomenon was observed
by Sweitzer and Berger (1993) in Nevada,
where porcupine body condition decreased
significantly throughout the winter season.
Those authors suggested the change in body
mass was an indication that porcupines deplet-
ed energy reserves early in the winter and were
stressed nutritionally during late winter. The
heavy accumulation of fat serves as an energy
reserve for porcupines to draw upon through-
out the winter, allowing them to concentrate
on a food source relatively high in crude pro-
tein. The reduced capabilities of protein
digestibility associated with a high-fiber diet
may have encouraged our study animals to
maximize dietary protein by selecting oak.

Porcupine herbivory was generally noted on
small branches. In large trees porcupines fed
high in the canopy where limbs are smaller.

Great Basin Natufuijst

[Volume 55

Table 3. Chi

i-square

analysis

per

cent

occunenct

■ and uti

lization

of trees by

porcnpines.*

occurrence

% nsecl

Chi -square
value

/; \ alue

Oak
Conifer

43.5
2.7

82.1
16.4

3.23

0.10

Maple
Conifer

.52.1
2.7

1.5
16.4

52.41

0.01

Oak
Maple

43.5
,52.1

82.1
1.5

59.14

0.01

*n values differ from those reported in Table 1 due to the extended use of some feed trees by porcupines. Occupancy of the same feed tree during more than
one sampling event counted as multiple utilization of oak but not double sampled for chemistry data. Df = 1.

We obsei-ved only two instances in which por-
cupines chipped bark of large tree boles and
fed on tissue from large dbh limbs or trunks.
Selection of larger feed trees by porcupines
may be related to the texture of bark and ease
of climbing (Roze 1989) rather than chemistry.

Deciduous Food and Roosting Resource

Roze (1989) discussed the thermal advan-
tages of dens and/or conifer roost trees in rela-
tion to maintenance of a core body temperature.
Citing Ii-ving et al. (1955) and Clarke (1969), he
indicated that the critical external tempera-
ture below which porcupines must increase
their metabolic rates to maintain a core body
temperature is a range between -12 and -4°C.
He suggested dens are temperature-averaging
devices that protect porcupines against convec-
tional and radiational heat loss. Station trees
provide thermal advantages to porcupines
(Clarke and Brander 1973) and may serve as a
substitute for rock caves and snow dens.
However, none of these are requisite to porcu-
pine survival. Roze (1989) noted that porcu-
pines may spend winters in trees away from
dens and that in every report the tree species
have been evergreens.

Our data conflict with this observation.
Porcupines throughout western North America
are able to survive using a variety of deciduous
species as food and roost tree resources.
Despite the prominence of literature concern-
ing dens and conifer station trees, use of a
deciduous food and roosting resource without
dependence on caves or snow dens is not an
anomaly for porcupines. Craig and Keller s
(1986) study site in southern Idaho was at an
elevation of 1525-2089 m in desert shrub
habitat. Animals in this study were not

observed using dens during the winter or fol-
lowing runways in feeding areas. They re-
mained in the tops of hawthorne {Crataegus
douglasii) thickets or utilized other deciduous
food sources throughout the winter. Sweitzer
and Berger (1993) identified buffalo-berry
{Shepherdia argentea), willow {Salix spp.), bit-
terbrush {Purshia tridentata), and juniper
{Jiiniperus osteosperma) as primary winter
food sources of porcupines in Nevada. We
have also observed the extensive use of hack-
beny {Celtis occidentalis) and green ash {Fraxi-
nus pennsylvanica) by porcupines as a food
and roosting resource in the Sand Hills of
Nebraska and the Missouri River Breaks of
South Dakota. Caves and conifers (except plan-
tation forests and eastern red cedar [Juniperus
virginiana]) are not available in the Sand Hills
(Swinehart 1989). Oveson (1983) reported that
a porcupine remained virtually motionless
while perched in a gambel oak tree for a 24-h
period when the ambient temperature was as
low as -37°C. During a 13-d period from 30
January through 11 Februaiy, when the mean
low temperature was -17°C, 3 of 25 (12%) loca-
tions of our study animals were in conifers, 4
(16%) were in rock or snow caves, and 18
(72%) were in oak. Although porcupines did
select trees with a larger dlih as roosting/feed-
ing sites, they were also often found in smallish
shrubs even though large trees were readily
available. It is therefore difficult to link possi-
ble benefits presumed to be available to porcu-
pines that roost in larger trees, such as protec-
tion from the elements or from predators, to
the selection shown by animals in this study.

Despite the availabilit\' of snow caves, dens,
and conifer species that could provide thermal
advantages, the study population was heavily

1995]

Wintering Porcupines in Gambel Oak

dependent on gambel oak for a roosting and
feeding resource. Considering that this re-
hance was during a season of energetic stress,
it is hkely that remaining motionless in the
canopy of oak trees to consei-ve energy while
exploiting a high-protein food source is an
adaptive strategy.

Movements and Predation

The availability of conifer feed trees was not
limiting since the average distance between
locations of study animals was significantly
greater than the mean distance of a move from
any roost tree to a conifer roost tree (Table 1).
It does not appear that spatial relationships of
the various feed tree species played a role in
feed tree selection by our study population.
The relatively large overlapping winter home
ranges of animals in this study differ from
reports of other researchers. Home ranges for
porcupines in northwestern Minnesota were
small enough to be reported in square meters
(Tenneson and Oring 1985). Curtis (1941),
Dodge (1967), Brander (1973), Roze (1987,
1989), and others have documented that por-
cupines move short distances from dens to
feed trees, sometimes along permanent trails
in the snow. Craig and Keller (1986) and
Smith (1979) also reported reduced ranges in
the winter. However, Dodge and Barnes
(1975) did not indicate a similar restriction in
winter movements. Roze (1987) suggested the
reason may be crusted snows that bear the
weight of the animals. Porcupines in our study
did adeptly toboggan on crusted snows down
extreme slopes in an attempt to avoid capture.
However, one female moved over 450 m in
fresh snow. Trails in powdery snow were often
direct and suggested that a destination may
have been predetermined.

Common use of oak and conifer feed trees
by different porcupines occurred several times
during the study, sometimes concurrently.
Hedging in the canopies of gambel oak trees in-
dicated that some trees were used consistently
over time by porcupines while others were
not. Consistent foraging in common trees over
time may indicate a learned behavior such as
that described by Glander (1981) for howler
monkeys, but we hesitate to attribute it to
such because porcupine young-of-the-year
were usually separated from their mothers
during the winter. It is possible that some
young accompanied their mothers for limited

periods in the winter or that more subtle cues
were used to transfer the information.

Long movements between feed trees in
dense oak cover by some study animals sug-
gest that predator-prey relationships may have
influenced movements. Sweitzer and Berger
(1992) found that habitat use was related to
the age or size class of porcupines, presum-
ably in response to increased risk of predation
to smaller porcupines. Our observations gen-
erally agree with their findings. Mountain lion
and coyote tracks were seen regularly in the
study area. Both species are known to prey on
porcupines (Keller 1935, Robinette et al. 1959,
Toweill and Meslow 1977, Maser and Rohweder
1983). The strong urine scent at station trees or
dens makes porcupines readily detectable.
Mountain lions are capable of knocking porcu-
pines from the canopies of trees (Taylor 1935).
If long moves decreased the predictability of
mountain lions locating porcupines in station
trees, it would be an adaptive strategy. How-
ever, long moves expose porcupines to terres-
trial predation by mountain lions, coyotes, and
wolves {Canis lupis, which are now extirpated
fi-om the study area) and would presumably be
nonadaptive. Since ample forage exists
throughout the study site and long moves to
locate food resources do not appear to be a
dietary necessity, long movements may be an
adaptive strategy to avoid arboreal predation
by mountain lions. This hypothesis deserves
further examination.

Acknowledgments

We thank S. H. Jenkins and two anony-
mous reviewers for helpful suggestions to this
manuscript.

Literature Cited

Bernays, E. a. 1981. Plant tannins and insect herbivores:

an appraisal. Ecological Entomology 6: 353-360.
Brander, R. B. 1973. Life historv' notes on the porcupine

in a hardwood-hemlock forest in upper Michigan.

Michigan Academician 5: 425—433.
Clarke, S. H. 1969. Thermoregulatory' response of the

porcupine, Erethizon dorsatiim, at low temperatures.

Special report. Department of Forestry' and Wildlife

Management, University of Massachusetts, Amherst.
Clarke, S. H., and R. B. Brander. 1973. Radiometric

determination of porcupine surface temperature

under two conditions of overhead cover.

Physiological Zoology 46: 230-237.
Cottam, C, and J. T. Curtis. 1956. The use of distance

measures in phytosociological sampling. Ecology 37:

451-460.

Great Basin Naturalist

[Volume 55

Craic;, E. H., and B. L. Kkller. 1986. Movements and
home range of porcupines {Erethizoii dorsutwn) in
Idaho shnib desert. Canadian Field-Naturahst 100;
167-173.

Curtis, J. D. 1941. The silvicuhural sigiiilicanee of the
porcupine. Journal of Forestry 39: 583-594.

D.^SiLVEiiu, A. J., K K Fkitosa Teles, and J. W. Stull.
1978. A rapid teclini(jue lor total nonstructural car-
bohydrate determination of plant tissue. Journal of
Agricultural and Food (]hemistr>' 26: 771-772.

Dodge, \V. E. 1967. Life histor\ and biology of the porcu-
pine {Erethizon dorsatum) in western Massachusetts.
Unpublished doctoral dissertation. University of
Massachusetts, Amherst. 167 pp.

Dodge, W. E., and V. G. Barnes. 1975. Movements,
home range, and control of porcupines in western
Washington. U.S. Department of Interior, Fish and
Wildlife Service Leaflet 507.

Gambliel, H. a., R. G. Gates, M. K. Gaffey-Moquin,
and T. D. Paine. 1985. Variation in the chemistry of
loblolly pine in relation to infection by the blue-stain
limgus. Pages 177-185 ;/i S. Branhani and R. Thatcher,
eds.. Proceedings of the Integrated Pest Mangement
Symposium, Asheville, NG. USDA Forest Service,
Southern Experiment Station, New Orleans, LA.

Glander, K. E. 1981. Feeding patterns in mantled howl-
ing monkeys. Pages 231-257 in A. G. Kami!, and T
D. Sargent, eds.. Foraging behavior: ecological, etho-
logical, and psychological approaches. Garland
STPM Press, New York, NY.

Glover, J., and D. W Duthie. 1958a. The nutritive ratio/
crude protein relationship in nmiinant and nonnmii-
nant digestion. Joinnal of Agricultural Science 50:
227-229.

. 1958b. The apparent digestibility of crude protein

by nonruminants and nmiinants. Journal of Agricul-
tural Science 51: 289-293.

Hagerman, a. E. 1987. A radial diffusion method for
determining tannin in plant e.xtiacts. Journal of Chem-
ical Ecologv' 13: 437-449.

HORWITZ, W, ED. 1980. official methods of analysis of the
Association of Official AnaKtical Chemists. Association
of Official Analytical Chemists, Washington, DC.
1018 pp.

Hume, I. D. 1982. Digestion physiology and nutrition of
marsupials. Cambridge University Press, Cambridge,
MA. 256 pp.

Irving, L. H., H. Krog, and M. Monson. 1955. The
metabolism of some Alaskan animals in winter and
summer. Physiological Zoology 28: 173-185.

Keller, E L. 1935. Porcupines killed and eaten by a coy-
ote. Journal of Mammalogy 16: 232.

Kufeld, R. C, M. Stevens, and D. G. Bowden. 1981.
Winter variation in nutrient and fiber content and in
vitro digestibility of gambel oak {Qucrcus gamhelii)
and big sagebrush {Artemisia tridentata) from diver-
sified sites in Colorado. Journal of Range Manage-
ment 34: 149-151.

Martin, J. S., and M. M. Martin. 1984. Surfactants: their
role in preventing the precipitation of proteins by
tannins in insect guts. Oecologia 61: 342-345.

Martin, M. M., D. G. Rockholm, and J. C. Martin.
1985. Effects of surfactants on precipitation of pro-
teins by tannins. Journal of Chemical Ecology 11:
485-494.

Maser, C, and R. S. Rohvveder. 1983. Winter food
habits of cougars from northeastern Oregon. Great
Basin Naturalist 43: 42.5-428.

Masslich, W J. 1985. Aspen-beaver relationships in the
Strawberry Valley of central Utah. Unpublished
master's thesis, Brigham Young University', Provo,
UT 34 pp.

MiNITAB. 1982. Release 82.1. Copyright, Penn State Uni-
versity, State College, PA.

Oveson, M. C. 1983. Behavioral and metabolic adapta-
tions of porcupines (Erethizon dorsatum) to winter
stress. Unpublished master's thesis, Brigham Young
University, Provo, UT. 20 pp.

Rhoades, D., and R. G. Gates. 1976. Toward a general
theory of plant antiherbivore chemistry. Pages
168-213 in Biochemical interaction between plants
and insects. J. W Wallace and R. W. Marshall, eds..
Recent Advances in Phytochemistry. Volume 10.
Plenum Press, New York-London.

ROBINETTE, W L., J. S. GaSHWILER, AND O. W. MORRIS.
1959. Food habits of the cougar in Utah and Nevada.
Journal of Wildlife Management 23: 261-273.

ROZE, U. 1987. Denning and winter range of the porcu-
pine. Canadian Journal of Zoology 65: 981-986.

. 1989. The North American porcupine. Smithsonian

Institution Press, Washington, DC. 261 pp.

SAS. 1985. Copyright, SAS Inc., Gary, NG.

Smith, A. D. 1957. Nutritive value of some browse plants in
winter. Journal of Range Management 10: 162-164.

Smith, G. W 1979. Movements and home range of the
porcupine in northeastern Oregon. Noithwest Science
53: 277-282.

SWEITZER, R. A., AND J. Berger. 1992. Size-related effects
of predation on habitat use and behavior of porcu-
pines [Erethizon dorsatum). Ecology 73: 867-875.

. 1993. Seasonal dxnamics of mass and body condi-
tion in Great Basin porcupines {Erethizon dorsatuni).
Journal of Mammolog>' 74: 198-203.

SwiNEHART, J. B. 1989. Wind-blown deposits. Pages 43-56
in A. Bleed and G. Flowerday, eds.. An atlas of the
Sand Hills. Resource Atlas No. 5, Conservation and
Survey Division, Institute of Agriculture and
Natural Resources, Lhiiversitv' of Nebraska-Lincoln.

Taylor, W R 1935. Ecology and life history- of the porcu-
pine {Erethizon epixanthum) as related to the forests
of Arizona and the southwestern United States.
University' of Arizona Bulletin 6: 1-177.

Tenneson, C., and L. W Oring. 1985. Winter food pref-
erences of porcupines. Journal of Wildlife Manage-
ment 49: 28-33.

TowEiLL, D. E., and C. E. Meslow. 1977. Food habits of
cougars in Oregon. Journal of \Vildlife Management
41: 576-578.

Underwood, E. J. 1966. The mineral nutrition of live-
stock. Central Press Ltd., Aberdeen, Great Britain.
237 pp.

U.S. Glimatological Data (Utah). 1984 and 1985.
National Climactic Center, Asheville, NG. 86: 12 and
87: 13.

Welgh, B. L. 1989. Nutritive value of shrubs. Pages
405-^24 in G. M. McKell, ed., Biology and utilization
of shrubs. Academic Press, Inc., New York, NY.

Received 18 August 1993
Accepted 30 September 1994

HISTORIC EXPANSION OF JUNIPERUS OCCIDENTALIS
(WESTERN JUNIPER) IN SOUTHEASTERN OREGON

Richard E Miller^ and Jeffeiy A. Rosel

Abstract. — The chronology of Junipents occidentalis (western juniper) expansion in eastern Oregon, the effect of
plant canopy and interspace on / occidentalis seedling establishment and growth rates, and the age of/, occidentalis
maximum reproductive potential were determined. Measurements were recorded in twenty-two 0.4-ha plots estab-
lished in sagebrush-grassland communities and six 0.1-ha plots in Populus tremuloides (quaking aspen) communities. /.
occidentalis began increasing during the 1880s in stands containing trees >130 yr old. Relatively steady establishment
ensued into the 1950s and then began to progress at a geometric rate in the 1960s. / occidentalis encroachment into
aspen stands began between 1910 and 1920. The largest proportion of juvenile trees established beneath Artemwia
species in sagebrush-grassland communities./, occidentalis trees appeared to reach full reproductive potential at >50 yr
of age. The ratio of male;female trees increased from 1.7 in scattered/, occidentalis stands to 3.8 in closed stands. The
initiation of/, occidentalis encroachment during the late 1800s coincides with optimal climatic conditions for Juniperus
beiTy production and establishment, reduced fire-retum intervals, and heavy livestock grazing. The accelerated increase
in /. occidentalis expansion since 1960 may be due to the continued absence of fire, abundant woody plant cover, and
the large increase in/, occidentalis seed production.

Key words: western juniper, Juniperus occidentalis, expansion. Great Basin, intennountain shrub steppe, aspen,
Populus tremuloides, succession.

One of the most pronounced plant commu-
nity changes in the 20th century has occurred
in the juniper and pinyon-juniper woodlands,
a major vegetation type characterizing the
Intermountain Region. These woodlands,
sometimes described as pygmy forests, cur-
rently occupy 17 million ha throughout this
region (West 1988). Juniperus occidentalis ssp.
occidentalis Hook, (western juniper) is consid-
ered the Northwest representative of the pin-
yon-juniper zone in the Intermountain Region
(Franklin and Dyrness 1973) and occupies
over 1 million ha (Dealy et al. 1978) in eastern
Oregon, southwestern Idaho, and northeastern
California (Cronquist et al. 1972). This sub-
species of/, occidentalis is found primarily
north of the polar front gradient (Neilson
1987; parallel to the Oregon and Nevada border,
latitude 42°) where temperatures are cooler,
summer precipitation decreases, and winter
precipitation increases (Mitchell 1976).

Relict juniper woodlands, tree-age class dis-
tribution, fire scars, and historical documents
indicate presettlement pinyon-juniper and
juniper woodlands were usually open, savan-

nah-like (Nichol 1937, West 1988), or confined
to rocky surfaces or ridges (Cottam and Stewart
1940, Barney and Frishknecht 1974, Hopkins
1979, Johnson and Simon 1987). /. occidentalis
began increasing in both density and distribu-
tion in the late 1800s (Burkhardt and Tisdale
1976, Young and Evans 1981, Eddleman 1987),
invading Artemisia tridentata subsp. vaseyana
(mountain big sagebrush), Artemisia arhuscula
(low sagebrush), Populus tremuloides (quaking
aspen), and riparian communities. Although /.
occidentalis is long lived (Vasek 1966, Lanner
1984), less than 3% of the woodlands in Oregon
are characterized by trees >100 years old
(USDI-BLM 1990). In 1825, Ogden' observed
only occasional /. occidentalis (reported as
cedars) growing on hillsides while traveling
through the Crooked River drainage in central
Oregon (Rich et al. 1950). Today these hill-
sides are covered by dense /. occidentalis
woodlands. In a nearby area J. W Meldrum's
1870 survey notes describe a gently rolling
landscape covered with an abundance of peren-
nial bunchgrasses and a wide scattering of/.
occidentalis trees (Caraher 1977). Today, /

'Eastern Oregon Agricultural Research Center, HC 71 4.51 Hwy 20.5, Bums, OR 97720. The Eastern Oregon Agricultural Research Center, including the
Bums and Union stations, is jointly operated by the Oregon Agricultural E.\perinient Station of Oregon State University and USDA Agricultural Research
Service.

CiHEAT Basin Natuiulist

[Volume 55

occidentalin densities on this site range
between 125 and 250 ha"^. In Silver Lake,
Oregon, /. occidentalis density increased from
62 ha-1 in 1890 to over 400 ha-1 by 1970
(Adams 1975). On another site in central
Oregon where trees were absent prior to
1880, /. occidentalis increased to 1018 ha~^ by
1980 (Eddleman 1987). Recent expansion is
similar to increases in other Juniperus species
throughout western United States (Ellis and
Schuster 1968, Tausch et al. 1981, West 1984,
Tausch and West 1988).

The objectives of our study were to (1)
describe the chronology of/, occidentalis
expansion during the past several centuries in
southeastern Oregon, (2) determine the effect
of plant canopy and interspace on / occiden-
talis seedling establishment and growth rates,
and (3) determine the age when /. occidentalis
reaches maximum reproductive potential.

Methods

Study Area

The study area is located on Steens Moun-
tain in southeastern Oregon, approximately 80
km south of Burns. This isolated volcanic
fault-block, which lies in the extreme north-
west Basin and Range Province (Fenneman
1931), is about 80 km long and oriented in a
northeast direction (Baldwin 1981). The eleva-
tion of Steens Mountain ranges from 1268 to
2949 m, with a steep east-facing escarpment
and a gentle west-facing slope. Climate is cool
and semiarid, characteristic of the northern
Great Basin. Annual precipitation at the lower
elevations averages 220-280 mm, increasing to
> 700 mm at higher elevations (NOAA 1993).
Most moisture is received as snow in Novem-
ber, December, and January and as rain in
March through June.

/. occidentalis woodlands on Steens Moun-
tain form a discontinuous belt between 1450
and 2100 m in elevation. Severe winter condi-
tions probably restrict /. occidentalis from ex-
panding into higher elevations (Billings 1954,
Mehringer 1987). Limited distribution below
1500 m is possibly due to a combination of late
spring frosts (Billings 1954) and limiting mois-
ture. Tree canopy cover varies from open to
30% cover, except on mesic P. treniuloides sites
where /. occidentalis cover approaches 100%.
However, based on age structure and canopy
leader growth, tree canopies are still actively

expanding on the majority of sites measured.
Early observations on Steens Mountain indi-
cate the landscape contained only scattered
stands of/ occidentalis (Griffiths 1902). Since
1900 the abundance of/, occidentalis pollen in
the Steens Mountain area has increased five-
fold (Mehringer and Wigand 1990).

Plant communities characteiistic of/ occi-
dentalis woodlands are Artemisia tridentata
ssp. vaseyana/Festuca idahoensis (Idaho fes-
cue), Artemisia arhuscula/E idahoensis, and P.
tremuloides. P. tremuloides communities on
Steens Mountain range in elevation from 1760
to 2400 m. At lower elevations, in the / occi-
dentalis woodland belt, P. tremuloides stands
form long, narrow communities along north
aspects, which capture windblown snow and
runoff.

Plot Layout

Plot locations were selected in an attempt
to reflect sagebrush-grassland communities in
different stages of/ occidentalis invasion on
the west slope of Steens Mountain. Old stands
on the rocky outcrops, which make up only a
small percentage of present-day woodlands,
were not measured. Sites selected support, or
have the potential to support, sagebrush-grass-
land communities. Currently these sites are
occupied by varying numbers and sizes of/
occidentalis dominance, creating a woodland
structure of dispersed, intermediate, and
closed tree stands (Table 1). Twent\'-two 0.4-ha
plots were located within the / occidentalis
belt of Steens Mountain; tliey ranged from 1500
to 2000 m in elevation and were distributed 32
km along the mountain range. Plots were situ-
ated along an elevation gradient representing
communities from the lower- to upper-eleva-
tion / occidentalis woodland belt. Dominant
understory vegetation in the dispersed and
intermediate plots was A. tridentata spp.
vaseijana and Festuca idahoensis (13 stands),
A. arbuscula and F idahoensis (4 stands), and a
mosaic of A. arbuscula and A. tridentata ssp.
vaseijana (2 stands). Understoiy vegetation in
the closed stands (n = 3) comprised a few
remnant deep-rooted perennial grasses, skele-
tons of dead A. tridentata ssp. vaseijana, and
70% bare ground (EOARC data file).

An additional six 0.1-ha plots were estab-
lished in six separate P. tremuloides stands.
Three stands were in advanced stages of/
occidentalis invasion with few to no adult P.

1995]

Western Juniper Expansion

Table L Juniperus occidcntalis stand maturity classes
in Artemisia communities (modified fi-om Blackburn and
Tueller 1970).

Closed Abundant adult trees generally >5 ni tall

and usually several trees > 130 yr of age,
with little understory, particularly on
south slopes.

Intermediate Abundant/ occidentalis of all age classes,
with a more open tree canopy and an
understory beginning to decline; trees
> 130 yr of age are rare.

Dispersed Abundant young trees <2 m tall, a few

adult trees but old trees absent, and a
well-developed understory.

tremuloides trees and dead P. tremuloides
trunks on the ground. The remaining three
stands were characterized by a dominant P.
tremuloides overstory and an understory of
young/, occidentalis. Elevation for the/, occi-
dentalis-P. tremuloides plots ranged from 1930
to 2000 m, all with a similar northeast aspect.

Measurements

Prior to sampling, string was stretched along
the contour of each 0.4-ha plot at 1-m intervals
to keep track of measured trees. / occidentalis
density (trees ha~^) was recorded for trees
<0.5 m tall, defined as adult, across the entire
plot. Tree height, minimal and maximal crown
diameters, and basal area just above the trunk
swell at the stem base near the litter layer
were recorded. Tree height was measured
with a tape for trees < 2 m and a clinometer
for trees >2 m tall. Tree canopy cover was
estimated by adding crown area measure-
ments of all trees for each plot. Similar mea-
surements were recorded on juvenile trees
(defined as trees < 0.5 m tall), but only those
on the lower left quarter (0.1 ha) of each 0.4-
ha plot. Current-year / occidentalis seedlings
(any plant with cotyledons still attached) were
not recorded. Establishment location of each
juvenile tree was recorded: beneath the canopy
of/ occidentalis, Artemisia, other shrubs, tus-
sock grass, or in the interspace. Less than 1% of
juveniles were located beneath other shrubs or
grasses; therefore, only / occidentalis, Artemisia,
and interspace are reported.

/ occidentalis is considered submonoecious
(Vasek 1966). Male and female reproductive
status was detemiined by estimating abundance
of cones and berries for each tree. Abundance
was ranked in four classes: (0) absent, (1) scarce,

(2) obvious but not abundant, and (3) abundant.

In each plot a 10-tree subsample was ran-
domly selected for aging in each of four height
classes: (1) <0.5 m, (2) 0.5-1.8 m, (3) 1.8-3 m,
and (4) >3 m. In several of the dispersed plots,
sample size for trees >3 m was smaller than
10, due to a lack of trees. We also sampled all
old trees on plots when they occun^ed (n = 0-5
ha~l). Old trees were easily identified by their
growth form, containing rounded tops and
heavy limbs, and lacking strong terminal
leader growth (Burkhardt and Tisdale 1969). A
cross section was removed approximately 30
cm above ground level from each tree >0.5 m
tall and at ground level for trees < 0.5 m, and
then brought back to the lab for aging. Two
radii from each cross section were polished,
stained, and counted. Age was estimated by
averaging both radii and adding 10 yr to cor-
rect for the 30-cm base. Mean differences
between radii were 4% for trees >50 yr and
1% for trees < 50 yr of age. Adams (1975)
reported that growth-ring characteristics of/
occidentalis are useful in dendrochronological
studies. The presence of false and missing
rings was similar to that for Pinus ponderosa.
Over 1200 trees were aged and approximately
14,000 counted and measured. In the six P.
tremuloides stands, density of both /. occiden-
talis and P. tremuloides and age and height for
/ occidentalis were measured across the entire
0.1-haplot.

Evidence indicated minimal / occidentalis
mortality has occurred on Steens Mountain
during the past 120 years. We observed very
few dead or dying trees for all age classes
(excluding seedlings), except where individual
/ occidentalis trees had been cut or burned.
Mortality of Jimiperus species rapidly declines
following the seedling stage (Van Pelt et al.
1990). Juniperus has few pests that prove fatal
to the tree (Lanner 1984). We avoided recently
cut or burned stands, which constituted a
small percentage of/, occidentalis-occupied
stands. Where remains of dead trees were
observed, we noted they persisted for a long
period of time. By recutting several stumps
adjacent to one of our plots and aging and
matching ring widths with adjacent live trees,
we determined these trees were harvested
around 1920. Others have also observed the
persistence of Juniperus stumps (Young and
Budy 1979).

Great Basin Naturalist

[Volume 55

Statistical Analysis

Height growth data for adult trees were
analyzed using a randomized complete block-
design in PROG GLM of SAS (SAS 1986).
Means were separated using Duncan's
Multiple Range Test at /; < .05 level. A split-
plot design was used in the analysis of juvenile
height growth. Main plots were sites and sub-
plots were location of establishment (interspace,
Artemisia, J. occidentalis). A Duncan's Multiple
Range Test was used to separate the means.

Results

Little change in /. occidentalis density
appeared to occur between the early 1700s
and the 1880s (Fig. 1). We encountered old
trees (standing trees >130 years old, large
stumps, and burned-out trunks) on several A.
arhusciila flats and A. tridentata ssp. vaseyana
communities. However, data indicated preset-
tlement tree densities in these Artemisia com-
munities were < 5 trees ha~^ suggesting very
open /. occidentalis stands. The first evidence
of an increase in tree densities occurred in the
1880s, with relatively steady establishment
ensuing into the 1950s, similar to that
observed by Tausch and West (1988). In the
1960s /. occidentalis establishment began
occurring at a geometric rate.

Glosed /. occidentalis stands, which once
supported A. tridentata ssp. vaseyana, were

characterized by an abundance of adult trees
(> 3 m tall), a tree canopy cover of 18-28%
(Table 2), and the presence of a few old trees
(130+ yr; 2 to 5 ha~^). /. occidentalis densities
began increasing in these stands between
1878 and 1890. In the intermediate/, occiden-
talis stands, trees >130 yr were rare. Tree
canopy cover ranged from about 8 to 16%, and
densities of adult trees varied from 35 to 100
ha~l. Trees <3 m in height, particularly juve-
niles, were abundant. /. occidentalis expansion
in these sagebrush-grassland communities
began between 1890 and 1910. In the dis-
persed stands few trees were >60 yrs old, and
we aged no trees > 100 yr. Tree canopy cover
was usually <5% in the dispersed stands and
densities of large adult trees <35 ha~l.
Invasion of/, occidentalis into these sage-
brush-grassland communities began after
1930.

Greatest densities of/, occidentalis trees
measured on Steens Mountain occurred in P.
tremuloides sites (Table 3). In the late stages of
/. occidentalis succession on these sites, tree
canopy cover approached 100%. Live P.
tremuloides occurred only on one of the three
sites, and almost all trees were <0.5 m tall. In
the remaining two stands only the remnants of
large P. tremuloides trunks decaying in the
understoiy were present. /. occidentalis inva-
sion in these P. tremuloides sites began be-
tween 1910 and 1920. No/, occidentalis trees

T \ \ 1 1 1 r

1700 1720 1740 1760 1780 1800 1820 1840 1860 1880 1900 1920 1940 1960 1980

YEAR

Fig. 1. Years of establisliiiient i'or J uuipcrus occidentalis trees on Steens Mountain, Oregon (n = 1200).

1995]

Western Juniper Expansion

Table 2. General description of closed, inteniiediate, and dispt'rscd J unipcriis occidentalis stands on Steens Mountain
in Artemisia tridentata ssp. vaseijami and A. arhuscula comminiities, and the percentage of juveniles located beneath /
occidentalis, Ai-teinisia. and interspace. Canop\' cover, basal area, and density means are followed by range in paren-
theses ( ).

Establishment site %

Canopy

Basal area

Density (#ha-l)

(for

juveniles)

Adults

JuNcniles

# Site

s cover '7i

(ni-ha-^)

>0.5 ni ht

<0.5 III ht J

\. occidentalis

Aiiernisia

Interspace

A. tridentata

ssp. vaseyana

closed

22 (18-28)

5.2 (3.1-9.8)

296 (217-496)

580 (118-1226)

86^'l

5''

intermediate

6 (5-10)

1.8 (0..5-4.7)

95 (50-165)

815 (335-1423)

291'

58^

13^-

dispersed

2 (1-3)

0.4 (0.2-0.6)

.52 (31-70)

188 (96-280)

3''

50^

47a

A. arhuscula

closed

15 (12-20)

3.5 (1.8-5.4)

158 (74-247)

99 (20-198)

27''

67"

6''

intermediate

6 (4.5-6.7)

1.8 (0.9-3.2)

104 (77-153)

375 (167-790)

11''

28l'

'Sites of establishment means (%) folloux-d 1

.V similar limercast

■ IrttiTs are not simnl:

le.mtiv iliflerent I.etNveei

. establishment si

les within / m

rithnhdis stand

maturity classes (p ■

i .05).

>80 yr were encountered. In stands with a P.
tretniiloides overstoiy, P. trcmiiloidcs density of
small shoots was greater than that of/ occi-
dentalis. However, P. tremuloides size classes
between 0.5 m and large adults were absent,
indicating a lack of P. tremuloides stand reju-
venation. On these sites /. occidentalis inva-
sion began between 1930 and 1940.

Height growth for young /. occidentalis
trees (<20 yr) across all sites averaged
2.9 cm yir^. Based on growth rates and height
of trees between 10 and 20 yr of age (n = 200)
across all Artemisia sites, 90% of trees 15 yr
old were <1 m tall (64% were <0.5 m tall).
Surprisingly, height growth rates of juvenile
trees did not significantly differ between A.
arhuscula and A. tridentata ssp. vaseyana
communities. However, location of establish-
ment within communities significantly influ-
enced growth rates of young /. occidentalis
trees (Table 4). Trees establishing beneath an
Artemisia canopy grew faster than young trees
growing in the interspace.

Shrub and tree canopies also significantly
influenced location of/, occidentalis seedling
establishment in Artemisia communities. The
largest proportion of juvenile trees was usually
located beneath canopies of A. tridentata ssp.
vaseyana or A. arhuscula and /. occidentalis
(Table 2). Less than 20% of juveniles across all
22 Artemisia sites established in the interspace.

On Steens Mountain, for trees >0.5 m tall,
32% expressed predominantly only male or
only female characteristics, 38% both male

and female, and 30% contained neither fnjits
nor cones. /. occidentalis trees producing abun-
dant crops of cones or berries were either
male or female dominant. No trees were mea-
sured which contained an abundant crop of
both berries and cones. Sixty-five percent of/.
occidentalis trees with an abundant crop of
berries contained no male cones. The remain-
ing 35% contained only a scarce number of
cones. The majority of trees producing abun-
dant crops of male cones contained only
scarce numbers of berries. Approximately 75%
of trees producing heavy crops of berries or
cones were >50 yr old. Trees <20 yr old ex-
pressing reproductive effort were rare and
produced only a few cones or berries. The
ratio of trees producing large crops of cones
versus berries (cones :berries) increased from
1.7 in the scattered/, occidentalis stands to 3.8
in the closed stands.

Discussion

Low densities and limited distribution of/
occidentalis trees >130 yr and limited num-
bers of dead trees or old stumps suggest /.
occidentalis has greatly expanded on Steens
Mountain during the past 100 yr. Distribution
of old trees was generally limited to rocky
ridges and A. arhuscula communities. Old
trees were found only occasionally growing in
deeper, well-drained soils such as A. tridentata
ssp. vaseyana grassland communities and
were absent in P. tremuloides communities. In
northeastern California, Barbour and Major

Great Basin Naturalist

[Volume 55

Table 3. Mean dcnsitifs (# ha"') follow i-d 1)\ rant^f in ( ) oi F()i)iilus tretnuloidcs nud Junipcriis occidentalis in P.
treinuloides sites.

Stage of
succession

P. tremuloides

Adult

Juvenile

J-

occidei

Italia

Adnlt

Juvenile

1392

(929-2203)

1090
(632-1739)

9462
(4327-18,791)

2816
(622-5968)

Late (n = 3)

Intermediate

(n=3)

(0-50)

1060

(476-1670)

1316
(0-3952)

6553
(5266-9480)

(1977) found a similar distribution of old and
young /. occidentalis trees. A. tridentata ssp.
vaseyana and A. arhtiscxda commimities, which
contained a low density of/, occidentalis trees
prior to settlement, were the earliest sites to
initiate an increase in /. occidentalis. Dates of
initial establishment of closed and intermedi-
ate stands were similar to periods of early
stand development reported by Young and
Evans (1981) in northeastern California and
Eddleman (1987) in central Oregon.

Expansion of/, occidentalis coincides with
Euro-American settlement in this portion of
the Great Basin. Although no direct cause-and-
effect relationship can be drawn, we hypothe-
size that climate, altered fire frequencies, and
grazing in the late 1800s were primary factors
initiating the recent expansion of/, occidentalis.
Following the end of the Little Ice Age in the
mid 1800s (Biyson 1989), winters became more
mild and precipitation increased above the
present long-term average in the northern half
of the Great Basin between 1850 and 1916
(Antevs 1948, Graumlich 1985). Mild, wet win-
ters and cool, wet springs promote vigorous
growth in /. occidentalis (Earle and Fritts
1986, Fritts and Xiangdig 1986).

Presettlement fire-return intervals in A. tri-
dentata ssp. vaseyana communities have been
reported to vary from 15 to 25 yr (Houston
1973, Burkhardt and Tisdale 1976, Martin and
Johnson 1979). Burkhardt and Tisdale (1976)
concluded that fire-frequency intervals of
30-40 yr would be adequate to keep /. occi-
dentalis from invading a sagebrush-grassland
community. Following settlement, frequency
of fire in sagebrush grasslands has greatly
declined. The reduction of fine fuels by high
densities of domestic livestock greatly reduced
the potential for fire in the Intermountian
Shrub Region (Burkhardt and Tisdale 1976,
West 1988). Griffiths' (1902) observations of
the overgrazed landscape on Steens Mountain

support this hypothesis. Fires set by Native
Americans also declined in the 19th century
due to large reductions in their populations
caused by European diseases (Thompson
1916, Grossman 1981) and relocation to reser-
vations in the 1870s.

The invasion of conifers into P. tremuloides
communities is a common occurrence through-
out the western U.S. However, conifers report-
ed to typically invade P. tremuloides stands are
species adapted to more mesic sites such as
Piniis contorta (lodgepole pine), P. ponderosa,
Pseudotsuga inenziesii (Douglas-fir), Abies
concolor (white fir), Abies lasiocarpa (sub-
alpine fir), Picea engelmannii (Engelmann
spiTice), and Picea piingens (blue spnice) (Bartos
1973, Mueggler 1985). Invasion of the more
drought-tolerant /. occidentalis into P. tremu-
loides stands is not well documented.

P. tremuloides is frequently considered a
fire-induced species, replaced by less fire tol-
erant conifers (Baker 1925, Daubenmire 1943,
Mueggler 1976). Prior to settlement, lightning
and human-set fires probably helped maintain
many P. tremuloides communities. However,
the occurrence of fire in P. tremuloides stands
in the Rocky Mountains has been greatly
reduced since the late 1800s (Jones and DeByle
1985). Mueggler (1985) suggested the combi-
nation of fire suppression and heavy grazing in
P. tremuloides communities may favor tlie estab-
lishment of conifers.

An increase in Artemisia cover may also
enhance the invasion of/, occidentalis. As a
sagebrush-grassland community shifts towards
a greater dominance of shrubs, the number of
safe sites for /. occidentalis seedling establish-
ment increases. Others have also reported the
majorit>' of/, occidentalis seedlings established
beneath Artemisia canopies (Burkhardt and
Tisdale 1976, Eddleman 1987). In west Texas,
/. pinchotii frequently establishes beneath
mesquite plants (McPherson et al. 1988).

1995]

Western Juniper Expansion

Table 4. Mean growth rates for juvenile /w;ii;jf'n/.s occi-
dentalis trees (2-30 yr old) in three different establish-
ment sites.

Establishment site

cm yi

Artemisia

J. occidentalis
Interspace

3.3A

2.7AB

2.4B

Means followed by similar uppercase letters are not significantly different (/;
< .05).

Shading by nurse plants may benefit /. occi-
dentalis seedlings (Johnsen 1962) by reducing
summer surface temperatures by 45-57% of
bare ground surface temperatures (Burkhardt
and Tisdale 1976). Enhanced growth rates of
young trees growing beneath A. tridentata ssp.
vaseijana suggest microclimates beneath shrub
canopies are more beneficial than conditions
in the interspace. Burkhardt and Tisdale
(1976) reported/, occidentalis seedling growth
rates were correlated positively with Artemisia
and correlated negatively with bare ground.

/. occidentalis approached hill reproductive
potential near 50 yr As /. occidentalis densities
increased, the proportion of trees became pre-
dominantly male across sites. Highly fecund
female trees appeared to be most important in
open stands where /. occidentalis was actively
expanding. In central Oregon, Eddleman
(1984) observed that trees in the interior
woodlands were strongly dominated by male
cone production while trees growing in the
open produced more female cones. He also
reported trees did not produce significant
quantities of fruit until 50-70 yr of age.

Conclusion

Optimal climatic conditions around the
turn of the century, reduced fire retuiTi inter-
vals, and the indirect effect of livestock
through the reduction of fine fuels and an
increase in Artemisia cover are probably pri-
mary factors that have contributed to the rapid
expansion of/, occidentalis in southeast
Oregon during the late 1800s and early 1900s.
The accelerated increase in /. occidentalis
density and invasion during the last 30 years
into new communities is probably largely due
to the continued absence of fire, abundant
woody plant cover, and the large increase in /
occidentalis seed rain.

Acknowledgments

This is Technical Report 10,494 of the
Eastern Oregon Agricultural Research Center,
Oregon State University.

Literature Cited

Ada.ms, a. W. 1975. A brief histoiy of juniper and shnib
populations in southern Oregon. Wildlife Research
Report 6, Oregon State Wildlife Commission,
Corvallis.

Antevs, E. 1948. Climatic changes and pre-white man.
Pages 168-191 in The Great Basin, with emphasis on
glacial and postglacial times. University of Utah
Biological Service Bulletin 10(7).

Baker, E S. 1925. Aspen in the central Rocky Mountain
region. US DA Bulletin 1291.

Baldwin, E. M. 1981. Geology of Oregon. Kendal/Hunt
Publishing Co., Dubuque, lA.

B,\RBOUR, M.G., AND J. Major. 1977. Terrestrial vegeta-
tion of California. Wiley Interscience, New York, NY.

Barney, M. A., and N. C. Frishknecht. 1974. Vegetation
changes following fire in the pinyon-juniper type of
west-central Utah. Journal of Range Management
74; 91-96.

Bartos, D. L. 1973. A dynamic model of aspen succes-
sion. Pages 1.3-25 in Proceedings, lUFRO Biomass
Studies. lUFRO.

Billings, W. D. 1954. Temperature inversions in the
pinyon-juniper zone of a Nevada mountain range.
Butler University Botanical Studies 12.

Blackburn, W H., and P T. Tueller. 1970. Pinyon and
juniper invasion in black sagebrush communities in
east-central Nevada. Ecology 51: 841-848.

Bryson, R. a. 1989. Late Quaternary volcanic modulation
of milankovith climate forming. Theoretical Applied
Climatology 39: 115-125.

Burkhardt, J. W, and E. W Tisdale. 1969. Nature and
successional status of western juniper vegetation in
Idaho. Journal of Range Management 22: 264-270.

. 1976. Causes of juniper invasion in southwestern

Idaho. Ecology 76: 472-484.

Caraher, D. L. 1977. The spread of western juniper in
central Oregon. Pages 3-8 in R. E. Martin, J. E. Dealy,
and D. L. Caraher, eds., Proceedings, Western
Juniper Ecology and Management Workshop.
US DA Forest Service, General Technical Report
PNW-4.

Cottam, W. E, and G. Stewart. 1940. Plant succession as
a result of grazing and of meadow desiccation by
erosion since settlement in 1892. Journal of Forestry
38: 613-626.

Cressman, L. S. 1981. The sandal and the cave. Oregon
State University Press, Corvallis.

Cronquist, a., a. H. Holmgren, N. H. Holmgren, and
J. L. Reveal. 1972. Intermountain flora: vascular
plants of the Intermountain West, U.S.A. Volume 1.
Hafiier Publishing Company, New York, NY.

Daubenmire, R. F 1943. Vegetational zonation in the
Rock>' Mountains. Botanical Review 9: 325-393.

Dealy, J. E., J. M. Geist, and R. S. Driscoll. 1978.
Western juniper communities on rangeland of the

Great Basin Natuiulist

[Volume 55

Pacific Northwest. Pages 201-204 in D. K. ihiler,
ed., Proceedings, First International liangeland
Congress, Denver, CO.

Earle, C. J., AND H. C. Fritts. 1986. Reconstructing
river flow in the Sacramento Basin since 1560.
Report. California Department of Resources, Agree-
ment DWR B-55395. Laboratory of Tree-ring
Research, University of Arizona, Tucson.

Eddleman, L. E. 1984. Ecological studies on western
juniper in central Oregon. Pages 27-35 in T. E. Bedell,
ed.. Proceedings, Western Juniper Management
Short Course. Oregon State University and Extension
Service, Corvallis.

. 1987. Establishment and stand development of

western juniper in central Oregon. Pages 255-259 in
R. L. Everett, ed.. Proceedings, Pinyon-Juniper
Conference. US DA Forest Service, General Technical
Report INT-215.

Ellis, D., and J. C. Schuster. 1968. Juniper age and dis-
tribution on an isolated butte in Garza County, Texas.
Soutliwestem Naturalist 13: 343-.348.

Fenneman, N. M. 1931. Physiography of the western
United States. McGraw-Hill, New York, NY.

Fr\nklin, J. F, and C. T. Dyrness. 1973. Natural vegeta-
tion of Oregon and Washington. USDA Forest
Service, General Technical Report PNW-8. Portland,
OR.

Fritts, H. C, and W Xiangdig. 1986. A comparison
between response-function analysis and other re-
gression techniques. Tree-ring Bulletin 46: 31^6.

Gr.\UMLICH, L. 1985. Long-term records of temperature
and precipitation in the Pacific Northwest derived
from tree rings. Unpublished doctoral dissertation.
University' of Washington, Seattle.

Griffiths, D. 1902. Forage conditions on the northern
border of the Great Basin. Bureau of Plant Industry',
USDA, Bulletin 15.

Hopkins, W. E. 1979. Plant associations of the Fremont
National Forest. USDA Forest Service, Pacific
Northwest Region, R6-ECOL-79-004.

Houston, D. B. 1973. Wildfires in northern Yellowstone
National Park. Ecolog\' 54: 1109-1117,

JOHNSEN, T. N. 1962. One-seed juniper invasion of north-
em Arizona grasslands. Ecological Monographs 32:
187-207.

Johnson, C. G., Jr., and S. A. Simon. 1987. Plant associa-
tions of the Wallowa-Snake Province, Wallowa-
Whitman National Forest. USDA Forest Service,
Pacific Northwest Region Report R6-ECOL-TP-
255B-86. Portland, OR.

Jones, J. R. and N. V. DeByle. 1985. Fire. Pages 77-81 in
N. V. DeByle and R. E Winokur, eds.. Aspen: ecology
and management in the western United States.
USDA Forest Service General Technical Report
RM-119.

Lanner, R. M. 1984. Trees of the Great Basin: a natural
history. University of Nevada Press, Reno.

Martin, R. E., and A. H. Johnson. 1979. Fire management
of Lava Beds National Monument. Pages 1209-1217
in R. M. Linn, ed.. Proceedings, First Conference of
Science and Research in the National Parks. USDI
National Park Service, Transactions Proceedings
Serial 5.

McPherson, G. R., H. a. Wright, and D. B. Wester.
1988. Patterns of shrub invasion in semiarid Texas

grasslands. American .Midland Naturalist 102:
391-397.

.Vli'HRlN(;ER, R J., Jr. 1987. Late Holocene environments
on the northern periphei-y of the Great Basin. Final
report. Bureau of Land Management, Portland, OR.

Mehringer, R J., Jr., .vnd P E. Wigand. 1990. Compari-
son of late Holocene environments from woodrat
middens and pollen: Diamond Craters, Oregon. Pages
294-325 in J. L. Betancourt, T. R. Van Devender,
and P. S. Martin, eds., Packrat middens: the last
40,000 years of biotic change. University of Arizona
Press, Tucson.

Mitchell, V. L. 1976. The regionalization of climate in
the western United States. Journal of Applied
Meteorology 15: 920-927.

Muegc;ler, W F 1976. Type variability and succession in
Rocky Mountain aspen. Pages 16-19 in Proceedings,
Utilization and Marketing Tools for Aspen Manage-
ment in the Rocky Mountains. USDA Forest Service,
General Technical Report RM-29.

. 1985. Vegetation associations. Pages 4.5-55 in N. V.

DeByle and R. P Winokur, eds.. Aspen: ecology and
management in the western United States. USDA
Forest Service, General Technical Report RM-119.

Neilson, R. E 1987. On the interflice between current
ecological studies and the paleobotany of pinyon-
juniper woodlands. Pages 93-98 in R. Everett, ed..
Proceedings, Pinyon-Juniper Conference. USDA
Forest Service, General Technical Report INT-215.

NiCHOL, A. A. 1937. The natural vegetation of Arizona.
University of Arizona Technical Bulletin 68.

NOAA. 1993. National Climatic Data Center. Federal
Building, Asheville, NC.

Rich, E. E., A. M. Johnson, and B. R. Baker, eds. 1950.
Peter Skene Ogden's Snake Country journals: 1824—25
and 1825-26. The Hudson Bay Society, London.

SAS. 1986. SAS-STAT user's guide (release 603). SAS
Institute, Inc., Gary, NC.

Tausch, R. J., N. E. West, and A. A. Nabi. 1981. Tree age
and dominance patterns in Great Basin pinyon-
juniper woodlands. Journal of Range Management
34: 259-264.

Tausch, R. J., and N. E. West 1988. Differential establish-
ment of pinyon and juniper following fire. American
Midland Naturalist 119: 174-184.

Thompson, D. 1916. David Thompson's narrative. J. B.
Tyrrel, ed. The Champlain Society, Toronto, Ontario,
Canada

USDI-BLM. 1990. The juniper resources of eastern
Oregon. USDA, Bureau of Land Management Infor-
mation Bulletin OR-90-166.

Van Pelt N., R. Stevens, and N. E. West 1990. Sun'ival
and growth of immature Juniperus osteospenna and
Piniis edulis following woodland chaining in central
Utah. Southwestern Naturalist 35: 322-328.

Vasek, F C. 1966. The distribution and tiixonomy of three
western junipers. Brittonia 18: 350-372.

West, N. E. 1984. Successional patterns and productivity
of pinyon-juniper ecosystems. Pages 1301-1332 in
Developing strategies for range management.
Westview Press, Boulder, CO.

. 1988. Intermountain deserts, shrub steppes, and

woodlands. Pages 209-230 in M. B. Barbour and W D.
Billings, eds.. North American terrestrial vegetation.
Cambridge University Press, Cambridge, MA.

1995] Western Juniper Expansion 45

Young, J. A., and J. D. Budy. 1979. Historical use of Received 7 February 1994

Nevada's pinyon-juniper woodlands. Journal of Forest Accepted 8 June 1994

History 23: 113-121.
Young, J. A., and R. A. Evans. 1981. Demography and

fire history of a western juniper stand. Journal of

Range Management 34: 501-506.

RANGELAND ALPHA DIVERSITIES: HARVEY VALLEY,
LASSEN NATIONAL FOREST CALIFORNIAl

Raymond D. Ratliff2

Abstract. — Monitoring diversib,- usually begins by estimating alpha diversity of a plant community on a specific-
site. The objectives of this study were to provide alpha diversity benchmarks and to determine whether rangeland com-
munity basal cover characteristics explained variation in diversity estimates. Plant and surface component cover per-
centages were estimated on 51 plots (representing four vegetation types) on the Lassen National Forest, CA. Each plot
was sampled with 30 random, 102 basal point transects. Jackknife procedures were used to compute means and standard
errors for Margalef's diversity' index (D,„), which stresses species richness, and Simpson's index (Dj, which stresses
species dominance. Within vegetation tvpes, D,„ and D, did not rank all plots in the same order Highest D^, values
occurred with the most species. Highest D^ values occurred with comparatively few species but more uniform cover.
With either index, average diversity declined from the meadow to grassland to open shrub-grass to timber-bunchgrass
t>'pes. All possible subset regressions of diversity on the basal cover characteristics were computed. Portions of the vari-
ance accounted for by the best models were too low to allow prediction of D,„ and D,. The relation of alpha diversity to
rangeland health is discussed.

Kei/ worch: ecology, plant communities, Margalef's index, Simpson's index, monitoring, basal cover

Biological diversity (hereafter called diver-
sity) involves ecological processes, structures,
and functions and may occur at any spatial scale
(Society of American Foresters 1992). Diversity
refers to variety and abundance; it is variety or
multiformity — of different forms or kinds
(Stein and Urdang 1966). There are alpha,
beta, and gamma diversities (Whittaker 1972).
Alpha diversity is the variety that occurs with-
in a plant community of a specific site. A site
or stand is defined as an individual unit that is
homogeneous in vegetation, soil, topography,
microclimate, and history (West 1993). Beta
diversity is the variety of communities along a
gradient (e.g., topography, soil acidity, or mois-
ture regime) or on a given site through time.
Gamma or large-scale diversity is the variety
of plant communities, or the total number of
species present, or both in a specific geo-
graphic area (e.g., grazing allotment or water-
shed).

Diversity has two components, richness
and evenness (Ludwig and Reynolds 1988,
Magurran 1988). Richness refers to variety
(numbers) of species, for example. Evenness
refers to equality (abundance or numbers) of
species botanical composition, for example.

Diversity may or may not follow traditional
concepts of succession and increase from pio-
neer to climax plant communities or decrease
with rangeland deterioration. Over large areas
diversity may be higher if communities are at
several serai stages than if the entire area is at
a single serai stage. Within specific sites phys-
ical/chemical factors or intense competition or
both may work to reduce diversity (Odum
1959). Absence of an expected species may be
due to fi-equent disturbances, a low immigration
potential, an immature soil, or an inhospitable
moisture regime (del Moral and Wood 1988).

Nevertheless, because it may change with
the kind of management, diversity should be
assessed as part of range health evaluations.
Diversity indices provide information that
may not be immediately apparent from basic
measures of the plant community such as
cover and composition. High diversity of plant
species is important in maintaining processes
and flow pathways for energy and nutrients
within and among communities. Higher diver-
sity implies a greater number of occupied
niches (Whittaker 1972).

Protecting or enhancing diversit); or both,
are goals commonly set by policy or law. West

'This iirtit'li- was written and prepared by U.S. governnK'iit employees on official time; it is therefb
^Pacific- Southwest Research Station, USDA Forest Service, 2081 E. Sierra, Fresno, CA 93710.

I tlie pubhc domain and not subject to copyright.

1995]

Alpha DivERSiri'

(1993) gave four reasons for having diverse
plant communities: a sense of moral obligation
to living things, an aesthetic appreciation of
nature, economic benefits possible from them
(e.g., the gene pool for cultivated crops), and
the important array of services they provide
(e.g., maintaining oxygen levels and cycling
nutrients).

A major cause of rangeland deterioration is
selective grazing of preferred plants and sites
in similar patterns each year (Hormay 1970).
Even with conservative grazing, populations
of preferred plants on preferred sites may dis-
appear, thereby reducing the overall diversity
of vegetation. If such populations are ecotypes
(Odum 1959), the ability of the species to
recapture site resources is reduced.

Because nature abhors a vacuum, other
species may increase or invade as those pre-
ferred by livestock decrease in abundance
(Dyksterhuis 1949). As a result, plant species
diversity may be higher rather than lower
under grazing, at least initially. As preferred
species decrease and less preferred ones in-
crease, their abundances tend to become more
even (Dyksterhuis 1949). With continued
deterioration, species not previously able to
compete tend to invade and become established
and thereby increase species richness. The new
plant community, though possibly comprising
more species that are more evenly abundant,
may cover less total area, and higher diversity
may be associated with greater amounts of
bare soil.

Increasingly, land managers are asked to
monitor and determine change in diversity.
Monitoring diversity usually starts with an esti-
mation of alpha diversity for plant communities
on specific sites. Such estimates are rare for
rangelands. To derive the greatest benefit from
monitoring efforts, managers must know what
constitutes high and low diversity in given situa-
tions. They need to know how diversity changes
when other commonly estimated properties of
the site change (e.g., litter cover and amount
of bare soil).

Seldom will examples of pristine or climax
plant communities be available for developing
diversity guides. Current plant communities
represent the sums of all past influences.
Current vegetation and site characteristics,
therefore, must serve as benchmarks from
which to develop guides and evaluate future
change.

The objectives of this study were (1) to pro-
vide local rangeland managers with indices of
alpha diversity from plant communities to use
as guides of expected diversity for similar
sites, and (2) to question whether variation in
basal cover percentages of common and
important indicators of rangeland health could
explain variation in diversity. Although the
findings are specific to the study area, it is
hoped they may assist others dealing with
questions of plant species diversity on range-
lands.

Methods

Study Plots

During 1964 and 1965, 51 plots were estab-
lished on the Harvey Valley and neighboring
grazing allotments of the Lassen National
Forest, CA (Radifif et al. 1972). The plots were
either 0.1 ha or 0.2 ha and unevenly distrib-
uted among meadow (8), open grassland (13),
open shrub-grass (12), and timber-bunchgrass
(18) vegetation types. These plots were used
for evaluating range condition (health) at
Harvey Valley relative to the neighboring
allotments.

Meadows ranged from ephemeral lake sites
with hardpans to deep, organically rich soil of
drainage bottoms. Open grasslands included
those dominated by shorthair sedge {Carex
exserta) and those where shorthair sedge had
been replaced by grasses. Open shrub-grass
areas included silver sagebrush {Artemisia
cana), black sagebrush (A. arhuscida), big sage-
brush (A. tridentata), and bitterbrush {Purshia
tridentata) subtypes. The timber-bunchgrass
types were all in second-growth ponderosa
pine {Pinus ponderosa). Some of them had bit-
terbrush and big sagebrush along with grasses
in the understory.

Data Collection

Data used to estimate alpha diversity on
each plot were actual point contacts (hits) with
plant bases or soil surface components (gravel,
litter, rock, bare soil, and large woody debris)
and shrub crown area. A hit on a shrub was
recorded when a point contacted the shrub
crown or was within its projected crown area
at the soil surface. For each plot 3060 hits were
recorded, consisting of 102 points (in regularly
spaced 3-point quadrats) on each of 30 ran-
domly placed transects. Points in a quadrat

Great Basin Naturalist

[Volume 55

were at 23-cni eenters and projected vertical-
ly. Within transects, quadrat spacing was
either 0.6 ni or 0.9 m, depending upon plot
width. Basal cover percentages (proportions of
the surface occupied by different plants and
surface components) were calculated from the
hits and summarized (Radiff et al. 1972).

Diversity Indices

Two indices of diversitv were used: (1)
Margalef's {D„, = (S - l)/ln N}, where S is
the number of species and N is the total num-
ber of individuals (hits) for all species and (2)
Simpson's

(D^ = 1/D), where D = I {nj(nj- - 1)/

i = l

N(N - 1)} and n^- is the number of individuals
(here the percentage cover) of the ith species
(Magurran 1988). D,^^ was selected for its sim-
plicity and because it stresses the species rich-
ness component. D^. was selected because it is
well known and stresses the species evenness
(dominance) component. In addition, these
indices were selected because they do not
require testing assumptions regarding the
underlying distributions of species abundance.
An overall estimate of diversity was com-
puted for each plot using each index. Then 30
new diversity estimates were computed using
the jackknife procedure. This procedure con-
sisted of deleting each transect in turn from
the data set. From each new estimate and the

overall estimate, a pseudovalue (related form)
was computed. From the pseudovalues, means
and standard errors for the two indices were
derived for each plot. Use of the jackknife pro-
cedure to improve estimates of diversity and
piovide a way of calculating confidence inter-
vals was suggested by Magurran (1988).

Basal Cover Relationships

Contributions of basal cover of various char-
acteristics to the variance in estimates of alpha
diversity were examined. Characteristics for
each plot were basal covers of grasses, grass-
like herbaceous plants, forbs, shrubs, and soil
surface components. All possible subset re-
gressions of D,,^ and D^^ on the characteristics
were computed using the Mallow's-Cp criteri-
on of the REG procedure (a multiple linear
regression program) of the SAS Institute, Inc.
(1982). Subset regression models explaining
most variation in the indices were selected for
study. The Pearson correlation matrix was com-
puted, using the correlations (CORR) module
of SYSTAT (Wilkinson 1989), to help assess
the influence of individual characteristics on
the indices.

Results

Alpha Diversity Indices

Diversity indices and basal cover values are
available for all 51 plots. Here, only those plots
within each vegetation type ranking lowest

Table 1. Numbers of species, dominant species and percentage composition, and jackknifed means and standard
errors (SE) for Margalef's and Simpsons diversitv' indices^ for vegetation tvpe^ benchmarks in 1964-65, Eagle Lake
Ranger District, Lassen National Forest, CA.

Dominant

Diversi

ity index

Margalef"

Simpson';

Veg.

No. of

Composi

tion

type

species

Species

percentage

Mean

Eleocharis paltt.stris

1.0

0.2

2.8

0.2

Deschampsia caespitosa

4.0

0.4

2.4

0.2

Juncus balticu,s

2.3

0.2

9.9

0.7

Carex exserta

1.6

0.2

1.9

0.1

C. exserta

1.6

0.2

1.6

0.1

C. exserta

3.2

0.3

3.8

0.3

Festiica idiihocnsis

2.0

0.3

5.8

0.3

Artemisia tridcntata

0.8

0.2

1.0

0.0

A. arhusruhi

3.6

0.5

2.6

0.4

Leptodactijlon pun<s,ens

1.9

0.3

5.0

0.3

Purshia tridcntata

0.8

0.0

1.4

0.1

Artemisia tridcntata

3.2

0.4

1.8

0.2

Carex rossii

1.1

0.2

4.7

0.4

'Highlighted values

are the highest and lowest for each index within vegetation types.

2Vegefad(

m types follow Ratliff et al. (1972): MD = meadow, GR

= open grassi

land, SG =

= open shnib-grass, TB

= timber-hui

nchgrass.

1995]

Alpha Diversity

Fig. 1. Meadow di\c'rsit>- benchmarks: (a) Elencharis ixihislris-, (h) Deschampsia caespitosa-, and {c) Jiincus balticus
-dominated plots; Eagle Lake Ranger District, Lassen National Forest, CA.

Great Basin Naturalist

[Volume 55

and highest for D„j and D^. are specihcalK' dis-
cussed. Those plots are considered diversity
benchmarks for their vegetation types in and
near the Haney Valley allotment.

Meadow. — D,^ in the meadows was lowest
on a plot with just six species and demonstrat-
ed the effect of lack of richness (Table 1). The
site was an ephcnicral lake meadow (Fig. la)
where dominant species covered 3.3% of the
surface. Among the meadow plots, percentage
litter cover was lowest and percentage bare
soil was highest (Table 2).

D,„ was highest, but D^ was lowest on a
meadow with 19 species. That finding demon-
strated the effect of good variety with uneven
abundance. The site was a basin meadow, pos-
sibly an ancient lake (Fig. lb). There the domi-
nant species covered 5.7% of the surface. Only
one species, among the others, contributed as
much as 5% to the composition. Percentages
of litter and bare soil were higher and lower,
respectively, than averages for tlie meadow plots
(Table 3).

D^ was highest on a plot with 14 species.
The site was a groundwater-fed meadow (Fig.
Ic). Evenness in species abundance with mod-
erate variety was demonstrated. Four species
(including the dominant) each constituted
more than 10% of the composition but less
than 1% of the basal cover Only one species,
among the others, contributed less than 1% to

the composition. Total live plant cover was
below average, but percentages of litter and
soil cover were well above and below the
averages, respectively.

Grassland. — Both D„^ and D^ were lowest
on grassland plots, with nine species (Figs. 2a,
2b), respectively. Shorthair sedge was the
main contributor to the composition. In the
case of Dj^^, three species each contributed 5%
or more, and five species each contributed 1%
or less. In the case of D^, only one species,
other than shorthair sedge, contributed as
much as 5% of the composition. For the plot
with low D„^ the evenness component was
better, litter cover was higher, and bare soil
cover was lower than for the plot with low D^,.

D,^^ was highest on a plot with 19 species
(Fig. 2c). Shorthair sedge, Idaho fescue {Festiica
idahoensis, 20%), and Sandberg bluegrass {Poa
sandbergii, 9%) were main contributors to the
composition. Sixteen species contributed less
than 5% each. Among the grassland plots, this
plot had the highest live plant cover and was
well above average in litter cover and well be-
low average in percentage of bare soil.

Idaho fescue dominated the plot with high-
est D^ (Fig. 2d). Four of the other 10 species
present each made up more than 10% of the
composition; two species each made up about
6%. While the evenness component of diversi-
ty was good and total live plant cover was

Table 2. Percentages of liasal cover for plant gronps and surface components for vegetation type benchmarks
1964-65, Eagle Lake Ranger District, Lassen National Forest, CA.

Perc

entage basal

cover

Plant

groups'

Surface

components-

Vegetation t\'pe'^

,^1"

Nieadow

4.6

0.6

6,2

47.1

—

46.6

—

6,1

2.2

0.9

—

0,7

9,2

0.1

80.2

—

9.8

—

1..3

2.6

1.2

—

5,1

—

90.3

0.1

4.5

—

Open grassland

1.5

5.4

0.5

0.2

15.0

7,6

3,2

.57,1

—

17.1

—

0.7

4.7

0,4

0.2

12.0

6,0

6.0

.50,6

0.5

24.9

—

4.8

5.3

0,6

0.3

7,6

11,1

1.3

60,1

19.9

4.4

2,5

0,3

3.0

1,7

10,2

11.3

26.1

2.6

48.1

—

Open shrub-grass

0.7

—

25.4

16,2

26,1

18.9

28.4

—

10.4

—

1.8

0.5

0,8

4.6

8,2

7.6

4.0

.38.0

42.1

—

3.6

3.4

1,1

4.5

3,1

12.5

8.6

21.0

1.2

53.7

—

Timl)er-bunchgrass

1.2

0.6

—

21.6

11,8

23.3

1.2

,54.0

0.5

6.5

2.7

1.0

0.4

0.1

4.5

5,6

5.9

2.9

76.2

4.0

4.6

0.7

I.O

0.6

—

0.6

3,1

2.1

2.8

62.3

9.4

13.8

6.6

'gr = grasses, gl = grasslike herbaceous plants, bl = hroadleaf heibaceous plants (forbs), sli = shrubs, dp

+ bl + sh).

^Gr = gravel, Li = litter, Ro = rock. So = soil. Wo = large woody debris,

^Types follow Ratliffet al. (1972), and plot order is (he s.une as in Table 1.

■'t = less than 0.1% of basal cover

deail attached eo\er. Ip = li\e plant co\er (g,r + gl

1995]

Alpha Diversity'

Table 3. Average diversiW indices and percentages of basal cover for plant groups, and surface components by vege-
tation type. Eagle Lake Ranger District, Lassen National Forest, CA, 1964-65.

Diversity
inde.x'

Percentage basal cover

Plant

groups^

Surface

components'^

Vegetation type

D,„

■•-i'

Meadow

Grassland

Open shrub-grass

Tiniber-bunchgrass

2.29
2.28
2.28
L75

4.38
3.63
2.78
2.39

2.1
3.3
1.6
1.1

3.6
3.6
1.3
1.1

0.8
1.2
0.6
0.1

0.8
14.4

7.7

0.3

5.6

10.7

4.0

6.4

8.9

18.0

10.0

0.2
9.3

7.2
2.5

77.2
39.0
30.2
60.2

0.1
1.5
0.2
3.0

15.8
35.7
33.8
16.4

t
3.9

'D„, = Margalef s index, D^ = Simpson s index.

^gr = grasses, gl = grasslike herbaceons plants, hi = broadlcaf herbaceous plant,s (forbs).

+ bl + sh).

''Gr = gravel, Li = litter, Ro = rock. So = soil. Wo = wood.

= shrubs, dp = dead attached coven Ip = live plant cover (gr + ]

above average, Idaho fescue covered only 3%
of the surface and htter cover was well below
but bare soil was well above average.

Shrub-grass. — Both indices were lowest
on an open shrub-grass plot where big sage-
brush contributed over 95% of the composi-
tion (Fig. 3a). Only one other species, bottle-
brush squirreltail {Sitanion hystrix), made up
as much as 1%, and only five species occurred
on that plot. This finding demonstrates the
effects of both low variety and low evenness
on diversity. Among the shrub-grass plots, this
plot was second highest in total live plant
cover (nearly all sagebrush), highest in gravel
cover, and lowest in bare soil. This suggests
soil loss and formation of pavement.

Black sagebrush dominated the plot with
highest D„^ (Fig. 3b). Of the 17 species on that
plot, 12 of them each contributed less than 3%
of the composition. The plot was above aver-
age in both litter and soil cover, but lowest in
total live plant cover

The plot with highest D^, (Fig. 3c) had just
11 species and was dominated by false phlox
{Leptodacttjlon ptmgens). Five other species
combined contributed nearly 62% of the com-
position. Among the shrub-grass plots, this
plot was well below average in litter cover but
highest in bare soil.

TiMBER-BUNCHGRASS. — Both indices were
lowest (Fig. 4a) on a timber-bunchgrass plot
with seven species. Bitterbrush contributed
over 80% of the composition. Three species
contributed 2% or more and three species
contributed less than 1% of the composition.
While total live plant cover was above aver-
age, litter was near average and bare soil was
well below average; there were few species,
and they were unevenly abundant. This plot
was similar in diversity to the shrub-grass plot
with D„, and D^. both low.

D„^ was highest on a plot with 13 species
(Fig. 4b). Nine of them contributed 1% or less
of the composition, thereby demonstrating
that high evenness is not required when vari-
ety is the main component of diversity' consid-
ered. Big sagebrush dominated the understory
and covered 4.3% of the surface. Litter cover
was well above and bare soil was well below
average for the timber-bunchgrass plots.

By contrast, D^ was highest on a plot with
just six species (Fig. 4c). Ross sedge {Carex
rossii) contributed most of the composition
(0.6% of the surface cover), three species con-
tributed 15-26% each, and two species con-
tributed 3% each, thereby demonstrating that
high variety is not required when evenness is
the main component of diversity considered.
Percentages of soil and litter cover were near
average for the timber-bunchgrass plots.

Beta Diversity Indices

Statistical comparisons of diversity among
communities and vegetation types were not
made. Nevertheless, average values for both
indices declined from meadow to grassland to
open shrub-grass to timber-bunchgrass tvpes
(Table 3).

Relative plot ranking (high to low diversity)
depends on the index used, and inconsistent
ranking by D,^^ and D^ was expected. Among
the open shrub-grass and timber-bunchgrass
types only two plots ranked the same, those
with lowest diversity by both indices. Rankings
by D,.,.j and D^ were the same for 3 of the 8
meadow plots and 2 of the 13 open grassland
plots.

Basal Cover Relationships

Meadow and grassland plots had higher
average diversity indices tlian open shrub-grass

Great Basin Natuiulist

[Volume 55

Fig. 2. Open grassland diNeisity Ijenchniaiks: (a, h, c) Carex exserta- and (d) Festiica i^ffl/ioensis-dominated plots;
Eagle Lake Ranger District, Lassen National Forest, CA.

or timber-bunchgrass plots, but lower average
percentages of live plant cover (Table 3). Total
live plant cover was largely ci property of shrub
cover because projected crown hits were in-
corporated into the data base.

Significant portions of variances in the
diversity indices (all 51 plots included) were
accounted for by variation in percentages of
some basal cover characteristics. Forty-seven
percent of the variation in D„-, and 27% of the
variation in Dj. were explained by the best
models (Table 4).

Dn, = a + grbi + glb2 + shb3 + Grl)4 +

Sob5 + Wobg + error) and D^ = a + shb^ +
Grb2 + error, where a, gr, gl, sh, Gr, So, and
Wo are explained in Table 4; and the (bj)'s are
the coefficients.

Although gravel and bare soil were includ-
ed in the model for D„^, they did not signifi-
cantly correlate with D„^. Also, while in the
model for D^, gravel was not significantly cor-
related with D^.

Individually, correlation with D,^^ was posi-
tive for grasses (r = .471) but negative for grass-
like plants (r = -.014), shrubs (r = -.320), and
wood (r = -.348). Correlation of Dj. with
shrubs was negative (r = -.507), also.

1995]

Alpha Diversity

Fig. 2. Continued.

Discussion

Alpha Diversity

Many diversity indices are available to the
land manager. Although a particular diversity
index may be preferred, it is generally best to
use one that stresses species richness and one
that stresses evenness (dominance), such as D„,
and Dj, respectively. Doing so allows the man-
ager to consider both components of diversity.
The richness component of diversity may in-
crease at the expense of the evenness compo-
nent, or vice versa. Also, those indices that

stress richness and those that stress evenness
tend to be poorly correlated (Magurran 1988).

Beta Diversity

Data used in this study represent single-
time samples and were not designed to esti-
mate beta diversity. Testing for differences in
diversity using such data was not considered
reliable (West and Reese 1991).

Nevertheless, diversity indices for different
but closely similar plots or communities, when
computed by the same methods, should be
nearly equal. With time or different treatment,
wade divergence of the indices may occur.

Great Basin Naturalist

[Volume 55

Fig. 3. Open shrub-grass diversity- benchmarks: (a) Artemisia tridentata-. (b) A. arbuscuki-, and (c) Leptodactyh.
pun^ens-domuvdh'd plots; Eagle Lake Ranger District, Lassen National Forest, CA.

1995]

Alpha Diversity

'Jr*^^?

Fig. 4. Timber-bunchgrass diversih lienchmarks: (a) Ptirshki trklentata- (b) Artemisia tridentata-, and (c) Carex rossii
-dominated plots; Eagle Lake Ranger District, Lassen National Forest, CA.

Great Basin Naturalist

[Volume 55

Table 4. Best model multiple linear regre.ssion coeffieients, tests of siKnificance (T), and pn)l)al)ilities of signifieance
(P) for Margalef's and Simpson's diversit>- indices; Eagle Lake Ranger District, Lassen National Forest, CA, 1964-65.

Symbol

Diversity

index

.\largalef\s

Simpson s

Variable

Coeff".'

CoeO:

Constant

2.696

8.436

.000

3.584

12.372

.000

Grasses

0.238

3.522

.001

Grasslike

plants

-0.161

-3.285

.002

Shrubs

-0.036

-2.722

.009

-0.102

-4.194

.000

Gravel

-0.026

-1.505

.139

0.033

0.970

.337

Bare soil

-0.007

-1.197

.238

Wood

-0.128

-3.173

.003

'Regression coefficient

Permanent plots represent a resource for
assessing beta diversity responses to land man-
agement practices. Although sampling a site to
include within- and between-season variation
is desirable, doing so is seldom possible, given
time and monetaiy constraints. As an alternative,
one might restrict sampling to times when
selected species indicators are in specific phe-
nologic stages (e.g., budding or flowering).

Basal Cover

Because of the usual dominance of a single
species and because that species tends to
occupy high proportions of an area, reductions
in diversity indices with increases in shrub
cover may be expected.

Both diversity indices may be related posi-
tively or negatively to characteristics of basal
cover or to soil properties. Nevertheless, D,^^
was related to a greater number of characteris-
tics than D^, suggesting that D„, may be the
more desirable index for comparing plant com-
munities of different sites or plant communities
present through time on a given site.

Conclusions

For similar communities we can expect
plant species diversity to be highest in the
meadow and lowest in the pine-bunchgrass
types. High and low values of Margalef's and
Simpson's diversity indices are available for
benchmark plots of different vegetation types in
and near the Harvey Valley allotment. Diversity'
indices for and averages among 51 plots are
available by vegetation types.

The influence of species richness on D„^
was clearly evident. D„, tended to be highest
with the greatest numbers of species. Frequently
that occurred when one species was clearly

dominant and the others contributed little
plant cover The inff uence of evenness in abun-
dance on D^ was clearly evident. D^ tended to
be highest when species were more or less
evenly abundant. Frequently that occun-ed with
relatively few species. Few species with one
contributing a high percentage of the compo-
sition produced low values of both indices.
Situations with many species, all contributing
equally to the composition, were not encoun-
tered, but such situations should give high val-
ues of D,^^ and Dj,.

Higher diversity did not necessarily mean
greater plant cover or greater forage cover or
more litter or less bare soil. While some rela-
tionships between diversity and basal cover
values were significant, coefficients of deter-
mination were too low to allow either of the
best models to be used to predict diversity'.

Neither index should be relied on apart
from other information for evaluating range-
land health. Nevertheless, plants capture the
sun's energy and pass it as food for other orga-
nisms, and a high degree of plant diversity
may equate with high diversity in other parts
of the biotic community.

Literature Cited

DEL xMoiUL, R., and D. M. Wood. 1988. The higli elevation

flora of Mount St. Helens, Washington. Madrofio 35:

309-319.
Dvksterhuls, E. J. 1949. Condition and management of

range land based on quantitative ecolog>-. Journal of

Range Management 2: 104-115.
HORNLAV, A. L. 1970. Principles of rest-rotation grazing

and multiple-use land management. Training te.xt

4(2200). USDA, Forest Service, Washington, DC.

26 pp.
Ludwk;, J. A., .\ND J. F REYNf)LDS. 1988. Statistical ecology.

John Wiley & Sons, New York, NY.

1995]

Alpha DivERSiTi'

Magurran, a. E. 1988. Ecological diversity and its mea-
surement. Princeton University Press, Princeton, NJ.

Odum, E. R 1959. Fundamentals of ecology. W. B. Saunders
Co., Philadelphia, PA.

Ratliff, R. D., J. N. Reppert, .vnd R. J. McConne.n. 1972.
Rest-rotation grazing at Harvey Valley . . . range
health, cattle gains, costs. USDA, Forest Service,
Pacific Southwest E.xperiment Station. Research
Paper PS\V-77. 24 pp.

SAS Institute, In'c. 1982. SAS user's guide; statistics. 1982
edition. SAS Institute, Inc., Gary, NC.

Society of American Foresters. 1992. Biological diver-
sity in forest ecosystems, a position of the Society of
American Foresters. Journal of Forestn 90: 42—43.

Stein, J., and L. Urdang, eds. 1966. The Random House
dictionary of the English language. Random House,
New York, NY.

West, N. E. 1993. Biodiversity of rangelands. Journal of
Range Management 46: 2-13.

West, N. E., and G. A. Reese. 1991. Gomparison of some
methods for collecting and analyzing data on above-
ground net production and diversity of herbaceous
vegetation in a northern Utah subalpine context.
Vegetatio 96: 145-163.

Whittaker, R. H. 1972. Evolution and measurement of
species diversity. Taxon 21: 213-251.

Wilkinson, L. 1989. SYSTAT: the system for statistics.
SYSTAT, Inc., Evanston, IL.

Received 26 July 1993
Accepted 26 May 1994

EFFECTS OF SALINITY ON ESTABLISHMENT OF POPULUS

FREMONTII (COTTONWOOD) AND TAMARIX RAMOSISSIMA

(SALTCEDAR) IN SOUTHWESTERN UNITED STATES

Patrick B. Shafroth', Jonathan M. Friedman', and Lee S. Ischinger^

Abstract. — The e.xotic shrub Tatnarix minosissimu (saltcedar) has replaced the nati\t' I'opulus jninontii (cottonwood)
along many streams in southwestern United States. We used a controlled outdoor experiment to examine the influence
of river salinity on germination and first-year survival of P. fremontii van wislizenii (Rio Grande cottonwood) and T.
ramosissima on freshly deposited alluvial bars. We grew both species from seed in planters of sand subjected to a declin-
ing water table and solutions containing 0, 1, 3, and 5 times the concentrations of major ions in the Rio Grande at San
Marcial, NM (1.2, 10.0, 25.7, and 37.4 meq l-l; 0.11, 0.97, 2.37, and 3.45 dS m-i). Germination of P. fremontii declined
by 35% with increasing salinity (P = .008). Germination of T. ramosissima was not affected. There were no significant
effects of salinity on mortality' or above- and belowground growth of either species. In laboratory tests the same salini-
ties had no effect on P. fremontii germination. P. fremontii germination is more sensitive to salinity outdoors than in cov-
ered petri dishes, prolwbly because water scarcity resulting from e\aporation intensifies the low soil water potentials
associated with high salinity. River salinity appears to play only a minor role in determining relative numbers of P. fre-
montii and T. ramosissima seedlings on freshly deposited sandbars. However, over many years salt becomes concentrat-
ed on floodplains as a result of evaporation and salt extrusion from saltcedar leaves. T. ramosissima is known to be more
tolerant of the resulting extreme salinities than P. fremontii . Therefore, increases in river salinities could indirectly con-
tribute to decline o{ P. fremontii forests by exacerbating salt accumulation on floodplains.

Key words: exotic species, Tamarix ramosissima, Populus fremontii, river salinity, seedling estahlisliment, Rio Grande,
riparian vegetation, Bosque del Apache National Wildlife Refuge.

In the last century the exotic shi-ub saltcedar
{Tamarix ramosissima Ledebour) has spread
throughout southwestern United States, where
it now dominates many riparian ecosystems
(Bowser 1958, Robinson 1965). In many areas
T. ramosissima has replaced stands dominated
by the native Fremont cottonwood {Populus
fremontii Wats.; Campbell and Dick-Peddie
1964, Ohmart et al. 1977), decreasing the habi-
tat of Neotropical migrant birds (Anderson et al.
1977, Cohan et al. 1978) and altering fluvial
processes (Graf 1978, Blackburn et al. 1982).
Understanding the factors controlling estab-
lishment of T. ramosissima and P. fremontii can
aid in managing these species.

Successhil invasion by Tamarix in the South-
west has been attributed to many factors. Much
of the early spread probably resulted from the
coincidental timing of clearing of P. fremontii
stands by early settlers and the availability of
Tamarix seed (Campbell and Dick-Peddie
1964, Harris 1966, Horton and Campbell
1974, Ohmart et al. 1977). Subsequent spread
resulted largely from effects of damming and

channelizing southwestern watercourses.
Reductions in the magnitude of high flows and
associated reductions in channel movements
decreased the formation of bare, moist alluvial
bars, which provide ideal P. fremontii seedling
habitat (Ohmart et al. 1977, Stromberg et al.
1991). Smaller peak flows have also reduced
leaching of salts from floodplain soils (Busch
and Smith in press), perhaps favoring the salt-
tolerant Tamarix (Everitt 1980, Brotherson
and Winkel 1986, Jackson et al. 1990). Flow
regulations that have altered the historical
timing of peak flows may have inhibited P. fre-
montii regeneration because of its short period
of seed dispersal and viability in early summer
(Horton 1977, Everitt 1980), but they have
enhanced Tamarix regeneration because of its
abundant seed production throughout the
growing season (Merkel and Hopkins 1957,
Tomanek and Ziegler 1962, Wanen and Turner
1975, Horton 1977). Finally, successful inva-
sion of T ramosissima has been attributed to
its superior ability to resprout following fire
(Busch and Smith 1993).

'National Biological Siii-vcy, Miclcoiitinenf Ecological Science Center, Fort Collins, CO 80525-3400.

1995]

SALiNiTi' Effects on Populus and Tamarix

We conducted experiments to examine the
influence of river salinity on germination, sur-
vival, and growth of Popuhis fremontii var wis-
lizenii (Rio Grande cottonwood) and T. ramo-
sissima on freshly deposited alluvial bars, the
principal habitat for seedling establishment of
both species. Field observations have suggest-
ed that P. fremontii is more negatively affected
by high salt concentrations than T. ramosissi-
ma (Brotherson and Winkel 1986, Anderson
1989). Laboratoiy studies have confirmed this
difference by exposing seedlings and cuttings
of these species to varying concentrations of
NaCl and CaCl2 (Jackson et al. 1990, Siegel
and Brock 1990). Two factors potentially con-
found the relationship of results of laboratory
studies to field conditions. First, the mix of
salts found in riparian ecosystems typically
includes many constituents other than Na, Ca,
and Cl. In many plants, salinit>' effects result
from toxicity of specific ions as opposed to
osmotic stress (Greenway and Munns 1980).
Second, moisture availability is lower and
more variable in the field than in these labora-
tory studies. This factor is important because
low soil water potential caused by high salinity
is exacerbated by low soil moisture content.
We addressed these concerns by exposing T.
ramosissima and P. fremontii seedlings to four
different concentrations of a mix of salts
designed to mimic ion concentrations in the
Rio Grande. The experiment was conducted
outdoors in planters subjected to a controlled
water-table drawdown. Experimental condi-
tions were designed to simulate alluvial bars
along the Rio Grande in central New Mexico,
where once-extensive P. fremontii forests have
largely been replaced by T. ramosissima thick-
ets (Campbell and Dick-Peddie 1964). Our
outdoor experiments were supplemented by
studies of germination under similar salinity
treatments in the laboratoiy.

Methods

Seedling establishment experiments were
conducted outdoors in 1993 near Fort Collins,
CO, at latitude 40° 35' north, longitude 105° 5'
west, and elevation 1524 m. Twelve 122 x 92-
cm (diameter X depth) epoxy-lined steel tanks
contained six 30 X 100-cm planters made of
PVC pipe. Holes 1.26 cm in diameter were
drilled into the lower 10 cm of each planter to
allow water exchange, and the planters were

filled to 92 cm with washed coarse sand
(approximately 6% gravel [>2000 fim], 78%
sand [> 300-2000 /am], 16% fine sand
[> 75-300 ^im], and <1% silt and clay).

Four salinity treatments were each replicat-
ed in three tanks (12 tanks total). Each tank
contained three planters of P. fremontii var.
wislizenii and three of T. ramosissima. Thus,
the experimental unit for each species was a
group of three planters within a tank. To avoid
pseudoreplication, responses were measured
as the mean value of the three planters. The
results for the two species were analyzed as
separate, completely randomized experiments
with four treatments and three replicates per
treatment.

The tanks were filled with water from the
Cache la Poudre River (a snowmelt stream low
in dissolved solids), and solutions containing
multiples (0, 1, 3, and 5 times) of the mean con-
centration of all major ions in the middle Rio
Grande were made. These four solutions con-
stitute treatments Ox, Ix, 3x, and 5x. Mean ion
concentrations were derived from eight mea-
surements from the conveyance channel at
San Marcial, NM, between October 1989 and
September 1991 (U.S. Geological Survey 1991,
1992). The following salts were added to make
treatment Ix: 309.9 mg h^ CaS04*2H20; 302.4
mg 1-1 NaHCOg; 122.0 mg H MgCf2*6H20;
70.1 mg 1-1 NaCl; 13.9 mg l-l K2S04.''Because
the coarse sand substrate was low in nutrients
(c£ Segelquist et al. 1993), 15 mg 1-1 of Fisons
Technigro fertilizer (16% N, 17% I^ 17% K) was
added to every tank.

At the time of planting and for 1 wk there-
after, the water level was 10 cm below the soil
surface. A 3.5-cm-week-l drawdown rate was
applied for the remainder of the growing sea-
son (17 June to late September). Water-table
drawdowns are associated with summer
declines in discharge along western streams.
The 3.5-cm-week-l drawdown rate was select-
ed because a previous study (Segelquist et al.
1993) indicated that it is within the optimal
range for establishment and growth of plains
cottonwood {Populus deltoides ssp. monilifera).

Flowering panicles of T. ramosissima were
collected on 17 May at the Bosque del Apache
National Wildlife Refuge (latitude 33° 46'
north, longitude 106° 54' west, elevation 1375
m). The panicles were air-dried for 48 h to
enhance opening of seed capsules. Collected
material was sifted through a series of soil

Great Basin Naturalist

[Volume 55

screens until clean samples of seeds were ob-
tained. Catkins of P. fremontii were collected
at the Bosque del Apache on 1 June. The cat-
kins were air-dried for 72 h to enhance open-
ing of seed capsules. Capsules were placed
between soil screens and seeds were separat-
ed from the cotton and capsules using forced
air. Seeds of both species were sealed in plas-
tic containers and refrigerated at 5°C (Zasada
and Densmore 1977). On 10 June, 100 P. fre-
montii seeds were planted in each of three
planters per tank, and 200 T. ramosissirna
seeds were planted in each of the other three
planters.

Electrical conductivity (EC) and tempera-
ture were measured using a Yellow Springs
Instrument Co., Inc., Model 33 S-C-T meter,
and pH was measured using a Corning 105
hand-held pH meter in conjunction with a
Coming ATC temperature probe and a Coming
general purpose combination electrode. EC
was measured weekly in every tank begining
12 June (17 measuring dates). Whenever EC
was measured, a representative water temper-
ature for that day was determined by averag-
ing the temperature values from hve randomly
selected tanks. All EC measurements were
corrected for temperature and reported at
25°C. Fourteen weekly measurements of pH
were made beginning 30 June. On 16 June, 14
July, 18 August, and 17 September, water sam-
ples from one randomly selected tank per
treatment were analyzed to determine con-
centrations of Ca, Mg, Na, K, CO3, HCO3, Cl,
SO4, and NO3. Ca, Mg, Na, and K were deter-
mined by inductively coupled plasma emis-
sion spectroscopy (ICP; EPA method 200.0,
United States Environmental Protection
Agency 1983); CO3 and HCO3 were deter-
mined by titration (EPA method 310.1, United
States Environmental Protection Agency
1983); Cl, SO4, and NO3 were determined by
ion chromatography. Concentrations are
reported in meq 1~^ to facilitate comparison of
our solutions to solutions in other studies and
because meq 1~^ can be related easily to elec-
trical conductivity, which is commonly report-
ed in the context of salinity studies.

On 29 September 1993 (day 112) we mea-
sured the shoot length of every living seedling.
We harvested all live seedlings in early October
To harvest, we lifted a planter and laid it hori-
zontally in a water-filled basin. The planter
was then slowly lifted upside down, leaving

the substrate column and seedlings in the
basin. We gently separated seedlings from the
sand and water and measured total length of
every harvested seedling. Mean root lengths
were determined by subtracting the mean
shoot length for a planter from the mean total
length in that planter Roots and shoots were
separated for both species, and P. fremontii
leaves were stripped from the stems. Roots,
shoots, and leaves were dried at 60 °C for 72 h
and weighed.

One-way analysis of variance (SAS Institute,
Inc. 1990) was used to assess the significance
of treatment differences within the tvvo species
for five variables: percent of planted seeds alive
at the end of the experiment ("end-of-season
survival"), shoot length, root length, per-plant
aboveground biomass, and per-plant root bio-
mass. For all variables the mean value of the
three planters in a tank was the unit of analysis.
The arcsine transformation was applied to end-
of-season survival values to meet die equal vari-
ance assumption (Snedecor and Cochran 1980).

Data from the Colorado Climate Center
were used to determine the difference be-
tween precipitation and open-pan evaporation
(adjusted with pan coefficient = 0.73) for the
period 1 June-30 September 1993 in Fort
Collins. Evaporation at Fort Collins exceeded
precipitation by 26.2 cm during this period. The
same calculation was made for the Bosque del
Apache using data from the Western Regional
Climate Center for the years 1975 through
1990. Precipitation data are from the Bosque
del Apache National Wildlife Refuge, and
open-pan evaporation data are from Socorro,
NM (latitude 34°5' north, longitude 106° 53'
west, elevation 1399 m; pan coefficient = 0.73).
Growing-season evaporation at the Bosque del
Apache exceeded precipitation by an average
of 40.6 cm; n = 16, maximum = 51.0 cm, and
minimum = 32.3 cm during these 16 years.

We performed laboratoiy gemiination exper-
iments in January 1994. Five 25-seed repli-
cates of five salinity treatments were com-
pletely randomized for both T. ramosissirna and
P. fremontii. Seeds were sowed in 7.5-cm petri
dishes containing a Whatman #3 filter and 7
ml of a treatment solution. Petri dishes were
placed in a Percival Model 1-35 biological
incubator after sealing the dish tops with Para-
film. Temperature in the incubator was 20 °C
throughout the experiment, and petri dishes
were exposed to 16 h of light and 8 h of dark-
ness each day. Four of the treatment solutions

1995]

SALiNiTi' Effects on Populus and Taaiarix

were the same as those used in the estabhsh-
ment experiment (0, 1, 3, and 5 times the con-
centration of the Rio Grande at San Marcial,
NM); the fifth sokition contained 7 times the
concenti'ation of the Hio Grande. Genninants in
every petri dish were counted after seven days.
A seed was considered gemiinated if it exliibited
expanded cotyledons and an elongated radicle.
The arcsine transformation was applied to per-
cent germination values to meet the equal
variance assumption, and one-way analysis of
variance was performed on the transformed
values (SAS Institute, Inc. 1990). When germi-
nation equaled 100%, the proportion was
counted as (n - 0.25)/n, where n = the num-
ber of seeds planted (Snedecor and Cochran
1980).

Results

EC and pH in the tanks varied little within
treatments over the course of the experiment
(Table 1). Mean temperature in the tanks was

21.7°C (standard en-or = 0.8, n = 17). Concen-
trations of measured chemical constituents in
different treatments did not increase propor-
tionally to the quantities of salt originally
added, indicating that salts (especially CaC03)
precipitated at higher concentrations (Table 1).
Nevertheless, concentrations increased across
treatments, with total concentrations ranging
from 0.7 meq h^ (0.11 dS m"^) in treatment Ox
to 37.4 meq \~^ (3.45 dS m~l) in treatment 5x
(Table 1).

For P. fremontii there was a significant
treatment effect (F = .003) on end-of-season
sui-vival, but not on any of the four measured
growth variables (Table 2). End-of-season sur-
vival was negatively associated with increasing
salinity: survival was greatest in treatment Ox
and lowest in treatment 5x. Because the end-
of-season sui"vival variable combines germina-
tion and mortality, we analyzed the arcsine-
transfomied number of seedlings 7 d after plant-
ing (germination), and the arcsine-transformed
difference between germination and end-of-

Table 1. Chemical analysis of tank water for four treatments in the outdoor establishment experiment in Fort Collins,
CO. For ion concentrations (n = 4), minimum and ma.\imum values are presented in parentheses below treatment
means. For electrical conductivity' {n = 51) and pH (n = 42), means ± 1 standard error are presented.

Treatment

Factor

Ca (mmol h^)

0.36
(0.20, 0.52)

1.82
(1.71, 2.00)

4.02
(3.49, 4.83)

4. .54

(3.02, 7.02)

Mg (mmol 1-1)

0.11
(0.08,0.16)

0.60
(0.46, 0.75)

1.65
(1.47, 1.97)

2.62

(2.28, 2.97)

Na (mmol l^)

0.17
(0.09, 0.28)

4.85
(4.41,5.11)

13.87
(11.91, 15.49)

22.24
(19.33, 24.65)

K (mmol 1-1)

0.08
(0.06, 0.09)

0.26
(0.20, 0.34)

0.51
(0.44, 0.,55)

0.79
(0.72, 0.90)

HCO3 (mmol 1-1)

1.04

(0.62, 1.44)

3.92
(3.24, 4.44)

8.34
(7.29, 9.96)

9.60

(5.87, 15.74)

CI (mmol 1-1)

0.10

(0.07,0.14)

2.47

(1.88, 2.82)

7.10
(6.96, 7.31)

12.12
(10.88, 13.21)

SO4 (mmol 1-1)

0.04
(0.04, 0.05)

1.66

(1.32, 1.86)

5.06

(4.76, 5.32)

7.73
(7.13, 8.33)

NO3 (mmol 1-1)

0.03
(0.002, 0.08)

0.03
(0.006, 0.09)

0.05
(0.01, 0.08)

0.06
(0.02,0.15)

Total cations (meq l-i)

1.2
(0.7, 1.6)

10.0

(9.2, 10.8)

25.7
(23.8, 26.7)

37.4
(34.5, 41.5)

EC (dS m-i)

1.09 ±0.03

0.97 ±0.11

2.37 ±0.23

3.45 ± 0..39

7..54 ± 0.03

8.10 ±0.02

8.29 ± 0.02

8.05 ± 0.03

Great Basin Naturalist

[Volume 55

T.'VBLE 2. Survival and growth oi P(>))iilus frcmontii and Tainarix raiiio.sis.siiim se(
ty treatnient.s for one growing sea.son outdoors in Fort Collins, CO. Higli and low
ses below the treatment means in = 3). Treatment effeets were anaKzed 1)\ con
Survival .\.\0\'A was performed on arcsine-transformed data.

'dlings e.\posed to four different salini-
replicate means are given in parenthe-
ipleteK randomized one-way ANOVA.

Species
Variable

Trc;

:itinent

1.x

5.\

Cottonwood
Survival
(% of planted seed)

57.0
(50.0, 63.0)

49.3
(45.7, 54.0)

46.6
(41.0,51.0)

29.0
(20.7, 35.0)

Shoot height (mm)

33.9 ■

(32.8, 34.5)

36.3

(.34.5, 38.5)

39.6

(.36.5, 43.9)

.38.3
(34.7, 40.8)

Hoot length (mm)

239.2

(227.1,258.4)

280.9
(257.8, 309.3)

286.9
(253.6,311.7)

247.4
(206.3, 274.6)

Per-plant shoot
biomass (mg)

14.1
(13.7, 14.4)

14.6
(11.2, 16.6)

21.4
(18.9, 25.8)

19.8
(14.3, 25.9)

Per-plant root
biomass (mg)

26.8

(21.2, 35.5)

19.6

(16.4,21.3)

31.8
(21.6, 43.2)

31.2

(17.2, 42.9)

Saltcedar
Sunival
(% of planted seed)

42.3
(29.5, 51.6)

37.8
(33.8, 42.0)

37.3

(31.8, 40.8)

29.5

(22.8, 35.2)

Shoot height (mm)

18.1

(17.3, 18.8)

17.7
(15.5, 19.8)

18.2
(15.6, 22.2)

18.3

(18.2, 18.3)

Root length (mm)

174.4
(166.4, 184.9)

173.6
(154.8, 192.9)

179.0
(128.1,243.6)

162.0
(147.2, 169.5)

Per-plant shoot
biomass (mg)

5.5

(4.8, 6.2)

5.5

(4.1,6.4)

6.3
(4.3, 9.6)

6.2
(5.8, 6.4)

Per-plant root
biomass (mg)

7.7
(7.1. 8.9)

7.3
(5.5, 9.2)

9.9

(7.0. 14.7)

9.5

(7.6. 12.1)

11.4

2.6

2.1

2.8

1.1

1.6

0.04

0.15

0.22

0.74

.003

.13

.17

.11

.41

.26

.99

.92

.88

.56

season sui-vival (mortality). There was a signif-
icant treatment effect on germination (F =
.008), but not on mortality [P = .45), indicat-
ing that the effect on end-of-season survival
was predominantly due to lower germination
at higher salt concentrations. For T. ramosimma
there were no significant treatment effects
(Table 2).

Although P. fremontii germination in out-
door tanks was significantly decreased at high
salinity, laboratory germination was not simi-
larly affected even at seven times the salinity
of the Rio Grande, total concentration 48.4
meq \~^ (4.56 dS m"^; Table 3). There was a sig-
nificant positive effect of increasing salinity on
T. ramosissima germination (P = .03) (Table 3).

Discussion

The absence of a negative effect of salinity
on P. fremontii germination in the laboratory at
concentrations as high as 48.4 meq 1~^ (4.56

dS m~l) is consistent with results of earlier
studies. Jackson et al. (1990) found that P fre-
montii germinated in the laboratoiy at salini-
ties of 0, 27, and 106 meq 1~^ using a mixture
of NaCl and CaCU, but not at 319 meq h^ or
above. Siegel and Brock (1990) observed high-
er percent germination of P. fremontii in the
laboratoiy in NaCl solutions of 0, 25, and 50
meq h^ than at 100 meq \~^ and above. There-
fore, P. fremontii is no more sensitive to the
mix of salts present in the Rio Grande than to
NaCl and CaCl2 solutions of equal strength.
Tests at higher salinities with the same ionic
ratios were not possible with our Rio Grande
mix because of low solubilities of some of the
constituent salts. The decrease in T. ramosissi-
ma germination at low salinit}' in the laborato-
ry (Table 3) is consistent with the finding by
Jackson et al. (1990) that germination increas-
es between 0 and 106 meq 1~^.

Our results indicate that a given water salin-
ity may negatively affect germination of P.

1995]

SALiNiTi' Effects on Populus and Tamarix

Table 3. Percent germination of Populus fremontii and Tamarix ramosissitna seedlings exposed to five salinity treat-
ments in covered petri dishes. High and low replicate values are given below the treatment mean {n = 5). Treatment
effects were analyzed by completely randomized one-way ANOVA using arcsine-transformed data.

Species

Treatment

Cottonwood

Saltcedar

90.4
(80.0, 100.0)

69.6
(60.0, 88.0)

96.0
(92.0, 100.0)

(56.0, 80.0)

96.0
(92.0, 100.0)

78.4
(68.0, 92.0)

92.8
(84.0, 96.0)

84.8
(76.0, 92.0)

96.0
(92.0, 100.0)

84.0
(76.0, 92.0)

1.2

3.3

.35

.03

fremontii seeds under ambient conditions but
not under laboratory conditions. This may have
resulted from an interaction between the
effects of salinity and soil moisture content, or
from vapor-pressure deficit differences. In
outdoor planters, but not laboratoiy petri dish-
es, evaporation of water may have resulted in
lower soil moisture and higher salt concentra-
tion at the soil surface. These factors would
both tend to reduce soil water potential,
thereby increasing plant water stress. Because
the difference between evaporation and pre-
cipitation is somewhat greater at the Bosque del
Apache than in Fort Collins, the effect of salin-
ity might be stronger at the Bosque, especially
in dry years. Finally, greater vapor-pressure
deficits in the field relative to the laboratory
may have exacerbated plant water stress.

Salinity appears to be a relatively minor fac-
tor regulating numbers of P. fremontii and T.
romosissinia seedlings on freshly deposited
sandbars along the Rio Grande. The only signif-
icant effects of increasing salinity were a small
decrease in P. fremontii germination in out-
door planters and a small increase in T. ramo-
sissima germination in the laboratory. There
were no significant effects on survival after
germination or above- or belowground growth
for either species, even at water salinities sev-
eral times that of the Rio Grande. The presence
of abundant seedlings of P. fremontii and T.
ramosissima on sandbars along the Rio Grande
in most years is consistent with our results.

Although salinity may play only a minor role
in the colonization of newly deposited alluvial
bars by T. ramosissima and P. fremontii, this
factor can become more important over time.
Over many years salt becomes concentrated
on some floodplains as a result of evaporation
and salt extrusion from T. ramosissima leaves.
EC readings as high as 10.0 dS m~^ have been
reported in floodplain sediment at the Bosque

del Apache (John Taylor, Bosque del Apache
National Wildlife Refuge, personal communica-
tion), and soil salinity levels as high as 60,000
mg 1~1 occur on floodplain sites along the
lower Colorado River (Jackson et al. 1990).
Soil EC above 2.0 dS m"^ can reduce the
growth of P. fremontii pole plantings (Anderson
1989). T. ramosissima has been shown to be
less susceptible than P. fremontii to many of
the negative effects of higher salinities
(Brotherson and Winkle 1986, Jackson et al.
1990). Tamarix species avoid harmful effects of
salts through extrusion from leaves and cellu-
lar compartmentation (Berr\' 1970, Kleinkopf
and Wallace 1974, Waisel 1991).

Our results could be applied to efforts to
revegetate riparian areas from seed. Riparian
revegetation in the Southwest has largely con-
sisted of planting poles or potted shoot cut-
tings. Although these approaches have been
successhil in some areas (Anderson et al. 1990),
they can cost up to $10,000 per hectare (Ohmart
et al. 1988). Furthermore, they require the
destruction of parts of existing trees, and often
entire trees or stands. Finally, these approach-
es may require importing cuttings or poles
adapted to different site conditions. One alter-
native is regeneration of native cottonwoods
and willows using natural seedfall (Friedman
1993, John Taylor personal communication).
This approach generally involves clearing and
irrigating an area so that seeds from nearby
trees can colonize it. Our results suggest that
water as saline as 37.4 meq 1~^ (EC 3.45 dS
m~l) can be used to grow P. fremontii from
seed on sand (Tables 1, 2). However, care must
be taken to prevent long-term salt accumula-
tion through evaporation (e.g., through period-
ic flooding to flush salts) and to avoid sites
with preexisting high salinity. Use of water
with low salinity can help prevent negative
effects on P. fremontii and may decrease the

Great Basin Naturalist

[Volume 55

germination rate of T. ramosis.mna (Table 3).
However, in a restoration effort along the
Cache la Poudre Rixer, T. ratn(>.sissi)na became
established in large numbers along with P. del-
toides in spite of use of water of low salinity
(Douglas Gladwin, National Biological Survey,
personal communication). Therefore, low salin-
it\' will not prevent establishment of T. ramo-
sissima from seed when moisture, a bare sedi-
ment, and a seed source are present.

Acknowledgments

G. T Auble, D. E. Busch, and an anonymous
reviewer provided constructive comments on
the manuscript. We thank E. R. Auble, G. T
Auble, J. Back, E. D. Eggleston, M. Jordan,
and M. L. Scott for invaluable assistance with
the experiments. D. Smeltzer, B. Upton, and
the Colorado Division of Wildlife generously
provided access to the Bellvue-Watson Fish
Rearing Unit where the outdoor experiment
was conducted. T Kem and P Soltanpour pro-
vided useful advice regarding the salinity
treatments. Concentrations of ions in solutions
were measured by the Soil, Plant and Water
Testing Laboratoiy at Colorado State University,
Fort Collins, CO.'

Literature Cited

Anderson, B. W. 1989. Research as an integral part of
revegetation projects. Pages 413—119 in D. L. Abell,
technical coordinator. Proceedings of the California
Riparian Systems Conference: protection, manage-
ment, and restoration for the 1990s. USDA Forest
Service General Technical Report PSW-110.
Berkeley, CA.

ANDER.SON, B. VV., A. HiGGINS, AND R. D. Ohmart. 1977.
Avian use of saltcedar communities in the lower Colo-
rado River Valley. Pages 128-136 in R. R. Johnson
and D. A. Jones, technical coordinators. Importance,
Presei^vation, and Management of Riparian Habitat:
a symposium. USDA Forest Service General
Technical Report RM-43. Fort Collins, CO.

Anderson, B. VV, E. R. Miller, and J. E Washlngton.
1990. Revegetation on the Kern River preserve
1986-1989. Research report prepared for The
Nature Conservancy and California Department of
Fish and Game. Revegetation and Wildlife Vlanage-
ment Center, Inc., Blythe, CA. 19 pp.

Berry, W. L. 1970. Characteristics of salts secreted by
Tamarix aphylhi. American Journal of Botany 57:
1226-1230.

Blackburn, W. H., R. W. Knight, and J. L. Schuster.
1982. Saltcedar influence on sedimentation in the
Brazos River. Journal of Soil and Water Conserva-
tion, 37: 298-301.

Bowser, C. W 1958. Introduction and spread of the un-
desirable tamarisks in the Pacific southwestern sec-

tion oi the United States and comments concerning
the plants influence upon the indigenous vegetation.
Pages 12—16 in Symposium on Phreatophytes. Ameri-
can Geophysical Union, Sacramento, CA.

Brotherson, J. D., AND V. WiNKEL. 1986. Habitat rela-
tionshijis o{ saltcedar {Tamarix ramosissiina) in cen-
tral Utah. Great Basin Naturalist 46: 53.5-541.

BUSGH, D. E., AND S. D. Smith. 1993. Effects of fire on
water and salinity relations of riparian woody taxa.
Oecologia 94: 186-194.

. In press. Decline, persistence, and competition of

woody ta.xa in riparian ecosystems of the southwest-
em U.S. Submitted to Ecological Monographs.

Campbell, C. J., and VV. A. Dick-Peddie. 1964. Compari-
son of phreatophyte communities on the Rio Grande
in New Mexico. Ecology 45: 492-502.

Cohan, D. R., B. W Anderson, and R. D. Ohmart. 1978.
Avian population responses to salt cedar along the
lower Colorado River. Pages 371—382 in R. R. Johnson
and J. F McCormick, technical coordinators.
Strategies for Protection and Management of Flood-
plain Wetlands and Other Riparian Ecosystems: pro-
ceedings of the symposium. USDA Forest Service
General Technical Report WO-12. Washington, DC.

EvERITT, B. L. 1980. Ecology of saltcedar — a plea for
research. Environmental Geolog}' 3: 77-84.

Friedman, J. M. 1993. Vegetation establishment and
channel narrowing along a Great-plains stream fol-
lowing a catastrophic flood. Unpublished doctoral
dissertation. University of Colorado, Boulder.

Graf, W L. 1978. Fluvial adjustments to the spread of
tamarisk in the Colorado Plateau region. Geological
Society of America Bulletin 89: 1491-1501.

Greenway, H., and R. Munns. 1980. Mechanisms of salt
tolerance in nonhalophytes. Annual Review of Plant
Physiolog>'31: 149-190.

Harris, D. R. 1966. Recent plant invasions in the arid and
semi-arid southwest of the United States. Annals of
the Association of American Geographers 56: 408-422.

HORTON, J. S. 1977. The development and peipetuation of
the permanent tamarisk type in the phreatophyte
zone of the Southwest. Pages 124-127 in R. R.
Johnson and D. A. Jones, technical coordinators.
Importance, Presei'vation, and Management of Ripar-
ian Habitat: a symposium. USDA Forest Service
General Technical Report RM-43. Fort Collins, CO.

HoRTON, J. S., and C. J. Campbell. 1974. Management of
phreatophyte and riparian vegetation for maximum
multiple use values. USDA Forest Service Research
Paper RM-1 17. 23 pp.

Jackson, J., J. T. Ball, and M. R. Rose. 1990. Assessment
of the salinity tolerance of eight Sonoran desert
riparian trees and shrubs. Find report, U.S. Bmeau of
Reclamation Contract No. 9-CP-30-0717(). Biological
Sciences Center, Desert Research Institute, Univer-
sity of Nevada System, Reno. 102 pp.

Kleinkopf, G. E., and A. Wallace. 1974. Physiological
basis for salt tolerance in Tamarix ramosissima. Plant
Science Letters 3: 157-163.

Merkel, D. L., and H. H. Hopkins. 1957. Life histoiy of
salt cedar {Tamarix gallica L.). Transactions of the
Kansas Academy of Science 60: 360—369.

Ohmart, R. D., B. VV. Anderson, and W C. Hunter.
1988. The ecolog>' of the lower Colorado River from
Davis Dam to the Mexico-United States Interna-
tional Boimdan': a commimitv' profile. U.S. Fish and
Wildlife Service Biological Report 85(7.19). 296 pp.

1995]

Salinity Effects on Populus and Tamarix

Ohmart, R. D., W. O. Deason, and C. Burke. 1977. A
riparian case history: the Colorado River. Pages
35-47 in R. R. Johnson and D. A. Jones, technical
coordinators, Importance, Preservation, and Man-
agement of Riparian Habitat: a symposium. USDA
Forest Service General Technical Report RM-43.
Fort Collins, CO.

Robinson, T. W. 1965. Introduction, spread and areal
extent of saltcedar {Tamarix) in the western states.
United States Geological Survey Professional Paper
491-A. 13 pp.

SAS Institute, Inc. 1990. SAS/STAT user's guide, ver-
sion 6, 4th edition. SAS Institute, Inc., Cary, NC.

Segelquist, C. A., M. L. Scott, and G. T. Auble. 1993.
Establishment of Populus deltoides under simulated
alluvial groundwater declines. American Midland
Naturalist 130: 274-285.

SlEGEL, R. S., AND J. H. Brock. 1990. Gemiination require-
ments of key Southwestern woody riparian species.
Desert Plants 10: 3-8, 34.

Snedecor, G. W., and W. G. Cochran. 1980. Statistical
methods. Iowa State University Press, Ames. 507 pp.

Stromberg, J. C, D. T. Patten, and B. D. Richter. 1991.
Flood flows and dynamics of Sonoran riparian forests.
Rivers. 2: 221-235.

Tomanek, G. W., and R. L. Ziegler. 1962. Ecological
studies of salt cedar. Unpublished report. Division of

Biological Sciences, Fort Hays Kansas State College,
Hays. 128 pp.

United States Environmental Protection Agency.
1983. Methods for chemical analysis of water and
wastes. Publication identification: EPA-600 4-79-020.
Environmental Monitoring and Support Laboratory,
Office of Research and Development, Cincinnati,
OH.

United States Geological Survey. 1991. Water re-
sources data. New Mexico, water-year 1990. U.S. Geo-
logical Survey water-data report NM-90.

. 1992. Water resources data. New Mexico, water-
year 1991. U.S. Geological Survey water-data report
NM-9I.

Waisel, Y. 1991. The glands oiTaiiuirix aphylla: a system for
salt secretion or for carbon concentration? Physiologia
Plantarum 83: 506-510.

Warren, D. K., and R. M. Turner. 1975. Saltcedar {Tama-
rix chinensis) seed production, seedling establishment,
and response to inundation. Journal of the Arizona
Academy of Science 10: 135-144.

Zasada, J. C, AND R. Dens.more. 1977. Changes in Sali-
caceae seed viability during storage. Seed Science
and Technology 5: 509-517.

Received 14 March 1994
Accepted 12 August 1994

NAMES AND TYPES OF HEDYSARUM L. (FABACEAE) IN NORTH AMERICA

Stanley L. Welsh ^

Abstract. — The names and t\pes of Hedysaruiii L., sensu stricto, for North America are included, alon^ with biblio-
graphic citations, type information and place of deposit of types, and all synonyms. Lectotypes are designated for
Hechjsarum auriculatum Eastvv., H. carnulosum Greene, H. marginatum Greene, H. pabulare A. Nels., and H. truncatum
Eastw'.

Key wards: Hed\ sarum, types, nomcnclatttrc.

The following list of names and types in
Hedysannn L. was prepared preliminary to
submittal of a summary revision to the Flora
North America Project. The genus Hedysannn
L. as here inteipreted for American taxa extends
from the Bering Strait to Newfoundland and
Vermont, and from the Polar Sea and the Cana-
dian Arctic Archipelago south through the
mountains and plains of western North Anerica
to Oklahoma, New Mexico, Arizona, and
Nevada. Excluded from this treatment are
those taxa originally included in Hedysannn,
which are now interpreted as belonging to
other genera, i.e., to Desmodiinn. The genus in
the restricted sense consists of two complexes,
i.e., those with leaflets thickened and veins
obscured (the boreale complex) and those with
relatively thin leaflets in which the veins are
rather readily apparent (the alpinum complex).
The earliest taxon within Hedysanim alpinum
complex is that by Michaux (1803), who estab-
lished the trinomial Hedysannn alpinum ameri-
canum Michx. The boreale complex was initi-
ated by Nuttall (1818) with the publication of
H. boreale.

Taxa in the two complexes demonstrate re-
markable morphological and geographical par-
allelism. Each consists of additional taxa sepa-
rable generally into two geographical sub-
groups juxtaposed at or near the 50th parallel
of longitude (somewhat north of the Canada-
U.S. boundary). North of that parallel lies most
of H. boreale ssp. mackenzii (Richards.) Welsh,
and most of H. alpinum sensu stricto. To the
south occurs H. occidentale Greene, most of H.
sulphurescens Rydb., limited extensions of H.
alpinum L., and most of H. boreale ssp. bore-

ale. Glacial events during the Pleistocene have
been suggested as having separated the sub-
sets, allowing them to achieve the degree of
morphological and genetic integrity of the
present populations. The present juxtaposition
is suggested to have resulted by expansion of
the respective entities into areas previously
occupied by glaciers.

The rather large number of names involved
in the genus is indicative of variation inherent
in the various taxa. Flower size, plant size,
leaflet size, and pubescence are features vari-
able in both complexes. Apparent correlation
of two or more of these features has served as
justification for several names. Indeed, when
one observes dwarf, large-flowered plants in
either complex, there appears to be a compel-
ling need for their recognition. However,
much, if not all, of the variation is haphazard,
or the attempt at segregation devolves to use
of a single characteristic, such as presence or
absence of pubescence, which fails also. There
are few truly diagnostic characteristics once
the two complexes are separated. The taxono-
mist ultimately must rely on a series of varying
features to identify a particular specimen.
Fortunately, the taxa are, with some notable
exceptions, disjunct from each other. If the
disjunction is not apparent from examination
of a distribution map, it is often apparent in
the field where the plants grow in different
habitats. For example, the range of yellow-
flowered H. sulphurescens apparently overlaps
that of pink-purple-flowered H. occidentale in
large part; yet, they seldom occur together,
and only an exceptional intermediate is
known.

'Department ol Botany ami Life Science Mnscuni, Brii^hani Younf; University-. Provo. UT 84602.

1995]

North American Hedysarum

There are, in spite of gross similarities of
the taxa within the respective complexes, few
recorded intermediates.

Adding to the difficulties of interpretation
of the North American materials is the inter-
rupted circumboreal distribution of H.
alpinum, a species with several close allies in
Siberia. The initial interpretation by Michaux
of North American H. alpinum as being taxo-
nomically different {'H. alpinum: americanwn)
from that of the Old World has paraded appar-
ition-like through most subsequent treatments
of the genus. Unpublished work by Northstrom
(1974) refuted the claim to difference between
Siberian and American phases of the species,
at least as far as broad categories were con-
cerned. The claim that North American mate-
rials constitute a separate entity is likewise re-
futed by comparison of specimens from Siberia
and North America in the present study. Other
workers have asserted that large-flowered,
low-growing plants of the species are identical
with substantial Asian taxa [i.e., H. hedysari-
oides (L.) Schinz & Thellung {Astragalus hedy-
sarioides L.)]. Such claims were investigated
by Northstrom (1974), who determined that
tliere is little basis for such assertions. Evidence
to support the conspecific nature of the sup-
posed entities is apparent when localities of
such supposed taxa in western Alaska are
examined and plants with larger flowers are
found to occur within populations having
small flowers, and that flower size within the
species in a broad sense forms a continuum.
And, occasional tall specimens within the
alpinum complex also bear large flowers.

Another factor leading to the creation of a
large number of synonyms was the early mis-
inteipretation of specimens of H. alpinum under
the name H. boreale. This switching of names,
while not uniquely a problem in this genus,
became of great importance to those workers
who encountered the genus piecemeal and
treated the variants as though they had not
already been named. It was not helpful, per-
haps, that the most ardent authors of western
American plant names should be involved with
the genus (i.e., Edward L. Greene, Per Axel
Rydberg, and Aven Nelson). Greene, as the
record indicates, was prone to name the same
species several times in this and other genera,
not recognizing, or possibly not caring, that he
was renaming the same taxon.

Still another trend resulting in the forma-
tion of inconsequential names was the well-
intentioned effort to provide epithets for spec-
imens differing in insubstantial ways, i.e., the
naming of white-flowered or teratological speci-
mens as formae.

The following list is thought to be exhaus-
tive for Hedysarum names in North America.
Pertinent types have been received on loan
through the kindness of curators of herbaria
cited with the specimens. Abbreviations for
herbaria are those standard ones cited in
Index Herbariorum. Type information is pre-
sented below in dual fomiat for some taxa, with
type information (type locality) as recorded
with the protologue cited first and label data
of the type specimen (type) cited second where
there is a substantial difference in the two
accounts.

Hedysarum albiflorum (Macoun) Fedtsch., Acta Hort.
Petrop. 19: 252. 1902.

Basionym: H. boreale var. ulhiflorum Macoun

= H. sulphurescens R\db.

Hedysarum alpinum L., Sp. Pi. 750. 1753.

Type locality: "Habitat in Siberia" (Linnaeus I.e.).
Tvpe: Possible lectotvpe 921.54 LINN (microfiche
BRY!).

Hedysarum alpinum var. americanum Michx., Fl. Bor.
Anier. 2: 74. 1803.

= H. alpinum L.

Synonyms: H. alpinum ssp. americanum (Mich.x.)
Fedtsch., Acta Hort. Petrop. 19: 2.55. 1902, in part;
H. americanum (Mich.x.) Britt., Mem. Torrey Bot.
Club 5: 201. 1894.

Type locality': "In borealibus Canadae, et in cataractis
montium alleghanis."

Type: "Hedys. J ii p. 74-75. Herb. M.x" (isotype NY!).

There is a mounted half herbarium sheet at NY
Torrey! bearing a large portion of a stem with a leaf
and mature, strigose fiaiit of H. boreale var. boreale.
This specimen is apparentlx' superfluous (probably
having been added later when additional material be-
came available to Dr Torrey from western American
collections), but more pertinent to the present case
the sheet also has an attached fragment envelope
on which is written the type information noted
above. The envelope contains a portion of an inflo-
rescence, a flower, and several immature loment
segments. The segments are glabrous, have a defi-
nite winged margin, and are identifiable as H.
alpinum L. It is probable that the specimen from
which the fragments were removed is with the
Michaux herbarium at R

Continued recognition of the American materi-
als of H. alpinum at any infraspecific rank is fraught
with difficulties; there are no diagnostic features
known that will allow segregation of the American
specimens from the Asiatic ones.

Great Basin Naturalist

[Volume 55

Hedysarum alpinum van americanuni f. alhiflorum
(Standi.) Fern., Rhodora35: 275. 1933.

BasioiiNin; //. americanum f. alhiflorum Staiidl.

= H. alpiiniin L.

The publication by Fernald (1933) recognizes
white-flowered plants from Newfoundland.

Hedysarum alpinum van grandiflorum Rollin.s, Rhodora
42: 233. 1940.

Type: "Newfoundland, Pistolet Bay, Mo.ssy and turfy
trap cliffs and talu.s, An.se aux Sauvages, M. L.
Fernald, K. M. Wiegand and Bayard Long 28625,
August 11, 1925"; holotype GH!
Paratype: "New Fonndland, Region of Port a Port Bay,
No. 10849. In humus or turf on the limestone table-
land, altitude 200-300 m.. Table .Mountain, M. L.
Fernald and H. St. John, July 16 6c 17. 1914" (CAN!;
BM!).
This name is based on large-flowered (about 16
mm long), low-growing specimens from Newfound-
land. There are other similar plants scattered
through much of the distribution of H. alpinum in
North America, but they are more consistently rep-
resented in frigid or other inhospitable arctic or
subarctic sites. Even in the type series cited with
the protologue there is considerable variation. The
paratype cited above differs significantly from the
holotype specimen; it is much taller and has flowers
of a size intermediate with those of specimens more
usual for var alpinum in a more strict sense. Indeed,
the low-growing, larger-flowered phase appears to
be a phenotypically recurring recombinant form
within a complex exhibiting much variation in
flower size and other features. However, size of
flower is not always conelated with plant height or
flower number. All possible combinations of flower
size, flower number, and plant height are represent-
ed in the species as a whole. It is possible to write a
key that will separate these plants, but it seems that
such a key will not then be segregating natural ta.\a.

Hedysarum alpinum ssp. philosocia (A. Nels.) Love &

Love,Ta.\on31:.347. 1982.

Basionym: H. philosocia A. Nels.
= H. alpinum L.

Hedysarum alpinum var. philosocia (A. Nels.) Rollins,
Rhodora 42: 224. 1940.

Basionym: H. philosocia A. Nels.

= H. alpinum L.

Hedysarum americanum (Michx.) Britt., Mem. Torrey
Bot. Clul) 5: 202. 1894.

Basionym: H. alpinum var americanum Michx.

= H. alpinum L.

Hedysarum americanum f. alhiflorum Standi., Incld Mus.
Pub. Bot. 8: 15. 1930.
= H. alpinum L.

Synonym: H. alhiflorum (Macoun) Fedtsch.
Type: "Alaska: Davidson Glacier July 4, 1929, William
S. Cooper & Frances E. Andrews 95 (Herb. Field
Mus. No. 598,264, type)"; holotype F!

White-flowered specimens occur sporadically
through populations of taxa with generally pink-
purple flowers. Their recognition at any taxonomic
rank is probably moot, and the publication of the
tiixon by Standley (1930) is therefore inconsequential.

Hedysarum americanum van mackenzii (Richards.) Britt.,

Mem. Torrey Bot. Club 5: 202. 1894.
Basionym: H. mackenzii Richards.
— H. horeale ssp. mackenzii (Richards.) Welsh

Hedysarum auriculatum Eastw., Bot. Gaz. 33: 205. 1902.

= H. alpinum L.

Type: Alaska, Cape Nome, Blaisdell s.n. sununer 1900
(lectotype selected here; GH!, isolectotype US!).

Specimens on which H. auriculatum is based
were distributed from the California Academy of
Sciences herbarium with collection information
recorded on labels of that institution. The only
known specimens in contemporary collections are
those at GH and US. The two specimens consist of
almost identical branches of H. alpinum, with both
flowers and fruit, although that at GH is designated
on the label as a duplicate of the type, which was
presumably at GAS prior to the San Francisco
earthquake early in this centuiy Fire resulting from
that devastating tragedy destroyed much of the
early Academy herbarium.

Hedysarum bakeri Greene ex Rydb., Bull. Agric. Exper
Sta. Colorado, 100: 215. 1906. pro syn.

= H. horeale Nutt. var horeale

Intended type: "Flora of Colorado. Plants the Gunnison
Watershed, Cimarron, June 28. Stems in large clus-
ters 8 in. to 1 1/2 ft. on dr>' open slopes. Collected in
1901 by C. F Baker, No. 274" (NDG!).

Evidently the name was never published by E. L.
Greene but was cited as a synonym of H. pabulare
A. Nelson by Rydberg in his Flora of Colorado. The
intended type has three mounted stems showing
flowers and maturing fruit; they are strigose both
on herbage and on the loments. The plants differ in
no material way from a great many specimens from
Colorado. Perhaps Greene also realized as much.

Hedysarum horeale Nutt., Gen. N. Amer Pi. 2: 110. 1818.

Type locality: North Dakota, "around Fort Mandan, on
the banks of the Missouri," Nuttall (I.e.).

Type: "Hedysanmi horeale — Sources of the xMissouri,"
Nuttall (probably late June) 1811; holotype BM!

The name H. horeale was early transfeired to the
concept of H. alpinum, and part of the synonymy
reflects attempts by various authors to resolve the
apparent lack of a name for this wide-ranging and
highly variable species. Nuttall (Torrey and Gray
1838) named the species a second time, as H.
canescens, based on specimens from along the
Snake River in present Idaho taken in 1834. He
was in the vicinity of Fort Hall, Idaho, from 14 July
to 6 August 1834 (McKelvey 1955: 602). Whether he
noted the similarity between earlier- and later-
named materials is not known. It seems likely that

1995]

North American Hedysarum

he did not have authentic material of the earher-
named taxon at hand for comparison with his col-
lections on the 1834 Wyeth expedition. The t\'pe at
BM, a solitary' flowering stem, is mounted with sev-
eral flowering stems of H. alpinum of unknown col-
lector.

Hedysarum boreale var. alhiflorum Macoun, Cat. Canad.
PI. 1; 510. 1S84. noni. nud.
= H. sulphurescens R\db.
Syn: H. alhiflorum (Macoun) Fedtsch.
T>pe localib,': "This form is peculiar to the foothills and

drier mountain slopes, and is abundant from the

Kananaskis through the Rocky Mountains to the

Columbia valley at Donald, Lat. 51° (Macoun).

Eastern summit of the North Kootanie Pass, Rocky

Mountains" (I.e.).
Type; "Geological and Natural History of Canada. No.

1111.5390. Hedysarum boreale var alhiflorum.

Dry soil. East summit of North Kootanie Pass, R.

Mts. Dawson. July 29th 1883" and "Geological and

Natiual Histor\' Survey of Canada. No. 533, 5389.

Hedysarum boreale Nutt. var. alba. Macoun.

Mountain slopes. Kananaskis. Rocky Mts. Macoun.

June 24th 1885"; syn txT^es CAN!
There is no description aside fiom the designation
"alhiflorum" proposed as an epithet. The only other
information provided hy Macoun aside from that
related with the locality data is the statement: "This
fine plant is closely related to H. boreale, but is cer-
tainly distinct." The name is regarded as a nomen
nudum. The collection by Dawson is, nevertheless,
an excellent flowering example of H. sulphurescens,
and the Macoun sheet consists of two plants with
both flowers and immature to mature fruits, both
also H. sulphurescens. The indication by Macoun of
relationship of var. albiforum to H. boreale reflects
the general misapplication by many American
botanists of H. boreale to the alpinum complex in
North America, of which H. sulphurescens is a por-
tion. Macoun used the number 533 for several col-
lections of Hedysarum taken from 1883 to 1885.

Hedysarum boreale var. cinerascens (Rydb.) Rollins,

Rhodora 42; 234. 1940.

Basionym; H. cinerascens Rydb. et H. canescens Nutt.

in seq.
= H. boreale Nutt. var boreale

Hedysarum boreale var. cinerascens f. album Boivin,

Naturaliste Canad. 87: 34. 1960.
= H. boreale Nutt. van boreale

Type: "Canada, Saskatchewan, Maple Creek District,
Eastend, hillside along river valley, 19 July 1950,
R. C. Russell S 5075" (holotype at DAO!).

Hedysarum boreale var. flavescens (Coult. & Fisher)
Fedtsch., Bull. Herb. Boiss. 7: 256. 1899.

Basionym; H. flavescens Coult. & Fisher

= H. sulphurescens Rydb.

Hedysarum boreale var. gremiale (Rollins) Northstrom &
Welsh, Great Basin Nat. 30; 125. 1970.
Basionym; H. gremiale Rollins

Hedysarum boreale var. leucanthum (Greene) M. E. Jones,

Proc. Calif Acad. Sci. 5; 677. 1895.

Basionym; H. mackenzii var. leucanthum Greene
= H. boreale ssp. mackenzii (Richards.) Welsh

Hedysarum boreale ssp. mackenzii (Richards.) Welsh,

Great Basin Nat. 28; 152. 1968.

Basionym; H. mackenzii Richards.

Hedysarum boreale var. mackenzii (Richards.) C. L.
Hitchc, Vase. Pi. Pacific N. W 3; 275. 1961.
Basionym; H. mackenzii Richards.

Hedysarum boreale var. mackenzii f. niveum (Boivin)
Boivin, Naturaliste Canad. 93; 433. 1966.

Basionym; H. mackenzii var. mackenzii f. niveum
Boivin

= H. boreale ssp. mackenzii (Richards.) Welsh

Hedysarum boreale var. obovatum Rollins, Rhodora 42:
235. 1940.

= H. boreale Nutt. van boreale

Type: Nevada, Elko County, Thorpe Creek, E of
Lamoile, 25 July 1928, H. H. Price 168 (holotype
photo RM!).

Hedysarum boreale f. proliferum (Dore) Boivin, Naturaliste

Canad. 94; 630. 1967.

Basionym; H. mackenzii f proliferum Dore
= H. boreale ssp. mackenzii (Richards.) Welsh

Hedysarum boreale var. utahense (Rydb.) Rollins, Rhodora
42; 2.35. 1940.

Basionym; H. utahense Rydb.

— Hedysarum boreale Nutt. var boreale

Hedysarum canescens Nutt., in Torr. & Gray, Fl. N. Amer
1: 357. 18.38. Not H. canescens L.

Basionym; H. cinerascens Rydb.; H. boreale var. cin-
erascens (Rydb.) Rollins

= H. boreale Nutt. var boreale

Type locality; Idaho, "Plains of the Rocky Mountains,
particularly near Lewis's River," Nuttall (I.e.)

Tvpe; "HedysaiTim * canescens. H. mackenzii? Hook.
R. Mts. Lewis [Snake] R." Nuttall s.n. (probably in
July) 1834 (holotype PH!; isotypes GH!, BM!, 2
sheets?).

The specimen at PH (which is mounted on a
sheet with two other superfluous collections) bears
the date "July 12," with the incorrect year date
1833 obviously added later. Nuttall was with the
Wyeth Expedition in 1834, and on 12 July was a
short distance east of where Fort Hall would be
constructed subsequently. Despite the existence of
the earlier-named H. boreale, with which H.
canescens is synonymous, this name or its substi-
tutes would be featured prominently in 19th-centu-
ry accounts of the genus in the American West.
There are two of Nuttall's specimens on the sheet
at GH, each provided with a label — both with flow-
ers and both representing the same taxon. The label
information consists of the following: "Hedysarum *
canescens. H. mackenzii? R. Mts.," and Hedysarum *
canescens R. Mts." Since no additional locality

Great Basin Naturalist

[Volume 55

iiiforniation or date accompanies the labels, the sta-
tus as exact dupheates is unknown. It seems hkely
that both were included within the concept of H.
canescens by Nuttall, and both can be regarded as
isotypes. There is a second possible isotype of H.
canescens at BM, "Hedysarum mackenzii? Fort
Hall. Prairie, common. Aug." It lacks the * usual for
Nuttall's labels, and his name is not in evidence,
but the handwriting appears to be his.

Hedijsarum carnulosum Greene, Pittonia 3: 212. 1897.
= //. horcdic \iitt. \ ar. borcale
Type localit\'; "Common in claye\' soil about the mouth

of the Canon of the Arkansas, in southern Colorado'

(Greene I.e.).
Type: Colorado, Fremont Co., "Plants of Colorado,

Canon Cit>', 8 Sept. 1896, Edw. L. Greene" (lecto-

tvpe here designated: NDG!, 2 isolectotypes also

NDG!).
While no specimens were cited with the original
description, the three specimens so named in
Greene's handwriting at NDG are most certainly
type material. All bear the same date and locality in-
formation. The specimen bearing the "Greeneanum
Herbarium ' number 35686 is here chosen as lecto-
type; the others, 35687 and 35688, are considered
isolectotypes. The lectotype has both flowers and
fruit; the other two are in fruit and flower (with
immature fruit), respectively. All have strigose
herbage and foments. Usual flowering time for the
species is April to late July. Is it possible that the
species flowered again following late summer rains
at Canon City in September 1896?

Hedysarum cinerascens Rydb., Mem. N. Y. Bot. Card. 1:
257. 1900. nom. no\. pro H. canescens Nutt.

— H. horeale Nutt. var boreale

Basionym: H. canescens Nutt.

Syn: H. boreale var cinerascens (Rydb.) Rollins

This material was retained liy Northstrom (1974)
at varietal rank. The taxon stands on the sole char-
acter of pubescence, and a plotting of the distribu-
tion of hairy versus glabrous plants demonstrates
much overlap. The specimens can be separated, but
do they represent taxa?

Hedysarum flavescens Coult. & Fisher, Bot. Gaz. 18: 300.
1893, non Regel & Schmalh.

Basionym of: H. sulphurescens Rydb.

Type: Montana, near Helena, May 1892, F D. Kelsey
s.n. (holotype F!).

It is unfortunate that the epithet /7r/r('.sf(?n.s was
occupied; it fits well the description of flower color
in this ta.xon. xMany of the Kelsey collections are in
the U.S. National Museum (Elisens 1985), but the
type of H. flavescens is at F, where Coulter's
herbarium is deposited.

Hedysarum gremiale Rollins, Rhodora 42: 230. 1940.

= H. boreale var gremiale (Rollins) Northstrom &

Welsh
Type: Utah, Uintah County, "ca 14 mi \V. of Vernal, 16

June 1937," R. C. Rollins 1733 (holot\pe GH!, iso-
types RMI, US!, CAS!, UTC!, MONTU!, PH!, F!).
This taxon stands on the feature of lateral spines
on the foment segments; it is otherwise indistin-
guishable from plants of var horeale by which it is
suiTOunded (Northstrom and Welsh 1970).

Hedysarum lancifolium Rydb., Mem. New York Bot.
Card. 1: 256. 1900.

= H. occidentale Greene

Type: "Mountain woods near head waters of Jocko

River, Montana, — flowers pale purple, W. M.

Canby 93, July 15, 1883" (holotype NY!).

The type specimen consists of a folded plant

some 75 cm tall bearing leaves and flowers, and a

stem fragment bearing immature fruit. Mature

flowers are about 16 mm long, on the short side of

the variation in H. occidentale, but the fruit, even

though immature, is of the size and form of that

species.

Hedysarum leucanthum (Greene) Greene, Pittonia 3: 213.

1897.

Basion\ni; H. mackenzii \ar leucanthum Greene
= H. boreale ssp. )nackenzii (Richards.) Welsh

Hedysarum mackenzii Richards., in Franklin, 1st Journey
Bot. Append. 745. 1823.

— H. boreale ssp. mackenzii (Richardson) Welsh
Type: Canada, Barren grounds. Point Lake to the
Arctic Sea, Richardson s. n. [holot\pe BM (photo
CAN!), isotype NY Columbia! "Hedysarum macken-
zii. Franklin's Journey. Dr Hooker isotype PH!;
"Astr (crossed out) Hedys. Mackenzii Richard. N.
W Amen Fn Exp. Dr Ho." and "Herb. A. Gray. H.
Mackenzii. Torn & Gray, Fl. N. Amer Richardson
(B.D.G.)," isotypes GH!]'.
Evidently Dr William Jackson Hooker sent rep-
resentative material obtained b\' John Richardson,
botanist with the Franklin expedition, to the
Philadelphia Academy and to John Torrey and Asa
Gray. Collections from the Franklin expedition
demonstrate the variability represented in a rather
large set of specimens, each matched by modern
collections of the ta,xon. The second sheet cited at
GH is doubly moimted with a mere fragment pre-
sumed to have come from the Franklin expedition
in the lower portion and a second fragmentar)' col-
lection by Burke (apparently a phase of H. boreale)
from the Rocky Mountains. The latter material is
not a portion of the type of H. nuickenzii.

Hedysarum mackenzii var. fraseri Boivin, Canad. Field-
Nat. 65: 20. 1951.

= H. boreale Nutt. var boreale

Type locality: Canada: "Saskatchewan: W. R Eraser,

Langham, river valle\', June 12 and 26, 1938 " (I.e.)

Type: "Hedysarum Mackenzii Richards. River valley,

Langham, Sask., W.RF[raser]., June 12 & 26, 1938"

(holot\pe DAO!).

Boivin (I.e.) cites the revisionary treatment of

Rollins (1940) as indicating that H. mackenzii var

u}ackenzii has flowers 18-21 mm long, but with

1995]

North American Hedysarum

some 16-20 mm. Prairie plants fiom Saskatchewan
and Alberta, however, have flowers 13-15 mm long.
These latter plants are the basis of his \Ar. fraseri.
The type of var. fraseri represents H. horeale ssp.
horeale var. horeale, which is present along the
plains and foot slope of the Rocky Mountains, and
has flowers of the size indicated for the variety.

Hedysarum mackenzii var. leticanthum Greene, Pittonia

2: 294. 1892.

Basionym of; H. leucanthum (Greene) Greene

— H. horeale ssp. maekenzii (Richards.) Welsh

Type locality: "On the Porcupine River, northern
Alaska, Mr. J. J. Turner" (Greene I.e.).

Type: "Plants of Alaska, collected on the Porcupine
River, 1891, Mr J. Turner" (holotv'pe NDG!).

Greene (1892) notes that this is "far more than
an albino state of H. Mackenzii; perhaps identical
with some asiatic species; but the plants were just
coming into flower when gathered, in that there is
no trace of the loment." He later (1897) elevated it
to species rank. However, except for white flowers,
which occur with some frequency in the species,
the plant differs in no respect from numerous other
plants fi'om the arctic range of the ta.\on generally.

Hedysarum mackenzii var. mackenzii f. niveum Boivin,

Canad. Field-Nat. 65: 20. 1951.

Basionym of: H. horeale var. mackenzii f. niveum
(Boivin) Boivin

= H. horeale ssp. mackenzii (Richards.) Welsh

Type: "Yukon Territory: J. W. Abbott 17a, Pine Creek,
sandy land, June 7, 1946" (holotype DAO!).

The collection consists of five flowering stems of
H. horeale ssp. mackenzii, all with white flowers.
The condition of white flowers is occasional
throughout the subspecies and hardly worthy of
taxonomic consideration.

Hedysarum mackenzii var. pahulare (A. Nels.) Kearney &
Peebles, J. Wash. Acad. Sci. 29: 485. 1939.

Basionym: H. pahulare A. Nels.

= Hedysarum horeale Nutt. van horeale

Hedysarum horeale var. leucanthum (Greene) M. E. Jones,

Proc. Calif. Acad. II. 5: 677.

= H. horeale ssp. mackenzii (Richards.) Welsh
Syn: H. leucanthum (Greene) Greene

Hedysarum mackenzii f. proliferum Dore, Canad. Field-
Nat. 73: 151. 1959.

Basionym of: H. horeale f. proliferum (Dore) Boivin

— H. horeale ssp. mackenzii (Richards.) Welsh

Type: "Plants of Alaska. A single clump in shallow soil
over broken rock (growing beside common petalif-
erous plant, cf 4983). North Side of Tanana River,
Mile 277, Richardson Highway, 64°10'N, 145°52'W,
W J. Cody & T J. M. Webster 4984, June 3, 1951"
(holotvpe DAO!).
This name is based (Dore 1959) on a teratologi-
cal specimen of H. horeale ssp. mackenzii, a recur-
ring variant induced by a pathogen, likely a smut-
fungus. Teratology occurs in several if not all boreal

legumes native to Alaska. In certain of those in-
stances the inflorescence typically elongates, flow-
ers become erect on attenuated pedicels, petals are
deformed, and the ovary is typically exserted from
the flower. In some specimens at least the ovary is
filled with black spores. The type specimen of f.
proliferum exhibits another variant than that typi-
cally encountered. The inflorescence is shortened
and modified flower buds are in tight clusters.
Whether elongate or compact, specimens on which
such aberrations are based are not taxa, and the
need to name them is therefore moot.

Hedysarum macquenzii f. canescens (Nutt.) Fedtsch.,
Acta Hort. Petrop. 19: 272. 1902.

Basionym: H. canescens Nutt.

= H. horeale Nutt. var. horeale

Hedysarum marginatum Greene, Pittonia, 4: 138. 1900.
= H. occidentale Greene

Type locality: "Mountains above Cimarron, southern
Colorado, collected by the writer, 30 Aug. 1896;
also near Pagosa Springs, Colo., 26 July, 1899, C. E
Baker" (Greene I.e.).
Type: Colorado, "Plants of Colorado, Near Cimarron,
30 Aug. 1896, Edw. L. Greene" (lectotv^je NDG!,
here chosen); "Plants of Southern Colorado, Pagosa
Springs, 26 July 1899, C. E Baker" (syntypes NDCl,
NY!. RM!, GH!, F!).
The lectotype collected by Greene is in fruit;
syntypes at NDG and NY have both flowering and
fruiting branches. In both the loment articles are
markedly winged and strigose; herbage is strigose
also (see Greene 1900).

Hedysarum occidentale Greene, Pittonia 3: 19. 1896.

Type locality: "Olympic Mountains, Washington, 1890,
C. V. Piper" (Greene I.e.).

Type: Washington, "Olympic Mts., C. V' Piper 905,
flowers 11 August 1890, fruit 30 Sept. 1890" (holo-
type NDG!).

Greene (1896) provides a description and a short
note: "Plant like H. horeale when in flower, though
with broader leaflets and widely different fruit." A
second sheet fi-om the Olympic Mountains at NDGl,
Piper 2227 (August 1895), has the epithet "occiden-
tale" in Greene's hand, but it was not cited by him.
It is much better material than the type. For a long
time the name H. horeale was included within the
concept of H. alpimnn. It is likely that Greene was
under a similar misconception. The general aspect
of H. occidentale (i.e., conspicuously veined leaflets
and large loments with prominently reticulate
venation), which occurs from Vancouver Island,
British Columbia, and the Olympic Peninsula,
Washington, disjunctly eastward to northern and
eastern Idaho, western Montana, western Wyoming,
northeastern Utah, and montane southern Colorado,
is that of H. alpinum; and it differs generally in the
manner indicated by Greene.

The most distinctive feature separating most, if
not all, specimens of H. occidentale from H.

Great Basin Naturalist

[Volume 55

alpintirn is the much larger, rather conspicuously
wing-margined loment segments. Flowers are gen-
erally larger, often much larger. However, plants
from the Absaroka Range of northwestern
Wyoming approach H. alpinum in occasionally hav-
ing small flowers, but when collected at maturity,
the fruit is that of H. occidentale . Additional collec-
tions might demonstrate that H. alpinum per se is
indeed in the Absarokas. Large-flowered plants of
H. alpinum, mainly of frigid sites in the arctic,
approach the size of flowers of some H. occidentale
specimens, but the fruit there is that of H. alpinum.
H. occidentale has loments very similar to those of
the closely allied H. sulphur escens.

Hedysariim occidentale Greene var. canone Welsh, Great
Basin Nat. 38: .314. 1978.

Type: Utah, Carbon County, "ca 14 mi due ENE of
Helper, Soldier Creek,' 30 June 1977, Welsh &
Taylor 15256" (holotype BRY!; isotype at NY!).

The syndrome of characters associated with this
taxon is shared individually elsewhere within the
species as a whole. However, specimens from
Duchesne, Carbon, and Emeiy counties, Utah, and
Gunnison County, Colorado, are recognizable by
their large, thick, ovate to ovate-lanceolate, yellow-
green leaflets, and large pale flowers. Plants are
known from rather xeric sites in pinyon-juniper and
mountain brush communities, whereas plants of the
type variety are mainly of more mesic sites.
Although the taxon is segregated on weak diagnos-
tic features, it seems to be at least a trend worthy of
taxonomic recognition. It has long been known in
collections.

Hedysarum pabiilare A. Nels., Proc. Biol. Soc. Wash. 15:
185. 1902.

= H. horeale Nutt. var. horeale

Type: Wyoming, Wind River, Dubois, A. Nelson 752,
1894 (lectotype here designated RM!).

This name is based on several Wyoming,
Colorado, and Utah syntypes: i.e., M. E. Jones
5592, Soldier Summit, Utah, in 1894, POM?, BM!,
F!; Snake River, Wyoming, A. Nelson 3496, 19
August 1899 RM!; Wyoming, Natrona Co., Bates
Creek, L. N. Goodding 201, 5 July 1901, RM!, F!

Hedysarum palndare var. rivulare L. O. Williams, Ann.
Missouri Bot. Card. 21: 344. 1934.
— H. horeale Nutt. van horeale

Type: Wyoming, Teton County, along the Snake River,
31 July 1932, L. O. Williams 975 (holotype RM!;
isotypes GH!, CAS!).

Hedysarum philoscia A. Nels., Proc. Biol. Soc. Wash. 15:
185. 1902.

Basionyin: H. alpinum var. philoscia (A. Nels.) Rollins;

H. alpinum ssp. philoscia (A. Nels.) Love & Love
= H. alpinum L.

Type: Wyoming, Albany County, Head of Crow Creek,
Laramie Mountains, 1896, A. Nelson 2034; holotype
RM!

Material from the Black Hills of South Dakota
and from southeastern Wyoming is morphologically
similar and has been recognized as belonging to a
taxon that sui-vived south of the major glacial events
of the Pleistocene. The main diagnostic criterion is,
however, loment pubescence. That feature is incon-
stant within the southern material and often is pres-
ent in plants far beyond its supposed range (which
has been plotted to include plants as far north as
the 50th parallel). Recognition of plants at any taxo-
nomic rank is, therefore, problematical.

Hedysarum roezlianum Prantl, Ind. Sem. Hort. Wirceb.
8. 1873.

= H. horeale Nutt. var. horeale?

I have been unable to find any reference to this
taxon aside from its citation by Rollins (1940).

Hedysarum sulphurescens Rydb., Bull. Torrey Bot. Club.
24:251.1897.

Basionyin: H. flavescens Coult. & Fisher, not Regel &
Schmalh.

Yellow to yellowish flowers easily distinguish
this entity, which shares the peculiar loment fea-
tures of H. occidentale. The species ranges from the
southern British Columbia-Alberta Rockies south
through north central Washington, northern Idaho,
western Montana, and northwestern Wyoming.

Hedysarum truncatum Eastw., Bot. Gaz. 33: 205. 1902.

= H. alpinum L.

Type: Alaska, Nome, Dr E E. Blaisdell s.n. summer
1900 (lectotype NY! here designated; isolectotype
GH!).

Type material is low, about 2-2.5 dm tall, has
mature flowers about 12 mm long, and has fruit the
size and conformation of H. alpinum. It is identical
for all practical purposes with material named by
Eastwood simultaneously as H. auriculatum and
taken at the same place and time by the same col-
lector in 1900.

Hedysarum uintahense A. Nels., Proc. Biol. Soc. Wash.
15: 186. 1902.

= H. occidentale Greene

Type: Wyoming, "In draws of the foothills, Evanston,
A. Nelson 7198, 14 June 1900" (holotype RM!; iso-
types RM!, NY!, GH!).

Type sheets uniformly bear thick, lanceolate to
lance-ovate leaflets similar to var. canone, but with
flowers of typical H. occidentale. Plants from south-
west Wyoming are not uniformly of the uintahense
type, but vary from one population to another, with
most being similar to traditional H. occidentale.

Hedysarum utahense Rydb., Bull. Torrey Bot. Club 34:

424. 1907.

— H. horeale Nutt. var. horeale

Type: Utah, Salt Lake County, "vicinity of Salt Lake

City Utah," Leonard 55, 26 May 1883 (holotype

NY!).

The type consists of two complete stems and a

fragmentary branch; it is typical of the material

1995]

North American Hedysarum

growing through much of Utah and elsewhere in
the West.

References

DORE, W. G. 1959. Some inflorescence form.s in clovers
and other legumes. Canadian Field Naturalist 73;
147-154.

Elisens, W. J. 1985. The Montana collections of Francis
Duncan Kelsey. Brittonia37: 382-391.

Fedtschenko, B. 1902. A revision of the Genus Hedy-
sarum. Acta Hortus Petropolitani 19: 185-325.

Fernald, M. L. 1933. Recent discoveries in the Newfound-
land Flora. Brittonia 35: 26.5-283.

Gree.ne, E. L. 1892. New or noteworthy species. — XIV.
Pittonia 2: 293-298.

. 1896. New or noteworthy species. — XV Pittonia

3: 12-28.
. 1897. New or noteworthy species.

-XIX. Pittonia

3: 212-230.

. 1900. A fascicle of new Papilionaceae. Pittonia 4:

132-139.

McKelvey, S. D. 1955. Botanical e.xploration of the Trans-
Mississippi West, 1790-1850. Arnold Arboretum,

Jamaica Plains. 1144 pp.
NoRTHSTROM, T. E. 1974. The Genus Hedysarum in North

America. Unpublished master's thesis, Brigliam Young

University, Provo, UT
NoRTHSTROM, T. E., AND S. L. WELSH. 1970. Revision of the

Hedysarum horeale complex. Great Basin Naturalist

30: 109-130.
NUTTALL, T 1818. The genera of North American plants.

Volume 2. D. Heartt, Philadelphia, PA.
MiCHAUX, A. 1803. Flora Boreali-Americana. 2: 1-340.
Rollins, R. C. 1940. Studies in the genus Hedysarum in

North America. Rhodora 42: 217- 239.
Standley, P C. 19.30. Studies of American plants — III.

Field Museum of Natural Histor\' Botanical Series 8:

3-73.
TORREY, J., AND A. GRAY. 1838. Flora of North America.

Volume 1. Wiley & Putnam, New York, NY.

Received 28 February 1994
Accepted 3 June 1994

'4-77

WHIPWORM (TRICHURIS DIPODOMYS) INFECTION IN KANGAROO RATS
{DIPODOMYS SPE): EFFECTS ON DIGESTIVE EFFICIENCY

James C. Munj^^er' and Todd A. Slicliterl

Abstract. — To determine whether infections by whipworms (Trichuris dipodornijs [Nematoda: Trichurata:
Trichuritlae] ) might affect digestixe efficiency and therefore energ>' budgets of two species of kangaroo rats {Dipodomijs
micrups and Dipodonujs ordii [Rodentia; Ileteromyidae]), we compared the apparent dry matter digestibility' of three
groups of hosts: those naturally infected with whipworms, those naturally uninfected with whipwoiTns, and those origi-
nally naturally infected but later deinfected by treatment with the anthelminthic Ivermectin. Prevalence of T.
dipodotnys was higher in D. tnicrops (53%) than in D. ordii (14%), Apparent dr\' matter digestibility was reduced by
whipworm infection in D. microps but not in D. ordii. Although a statistically significant effect was shown, its small mag-
nitude indicates that whipworm infection is unlikely to have a biologically significant impact on the energy budgets of
host kangaroo rats.

Key words: parasite, digestive ejficieney, whipwonn, kangaroo rat, Trichuris, Dipodomys, energy budget.

Parasites inhabiting the gastrointestinal
tract of a host may reduce the efficiency of the
organs they inhabit either through direct com-
petition for nutrients or through damage to
absoqDtive surfaces. Because decreased diges-
tive efficiency may reduce the rate of energy
input into a host, gastrointestinal parasites have
the potential to cause a change in host energy
allocation (e.g., reduced activity or reduced
reproduction), and thereby impact the ecology
of the host (Munger and Karasov 1989).

Tapeworm infections have a measurable
effect on digestive efficiency, but a biologically
unimportant effect on the energy budget of
host white-footed mice [Peromyscus leucopus;
Munger and Karasov 1989). The present study
was designed to determine if infection by a
nematode, the whipworm Trichuris dipodomys,
has a substantial effect on one aspect of the
energy budget, digestive efficiency, of host
kangaroo rats {Dipodomys microps and D. ordii).

Materials and Methods

Our study site, located 2 km N of Muq^hy,
Owyhee County, ID, is in desertscrub habitat
with sandy loam substrate. Primary shrub
species of the study area are Artemisia
spinescens, Artemisia tridentata, Atriplex
canescens, Atriplex confertifolia, Atriplex spin-
osa, and Chrysothatnniis nauseosus. Six rodent

species were captured at tlie site, Ammospermo-
philus leucurus, Neotoma lepida, Perognathus
flaviis, Peromyscus manicuhitus, and two
species of kangaroo rats, Dipodomys ordii and
Dipodomys microps. Dipodomys ordii ranges
from 42 to 72 g and consumes a diet consisting
primarily of seeds (Zeveloff 1988). Dipodomys
microps is larger, 72-91 g, and is unique among
kangaroo rats in that it relies heavily on leaves
of Atriplex confertifolia for forage (Kenagy
1972, Zeveloff 1988). Both species are liable to
infection by the whipworm Trichuris dipodo-
mys, a nematode that inhabits the cecum of
infected hosts (Giimdmann 1957, Whitaker et
al. 1993).

On the study site we established a 13 X 13
grid of 169 Sherman live traps baited with
millet and placed at 15 m intervals. During
two trapping sessions, 14-22 June and 15-18
August 1990, kangaroo rats (30 individuals of
D. microps and 85 of D. ordii) were captured
and brought into the laboratory. Fecal speci-
mens from each animal were anaK zed for the
presence of parasite eggs by standard centrifu-
gal flotation techniques using saturated sucrose
solution (Pritchard and Kruse 1982). Six in-
fected but untreated animals from the June
experiment were included in the pool of ani-
mals used in the August experiment. The few
animals that failed to thrive in the lab were
removed from the experiment; data from a

'Department of UiologN, Boise State University, 1910 University Drive, Boise, ID 83725.

1995]

Whipworms in Kangaroo Rats

total of 29 D. microps individuals and 56 D.
ordii were analyzed.

Each month's set of captures was subjected
to the following protocol:

(1) Kangaroo rats were acclimated to a diet
of millet seed for 3-11 d.

(2) A pretreatment feeding trial was per-
formed: Animals were placed in wire-bottomed
cages with a measured amount of whole millet
seed. At the end of 5 d, fecal pellets were sep-
arated from spilled food and dried >24 h at
50° C. Initial digestive efficiency of each ani-
mal was measured as apparent dry matter
digestibility (i.e., the proportion of mass con-
sumed but not lost as waste), which was calcu-
lated as (Mpo -Mpe) / Mpo, where Mpo and
Mpp are the mass of food consumed and feces
produced, respectively.

(3) Half of the infected animals were then
injected subcutaneously with a solution of
Ivermectin (a svstemic anthelminthic; Ivomec
brand, from MSD AG VET, Rah way, NJ).
Figure I gives sample sizes of treatment groups.
June captures received, on each of two con-
secutive days, a 0.2-cc injection of Ivermectin
in 40% glycerol formal and 60% propylene
glycol; each injection delivered ca 350 /xg
Ivermectin / kg body mass. Controls received
equal-volume injections of the glycerol for-
mal-propylene glycol carrier. This dosage had
little effect on the presence of whipworm eggs
in feces of injected animals. Therefore animals
received 8 d later a second set of two injec-
tions, each of 0.15 cc and delivering ca 2 mg
Ivermectin / kg body mass; control animals
received the carrier August captures received,
on each of two consecutive days, an injection
of 0.15 cc volume delivering ca 2 mg
Ivermectin / kg body mass. Control animals
received the carrier. To control for possible
side effects of Ivermectin, half of the uninfect-
ed animals captured in August were also
injected with a solution of Ivermectin.

(4) Two days after each set of injections a
posttreatment feeding trial was conducted
using techniques in (2) above. Only results of
the pretreatment feeding trials and feeding
trials following the 2-mg Ivermectin / kg body
mass injection will be presented below.

Results and Discussion

Adult worms (seven of each gender) taken
from a Dipodomys microps at our site were
identified as Trichuris dipodomys. Although

m a

.02-

Deinfecled

•

Infected

Uninfected

,01 •

..s

0 -

«>9

) 40

- ,01 •

r 6,0

l|7

110

— np.

P. microDS

D. ordii

Fig, 1. Effects of variation in parasite load on propor-
tional change in dry matter digestibility. Means ± SE.
Numbers represent sample sizes.

some minor morphological differences from
the original species description (Read 1956)
do exist, perhaps as a result of geographical
variation, the specimens most closely match
Read's description of T. dipodomys (A. Shostak
personal communication). Measurements of
several key morphological characters are as
follows (X ± SD): total length: S 25.6 ± 0.8
mm, 9 41.3 ± 2.9 mm; hindbody length: 6
12.7 ± 0.4 mm, 9 23.7 ± 1.9 mm; spicule
length: 850 ± 85.1 yitm; egg length: 64.8 ± 5.0
fxm; egg width: 33.5 ± 1.0 /xm. Voucher speci-
mens were deposited with the University of
Alberta Parasite Collection (#'s UAPC 11464
and UAPC 11465). Although we did not identi-
fy whipworms from D. ordii, we are confident
they are T. dipodomys; the type host for T.
dipodomys is D. ordii, and T. dipodomys is
known only from D. ordii and D. microps
(Whitaker et al. 1993).

Prevalence in Host Species.

Trichuris dipodomys occurred at substan-
tially higher prevalence in D. microps than in
D. ordii (Table 1), a result similar to that of
Grundmann (1957). We can speculate as to
three possible explanations for this pattern.
The first is that eggs produced by adult worms
in D. microps may become embryonated more
easily than those in D. ordii. Freshly produced
fecal pellets of D. microps appear moister than
those of D. ordii (Munger personal observa-
tion), probably because of the higher amount
of green or leafy vegetation in the diet of D.
microps. If moisture is necessary for embiyona-
tion of the eggs (as is implied by Parry 1968),

Great Basin Naturalist

[Volume 55

Table 1. Infection oltwo species of kangaroo rat with the
nematode Trichuris dipodornijs.

D. microps

D. ordii

Infected Uninfected

June trapping 10 5
August trapping 6 9

5 39

7 34

moister feces may lead to higher embryonation
rates and therefore higher prevcilence among D.
microps. The second explanation is that social
and burrow use behavior may differ between
these species. For example, perhaps D. microps
individuals visit one another's burrows (and
thereby become exposed to parasite eggs) at a
substantially higher fiequency tlian do D. ordii.
Also, D. microps inhabits a mound up to 2 m
in diameter while D. ordii inhabits less sub-
stantial individual holes. Studies of another
system of two species of kangaroo rats has
shown that the larger, mound-inhabiting D.
spectahilis uses its burrow system for pro-
longed periods, while the smaller D. merriami
rotates among several burrows (Jones 1989).
This latter behavior would tend to reduce
reinfection of individuals; it would be interest-
ing to see if behaviors differ similarly between
D. microps and D. ordii. The third explanation
is that resistance to infection may differ
between these two host species.

Effects on Digestive Efficiency

Apparent dry matter digestibility (ADMD)
of millet seed was quite high, >95% on aver-
age (Table 2), a figure comparable to that
found by Schrieber (1979) for granivorous
rodents. Injection of Ivermectin did not
appear to affect ADMD of animals uninfected
by whipworms, an effect that might occur
through the removal of other symbionts, or
through some direct effect (proportional
change in ADMD, X ± SE; untreated:
-0.0043 ± 0.0035, treated: -0.0058 ± 0.0037).
Therefore, in the following analyses all natu-
rally uninfected animals are combined into one
class.

The effect of whipworm removal on ADMD
was analyzed with a two-way analysis of vari-
ance (AN OVA). One factor analyzed was the
treatment: deinfected (naturally infected but
treated with Ivermectin) vs. infected (naturally
infected but not treated with Ivermectin) vs.
naturally uninfected. The other factor was

Table 2. Effects of whipworm infection on apparent
dry matter digestibility (ADMD). Standard errors are in
parentheses. Figures on change between initial and final
feeding trials, as well as sample sizes, are in Figure 1. See
text for a description of treatments.

Treatment

Deinfected Infected Uninfected

Dipodomys microps

Initial ADMD .9.56 .96.5 .9.55

(.0051) (.0029) (.010.3)

Final ADMD .961 .9.50 .953

(.0039) (.0026) (.00.52)
Dipodomijfi ordii

Initial ADMD .967 .957 .961

(.0107) (.0076) (.0022)

Final ADMD .9.55 .9,58 .9.57

(.0034) (.0037) (.0014)

species. Experimental period (July vs. August)
was included as a blocking factor. The depen-
dent variable in the analysis was proportional
change between pretreatment and posttreat-
ment ADMD ([post-pre]/pre); this measure
should be more sensitive than posttreatment
ADMD in expressing treatment effects because
it takes account of initial differences in ADMD
among hosts.

Although there were no statistically signifi-
cant main effects of treatment or species on
ADMD, there was a significant interaction
between these factors (Table 3), indicating that
the two host species differ in their response to
treatment. This difference between species was
explored using a separate AN OVA for each
species, which revealed that treatment with
Ivermectin had a significant effect on change
in ADMD in D. microps, but not in D. ordii
(Table 4, Fig. 1). A Tukey's a posteriori multiple
sample test revealed that, within D. microps,
the change in ADMD of the deinfected group
differed significantly from the change in
ADMD of both the infected group and the
uninfected group. These results can be inter-
preted as showing that the deinfected group
had 1.9% higher ADMD than the other two
groups.

Of interest is the lack of effect Trichuris
causes in D. ordii. This may be due to what
appears to be a higher intensity of infection
(more parasites per infected host) in D. microps:
fecal floats of D. microps in general contained
more eggs than did floats of D. ordii J^D.
microps X = 254, SE = 115.2; D. ordii X =
63.5, SE = 21.0; Mann-Whitney U-test, U =
79, P = .1). If fewer worms were present in D.

1995]

Whipworms in Kangaroo Rats

Table 3. F values and probability values (P) from three-way analyses of variance on effects of species, month, and
treatment (deinfected, infected, or uninfected) on apparent dry matter digestibility (ADMD).

Dependent

variable

Proportion

change

Initial ADMD

Final ADMD

in ADMD

Source

Treatment

.15

.86

.72

.49

.47

.63

Species

.46

.50

.82

.37

.33

.57

Treatment * Species

1.33

.27

1.78

.18

4.74

.012

Block (= Month)

.51

.48

9.11

.003

.00

.95

Error

Table 4. Results from one-way analyses of variance on
the effect of treatment (deinfected, infected, and uninfect-
ed) on % change in dry matter digestibility in D. microps
and D. ordii.

Species Source

D. microps

D. ordii

Treatment 2 .00106 4.64 .019

Error 27 .000229

Treatment

Error

2
52

.00034
.01442

1.21

.31

ordii, the effect of eradicating those worms
would have been less apparent.

One might question the biological impor-
tance of the slight, albeit statistically significant,
decrease in ADMD caused by Trichuris infec-
tion. Munger and Karasov (1989) showed an
effect of similar magnitude resulting from tape-
worm infection [Hijrnenolepis citelli) in white-
footed mice {Perotnyscus leucopus). They
argued that hosts can easily compensate for
such small effects by slight increases in food
consumption or decreases in expenditures, or
by changes in gut morphology (Mettrick 1980),
and concluded that such effects on ADMD are
unlikely to affect host energy budgets or to
translate through to population-level effects.
The same conclusion is likely to apply to the
kangaroo rat-whipworm system.

Acknowledgments

We thank Allen Shostak of the University of
Alberta's Parasitology Museum for measuring
specimens of the parasite and for its identifica-
tion. Kay Kesling helped both in the field and

in the lab. Sara Murray and Aaron Munger
helped in the field. Discussion with Mary Price
was helpful, as were comments from anony-
mous reviewers. This research was supported
by an Intramural Faculty Research Grant from
Boise State University.

Literature Cited

Grundmann, a. W. 1957. Nematode parasites of mam-
mals of the Great Salt Lake Desert of Utah. Journal
of Parasitology 43: 105-112.

Jones, W. T. 1989. Dispersal distance and the range of
nightly movements in Merriam's kangaroo rats.
Journal of Mammalogy 70: 27-34.

Kenagy, G. J. 1972. Saltbush leaves: excision of hypersaline
tissue by a kangaroo rat. Science 178: 1094—1096.

Mettrick, D. E 1980. The intestine as an environment
for Hijrnenolepis diminuta. Pages 281-356 in H. P
Arai, ed.. Biology of the tapeworm Hijrnenolepis
diminuta. Academic Press, New York, NY.

Munger, J. C., and W. H. K.\r\sov. 1989. Sublethal para-
sites and host energy budgets: tapeworm infection in
white-footed mice. Ecology 70: 904-921.

Parry, J. E. 1968. Transmission studies of nematodes with
direct life histories in selected Utah mammals.
Unpublished doctoral dissertation, University of
Utah, Salt Lake City.

Pritchard, M., and G. Kruse. 1982. The collection and
preservation of animal parasites. University of
Nebraska Press, Lincoln. 141 pp.

Read, C. P 1956. Trichuris dipodomijs, n. sp., from Ord's
kangaroo rat. Proceedings of the Helminthological
Society of Washington 23:119.

Schrieber, R. K. 1979. Coefficients of digestibility and
caloric diet of rodents in the northern Great Basin
desert. Journal of Mammalogy 60:416-420.

Whitaker, J. O., Jr., W. J. Wrenn, and R. E. Lewis.
1993. Parasites. In: H. H. Genoways and J. H.
Brown, eds.. Biology of the Heteromyidae. American
Society of Mammalogists Special Publication 10.

Zeveloff, S. I. 1988. Mammals of the Intermountain
West. University' of Utah Press, Salt Lake City. 365
pp.

Received 27 July 1993
Accepted 20 June 1994

LOCAL DISTRIBUTION AND FORAGING BEHAVIOR OF THE

SPOTTED BAT {EUDERMA MACULATUM) IN NORTHWESTERN

COLORADO AND ADJACENT UTAH

Jay F Stoiz^

Abstract. — This study investigated local distribution and foraging behavior of the spotted bat {Eudenna maculatum)
in Dinosaur National Monument, Colorado-Utah, by monitoring audible echolocation calls. The occurrence of this
species was verified in a variety of habitat types in canyon bottoms and other relatively low elevation sites, indicating
that the animals are widely distributed and locally common in the area. Foraging spotted bats concentrated flight activi-
ty in the open-air space above meadows and occasionally exploited near-canopy habitat (within 8 m of foliage). Bats
began to forage shortly after dark, and activity levels were relatively constant throughout the night. Foraging spotted
bats attacked airborne prey every 2.15 min on average. Consistent with published observations, spotted bats maintained
exclusive foraging areas. Distinct vocalizations indicating agonistic encounters occurred when a bat encroached on the
foraging area of a conspecific.

Key words: spotted bat, Euderma maculatum, Colorado, Utah, Dinosaur, National Monument, foraf:,ing, habitat me,
attack rates, echolocation.

The spotted bat {Eudenna maculatum) is
widely distributed across western North
America and apparently exists in low popula-
tion numbers throughout its range (Fenton et al.
1987). The species is rare in collections, and
viable populations have been documented in
only a few widely separated localities (Watkins
1977, O'Fan-ell 1981). Findings presented here
and those of Navo et al. (1992) indicate that E.
maculatum is locally common in canyon bot-
toms and other low-elevation sites in Dinosaur
National Monument, Colorado-Utah, and
occurs throughout a diverse range of habitat
types.

Population studies (e.g., Leonard and Fenton
1983) in south central British Columbia have
demonstrated that foraging spotted bats exhibit
considerable habitat specificity; radiotracking
in this same area (Wai-Ping and Fenton 1989)
has demonstrated that individuals are faithful
to specific sites over several consecutive nights.
However, no clear association between forag-
ing activity and any specific habitat conditions
is apparent. In British Columbia, spotted bats
forage over clearings in ponderosa pine {Pinus
ponderosa) forests, open fields, and marshes
(Leonard and Fenton 1983, Wai-Ping and
Fenton 1989). There is little information about

foraging habitat throughout the remainder of
the geographic range of E. maculatum.

The purpose of this study was (1) to investi-
gate local distribution of E. maculatum by
monitoring echolocation calls, (2) to identify
and describe foraging habitat, and (3) to make
a preliminary examination of spatial and tem-
poral patterns of habitat use by spotted bats in
the study area.

Methods

This study was conducted in the canyon
bottoms and other relatively low elevation
sites in Dinosaur National Monument (109°W,
40°31'N), northwestern Colorado and north-
eastern Utah, from 17 May to 9 June 1993.
Navo et al. (1992) provided a description of
the physiography and vegetation of Dinosaur
National Monument. In each study site where
spotted bats occurred, I monitored movement
patterns and foraging behavior by listening to
the low-frequency (15-9 kHz; Leonard and
Fenton 1984) echo-location calls of this
species, which are clearly audible to the unaid-
ed human ear (Woodsworth et al. 1981).

As reported previously (Navo et al. 1992),
E. maculatum is readily identifiable because it
has the lowest-frequency echolocation calls of

'Environmental, Population, and Organisniic Biologv', Universit>' of Colorado. Boulder. CO 80309-0.334. Present address: Department ofBiolog>'. Boston
University. Boston. M.\ 0221.5.

1995]

Spotted Bats in Colorado-Utah

any bat species in the study area. Nyctinomops
macrotis and Idionycteris phyllotis also pro-
duce orientation sounds that are partly audible
to humans, with frequencies of 25-17 kHz for
N. macrotis (Fenton and Bell 1981) and 24-12
kHz for I. phyllotis (Simmons and O'Farrell
1977). These two species inhabit southern
parts of the Colorado Plateau and the Great
Basin (Milner et al. 1990, Tumlison 1993), but
neither is known to occur as far north as
Dinosaur National Monument, extralimital
records of N. macrotis notwithstanding (Milner
et al. 1990). To further ensure conect call identi-
fication, I referred to recordings of known E.
maciilatum calls. I also visually identified free-
flying individuals (based on conspicuous white
venter and large ears) at close range in the
beam of a high-intensity flashlight after locat-
ing the animals by listening to orientation
sounds. It should be noted that /. phyllotis is
often buff-colored ventrally and therefore could
be visually misidentified as E. maculatum in
areas of sympatry.

Sampling Locations

To investigate the ecological distribution of
E. maculatum, I sampled 15 sites at 12 loca-
tions representative of common low-elevation
habitat types in the area (Fig. 1). Riparian sites
(Jenny Lind Rock, 1603 m; Echo Park, 1553
m; Split Mountain Gorge, riverbank and sand-
bar, 1439 m) are characterized by wide chan-
nels and reaches of calm water bounded by
steep sandstone cliffs. Isolated stands of box-
elder {Acer negundo) and cottonwood {Poptdiis
fremontii) line the riverbanks along with thick-
ets of tamarisk {Tamarix sp.).

Orchid Draw (1484 m) and Red Wash (1537
m) are dry desert washes characterized by rab-
bitbrush {Chrysothamnus nauseosus), sage-
brush {Seriphidium tridentata), greasewood
[Sarcobatus vermiculatus), and shadscale {Atri-
plex confertifolia), with tamarisk dominating
drainage bottoms.

Echo Park Meadow (1548 m) and Pool Creek
(1635 m) are both open meadows with domi-
nant ground cover of cheatgrass {Anisantha
tectortim), various bunchgrasses, and isolated
clumps of boxelder. Echo Park Meadow en-
compasses an area of ca 18 ha, bounded by the
Green River to the west and high (150-230 m)
sandstone cliffs on remaining sides. The mead-
ow at Pool Creek (ca 8 ha) is situated at the
mouth of a narrow canyon; boxelder and cotton-

wood form a dense, continuous canopy over
much of the adjacent creek.

Remaining locations consist of a moist mead-
ow (Hog Canyon, 1635 m), open sagebrush
shrublands (Rainbow Park, 1488 m; Island
Park, 1512 m), and a narrow canyon with thick
riparian vegetation (Jones Hole, 1585 m).

Sampling Methods

At all locations I remained at a single site
during each night of sampling. By pacing from
a boxelder, which sei"ved as a focal point of bat
foraging activity in Echo Park Meadow, I esti-
mated that calls of E. maculatum were detect-
able at a distance of roughly 100 m. Therefore,
the area sampled at each site is here defined
as the air space within a hemisphere of radius
100 m. On several nights periods of high wind
and/or rain reduced this range of detectability,
with an attendant underestimation of bat activ-
ity. Furthermore, sites differed slightly in levels
of background noise from nearby streams, the
amount of obstructive vegetation, and various
atmospheric conditions such as relative
humidity, all of which affect the propagation of
sound (Lawrence and Simmons 1982).

Study sites were situated either at the
mouths of canyons or draws or in the middle
of open areas where movement patterns of bats
could best be assessed and the range of detect-
ability was maximized. In locations character-
ized by expansive terrain (open meadows or
shrublands), I monitored two different sites
separated by >300 m on consecutive nights to
assess uniformity of activity levels over large
areas. All sites were monitored from 2000 to
0200 h with the exception of Echo Park
Meadow, which was monitored from 2000 to
0400 h for seven consecutive nights (19-26
May) to assess temporal patterns of foraging
activity.

At locations where I observed high levels of
foraging activity (e.g., Echo Park Meadow and
Pool Creek), bat activity was quantified by
timing the duration of individual foraging ses-
sions and recording the number of feeding
buzzes (the increased rate of echolocation
pulse repetition associated with attacks on air-
borne prey; Griffin et al. 1960). Following
Leonard and Fenton (1983), the occurrence of
feeding buzzes indicates foraging activity, and
a foraging session is defined as the time during
which a single spotted bat hunted continuous-
ly within the study site. To permit comparison

Great Basin Naturalist

[Volume 55

1
1

5 km

— 40° 35' N

-^. 1

Jones ' '
Hole t

Echo
Park

Island
fe^ Park

Rainbow^,^
Park^^
• /

^ 1

Pool Creek
1

Jenny Lind Rock

Echo Park C.
Meadow 7

Orchid R^d
Draw ^^sh

• •

C_ Split Mtn. Gorge (sandbar)

Yampa River

f|> Split Mtn. Gorge (riverbank)
'In. Hog Canyon

109° W

~~~^^

UTAH COLORADO

c V

1 Green River

Fig. 1. Map showing sampling locations for monitoring activity oi Euderma maculatum in Dinosaur National
Monument in late spring 1993. Circles = locations at which transient occurrences of commuting or foraging bats were
recorded; triangles = foraging areas (see te.xt for details).

of relative levels of activity throughout the
night, the time spent by spotted bats in the
study site was totaled for every 15-min period
sampled. Sampling periods during which
heavy rain occurred were not considered.

To assess spatial patterns of habitat use, I
described the foraging flights of spotted bats
into a minicassette recorder and noted flight
patterns and use of available foraging space
relative to a near-canopy habitat zone (within
8 m of tree canopies) and an open-area zone
(the clutter-free air space over the open mead-
ow). These habitats con^espond to habitat zones
1 and 4, respectively, proposed by Aldridge
and Rautenbach (1987). I recorded the dura-
tion of foraging activity occurring within each
zone as well as the number of bats simultane-
ously present within the study site and inter-
actions between them.

I recorded the number of feeding buzzes
heard during each foraging session for each
night of observation at Echo Park Meadow
and Pool Creek for the purpose of calculating
attack rates (feeding buzzes/min) of foraging
spotted bats. I considered only those foraging

sessions of duration >3 min during which all
feeding buzzes produced by a single individ-
ual within the study site could be counted
accurately.

At other locations where I observed only
transient occurrences of foraging or commut-
ing spotted bats, activity was quantified by
recording the number of bat passes (sensu
Fenton 1970) per 15-min sampling period.

Results and Discussion

I observed spotted bats in 13 of 15 sites
sampled (Table 1). At 8 of these locations I
observed only commuting bats. Passes of com-
muting spotted bats occurred sporadically
throughout the night. At locations where two
separate sites were monitored on consecutive
nights, the number of passes remained fairly
constant (passes/night: Echo Park, 5, 4; Hog
Canyon, 5, 6; Island Park, 6, 10), and direc-
tions of travel appeared similar for bats on
both nights.

Availability of cliff roosting sites has been
suggested as a limiting factor in the distribu-

1995]

Spotted Bats in Colorado-Utah

Table 1. Number of passes of Eiiderma maciilatiim per 15-min sampling period between 2000 and 0200 h at Dinosaur
National Monument (16 May-8 June 1993). See text for general description of habitat types.

Number of

Number

sampling

Sampling location

of nights

periods

Range

Jenny Lind Rock

—

Echo Park

.19

0-2

Hog Canyon

.23

0-3

Orchid Draw

.75

0-6

Red Wash

1.33

0-1

Split Mountain Gorge

(sandbar)

.04

0-5

Split Mountain Gorge

(riverbank)

.33

0-8

Rainbow Park

.25

0-1

Island Park

.33

0-2

Jones Hole

—

All sites

280

.35

0-8

tion of E. maculatiun (Easterla 1973). The
abundance of high ehfifs in Dinosaur National
Monument as well as transient occurences of
commuting bats throughout a variety of wide-
ly separated low-elevation sites (Navo et al.
1992, this study) suggests that suitable roost-
ing habitat is widespread throughout the area.
However, information about microclimate re-
quirements of this species is needed to fully
assess actual availability of suitable roost sites.

I obsei-ved foraging spotted bats by sight and
sound at five locations, three of which (Echo
Park, Orchid Draw, Red Wash) involved only
transient occurrences of bats that were
observed executing steep dives and other
abrupt flight maneuvers coincident with feed-
ing buzzes as they passed through the area. I
observed a single spotted bat foraging over a
sand-and-gravel bar at Echo Park, but activity
levels at this location were lower than those
reported by Navo et al. (1992), who sampled
this same site previously. Fairly high levels of
activity occurred at Orchid Draw and Red Wash
(Table 1), and I heard three feeding buzzes at
each site. However, because spotted bats
apparently capture prey opportunistically
while commuting to specific foraging sites
(Wai-Ping and Fenton 1989), observations of
foraging bats passing through an area cannot
be considered indicative of habitat preferences.

At Echo Park Meadow, spotted bats first
arrived at the study site at 2123 h ± 11 min
Mountain Daylight Time (n = 6 rain-free
evenings), always after dark, and remained
active throughout the night (Fig. 2). Spotted
bats foraged within the study site for 6.22 ±
2.40 min out of every 15-min sampling period
between 2100 and 0400 h {n = 2490 min; Fig.

2), and foraging sessions lasted 5.48 ± 2.74
min {n = 187). At Pool Creek, spotted bats
hunted within the study site for 6.82 ± 5.03
min out of every 15-min sampling period
between 2100 and 0200 h (n = 525 min), and
foraging sessions lasted 8.97 ± 8.78 min (n =
30). These activity levels offer strong evidence
that open meadows represent important forag-
ing habitat for E. maculatum in this area.

Comparatively low levels of activity were
recorded at riparian sites adjacent to Echo
Park Meadow (Echo Park, Jenny Lind Rock).
Because no physiographic barriers are present
that might restrict accessibility to the bats, it
appears that open water courses do not repre-
sent foraging areas of choice. These observa-
tions agree with those of Leonard and Fenton
(1983), who reported that in British Columbia
spotted bats foraged in forest clearings and
open fields to the exclusion of a nearby river

The temporal pattern of foraging activity in
Dinosaur National Monument is similar to
that reported from British Columbia (Leonard
and Fenton 1983), where spotted bats were
active throughout the night. Because radio-
tracking (Wai-Ping and Fenton 1989) has
demonstrated that individual spotted bats
hunt on the wing >300 min per night, reports
of apparent peaks in nightly activity (which
have been especially pronounced in mistnet-
ting studies, e.g., Easterla 1973) are likely arti-
facts related to the proximit)' of sampling sites
to diurnal roosts and/or drinking sites.

At Echo Park Meadow and Pool Creek, for-
aging spotted bats typically flew in large circu-
lar or elliptical orbits at heights of 10-30 m
above the ground. In 1088.8 min of observa-
tion of foraging spotted bats at Echo Park

Great Basin Naturalist

[Volume 55

15.0-

•- 10.0

5.0-

Time (h)

Fig. 2. Foraging activity patterns ol Euderma nuwiihi-
tum at Echo Park Meadow (19-26 May 1993). Bars repre-
sent mean time (+ SD) spent by bats in the study site per
15-min sampHng period from 2000 to 0400 h (n = 6 for
each 15-min period in the inter\'al 2000-0345 h, n = 4 for
0345-0400 h).

Meadow, 81.5% of activity occurred over the
open meadow, which constituted roughly 85%
of the site, while 18.5% of activity occurred
within 8 m of the foliage of fully leafed box-
elders at mid- to upper-canopy level. Such
activity consisted of bats circling closely above
and around individual trees or isolated clumps
of trees. I rarely obsei"ved bats within 0.5 m of
the canopy, and I never observed hovering
flight or other evidence of foliage gleaning. In
290.8 min of obser\'ation of foraging spotted
bats at Pool Creek, all activity occurred over
the open meadow, although a much larger
percentage of the study site area comprised
canopies of boxelder and cottonwood than at
Echo Park Meadow.

The predilection of £. maciilatiim for forag-
ing over open terrain in Dinosaur National
Monument agrees with the pattern observed
in previous studies (e.g., Woodsworth et al.
1981, Leonard and Fenton 1983). Low-fre-
quency echolocation calls and long intercall
intei-vals suggest that spotted bats use a forag-
ing strategy based on long-range prey detec-
tion and high-level flight (Simmons and Stein
1980, Woodsworth et al. 1981, Barclay 1986).
This strategy likely is best suited to open areas
(Neuweiler 1984). Although 1 never directly

observed these bats gleaning prey from foliage
during this study, observations of near-canopy
foraging contrast with those of some other
workers (e.g., Wai-Ping and Fenton 1989) who
have reported that this species never attacked
insects near foliage or any other type of sur-
face. Information about individual variability
in foraging behavior is needed before drawing
conclusions about variabilit)' between popula-
tions related to different ecological conditions.

At both Echo Park Meadow and Pool Creek,
there were 118 instances in which two or three
E. macidatum were present within the study
site simultaneously. Leonard and Fenton
(1983, 1984) estimated that spotted bats in
British Columbia maintain a distance of at least
50 m between adjoining foraging areas and
suggested that this spacing is accomplished
through a combination of mutual avoidance and
active monitoring of encroachments by con-
specifics. This same system appears to be oper-
ating at foraging areas in Dinosaur National
Monument. Consistent with observations of
Leonard and Fenton (1983), foraging spotted
bats often produced agonistic vocalizations
when the 50-m buffer zone was breached by
an intruding bat. Such vocalizations sounded
qualitatively different from feeding buzzes and
occurred only during close-range encounters
between conspecifics. Information about
known individuals and resource availability is
needed to elucidate the role of agonistic inter-
actions in the foraging ecology of E. niacidatum.

During this study I heard a total of 247 feed-
ing buzzes, and never more than one per min
from the same individual. In a sample of 37
foraging sessions, spotted bats attacked an
insect eveiy 2.15 min on average (0.466 ± 0.294
attacks/min, range 0.16-0.94; n = 152 feeding
buzzes). These rates generally agree with val-
ues reported in previous studies (Leonard and
Fenton 1983, Wai-Ping and Fenton 1989), fur-
ther confirmation that this species attacks prey
at a rate much lower than is typical of bats that
forage from continuous flight (Barclay 1985,
Hickey and Fenton 1990).

Density of clutter in an environment im-
poses differential constraints on the maneuver-
ability and perceptual capacities of bats, there-
by determining the accessibility of different
habitats b\' influencing foraging efficiency (Neu-
weiler 1984, Aldridge and Rautenbach 1987,
Fenton 1990). Spotted bats appear to forage
preferentially in open areas, which may be

1995]

Spotted Bats in Color.\do-Utah

related to the use of a long-range foraging stiat-
egy (Barclay 1986), and the ability to exploit
edge situations may reflect a measure of
behavioral flexibility in this regard. Because
spotted bats are obviously not greatly restrict-
ed in foraging habitat with regard to vegeta-
tion associations (Wai-Ping and Fenton 1989,
Navo et al. 1992), structural features of the
environment related to density of clutter may
be more predictive of habitat suitability and
the use of available foraging space. However,
information on individual variability is needed
before drawing conclusions about the foraging
strategy of this species.

Acknowledgments

I am grateful to D. M. Armstrong, K. W.
Navo, M. B. Fenton, M. L. Leonard, M. A.
Bogan, C. E. Bock, J. A. Gore, and G. T. Skiba
for advice regarding study site locations and
sampling methods. I thank the personnel of
Dinosaur National Monument, and especially
S. J. Petersburg, for cooperation and for shar-
ing knowledge of the area. Critical comment
on the manuscript fiom D. M. Amistrong, M. B.
Fenton, R. M. Timm, M. J. O'Farrell, and two
anonymous reviewers was much appreciated.
Funding was provided by the Undergraduate
Research Opportunities Program, University
of Colorado at Boulder.

Literature Cited

Aldridge, H. D. J. N., AND I. L. Rautenbach. 1987.
Morphology, echolocation, and resource partitioning
in insectivorous bats. Journal of Animal Ecology 56:
763-778.

Barclay, R. M. R. 1985. Long- versus short-range forag-
ing strategies of hoary {Lasiunis cinereus) and silver-
haired {Lasioiujcteris noctivagans) bats and the con-
sequences for prey selection. Canadian Journal of
Zoology 63: 2507-2515.

. 1986. The echolocation calls of hoaiy {Lasiunis

cinereus) and silver-haired [Lasiomjcteris noctiva-
gans) bats as adaptations for long- versus short-range
foraging strategies and the consequences for prey
selection. Canadian Journal of Zoology 64:
2700-2705.

Easterla, D. a. 1973. Ecology of the 18 species of
Chiroptera at Big Bend National Park, Te.\as. North-
west Missouri State University- Studies 349: 1-165.

Fenton, M. B. 1970. A technique for monitoring bat
activity with results obtained from different environ-
ments in southern Ontario. Canadian Journal of
Zoology 48: 47-51.

. 1990. The foraging behaviour and ecologx' of ani-
mal-eating bats. Canadian Journal of Zoology 68:
411-422. '

Fenton, M. B., and G. P Bell. 1981. Recognition of
species of insectivorous bats by their echolocation
calls. Journal of Mammalogy 62: 233-243.

Fenton, M. B., D. C. Tennant, and J. Wyszeckl 1987.
LJsing echolocation calls to measure the distribution
of bats: the case of Euderma maculatum. Journal of
Mammalogy 68: 142-148.

Griffin, D. R., F A. Webster, and C. R. Michael. 1960.
The echolocation of flying insects by bats. Animal
Behavior 18: .5.5-61.

HiCKEY, M. B. C, and M. B. Fenton. 1990. Foraging by
red bats [Lasiunis boreaUs): Do intraspecific chases
mean tenitoriality? Canadian Journal of Zoology 68:
2477-2482.

Lawrence, B. D., and J. A. Simmons. 1982. Measurements
of atmospheric attenuation at ultrasonic frequencies
and the significance for echolocation by bats. Journal
of the Acoustical Society of America 71: 585-590.

Leonard, M. L., and M. B. Fenton. 1983. Habitat use by
spotted bats {Euderma maculatum, Chiroptera:
Vespertilionidae): roosting and foraging behavior
Canadian Journal of Zoologv' 61: 1487-1491.

. 1984. Echolocation calls of Eudenna maculatum

(Chiroptera: Vespertilionidae): use in orientation and
communication. Journal of Mammalogy' 65: 122-126.

Milner, J., C. Jones, and J. K. Jones, Jr. 1990. Nijcti-
nomops macrotis. Mammalian Species 351: 1—1.

Navo. K. W, J. A. Gore, and G. T Skiba. 1992. Obsewa-
tions on the the spotted bat, Eudenna maculatum, in
northwestern Colorado. Journal of Mammalogy 73;
547-551.

Neuweiler, G. 1984. Foraging, echolocation, and audi-
tion in bats. Natunvissenschaften 71: 446—455.

O'Farrell, M. J. 1981. Status report: Eudenna maculatum
(J.A. Allen). United States Fish and Wildlife Sei-vice,
Office of Endangered Species. 29 pp.

Simmons, J. A., and M. J. O'Farrell. 1977. Echolocation
by the long-eared bat, Plecotus phijllotis. Journal of
Comparative Physiology 122: 201-214.

Simmons, J. A., and R. A. Stein. 1980. Acoustic imaging
in bat sonar: echolocation signals and the evolution
of echolocation. Joumal of Comparative Physiology
135: 61-84.

Tumlison, R. 1993. Geographic variation in the lappet-
eared bat, Idiomjcteris phijllotis, with descriptions of
subspecies. Joumal of Mammalogy 74: 412^21.

Wai-Plng, v., and M. B. Fenton. 1989. Ecology of spot-
ted bats {Eudenna maculatum): roosting and forag-
ing behavior. Journal of Mammalogy 70: 617-622.

Watkins, L. C. 1977. Eudenna maculatum. Mammalian
Species 77: 1-4.

Woodsworth, G. C, G. R Bell, .-vnd M. B. Fenton. 1981.
Obsei-vations of the echolocation, feeding behavior,
and habitat use of Eudenna maculatum (Chiroptera:
Vespertilionidae) in southcentral British Columbia.
Canadian Journal of Zoology- .59: 1099-1102.

Received 7 Febnianj 1994
Accepted 20 June 1994

THE CHRYSOTHAMNUS-ERICAMERIA CONNECTION (ASTERACEAE)

Ijoran C. Anderson ^

AlJSTfUCT. — The geniis Chrysothdiiuiiis (Asteraceae) contains 16 species. Recently, 4 species were transferred to
Ericameria, and the remaining 12 were left in Chnjsothamnus. The remaining species are now transferred to Ericameria
as £. albida, E. depressa, E. eremobia, E. graminea, E.filifolia (formerly C. greenei). E. hwnilis, E. linifolia, E. molesta, E.
pulchella, E. pulclielloides (a fossil species), £. spathulaia, E. vaseyi, and E. liscidiflora. Section alignments are given,
and some infraspecific combinations are also made.

Key words: Chr>sothamnus, Ericameria, rahhithnish, nomenclature transfers.

The Asteraceae are a relatively young group,
and yet they have experienced rapid evohition
into a great number of species. One result is
that many taxa appear more distant moq^holog-
ically (phenotypically) than they actually are
genetically, and, conversely, some taxa may
appear more closely related than they are.
These situations have created havoc amongst
taxonomists in their attempts to circumscribe
genera. This is particularly evident in the tribe
Astereae. In 1894, E. L. Greene stated:

In North America the Astereae are excessively
numerous, and no natural assemblage of plants has
seemed to present such difficulties to the systema-
tist; and the widest conceivable diversities of opin-
ion as to the limits of genera have found expression
among botanists when undertaking to classify
them.

The situation continues a century later

The genus Haplopapptis was thought to be
an unnatural, polyphyletic assemblage by
many (e.g., Shinners 1950, Anderson 1966,
Johnston 1970, Turner and Sanderson 1971,
Clark 1977, Urbatsch 1978). Nevertheless,
because there was no suitable taxonomic reor-
ganization of the group, I continued to describe
new taxa in Haplopappiis (Anderson 1980a,
1983b), even though the species would probably
be placed in some other genera at a later date.
Recently, additional data have contributed to a
clearer understanding of the relationships in
this and related groups (Morgan and Simpson
1992), and several genera have been recog-
nized for North American Haplopappi.

In a 1976 presentation at national meet-
ings, I discussed the close affinity of Chnjso-

thamnus with woody elements of Haplopappiis
and suggested that the Asiris-Ericamcria-
Macronema complex of Haplopappiis probably
should be included in Chnjsothamnus. But,
given the state of knowledge at that time, I de-
ferred. In 1990, Nesom reorganized Ericameria
as a genus to include Asiris and Macronema.
Recently, based on occurrences of intergeneric
hybrids (Anderson and Reveal 1966, Anderson
1970) and DNA data (Morgan and Simpson
1992), Nesom and Baird (1993) transferred
four species of Chnjsothamnus into Ericameria
(C. nauseosus and C. parnji of section Nauseosi
and C paniculatus and C. teretifolius of section
Piinctati). They continued to recognize Chnjso-
thamnus as a distinct (but smaller) genus and
gave arguments for separating the two.

A problem in separating Ericameria and
Chnjsothamnus (sensu Nesom and Baird) is the
occurrence of hybrids (Anderson 1970, 1973)
between C. nauseosus (their Ericameria) and
C. alhidiis (their Chnjsothamnus). After study-
ing a specimen of only one of the three collec-
tions involved, Nesom and Baird (1993) deval-
ued the connection by stating that "the plant
in question [is] characteristic of C. nauseosus, and
we identify it as C. nauseosus, finding no strong
reason to implicate C. albidus in its parentage."
They stated that achenes of C. albidus are linear
and consistently producing 10 slightly raised
nerves, whereas those of C. nauseosus are nar-
rowly obovate with 5-7 nerves. Actually, ach-
enes of both species can be characterized as
being narrowly cylindrical. The number of
vascular bundles (associated with the nerves)
in the achenes averages approximately 7 and

'Dcpartuifiit ol Biological Science, Florida State University. Tallahassee, FL 32306-2043.

1995]

Chrysothminus-Ericameria Connection

ranges from 5 to 10 (but mostly 6-8 in Ash
Meadows) for C. alhidus (Anderson 1970,
1973), whereas achene bundle number in C.
nauseosus ranges from 5 to 12 (but is restrict-
ed to 5 for those in Ash Meadows).

The interspecific hybrid examined by Nesom
and Baird {Beatleij 11894, KSC) was studied
anatomically by Anderson (1973); its hybridity
is indicated by low pollen fertility and by mor-
phological intermediacy between the two
species in its revolute leaves, in vascular bundle
number in the ovary wall, in corolla lobe
length, and in anther appendage length. It has
secretoiy canals in the ovaiy wall and glandular
trichomes on the corolla tube (like C. nauseo-
sus, unlike C. alhidus) and ovaiy wall (unlike
C. nauseosus, like C. alhidus). Further, proge-
ny from one of my C. alhidus garden plants
also has low pollen fertility and looks interme-
diate between its seed parent and C. nauseo-
sus (Anderson 1970). Its flowers have secreto-
ry canals in the ovar>' wall and glandular tri-
chomes on the corolla tube but lack glandular
trichomes on the ovary wall; those three fea-
tures are characteristic of C nauseosus but not
of C. alhidus (the seed parent), clearly suggest-
ing hybridity. If existence of interspecific
hybrids is used to justify transferring C. nau-
seosus to Ericameria, then this feature also
argues for bringing the remainder of Chryso-
thamnus into Ericameria.

The warranted position of Chnjsothamnus
teretifolius in Ericameria is taken by Nesom
and Baird (1993: 80) because, like many Eri-
cameria species (sensu strictum), that species
has the tendency for the "resiniferous ducts
that are almost always distinctly associated
with the phyllaiy midvein to expand near the
apex of the phyllary." This characteristic also
occurs in many species of Chnjsothamnus
(sensu Nesom and Baird) as illustrated for C.
vaseiji (Anderson 1963: 660) and cannot be
used to distinguish the two groups. I have ob-
served adjacent populations of C. viscidiflorus
subsp. puherulus in which plants of one had
prominently enlarged resin ducts at the phyl-
lary tips and plants of the other did not.

With the transferral of four species from
Chnjsothamnus to Ericameria, Nesom and
Baird (1993) separate the two newly struc-
tured genera with six criteria. (1) Leaves 3-
nerved for Chrysothamnus and 1-nerved for
Ericameria — but many of their Ericameria
have prominently 3-nerved leaves. Hall and

Clements (1923) used nei"ve number to distin-
guish C. nauseosus ssp. graveolens from spp.
consimilis (so the character is variable even
within a species). Many of the latter group,
such as C. alhidus, C. greenei, and some forms
of C. viscidiflorus, appear to have 1-nerved
leaves. Actually, all species of Ericameria and
Chrysothamnus have trilacunar, 3-trace nodal
anatomy (personal obsei^vation); thus, the char-
acter of 1 versus 3 nerves is a matter of per-
ception, not of fact. (2) Leaf margins ciliate in
the former and never in the latter — but C.
alhidus, C. eremohius, and C. viscidiflorus
subsp. planifolius of the former have entire
leaf margins; also in that group, C. pulchellus
subsp. pulchellus has entire leaf margins,
whereas subsp. haileyi has ciliate leaf margins,
and some populations of C. gramineus and C.
vaseyi have entire leaf margins, but others do
not. Ericameria (sensu Nesom 1990) has sev-
eral species that have leaves with ciliate leaf
margins, fairly prominent in E. cooperi and
less so in several other species (e.g., E. cervina,
E. nana, E. ophitidis, and E. zionis). (3) Corollas
more or less abruptly broadened from the
tube into the throat with long, recurving or
coiling lobes in the former and corollas tubu-
lar with short, erect or spreading lobes in the
latter — but corollas of C. spathulatus (of the
latter) have relatively broad tubes that lack
noticeably flaring throats, C. humilis (of the
former) has tubular corollas with short, erect
lobes (Anderson 1964: 226), and C. nauseosus
ssp. ceruminosus (of the latter) has corollas
that are abruptly broadened from the tube
into the throat with long, spreading lobes. (4)
Style appendage collecting hairs merely papil-
late in the former, whereas they are long and
sweeping in the latter — but C. alhidus, C.
molestus, C. pulchellus, and certain popula-
tions of C. viscidiflorus (all of the former) have
style appendages with moderately long, sweep-
ing hairs. Diversity in collecting hairs is greater
in Chrysothamnus (sensu Anderson 1986) than
Nesom and Baird (1993) imply and does not
fall into two groups. Collecting hair length
may be correlated with other floral features;
namely, the corollas, style lengths, and pollen
volumes of the former group (Anderson 1966)
are generally smaller than those of the latter
(5) Involucral bracts in vertical files in the for-
mer (caveat noted) and usually not in vertical
files in the latter — but, perhaps the most
strongly aligned bracts occur in C. nauseosus

Great Basin Natur^vlist

[Volume 55

ssp. arenarius (of the latter). (6) Achenes glan-
dular with nonresinous nerves in the former
and eglandular (with duplex hairs) and
resinous nerves in the latter — but only five
species of the former have glandular achenes
(in some they are hidden by duplex hairs) and
the other seven do not, having either glabrous
achenes or achenes with duplex hairs exclu-
sively (Anderson 1970, 1983a), and many have
resin canals associated with the bundles of the
achenes, admittedly fewer than in those of the
latter but well developed in C. molestus of the
former. Also, C. paniculatus (of the latter group-
ing) lacks resin canals in its achenes (Anderson
1970). None of these six sets of characteristics
can be used to consistently separate the two
groups.

Clearly, Chrysothamnus (sensu Anderson
1986, not Nesom and Baird 1993) is fairly
homogeneous and should not be dismem-
bered. If some are to go into Ericameria (and
DNA data suggest they should), then all should
go into Ericameria. Therefore, the remaining
12 species of Chrysothamnus are transfened to
Ericameria, and new combinations are made
here.

1. Ericameria albida (M. E. Jones ex A. Gray)
L. C. Anders., comb. nov. Basionyni: Bigelovia albi-
da M. E. Jones ex A. Gray, Proc. Amer. Acad. Aits
17: 209. 1882. Chrysothamnus albidus (M. E. Jones
ex A. Gray) E. Greene, Eiythea 3: 107. 1895.

2. Ericameria depressa (Nutt.) L. C. Anders.,
comb. nov. Basionym: Chrysuthainnus deprcssus
Nutt., Proc. Acad. Nat. Sci. Philadelphia 4: 19.
1948. Linosyris depressa (Nutt.) Ton., in Stigreaves,
Kept. Exped. Zuni & Colorado Rivers 161. 1853.
Bigelovia depressa (Nutt.) A. Gray, Proc. Amer.
Acad. Arts 8: 643. 1873.

3. Ericameria eremohia (L. C. Anders.)
L. C. Anders., comb. nov. Basionym: Chrysothamnus
eremobius L. C. Anders., Brittonia 35: 2.3. 1983.

4. Ericameria graminea (H. M. Hall) L. C.
Anders., comb. nov. Basionym: Chrysothamnus
gramineus H. M. Hall, Muhlenbergia 2: 342. 1916.
Petradoria discoidea L. C. Anders., Trans. Kansas
Acad. Sci. 66: 676. 1964.

5. Ericameria filifoUa (Rydb.) L. C. Anders.
comb. nov. Basionym: Chrysothamnus filijolius
Rydb., Bull. Toney Bot. Club 28: 503. im\. Bigelovia
greenei A. Gray, Proc. Amer. Acad. Arts 11: 75. 1876
[not Ericameria greenei (A. Gray) Nesom].
Chrysothamnus greenei (A. Gray) E. Greene,
Erythea 3: 94. 1895. Chrysothamnus pumilus var
acuminatus A. Nels., Bot. Gaz. 28: 376. 1899.
Chrysothamnus scoparius Rydb., Bull. Torrey Bot.
Club 28: 504. 1901. Chrysothamnus laricinus E.
Greene, PittoniaS: 110. 1903.

6. Ericameria humilis (E. Greene) L. C. Anders.,
comb. nov. Basionym: Chrysotlunnnus humilis E.
Greene, Pittonia 3: 24. 1896.

7. Ericameria linifolia (E. Greene) L. C. Anders.,
comb. nov. Basionym: Chrysothamnus linifolius E.
Greene, Pittonia 3: 24. 1896.

8. Ericameria molesta (Blake) L. C. Anders.,
comb. nov. Basionym: Chrysoihamnus viscidiflorus
var. molestus Blake, J. Wash. Acad. Sci. 30: 368.
1940. Chrysothamnus molestus (Blake) L. C.
Anders., Madroiio 17: 222. 1964.

9a. Ericameria pulchella (Gray) L. C. Anders.,
comb. nov. Basionym: Linosyris pulchella A. Gray,
Pi. Wright. [Smidis. Contr. Know!.] 3(5): 96. 1856.
Bigelovia pulchella (A. Gray) A. Gray, Proc. Amer
Acad. Arts 8: 643. 1873. Chrysothamnus pulchellus
(A. Gray) E. Greene. Enthea 3: 107. 1895.

9b. Ericameria pulchella subsp. baileyi (Woot.
& Standi.) L. C. Anders., comb. nov. Basionym:
Chrysothamnus baileyi Woot. & Standi., Contr. U.S.
Nati. Herb. 18: 181. 1913.

9c. Ericameria pidchella subsp. pulchella var.
elatior (Standi.) L. C. Anders., comb. nov. Basionym:
Chrysothamnus elatior Standi., Proc. Biol. Soc. Wash.
26: 118. 1913. This variety with uniformly pubes-
cent leaves occurs sporadically in a few populations
of the typically glabrous-leaved subspecies pulchel-
lus and does not warrant a higher taxonomic status
than this quadrinomial affords.

10. Ericameria spathulata (L. C. Anders.) L. C.
Anders., comb. nov. Basionym: Chrysothamnus
spathulatus L. C. Anders., Madroiio 17: 226. 1964.
Chrysothamnus viscidiflorus var. ludens Shinners,
Sida 1: 374. 1964.

11. Ericameria vaseyi (A. Gray) L. C. Anders.,
comb. nov. Basionym: Bigelovia vaseyi A. Gray,
Proc. Amer. Acad. Arts 12: 58. 1876. Chrysothamnus
vaseyi (A. Gray) E. Greene, Erythea 3: 96. 1895.
Chrysothamnus bakeri E. Greene, Pittonia 4: 152.
1900.

12a. Ericameria viscidiflora (Hook.) L. C.
Anders, comb. nov. Basionym: Crinitaria viscidiflo-
ra Hook., Fl. Bor. Am. 2: 24. 1834. Chrysothanmus
viscidiflorus (Hook.) Nutt. Trans. Amen Philos. Soc.
11, 7: 324. 1840. Bigelovia douglasii A. Gra\', Proc.
Amer Acad. Arts 8: 645. 1873. Chrysotha)nnus dou-
glasii (A. Gray) Clements & Clements, Rocky Mtn.
Els. 226. 1914. Chrysothamnus pumilus Nutt.,
Trans. Amer Philos. Soc. II, 7: 323. 1840. Linosyris
serrulata Torr, Stansbuiy Rep. 1: 389. 1851. Chryso-
thamnus serrulatus (Torn) Rydb., Bull. Tonxy Bot.
Club 33: 152. 1906. Chrysothanmus tortifolius E.
Greene, Fl. Fran. 368. 1897. Chrysodianmus Icuco-
cladus E. Greene, Pittonia 5: 59. 1902. Chrysodiam-
nus stenolepis Rydb., Bull. Torrey Bot. Club 37:
131. 1910.

12b. Ericameria viscidiflora subsp. viscidiflora
var. latifolia (D. C. Eaton) L. C. Anders., comb,
nov. Basionxni: Linosyris viscidiflora van latifolia

1995]

Chrysothamnus-Ericameria Connection

D. C. Eaton, Bot. King Expl. 157. 1871. Chnjsotluiin-
mts latifolins (D. C. Eaton) Rydh., Bull. Toirey Bot.
Club 33: 152. 1906.

12c. Ericameria viscidiflora subsp. viscidiflora
van stenophylla (A. Gray) L. C. Anders., comb. nov.
Basionym: Bigelovio doiiglasii var. stenophylla A.
Gray, Proc. Amer Acad. Arts 8: 646. 1873. Chryso-
thamnus stenophyllus (A. Gray) E. Greene, Eiythea
3: 94. 1895. These quadrinomials (12b and 12c)
identify sporadic but rather distinctive morpho-
types that occur in the northern regions of this sub-
species (for conceptual distinction between sub-
species and variety, see Anderson 1980b)

12d. Ericameria viscidiflora subsp. axillaris
(Keck) L. C. Anders., comb. nov. Basionym: Chryso-
thainnii.s (ixillaris Keck, Aliso 4: 104. 1958.

12e. Ericameria viscidiflora subsp. lanceolata
(Nutt.) L. C. Anders., comb. nov. Basionym: Chryso-
thamnus lanceolatus Nutt., Trans. Amer Philos. Soc.
II, 7: 324. 1840. Chrysothainiuts elegons E. Greene,
Erythea 3: 94. 1895. Bigelovia doiiglasii var spathii-
lata Jones, Proc. Calif Acad. Sci. II 5: 690. 1895.
Chrysothamnus glaticus A. Nels., Bull. Torrey Bot.
Club 25: 377. 1898. Chrysothamnus pumilus var
latus A. Nels., Bot. Gaz. 54: 413. 1912.

12f. Ericameria viscidiflora subsp. planifolia (L.
C. Anders.) L. C. Anders., comb. nov. Basionym:
Chrysothamnus viscidiflorus subsp. planifolius L. C.
Anders., Madrono 17: 223. 1964.

12g. Ericameria viscidiflora subsp. puberula
(D. C. Eaton) L. C. Anders., comb. nov. Basionym:
Linosyris viscidiflora var puberula D. C. Eaton, Bot.
King Expl. 158. 1871. Chrysothamnus puherulus
(D. C. Eaton) E. Greene, Erythea 3: 93. 1895.
Chrysothamnus marianus Rydb., Bull. Torrey Bot.
Club 37: 131. 1910.

The following sections in Ericameria are
proposed to accommodate these species trans-
fers: Ericameria section Chrysothamnus (A.
Gray) L. C. Anders., comh. nov. Basionym: Bige-
lovia section Chrysothamnus A. Gray., Proc.
Amer. Acad. Arts 8: 641. 1873. This section in-
cludes E. albida, E. filifolia, E. humilis, E. lini-
folia, E. spathulata, and E. viscidiflora. Ericam-
eria section Gramini (L. C. Anders.) L. G.
Anders., comb. nov. Basionym: Chrysothamnus
section Gramini L. G. Anders., Proc. Symp.
Biology o{ Artemisia and Chrysothamnus 29.
1986. This section includes E. eremobia and E.
graminea. Ericameria section Pulchelli (Hall &
Clements) L. C. Anders., comb. nov. Basionym:
Chrysothamnus section Pulchelli Hall &
Clements, Carnegie Inst. Publ. 326: 175, 193.
1923. This section includes E. depressa, E.
molesta, E. pulchella, and E. vaseyi.

Additionally, there is a fossil species, Eri-
cameria pulchelloides (L. G. Anders.) L. G.

Anders., comh. nov. Basionym: Chrysothamnus
pulchelloides L. G. Anders., Great Basin
Naturahst 40: 351. 1980.

Nesom and Baird (1993) suggest the Chnj.so-
thamnus taxa that I have just transferred to
Ericameria should be placed in a restructured
genus to include elements of Hesperodoria,
Petradoria, and Vanclevea. They conclude that
chloroplast DNA data (Suh 1989) show Petra-
doria to be integrally related to the Solidago
lineage and far removed from Ericameria.
However, they note that neither Suh (1989)
nor Morgan and Simpson (1992) sampled any
taxa o( Chrysothamnus sensu Nesom and Baird.
These taxa need DNA profiles determined
because they certainly do not make a morpho-
logically compatible grouping with Petradoria
or Vanclevea. For example, Petradoria (Anderson
1963) has radiate heads with disk flowers that
lack stigmatic areas on the style branches and
have abortive ovaries, and Vanclevea (Anderson
and Weberg 1974) has large turbinate heads
with many phyllaries, many flowers, and a
tardily deciduous pappus of paleacous awns —
none of these conditions are found in Chryso-
thamnus sensu Nesom and Baird. The cohe-
siveness of Chrysothamnus sensu Anderson is
further illustrated in that C. spathulatus twigs
emit odor similar to that of C. nauseosus
(Anderson 1964: 227).

Two alternate taxonomies are now available:
one for Chrysothamnus as a genus (Anderson
1986) or as a component of Ericameria
(Nesom and Baird 1993, and here); both are
preferable to merging some elements of
Chrysothamnus with Petradoria or Vanclevea.

Acknowledgments

James Reveal and Arnold Tiehm offered
constructive comments on the manuscript.

Literature Cited

Anderson, L. C. 1963. Studies on Petradoria (Compos-
itae): anatomy, cytology, and ta.xonomy. Transactions
of the Kansas Academy of Science 66: 632-684.

. 1964. TiL\onomic notes on the Chrysothamnus vis-
cidiflorus complex (Astereae, Compositae). Madroiio
17; 222-227.

. 1966. Cytota.xonomic stndies in Chrysothamnus

(Astereae, Compositae). American Journal Botany 53:
204-211.

. 1970. Floral anatomy of Chrysothamnus (Astereae,

Compositae). Sida 3: 466-503.

. 1973. Unique Chrysothamnus hybridizations in Ash

Meadows, Nevada. Bulletin of the ToiTe\^ Botanical
Club 100: 171-177.

Great Basin Naturalist

[Volume 55

. 1980a. ll(i})lopa})pus iil}Hmis (Asteraceae): a new

species fVorii Nevada. Great Basin Naturalist 40;
73-77.

. 19S0h. Identity of narrow-leaved ahnjsofhaiiiiius

viscidifloni.s (Asteraceae). Great Basin Naturalist 40:
117-120.

. 1983a. CJinjsothammt.s ereinobius (Asteraceae): a

new species troiu Nevada. Brittonia 35: 23-27.

. 19831). Haplopappus crispii.s and //. zionk (Aster-
aceae): new species from Utali. (Ireat Basin Naturalist
43: 358-364.

1986. An overview of the genus Chnjsotliatnnua

(Asteraceae). Pages 29-45 iii E. D. McArtliur and B. L.
Welsh, eds.. Proceedings, Symposium on the Biol-
ogy of Artemma and Chnjsothamniis. USDA, Forest
Service, Intermountain Research Station, Ogden,
UT 398 pp.

Anderson, L. C., .and J. L. Reveal. 1966. Chnjsothainnus
bolanderi, an intergeneric hybrid. Madrono 18:
225-233.

Anderson, L. C., and R S. Weberg. 1974. The anatomy
and taxonomy of Vanclevea (Asteraceae). Great Basin
Naturalist 34: 151-160.

Clark, W. D. 1977. Chemosystematics of the genus
Hazardia (Compositae). Journal of the Arizona
Academy of Science 12: 16.

Greene, E. L. 1894. Observations on the Compositae. IV.
Erythea 2: 53-60.

Hall, H. M., and E E. Clements. 1923. The phylogenet-
ic method in tiLxonomy: the North American species
of Artemisia, Chrijsothammis, and Atriplex. Carnegie
Institute Publication 326: 1-355.

Johnston, M. C. 1970. Compositae. Pages 152.3-1744 in
D. S. Correll and M. C. Johnston, Manual of the vas-
cular plants of Texas. Texas Research Foundation,
Renner, TX. 1881 pp.

.Morgan, D. R., and B. B. Slmpson. 1992. A systematic
study of Machaeranthera (Asteracaee) and related
groups using restriction site analysis of chloroplast
DNA. Systematic Botany 17: 511-531.

Nesom, G. L. 1990. Nomenclatural summary of
Ericaineria (Asteraceae: Astereae) with the inclusion
of Haplopappus sects. Asiris and Macronema. Ph\ tol-
ogia 68: 144-155.

Nesom, G. L., and G. I Baird. 1993. Completion of
Ericaineria (Asteraceae: Astereae), diminution of
Chnjsothainnus. Phytologia 75: 74-93.

Shinners, L. II. 1950. Notes on Texas Compositae, IV, V.
Field and Laboratory 18: 25^2.

SuH, Y. 1989. Phylogenetic studies of North American
Astereae (Asteraceae) based on chloroplast DNA.
Unpublished doctoral dissertation. University of
Texas, Austin.

Turner, B. L., and S. Sanderson. 1971. Natural hybridi-
zation between the Composite "genera" Machaeran-
thera and Haplopappus (sec. Blepharadon). American
Journal of Botany 58: 467.

Urbatsch, L. E. 1978. The Chihuahuan Desert species of
Ericameria (Compositae, Astereae). Sida 7: 298-303.

Received 7 Felniranj 1994
Accepted 2 June 1994

REPRODUCTIVE BEHAVIOR IN MERRIAM'S CHIPMUNK
{TAMIAS MERRIAMI)

Stephen B. Compton^ and J. R. Callahan^
Key words: Tamias, Eutamias, Neotamias, chipmunk, copulation, olfaction.

The literature contains little information
regarding mating chases and copulation in any
of the western chipmunks {Tamias, subgenus
Neotamias). Callahan (1981) reported mating
chases for Merriam's {T. merriami) and dusky
chipmunks {T. ohscurus), but noted (unpub-
lished) that both copulating pairs were partly
concealed by fohage. Larson (1981) described
two copulations for Merriam's chipmunk, but
a careful reading suggests that one of these
was a mounting attempt by an immature male
and the other was observed from a consider-
able distance. Best and Granai (1994) found no
references on this subject other than Callahan
(1981) and Larson (1981).

There has been some speculation and dif-
ference of opinion regarding reproductive iso-
lating mechanisms in parapatric species of
western chipmunks. Blankenship and Brand
(1987) reported differences in vocal behavior
between Tamias merriami and T. ohscunis at
Black Mountain (Riverside County, CA) and
noted a possible role in reproductive isolation.
One of us (JRC), however, had previously con-
ducted a more extensive study of vocal behavior
in these two cryptic species at Black Mountain
from 1975 to 1980. Vocalizing individuals were
collected to confirm species identity, and sono-
grams were prepared and measured; yet no
statistically significant vocal differences were
found (Callahan 1981, and in preparation).
Ecological, olfactory, and mechanical barriers
to hybridization also have been suggested
(Callahan 1977, 1981, Patterson 1984). These
hypotheses cannot be tested without more
data on chipmunk reproductive behavior.
Accordingly, this note provides the first
detailed description of western chipmunk cop-
ulation that has been published, to the best of
our knowledge. Comparative data for other
western chipmunk species would be of interest.

The obsei-vation was made 1 April 1994 in a
wooded residential area in Idyllwild, Riverside
County, CA (elevation 1590 m), between 1000
and 1130 h. The habitat is mixed-conifer forest
dominated by incense cedar {Calocedrus
decurrens), yellow pine {Pimis ponderosa), live
oak {Quercus chrysolepis), and black oak {Q.
kelloggii), with a sparse understory of chapar-
ral shrubs. When the observer arrived at 1000
h, six to seven Merriam's chipmunks, many of
them males, were running over, around, and
through a large woodpile while performing
conspicuous leaping maneuvers. No agonistic
interaction was observed. It was not possible at
this stage to identify the female(s) or to tell in
which direction the "chase " was headed. The
overall effect was somewhere between a
Sciurus-hke mating chase (e.g., Thompson
1977), in which several males follow one
female, and a lek, involving male display. The
chase covered an area 13-15 m in diameter
but centered on the woodpile and a nearby
heap of smaller pine branches.

After about 20 min, one chipmunk (later
identified as female) ran up on one of the piled
branches. A second chipmunk approached and
they ran around for a few seconds. The female
stopped on a branch and the other chipmunk,
a male, ran up beside her. His entire right side
was in contact with her left side for about 1 sec,
during which he made a nuzzling motion with
the right side of his face on the rear left portion
of her face. The expected nasal/genital contact
was not observed, but the pair had been out of
sight for a short time previously and this could
have occurred. The female then jumped to
another branch, which was 5 cm in diameter
and 20 cm above the ground, sloping at a 25°
angle so that the female was facing downhill.
Copulation then occurred only 2 m from the
observer (who was inside a parked vehicle).

'Send reprint requests to Box 3140, Hemet, CA 92546.

^Museum of Southwestern Biology, University of New Mexico, Albuquerque, NM 871.31.

Great Basin Natufl\list

[Volume 55

The female crouched on the branch as if
resting. The male jumped to that branch and
quickly grasped the female from the rear with
his forelimbs around her pectoral region. The
female's tail was deflected to the side and
slightly raised, and the male's tail was extend-
ed to the rear. Copulation consisted of four
series of pelvic thrusts. Each series (except the
last) lasted about 4 sec and comprised an esti-
mated 12-24 thrusts, at a rate of 3-6 per sec.
Each series of rapid thrusts was followed by a
short resting period, during which the male
stopped thrusting and brushed his face
(mouth, nose, and chin) from side to side 2-4
times against the back of the female's neck.
The fourth and last series of thrusts was short-
er than the first three. The male then released
his grip on the female, dismounted, and ran
off into the woodpile. The female, who had
remained motionless during the act, remained
on the branch about 1 sec and then also ran to
the woodpile. The entire copulation lasted
about 18 sec.

Although several male chipmunks partici-
pated in the chase, none of them approached
the copulating pair. No chipmunks were heard
vocalizing during the mating chase or copula-
tion. We did not note any pre-mating vocal dis-
play or Lockrufe by the estrous female
(Callahan 1981), but we were not present on
the days when the display (if any) would have
taken place. The Tamias vocal display has
been reported for a few species of chipmunks
by Callahan (1981), Blake (1992), and others.
It is not clear whether this vocal display is
universal or occurs only at low population
densities, when the female benefits by attract-
ing more distant males.

No further copulations were seen, but as
many as seven male chipmunks continued to
mn around the same woodpile for another hour.
The level of activity appeared to decrease, and
there were none of the prodigious leaps seen
earlier. The group then gradually dispersed as
individuals headed for an adjacent area where
other chipmunks were heard giving occasional
"chipper" vocalizations (not the long series of
chips that characterizes the Lockrufe).

The behavior described above suggests that
scent glands play a key role in reproductive be-
havior of this species. Larson (1981) and others
have noted that male chipmunks have scent
glands near the chin and angle of the jaw (oral
glands) that become enlarged during the

breeding season. Scent marking is prevalent in
sciurids, but usually this means marking the
ground or a branch, not marking another ani-
mal. The "nuzzling" and "brushing" behavior
of the male Merriam's chipmunk, before and
during copulation, suggests that he was scent
marking the female.

Conspecific marking has been described for
various mammals, such as rabbits (Mykytowycz
1965), but not for sciurids. Gurnell (1987)
describes "face-wiping" behavior by various
tree squirrels, but only in the context of sub-
strate marking and (in Paraxenis) self-groom-
ing; his description of copulation in Sciiirus
and Tamiasciurus says nothing about the male
marking the female. With reference to olfacto-
ly communication in ground squirrels, Halpin
(1984) wrote that "there is no experimental
evidence that conspecific marking . . . actually
occur[s] among the sciurids."

Our obsei-vation indicates that conspecific
marking does occur in Merriam's chipmunk as
a component of reproductive behavior Without
experimental data, it is not possible to deter-
mine the significance of this marking. Pair
bonding comes to mind, but there is no good
evidence of long-term pair bonding in
Merriam's or any other species of western
chipmunk, despite many years of field obser-
vation. Other possibilities include the follow-
ing: (1) the marking induces some required
physiological state in the female; (2) the mark-
ing tells other males that the female has
already mated (before the copulation plug
forms and the message becomes redundant);
or (3) the marking reinforces a short-term pair
bond to ensure that subsequent copulations (if
any) on the day of estrus will be with the same
male. Larson (1981) indicated that the same
estrous female sometimes copulates more than
once.

Mortality from all causes is higher for male
than for female chipmunks (Smith 1978), per-
haps due in part to the dispersal and exposure
associated with the breeding season (Callahan
1981). After incurring the risk of predation
and expending considerable energy on the
mating chase, it should be to the male's advan-
tage to ensure that his genes are passed to all
the female's offspring of the season.

Literature Cited

Best, T. L., and N. J. Granai. 1994. Tamias meniami.
Mammalian Species 476: 1-9.

1995]

Notes

Blake, B. H. 1992. Estrous calls in captive Asian chip-
munks, Tamias sibiriciis. Journal of Mannnalosy 73:
597-603.

Blankenship, D. J., AND L. R. Brand. 1987. Geographic
variation in vocalizations of California chipmunks
Tamias obscunis and T. merriami. Bulletin of the
Southern California Academy of Sciences 86:
126-135.

Callahan, J. R. 1977. Diagnosis of Eutainias ohscurus
(Rodentia: Sciuridae). Journal of Mammalogy 58:
188-201.

Callahan, J. R. 1981. Vocal solicitation and parental
investment in female Eutainias. American Naturalist
118: 872-875.

Cornell, J. 1987. The natural histoiy of squirrels. Ricts
on File, New York.

Halpin, Z. T. 1984. The role of olfactoiy communication
in the social systems of ground-dwelling sciurids.
Pages 201-225 in J. O. Murie and C. R. Michener,
eds.. The biology of ground-dwelling squirrels.
University of Nebraska Press, Lincoln.

Larson, E. A. 1981. Merriam's chipmunk on Palo Escrito
in the Santa Lucia Mountains of California. Part L
Regimen with recorded episodes of natiualistic be-
havior. Enid A. LarsonAVacoba Press, Big Pine, CA.

Myk'VTOWTCZ, R. 1965. Fiuther obsei-vations on the terri-
torial function and histology of the submandibular
cutaneous (chin) glands in the rabbit, Orijctolagus
cimiculiis (L). Animal Behaviour 13: 400-412.

R'VTTERSON, B. D. 1984. Geographic variation and ttixonomy
of Colorado and Hopi chipnumks (genus Eutainias).
Journal of Mammalogy 65: 442^56.

Smith, S. E 1978. Alarm calls, their origin and use in
Eutainias sonomae. Journal of Mammalogy 59:
888-893.

Thompson, D. C. 1977. Reproductive behavior of the grey
squinel. Canadian Journal of Zoology 55: 1176-1184.

Received 11 April 1994
Accepted 1 9 October 1 994

ADDITIONAL RECORDS OF FLEAS (SIPHONAPTERA) FROM UTAH

James R. Kucera'

Key words: Si^hunaptcrci. Jleas, Utah. Mcgabothris asio megacolpus, Euhoplops>lliis glacialis Knx.

Subsequent to the important work of Stark
(1959), few publications have given flea collec-
tion records from Utah. These include Jellison
and Senger (1976) and Kucera and Haas (1992);
but most effort in this area has been that of
Egoscue (1966, 1976, 1977, 1988, 1989).

Herein is presented information for 10
species of Siphonaptera for Utah. A number of
important records were obtained from the flea
collection at the Monte L. Bean (MLB) Life
Science Museum, Brigham Yoinig University,
Provo, UT. Catalog numbers of host specimens
deposited in the University of Utah Museum of
Natural Histoiy (UU) mammal collection and
flea specimens in the MLB Museum (BYU)
insect collection are given in parentheses when
available. Unless indicated otherwise, speci-
mens were collected by me and are retained in
my personal collection.

Carter ett a chivata Good 1942

Washington Co.: west slope Beaver Dam
Mts., vie. Welcome Spring, 1220 m, 20 March
1988, 1 9 ex Chaetodipus formosus. "Beaver
Dam," 23 Februaiy 1952, 1 9 (BYU #3462) ex
Perognathus [ = Chaetodipiis] formosus, coll.
C. L. Hayward. Beaver Dam Wash, 17 April
1952, Ic? (BYU #3607) ex Peromysciis truei,
coll. Grace Grant et al.

Few collections of this species are known
from Utah (Tooele County: Stark 1959 [Id],
Egoscue 1976 [1 specimen, sex unknown];
Washington County: Jellison and Senger 1976
[2d (5, 2 9 9]). It has also been collected in
Clark County, NV (the type locality, Good 1942),
and Mohave County, AZ (Augustson and Dur-
ham 1961). It is likely a nest flea o{C. formosus.

Nearctopsylla brooksi (Rothschild 1904)

Utah Co.: Provo, 21 August 1951; M, 19
(BYU #1366 & 1365) e\ Mustek frenata, coll.
D. Brown. Provo Canyon, 16 June 1959; 26 6 ,
79 9 ex Spilogale gracilis, coll. D E. Beck.

This species was previously known in Utah
from a single collection in Sevier County
(Stark 1959). It is usually found on weasels
{Mustela spp.).

NearctopsijIIa hi/rtaci
(Rothschild 1904)

Salt Lake Co.: Wasatch Mts., Big Cotton-
wood Canyon, vie. Redman campground, 2560
m (spruce-fir), 21 October 1990, 19 ex Sorex
monticolus (UU #29163). Same locality, 28
October 1990, 19 ex Sorex monticolus (UU
#29164).

Stark (1959) reported this species from
Cache County. My collections extend the
known range of this species further south in
Utah along the Wasatch Cordillera. It is foimd
on shrews [Sorex spp.) and Mustela spp.

Delotelis telegoni
(Rothschild 1905)

Salt Lake Co.: Wasatch Mts., Big Cotton-
wood Canyon, vie. Redman campgroimd, 2560
m (spruce-fir), 21 October 1990, Id ex
Clethrionomys gapperi. Same date & locality,
1 9 ex Tamiasciurus hudsonicus nest. [The
nest, about 3 m above ground level, also con-
tained many red squirrel fleas {Orchopeas c.
caedens). The squirrel probably carried this
vole flea to its nest.] Same locality, 15
September 1991, 19 ex Peromyscus manicula-
tus. Same locality, 5 October 1991, Id ex
Cleth rionomys gapperi.

'Associated Regional and Hinversih' Palliologists, liie . Salt Uke City, UT 84108. Address for correspondence: .59.30 S. Siillan Circle, Murra\'. UT 84107-H930.

1995]

Notes

Delotelis telegoni has rarely been found in
Utah; single specimens have been collected in
Sanpete County (Stark 1959) and in Utah
County (Egoscue 1988). It is significant that
these collections were made in summer months
(August 1951 and July 1985, respectively).
Many more collections will likely be made if
this species is searched for during the cooler
months of the year Also, it presumably would
be profitable to search nests of Microtus and
Clethrionomys for this species.

Meringis shannoni (Jordan 1929)

Eads et al. (1987) listed two collections of
this species from Utah. The specimens are
present in the BYU collection. However, the
records are erroneous because the collection
locality (Douglas County) does not exist in
Utah. All other cited records of this species
are from the states of Washington and Oregon,
except a lone locality record in Humboldt
County, NV (Lewis et al. 1988).

Stenistomera hubbardi
Egoscue 1968

This rare species was listed by Tipton and
Saunders (1971) as occurring in Utah although
no specific records were cited. Egoscue (per-
sonal communication) knows of no records from
Utah, and no specimens were present in the
MLB Life Science Museum, the main reposi-
tory of Tipton's Utah collections. In addition to
the type specimens from Oregon (Egoscue
1968), the only other published record of S.
hubbardi is that of Lewis et al. (1988), also
from Oregon. It is unlikely that the species
has been collected in Utah.

Megarthroglosstis becki
Tipton & Allred 1951

Salt Lake Co.: Wasatch Mts., mouth of
Little Cottonwood Canyon, 1676 m (scrub
oak), 3 December 1989; 1 c? , 1 9 ex Neotoma
cinerea nest.

The species has been collected only in
Utah (Kane, Piute, Utah, and Wayne counties;
Tipton et al. 1979) and Arizona (Augustson and
Durham 1961). This is the northernmost record
known, some 37 km north of the type locality
in the Wasatch Mountains. Megarthroglossus
becki is a nest flea of woodrats, principally the
bushy-tailed woodrat Neotoma cinerea.

Megabothris asio megacolpus
(Jordan 1929)

Rich Co.: Laketown, 22 August 1952;
nSS, 26$ $ (BYU #5097-5099, 5101-5103,
5105-5119, 5121-5131, 5135, 5137, 5138,
5737, 5738) ex Microtus montanus nests [3 ex-
amined], coll. D E. Beck & L. Beck. Laketown,
26 June 1953; Id, 49 9 (BYU #7823-7827) ex
Microtus nests [3 examined], coll. Beck et al.
Sevier Co.: Fish Lake [south end], 5 August
1952, 19 (BYU #5622) ex Microtus sp., coll.
Coffey & Killpack. Monroe Mt., 7 mi. [= 11
km] W of Koosharem, 30 July 1958; M, 29 9
ex Microtus sp. [3 examined], coll. unknown.

The range of this boreal vole flea extends
deep into south central Utah. Only two speci-
mens are known from Idaho, including one
from Bear Lake County adjacent to Rich
County (Baird and Saunders 1992). Collections
from Ravalli and Beaverhead counties, MT,
are documented (Holland 1950), and speci-
mens from Ravalli County are present in the
Natural History Museum, London (T. M.
Howard personal communication). A point-
mapped record roughly on the Utah-Wyoming
border given in Haddow et al. (1983) is evi-
dently meant to be Laketown because four
specimens with the same collection data as the
22 August 1952 series are present in the
Natural History Museum, London (Howard
personal communication).

Chaetopsylla stewarti Johnson 1955

Utah Co.: Wasatch Mts., "near summit of
Alpine Loop" [American Fork or Provo Can-
yons], 24 November 1965; 666,699 ex
Mustela frenata, coll. D. Andrews. Summit Co.:
Uinta Mts., 1/2 mi. [0.8 km] E Bald Mt., 8
August 1957, 16 ex Martes sp., coll. D. Allred
& M. Killpack.

These specimens are the only ones known,
other than the type series (from Cache County,
Lewis and Lewis 1994). Weasels seem to be
the preferred host.

Euhoplopsyllus glacialis hjnx
(Baker 1904)

Salt Lake Co.: Wasatch Mts., Big Cotton-
wood Canyon, vie. Redman Campground, 2560
m, 17 August 1988; 39 9, 16 ex Lepus ameri-
canus (UU #28674). Big Cottonwood Canyon,
2280 m, 30 August 1988, 1 9 ex Lepus ameri-
canus. Big Cottonwood Canyon, vie. Butler

Great Basin Natur.\list

[Volume 55

Fork trailhead, 2182 m, 19 May 1991, 26 6 ex
Lepus atnericanus.

Prc'Niously unknown in Utah, the nearest
published records are for Ravalli County, MT
(Kohls 1940), more than 570 km to the north.
The t\pe lociilitv' is Moscow, ID, about 790 km
to the northwest (Baker 1904). This flea is con-
sistently found on the snowshoe hare [Lepus
americanus) and its predator, the hnx {Lynx
canadensis). The foim E. glacialis affinis is com-
mon in Utah and surrounding states on rabbits
and jackrabbits (Sylcilagiis spp. and Lepus
spp. other than L. americanus). Questionable
records of E. g. lynx from the states of
Tamaulipas and Veracruz, Mexico (ex Sylvilagus
floridanus and unidentified Sylvilagus sp.), are
listed by Ayala et al. (1988).

Acknowledgments

Comments by Glenn E. Haas, B. C. Kondra-
tieff, and an anonymous reviewer improved
the manuscript. Flichard W. Baumann, Curator
of Insects at the M. L. Bean Life Science
Museum, Brigham Young University, kindly
allowed me to examine specimens kept there.
Harold J. Egoscue confirmed identification of
the E. g. lynx. Theresa M. Howard of The
Natural History Museum, London, sent data
on specimens in the Rothschild Collection.

Literature Cited

AuGUST.soN, G. E, AND E E. Durham. 1961. Records of
flea.s (Siphonaptera) from northwestern Arizona.
Southern CaHfornia Academy of Sciences Bulletin
60: 100-105.

Ayala, R., J. C. Morales, N. Wilson, J. E. Llorente, and
H. E. Ponce. 1988. Catalogo de las pulgas (Insecta;
Siphonaptera) en el Museo de Zoologia, Ricultad de
Ciencias Universidad Nacional Autonoma de Me.xico
1: Coleccion Alfredo Barrera. Serie Catalogos del
Museo de Zoologia "Alfonso L. Herrera" Catalogo
No. 1. 102 pp.

Baird, C. R., and R. C. Saunders. 1992. An annotated
checklist of the fleas of Idaho (Siphonaptera). Idaho
Agricultural E.xperiment Station, Bulletin 148.

Baker, C. E 1904. A revision of the American Siphonap-
tera, or fleas, together with a complete list and bibli-
ograph>' of tlie group. Proceedings of die U.S. National
Museum 27: 365-469.

Eads, R. B., E. G. Campos, and G. O. Maupin. 1987. A
review of the genus Meringis (Siphonaptera: Hystri-
chopsyllidae). Jouj^nal of Medical Entomology 24:
467-476.

EgoscuE, H. J. 1966. New and additional host-flea associ-
ations and distributional records of fleas from Utah.
Great Basin Naturalist 26: 71-75.

. 1968. A new species of the genus Stenistomcra

(Siphona])tera: Ilystrichopsyllidae). Southern Cali-
fornia Academy of Sciences Biflletin 67: 1.38-142.

. 1976. }""lea exchange between deer mice and some

associated small mannnals in western Utah. Great
Basin Naturalist 36: 475-480.

. 1977. The sagebrush vole flea, Megahothri.s cliin-

toni princeu in western Utah, with comments on the
distribution o{ Megabothris in the Bonneville Basin.
Great Basin Naturalist 37: 75-76.
. 1988. Noteworthv flea records from Utah, Nevada,

and Oregon. Great Basin Naturalist 48: 530-.532.

. 1989. A new species of the genus Traiibella (Sipho-
naptera: Ceratophyllidae). Southern California
Academy of Sciences Bulletin 88: 131-134.

Good, N. E. 1942. Carteretta carteri clavata, a new sub-
species from Nevada, and notes on s>Tionymy (Sipho-
naptera). Annals of the Entomological Society of
America .35: 110-113.

Haddovv, J., R. Traub, and M. Rothschild. 1983. Dis-
tribution of ceratophyllid fleas and notes on their
hosts. Pages 42-163 in R. Traub, M. Rothschild, and
J. E Haddow, The Rothschild collection of fleas — the
Ceratophyllidae: keys to the genera and host relation-
ships with notes on their evolution, zoogeography
and medical importance. 288 pp. [Privately published.]

Holland, G. P. 1950. Notes on Megabothris asio (Baker)
and M. calcurifer (Wagner) with the description of a
new subspecies (Siphonaptera: Ceratophyllidae).
Canadian Entomologist 82: 126-133.

Jellison, W. L., and C. M. Senger. 1976. Fleas of west-
ern North America e.xcept Montana in the Rocky
Mountain Laboratory collection. Pages 55-136 in
H. C. Taylor, Jr., and J. Clark, eds.. Papers in honor
of Jerry Flora. Western Washington State College,
Bellingham.

Kucera, J. R., and G. E. Ha.\s. 1992. Siphonaptera (fleas)
collected from small mammals in montane southern
Utah. Great Basin Naturalist 52: 382-384.

Kohls, G. M. 1940. Siphonaptera — a study of the species
infesting wild hares and rabbits of North America
north of Mexico. National Institute of Health Bulletin
175.

Lewis, R. E., and J. H. Lewis. 1994. Siphonaptera of
North America north of Mexico: Vermips>'llidae and
Rhopalopsyllidae. Journal of Medical Entomology
31: 82-98. '

Lewis, R. E., J. H. Lewis, and C. Maser. 1988. The fleas of
the Pacific Northwest. Oregon State Universit>' Press,
Coi^vallis. 296 pp.

Stark, H. E. 1959. The Siphonaptera of Utah. U.S.
Department of Health, Education and Welfare,
Communicable Disease Center, Atlanta, GA. 239 pp.

Tipton, V J., and R. C. Saunders. 1971. A list of arthro-
pods of medical importance which occur in Utah
with a review of arthropod-bome diseases endemic in
the state. Brigham Young University' Science Bulletin,
Biological Series 15: 1-31.

Tipton, V. J., H. E. Stark, and J. A. Wildie. 1979.
Anomiopsyllinae (Siphonaptera: Hystrichops> llidae),
n. The genera CullistopsijUus, ConorhinopsyUa,
Mcgarthroglossiis, and Sti'ni.sto)nera. Great Basin
Naturalist 39: 351-418.

Received 25 May 1994
Accepted 10 August 1994

INFORMATION FOR AUTHORS

The Great Basin Naturalist welcomes previously
unpublished niauuscripts pertaining to the biologi-
cal natiual histor\ of western North America. Pref-
erence will be given to concise manuscripts of up to
12,000 words. Simple species lists are discouraged.

SUBMIT MANUSCRIPTS to Richard W. Baumann,
Editor, Great Basin Naturalist, 290 MLBM, PO Box
20200, Brighani Young University, Provo, UT
84602-0200. A cover letter accompanying the man-
uscript must include phone number(s) of the author
submitting the manuscript, and FAX number and
E-mail address when applicable; the letter must
also provide information describing the extent to
which data, text, or illustrations have been used in
other papers or books that are published, in press,
submitted, or soon to be submitted elsewhere.
Authors should adhere to the following guidelines;
manuscripts not so prepared may be returned for
revision.

MANUSCRIPT PREPARATION. Consult Vol. 51, No.
2 of this journal for specific instructions on style
and format. These instructions. Guidelines for
Manuscripts Submitted to the Great Basin
Naturalist, supply greater detail than those pre-
sented here in condensed form. The guidelines are
printed at the back of the issue; additional copies
are available from the editor upon request. Also,
check the most recent issue of the Great Basin Nat-
uralist for changes and refer to the CBE Style Man-
ual, 5th edition (Council of Biology Editors, One
Illinois Center, Suite 200, 111 East Wacker Drive,
Chicago, IL 60601-4298 USA, phone: (312) 616-
0800, F.w: (312) 616-0226; $24).

TYPE AND DOUBLE SPACE all materials, including
literatiu'e cited, table headings, and figure legends.
Avoid hyphenated words at the right-hand margins.
Underline words to be printed in italics. Use stan-
dard bond (22x28 cm), leaving 2.5-cm margins on
all sides.

SUBMIT 3 COPIES of the manuscript and the origi-
nal on a 5.25- or 3.5-inch disk utilizing WordPerfect
4.2 or above. Number all pages and assemble each
copy separately: title page, abstract and key words,
text, acknowledgments, literature cited, appen-
dices, tables, figure legends, figures.

TITLE PAGE includes an informative title no longer
than 15 words, names and addresses of authors, a
running head of fewer than 40 letters and spaces,
footnotes to indicate change of address and author
to whom correspondence should be addressed if
other than the first author.

ABSTRACT states the purpose, methods, results,
and conclusions of the research. It is followed by
6-12 key words, listed in order of decreasing
importance, to be used for indexing.

TEXT has centered main headings printed in all
capital letters; second-level headings are centered
in upper- and lowercase letters; third-level head-
ings begin paragraphs.

VOUCHER SPECIMENS. Authors are encouraged to
designate, properly prepare, label, and deposit
high-quality voucher specimens and cultures docu-
menting their research in an established permanent
collection, and to cite the repository in publication.

REFERENCES IN THE TEXT are cited by author and
date; e.g., Martin (1989) or (Martin 1989). Multiple
citations should be separated by commas and listed
in chronological order. Use "et al." after name of
first author for citations having more than two
authors.

ACKNOWLEDGMENTS, under a centered main
heading, include special publication numbers when
appropriate.

LITERATURE CITED, also under a centered main
heading, lists references alphabetically in the fol-
lowing formats:

Mack, G. D., and L. D. Flake. 1980. Habitat rela-
tionships of waterfowl broods on South Dakota
stock ponds. Journal of Wildlife Management
44: 695-700.

Sousa, W. P 1985. Distiu-bance and patch dynamics
on rockv intertidal shores. Pages 101-124 in
S. T A. Pickett and P S. White, eds.. The ecolo-
gy of natural disturbance and patch dynamics.
Academic Press, New York.

Coulson, R. N., and J. A. Witter. 1984. Forest ento-
mology: ecology and management. John Wiley
and Sons, Inc., New York. 669 pp.

TABLES are double spaced on separate sheets and
designed to fit the width of either a single column
or a page. Use lowercase letters to indicate foot-
notes.

PHOTOCOPIES OF FIGURES are submitted initially
with the manuscript; editors may suggest changes.
Lettering on figures should be large enough to
withstand reduction to one- or two-column width.
Originals must be no larger than 22x28 cm.

NOTES. If the manuscript would be more appro-
priate as a short communication or note, follow the
above instructions but do not include an abstract.

A CHARGE of $50 per page is made for articles
published; the rate for individual subscribers will
be $35 per page. However, manuscripts with com-
plex tables and/or numerous half-tones will be
assessed an additional charge. Reprints may be pur-
chased at the time of publication (an order form is
sent with the proofs).

FINAL CHECK:

• Cover letter explaining any duplication of
information and providing phone number(s),
FAX number, and E-mail address

• 3 copies of the manuscript and WordPerfect
diskette

• Conformity with instructions

• Photocopies of illustrations

(ISSN 001 7-3614)

GREAT BASIN NATURALIST Vol 55, no l, January 1995

CONTENTS

Articles

Life histories of stoneflies (Plecoptera) in the Rio Conejos of southern Colorado

R. Edward DeWalt and Kenneth W. Stewart 1

Polhnator sharing by three sympatric milkvetches, inchiding the endangered

species Astragalus montii S. M. Geer, V. J. Tepedino, T. L. Griswold,

and W. R. Bowhn 1 9

Factors affecting selection of winter food and roosting resources by porcupines

in Utah Dave Stricklan, Jerran T. Flinders, and Rex G. Cates 29

Historic expansion of Jiiniperus occidentalis (western juniper) in southeastern

Oregon Richard F Miller and Jeffery A. Rose 37

Rangeland alpha diversities: Hai-vey Valley, Lassen National Forest, California ...

Raymond D. Ratliff 46

Effects of salinity on establishment of Populus fretnontii (cottonwood) and

Tamarix ramosissima (saltcedar) in southwestern United States

Patrick B. Shafroth, Jonathan M. Friedman, and Lee S. Ischinger 58

Names and types of Hedijsarum L. (Fabaceae) in North America

Stanley L. Welsh 66

Whipwonii {Trichiihs dipodomys) infection in kangaroo rats {Dipodomys spp.):

effects on digestive efficiency James C. Munger and Todd A. Slichter 74

Local distribution and foraging behavior of the spotted bat {Eudertna maculatum)

in northwestern Colorado and adjacent Utah Jay F Storz 78

The Chrysothamnus-Ericameria connection (Asteraceae) Loran C. Anderson 84

Notes

Reproductive behavior in Merriam's chipmunk {Tamias merriami)

Stephen B. Compton and J. R. Callahan 89

Additional records of fleas (Siphonaptera) from Utah James R. Kucera 92

H E

GREAT BASIN

NATURALIST

VOLUME 55 NO 2 — APRIL 1995

BRIGHAM YOUNG UNIVERSITY

GREAT BASIN NATURALIST

Editor

Richard W. Baumann

290 MLBM

PO Box 20200

Brigham Young University

Provo, UT 84602-0200

801-378-5053

FAX 801-378-3733

Assistant Editor

Nathan M. Smith

190 MLBM

PO Box 26879

Brigham Young University

Provo, UT 84602-6879

801-378-6688

E-mail: NMS@HBLL1.BYU.EDU

Associate Editors

Michael A. Bowers

Blandy Experimental Farm, UniversiU' of

Virginia, Box 175, Boyce, VA 22620

J. R. Callahan

Museum of Southwestern Biology, Universit) of

New Mexico, Alhuquerque, NM

Mailing address: Box 3140, Hemet, CA 92546

Jeffrey J. Johansen

Department of Biology, John Carroll University'

University Heights, OH 44118

Boris C. Kondratieff

Department of Entomology, Colorado State

University', Fort Collins, CO 80523

Paul C. Marsh

Center for Environmental Studies, Arizona

State University, Tempe, AZ 85287

Stanley D. Smith
Department of Biology
University of Nevada-Las Vegas

Las Vegas, NV 89154-4004

Fault. Tueller

Department of Environmental Resource Sciences
University of Nevada-Reno, 1000 Valley Road
Reno, NV 89512

Robert C. Whitmore

Division of Forestry, Box 6125, West Virginia

University, Morgantown, WV 26506-6125

Editorial Board. Jerran T Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoology;
William Hess, Botany and Range Science. All are at Brigham Young University. Ex Officio Editorial Board
members include Steven L. Taylor College of Biology and Agriculture; H. Duane Smith, Director Monte L.
Bean Life Science Museum; Richard W Baumann, Editor Great Basin Naturalist.

The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University
Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas
in western North America are accepted for publication.

Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin
Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian,
6385 HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889.

Editorial Production Staff

JoAnne Abel Technical Editor

Jan Spencer Assistant to the Editor

ISSN 0017-3614
4-95 750 13821

MCZ

The Great Basin Natfli-alist

Published AT Provo, Utah, BY (j j jv^'vy5
Brigham Young University

ISSN 0017-3614 H A R V A ^"^ ^^

UNiVER-

Volume 55 30 April 1995 No. 2

DIETS OF YOUNG COLORADO SQUAWFISH AND OTHER SMALL FISH
IN BACKWATERS OF THE GREEN RIVER, COLORADO AND UTAH

Robert T. Muthl and Barrel E. Snyderl

Abstract. — We compared diet of young-of-year Colorado squawfish {Pfijchocheihis hiciits), an endangered cyprinid,
with diets of other fish <75 mm total length (TL) collected fiom backwaters of the Green River between river kilome-
ters 555 and 35 during summer and autumn 1987. Species included native Wiinichthys osciiliis, Catostomus discobolus,
and C. latipinnis, and nonnative Cyprinella hitrensis, Notropis stramineiis, Pimephales promelas, Ictalunis pimctatus, and
Lepomis cijanellus. For each species, diet varied with size and between upper and lower river reaches but not between
seasons for fish of similar size. Larval chironomids and ceratopogonids were principal foods of most fishes. Copepods
and cladocerans were important in diets of E lucius <21 mm TL and L. cijanellus <31 mm TL. Catostomus discobolus
was the only species that ate moderate amounts of algae. Fish (all larvae) were in digestive tracts of only 10 P. lucius
(21-73 mm TL), about 1% of P. lucius analyzed. High diet overlap occuired between some size-reach groups of P. lucius
and C. hitrensis, R. osculus, C. lotipinnis, I. punctatus, and L. cijanellus. Potential for food competition between young-
of-year P. lucius and other fishes in backwaters appeared greatest with the ver>' abundant C. hitrensis.

Key words: Pt>'chocheilus lucius, CNTDrinella lutrensis, nonnative fishes, young-of-yean diets, diet overlap, backwaters.
Green River

Wild populations of federally endangered Colorado squawfish (Nesler et al. 1988, Haines

Colorado squawfish {Ptychocheilus lucius) per- and Tyus 1990, Tyus and Haines 1991). Ichdiyo-

sist only in the upper Colorado River basin, fauna of these backwaters is dominated by

They are most abundant in the Green and nonnative fishes, especially red shiner {Cypri-

Yampa rivers of eastern Utah and northwest- nella hitrensis; Tyus et al. 1982, Haines and

em Colorado (Tyus 1991a). Decline of this and Tyus 1990). This observation has led to a hy-

other native fishes in the Colorado River basin pothesis that nonnative fishes adversely affect

has been attributed to habitat alterations survival of young Colorado squawfish through

caused by water development and introduc- competition or predation. Stanford (1993) sug-

tion and proliferation of nonnative fishes gested that strong food-web interactions be-

(Carlson and Muth 1989, Minckley 1991). tween native and nonnative fishes probably

Backwaters of the Green River below its occur, but dietary relationships have not been

confluence with the Yampa River are impor- adequately documented (Haines and Tyus

tant nursery areas for young-of-year (YOY) 1990, Ruppert et al. 1993). Our objectives

'Lanal Fish Lal)orator\, Department of Fishery and Wildlife BioloKV". Colorado State University, Fort Collins, CO 80523.

Great Basin Naturalist

[Volume 55

were to (1) describe diets of YOY Colorado
squawfish and other small fish in backwaters
of the Green River and (2) examine diet oxerlap
and potential for competition with Colorado
squawfish.

Methods

Samples of small fish were provided by the
U.S. Fish and Wildlife Service Field Station at
Vernal, UT. These were collected from back-
waters of the Green River during summer (30
June-27 August) and autumn (22 September-
10 Decembei) 1987. The study area extends
from confluence of the Green and Yampa
rivers in Echo Park, Dinosaur National Monu-
ment, CO, to Turks Head in Canyonlands
National Park, UT — river kilometer (RK) 555
to 35 above confluence with the Colorado
Fliver. Upper and lower reaches are divided at
Sand Wash (RK 346), UT, a convenient access
point just above Desolation Canyon. Each
river reach began with a rocky, high-gradient
(1.3-2.1 m/km) segment and continued with a
sand- and silt-laden, low-gradient (0.2-0.4
m/km) segment known for relatively high
catches of YOY Colorado squawfish (Haines
and Tyus 1990, Tyus and Haines 1991). The
river was further divided into 8-km sections
starting from a random location within each
reach to help assure an even distribution of
collection sites.

Back-waters were defined as shallow (typi-
cally <0.5 m maximum depth), ephemeral
embayments with negligible water velocity.
Substrates consisted of silt and sand or silt and
mud, sometimes overlaying or interspersed
with gravel or cobble. Backwaters had little or
no rooted aquatic vegetation, but some had
dense mats of algae. Two backwaters were
sampled weekly in each 8-km section during
daylight (1000-1800 h) using l-m^ seines (0.8-
mm^ mesh) in summer and 1-m X 3-m seines
(3.2-mm X 4.8-mm mesh) in autumn. Fish
were killed and fixed in 10% formalin immedi-
ately after collection.

Up to five specimens < 20 mm total length
(TL) and five >20 mm TL of each fish species,
representing graded size series, were selected
from each sample. Each digestive tract (from
esophagus to vent) was removed, opened, and
visually assessed for percent fullness. Food
items were identified to lowest practical taxon,
and a visual estimate was made of percentage

contributed by each taxon to total \'olume of
food in each digestive tract (Larimore 1957,
Mathur 1977). For diet analyses, food-item
taxa (total of 124) were grouped into 20 family,
order, or liroader-based categories, sometimes
divided according to habitat (e.g., aquatic or
terrestrial).

Data for each fish species were stratified
according to length (10-mm TL or larger inter-
vals) by season (summer or autumn) within river
reach (upper or lower). Only subsets with at
least six fish containing food were included in
analyses. Diet measures calculated for each
subset were (1) mean percentage each food
categoiy contributed to total volume of food in
each digestive tract (mean of volume percent-
ages) and (2) percentage of all digestive tracts
in which each food category occurred (per-
centage of occurrence). Wallace (1981) evalu-
ated several diet measures and concluded that
mean of volume percentages is the best mea-
sure for calculating overlap. However, per-
centage of occurrence is useful for describing
general variations in diet (Wallace 1981,
Bowen 1983).

Similarities in diet by subset between Colo-
rado squawfish and other fishes were evaluat-
ed by Schoener's (1970) resource-overlap
index:

n
a = 1-0.5(I|P.17-F?//|),

/=1

where n is the number of food categories, Pxi
is the proportion of food category / (expressed
as mean of volume percentages) in the diet of
species x (Colorado squawfish), and Piji is the
proportion of food category / in the diet of
species y (other fishes). Values range from 0.0
(no overlap) to 1.0 (complete overlap). When
data on resource availability are absent,
Schoener's index is one of the best indices
available for calculating resource overlap
(Hurlbert 1978, Linton et al. 1981, Wallace
1981). Diet overlap is useful in helping to elu-
cidate food relationships among species and
has been considered "biologically important
when values exceed 0.60 (Zaret and Rand
1971, Matthews and Hill 1980, Galat and
Vucinich 1983).

Results

Digestive tracts from 2554 fish represent-
ing 15 species were examined for food items;

1995]

Diets of Fishes in Backwaters

<3% were empty, mostly from fish < 13 mm
TL. After subsets with <6 specimens contain-
ing food were ehminated from the data set,
2297 specimens representing nine species
remained for diet analyses. Native fish includ-
ed 972 Colorado squawfish (7.5-73.0 mm TL,
mean = 19.1), 35 speckled dace {Rlunichthys
osciihis\ 23.1-39.8 mm TL, mean = 28.1), 42
bluehead sucker [Catostomus discoholus;
23.0-58.9 mm TL, mean = 35.9), and 21 flan-
nelmouth sucker (C. latipinnis\ 32.0-64.3 mm
TL, mean = 47.9). Nonnative fish included
729 red shiner (11.3-74.5 mm TL, mean =
29.1), 92 sand shiner {Notropis stramineus;
22.2-53.2 mm TL, mean = 31.0), 330 fathead
minnow {Pimephales promelas; 11.0-65.9 mm
TL, mean = 32.5), 58 channel catfish
{Ictaluriis pimctatiis; 22.5-70.0 mm TL, mean
= 42.9), and 18 green sunfish {Lepomis
cyanellus; 20.7-56.8 mm TL, mean = 39.6).

Characterization of Diets

No major or consistent seasonal differences
in diet measures were obsei^ved within species
for fish of similar size. Accordingly, summer
and autumn data were combined for species
and lengths by river reach. Trends in values of
proportional importance of each food categoiy
were similar between the two diet measures
for all fishes; therefore, only means of volume
percentages are reported.

Diets consisted mostly of insects, zooplank-
ton, algae, seeds, and organic and inorganic
debris; but relative importance of these food
categories varied among fishes or subsets
within species (Table 1). Based on total num-
ber of food categories included in the diet of
each fish species, diets of Colorado squawfish
and red shiner were the most varied (18 and
17 food categories, respectively), followed by
speckled dace (15), fathead minnow, channel
catfish, and green sunfish (12 each), sand shin-
er (11), flannelmouth sucker (9), and bluehead
sucker (6). Variety of food consumed was
greater in the lower than upper reach for red
shiner, Colorado squawfish, flannelmouth
sucker, channel catfish, and green sunfish,
whereas diets of sand shiner, fathead minnow,
and speckled dace were more varied in the
upper reach (diet of bluehead sucker was ana-
lyzed for fish from the upper reach only). Diet
variety relative to fish length was greatest in
red shiner, sand shiner, fathead minnow,
Colorado squawfish, speckled dace, and blue-

head sucker 21-30 or 31-40 mm TL and in
flannelmouth sucker, channel catfish, and
green sunfish >40 mm TL. Mean percent full-
ness of digestive tracts was highest in fish
21-30 or 31—40 mm TL for all species.

Aquatic insects were a principal part of
diets for all fishes except fathead minnow and
bluehead sucker. Of identifiable insects,
immature dipterans (especially larval chirono-
mids) were predominant in digestive tracts.
Larval chironomids were represented by at least
21 genera, the most common being Chironomiis
followed by Wieotanytarsii^, Eukiefferiella, Polij-
pedilum, Tanytarsus, Cricotopus, and Microp-
sectra. Representative families of other imma-
ture dipterans were (in order of importance)
Ceratopogonidae, Simuliidae, Dolichopodidae,
Empididae, Muscidae, and Tipulidae. Propor-
tional contribution of immature dipterans to
diets of red shiner, sand shiner, speckled dace,
and flannelmouth sucker was higher in the
lower than upper reach. Relative importance
of immature dipterans in diets of red shiner,
sand shiner, and speckled dace decreased and
utilization of other insects increased as fish
length increased. Conversely, relative impor-
tance of immature dipterans in diets of
Colorado squawfish and channel catfish
increased or remained high with increasing
fish length. Corixids, lai"val and adult aquatic
coleopterans (predominantly Dytiscidae, Elmi-
dae, Haliplidae, and Hydrophilidae), trichopter-
an lai^vae (mainly Hydropsychidae and Hydrop-
tilidae), and ephemeropteran nymphs (pre-
dominantly Baetidae and Heptageniidae) were
minor components of diets for all fishes (<10%
of food volume) except larger red shiner,
speckled dace, and green sunfish.

Red shiner and sand shiner ate more semi-
aquatic or terrestrial insects than other fishes.
Semiaquatic insects consumed were primarily
larval and adult coleopterans (predominantly
Heterocercidae and Staphylinidae) and adult
hymenopterans (Scelionidae). Terrestrial
insects consumed were primarily hemipterans
and formicids.

All fishes ate zooplankton, but it was partic-
ularly important in diets of Colorado squaw-
fish <31 mm TL (especially <21 mm TL),
green sunfish <31 mm TL, and, to a lesser
extent, red shiner and channel catfish <31 mm
TL and flannelmouth sucker Cladocerans (many
identified as Daphnia, Eurycercus, and Macro-
thrix) and especially cyclopoid copepods

Great Basin Naturalist

[Volume 55

Table 1. I^iets by tot;
sure is mean percentage

il-Iengtli intervals (mm) of nine fish species collected during sunnner and autunm 1987 from
contributed by each food categoiy to total volume of food in each digestive tract (mean of vol-

Colorado

Red si

liincr

lid shiner

Fatheai

1 minnow

squawfish

Food categon

1 1-20

21-30

31-40

>40

21-30

31-40

>40

11-20

21-30

31-10

>4()

<11

11-20

- Upper n

eaclv —

Insects

Unidentitial)lc parts

Semiaeiuatic or terrestrial

Diptera imnuitiires

Chirononiidac adults

Anisoptera nymphs

Aquatic Coleoptera

Corixidae

Trichoptera larvae

Ephenieroptera uyniplis

Zooplankton

Cladocera and Copepoda

Rotifera

Ostracoda

Ganiniaridae

Hydracarina

Invertebrate eggs

Nematoda

Fish

Plant seeds

Algae

Organic-inorganic dehris

Number offish

140

108

305

Mean % fullness of

digestive tracts

■ Lower rp"~"^ T. _ _ _

Insects

Unidentifiable parts

Semiaquatic or terrestrial

Diptera immatures

Chironomidae adults

Aquatic Coleoptera

Corixidae

Trichoptera larvae

Ephemeroptera nymphs

Anisoptera nymphs

Zooplankton

Cladocera and Copepoda

Rotifera

Ostracoda

Gammaridae

Hydracarina

Invertebrate eggs

Nematoda

Fish

Plant seeds

Algae

Organic-inorganic debris

Number of fish

208

138

301

Mean % fullness of

digestive tracts

"Upper reach = conflueiice of Green and Vanipa nvers at RK 555 (river kilometers al>o\e confluence of tlie Green and Colorado i
Monument, CO. to Sand Wash, UT (RK .346); lower reach = Sand Wash to Turks Head. Canyonlands National Park. UT (RK 35).

in Echo Park, Dinosaur National

1995] Diets of Fishes in Backwaters 99

backwaters in two reaches of the Green River below its confluence with tlie Yampa River, Colorado and Utah. Diet mea-
ume percentages).

Colorado Fliinnelniouth Green

squavvTish Speckled dace Bluehead sucker sucker Channel catfish sunfish

21-30 3I-K) >4() 21-30 31^0 21-.30 31-40 >40 31-40 >40 21-.30 31-40 >40 21-,30 >40

Upper reach"

10 34 4

2 1

72 52 54 31 5 1 28 22 70 66

8 13

3 12

4 2

<1 8 13 1
1 <1

6
6

8
78

6
71

174

2 <1 <1

11 11 10 <1

82 88 90 64 52

14 19 9 8 6

88 78 81 77 63

■ Lower reach*

9 8

65 49 52

7 15 1

3 34

7 13 2 44 2

68
5

3 14

<1 3 <1 <1

100

Great Basin Naturalist

[Volume 55

(many identified as Cyclops) represented most
zooplankton found in digestive ti'acts. Identified
genera of rotifers ineluded Brachionus,
Cephalodellu, Keratelhi, Lecane, Monostijla,
Polyarthra, and Trichocerca. Proportion of
zooplankton in diets of all fishes tended to
decrease with increasing fish length.

Bluehead sucker was the only species that
ate moderate amounts of algae (10-30% of
food volume); other fishes consumed minor
amounts. Algae consisted mostly of six diatom
genera {Cymhella, Fragilaria, Gyrosi^iina,
Naviciila, Surirella, and Synedra), one desmid
genus {Closteritim), and, to a lesser extent,
Pediastnim (a colonial green alga). Most diges-
tive tracts contained debris that accounted for
moderate or large proportions of gut contents
(>30%) in all fishes except speckled dace and
green sunfish. It was over 80% of gut content
in fathead minnow and bluehead sucker. Debris
consisted of fibrous particles of vascular plant
tissue usually mixed with large amounts of
clay particles and sand grains, suggesting bot-
tom feeding. Seeds (many identified as tama-
risk [Tarnarix gallica]) were eaten by all fishes,
especially red shiner <31 mm TL.

Two obsenations were unique to Colorado
squawfish. Fish larvae were found in digestive
tracts of 10 Colorado squawfish (about 1% of
total examined); 1 was 21 mm TL, 8 were
36-48 mm TL, and 1 was 73 mm TL (probably
a yearling). No fish were detected in digestive
tracts of other species. Of the 18 fish larvae
found, most were too digested for species
identification or accurate length measurement,
but all were cypriniforms (mostly cyprinids)
and probably < 10 mm TL. Six fish larvae (6-9
mm TL) were identified as red shiner, and one
(about 8 mm TL) as fathead minnow. Interest-
ingly, the smallest Colorado squawfish had
four prey fish (all red shiner), whereas only
one or two fish were found in digestive tracts
of the others. Gut contents of six Colorado
squawfish, 36-48 mm TL, and the 73-mm-TL
specimen were exclusively fish; those for the
remaining specimens were 70-80% fish.
Digestive tracts of six Colorado squawfish
contained 2-6 cestode parasites (probably
Proteocephalus ptychocheilus; Flagg 1982);
cestodes were not found in guts of other fish-
es. Colorado squawfish infested with cestodes
were larger than 27 mm TL and were collect-
ed from both river reaches in autumn.

Diet Overlap

Degree of diet overlap between YOY Colo-
rado squawfish and other fishes was influenced
mainly by zooplankton and especially imma-
ture dipterans (Table 2). Within each reach,
diet overlap for all length intervals of
Colorado squawfish generally decreased as
lengths of other species increased. Degree of
diet overlap among fish of similar size was
generally greater in the lower than upper
reach. Overlap values were <0.60 (range =
0.10-0.59) for most comparisons; generally,
values were lowest for comparisons with fat-
head minnow and bluehead sucker (range =
0.10-0.44). Biologically important overlap
(values >0.60) occurred only between
Colorado squawfish > 10 mm TL and some
size-reach groups of native speckled dace and
flannelmouth sucker and nonnative red shiner,
green sunfish, and especially channel catfish.
These higher overlap values were primarily
attributed to high proportions of larval chi-
ronomids in diets and, secondarily, especially
for diet overlap with green sunfish >40 mm
TL (upper reach) and 21-30 mm TL (lower
reach), to proportions of zooplankton. Degree
of diet overlap was greatest with channel cat-
fish and green sunfish.

Discussion

Comparisons among food-habits investiga-
tions are difficult because of differences in
study design, location, and season. However,
our observations on diets of native and nonna-
tive fishes in back"waters of the Green River
generally agree with results of prior studies in
the upper Colorado River basin (e.g., Vanicek
and Kramer 1969, Jacobi and Jacobi 1982,
McAda and Tyus 1984) and reported food
habits of the nonnative species within their
native ranges (e.g., Carlander 1969, 1977,
Pflieger 1975, Harlan et al. 1987). Larger YOY
or yearling red shiner, sand shiner, speckled
dace, flannelmouth sucker, channel catfish,
and green sunfish eat mainly immature aquatic
insects. Diets of larger YOY or yearling fathead
minnow and bluehead sucker consist mostly of
algae and organic debris. Diet of YOY Colo-
rado s(|uawfish consists primarily of zooplank-
ton and immature insects (especially chirono-
mid larvae) and occasionally includes fish.

Reported size at which wild Colorado
squawfish shift to a more piscivorous diet

1995]

Diets of Fishes in Backwaters

101

varies, but generally fish become an important
food item after Colorado squawfish attain a
length of >40 mm. Osmundson and Kaeding
(1989) suggested that slower growth and poor-
er condition of YOY and especially yearling
Colorado squawfish in grow-out ponds with
lower densities of appropriate-size forage fish
might have been caused by higher reliance on
insect forage. Identifiable fish reported in
digestive tracts of YOY Colorado squawfish
here and by McAda and Tyus (1984) and
Grabowski and Hiebert (1989) were either red
shiner or fathead minnow larvae. These non-
native species are short-lived fractional spawn-
ers (Gale and Buynak 1982, Gale 1986) and
are typically present in high numbers and at
appropriate forage sizes in back-waters of the
Green River throughout summer and autumn
(Tyus et al. 1982, Karp and Tyus 1990). Kaip
and Tyus (1990) suggested that although the
abundance of small nonnative prey fishes in
the Green River might benefit growth of
young Colorado squawfish, the benefit might
be countered by the aggressive nature of some
nonnative fishes, which could have negative
effects on growth and survival of young
Colorado squawfish. In their laboratory exper-
iments on behavioral interactions, Karp and
Tyus observed that red shiner, fathead min-
now, and green sunfish shared activity sched-
ules and space with Colorado squawfish and
exhibited antagonistic behaviors toward small-
er Colorado squawfish.

We could not effectively evaluate competi-
tion for food between YOY Colorado squaw-
fish and other fishes because study design did
not provide for estimation of resource abun-
dance and availability, intraspecific diet selec-
tivity, and effects of interspecific use of impor-
tant resources. Direct evidence for interspecific
competition should be determined through
experiments demonstrating that shared use of
a limited resource negatively affects one or
more of the species (Schoener 1983, Under-
wood 1986, Wiens 1992). Additionally, we
assume gut contents represented iood con-
sumed in the backwaters of capture, but this
might not always have been the case. Tyus
(1991b) observed that although young Colo-
rado squawfish in the Green River were found
mostly in backwaters, some moved to or from
other habitats during 24-h periods. We found
that diet overlap for most comparisons with
Colorado squawfish was below the level gen-

erally considered biologically important (Table
2). Although not conclusive, these compar-
isons suggest either general resource parti-
tioning or differences in diet preferences. Diet
overlap values were considered biologically
important only for comparisons with certain
size-interval, river-reach groups of five fishes.

Because interspecific demand for resources
might not exceed supply, Bowen (1983) noted
that even extensive diet overlap is not conclu-
sive evidence for competition. Accordingly,
McAda and Tyus (1984), who also used
Schoener's index to examine diet overlap
between YOY Colorado squawfish and nonna-
tive fishes in the Green River, suggested that
high diet overlap they observed between
Colorado squawfish 22-40 mm TL and chan-
nel catfish 19-55 mm TL (overlap value =
0.60) and especially red shiner 15-69 mm TL
(overlap values 0.70-0.80) might reflect shared
use of abundant resources, primarily imma-
ture dipterans, rather than competition. The
same may be true for higher diet overlaps we
obsei-ved. Ward et al. (1986) reported that chi-
ronomids, the principal food category result-
ing in high diet overlap, were among the more
common benthic invertebrates in the Colo-
rado River basin.

We observed that overlap values were gen-
erally higher and, for most fishes, diet variety
was greater in the lower than upper reach,
perhaps because food resources were more
abundant and diverse in backwaters of the
lower reach. Based on observations during
summer and autumn 1979-1988, Haines and
Tyus (1990) found that backwaters in the
upper and lower reaches were similar in mean
surface area, but that those in the lower reach
were shallower and warmer, conditions that
may favor higher productivity. Also, within the
upper reach, Grabowski and Hiebert (1989)
noted that during summer and autumn
1987-88 concentrations of backwater nutri-
ents, particulate organic matter, phytoplank-
ton, zooplankton, and benthic macroinverte-
brates (particularly chironomid larvae) in-
creased progressively downstream. They sug-
gested this trend was due to attenuation of
flow releases from Flaming Gorge Reservoir
(located near the Wyoming-Utah border) at
downstream sites that reduced the degree of
water exchange between the main channel and
backwaters and allowed for greater backwater
warming and stability.

102

Great Basin Naturalist

[Volume 55

Tahle 2. Diet overlap by total-length (TL) intei-vals (mm) beUveen yoiiiiK-of'-year Colorado squavvfish and eight other
confluence with the Yampa River, Colorado and Utaii. Overlap \'alues were calculated using Schoener's (1970) index
asterisk (*).

Red sliiiKT

Saiul sliiiu

Fathfad
iiiiiiiKm

Ippcr reach'

LouiT reacii'

Upper

TL of
Colorado

squawfish 11-20 21-30 31-40 >40 11-20 21-30 31-40 >40 21-30 31-40 >40 21-30 31-40 >40 21-30 31-40 >40

^71 aiij 054 040 0.43 0.42 0~49 oio 047 0~46 047 035 053 055 052 O40 037 0.38

11-20 0.49 0.45 0.43 0.31 0.63* 0.53 0.47 038 0.37 0.31 0.14 0.49 045 0.38 0.18 0.15 0.15

21-30 0.55 0.51 053 0.41 0.74* 0.57 0.51 0.42 0.42 0.43 0.27 0.49 0.47 0.38 0.23 0.19 0.23

.31_1() 0.40 0.39 0.40 0.27 0.73* 0.57 0.52 0.42 0.35 0.29 0.12 0.50 0.45 0.38 0.17 0.13 015

>40 039 0.37 0.39 0.35 065* 047 0.44 0.37 0.34 0.28 0.12 0.53 0.47 037 0.17 0.13 0.14

"Upper reach = eonfliience of Green and Yampa rivers at RK 555 (river kilometers above confluence of Green and Colorado ri\ers) in Echo Park, Dinosaur National Moiinriicii
CO, to Sand Wash. UT (RK 346); lower reach = Sand Wash to Turks Head, Canyonlands National Park. UT (RK 35)

Alternatively, greater diet overlap and vari-
ety in the lower reach might have been a
reflection of a difference in backwater avail-
ability between the upper and lower reaches.
Tyus and Haines (1991) reported about 150%
more backwaters per kilometer in the upper
than lower reach. Fishes in the lower reach
might have been more crowded in available
backwaters, resulting in greater shared use
and broader intraspecific use of available food.

McAda and Tyus (1984) attributed reduc-
tions in diet overlap between Colorado squaw-
fish >40 mm TL and red shiner or channel
catfish to decreased consumption of immature
dipterans and increased consumption of fish
by Colorado squawfish. However, Ruppert et
al. (1993) reported fish larvae in digestive
tracts of 15% of adult red shiner (36-79 mm
TL) from ephemeral shoreline embayments
near confluence of the Green and Yampa
rivers. Unlike our study, they sampled on a
diel basis and killed fish with an overdose of
anesthetic before preservation to minimize
possible regurgitation. Their results suggest
that high diet overlap between young
Colorado squawfish > 40 mm TL and red
shiner might reoccur or continue with larger,
piscivorous red shiner. Although we docu-
mented high diet overlap between young
Colorado squawfish >10 mm TL and other
fishes in backwaters of the Green River, espe-
cially channel catfish (Table 2), only red shiner,
because of its extreme abundance (Haines and
Tyus 1990), is likely to be a serious competitor
for food with young Colorado squawfish. Red
shiner has often been implicated in decline of
native fishes of the American Southwest (e.g..

Minckley 1973, Greger and Deacon 1988,
Rinne 1991).

Competition might also be a factor between
smaller specimens of both Colorado squawfish
and other fishes. Few specimens <21 mm TL,
other than red shiner and fathead minnow
11-20 mm TL, were available for comparisons
with Colorado squawfish. However, as for
smaller Colorado squawfish, zooplankton
would likely be an important component of
their diets (Joseph et al. 1977), and corre-
sponding overlap values would be high, espe-
cially for specimens <11 mm TL. Although
dense populations may develop in backwaters,
zooplankton may be limited under certain
conditions because plankton communities in
rivers are subject to dramatic spatial or tempo-
ral fluctuations in abundance and diversity
(Hynes 1970, Welcomme 1985, Ward 1989). In
support of this generalization, Grabowski and
Hiebert (1989) reported that zooplankton den-
sities were higher in back-waters than in main-
channel habitats within the upper reach and
documented both spatial and temporal fluctu-
ations in zooplankton abundance. They also
observed higher concentrations of zooplank-
ton in more confined backwaters than those
with a broad connection to the river and sug-
gested that densities were influenced by
extent of water exchange between backwaters
and the main river.

In conclusion, we found high diet overlap
between YOY Colorado squawfish and several
small size groups of other fish species in Green
River backwaters. Because of the extreme
abundance of red shiner, we speculate that
diet overlap could result in food competition

1995]

Diets of Fishes in Backwaters

103

fish species collected during suniiiier and aiitiunn 1987 from backwaters in two reaches of the Green River below its
with mean of volume percentages as the diet measure; values >0.60 (biologically important overlap) are marked with an

Fathead
minnow

Flannelmouth
Speckled dace Bluehead sucker sucker

C>'liannel cattish

Lower

Upper Lower

Upper

Upper Lower Upper

Lower

Green sunHsh
Upper Lower

11-2021-30 31-40 >40 21-30 21-30 31-1() 21-30 31-40 >40 31-40 >40 >40 >40 21-30 31-40 >40 >40 21-30 >40

0.44 0.43 0.37 0.33 0.21 ().3(i 0..34 0.39 0.3.5 0.34 0.52 0.57 0.59 0.34 0.40 0.34 0.30 0.21 0.56 0.25

0.22 0.21 0.15 0.11 0.37 0.59 0.57 0.16 0.12 0.11 0.46 0.45 0.65* 0.83* 0.72* 0.61* 0.57 0.73* 0.91* 0.26

0.21 0.21 0.15 0.12 0.52 0.76* 0.61* 0.20 0.16 0.19 0.42 0.36 0.69* 0.89* 0.81* 0.75* 0.75* 0.73* 0.75* 0.27

0.21 0.21 0.14 0.10 0.42 0.78* 0.61* 0.14 0.11 0.11 0.38 0.31 0.69* 0.68* 0.81* 0.79* 0.82* 0.61* 0.69* 0.27

0.21 0.20 0.14 0.10 0..35 0.74* 0.58 0.14 0.10 0.10 0.38 0.31 0.63* 0.64* 0.69* 0.89* 0.77* 0.68* 0.57 0.24

and might have a negative impact on Colorado
squawfish growth, condition, or survival.
Studies are needed to better assess the type
and strength of interactions between native
and nonnative fishes in backwater food webs
under present regulated flow regimes and to
define factors affecting these interactions.

Acknowledgments

H. Tyus, C. Karp, and S. Lanigan initiated
this study and provided samples and field
data. H. Copeland, J. Piccolo, and E Sikoski
assisted with analysis of gut contents. H. Tyus
and C. Karp reviewed data analyses. K.
Bestgen, D. Beyers, J. Deacon, G. Haines, J.
Hawkins, C. Karp, H. Tyus, and R. Valdez
reviewed drafts of the manuscript. This proj-
ect was funded by the Recovery Implemen-
tation Program for Endangered Fish Species
in the Upper Colorado River Basin. The pro-
gram is a joint effort of the U.S. Fish and
Wildlife Service, U.S. Bureau of Reclamation,
Western Area Power Administration, states of
Colorado, Utah, and Wyoming, upper basin
water users, and environmental organizations.
This paper is Contribution No. 75 of the Colo-
rado State University Larval Fish Laboratory.

Literature Cited

BowEN, S. H. 1983. Quantitative description of diet.

Pages 325-336 in L. A. Nielsen and D. L. Johnson,

editors, Fisheries techniques. American Fisheries

Society', Bethesda, MD.
Carlander, K. D. 1969. Handbook of freshwater fishery'

biology. Volume 1. Iowa State University Press,

Ames. 752 pp.

. 1977. Handbook of freshwater fishery biology.

Volume 2. Iowa State University Press, Ames. 431
pp.

Carlson, C. A., and R. T. Muth. 1989. The Colorado
River: lifeline of the American Southwest. Pages
220-239 in D. P Dodge, editor. Proceedings of the
International Large River Symposium. Canadian
Special Publication of Fisheries and Aquatic
Sciences 106.

Flagg, R. 1982. Disease survey of the Colorado River
fishes. Pages 177-184 in Colorado River fishery proj-
ect final report. Part 3, Contracted studies. U.S. Fish
and Wildlife Service and Bureau of Reclamation,
Salt Lake City, UT

Galat, D. L., and N. Vuginich. 1983. Food partitioning
between young of the year of two sympatric tui chub
morphs. Transactions of the American Fisheries
Society 112:486-497.

Gale, W. F 1986. Indeterminate fecundity and spawning
behavior of captive red shiner — fractional, crevice
spawner. Transactions of the American Fisheries
Society' 115: 429-437.

Gale, W. F, and G. L. Buynak. 1982. Fecundity and
spawning frequency of the fathead minnow — frac-
tional spawner. Transactions of the American
Fisheries Society' 111: 35-40.

Grabowski, S. J., AND S. D. HiEBERT. 1989. Some aspects
of trophic interactions in selected backwaters and
the main channel of the Green River, Utah. Final
report of U.S. Bureau of Reclamation, Research and
Laboratory Services Division, Applied Sciences
Branch, Environmental Sciences Section, Denver CO,
for U.S. Bureau of Reclamation, Upper Colorado
Regional Off"ice, Salt Lake City, UT 131 pp.

Greger, P D., and J. E. Deagon. 1988. Food partitioning
among fishes of the Virgin River. Copeia 1988:
314-323.

Haines, G. B., and H. M. Tius. 1990. Fish associations and
environmental variables in age-0 Colorado squaw-
fish habitats. Green River, Utali. Journal of Freshwater
Ecology 5: 427-435.

Harlan, J. R., E. B. Speaker, and J. Mayhew. 1987. Iowa
fish and fishing. Iowa Department of Natural
Resources, Des Moines. 323 pp.

Hurlbert, S. H. 1978. The measurement of niche overlap
and some relatives. Ecology 59: 67-77.

104

Great Basin Naturalist

[Volume 55

Hynes, H. B. N. 1970. The ecoloj^y of rumung water.
University of" Toronto Press, Ontario, Canada. 555
pp.

JacOBI, G. Z., .\nd M. D. J.\cobi. 1982. Fish stomach con-
tent analysis. Pages 285-324 in Colorado River fish-
ery project final report. Part 3, Contracted studies.
U.S. Fish and Wildlife Service and Bureau of
Reclamation, Salt Lake City, UT.

Joseph, T. W., J. A. Sinning, R. J. Behnke, and R B.
IIoi.DEN. 1977. An evaluation of the status, life histo-
ry, and habitat reciuirements of endangered and
threatened fishes of the Upper Colorado River
System. U.S. Fish and Wildlife Service, FWS/OBS-
77/62. 169 pp.

K.\RP, C. A., AND H. M. Tvus. 1990. Behaxioral interac-
tions between young Colorado squawtish and six fish
species. Copeia 1990: 25-34.

Larimore, W R. 1957. Ecological life histoi-y of the war-
mouth (Centrarchidae). Illinois Natural History
Survey Bulletin 27: 1-83.

Linton, L. R., R. W Davies, and E J. Wrona. 1981.
Resource utilization indices: an assessment. Journal
of Animal Ecologv' 50: 283-292.

Mathur, D. 1977. Food habits and competitive relation-
ships of the bandfin shiner in Halawakee Creek,
Alabama. American Midland Naturalist 97; 89-100.

Matthews, W. J., and L. G. Hill. 1980. Habitat partition-
ing in the fish community of a southwestern river
Southwestern Naturalist 25: 51-66.

McAda, C. W, and H. M. Tvus. 1984. Resource overlap
of age-0 Colorado squawfish with other fish species
in the Green River, fall 1980. Proceedings of the
Bonneville Chapter American Fisheries Society
1984: 44-54.

MiNCKLEY, W. L. 1973. Fishes of Arizona. Arizona Came
and Fish Department, Phoeni.x. 293 pp.

. 1991. Native fishes of the Grand Canyon region:

an obituaiy? Pages 124-178 in Colorado River ecolo-
gy and dam management. National Academy Press,
Washington, DC.

Nesler, T. R, R. T. Muth, and A. E Wasowicz. 1988.
Evidence for baseline flow spikes as spawning cues
for Colorado squawfish in the Yampa River,
Colorado. American Fisheries Society Symposium 5:
68-79.

Osmund.son, D. B., and L. R. K,\eding. 1989. Colorado
squawfish and razorback sucker grow-out pond
studies as part of conservation measures for the
Green Mountain and Ruedi Reservoir water sales.
Final report of U.S. Fish and Wildlife Sei^vice, Grand
Junction, CO. 57 pp.

Pflieger, W. L. 1975. The fishes of Missouri. Missouri
Department of Conservation, Jefferson City. 343 pp.

Rlnne, J. N. 1991. Habitat use by spikedace, Meda fulgida
(Pices: Cyprinidae), in southwestern streams with
reference to probable habitat competition by red
shiner, Notropis lutrensis (Pices; CyiDrinidae). South-
western Naturalist 36: 7-13.

Ruppert, J. B., R. T. Muth, and T. R Nesler. 1993.
Predation on fish larvae by adult red shiner, Yampa
and Green rivers, Colorado. Southwestern Naturalist
38: 397-399.

SCHOENER, T. W. 1970. Non-synchronous spatial overlap
of lizards in patchy habitats. Ecology 51: 408-418.

. 1983. Field experiments on interspecific competi-
tion. American Naturalist 122: 240-285.

Stanford, J. A. 1993. Instream flows to assist the recov-
ery of endangered fishes of the upper Colorado
River basin: review and synthesis of ecological infor-
mation, issues, methods, and rationale. Final report
of Flathead Lake Biological Station, University of
Montana, Poison, for U.S. Fish and Wildlife Service,
Region 6, Denver, CO. 89 pp -I- appendices.

Tvus, H. M. 1991a. Ecology and management of Colorado
squawfish. Pages 379-402 in W. L. Minckley and
J. E. Deacon, editors. Battle against extinction.
University of Arizona Press, Tucson.

. 1991b. Movements and habitat use of young Colo-
rado squawfish in the Green River, Utah. Journal of
Freshwater Ecology 6: 43-51.

Tvus, H. M., AND G. B. Haines. 1991. Distribution, habi-
tat use, and growth of age-0 Colorado squawfish in
the Green River basin, Colorado and Utah.
Transactions of the American Fisheries Society 120:
79-89.

Tvus, H. M., B. D. Burdick, R. A. Valdez, C. M.
H.-wnes, T. a. Lvtle, and C. R. Berrv. 1982. Fishes
of the upper Colorado River basin: distribution,
abundance, and status. Pages 12-70 in W. H. Miller,
H. M. Tyus, and C. A. Carlson, editors. Fishes of the
upper Colorado River system: present and future.
Western Division of the American Fisheiy Society,
Bethesda, MD.

Underwood, T. 1986. The analysis of competition by field
experiments. Pages 240-268 in J. Kikkawa and D. J.
Anderson, editors, Commiuiity ecology: pattern and
process. Black-well Scientific Publications, Oxford,
England.

Vanicek, C. D., and R. H. Kramer. 1969. Life histoiy of
the Colorado squawfish, Ptijchocheilus liicius, and
the Colorado chub, Gila robusta, in the Green River
in Dinosaur National Monument, 1964-1966.
Transactions of the American Fisheries Society 98:
193-208.

Wallace, R. K., Jr. 1981. An assessment of diet-overlap
indexes. Transactions of the American Fisheries
Society 110: 72-76.

Ward, J. V. 1989. Riverine-wetland interactions. Pages
385-400 in R. R. Sharitz and J. W Gibbons, editors.
Freshwater wetlands and wildlife. U.S. Department
of Energy Symposium Series 61. U.S. Department of
Energy Office of Scientific and Technical
Information, Oak Ridge, TN.

Ward, J. V, H. J. Zimmerman, and L. D. Gline. 1986.
Lotic zoobenthos of the Colorado system. Pages
403-422 in B. R. Davies and K. E Walker, editors.
The ecology of river systems. Dr W. Jimk, Dordrecht,
The Netherlands.

WiENS, J. A. 1992. The ecology of bird communities.
Volume 2. Cambridge University Press, New York,
NY 316 pp.

Welcomme, R. L. 1985. River fisheries. FAO Fisheries
Technical Paper 262. 330 pp.

Zaret, T. M., and a. S. R\nd. 1971. Competition in tropi-
cal stream fishes: support for the competitive exclu-
sion principle. Ecology 52: 336-342.

Received 21 April 1994
Accepted 15 September 1994

INVERTEBRATE FAUNA OF WASTEWATER PONDS
IN SOUTHEASTERN IDAHO

Karen L. Cieminskil'2 and Lester D. Flake^"^

Abstract. — Water column invertebrates were sampled with .3.8-L activity traps in 15 sewage, industrial, and
radioactive wastewater ponds at the Idaho National Engineering Laboratoiy in southeastern Idaho. One collection was
made per pond, per month, during all months the ponds were ice-free from June 1990 through July 1991. In addition,
nutrient and selected heavy metal concentrations in pond water were determined in July 1991. Arsenic, barium, boron,
lead, selenium, and mercuiy were detected in ponds. Sewage ponds generally had higher nitrogen and phosphorus lev-
els than industrial and radioactive ponds. Of the .30 aquatic invertebrate taxa collected, the most ubiquitous were
Rotifera, Daphnidae, Eucopepoda, Ostracoda, Acari, Baetidae, Corixidae, Notonectidae, Dytiscidae, and Chironomidae.
Activity trap samples from sewage ponds contained more Rotifera, Daphnidae, and Notonectidae, whereas industrial
ponds yielded more Chydoridae, Acari, and Baetidae. Numbers of Oligochaeta, Eucopepoda, Ostracoda, Corixidae,
Dytiscidae, and Chironomidae collected were not significantly different between sewage and industrial ponds.
Compared with natural systems, these ponds had fewer taxa, but a greater number of individuals of most taxa. The high
number of invertebrates collected is attributed to the lack of fish in wastewater ponds and the high levels of nitrogen
and phosphorus, particularly in sewage ponds.

Key words: aquatic invertebrates, sanitarij wastewater, industrial wastewater, Idaho National Engineering
Laboratory.

Constructed ponds have been a common
tool in wastewater treatment for decades
(Gloyna et al. 1976). Wastewater ponds are
constructed in a variety of manners and used
in various treatment procedures, from settling
ponds to ponds with various aquatic macro-
phytes that enhance removal of nutrients and
break down organic materials (Brix 1993).
Recently, constructed wetlands have also been
incorjDorated into many wastewater treatment
systems associated with municipalities and
industry (Task Force on Natural Systems 1990,
Moshiri 1993). Wastewater ponds and wet-
lands are also associated with federal research
sites such as the Idaho National Engineering
Laboratory (INEL) in southeastern Idaho and
the Hanford Site in south central Washington.

Wastewater ponds at INEL receive sani-
tary, industrial, and radioactive waste pro-
duced at the facility. Other than wildlife
watering cisterns and ephemeral rain pools,
waste disposal ponds are usually the only sur-
face water at INEL and, as such, attract
wildlife (Halford and Millard 1978, Howe and
Flake 1989, Millard et al. 1990, Cieminski
1993). Migrating and resident waterfowl, shore-

birds, blackbirds, and swallows use the ponds
heavily, feeding partially or exclusively on
aquatic invertebrates, and on invertebrates
that have emerged from the ponds (Millard et
al. 1990, Cieminski 1993).

Most studies of macroinvertebrates, espe-
cially insects, in conjunction with waste treat-
ment have been limited to studies of benthic
invertebrate assemblages in streams receiving
raw sewage or effluent from sewage treatment
plants (e.g., Klotz 1977, Kownacki 1977, Duda
et al. 1982, Kondratieff and Simmons 1982,
Kondratieff et al. 1984, Chadwick et al. 1986,
Lewis 1986, Crawford et al. 1992). Literature
on plankton and nekton in constructed ponds
focuses mainly on pathogens, and microscopic
flora and fauna important in waste decomposi-
tion, such as bacteria, protozoa, and algae
(Goulden 1976, Task Force on Natural
Systems 1990).

Because the invertebrate fauna of waste-
water ponds attracts wildlife, it is important to
understand invertebrate communities of the
ponds, as well as if and how they differ from
natural communities. Our objectives were to
(1) provide baseline data on invertebrate

'Department of Wildlife and Fisheries Sciences, South Dakota State Universitv", Box 2140B, Brookings, SD 57007
^Present address: National Park Service, 1302.5 Riley's Lock Road, Pooles\ille, MD 208.37.
■^Address reprint requests to this author

105

106

Great Basin Naturalist

[Volume 55

resources available to migrating birds in con-
structed waste ponds and (2) determine if
nutrients and selected heax')' metals in ponds
influence invertebrate populations.

Study Site

The 231,600-ha INEL lies in Butte,
Bonneville, Bingham, Clark, and Jefferson
counties, ID, on the western edge of the
Snake River plain near the foothills of the Lost
River, Lemhi, and Bitterroot mountain ranges
(Fig. 1). Topography at INEL is flat to rolling,
with elevation ranging from 1463 m to 1829 m.

Big Lost River, Little Lost River, and Birch
Creek drainages terminate in playas on or
near INEL; flow is intermittent and largely
diverted for agriculture. During this study no
surface water flowed onto INEL. Plant com-
munities are dominated by big sagebrush
{Artemisia tridentata), low sagebrush (A. arbiis-
cula), and three-tipped sagebrush (A. triparti-
ta) (McBride et al. 1978).

INEL lies in a semiarid, cold desert. Annual
temperatures range from -42 °C to 39 °C.
Average annual precipitation is 19.1 cm, 40%
of which falls fiom April through June (Clawson
et al. 1989). Precipitation levels are lowest in
July. Snowfall averages 71.3 cm per year, and
snow cover can persist fi-om December through
March.

Wastewater ponds on INEL contained san-
itary waste (eight ponds), industrial waste
(four ponds), or radioactive waste (three
ponds) (Fig. 1). Because two radioactive ponds
also contained industrial waste, in most analy-
ses radioactive ponds were grouped with in-
dustrial ponds (as "industrial ponds") for com-
parison with sewage ponds.

Ponds were grouped around INEL facili-
ties, which were 4-36 km apart. Generally,
each facility had between one and four sewage
ponds and an industrial waste pond. Sewage
ponds ranged from 0.04 to 2.20 ha and were
0.6—2 m deep. Industrial waste ponds ranged
from 0.20 to 2.24 ha and were 0.3-4.5 m deep.
Seven of the sewage ponds and one industrial
pond were lined to prevent infiltration into
surrounding soil. Four ponds (all industrial
and/or radioactive) supported emergent plant
growth. A more thorough description of the
ponds can be found in Cieminski (1993).

Methods

Water samples were collected at ponds in
July 1991 and analyzed for nutrients (nitrogen
and phosphorus) and selected heavy metals
(arsenic, barium, ber\'llium, boron, lead, sele-
nium, and mercury) that could influence pres-
ence of invertebrates. Water pH was taken
once at each pond at the same time water
samples were collected. Further heavy metal
and nutrient sampling was prohibitively
expensive and time consuming. Water samples
were analyzed at the U.S. Geological Sui-vey's
National Water Qualit\' Laboratoiy at Ai-vada,
CO. Collection and analysis methods were as
per Brown et al. (1970) and Fishman and
Friedman (1989). Data on heavy metals for
pond ANLi (acronyms and names of pools are
included in Tables 1 and 5) were taken from
analyses conducted in 1988.

Benthic samples were not taken because
most ponds had lined bottoms, or because
sediment sampling was not pennitted for other
reasons. We collected water column inverte-
brates once each month to obtain gross esti-
mates of invertebrate populations. Additional
collections and identification were time- and
cost-prohibitive, given our concunent collection
of bird and mammal count data at these ponds
for a related project. Nevertheless, we felt that
invertebrates influenced bird use of ponds,
thus the need for estimates of invertebrate
abundance.

Water column invertebrates were collected
at all nonradioactive ponds in months the
ponds were ice-free from June 1990 through
May 1991. Because of restricted access to
radioactive waste ponds, they were sampled
only once during July 1991. Invertebrates
were collected in 3.8-L activity traps (Ross
and Murkin 1989) suspended horizontally 5.3
cm under the water surface for approximately
24 h. Modifications on the technique of Ross
and Murkin (1989) were necessary' since most
ponds had artificial liners; therefore, jars could
not be suspended from a pipe driven in the
pond bottom. Instead, jars were suspended
from floats and attached to a 50- to 300-cm-
long piece of PVC pipe anchored on the
pond's shore. The first sample was taken at the
southeast corner of each pond. Subsequent
monthly sample locations were chosen ran-
domly based on a single-digit number of paces

1995]

Invertebrates in Wastewater Ponds

107

Bitteroot
Rangoy

Containment Test Facility

disposal pond (CTFi)

Technical Support Facility

disposal pond (TSFir)

Naval Reactors Facility

industrial waste
ditch (NRFi)

sewage pond (NRFs)

Argonne National Laboratory - West

m. — secondary sewage
' pond (ANLs2)

primary sewage
pond (ANLsl)

industnal waste {>ond (ANIi)

-north cold waste
pond(TRAi2)

~ south cold waste
pond(TRAil)

■ ■- east percolation
y^ pond (CPPirl)

west percolation pond (CPPirl)

Fig. 1. Map of the Idaho National Engineering Laboraton; indicating location of facilities and wastewater ponds
where invertebrate fauna was sampled. Waste type is indicated by lowercase letter in the pond code: s = sewage, i =
industrial, r = radioactive.

counterclockwise from the previous sample
site. Where dense emergent vegetation cov-
ered the near-shore zone, the activity trap was
placed in the nearest open water.

Activity trap contents were strained tiirough
a 75-/x.m (No. 200) sieve and preserved in 80%
propanol. In the laboratory, macroinverte-

brates were removed first. Samples from shal-
low ponds with unlined bottoms often con-
tained sediment. To these, rose bengal stain
was added to aid in sorting microinvertebrates
(Mason and Yevich 1967). Samples in which
zooplankton was estimated to exceed 300 indi-
viduals were subsampled. To subsample.

108

Great Basin Naturalist

[Volume 55

samplers were diluted to 500 or 1000 ml and
stirred while 1% of the volume was drawn out
with 1- and 2-ml Henson-Stemple pipettes.

Invertebrate fauna were counted and iden-
tified to family, with the exception of the
orders Oligochaeta, Acari, Araneae, Eucope-
poda, Ostracoda, and Lepidoptera, and the
phyla Nematoda and Rotifera. Invertebrates
were identified using keys in Pennak (1989)
for non-insects, Merritt and Cummins (1984)
for aquatic insects, and Borror and DeLong
(1971) for terrestrial insects. B. McDaniel
(Plant Science Department, South Dakota
State University, Brookings) identified terres-
trial invertebrate families and verified other
identifications.

Because data were not normally distrib-
uted, nonparametric analysis methods were
used. A median test was conducted on the
dozen most common invertebrate taxa to
determine if their abundance in sewage ponds
differed from that in industrial ponds. For
each taxa, numbers of individuals collected in
each sample were used in analysis. Data were
pooled over all ponds, years, and months with-
in each of the two groups: sewage ponds and
industrial ponds. Pooling samples for years
and ponds allowed ample sample size for com-
parison of gross invertebrate population differ-
ences between pond types. A median test was
also run on the total number of species collect-
ed per pond during the entire sampling period
to determine if species richness was greater at
sewage ponds or industrial ponds. A third
median test was conducted to compare inver-

tebrate numbers between ponds with heavy
metal concentrations greater than EPA criteria
and those with heavy metal concentrations
within EPA chronic exposure standards. Data
were again pooled over all ponds, years, and
months. Radioactive waste ponds were elimi-
nated from median tests because only one
sample was taken from them.

Results

Water Chemistry

Heavy metal concentrations in most ponds
were below criteria established by the EPA
(U.S. Environmental Protection Agency 1987)
(Table 1). Mercury was the only metal found in
concentrations that might affect aquatic life
(ponds TRAr and NRFi). However, in TRAr
and NRFi mercury concentration was below
the acute value of 2.4 ^tg/L (U.S. Environ-
mental Protection Agency 1987).

Sewage ponds had higher nitrogen and
phosphorus concentrations than industrial and
radioactive ponds (Table 2). Ammonia
(NH4-N) concentrations in most ponds were
within the range found in unpolluted surface
water (Wetzel 1983); however, NH4-N con-
centrations at ICPP sewage ponds were well
above those usually found in eutrophic lakes.
Nitrite (NO2-N) concentrations indicated
high organic pollution at all sewage ponds
except NRFs, which was the only sewage
pond where NO2-N concentrations did not
exceed those of industrial and radioactive

Table L Selected heav\' metal concentrations (p-g/h) in wastewater ponds at INEL, Idiilio, August 1991, and EPA
criteria''.

Fond''

Criteria

Metal

ANLi^

CPPir2

TR.\r

TR\il

NRFi

CTFi

TSFir

(Mg/L)

Arsenic

9.4

<1'1

190'"

Barium

< 100

<100

100

50,000

Bervllium

<10

5.3

Boron

—

.50

120

5000

Lead

<2.1

3.2f

Selenium

Mercun

<2()

<0.1

0.2

<0.1

1.4

<0.1

0.0 12«

''Concentrations at or below these le\els should have no adverse effects on freshwater systems. Naval Reactor Facilities officials suggested die following clarifica-
tion: "The criteria in the last column have questionable applicabilitv' to die NRFi. The EPA maximum contaminant level for mercur\- in public community- drink-
ing water systems is 2.0/ig/L."

"ANLJ = Argonne National Laboratory-west industrial waste pond, CPPir2 = Idaho Chemical Processing Plant east percolation pond (industrial and radioactive),
TRAr = Test Reactor Area warm waste pond (radioactive), TRAil = Test Reactor .\rea south cold waste pond (industri;d), NRFi = Naval Reactors Facility industrial
waste ditch, CTFi = Containment Test Facility disposal pond (industrial), TSFir = Technical Support Facilitx disposal pond (industrial and radioactive).
•^ANLi water sample tested at Envirodyne Engineers. St. Louis. MO, February 1988.
"< symbol means water sample contained less than the detection level, which follows the < svnibol.
''Arsenic (III)

'At wafer hardness of 100 nig/L. Value is 1.3 at water hardness of 50 mg/L.
KMercury (II)

1995]

Invertebrates in Wastewater Ponds

109

Table 2. Nutrient concentrations in wastewater ponds at INEL, Idaho, August I99I.'*

Nitrogen

PliosphoiTiS

pH'i

(mg/Las N)

(nig/L as P)

Pond

NH4+

NO2-

NO2-+NO3

NO3

N03:NH4+

POj-^

Sewage ponds

ANLs2

9.02

0.19

0.17

0.46

0.29

1.50

1.20

CPPsl

7.52

11.00

2.20

4.60

2.40

0.21

4.00

CPPs2

7.23

17.00

0.69

2.40

1.71

0.10

4.80

CPPs3

7.33

17.00

0.15

0.46

0.31

0.02

6.40

CPPs4

7.43

17.00

0.14

0.43

0.29

0.02

6.10

TRAs

6.87

0.41

0.13

5.10

4.97

12.12

0.79

NRFs

9.90

0.40

0.02

0.14

0.12

0.30

3.00

Nonsewage ponds

AN Li

7.42

0.97

0.09

0.74

0.65

0.67

1.40

CPPir2

8.80

0.04

0.05

1.30

1.25

30.49

0.01

TRAil

7.60

0.01

0.06

1.10

1.04

104.00

0.07

TRAr

8.43

0.15

0.01

0.27

0.26

1.73

0.01

NRFi

7.42

0.01

O.OI

1.60

1.59

159.00

0.40

CTFir

9.97

0.01

0.45

0.44

44.00

0.09

TSFir

9.75

0.04

0.02

0.11

0.09

2.17

0.12

-'Samples were collected between 0800 and 1400 h, Mountain Standard Time.

''Water pH values fluctuate readilv. According to the INEL Industrial Waste Management Information System, 1989 effluent pH ranges and numbers of months
pH was sampled ( ) were as follows: ANLsl, 7.8-9.8 (7); CPPsl-4, 7.5-8.6 (12); TRAsl-2. 7.1-8.0 (10); NRFs, 7.4-11.0 (12); TRAil-2, 7.5-8.0 (6); TRAr 6.3-6.8
(2); NRFi, 6.9-7.5 (12); TSFir, 7.1-7.9 (12).

ponds. Nitrate (NO3-N) concentrations were
not noticeably different between sewage
ponds and industrial/radioactive ponds, and
NO3-N levels of all ponds were within ranges
commonly found in unpolluted freshwater
(Wetzel 1983).

The N03-N:NH4-N ratio is an indication
of organic pollution, a lower number indicat-
ing greater pollution (Wetzel 1983). The
N03-N:NH4-N ratio was <1 at all sewage
ponds except ANLs2 and TRAs, and >1 at all
industrial and radioactive ponds except ANLi.
However, only in ICPP sewage ponds were
ratios small enough to be considered organi-
cally contaminated (Wetzel 1983). Phosphorus
concentrations at most sewage ponds were
much higher than the concentration in the
highest industrial/radioactive pond. Compared
with maximums in uncontaminated surface
waters, phosphorus concentrations in sewage
ponds were 4-30 times greater, but of the
industrial and radioactive ponds only concen-
trations in ANLi and NRFi were substantially
greater (7 and 2X) (Wetzel 1983).

Invertebrate Fauna

Forty-nine taxa of invertebrates were col-
lected from waste ponds, of which 30 were
aquatic (Table 3). Most nonaquatic forms were
found in small numbers. Collembola, however,
were found regularly and were probably on

the water surface or shaken from emergent
vegetation in the collection process. In order
of decreasing abundance, the main taxa col-
lected were Rotifera, Daphnidae, Ostracoda,
Eucopepoda, Chydoridae, Corixidae, Chirono-
midae, Oligochaeta, Baetidae, Psychodidae,
Acari, Dytiscidae, and Notonectidae. The
above taxa were also the most ubiquitous,
except Chydoridae, Oligochaeta, and Psycho-
didae, which were found in large numbers but
in few samples.

The number of invertebrate taxa collected
per pond ranged from 5 to 22. Excluding ter-
restrial taxa, the number of aquatic taxa col-
lected ranged from 4 to 16 per pond.
Radioactive ponds were sampled only in July,
but the number of taxa collected was almost
identical to July samples from nonradioactive
industrial ponds (Table 4). Statistical analyses
were not performed on radioactive ponds
because only one activity trap sample was col-
lected. Industrial (ANLi, TRAil and 2, NRFi,
and CTFi) and sewage ponds had similar (P =
.11) numbers of taxa per sample.

Within most taxa, the number of individu-
als collected varied greatly from pond to pond
(Table 5). A median test revealed that activity
trap samples from sewage ponds contained
more Rotifera (P < .01), Daphnidae (F < .01),
and Notonectidae (P = .04), whereas industri-
al ponds yielded more Chydoridae (P < .01),

Great Basin Naturalist

[Volume 55

Table 3. Invertebrate taxa and mean number collected
from 15 wastewater ponds at INEL, Idaho, I99()-91-'.

x/24 h

TiLxa

in = 96)

Phylinn Rotifera

1471.14

Phylum Nematoda

0.05

Phylum Annelida

Class Oligocliacta (ac|uatic earthworms)

6.32

Class Iliiiidinea (leeches)

( )rder Rh\ nchohdellida

FamiK (Mossiphoniidae

0.02

Pin lum Arthropoda

Class Crustacea

Order Cladocera (water fleas)

Famih' Daphnidae

1351.26

FamiK Ch\ doridae

102.88

Family Sididae

0.09

Order Eucopepoda (copepods)

151.45

Order Ostracoda (seed shrimps)

317.17

Order Amphipoda (scuds)

FamiK' Talitridae

0.45

Class Arachnoidea

Order Acari (mites)

1.51

Order Araneae (spiders)''

0.04

Class Insecta

Order Collembola (springtails)

Family Entomobiyidae''

0.57

Family Onychimidae''

0.30

Order Ephemeroptera (mayflies)

Family Baetidae

5.71

FamiK' Caenidae

0.01

Order Odonata

Suborder Anisoptera (dragonflies)

Family Aeshnidae

0.01

Suborder Zygoptera (damselflies)

Family Coenagrionidae

0.31

Order Thysanoptera (thrips)''

Family Thripidae (common thrips)''

0.11

Family Aeolothripidae (l:)anded thrips)''

0.02

Order Hemiptera (tnie bugs)

Family Corixidae (water boatmen)

39.76

Family Notonectidae (backswimmers)

0.53

Order Homoptera

Family Aphidae (aphids)'' 0.05

Family Cercopidae (spittlebugs)'' 0.01

Family Cicadellidae (leaflioppers)'' 0.03

Family unidentified'' 0.25

Order Coleoptera (beetles)

Family Chr\'somelidae (leaf beetles) 0.03

Family Coccinellidae (ladybird beetles)'' 0.01
Family Dytiscidae (predaceous

diving beetles) 0.65

Family Elmidae (riffle beetles) 0.0 1

Family Gyrinidae (whirligig beetles) 0.01
FamiK Haliplidae (crawling water beetles) 0.02
Family Hydrophilidae (water

scavenger beetles) 0.02

Family Ptiliidae (feather-winged beetles) 0.01

Family Staphylinidae (rove beetles) 0.02

Order Trichoptera (caddisflies)

Family Leptoceridae 0.05

Order Lepidoptera (Ijuttei-flies and moths)'' 0.02

Order Diptera (flies)

Family Ceratopogonidae (biting midges) 0.01
Family Psychodidae (moth flies and

sand flies) 1.68

Family Chironomidae (midges) 11.52

Family Tipulidae (crane flies) 0.02

Family imidentified, adults'' 0.80

Family imidentified, pupae 0.99

Order Ilymenoptera

Family Formicidae (ants)'' 0.03

Family Platygasteridae'' 0.01

Family Braconidae'' 0.01

Rimily Encyrtidae'' 0.01

Rimily Pteromalidae'' 0.01

Family Scelionidae'' 0.01

Family Sphecidae (sphecid wasps)'' 0.01

"Iiix'ertebrates were collected in 3.8-L activitv' traps suspended in the water

column tor 24 h, one per pond, per month. Collections were June-October

1990 and March-May 1991 for 12 ponds, and July 1991 for .3 radioactive

ponds.

"Individuals found were mosth or e.\clusiveK' terrestrial.

Acari (P = .01), and Baetidae (P = .01).
Numbers of Oligochaeta (P = .44), Eucope-
poda (P = .50), Ostracoda (P = .09), Corixidae
(P = .08), Dytiscidae (P = .54), and Chirono-
midae (P = .70) collected were not significant-
ly different between sewage and industrial
ponds.

Invertebrate numbers in pond NRFi, which
had a high mercury content, were compared
to those in the remaining industrial ponds,
where mercury was not detected. Samples
from NRFi contained more Chironomidae (P
= .02) and Oligochaeta (P < .01), and fewer
Chydoridae (P = .03) and Ostracoda (P = .03)
than ponds ANLi, TRAi, and CTFi. Numbers
of Rotifera (P = .10), Daphnidae (P = .10),
Eucopepoda (P = .10), Acari (P = .15), Baeti-
dae (P = .55), Cori.xidae (P = .07), Notonectidae

(P = .45), and Dytiscidae (P = .07) were simi-
lar between the pond with mercuiy and those
without.

Discussion

Wastewater ponds at INEL were nutrient-
rich, especially sewage ponds. Organic enrich-
ment may be the cause of high abundance and
low number of invertebrate taxa found. Species
richness at sewage ponds was similar to that at
industrial ponds. However, species composi-
tion differed between sewage and industrial
ponds. Differences were probably due to the
greater organic enrichment in sewage ponds.

Activity trap samples from INEL ponds
contained fewer invertebrate taxa than compa-
rable samples fi-om natural waters (Gordon et al.

1995]

Invertebrates in Wastewater Ponds

111

Table 4. Number of aquatic invertebrates per collec-
tion (activity trap set for 24 h) from radioactive waste
ponds at INEL. Idaho, Jul> 199 b>.

CPPir2l'

TIUi

TSFir

TlLxa

(h = 1)

(n = 1)

(/i = I)

Daphnidae

Chydoridae

129

Eucopepoda

818

Ostracoda

1620

Amphipoda

Baetidae

Corixidae

Dytiscidae

Chironomidae

■'Data troin iadioacti\e vxaste ponds were not anal\ zed \\ itli tlioie troni sewage

and industrial ponds because onK one sample was taken from radioactive

ponds.

''CPPir2 = Idalio Chemical Processing Plant east percolation pond (industrial

and radioactive), TRAr = Test Reactor Area warm waste pond (radioactive),

TSFir = Technical Support Facilitv- disposal pond (industrial and radioactive).

1990, Neckles et al. 1990). Dominant taxa col-
lected from study ponds were similar to domi-
nant taxa collected in activity traps at natural
wetlands in Nebraska (Gordon et al. 1990) and
Manitoba (Neckles et al. 1990), with the
exception of Culicidae, Turbellaria (Neckles
1990), and Gastropoda (Gordon et al. 1990,
Neckles et al. 1990), which were not collected
from wastewater ponds. In our study fewer
taxa per sample were collected compared to
activity trap samples from seasonal wetlands
(Cowardin et al. 1979, Neckles et al. 1990);
seasonal wetlands, like organically enriched
systems of sewage ponds, tend to have low
invertebrate taxa diversity (Wiggins et al.
1980).

The reduced number of taxa in wastewater
ponds may be due to lack of emergent vegeta-
tion in most ponds. Odonate families Libelluli-
dae and Lestidae, which were collected by
Gordon et al. (1990) but not from wastewater
ponds, are commonly associated with vascular
hydrophytes (Merritt and Cummins 1984).
Vegetation has been found to be correlated
with macroinvertebrate species richness
(Gilinsk-y 1984).

Another possible cause of low species rich-
ness in wastewater ponds is high organic
waste content. Streams and wetlands receiv-
ing organic waste typically exhibit low inverte-
brate taxa diversity (Olive and Dambach 1973,
Brightman and Fox 1976, Kondratieff and
Simmons 1982, Kondratieff et al. 1984, Victor
and Dickson 1985, Pearson and Penridge 1987).
Hilsenhoff (1988) assigned arthropod families
from streams in the Great Lakes region a tol-

erance value from 0 (lowest tolerance to
organic pollution) to 10 (highest). Eleven of the
families for which Hilsenhoff (1988) presented
tolerance values were found in INEL ponds,
and only 2 had tolerance values of less than 4.
Those 11 families and tolerance values are as
follows: Aeshnidae and Tipulidae (3), Baetidae,
Elmidae, and Leptoceridae (4), Ceratopogon-
idae (6), Caenidae (7), Chironomidae and
Talitridae (8), Coenagrionidae (9), and Psycho-
didae (10). The two families with a 3 tolerance
rating were represented by only single speci-
mens in INEL wastewater ponds.

Low invertebrate diversity in industrial
ponds may be caused by organic or chemical
constituents. Although nutrients in industrial
waste ponds were within ranges found in nat-
ural waters, most industrial ponds at INEL
would be considered eutrophic (Wetzel 1983).
Additional organic enrichment in sewage ponds
did not affect species richness compared to
industrial ponds; however, species composi-
tion (%) was different between the two pond
types. Metal and saline pollution has also been
found to decrease aquatic invertebrate diversi-
ty (Savage and Rabe 1973, Seagle et al. 1980,
Euhss 1989).

In most instances, the seven heavy metals
tested did not occur in concentrations great
enough to affect aquatic life. Only mercury was
found at concentrations over chronic exposure
levels. At concentrations below chronic levels,
freshwater organisms should show no chronic
toxic effects (U.S. Environmental Protection
Agency 1987). Chydoridae and Ostracoda were
scarcer, and Chironomidae and Oligochaeta
more abundant, in samples from pond NRFi,
wherein mercury was detected. Other toxins
may occur in the water, and no other ponds
with elevated mercury concentrations were
available for comparison. Therefore, we do not
know if mercury caused tlie difference detected.

Although species richness of INEL ponds
was low, comparison with natural wetlands
(Gordon et al. 1990, Neckles et al. 1990)
revealed that study ponds exhibited high
invertebrate abundance. Of the taxa that waste-
water pond and Nebraska wetland collections
had in common, wastewater pond samples con-
tained higher densities of all except Gyrinidae,
Ceratopogonidae, and Hirudinea (Gordon et al.
1990). Gyrinidae and Ceratopogonidae were
collected in almost identical amounts, and
Hirudinea were more abundant in Nebraska

112

Great Basin Naturalist

[Volume 55

O --I o

a: "
7. 3

Z
< 3

^'i

U 3

o-i II

^ in

CO iiij

—I lO lO

o o o

r^ CO ^^

Lo m ic

i7 M

^ ^ -i<

U 3

-^ II

^ t~-

^ CO

,— , O CO

)n. .-', ^

O — I o

n T2

,u _= 5

Q Z

1j '-^- ^

CLi o

>. _C -C

O W O U D U

U < CQ

■= II

.2 H

■:r ~ t^ -= -

"c p- T ^ f-
= 1^ -a fe

o s,^

A 1 2- ^

J: S J c z

11. ill

= z

'S £ z tt- T

1 -P

O 5i lO

-^ T3 -o -a

^, in C C C

.OJ Q- O O C

i, c H a.

-^ M tX 61

5-5

■II s "S ■« '

II O C; 0;

P u 2 ^ _aj _ai

f U o.'g 1 "g "g

o <

a; Z >

3 »i K' " t; t;
- p <"S "S "I

1 cl = S S S

1995]

Invertebrates in Wastewater Ponds

113

wetlands, compared to our study ponds (Gordon
et al. 1990). Also, in our study, more Cladocera
and Ostracoda were collected compared to
activity trap samples from seasonal wetlands
(Neckles et al. 1990), which tend to have a
high invertebrate abundance (Wiggins et al.
1980). Nutrient-polluted natural waters also
have invertebrate communities containing
many individuals of a few species (Brightman
and Fox 1976, Lubini-Ferlin 1986); Brightman
and Fox (1976) attribute this partially to a
reduction in competition from pollution-intol-
erant forms.

High invertebrate growth and abundance
have been associated with high algal produc-
tivity (Wallace and Merritt 1980, Richardson
1984), which in turn has been associated with
high phosphorus and nitrogen concentrations
(Liao and Lean 1978, Wetzel 1983). Most INEL
wastewater ponds were eutrophic or highly
eutrophic (Wetzel 1983). Therefore, wastewater
ponds, which are higher in nutrients than nat-
ural wetlands, would be expected to produce
more invertebrate biomass.

The absence of fish in study ponds proba-
bly also contributed to high invertebrate den-
sities. Fish have been shown to decrease aquatic
invertebrate densities (Gilinsky 1984). For most
taxa, collections from industrial ponds also had
more individuals than collections from natural
systems (Gordon et al. 1990, Neckles et al.
1990), even though industrial ponds were not
as nutrient-rich as sewage ponds.

In certain systems a large abundance of in-
vertebrates has also been attributed to a paucity
of insect predators (Brightman and Fox 1976,
Williams 1985, Dodson 1987). However, sev-
eral predaceous taxa were collected from
waste ponds, most notably Dytiscidae and
Notonectidae. Because these taxa were col-
lected in greater numbers from wastewater
ponds than from natural wetlands (Gordon et
al. 1990), and because Notonectidae were most
numerous in sewage ponds where many prey
taxa were also most numerous, we surmise the
large number of invertebrates collected from
waste ponds resulted mostly from a reduction
in competition from pollution-intolerant taxa,
high algal productivity, and the absence of
fish, rather than from lack of invertebrate pre-
dation.

Comparison of our results on water column
invertebrates with other studies of sewage

ponds is limited due to a scarcity of published
papers. Porcella et al. (1972) noted large popu-
lations of Daphnia in a reservoir fed mostly by
treated sanitary wastewater. Daphnidae,
Rotifera, and Notonectidae were more com-
mon in INEL sewage ponds than in industrial
ponds. All three species, as well as Oligochaeta,
Eucopepoda, Ostracoda, and Corixidae (Sinclair
1975), are common inhabitants of sanitary
wastewater Oligochaeta, Eucopepoda, Ostra-
coda, Corixidae, and Chironomidae were
abundant in sewage ponds, but not more so
than in industrial ponds. Cladocera, Euco-
pepoda, Ostracoda, Corixidae, and Chironomi-
dae were also common in evaporation ponds
in California, which contain salts and heavy
metals (Euliss et al. 1991).

Invertebrate communities in INEL sewage
ponds differed from those in organically pol-
luted streams. However, in making these com-
parisons we note that our sampling methods
did not target benthic organisms. In nutrient-
enriched stream reaches, oligochaetes and
chironomids are dominant (Duda et al. 1982,
Pearson and Penridge 1987, Crawford et al.
1992), but we found no difference in numbers
between sewage and industrial ponds. Some
chironomid species (Kownacki 1977) and
oligochaete families (Lewis 1986) are charac-
teristic of clean waters, and it is possible the
species inhabiting sewage ponds differed from
those in industrial ponds. Ostracoda have also
been described as pollution tolerant (Kownacki
1977), but we found no difference in their
numbers at the .05 level of significance; at the
.10 level, sewage pond samples contained more
ostracods. Baetidae may be either pollution
tolerant (Savage and Rabe 1973, Victor and
Dickson 1985) or intolerant (Kownacki 1977)
depending upon the species. We found more
Baetidae in industrial ponds, indicating they,
as well as Chydoridae and Acari which were
also more abundant in industrial pond sam-
ples, may be less tolerant of low oxygen con-
centrations than the other common taxa.

Taxa found in greater abundance in sewage
ponds than in industrial ponds were those that
could take advantage of the unique and difficult
living conditions. Eutrophic waters typically
exhibit lower dissolved oxygen concentrations
and greater fluctuations in dissolved oxygen
and pH than less organically enriched waters.
Some cladoceran species can form hemoglo-
bin when dissolved oxygen concentrations are

114

Great Basin Natuiulist

[Volume 55

low; thus, oxygen levels are rarely a limiting
factor (Pennak 1989). The same is true of"
rotifers; certain genera are capable of with-
standing anaerobic conditions for a short time
and \ery low ox\'gen concentrations for
extended periods (Pennak 1989). Since
Notonectidae breathe at the water surface
(Merritt and Cummins 1984), they are unaf-
fected by dissolved oxygen concentrations.
Most Cladocera are less affected by pH fluctu-
ations than some taxa because they typically
occur over a wide pH range (Pennak 1989). If
pH levels are too high or too low, Cladocera
and Rotifera can withstand temporarily unfa-
\'orabIe environmental situations by producing
resting eggs that are resistant to adverse
chemical conditions. Under more favorable-
conditions, Cladocera and Rotifera life cycles
allow them to respond (quickly to improving
conditions (Pennak 1989).

Regarding the feeding habits of taxa that
were more abundant in sewage ponds,
Notonectidae were possibly taking advantage
of the reduced competition from other preda-
tors. Both rotifers and Daphnia are omnivo-
rous and feed on any suitable-sized food parti-
cle; therefore food was abundant for them in
sewage ponds (Sinclair 1975). Daphnia can
alter their body structure in response to algal
concentrations, which is thought to be a
mechanism for sui-viving algal blooms (Pennak
1989). Thus, while conditions in sewage ponds
are hostile to many species, those that can tol-
erate the conditions flourish due to an abun-
dant food supply and the absence offish.

In summary, wastewater ponds had low
invertebrate diversity, which we attribute to
lack of vegetation and inability of many
species to withstand the environmental condi-
tions. Wastewater ponds also had high inverte-
brate abundance, which we attribute to reduc-
tion of competing taxa, organic enrichment,
and absence of vertebrate predators. There
was no indication that heavy metal concentra-
tions were high enough to reduce water column
invertebrate concentrations in most ponds.

High invertebrate concentrations in INEL
wastewater ponds provided an abimdant food
source for many bird species, migrator) and
resident, which used INEL wastewater ponds.
Bacteria, protozoa, and algae are important in
waste treatment because they reduce the
organic load of wastewater and convert waste
into a form useable by organisms in the receiv-

ing water body (Goulden 1976). In systems
like some at INEL where water loss is
through evaporation, all waste processing
occurs in the pond. Zooplankton are also
important in waste elimination and transfer
(Goulden 1976, Patrick 1976, Bogatova and
Yerofeyeva 1980). Other aquatic invertebrates
that consume algae or bacteria, or feed on zoo-
plankton, and are then eaten by birds also
influence the reduction and transformation of
organic waste and its dissipation out of the
system.

Acknowledgments

We thank O. D. Markham for suggestions
from initiation through project completion.
We appreciate the assistance of L. Knobel and
R. Bartholomay of the U.S. Geological Survey,
which provided water chemistiy analysis. We
thank W L. Tucker, Experiment Station statis-
tician. South Dakota State University, for pro-
viding statistical advice, and B. McDaniels
and W G. Duffy of South Dakota State Uni-
versity for assisting in invertebrate identifica-
tion. W G. DuffV, O. D. Markham, and R. C.
Morris reviewed the manuscript. Field and lab
assistance was provided by L. Maddison, N.
Anderson, P Saffel, S. Allen, and C. Birkelo.
This research is a contribution from the INEL
Radioecology and Ecology Program and was
funded by the New Production Reactor
Office, Idaho Field Office, and the Office of
Health and Environmental Research, U.S.
Department of Energy.

Literature Cited

Bogatova, I. B., and Z. I. Yerofeyeva. 1980. The use of
container-reared cultures of Cladocera in polishing
fish farm effluents. Hydrobiological Journal 16:
,56-61.

BoRROR, D. J., AND D. M DeLong. 1971. An introduction
to the study of insects. 3rd edition. Holt, Rinehart
and Winston, New York, NY. 812 pp.

Brightman, R. S., AND J. L. Fox. 1976. The response of
benthic in\ertehrate populations to sewage addition.
Pages 29.5-308 in Third annual report on cypress
wetlands. Florida University, Center for Wetlands,
Gainesville.

Brix, H. 1993. Wastewater treatment in constructed wet-
lands: system design, removal processes, and treat-
ment performance. Pages 9-22 in G. A. Moshiri,
editor. Constructed wetlands for water quality im-
provement. Lewis Publishers, Ann Arbor, Ml.

Brown, E., M. W Skougstad, and M. J. Fishman. 1970.
Methods for collection and analysis of water samples
for dissolved minerals and gases. Techniques of

1995]

Invertebrates in Wastewater Ponds

115

water-resources investigations of the United States
Geological Survey, Book 5, Chapter Al. U.S. Goxern-
ment Printing Office, Washington, DC. 160 pp.

Ch,'\dvvick, J. W., S. E Canton, .^nd R. L. Dent. 1986.
Recovery of benthic invertebrate communities in
Silver Bow Creek, Montana, following improved
metal mine wastewater treatment. Water, Air, and
Soil Pollution 28: 427-438.

CiEMiNSKi, K. L. 1993. Wildlife use of wastewater ponds
at the Idaho National Engineering Laboratory.
Unpublished master's thesis. South Dakota State
University, Brookings. 311 pp.

Cl.wson, K. L., G. E. Start, and N. R. Ricks. 1989.
Climatography of the Idaho National Engineering
Laboratory. 2nd edition. U.S. Department of Com-
merce, National Oceanic and Atmospheric Adminis-
tration, Idaho Falls, ID. DOE-ID-12118. 155 pp.

Cow ardin, L. M., V Carter, E C. Golet, and E. T LaRoe.
1979. Classification of wetlands and deepwater habi-
tats of the United States. U.S. Fish and Wildlife
Service, Office of Biological Sei^vices, Washington,
DC. FWS/OBS -79/31.

Cr.\wford, C. C, D. J. Wangsness, and J. D. Martin.
1992. Recover)' of benthic-invertebrate communities
in the White River near Indianapolis, Indiana, LISA,
following implementation of advanced treatment of
municipal wastewater. Archi\' fiir H\drobiologie
126: 67-84.

Dodson, S. I. 1987. Animal assemblages in temporary
desert rock pools: aspects of the ecology oi Dasijhelea
sublettei (Diptera: Ceratopogonidae). Journal of the
North American Benthological Society 6: 65-71.

Duda, a. M., D. R. Lenat, and D. L. Penrose. 1982.
Water quality in urban streams — what we can e.xpect.
Journal of the Water Pollution Control Federation
54: 1139-1147.

EULISS, N. H., Jr. 1989. Assessment of drainwater evapo-
ration ponds as waterfowl habitat in the San Joaquin
Valley, California. Unpublished doctoral dissertation,
Oregon State Universit\, Conallis.

EuLiss, N. H., Jr., R. L. Jarvis, .\nd D. S. Gilmer. 1991.
Feeding ecology of waterfowl wintering on exapora-
tion ponds in California. Condor 93: 582-590.

FisH.\iAN, M. J., AND L. C. Friedman, editors. 1989.
Methods for determination of inorganic substances
in water and fluvial sediments. 3rd edition. Tech-
niques of water-resources investigations of the
United States Geological Survey, Book 5, Chapter Al.
U.S. Government Printing Office, Wishington, DC.
545 pp.

GlLlNSKT, E. 1984. The role offish predation and spatial
heterogeneity in determining benthic community
structure. Ecology 65: 455-468.

Gloyna, E. F, J. F Malina, Jr., .\nd E. M. D.wis, edi-
tors. 1976. Ponds as a wastewater treatment alteiTia-
tive. Center for Research in Water Resources, Unixer-
sity of Te.xas at Austin. 447 pp.

Gordon, C. C, L. D. Flake, and K. F Higgins. 1990.
Aquatic invertebrates in the Rainwater Basin area of
Nebraska. Prairie Naturalist 22: 191-200.

GouLDEN, C. E. 1976. Biological species interactions and
their significance in waste stabilization ponds. Pages
57-67 in E. F Gloyna, J. F Malina, Jr., and E. M.
Davis, editors. Ponds as a wastewater treatment
alternative. Center for Research in Water Resources,
University' of Texas at Austin. 447 pp.

Halford, D. K., and J. B. Millard. 1978. Vertebrate
fauna of a radioactive leaching pond complex in
southeastern Idaho. Great Basin Naturalist 38:
64-70.

HiLSENHOFK, W L. 1988. Rapid field assessment of organ-
ic pollution with a family-level biotic index. Journal
of the North American Benthological Society 7:
65-68.

Howe, F P, and L. D. Fl\ke. 1989. Mourning dove use
of man-made ponds in a cold-desert ecosystem in
Idaho. Great Basin Naturalist 49: 627-631.

Klotz, L. 1977. The effects of secondarily treated sewage
effluent on the Willimantic/Shetueket River. Uni-
versity of Connecticut Institute of Water Resources,
Storrs, Report 27. 85 pp.

KONDR/VTIEFF, R F, R. A. MATTHEWS, AND A. L. BUIKEMA,
Jr. 1984. A stressed stream ecosystem: macroinver-
tebrate community integrity and microbial trophic
response. Hydrobiologia 111: 81-91.

KONDR.^TIEFF, P F, AND G. M. SiMMONS, JR. 1982.
Nutrient retention and macroinvertebrate community
structure in a small stream receiving sewage efflu-
ent. Archiv fiir Hydrobiologie 94: 83-98.

KovvNACKl, A. 1977. Biocenosis of a high mountain stream
under the influence of tourism. 4. The bottom fauna
of the stream R\'bi Potok (the High Tatra Mts.). Acta
Hydrobiologica 19: 293-312.

Lewis, M. A. 1986. Impact of a municipal wastewater
effluent on water quality, periphyton, and inverte-
brates in the Little Miami River near Xenia, Ohio.
Ohio JouiTial of Science 86: 2-8.

LlAO, C. F-H., AND D. R. S. Lean. 1978. Nitrogen trans-
formations within the trophogenic zone of lakes.
Journal of Fisheries Research Board of Canada 35:
1102-1108.

Lubini-Ferlin, V. 1986. The influence of sewage treat-
ment plant effluents on benthic invertebrates in
Lake Zurich Switzerland. Schweitzerische Zeit-
schrift fiir Hydrologie 48: 53-63.

Mason, W. T, Jr., and P P Yevich. 1967. The use of
phloxine B and rose bengal stains to facilitate sorting
benthic samples. Transactions of the American
Microscopical Societ\' 86: 221-223.

McBride, R., N. R. French, A. H. Dahl and J. E.
Detmer. 1978. Vegetation tv'pes and surface soils of
the Idaho National Engineering Laboratory Site.
lDO-12084. U.S. Department of Energy, Idaho
Operations Office, Idaho Falls. 29 pp.

Merritt, R. W, and K. W. Cummins. 1984. An introduc-
tion to the aquatic insects of North America. 2nd
edition. Kendall/Hunt Publishing Co., Dubuque, lO.
722 pp.

Millard, J. B., F W Whicker, and O. D. Markha.m.
1990. Radionuclide uptake and growdi of barn swal-
lows nesting by radioactive leaching ponds. Health
Physics 58: 429-439.

MOSHIRI, G. A., EDITOR. 1993. Constructed wetlands for
water quality improvement. Lewis Publishers, Ann
Arbor, MI. 632 pp.

Neckles, H. a., H. R. Murkin, and J. A. Cooper. 1990.
Influences of seasonal flooding on macroinverte-
brate abundance in wetland habitats. Freshwater
Biology 23: 311-322.

Olive, J. H., and C. A. Dambach. 1973. Benthic macroin-
vertebrates as indexes of water quality in whetstone
Creek, Monow Count); Ohio (Scioto River Basin).
Ohio Journal of Science 7.3: 129-149.

116

Great Basin Naturalist

[Volume 55

Patrick, R. 1976. The effect of a stabilization pond on the
Sabine Estuary. Pages 33-55 in E. F Gloyna, J. E
Mahna, Jr., and E. M. Davis, editors. Ponds as a
wastewater treatment alternative. Center for
Research in Water Resources, University of Texas at
Austin. 447 pp.

Pearson, R. G., and L. K. Penridce. 1987. Effects of pol-
lution by organic sugar mill effluent on the macro-
invertebrates of a stream in tropical Queensland,
Australia. Journal of Environmental Management
24: 205-215.

Pennak, R. W. 1989. Fresh-water invertebrates of the
United States: Protozoa to Mollusca. 3rd edition.
John Wiley & Sons, Inc., New York, NY. 628 pp.

PoRCELLA, D. B., P H. McGauhey, and G. L. Dugan.
1972. Response to tertiai-y effluent in Indian Creek
Reservoir. Journal of the Water Pollution Control
Federation 44: 2148-2161.

Richardson, J. S. 1984. Effects of seston quality on the
growth of a lake-outlet filter-feeder. Oikos 43:
386-390.

Ross, L. C. M., AND H. R. Murkin. 1989. Invertebrates.
Pages 35-38 in E. J. Murkin and H. R. Murkin, edi-
tors. Marsh ecology research program: long term
monitoring procedures manual. Delta Waterfowl
Wetlands Research Station, Technical Bulletin 2.

Savage, N. L., and E W. Rabe. 1973. The effects of mine
and domestic wastes on macroinvertebrate commu-
nity structure in the Coeur d'Alene River Northwest
Science 47: 159-168.

Seagle, H. H., Jr., A. C. Hendricks, and J. Cairns, Jr.
1980. Does improved waste treatment have demon-

strable biological benefits? Environmental Manage-
ment 4: 49-56.

Sinclair, R. M. 1975. Fresliwater biology and pollution
ecology training manual. EPA-430/1 -75-005. National
Technical Information Service, Springfield, VA.

Task Force on Natural Systems. 1990. Natural systems
for wastewater treatment. Water Pollution Control
Federation, Alexandria, VA. 270 pp.

United States Environmental Protection Agency.
1987. Quality criteria for water 1986. EPA 440/5-86-
001. Office of Water Regulations and Standards, U.S.
Government Printing Office, Washington, DC.
1987/1302-M/60645.

Victor, R., and D. T. Dickson. 1985. Macrobenthic
invertebrates of a perturbed stream in southern
Nigeria. Environmental Pollution (Series A) 38:
99-107.

Wallace, J. B., and R. W Merritt. 1980. Filter-feeding
ecology of aquatic insects. Annual Review of Entom-
ology 25: 103-132.

Wetzel, R. G. 1983. Linniology. 2nd edition. Saunders
College Publishing, Chicago, IL. 767 pp.

Wiggins, G. B., R. J. Mackay, and I. M. Smith. 1980.
Evolutionaiy and ecological strategies of animals in
annual temporary pools. Archiv fiir Hydrobiologie
Supplement 58: 97-206.

Williams, W. D. 1985. Biotic adaptations in temporary
lentic waters, with special reference to tliose in semi-
arid and arid regions. Hydrobiologia 125: 85-110.

Received 14 January 1994
Accepted 7 September 1994

GROWTH AND REPRODUCTION IN AN ALPINE CUSHION PLANT:
ASTRAGALUS KENTROPHYTA VAR. IMPLEXUS

Wayne R. Owenl

Abstract. — A two-year field experiment was conducted to investigate factors hypothesized to affect the reproduc-
tive potential of Astragalus kentwphyta van implexus and to test the importance of trade-offs between growth and repro-
duction in this species. Levels of mineral nutrients, water, herbivory, and competition were manipulated. Seed output
and growth of individuals in treatment groups were compared against control plants. Neither water nor mineral nutri-
ents alone were shown to affect growth or reproduction. Herbivoiy was shown to be similarly unimportant in affecting
growth and reproduction. Competition with other species influenced growth but not reproduction. No significant trade-
offs between growth and reproduction were detected within \'ears. However, there did appear to be a trade-off between
these major fitness components when compared between years.

Key words: Astragalus, alpine, competition, fecundity, trade-ojf. White Mountains.

The impact of resource availability on the
reproductive output of plants is well estab-
lished (Harper 1977, Schoener 1983, Fowler
1986, Welden and Slausen 1986). Plants may
experience resource limitation as a result of
competition (inter- or intraspecific) or poor
habitat quality. Resource limitations can also
occur when a portion of a plant's photosyn-
thetic organs are removed (e.g., by herbivory),
damage which clearly interferes with the plant's
ability to provision its offspring (Marquis
1991). A number of authors (Cody 1966,
MacArthur and Wilson 1967, Harper 1977,
Grime 1979, Tilman 1982, Weiner 1988, 1990)
have considered the ecological consequences
of resource limitation for individuals and pop-
ulations and have described various strategies
that plants might be expected to pursue to
optimize the allocation of limited resources.

This study tests whether the availability of
resources limits the fecundity of Astragalus
kentrophyta Gray var. implexus (Canby)
Barneby (hereafter, simply A. kentrophyta) and
to what extent trade-offs between growth and
reproduction might influence patterns of
reproduction observed in this species. A. ken-
trophyta is an alpine cushion plant indigenous
to high elevations throughout the Intermoun-
tain West of North America (Barneby 1964).

Many lines of evidence suggest that repro-
duction in A. kentrophyta might be resource

limited. Experiments involving other organ-
isms from this habitat have shown that avail-
ability of resources influences the competitive
ability and distribution of species (Wright and
Mooney 1965, Mooney 1966, Marchand 1973),
though this is not generally true of all alpine
habitats (Korner 1989). Second, standing bio-
mass and percent cover are substantially lower
on dolomitic soils than on adjacent sandstone-
and granite-derived substrates, suggesting that
plants on the dolomite barrens might be rela-
tively resource limited (Mooney 1966, Owen
1991). Third, A. kentrophyta plants routinely
abort the majority of flowers they produce
each year (Owen 1991), a pattern that has been
attributed to resource limitations in a broad
spectrum of species (Lovett Doust and Lovett
Doust 1988).

An experiment was designed (1) to test
whether there are resource constraints on the
reproduction and growth of A. kentrophyta
and (2) to assess the interactions between two
major components of fitness (i.e., growth and
reproduction) under different regimes of
resource availability. To do this, a factorial
field experiment was established in which sep-
arate groups of plants would receive either (1)
water or (2) nutrient supplements, (3) protec-
tion from herbivory, or (4) relief from the
potentially competitive influence of neighbors.

'University of California, Davis, and White Mountain Research Station. University- of California, Los .\ngeles. Present address; Boise National Forest. 17.50
Front Street, Boise, ID 83702.

117

118

Great Basin Naturalist

[Volume 55

Study Area

The study was condueted on the alpine
dolomite barrens of Sheep Mountain Pass
above the Patriareh Grove bristlecone pine
forest, in the White Mountains of Mono
County, CA. Elevations at the site range from
3535 m (11,600 ft) to 3660 m (12,000 ft), and
topographie relief of the site is minimal. In the
White Mountains A. kentrophyta occurs only
on dolomitic soils (Lloyd and Mitchell 1973,
Hall 1991).

Weather data were obtained from the White
Mountain Research Station, Mt. Barcroft
Laboratory, located 6 km north of the study
site at an elevation of 3800 m. Soils on the dolo-
mite barrens have a high cation exchange
capacity and are depauperate in nitrogen,
phosphorus, and potassium (Mooney et al.
1962, Wright and Mooney 1965, Brayton and
Mooney 1966, Mooney 1966, Marchand 1973,
1974). The moisture-holding capacity of
dolomite-derived soils is equivalent to that of
adjacent granitic soils (Mooney et al. 1962,
Wright and Mooney 1965, Marchand 1973).
Vegetation of the White Mountains is general-
ly xerophytic; this trend is especially prevalent
on the dolomite barrens (Lloyd and Mitchell
1973).

Materials and Methods

In June 1989, 195 healthy A. kentrophyta
plants were selected randomly from within an
area of approximately 0.2 ha. Decadent (senes-
cent) plants were disqualified from inclusion
in this experiment. The specific location of the
site was chosen for its apparent homogeneity
with respect to soil physical characteristics,
vegetation, and topographic profile. Plants
were randomly allocated to five treatment
regimes: (1) 50 plants were provided with
three separate 1-L applications of water dur-
ing the 1989 growing season. Plants were
watered during the driest part of the summer
(4 July, 2 August, and 19 August) to maximize
the beneficial impact of the treatment. Water
was applied slowly (to maximize infiltration) in
a radius of 12.5 cm around each plant. This
treatment supplied 6.1 cm of moisture to each
plant. Expected precipitation for the three-
month growing season is 8.7 cm (Pace et al.
1968). The 1989 summer precipitation was 1.1
cm. This treatment group will be referred to

as "Water. " (2) Another 50 plants received sup-
plemental nutrients. These plants were given
approximately 17 g of a balanced general-pur-
pose fertilizer (Scott's All- Purpose Builder,
12:10:12 N:P:K), providing each plant with 2.0 g
N (in the form of ammoniacal nitrogen, ureas,
and water soluble nitrogen), 1.7 g P (from
phosphoric acid, P2O5), and 2.0 g K (from sol-
uble potash, K2O). These quantities are equiv-
alent to application rates of 13.8, 11.7, and
13.8 kg ha~^ respectively. A balanced fertilizer
was chosen because experiments by Chambers
et al. (1987) and Shaver and Chapin (1980) have
shown that plants in cold environments re-
spond most vigorously to resource augmenta-
tion with fertilizer containing a balance of
essential nutrients. The diy fertilizer was scat-
tered in an approximately 2-cm-wide ring
around the perimeter of each test plant.
Summer seasonal precipitation in 1989 was
apparently sufficient to solubilize the fertilizer
and deliver it to the soil profile, as the granules
had completely disappeared from the surface
in approximately one month. This treatment
group will be referred to as "Fertilized." (3) A
third treatment was designed to protect plants
from herbivoiy and predation on flowers and
young fruits. Two locally common insects ha-
bitually consume the reproductive parts of A.
kentrophyta. The more common of these in-
sects, a darkling beetle (Tenebrionidae: Coleop-
tera), consumes flowers. Larvae of a locally com-
mon Lycinid butterfly species (Lycaenidae:
Lepidoptera) occasionally consume immature
A. kentrophyta fruits. "Tangle-foot" brand
sticky-trap was applied in a circle around each
of 25 plants to exclude potential herbivores.
Tanglefoot barriers were repaired as needed.
This treatment group will be called "No
Predation." (4) The fourth treatment sought to
relieve a group of 20 A. kentrophyta plants
from neighborhood competition. A 0.25-m-
radius circle around a central target A. kentro-
phyta plant was cleared of all other plants by
cutting them off at ground level. This method
minimized ground surface disturbance.
Clearings were 0.2 m^ in area. The average
number of neighbors (ramets) removed was 63
(mostly tillers of Poa rupicola), covering an
average of 15% of the ground surface.
Excavations of A. kentrophyta plants show diat
its roots grow straight downward into the soil
with minimal lateral root spread (Owen 1991).
Roots of the target plants were therefore

1995]

Astragalus Growth and Reproduction

119

thought to be well isolated from interactions
with actively assimilating roots of other plants.
Plants clipped in the cleared areas were
trimmed if they resprouted. Plants in this treat-
ment group are referred to as the "Target "
group. (5) A final group of 50 unmanipulated
plants was marked as a "Control" group. Size
of the experimental groups was based on an
analysis of expected variances in responses to
the treatments; lower expected variances re-
quire smaller necessarv samples (Sokal and
Rohlfl981).

Plant sizes (cushion area) were measured
and recorded on 23 Jime 1989, shortly after
initiation of growth for the season. Treatments
were initially applied on 4 July 1989. In Sep-
tember 1989 all plants were remeasured, and
the entire fruit and seed crop produced by
each of the 195 plants was han'ested. Since A.
kentropJnjta forms a tight cushion that never
exceeds 1 cm in height and seeds are not
released from the plant before the end of the
growing season, there was great confidence
that the entire seed crop of each individual
was retrieved. In early June 1990 I again mea-
sured the area of all plants just as they were
initiating growth for the season. Fertilized and
Water treatments were not repeated in 1990
so as to evaluate the potential for lags in the
effectiveness of resource supplementation.
Tanglefoot barriers were maintained during
1990 to test for interannual variation in the
effects of herbivores and predators. Clear zones
around Target plants were maintained in 1990.
All plants were allowed to grow through the
season, and in September 1990 all 195 plants
were remeasured and all fruits and seeds har-
vested. No attempt was made to quantify' flower

production, but previous experience (Owen
1991) had shown that seed production is a sig-
nificant function of flower production (Owen
1991). Flowers, when aborted, are dropped at
a very early age (Owen 1991) and probably
represent a minimal per-unit cost in resources
to the plant (Bookman 1983, Stephenson
1984). Therefore, the cost of flowers should be
proportional to a plant's seed output and can
safely be disregarded for the purpose of this
work. Fruits and seeds were cleaned and sepa-
rated in the laboratoiy, counted, and weighed.

Results

Weight of individual reproductive struc-
tures (seeds and fruits) was independent of
total numbers of those items produced per
plant in both years (Table 1). Average seed and
fruit weights were significantly correlated [R
= .429 in 1989, R = .443 in 1990). There were
no significant differences between treatment
groups for the weight of individual seeds or
fruits (results not presented). Because seed
production is well correlated with other possi-
ble measures of fitness in A. kentrophyta and
weights of those seeds are independent of the
numbers of reproductive structures produced
on a plant (Table 1), seed output was used as
an index of total reproductive effort.

In a comparison of slopes of regression
analyses, growth was a significant function of
plant size in both 1989 and 1990 (Table 2),
though the relationship was weaker in 1990.
The weight of individual seeds and fruits was
independent of seasonal growth (Table 2). The
amount of growth across years was significant-
ly but poorly correlated.

Table 1. Con-elation matrix for selected demographic traits. Values above the diagonal are conelation coefficients (R)
based on 1990 data; those below the diagonal are derived from 1989 data.

Seed

Fruit

Reproductive

Seeds

weight

Fniits

weight

produced

(a\erage)

(total)

produced

(ax'erage)

(total)

Seeds produced**

.003

.976*

,964*

.143*

.920*

.966*

Seed weight (average)

.042

.139*

-.001

.433*

.081*

.115*

Seed weight (total)

.977*

.200*

.94,5*

.229*

.937*

.987*

Fruits produced**

.963*

.024

.033*

.106

.963*

.968*

Fruit weight (average)

.136*

.429*

.215*

.074

.289*

.260*

Fruit weight (total)**

.943*

.120*

.949*

.9,52*

.284*

.981*

Total reproductive weight**

.973*

.16,56*

.989*

.9,54*

.249*

.985*

♦Kendall Rank Correlation is significant at P < .0,5.
**Treatment differences noted with one-way .\NOVA. Tlies

■ diflcrences do not affect the magnitude of significance of the coiTelations.

120

Great Basin Naturalist

[Volume 55

Table 2. Slopes of regressions for selected demographic
traits on growth in 1989 and 1990 using the total data set
(i.e., not partitioned by treatment). Where the overall
regressions are not significant, there were also no treat-
ment differences.

Ciowth in 1989 Growth in 1990

Table 3. Result of an ANCOVA on seed production and
growth h\' treatment group. The covariate is plant size.
The treatments are those listed in the text (see also Table 4).

Growth in 1990

.168*

—

Plant size

.340*

.110*

Seed weight

-.038

-.054

Fruit weight

.035

.036

'Regressions are significantK positise (P < .0.5). f)ne-wa\- .WOVAs suggest
differences between treatment groups for values ol these traits (P < .0.5).

Seed production (square root transformed)
was a positive linear function of plant size.
Overall values of R- for regressions of seed
production on plant size were .206 in 1989
and .182 in 1990. Slopes of individual regres-
sions for each treatment for seed production
on plant size did not differ from the slope for
control plants.

Plant size was a minor but important factor
influencing both growth and reproduction in
A. kentrophyta and indicates that size should
be considered as a covariate in an analysis of
variance of treatment effects in this experi-
ment. Analyses of covariance (ANCOVA) and
experimental results are presented in Tables 3
and 4, respectively. Plant size was a significant
covariate in three of four analyses. There were
no differences among treatment groups in
seed production (reproduction) for either year
Growth did not differ among treatment groups
in 1989, but there was a significant difference
between groups in 1990 (P = .047). A protect-
ed least-significant-difference (LSD) test indi-
cates that growth in the Target group was
greater than that of individuals in other treat-
ment groups (Table 4).

Table 5 gives the results of two-tailed t tests
comparing mean reproduction and growth
across years within treatment groups. There
were no significant differences for seed pro-
duction among treatment groups between
1989 and 1990. Average size for plants in 1990
was consistently significantly greater than the
size of the same plants the previous year (i.e.,
on average, plants grew larger over the course
of the experiment). The No Predation treat-
ment grew significantly less in 1990 than
1989, whereas plants in the Target group grew
significantly more in 1990. There were no sig-
nificant differences in growth across years for
plants in the Control, Fertilized, or Water
groups.

Covariate

Treatment

1989 Seed production 37.164 <.001 1.358 .25

1990 Seed production ,39.818 <.001 1.854 .12
1989 Growth 27.207 <.001 0.822 ..583

1990 Growth

0.893 .346 2.453 .047

A series of simple linear regressions was
used to compare seed production with growth
to test for the presence of a trade-off between
these two primary components of fitness.
When the data are corrected for the fact that
larger plants are inherently more capable of
producing more flowers and fruits, the analy-
sis finds no significant differences among
treatment groups (by virtue of overlapping
95% confidence intervals); and, therefore, no
trade-off between growth and reproduction
within a given year was detected.

To compare trade-offs across years, the
ratio of 1990 to 1989 data was used (Table 5).
This provides a number >1.0 when 1990 data
values exceed 1989 values; the converse is
true when results are <1.0. Seed production
was greater in 1990 than in 1989 regardless of
treatment group. In contrast, growth in 1990
was less than that experienced in 1989 with
the notable exception of Target plants. The
results can be interpreted as evidence for a
trade-off between growth and reproduction.
They indicate that, in general, increased seed
production is associated witli decreased growth.
Furthermore, plants may be relieved of trade-
off constraints by removing competitors,
which should increase availability of mineral
resources to the remaining (target) plant.

Discussion

Resource supplementation or alleviation of
resource competition did not significantly
influence the reproductive output of A. kentro-
phyta. Instead, seed production was more close-
ly related to the individual s past record of seed
output (Tables 1, 3, 5). Plants that produced
many seeds in 1989 tended to produce many
seeds in 1990, regardless of treatment. Growth,
while similarly unresponsive to the addition of
single resources, increased significantly when
potential competitors were removed (Tables 4,

1995]

Astragalus Growth and Reproduction

121

Table 4. Treatment means (SD) in both 1989 and 1990 for important demographic traits.

Control

No bugs

Fertilized

Water

Target

1989 Seed production

1990 Seed production

25.8 (25.2
32.2 (32.28)

16.1 (11.8)
20.5 (16.7)

30.6 (24.5)

39.7 (37.3)

25.1 (22.2)
30.7 (27.9)

44.2 (41.4)
54.5 (58.4)

1989 Plant size

1990 Plant size

5997.1 (2851.7)
7247.3 (3128.8)

4594.6 (1871.8)
5596.3 (2156.6)

6833.9 (2892.7)
7934.0 (3242.6)

6333.2 (2891.4)
7418.2 (3627.4)

7683.2 (3683.8)
8393.0 (4159.9)

1989 Growth

1990 Growth*

1478.4 (1329.7)
1156.1 (1529.9)

1530.0 (987.7)
808.4 (1000.4)

1772.1 (1634.2)
1587.8 (2044.5)

1797.9 (1486.9)
1395.0 (1760.3)

1503.1 (988.6)

2433.2 (1749.0)

*Groutli in 1990 \aried significantK among treatments (see Table 3). The Target groups grew more, on a\erage, tlian did plants in an\ other treatment group.
No other differences were significant.

5). These results differ from those of Wright
and Mooney (1965), Mooney (1966), and
Marchand (1973), which show that mineral
nutrients were the primary factors limiting
other species that occur on dolomite in the
White Mountains {Artemisia tridentata, two
Erigeron species, and Liipinus argenteus,
respectively). Korner (1989) reports that the
effect of fertilization on the growth of species
from nutrient-poor environments is often diffi-
cult to detect. He does not cite studies that
address the relationship between growth and
reproduction in nutrient-supplementation
experiments.

The addition of mineral nutrients or water
alone may have been insufficient stimuli for A.
kentrophyto to increase either reproduction or
growth if both factors were limiting. Multiple
limiting factors have been reported in a vari-
ety of species (Harper 1977) and are specifi-
cally predicted by Tilman's (1980, 1982) mod-
els of optimal resource consumption. That
there may be multiple resource limits to A.
kentrophyta growth and reproduction is sup-
ported by the response of A. kentrophyta to
the removal of competitors in this study.

Tanglefoot barriers were very effective at
excluding ground-moving herbivores and
predators. This was evidenced by the lack of
foliar damage or partially eaten fruit and the
capture of many insects in the traps. Flowers
of A. kentrophyta are produced in sufficient
excess to buffer individuals against the levels
of flower and fruit predation observed in this
population.

Growth in A. kentrophyta, as has been re-
ported for a number of species from arid regions
throughout the world (Fonteyn and Mahall
1981, Robberecht et al. 1983, Ehleringer
1984, Parker and Salzman 1985, Shaw 1987,
Manning and Barbour 1988, and Chapin et al.

1989), is most sensitive to the pro.ximity of its
neighbors. It is unclear, however, why repro-
duction among such species is rarely similarly
influenced (as is the case with A. kentrophyta).
The buffering of fitness components against
environmental stochasticity is characteristic of
density-vague demographics as described by
Strong (1986). Under density-vague condi-
tions, selection favors demographic functions
with indeterminate functional thresholds. That
is, current allocation decisions are only loosely
linked to current environmental conditions
(Strong 1986).

Trade-offs between growth and reproduc-
tion within years were not observed in this
experiment under any conditions. A weak
trade-off between growth and reproduction
was identified in most treatment groups when
data were compared across years (Table 5). It
is of great interest that the Target group alone
experienced an increase in both seed produc-
tion and growth in 1990 compared to 1989 val-
ues (and thus did not experience a trade-off).
The absence of well-defined trade-offs between
primaiy components of fitness could be due to
one of several reasons. Lack of a discernible
trade-off would be noted if resources were not
truly limiting. It may also be that growth and
reproduction are not co-limiting for this
species in this environment. If this were true,
factors that influence growth and reproduction
are likely to be independent (e.g., one fitness
component might be canalized and the other
dependent on environmental conditions).
Finally, a trade-off between growth and repro-
duction would not be detected if a resource
other than one provided in this experiment
were limiting.

Adult A. kentrophyta mortality at the Sheep
Mountain study site is low, juvenile mortality
is extremely high (even though germination

122

Great Basin Naturalist

[Volume 55

Table 5. Cross-year comparisons of fitness components. 1990 \ akies represented as a fraction of 1989 trait valnes.
Values of t and tiie associated prol)al)ilities (P) represent results of two-tailed / tests for differences in values between
years. Refer to Table 4 lor raw tlata.

Control

No Inigs

Fertilized

Water

Target

Seed production

90/89*

1.25

1.16

1..32

1.15

1.18

1.41

1.71

1.80

1.39

0.71

.17

.10

.08

.17

.49

Plant size

7.06

5.02

4.90

5.05

3.50

<.01

(wowth

90/89*

0.85

0.98

0.56

0.86

2.07

1.13

2.50

0.40

1.42

2.12

.26

.02

.70

.16

.05

*\'alues listed represent the ratio ol 1990 trait \ allies to those ol 19S9

tests under controlled conditions show seed
viability of greater than 95%), and recruitment
is low (Owen 1991). These demographic attri-
butes would certainly favor a strategy that
routes resources away from the risky business
of reproduction toward growth. The small but
consistent portion of A. kentrophyta's annual
accumulation of biomass allocated to repro-
duction guarantees that each plant will proba-
bly produce at least a few seeds each year
while being able to dedicate most of each sea-
son's accumulated resources to growth and
survival. That the allocation of resources to
reproduction, but not growth, in this species is
constant over a broad range of resource avail-
abilities is consistent with a bet-hedging life-
history strategy (Kozlowski and Stearns 1989,
Philippi and Seger 1989, Stearns 1989).

Resource limitations on organisms are rarely
simple or solitary. While fruit and flower pre-
dation can be an important limit on fecundity,
such an effect was not noted here. Similarly,
the reproductive output of plants growing on
the Sheep Mountain dolomite barrens would
appear to be resource limited, although single
resource augmentation had no direct effect on
seed production. In combination, however,
resources can influence the amount of realized
growth that in subsequent years will affect
reproduction.

Acknowledgments

I would like to thank the White Mountain
Research Station for providing logistic and
financial support for this project, especially
the crew at the Mt. Barcroft Laboratory. T.
Holmes, E. Nagy, A. Fitter, and two anony-
mous reviewers made significant improve-
ments on earlier drafts of this manuscript.

Literature Cited

B.^RNEBY, R. C. 1964. Atlas of North American Astragalus.
Memoirs of the New York Botanical Garden 13:
1-1187.

Bookman, S. S. 1983. Costs and benefits of flower abscis-
sion and fruit abortion in Asclepias speciosa. Ecology
64; 264-273.

Brayton, R., and H. a. Mooney. 1966. Population vari-
ability of Cercocarpus in the White Mountains of Cali-
fornia as related to habitat. Evolution 20: 383-391.

Chambers, J. C, J. A. MacMahon, and R. W. Brown.
1987. Response of an early serai dominant alpine
grass and a late serai dominant alpine forb to N and P
availability. Reclamation and Revegetation Research
6: 219-234.

Chapin, E S., J. B. McGraw, and G. R. Shaver. 1989.
Competition causes regular spacing of alder in
Alaskan shrub tundra. Oecologia 79: 412-416.

Cody, M. L. 1966. A general theory of clutch size.
Evolution 20: 174-184.

Ehleringer, J. R. 1984. Intraspecific competitive effects
on water relations, growth, and reproduction in
Encelia farinosa. Oecologia 63: 153-158.

EONTEYN, P J., AND B. E. Mahall. 1981. An e.xperimental
analysis of structure in a desert plant community.
Journal of Ecolog>' 69: 883-896.

Fowler, N. 1986. The role of competition in plant commun-
ities ill arid and semiarid regions. Annual Review of
Ecology and Systematics 17: 443-464.

Grime, J. P 1979. Plant strategies and vegetation process-
es. John Wiley and Sons, New York, NY.

Hall, C. A., editor. 1991. Natural histoiy of the White-
Inyo Range. University of California Press, Berkeley.

Harper, J. L. 1977. Population biolo,g\ of plants. Academic
Press, New York, NY.

KoRNER, C. 1989. The nutritional status of plants from
higher altitudes. Oecologia 81: 379-391.

Kozlowski, J., and S. C. Stearns. 1989. Hypotheses for
the production of excess zygotes: models of bet-
liedging and selective abortion. Exolution 43:
1369-1377.

Lloyd, R. M., and R. S. Mitchell. 1973. A flora of the
White Mountains, Ctilifornia and Nevada. University
of California Press, Berkeley.

Lovett Doust, J., and L. Lovett Dolst. 1988. Plant
reproductive ecolog>'. O.xford Universit)' Press, New
York, NY

1995]

Astragalus Growth and Reproduction

123

MacArthur, R. H., and E. O. Wilson. 1967. The theow
of island biogeography. Princeton University Press,
Princeton, NJ.

Manning, S. J., and M. G. Barboi r. 1988. Root systems,
spatial patterns, and competition for soil moisture
between desert subshrubs. American Journal of
Botany 75: 885-893.

Marchand, D. E. 1973. Edaphic control of plant distribu-
tion in the White Mountains, eastern California.
Ecolog>- 54: 233-250.

. 1974. Chemical weathering, soil development, and

geochemical fractionation in a part of the White
Mountains, Mono and Inyo counties, California.
uses Professional Paper 352-J.

Marquis, R. J. 1991. Evolution of resistance in plants to
herbivores. Evolutionar\' Trends in Plants 5: 23-29.

MooNEV, H. A. 1966. Influence of soil t}'pe on the distribu-
tion of Kvo closeK' related species oi Erigeron. Ecolog\'
47:950-958.

MooNEY, H. A., G. St. Andre, and R. D. Wright. 1962.
Alpine and subalpine vegetation patterns in the
White Mountains of California. American Midland
Naturalist 68: 257-273.

Owen, W R. 1991. The reproductive ecology of an alpine
legume: A. kentrophijfa var iinplexiis. Unpublished
dissertation, Universit\' of California, Davis. 226 pp.

Pace, N., D. W Kiepert, and E. M. Nissen. 1968. Clima-
tological data summaiy for die Crooked Creek Labora-
tory, 1949-1967, and the Barcroft Laboraton; 1953-
1967. University of California, White Mountain
Research Station Publication, Berkeley.

Parker, M. A., and A. G. Salzman. 1985. Herbivore
exclosure and competitor removal: effects on juve-
nile survivorship and growth in the shrub
Giitierrezia microcephaUi. Journal of Ecology 73:
903-913.

Philippi, T, and J. Seger. 1989. Hedging one's evolution-
ary bets, revisited. Trends in Ecolog}- and Evolution
4:41-44.

Robberecht, R., B. E. Mahall, and R S. Nobel. 1983.
E.xperimental removal of intraspecific competitors —
effects on water relations and productivity of a
desert bunchgrass Hilaria rigida. Oecologia 60:
21-24.

ScHOENER, T. W. 1983. Field experiments on interspecific
competition. American Naturalist 122: 240-285.

Shaver, G. R., and E S. Chapin. 1980. Response to fertil-
ization by various plant growth forms in an Alaskan
timdra: nutrient accumulation and growth. Ecology
61: 662-675.

Sh.wv, R. G. 1987. Density dependence in Salvia Iijrata:
experimental alterations of densities of established
plants. Journal of Ecology 75: 1049-1063.

SOKAL, R. R., and E J. Rohlk 1981. Biometry. W.H.
Freeman and Company, New York, NY.

Stearns, S. C. 1989. Trade-offs in life-histor\- evolution.
Functional Ecolog>' 3: 259-268.

Stephenson, A. G. 1984. The cost of over-initiating fruit.
American Midland Naturalist 112: 379-386.

Strong, D. R. 1986. Density vagueness: abiding the vari-
ance in the demography of real popidations. Pages
257-268 in J. Diamond and T J. Case, editors,
Communitv ecologv. Harper and Row Publishers,
New York, NY

TiLMAN, D. 1980. Resources, a graphical-mechanistic
approach to competition and predation. American
Naturalist 116: 362-393.

. 1982. Resource competition and communit\' stnic-

ture. Princeton University Press, Princeton, NJ.

Weiner, J. 1988. Variation in the performance of individ-
uals in plant populations. Pages 59-81 in A. J. Davey,
M. J. Hutchings, and A. R. Watkinson, editors. Plant
population ecolog\'. Black-well Scientific Publications,
London.

. 1990. Resource competition and commimit>- stmc-

ture. Trends in Evolution and Ecology 5: 360-364.

Welden, C. W, and W L. Slausen. 1986. The intensity
of competition versus its importance: an overlooked
distinction and some implications. Quarterly Review
ofBiology 61: 23-44.

Wright, R. D., and H. A. Mooney. 1965. Substrate-ori-
ented distribution of bristlecone pine in the White
Mountains of California. American Midland Naturalist
73; 257-284.

Received 21 January 1994
Accepted 28 October 1994

CALILEUCTRA, A NEW GENUS, AND TWO NEW SPECIES

OF STONEFLIES FROM CALIFORNIA

(PLECOPTERA: LEUCTRIDAE)

W. D. Shepard' and R. W. Baumann^

Abstract. — Calileuctra is proposed as a new genus in the family Leuctridae, with Calileuctra ephemera designated
as the type species. All stages of Calileuctra ephemera are described. Calileuctra dohnji is described in the male and
female stages. Both species inhabit the Mediterranean climatic region of California. A phylogenetic analysis of the gen-
era in the family Leuctridae is given, which places Calileuctra near the genus Perlomijia.

Key words. — Insecta, Plecoptera, Leuctridae, Calileuctra, description, distribution, plnjlogemj.

Both of us have been collecting stoneflies
from streams all across California. Several years
ago one of us (WDS) collected a small and
poorly sclerotized stonefly nymph from an inter-
mittent Napa Valley stream. The male adult
that was reared from the nymph could not be
determined using existing keys by WDS. The
specimen was then given to RWB for identifi-
cation. His identification kept us collecting at
the same site for nine years. The single male
specimen was first thought to be a new species
in the Asian genus Rhopalopsole. However,
recent work indicates that the male represents
a new genus in the family Leuctridae. Despite
extensive searching in surrounding areas, only
the Napa Valley population has been found.

A few years after discovery of the first new
species, RWB found, in the Natural History
Museum of Los Angeles County, a small series
of an interesting new leuctrid from the San
Cabriel Mountains. Later, two additional fe-
males of this species were collected in the
Santa Ana Mountains. However, we decided
that fresh male specimens were needed before
a description could be undertaken.

Keith Dobry, who was doing fieldwork in
the Los Angeles area, was encouraged to look
for additional specimens of this leuctrid species.
He was successful in locating two additional
populations, one in the San Cabriel Moun-
tains, the other in the Santa Monica Mountains.
This species is known from only four popula-

tions, all from mountains surrounding the Los
Angeles basin.

Calileuctra, new genus

Type species. — Calileuctra ephemera, new
species

Adults. — Body brownish, weakly sclero-
tized; setation sparse, except for abundant tiny
setae, "clothing hairs" (Figs. I, 10). Wings
macropterous or brachypterous; venation as
illustrated (Fig. 3). Prosternum with prester-
num separate, furcasternum fused to base of
triangular basisternum; meso- and metaster-
num similar except basisternum rectangular
(Fig. 2).

Male. — Tergum IX with posterior border
heavily sclerotized and irregularly serrate or
dentate; tergum X with posterolateral corners,
each with one or two elongate horns project-
ing posteriorly (Figs. 4, 11); sternum IX pro-
jecting posteriorly to cover base of paraprocts,
with vesicle broadening posteriorly (Figs. 6,
13); paraprocts fused into a complex, T-
shaped, subanal probe, with two ventromedial
projections off subanal probe (Figs. 8, 9, 13).

Female. — With weak abdominal scleroti-
zation; sternum VII completely sclerotized;
sternum VIII largely membranous; sternum
IX completely sclerotized; subgenital plate
poorly produced; sternum X incompletely
sclerotized (Figs. 7, 14). Cerci one-segmented;
elongate in male, poorly sclerotized on sides.

' Depaitnu-iit of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, CA 94nf). Mailing address: 6824 Linda Sue Way, Fair
Oaks. CA 9.5628.

-Department of Zoology, Monte L. Bean Life Science Musenni, Brigliam Young University, Provo, UT 84602.

124

1995]

Calileuctra New Genus

125

apically flat and membranous (Figs. 4, 11);
simple in female (Figs. 7, 14).

Nymph. — Mature nymph weakly sclero-
tized; body elongate; setation scarce.
Abdominal segments I-VII with membranous
pleural fold. Mesosternal Y-ridge with double
stem; arms meeting furcal pits at posterior
ends. Paraprocts fused basally with no visible
suture; sparse setation. Cereal segments each
with apical fringe of 10-15 setae; setae approx-
imately one-half length of cereal segments.

Distribution. — Napa Valley and Los
Angeles basin, CA.

Diagnosis. — Males are best characterized
by their unique elongate, flat-topped cerci.
Females are characterized by sternum VIII
being incomplete, and the lack of a posteriorly
projecting subgenital plate. Nymphs are char-
acterized by abdominal segments I-VII hav-
ing a membranous pleural fold, the subanal
lobes having basal fusion but no distinct fusion
line, and the cereal segment setae being one-
half length of the cereal segment.

Eri'MOLOGY. — The prefix Cali- was select-
ed to denote California, the origin of the spec-
imens. The suffix -leuctra was selected to indi-
cate placement of the genus in the family
Leuctridae. Gender of the name is neuter.

Key modification. — Modifications are
given for the following identification keys for
Nearctic leuctrid genera: Hai-per and Stewart
(1984) — nymphal and adult keys; Stewart and
Stark (1988) — nymphal key. Wording, style,
and figure citations are as presented in the
original keys.

Haiper and Stewart (1984) — Nymphal Key

38 (37) Body robust, length less than 8 times width;
body conspicuously clothed with hairs
about one-fifth the length of middle Ab
segment; subanal lobes of mature male a
fused, strongly keeled plate, much pro-
duced with no posterior notch (fig.
13.44) Megaleiictra

38' Body more elongate, Hue hair pile incon-
spicuous, appearing naked; subanal lobes
of mature male fused one-half to two-
thirds length, leaving a notch at tip 38A

38A (38) Subanal lobes fused but with complete
suture; apical setae on cereal segments
usually less than one-half length of cer-
eal segments Peiiomijia

38A' Subanal lobes basally fused, no suture in

basal half; apical setae on cereal segments

one-half length of cereal segments
Calileuctra

Harper and Stewart (1984)— Adult Key

58 (56) In hind wing, Cuj not forked . . . Calileuctra
58' In hind wing, Cuj forked 58A

58A (58') In hindwing, m-cu joining Cuj beyond fork

of Cui ' 59

58A' In hindwing, m-cu joining Cuj before fork

of Cui . 60

Stewart and Stark (1988)— Nymphal Key

5. Pronotum with no long, marginal setae (Fig.
8.14A); paraprocts of both sexes fused
basally, with no distinct medial line of
separation 5A

Pronotum with 2-4 long hairs on anterior
and posterior margins (Fig. 8.2A,
8.12A); paraprocts of both sexes fused
with distinct medial line of separation
or slightly separated medially (Fig.
8.2H,I; 8.12H,I) 6

5A. Abdominal segments 1-6 divided by ventro-
lateral membrane; ENA and SW
Zealeucfra

5A' Abdominal segments 1-7 divided by ventro-
lateral membrane; WNA Calileuctra

Key to Adults o( Calileuctra

la Wings macropterous; male with epiproct bifurcate
dorsally, tergite IX posteriorly emarginate with
two large heavily sclerotized teeth, cerci with api-
cal tooth, tergite X with one tooth on each postero-
lateral corner, sternite IX with posterior projection
broadly rounded; female sternite VII broad with a
posteriorly projecting rectangular lobe, sternite
VIII membranous C. dobryi

lb Wings brachypterous; male with epiproct with one
dorsal hook, tergite IX with single sclerotized pos-
terior plate bearing numerous teeth, cerci without
apical tooth, tergite X with two teeth on each pos-
terolateral corner, sternite IX with posterior pro-
jection broadly angulate; female sternite VII elon-
gate with lateral constrictions, sternite VIII with
two arcuate sclerotized plates C. ephemera

Calileuctra ephemera, new species

Figs. 1-9

Male. — General color brown; dark brown
pattern as illustrated (Fig. 1). Length of body
4.5 mm. Brachypterous, length of forewing
2.5-3.0 mm; wings light brown, venation simi-
lar to the genus Perlomyia (Fig. 3). Frothoracic
basisternum triangular in shape (Fig. 2).
Abdominal tergum IX with posterior border
complete, projecting and serrate; tergum X

126

Great Basin Naturalist

[Volume 55

->^

"imtm-^-r-^

'm^^jjp

Fig. 1. Calileuctra ephemera. Habitus.

incomplete medially, posterolaterally with two
elongate projections (Figs. 4, 5). Sternum IX
with basal pear-shaped vesicle; posterior bor-
der extending to base of subanal probe (Fig.
6). Cerci extending beyond genitalia, with api-
cal membranous area expanded and flattened
(Figs. 4, 5, 6). Epiproct small and hook-shaped
(Fig. 5). Subanal probe large, elongate, both

membranous and sclerotized, expanded near
apex (Figs. 8, 9).

Female. — General color and wing vena-
tion similar to male. Brachypterous, length of
forewing 3.5-4.0 mm. Sternum VII constrict-
ed laterally, projecting slightly over sternum
VIII; sternum VIII reduced to 2 small arcuate
sclerotized plates (Fig. 7).

1995]

Calileuctra New Genus

127

Figs. 2, 3. Calileuctra ephemera: 2, ventral view of thorax; 3, wings.

Nymph. — Body lightly sclerotized; light in
color; setation sparse except on labrum, legs,
and cerci; size small — 7.2 mm long. Head
slighth' broader than thoriix; color pattern faint.
Mouthparts of the herbivorous/detritivorous
type [Type I (Stewart and Stark 1988)]. Labrum
and clypeus with numerous long setae. Man-
dible typical for Leuctridae: 4 dorsal cusps (2
large, 2 small), and 1 small ventral cusp on
side of first large dorsal cusp; bowl-shaped
molar region, with transverse ridges in the
"bowl," and with a pectinate scraping ridge on
the ventromedial edge. Maxillary palpi 5-seg-
mented. Labial palpi 3-segmented; glossae
and paraglossae short and subequal in size
(paraglossae slightly larger). Pronotum quad-
rangular; transverse anterior and posterior
sclerotized bands; median longitudinal suture
unsclerotized; color pattern weak. Mesonotum
with two sclerites; anterior sclerite transverse
and roughly trapezoidal; posterior sclerite
roughly U-shaped. Metanotum like mesono-
tum. Wing pads three or more times as long as
wide; posterior wing pads a little shorter than
anterior wing pads; longitudinal axes of wing
pads almost parallel to axis of body. Prostenium
naked and membranous except for two small
sclerites between the coxae; sclerites forming

a posteriorly directed U-shape. Mesosternum
narrowest anteriorly, widest by coxae; mem-
branous except for weak sclerotization of the
furcal pits and the Y-ridge; Y-ridge with faintly
sclerotized double stem, arms connecting to
posterior ends of furcal pits; transverse ridge
connecting anterior ends of furcal pits.
Metastemum similar to mesosternum; sclero-
tization only in a transversely rectangular area
limited by the furcal pits, a transverse ridge
connecting the anterior ends of the furcal pits,
and the area between the pits and the ridge.
All legs similar but increasing in size posteri-
orly; setation consists of abundant ver>^ small
setae ("clothing hairs") and sparse longer
setae; tibiae and femora with setal fringes;
apex of tibiae with a pair of spines; tarsi 3-seg-
mented, first segment short and conical, sec-
ond veiy short, ringlike with apex cleft, third
elongate and cylindrical, suture between first
and second tarsomeres very narrow and hard
to see; tarsomeres with ventral pad of numer-
ous fine setae; tarsal claws slender. Abdominal
terga very weakly sclerotized; setation sparse
except on end of tenth segment. Abdominal
fold present on segments I-VIL Subanal lobes
incompletely fused. Cereal segments with api-
cal fringe of 10-15 setae; setae about one-half
length of the segments.

128

Great Basin Naturalist

[Volume 55

M!>%M.W'§

Figs. 4-7. Calileuctra ephemera: 4, male terminalia, dorsal view; 5, male terminalia, lateral view; 6, male terminalia,
ventral view; 7, female terminalia, ventral view.

1995]

Calileuctra New Genus

129

Figs. 8, 9. Calileuctra ephemera: 8, male subanal probe,
right lateral view; 9, male subanal probe, ventral view.

Egg. — Shape oval; size uniform, 0.133 mm
in length, 0.095 mm in width. Surface coarsely
rugose; large, coarse punctures present in an
irregular distribution.

Type locality. — California: Napa Co.,
3.36 km (2.1 mi) N on Hw\' 128 from the inter-
section of Hwy 128 and Hwy 121, unnamed
tributary to Capell Creek (ca 300 m [275 ft]
elevation).

This intermittent stream has water present
only a few months each year; some years there
is no water (i.e., 1987 and 1990). When water
is present, it flows down a small, steep canyon,
across a grassy flat, under Hwy 128, and down
a short cliff into Capell Creek. The stream
course appears to be less than 350 m (1000 ft)
long. All specimens have been collected in the
grassy flat or just downstream. The stream
course has a substrate of either bedrock or rocks
on a clay soil. There is no obvious hyporheic
zone. Detrital input is usually leaves from
trees (mainly live oak), grass, and star-thistle.

Type specimens. — Holotype: male, type
locality, 19 II 1983, WDS-A-160, reared from
nymph. To be deposited in the entomology
collection at the California Academy of
Sciences, San Francisco, CA. Allotype:
female, type locality, 25 II 1984, WDS-A-240.
To be deposited with the holotype. Paratypes:
1 male, type locality, 18 II 1984, WDS-A-234,
reared from nymph (deposited at Monte L.
Bean Life Science Museum, Brigham Young
University, Provo, UT); 3 females, type locality,
18 II 1984, WDS-A-234 (deposited with male
paratype). Additional specimen: 1 nymph,
type locality, 27 II 1988, WDS-A-527 (deposit-
ed with the holotype and allotype).

Etymology. — The trivial name was select-
ed to indicate the temporary nature of the
stream at the type locality, and also to indicate
the difficulty encountered when trying to col-
lect specimens.

Biology. — All specimens were collected
during the last two weeks of February, when
the stream was flowing. Streams in this area of
the coastal mountains experience a Mediter-
ranean climate with a December-to-Februaiy
rainy season. Local intennittent streams usually
have surface flow only from January through
March.

When first collected, all specimens were
either late-instar nymphs (3) or adults (4). All
field-collected adults (4 females) were swept
from vegetation overhanging the stream. Two
of the three nymphs collected were held in
styrofoam containers until they molted to the
adult stage. Both individuals were males.

The bowl- shaped molar region of the man-
dibles is similar to molar modifications found
in beetle lawae that feed on fungal tissues, in
general, and fungal spores, in particular
(Lawrence 1977, Lawrence and Hlavac 1979,
Lawrence and Newton 1980). Since fungal

130

Great Basin Naturalist

[V'olume 55

tissues are high in protein (Martin 1987), use
of them as a food would aid the fast growth
and development of nymphs.

Present information suggests that Calileuc-
tra ephemera has a faeultati\ely long egg cha-
pause, very fast nymplial deveh)pment, and
short stadium for liotli nymplis and adults.
The high protein content of fungal tissues
(Martin 1987) may aid in the fast growth and
development of nymphs of this unique species.
These characteristics are similar to those of
Zealeiictra (Snellen and Stewart 1979), an
eastern North American genus and another
inhabitant of intermittent streams.

Calileuctra dohryi, new species

Fi^js. 10-14

Male. — General color brown; dark pattern
as illustrated (Fig. 10). Length of body 4.0-6.0
mm. Macropterous, length of forewing 4.5-5.5
mm, wings light brown, venation similar to the
genus Perlomijia. Tergum IX with membra-
nous median band dividing tergum into two
sclerotized halves, each half bearing a small,
nipplelike projection and a large, earlike pos-
terior projection. Tergum X also divided into
two halves, each half with a gently rounded,
knoblike lobe and an enlarged, lateral posteri-
or lobe which ends in a sclerotized prong that
extends about one-third the way up the cercus
(Figs. 11, 12). Sternum IX broadly rounded
posteriorly, extending only to base of subanal
probe, large vesicle present at median anterior
margin, vesicle with truncate apex (Fig. 13).
Cerci enlarged, elongate, extending beyond
genitalia posteriorly, sclerotized on lateral
margins, apex rounded, ending in a sclerotized
lateral prong (Figs. 11, 12, 13). Epiproct with
narrow bifurcate apex (Fig. 11). Subanal probe
large, elongate, broadest medially, apex pointed
(Fig. 13).

Female. — General color and wing vena-
tion similar to male. Length of body 5.0-6.0
mm. Macropterous, length of forewing 5.0-6.0
mm. Abdominal sternum VII enlarged,
expanded slightly over VIII; posteromedial
area formed into a narrow, medially roiuided
lobe. Sternum VIII small and only lightly scle-
rotized (Fig. 14).

Nymph. — Unknown.

Egg. — Unknown.

Type locality. — California: Los Angeles
Co., South Fork Elsmere Canyon, San Gabriel
Mountains. The type locality is a very small

headwater tributaiy of Elsmere Creek. It has
an extremely steep gradient and is hard to
access. Thus, the habitat has been presented
more than the surrounding drainage area.

Type specimens. — Holotype: male, tvpe
locality, 22 IV 1991, K. F Dobry. To' be
deposited in the entomology collection at the
California Academy of Sciences, San Francisco,
CA. ALLOTi'PE: female, same data as holotype.
To be deposited with holotype. Paratypes: 1
male and 1 female, same data as holotype; 2
males and 1 female, CA: Los Angeles Co.,
Santa Monica Mountains, East Fork Arroyo
Sequit, 5 mi NW Pacific Coast Highway off
Mulholland Highway 28 II 1992, K. F Dobiy;
2 females, CA: Orange Co., Santa Ana Moun-
tains, Trabuco Canyon, 1300', 11 1 1988, R. W.
Baumann, B. J. Sargent, B. C. Kondratieff, and

C. R. Nelson; 3 males and 1 female, CA: Los
Angeles Co., San Gabriel Canyon, 23 IV 1960,

D. Gibbo (LACM). Remaining paratypes to be
deposited at Monte L. Bean Life Science
Museum, Brigham Young University, Provo,
UT

Etymology. — The trivial name honors
Keith E Dobiy, Los Angeles, CA, who collect-
ed many of the specimens.

Biology. — Specimens were collected as
adults between January and April. All popula-
tions are from the Los Angeles basin and
experience a Mediterranean climate.

Phylogeny

Leuctrid phylogeny has been examined
from a cladistic point of view only two times.
The two studies (Ricker and Ross 1969, Nelson
and Hanson 1973) are somewhat contradictoiy.
However, examination of the analysis given in
both studies shows Calileuctra to possess many
character states that are termed primitive or
ancestral. Following Nelson and Hanson's
more comprehensive analysis, the character
states present in Calileuctra are as follows: 1-0,
2-0, 3-0, 4-0, 5-0, 6-0, 7-1, 8-0, 9-0, 10-2, 11-0,
12-0, 13-0, 14-2, 15-0, 16-0, 17-0, 18-0,19-0,
20-0, 21-1, 22-0, 23-0, 24-1, 25-1, 26-1, 27-1,
28-1, 29-0, 30-1, and 31-2 (first number =
character; second number = character state).
See Nelson and Hanson (1973) for a key to the
characters and character states. Character
states for Calileuctra and those cited in Nelson
and Hanson (1973) for other leuctrid genera
were run through the PAUP 3.1.1 program

1995]

Calileuctra New Genus

131

Fig. 10. Calileuctra dobryi. Habitus.

132

Great Basin Naturalist

[Volume 55

Figs. 11-14. Calileuctra dohnji: 11, male terminalia, dorsal view; 12, male terminalia, lateral view; 13, male terminalia,
ventral view; 15, female terminalia, ventral view.

1995]

Calileuctra New Genus

133

Ancestor

Moselia

Leuctra

Pachyleuctra

Despaxia

Paraleuctra

Zealeuctra

Rhopalopsole

Perlomyia

Calileuctra

Tyrrhenoleuctra

Megaleuctra

Fig. 15. Phylogeny of the Leuctridae.

using the branch and bound algorithm with all
characters unordered. This analysis found one
minimum-length tree (Fig. 15) with a length of
68, a consistency index of 0.82, and a retention
index of 0.80.

This new tree is not considerably different
from that given by Nelson and Hanson (1973).
It differs only in the collapse of the sister
group relationship between Rhopalopsole and
Zealeuctra and the exclusion of Euleiictra and
"Leuctra" divisa from consideration. The sta-
bility of diis tree with Calileuctra added is taken
as evidence of the consistency of this data set
and the overall stability of this new classifica-
tion. It is heartening to find the cladogram of
Nelson and Hanson (1973) stable despite the
previous "extinction" (i.e., absence) o( Calileuc-
tra. In tliis particular case, an "extinct" taxon did
not particularly influence the overall topology
of the cladogram. Hence, there is hope in our
search for relationships among living taxa
despite "known" extinction events.

In this tree, Calileuctra is a the sister-taxon
of the group containing Perlomyia, RJiopalop-
sole, Zealeuctra, Paraleuctra, Despaxia, Pachy-
leuctra, Leuctra, and Moselia. In leuctrid phy-
logeny, Calileuctra occupies a near basal posi-
tion and as such gives an important addition to
our knowledge of the group.

Acknowledgments

Many thanks go to Charles H. Nelson, Uni-
versity of Tennessee at Chattanooga, for run-

ning phylogeny programs for us and for his
many helpful comments. C. Riley Nelson, Uni-
versity of Texas, Austin, provided a review and
made valuable suggestions. Boris C. KondratiefF,
Colorado State University, also offered many
helpful suggestions as well as helped collect
specimens. The late Charles L. Hogue kindly
loaned specimens fiom the Los Angeles County
Museum (LACM). Keith F. Dobry helped
greatly in the collection of additional speci-
mens. Jean A. Stanger made the many excel-
lent illustrations.

Literature Cited

Harper, P. P., and K. W. Stewart. 1984. Chapter 13.
Plecoptera. In: R. W. Menitt and K. W. Cummins,
editors. An introduction to the aquatic insects of
North America. Kendall/Hunt Publishing Co.,
Dubuque, lO. 722 pp.

Lawrence, J. E 1977. The family Pterogeniidae, with
notes on the phylogeny of the Heteromera. The
Coleopterists' Bulletin 3i: 25-26.

Lawrence, J. E, and T. E Hlavac. 1979. Review of the
Derodontidae (Coleoptera: Polyphaga) with new
species from North America and Chile. The
Coleopterists' Bulletin 33: 369-414.

Lawrence, J. E, and A. E Newton. 1980. Coleoptera
associated with the fruiting bodies of slime molds
(Myxomycetes). The Coleopterists' Bulletin 34:
129-143.

Martin, M. M. 1987. Invertebrate-microbial interactions.
Cornell University Press, Ithaca, NY. 148 pp.

Nelson, C. H., and J. E Hanson. 1973. The genus Per-
lomyia (Plecoptera: Leuctridae). Journal of the Kansas
Entomological Societ>' 46: 187-199.

RiCKER, W. E., and H. H. Ross. 1969. The genus
Zealeuctra and its position in the family Leuctridae

134 Great Basin Naturalist [Volume 55

(Plecoptera: l^uctridaej. Canadian Jonnial ofZoology STEWART, K. W., AND B. P Stark. 1988. Nymphs of North

47: 1113-1127. American stonefly genera (Plecoptera). Thomas Say

Snellen, R. K., and K. W. Stewart. 1979. The life cycle Foundation, Entomological Society of America 12:

and drumming behavior of Zealeuctra claasseni 1-460.
(Prison) and Zealeuctra hitei Ricker and Ross

(Plecoptera: Leuctridae) in Te.xas, USA. Aquatic Received 27 Septe)nber 1994

Insects 1:6.5—89. Accepted 17 January 1995

CARBON ISOTOPE DISCRIMINATION IN THE C4 SHRUB ATRIPLEX
CONFERTIFOLIA ALONG A SALINITY GRADIENT

Darren R. Sandquistl and James R. Ehleringer^

Abstract — Carbon isotope discrimination (A) was measured for leaves ofAtriplex confertifolia along a salinity gradi-
ent in northern Utah. Over this gradient, the variation of A values was high for a C4 species, and the A values were posi-
tively correlated with salinity in both years of the study. Of the possible explanations for this pattern, the A results are
consistent with tlie notion that salinity' induces an increase in the bundle sheath leakiness of tliese C4 plants.

Key tvords: carbon isotope ratio, salt stress, bundle sheath leakiness. halophyte. desert ecology.

The analysis of carbon isotope ratios
(I'^C/l^C) has become a useful tool for under-
standing various integrated aspects of plant
metabolism, including numerous investiga-
tions of plant-environment interactions. The
impact of environmental factors on carbon iso-
tope discrimination (A) by plants with C3 pho-
tosynthesis has been well studied; however,
only a limited number of studies have exam-
ined variation of A values in C4 plants
(O'Leary 1988, Farquhar et al. 1989, Peisker
and Henderson 1992). In part, this disparity
stems from C4 plants having much smaller
variation of A values than C3 plants.
Additionally, A values in C3 plants have been
correlated with water-use efficiency, and this
has lead to an emphasis on applying carbon
isotope analyses to breeding programs
(Farquhar et al. 1989, Ehleringer et al. 1993).
However, a few recent studies have demon-
strated that variation of A values in C4 plants
may reflect environmental influences on phys-
iological function (Bowman et al. 1989,
Meinzer et al. 1994). In this study we exam-
ined variation of A values in a C4 perennial
shrub, Atriplex confertifolio (Torr. & Frem.)
Wats., and its relationship to natural condi-
tions of soil salinity.

The A value of a C4 plant integrates two
factors that can impact productivity: (1) the
ratio of intercellular to ambient CO2 concen-
tration (Cj/cJ, which can reduce photosynthe-
tic activity when low, and (2) bundle sheath
leakiness (0), which reduces photosynthetic
efficiency when high. Farquhar (1983) mod-
eled the relationship between these factors

and carbon isotope discrimination in C4 plants
as

A = o + (Z?4 + Z?30 - a) c^/c.^^.

(1)

where a (4.4%c) is discrimination against the
heavier ^'^C02 molecule relative to the lighter
1^C02 based on differential rates of diffusion,
Z?3 (29%c) is the discrimination due to a
greater affinity for 12CO2 relative to ^'^C02 by
ribulose bisphosphate carboxylase (Rubisco),
and b^ (typically = -5.2%c) is discrimination
based on the steps leading to, and including,
CO2 fixation by phosphoenol pyruvate car-
boxylase (PEPC) after atmospheric CO2
enters the leaf The b^ term varies slightly as a
function of temperature and is negative
(greater proportion of l'^C02) due to fractiona-
tion associated with the hydration of CO2 to
HC03~ (Mook et al. 1974). The discrimination
terms of Equation 1 [a, b^, and b_^) are con-
stants, for the most part, and thus differences
among A values are the result of changes in 0
and/or c^/c.^ during CO2 assimilation.

In C4 plants, COg is'^initially fixed by PEPC
in the mesophyll cells, transported and decar-
boxylated in the bundle sheath cells, and then
refixed by Rubisco. However, before the assimi-
lation by Rubisco a fraction of the CO2 may
diffuse out through apoplastic portions of the
bundle sheath cells. This is known as "leaki-
ness" and is thought to be reduced by suber-
ization of bundle sheath surfaces (Farquhar
1983). This leakiness, however, may be in-
creased by environmental stresses, such as
salinity (Bowman et al. 1989), and an increase

' Department of Biology, UniversiU' of Utah, Salt Lake Cit\-, UT 84112

135

136

Great Basin Naturalist

[Volume 55

in leakiness represents an energetic cost to the
plant as a result of incomplete carbon assimi-
lation or overcycling (Ehleringer and Pearcy
1983, Jenkins et al. 1989, Henderson et al.
1992).

Leakiness affects A because it causes the
bundle sheath cell to become an open system
and therefore allows expression of discrimina-
tion by Rubisco (^3). The proportion of CO2
that leaks out of the bundle sheath cell (0)
modifies the degree to which b^ is expressed
and thereby determines the relationship
between A and c^/c.^ (Eq. 1). At low 0 values
the relationship between A and c^/c.^ is nega-
tive, at high 0 the relationship is positive, and
at 0 = 0.32, A is constant at 4.4%o regardless of
Ci/c.^. Equation 1 also predicts that for any
given Cj/c^, an increase in 0 results in an
increase in A. Given these relationships, varia-
tion of A values in G4 plants can provide an
indication of bundle sheath leakiness and its
relationship to environmental stresses.

To date, much work investigating variation
of A in C4 plants has come from either labora-
toiy gas exchange studies (Evans et al. 1986,
Bowman et al. 1989, Henderson et al. 1992) or
theoretical models (Peisker 1982, Farquhar
1983, Peisker and Henderson 1992). There is
little direct information on environmental
stresses that influence A in natural popula-
tions of C4 plants (except see Walker and
Sinclair 1992). Here we report on changes in
A values for the G4 species Atriplex confertifo-
lia found along a natural salinity gradient in
Utah. The purpose of this study was to deter-
mine if A values changed in relation to soil
salinity under field conditions, and if these
changes corresponded to variation in 0 values.
Two previous laboratory studies have shown
that higher soil salinity does increase A values
in G4 plants and that this change is a result of
greater 0 (Bowman et al. 1989, Meinzer et al.
1994). For A. confertifolia, we hypothesized
that the same trend would be found over a
transect of naturally increasing soil salinity.

Methods

Study Sites

Four study sites of increasing salinity were
chosen along a south-to-north transect in the
northern end of Skull Valley (Tooele Gounty,
UT) flanking the western slope of the Stans-

buiy Mountain Range. The four sites range in
elevation fiom 1366 m to 1286 m (Fig. 1). Site 1
(1366 m) is dominated by sagebiiish {Atiemisia
tridentata) with low densities oi Atriplex con-
fertifolia, Jiiniperus osteosperma, and Tetrady-
mia spinosa. Weedy grasses and annual species
of the Ghenopodiaceae are also found within
disturbed areas of this and all other sites.
Greasewood {Sarcohatus verrniculatus) is the
dominant species at sites 2 (1317 m) and 3
(1294 m) with A. confertifi)lia co-occurring in
low frequency. Site 4 (1286 m), along the mar-
gins of the salt flats, is a heterogeneous site
with a mixed community of salt-tolerant
species. S. vermiculatits is the dominant
species with moderate densities of Allenrolfea
occidentalis, Atriplex gardneri, A. confertifolia,
Chnjsothaninus viscidiflorus, Kochia ameri-
cana, and Suaeda torreyana.

Weather data for this transect are taken
from the Grantsville weather station (Grants-
ville, Tooele County, UT, 1307 m) located 17.3
km E and 8.2 km S from the center of our
study transect.

Leaf and Soil Samples

Leaves of Atriplex confertifolia and soil
samples were collected from each of the four
transect sites in October 1991 and 1992, with
the help of the 1991 and 1992 Plant Ecology
classes from the Universit>' of Utah. Recently
matured leaves of A. confertifolia were collect-
ed from five to eight individuals per site in
1991 and three per site in 1992. Leaf samples
were oven-dried (70 °G, 7 d), ground with
mortar and pestle, and analyzed for carbon
isotopic composition (Windy Ike, Delta S mass
ratio spectrometer, Finnigan-MAT, San Jose,
GA) relative to the Pee Dee Belemnite stan-
dard. Analyses were done at the Stable
Isotope Ratio Facility for Environmental
Research (SIRFER, University of Utah, Salt
Lake Gity, UT). Garbon isotope ratio values (6)
were transformed to discrimination (A) values

^ = (Sa-5pMl + Sp)

(2)

where 5p is the measured carbon isotope ratio
of the plant, and 5^ is the carbon isotope ratio
of GO2 in the atmosphere (-.008 or -8%c;
Farquhar et al. 1989). The standard per mil
{%c) notation is used throughout for ease of
presentation, and the overall, long-term error

1995]

Carbon Isotope Discrimination in Atriplex

137

10 15 20 25 30

Transect distance from site 1 (km)

Fig. 1. Study transect in cross section. Shown is the topography over the transect and locahties of each study site
based on the appro.ximate hnear distance from site 1.

associated with carbon isotope determination
is±0.11%o.

Soil samples were collected from two depths
(15-20 cm and 40-60 cm) in two to six excava-
tion pits at each site. Approximately 200 g of
freshly extracted soil from each hole and depth
was placed immediately into soil canisters,
sealed, and kept cool until analysis in the labo-
ratory. In the lab one subsample per canister
was removed for salinity' analyses. The remain-
ing soil was used for gravimetric water content
detennination based on the difference between
soil fresh (wet) weight and dry weight (i.e.,
water content) relative to the soil dry weight.
Soils were dried at 70°C for 7 d.

In 1991 the soil salinity analysis was based
on electrical conductivity (EC) of an aqueous
solution extracted from a 1:2 soihdeionized
water mixture, and in 1992 from a 1:5 soil:
deionized water mixture. There was no evi-
dence that the 1:2 mixture was ion saturated;
thus, to standardize these ratios, the ECs of
samples using a 1:2 solution were extrapolated
to EC based on a 1:5 ratio assuming a linear
dilution relationship. Tests confirmed that this
extrapolation was valid even for EC values
higher than those found in actual field samples.

Although a more standard procedure for
salinity determination is the "soil paste "
method, the 1:5 ratio method we used is rec-
ommended as a simpler technique to deter-
mine relative salinity contents (Rhoades 1982)
and is suitable for the purposes of this study
(i.e., standardized comparison of relative salin-
ities among sites). Additionally, the ECs of 1:5

ratio extracts are highly correlated with soil
paste ECs for soils within and near our tran-
sect (D. G. Williams unpublished data).
Electrical conductivity is reported in ^mhos
cm~l (1 jUmhos cm"^ = 0.1 mS m"^ = 0.502
mM NaCl), and the data were log transformed
for statistical analyses. Interannual compar-
isons of means for each soil trait were done by
t tests, and correlations between soil trait and
plant carbon isotope discrimination means
were determined by Pearson product-moment
correlation.

Results

Transect Characterization

Salinity increased across the gradient in
both the 1991 and 1992 samples; electrical
conductivity increased by two orders of mag-
nitude over the entire transect (Table 1). Site 1
was the least saline, and salinity progressively
increased toward the highly saline site 4.

There were few differences between years
in soil electrical conductivity. Significant dif-
ferences were found at only two sites and at
only one depth per site. Furthermore, sites
gave opposite results: soils of site 3 at the
15-20-cm depth had greater conductivity in
1991 than 1992 {t = 4.33, P < .01), and soils
from site 1 at the 40-60-cm depth had higher
conductivity in 1992 than in 1991 {t = 4.60, P
< .01).

Cravimetric water content also increased
over the transect from site 1 to site 4 (Table 1).
Soil water content was somewhat greater in

138

Great Basin Naturalist

[Volume 55

Table 1. Soil propfrtie.s at two tleptli.s for site.s 1—4 along tlic .study transect (» = number of pits; one sample for each
depth per pit). Soil water content was measured as gravimetric water content, and electrical conductivity is of an aque-
ous extract from 1:5 soihwater mi.xture (extrapolated for 1991 from 1:2 ratio; see te.xt).

Electrical

ectrical

Soi

1 water

Soi

1 water

conductivity

com

ductivity

coni

tent (%)

content (%)

(/Ainhos/cm)

(jun

ihos/cm)

@ 15-20 cm

@40-60 cm

@ 15-20 cm

@40-60 cm

Mean

Mean SE

Mean

October

1991

Sitel

4.66

0.300

5.34

0.234

89 15.7

3.2

Site 2

4.15

0.687

7.23

0.360

91 7.6

324

81.1

Site 3

11.79

1.446

17.24

0.892

2309 114.1

2066

657.7

Site 4

24.84

7.578

.39,41

7.841

3596 587.6

3382

530.7

October 1992

Sitel

2.89

0.454

3.81

0.402

84 6.3

3.7

Site 2

4.79

0.226

5.56

0.499

144 29.2

324

111.0

Site 3

2.46

0.270

10.26

3.672

546 459.3

984

858.5

Site 4

10.66

0.950

—

1640 1440.0

3250

350.0

1991 than in 1992, but significant differences
at both depths were found only at site 1
(15-20-cm depth, t = 3.34, P < .05; 40-60-cm
depth, t = 3.52, P < .05). Rainfall over the 10-
wk period prior to sampling in 1991 was much
greater than that of 1992 (82.5 mm vs. 18.8
mm), which likely accounts for the trend of
greater water content in the soils during the
1991 sample period.

Carbon Isotope Discrimination

Along the transect the carbon isotope dis-
crimination for Atriplex confertifolia ranged
from a low of 4.74 ± 0.96%^ at site 1 in 1992,
to a high of 6.55 ± 0.1 l%c at site 3 in 1991
(Fig. 2). This range of nearly 2%c is high for C4
plants (Farquhar et al. 1989). The mean A
value was always greater than 4.4%c, and for
only a single sample was the individual shrub
value less than 4.4%c. These high A values
indicate that the mean 0 values were always
greater than 0.32 (Eq. 1).

With respect to the environmental parame-
ters examined along the transect, mean leaf A
was not significantly correlated with water
content during any obsei-vation, but was posi-
tively correlated with log EC (Fig. 2).
Inclusion of the notably low A value of site 4
in 1991 resulted in a nonsignificant, positive
trend (but when excluded, A was significantly
correlated with log EC in 1991 at the deeper
soil depth, R = 1.0, P < .01). In 1992 there
was a highly significant, positive correlation ol
A and log EC for both the shallow soils {R =
.978, P < .05) and deeper soil depths (R =
.999, P < .001) (Fig. 2).

Discussion

Variation in carbon isotope discrimination
values of C4 plants is, in part, dependent upon
the proportion of CO2 that is initially fixed by
PEPC and ultimately diffuses out of the bun-
dle sheath cells without being refixed (i.e., the
leakiness, 0). Leakiness might be influenced
by environmental stresses, such as salinity
(Bowman et al. 1989, Meinzer et al. 1994),
because such stresses could disrupt mem-
brane properties or the biochemical coordina-
tion between the C4 and C3 cycles operating
in the mesophyll and bundle sheath cells,
respectively (Peisker and Henderson 1992).
The other component influencing variation of
A in C4 plants is Cj/c^. Figure 3 illustrates how
the relationship between A and c^/c^ depends
upon the value of 0 (from Eq. 1), and provides
a model for how changes in 0 and c^/c.^ can
account for the changes in A values we
observed.

We found that A values of A. confertifolia
increased by 2%c in concordance with increas-
ing salinity (Fig 2). These A values were
always greater than 4.4%o; therefore the 0 val-
ues must be greater than 0.32 (cf Fig. 3). A
2%c increase in A values, at 0 > 0.32, cannot
be explained solely by changes in c^/c.^^ given
the typical range of c^/c.^^ values for C4 plants
under ambient conditions (0.20-0.40; Pearcy
and Ehleringer 1984). To do so would require
either extreme leakiness values (0 > 0.6) or an
increase of c\/c.^ with increasing salinity since
A and Cj/c.j are positively related when 0 >
0.32. Leakiness values greater than 0.6 have

1995]

Carbon Isotope Discrimination in Atriplex

139

■■E
©

o
o

6.5

Soil depth = 15 -20 cm

•

i\ 1

H -L

+ C-

5.5

•

4.5
3.5

Soil depth = 40-60 cm

i-in

6.5

-^ ^•l

;

0—\ -L
^ -r

•

5.5

l-o^

4.5

-a c

100

1000

10000

Log Soil Electrical Conductivity
(^imhos cm"'')

Fig. 2. Relationship between carbon isotope discrimina-
tion (A) of Atriplex conferfifolia leaves and log electrical
condnctivit>' (log EC) of soil at Kvo depths, 15-20 cm and
40-60 cm, for sites 1-4 along the transect. Closed symbols
(•) are study site means for 1991, and open s\'mbols (O)
are those for 1992. Error bars are ± ISE.

never been reported, and the latter explana-
tion is unlikely since salt stress typically
decreases or does not change c^/c.^ (Long and
Baker 1986, Flanagan and Jefferies 1988). A
simpler explanation for the change in A values
is that 0 increases with higher salinity. A 2%c
increase based on changes in 0 values can be
easily accommodated within the limits of Cj/c^
found for C4 plants (Fig. 3). Thus, changes in
A values for A. conferfifolia are more likely
due to an increase of 0 associated with a
change in salinity; consequently, the presence
of a significant relationship between A values
and EC (Fig. 2).

The trend of increasing A values with
increasing salinity held in all but one site in
the two-year study (site 4 in 1991). This devia-

9.0
8.0
7.0
6.0

< 4.0

3.0

2.0

1.0

' //'

-^' /^

0 = 0.6/ /

/•

/ / y

-^ 0 = 0.5

y
^

/y ^ -^

^^.-'--^^

/ y^" ^

^^^^

-""^ 0 = 0.4

y /> ^^---^

, ■

/V ^^^-^

^^>K^ ^„— — "^ ■ — —

- J€^

--^ir^—"

■ ^o^^>^

"--- — ____

;:>^^ ~~~~--

0 = 0.3'

\ ^^ ~^

-^^^^

^^"-~-^,^^^

\ \

■^ \0 = 0.2^

\ \

0 = 0.1 \ \

^,.^

0.2

0.4 0.6 0.8 1

C/Ca

Fig. 3. Model for the relationship between carbon iso-
tope discrimination (A) and Cj/c^ (ratio of intercellular to
ambient COo) based on Equation 1 and for 0 values rang-
ing from 0.1 to 0.6. Dashed and solid lines represent the
range of A values for each 0 value depicted, based on a
high leaf temperature (34°C) where ^4 = -4.8%c (solid
line) and a lower leaf temperature (25 °C) where h^ —
-5.77cc (dashed line).

tion could be due simply to the high degree of
edaphic variability at site 4; this location had
the greatest topographic variability, highest
species diversity, and greatest overall variance
for soil conductivity and water content (Table
1). Site 4 was also extremely wet in 1991 (near
40% water content at 40-60-cm depth), which
may have diluted the salinity of these soils,
thereby reducing the salinity experienced by
the plants. Without a more detailed study, how-
ever, this deviation remains unexplained.

Previous studies have found contrasting
patterns of the relationship between A and
salinity. In a laboratory study with 11 C4
species, Henderson et al. (1992) found that 0
values were invariable and low, remaining at 0
~0.21, thereby resulting in a negative rela-
tionship between A and c^/Cg^ (Fig. 3). The
small variation they observed in A values was
attributed to changes in Cj/c^ values. However,
in an earlier study with the C4 monocots Zea
mays and Andropogon glomeratus. Bowman et
al. (1989) found that A values of salt-stressed
plants were more dramatically influenced by
changes in c^/c.^ than were control plants. The
increase of A values with salinity was ex-
plained by a changing relationship between A

140

Great Basin Naturalist

[Volume 55

and Cj/c.j due to increasing 0 values as the
water status of salt-stressed plants declined
through the day (Bowman et al. 1989).

Recently, Mcinzer et al. (1994) also obseived
that increasing salinit)' resulted in increases of
A values. Using two sugarcane cultivars, they
showed that change in A value could be
ascribed to greater 0 values as salinity in-
creased, and that variability of Cj/c., had much
less impact on the increase of A values. In
contrast, Walker and Sinclair (1992) reported
that A values of two Australian C4 Atriplex
species decreased at sites with increased salin-
ity. The A values of these Australian Atriplex
leaves were greater than 4.4%o, which could
have been achieved only with a bundle sheath
leakage greater than 0.32 (Fig. 3). Since the
relationship between A and c^/c..^ is positive at
0 > 0.32 (Fig. 3), the Walker and Sinclair data
suggest that salinity affected a decrease of Cj/c.^
and, therefore, a decrease of A.

Our findings of a positive correlation be-
tween A values oi Atriplex confertifolia and
salinity are in contrast to findings of Walker
and Sinclair (1992). Our observations, like
those of Bowman et al. (1989) and Meinzer et al.
(1994), suggest that changes in leaf carbon iso-
tope discrimination result from an increased
bundle sheath leakage when plants are exposed
to a salinity stress. The mechanism of change
in 0 values is likely to be associated with phys-
ical changes in the bundle sheath permeability
to CO2 (or to HC03~) and/or biochemical
changes in the coupling of Rubisco and PEPC
activity. Such biochemical changes due to
salinity have been previously found. Guy and
Reid (1986) have shown that salinity may
reduce Rubisco activity in C3 plants without a
concomitant decrease in PEPC activity.
Increased salinity (NaCl) has also been shown
to increase PEPC activity in some C4 halo-
phytes (Shomer-Ilan et al. 1985). Any such
increase in the activities of C4 carboxylation
enzymes relative to those of C3 carboxylation
enzymes in C4 plants should increase 0 values
(Peisker and Henderson 1992). Thus, under
natural conditions it appears that salinity
could increase A values of A. confertifolia by
influencing an increase in 0 values.

The relationship between salt stress and 0
of C4 plants may be species specific or even
population specific and may account for dis-
crepancies among different studies of A values
in C4 plants. For example, there is high vari-

ability among pre\'ious studies of carbon iso-
tope discrimination in Atriplex confertifolia;
mean A values range from 4.4%c (Marino et al.
1992) to 6.9%c (Troughton et al. 1974). Yet,
each of these observations is consistent with
the notion that 0 values exceed 0.32 and are
therefore high compared to nonhalophytic C4
species (Henderson et al. 1992).

In the present study we have shown that
salinity may be one factor that significantly
influences variation of A values in C4 plants,
most likely through an effect on bundle sheath
leakiness. While variation in A values of C4
plants may provide new insights into plant-
salinity dynamics along environmental gradi-
ents, results also suggest that caution is neces-
saiy when using A values of C4 plants to inter-
pret historical changes in atmospheric CO2
concentrations and ^'^C values, as has been
proposed by Marino et al. (1992).

Acknowledgments

We thank University of Utah students in
1991 and 1992 Plant Ecology classes for assis-
tance in sample collection, Craig Cook for
assistance in carbon isotope analyses, and Dr.
David Williams for salinity analyses compar-
isons. Dr. Williams and two anonymous
reviewers also provided helpful comments on
a previous version of this manuscript.

Literature Cited

Bowman, W. D., K. T. Hubick, S. von Caemmerer, and
G. D. Farquhar. 1989. Short-term changes in leaf
carbon isotope discrimination in salt- and water-
stressed C4 grasses. Plant Physiologv 90: 162-166.

Ehleringer, J., and R. W. Pearcy. 1983. Variation in
quantum yield for CO2 uptake among C3 and C4
plants. Plant Physiolog\' 73; 555-559.

Ehleringer, J. R., A. E. Hall, and G. D. Farquhar.
1993. Stable isotopes and plant carbon-water rela-
tions. Academic Press, San Diego, CA. 555 pp.

Evans, J. R., T. D. Sharkey, J. A. Berry, and G. D.
Farquhar. 1986. Carbon isotope discrimination
measured concurrentK' with gas exchange to investi-
gate CO2 diffusion in leaves of higher plants.
Australian Journal of Plant Physiology 13: 281-292.

Farquhar, G. D. 1983. On the nature of carbon isotope
discrimination in C4 species. Australian Journal of
Plant Physiology 10: 205-226.

Farquhar, G. D., J. R. Ehleringer, and K. T. Hubick.
1989. Carbon isotope discrimination and photosyn-
thesis. Annual Review of Plant Physiology and
Molecular Biology 40: 503-537.

Flanagan, L. B., and R. L. Jefferies. 1988. Stomatal
limitation of photos>'nthesis and reduced growth of

1995]

Carbon Isotope Discrimination in Atriplex

141

the halophyte, Phmtiifio moritima L., at higli salinitv.
Plant, Cell and Environment 11: 239-245.

Guy, R. D„ and D. M. Reid. 1986. Photosynthesis and the
influence of COq enrichment on 5^-^C values in a C3
halophyte. Plant, Cell and Environment 9: 65-72.

Henderson, S. A., S. von Cafmmerer, and G. D.
Farquhar. 1992. Short-term measurements of car-
bon isotope discrimination in several C4 species.
Australian Journal of Plant Physiology 19: 263-285.

Jenkins, C. L. D., R. T. Furbank, and M. D. Hatch.
1989. Mechanism of C4 photosynthesis. A model
describing the inorganic carbon pool in bundle
sheath cells. Plant Physiology 91: 1372-1381.

Long, S. E, and N. R. Baker. 1986. Saline terrestrial
environments. Pages 63-102 in N. R. Baker and S. E
Long, editors. Photosynthesis in contrasting envi-
ronments. Elsevier Scientific Publishers, New York,
NY.

Marino, B. D., M. B. McElroy, R. J. Salawitch, and
W. G. Spaulding. 1992. Glacial-to-interglacial varia-
tions in the carbon isotopic composition of atmos-
pheric COo. Nature 357: 461-466.

Meinzer, F C, Z. Plaut, and N. Z. Saliendra. 1994.
Carbon isotope discrimination, gas exchange, and
growth of sugarcane cultivars under salinity. Plant
Physiology 104: 521-526.

MooK, W. G., J. C. Bommerson, and W. H. Staverman.
1974. Carbon isotope fractionation between dis-
solved bicarbonate and gaseous carbon dio.xide.
Earth and Planetary Science Letters 22: 169-176.

O'Leary, M. H. 1988. Carbon isotopes in photosynthesis.
BioScience 38: 325-336.

Pearcy, R. W. , and J. Ehleringer. 1984. Comparative
ecophysiology of C3 and C4 plants. Plant, Cell and
Environment 7: 1-13.

Peisker, M. 1982. The effect of CO2 leakage from bundle
sheath cells on carbon isotope discrimination in C4
plants. Photosynthetica 16: 53.3-541.

Peisker, M., and S. A. Henderson. 1992. Carbon: terres-
trial C4 plants. Plant, Cell and Environment 15:
987-1004.

Rhoades, J. D. 1982. Soluble salts. Pages 167-179 in
Methods of soil analysis, part 2. Chemical and micro-
biological properties. ASA-SSSA, Madison, WL

Shomer-Ilan, a., D. Moualem-Beno, and Y. Waisel.
1985. Effects of NaCl on the properties of phospho-
enolpyiaivate carboxylase from Suaeda nwnoica and
Chloris gcnjana. Physiologia Plantaruin 65: 72-78.

Troughton. J. H., E V Wells, and H. A. Mooney. 1974.
Photosynthetic mechanisms and paleoecology from
carbon isotope ratios in ancient specimens of C4 and
CAM plants. Science 185: 610-612.

Walker, C. D., and R. Sinclair. 1992. Soil salinity is cor-
related with a decline in I'^C discrimination in leaves
oi Atriplex species. Australian Journal of Ecology 17:
83-88.

Received 20 May 1994
Accepted 16 August 1994

DEMOGRAPHY OF ASTRAGALUS SCAPHOIDES AND EFFECTS OF
HERBI\ ORY ON POPULATION GROWTH

Peter Lesica^

Abstract. — Losses in feeinulit\ due to predispersal lierbivon' can lie large; however, the effects of this loss on long-
term population viabilit\' ha\'e rarely been investigated. I conducted a demographic study of Astragalus scaphoides
(Fabaceae), a long-lived perennial endemic to east central Idaho and adjacent Montana, b\- following mapped individu-
als at two sites from 1986 to 1993. Astragalus scaphoides suffers losses of predispersal fecundity averaging nearly 50%
from insect seed predatioii and inflorescence predation by insects and livestock. Cattle reduced fecundit) by 0-85%.
Nonetheless, estimates from matri.x projection models indicate that both sample populations had positive growth in
most years. Elasticity analyses revealed that population growth occurred in spite of relatively small contributions by
recniitment compared to growth and survival of nonreproductive plants. Results suggest that populations of this long-
lived perennial depend little on reproduction and recruitment for growth and can persist in association with seasonal-
rotation livestock grazing.

Key words: demography, Iierbivory, livestock grazing, predation, matrix projection models, elasticity analysis.
Astragalus, rare plant.

The importance of herbivory in determining
plant population dynamics and composition of
vegetation has long been debated (Ehrlich and
Birch 1967, Slobodkin et al. 1967, Belsky
1986). A great deal of evidence suggests a neg-
ative impact of herbivory on the host plant
(Harper 1977, Crawley 1983, Dirzo 1984);
however, researchers have recently presented
evidence for positive interactions (McNaughton
1986, Paige and Whitham 1987).

A plant s life history plays an important role
in determining the effects of herbivory. Loss of
reproductive output from seed predators can
be disastrous for an annual or biennial but
may have little effect on a long-lived perenni-
al. Furthermore, effects of herbivory will
depend on the age or stage (e.g., seeds, adults)
at which it occurs (Dirzo 1984). Most studies
have focused on the effects of herbivores on
particular components of fitness over relatively
short time spans. This is unfortunate because
it is the long-term effect on population growth
that determines the importance of herbivory
to population viability. Few studies have inte-
grated the effects of herbivory on population
dynamics and growth (Harper 1977; but see
Louda 1982, 1983).

Predation, particularly by exotic species,
has often been cited as a threat to endangered
plant populations (Greig-Smith and Sagar 1981,
Parsons and Browne 1982, Willoughby 1987,
Norton 1991, Pavlik et al. 1993). Negative im-
pacts of herbivores were shown, but a causal
link to declining population size has rarely been
demonstrated.

Astragalus scaphoides (Jones) Rydb. is
endemic to a small area of east central Idaho
and adjacent Montana (Barneby 1964). It was
formerly a candidate for listing as a threatened
or endangered species by the U.S. Fish and
Wildlife Service (Category 3C; USDI-FWS
1993) and is currently listed as sensitive in
Idaho (Moseley and Groves 1990) and Montana
(Lesica and Shelly 1991). Most populations of
A. scaphoides occur on public lands subject to
livestock grazing. High levels of inflorescence
and seed predation have been observed in
some populations (Lesica and Elliott 1987a).
Here I report the results of an eight-year
demographic study of A. scaphoides at two sites.
The puipose of the study is to document levels
of herbivory and to assess its importance to
population growth using stage-based transi-
tion matrix models and elasticity analysis (de
Kroon et al. 1986, Caswell 1989).'

' Di\ ision ol Biolowcal Science, Universih,' of Mdiitana, Missoula, MT 59812, ;incl Conservation BiologN Research, 929 Locust, Missoula, MT 59802.

142

1995]

Demography and Herbivory in Astragalus

143

Methods

Species Studied

Astragalus scaphoides is a caulescent peren-
nial with a taproot surmounted by a branched
caudex. Reproductive individuals are 20-50
cm high with a cluster of pinnately compound
basal leaves and 3-10 leaves at intervals along
the erect stem. The inflorescence is composed
of 1—4 racemes arising from the axils of the
upper leaves. Each raceme is composed of a
naked peduncle, 5-15 cm long, surmounted
by a tight cluster of 10-30 flowers that
expands in fruit. Nonreproductive individuals
generally have 1-4 basal leaves and may have
a sterile stem less than 15 cm tall with 1-5
leaves. The branching caudices of reproduc-
tive plants may bear up to four stems and
more than a dozen racemes (Barneby 1964,
Lesica unpublished data).

Astragalus scaphoides generally flowers dur-
ing the first three weeks of June. The most
conspicuous form of herbivoiy of these plants
is the removal of inflorescences during flower-
ing. Inflorescence predation has two principal
sources: insects and livestock. Ants (subfamily
Formicinae) and moth larvae [Melacosoma spp.,
family Lasiocampidae) were observed remov-
ing inflorescences at a site near Haynes Creek
in Idaho. Peduncles below the flowers were
girdled, and withered inflorescences were
often still present near the base of the plant.
Inflorescence predation by livestock also
occurred but differed from insect predation in
that peduncles were all removed at the same
height, and severed inflorescences were not
found below the plants. In either case the
cluster of basal leaves was usually left intact. It
was possible to assign primary responsibility
for inflorescence herbivory at a site in a partic-
ular year to either insects or ungulates based
on the appearance of damaged plants and the
presence or absence of droppings, hoof prints,
or trampled vegetation. However, it was not
possible to unambiguously assign each case of
herbivory to one or the other source. Inflor-
escence predation by insects was obsei^ved at
both study sites in all years that inflorescences
were produced, but ungulate predation was
common only at Sheep Corral Gulch.

Predispersal seed predation occurred at
both sites in most years. Lai-vae were collected
from developing legumes in 1986 and identi-

fied as weevils, small beetles in the family
Curculionidae. Weevil larvae feed on maturing
seeds and leave the developing or mature
legume by creating a small hole in the outer
wall. Seed predation by weevil larvae was
inferred from the presence of fecula and/or an
exit hole in the legume.

Study Sites

The Sheep Corral Culch population occurs
in southern Beaverhead County, MT, on a gen-
tle south-facing slope at 1920'm (T8S R12W
S16). Mean July and Januaiy temperatures at
Diflon, 32 km NW and 275 m lower, are 19.0°
and -6.6° C, respectively. Mean annual precipi-
tation is 241 mm. Vegetation is dominated by
Artemisia tridentata and Agropijron spicatum.
Aster scopulorum and Phlox hoodii are com-
mon forbs. Livestock were managed on a rest-
rotation system by which grazing occurred in
different seasons in most consecutive years.
Evidence of heavy spring grazing by livestock
was observed in 1989, 1990, and 1993.

The Haynes Creek population is in central
Lemhi County, ID, approximately 48 km W of
Sheep Corral Gulch. It occurs on a moderate
southeast-facing slope at 1555 m (T19N R23E
S2). Mean July and January temperatures at
Salmon, 24 km NW and 365 m lower, are
16.2° and -6.7° C, respectively. Mean annual
precipitation is 252 mm. Vegetation is domi-
nated by Artemisia tridentata, Agropijron spica-
tum, and Bromus tectoruni. This site was not
grazed by livestock before early July during
the course of the study.

Field Methods

Two permanent monitoring transects were
established at each of the study sites in early
July 1986 following methods outlined in Lesica
(1987). Transects were located subjectively to
represent the populations and were read in
early July because fruits were mature or near-
ly so, but seed dispersal had not yet begun. At
each site the transects were parallel to each
other and the slope. Each transect consisted of
50 adjacent l-m^ mapping quadrats placed
along the transect line. The position of each A.
scaphoides plant encountered in the quadrats
was mapped and classified for three traits: (1)
size, (2) inflorescence production, and (3)
fecundity using the following classification:

144

Great Basin Naturalist

[Volume 55

(1) Size classes:

D Dormant (no abovegroiuid parts

ohserved)
S Small nonreproclucti\'es

(1-3 leaves)
L Large nonreproductives

(> 4 leaves)
R Reproductive

(2) Inflorescence production:

A Inflorescence produced no Iruit
P Inflorescence was removed due

to predation
I Inflorescence produced at least

one mature fruit

(3) FecunditN': total number of mature
fruits

When stems were removed below the point of
inflorescence articulation, I made a consei-va-
tive estimate of the number of inflorescences
removed based on the size of the remaining
plant. Evidence of livestock and native ungu-
lates (e.g., droppings, hoof prints, trampled
\ egetation) was noted along each transect and
for the site as a whole.

I found that some plants would go unde-
tected for one to several years but reappear in
subsequent years (Lesica and Steele 1994).
These "dormant" plants may have produced
small leaves that had senesced and disap-
peared by early July; however, my observa-
tions in May and June suggest that most of
them produced no vegetation on the years in
(luestion. The presence of dormant plants can
be inferred by comparing transect maps from
tlie full sequence of years. The proportion of
dormant plants ranged from 1% to 23%, with a
mean of 10% in 1987-1991. Plants have "'dis-
appeared" for as many as five years before re-
appearing. However, in 1986-1992 at the two
sites, 71% of dormant plants reappeared after
one year, and 88% reappeared after two years
(Lesica and Steele 1994). As a result, ca 10%
of the plants were undetected in the first and
last years of the study, and ca 3% were imde-
tected in the second and second from last
years. Thus, I have chosen to eliminate the
first and last years (1986, 1993) of the study
from demographic analysis, recognizing that a
small (ca 3%) error still remains in mortality
and recruitment estimates in 1987 and 1992.

On years when fruit production was ade-
quate, I collected 50 randomly selected
mature fruits from at least 25 plants. I opened

the pods, counted intact seeds, and recorded
evidence of insect predation.

Data Analysis

Stage-structined transition matrix projec-
tion models summarize the way in which sur-
vival, growth, and reproduction at various life-
history stages interact to determine population
growth (van Groenendael et al. 1988, Caswell
1989). Matrix projections assume fixed transi-
tion probabilities between stages in a popula-
tion through time (Lefkovitch 1965, Menges
1990). They also assume density-independent
population growth and thus do not give an
accurate projection of long-term population
future. Nonetheless, they can be used to sum-
marize short-term population dynamics
(Caswell 1989). One-year transition probabili-
ties were estimated as the number of plants in
life-stage class i moving into class j over the
course of one year divided by the number of
plants in stage / at the beginning of the year.
This method assumes that an individual's tran-
sition depends only on its life-stage class at
the beginning of the period and is indepen-
dent of its transition the previous year. The
equilibrium growth rate (A,) is the dominant
eigenvalue of the transition matrix (Lefkovitch
1965, Caswell 1989). }i > 1.0 indicates popula-
tion increase, while X < 1.0 indicates
decrease. X integrates the effects of sui-vival,
growth, and fecundity of the different life-his-
toiy stages into a single parameter There are
two ways in which a reproductive plant can
undergo a transition: (1) the plant itself moves
into a different class or stays the same and (2)
the plant produces progeny in one or more
classes. These two prol^abilities (Recniit, Repro)
are presented separately in the matrices but
must be added together to solve for X. Details
on the construction and use of matrix popula-
tion models can be found in Caswell (1989)
and Menges (1990). X was calculated using
RAMAS/stage (Person 1991).

Elasticity measures the relative change in
the value of X in response to changes in the
value of a transition matrix element. Elasticity
matrices allow comparison of the relative con-
tributions of various life-history transitions to
population growth and fitness (de Kroon et al.
1986). Elasticities sum to unity, and regions of
the matrix may be summed to compare the im-
portance of growth and sui-vival to recruitment

1995]

Demography and Herbivory in Astragalus

145

(Caswell 1989). Elasticities for nonreproductive
plants are sums from the small (S) and large
(L) classes. Elasticities were calculated using
RAMAS/stage (Ferson 1991).

When the majority of seeds pass directly
from production to germination in less than
one year, seeds should not appear as a sepa-
rate stage in matrix models (Caswell 1989,
Silvertown et al. 1993). Seeds oi' Astragalus
scaphoides germinate readily without stratifica-
tion (Lesica and Elliott 1987b), suggesting that
most seeds germinate the same year they are
produced. Nonetheless, A. scaphoides may fonii
a seed bank. Not including a seed bank in the
matrix model may affect the value of X (Kalisz
and McPeek 1992), especially when it is <1.0.
However, it will have little effect on analyses
based on elasticities (Silvertown et al. 1993). I
calculated separate elasticities for reproduc-
tive transitions and recruitment by dividing
the reproductive -I- recruitment elasticities
proportionately between the two components.

Losses to predation were estimated from
the number of inflorescences lost using the
calculated means for seeds/fruit and fruits/
inflorescence. Cumulative fecundity losses
were calculated by multiplying the propor-
tions of inflorescences and seeds remaining
after predation and subtracting from one.

Results

Population Growth

The number of Astragalus scaphoides
plants in the transects at both sites increased
by about one-third between 1986 and 1993
(Fig. 1). Equilibrium population growth rate
(k) was >1.0 at both sites over the course of
the study and was >2.5 at Sheep Corral Gulch
in 1988-89 and 1990-91. At no time during
the study was X < 0.8 at either site (Table 1).

Survivorship

Between 40% and 50% of the Astragalus
scaphoides plants observed at the start of the
study in 1986 were still alive in 1993 (Fig. 2).
Approximately 50% of the 1989 cohort (the
first large cohort recruited during the study)
survived for more than 3-4 years. Taken
together these results indicate that A.
scaphoides is a long-lived perennial, with ca
50% mortality occurring in the first 3-4 years,
but a large proportion of plants living to be
> 10 years.

Predation

Inflorescence predation attributable to ungu-
lates was virtually absent from the Haynes
Creek population. Droppings and hoof prints
of cattle were the only signs of ungulates at
Sheep Corral Gulch. Droppings occurred in
3-9% of the mapping quadrats during the
study. Inflorescence predation by insects
occurred at both sites in all years.

A significant number of inflorescences were
produced in six of eight years at Haynes Creek,
and inflorescence predation accounted for
fecundity losses of 14-50% over the course of
the study (Fig. 3). Most of this herbivory was
attributable to insect damage. At Sheep Corral
Gulch reproductive plants were common in
only four of eight years. Inflorescence predation
resulted in fecundity losses of 19-90%, and
the proportion of inflorescences lost to preda-
tion was highest in 1989, 1990, and 1993,
years in which predation was due mainly to
livestock (Fig. 3).

Seed predation occurred at both sites in
nearly every year in which significant fruiting
occurred (Fig. 3). Overall, loss of seeds to wee-
vil predation ranged from 0 to 33% with a
mean of 18%. Insect seed predation was gen-
erally higher at Sheep Corral Gulch than at
Haynes Creek (Fig. 3).

Losses of fecundity due to the combined
effects of inflorescence and seed predation
were 19-90% in 1986-1993, with means of

250
225

200

D 175

o 150

(U
XI

E 125

-z.

100

87 88 89 90 91 92

Year

Fig. 1. Density of Astragalus scaphoides plants in the
two sample populations, 1987-1992.

146

Great Basin Naturalist

[Volume 55

Table 1. Stage-based transition matrices for Aslra^ahis scaphoides at two sites in 1987-1992. Four stages are recog-
nized: dormant (D), small nonreproductive (S), large nonreproductive (L), and reproductive (R). The reproductive and
recruitment (Re) columns must lie added together before solving for X, the dominant eigenvalue (see Methods).

>;Lo„,-, r^^rval r:„lr.lT

1987-88

1990-91

From

R Re

S L

R Re

.67

.18

.20

0 + 0

.14

.06 0

0 +0

.11

.55

.24

0 + 0

.21

.23 .06

0 + 9.86

.22

.06

.36

0 +0

.50

.42 .26

.29 + 2.42

0 0

.03

1.0 + 0

.14

.12 .57

1.0 + .14

X- 1.18

X = 2.69

1988-89

1991-92

From

R Re

S L

R Re

.23

.04

.02

0 + 0

.70

.24 .21

.25 + 0

.17

.27

.05

0 + 4.0

.30

.27 .37

.33 + .20

.43

.53

.45

0 + 7.0

0 .14

.22 + 0

.17

.08

.43

1.00 + .25

0 0

0 +0

X = 2.51

X = 0.83

1989-90

From

R Re

.80

.17

.14

.06 + 0

.10

.73

.37

.22 + .91

.10

.02

..35

.56 + .03

X = 0.97

.02

.01

.16 + 0

Haynes Creek -

1987-88

1990-91

From

R Re

S L

R Re

.50

.03

.04

0 +0

.21

.03 0

0 +0

.40

.45

.04

0 + 5

.21

.34 .03

0 + .95

.24

.37

0 + .16

.50

.31 .21

.18 + .27

.10

X = 1.88

.06

.52

.60 + .20

.07

x= :

1.31

.19 .66

.64 + 0

1988-89

1991-92

From

R Re

S L

R Re

.57

.13

.04

0 +0

.75

.03 .06

.05 + 0

.14

.42

.24

.05 + 1.10

.44 .30

.15 + .30

.14

.16

.28

.14+ .14

.25

.08 .36

.53 + .03

.14
X= 1.13

.04

.32

.67 + .05

X = (

J.83

0 .11

.20 +0

1989-90

From

R Re

.64

.12

.05

0 +0

.27

.38

.14

.04 + .23

.20

.48

.27 + .04

.09 (
X = 0.97

.33

.58 + 0

1995]

Demography and Herbivory in Astragalus

147

100
90
80
70
60
50
40
30
20

• Sheep Corral
V Haynes

Sheep Corral Gulch

56 87 88 89 90 91 92 93
Year

Fig. 2. Depletion curves for the 1986 sample popula-
tions of Astragalus scaphoides at the two study sites.

51% and 44% at Haynes Creek and Sheep
Corral Gulch, respectively (Fig. 3).

Elasticity Analysis

Elasticity gives the proportional impor-
tance of demographic transitions to population
growth. Elasticity matrices for five years of
transitions for the two study sites are given in
Table 2. Elasticities were summed into four
life-history transition categories: (1) recruit-
ment and sui-vival and growth of (2) dormant,
(3) nonreproductive, and (4) reproductive
plants (Fig. 4). Growth and sui'vival of nonre-
productives was consistently important at both
sites, with mean elasticities of 42% and 36% at
Haynes Creek and Sheep Corral Gulch,
respectively. Survival of dormant plants was
important in two years at Sheep Corral Gulch
and one year at Haynes Creek, with mean
elasticities of 19% and 29%. Survival of repro-
ductives had mean elasticities of 23% and 20%
for the two sites, and mean elasticities for
recruitment were 16% and 17%.

Discussion

Loss of Astragalus scaphoides fecundity
due to inflorescence and predispersal seed
predation was high at both sites, ranging from
ca 20% to >90%. Further losses in reproduc-
tive output due to ants or rodents may have
occurred following dispersal. Recixiitment was
the least important stage transition in the life
history of A. scaphoides during my study.

0.6

0.4

0.2

0.0

{//} Inflorescence
P^ Seed
^H Combined

/X ,

89 91
Year

Haynes Creek

0.4

0.2

0.0

\//} Inflorescence
^ Seed
^H Combined

Fig. 3. Proportion of Astragalus scaphoides reproduc-
tive output lost to inflorescence predation, predispersal
seed predation, and the combination of the two in those
years when significant flowering occurred at two study
sites. Numbers of inflorescences in samples are given
above bars.

accounting for an average of less than 17% of
population growth at both sites. High levels of
inflorescence and seed herbivoiy are undoubt-
edly one of the main reasons for the low con-
tribution of recruitment to X in this species.
Nonetheless, both sample populations became
larger during the study. Furthermore, popula-
tion growth rate was >1.0 in four of five years
at both sites and never <0.8. Growth and sur-
vival of dormant and nonreproductive plants
contributed >60% to population growth at
both sites. These results suggest that popula-
tions of A. scaphoides can persist and even
grow larger in spite of heavy losses in repro-
ductive output and low recruitment.

Large reductions in fecundity due to herbi-
vores have been documented for Astragahis
species (Green and Palmbald 1975) as well as

148

Great Basin Naturalist

[Volume 55

Table 2. Elasticities for Astrufialus scaphukh's stage transition matrices at two sites for 1987-1992. The left three
columns (D, S, L) represent nonreproductive growth and survival. The reproductive (R) column represents growth and
survival of reproductives. The recniitment column (Re) represents recruitment from seed.

. - - ^nf**^r\ \ cwv'^' C^ \\\t^'\-t _

1987

-88

■ - - kjiiccrp \_>uii

1990-91

.048

.022

,015

.001

.005

.004

.032

.009

.024

.001

.018

.002

.185

.0.32

.015

.055

.077

.002

.082

.023

.014

.118

1988

.101
-89

.568

.003

.099

.215
1991-92

.204

.029

.001

.003

.001

.686

.126

.001

.017

.002

.139

.126

.061

.002

.081

.049

.143

.003
D

.057

1989

.222

-90

.224
R

.056
Re

.413

.074

.015

.001

.048

.292

.037

.002

.011

.042

.007

.030

.005

.001

.017

.002

.004

—

Haynes

Creek - -

1987

-88

1990-91

.002

.003

.002

.001

.005

.001

.045

.002

.141

.001

.058

.004

.160

.082

.058

.153

.004

.080

.038

.046

.068

.003

.059
1988

.231
-89

.164

.055

.001

.079

.193
1991-92

.264

.055

.048

.006

.617

.020

.041

.006

.007

.086

.020

.005

.113

.042

.030

.003

.005

.015

.064

.044

.028

.068

.018

.082

.021

.001

.031
D

.034

1989

.109

-90

.287
R

.021
Re

.036

.011

.081

.028

.013

.023

.061

.024

.007

.039

.065

.170

.093

.014

.018

.134

.230

many other plant.s (Janzen 1971, Hendrix 1988,
Louda 1989). Louda (1982, 1983) has shown
that seed predation can lead to lowered
recruitment; however, reductions in seed out-
put will not necessarily lead to reduced
recruitment if germination safe sites are limit-
ing (Harper 1977). Analysis of the matrix pro-
jection models suggests that recruitment is not
limiting population growth of A. scaphoides.

Recruitment from seed is likely to be im-
portant to population growth for short-lived
species and is essential for semelparous ones.
Furthermore, successful reproductive episodes
are rare for some perennial species in rigorous

environments (Jordan and Nobel 1979). Signi-
ficant reductions in a single reproductive bout
could greatly increase chances of population
extirpation for these sorts of species. On the
other hand, many populations of long-lived
plants will have more stable populations
whose persistence is more dependent on the
growth and survival of established plants
(Silvertown et al. 1993). Survivorship curves
indicate that Astragalus scaphoides is a long-
lived species, and elasticity analysis suggests
that recruitment is indeed less important to
population persistence than growth and sur-
vival of nonreproductive plants.

1995]

Demography and Herbivory in Astragalus

149

Sheep Corral Gulch

0.8

>- 0.6

CD 0.4

0.2

0.0

^B Dormant

I I Non-reproductive

)^\\| Reproductive

lAAJ Recruitment

imJ

90
Year

1.0

0.8

Haynes Creek

^B Dormant
I I Non-reproductive
- L\N Reproductive
[XX] Recruitment

>^ 0.6

C 0.4 -

0.2 -

0.0

KL^

90
Year

Fig. 4. Elasticities summed into four life-history transi-
tion categories (recruitment and sur\'ival and growth of
dormant, nonreproductive, and reproductive plants) for
Astragalus scaphoides at two stud\' sites, 1987-1992.

Inflorescence predation of Astragalus sca-
phoides was greatest in years when livestock
were present. In 1993 inflorescence predation
was greater than 85%, and A. scaphoides was
grazed in preference to the highly palatable
grass, Agropyron spicatum (P Lesica personal
observation). These observations suggest that
livestock could nearly eliminate reproductive
output under high stocking rates and repeated
heavy spring grazing if carried on over a long
enough period of time. However, results of my
study suggest that A. scaphoides populations
can persist if predation is moderate, at least in
some years. Rotation grazing systems in which
spring grazing occurs only one in three years
appear to be compatible with the long-term
persistence of A. scaphoides populations.

These results have implications for other
long-lived perennials exposed to livestock pre-

dation. Upper portions of plants are most
accessible to livestock, and newer growth is
generally selected by livestock (Arnold and
Dudzinski 1978, Vcilentine 1990). Furthemiore,
sugars, such as found in flower nectar, also
increase palatability (Arnold and Dudzinski
1978, Valentine 1990). Thus, livestock often
remove only the upper portions of broad-
leaved plants. Predation that mainly affects
fecundity is likely to endanger populations
only when grazing removes most inflores-
cences consistently for many years because
population growth is not likely to be limited
by recruitment. On the other hand, grazing
that lowers growth and survival (e.g., high-
density stocking during periods of growth) will
have a much more detrimental effect on popu-
lation viability.

Acknowledgments

I am grateful to Joe Elliott, Anne Garde,
and Lou Hagener for help in the field. James
Liebherr of the Comstock Museum, Ithaca,
NY, and Will Lanier of the Entomology
Research Lab, Bozeman, MT, identified
insects. Kimball Harper and an anonymous
reviewer gave helpful comments on the manu-
script. Funding was provided by the Idaho
and Montana Bureau of Land Management
and the Montana Natural Heritage Program.

Literature Cited

Arnold, G. W, and M. L. Dudzinski. 1978. Ethology of
free-ranging domestic animals. Elsevier, Amsterdam.
198 pp.

BEL.SKY, A. J. 1986. Does herbivory benefit plants? A
review of the evidence. American Naturalist 127:
870-892.

Barneby, R. C. 1964. Atlas of North American Astragalus,
parts 1 and 2. Memoirs of the New Ybrk Botanical
Garden 13: 1-1188.

Caswell, H. 1989. Matri.x population models. Sinauer
Associates, Sunderland, MA. 328 pp

CR.WLEY, M. J. 1983. Herbivoiy, the dynamics of animal-
plant interactions. University of California Press,
Berkeley. 437 pp.

DiRZO, R. 1984. Herbivory: a phytocentric viewpoint.
Pages 141-165 in R. Dirzo and J. Saruklian, editors.
Perspectives on plant population ecology. Sinauer
and Associates, Sunderland, MA.

Ehrlich, F R., and L. C. Birch. 1967. The "balance of
nature" and "population control." American Naturalist
101: 97-108.

Ferson, S. 1991. RAMAS/stage. Generalized stage-based
modeling for population dynamics. Applied Bio-
mathematics, Setauket, NY.

150

Great Basin Natl iulist

[Volume 55

Green, T. W., and 1. C. Palmualu. 1975. En'fcts of insect
seed predators on Astruflfilus ciharius and Astragalus
utahensis (Leguminosae). Ecology 56; 1435-1440.

CREic-Svirm, J., and G. R. Sacar. 1981. Biological causes
of rarity in Carlina vulgaris. Pages 389-399 in H.
Synge, editor, Biological aspects of rare plant conser-
vation. John Wiley and Sons, Chichester, England.

VAN Groenendael, J. M., H. DE Kroon, and H. Caswell.
1988. Projection matrices in population biology.
Trends in Ecolog\' and Evolution 3: 264-269.

Harper, J. L. 1977. Population biology of" plants. Aca-
demic Press, London. 892 pp.

IIendri.X, S. D. 1988. Herbivow and its impact on plant
reproduction. Pages 246-263 in J. Lovett-Doust and
L. Lovett-Doust, editors. Plant reproductive ecology.
O.xford University Press, New York, NY.

Janzen, D. H. 1971. Seed predation. Annual Review of
Ecolog\' and Systematics 2: 465^92.

Jordan, R W., and P S. Nobel. 1979. Infrequent estab-
lishment of seedlings of Agave deserti (Agavaceae) in
the northwestern Sonoran Desert. American [ounial
ofBotany 66: 1079-1084.

IC\Llsz, S., and M. A. McPeek. 1992. Demography of an
age-structured annual: resampled projection matri-
ces, elasticity analyses, and seed bank effects.
Ecologn' 73: 1082-1093.

DE Kroon, H., A. Plaiser, J. \1. van Groenendael, and
H. Caswell. 1986. Elasticity: the relative contribu-
tion of demographic parameters to population
growth rate. Ecology 67: 1427-1431.

Lefkovitch, L. P 1965. The study of population growth
in organisms grouped by stage. Biometrics 21: 1-18.

IjESICA, P 1987. A technique for monitoring nonrhizoma-
tous, perennial plant species in permanent i)elt tran-
sects. Natural Areas Journal 7: 65-68.

Lesica, P, and J. C. Elliott. 1987a. Distribution, age
stiTicture, and predation of Bittenoot milk^ etch pop-
ulations in Lemhi County, Idaho. Report submitted
to the Bureau of Land Management, Boise, ID.

. 1987b. 1987 monitoring study of Bittenoot milk-
vetch populations in Lemhi Count\, Idaho. Report
submitted to the Bureau of Land Management,
Boise, ID.

Lesica, P, and J. S. Shelly. 1991. Sensitive, threatened and
endangered vascular plants of Montana. Montana
Natural Heritage Program Occasional Publication 1,
Helena. 88 pp.

Lesica, E, and B. M. Steele. 1994. Prolonged dormancy
in vascular plants and implications for monitoring
studies. Natural Areas Journal 14: 209-212.

Louda, S. M. 1982. Distribution ecology: variation in
plant recruitment over a gradient in relation to
insect seed predation. Ecological Monographs 52:
25^1.

. 1983. Seed predation and seedling mortality in

the recruitment of a shrub, Haplupappus venetus
(Asteraceae), along a climatic gradient. Ecologv 62:
511-521.

. 1989. Predation in the cKnamics of seed regener-

ation. Pages 25-51 in M. A. Leek, V. T Parker, and
R. L. Simpson, editors, Ecolog\' of soil seed banks.
Academic Press, New York, NY.

McNalghton, S. J. 1986. On plants and herbivores.
American Naturalist 128: 765-770.

Menges, E. S. 1990. Population viability analysis for an
endangered plant. Consei-vation Biology 4: 52-62.

MosELEY, R., and C. Gro\ es. 1990. Rare, threatened and
endangered plants and animals of Idaho. Idaho
Natural Heritage Program, Boise. 33 pp.

Norton, D. A. 1991. Trilcpidca udumsii: an obituary for a
species. Conservation Biology 5: 52-57.

P.\IGE, K. N., .\ND T. G. Whitham. i987. Overcompensation
in response to mammalian herbivoiy: the advantage
of being eaten. American Naturalist 129: 407-416.

Parsons, R. E, and J. H. Browne. 1982. Causes of plant
species rarity in semi-arid southern Australia. Bio-
logical Consei"vation 24: 183-192.

Pavlik, B. M., N. Ferguson, and M. Nelson. 1993.
Assessing limitations on the growth of endangered
plant populations, II. Seed production and seed bank
d\namics of Erysimum capitafum ssp. angustatum
and Oenothera deltoides ssp. howellii. Biological
Conservation 65: 267-278.

Silvertown, J., M. Franco, I. Pisanty; and A. Mendoza.
1993. Comparative plant demography — relative
importance of life-cycle components to the finite
rate of increase in woody and herbaceous perenni-
als. Journal of Ecology 81: 465-476.

Slobodkin, L. B., F E. Smith, and N. G. Hairston. 1967.
Regulation in terrestrial ecosystems and tlie implied
balance of nature. American Naturalist 101: 109-124.

USDI-FisH AND Wildlife Sermce. 1993. Endangered
and threatened wildlife and plants; review of plant
ta\a for listing as endangered or threatened species;
notice of review. Federal Register 58: 51144-51190.

Valentine, J. F 1990. Grazing management. Academic
Press, San Diego, CA. 533 pp.

Willoughby, J. W 1987. Effects of livestock grazing on
two rare plant species in the Red Hills, Tuolumne
Count>', CiJifomia. Pages 199-208 in T S. Elias, editor.
Conservation and management of rare and endan-
gered plants. California Native Plant Society,
Sacramento.

Received 1 April 1994
Accepted 7 September 1994

LAHONTAN SAGEBRUSH {ARTEMISIA ARBUSCULA SSR LONGICAULIS):

ANEWTAXON

Alma H. Winwardl and E. Durant McArthur^

Abstract. — A new subspecies of Artemisia arbuscula is described, A. arbuscida ssp. lonp^icaiilis Winward &
McArthur, ssp. nov. This ta.\on is a landscape dominant in portions of northwestern Nevada and adjacent California and
Oregon at elevations from 1050 to 2000 m on shallow or argillic (clayey) soils. It differs from A. arbuscula ssp. arbuscula
in its long floral stalks and large leaves. Moiphological, chemical, ecological, and cytological data suggest that it is of
hybrid origin. It is he.xaploid (6x). We hypothesize that 2.x A. arbuscula ssp. arbuscula and 4x A. tridentata ssp. wijomin-
gensis are its parents.

Key words: Nevada, taxonomy, chemotaxonomy, allopolyploid, hybrid, Tridentatae.

In preliminaiy repoi^ts we (Winward et al.
1986, 1991) provided a brief description of a
new taxon of Ai~temisia found in northwestern
Nevada and adjacent California and Oregon.
We suggested referencing it by the common
name Lahontan sagebrush pending a formal
description. This paper provides that formal
description and details concerning its ta.xono-
my, distribution, general ecology, and origin.

Taxonomy

The new taxon is a member of subgenus
Tridentatae of Artemisia, the true sagebrushes
(Beetle 1960, McArthur et al. 1981, Shultz
1986). We suggested (Winward et al. 1986) that
this taxon may have originated as a hybrid with
parental lines consisting of low and big sage-
brush (A. arbuscula and A. tridentata). Because
of its general morphology and ecology, we sug-
gested that it be considered a subspecies of A.
arbuscula. Furtlier studies indicate that this new
taxon is in fact best treated as a subspecies of
A. arbuscula.

The type specimen of A. arbuscula came
from a collection along the arid plains of the
Lewis (now known as the Snake) River (Nuttall
1841). Subsequent workers have submerged the
species as a subspecies of big sagebrush, A.
tridentata ssp. arbuscula (Hall and Clements
1923), or, in contrast, recognized a number of
races and subspecies within the species (Ward
1953, Beetle 1960). Ward proposed that black

sagebrush (A. nova) was best treated as a sub-
species of A. arbuscida, but Beetle (1960)
restored it to Nelson's (1900) original species
status. Beetle (1960) recognized two sub-
species of A. arbuscula, arbuscula and ther-
mopola. His treatment has been generally
accepted (Winward and Tisdale 1977, McArthur
et al. 1981, Shultz 1986), although Winward
(1980) has observed an unusual variant of A.
arbuscula in eastern Oregon that reaches a
height of 1 m. He suggested that further taxo-
nomic treatment of A. arbuscula would be
appropriate.

During the past few decades fieldworkers
in western Nevada have obsei^ved a sagebrush
that does not fit the existing Artemisia taxo-
nomic keys. Brunner (1972) termed this sage-
brush "wide-lobe" with the comment, "Dr.
Beetle feels this may be an ecotype of A. triden-
tata ssp. wijomingensis. I concur." Others have
referred to it as "wonder sagebrush," "junk
sagebrush," or "N" sagebrush (Winward et al.
1986). Accessions of two populations (Trough
Springs, Humboldt County, NV, cultures Ul
and U58 and Leonard Creek, Humboldt
County, NV, culture U55) of this taxon were
established in common gardens of the Forest
Service's Shrubland Biology and Restoration
Research Work Unit at several locations
around central Utah; there they were treated
as an ecotv'iDe of A. tridentata ssp. wijomingensis
following Beetle and Brunner (Brunner 1972;

'Range and Watershed Management, Intennountain Region, Forest Service, U.S. Department of Agriculture, Ogden, UT 84401,
^Shrub Sciences Laboraton-, Intermountain Research Station. Forest Senice, U.S. Department of Agriculture, Pro\'0, I'T 84606.

151

152

Great Basin Naturalist

[Volume 55

e.g., McArthur and Pliimnier 1978, Welch and
McArthur 1979, 1981, 1986, McArthur et al.
1981, 1985, McArduir and Welch 1982, Welch
et al. 1986, 1987). The new taxon is a landscape
dominant over much of its range (Winward et al.
1986), and both domestic and wild animals
feed e.xtensivcK on it (Bmnner 1972, Welch and
McArthur 1986, Winward et al. 1986, Welch
et al. 1987).

Description

Artemisia arhuscula ssp. lonfiicaidis Win-
ward & McArthur ssp. nov. Similis A. arhuscula
ssp. arhuscula sed ramis floralibus nuilto lon-
gioribus et foliis magnioribus differt (Similar
to A. arhuscula ssp. arhuscula except flower
stalks are nuich longer and leaves are larger).

The longer flower stalks and larger leaves
also differentiate ssp. longicaulis from ssp.
thennopola, which differs from ssp. arhuscula
and longicaulis by having deeply trifid leaves
(Beetle 1959).

We chose the common name Lahontan sage-
brush because the old shorelines of Pleisto-
cene Lake Lahontan are one of the centers of
its current distribution and may have provided
the ecological setting for the taxon's origin and
development (Winward et al. 1986, 1991).

Type: Toulon, Pershing County, Nevada,
USA, 1053 m, S. C. Sanderson and E. D.
McArthur 1593, 21 August 1986. Holotype:
BRY. Isotypes: OGDF, RENO, SSLR and
UTC. Other specimens examined:

• Nevada, Douglas Co., Topaz Lake,
Sanderson & McArthur 1594, (SSLP
four sheets);

• Nevada, Humboldt Co., Golconda,
Plummers.n., 1985, (SSLP);

• Nevada, Humboldt Co., Leonard
Creek, Plummer & McArthur, s.n., 3
October 1975, culture U55, (SSLP);

• Nevada, Humboldt Co., Trough
Springs, Jackson Mountains, Plummer,
Brunner, & McArthur, s.n., 3 October
1975, culture Ul, (SSLP);

• Nevada, Humboldt Co., Trough
Springs, Jackson Mountains, McArthur
1532, culture Ul, (SSLP);

• Nevada, Humboldt Co., Trout Creek
Basin, Jackson Mountains, McArthur
1501, (SSLR two sheets);

• Nevada, Lyon Co., Dayton, Sanderson
& McArthur 1595, (SSLI^ two sheets);

• Nevada, Pershing Co., 6.4 km west of
Toulon, McArthur & McArthur 1683,
(SSLP two sheets);

• Nevada, Washoe Co., Mustang,
McArthur & McArthur 1684, (SSLP);

• Oregon, Lake Co., 32 km east of Adell,
Sanderson & McArthur 1590, (SSLP);

• Oregon, Malheur Co., near McDermitt,
Nevada, Winward, s.n. 31 October
1986, (OGDF; two sheets, SSLP).

Distribution and Ecology

Artemisia arhuscula ssp. longicaulis occurs
on several hundred thousand hectares in
northwestern Nevada and in adjacent areas of
California and Oregon at elevations from
about 1050 to 2000 m (Fig. 1). It often occurs
in pure stands. It may also share dominance
with other sagebrush taxa such as big sage-
brush (A. tridentata ssp. tridentata and wyo-
mingensis), low sagebrush (A. arhuscula ssp.
arhuscula), and black sagebrush (A. nova). At
lower elevations it is interspersed with salt
desert shrub species such as shadscale
{Atriplex confertifolia), Bailey greasewood {Sar-
cohatus hadeyi). Mormon tea {Ephedra spp.),
budsage {Artonisia spinescens), Shockleys
desert thorn {Lycium shockleyi), and horse-
brush {Tetradytnia spp.). Except for Artemisia,
our taxonomy follows Welsh et al. (1993) and
Mozingo (1987). The most common grass
understoiy species at upper-elevation Lahontan
sagebrush sites is bluebunch wheatgrass
{Elymus spicatus). At lower elevations Thurber
and desert needlegrasses {Stipa thurheriana
and S. speciosa), and Indian ricegrass {Stipa
hymenoides), bottlebrush squirreltail {Elymus
elymoides), and Sandberg bluegrass {Poa
secunda) are more common. Areas supporting
A. arhuscula ssp. longicaulis receive between
175 and 350 mm of precipitation annually with
most as wdnter precipitation. The frost-free
season ranges from 90 to 110 days. Lahontan
sagebrush grows most commonly on Aridisols,
but at upper elevations it also occurs on
MoUisols. Soil Conservation Service, U.S. De-
paitment of Agricultin-e, personnel have located
A. arhuscula ssp. longicaulis on at least 17 soil
series. Generally, these soils have low available
water-holding capacities and a shallow depth
to an argillic horizon and/or bedrock. These
soils are similar to those of low sagebrush (A.
arhuscula ssp. arhuscula) communities

1995]

Lahontan Sagebrush, A New Taxon

153

Susanville

Austin

NEVADA

Fig 1. Extent of the known distribution o{ Artemisia arhuscula ssp. longicaulis.

154

Great Basin Naturalist

[Volume 55

(Fosberg and Hironaka 1964, Zaniora and
Tueller 1973, G. K. Brackley and C. A.
Plumnier personal commnnication).

General distributions of the three sub-
species of A. arhuscula are as follows: ssp.
arhuscula, western Wyoming and eastern
Utah to eastern Washington and northeastern
California; ssp. thennopola, western Wyoming
and adjacent Idaho and northern Utah to
northern Nevada and eastern Oregon; ssp.
J()u<i.ic(iiilis, western Nevada, extending into
adjacent California and Oregon.

Supporting Data and Discussion

Morphological, chemical, and cytological
data are consistent with the hypothesis that A.
arhuscula ssp. longicaulis is of hybrid origin,
with A. arhuscula ssp. arhuscula as one parent
and A. tridentata ssp. wyomingensis as the
other parent. Hybridization and introgression
are thought to have been important in the evo-
lutionary development and differentiation of
Tridentatae taxa (Ward 1953, Beetle 1960,

McArthur et al. 1981, 1988). The following
taxa are thought to have originated as hybrids:

• A. argillosa (A. cana ssp. viscidula x A.
longiloha— Beetle 1959),

• A. tridentata ssp. spicifonnis (A.
tridentata ssp. vaseijana X A. cana ssp.
viscidula — Beetle 1959, Goodrich et al.
1985, McArthur and Goodrich 1986),

• A. tridentata ssp. xericensis (A. tridentata
ssp. tridentata X A. tridentata ssp.
vaseijana — Winward 1970, Rosentreter
and kelsey 1991),

• A. tridentata ssp. wijomingensis (A.
tridentata ssp. tridentata X A. tridentata
ssp. vaseijana with perhaps some
involvement with A. nova as well —
Beetle and Young 1965, Winward 1975,
McArthur 1983).

Figure 2 shows a polygonal representation of
morphological features of the new subspecies
and its putative parents, including permanent
leaf width, length, and length/width ratio and
flower and vegetative stalk lengths. Values are
shown in Table 1. Morphological differences

A. arhuscula ssp. longicaulis

A. arhuscula ssp. arhuscula

A. tridentata ssp. wyomingensis

Fig. 2. Polygonal graph comparing niorpliological features of Artcini.siu arhtisciila ssp. loti^icaulis and its putati\e par-
ents, A. arhuscula ssp. arhuscula and A. tridentata ssp. wyomingensis. Data from Table I; A = flower stalk length; B =
vegetative stalk length; C = leaf width; D = leaf 1/w ratio; E = leaf length.

1995]

Lahontan Sagebrush, A New Taxon

155

Table 1. Morjihological measurements (means ± SD) of A. arbiisctila ssp. arhuscula (AFL\R^), A. arhuscula ssp. lonfii-
caulis (ARA.R'), and A. tridentafa ssp. wyoiningensi.'i (ARTR")"'.

ARAR^'

ARARl

ARTR"

Flower stalk lenf;;tli (cm)
Vegetative stalk length (cm)
Leaf length (mm)
Leaf width (mm)
Leaf 1/w ratio

16.411.7

3.9 ±0.8

5.9+1.1

3.0 ±0.4

2.0

24.5 ± 3.4
5.7 ±1.2

7.3 + 1.0

3.4 ±0.4

2.2

10.4 ±4.2

4.7 ±2.4

8.3 ±2.4

2.7 ±0.8

3.1

"Data for ARAR" are from the follouing colleetions: California, Lassen Co., Shaffer Mountain, Sanderson & McArtlmr 1591 (SSLP); Nevada, Eureka Co., Tusearora
Mountains, 40 km north of Carlin, Frischkneclit 210, (SSLP); Nevada, Humboldt Co., 1.3 km northwest of Paradise Valley Town, Holmgren 128, (SSLP); Nevada,
Lander Co., Bade Creek, Toiyabe National Forest, Goodrich 8868, (OGDF); Nevada, Lander Co., Toiyabe Range, Toiyabe National Forest, Goodrich 9966,
(OGDF); Nevada, Nye Co.. Toiyabe Range, Toiyabe National Forest. Goodrich 12201, (OGDF): .Nevada, Pershing Co.. 40 km northeast of Winnemucca, Winward
s.n.. 25 October 1984, (OGDF); Nevada, Washoe Co.. Buffalo Hills, 40 km northwest of Gerlach, Sanderson & .VIcArthur 1592, (SSLP).

Data for AflAR' are from collections: .Nevada. Humboldt Co.. Golconda, Plummer s.n.. 1985, (SSLP); Nevada, Humboldt Co., Trough Springs, McArthur 1532,
(SSLP); Nevada, Humboldt Co., Trout Creek Basin, McArthur 1501, (SSLP); Nevada, Pershing Co., Toulon, Sanderson & McArdiur 1593, (SSLP); Nevada,
Pershing Co., 6.4 km. west of Toulon, McArthur & McArthur 1683, (SSLP); Nevada, Washoe Co., Mustang, McArthur & McArthur 1684, (SSLP); Oregon, Lake
Co., 32 km east of Adell, Sanderson & .McAi-thur 1590, (SSLP); Oregon. Malheur Co., near .McDermitt, Nevada, Winward s.n., 31 October 1986, (OGDF).

Means for each ta\a ai'e based on N = 8. Differences are significant by one-way analysis of variance (Woolf 1969) for flower stalk length {P < .01), vegetative
stalk length (P < .05), and leaf length (P < .01) between AK\R^ and ARAR'. Data for ARTR" are from Winward (1970). Means from ARTR" were not statistically
compared with die odier ta.\a.

may be summarized as follows: (1) Flower
stalk length averages 50% longer for A. arhus-
cula ssp. longicaulis than for A. arhuscula ssp.
arhuscula and over twice as long as A. triden-
tafa ssp. wijomingensis. (2) A. tridentata ssp.
wyoniingensis has longer, narrower leaves than
the other taxa. (3) A. arhuscula ssp. longicaulis
has larger leaves and a longer vegetative stalk
than A. arhuscula ssp. arhuscula.

Wilt et al. (1992) compared 34 phenolic
compounds from six taxa of Tridentatae in-
cluding A. tridentata ssp. wijomingensis, A.
arhuscula ssp. longicaulis (undescribed at the
time and called by them "Lahontan" sage-
brush), and A. arhuscula ssp. arhuscula. Their
work demonstrated that typical high-pressure
liquid chromatography patterns for A. arhuscula
ssp. longicaulis and A. tridentata ssp. wijo-
mingensis are remarkably similar (their Fig. 2).
Representative chromatograms of those two
taxa are the most similar of the six taxa studied
(A. arhuscula ssp. arhuscula, A. nova, A. triden-
tata ssp. tridentata, A. t. ssp. vaseijana, A. t. ssp.
wijomingensis, and "Lahontan " sagebrush; Wilt
et al. 1992). We analyzed the Wilt et al. (1992)
frequency of detection of individual phenolics
by assigning percent similarity values for each
shared phenolic (Table 2). Those values are
not dramatically different for A. arhuscula ssp.
longicaulis in respect to the other five taxa, but
A. arhuscula ssp. arhuscula and A. tridentata
ssp. wijomingensis are the two most similar (at
54%) to A. arhuscula ssp. longicaulis. For total
phenolics the Wilt et al. (1992) data are ver)'
close for A. arhuscula ssp. longicaulis and A. t.
ssp. wijomingensis, closer than for any other
two taxa examined.

A. arhuscula ssp. longicaulis is hexaploid,
6x = 2n = 54. McArthur et al. (1981) reported
this taxon as A. tridentata ssp. wijomingensis
"wide-lobe" and gave chromosome counts as
2n = 36, 54 for two transplanted population
samples growing in a common garden. We
now suspect both bona fide A. tridentata ssp.
wijomingensis and A. arhuscula ssp. longi-
caulis were growing in the accessional rows.

A. tridentata ssp. wijomingensis is uniform-
ly tetraploid, 2n = 36 for the ca 70 popula-
tions that have been examined, and A. arhus-
cula ssp. longicaulis is uniformly hexaploid, n
= 27 for seven populations including Trout
Creek (McArthur 1501), Adell (Sanderson and
McArthur 1590), Toulon (Sanderson and
McArthur 1593), Topaz Lake (Sanderson and
McArthur 1594), Dayton (Sanderson and
McArthur 1595), and Mustang (McArthur and
McArthur 1684) populations cited earlier in
this report (McArthur et al. 1981, McArthur
and Sanderson in review).

Morphological, chemical, and cytological
evidence reveals affinities of A. arhuscula ssp.
longicaulis for both A. a. ssp. arhuscula and A.
tridentata ssp. wijomingensis. In a series of
common garden experiments (Hanks et al.
1973, McArthur and Plummer 1978, Welch
and McArthur 1979, 1981, 1986, Nelson and
Krebill 1981, McArthur et al. 1981, McArthur
and Welch 1982, Welch et al. 1987), Lahontan
sagebrush accessions (Ul and U58 — Trough
Springs, NV, and U55 — Leonard Creek, NV)
were treated as Wyoming big sagebrush. They
fit with that group in those experiments.

Sagebmsh taxa hybridize natunilly (McAithur
et al. 1988). We propose that A. arhuscula ssp.

156

Great Basin Naturalist

[Volume 55

Table 2. Percent frequency of shared phenolics for A. nova (ARNO), A. tridentata ssp. tridentata (ARTR'), A. tridenta-
ta ssp. vaseijana (ARTR^), A. tridentata ssp. wyoiniii^cn.sis (ARTR^^), A. arhuscula ssp. arbii.scula (ARAR"), and A. arbus-
cula ssp. Ion<iicaulis (ARAR'); data from W^ilt et al. (1992).

ARNO

ARTR'

ARTR^

ARTR"

ARAR^'

ARARl

ARNO
ARTRf
ARTR^
ARTR^^
AR.\R^

54
54

longicaulis is an allopolyploid derivative from
diploid, n = 9, A. arhuscula ssp. arhuscula, and
tetraploid, n = 18, A. tridentata ssp. wyoinin-
gensis. A. arhuscula ssp. arhuscula occurs in
both diploid and tetraploid populations in the
northwestern Nevada area. Tetraploid A. triden-
tata ssp. luyotningensis also occurs there
(McArthur and Sanderson in review). Popula-
tion dynamics of sagebioish populations migrat-
ing in response to climatic change during the
various phases of ancient Lake Lahontan and
other pluvial lakes of the Lahontan basin
(Morrison 1965) could have provided the
opportunity for the origin and establishment
of A. arhuscula ssp. longicatdis. Morphological,
cytological, and chemical data are consistent
with the hybrid origin hypothesis.

Acknowledgments

We thank Gaiy Brackley, Jim Brunner, Chu
Ge-lin, Sherel Goodrich, Craig Plummer,
Stewart Sanderson, Marty Wilt, and the late
Peny Plummer for their help in various phas-
es of this study. We thank field personnel of
the Soil Conservation Sei-vice, U.S. Department
of Agriculture, Nevada offices, who provided
much of the information for Figure 1. We also
appreciate assistance provided by the
Pittman-Robertson W82R Wildlife Habitat
Project and U.S. Department of Agriculture
Cooperative State Research Service Grant 91-
38300-6157.

Literature Cited

Beetle, A. A. 1959. New names within the section
Tridcntatae o{ Artemisia. Rhodora 61: 82-85.

. 1960. A study of sagebrush, the section Tridcnta-
tae of Arienmia. Wyoming Agricultural Experiment
Station Bulletin 368: 1-83.

Beetle, A. A., and A. Young. 1965. A third subspecies in
the Artemisia tridentata complex. Rhodora 67:
405-406.

Brunner, J. R. 1972. Obsei-vations on Artemisia in Nevada.
Journal of Range Management 25: 205-208.

FOSBERG, M. A., AND M. HiRONAKA. 1964. Soil properties
affecting the distribution of big and low sagebiiish
comnumities in southern Idaho. Forage plant physi-
ology and soil-vegetation relationships. Special
Publication of the American Societv' of Agronomy 5:
230-236.

Goodrich, S., E. D. McArthur, and A. H. Wlnward.
1985. A new combination and a new variety' in Artem-
isia tridentata. Great Basin Naturalist 45: 99-104.

Hall, H. M., and F E. Clements. 1923. The phylogenet-
ic method in taxonomy, the North American species
of Artemisia, Chrijsothamniis, and Atriplcx. Publica-
tions of the Carnegie Institute of Washington 326:
1-3,55.

Hanks, D. L, E. D. McArthur, R. Stevens, and A. P
Plummer. 1973. Chromatographic characteristics
and phylogenetic relationships of Artemisia, section
Tridentatae. USDA Forest Service Research Paper
INT-141. U.S. Department of Agriculture, Forest
Service, Intermountain Forest and Range Experi-
ment Station, Ogden, UT. 24 pp.

McArthur, E. D. 1983. Taxonomy, origin, and distribu-
tion of big sagebrush {Artemisia tridentata) and allies
(subgenus Tridentatae). Pages 3-13 in K. L. Johnson,
editor. Proceedings of the first Utah shrub ecology
workshop. Utah State University, Logan. 50 pp.

McArthur, E. D., and S. Goodrich. 1986. Artemisia tri-
dentata ssp. spiciformis: distribution and t;t\onomic
placement. Pages 55-57 in E. D. McArthur and B. L.
Welch, compilers. Proceedings — symposium of the
biology of Artemisia and Chrysothamnus. USDA
Forest Service General Technical Report INT-200.
Intermountain Research Station, Ogden, UT. 398
pp.

McArthur, E. D., and A. P Plummer. 1978. Biogeography
and management of important western shrubs: a
case study, section Tridentate of Artemisia. Great
Basin Naturalist Memoirs 2: 229-243.

McArthur, E. D., and S. C. Sanderson. In review.
Additional chromosome counts in subgenus Tridenta-
tae of Artemisia with implications.

McArthur, E. D., and B. L. Welch. 1982. Growth rate
differences among big sagebrush {Artemisia tridenta-
ta) subspecies. Journal of Range Management 35:
396-401.

McArthur, E. D., C. L. Pope, and D. C. Freeman. 1981.
Chromosomal studies of subgenus Tridentatae of
Artemisia: evidence for autopolyploidy. American
Journal of Botany 68: 589-605.

McArthur, E. D., B. L. Welch, and D. L. Nelson.
1985. Developing improved cultivars of sagebnishes
and other composite shrubs. Pages 188-196 in J. R.
Carlson and E. D. McArthur, editors. Symposium:

1995]

Lahontan Sagebrush, A New Taxon

157

range plant improvement, proceedings — selected
papers presented at the 3Stli annual meeting. Society
for Range Management, Denver, CO. 220 pp.

McArthur, E. D., B. L Welch, and S. C. Sani^erson.
1988. Natural and artificial hybridization between
big sagebrush {Artemisia tridentata) subspecies.
Journal of Heredity 79: 268-276.

Morrison, R. B. 1965. Quatemaiy geology of the Great
Basin. Pages 265-285 in H. E. Wright', Jr., and D. G.
Frey, editors. The Quatemai-y of the United States.
Princeton Universit)' Press, Princeton, NJ. 922 pp.

MoziNGO, H. N. 1987. Shnibs of the Great Basin. Univer-
sity of Nevada Press, Reno. 342 pp.

Nelson, A. 1900. New plants from Wyoming — XII. Bul-
letin of the Ton-ey Botanical Club'27: 258-274.

Nelson, D. L., and R. G. Krebill. 1981. A sagebrush wilt
disease of unknown cause. Great Basin Naturalist 41:
184-191.

NUTTALL, T. 1841. Descriptions of new species and genera
of plants in the natural order of the Compositae.
Transactions of the American Philosophical Societv
II, 7: 357-453.

Rosentreter, R., and R. G. Kelsey. 1991. Xeric big
sagebrush, a new subspecies in the Artemisia triden-
tata complex. Journal of Range Management 44:
330-335.

Shultz, L. M. 1986. Ta.\onomic and geographic limits of
Artemisia subgenus Tridentatae (Beetle) McArthur
(Asteraceae: Anthemideae). Pages 20-28 in E. D.
McArthur and B. L. Welch, compilers. Proceed-
ings— symposium of the biology of Artemisia and
Chrysothamniis. USDA Forest Sendee General Tech-
nical Report INT-200. Intermountain Research
Station, Ogden, UT 398 pp.

Ward, G. H. 1953. Artemisia, section Seriphidiiim, in
North America, a cytotaxonomic study. Contributions
of the Dudley Herbarium 4: 155-205.

Welch, B. L., and E. D. McArthur. 1979. Variations in
winter levels of crude protein among Artemisia tri-
dentata subspecies grown in a unifomi garden. Journal
of Range Management 32: 467—169.

. 1981. Variation of monoterpenoid content among

subspecies and accessions of Artemisia tridentata
grown in a uniform garden. Journal of Range Manage-
ment 34: 380-384.

1986. Wintering mule deer preference for 21 acces-

a superior selection of low-elevation mountain big
sagebrush. USDA Forest Service Research Paper
INT-370. Intermountain Research Station, Ogden,
UT 10 pp.

Welch, B. L., E. D. McArthur, .^nd R. L. Rodriguez.
1987. Variation in utilization of big sagebrush acces-
sions by wintering sheep. Journal of Range Manage-
ment 40: 113-115.

Welsh, S. L., N. D. Atwood, S. Goodrich, and L. C.
HiGGiNS. 1993. A Utah flora. 2nd edition, revised.
Brigham Young Universitv Print Services, Provo,
UT 986 pp.

Wilt, F M., J. D. Geddes, R. V Tamma, G. C. Miller,
and R. L. Everett. 1992. Interspecific variation of
phenolic concentrations in persistent leaves among
six taxa from subgenus Tridentatae of Artemisia
(Asteraceae). Biochemical Svstematics and Ecology
20: 41-52.

WiNWARD, A. H. 1970. Taxonomic and ecological relation-
ships of the big sagebmsh complex in Idaho. Unpub-
lished dissertation. University of Idaho, Moscow. 80
pp.

. 1975. Evolutionaiy development oi' Artemisia tri-
dentata taxa. Page 163 in H. C. Stutz, editor.
Proceedings of the symposia and workshop, wild-
land shrubs. Brigham Young Universit\', Provo, UT.
168 pp.

. 1980. Taxonomy and ecology of sagebrush in

sions of big sagebrush. Great Basin Naturalist 46:
281-286.
Welch, B. L., E. D. McArthur, D. L. Nelson, J. C.
Pederson, and J. N. Davis. 1986. 'Hobble Creek' —

Oregon. Oregon Agricultural Experiment Station
Bulletin 642: 1-15.

WlNWARD, A. H., and E. W. Tisdale. 1977. Taxonomy of
the Artemisia tridentata complex in Idaho. Univer-
sity of Idaho College of Agriculture. Idaho Agricul-
tural Experiment Station Bulletin 19: 1-15.

WiNWARD, A. H., E. D. McArthur, and G. K. Brackley.
1991. Lahontan sagebrush — another sagebrush in
Nevada. Page 13 in Abstracts, 44th annual meeting,
Society' for Range Management, Denver, CO. 45 pp.

WiNWARD, A. H., E. D. McArthur, D. A. K.\ffer, C. A.
Plummer, and G. K. Br^ackley. 1986. Another sage-
brush in Nevada. Technical Note: Range Nevada 44.
USDA Soil Conservation Sei-vice, Reno, NV 2 pp.

WooLF, C. M. 1969. Principles of biometry. D. Van
Nostrand Co., Inc., Princeton, NJ. 359 pp.

Zamora, B., and R T Tueller. 1973. Artemisia arbuscula,
A. longiloba, and A. nova habitat t\'pes in northern
Nevada. Great Basin Naturalist 33: 225-232.

Received 1 March 1994
Accepted 11 August 1994

DOUGLAS-FIR TUSSOCK MOTH {ORGYIA PSEUDOTSUGATA
MCDUNNOUGH) ON SUBALPINE FIR IN NORTHERN UTAH

E. Matthew Hansen'

Abstract. — Douglas-fir tussock moth (Or<^yi(i pscitdutsu^ata McDunnough) defoliation was detected by aerial sur-
vey on three areas of the Wasatch-Cache National Forest in 1990 and 1991. These are the first documented tussock-
moth outbreaks in Utah. Ciround surveys revealed that subalpine fir (Abies lasiocarpa [Hook.] Nutt.) was heavily defoli-
ated dvn-ing the oud^reak. Douglas-fir (Pseiidotsufia inenziesii [Mirb.] Franco), though a minor component in the affected
areas, had noticeably less defoliation and mortalib,'. Adjacent stands of Douglas-fir had little or no visible tussock moth
activit)'. Defoliation on subalpine fir was b,'picalK' found e\enK distributed throughout the crown rather than concen-
trated at the top. Ninety-four percent of subalpine fir with defoliation ratings of 90% or more were killed. Top-kill
occuned on nearly one-half of subalpine firs defoliated 25-89%. Heavily defoliated trees tended to occur in pockets
bounded by areas of light defoliation. After three consecutive \'ears of defoliation, tussock moth populations collapsed.
No life stages were detected in 1993 from visual inspections (jf foliage or in pheronione traps.

Key words: tmsock moth, suhcdpine fir. dejolhitors. Doiighis-fin Utcd^ fi)rests. jorest insects.

The Douglas-fir tussock moth {Or^yia
pseiidotsugata McDunnough) is a significant
defohator of Douglas-fir and true firs through-
out its host range in western North America.
Tussock moth outbreaks on the Wasatch-
Cache National Forest, UT from 1990 to 1992
were unique because subalpine fir {Abies lasio-
carpa [Hook.] Nutt.) experienced greater defo-
liation and mortalit)' than Douglas-fir {Psciido-
tsuga menziesii [Mirb.] Franco). These tussock
moth infestations are the first documented
outbreaks in Utah (Ollieu 1978, Tunnock et al.
1985).

Previous tussock moth outbreaks indicate
three primary hosts depending on location. In
British Columbia and northern Washington,
Douglas-fir is preferred; in southern Washing-
ton, Oregon, and Idaho, Douglas-fir, white fir
[Abies concolor [Cord, and Clend.]), and grand
fir (A. grandis [Dough] Lindl.) are preferred;
in California, Nevada, Arizona, and New
Mexico, white fir is preferred (Wickman et al.
1981). In these areas subalpine fir and other
members of the pine family are t\picall\' defoli-
ated after the preferred host is consumed. The
Wasatch-Cache National Forest outbreaks
experienced light defoliation of Douglas-fir
within stands of completely defoliated sub-
alpine fir.

InterestingK', the first record of tussock moth
defoliation in the United States was on sub-
alpine fir at larbidge, NV, in 1927 (Balch 1930).
Affected areas were of subalpine fir occurring
either purely or in mixture with limber pine
{Piniis flexilis James). These two conifers and
quaking aspen {Popuhis tremuloides Michx.)
formed virtually the entire forest at Jarbidge
(Balch 1932) in contrast to the Utah sites that
contain or are surrounded by substantial
Douglas-fir. Balch s records are the only litera-
ture indicating subalpine fir to be a principal
host, while more recent authors consider the
species to be secondary (Wickman et al. 1981,
Benyman 1988).

The objective of this study was to docu-
ment the effects of these outbreaks on sub-
alpine fir. Information will be used to predict
future impacts of tussock moth defoliation on
subalpine fir resource. This report summarizes
stand conditions following infestation and
attempts to characterize the effects of defolia-
tion on individual trees.

Methods

Aerial detection surveys delineated tussock
moth activity near the Baxter Sawmill site,
Ogden Ranger District, Wasatch-Cache Nation-
al Forest, UT, in 1990. Additional infestations

'USDA Forest Senice. Forest Pest Management. 4746 South 1900 East, OKclen, UT 84403.

158

1995]

Tussock Moth on Subalpine Fir

159

were detected at Blind Hollow and Amazon
Hollow, Logan Ranger District, Wasatch-Cache
National Forest, UT, in 1991. Subsequent
ground sui-veys confirmed tussock moth popu-
lations at these locations.

A total of 35 pairs of 20 basal-area-factor
variable plots and 1/300 ac seedling/sapling
fixed plots were established in July 1992 at the
three areas. Plots were installed along a tran-
sect at 100-m intervals starting from a refer-
ence point in areas with visible defoliation. No
other attempt was made to randomize plot
locations or to cover the entire affected area.
Tree data collected include species, diameter
at breast height (DBH), height, age, 5- and 10-
year radial growth, insect/disease damage, and
an ocular estimate of percent defoliation.
Additionally, the distal 18 inches of three ran-
domly selected lower crown branches on all
host species were examined for pupae or egg
masses.

Plots were remeasured in July 1993 to eval-
uate changes in insect-related tree injury and
mortality, percent defoliation, and presence of
tussock moth life stages. One pair of plots at
Baxter Sawmill was cut and lost during salvage
operations. To maximize data-collection con-
sistency, the same individual conducted defoli-
ation estimates on 32 of 35 plots for plot estab-
lishment and remeasurement.

INDIDS, the insect and disease damage
survey program (Bousfield et al. 1985), was
used to calculate per-acre average stand char-
acteristics. Individual tree defoliation ratings
were classified into the five categories of
Weatherby et al. (1992): (1) undamaged, (2)
lightly defoliated (1-24% of total needle com-
plement defoliated), (3) moderately defoliated
(25-74%), (4) heavily defoliated (75-89%), and
(5) very heavily defoliated (90+%). Defoliation

ratings from 1992 and 1993 were used to de-
termine defoliation effects on tree condition.

Results

Blind Hollow

Prior to the outbreak, composition of all
live trees greater than 5 inches DBH was 69%
subalpine fir, 28% aspen, and 3% Douglas-fir
Total pre-outbreak live basal area was 178.9 sq
ft/ac. By 1993 live basal area was reduced to
166.8 sq ft/ac. Site elevations range from 7800
to 8100 ft. Aspect ranges from southwest to
south to east on slopes varying from 10 to
50%.

Subalpine fir seedlings and saplings (0-4.9
inches DBH) were significantly affected by
tussock moth. Sixty stems per acre in this size
class representing 25% of the stocking were
killed (Table 1). Pole-sized subalpine fir (5-8.9
inches DBH) were more frequently defoliated
than larger diameter classes. Fifty-eight per-
cent of stems 5-8.9 inches DBH, 34% of stems
9-11.9 inches DBH, and 39% of stems >12
inches DBH were defoliated. Stems exhibiting
top-kill increased proportionately with per-
cent defoliation. Four percent of subalpine fir
stems over 5 inches DBH were killed by tus-
sock moth.

In the >12-inches diameter class, none of
7.3 Douglas-fir per acre were visibly defoliat-
ed (Table I). Among subalpine fir in this class,
3% of 65.1 per acre were defoliator killed.
Twenty-eight percent survived defoliation,
while 69% were not visibly defoliated.

Western balsam bark beede [Dryocoetes con-
fiisiis Swaine) killed 4.9 subalpine fir stems per
acre. These trees were attacked in 1991, coin-
ciding with the peak of the tussock moth out-
break.

Table 1. Trees per acre condition sunimaiy of subalpine fir and Douglas-fir following a Douglas-fir tussock moth out-
break. Blind Hollow, Wasatch-Cache National Forest, July 1993. Summary calculated from 10 variable/fi.xed plot pairs.
SAF = subalpine fir, DF — Douglas-fir.

Undamaged
SAF DF

Defoliation

class

Diameter

Light

Moderate
SAF DF

Heav)'

VePi'
SAF

heav)-
DF

Mortality'

class

SAF DF

SAF

SAF DF

0-4.9"

5-8.9"
9-11.9"
12" -H

30.0
39.0
25.9
40.0

0.0
0.0
2.9

7.3

120.0 0.0

16.0 0.0

7.0 0.0

13.4 0.0

30.0

18.5

6.0

4.8

0.0
0.0
0.0
0.0

0.0
6.0
0.0
0.0

0.0
0.0
0.0
0.0

0.0
6.2
0.0
0.0

0.0
0.0
0.0
0.0

60.0 0.0

6.0 0.0
0.0 0.0

2.1 0.0

160

Great Basin Natuiulist

[Volume 55

The 1992 survey found an average ol 3.1
pupae and 0.5 egg masses per three-branch
samples. The 1993 survey found no current
life stages on any sample tree, and no life
stages were visible in the area.

Baxter Sawmill

Prior to the outbreak, composition for all
live trees greater than 5 inches DBH was 65%
subalpine fir, 25% aspen, and 10% Douglas-fir.
Total live basal area was 176.1 sq ft/ac at the
onset of the outbreak. Live basal area in 1993
was 112.8 sq ft/ac. Site elevations range from
7400 to 7900 ft. Aspect is south, southwest,
west, and northwest on slopes vai^ving from 10
to 30%.

Subalpine fir seedlings and saplings had
considerable defoliator damage. More than
250 seedlings and saplings per acre, or 55% of
stocking in this size class, died (Table 2). Most
surviving seedlings and saplings were only
lightly defoliated. Forty-nine percent of sub-
alpine fir stems 5.0-11.9 inches DBH were
killed by tussock moth. Trees with top-kill
increased proportionately with percent defoli-
ation. Only 3% of subalpine fir stems in the
lightly defoliated category experienced top-
kill, compared to 92% of surviving trees in the
heavily and very heavily defoliated classes.

In the >12-inches diameter class, Douglas-
fir had 10% of 22.6 trees per acre defoliator
killed. Fift\'-seven percent were not defoliated,
with another 33% defoliated but surviving
(Table 2). Among 38.5 subalpine fir per acre in
this size class, 7% were defoliator killed and
77% were defoliated but sui-vived.

Western balsam bark beetle has also been
active at Baxter Sawmill, killing 38.2 subalpine
fir per acre, mostly in 1990 or 1991. Annosus
root disease {Heterobasidion anuosiiin [Fn] Bref )
was found on 4.6 subalpine fir per acre.

The 1992 survey found an average of 4.8
pupae and 1.2 egg masses per three branches
sampled. No cmrent life stages were found in
1993 on the plots or in the area. Additionally, no
tussock moths were caught in pheromone traps
placed in the Baxter Sawmill area in 1993.

Amazon Hollow

Prior to the outbreak, composition of all
live trees greater than 5 inches DBH was 73%
subalpine fir, 24% aspen, 2% Douglas-fir, and
1% lodgepole pine. Total live basal area was
125.5 sq ft/ac at the onset of the outbreak.
Live basal area in 1993 was 72.2 sq ft/ac. Site
elevations range from 7500 to 7800 ft. Aspect
is east on slopes vaiying from 10 to 25%.

One-hundred subalpine fir seedlings and
saplings per acre, or 10% of stocking in that
class, were killed (Table 3). Mortality in the
three size classes greater than 5 inches DBH
i-anged from 50 to 62%. Top-kill was common
for all defoliation intensities. Of the sui'viving
defoliated subalpine fir (>5 inches DBH),
60% had top-kill, including 63% of stems clas-
sified as lightly defoliated.

In the >12-inches size class, 28% of 4.3
Douglas-fir per acre were defoliator killed
with another 16% defoliated but surviving
(Table 3). Among 29.7 subalpine fir per acre in
that class, 50% were defoliator killed and
another 31% were defoliated but sui^vived.

Western balsam bark beetle killed 2.6 sub-
alpine fir per acre. Annosus root disease was
found on 4.2 trees per acre.

The 1992 sin-vey found 2.0 pupae and 0.6 egg
masses per three branch samples. The 1993
sui'vey failed to detect any current life stages.

Sample Tree Summaiy

Two-hundred ninety-one host sample trees
were rated for defoliation and monitored for

Table 2. Trees per acre condition suniman' ot subalpine tir and Douglas-fir following a Douglas-fir tussock moth out-
break, Baxter Sawmill, Wasatch-Cache National Forest, July 1993. Summaiy calculated from 13 \ariable/fixed plot pairs.
SAF = subalpine fir, DF = Douglas-fir

Undamaged
SAF DF

Defoli

ation class

Morta
SAF

Diameter

Light
SAF DF

Moderate

avy

Veiy heavy
SAF DF

lit)'

class

SAF

0-4.9"

23.1

0.0

13S.5

0.0

23.1

0.0

23.1

0.0

2,53.5

0.0

5-8.9"

12.9

0.0

29.3

0.0

9-11.9"

2.5

0.0

7.8

0.0

2.5

0.0

2.1

0.0

25.0

0.0

12" -h

3.4

12.S

17.9

6,5

2.7

1.0

0.0

1.3

0.0

2.7

2.2

1995]

Tussock Moth on Subalpine Fir

161

Table 3. Trees per acre condition suninian of subalpine tir and Douglas-fir following a Douglas-fir tussock moth out-
break, Amazon Hollow, Wasatch-Cache National Forest, July 1993. Summai-y calculated from 12 variable/fixed plot
pairs. SAF = subalpine fir, DF — Douglas-fir

iameter

IISS

Undamaged
SAF DF

Defol

iation class

Morta
SAF

Lii

^ht

Mod.

erate

Heavy
SAF DF

Veiy heavy
SAF DF

lity

cl;

SAF

0-4.9"

500.0

0.0

275.0

0.0

75.0

0.0

25.0

0.0

100.0

25.0

-8.9"

6.8

0.0

13.4

0.0

4.3

0.0

40.4

0.0

-11.9"

5.0

0.0

9.5

0.0

3.1

0.0

23.0

0,0

'." +

5.7

2.4

5.5

0.4

0.7

0.3

1,2

0.0

1.8

0.0

14.8

1.3

sur\i\'al (Tables 4, 5). Defoliator-caused mor-
tality was found to increase with the degree of
defoliation. In the very heavily defoliated class,
94% of subalpine firs and 100% of Douglas -firs
were killed. None of the sample trees in the
lightly defoliated class were killed. Incidence
of top-kill also increased with degree of defoli-
ation, although trees in the heavily and very
heavily defoliated classes were more likely to
succumb than exhibit top-kill. This parallels
other tussock moth study results, where
degree and incidence of top-kill is related to
severity of defoliation (Wickman 1978).

Sui-viving defoliated trees began to recover
by 1993 (Tables 4, 5). Average defoliation rat-
ing for lightly defoliated subalpine fir in 1992
was 7.7%. In 1993 the same trees had an aver-
age rating of 3.9% with no visible defoliation
of that year's needles. The other defoliation
classes for subalpine fir and Douglas -fir had
similar recoveries. Some of the most dramatic
recoveries, however, can be partially attrib-
uted to the most heavily defoliated trees of
their respective classes succumbing and there-
fore not being rated in 1993.

Discussion

Although Douglas-fir tussock moth had been
previously captured in pheromone traps in
Utah, the Wasatch-Cache outbreaks are the
first to be documented in the state (Tunnock et
al. 1985). More significantly, a literature review
revealed the Wasatch-Cache outbreaks to be
unique in that subalpine fir is apparently the
preferred host type. Balch's (1930, 1932) stud-
ies are the only that list subalpine fir as a pri-
mary host. More recent literature indicates sub-
alpine fir to be secondaiy to Douglas -fir, white
fir, or grand fir (Wickman et al. 1981, Johnson
and Lyon 1988). At the Wasatch-Cache out-
breaks, subalpine fir appears to be preferred

over Douglas-fir. All three study sites are in
close proximity to stands where Douglas -fir is
the primary overstory component. These Doug-
las-fir stands experienced little or no visible
defoliation. This contrasts to Balch's Jarbidge,
NV, site where subalpine fir, limber pine, and
quaking aspen "form practically the whole of
the forest" (Balch 1932).

Another exception to the tussock moth's pref-
erence for Douglas-fir, white fir, or grand fir
has been observed in urban areas along the
Colorado Front Range. In these cases blue
spruce {Picea piingens Engelm.) has been the
preferred host over white fir and Douglas-fir
(D. Leatherman-, personal communication).
In Colorado's native forests Douglas-fir is the
principal host.

The defoliation pattern seen on the Wasatch-
Cache National Forest outbreaks differed great-
ly from previously recorded patterns, such as
in Oregon's Blue Mountains. Wickman (1978)
recommends estimating defoliation "according
to the percent of crown totally defoliated from
the top down. " That technique was abandoned
for this study because most needle loss was
distributed evenly throughout the crown rather
than concentrated at the top. Application of
Wickman's method would have misrepresent-
ed many trees with significant defoliation by
having them rated at < 10% defoliation. In
other words, the Wasatch-Cache National
Forest outbreaks did not fit the "top down"
defoliation pattern observed in other outbreaks
(J. Weatherby^, personal communication).

This study indicates that subalpine fir may
be locally more susceptible to tussock moth
mortality than either grand fir or Douglas-fir.
Despite the difference in percent defoliation

"Entomologist, Colorado State Forest Service,

■^Entomologist, USDA Forest Service, Forest Pest Management. Boise,
Idaho.

162

Great Basin Naturalist

[Volume 55

Table 4. Condition of subalpine fir sample trees within tussock moth nionitorinij plots at Blind Hollow. Baxter
Sawmill, and Amazon Hollow, Wasalch-Caehe National Forest, July 1993.

Average 1992

Average 1993

Class limits

defoliation

defoliation-

Total no.

Top-

-kill

Mortality

Defoliation class^

(% defoliation)

of trees

# ^/c

Undamaged

0.0

0 0

Light

1-24

7.7

3.9

0 0

Moderate

25-74

39.5

36.3

3 10

Heav"\

75-89

77.5

5 55

Very heavy

>9()

91.3

72.5

59 94

'Trees assigned defoliation class based on 1992 deluliation ratings.
^Surviving trees from 1992 defoliation class.

T.VBLE 5. Condition of Douglas-fir sample trees within tussock moth monitoring plots at Blind Hollow, Baxter
Sawmill, and Amazon Hollow, Wasatch-Cache National Forest, July 1993.

Average 1992
defoliation

Average 1993
defoliation- Total no.

Top-kill

Mortality

Defoliation classl

(% defoliation)

('7c defoliation)

(% defoliation)

of trees

Undamaged

0.0

Light

1-24

6.5

2.0

Moderate

25-74

40.0

21.7

Heavs'

75-89

80.0

65.0

Very heav\

>90

95.0

—

100

'Trees assigned defoliation class based on 1992 defoliation ratings.
^Surviving trees from 1992 defoliation class.

estimation techniques, the Wasatch-Cache
results can be compared to those of Wickman
(1978). At the 90% defohation level, Wickman
found 24% grand fir mortality and 30% Douglas -
fir mortality (90% defoliation in Wickman's
study means complete defoliation in the top
90% of the live crown). At the Wasatch-Cache
outbreaks, 57% of subalpine fir defoliated at
90% were killed (90% defoliation using the
methodology of this study means that 90% of
the estimated total needle complement was
consumed). At the 99% defoliation level, Wick-
man found that grand fir died at 53% and
Douglas-fir at 46%. This compares to 96%
mortality on Wasatch-Cache subalpine fir
rated at 95% defoliation.

Within the infested stud\' areas, the degree
of damage varies greatly from one plot to the
next. One plot at Amazon Hollow had all host
type defoliator killed, while a plot 100 m dis-
tant was only lightly defoliated. Although the
very heavily defoliated areas are restricted in
size (usually less than 5 ac), the amount of mor-
tality in these pockets is substantial. An area
not sampled, at Ba.xter Sawmill due to salvage
logging operations, included over 20 ac where

virtually all host t\'pe was killed. Many of these
areas are bounded by stands of similar compo-
sition and density that were only lightly defoli-
ated. In a study of five case histories in Oregon
and California, Wickman et al. (1973) found
almost one-half of tree mortality occurring in
patches coinciding with high moth population
centers.

Douglas-fir tussock moth outbreaks typical-
ly span two to four years. Moth populations
develop rapidly and then abruptly subside
after only one to two years of outbreak popula-
tions (Wickman et al. 1981). The Wasatch-
Cache outbreaks have followed this pattern.
Moderate to heavy defoliation at Baxter Sawmill
was first detected from aerial suney in 1990;
defoliation was very heavy in 1991. In 1992
moth activity dramatically declined, and in
1993 no life stages were discovered by either
visual inspection or pheromone trapping.

While it is be>'ond this stud\ s scope to iden-
tify causal agents that initiated the Wasatch-
Cache outbreaks, it should be noted that a
prolonged drought coincided with the infesta-
tion. Most damage occurred on drier sites,
such as ridge tops and southerly facing slopes.

1995]

Tussock Moth on Subalpine Fir

163

This corresponds to patterns seen in other
outbreaks (Bergstrom 1980). The affected
trees were apparently drought stressed at the
time of defohation. The sudden moth popula-
tion collapse mimics that of other outbreaks
where a nuclear polyhedrosis virus appears to
be the major mortality factor (Wickman et al.
1973).

Conclusion

Although uncommon, Douglas-fir tussock
moth can cause considerable damage to sub-
alpine fir. While damage in the three study
areas was variable, pockets of heavy defoliation
had substantial subalpine fir mortality. Larger-
diameter trees are apparently less susceptible
to mortality except in these pockets where vir-
tually all host tv'pe was killed. Although a minor
component in the heavily defoliated areas, local-
ly Douglas-fir appears to be less-preferred
host material. All study areas are in close prox-
imity to Douglas-fir stands that exhibited little
or no tussock moth activity. Western balsam
bark beetle and annosus root disease con-
tributed to subalpine fir mortality, though visi-
bly minor relative to defoliator impacts. While
forecasting losses in volume would be difficult
based on this study, the fate of individual trees
can be reasonably predicted given degree of
defoliation.

Acknowledgments

I am grateful for the many people who
helped with this project. David Leatherman
(Colorado State Forest Sendee), Julie Weather-
by, Steve Munson, and John Anhold (all Forest
Pest Management, Intermountain Region) pro-
vided critical review of the manuscript. Alan
Dymerski, John Guyon, Dawn Hansen, John
Anhold, Valerie DeBlander (all Forest Pest
Management, Intermountain Region), Jill
Ansted, Craig Yanase, and Lisa Robinson (all
Utah Department of Agriculture) assisted with
data collection. Julie Weatherby and John
Anhold provided input for the sui-vey design.

Dawn Hansen, Cindy Hampton, John Cuyon,
and Bent Olsen (all Forest Pest Management,
Intermountain Region) helped with data pro-
cessing, table preparation, and editing. Irene
Voit (Intermountain Research Station) assisted
with the literature search.

Literature Cited

Balch, R. E. 1930. The fir tussock moth reveals abihty to
cause serious damage. Forest Worker 6(2): 17-18.

. 19.32. The fir tussock moth (Hemerocmnpa psetido-

tsiigata McD.). Journal of Economic Entomology 25;
1143-1148.

Bergstrom, D. 1980. New lessons from old tussock moth
outbreaks. USDA Forest Service, Pacific Northwest
Research Station. 4 pp.

Berryman, a. a. 1988. Dynamics of forest insect popula-
tions. Plenum Press, New York, NY.

BousFiELD, W, R. Eder, and D. Bennett. 1985. User's
guide and documentation for insect and disease
damage survey (INDIDS). Rl-85-19. USDA Forest
Service, Northern Region, Missoula, MT.

Johnson, W. T, and H. H. Lyon. 1988. Insects that feed
on trees and shrubs. 2nd edition. Cornell University
Press, Ithaca, NY.

Ollieu, M. 1978. Detection of Douglas-fir tussock moth
in the Intermountain Region using baited sticky
traps. USDA Forest Service, Intermountain Region,
Ogden, UT. 7 pp.

TuNNOCK, S., M. Ollieu, .\nd R. W Thier. 1985. Histoiy
of Douglas-fir tussock moth and related suppression
efforts in the Intermountain and northern Rocky
Mountain regions — 1927 through 1984. USDA Forest
Service, Report 8.5-13. Intermountain and Northern
Regions, Missoula, MT.

Weatherby, J. C, K. A. Knapp, B. R. G.\rdner, J. Roberts,
and P Mocettlnt. 1992. A biological evaluation of
the Douglas-fir tussock moth outbreak in southern
Idaho, 1991. USDA Forest Sei-vice, Report R4-92-
01. Intermountain Region, Ogden, UT.

WiCKNL^N, B. E. 1978. How to estimate defoliation and
predict tree damage. USDA Agriculture Handbook
550.

Wickman, B. E., R. R. Mason, and C. G. Thompson.
1973. Major outbreaks of the Douglas-fir tussock
moth in Oregon and California. USDA Forest
Service. General Technical Report PNW-5.

Wickman, B. E., R. R. Mason, and G. C. Trostle. 1981.
Douglas-fir tussock moth. USDA Forest Service.
Forest Insect and Disease Leaflet 86.

Received 21 April 1994
Accepted 14 November 1994

SEASONAL NUTRIENT CYCLING IN POTAMOGETON PECTINATUS
OF THE LOWER PROVO RIVER

C. Mel Lytle' and Bruce N. Sniitlii-2

Abstiuct. — A common submersed aquatic plant of Great Basin wetland and riverine systems, Potamogeton pectina-
tiis L. (sago pondweed) is a key waterfowl food. Nutritional qualities of submersed aquatics in the Great Basin are little
understood. The puipose of this study was to determine the seasonal element cycling and nutritional qualities of P.
pcctiiialus drupelet, leaf and root tissues from the lower Provo River. Leaf tissue protein was 27% (dry weight) in July,
hut declined to 15% by December Diiipelet protein content was 9% in July and 6.5% in October. Lignocellulose in leaf
tissue was lowest in July at 34% and increased as the season progressed. Percent fat was highest in leaf tissue at 12% in
|ulv Sugars were highest in P pectinatus leaf tissues in December and July. Calcium and magnesium concentrations
increased in P pectinatus tissues over the entire season. Leaf tissue zinc was 329 ppm (diy weight) in October Leaf iron
concentration was highest in September at 1184 ppm, while root tissue iron was 7166 ppm. Manganese content in leaf
tissue peaked in October at 4990 ppm. Copper concentrations in leaves and roots were variable. High protein in leaf tis-
sue would benefit local nesting and brooding waterfowl populations that feed on this aquatic. Trace metal concentrations
in leaf and root tissues, fi-om possible anthropogenic activities, appear veiy high during fall migratoiy months. Metal
bioaccumulation by this species in other Great Basin wetlands and possible metal toxicity in waterfowl warrant ftuther
study.

Key words: sago pondweed. Potamogeton pectinatus, nutritional qualities, trace element cycling, metal bioaccumulation,
waterfowl.

A common submersed aquatic plant of the
Great Basin, Potamogeton pectinatus is a key
priman' producer in fresh and sahne wetlands
(Kantrud 1990). Waterfowl feed on all plant
parts including drupelet, leaf, and root tissues
(Anderson and Ohmart 1988, Korschgen et al.
1988). Sherwood (1960) noted that whisding
swans {Olor columhianus) fed heavily on tubers
and drupelets during fall migration in the Bear
River Migratory Bird Refuge and Ogden Bay
Refuge. Other waterfowl species — Canada
geese {Branta canadensis), mallards {Anas platy-
rhynchos), pintails {Anas acuta), gadwalls {Anas
strepera), canvasbacks {Aytha vallisneria), and
redheads {Aytha americana) — also fed on P.
pectinatus leaf and root tissues. Localized inter-
mountain trumpeter swan {Cygnus buccinator)
populations are also largely dependent on sub-
mersed aquatic plants as food (Anderson et al.
1986, Henson and Cooper 1993).

Little is known concerning nutrient dynam-
ics and seasonal element cycling of P. pectina-
tus from Great Basin wetlands (Kadlec and
Smith 1989). Consequently, how this aquatic
species may affect waterfowl nutrition is poorly
understood. Most assumptions concerning body

condition and nutritional requirements are
based on studies from other areas of North
America. Yet, energy and sustenance required
by waterfowl species that frecjuent the Great
Basin are largely provided by resident aquatic
plants. Of these, P. pectinatus, Ruppia mariti-
ma L. (widgeon grass), Scirpus mahtimus L.
(alkali bulrush), Scirpus pungens L. (Olney
three-square), Scirpus acutus L. (hardstem
bulrush), and Zannichellia palustris L. (horned
pondweed) are common plant species man-
aged in national refuges and waterfowl man-
agement areas. Potamogeton pectinatus is con-
sidered the most important of these species
for diving and dabbling ducks (Kadlec and
Smith 1989). The purpose of this study was to
determine the seasonal element concentra-
tions and nutritional qualities of P. pectinatus
from a local Great Basin river drainage.

Methods

Plant harvests were conducted monthly
from three locations within the lower Provo
River drainage from July 1991 to December
1991: (1) just below Deer Creek dam (40°24'N,

'Department of Botany and Bange Science. 401 WIDE, Brinliarii VounK I iii\ersit>. Provo. UT 84602.
^Author to whom correspondence should be addressed.

164

1995]

Seasonal Nutrient Cycling in P. pectinatus

165

Table 1. A range of measured water column and sedi-
ment characteristics, pH, and electrical conducti\'it\' (EC)
from the lower Provo River drainage.

Water

Claritv

clear-opaque

Velocity (m/sec)

0-0.4

Depth (cm)

5-60

Temperature (°C)

3-14

7.4

EC (/j,mhos/cm'^)

42.5

Sediment

>120

3-12

6.9

1570

Table 2. Mean exchangeable Fe and Mn from lower
Provo River drainage sediments (ppm dr>' weight ± S.E.,
n > 3). Means sharing the same letter are not significantly
different [P < .05).

Depth (cm)

0-7
7-15
15-22
22-30

61.6 ± 0.7a
56.5 ± 1.4a
61.3 ± 1.7a
57.1 ± 0.8a

19.2 ± 1.2a

11.4 ± 0.8b

9.4 ± 2.1b

12.0 ± 1.7b

lll°3rw, elev. 1603 m), (2) near the Sundance
turnoff (40° 22'N, lir34'W, elev. 1560 m), (3)
=200 m from the mouth of the Provo River
near Utah Lake (40°14'N, 111° 44 'W, elev.
1347 m). Water column and sedhiient charac-
teristics measured in the lower Provo River
are found in Table 1. Sediment conditions
ranged from stony with gravelly patches to
silty-clay mud. Stands of P. pectinatus were
most abundant on muddy sediments.

Whole plants (leaf and root tissues) of P. pec-
tinatus were sampled in replicate from each
location. Drupelets, shoot (stems and leaves)
tissues, and belowground (root, rhizomes, and
turions) tissues were separated from plant litter
and sediments. Invertebrates were removed
from samples when rinsed in warm water
(38 °C). Cleaned samples were rinsed in
deionized water and dried in a forced-air oven
at 70 °C. Plant, sediment, and water samples
were analyzed at Brigham Young University,
Department of Agronomy and Horticulture,
Plant and Soil Analysis Laboratoiy Diy plant
tissue samples were weighed and ground in a
Wiley Mill to pass a 40-mesh screen, and 0.25-
g samples were digested in Folin-Wu tubes
with 5 ml of concentrated HNO3. Samples
were left covered for 16 h before digestion in
an aluminum block for 1 h at 100 °C. Three ml
of 70% HCIO4 was added, and samples were
refluxed at 200 °C until the solution cleared
(approx. 30 min). Samples were then brought to
50-ml volume with deionized water (Orson et
al. 1992). Element contents were detected by
direct aspiration into a Perkin-Elmer Model
5000 Atomic Absorption Spectrophotometer.
All blanks and standards were run with the
same procedures. Percent total nitrogen and
phosphorus were determined using a Kjeldahl
digestion followed by analysis with an ALP-
KEM rapid-flow analyzer.

Sediment (0-30 cm) and water (1000 ml)
samples were obtained from the same loca-

tions and intervals as plant samples.
Sediments were air-dried and extracted for
exchangeable iron (Fe) and manganese (Mn)
with diethylenetriaminepenta-acetic acid
(DTPA) and detected by atomic absorption
spectroscopy. Water samples were analyzed
for pH, electrical conductivity (;Umhos/cm^),
and available Fe and Mn with an Orion Micro-
processor Ion-analyzer/901 pH meter, a wheat-
stone bridge, and by atomic absorption spec-
troscopy.

Mean concentrations and standard errors
(S.E.) were determined for each plant, sedi-
ment, and water sample. To determine if sig-
nificant variation in plant tissue nutrient and
element concentrations existed between the
different months, we used analysis of variance
(ANOVA) where month was considered the
fixed effect and sample site the experimental
unit in a repeated measures design. If signifi-
cance (P < .05) was found, Tukey's multiple
comparison procedures were used to separate
means.

Results and Discussion

Available Fe and Mn concentrations in water
samples were 0.06 ± 0.01 and 0.001 ppm.
Sediment exchangeable Fe and Mn contents
were found between the normal soil range of
5-65 ppm. Yet, under anoxic conditions that
are common in sediments, Fe and Mn may be-
come more available for root uptake (Spencer
and Brewer 1971, Tisdale et al. 1985; Table 2).
Significant differences in sediment exchange-
able Mn were found between surface sedi-
ments (0-7 cm) and the rest of the sampled
profile (Table 2).

Element concentrations and forage quali-
ties were determined for P. pectinatus tissues
from July to December. Leaf and root tissue
dry matter, as a percentage of fresh weight,
remained constant at 6-7%, with the highest

166

Great Basin Naturalist

[Volume 55

dry matter content observed in October.
Throughout the season, P. pectinatus element
and forage composition varied with growth
stage. Significant variation in leaf tissue pro-
tein was found (F = 21.69; d.f = 4,14; P <
.001) between July, September, October, and
December (Table 3). Drupelet protein content
was higher in July than in October. In all
months sampled, leaf tissue piotein was high-
er than drupelet protein. Percent protein in
leaf tissue was higher than values reported in
other studies (Linn et al. 1975, Kantrud 1990).
Acid detergent fiber (ADF) analysis revealed
that leaf tissue was lowest in lignocellulose
(fiber) in July, but significant differences (F =
3.03; d.f = 4,14; P = .07) in fiber content
were not observed as the year progressed.
Linn et al. (1975) found P. pectinatus leaf fiber
content of 33% that is similar to values
obsei-ved in this study. Increased fiber content
would decrease the overall forage quality of
leaf tissue. Significant variation did exist (F =
177.40; d.f = 4,14; P < .001) in leaf tissue fat
content and was highest in July. Total non-
structural carbohydrates (sugars) in leaf tissues
were highest in December and differed from
all other months (F = 42.19; d.f = 4,14; P <
.001). By October, drupelet fat and sugar con-
tent were both higher than values found in
July.

Percent nitrogen (N) and phosphorus (P) in
leaf tissue reached peak concentrations in July
but were significantly lower by December (F
= 23.37; d.f = 4,14; P < .001) (F = 79.30; d.f
= 4,14; P < .001; Tible 4). Veniiaak et al. (1983)
stated that P. pectinatus played an important
role in P cycling in aquatic systems. Cultured
P. pectinatus grown in water relatively high in
phosphate (PO4-P) (0.3 ppm) bioaccumidated
p32 jq 4738 times the amount found in the
water column. Nitrogen and P content in P.
pectinatus can be well above that required for
plant growth; this would indicate luxuiy con-
sumption of these elements (Jupp and Spencer
1977, Ho 1979, Madsen 1986). Significant
concentrations of calcium (Ca) and magnesium
(Mg) accumulated (F = 29.12; d.f = 4,14; P <
.001) (F = 278.71; d.f = 4,14; P < .001) in
leaf tissue between Jul\' and December. This
may indicate abiotic deposition, though no vis-
ible encrustation on exterior leaf or stem sur-
faces was obsei-ved. Hutchinson (1975) report-
ed that P. pectinatus leaves were higher in Ca,
Fe, K, Mg, Na, and several micronutrients than
other aquatic plants. Yet, no mention of time
sampled was given for these mineral concen-
trations. Therefore, no knowledge of seasonal
accumulation was determined. Potassium (K)
content was highest in September and differed
significantly from percent K content in July (F

Table 3. Mean piotein, fiber, fat, and sugar content in P. pectinatus drupelet and leaf tissue over fi\'e months. Forage
quality constituents expressed as % diy weight ± S.E., n > 3. Means sharing the same letter are not significantly differ-
ent (P < .05).

Monti 1

Tissue

Protein

ADF^' Fat

TNC'^

July

Leaf

27.4 ± 0.3a

.34.2 ± 0.9a 12.2 ± 0.1a

8.3 ± 0.1a

Aug.

Leaf

24.9 ± 0.3ab

35.6 ± 2.8a 6.5 ± 0.2b

8.1 ± 0.4a

Sept.

Leaf

2L4±0.3b

39.7 ± 0.4a 6.8 ± 0.2b

7.9 ± 0.2a

Oct.

Leaf

20.3 ±L4b

.37.9 ± 1.3a 7.1 ± 1.1b

8.6 ± 0.1a

Dec.

Leaf

15.1 ± 0.3c

38.1 ± 0.5a 5.9 ± 1.1c

11.0±0.1b

July

Dnipelet

9.0 ± 0.5

33.4 ±0.6 6.1 ±0.7

12.0 ± 0.3

Oct.

Drupelet

6.5 + 0.8

36.3 ±1.3 7.4 ±0.8

16.3 ±1.2

"Acid detergent fiber (ADF), a measure of percent lignocellulose or fiber
''Total nonstructural carbohydrate (TNC), a measure of sugars

Table 4. Mean mineral element concentration in P. pectinatus leaf tissue over five months. Element content
expressed as % dr\' weight ± S.E., n > 3. Means sharing the same letter are not significantly different {P < .05).

Month

Tissue

(% j,-^-

wt.)

1,3 ± 0.1a

July

Leaf

4,4 ± 0.7a

0.6 ± 0.1a

1.9 ± 0.2a

0.3 ± ().04a

L2±0.1a

Aug.

Leaf

2.8 ± 1.0a

0.5 ± 0.1a

3.5 ± 0.1b

1.2 ± 0.1a

0.5 ± 0.01b

Sept.

Leaf

3.4 ± 0.1a

0.5 ± 0.2a

3.7 ± 0.1b

1.4±0.04ab

0.6 ± 0.02c

1.8 ± 0.2b

Oct.

Leaf

3.3 ± 0.2a

0.5 ± 0.1a

3.1 ± 0.1b

1.7 ± 0.1b

0.6 ± 0.01c

Dec.

Leaf

2.4 ± 0,1b

0.2 ± 0.1b

3.0 ± 0.3b

2.3 ± 0.1c

0.7 ± 0.02c

0.6 ± 0.1c

1995]

Seasonal Nutrient Cycling in P. pectinatus

167

Table 5. Mean trace element concentration in P. pectinatus leaf tissue over five months. Element content expressed
as ppni cli"y weight ± S.E., n > 3. Means sharing the same letter are not significantly different (P < .05).

Month

July

Leaf

Aug.

Leaf

Sept.

Leaf

Oct.

Leaf

Dec.

Leaf

(ppni diy

wt.)

213 ± 14a

633 ± 67a

122 ± 6a

21±4a

185 ± 10a

1097 ± 58b

1744 ± 101b

10±lb

211 ± la

1184 ± 75b

3861 ± 117c

10 ± lb

329 ± 4b

963 ± 73b

4990 ± 48d

ll±Ob

295 ± 13h

1038 ± 63b

21.30 ± 65b

8±0b

= 26.40; d.f. = 4,14; P < .001). Percent sulfur
decreased between July and December (F =
13.41; d.f. = 2,10; P = .03; Table 4).

Zinc (Zn) concentration in leaf tissue was
significandy higher (F = 36.56; d.f = 4,14; P
< .001) in October and December than in all
other months (Table 5). Mean Fe content was
higher in August leaf tissue than in July (F =
12.59; d.f = 4,14; P = .001), after which Fe
content remained fairly constant throughout
the remainder of the sample period. Leaf tissue
Mn content increased through the season and
was highest in October (F = 587.38; d.f =
4,14; P < .001; Table 5). Dudkin et al. (1976)
found that P. pectinatus, growing in polluted
coastal waters of the Black Sea, accumulated
Mn to 0.5% (dry weight). This Mn concentra-
tion corresponds to values found in this study.
Yet, Mn concentrations in water and sediment
from the lower Frovo River appear normal.
Copper (Cu) in leaf tissue varied significantly
(F = 44.48; d.f = 4,14; P < .001), with high
concentrations in July followed by lows in
August through December (Table 5).

Root tissues (root, rhizomes, and turions) of
P. pectinatus were not separated for analysis.
Mean root tissue forage qualities, compared to
leaf tissues, were lower in percent protein but
higher in fat content (Table 6). Phosphorus
was the only mineral element with a concen-
tration higher in root tissues than in leaf tissues.
Mineral (N, P K, Ca, and Mg) contents of root
tissues in this study were similar to contents
found in other studies (Kollman and Wali
1976, Van Vierssen 1982). High trace metal
concentrations were also found in root tissues.
Like leaf tissues, mean Fe and Mn concentra-
tions in root tissues appear inordinately high.

Conclusions and Future Research

Seasonal variation did exist in forage quali-
ties and nutrient concentrations in P. pectina-
tus. Protein content in leaf tissue was highest

in the summer months when P. pectinatus was
growing rapidly. By fall and early winter, protein
content decreased but was still higher than
concentrations found in drupelets. Apparently,
protein content in P. pectinatus leaf tissue
from the lower Provo River was higher than
concentrations reported elsewhere. High pro-
tein content in leaves and stems in the sum-
mer months would greatly benefit nesting and
brooding waterfowl that feed on this aquatic
species. Drupelet fat and sugar content was
higher than that for leaf or root tissues in
October. This would tend to confirm why dru-
pelets are so eagerly sought after by staging
and migrating waterfowl. Trace metal (Fe and
Mn) contents in leaf and root tissues accumu-
lated over the season and were very high by
fall. However, water and sediment concentra-
tions appear normal. It should be determined
whether the trace metal concentrations
observed are of natural or anthropogenic ori-
gin. Future research should develop a greater
understanding of heavy metal accumulation in
this and other key Great Basin aquatic plant
species.

Table 6. Forage quality, mineral and trace element con-
centration of P. pectinatus root tissue (root, rhizome, and
turions) averaged over five months. Forage quality con-
stituents and mineral content expressed as % and ppm dry
weight! S.E., n > 3.

Protein ADF Fat TNC^

(% dn' wt.)

13.0 ±1.0 ndl' ' 10.9 + 3.0 11.9 ±1.3

N P K Ca Mg
(% dry wt.)

2.1 ±0.2 0.4 ±0.1 2.9 ±0.2 1.5 ±0.2 0.3 ±0.1

Zn Fe Mn Cu
(ppm di-y wt.)

167 ± 25 7166 ± 1438 ' 2051 ± 570 14.8 ± 3.3

^Total nonstructural carbohydrate (TNC), a measure of sugars
"Not detemiined

168

Cheat Basin Natur.\list

[Volume 55

Acknowledgments

Funding and materials foi- this stud>' were
provided by the Department of Botany and
Range Science at Brigham Young University
and the Utah Chapter of the Wildlife Society.

Literature Cited

Anderson, B. W, and R. D. Ohmart. 1988. Stmcture of
the winter duck commiinit\' on the Lower Colorado
River: patterns and processes. Waterfowl in winter
University- of Minnesota Press, Minneapolis. 624 pp.

Anderson, D. R., R. C. Herron, and B. Reiswk;. 1986.
Estimates of annual survival rates of tnmipeter swans
handed 1949-82 near Red Rock Lakes National Wild-
life Reflige, Montana. Journal of Wildlife Management
50:218-221.

DuDKiN, M. S., I. V. Areshidze, and G. D. Lukina. 1976.
Chemical composition of seaweed in the coastal
waters of the Black Sea Ukrainian-SSR USSR.
Rastitel' nye Resursy 12: 133-137.

Henson, R, and J. A. Cooper. 1993. Trumpeter swan
incuhation in areas of differing food quality. Journal
of Wildlife Management 57: 709-716.

Ho, Y. B. 1979. Inorganic mineral nutrient level studies in
Potamogeton pectbuitus L. and EnteromoiyJw prolifera
in Forfar Loch, Scodand. Hydrohiologia 62; 7-15.

Hutchinson, G. E. 1975. A treatise on limnolog>'. Volume
III: Linmological botany. John Wiley & Sons, New
York, NY, 660 pp.

JUPP, B. P, and D. H. Spencer. 1977. Limitations on macro-
phytes in a eutrophic lake. Loch Leven. I. Effects of
phytoplankton. Journal of Ecology 65: 175-186.

Kadlec, J. A., and L. M. Smith. 1989. The Great Basin
marshes. In: L. M. Smith, R. L. Pedersen, and R. M.
Kaminski, editors. Habitat management for migrating
and wintering waterfowl in North America. Te.xas
Tech University Press, Lubbock. 651 pp.

Kantrud, H. a. 1990. Sago pondweed (Potamogeton
pectinatus L.): a literature review. U.S. Fish Wildlife
Service Resource Publication 176. 89 pp.

Kollman, a. L., and M. K. Wall 1976. Intraseasonal
variations in environmental and productixity rela-

tions of Potamagcton pectinatus communities. Archiv
fuer Hydrobiologie, Supplementband 50: 439-472.

KoRSCHGEN, C. E., L. S. Georce, .\nd W. L. Green. 1988.
Feeding ecology of canvasbacks staging on Pool 7 of
the Upper Mississippi River Waterfowl in winter.
University of Minnesota Press, Minneapolis. 624 pp.

Linn, J. G., E. j. Staba, R. D. Goorich, J. C. Meiske, and
D. E. Otterby. 1975. Nutritive value of dried or
ensiled aquatic plants. I. Chemical composition.
Journal of Animal Science 41: 601-609.

Madsen, J. D. 1986. The production and physiological
ecolog}' of the submerged acjuatic macroph\ te com-
munity in Badfish Creek, Wisconsin. Unpublished
doctoral dissertation. University of Wisconsin,
Madison. 449 pp.

Orson, R. A., R. L. Simpson, and R. E. Good. 1992. A
mechanism for the accumulation and retention of
heavy metals in tidal freshwater marshes of the
LJpper Delaware River Estuaw. Estuarine, Coastal,
and Shelf Science 34: 171-186.'

Sherwood, G. A. 1960. The whistling swan in the West
with particular reference to Great Salt Lake Valley,
Utah. Condor 62: 370-377.

Spencer, D. W, and P. G. Brewer. 1971. Vertical advec-
tion, diffusion and redox potentials as controls on the
distribution of manganese and other trace metals
dissolved in water of the Black Sea. Journal of Geo-
physical Research 76: 5877-5892.

Tisdale, S. L., W. L. Nelson, and J. D. Beaton. 1985.
Soil fertility and fertilizers. 4th edition. Macmillan
Publishing, New York, NY 753 pp.

Van Vierssen, W. 1982. The ecology of communities
dominated by ZannicJieUia ta.xa in western Europe.
III. Chemical ecology. Aquatic Botany 14: 259-294.

Vermaak, J. E, J. H. Swanepoel, and H. J. Schoonbee.
1983. The phosphorus cycle in Germiston Lake with
special reference to the role of Potamogeton pectinatus
L. Pages 317-321 in ProceecUngs of the international
symposium on aquatic macrophytes, Nijmegan,
Netherlands.

Received 31 May 1994
Accepted 3 Jamianj 1995

FACTORS INFLUENCING FISH ASSEMBLAGES OF A HIGH-ELEVATION
DESERT STREAM SYSTEM IN WYOMING

Bernard Carter^ and Wayne A. Hubert^

Abstract. — Seven fish species were found in the Bitter Creek drainage of southwest Wyoming, but only speckled dace
(RJiinichthijs oscuhis), flannelmouth sucker [Catostomus latipinnis), and mountain sucker (Catostomtis phityrhynchus)
were indigenous. No relationships were foimd between fish standing stocks and habitat features, but species richness was
related to elevation and stream width. No fish were found above an elevation of 2192 m. Only the most downstream study
reach had more than three species present. Two indigenous species, speckled dace and moimtain sucker, and a nonnative
species, fathead minnow (Pimephales promelas), were predominant fishes in the drainage. These three species withstand
intermittent stream flows that are common in the drainage.

Key words: fish, streams, desert, Wyoming, habitat, distribution.

Fish communities in streams become more
complex as habitat diversity increases along
the length of a stream. Variation in fish com-
munity' stnicture within a stream system can fol-
low patterns of zonation or addition. Specific
fish communities can be associated with zones
defined by water temperature or geomoi-pho-
logic features, or community complexity can
increase with progression downstream by
addition of species (Moyle and Nichols 1973,
Guillory 1982, McNeely 1986, Hughes and
Gammon 1987, Platania 1991, Rahel and
Hubert 1991). However, such patterns may
differ in arid drainages of the western United
States with depauperate ichthyofauna (Cross
1985).

Little is known about the fish communities
in high-desert stream systems in southwestern
Wyoming. Annual precipitation over most of
these drainages is < 16 cm, with much of it as
snow in headwater areas during late winter and
thunderstorms during late summer Discharge
is highest during spring runoff, and streams
frequently become intermittent during sum-
mer and winter. Because these systems in
Wyoming are at high elevations (>1800 m
above mean sea level), water temperatures are
cool compared with other desert streams. The
climate in these areas typically consists of diy,
moderately warm summers with long, cold
winters.

The puipose of this study was to (1) describe
fish species present in a high-desert stream

system in southwestern Wyoming and (2)
determine the factors that influence fish abun-
dance and community structure within the
drainage.

Study Area

The study was conducted in an intermittent
drainage. Bitter Creek, a tributary to the Green
River in the Red Desert of southwest Wyo-
ming (Fig. 1). The study area consists of Bitter
Creek and four tributaries — Little Bitter, Salt
Wells, Bean Springs, and Gap creeks. Frequent-
ly, no measurable surface flow occurs in Bitter
Creek at Bitter Creek, WY, during midsum-
mer and midwinter (flow data available in the
Water Resources Data System at the Wyoming
Water Resources Center, University of Wyo-
ming, Laramie). Bitter Creek at Salt Wells,
WY, generally has no measurable surface flow
from July to February. Salt Wells Creek has
more persistent flows near its mouth, but
records of no measurable flows occur in mid-
summer and midwinter When no measurable
flow occurs in these streams, isolated pools of
standing water can be found in the stream
channels. Elevation of the study area ranges
from 1800 to 2400 m.

Streams in the Bitter Creek drainage typi-
cally are downcut by at least 1.5 m, with steep
clay banks having no vegetation. Riparian vege-
tation consists of grasses and sagebrush {Artem-
isia spp.); upland vegetation is primarily the
latter

^Wyoming Cooperative Fish and Wildlife Research Unit, University of Wyoming, Laramie. \V\' 82071-.3166. The unit is jointK- supported by the University
of Wyoming, Wyoming Game and Fish Department, and National Biological Survey.

169

170

Great Basin Naturalist

[Volume 55

Salt Wells,
Wyoming

Rock Springs,
Wyoming

?>^'

iixef

G<'

,e^

Bitter Creek,
Wyoming

WY0MIN6

Research location

0^-

0^'

>#>

5> ) 95^1

20 km

Fig. 1. Map of the Bitter Creek drainage, \\T, showing the location of the 16 study readies.

Baxter and Simon (1970) reported four fi.sh
species in collections at two sites in Bitter
Creek. Speckled dace {Rhinichthijs osculus),
fathead minnow {Pimephales promelas), and
mountain sucker {Catostomus platijrhtjnchus)
were reported from a site al:)out 10 km upstream
from the mouth. Bluehead sucker {Catostomus
discobolus) was the only species reported from
a site near Salt Wells.

Methods

Sixteen 100-m-long study reaches were
selected to represent variation in stream size
and habitat in the drainage during summer
1993. Wetted width, mean depth, and sub-
strate were determined across transects at 10-
m intervals. Dominant substrate at each tran-
sect was visually determined following Bain et
al. (1985): sand-silt (<2 mm diameter), gravel
(2-16 mm), pebble (17-64 mm), cobble
(65-256 mm), and boulder (>256 mm). Water

velocity was determined within each reach
using the dye flow mediod (Binns 1982). Stream
discharge at time of sampling was computed
from width, depth, and velocity.

Alkalinity, hardness, and pH were mea-
sured at the time of sampling. Alkalinity and
hardness were determined with field test kits
(Hach Model A1-36DT), pH with an electron-
ic meter Mean elevation and channel slope at
each study reach were estimated from 7.5-
minute topographic maps.

Fish were sampled in each 100-m reach by
electrofishing. Small-mesh (6.4-mm) block nets
were placed at each end, and two or three elec-
trofishing passes were made over the entire
reach. Three-pass depletion estimates of species
abundance were made in most reaches. Two-
pass depletion estimates were used when
>80% offish captured by the first two passes
were captured during the first pass. Fish
abundance was computed using the Zippin
method (Platts et al. 1983). All fish were

1995]

Desert Fishes

171

weighed to enable computation of standing
stock estimates.

Standing stocks of individual species, total
standing stock of all species, and number of
species in a reach were evaluated for their
relation to nine habitat variables using simple-
linear and multiple-regression analyses.
Independent variables were included in
regression models if they were significant at
P < .05. We further limited inclusion of de-
pendent variables in multiple-regression mod-
els to ones that were not correlated at P < .05.
Computations were performed using Statistix
4.0 (Analytical Software 1992).

Results

Seven fish species were collected: speckled
dace, fathead minnow, Utah chub {Gila atrarici),
Bonneville redside shiner {Richardsoniiis
balteatus hydrophlox), mountain sucker, white
sucker {Cotostomus commersoni), and flannel-
mouth sucker (C. latipinnis). Abundance varied
substantially among study reaches (Table 1).
Mean total standing stock of all species was
3.0 g/m'^ and ranged from 0 to 21.3 g/ni'^. No
fish were found in the four reaches above
2192 m.

Habitat features varied among the 16 study
reaches (Table 2). Flow was measurable at all
reaches. Stream width, water velocity, and dis-
charge increased downstream. Sand-silt sub-
strate occurred over >90% of almost all study
reaches. Alkalinity, pH, and hardness also
increased downstream.

No significant relations were found
between any of the nine habitat variables and
standing stocks of individual species or total
standing stock of all species. However, there
were significant relations between the number
of species and four habitat variables:

NS = 20.88 - 0.0091 E (F = .0003, R^ = .61),
NS = 0.13 + 0.812 W (F = .0010, R2 = .52),
NS = 3.40 - 11.008 V (F = .029, fi2 = .33), and
NS = 0.57 + 31.245 F (F = .022. R^ = .32),

where NS = number of species, E = eleva-
tion in meters, W = mean wetted width in
meters, V = water velocity in meters per sec-
ond, and F = flow in cubic meters per second.
The best two-variable model was

NS = 14.36 - 0.0065 E + 0.53 VV (F < .0001, R'- = .80).

As study reaches declined in elevation and as
width, water velocity, and discharge increased,
the number of species increased.

Because the most downstream reach on
Bitter Creek had twice as many species as any
other reach and flow at the reach was en-
hanced by discharge from a sewage treatment
plant, we assessed relations with the omission
of that reach. Again, no relationships were
found between any of the habitat variables and
standing stocks of fish, but the number of
species (NS) was significantly related to eleva-
tion (E) and water velocit)' (V):

NS = 15.95 - 0.0068 E (F = .0014, R- = .55), and
NS = 3.00-10.11 V (F = .0018, R^ = .51).

Among the 15 study reaches with a maximum
of three species present, the number of species
increased with decline in elevation and water
velocit}'.

Discussion

Of the seven fish species in the Bitter Creek
drainage, only three — speckled dace, flannel-
mouth sucker, and mountain sucker — are
indigenous (Baxter and Simon 1970). Absence
of fish above 2192 m is probably due to a cli-
mate that is too cold for warmwater fishes.
Additionally, no trout occur naturally or have
become naturalized in the watershed.

The number of species increased with pro-
gression from headwater to downstream
reaches (Table 1). With the exception of the
most downstream reach on Bitter Creek, no
more than three species — specked dace,
mountain sucker, and fathead minnow — were
found in any of the study reaches. The high-
elevation reaches with fish tended to have
predominantly or exclusively speckled dace
and mountain sucker.

Much of the increase in species richness
with downstream progression was due to the
most downstream reach on Bitter Creek
where six species were found (Table 1). Four
of six species were not natives — fathead min-
now, white sucker, Utah chub, and Bonneville
redside shiner Mountain sucker was not found
in this reach, but it was common throughout
most of the Bitter Creek drainage. While this
study reach was lowest in elevation among the
16 study reaches, it also was downstream from
the outfall of the wastewater treatment facility

172

Great Basin Naturalist

[Volume 55

c£>

»o

<M cc in

d Lo d

V V '^ '^ V V

d c^i

—I o
V

"^ ^ 'i 2 -o ^ ^

"^ ■= " £ = 3 ■:= •_

c/2 «>. fc u, ;> D oa

O ,^

-1^

■^

;>5

CO o) — O'

— ^ d d d

— : p — ; q

— < o o

-— I 00 00 in t~-
d ^ CO d 00

05 t^ Ol f>]

-H oj CO r-^

p in — ; q
oi t>i oq o6

oq oq oi

q 00 t-; CO

oi oi ^ d>

^ ys O^
oq oq — <

a 1-

5 u

X 1000)
inity

g/L X 10
Iness

o
X

^
~

■|

0; !Ll

j;"!

s §

S3 E^

,-*

""'^

-^^ — rt

cii

^ X

1995]

Desert Fishes

173

for Rock Springs, WY, and was only 13 km
upstream from the confluence of Bitter Creek
and the Green River. The more permanent
flows due to the wastewater treatment facihty
may have enabled fish not adapted to inter-
mittent flows to persist in this reach. Also, the
relatively short distance to the Green River
may enable upstream migration of fish to this
reach, contributing to higher species diversity.
Repeated invasion of nonnative species from
downstream reservoirs maintains species
diversit)' in the Virgin River, UT (Cross 1985).
Also, human disturbances have been found to
create environmental conditions favorable to
nonnative fish in California (Moyle and Nichols
1973, 1974). Therefore, enhanced flows due to
the wastewater treatment facility and invasion
of nonnative species from the Green River
probably contribute to the diversity of fish in
the downstream portion of Bitter Creek.

During summer 1993, flowing water
occurred at all study reaches when they were
sampled. Precipitation in spring and summer
1993 was substantially greater than normal,
enabling measurable surface flows to persist
during summer. However, study reaches up-
stream from the outfall of the Rock Springs,
WY, wastewater treatment facility are frequent-
ly intermittent during late summer. Fathead
minnow has been described previously as a
species associated with intermittent streams
(Baxter and Simon 1970, Pflieger 1975). Our
observations indicate that two indigenous
species — speckled dace and mountain suck-
er— and one introduced species — fathead
minnow — can survive in the frequently inter-
mittent streams. Consequently, these three
fishes are the only species occuning over most
of the Bitter Creek drainage, but fathead min-
nows tend to be limited to lower elevations
than the two native species. It is not known
how the invasion by fathead minnow may
affect the native speckled dace and mountain
sucker in this desert stream system.

Acknowledgments

We thank M. Fowden, K. Johnson, W.
Wengert, and R. Wiley, Wyoming Game and

Fish Department, for their assistance, and H.
Li, T. Patton, F Rahel, R. Wiley, and an anony-
mous referee for review of the manuscript.
The project was supported by the Wyoming
Game and Fish Department.

Literature Cited

Analytical Software. 1992. Statistix Version 4.0 user's
manual. Analytical Software, St. Paul, MN.

Baln, M. B., J. T. Finn, and H. E. Booke. 198.5.
Quantifying stieam substrate for habitat analysis stud-
ies. North American Journal of Fisheries Manage-
ment .5: 499-500.

Baxter, G. T, and J. R. Simon. 1970. Wyoming fishes.
Bulletin 4. Wyoming Game and Fish Department,
Cheyenne.

Binns, N. a. 1982. Habitat quality inde.x procedures manu-
al. Wyoming Game and Fish Department, Cheyenne.

Cross, J. N. 1985. Distribution offish in the Virgin River,
a tributary of the lower Colorado River Environ-
mental Biology of Fishes 12: 13-21.

GuiLLORV, V. 1982. Longitudinal gradients of fishes in
Thompson Creek, Louisiana. Southwestern Naturalist
27: 107-115.

Hughes, R. M., and J. R. Gammon. 1987. Longitudinal
changes in fish assemblages and water qualit>' in the
Willamette River, Oregon. Transactions of the
American Fisheries Society 116: 196-209.

McNeeley, D. L. 1986. Longitudinal patterns in fish
assemblages of an Ozark stream. Southwestern
Naturalist 31: 375-380.

Moyle, R B., and R. D. Nichols. 1973. Ecology of some
native and introduced fishes of the Sierra Nevada
foothills in central California. Copeia 1973: 478-490.

. 1974. Decline of native fish fauna of the Sierra

Nevada foothills, central California. American Mid-
land Naturalist 92: 72-83.

Pflieger, W. L. 1975. The fishes of Missouri. Missouri
Department of Conser\'ation, Columbia.

Platania, S. R 1991. Fishes of the Rio Chama and upper
Flio Grande, New Mexico, widi preliminan' comments
on their longitudinal distribution. Southwestern
Naturalist 36: 186-193.

Platts, W. W., W. F Megahan, and G. W Minshall.
1983. Methods for evaluating stream, riparian, and
biotic conditions. LISDA Forest Service General
Technical Report INT-138. 70 pp.

R\hel, F J., AND W. A. Hubert 1991. Fish assemblages
and habitat gradients in a Rocky Mountain-Great
Plains stream: biotic zonation and additixe patterns
of community change. Transactions of the American
Fisheries Society 120: 319-332.

Received 21 April 1994
Accepted 3 October 1994

SPECIATION BY ANEUPLOIDY AND POLYPLOIDY IN MIMULUS
(SCROPHULARIACEAE)^

Robert K. Vickcn', Jr.2

Key wards: Miimiliis. speciation. cvoluthni, aiwuploidy, polyploidy.

Speciation by aneuploid and polyploid
changes in chromosome numbers is so common
in flowering plants as to be almost a character-
istic of the angiosperms. Elegant examples of
these patterns of evolution are exhibited by
monkey flowers of the genus Mimiihis (Scro-
phulariaceae).

The genus Mimulus contains some 150
species occurring in western North and South
America with a few outlying species in eastern
North America, Japan, Vietnam, the Himalayas,
New Zealand, Australia, and South Africa. The
center of diversity is California, with a second-
ary center in Chile. Some species are annuals of
deserts, grasslands, or forests; some are biennials
of marshy places; some are herbaceous peren-
nials from springs, streamsides, or lake-shore
habitats; and others are woody shrubs of the
dry California chapanal. The species fonii clus-
ters reflecting these various life forms. There
are 8-10 such clusters commonly recognized
as sections of the genus Mimuhis (Crant 1924,
Fennell 1951, Chuang and Heckard personal
communication).

Chromosome numbers of over 50 species
(Table 1), that is, approximately one-third of
the Mimulus species, have been ascertained
by Vickery and his co-workers (Vickeiy 1978,
Vickery, Chu et al. 1981, Vicken, Simpson et al.
1981, Vickery et al. 1982, 1985,' 1986, 1990, un-
published) and by Chuang and Heckard (per-
sonal communication). Chromosome numbers
reveal intriguing patterns of evolution by aneu-
ploidy and polyploidy.

First, let us consider the base chromosome
numbers of the eight main sections of the
genus. Section Munulastrutn has a ])ase num-
ber of X = 7; Eunanus and Erytliranthe have
base numbers of x = 8; Paradanthus 8, 9, 10;
Eumimulus 8, 11, 12; Oenoe 9; DipJacus 10;

and Simiolus 14, 15, 16, 30. Base numbers of
the sections suggest extensive evolution by
both aneuploidy and poKploidv'. For the genus
as a whole, the base number appears to be x =
8, inasmuch as the other plausible base num-
ber, X = 7, is found only in one, apparently de-
rived, desert species, M. moJuiven.sis Lemmon
(Table 1).

Next, let us consider the chromosome num-
bers by individual species. All species counted
thus far are the same in each of several sections,
specifically, in Mitnulastrum, Erythranthe,
Oenoe, and Diplacus. The other sections are
polymorphic for their species' chromosome
numbers and frequently exhibit speciation by
aneuploidy and/or polyploidy, often in com-
plex combinations. For example, the various
species of section Eumimulus exhibit n = 8,
11, and 12; species of section Eunanus exhibit
n = 8, 10, and 16; species of section Paradan-
thus exhibit n = 8, 9, 16, 17, 18, and 30; and
species of section Simiolus exhibit n = 13, 14,
15, 16, 24, 28, 30, 31, 32, 46, and 48 (Table 1).

Section Simiolus, which shows by far the
most speciation by aneuploidy and/or poly-
ploidy of all sections of the genus, consists of
six species groups, that is, complexes of related
species and varieties. First is the M. guttatus
complex, centered in California; it has as its
base number x = 14, with aneuploid forms at
n = 13 and n = 15 (Table 1), as well as tetra-
ploid forms with n = 28. Second is the alpine
(western United States) M. tdingii complex
with its base number of x = 14 and aneuploid
forms at n = 15, /] = 16, and an unusual pol> -
ploid form at n = 24. The third species group
is the M. dentilobus complex of southwestern
United States and northwestern Mexico with
its base number of x = 16 and an aneuploid
form at n = 15. Fourth is the M. luteus complex

'a talk presented 4 September 1993 as part of the symposium, "Plant Evolution.
^Biology Department, University of Utali, Salt Lake City, UT 84112 USA.

lit tlie National Institute ol't^enetics, Mishima. Japan.

174

1995]

Notes

175

T\BLE 1. Chromosome mmibers in tlie genus Miimihis li\'
sections (counts by Chuang and Heckard and 1)\' Vicken
and co-workers; see text for references).

Taxon n =

Mimiilastrum Gray (.v = 7)

M. mohavensis Lemnion 7

Eumiimilus Gray {x = 8, 11, 12)

M. alatus Aiton 11

M. gracilis R. Br 8

M. ringens L. 8, 12

Eunanus Gray (.r = 8)

M. bolanderi Gray 8

M. layneae (Greene) Jepson 8

M. brevipes Bentham 8

M. cusickii (Greene) Piper 8

M. nanus Hook. & Arn. 8

M. torreyi Gray 10

M. biglorii Gra\ 16

Paradantlms Grant (.v = 8, 9, 10)

M. bicolor Haitweg ex Bentham 8

M. filicaulis Watson 8

M. breweri (Greene) Coville 16

M. floribundus Douglas 16

M. moschattis Douglas 16

M. laiidens (Gray) Greene 16

M. arenarius Grant 16
M. primidoides Rydb. 9, 17, 18

M. repens R. Br 10

M. nepalensis Bentham 16, .30

Enjtbrandw Greene {x = 8)

M. cardinally Douglas 8

M. eastwoodiae Rydb. 8

M. lewisii Pursh 8

M. nelsonii Grant 8

M. ntpestrh Greene 8

M. verbenaceits Greene 8

Oenoe Gray (x = 9)

M. picttis (Cunan) Gray

M. tricolor Lindl.

M. pygmaeus Grant

9 (or 10?)

M. pilosellus Greene

Diplacus Gray (.t = 10)

M. aridiis (Abrams) Grant

M. aurantiacus Curt.

M. calycinus Eastw.

M. clevelandii Brandg.

M. fasiculatus (Pennell) McMinn

M. longiflorus (Nutt.) Grant

M. puniceus (Nutt.) Steud.

Simiohts Greene (.t = 14, 1.5, 16)

M. gtttatus Fischer e.\ DC.

14, 1.5, 28

M. laciniatus Gra\'

M. nasutus Greene

13, 14

M. glaucescens Greene

M. platycalyx Pennell

M. tilingii Regel

14, 15, 24, 28

M. gernnipanis Weber

M. dentilobus Rols. & Fern.

15, 16

M. wiensii Vicker>'

M. glabratus HBK

1.5, .30, 31

Af. andicolus HBK

M. pilosiuscidus HBK

M. extemiis (Skotts.) Skotts

M. luteus L.

30, 31, .32

A/, cupreus Dombrain

Undescribed

n. sp #A

n. sp #B

n. sp #C

.32, 48 ± 1-4

from the central and southern Andes of South
America. Its base number is x = 30, but there
are n = 31 and n = 32 forms as well. Fifth,
there is the M. glabratus complex that ranges
from Canada to Patagonia. Its varieties in cen-
tral North America exhibit the base number of
the complex, x = 15. In the Rio Grande
drainage we find tetraploids with n = 30.
From northern Mexico to southern Colombia
we find the aneuploid tetraploid n = 31 vari-
eties of the complex. From Ecuador south to
southern Argentina and including the Juan
Fernandez Islands off the coast of Chile, we
find the aneuploid hexaploid species and vari-
eties with n = 46 chromosomes. Apparently,
each change in chromosome number facilitat-
ed an adaptive radiation further south. Last is
the M. wiensii complex of the mountains of
western Mexico with its base number of .t =
16 and three apparent new species that are
morphologically distinct and reproductively
isolated (Vickeiy et al. unpublished). One has
n = 16 chromosomes, one has n = 32 chromo-
somes, and the third has two forms — one with
n = 32 chromosomes and the other with n = 48
± 1—4 chromosomes (incipient aneuploidy?).

How does speciation by aneuploidy and
polyploidy occur? We carefully examined meiosis
in M. glabratus var. utahensis and M. glabra-
tus var. fremontii, two of the widespread
diploid varieties of the M. glabratus complex,
and their intervarietal F^ hybrids. First, of
1317 cells examined in diakinesis or meta-
phase of first meiosis (Tai and Vickery 1970,
1972), 1090 exhibited regular 15 bivalent
chromosomes. Another 23 cells, or 1.7%, had
aneuploid numbers of chromosome pairs rang-
ing from only 6 to as many as 13, plus 4-18
univalents. These cells presumably could pro-
duce aneuploid gametes, at least in some cases.
A sizeable minority, 204 cells, exhibited 14 II
and 2 I, or 13 II and 1 IV, or complement frac-
tionation with its uneven groupings of chro-
mosomes. These cells might produce aneu-
ploid gametes also. Second, of 782 additional
cells observed in Anaphase I, 294 (37.5%)
exliibited unequal disjunction, laggard chromo-
somes, or chromatin bridges. These cells also
could result in aneuploid gametes. Some 47 of
these abnormalities occurred in M. glabratus
wax. fremontii, only 18 occurred in M. glabratus
var utahensis, but most, 229, occuiTcd in die in-
tervarietal hybrids. Thus, varieties differ in their
potential for producing aneuploid gametes.

176

Cheat Basin Naturalist

[Volume 55

and intcrvarictal hybrids are particular!)'
prone to do so. This suggests to me that natur-
al hybridization probably plays a significant
role in evolution in monkey flowers. Finding
occasional plants in various populations with
aneuploid chromosome numbers indicates
that aneuploid gametes not only are produced,
but actually function. Third, of 95 cells exam-
ined in Anaphase II, 22 were polyploid and
could presumably lead to polyploid gametes.
Thus, we see significant numbers of the veiy
cytological abnormalities in the basic diploid
varieties that could lead to evolution by aneu-
ploidy and polyploidy, that is, to the veiy pat-
terns of evolution that we actually see in the
M. glabratus complex.

Literature Cited

Grant, A. L. 1924. A monograph of the genus Miinulus.
Annals of the Missouri Botanical Garden 11; 99-389.

Pennell, F. W. 19.51. Mimuhis. Pages 688-731 in L.
Abrams, Illustrated flora of the Pacific States.
Volume 3. Stanford University Press, Stanford, CA.

Tai, W., and R. K. Vickery, Jr. 1970. Cytogenetic rela-
tionships of key diploid members of the Mimuhis
glahratus complex (Scrophulariaceae). Evolution 24:
670-679.

. 1972. Unusual cytological patterns in microsporo-

genesis and pollen development of evolutionaiy sig-

nificance in the Mimulu.s glabratu.s complex (Scrophu-
lariaceae). American Journal of Botany .59; 488-493.

Vickery, R. K., Jr. 1978. Case studies in the evolution of
species complexes in Mimuhis. Evolutionarv' Biolog\
11; 404-506.

Vickery, R. K., Jr., Y. E. Chu, K. Fine.man, and S.
Pt'RC:ELL. 1981. Chromosome number reports on the
Scrophulariaceae in lOPB Chromosome number
reports LXX presented by Askell Love. Taxon 30: 68.

Vickery, R. K., Jr., M. Simpson, and M. Nellestein.

198 1. Chromosome number reports on the Scrophu-
lariaceae in lOPB Chromosome number reports
LXX presented by Askel Love. Taxon 30: 68-69.

Vickery, R. K., Jr., S. A. Werner and E. D. MacArthur.

1982. Chromosome number reports on the Scrophu-
lariaceae in lOPB Chromosome number reports
LXXV presented by Askell Love. Taxon 31; 360.

Vickery, R. K., Jr., S. A. Werner , D. R. Phillips, and
S. R. Pack. 1985. Chromosome counts in section
Siiniohis of the genus Mimuhis. X. The M. gkibnifiis
complex. Madrono 32: 91-94.

Vickery, R. K., Jr., B. Y. Kang, T K. Mac, S. R. Pack, and
D. A. Phillips. 1986. Chromosome counts in Mimuhis
sect. Enjthranthc (Scrophulariaceae). III. Madrono
.33; 264-270.

Vickery, R. K., Jr., E R^^hmen, S. R. Pack, and T. Mac.
1990. Chromosome coimts in section Simiohis of the
genus Mimuhis (Scrophulariareae). XI. M. ghibratus
complex (cont.). Madrono 37; 141-144.

Received 6 J uhj 1994
Accepted 24 September 1994

SPECIATION IN MIMULUS, OR, CAN A SIMPLE FLOWER COLOR
MUTANT LEAD TO SPECIES DIVERGENCE?^

Robert K. Vickeiy Jr.-

Key words: Mimulus, Eiythranthe, speciation, reproductive isolation, flower color mutations, pollinators, bumble-
bees, hummingbirds.

The general pattern of speciation in nature
has been clear for a long time — the diver-
gence of portions of a population, usually small
(Levin 1993), usually in geographic isolation
(Mayr 1976), and the accumulation of genetic
changes by selection and/or genetic drift (Crow
and Kimura 1970) that produce reproductive
isolation and normally character divergence as
well. The critical step is reproductive isolation,
and yet that step — except for polyploid forma-
tion which in itself is not always effective
(DeWet 1980) — has rarely been observed
actually happening in nature. A promising
group in which to study speciation events in
progress is section Erythranthe of the genus
Mimidus (Vickeiy 1978).

The six species of monkey flowers compris-
ing section Erijthranthe are moisture-requiring,
herbaceous perennials 1-10 dm in height, with
variously shaped, opposite leaves and bilabiate
flowers that have four stamens, one style with
a bilobed sensitive stigma, and five corolla
lobes that range in color from orange to red —
rarely yellow — and from lavender-pink to
magenta-pink — rarely white. See Grant (1924)
for further details. When considered species
by species, corollas of M. cardinalis Douglas
vary from orange to red — rarely yellow — and
are sharply and fully reflexed, hummingbird-
pollinated flowers. Corollas of M. verbenaceus
Greene are partially reflexed; that is, the
upper two corolla lobes are reflexed, whereas
the lower three are gently recurved. Flowers
are orange -red to red — rarely yellow — and
also are hummingbird-pollinated. Corollas of
M. nelsonii Grant are partially re-flexed also
and have orange-red to red flowers, which are
longer than those of M. verbenaceus (6-7 cm

versus 4-5 cm). Corollas of M. eastwoodiae
Rydberg and M. rupestris Greene, the two
cliff-dwelling species, are partially reflexed,
red, and typically hummingbird-pollinated
also. And last, flowers of the Rocky Mountain
variety of M. lewisii Fursh are magenta-pink
with all five corolla lobes gently recurved
rather than reflexed, thus forming a bee-land-
ing platform; flowers of the Sierra Nevada
variet\' of Al lewisii are lavender-pink — rarely
white — with corolla lobes thrust foi"ward. Both
varieties of M. lewisii are bee-pollinated.
Mimulus lewisii flowers and those of M. east-
woodiae and M. rupestris produce only modest
amounts of nectar, whereas the other species
produce abundant nectar (Table 1). Thus, the
species differ markedly in flower shape, flower
color, nectar production, and, consequently, in
pollinators sei'vicing the flowers. In the forma-
tion of the six species, evolution appears to
have responded to selection imposed by polli-
nator preferences and ecological opportunities.
The result is that members of the complex
have radiated into a wide variety of different
habitats and niches.

A bright yellow-flowered mutant has
appeared on the scene in this setting of polli-
nator-driven, ecologically opportunistic evolu-
tion. In two populations of M. cardinalis bright
yellow-flowered morphs have become well
established. One population is in the Siskiyou
Mountains of Oregon, which is the northern
limit of the range of M. cardinalis (Grant 1924).
The other population is on Cedros Island, Baja
California, and is at the southern limit of the
species range. As Mayr (1976) suggests, new
forms often evolve from isolated populations
such as these on the periphery of a species

'The opening talk in the s\niposiuni, "Mechanisms of Speciation in Higher Plants," given 1 September 1993 at the XV International Botanical Congress,
Yokohama, Japan.

^Biology Department, University of Utah, Salt Lake City, UT 84112 USA.

177

178

Great Basin Natur/VLIst

[Volume 55

Table 1. Nectar production in the species of section
Enjthranthe measured at OSOO h in the wild (Vickery- and
Sutherland 1994). Averages are based on 20 or more mea-
surements from a population representative of each
species or variet>'.

Species

\olinne in

Cf sut^ar

M. canlinalis

3.9

11.5

M. vcrbenaceus

6.7

5.8

M. nipestris

0.9

19.0

M. eastwoodiae

1.5

13.7

M. nelsonii

18.3

19.2

M. lewisii

Roclcv Mountains

0.5

Sierra Ne\ada

0.7

11.3

range adjacent to new ecological opportuni-
ties. A bright yellow-flowered morph of M.
verhenaceus has appeared also and become
well established in a population growing in an
isolated spring area, Vasey's Paradise, at the
bottom of the Grand Canyon of the Colorado
River, AZ, that species' northwesteiTi limit.

Flower colors in section Enjthranthe are
due to various combinations of six anthocyanin
pigments — three pelargonidins (apricot-pink)
and three cyanidins (lavender-pink) — and at
least one carotene pigment (Pollock et al. 1967).
The lavender to magenta flowers of M. lewisii
are due to various combinations of the pelar-
gonidin and cyanidin anthocyanin pigments
without die yellow carotene. Flowers of the red-
flowered species have all or most of the six
anthocyanin pigments plus the carotene pig-
ment. Red color results from a visual blend of
pink pigments and yellow pigment. Yellow-
flowered plants have a pair of recessive genes
at one locus that suppresses anthocyanin pro-
duction (pink pigments), leaving just the yel-
low carotene pigment showing. So, a sin^h'
mutation, when homozygous, changes flower
coh)rfrom red to yellow.

If the change from red to yellow flowers
leads to a change in pollinators, for example,
from hummingbiids to bumblebees or hawk-
moths, then the first major step in reproduc-
tive isolation has been established by a single
gene change (when homozygous)! Once repro-
ductive isolation has been established by color
differences, presumably selection would fine-
tune it, e.g., by favoring more tubular flowers
for hummingbird-pollinated flowers and by
favoring a landing platform morphology and
nectar guides for bee-pollinated flowers.

Are pollinators required for seed set in
Mimulus cardinalis or do the flowers self-polli-

nate? To test these hvo questions, I used the fact
that M. cardinalis flowers are borne in pairs. I
grew plants of red- and of yellow-flowered M.
cardinalis from Cedros Island in the green-
house of the Biology Department, University
of Utah. The greenhouse is free of pollinators.
I carefulK hand-pollinated one flower of each
of ten pairs of red flowers and of ten pairs of
\'ellow flowers. The hand-pollinated flowers of
both the red-flowered and \'ellow-flowered
plants set moderate numbers of seeds per cap-
sule (50-150), while the unpollinated flowers
set no seeds at all. This finding corroborates
my earlier observations on the Cedros Island
M. cardinalis (Vickery 1990) that flowers do not
self-pollinate and that pollinators are required
for seed set.

Are the rewards for pollinators the same in
yellow flowers as in red? That is, do yellow
flowers and red flowers produce equal volumes
of nectar with the same concentrations of sug-
ars? Red flowers of the Cedros Island M. car-
dinalis produced an average (based on flowers
from 30 greenhouse-grown plants) volume of
9.5 fx\ of nectar with 18.2% sugar Yellow flow-
ers produced an average (based on measure-
ments of flowers from 40 greenhouse-grown
plants) of 10.9 (jl\ of nectar with 23.0% sugar.
There is so much variation that these values
are not significantly different.

Finally the key c|uestion, do pollinators show
a preference for red or yellow flowers? To study
this question, I placed 24 red-flowered and 24
yellow-flowered plants in a random arrange-
ment in a meadow in the Red Butte Can\'on
Natural Area in the Wasatch Mountains be-
hind the University of Utah and observed pol-
linators that visited this experimental popula-
tion. Pollinators that came were humming-
birds and bumblebees, with rare visits from
flies, but no hawkmoths or honey bees.
Pollinators were observed for three 50-min
periods on each of 5 d. On 28 July 1987 there
were 55 hummingbird visits to the 39 red
flowers present and 20 visits to the 35 yellow
flowers. Chi-square = 14.379, p <' .001,
which indicates a significant preference for
red flowers. That da\' there were 10 bumble-
bee visits to red flowers and 12 to yellow flow-
ers. Chi-square = 0.1818, no significant pref-
erence. On 31 July there were 176 humming-
bird visits to the 42 red flowers in bloom that
day in the population and 40 \'isits to the 21
yellow flowers. Chi-square = 70.246, p < .001,

1995]

Notes

179

which indicates a significant preference for
red. That day there were six l:)umhlebee visits
to red and one to yellow. There were too few
bumblebee visits for a meaningful .v- value to
be calculated. The same pattern of three
observation periods was continued on 2-4
August, but once again there were too few pol-
linator visits to obtain meaningful .v^ values.
Apparently, most hummingbirds had migrated
south and there were few bumblebees all sea-
son that year. On die first day of the experiment
when the plants had just been placed in the
meadow all pollinators would be naive for both
red- and \ellow-flowered M. cordinalis plants
inasmuch as Red Butte Canyon is hundreds of
miles from the nearest M. cardinalis popula-
tion in northern Arizona. Therefore, the highly
significant preference for red appears to be
real and not the result of learned behavior.
Apparently, hummingbirds strongly preferred
the red flowers but also visited the yellow flow-
ers to some extent. The few bumblebee visits
did not suggest a preference.

Results show that the change in flower color
from red to yellow did affect the frequencies
of pollinator visits, but not in an all-or-none
way that would immediately establish repro-
ductive isolation. However, the change would
probably be enough to initiate partial, incipi-
ent reproductive isolation.

Would M. verbenaceus with its normal red
morph and mutant yellow morph produce the
same reactions in pollinators? The flowers of M.
verbenaceus differ from those of M. cardinalis
in that only the upper two corolla lobes are
reflexed, whereas all five of those of M. cardi-
nalis are reflexed. Both species sometimes
have wild populations with orange-red flowers
instead of the typical red flowers.

For the M. verbenaceus experiment, plants of
red-flowered and yellow-flowered individuals
from Vasey's Paradise in the Grand Canyon
plus plants of an orange-red-flowered popula-
tion from Yecora, Sonora, Mexico, were placed
on a lawn by clumps of native Gambel oak at
the mouth of Parley's Canyon, Salt Lake City,
UT This location had an abundance of pollina-
tors in contrast to the paucity of pollinators in
the Red Butte Canyon meadow used previ-
ously. The test population was observed for 15
periods of 1 h each at different times of day
from 26 July through 8 August 1988. On aver-
age, there were 73 red flowers, 87 orange flow-
ers, and 136 vellow flowers (see Vicken' 1990

for daily details of numbers and chi-square
calculations). On average, bumblebees visited
them 24, 56, and 128 times, respectively; and
hummingbirds 43, 98, and 52 times, respective-
ly (Vickery 1990). Bumblebees significantly
eschewed red and orange flowers and prefer-
entially visited yellow flowers. Hummingbirds
significantly preferred orange, visited red
flowers in proportion to their fi-equency in the
population, and significantly eschewed yellow
flowers. Results for M. verbenaceus are much
clearer than those for M. cardinalis. There is a
definite preference for yellow by bumblebees
and a clear avoidance of yellow by humming-
birds. Thus, this color change has lead to sig-
nificant, partial isolation between the normal
orange- and red-flowered morphs and the yel-
low-flowered mutant morph under the condi-
tions of this experiment.

Would M. cardinalis react like M. verbe-
naceus in the better experimental locality at
the mouth of Parley's Canyon? To probe this
question, I added red-, orange-, and yellow-
flowered morphs of M. cardinalis to the M.
verbenaceus red-, orange-, and yellow-flow-
ered moiphs of the previous experiment. The
new experiment was run 8-17 August 1988,
with the population being observed for 15
periods of 1 h each at different times of day.
On average there were 61 red, 57 orange, and
22 yellow flowers of M. cardinalis (see Vickeiy
1990 for day-to-day numbers and chi-square
calculations). On average, bumblebees visited
them 28, 30, and 29 times, respectively, and
hummingbirds 59, 60, and 6 times, respectively.
Bumblebees eschewed red and orange flowers
and significantly preferred yellow flowers
despite their low numbers in the population.
Hummingbirds significantly eschewed yellow
flowers and preferentially visited orange flow-
ers. M. verbenaceus plants were run again at
this time with M. cardinalis plants and exhibit-
ed the same attractiveness or lack of attrac-
tiveness to the pollinators as before. The pres-
ence of M. cardinalis flowers did not alter pol-
linator response to M. verbenaceus flowers.
The color shift from red (or orange) to yellow
leads to marked, partial reproductive isolation
in M. verbenaceus as well as in M. cardinalis.

How effective is the partial reproductive
isolation? To test this, I placed 198 plants of
M. verbenaceus — one-sixth yellow-flowered
and five-sixdis red-flowered to simulate a popu-
lation with a well-established mutant — in four

180

Great Basin Natuiulist

[Volume 55

experimental areas: the experimental garden
on the University of Utah campus. Red Butte
Canyon Natural Area, the mouth of Parley's
Canyon, and at Silver Fork, Big Cottonwood
Canyon, Salt Lake County, UT. 1 harvested
seeds of each plant and planted seeds han'est-
ed from 20 yellow-flowered plants and grew
them to flowering. If pollinators were visiting
the flowers at random, then they should pick
up and cany pollen from red flowers five times
more often than pollen from yellow flowers.
Pollen loads and resulting seed sets were well
below the 500-1500 seeds per capsule that
may occur in M. verbenaceus. So, results were
not skewed by saturation of the stigma. Also,
assuming all else to be neutral such as relative
growth rates of yellow- and red-pollen tubes,
speed of flowering of red- and yellow-flow-
ered plants, randomness of placement of red-
and yellow-flowered plants, and sample size of
red- and yellow-flowered plants, then the
expected five-to-one visitation rate should
hold. Inasmuch as red is genetically dominant
to yellow, then five-sixths of the seedlings
should be red-flowered and one-sixth yellow-
flowered; that is, of the 214 seedlings grown,
178 should be red-flowered and 36 yellow-
flowered. In fact, there were 86 red-flowered
seedlings and 128 yellow-flowered seedlings.
The ratio is 2 red to 3 yellow flowers, which is
far from the expected ratio of 5 red flowers to
1 yellow flower. This suggests considerable
pollinator faithfulness to one color or the
other. However, in addition to pollinator faith-
fulness there could be self-pollination.
Mimulus cardinalis does not self-pollinate but
M. verbenaceus does at the average rate of 10
seeds per capsule. Average normal seed set is
110 seeds per capsule. Therefore self-pollina-
tion would account for 9% of the yellow-flow-
ered seedlings; i.e., 9% of the 214 seedlings, or
19 seedlings, would be expected to be yellow-
flowered as a result of self-pollination. Of the
remaining 195 seedlings, five-sixths, or 162,
would be expected to be red, and one-sixth, or
33, would be expected to be yellow. Therefore,
I should expect to observe 162 red-flowered
seedlings and 52, i.e., 33 + 19 (the results of
self-pollination), yellow-flowered seedlings

instead of the 86 red-flowered and 128 yellow-
flowered seedlings actually obsened. This is a
highly significant difference (.r^ = 146.730,
p < .0001) and greatly strengthens the point
of pollinator faithfulness. Clearly, pollinator
preference for yellow and faithfulness to yel-
low are having a large effect, though not an
all-or-none effect. We are seeing strong incipi-
ent reproductive isolation due to color change.
In different areas with different conditions
and different guilds of pollinators the effect
might be less or might be stronger, even lead-
ing eventually to effective reproductive isola-
tion and speciation.

Acknowledgments

I appreciate the financial support of the U.S.
National Science Foundation, Grant BSR-
8306997. I thank Dr. Stephen Sutherland for
nectar measurements and for carrying out the
Red Butte Canyon experiment on M. cardinalis.

Literature Cited

Crow, J. E, and M. Kimura. 1970. An introduction to

population genetics theory. Harper & Row, New

York, NY. 591 pp.
DeWet, J. M. J. 1980. Origins of polyploids. Pages 3-15 in

W. H. Lewis, ed.. Polyploidy. Plenum Press, O.xford.
Grant, A. L. 1924. A monograph of the genus Mimulus.

Annals of the Missouri Botanical Gardens 11; 99-389.
Levin, D. A. 1993. Local speciation in plants: the rule not

the exception. Systematic Botany 18: 197-208.
Mayr, E. 1976. Evolution and the diversity of life.

Harvard University Press, Cambridge, MA. 721 pp.
Pollock, H. G., R. K. Vickery, Jr., and K. G. Wilson.

1967. Flavonoid pigments in Mimulus cardinalis and

its related species. I. Anthocyanins. American Joumiil

of Botany 54: 695-701.
Vickery, R. K., Jr. 1978. Case studies in the e\ olution of

species complexes in Mimulus. Evolutionai-y Biology

11:404-506
. 1990. Pollination experiments in the Mimulus car-

(linalis-M. Icwisii complex. Great Basin Naturalist 50:

153-159.
Vickery, R. K., Jr. and D. Sutheri^vnd. 1994. Variance and

replenishment of nectar in wild and greenhouse

populations of Mimulus. Great Basin Naturalist 54:

212-227.

Received 6 July 1994
Accepted 27 September 1994

FALL LAMB PRODUCTION BY A CALIFORNIA BIGHORN SHEEP

Matthew McCoyl, Wcilt Bodie^, and EiRoy Taylor^
Key words: paiiiirition. California bighorn sheep, Ovis canadensis, Idaho.

Parturition is timed to maximize sui^vival of
offspring (Thompson and Turner 1982).
Parturition occurring outside an optimum time
period lowers reproductive fitness and, there-
fore, should be selected against. Timing of
parturition in bighorn sheep {Ovis canadensis)
has been related to resource abundance (Geist
1974, Bunnell 1982, Thompson and Turner
1982, Risenhoover and Bailey 1988) and climat-
ic conditions (Stewart 1982). Parturition varies
by latitude between subspecies (Thompson
and Turner 1982) and by elevation within sub-
species (Risenhoover and Bailey 1988). Peak-
lambing periods occur in March for desert
bighorn sheep (O. c. nelsoni; Hanson 1960,
Sandoval 1980, Witham 1983), May for Cali-
fornia bighorn sheep (O. c. californiana; Jones
1950), and early June for Rocky Mountain
bighorn sheep (O. c. canadensis; Bunnell 1982,
Thompson and Turner 1982). Unusual lambing
periods such as January for desert bighorn
(Russo 1956, Welles and Welles 1961) and July
for Rocky Mountain bighorn (Stewart 1982)
were attributed to extremes in climatic condi-
tions and elevations.

Vegetation in the Big Jacks Creek drainage,
Owyhee County, ID, is dominated by sage-
brush {Artemisia spp.), bluebunch wheatgrass
{Pseudoroegneria spicata), and Sandberg blue-
grass {Poa sandbergii). Climatic conditions are
characterized by warm, dry summers and cool
winters. Total precipitation from November
1988 through March 1989 was equal to the 10-
year average; however, most precipitation
occurred as rain in November and March. Ten
ewes, one ewe lamb, and three ram lambs
from Chilcotin, BC, and two rams from East
Fork Owyhee River, ID, were transplanted to

Big Jacks Creek during February and March

1988. Four ewes were fitted with radio-collars
and periodically located to monitor their
movements and status. Three radio-collared
ewes were observed with lambs in May 1988.
A fourth collared ewe (34) was obsei-ved with a
lamb (A) less than two weeks old (based on size
and behavior) on 26 October 1988. An average
gestation period of 174 days for bighorn sheep
(Shackleton et al. 1984) indicated conception
occurred about 25 April 1988. Ewe 34 and
lamb A were located monthly through March

1989. During 1987-1990, onset of parturition
occurred from 11 April to 3 May, and mating
activity was observed between October and
December in an adjacent drainage. The birth
of lamb A occurred approximately six months
out of cycle. Ewe 34 was observed 4 January
1990 with a lamb (B) that appeared to have
been born during the normal lambing period
(April-June 1989). Lamb B was conceived be-
tween October and December 1988 while ewe
34 was nursing lamb A.

Ewe 34 may not have bred in 1987, or stress
related to transplanting may have caused her
to abort. Stress can affect any aspect of repro-
duction (DeForge 1976). Contact with rams
during March and April 1988 may have caused
ewe 34 to come into estrus. Presence of males
has been found to induce estrus in female
merino sheep (Watson and Radford 1960) and
feral goats (Coblentz 1980). Recurrent estrus
was observed in a cow elk {Cervus elaphiis)
that was associated with bulls but not bred
during previous estrus periods (Morrison
1960).

Lamb survival has been related to forage
quality (Wehausen et al. 1987, Festa-Bianchet

^Idaho Department of Fish and Game, 3101 S. Powerline Road, Nampa, ID 83686. Present address: Bureau of Land Management, 3948 Development
Avenue, Boise, ID 83705.

2ldaho Department of Fish and Game, 3101 S. Powerhne Road, Nampa, ID 83686.
•^Bureau of Land Management, 3948 Development Avenue, Boise, ID 83705.

181

182

Gkeat Basin Natuk.\list

[Volume 5.'

1988a), precipitation patterns (as they affect
plant growth; Douglas and Leslie 1986), popu-
lation density (Douglas and Leslie 1986), and
mother's age (Festa-Iiianchet 1988a). Cheatgrass
hrome {Broiniis tectormn) seedlings were avail-
able in November, and Sandberg bluegrass
greenup was observed in Januaiy South-facing
slopes were generally free of snow soon after
storms. Cattle grazing occurred in riparian
areas and on plateaus adjacent to drainages,
areas that received limited use by bighorns
during summer and lambing periods. Bighoin
and mule deer {Odocuileiis hemionus) popula-
tions were at low densities. Competition for
forage was probably not a limiting factor.
Festa-Bianchet (1988b) reported that lambs
born to ewes four to nine years old had signifi-
cantly higher survival rates than those born to
two- to three-year-old ewes. Ewe 34 was esti-
mated to be five years old in 1988. Mild cli-
matic conditions, availability of green forage
during weaning, limited competition for for-
age, and probable previous lambing experi-
ence allowed ewe 34 to raise an out-of-season
lamb and survive concurrent fall/winter lacta-
tion and gestation periods. This observation
suggests that under favorable conditions
bighorn sheep may be able to successfully
reproduce outside generally observed repro-
ductive periods.

Literature Cited

Bunnell, E L. 1982. The lambing period of mountain
sheep: synthesis, hypotheses, and tests. Canadian
Journal of Zoology 560: 1-14.

CoBLENTZ, B. E. 1980. A unique ungulate breeding pat-
tern. Journal of Wildlife Management 44: 929-933.

Deforce, J. R. 1976. Stress: Is it limiting bighorn?
Desert Bighorn Council Transactions 19: 30-31.

Douglas, C. L., and D. M. Leslie. 1986. Influence of
weather and density on lamb sur\i\al of desert
mountain sheep. Journal of Wildlife Management
50: 153-156.

Eesta-Bianchet, M. 19S8a. Nursing behavior of bighorn
sheep: correlates of ewe age, parasitism, lamb age,
birthdate and sex. Animal Behavior 36:1445-1454.

. 1988b. Age-specific reproduction of bighorn ewes

in Aliierta, Canada, Journal of Mammalogy 69:
157-160.

Geist, V. 1974. On the relationship of ecology and behav-
ior in the evolution of ungulates. Pages 235-246 in \.
Geist and E Walthers, editors. The behavior of imgu-
lates and its relation to management. International
Union Consenation Nature Publication.

Hanson, G. 1960. Lamb survival on the Desert Game
Range. Desert Bigliom Council Transactions 4: 60-61.

Jones, E L. 19.50. A survey of the Sierra Nevada bighorn.
Pages 29-76 in Sierra Club Bulletin 1950.

.MoRRLSON, J. A. 1960. Characteristics of estrus in captive
elk. Behaviour 16: 84-92.

Risenhoover, K. L., AND J. A. Bailey. 1988. Growth-
rates and birthing period of bighorn sheep in low-
elevation environments in Colorado. Journal of
Mammalogy 69: 592-597.

RUSSO, J. P 1956. The desert bighorn sheep in Arizona.
Arizona Game and Fish Department, Wildlife
Bulletin 1.

Sando\'AL, a. V 1980. Preferred habitat of desert bighorn
sheep in the San Andres Mountains, New Mexico.
Unpublished thesis, Colorado State University, Fort
Collins. 282 pp.

SlIACKLETON, D. M., R. G. PETERSON, J. HA'iAVOOD, AND A.

Bottrell. 1984. Gestation period in Ovis canaden-
sis. Journal of Mammalogy 65: ■337-.338.

Stewart, S. T. 1982. Late parturition in bighorn sheep.
Journal of Mammalogx' 63: 154-1.55.

Thompson, R. W, .\nd J. C. Turner. 1982. Temporal geo-
graphic variation in the lambing season of bighorn
sheep. Canadian Journal of Zoology 60: 1781-1793.

W.^TSON, R. H., and H. M. R\DF()RD. 1960. Influence of
rams on the onset of oestrous in merino ewes in the
spring. Australian Joimial of Agricultural Research 2:
65-71.

Wehausen, J. D., V. C. Bleicii, B. Bloxg, and T. L.
RUSSI. 1987. Recruitment dynamics in a southern
California mountain sheep population. Joinnal of
Wildlife Management 51: 86-98.

Welles, R. E., and F B. Welles. 1961. The bighorn of
Death Valley. Fauna of the National Parks of the
United States, Fauna Series 6. 242 pp.

WiTHAM, J. B. 1983. Desert bighorn sheep in southwest-
ern Arizona. Unpublished dissertation, Colorado
State University, Fort Collins. 93 pp.

Received 22 November 1993
Accepted 20 June 1994

AGE, GROWTH, AND REPRODUCTION OF LEATHERSIDE CHUB

{GILA COPEI)

Jerald B. Johnsonl-^, Mark C. Belk'i, and Dennis K. Shiozawa^
Key words: Gila copei, leathersicle chub, life history, reproduction, age. growth.

The leatherside chub {Gila copei) is a small
cyprinid native to eastern and southern areas
of the Bonneville Basin of Utah, Idaho, and
Wyoming, to Wood River, Idaho, and to regions
of the Snake River, Idaho and Wyoming,
above Shoshone Falls (Baxter and Simon 1970,
Simpson and Wallace 1982, Sigler and Sigler
1987). Gila copei is currently listed as a can-
didate for federal protection under the
Endangered Species Act.

Conservation efforts for G. copei would
benefit from accurate life histoiy data, yet the
life history of G. copei is not well known. This
species was thought to live less than five years
(Sigler and Sigler 1987). Based on bright col-
oration of males and abdominal distension in
females, Sigler and Miller (1963) concluded G.
copei spawns between lune and August. Using
similar coloration criteria on males from
Sulphur Creek, WY, Baxter and Simon (1970)
suggested breeding occurred in late summer;
Simon (1951) found females distended with
eggs in early August. Other than these few
obser\'ations, no data on age, growth, or repro-
duction are available.

We present data on age, growth, and repro-
duction of leatherside chub from central Utah;
these data were generated as a first step to
understanding and protecting this potentially
threatened, endemic fish species.

Study Site

Age and growth data were obtained from
36 G. copei collected from Thistle and Main
creeks, both tributaries to larger rivers that
flow into Utah Lake. Thistle Creek, a tributaiy
to Spanish Fork River, was sampled in May,
September, and October 1993 {n = 25)
(39°52'N, lir32'W) at an elevation of approx-

imately 1700 m. Main Creek flows directly
into Deer Creek Reservoir (an impoundment
of Provo River) and was sampled in Inly 1993
(n =11) 500 m upstream from the reservoir
(40°24'N, lir28'W) at an elevation of 1650 m.
Chubs used for determining reproductive pat-
terns (below) were collected from the latter
site in 1978-79. Creeks at both locations flow
slowly at low gradient through meadows. The
Main Creek site is downstream from beaver
dams; tall grasses and small trees grow along
banks. Collections for both creeks were made
from vegetated pools separated by shallow
riffles; stream substrate is silt, gravel, and
boulders.

Materials and Methods

Because Gila copei is a species of special
concern, our permit was limited to 40 speci-
mens, and care was taken to collect the entire
size range. Following collection by elec-
troshocking, fish were placed on ice and trans-
ported to Brigham Young University (BYU)
where they were stored frozen. Individuals
were then thawed, rinsed in water, blotted diy,
and weighed (0.001 g) on a Denver Instmment
XD-1200D® electronic balance; standard
length (SL) was determined (0.01 mm) using
Fowler Ultra Cal III® electronic calipers.

Ages were determined by grinding otoliths
(lapilus) to a thin section and counting opaque
bands under a Leica dissecting microscope
(40X). Opaque bands were validated as annuli
using a marginal increment analysis; because
juvenile (ages 1-2) and adult fish (ages 3-8)
demonstrated distinct growth rates, they were
evaluated separately. Identification of annuli
was facilitated by generating digitized images
of otoliths on a video monitor using a Hitachi

'Department of Zoolog>'. Brigham Young University', Provo, UT 84602.
^Address correspondence to this author

183

184

Great Basin Natufl^list

[Volume 55

CCTV® camera fitted to a Heerliru^g Wild®
dissecting microscope. Annual growth incre-
ments along the longest axis of the otolith
were then measured (0.001 mm) using video
image analysis software (Mocha release 1.0,
Jandel Scientific; Rundel 1993), which reduces
measurement errors introduced when reading
otoliths directK under a microscope (McGowan
et al. 1987).

Size at age was back-calculated from otolith
measurements using a modified Fraser-Lee
formula (Campana 1990):

L, = L„ + (L,, - LJ(R, - R„)/(R, - R„),

where L,. is estimated SL at age x, L^. is length
at capture, R^ is otolith radius at age x, and R^.
is otolith radius at capture. L^ is estimated
length at swim-up (estimated at 4 mm from
data on Gila atraria; Varley and Livesay 1976),
and R„ is otolith radius at swim-up (estimated
from otoliths at 0.01 mm).

There was no significant difference in back-
calculated lengths at age I between Main and
Thistle creek chubs (Main Creek, n = 11;
Thistle Creek, n = 25; T = 1.96, d.f = 34, P
= .06) . Numbers of age II (n = 2) and age III
(n = 3) fish from Main Creek precluded statis-
tical comparisons; however, back-calculated
lengths at age II and age III for Main Creek
fish were within the range of comparably aged
fish from Thisde Creek. Hence, growth data for
the two populations were combined. An age-
growth curve was generated for the combined
samples by averaging back-calculated sizes at
age.

Leatherside chub collected in 1993 were
sexed by dissection and examination of gonads;
individuals lacking mature gonads were classi-
fied as juveniles. Immature testes were trans-
lucent and threadlike, while mature testes
were opaque (white or pinkish) and firm.
Reproductive states of ovaries were deter-
mined according to criteria in Holden and
Berry (1983); immature ovaries were small,
translucent, and lacked yolked ova; mature
ovaries were larger and contained both imma-
ture ova and firm, yolked ova.

Reproductive data were obtained from a
collection of 176 adult leatherside chubs
archived in the Monte L. Bean Museum at
BYU (#5592-5619, 5629-5686, 5688-5775).
Monthly collections from Main Creek (August
1978 to September 1979) were made using

minnow traps, hand nets, and electrofishing
gear, and preserved in formalin. Daily temper-
ature was recorded from September 1978 to
July 1979.

Standard length was measured (mm), and
presei-ved wet mass (0.01 g) was recorded, for
each specimen. Gonads from all {n = 176)
individuals >50 mm SL were removed and
weighed (0.001 g). No fish <50 mm SL had en-
larged gonads. A gonadosomatic index (GSI)
was generated for each fish using the follow-
ing formula (Andreasen and Barnes 1975):

GSI = (gonad weight / body weight) X 100.

Mean monthly GSI values were used to deter-
mine onset and duration of spawning. Ova
counts were made on nine fish collected in
May 1979. The relationship between number
of ova present and SL was evaluated by linear
regression.

Results

Opaque bands on leatherside chub otoliths
appear to be valid annuli as demonstrated by
an increase in the marginal growth increment
throughout the growing season for both adult
and juvenile fish (Fig. 1).

U.iiO -

■ = Ages 1-2

E 0.20 -

• = Ages 3-8

\ /*

E 0.15

^
r

- 0.10 -
CD

/ J

TO 0 05

0.00 -

{

1 1 1 1

10 11

Month

Fig. 1. Mean inarginal increment widths (±2 S.E.) mea-
sured from otoliths in Gila copei (n = 36). Immature age
classes (1-2) and mature age classes (3-8) plotted sepa-
rately.

1995]

Notes

185

Ages of 36 G. copei collected in 1993 ranged
from one to eight years, with SL of 38-110
mm (Table 1). Chubs grew rapidly to -58 mm
SL at about age II (Fig. 2, Table 1). From age
II on, annual growth was slower and fairly
uniform. Mean GSI values for males and
females (Fig. 3a) were highest for both sexes in
spring with maxima in May (female GSI =
12.3, male GSI = 2.7). Increasing water tem-
peratures from Januaiy through May (Fig. 3b)
were associated with increased GSI values for
both sexes. Average water temperature in May,
corresponding to GSI maxima, was 9.4° G.

Fecundity (as measured by ovimi counts) in-
creased with SL for females collected in May

1979 and ranged from 938 in a 67-mm-SL,
5.9-g female to 2573 in a 92-mm-SL, 14.6-g
female. Average count for leatherside chubs
collected in May 1979 was 1813. Significant
correlations existed between SL and fecundity
(R2 = .82, P < .05, n = 9) and weight and
fecundity (fi2 = .72, P < 0.05, n = 9).

Discussion

A maximal age of eight years in our sample
of G. copei indicates a life span much longer
than previously thought (Sigler and Sigler
1987). Longevity in G. copei may be a life his-
toiy trait that has evolved in response to living

Table 1. Capture and back-calculated standai^d lengths (SL) of Gila copei from Thistle and Main creeks, central Utah.

SLat

capture

Mean

back-calculated SL at

annulus

Age

Mean

Range

38-49

65-85

71-104

—

83-110

88-105

—

Overall

means

100 r

■o
c

Fig. 2. Mean back-calculated standard lengths at age (±2 S.E.) for Gila copei {ii
indicates estimated age at first reproduction.

36) in central Utah. Shaded block

186

GiucAT Basin Naturalist

[Volume 55

a 18

1 I I r
A S 0 N D J

F M A M

Month

J J A

JASONDJ FMAMJ JA

Month

Fig. 3a. Mean gonadosomatic indices (±2 S.E.) for male
and female Gihi copei (n = 176); b, mean monthly temper-
atures (±2 S.E.) from August 1978 to July 1979 in Main
Creek, Wasatch County, UT.

in an environment where annual precipitation
and stream flow vaiy considerably. Successful
chub reproduction and recruitment may be
uncertain in any given year. An extended life
span would increase the likelihood that appro-
priate environmental conditions for reproduc-
tive success would be met at some time in an
individual's life; thus, longevity may be a "bet-
hedging" strategy (Stearns 1976) for living in
unpredictable conditions.

The growth pattern of G. copei is typical of
other fishes in which rapid juvenile growth
decreases at the onset of sexual maturity as
finite energy resources are allocated to both
growth and reproduction (Roff 1984). The
inflection point in the growth curve (Fig. 1),

coupled with the facts that the smallest fish
with dexeloped gonads collected in 1993 was
65.2 mm SL and no fish in the museum collec-
tion <50 mm SL had enlarged gonads, sug-
gests that first reproduction in G. copei occurs
at age II.

High OS I in May followed by decreased
GSI in June and minimal values in July and
August (Fig. 3a) indicates that peak spawning
occurred in May, with some activity possibly
extending into early June. Gila copei appar-
entl)' follows a pattern of reproduction common
to various cyprinids living in temperate climates
(Munro et al. 1990). This pattern is character-
ized by the onset of spawning in late spring
followed by a period of gonadal recrudescence
and inactivity; size of gonads begins to increase
in autumn and continues through winter, with
final maturation occurring in early spring.

If temperature influences the onset of
spawning, differences in temperature (as a
function of latitude) between Main Creek (this
study) and southwestern Wyoming (Simon
1951) could explain the discrepancy between
onset of chub spawning at these locations (May
vs. August). A more detailed investigation of
G. copei will be required to resolve questions
of differences in reproductive and life histoiy
characteristics among populations.

Literature Cited

Andreasen, J. D., .WD J. R. B.\rnes. 1975. Reproductive
life histoiy oi Catostomu.s aniens and C. discobolus in
the Weber River, Utah. Copeia 1975; 643-648.

Baxter, G. T, and J. R. Si.vion. 1970. Wyoming fishes.
Wyoming Game and Fish Department, Cheyenne.
168 pp.

Camfana, S. E. 1990. How reliable are growth back-calcu-
lations based on otoliths? Canadian Journal of Fish-
eries and Aquatic Science 47; 2219-2227.

HOLDEN, M. A., AND C. R. Berry. 1983. Vitellogenesis in
the Utah chub {Gila atraria) and its use in evaluating
reproduction in a transferred population. Encyclia
60: 32-42.

McGowAN, W E, E. D. Prince, and D. W Lee. 1987. An
ine.vpensive microcomputer-based system for making
rapid and precise counts and measurements of zona-
tion in \'ideo displa\ed skeletal structures in fish.
Pages 385-395 in R. C. Summerfelt and G. E. Hall,
editors. Age and growth of fish. Iowa State Univer-
sity Press, Ames.

Munro, A. D., A. R Scott, and T J. Lam. 1990. Repro-
ductive seasonality in teleosts; environmental influ-
ences. CRC Press, Inc., Boca Raton, FL. 254 pp.

Roff, D. A. 1984. The evolution of life histoiy parameters in
teleosts. Canadian Journal of Fisheries and Aquatic
Science 41: 989-1000.

Rundel, R. 1993. Mocha image analysis software: user's
manual. Jandel Scientific, San Raliiel, CA. 189 pp.

1995]

Notes

187

SiGLER, W. E, AND R. R. MiLLER. 1963. Fishes of Utah.

Utah Department of Fish and Game, Sah Lake Cit\'.

203 pp.
SiGLER, W. F, AND J. W. SiGLER. 1987. Fishes of the Great

Basin: a natural histor>'. University of Nevada Press,

Reno. 425 pp.
Simon, J. R. 1951. Wyoming fishes. Wyoming Game and

Fish Department, Cheyenne. 129 pp.
Simpson, J. C., and R. L. W.all.\ce. 1982. Fishes of Idaho.

University Press of Idaho, Moscow. 238 pp.

Stearns, S. C. 1976. Life history tactics: a review of the
ideas. Quarterly Review of Biology 51: 3—47.

Varley, J. D., AND J. C. Ll\ESAY. 1976. Utah ecology and
life history of the Utah chub, Gila atraria, in Flaming
Gorge reservoir, Utah-Wyoming. Utah Division of
Wildlife Resources, Publication 76-16, Salt Lake
Cit\'. 29 pp.

Received 15 June 1994
Accepted 7 September 1994

CONSUMPTION OF A TOXIC PLANT {ZIGADENUS
PANICULATUS) BY MULE DEER

William S. Longland'-^ and Charlie Clements^
Key words: death eamas. Zitiadfiius, iiinle deer, poisonous plants, cocrolution.

The abundance of green vegetation in nature
can yield false impressions of the availability
of food resources to herbivores because many
plants have evolved anti-herbivore defenses.
Defensive mechanisms commonly include
incorporation of distasteful or toxic secondary
chemical compounds into plant tissues. Effects
of different compounds on consumers range
from mild (unpalatable) to severe (illness or
death fi-om poisoning). Herbivores have conse-
quently evolved a host of means for coping with
defensive compounds, resulting in an evolu-
tionary arms race between plants and herbi-
vores (Freeland and Janzen 1974). Although
evidence of plant/herbivore coevolution can
be found for herbivores ranging from phyto-
phagous insects to mega-vertebrates, we con-
centrate specifically on mule deer {Odocoileus
hemoniiis) feeding on toxic plants.

Because domestic grazing animals lack a
coevolutionaiy history with the plant commu-
nities in which they forage, they are often
affected by toxic secondary compounds to a
greater degree than native herbivores. This
has significant economic impact on the range
livestock industry due to direct losses, such as
death, reduced fecundity, or reduced weight
gain, and to indirect costs of minimizing such
losses (Nielsen et al. 1988, James et al. 1992).
Historical familiarity with local plant assem-
blages has provided herbivores foraging in
their native ranges with two advantages over
introduced domestic counterparts (Freeland
and Janzen 1974, Laycock 1978, Laycock et al.
1988). First, native mammals often avoid eating
toxic plant species that are eaten by domestic
grazers. For example, toxic plants eaten by
livestock, such as azalea {Azalea spp.) and lark-
spur {Delphinium spp.), are avoided by mule
deer even when these plants are abundant

(Dixon 1934). Second, in most cases of native
ungulates eating a plant species that is toxic to
domestic animals, the plant does not produce
noticeable toxic effects in the fomier, indicat-
ing that native herbivores may possess detoxi-
fication mechanisms for some plant toxins (Lay-
cock 1978). Thus, deer consume without adverse
effects a variety of plants poisonous to live-
stock (Stoddart and Rasmussen 1945, Dean
and Winward 1974). Reciprocal examples in
which native plants are toxic to native herbi-
vores, but benign to domestic animals, are
lacking in the literature.

Herein we report on four years of obsei-va-
tions of an eastern Sierra Nevada mule deer
herd feeding on substantial quantities of foot-
hill death camas {Zigadenus paniculatus), a lilia-
ceous bulb plant that is toxic to domestic sheep,
cattle, and horses (Fleming et al. 1921, Kings-
buiy 1964, James et al. 1980, Panter et al. 1987).
The genus Zigadenus includes several species,
all containing toxic steroidal alkaloids (James
et al. 1980). Death camas emerges earlier than
most plants, making it particularly hazardous
for spring grazing of livestock (Panter and
James 1989). These plants have been variously
described as "the most important poisonous
plants in the western U.S." (Kingsbury 1964)
and "the most dangerous poisonous plants in
North America" (Clarke and Clarke 1975).
Foothill death camas has been described as
one of the more toxic Zigadenus species
(Kingsbuiy 1964, James et ah 1980).

Our study site is located at T20N, R18E,
S36 just west of Reno, NV, on an alluvial fan at
the southern base of Peavine Mountain. Woody
vegetation is dominated b\' basin big sage-
brush {Aiiemisia thdentata tridentata) and bit-
terbrush {Purshia tridentata). Death camas
emerges at this site in mid- March, flowers in

lUSDA, Agricultural Research Service, 920 Valle\ Koad, Reno, NV 89512.
^Address correspondence to this author.

188

1995]

Notes

189

April, and remains green into May. A herd of
mule deer, usually numbering 20-25 animals,
has foraged extensively in this area from
October to May since we began making obser-
vations in fall 1988.

We first noticed deer consuming death
camas on 28 March 1989 (before plants flow-
ered) and confirmed this with additional obser-
vations in all subsequent years. Examination
of death camas foliage immediately after deer
left the foraging patches consistently revealed
fresh herbivore damage. We found that deer
herbivoiy left a characteristic leaf damage pat-
tern, with most or all leaves of a foraged plant
cleanly bitten off perpendicular to their long
axes. In addition to direct obsei^vations of deer
consuming death camas, fresh deer pellet
groups were found in patches of plants ex-
hibiting this characteristic damage pattern
during all five springs (1989-1993). During
observation periods we found no evidence of
deer exhibiting toxic effects from death camas
consumption, and neither we nor personnel
from the Nevada Department of Wildlife
(which surveys deer in the area by air) have
found any fresh deer carcasses in the vicinity.

Each year from 1990 through 1993 we
walked 10-12 permanently located, parallel
transects and categorized all death camas
plants seen as either eaten or uneaten by deer
Transects were 500 m long, 20 m wide (i.e., we
generally saw all plants occurring < 10 m from
the transect lines), and spaced 30 m apart.
Usually, deer removed only the distal 2-5 cm
of leaves, but on several occasions we found
plants eaten to within 2 cm of ground level.
Plants were considered eaten regardless of the
amount of leaf removed. We tested these data
for temporal differences in frequency of death
camas consumption by comparing numbers of
eaten versus uneaten plants among the four
years of the study using a G-test of indepen-
dence. We similarly tested for spatial effects
on consumption by comparing eaten versus
uneaten plant counts among individual tran-
sect lines within years.

There are at least two potential explana-
tions for the partial consumption of leaves that
we noted. Perhaps ends of leaves are less toxic
than leaf bases, and deer preferentially con-
sume less-toxic plant parts. Kingsbuiy (1964)
suggests that death camas bulbs are the most
toxic part of the plants, and a gradient of
decreasing toxicity could occur from bulbs to

ends of leaves. Alternately, deer may occasion-
ally sample plants in their environment (Free-
land and Janzen 1974), and removal of short
leaf segments may represent cautious sam-
pling of a plant deer find undesirable. The lat-
ter possibility (sampling) seems less likely
than the former (selectivity) because we have
observed individual deer feeding on several
death camas plants consecutively. Furthermore,
total numbers of plants consumed on our tran-
sects were several orders of magnitude greater
than the number of deer foraging in the study
area, and it seems unlikely that deer would
have to sample repeatedly so many plants to
discover they are undesirable.

We found significant annual variation in the
frequency of death camas consumption, rang-
ing from 3.8% to 18.9% of total plants counted
showing evidence of deer herbivory (G =
232.8, df = 3, F < .0001; Table 1). Maximum
and minimum percentages of plants eaten
(Table 1) illustrate that frequency of herbivory
also varied spatially; in each of the four years
we sampled there was significant variation
among transects in numbers of plants eaten
(F < .001 for all years). While the minorit)' of
plants in the local death camas population
were eaten, the values in Table 1 also represent
a surprisingly high frequency of herbivory on
a plant species with such a notorious reputation.

The relatively low proportions of damaged
plants indicate that deer may be selective for
particular death camas plants. This is support-
ed by the fact that deer generally ate only a
few non-neighboring plants from large patches
of death camas; rarely did the majority of
plants within a patch show evidence of her-
bivory. The apparently selective use of indi-
vidual death camas plants, significant tempo-
ral and spatial variation in death camas use,
and infrequent extensive herbivory on small
patches of plants could be due to variation
among plants or patches in toxicity or to dif-
fering availabilities of superior foods leading
to variation in the use of toxic foods.

Our observations suggest that death camas is
more palatable to deer than to domestic cattle
or sheep. Domestic animals must be force-fed
death camas in captivity experiments (Fleming
1918, Fleming et al. 1921, Panter et al. 1987)
and must be stressed or left with few alterna-
tive foods in nature before they consume it
(Panter et al. 1987). Mule deer at our study site,
however, occur at a low density and consume

190

Great Basin Naturalist

[Volume 55

Table 1. Numbers and percentages of foothill death canuis plants consiinucl in nuile deer along 500-ni transects,
1990-1993, at Peavine Mountain (Washoe County, NV).

Xuniher of

Number of

plants"

Pla

nts

eaten per transect {%)

Yeai-

transects

Total

f':aten

axinunii

inimuin

XtS.D.

1990

2646

501

29.0

7.3

18.6 ± 10.6

1991

2726

259

44.7

2.6

16.4 ±13.4

1992

3073

118

32.6

1.6

8.3 ± 9.6

1993

3799

202

15.4

2.5

8.0 ± 4.0

■'huliiilcs ccinihini-d data from all tiansccts

death camas each spring akliougli alternative
plants are available. Because bitterness is a
general property of alkaloids (Laycock 1978),
death camas is quite bitter Most herbivores
apparently find bitterness distasteful (Laycock
1978); howexer, bitterbrush {Purshia tridentata),
which is named for its bitterness, is a pre-
ferred browse plant of mule deer. Although
bitterbrush is also consumed by domestic
ungulates, it is not highly preferred by them,
perhaps because bitterness is a greater feed-
ing deterrent to domestic animals than to deer
Native herbivores have been observed con-
suming a variety of plant species known to be
toxic to domestic herbivores (Laycock 1978),
including an anecdotal report of mule deer in
Utah consuming death camas and several other
toxic plants (Stoddart and Rasmussen 1945).
Recent work stimulates the interesting possi-
bility that herbivores consume specific toxic
plants to rid themselves of gut parasites
(Barbosa et al. 1991, Gauld and Gaston 1992).
However, this hypothesis only addresses why
toxic plants are consumed lather than wh\' the
consumers are physiologically able to tolerate
the toxins. Although we can only speculate
about reasons mule deer are less affected by
death camas toxicity than domestic ruminants,
a likely explanation is that deer possess rumen
microflora that have acquired the ability
through natural selection to detoxify this plant
(Freeland and Janzen 1974, Laycock 1978).
Such selection is perhaps to be expected for
native ruminants because the microflora com-
mimity has seen prolonged exposine to native
toxic plants. It is certainly possible, however,
that deer are able to detoxify death camas by
some other mechanism. For example, since
deer are browsers, their diets include large
amounts of tannins (Cooper and Owen-Smith
1985, Bobbins et al. 1987) that may precipitate
the alkaloids in death camas into a harmless
tannate (Freeland and Janzen 1974).

Because even limited past exposure of a
herbivore to a particular toxin can result in
reduced toxic effects, selection for detoxifying
rumen microflora may also account for intra-
specific variation in toxicity among individuals
of a domestic species. Such individual \'aria-
tion in susceptibility to death camas toxicity
has been reported in force -feeding experi-
ments with domestic sheep (Fleming et al.
1921, Kingsbun' 1964). Perhaps it is possible to
utilize this indixidual variation in selectively
breeding for reduced vulnerability to particu-
lar toxins. Currently, most domestic grazing
animals are products of artificial selection for
productivity, rather than for resistance to envi-
ronmental challenges.

Another avenue for applied research con-
cerns the possibility of ameliorating effects of
toxic plants through the transfer of rumen
innocula from animals resistant to specific tox-
ins to those that are susceptible. Jones (1985)
reported that transfer of rumen cultures from
goats that were resistant to poisoning by Leii-
caena leucocephala to susceptible goats and
steers eliminated adverse effects of Leucaena
consumption in the previously susceptible ani-
mals. This example suggests that even inter-
specific transfer of rumen fluids may effective-
ly reduce toxic effects in some cases.

Deer herbivory we witnessed on Peavine
Mountain may affect the demography of the
local death camas population. Defoliation exper-
iments indicate that death camas probably suf-
fers reduced reproductive output after her-
bivory (Tepedino 1982, Knapp 1986). While
plants adapted to herbi\'or\' ma\' compensate
for loss of biomass by allocating additional
energy to growth and/or reproduction, highly
toxic species instead employ an evolutionary
strategy of defense against herbivory and thus
may not exhibit compensation (Gates 1975,
Laycock 1978). When such defenses are cir-
cumvented bv herbivores with detoxification

1995]

Notes

191

mechanisms, toxic plants should experience
reduced fitness.

Acknowledgments

We thank Dr. Jeanne Chambers, Dr. Kip
Panter, and two anonymous reviewers for
thoughtful reviews of the manuscript. This
paper is a contribution of the USDA, Agricul-
tural Research Senice, Consei-vation Biology
of Rangelands Unit, Reno, NV

Literature Cited

Barbosa, E, E Gross, and J. Kemper. 1991. Influence of
plant allelochemicals on the tobacco hornworn and
its parasitoid, Cotesia congregata. Ecology 72:
1567-157,5.

Gates, R. G. 1975. The interface between slugs and wild
ginger; some evolutionary aspects. Ecology 56:
.391-400.

Glarke, E. G. G., and M. L. Glarke. 1975. Veterinary
toxicologv'. Macmillan Eublishing, New York, NY.

GooPER, S. M., AND N. Owen-Smith. 1985. Gondensed
tannins deter feeding by browsing ruminants in a
South African savanna. Oecologia 67: 142-146.

Dean, R. E., and A. H. Winward. 1974. An investigation
into the possibility of tansy ragwort poisoning of
blacktailed deer Journal of Wildlife Disease 10:
166-169.

Dl.xON, J. S. 19.34. A study of the life history and food
habits of mule deer in California, part 2. Food habits.
Galifornia Fish and Game 20: 31.5-354.

Fleming, G. E. 1918. Range plants poisonous to sheep
and cattle in Nevada. Nevada Agricultural E.xperi-
ment Station Bulletin 95.

Fleming, G. E., N. F Eeterson, M. R. Miller, and L. H.
Wright. 1921. Death camas. Flants poisonous to
sheep cattle. Nevada Agricultural Experiment
Station Bulletin 101.

Freeland, W J., AND D. H. Janzen. 1974. Strategies in
herbivory by mammals: the role of plant secondaiy
compounds. American Naturalist 108: 269-289.

Gauld, I. D., AND K. J. Gaston. 1992. Flant allelochemi-
cals, tritrophic interactions and the anomalous diver-
sity of tropical parasitoids: the "nasty' host hypothe-
sis. Oikos 65: 3.53-357.

James, L. F, R. E Keeler, A. E. Johnson, M. G. Williams,
E. H. Gronin, and J. D. Olsen. 1980. Flants poison-

ous to livestock in the western states. USDA-SEA
Agricultural Information Bulletin 415.

James, L. F, D. B. Nielsen, and K. E. Fanter. 1992.
Impact of poisonous plants on the livestock industiy
Journal of Range Management 45: 3-8.

Jones, R. J. 1985. Leucaena toxicity and the ruminal
degradation of mimosine. Fages 111-119 in A. A.
Seawright, M. E Hegarty, L. E James, and R. F.
Keeler, editors, Elant toxicology. Eroceedings of the
Australia-U.S. Eoisonous Elant Symposium, Brisbane,
Queensland Department of Erimary Indirstries,
Yeerongpilly.

Kingsbury, J. M. 1964. Eoisonous plants of the United
States and Ganada. Erentice-Hall, Englewood Gliffs,
NJ.

Knapp, A. K. 1986. Ecophysiolog\' of Zigadenti.s luittallii, a
toxic spring ephemeral in a warm season grassland.
Oecologia 71: 69-74.

Laycock, W. a. 1978. Goevolution of poisonous plants and
large herbivores on rangelands. Journal of Range
Management 31: 335-342.

Laycock, W. A., J. A. Young, and D. N. Ueckert 1988.
Ecological status of poisonous plants on rangelands.
Eages 27-42 in L. F James, M. H. Ralphs, and D. B.
Nielsen, editors, The ecology and economic impact
of poisonous plants on livestock production.
Westv'iew Eress, Boulder, GO.

Nielsen, D. B., N. R. Rimbey, and L. F James. 1988.
Economic considerations of poisonous plants on live-
stock. Eages .5-15 in L. E James, M. H. Ralphs, and
D. B. Nielsen, editors. The ecology and economic
impact of poisonous plants on livestock production.
Westview Eress, Boulder, GO.

Eanter, K. E., and L. F James. 1989. Death camas — early
grazing can be hazardous. Rangelands 11: 147-149.

E\NTER, K. E., M. H. Ralphs, R. A. Smart, and B. Duelke.
1987. Death camas poisoning in sheep: a case report.
Veterinaiy and Human Toxicology 29: 4.5-48.

Bobbins, G. T, S. Mole, A. E. Hagerman, and T. A.
Hanley. 1987. Role of tannins in defending plants
against njminants: reduction in dn.' matter digestibil-
ity? Ecology 68: 1606-1615.

Stoddart, L. a., and D. I. Rasmussen. 1945. Deer man-
agement and range livestock production. Utah Agri-
cultural Experiment Station Circular 121. 17 pp.

Tepedino, V J. 1982. Effects of defoliation on reproduction
of a toxic range plant, Zigademis paniculatiis. Great
Basin Naturalist 42: .524-.528.

Received 12 October 1993
Accepted 30 August 1994

USE OF AN UNUSUAL FOOD SOURCE BY
ROCK WRENS (TROGLODYTIDAE)

PolK K. Piiillipsl and Allen F. Sanhoni^
Key nonls: Rock Wrens, foixl source, forci'^in'^i,. Salpinctes obsoletiis, Tro'Jodiithhte, jecdiiinhchavior

On 12 Jul>' 1993 we obsei-ved an interesting
I'xchange between an achilt Rock Wren
[Salpinctes obsoletiis) and two juveniles. While
at Toroweap Point on the north rim of the
errand Canyon we observed an adult wren
accompanied by two juveniles near our vehi-
cle. All three birds walked beneath the vehicle
by the rear wheel, but the adult moved imme-
diately to the front end whereupon it hopped
onto the front bumper and began to inspect the
grill. The adult fr)und and ate an insect that
had been trapped in the grillwork. While stand-
ing on the bumper, the adult began to vocalize
after consuming the insect. The juveniles
appeared to show a positive phonotactic
response to these calls, stopped foraging under
the rear of the vehicle, and moved to the front.
After the juveniles arrived at the front of the
vehicle, the adult continued collecting insects
from the grill. The adult ate none of these
insects but mereh' held them in its beak while
walking back and forth across the bumper The
adult continued to vocalize, periodically paus-
ing to face the juveniles. Then it continued for-
aging in the grill. It appeared to us that the
adult was showing the insects to the young.
Neither of the young birds joined the adult on
the bumper, however, and within a few min-
utes the adult and juveniles flew off, not to
return that afternoon. As far as we have been
able to determine, this sort of acquired or
derived behavior has not been reported previ-
ously for Rock Wrens nor for any member of
the famiU' Troglodytidae.

Other obsen'crs have noticed birds taking
advantage of unusual food sources, such as the
opening of milk bottles (Fisher and Hinde 1949).
There is generally a question, however, as to
whether the behavior was bv chance or learned.

One possible explanation for our obsei'vations
is that the parent was tutoring the offspring
about the availability of food in xehicle grill-
work. Tutoring and obsei^vational learning have
been documented in laboratory experiments in
blackbirds (Mason et al. 1984), tits (Sheriy and
Galef 1984, 1990), and pigeons (Palameta and
Lefebvre 1985), and have also been document-
ed in the wild in other birds using usual food
sources (Schaadt and Rymon 1982). We have
no way of knowing whether the adult we
observed was attempting to teach what we
believe were its offspring about an unusual
food source. We hope this observation will
stimulate further study of feeding in fledgling
birds with the possibility of discoveries in
social learning.

Literature Cited

Fisher, J., and R. A. Hindi:. 1949. The opening of milk
bottles by birds. British Birds 42; 347-357.

Mason, J. R., A. H. Arzt, and R. F. Reidinger. 1984.
Comparative assessment of food preferences and
aversions acquired by blackbirds via observational
learning. Auk 101: 796-803.

Palameta, B., and L. Lefebvre. 1985. The social trans-
mission of a food finding technique in pigeons: What
is learned? Animal Behaviour 33: 892-896.

Schaadt, C. E, and L. M. Rymon. 1982. Innate fishing
behavior of ospreys. Raptor Research 16: 61-62.

Sherry, D. F, and B. G. Galef, Jr. 1984. Cultural trans-
mission without imitation: milk bottle opening by
birds. Animal Behaviour 32: 937-938.

. 1990. Social learning without imitation: more about

milk bottle opening by birds. Animal Behaviour 40:
987-990.

Received 13 June 1994
Accepted 16 November 1994

'Biology Dc|)aitiMciit, Miami-Dack- Coii
2Sl'Ii()()I of Natural and Hcaltli Scienc-ef

inimity CIol
, Barr\' LInl

c-nc North Campus, 11:380 N.W. 27th Avenue, Miami, FL .331(17-349.5.

.(•rsity, 11.3()() N E, Second Avenue, Miami Shores, FL 33Kil-669.'5. Address reprint requests to this

192

INFORMATION FOR AUTHORS

The Great Basin Naturalist welcomes previously
unpublished manuscripts pertaining to the biologi-
cal natural history of western North America.
Preference will be given to concise manuscripts of
up to 12,000 words. Simple species lists are dis-
couraged.

SUBMIT MANUSCRIPTS to Richard W. Baumann,
Editor, Great Basin Naturalist, 290 MLBM, PO Box
20200, Brigham Young University, Provo, UT
84602-0200. A cover letter accompanying the man-
uscript must include phone number(s) of the author
submitting the manuscript, and FAX number and
E-mail address when applicable; the letter must
also provide information describing the extent to
which data, text, or illustrations have been used in
other papers or books that are published, in press,
submitted, or soon to be submitted elsewhere.
Authors should adhere to the following guidelines;
manuscripts not so prepared may be returned for
revision.

MANUSCRIPT PREPARATION. Consuk Vol. 51, No.
2 of this journal for specific instructions on style
and format. These instructions, GUIDELINES for
Manuscripts Submitted to the Great Basin
Naturalist, supply greater detail than those pre-
sented here in condensed form. The guidelines are
printed at the back of the issue; additional copies
are available from the editor upon request. Also,
check the most recent issue of the Great Basin
Naturalist for changes and refer to the CBE Stijle
Manual, 5th edition (Council of Biology Editors,
One Illinois Center, Suite 200, 111 East Wacker
Drive, Chicago, IL 60601-4298 USA, phonE: (312)
616-0800, FAX: (312) 616-0226; $24).

TYPE AND DOUBLE SPACE all materials, including
literature cited, table headings, and figure legends.
Avoid hyphenated words at the right-hand margins.
Underline words to be printed in italics. Use stan-
dard bond (22x28 cm), leaving 2.5-cm margins on
all sides.

ABSTRACT states the purpose, methods, results,
and conclusions of the research. It is followed by
-6-12 key words, listed in order of decreasing
importance, to be used for indexing.

TEXT has centered main headings printed in all
capital letters; second-level headings are centered
in upper- and lowercase letters; third-level head-
ings begin paragraphs.

REFERENCES IN THE TEXT are cited by author and
date: e.g., Martin (1989) or (Martin 1989). Multiple
citations should be separated by commas and listed
in chronological order. Use "et al. " after name of
first author for citations having more than two
authors.

ACKNOWLEDGMENTS, under a centered main
heading, include special publication numbers when
appropriate.

LITERATURE CITED, also under a centered main
heading, lists references alphabetically in the fol-
lowing formats:

Mack, G. D., and L. D. Flake. 1980. Habitat rela-
tionships of waterfowl broods on South Dakota
stock ponds. Journal of Wildlife Management
44: 695-700.

Sousa, W. P 1985. Disturbance and patch dynamics
on rocky intertidal shores. Pages 101-124 in
S. T A. Pickett and P S. White, eds., The ecolo-
gy of natural disturbance and patch dynamics.
Academic Press, New York.

Coulson, R. N., and J. A. Witter. 1984. Forest ento-
mology: ecology and management. John Wiley
and Sons, Inc., New York. 669 pp.

TABLES are double spaced on separate sheets and
designed to fit the width of either a single column
or a page. Use lowercase letters to indicate foot-
notes.

NOTES. If the manuscript would be more appro-
priate as a short communication or note, follow the
above instructions but do not include an abstract.

FINAL CHECK:

• Cover letter explaining any duplication of
information and providing phone number(s),
FAX number, and E-mail address

• 3 copies of the manuscript and WordPerfect
diskette

• Conformity with instructions

• Photocopies of illustrations

(ISSN 001 7-361 4)

GREAT BASIN NATURALIST voi 55 no 2 ap ii 1995

CONTENTS

Articles

Diets ot >'oiing Colorado sqiiavvfish and otlier small fish in backwaters of the Green

River, Colorado and Utah Robert T. Mnth and Darrel E. Snyder 95

Invertebrate fauna of wastewater ponds in southeastern Idaho

Karen L. Cieminski and Lester D. Flake 1 05

Growth and reproduction in an alpine cushion plant: Astragalus kentrophyta var

implexus Wayne R. Owen 117

Calileuctra, a new genus, and two new species of stoneflies from California

(Plecoptera: Leuctridae) W. D. Shepard and R. W. Baumann 1 24

Carbon isotope discrimination in the C4 shrub Atriplex confertifolia along a

salinity gradient Darren R. Sandquist and James R. Ehleringer 1 35

Demography of Astragalus scaphoides and effects of herbivory on population

growth Peter Lesica 1 42

Lahontan sagebrush {Artemisia arhuscula ssp. longicaulis): a new taxon

Alma H. Winward and E. Durant McArthur 151

Douglas-fir tussock moth {Orgyia pseudotsugata McDunnough) on subalpine fir

in northern Utah E. Matthew Hansen 1 58

Seasonal nutrient cycling in Potamogeton pectinatus of the lower Provo River . .

C. Mel Lytle and Bruce N. Smith 1 64

Factors influencing fish assemblages of a high-elevation desert stream system in

Wyoming Bernard Carter and Wayne A. Hubert 1 69

Notes

Speciation by aneuploidy and polyploidy in Mimulus (Scrophulariaceae)

Robert K. Vickery, Jr. 1 74

Speciation in Mimulus, or, Can a simple flower color mutant lead to species

divergence? Robert K. Vickery, Jr 177

Fall lamb production by a California bighorn sheep

Matthew McCoy, Walt Bodie, and ElRoy Taylor 181

Age, growth, and reproduction of leatherside chub {Gila copei)

Jerald B. Johnson, Mark C. Belk, and Dennis K. Shiozawa 1 83

Consumption of a toxic plant {Zigadenus panicidatus) by mule deer

William S. Longland and Charlie Clements 1 88

Use of an unusual food source by Rock Wrens (Troglodytidae)

Polly K. Phillips and Allen F Sanborn 1 92

6»r>^

H E

GREAT BASIN

NATURALIST

VOLUME 55 NO 3 — JULY 1995

BRIGHAM YOUNG UNIVERSITY

GREAT BASIN NATURALIST

Editor Assistant Editor

Richard W. Baumann Nathan M. Smith

290 MLBM 190 MLBM

PO Box 20200 PO Box 26879

Brigham Young University Brigham Young University

Provo, UT 84602-0200 Provo, UT 84602-6879

801-378-5053 801-378-6688

FAX 801-378-3733 E-mail: NMS@HBLL1.BYU.EDU

Associate Editors

Michael A. Bowers Paul C. Marsh

Blandy Experimental Farm, University of Center for Environmental Studies, Arizona

Virginia, Box 175, Bojce, VA 22620 State University, Tempe, AZ 85287

J. R. Callahan Stanley D. Smith

Museum of Southwestern Biology, University of Department of Biology

New Mexico, Albuquerque, NM University of Nevada- Las Vegas

Mailing address: Box 3140, Hemet, CA 92546 Las Vegas, NV 89154-4004

Jeffrey J. Johansen Paul T. Tueller

Department of Biologv; John CanoU University Department of Environmental Resource Sciences

University Heights, OH 44118 University of Nevada-Reno, 1000 Valley Road

Boris C. Kondratieff R^"«' ^V 89512

Department of Entomology, Colorado State Rorert C. Whitmore

University', Fort Collins, CO 80523 Division of Forestry, Box 6125, West Virginia

University', Morgantown, WV 26506-6125

Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoology;
William Hess, Botany and Range Science. All are at Brigham Young University. Ex Officio Editorial Board
members include Steven L. Taylor, College of Biologv' and Agriculture; H. Duane Smith, Director, Monte L.
Bean Life Science Museum; Richard W Baumann, Editor, Great Basin Naturalist.

The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University.
Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas
in western North America are accepted for publication.

Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1995 are $25 for individual sub-
scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back
issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi-
ness should be directed to the Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young
University Provo, UT 84602-0200.

Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin
Naturalist through a continuing exchange of scholarlv publications should contact the Exchange Librarian,
6385 HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889.

Editorial Production Staff

JoAnne Abel Technical Editor

Jan Spencer Assistant to the Editor

Official publication date: 21 July 1995 7-95 750 15072

The Great Basin Naturalist

Published at Provo, Utah, by
Brigham Young University

ISSN 0017-3614

VoLU M E 55 3 1 J u LY 1 995 N o. 3

BENTHIC COMMUNITY STRUCTURE IN TWO ADJACENT

STREAMS IN YELLOWSTONE NATIONAL PARK FIVE YEARS

AFTER THE 1988 WILDFIRES

G. Wayne MinshalU, Christopher T. Robinson^ Todd V Royerl, and Samuel R. Rushforth^

Abstract. — Ph\sical characteristics, benthic macroin\'ertebrates, and periph>'ton assemblages in two adjacent head-
water streams in Yellowstone National Park were evaluated five years after the 19SS wildfires. The catchment of one
stream was burned by wildfire (burned stream) while the other catchment was unburned (unbumed stream). Physical
measures revealed channel alteration in the burned stream relative to the unbin-ned stream. Periphyton biomass was
lower in the burned than the unburned stream (29.2 vs. 50.5 g/m^ AFDM, respectively), further demonstrating the
unstable physical conditions of that system. Kendall's coefficient of concordance (an index of similarity) between diatom
assemblages was 0.22, indicating distinct assemblage composition between streams. Navicida pennitis Hust. was the
most abundant diatom in the burned stream while Hannaea arcus (Ehr.) Patr was dominant in the unbumed stream.
Macroinvertebrate taxa richness, density, and biomass were all greater in the unburned stream, although Chironomidae
was the most abundant taxon in both streams. Results suggest the removal of ten-estrial/riparian vegetation by wildfire
can directly influence stream benthic assemblages by altering the inherent disturbance regime of the physical habitat
templet.

Key words: wildfire, streams, disturbance, inacroinvertebrates, diatoms, benthic habitat, Yellowstone National Park.

Physical disturbance, acting at various spa- and Minshall 1992, Robinson et al. 1994, Mihuc
tial and temporal scales, often is the predomi- et al. in press, Robinson and Minshall in press).
nant factor structuring stream benthic com- In lotic ecosystems, physical disturbance also
munities (Minshall 1988, Resh et al. 1988). may constrain the estabhshment of biotic con-
Further, physical disturbances may be viewed trols, such as competition and predation, on
in a hierarchical framework, with the effects of benthic community structure (McAuliffe 1984,
small-scale disturbances altered (intensified or Minshall and Petersen 1985, Resh et al. 1988).
mediated) by large-scale disturbance events Wildfire burned extensive portions of the
(sensu O'Neill et al. 1986). Wildfire, as a large- Greater Yellowstone Ecosystem during the
scale disturbance, directly influences stream summer of 1988. Over 32% of the streams in
biotic structure and function by affecting the Yellowstone National Park (YNP) were affected
physical habitat of stream ecosystems (Minshall to varying degrees by wildfires (Minshall et al.
et al. 1989, Minshall and Brock 1991, Richards 1989, Minshall and Brock 1991, Robinson et al.

'Streuni Ecolog\ Center. Department of Biological Sciences, Idaho State Uni\ersit>'. Pocatello, ID 8.3209.
^Department of Botan\' and Range Science, Brigham Young University. Pro\o, UT 84602.

193

194

Great Basin Naturalist

[Xbliinie 55

1994). Minshall and Brock (1991) sumniarized
the immediate effects of the fires on YNP's
stream ecosystems and h\'pothesized on the
mid-term (10-25 yr) and long-term (50-300 yr)
effects. They suggested that most adverse
short-term effects on streams resulted from
increased sediment load and channel erosion
caused by increased overland runoff following
precipitation events and snowmelt. The in-
tensity and frequency of short-term effects
were hxpothesized to decrease by year 5 as
riparian conditions improve (see Richards and
Minshall 1992). In general, mid- and long-
term effects on streams, including recovery to
prefire conditions, should coirespond to vegeta-
tive regrowth in burned catchments (Minshall
at al. 1989, Minshall and Brock 1991).

The present study compared physical char-
acteristics and benthic community structure in
two streams five years after the 1988 wildfires.
The streams are adjacent second-order (after
Strahler 1952) tributaries of the South Fork
Cache Creek. The catchment of one stream
was bunied during the 1988 wildfires, while the
catchment of the other was essentially un-
burned. The spatial arrangement of these
streams (adjacent basins) provided a treatment/
reference situation where confounding factors
of climate and geology are minimized when
comparing differences among the study
streams. However, the study lacks true repli-
cation of the burned and unburned treatments
(sensu Hurlbert 1984) and must be viewed as
a simple comparison study. Phenomenological
studies and/or two stream comparisons are
common in stream ecology (e.g., Wallace et al.
1986, Robinson et al. 1993, Scarsbrook and
Townsend 1993) and are capable of providing
valuable insights (Townsend 1989). The present
study can be viewed as a natural "experiment"
with observed differences between the two
streams attributed to the effects of wildfire. In
that context, the study provides insights on
general patterns of lotic ecosystem recoveiy to
an unpredictable, large-scale disturbance
(Townsend 1989, Lamberti et al. 1991).

Methods

The study streams, located in the northeast
comer of YNI^ were surveyed on 19 July 1993.
One stream had over 80% of its catchment
burned during the 1988 Yellowstone wildfire
(hereafter, burned stream; 110°0r30"W,

44°50'00"N), while less than 10% of the catch-
ment of the other stream was bunied (hereafter,
unburned stream; 110°01'00"\V, 44°49'30"N).
Climate of the area is typical of the northeiTi
Rocky Mountains, with precipitation primarily
occurring as snow during the winter months.
Both streams drain catchments primariK' veg-
etated (prior to the fire in the buiTied stream)
by coniferous forests of lodgepole pine {Pinus
conturta) and Engelmann spruce {Picea engel-
mannii). Riparian vegetation consisted of wil-
low {Salix), rose {Rosa), and alder {Alniis).

Sui^veys were conducted approximately 0.5
km above the confluence of the two streams.
Physical characteristics were measured in each
stream at five cross-sectional transects, each sit-
uated approximately 50 m apart. Measurements
made at each transect included stream width
at baseflow, stream width at bankfull discharge,
and stream cross-sectional profile (for calcula-
tion of width:depth ratios). Discharge was cal-
culated in each stream at the most suitable
transect following the methods of Platts et al.
(1983). In addition to measurements at each
transect, 100 randomly selected rock substrata
along a 100-m length of stream (located within
the outermost cross-sectional transects) were
measured for size (length of the longest axis)
and percent embeddedness. Embeddedness
was defined as the percent coverage of the
rock (three-dimensional surface) by fine sedi-
ments. Large boulders that protruded through
the water surface were not used in substratum
size measurements. Water depth and near-bed
water velocity also were recorded at each of
the 100 random locations. Near-bed water
velocity was measured with a small Ott C-1
current meter approximately 2 cm above each
substratum.

One periphyton sample was collected from
a suitable (flat-surfaced, medium-sized) rock
substratum at each cross-sectional transect
using a method described in Robinson and
Minshall (1986). Samples were frozen in the
field in a Taylor-Wharton 3DS dry shipper
charged with liquid nitrogen and returned to
the laboratory for processing. In the laboratoiy,
samples were extracted in 10 ml of methanol
for 24 h (Holm-Hansen and Riemann 1978).
One 3-ml subsample was then removed from
each sample and analyzed for chlorophyll a
using a Gilford Instruments (Model 2600) spec-
trophotometer. The remaining periphyton
material from each sample was used for algal

1995]

Wildfire and Benthic Communities

195

biomass determination, expressed as grams
ash-free dry mass (AFDM) per m^. The mater-
ial was dried at 50 °C for 24 h, weighed on a
Sauter balance (Model AR 1014), ashed at
550 °C for a minimum of 3 h, rehydrated,
redried at 50 °C, then cooled to ambient tem-
perature in a desiccator and reweighed. The
difference in weights equaled the AFDM of
the sample.

Diatom samples were collected in each
stream, after Robinson and Rushforth (1987),
from three to five rock substrata representing
the predominant habitat type (typically riffles).
Samples were composited, preserved with 5%
formalin, and returned to the laboratory. The
composite sample was boiled in concentrated
nitric acid, rinsed, mounted in Naphrax moun-
tant, and examined under lOOOX oil immersion
using a Zeis RA microscope with Nomarski
optics (St. Clair and Rushforth 1976). Relative
abundances of diatom taxa were determined
by counting a minimum of 1000 diatom valves
from each stream. Diatoms were analyzed in
terms of species richness, Simpson's index,
and Kendall's coefficient of concordance (an
index of similarity using all taxa with a relative
abundance >1%). Other algal groups such as
Chlorophyta (green algae) and Cyanobacteria
(blue-green algae) were not abundant at the
time of sampling and thus were not consid-
ered in the present study.

One benthic sample was collected from a
riffle/run habitat (pools were rare and not sam-
pled) near each transect and analyzed for
macroinvertebrates and benthic organic mat-
ter (BOM). Samples were collected using a
Surber sampler (250 /xm mesh), preserved
with 5% formalin, and returned to the labora-
tory. Woody debris >5 cm in length that was
collected in the benthic samples was rinsed of
invertebrates and removed from the samples.
In the laboratory, macroinvertebrates were
hand-sorted from the benthic detritus with the
aid of a 3X dissecting microscope, identified to
the lowest feasible level (usually genus), enu-
merated, dried at 50°C for a minimum of 48 h,
then cooled to ambient temperature in a des-
iccator and weighed. Dry weights, in milli-
grams, were determined on a Cahn (Model
25) electrobalance. The benthic detritus from
each sample was used for BOM determination.
The quantity of BOM, expressed as g AFDM/
m^, was determined as described above for
periphyton. Macroinvertebrates were analyzed

in terms of density (no./m-), biomass (mg/m^),
taxa richness, Simpson's index, and relative
abundances.

Chi-squared analysis was used to test for
statistical differences in median substratum
size between the two streams (Zar 1984).
Independent sample t tests were used to com-
pare the other characteristics for differences
between the two streams. Prior to the ^-test
analysis all data were log (x -I- 1) transformed,
except substratum embeddedness and the rela-
tive abundance of invertebrate taxa (both per-
centage measures), which were arcsine (square
root [x]) transformed (Zar 1984). Tabular re-
sults are presented as untransformed means
and standard deviations. All statistical analyses
were performed on SYSTAT (Wilkinson 1990).

Results

Baseflow discharge was equal in the two
streams (0.2 m'^/s), reflecting the similar catch-
ment size of the burned (22 km^) and unbumed
(26 km^) streams. Mean baseflow width, near-
bed water velocity, and BOM were not signifi-
cantly different between the two streams [P >
.05). Substratum embeddedness was signifi-
cantly greater in the burned stream (P = .01),
although the difference between mean values
was not large (burned = 62.9, unburned =
52.8). It is not known whether this statistical
difference was biologically meaningful or sim-
ply a reflection of the large sample size {n =
100).

Water depth at baseflow (F < .01) was lower
and stream width at bankfull discharge greater
(P = .03) in the burned stream than the un-
burned stream. Although not statistically sig-
nificant (P = .06), the ratio of stream width:
depth was greater in the burned than the un-
burned stream (216 and 91, respectively). The
general appearance of the two streams was dis-
tinctly different (Fig. 1; Minshall personal ob-
servation). Large, woody debris and streamside
riparian vegetation, which provide bank and
channel stability, were noticeably absent in the
burned stream.

Mean substratum size was not significantly
different between the two streams (P > .05) in
1993, possibly because large boulders were
not recorded in the measurements (see Fig. 1).
We collected additional data on substratum size
in August 1994 and included large boulders in
the measurements. Further, substrata within

196

Ghkat Basin Naturalist

[Volume 55

^-f.

_"**
■j^ '

Fig. I. HepresLMitatixe photographs of the huined (upper) and iinl)urned (lower) streams Fixe >ears after the 1988
wildfire. Note absence of large, woody debris and streaniside riparian xegetation in the l)inned stream.

1995]

Wildfire and Benthic Communities

197

the bankfull channel were measured in 1994,
in contrast to measures being recorded only
within the baseflow channel in 1993. The 1994
results showed that mean substratum size was
significantly larger in the unburned than in
the burned stream (F < .01).

A comparison of median substratum size
showed similar results to that of mean substra-
tum size. Median substratum size was not dif-
ferent between the two streams when mea-
surements excluded large boulders and were
confined to the baseflow channel (P > .05).
However, when measurements included large
boulders and encompassed the bankfull chan-
nel, the difference in median size was signifi-
cant (F < .01). Whether large boulders were
present in the burned stream prior to the wild-
fire has yet to be detemiined. However, in other
streams influenced by intensive wildfire, large
boulders were obsei-ved to be buried by inor-
ganic debris (primarily gravel and fine sedi-
ments) within five years following wildfire
(Minshall personal obsei'vation).

The burned stream contained less periphy-
ton chlorophyll a (F = .06) and AFDM (F <
.01) than did the unburned stream (Table 1).
Diatom species richness was greater in the
burned (34 taxa) than in the unburned stream
(27 taxa; Table 2). Simpson's index was lower for
the burned than the unburned stream (0.12
and 0.42, respectively). Kendall's coefficient of
concordance for the two diatom communities
was 0.22, suggesting distinct assemblage com-
position among sites. For example, Navicida
pennitis Hust. was the most abundant species
in the burned stream, constituting 24.7% of

the assemblage, while Hannaea orciis (Ehr.)
Patr. comprised 63.1% of the assemblage in
the unburned stream (Table 2).

Mean macroinvertebrate density and bio-
mass were lower in the burned than unburned
stream (Table 3), but the differences were not
significant (F > .05). For example, mean den-
sity' in the burned stream was 9960 individuals/
m^, while the unburned stream had 16,950 indi-
viduals/m^, and mean biomass (dry weight)
was 1960 and 3200 mg/m- in the burned and
unburned streams, respectively. Taxa richness
and Simpson's index both were reduced in the
burned stream, although the difference was
significant only for Simpson's index (F = .04)
(Table 3). The burned stream contained a
mean of 15 taxa per benthic sample compared
to a mean of 20 taxa for the unburned stream.
The mean Simpson's index was 0.57 for the
burned stream and 0.73 for the unburned
stream. Chironomidae was the most abundant
taxon in both streams (Table 4), although their
relative abundance was significantly greater (P
= .03) in the unburned stream. There were no
statistical differences (F > .05) in relative
abundances of other taxa common to both
streams (Hydracarina, Simuliidae, Baetis bi-
caudatiis, Cinygnnilo, and Zopada columhiana).

Discussion

Alterations of the surrounding terrestrial
landscape by major unpredictable disturbances
such as hurricanes, volcanic eruptions, or wild-
fire directly influence streams draining the

Table 1. Means (SD) and P values for physical characteristics measured in the study streams.

Burned

Unbi

P \'alue

Baseflow width (m)
Near-bed velocity (cm/s)
BOM(g/m2)
Embeddedness (%)
Baseflow depth (cm)
Bankfull width (m)
Bankfull width:depth ratio
Periphyton chl a (mg/m-)
Periphyton AFDM (g/m^)

Mean substratum size (cm)
Mean substratum size (cm)

Median substratum size (cm)
Median substratum size (cm

5.9

(2.6)

10.3

(0.1)

1.5

(0.8)

62.9

(28.5)

16.9

(11.0)

35.0

(11.9)

216

(101)

8.9

(5.5)

29.2

(3.7)

14.4

(10.5)

15.5

(15.5)

12.0

11.0

4.5

(1.8)

.46

10.7

(0.1)

.71

2.5

(1.2)

.20

52.8

(30.0)

.01

24.3

(12.5)

<.01

16.6

(1.4)

.03

(21)

.06

32.1

(19.5)

.06

.50.5

(9.3)

<.01

15.0

(14.2)

.26

27.6

(27.4)

<.01

11.0

.72

17.0

<.01

Baseflow channel, large boulders excluded.
Bankfull channel, large boulders included.

198

Great Basin Natuhaijst

[Volume 55

Table 2. Community measures and relative aliuudanees
(%) for the diatom assemblage of each study stream.

Tahi>K 3. Means (SD) and P values of macroinvertehrate
eomnnmity measures for the study streams.

Burned

Unburned

Species richness

Simpson s index (C)

0.12

0.42

Navictila pcnniti.s I hist.

24.7

5.1

Nitzsch id dissipatii

(Kuetz.) Grun.

17.3

4.2

Achnanthes lanceolata

(Breb.) Grun.

9.8

1.7

Nitzschia paleacea Grun.

7.6

3.1

Navicuhi (iriensis Hust.

4.8

2.4

Hannaca arciis (Ehr.) Patr.

2.1

63.1

Burned

Uiihunic'tl

P Mill

affected watersheds. For example, the Mt. St.
Helens eruption of 1980 dramatically changed
drainage patterns and river networks, elimi-
nated terrestrial vegetation, and caused major
debris flows that scoured stream channels
(Wilzbach et al. 1983, Hawkins 1988). However,
high spatial variation in the intensity of these
major disturbances may occur, causing tempo-
ral differences in recovery patterns (Yount and
Niemi 1990). In catchments of YNP the rela-
tive area burned ranged from <10% to >90%
(Minshall and Brock 1991). Further, the
degree of alteration of stream habitat was
highly correlated with percent of catchment
burned (Robinson and Minshall in press).

In the present study significant differences
were observed in the benthic habitat of the two
streams. The width:depth ratio of the burned
stream was greater than that of the unburned
stream. Anderson (1992) also observed in-
creased widthidepth ratios following major
disturbances in streams of the Cascade Moun-
tains. With large boulders included in the mea-
surements, the unburned stream exhibited sig-
nificantly greater substratum size. Gurtz and
Wallace (1984) demonstrated that large sub-
strata could mediate the effects of large-scale
disturbances by providing stable habitat for
benthic organisms. At the time of sampling,
the burned stream did not contain the larger-
sized substrata found in the unburned stream.
It is probable that the larger substrata in the
burned stream were buried by inorganic sedi-
ments following the wildfire, as has been
observed in other YNP streams (Minshall per-
sonal observation). Thus, one effect of the wild-
fire appeared to be alteration of the substrata
in such a manner as to make the benthic habitat
more susceptible to future disturbances (e.g.,
Gurtz and Wxllace 1984).

Densit\ (no./ni^) 9963 (47.30) 16,948 (899Si .31

Biomass (mg/m^) 1956(1056) 3198(1274) .19

Ta.xa richness 15.0 (3.7) 20.4 (2.6) ..58

Simpsons indf.xfC) 0..57 (0.09) 0.73(0.10) .04

Lamberti et al. (1991) found that faunal
densities and macroinvertehrate species rich-
ness had recovered within one year following
a major debris flow in an Oregon stream. In
central Idaho, however, streams disturbed by
wildfire and unburned reference streams
showed little similarity in macroinvertehrate
assemblages, even after five years of recovery
(Richards and Minshall 1992). Similarly, in the
present study the influence of wildfire was still
apparent after five years. Macroinvertehrate
community structure was not similar between
the two systems, despite their close proximity
to each other (0.5 km). Most researchers agree
that recovery of the benthic community will
correspond to recovery of the surrounding
landscape (Steinman and Lamberti 1988,
Minshall et al. 1989, Lamberti et al. 1991,
Minshall and Brock 1991, Anderson 1992,
Richards and Minshall 1992, but see Hawkins
1988).

Primary producers (lotic algae) may recover
sooner than consumers (macroinvertebrates and
fish) because of their much shorter life cycles,
and subsequenfl}' they may influence recoveiy
of the higher trophic levels (Steinman and
Mclntire 1990). In the present study, peri-
phyton biomass (as AFDM) in the unburned
stream was 1.7X greater than in the burned
stream, implying a present lack of recoven' by
primary producers in the burned system.
Macroinvertehrate taxa richness also was
greater in the unburned stream than in the
burned stream. How functional or structural
recovery of macroinvertebrates is related to
algal recoveiy following wildfire has yet to be
determined, but provides an interesting and
important avenue for future research. Algae
have shorter life cycles and reduced mobility
relative to macroinvertebrates, and possibly
the two groups respond differently to large-
scale disturbances.

1995]

Wildfire and Benthic Communities

199

Table 4. Mean (SD) density and relative abundance of
the 10 most abundant niacroin\ertebrate taxa from each
stream. These taxa constituted >909'f of the assemblage in
their respective streams.

Taxa

Density

Relative

(no./m-)

abimdance (%)

- - Burned

Chironomidae

5437 (2218)

59.7 1

(19.2)

Simuhidae

2737 (4583)

18.2 1

(24.1)

Baetis bicaitdatits

576 (346)

8.1

(6.3)

Hydracarina

148 (232)

1.5

(2.3)

Epeoriis olbertae

95 (123)

1.0 (1.2)

Zapada coluinbiana

90 (72)

1.0 (0.7)

Epcorits longimaniis

80 (125)

0.5

(0.7)

Ciiujgmula

75 (71)

0.8

(0.7)

Eluthrogena

37 (52)

0.3

(0.3)

Ameletus cooki

22 (15)
- Unbumed

0.3

(0.2)

Chironomidae

14,676 (8289)

84.7

(6.7)

H\dracarina

372 (253)

2.1

(1.6)

Cinygmula

314 (187)

1.7

(0.8)

Zapada cohiiubiana

310 (215)

1.6

(0.6)

Simuliidae

307 (181)

3.8

(5.3)

Baetis bicaitdatus

125 (60)

1.4

(1.7)

Dninella coloradensis

75 (64)

0.4

(0.5)

Wniacophila angelita

73 (55)

0.4

(0.2)

Rliyacoplula tiicida

52 (29)

0.3

(0.1)

Kogotiis

52 (35)

0.3 (0.2)

After five years of recovery, the channel of
the burned stream still appeared unstable as
indicated by different diatom assemblages
between the two streams. For example, the
small, adnate diatom Naviciila pennitis Hust.
was predominant in the burned stream but
was found in relatively low abundance in the
unburned stream. N. pennitis was predomi-
nant in other YNP streams influenced by the
1988 wildfires, and it has been suggested that
a diatom community with an abundance of N.
pennitis is indicative of more physically dis-
turbed stream environments (Robinson et al.
1994). Further, Robinson et al. (1994) showed
diatom recoveiy among 14 streams in Yellow-
stone was inversely related to degree of dis-
turbance by wildfire. Similarly, Steinman and
Lamberti (1988) found little recovery, after six
years, in the composition of algal communities
in intensively disturbed streams of Mt. St.
Helens. In summary, benthic community re-
covery patterns appeared to be related to the
recovery of stream physical habitat which, five
years after the 1988 wildfires, still displayed
evidence of instability. When examining the
recovery of benthic communities following

large-scale disturbance, one must remain aware
of the connections between the terrestrial land-
scape, lotic habitat, and benthic organisms.

Acknowledgments

We thank Vincent Archer, Michael Bray,
Justin Gill, and especially Scott Relyea for
assistance in the field. Cecily Nelson, Jason
Nelson, Mark Overfield, and Jeffrey
Varricchione assisted in the laboratory.
Suggestions from Dr. Richard Hauer and two
anonymous reviewers greatly improved the
manuscript. The research was partially sup-
ported by Grant No. 725 from the Faculty
Research Committee, Idaho State University.
Additional funding was provided through
Yellowstone Ecosystem Studies (Dr. Robert
Crabtree, Director) and the Department of
Botany and Range Sciences, Brigham Young
University.

Literature Cited

Anderson, N. H. 1992. Influence of disturbance on insect
communities in Pacific Northwest streams. Hydro-
biologia 248: 79-92.

GuRTZ, M. E., AND J. B. Wallace. 1984. Substrate-medi-
ated response of stream invertebrates to distm-bance.
Ecologv' 65: 1556-1569.

Hawkins, C. P 1988. Effects of watershed vegetation and
disturbance on invertebrate community structure in
western Cascade streams: implications for stream
ecosystem theoiy Verhandlungen der Internationale
Vereinigung fiir Theoretische und Angewandte
Limnologie 23: 1167-1173.

Holm-Hansen, O., and B. Riemann. 1973. Chlorophyll a
determination: improvements in methodology. Oikos
30: 438-447.

HURLBERT, S. H. 1984. Pseudoreplication and the design
of ecological field experiments. Ecological Mono-
graphs 54: 187-211.

La.mberti, G. a., S. V Gregory, L. R. Ashkenas, R. C.
Wildman, and K. M. S. Moore. 1991. Stream eco-
system recoveiy following a catastrophic debris flow.
Canadian Journal of Fisheries and Aquatic Sciences
48: 196-208.

McAuLlFFE, J. R. 1984. Competition for space, distur-
bance, and the stnicture of a benthic stream commu-
nity. Ecolog>' 65: 894-908.

MlHUC, T., G. W. iviiNSHALL, AND C. T. RoBlNSON. In press.
Responses of benthic macroinvertebrate populations
in Cache Creek Yellowstone National Park to the
1988 wildfire. In: D. G. Despain and P Schullery,
editors. The ecological implications of fire in Greater
Yellowstone: 2nd biennial conference on the Greater
Yellowstone Ecosystem. National Park Service,
Yellowstone National Park.

MiNSHALL, G. W. 1988. Stream ecosystem theoiy: a global
perspective. Journal of the North American Bentho-
logical Societv' 7: 263-288.

200

Great Basin Natur.\list

[Volume 55

MlNSHALL, G. VV., AND J. T. BROt:K. 1991. Ohscrvfd and
anticipated effects of forest fire on Yellowstone
stream ecosystems. Pages 123-135 in R. B. Keitcr
and M. S. Boyce, editors. The Greater Yellowstone
Ecosystem: redefining Americas wilderness heritage.
Yale University Press, New Haven, CT.

MlNSHALL, G. W, AND R. C. Peteksen. 1985. Towards a
theory of macroinvertebrate community structure in
stream ecosystems. Archives fiir Hxdrohiologia 104;
49-76.

MlNSHALL, G. W., J. T Brock, and J. D. Vahley. 1989. Wild-
fire and Yellowstone's stream ecosystems. BioScience
39; 707-715.

O'Neill, R. V, D. L. DeAngelis, J. B. Waide, .\nd T. H. E
Allen. 1986. A hierarchical concept of ecosystems.
Princeton University Press, Princeton, NJ. 257 pp.

I'lATis. W. S., W. E Megahan, and G. W. Minshall. 1983.
Methods for evaluating stream, riparian, and hiotic
conditions. USDA Intermountain Forest and Range
E.\periment Station, Ogden, UT. General Technical
Report I NT- 138. 70 pp.

Resh, V. H., et al. 1988. The role of distiubance in stream
ecology. Journal of the North American Benthological
Societx' 7: 433^55.

Richards, C., and G. W. Mlnshall. 1992. Spatial and
temporal trends in stream macroinvertebrate com-
munities; the influence of catchment disturbance.
HydrobioIogia241; 173-184.

Robinson, C. T, and G. W. Minshall. 1986. Effects of
disturbance fi^equency on stream benthic communi-
ty structure in relation to canopy cover and season.
Journal of the North American Benthological Society
5: 237-248.

_. In press. Physical and chemical responses of

streams in Yellowstone following the 1988 wildfire.
In D. G. Despain and P Schullery, editors. The eco-
logical implications of fire in Greater Yellowstone;
2nd biennial conference on the Greater Yellowstone
Ecosystem. National Park Service, Yellowstone
National Park.

R0BIN.SON, C. T, AND S. R. Rlshforth. 1987. Effects of
physical disturbance and canopy cover on attached
diatom community structure in an Idaho stream.
Hydrobiologia 154;' 49-59.

Robinson, G. T, G. W. Minshall, and L. Van Every. 1993.
Seasonal trends and colonization patterns of macro-
invertebrate assemblages in two streams with con-

trasting How regimes. Great Basin Naturalist 53;
321-331.

Robinson, G. T, S. R. iirsiiEOHTii, and G. VV. Minshall.
1994. Diatom assemblages of streams influenced by
wildfire. Journal of Phycology 30; 209-216.

ScARSBROOK, M. R., AND G. L. TowNSEND. 1993. Stream
community structure in relation to spatial and tem-
poral variation; a habitat templet study of two con-
trasting New Zealand streams. Freshwater Biology
29; 395-410.

St. Clair, L. L., and S. R. Rushforth. 1976. The diatom
flora of the Goshen Warm Springs Ponds and Wet
Meadows, Goshen, Utah, U.S.A. Nova Hedwigia 24;
353-125.

Steinman, a. D., and G. A. Lamberti. 1988. Lotic algal
communities in the Mt. St. Helens region six years
following the eruption. Journal of Phycology 24;
482-489.

Steinman, A. D., and G. D. McIntire. 1990. Recovery of
lotic periphyton communities. Environmental
Management 14; 589-604.

Strahler, a. N. 1952. Hypsometric (area-altitude) analy-
sis of erosional topography. Geological Society of
America Bulletin 63; 11 17-1 142.

Townsend, C. R. 1989. The patch dynamics concept of
stream community ecology. Journal of the North
American Benthological Society 8; 36-50.

Wall.\ce, J. B., D. S. Vogel, and't. E Cuffney. 1986.
Recovery of a headwater stream fi-om an insecticide-
induced community disturbance. Journal of the North
American Benthological Society 5; 115-126.

Wilkinson, L. 1990. SYSTAT; the system for statistics.
SYSTAT, Inc., Evanston, IL. 677 pp.

Wilzbach, M. a., T. H. Dudley, and J. D. Hall. 1983.
Recovery patterns in stream communities impacted
by the Mt. St. Helens eruption. Water Resources
Research Institute, Oregon State University, Gorvallis.
Report No. WRRI-83. 33 pp.

YOUNT, J. D., and G. J. NiEMl. 1990. Recovery of lotic
communities and ecosystems from disturbance — a
narrative review of case studies. Environmental
Management 14; 547-569.

Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall Inc.,
Englewood Cliffs, NJ. 718 pp.

Received 16 November 1994
Accepted 2 March 1995

EFFECTS OF BROWSING BY NATIVE UNGULATES ON THE SHRUBS IN BIG
SAGEBRUSH COMMUNITIES IN YELLOWSTONE NATIONAL PARK

Francis J. Singer^ and Roy A. Renkin^

Abstract. — The effects of elk {Cerviis elaphus), pronghorn (Antilocapra americuna), and mule deer (Odocoileus
liemionus) browsing on shi-ubs in big sagebrush {Artemisia tridentata) communities were monitored over a 31-year peri-
od in Yellowstone National Park. Ungulates were restricting Wyoming big sagebrush (spp. wyomingensis) heights, size,
and recruitment on the lower-elevation stratum only, while no such suppression was observed on the high-elevation
stratum. Parallel increases in mountain big sagebmsh (spp. vaseijana) densities and cover occuired over the study period
on both browsed and unbrowsed sites at the higher-elevation stratum, although big sagebrush, green rabbitbrush
(Chnjsothamnus viscidiflorus), and horsebnish (Tetradymia canescens) were slightly taller and crown sizes were slightlv
larger on unbrowsed than browsed sites. Wyoming big sagebrush utilization (percent leader use) was eight times higher
(x = 87 ± 7.2% by pronghorns, mule deer, and elk) on the low-elevation winter range stratum (the Boundary Line Area
[BLA] portion of the winter range), while mostly mountain big sagebrush with leader use averaged only 11 ± 4.1% (near-
ly all by elk) on the high-elevation range stratum. In addition, annual aboveground biomass production of big sagebrush
did not differ between browsed and unbrowsed study sites on the high-elevation stratum of the winter range. Population
turnover was higher on browsed big sagebrush at the high-elevation plots; seedling germination and survival rates were
higher on browsed plots versus unbrowsed plots. No difference was observed in percent dieback of big sagebmsh adult
plants between browsed and unbrowsed plots at the higher stratum. Browsing did not influence the number of leaves or
seedstalks per plant (P > .05), but leaves averaged 45% longer and seedstalks 42% longer on browsed big sagebmsh.
Ungulate browsing, however, apparently suppressed production, germination, and sur\'ival of Wyoming big sagebmsh
on the low-elevation stratum. Numbers of Wyoming big sagebmsh declined 43% and cover declined 29%, 1957-1990,
on browsed sites on the BLA. Annual biomass production on browsed sites at the low-elevation stratum was only 6-35%
that of unbrowsed sites, and big sagebrush recruitment was less on browsed sites. Percent leader use of big sagebrush
did not differ between the period of ungulate reductions, 1962-1969, and the 19S0s on the lower stratum (.v = 87%
leader use), but utilization was less on higher portions of the winter range during the period of elk reductions (x = 2%)
than during the 1980s following cessation of elk controls (.v = 11%).

Key words: big sagebrush browsing, noiihcrn Yellowstone elk, pronghorn. mule deer Cen'us elaphus.

Native populations of elk {Cervus elaphus), models suggest 8-15% fewer elk and 10-25%

bison {Bison bison), and pronghorn {Antilocapra fewer bison would occupy the system if wolves

americana) were artificially reduced in Yellow- were recovered (Carton et al. 1990, Boyce 1993,

stone National Park (YNP), particularly from Mack and Singer 1993).

1942 through 1967 (Meagher 1973, Houston Early workers expressed concern about
1982). Reductions were terminated in 1967 apparent overbrowsing and declines in big sage-
when an experimental management program brush {Artemisia tridentata) due to possible
of natural regulation was initiated (Cole 1971, overabundant populations of elk and prong-
Houston 1976, 1982). Elk and other ungulate horn. As early as the 1930s, Rush (1932) and
numbers tripled after cessation of controls, and Cahalane (1943) reported losses of big sage-
concerns were expressed over high ungulate brush over lower-elevation areas of the north-
densities (Chase 1986, Kay 1991). Appropriate ern winter range. Rush (1932) reported that
numbers of ungulates for the park are unknown less-palatable rabbitbrushes {Chnjsothamnus
since no similar control area exists where wolves spp.) were increasing. KJttams (1950) concluded
{Canis hipus) are present and where ungulate that big sagebrush numbers were declining at
migrations are completely unrestricted by both lower and higher elevations of the north-
humans (Cayot et al. 1979, Peek 1980). Ungulate em winter range. He felt that physical distur-
densities are likely slightly above natural con- bances of big sagebrush by elk during cold
ditions, in that three independent computer periods (shattering and trampling) and an

^Division of Research, Box 168, Mammoth Hot Springs, Yellowstone National Park, WT 82190, and Colorado State Universit\; Fort Colhns, CO 80523.
^Resources Management Division, Box 168, Yellowstone National Park, WT 82190.

201

202

Great Basin Naturalist

[Volume 55

absence of l)ig sagebrush reprocliielioii con-
tributed to the dechne. Dechnes in big sage-
brush at the lower-elevation Boundary Line
Area (BLA) were attributed by Kittanis (1950)
to excessive lexcls of browsing b\' pronghorn.
Park management established a goal to reduce
the pronghorn herd by 50% (Kittams 1959); by
1969 pronghorn numbers were artificially re-
duced from 600-800 to less than 200 dn-ough
a combination of artificial reductions and
severe winters (Barmore 1980).

Houston (1982) provided alternative inter-
pretations concerning big sagebrush. He re-
ported increases in big sageliiaish numbers over
all the northern winter range except the BLA
near Gardiner, MX where numbers declined.
Houston (1982) compared photos taken during
the 1860s to photos retaken in the 1970s. He
attributed the increase in big sagebrush at
higher elevations to fire suppression and the
decline in big sagebrush in the BLA to a
return to more natural conditions following
the removal of intense grazing by livestock in
the early 1930s when the area was added to
the park. In 1986 the U.S. Congress directed
the National Park Sei^vice (NPS) to conduct a
study to evaluate whether native ungulates
were overgrazing the northern winter range
(Congressional Record 1986).

Our objectives were to document trends in
big sagebrush abundance on a series of per-
manently marked plots from 1958 to 1990.
Height, canopy size, twig lengths, and annual
production of shrubs were compared between
browsed and unbrowsed sites.

Study Area

Shrub sampling was conducted on un-
browsed (exclosed) and paired browsed sites at
eight ungulate exclosures erected in 1958 and
1962 on Yellowstone's northern winter range.
The eight exclosures, 2 ha in size, were locat-
ed on gently rolling upland steppe ridge and
the intervening swale habitats (Fig. 1). We
divided the study area into a low-elevation
stratum (the BLA of Houston 1982) with two
exclosures, and a much larger, high-elevation
stratum (n = 6 exclosures), based on large dif-
ferences in ungulate species, elevation, snow-
pack, precipitation, and big sagebrush sub-
species (Fig. 1).

Underlying soils are typic calciborolls, aridic
haploborolls, and aridic calciborolls (Lane

1990). Precipitation averages 30 cm at the low-
elevation exclosures and 55 cm at the high-
stratum exclosures (Houston 1982, Despain
1991). The northern winter range is approxi-
mate!)' 1100 km- and is located along the
upper Yellow stone River drainage. Elevations
range from 1500 m at the low-stratum exclo-
sures to about 2200 m at the highest exclosure.
The northern winter range is lower, warmer,
and drier than the remaining higher plateaus
of YNP (Houston 1982). As a result, 80% of the
ungulates in the park during winter are found
on the northern winter range (Singer 1991).

Dominant shrubs at all the study sites
include two subspecies of big sagebrush at the
high-elevation stratum, nearly all mountain big
sagebrush (A. t. vaseijana) with some basin big
sagebnish {Aiicinisia tridentata tridentata), while
Wyoming big sagebrush (A. t. wyotningensis)
occurs in the low-stratum study sites. The big
sagebrush subspecies vary markedly in their
site requirements, growth, and preferences by
ungulates (Beetle 1960, Welch et al. 1981, Beede
and Johnson 1982, McArthur and Welch
1982). Rubber rabbitbrush {Chnjsothamnus
noiiseosiis), green rabbitbrush {Chn/sothcnnuiis
viscidiflorus), and horsebrush {Tetradymia
canescens) are found at all study sites. Grayia
spinosa and Athplex canescens occur at the
low-stratum study sites (Houston 1982).
Dominant grasses are bluebunch wheatgrass
{Pseiidoroegneria spicato). Idaho fescue {Festuca
idahoensis), junegrass {Koeleria pyramidata),
bluegrasses {Poa compresses P- sandbergii, or P.
pratensis), and thick-spike wheatgrass (A.
dasystachyum; Houston 1982, Wambolt et al.
1987, Despain 1991).

Pronghorn and mule deer occupy only the
low-elevation stratum of approximatcK' 52 km^
within Yellowstone National Park (Barmore
1980, Houston 1982, Singer 1991). About one-
half of the pronghorn population also sum-
mers in the low-elevation stratum. The lower-
elevation (about 1500 m) exclosures are locat-
ed in typical, nearly snow-free, rolling xeric
shrub and mixed grassland habitats. Elk occu-
p\' both strata and the entire winter range of
810-1000 km- (Houston 1982, Singer 1991).
Only elk and bison winter near the high-eleva-
tion stratum exclosure sites. Winter snow depths
near the higher exclosures (1639-2200 m) are
typically 0.4-0.6 m, which are excessive for
pronghorn and deer. The abrupt elevation rise
for Mt. Everts separates the high and low strata.

1995]

Ungulate Browsing in Yellowstone Park

203

YELLOWSTONE

NATIONAL PARK

Scale

I 1

5 km

Ungulate
• Exclosures

Fig. L Map of the northern Yellowstone ungulate winter range and the high-elevation stratum (six exclosures) and
low-elevation stratum (two exclosures) big sagebrush study sites. Pronghorn, mule deer, and the Wyoming subspecies of
big sagebrush were found only at the low-elevation stratum.

The bison winter range expanded from about
130 km^ in the 1960s in the higher stratum to
about 460 km^ in the late 1980s during a peri-
od of bison population and range expansion
(Meagher 1989, Singer and Norland 1995).
Periodic bison use of the low-elevation stratum
occurred following the population expansion
in the late 1980s (Meagher 1989), but vegeta-
tion measures reported here are nearly all prior
to any bison use of the low-elevation stratum.

Elk numbered about 8000 in 1958-1962
when initial monitoring of the sagebrush belt
transects began. Elk were subsequently further
reduced by artificial controls to less than 5000
in 1967 (Houston 1982). After cessation of con-
trols, elk steadily increased, with counts rang-
ing from 16,000 to 19,000 from 1982 to 1989
(Singer et al. 1989, Singer 1991). Bison were also
artificially controlled until 1967. After cessa-
tion of controls, bison on the northern range
increased from less than 100 in 1967 to 850 by

1988 (Houston 1982, Meagher 1989). Prong-
horn were artificially reduced from 600-800 to
<200 (Barmore 1980), and pronghorn num-
bers remained <200 until about 1981. During
the 1980s — apparently due to milder win-
ters— pronghorn increased to about 600 (Singer
1991). Mule deer counts increased from 1000
in 1985 to 2300 in 1988 over the entire deer
winter range, the majority of which lies north of
the park boundary (Singer 1991). Conversely,
mule deer counts just within the park bound-
aries declined from 230 in the 1960s (Barmore
1980) to about 100 in 1988, in spite of the
overall herd increase.

Methods
Ungulate Densities and Diets

Average ungulate densities near the exclo-
sures were based on actual aerial counts made
from fixed-wing aircraft as described in

204

Great Basin Naturalist

[Volume 55

Barmore (1980), Houston (1982), Meagher
(1989), Singer (1991), and Singer and Norland
(1995). Densities are uncorrected for visibilitv
bias (Samuel et al. 1988) and therefore repre-
sent minimum axerage densities for the study
periods — undoubtedly some animals were
missed on the counts (Singer et al. 1989). Diets
of all fom- ungulates found near the study sites
on the northern winter range were estimated
for each of three winters, December-March
1985-1988, from microhistological analysis of
fecal samples (Washington State University,
Wildlife Habitat Laboratory, Pullman). Each
sample was a composite of 5 g of fresh dung
material from 6-12 dung piles. Aggregate
average percentages are reported for signifi-
cant species and plant groups. To avoid confu-
sion between similar species, fresh samples
were collected for groups of animals immedi-
ately after the groups had vacated an area.
Bighorn sheep (Ovis canadensis) use steeper
tenain on the northern range, and moose {Alces
dices) are found at higher elevations; neither
species was observed near the study sites.

Shrub Utilization Rates

Winter ungulate herbivory rates were sam-
pled on the browsed transects in late winter-
spring before leaf emergence (usuallv late April)
in 1963-1969, 1987, 1989, and 1990. Percent
twig utilization was obtained from counts of all
browsed and unbrowsed twigs on each shrub
located in the transect. Diameters at basal point
and browsing point were measured on 20 ran-
dom shoots on every fifth browsed shrub of
each species, and bite sizes were estimated fol-
lowing Pitt and Schwab (1990).

Trends in Big Sagebrush,
1958-1990

Five exclosures were erected in 1957 and
three more in 1962 (n = 8 total). The exclosures
were placed in sites representative of mixed
big sagebrush/bunchgrass communities. Paired
belt transects (each 1.5 m x 30.5 m = 46.5 m^)
were permanently located inside and outside
eight of the exclosures (one per exclosure treat-
ment) in big sagebrush communities (Canfield
1941, Parker 1954). Each matched pair of
transects was as nearly comparable as possible
in terms of slope, aspect, elevation, shrub
species, and shrub cover (Barmore 1980,
Houston 1982); nevertheless, differences might
have occurred. Sampling of transects occurred

at the date of exclosure, which should reveal
any initial site differences. The transect for
exclosure was selected arbitrarily. Heights and
species of all shrubs found on the belt tran-
sects were recorded in 1958, 1962, 1967,
1974, 1981, 1986, and 1990. Numbers of indi-
vidual shrubs and any shrub seedlings were
tallied. Aerial cover of all shrubs was mapped
on graph paper, and shrub cover was later esti-
mated using a grid (Barmore 1980, Houston
1982).

Detailed Site Comparisons
in 1986 and 1987

Shruli belt transects were not replicated at
a site {n = 1 transect per treatment per loca-
tion, 13 transects total) and were useful pri-
marily for long-term trends and assessment of
pretreatment conditions (Parker 1954). In 1986
and 1987 more intensive and better replicated
measurements (n = 15 plots per treatment)
were gathered; 15 circular plots, each 1.7 m in
radius (9.3 m^), were randomly located in big
sagebrush stands both inside and outside six
e.xclosures. The tallest height, widest diameter,
and perpendicular diameter were recorded for
each shrub within each plot. The number of
totally dead shrubs was recorded. The percent-
age of dead material on partialK' li\ e shnibs was
estimated. At eveiy fifth shrub of each species,
lengths and diameters of 10 randomly sampled
twigs were measured, and eveiy vegetative twig
and reproductive stalk were counted. A mini-
mum of 100 twigs of each species fi'om each site
was collected, dried, and weighed. Canopy
area for each individual shrub was estimated
following Peek (1970) from the formula for the
area of an ellipse:

area

\\)cnd^

where d^ = largest diameter and d^ = its per-
pendicular diameter. All plots and long-term
transects were located more than 25 m from
exclosure fences to avoid the effects of snow-
drifts or ungulate trails along the fences.

Mean shrub height, largest crowii area, shoot
numbers and lengths, and total shrub cover
were compared using a two-way ANOVA, with
browsing and exclosure location as treatments.
The six exclosures should l)e considered repli-
cations of one treatment (browsing) with 5 d.f
used to test for differences among the treat-
ment and replications. Replications at a location

1995]

Ungulate Browsing in Yellowstone Park

205

included the 15 random plots in each treat-
ment (15 X 2 X 8 locations = 240 plots of 9.3
m^). Nonparametric procedures were used for
percent twig utilization comparisons between
the 1960s and 1980s and for other data that
were nonnormal or with unequal variances (F-
max tests; Sokal and Rohlf 1981). Frequency
distributions of shrubs in browsed and un-
browsed plots were compared using the
Kolomogorov-Smirnov test (Zar 1974). All dif-
ferences discussed are significant at the P <
.05 level unless othei^wise indicated.

Aboveground biomass production of shrubs
was estimated from the numbers of shrubs per
plot times the average number of reproductive
and vegetative shoots per plant times the aver-
age dry weight of shoots. Regression equations
for dry weight of shoots (independent vari-
able) were calculated following MacCracken
and Viereck (1990) from diameters at base (de-
pendent variable) and length (dependent vari-
able) of shoots. Separate regression equations
were calculated for reproductive and vegetative
shoots of big sagebrush, green rabbitbrush,
and rubber rabbitbiaish. The regressions on diy
weight were applied to the sample of all twig
diameters and lengths to estimate average
twig biomass.

Results

Ungulate Densities and Diets

Ungulate densities approximately doubled
during the study period on the low-elevation
stratum (Table 1), whereas they approximately
tripled on the high-elevation stratum during
the same period (Table 1). Pronghoni consumed
81% shrubs in their diet, followed by mule
deer 50%, elk 8%, and bison 1% (Table 2).
Pronghorn diets were 49% big sagebrush,
mule deer consumed 23%, and elk diets were
only 4%. The higher combined ungulate den-
sities and the presence of pronghorn and mule
deer, both of which eat more big sagebrush,
suggest that ungulate herbivoiy on big sage-
brush will be greater on the low-elevation
stratum study sites. Since pronghorn consume
12x more big sagebrush in their diets than elk
and 2x more than mule deer, and since prong-
horns also spend summers in the low-elevation
stratum, we suspect pronghorn were the most
important herbivore on big sagebrush on the
low-elevation stratum.

Table 1. Average minimum densities of elk, pronghorn,
mule deer, and bison near the lower- and higher-ele\'ation
exclosures on the northern winter range of Yellowstone
National Park. Reported densities are based upon actual
counts from fixed-wing aircraft (Houston 1982, Meagher
1989, Singer 1991) and are uncorrected for visibility bias.

Ungulate

density (no. /km-)

Lower

Higher

Ungulate

e.xclosures

1965-1968

Elk

Pronghorn

Mule deer

Bison

Total

1985-1988

Elk

16-19

Pronghorn

7-10

Mule deer

Bison

Total

25-31

18-21

Shrub Utilization Rates

Big sagebrush utilization rates were consis-
tently high (87%) and did not differ between
1963-1969 and 1985-1988 at the low-eleva-
tion stratum study sites dominated by the more
palatable (to pronghorn) Wyoming big sage-
brush (Table 3, Mann-Whitney U tests, P >
.05). Pronghorn and elk reductions during
1962-1967 apparently did not result in any
decrease in percent leader use of Wyoming
big sagebrush on the low stratum. Green rabbit-
brush was also used heavily at the low-stratum
sites where deer and pronghorn occuned (Table
3). Utilization rates of big sagebrush at the
higher sites dominated by mostly mountain big
sagebrush, however, increased about sixfold
after ungulates increased threefold (Table 3, P
< .05). Use of green rabbitbrush did not in-
crease significantly at the high stratum during
this period of ungulate increase. Percent leader
use of big sagebrush at the lower-elevation
sites averaged 87%, but leader use averaged
only 11% at the higher sites. Bite sizes averaged
73% of vegetative shoots and 83% of repro-
ductive shoots {n = 180 measured diameters
of browsed shoots and 540 unbrowsed vegeta-
tive and reproductive shoots). Consumption of
annual aboveground biomass of big sagebrush
by ungulates averaged about 68% at the low-
elevation stratum sites and 9% at higher-ele-
vation stratum sites.

206

Great Basin Naturalist

[Volume 55

Table 2. Mean percent ofslirnl);, in winter diets ol lour uny;ulates on Yellowstone's northern winter range, 198.5-1988,
determined by microhistological analysis of feces (x ± SE).

Ungulate (no.

aggregate

Big

Hahhit-

Eiirc

)tUl

Fri

nged

Total

samples)

sagebnish

hrus

I ana

\ta

:ige

shnibh

;''

Elk (28)

3.8

3.1

l.(i

2.7

0.4

1.0

2.0

7,8

4.1

Bison (25)

0.1

0.0

1.3

1.6

Mule deer (21)

23.2

15.1

7 2

4.(i

0.7

1.0

17.9

10.7

49.7

20.9

Pronghorn (20)

48.7

18.0

5.3

3.5

5.8

6.0

18.5

13.7

80.5

15.V

^Ral)l)itl)nisli (iiclmlt's Chnjsuthatnnm numcostt.s and C. lisklijldni.s
''Total slinilis also includes Poptilus spp.. Salix spp.. and Atripli-i spp

Big Sagebrush Trends in Densities,
Heights, and Cover, 1958-1990

Big sagebrush on belt transects in the lower
stratum differed at the time of exclosure in
1958-1962. Densities were similar, but average
heights were 50% and cover about 60% those
values on transects selected for exclosure (Table
4). Big sagebrush densities, heights, and cover,
however, were similar between browsed and
unbrowsed transects at the time of exclosure
on the high-elevation stratum.

Apparently, ungulates were suppressing
Wyoming big sagebrush on the low-elevation
stratimi during the study period. Wyoming big
sagebrush densities decreased 43% and big
sagebrush cover decreased 29% on the low-
stratum browsed site over the 31 -year period
(Table 4). Density and cover of Wyoming big
sagebrush increased dramaticalK' (350% and
830%, respectively) in the unbrowsed sites of
the low sti-atum. Big sagebrush individuals were
taller on unbrowsed sites (Table 4).

Herbivory effects were less on the high-
elevation stratum study sites, and all trends
were similar for browsed and unbrowsed sites.
Mountain and basin big sagebrush density
declined, and canopy cover increased on both
browsed and unbrowsed belt transects,
1958-1990 (Table 4). Wyoming and basin big
sagebrush density declined 39%, but cover
increased 39% on browsed sites over the 31
years. Heights of big sagebrush increased on
both browsed and unbrowsed sites, but more
on unbrowsed sites (Table 4).

Detailed Site Comparisons of Densities,

Cover, and Biomass Production in

1986 and 1987

Densities of big sagebrush (F = 50.9), total
canopy cover of big sagebrush (F = 8.1), indi-
vidual shrub crown area (F = 22.5), and heights

of big sagebrush (F = 79.8, P < .05) differed
between a much larger sample of browsed (n
= 15) and unbrowsed plots {n = 15 per loca-
tion, n = 180 total) sampled in 1986 and 1987.
In each case, however, location was also signif-
icant, and the interaction between location and
browsing was significant. For example, sage-
bmsh individuals were 59% taller on unbrowsed
plots at six e.xclosure sites, but at the Blacktail
exclosures sagebrush plants were taller on
browsed plots. Heights of big sagebrush,
green rabbitbrush, and horsebrush increased
with elevation in both treatments. As a conse-
quence of this exploratory analysis and signifi-
cant interactions with location, our division of
plots into a high and low strata appeared justi-
fied, and we anaK'zed data from the lower and
higher study sites separately in all subsequent
analyses.

Big sagebrush individuals were shorter and
crowns smaller in browsed versus unbrowsed

T.-\BLE 3. Percent of twigs browsed in big sagebrush
communities on ^'ellowstone's northern range. Total un-
gulate numbers increased twofold at the lower e.xclosures
and threefold at the higher e.xclosures between 1963-1969
and 198.5-1988. The same transects of 46.5 m- each {n = 5)
were sampled both periods; only these five browsed tran-
sects were sampled 1963-1969.

Location

196.3-

-1969

198.5-

-1988

Shmb species

(»i = transects)

Low elevation {n =2)

Big sagebrush

88.0

4.2

86.8

7.2

Green rabbitbrush-'

70.1

10.5

Spiny hopsage-'

14.8

4.3

High elevations {n - 3)

Big sagebrush

1.9

0.8

11.6

3.5*

Green rabbitbrush

6.7

3.6

8.9

2.9

Horsebrush

46.6

11.4

'Only big sagebrush utilization was sampled 1963-1969, and green rabbitbrush

at only the higher exclosures.

*P < .05, according to Mann-Whitney U tests.

1995]

Ungulate Browsing in Yellowstone Park

207

Table 4. Changes in densih', heights, and canopy cover of individnal big sagebnish shrubs between lime of exclosure
placement in 1958 and 1990 on permanently marked 46.5 m^ shioib transects, Yellowstone's northern winter range.

Heights

Canopy

cover

Density

of shrubs

:'m)

(m2/46.5 m-)

1958-1962

1990

1958-1962

1990

1958-1962

1990

Treatment

X SE

Lower exclosures"

Browsed

"■

0.7

0.3

0.5

0.1

16 4

Unbrowsed

103

1.8

0.6

16.7

0.9

50 9

Higher exclosures''

Browsed

1.7

1.9

0.9

5.9

1.6

42 7

Unbrowsed

37.S

6.2

1.6

0.2

8.6

0.9

82 9

*Big sagebrush subspecies in these transects, mostly A. t. wyomingensis, are apparent!) high!) palatable to pronghorns.
''Big sagebrush subspecies include mostly A. t. vaseyana.

exclosure sites at low elevations (F = 29.8,
14.3, respectively), but there was no difference
in heights or crowoi sizes due to browsing at
the high-elevation sites (F > .05, Table 5).
Horsebrush was shorter and crowns were
smaller on browsed and unbrowsed exclosure
sites at the higher elexations only (F = 14.5,
4.6, Table 5). Common rabbitbrush was short-
er on browsed plots at the lower elevations,
but it was taller on browsed plots at the higher-
elevation exclosure sites (Table 5). Density of
Wyoming big sagebrush was less on browsed
versus unbrowsed plots at the lower exclo-
sures (F = 14.7), but there was no effect of
browsing at the higher exclosures (Table 6).
No difference in the number of dead big sage-
brush individuals was observed between
browsed and unbrowsed plots at either eleva-
tion category (F > .05). Twenty-two times
more seedlings of the year were obserx^ed on
browsed than unbrowsed plots at higher ele-
vations (F = 2.7, Table 6).

Big sagebrush contributed 82-99% of annual
aboveground shrub production in these shrub
communities. Browsing did not consistently
influence the production of big sagebrush or
green rabbitbrush at higher exclosure sites,
but browsed rubber rabbitbrush produced
less biomass at higher-elevation sites (Table 6).
Botii Wyoming big sagebrush and rubber rabbit-
brush produced much less aboveground bio-
mass on browsed sites on the low study sites
(Table 6).

There was no influence from browsing on
the number of vegetative or reproductive shoots
per shrub for big sagebrush or green rabbit-
brush. Reproductive shoots averaged 42%
longer (Friedman test, Xr^ = 38, n = 6 locations,
F < .05), and vegetative shoots averaged 45%

longer on browsed versus unbrowsed big
sagebiTish (Friedman test, %j.- = 42, n = 6 loca-
tions, F < .05, Table 7). There was no effect of
browsing on length of reproductive shoots of
green rabbitbrush (F > .05).

Discussion

Other studies indicate mountain big sage-
brush is preferred and eaten at a higher rate
by mule deer and elk than Wyoming big sage-
brush, while basin big sagebrush is the least
preferred (Sheehy and Winward 1981, Welch
et al. 1981, Fersonius et al. 1987). Our obsen^a-
tions initially appear in constrast widi diis gener-
alization; we observed 70% more winter utili-
zation on Wyoming big sagebrush than moun-
tain big sagebrush. Too few basin big sage-
brush occurred on the study sites to draw any
conclusions. Our data do not constitute a pala-
tability test, however, in that mountain and
Wyoming subspecies did not occur at the same
study sites. We suspect pronghorn were the
primary herbivore on Wyoming big sagebrush
in lower study sites; pronghorn find the Wyo-
ming subspecies highly palatable (Beetle 1960,
Beetle and Johnson 1982), and that subspecies
was more available to all ungulates due to
shallow snows and more winds in the low stra-
tum. Ungulate preference for big sagebrush
subspecies also varies between locales (Welch
et al. 1981, McArthur and Welch 1982); for
example, Dietz and Nagy (1976) found
Wyoming big sagebiiish was prefen^ed by mule
deer in Colorado.

Mountain and basin big sagebrush seedling
germination, establishment, and survival were
apparently enhanced by browsing and ungu-
late grazing (possibly due to secondary effects

208

Great Basin Naturalist

[Volume 55

Table 5. Iiulixidual slinil) crown and liciiilits of shnihs in lirowscd aiitl iiiihrowsed (protected) sites on Yellowstone's
northern elk winter range. Samples were drawn Ironi in — ISO) plots oi 9.3 ni- eaeli located randoniK in browsed and
nnhrovvsed sites in 1986 and 1987.

( jown

area (cm-)

Heigh

ts (cm)

Shrill) sjiecies

Unbrowsed

■( )w sed

Unl

browsed

owsed

Location

Big sagebrush'

Lower (Artrwy)

678

347

66*

Higher

798

524

Horscbrush

Lower

Higher

575

222

93*

Common rabl)i thrush

Lower

287

109

278

Higher

196

881

590*

Green rabbitbrush

Lower

104

Higher

742

392

*Significant difference behveen grazed and control means using ANOVA, P < .05.

"Big sagebrush subspecies included lower exclosures — A. t. wijomingensis only; higher exclosures — nii.xed piiinihiliniis nf A. I tridriitiita ;

nearK' all A. t. vaseyanu.

id -A. (- vasvifana. but

Table 6. Estimated annual production (g/m-) of the most common shrubs in browsed and unbrowsed big sagebrush
communities at six e.xclosures on Yellowstone's northern winter range (n — 15 plots each in both browsed and
unbrowsed treatments at each site). Wyoming big sagebrush is found only at the lower-stratimi exclosures. and mi.xed
populations of nearly all mountain with some basin big sagebiiish at the higher exclosures.

Lower elev

ations

Higher

elevations

Unbrowsed

Browsed

Unbrowsed

owsed

Exclosure locatioTi

Estimated biomass

(g/m2)

Big sagebrush

1.9

73.6

72.6

Green rabbitbrush

0.1

1.5

3.1

5.8

Common rabbitbrush

0.6

0.3

5.9

4.1

No. big sagebrusli

indi\'iduals/9.3 m-

No. alive

No. dead

1.3

0.4

0.6

0.5

3.3

0.8

5.9

1.9

No. seedlings

0.8

0.3

0.2

0.1

0.2

0.1

4.4

1,4*

*P < .0.5 according to t tests. No tests were conducted on biomass since it was estimated from a product of no. of plants X average no. of shoots X average
weight of shoots. Tests were conducted on each of those parameters separately, however (see text and Table 7), suggesting statistically significant differences at
the lower elevations.

such as reductions of herbaceous vegetation
competition and ungulate hoof action) at the
higher winter range, but the opposite trend
was observed on lower sites. McArthur et al.
(1988) also observ'ed more big sagebrush seed-
lings on a site browsed by mule deer in winter
than on an imbrowsed site. The physical act of
ungulate grazing, with its accompanying hoof
action, greater soil disturbance, more bare
ground, and less standing dead vegetation and
Utter, may provide conditions more suitable to

big sagebrush gemiination. Big sagebrush indi-
viduals are smaller on browsed sites, which
may also benefit establishment and survival of
seedlings due to reduced competition for
light, soil moisture, and other resources.

Ungulate herbivory suppressed big sage-
brush on die lower-elevation sites, where almost
no recruitment of Wyoming big sagebrush
occuiTcd on browsed sites; apparenth' few seed-
lings survive the intense browsing. Wyoming
big sagebrush reproduces more successfully

1995]

Ungulate Browsing in Yellowstone Park

209

Table 7. Numbers and lengths of reproductive and vegetative stalks on shrubs in browsed and unbrowsed plots in big
sagebrush communities on Yellowstone's northern range (n = no. shnibs).

No. 1

reproductive flowers/shrub

Length (cm) of flower st,

alks

Unbrowsed

Brow

sed

ibrowsed

Browsed

Species
Location

Big sagebrush

Lower (Wyoming spp.
Higher (basin and

only)

0.2

0.6**

mountain spp.)
Green rabbitbrush

0.7

0.6**

Higher

0.4

1. vegetative

shoots/shn

Length (cm) i
2 0.1

of vegetative :
3

shoots

Big sagebrush
Lower

0.4

Higher
Green rabbitbrush

0.5

0.5**

Higher

3.3

0.2

0.4**

*P < .0.5.

**P < .01. Differences between numbers in browsed and unbrowsed plots were tested vrith f tests and lengths v\ith Mann-\\'hitne\ U tests.

than the other subspecies on xeric sites (Welch
and Jacobson 1988), and the xeric, sodic clay
soils of the low stratum are clearly more suit-
able to Wyoming big sagebrush. At tlie high lev-
els of ungulate herbivory we observed (rough-
ly 68% biomass removal), the Wyoming sub-
species is presently suppressed by ungulates.

The ability of Wyoming big sagebrush to
recover from herbivory is less than for moun-
tain and basin big sagebrush. Wyoming big
sagebrush is shorter (individuals often do not
exceed 0.3 m), seedling growth rates are lower,
and current annual growth is less than for the
other two subspecies (McArthur and Welch
1982, Booth et al. 1990). The approximately
66% decline in numbers of mule deer using the
lower stratum within the park over the past
two decades may be due to the localized Wyo-
ming big sagebrush decline. Pronghom did not
decline in the lower stratum during the same
period, but pronghom, unlike mule deer, were
artificially reduced well below carrying capac-
ity levels during the 1960s (Houston 1982),
and they may still be recovering from the
reductions.

Increases in height and cover of big sage-
brush are reported after protection from ungu-
lates. Robertson et al. (1970) reported big sage-
brush cover increased 76% after 30 years of
protection from browsing, although mean
heights declined 12%. Heights and crov^ni sizes
were similar, but live cover by big sagebrush
was greater on unbrowsed sites on a mule deer

winter range, primarily due to a greater die-
back of browsed big sagebrush (McArthur et al.
1988). Average crown dieback was 64% in the
browsed area and 17% in the unbrowsed area
(McArthur et al. 1988). Mule deer use was heavy
(370 deer-use days/ha), and dieback of big
sagebrush occurred after two successive win-
ters of heavy snowfall (McArthur et al. 1988).

Browsing by native ungulates stimulated
seedstalks and leaves of big sagebrush and
leaves of green rabbitbrush on the study sites.
Stagnation of shrubs occurred inside big game
exclosures after only two years of exclosure —
nonuse of big sagebrush resulted in an aver-
age 36% reduction in biomass production over
clipped plants (Tueller and Tower 1979).
Numbers of sprouts of green rabbitbrush were
similarly increased by clipping (30% herbage
removal), and new growth was longer, leaves were
larger, and leaves remained green for one
month longer (Willard and McKell 1978). On
the other hand, browsing of more than 80% of
the leaders of mountain big sagebrush by
mule deer resulted in a reduction of 50-93%
in total number of seedstalks per plant and a
reduction of 0-53% in length of seedstalks
(Wagstaff and Welch 1991). Grazed grasses
on the northern Yellowstone winter range
have higher protein levels (Coughenour 1991),
and grazing stimulates aboveground growth
of grasses (Frank and McNaughton 1993).
Increased vigor in new growth of browsed
shrubs on the Yellowstone northern winter

210

Great Basin Naturalist

[Volume 55

range is consistent with these observations of"
grasses, and shrub vigor may be the result of
increased rates of nutrient cycling due to
ungulate defecation and urination (McNaugliton
1979). In addition, plant competition is reduced
and water availability may be increased on
browsed sites on the northern winter range
due to smaller crown sizes and fewer transpir-
ing tissues for indixidual shrubs.

Historic mean duration between fires was
25 years on the Yellowstone northern winter
range, but due to fewer fire starts and active
fire suppression, no significant burning of
grasslands occurred between 1870 and 1988
(Houston 1973, Romme and Despain 1989). Big
sagebrush communities had not yet achieved
climax postfire state on the northern winter
range as indicated by increases in heights and
cover of both browsed and unbrowsed big
sagebrush individuals between 1958 and 1990.
Browsed big sagebnish communities on higlier-
elevation ranges were replacing themselves;
many successful seedlings and small individuals
were observed on browsed versus unbrowsed
sites. Lomasson (1948) observed almost no
reproduction for 40 years in a stand of big
sagebiTish, but then reproduction increased as
the original population began dying. Average
life span of big sagebrush is 53-72 years, and
in a mature, undisturbed stand, most big sage-
brush individuals are in the 55-59-year age
class (Roughton 1972). Sagebrush recovery
following fire varies from a few years to 30
years depending upon environmental condi-
tions for reestablishment (Sneva 1972, Harniss
and Murray 1973). If most big sagebrush com-
munities we studied on the northern winter
range last burned in the 1840-1890 period
(Houston 1973), then most big sagebrush pop-
ulations should have approached senescence
and population turnover at the time of the
1986-87 sampling.

Ungulate herbivory levels on the lower
study stratum restricted growth, establish-
ment, and survival of big sagebrush at the
time of this investigation, although browsed
big sagebrush communities were stable or in-
creasing at the higher elevation. Suppression
of growth and reproduction of plants by in-
creasing native ungulates can result in a new,
altered plant-ungulate equilibrium (Sinclair
1977, Caughley 1981). If unnatural (human-
caused) concentrations of ungulates cause
plant alterations, the situation is not accept-

able under NFS policy (U.S. Department of the
Interior 1988). Houston (1982) concluded
densities of ungulates in the BLA were unnat-
ural and artificially high due to animal avoid-
ance of hunting outside the park. If so, some
form of ungidate management — control, en-
couragement of migrations — is justified on the
BLA. Elk and pronghorn reductions in the
1960s, however, did not reduce percent leader
use or improve the declining status of big
sagebrush in the BLA. Either effective ungu-
late densities remained the same near the big
sagebrush study sites, the ungulate reductions
did not go on long enough, or high preference
for the Wyoming subspecies by pronghorn
maintained high levels of herbivoiy in the area
during the control period.

We caution that we were unable to calculate
appropriate or recommended ungulate herbi-
voiy levels or ungulate densities for the north-
ern winter range. Our data included two dichot-
omous periods in ungulate management. The
first period of our study, 1958-1968, was clear-
ly a period of ungulate underpopulation dur-
ing which time elk, bison, and pronghorn
were controlled far below ecological carrying
capacity (ECC) densities (Barmore 1980,
Houston 1982, Boyce 1993, Mack and Singer
1993, Singer and Norland 1995). The second
period of our investigations, 1986-1988, likely
was a time of ungulate densities in excess of
natural conditions, at least for elk and bison.
This statement is not based upon any compar-
isons to conti^ol conditions (no similar ecosystem
exists with wolves and nondisrupted migra-
tions for a comparison), but upon the conclu-
sions of Houston (1982) that elk concentra-
tions were unnaturally high in the low-eleva-
tion BLA stratum, and computer predictions
that elk and bison would number 8-25% less
following wolf restoration (Garton et al. 1990,
Boyce 1993, Mack and Singer 1993). Pronghorn
densities in relation to EGG are unknown —
one author feels coyotes {Caiiis kitrans) are
suppressing pronghorn on the northern Yellow-
stone winter range and that, following wolf
restoration, coyotes will decline and prong-
horn will further increase (Berger 1991). Wolf
restoration occurred on the study area during
tlie winter of 1994-95, providing an opportunity
to test the effects of wolves upon ungulate-
plant interactions in the Yellowstone ecosys-
tem (Gook 1993).

1995]

Ungulate Browsing in Yellowstone Park

211

Acknowledgments

The research was funded by the U.S. Depart-
ment of the Interior, National Park Service,
Natural Preservation Program, Washington
D.C., and Yellowstone National Park. The
authors acknowledge J. Varley and R. Barbee for
administrative support, and D. Frank, W. Wiens,
J. Whipple, G. Kittel, M. Hennen, J. Meek,
and M. Harter for field assistance. D. Swift,
M. Coughenour, A. Beetle, E. Durant McAithur,
and J. Whipple reviewed the manuscript.

Literature Cited

Barmore, W, J. 1980. Population characteristics, distribu-
tion and habitat relationships of six ungulates in
northern Yellowstone Park. Unpublished report,
Yellowstone National Park files. Mammoth, \VY.

Beetle, A. A. 1960. A study of sagebrush: the section
Tridentata of Artemisia. University of Wyoming,
Agricultural E.xperiment Station Bulletin .368. 68 pp.

Beetle, A. A., and K. L. Johnson. 1982. Sagebrush in
Wyoming. Wyoming Agricultural Experiment Station
Bulletin. 68 pp.

Berger, J. 1991. Greater Yellowstone's ungulates; myths
and realities. Consei^vation Biology 5: 353-363.

Booth, G. D., B. L. Welch, and T. L. C. Jacobson. 1990.
Seedling growth rates of 3 subspecies of big sage-
brush. Journal of Range Management 43: 432-436.

Boyce, M. S. 1993. Predicting the consequences of wolf
recovery to ungulates in Yellowstone National Park.
Pages 234-269 in R. S. Cook, editor. Ecological
issues on reintroducing wolves into Yellowstone
National Park. U.S. National Park Service Science
Monograph 22. Denver, CO.

Cahalane, V. H. 1943. Elk management and herd regula-
tion— Yellowstone National Park. Transactions of the
North American W'ildlife Conference 8; 95-101.

Canfield, R. 1941. Application of the line interception
method in sampling range vegetation. Journal of
Forestry 39: 388-394.

Caughley, G. 1981. Overpopulation. Pages 7-19 in P A.
Jewell, S. H. Holt, and D. Hart, editors, Problems in
management of locally abundant wild animals.
Academic Press, New York. 361 pp.

Cayot, L. J., J. Prukop, and D. R. Smith. 1979. Zootic cli-
max vegetation and natural regulation. Wildlife
Society Bulletin 7; 162-169.

Chase, A. 1986. Playing God in Yellowstone. Atlantic
Monthly Press, Boston. 446 pp.

Cole, G. F 1971. An ecological rationale for the natural or
artificial regulation of ungulates. Transactions of the
North American Wildlife Conference 36: 417-425.

Congressional Record. 1986. Senate S. 12613. U.S. House
of Representatives.

Cook, R. S., editor. 1993. Ecological issues on reintro-
ducing wolves into Yellowstone National Park.
National Park Service Science Monograph 22. Denver,
CO.

Coughenour, M. B. 1991. Biomass and nitiogen responses
to grazing of upland steppe on Yellowstone's north-
ern winter range. Journal of Applied Ecology 28:
71-82.

Despain, D. G. 1991. Yellowstone vegetation, conse-
quences of environment and historv' in a natural set-
ting. Roberts Rinehart Publishing, Boulder, CO. 239
pp.

DiETZ, D. R. and J. G. Nagy. 1976. Mule deer nutrition
and plant utilization. Pages 71-78 in C. W. Workman
and J. B. Low, editors. Mule deer decline in the
West — a symposium. Utah State University, College
of Natural Resources, Logan. 134 pp.

Frank, D., and S. J. McN,\ughton. 1993. Interactive
ecology of plants, large mammalian herbivores and
drought in Yellowstone National Park. Unpublished
doctoral dissertation, Syracuse University, Syracuse,
NY 150 pp.

Carton, E. O., R. L. Crabtree, B. B. Ackerman, and C.
Wright. 1990. The potential impact of a reintro-
duced wolf population on the northern Yellowstone
elk herd. Pages 3-59 in Yellowstone National Park,
U.S. Fish and Wildlife Service, University of Wyo-
ming, University of Idaho, Interagency Study Team,
University of Minnesota Cooperative Park Studies
Unit, editors. Wolves for Yellowstone? Report to the
United States Congress. Volume II. Research and
analysis.

Harniss, R. O., and R. B. Murray 1973. Thirty years of
vegetation change following burning in sagebrush-
grassland range. Journal of Range Management 26:
322-325.

Houston, D. B. 1973. Wildfires in northern Yellowstone
National Park. Ecology .54: 1111-1117.

. 1976. Research on ungulates in northern Yellow-
stone National Park. Pages 11-27 in Research in the
parks. Transactions of the National Park Symposium,
National Park Sei-vice Symposium No. 1.

. 1982. The northern Yellowstone elk: ecology and

management. Macmillan Publishing Co., Inc., New
York. 474 pp.

Kay, C. E. 1991. Yellowstone's northern elk herd: a critical
evaluation of the "natural regulation" paradigm. Un-
published doctoral dissertation, Utah State University,
Logan. 490 pp.

Kittams, W. H. 1950. Sagebrush on the lower Yellowstone
range as an indicator of wildlife stocking. Yellow-
stone National Park files. Mammoth, WY. 14 pp.

. 1959. Future of the Yellowstone wapiti. Naturalist

10: 30-39.

Lane, J. 1990. Characterization and comparison of soils
inside and outside of grazing e.xclosures on Yellow-
stone National Park's northern winter range. Un-
published master's thesis, Montana State University,
Bozeman.

Lomasson, T. 1948. Succession in sagebrush. Journal of
Range Management 1: 19-21.

MacCracken, J. G., AND L. a. Viereck. 1990. Browse
regrowth and use by moose after fire in interior
Alaska. Northwest Science 64: 11-18.

Mack, J. A., and F J. Singer. 1993. Using Pop-II models
to predict effects of wolf predation and hunter har-
vests on elk, mule deer, and moose on the northern
range. Pages 49-74 in R. S. Cook, editor. Ecological
issues on reintroducing wolves into Yellowstone
National Park. National Park Service Science Mono-
graph 22. Denver, CO.

McArthur, E. D., and B. L. Welch. 1982. Growth rate
differences among big sagebrush (Artemisia tridentata)
accessions and subspecies. Journal of Range Manage-
ment 35: 396-401.

212

Cheat Basin Natufl\list

[Volume 55

McArthur, E. D., a. C. Blauer, and S. C. Sanderson.
1988. Mule deer induced mortaliU' of mountain hi^
sagebrush. Journal of Range Management 41;
114-117.

McN.AUGHTON, S. J. 1979. Grazing as an optimi/.ation pro-
cess: grass-ungulate relationships in the Serengeti.
American Naturalist 113; 691-703.

MEAf;iiER, M. M. 1973. The bison of Yellowstone National
Park. National Park Service Science Monograph
Series 1. 161 pp.

. 1989. Range expansion by bison of Yellowstone

National Park. Journal of Mammalogy 70; 670-675.

Parker, K. W. 1954. A method for measuring trend in
range condition on National Forest ranges with sup-
plemental information for measurement of vigor,
composition, and browse. USDA Forest Service
Report. 37 pp.

Peek, J. M. 1970. Relation of canopy area and volume to
production of three woody species. Ecology 51;
1098-1101.

. 1980. Natural regulation of ungulates. Wildlife

Society Bulletin 8: 217-227.

Personius, T. L., C. L. Wambolt, J. R. Stephens, and
R. G. Kelsey. 1987. Crude terpenoid influence on
mule deer preference for sagebrush browse. Journal
of Range Management 40: 84—88.

Pitt, M. D., and F. E. Schwab. 1990. Assessment of a
nondestructixe method for estimating browse use.
Journal of Wildlife Management 54: 175-179.

Robertson, J. H., D. L. Neal, R. McAdams, and R T.
TUELLER. 1970. Changes in crested wheatgrass ranges
under different grazing treatments. Journal of Range
Management 23; 27-.34.

ROMME, W. H., and D. G. Despain. 1989. Historical per-
spective on the Yellowstone fires of 1988. BioScience
39; 695-699.

ROUGHTON, R. D. 1972. Shrub age structure on a mule
deer winter range in Colorado. Ecology' 53; 615-625.

Rush, W. M. 1932. Northern Yellowstone elk study. Mon-
tana Fish and Game Commission, Helena. 131 pp.

Samuel, M. D., E. O. Carton, M. W. Schlegel, and R. G.
Carson. 1988. Visibility bias during aerial sun'eys of
elk in north-central Idaho. Journal of Wildlife Man-
agement 51; 622-630.

Sheehy, D. R, and a. H. Winward. 1981. Relative palata-
l:)ility of seven Artemisia taxa to mule deer and sheep.
Journal of Range Management 34: .397-399.

Singer, E J. 1991. The ungulate prey base for wolves in
Yellowstone National Park. Pages .323-348 in R. B.
Keiter and M. S. Boyce, editors. The Greater Yellow-
stone Ecosystem: redefining America's wilderness
heritage. Yale University' Press, New Haven, CT.

Singer, F J., and J. Norland. 1995. Niche relationships
within a guild of ungulates following release from
artificial controls, Yellowstone National Park, Wyo-
ming. Canadian Journal of Zoology 72; In press.

Singer, F J., W. Sghreier, J. Oppenheim, and E. O.
Carton. 1989. Drought, fires, and large mammals.
Bioscience 39; 716-722.

Sinclair, A. R. E. 1977. The African buffalo; a study of
resource limitation of populations. University of
Chicago Press, Chicago, IL. 355 pp.

Sneva, F a. 1972. Grazing return following sagebrush
control in eastern Oregon. Journal of Range Manage-
ment 25; 174-178.

SOKAL, R. R., AND E J. RoHLE 1981. Biometry. W. H.
Freeman, San Francisco, CA.

TuELLER, R J., AND J. D. TowER. 1979. Vegetation stagna-
tion in three-phase big game exclosures. Journal of
Range Management 32; 258-263.

U.S. Department of Interior, National Park Service.
1988. Management policies. U.S. Government Print-
ing Office, Washington, DC.

Wagstaff, F J., AND B. L. Welch. 1991. Seedstalk pro-
duction of mountain big sagebiiish enhanced througli
short-term protection from heavy browsing. Journal
of Range Management 44: 72-74.

Wambolt, C. L., R. G. Kelsey, T. L. Personius, K. D.
Striby, a. F McNeal, and K. M. H.wstad. 1987.
Preference and digestibility of three big sagebrush
subspecies and black sagebnish as related to crude
terpenoid chemistry. Pages 71-73 in E D. Provenza,
J. T. Flinders, and E. D. McArthur, compilers. Sym-
posium on plant-herbivore interactions. USDA Forest
Service, General Technical Report INT-222. Ogden,
UT.

Welch, B. L., and T L. C. Jacobson. 1988. Root growth
of Artemisia tndentata. Joum;d of Range Management
41;. 3.32-334

Welch, B. L., E. D. McArthur, and J. N. Davis. 1981.
Differential preferences of wintering mule deer for
accessions of big sagebrush and black sagebrush.
Journal of Range Management 34: 409-411.

Will.\rd, E. E., and C. M. McKell. 1978. Response of
shrubs to simulated browsing. Journal of Wildlife
Management 42: 514-519.

Zar, J. H. 1974. Biostatistical analysis. Prentice-Hall Inc.,
Englewood Cliffs, NJ. 620 pp.

Received 13 September 1994
Accepted 19 January 1995

SOFT SEDIMENT BENTHIC MACRO INVERTED RATE COMMUNITIES

OF THE GREEN RIVER AT THE OURAY NATIONAL WILDLIFE

REFUGE, UINTAH COUNTY, UTAH

Eric R. Wolzl and Dennis K. Shiozawa^-^

Abstract. — Benthic macroinvertebrates from four habitat types (river channel, ephemeral side channel, river back-
water, and seasonally inundated wetland) were e.xamined from the Green River at the Ouray National Wildlife Refuge,
Uintah County, UT, June-August 1991. Four major taxa (Nematoda, Oligochaeta, Diptera: Ceratopogonidae, and
Chironomidae) were quantified. Cluster analysis of densities showed that habitat types with comparable flow conditions
were the most similar. Highest to lowest overall benthic invertebrate densities of the four habitats were as follows:
ephemeral side channel, river backwater, seasonally inundated wetland, and river channel. Nematodes were the most
abundant taxon in all habitat t\'pes and sample dates e.xcept the August sample of the river channel and river backwater
and the July sample of the seasonally inundated wetland.

Key words: benthic macroinvertebrates, Nematoda, Oligochaeta, Ceratopogonidae, Chironomidae, river benthos, wetland,
benthos. Green Riven

In 1962 Flaming Gorge Dam was completed
on the Green River in northeastern Utah. This,
in addition to dikes constnicted along the river's
course and the introduction of nonnative fishes,
has altered natural conditions such that many
native fishes have reached the brink of extinc-
tion and are now listed as endangered species.
Grabowski and Hiebert (1989) studied the
Green River below Flaming Gorge Dam and
noted the importance of backwaters as nursery
habitats to introduced and native fishes. They
found the most important food items to be ben-
thic macroinvertebrates, predominantly chiro-
nomid larvae. Their investigation was confined
to two habitats: the main channel and river
backwaters. We also studied benthic commu-
nities of the river channel and back-water habi-
tats and two additional habitats — seasonally
inundated wetlands and ephemeral side chan-
nels. No published information exists about
the community structure of benthic macro-
invertebrates in these latter two habitat types.

Benthic invertebrates of large rivers are
poorly known. Difficulty in sampling, the
amount of time needed to process samples,
identification of specimens after collection,
and heterogeneity of habitats make study diffi-
cult and often expensive. Studies of riverine
systems have utilized divergent methodologies.

Some studies randomly sample an entire river
cross section and do not attempt to quantify dif-
ferent river habitat types (Grzybkowska 1989,
Grzybkowska et al. 1990, Munn and Brusven
1991). Other studies have been directed toward
specific river habitats such as riffles (Rader
and Ward 1988, Morgan et al. 1991), floodplains
(Gladden and Smock 1990), or tailwaters of re-
servoirs (Swink and Novotny 1985). Relatively
few have simultaneously studied multiple
habitat types in a single river system (Beckett
et al. 1983, Grabowski and Hiebert 1989).

Our purpose was to determine densities
and community assemblages of the major ben-
thic macroinvertebrates in four Green River
habitats: river channel, ephemeral side channel,
river backwater, and seasonally inundated wet-
land. Benthic samples were taken from lune
through August 1991, in the Green River at the
Ouray National Wildlife Refiige, Uintali County,
UT USA.

Study Sites

The Green River originates in Wyoming and
flows south through eastern Utah to its conflu-
ence with the Colorado River (Fig. 1). It adds
more volume to the Colorado River system than
any other tributary. In eastern Utah, at river km
404, the Green River enters the Ouray National

'Chadwick & Associates, Inc., Littleton, CO 80120.

^Department of Zoology, Brigham Young University', Prove, UT 84602 USA.

■'Author to whom correspondence should be addressed.

213

214

Great Basin Naturalist

[Volume 55

Wildlife Refuge. This seetion of the river has
the lowest gradient of the entire Green River
system. Riparian vegetation consists of willow
and tamari.x with occasional cottonwoods. We
collected monthly samples in the Ouray
National Wildlife Refuge (see also Fig. 2). In
addition to benthic samples, water chemistry
was determined for each habitat type on each
sample date (Table 1). Salinity and conductivity
were recorded with a YSI meter (Yellowstone
Instruments); turbidity was measured with a
nephelometer; and hardness, pH, and alkalini-
ty were determined with a Hach Kit (Hach
Chemical Corporation). Water chemistry was
recorded at three locations per sample area on
each sample date. At each site, a min-max ther-
mometer was placed near the benthos-water
interface at the time of sampling and left for
10 days. Substrate composition was estimated
visually.

River Channel

The river channel was sampled approximate-
ly 1.3 km north of the United States Fish and
Wildlife Service (USFWS) hatchery on the
Ouray National Wildlife Refuge. Sampling was
adjacent to a sand bar that decreased water
tmbulence and prevented shifting sands. Water
chemistry values were relatively stable. Turbid-
ity was substantially higher during the August
sample. Substrate consisted mostly of sand with

little silt and detritus. Water levels were too
high during June (peak flow) to allow sampling.

Ephemeral Side Channel

During high flows the Green River will
occupy various smaller channels that are diy
during low-flow intervals. We have named
such habitats "ephemeral side channels." The
ephemeral side channel studied was approxi-
mately 2.75 km south of the USFWS hatchery.
For most of the year water levels in the main
channel were below the level of the ephemeral
side channel. However, during peak flow, water
filtered through a wooded area and gathered
into the channel, which was 10 m wide and
500 m long. As the river level dropped, flow
slowed and eventually stopped. Because the
side channel dried up shortly after the July
sample, no August sample was taken. Most
notable of the water chemistry measurements
was the increase of salinity and alkalinity when
comparing June to July. Water temperature
also deviated more during July. Substrate con-
sisted mostly of firm silt and detritus with little
sand. Sediment deposition contributed little to
the site during our study.

SEASONALLY I^aINDATED WETLAND

100 Kilometers

Fig. 1. Regional map showing the location of the Oura\'
National Wildlife Refuge.

Fig. 2. Local map of the Ouray National Wildlife
Refuge, Uintah Count\; UT, showing the location of sam-
pling sites.

1995]

Macroinvertebrates of the Green River

215

Table 1. Mean ± standard deviation water cheniistiy
temperature in °C, salinity in percent, condiicti\it\
CaCOg).

values from Green River sample sites, June-August 1991 (n — 3,
in /xmhos, turliidity in NTUs, hardness and alkalinity in ppm

Habitat type Date

M in. /max
temp.

pll

Salinity

Conductivity

Turbidity

Hardness

Alkalinity

River channel

7/15

8. 14 ±.09

.04 ± .0

753 ± 6

183 ±318

411±0

183 ± 10

8/12

20.5/26.5

8.48 ±.10

.04 ± .01

718 ± 8

402 ±41

320 ± 20

205 ±17

Ephemeral side channel

6/3

20.5/30.5

9.0 ±0

.03 ± .06

326 ± 10

57 ±6

183 ± 20

171 ±0

7/1

16/30.5

9.14 ±.16

.12 ±.03

445 ± 5

127 ±21

228 ± 10

240 ± 17

River back\\ater

7/10

20.5/29.5

7.98 ± .23

.01 ±.01

523 ± 23

57 ±9

228 ± 10

183 ± 20

8/8

19/26.5

8.59 ±.12

.03 ± .0

730 ±111

45 ±11

268 ± 40

228 ± 26

Seasonallv inundated we

tland

6/10

19.5/26.5

9.0 ±0

,02 ± .0

314 ±8

52 ±8

154 ±0

143 ± 10

7/12

22/32

8.37 ±.11

.02 ± .01

446 ± 20

36 ±8

205 ±0

223 ±0

8/15

22/29.5

8.93 ±.1

.01 ± .0

345 ± 13

195 ± 17

171 ±17

154 ± 0

*Theniioineter lost

River Backwater

River backwaters are submerged during high
flows and do not emerge as distinct entities
until the river drops. For this reason the river
backwater was not sampled during peak flow
(June). The river backwater we sampled, located
just upstream of the river channel site described
above, was approximately 10 m wide X 50 m
long and 1.3 m deep. Turbidity, alkalinity, and
pH were highest during the August sample.
Substrate consisted mostly of loose silt and
detritus with virtually no sand. Silt and detritus
were constantly being deposited during the
study period.

Seasonally Inundated Wetland

This site, commonly called "Old Charlie's
Wash," is a shallow floodplain wetland man-
aged by the USFWS for waterfowl and is
located approximately 4.3 km south of the
USFWS hatchery. As the river rises in the
spring, water enters Old Charlie's Wash and,
at peak flow, retaining structures are put in
place to create a 43-ha pond and to prevent
the impounded water from receding as rapidly
as the river. By early fall the water in Old
Charlie's Wash is nearly depleted by seepage
and evaporation. Turbidity increased dramati-
cally during the August sample, and conduc-
tivity, hardness, and alkalinity peaked during
the July sample. Substrate consisted of firm silt,
detritus, and sand.

Methods

Sampling

Samples were collected during the summer
of 1991 (Tables 2-5). Initial sampling of the
ephemeral side channel and seasonally inun-
dated wetland occurred just after river flow
peaked in early June, but samples for the river
channel and backwater habitats were not col-
lected because the water level was too high. All
four habitats were sampled during July and all
but the ephemeral side channel during August.
Fifty core samples were taken along a 30-m
transect at each site. Each sample was collect-
ed with a clear aciylic tube, 450 mm long x
47 mm in diameter (Shiozawa 1985), which
was pushed into the substrate to a depth of
60-80 mm. Sediment from each sample was
preserved in 5% formalin with rose bengal
stain added to aid in sample sorting.

Sample Processing

In the laboratory we washed each sample
to separate organisms from sediments using
the following procedure. First, the formalin
was drained and replaced with tap water. The
sample was then gently stirred to resuspend
the sediments and poured into a plastic tray
(36.5 cm X 31.5 cm X 6 cm) through which a
small volume of warm water flowed. The out-
flowing water, laden with small sand and clay
particles, detritus, and benthic invertebrates,
was filtered through a 63-^tm screen. Larger

216

Great Basin Naturalist

[Volume 55

Table 2. Densities of bentliic iiivertehrates (#/ni-) from tlic Green River, river channel hal)itat, Onray National
Wildlife Reftige, Ouray, UT

15 July 1991

12 August 1991

# of samples

# ot samples

Taxon

Density/m2 (95% C.L.)

processed

Density/m2 (95% C.L.)

processed

Nematoda

24,881

(13,107-47,302)

2421 (2063-2840)

Oligochaeta

3426

(2565-4570)

11,182 (7497-16,678)

Insecta

Ceratopogonidae

3608 (2731-4767)

13,026 (9316-18,215)

Chironomidae

4150 (2798-6155)

3516 (2454-5037)

Earlv instars

1037

3016

Chirottomus

346

Cijphomella

Lenziella

576

Paramerina

115

Paratendipes

Polijpedihim

1844

269

Procladiiis

115

Psectrocladius

115

StempellineUa

Tanijtarsus

T.\BLE 3. Densities of benthic invertebrates {#/m~) from the Green River, ephemeral side channel habitat, Ouray
National Wildlife Refuge, Oura>, UT

3 June 1991

1 July 1991

# of samples

Taxon

Density/m2 (95%^ C.L.)

processed

Densit

y/m2(95%C.L.)

processed

Nematoda

261,680 (88,934-769,968)

302,603 (215,886-424,154)

Oligochaeta

2728 (2096-3546)

12,796

(10,681-15,329)

Insecta

Ceratopogonidae

Chironomidae

2325 (1843-2927)

8185

(6385-10,491)

Earlv instars

979

2075

Chironomus

1134

3112

Cryptochironomus

115

Cryptotendipes

461

Lenziella

1383

Polypedilum

692

Procladius

346

Tanypus

Tanytarsits

.sediment particles (sands and structural clays)
that remained in the plastic tray were periodi-
cally examined for specimens. If none were
found, the sediments were discarded. Material
collected on the screen was stored in 70%
ETOH.

Samples sorted were randomly chosen from
the 50 samples taken at each site and date. Each
sample was placed in glass petri dishes (from
one to six dishes depending on the amount of
material) and sorted under a dissecting micro-
scope (see Tables 1-4 for number of samples
processed). Four major taxa (Nematoda, Oligo-

chaeta, Ceratopogonidae, and Chironomidae)
were counted. Only Chironomidae were iden-
tified to the generic level. Miscellaneous taxa
were also recorded but were not quantified
(see Table 5).

The number of samples sorted fiom each site
and sampling date was determined as follows:
5 of the 50 samples were randomly selected
and the four major taxa were counted. Because
of their contagious distribution (determined by
calculating variance to mean ratios), numbers
of individuals of each taxon were then log
transformed (x + 1). The variance and mean

1995]

Macroinvertebrates of the Green River

217

Table 4. Densities of benthic invertebrates (#/ni-) from the Green River, river backwater habitat, Ouray National
Wildhfe Refuge, Ouray, UT

10 July 1991

8 August 1991

# of sam

pies

# of samples

Taxon

Density/m- (95% C.L.)

process

ied

Density/rn- (95% C.L.)

processed

Nematoda

54,872 (24,350-123,650)

134,183 (94,656-190,542)

Ohgochaeta

26,642 (14,622-48,495)

164,731 (101,881-266,728)

Inseeta

Ceratopogonidae

96 (90-107)

461 (385-552)

Chironoinidae

31,125 (15,356-63,089)

22,863 (12,139-13,136)

Earlv instars

8877

7301

Chironomus

7032

6340

Lenziella

346

1249

Polypedilum

14,179

5860

Prochidiits

461

1345

Psectrodadius

115

Tamjtarsus

115

769

Table 5. Densities of benthic invertebrates (#/m2) from the Green River, seasonally inundated wetland habitat,
Ouray National Wildlife Reftige, Ouray, UT

10 June 1991

12 July 1991

15 August 1991

Densitv/m^

# of samples

Densit>7m- # of samples Density/m- #

of samples

Ta.\on

(95% C.L.)

processed

(95% C.L.) processed

(95% C.L.) ]

Drocessed

Nematoda

7133 (4534-11,266)

80,694

(38,595-168,713) 5

88,533 (83,125-94,784)

Oligochaeta

4573 (3402-6141)

87,150

(39,242-193,547) 10

22,249 (11,930-41,494)

Inseeta

Ceratopogonidae

0 14

2478 (1941-3165)

Chironomidae

903 (895-915)

23,055

; (13,707-38,780) 14

3977 (2816-5617)

Earlv instars

8769

2479

Ablabesmijia

124

Chironomus

154

576

Cricotopus

453

Cnjptochironomus 134

206

Cryptotendipes

947

346

Glyptotendipes

988

Lenziella

115

1112

Microtendipes

1029

Paratamjtarsiis

231

6505

Polypedilum

2388

173

Procladius

124

Psectrocladius

Tanypus

124

173

Tamjtarsus

206

115

Zavrelia

were used in the following formula to estimate
the number of samples to process (Elliot

1977):

N =

where N = number of samples to process, S =
variance, d = level of accuracy desired for the

sample (in this case 0.1), and x = the mean.
For our samples d was chosen to be 0.1, for an
accuracy within 10% of the mean. If, after five
samples were processed, N was <5 for a spe-
cific taxonomic group, no more samples were
processed for that group. Those taxa for which
N was >5 were counted in an additional sam-
ple. The mean and variance for taxa not elimi-
nated were again calculated using the addi-
tional sample value(s) and above formula. This

218

Great Basin Naturalist

[Volume 55

process continued until N was less than the
number of samples already processed for the
taxon. Because of time and financial constraints,
we never picked more than 30 samples for any
specific habitat and sample date. All sorted
samples were preserved in 70% ETOH.

Chironomids were removed from 70%
ETOM and placed in distilled water for 10-15
min prior to clearing. Individual specimens were
placed in hot (-80 °C) 10% KOH (Cranston
1982) for 5-15 min to clear (larger specimens
lequired more time to clear). After clearing,
specimens were transferred to distilled water
for at least 5 min. Each specimen was then
placed in glycerine on a microscope slide for
identification. Only late instars were identifi-
able. Representative specimens of each genus
encountered were permanently mounted.
Specimens were classified to the generic level
using keys by Mason (1968), Wiederholm (1983),
and Merritt and Cummins (1984).

Data Analysis

Average densities (#/m") and 95% confi-
dence limits for each of the four main taxa and
each genus of Chironomidae were calculated
for each sample site and date. Because density
distributions were contagious, 95% confidence
intervals were calculated for each of the four
main taxa using a logarithmic transformation
suggested by Elliot (1977; Tables 2-5). These

values were then applied to the arithmetic mean
(Shiozawa and Barnes 1977). Confidence inter-
vals were not calculated for each genus in the
Chironomidae because densities of some genera
were too low.

Cluster analysis was performed using the
statistical package NTSYS-pc (Rohlf 1992).
Several dissimilarit>' measures, including Bray-
Curtis, Canberra's, and Renkonen s, were used
to generate distance matrices. A comparison of
each of these matrices to the original data
showed that the Bray-Curtis measure (Bray
and Curtis 1957) provided the best "fit " of the
cluster analysis to the data. Average linkage
clustering of the Bray-Curtis distances, based
on the mean number of individuals/m^ of each
species between habitat types and sample dates,
was done with the unweighted pair-group
method using arithmetic averages (UPGMA;
Krebs 1989).

Results

Invertebrates

Nematodes occurred in eveiy sample pro-
cessed and were most abundant in the July
sample of the ephemeral side channel habitat
(302,603/m-) and least abundant in the river
channel August sample (2421/m-; Tables 2-5).
They comprised the majority of benthic inver-
tebrates in all habitats and sample dates except

Table 6. Functional group (Merritt and Cummins 19S4) and habitat association of Chironomidae genera from the
Green River, Ouray National Wildlife Refuge, Ouray, UT.

Fimctional

group

Taxon

Collectors

Predators

Shredders

Unknown

Habitat association*

Ahlahcsmijia

SIW

Chir(»u)intis

RC,ESC,RB,SIW

CladotaiHjtarsiis

RC,ESC,RB,SIW

Criartopus

SIW

Cnjptnchironomus

ESCSIW

('njptotendipes

ESCSIW

nr. CijphomeUa

C.lyptolcndipes

SIW

Microtcndipes

SIW

Paraiiichna

ParatcDiiitarsiis

SIW

Paratendipi's

Polijpcdilum

RC,ESC,RB,SIW

Pwcladius

RC,ESC,RB,SIW

Pscctroclddius

RC,RB,SIW

nr Steinpellinella

Tamjpus

ESC.SIW

Tanijtarsus

RC,ESC,RB,SIW

Zavrelia

SIW

lumifl. ESC = ephemeral side channel, RB = ri\ (. r l)ack\\atf r, SIW = seasonully iiiinitlalrd \

1995]

Macroinvertebrates of the Green River

219

the August river channel and river backwater
habitats and the July wetland sample.

Oligochaetes were present in all habitat
types and on all sample dates. Densities ranged
from a low of 2728/m^ in the June ephemeral
side channel sample to a high of 164,73 l/m^ in
the July river backwater sample (Tables 2-5).

The lowest abundance of Ceratopogonids
was observed in the July river backwater sam-
ple (96/m^). Their density was 136X greater in
the river channel August sample (13,026/m^;
Tables 2-5). Ceratopogonids were absent from
both June and July samples of the seasonally
inundated wetland and the ephemeral side
channel.

Ninteen chironomid genera were collected
during this study. Fourteen genera were found
in the July seasonally inundated wetland sam-
ples, and five genera occurred in the August
river channel and river backwater samples.
Seven genera occurred in only one habitat or
on only one date. Si.x genera were found in the
seasonally inundated wetland habitat only, and
four occuned onK' in the river channel. No chi-
ronomid genus was unique to the ephemeral
side channel or the river backwater. The genus
Polypedihim was collected in all habitat types
and on all sample dates. Total chironomid densi-

ties were least (903/m-) in the June sample of
the seasonally inundated wetland and greatest
(31,125/m^) in the July river backwater sample
(Tables 2-5). Unidentifiable early instars were
collected in all habitat types and in all sample
periods and comprised 86% of the river chan-
nel sample in August. The most common func-
tional group category of the Green River chi-
ronomids was collectors followed by predators
and shredders. Specific functional group and
Green River habitat association for each genus
are presented in Table 6.

Other insects found in the samples are list-
ed in Table 7. Density estimates would not be
valid for these taxa because of their ability to
avoid the core sampler.

Cluster Analysis

The UPGMA cluster analysis of the benthic
invertebrate communities in each habitat type
and sample date indicated that sites with similar
flow conditions tended to cluster together (Fig.
3). A matrix comparison of original distances
calculated using the Bray-Curtis coefficient
with distances implied from the dendrogram is
presented in Figure 4. Correlation between
the two was high {R = .907), implying that the
dendrogram is an accurate representation of

Table 7. Other insects encountered in the Green Ri\ or ecos> stem, June-August 1991.

River

Ephemeral

Riv

Seasonally inundated

channel

side channel

backwater

wetland

Taxon

July

August

June July

July

August

June Juh' August

Coleoptera

Hydrophihdae (larvae)

Diptera

Chironomidae (pupae)

X X

Empididae (larvae)

Simuliidae (lan'ae)

Ephemeroptera

Baetidae

Baetis (nymph)

X X

Callibaetis (nymph)

Caenidae

Caenis (nymph)

Tricorythidae

Tricorythodes (nymph)

Hemiptera

Corixidae

Odonata

Coenagrionidae

Ischnura (nymph)

X X

Gomphidae (nymph)

Plecoptera

Perlodidae (nymph)

Isoperla

220

Great Basin Natur.\list

[Volume 55

1.00

0.75

Bray-Curtis Distance

0.50
I

0.25

•0.00

Seasonal Wetland -June
River Channel -July
River Channel - August
Seasonal Wetland -July
River Backwater - August
Seasonal Wetland - August
River Backwater -July
Side Channel - June
Side Channel -July

Fig. 3. UPGMA cluster analysis of Green River habitat t)pes located in the Ouray National Wildlife Refuge.

the original Bray-Curtis distances. Ephemeral
side channel samples show the greatest simi-
larity (least distance), and wetland and back-
water sites are more similar to one another

Discussion
Nematoda

The importance of free-living nematodes in
aquatic systems has not been extensively stud-
ied. Aquatic nematodes are known to be micro-
botrophic, predaceous, and/or parasitic during
one or more of their life stages (Poinar 1991).
Due to the scarcity of adequate keys and their
small size, nematodes are seldom listed beyond
the phylum designation in most studies and
may not even be quantified. In studies of aqua-
tic systems where nematodes are quantified,
highest densities have been found in lakes.
Strayer (1985) and Nalepa and Quigley (1983)
reported that nematodes comprised 60% and
80%, respectively, of all benthic metazoans in
Mirror Lake, NH, and in Lake Michigan with
means of 680,000/m2 (Minor Lake) and 260,000/
m2 (Lake Michigan). In contrast. Palmer (1990)
in Goose Creek and Gladden and Smock (1990)
on the floodplain of Colliers Creek reported
that nematodes comprised a much smaller
percentage (6% of total invertebrates) and
occurred at diminished densities (1000-15,000/
m^ and 1746/m2, respectively) in lotic systems.

In our study nematode density estimates
from the seasonally inundated wetland June
sample (7133/m2) and the July and August
river channel samples (24,881/m- and 2421/m^,
respectively) are comparable to densities pre-
viously reported from lotic systems (Gladden
and Smock 1990, Palmer 1990). Density esti-

mates for all other sites and dates (54,872-
302,603/m^) are more similar to densities in
lentic habitats (see above). Greater densities
are achieved in the more stable benthic envi-
ronments provided by calmer waters and finer
sediment particle size. In their study of White
Clay Creek, Bott and Kaplan (1989) found that
nematode densities were greater in silt than in
sand. In our study the highest densities are
also associated with a low sand content in the
substratum. Low densities reported for the
June sample of the seasonally inundated wet-
land site reflect the relatively short time that
water had been on the sample site. Of the four
major invertebrate groups collected in this
study, nematodes accounted for 8% of the
individuals in the river channel August sample
and 98% in the June ephemeral side channel.
Nematodes accounted for 67.7% of all organisms
observed. Palmer (1990), using a 3.3-cm-dia.
core and 44-yam mesh, reported that nematodes
constituted only 4-15% of the Goose Creek
community, with a mean of 9%. Her data are
similar to our river channel values. High nema-
tode densities and their high percentage of the
total invertebrates that we report from the
ephemeral side channel, river backwater, and
seasonally inundated wetland are unusual and
should be compared to samples taken at similar
locations in this and other large rivers using
comparable methods.

Oligochaeta

Freshwater oligochaetes are a well-studied
and diverse group found in every type of estu-
arine and freshwater habitat. They feed mostly
on bacteria living in soft sediments (Brinkliurst
and Gelder 1991). The amount and quality of

1995]

Macroinvertebrates of the Green River

221

i.uu-

•

• ••

0.75-

•

«••

•

0.25-
n nn-

•

0.00

0.25 0.50

Original Distance

.75

1.00

Fig. 4. Comparison of original dissimilarih' matrix and implied matrix from the dendrogram.

organic matter found in the sediment are pri-
mary factors determining which species will
be present in a particular area (Brinkhurst and
Cook 1974). We identified our specimens only
to class level. Oligochaete densities in nonpol-
luted lakes are lower than those in organically
polluted waters. Densities in Mirror Lake
ranged fi'om .30,000 to 33,000/m2 (Strayer 1985).
Jonasson and Thorhauge (1976) reported oligo-
chaete densities in Lake Esrom, Denmark, of
6000-12,000/m2. Brinkhurst and Cook (1974)
found that densities of the three most common
tubificids in the more polluted areas of Toronto
Harbor ranged from 51,000 to 197,000/m2.
Oligochaete densities in nonpolluted lotic sys-
tems tend to be lower. Grzybkowska and
Witczak (1990) report oligochaete densities in
the lower Grabia River, Poland, ranging from
110 to 900/m2, and Palmer (1990) reports den-
sities from 5000 to 15,000/m2 in Goose Creek,
VA. Densities from polluted lotic systems can
approach 200,000/m2 (Koehn and Frank 1980).
Oligochaete densities in the seasonally in-
undated wetland June sample (87,150/m2) and
river backwater August sample (164,731/m2)
are comparable to values observed in polluted
systems described above. Densities from both
ephemeral side channel samples (2728 m^ and
12,796/m2) and both river channel samples
(3426/m2 and ll,182/m2) are comparable to
those in Goose Creek (Palmer 1990). In general,

oligochaete densities in our study were higher
in habitats with the least amount of water flow
(seasonally inundated wetland and river back-
water habitat types). Terrestrial vegetation
invades wetlands during dry periods, and when
the water returns the following spring, decaying
vegetation forms a rich food base. Backwater
habitats retain fine particles, including detri-
tus, being transported by the river; as summer
progresses, this creates an enriched food base.
These factors are the likely reason for the con-
vergence oligochaete densities in these two hab-
itats with those in organically polluted systems.

Ceratopogonidae

The study of ceratopogonids has mainly
centered on adults because of their economic
importance (Davies and Walker 1974). Larvae
inhabit a variety of habitats including tree
holes, leafiDacks, and pitcher plants, but are usu-
ally most numerous in shallow areas of streams,
lakes, and ponds (Bowen 1983). Aquatic forms
are mostly predaceous (Merritt and Cummins
1984), but several species are known to consume
algae and plant debris (Kwan and Morrison
1974).

Corkum (1990) investigated streams associ-
ated with different land-use types in south-
western Ontario and found densities of 50/m2
in "forested" sites, 480/m2 in "mixed" sites,
and 5300/m2 in "farmland" sites. Adamek and
Sukop (1992) found maximum densities of only

222

Great Basin Natur.\list

[Volume 55

1/m^ on over-flooded meadows in Czechoslo-
\akia. In Lake Norman, NC, Bowen (1983)
reported a mean lar\'al ceratopogonid density
of767/m2.

Ceratopogonid densities reached a peak in
tiie August river channel sample (13,026/ m^) —
much higher than any reported in the litera-
ture above. In their study of the Green River,
Grabowski and Hiebert (1989) did not report
densities, but did conclude that ceratopogo-
nids were more abundant in river channel
samples than in backwaters. Our study supports
this conclusion. Average densities for the river
channel July and August samples were 3608/m-
and 13,026/m2, respectively, compared to 96/m-
and 461/m2 for the backwater July and August
samples. Ceratopogonid lai^vae were complete-
ly absent from the ephemeral side channel as
well as the June and July seasonally inundated
wetland samples.

Chironomidae

Chironomidae are typically the most abun-
dant macroinvertebrates in lentic (Strayer 1985)
and lotic (Grzybkowska and Witczak 1990) sys-
tems. Studies of relatively small geographical
areas have reported impressive numbers of
species. For instance, Douglas and Murray
(1980) found 142 species in Killarney Valley,
Ireland. High diversity of chironomids makes
them important as indicators of environmental
condition (Wingard and Olive 1989). They are
also abundant and provide an important food
source for fish (Brown et al. 1980, Winkel and
Davids 1987, Grabowski and Hiebert 1989),
waterfowl (Titmus and Baddock 1980), and
other migratoiy birds (Bowman 1980).

We identified 19 chironomid genera from
our sites within the Green River ecosystem.
Other investigations of lotic systems have yield-
ed similar numbers — 12 genera in the upper
Tuscarawas River, OH (W^ingard and Olive
1989), 24 genera in the River Frome, England
(Finder 1980), 25 genera in the Mississippi
River (Beckett et al. 1983), and 36 genera in
Juday Creek, IN (Berg and Hellenthal 1991).
Grabowski and Hiebert (1989) studied the
Green River in the same general area consid-
ered in our study and also identified 19 genera.
However, only seven of the genera reported
by the latter authors were found in our study:
Chiron(»mis, Cricotopus, Cryptochirunouius,
Polijpediliiin, Procladius, Tamjpus, and
Tanytarsus.

Densities of chironomids in aquatic sys-
tems can van substantialK. In a study of Lake
Vissavesi, Finland, Paasivirta and Koskenniemi
(1980) reported densities of 64/m^ in a coarse
debris habitat and 2997/m- in a moss-grown
site. Jonasson and Lindegaard (1979) reported
59,000/m2 from Lake Myvatn, Iceland. Vari-
ability in lotic systems has also been docu-
mented. Finder (1980) reported densities from
a low of 48/m2 to 6273/m- in a chalk stream in
England, and Grzybkowska (1989) found
10,664/m^ in the River Grabia, Foland. While
no distinct trends e.xist when comparing chiro-
nomid densities in lentic and lotic SNstems, den-
sities are influenced by sediment size (Faasivirta
and Koskenniemi 1980, Beckett et al. 1983).

Chironomid densities from the July and
August river channel samples were 4148/m2
and 3516/m2, respectively. River backwater
samples were 31,125/m2 and 22,864/m2 for the
same times. Grabowski and Hiebert (1989)
reported maximum chironomid densities in
the same area of the Green River of less than
lOO/m^ for the river channel and 2800/m2 for
river back'waters — substantially less than our
estimates. It is possible that annual differences
in seasonal discharge, area of the sampling
device, and later sampling period all contrib-
uted to this discrepancy. However, because of
significant differences in mesh size (63-/u,m
ours, 600-/xm Grabowski and Hiebert's), data
of Grabowski and Hiebert and ours cannot be
considered equivalent. It is worth noting that
mesh sizes larger than 100 [xm have been shown
to negatively bias density estimates (Strayer
1985).

Community Similarity

Cluster analysis of the data showed that, in
general, habitat t\'pes clustered together inde-
pendent of sample date, suggesting that the
different habitat types studied in the Green
River are distinct. Beckett et al. (1983), for ex-
ample, studied five habitats in the Mississippi
River and also found them to remain composi-
tionally distinct regardless of flow and sample
date. Distril)ution and abundance of benthic
macroinvertebrates characteristic of these
habitat types have been attributed to flow con-
ditions and sediment size in our study. Since
flow conditions are the major determinant of
particle size, flow conditions are likely the
determining factor. This conclusion has also

1995]

Macroinvertebrates of the Green River

223

been reached by other investigators (Beckett
et al. 1983, Statzner and Higler 1986).

Grabowski and Hiebert (1989) conchided
that benthic macroinvertebrate densities in
backwaters of the Green River were higlier than
those of the river channeh Our data suggest
that the seasonally inundated wetland and
ephemeral side channel are also valuable habi-
tats and have the potential to contribute sub-
stantial biomass to the Green River system.
Oligochaete and chironomid densities report-
ed in our study are comparable to other lotic
systems (Koehn and Frank 1980, Finder 1980,
Grzybkowska 1989, Grzybkowska and Witczak
1990, Palmer 1990). High densities of nema-
todes and ceratopogonids imply that these
groups may be very important in the overall
energetics of the Green River system. Both
should be studied more intensely. The overall
dynamics of these communities is undoubtedly
associated with seasonal changes in flow as well
as year-to-year variability in annual discharge.
This study, while describing a backwater, river
site, side channel, and floodplain wetland over
a short time interval, does not allow a full
assessment of either annual or spatial variabil-
ity. It is clear that some sort of successional
colonization of various habitats occurs; for
instance, floodplain wetlands are maximum in
extent during highest spring-early summer
flows, but their faunal development lags peak
flooding. Back-waters do not exist during high
flows, but as floodplains diminish with reced-
ing water levels, back-water habitats develop.
Again their faunal assemblages tend to lag be-
hind the emergence of recognized back-waters.
While we documented what appears to be
seasonal succession within habitat type, such
changes should not be assumed the norm.
Until a detailed study is undertaken for the
Green River or Colorado River system with
replicate habitats over at least a full year period,
our observations must be considered tentative.
Further, annual discharge can vary tremen-
dously from year to year, depending upon fac-
tors such as drought cycles and their link with
El Nino dynamics in the Pacific. Thus, what is
seen in one year may not be representative of
all years. Such factors introduce additional
variables that should be considered when
attempting to understand the dynamics of the
benthos of the Green River.

Literature Cited

Adamek, Z., and I. SUKOP, 1992. Invertebrate communi-
ties of fomier soiitliem Moravian floodplains (Czecho-
slovakia) and impacts of regulation. Regulated Rivers:
Research and Management 7: 181-192.

Beckett, D. C, C. R. Bingham, and L. G. Sanders. 1983.
Benthic macroinvertebrates of selected habitats of the
lower Mississippi River. Journal of Freshwater
Ecology 2: 247-261.

Berg, M. B., and R. A. Hellenthal. 1991. Secondaiy pro-
duction of Chironomidae (Diptera) in a north tem-
perate stream. Freshwater Biology 25: 497-50.5.

BOTT, T. L., and L. a. K\pl.\n. 1989. Densities of benthic
protozoa and nematodes in a piedmont stream. Journal
of the North American Benthological Society 8:
187-196.

Bowen, T. W. 1983. Production of the predaceous midge
tribes Sphaeromiini and Palpomyiini (Diptera;
Ceratopogonidae) in Lake Norman, Nortli Carolina.
Hydrobiologia 99: 81-87.

Bowman, C. M. T. 1980. Emergence of chironomids from
Rosterne Mere, England. Pages 291-295 in D. A.
Murray, editor, Chironomidae: ecolog>', systematics,
cytology, and physiology. Proceedings of the 7th
International Symposium on Chironomidae, Dublin,
Ireland.

Brw; J. R., AND J. T. Curtis. 1957. An ordination of the
upland forest communities of southern Wisconsin.
Ecological Monographs 27: 325-349.

Brinkhurst, R. O., AND D. G. Cook. 1974. Aquatic earth-
worms (Annelida: Oligochaeta). Pages 143-156 in
C. W. Hart, Jr, and S. L. H. Fuller, editors. Pollution
ecology of freshwater invertebrates. Academic Press,
New York.

Brinkhurst, R. O., and S. R. Gelder. 1991. Annelida:
Oligochaeta and Branchiobdellidi. Pages 401-433 in
J. H. Thoip and A. P Covich, editors, Ecolog\ and
classification of North American freshwater inverte-
brates. Acedemic Press Inc., San Diego, CA.

Brown, A. E., R. A. Oldham, and A. Warlow. 1980.
Chironomid larvae and pupae in the diet of brown
trout {Salmo tnitta) and rainbow trout {Salmo gaird-
neri) in Rutland Water Leicestershire. Pages 323-329
in D. A. Munay, editor Chironomidae: ecology; sys-
tematics, cytology, and physiology. Proceedings of
the 7th International Symposium on Chironomidae,
Dublin, Ireland.

CORKUM, L. D. 1990. Intrabiome distributional patterns of
lotic macroinvertebrate assemblages. Canadian
Journal of Fisheries and Aquatic Sciences 47:
2147-2157.

Cranston, R S. 1982. A key to the Umae of the British
Orthocladiinae (Chironomidae). Freshwater Biological
Association. Scientific Publication 45.

Davies, E G., and a. R. Walker. 1974. The isolation of
ephemeral fever virus from cattle and Culiciodes
midges in Kenya. Veterinary Record 95: 63-64.

Douglas, D. J., and D. A. Murray. 1980. A checklist of
Chironomidae (Diptera) of the Killamey Valley catch-
ment area Ireland. Pages 12.3-129 in D. A. Murray
editor, Chironomidae: ecology, systematics, cytologx;
and ph\'siolog>-. Proceedings of the 7th International
Symposium on Chironomidae, Dublin, Ireland.

Elliot, J. M. 1977. Some methods for the statistical anal\-
sis of samples of benthic invertebrates. Freshwater
Biological Association, Scientific Publication 25.

224

Great Basin Naturalist

[Volume 55

Gladden, J. E., and L. A. SM(X:k. 1990. Macroinvcrte-
brate distribution and production on the floodplains
of two lowland headwater streams. Freshwater Biology
24: 533-545.

CiRABOWSKi, S. J., AND S. D. HiKBKivi. 1989. Some aspects
of trophic interactions in selected backwaters and the
main channel of the Green River, Utah, 1987-1988.
Bureau of Reclamation, Environmental Sciences
Section, Denver, GO.

GkzyhK{)\\sk:\, M. 1989. Production estimates of the dom-
inant ta.xa of Chironomidae (Diptera) in the modi-
fied, Hi\er W'idawka, and the natural. River Grabia,
central Poland. H>dn)biologia 179; 245-259.

Grzybkow.ska, M., and J. WiTCZAK. 1990. Distribution
and production of Ghironomidae (Diptera) in the
lower course of the Grabia River (central Poland).
Freshwater Biology 24: 519-531.

Gr7.vbk{)\vsic\, M., J. Hejduk, and P Zielinski. 1990.
Seasonal dynamics and production of Chironomidae
in a large lowland river upstream and downstream
from a new reservoir in central Poland. Archiv fiir
Hydrobiologie 119: 439-455.

JoNASSON, R M., and C. Lindegaard. 1979. Zoobenthos
and its contribution to the metabolism of shallow-
lakes. Ergebnisse der Limnologie (supplement to
Archiv fiir Hydrobiologie) 13: 162-180.

JONASSON, R M., AND E Thorhauge. 1976. Population
dynamics of Potamothrix hainmoniensis in the pro-
fundal of Lake Esrom with special reference to envi-
ronmental and competitive factors. Oikos 27: 19.3-203.

KOEHN, T, .\ND G. Fr^^nk. 1980. Effect of thermal pollution
on the chironomid fauna in an urban channel. Pages
187-194 in D. A. Murray, editor, Ghironomidae;
ecology, systematics, cytology, and physiology. Pro-
ceedings of the 7th International Symposium on
Ghironomidae, Dublin, Ireland.

Krebs, G. K. 1989. Ecological methodology. Harper
Gollins Publishing Go., New York.

KwAN, \V. E., AND E O. Morrison. 1974. A summan of
published information for field and laboratoiy stud-
ies of biting midges, Cidicoides species (Diptera;
Geratopogonidae). Annals of the Entomological
Society of Quebec 19: 127-137.

M.vsoN, W. T, Jr. 1968. An introduction to the identifica-
tion of chironomid larvae. Federal Water Pollution
Gontrol Administration, U.S. Department of the
Interior, Cincinnati, OH.

Merritt, R. W, AND K. W. Gummins, editors. 1984. An
introduction to the aquatic insects of North America.
Kendall/Hunt Publishing Go., Dubuque, lA.

Morgan, R. R, II, R. E. Jacobsen, S. B. Weisberg, L. A.
MgDowell, and H. T. Wilson. 1991. Effects of
flow alteration on benthic macroinvertebrate com-
miMiities below the Brighton hydroelectric dam.
Journal of Freshwater Ecology 6: 419-429.

MuNN, M. D., and M. a Brusven. 1991. Benthic macro-
invertebrate communities in nonregulated and regu-
lated waters of the Glearwater River, Idaho, U.S.A.
Regulated Rivers: Research and Management 6;
1-11.

Nalepa, T. F, and M. a. Quigley. 1983. Abundance and
biomass of the meiobenthos in nearshore Lake
Michigan with comparisons to the macrobenthos.
Journal of Great Lakes Research 9: 530-547.

Palmer, M. A. 1990. Temporal and spatial dynamics of
meiofauna within the hyporheic zone of Goose Greek,

Virginia. Journal of the North American Benthologi-
cal Society 9: 17-25.

P.\ASiviRTA, L., AND E. KosKENNlEMl. 1980. The Ghiro-
nomidae (Diptera) in two polyhumic reservoirs in
western Finland. Pages 23.3-238 in D. A. Murray,
editor, Ghironomidae: ecolog}-, systematics, cytology,
and physiolog). Proceedings of the 7th International
Symposium on Ghironomidae, Dublin, Ireland.

PiNDER, L. G. V. 1980. Spatial distribution of Ghironomidae
in an English chalk stream. Pages 15.3-161 in D. A.
Murra\', editor, Ghironomidae; ecology, systematics,
cytology, and physiology. Proceedings of the 7th
International Symposium on Ghironomidae, Dublin,
Ireland.

POINAR, G. O., Jr. 1991. Nematoda and Nematomoqiha.
Pages 249-283 in J. H. Thoip and A. P Govich, editors,
Ecology and classification of North American fi-esh-
water invertebrates. Academic Press Inc., San Diego,
GA.

Rader, R. B., AND J. V. Ward. 1988. Influence of regulation
on environmental conditions and the macroinverte-
brate community in the upper Colorado River. Regu-
lated Rivers: Research and IVlanagement 2; 597-618.

ROHLF, E J. 1992. NTSYS-pc. Numerical ta.xonomy and
multivariate analysis system. Version 1.70. Depart-
ment of Ecology and Evolution, State University of
New York, Stony Brook.

Shiozawa, D. K. 1985. The seasonal community structure
and drift of microcmstaceans in Valley Creek, Minne-
sota. Canadian Journal of Zoology 64: 16.55-1664.

Shiozawa, D. K., and J. R. Barnes. 1977. The microdis-
tribution and population trends of larval Tanypus
stellatu.s Cociuillett and Clunmomits frommeri Atchley
and Martin (Diptera: Chironomidae) in LUah Lake,
Utah. Ecolog>- .58; 610-618.

Statzner, B., and B. Higler. 1986. Stream hydraulics as a
major determinant of benthic invertebrate zonation
patterns. Freshwater Biolog\' 16: 127-139.

STR.AYER, D. 1985. The benthic micrometazoans of Mirror
Lake, New Hampshire. Archiv fiir Hydrobiologie
Supplement 72: 287-426.

SwiNK, W D., and J. E NovoTNY. 1985. Invertebrate colo-
nization rates in the tailwater of a Kentucky flood-
control resei-\'oir. Journal of Freshwater Ecology 3:
27-34.

TiTMUS, G., and R. M. Baddogk. 1980. Production and
emergence of chironomids in a wet gravel pit. Pages
299-305 Hi D. A. Murray, editor, Ghironomidae: ecol-
ogy, systematics, cytologv', and physiolog\'. Proceedings
of die 7di International Symposiimi on Ghironomidae,
Dublin, Ireland.

WiEDERHOLM, T, EDITOR. 1983. Ghironomidae of the
Holarctic region: keys and diagnoses. Part 1, Larvae.
Borgstroms Tryckeri AB, Motala.

WiNGARD, G. J., and J. H. Olive. 1989. Larval Ghirono-
midae (Diptera) of the upper Tuscarawas River of
northeastern Ohio, U.S.A. Journal of Freshwater
Ecology 5; 93-102.

VViNKEL, E. K., and T. Davids. 1987. Cyprinid fish and
water mite reducing chironomid populations. Entomo-
logica Scandinavica Supplementum 29: 26.5-267.

Received 24 January 1994
Accepted 14 November 1994

ALPINE VASCULAR FLORA OF THE TUSHAR MOUNTAINS, UTAH

Alan C. Tliyel

Abstract. — The Tushar Mountains of southwestern Utah rise to a maximum elevation of 3709 m, with tiniherline and
krummholz reaching maximum elevations of 3438 m and 3566 m, respectively. Voucher specimens were collected from
the alpine region during eight field seasons to inventor^' this largely unknown alpine flora. Listed are 171 vascular plant
species from 102 genera and 34 families that occur in eight types of plant communities within an alpine area of about
19.3 km-. The seven largest families are Asteraceae (29 species), Poaceae (20), Brassicaceae (13), Rosaceae (12),
Cyperaceae (11), Car}ophyllaceae (10), and Fabaceae (8). Thirteen species are restricted to the alpine area. The perenni-
al herb growth form accounts for 86.4% of the flora, .5.9% of the species are shrubs, and the remaining species are annu-
als to short-lived perennials. Bedrock in the alpine region is entirely of Tertiary igneous origin. Vegetation cover and
species richness are highest on an andesite ash-How tuff and latite flow and lowest on hydrothermally altered inter-
caldera rhyolites and tuffs. Forty-four species (26.0% of the indigenous flora) also occur in the Arctic, and 13 species are
at a southern margin of distribution. Eight taxa (4.7% of the flora) are local or regional endemics. The majority' of the
alpine species appear to have migrated to the range by way of the contiguous mountain system to the north; statistical
comparison with neighboring alpine floras shows the flora to be most similar to the floras of the Wasatch Mountains,
Uinta Mountains, and Teton Range, with Sorensen's similarity indices of 52.8, 50.2, and 48.8% respectively.

Key words: Utah, Colorado Plateau, Tushar Mountains, alpine vascular flora, alpine vegetation, plant geography.

The Tushar Mountains, located in south-
western Utah in the High Plateaus section of
the Colorado Plateau at the eastern margin of
the Great Basin (Fig. 1), reach a maximum ele-
vation of 3709 m at the summit of Delano Peak.
This elevation is surpassed within the state
only by peaks in the Uinta Mountains and La
Sal Mountains. A diverse alpine environment
and flora occur on the 11 peaks that rise above
the elevation of timberline. The alpine area is
isolated. Though minor patches of alpine vege-
tation occur on the Fish Lake Platetiu 66 km to
the northeast and Markagunt Plateau 81 km to
the south, the nearest extensive alpine area
occurs in the Wasatch Mountains (Mount Nebo)
157 km to the north. The purposes of this paper
are to document this isolated alpine flora,
briefly describe the alpine plant communities,
and determine possible migrational pathways
to the Tushars by means of statistical and qual-
itative comparisons with neighboring alpine
areas.

Study Area

The Tushar Mountains have a length of 60
km and a width of 36 km at the widest point.
Vertical relief exceeds 2000 m, with a low eleva-

tion of 1695 m at the confluence of Clear Creek
and the Sevier River. The range is located
within an area of large-volume Tertiary (Oligo-
cene to Miocene) volcanic activity known as
the Marysvale volcanic field and is composed
mostly of volcanic rocks (Cunningham and
Steven 1979). Structurally, the range consists
of a plateau-like, north-trending, up-faulted
block bordered by structural valleys formed
from down-faulted blocks; the High Plateaus
section is thus structurally transitional be-
tween the Basin and Range Province and the
Colorado Plateau Province (Hunt 1987). The
major faulting that produced the current linear
ranges of the High Plateaus occured between
8 million and 5 million years ago (Steven et al.
1984).

Topography and soil development in the
alpine area are strongly influenced by the two
volcanic formations exposed near timberline
and above (Fig. 2). The mostly plateau-like to
domelike ridges in the southern and eastern
portions of the alpine region (including Delano
Peak) are composed of calc-alkaline basaltic
andesite flows and tuffs of the Bullion Canyon
Volcanics (Cunningham et al. 1983) on which
two soil complexes consisting of mollic cry-
oborolls, argic pachic cryoborolls, pachic

1465 North 300 \\est #22. Pro\o, UT 84601.

225

226

Great Basin Naturalist

[Volume 55

38-

Fig. 1. Map of Utah showing location of Tushar
Mountains (in black) and other mountainous areas above
2286 m in elevation (generalized and adapted from
Smouse and Gurgel [1981]).

ciyohoroUs, litliic ciyoboroUs, and rock outcrops
are recognized (U.S. Forest Sei-vice 1993).

The more mountainous noilliem and western
portions of the alpine region are composed of
intercaldera siliceous alkali rhyolite lava flows,
lava domes, and ash-flow tuffs of the Mount
Belknap Volcanics which have been hydrother-
mally altered in many places (Cunningham
and Steven 1979) and which are resistant to
weathering. The summit pyramids of Mount
Belknap (3699 m) and Mount Baldy (3695 m)
consist of steep talus slopes and cliffs; portions
of these talus slopes lack a cliff at their head
and thus appear by definition (Washburn 1979)
to be block slopes formed from periglacial
frost action. Soil development in this region is
limited to areas too small to map, and all
alpine exposures of this formation as mapped
by Cunningham et al. (1983) are classified by
soil scientists as a cirqueland-rubbleland-rock
outcrop complex (U.S. Forest Sei-vice 1993).

Pleistocene glaciers produced several well-
defined cirques on the eastern side of the crest
where glacial ice descended to a low elevation
of about 2500 m (Callaghan 1973). Glaciation
also occurred on the western side of the crest
as evidenced by glacial striations in the Poison

[TT| Mount Belknap Volcanics
m Bullion Canyon Volcanics

Fig. 2. Map of the central Tushar Mountains with out-
lined portions approximating the areas above 3383 m in
elevation. Location of igneous formations is generalized
from Cunningham et al. (1983).

Creek drainage. Periglacial patterned ground
in the form of stone stripes, stone circles, and
stone nets occurs on the main ridgecrest be-
tween the elevations of 3432 and 3600 m.

Climate of the Rock\' Mountain alpine zone
is characterized by Billings (1988) to have a
mean temperature of less than 10° C for the
warmest month. Climatic data are unavailable
for the alpine area in the Tushar Mountains. A
weather station located in an Engelmann spiaice
{Piceo engehnannii) community 3.6 km south
of the alpine region at an elevation of 3136 m
has a mean annual temperature of 1.7°C, the
warmest months being June, July, and August
with mean monthly temperatures of 9.2, 11.7,
and 1().9°C, respectively (three years of rec-
ords). Most of the mean annual precipitation
of 909 mm (12 years of records) falls as snow.
Rainfall from summer thunderstorms is high-
est in August, with an average of 74 mm, while
the months of June and July receive averages

1995]

TusHAR Mountains Alpine Flora

227

of 46 and 45 mm, respectively (Soil Conserva-
tion Sei-vice 1993).

An alpine region is defined by Bliss (1985)
to be tlie area above die climatic limit ot upright
tree growth, although it may include patches
of krummholz. The average elevation of tim-
berline in the Tushar Mountains occurs at
about 3383 m (11,100 ft), with a corresponding
alpine area of about 19.3 km^ above this eleva-
tion. This alpine area, located between
38°20'04" and 38°27'47" North latitude and
112°19'32" and 112°26'42" West longitude,
extends from Signal Peak in the north to Lake
Peak at the south over a distance of 14.5 km.
Much of the alpine area, centered about 25
km ENE of the city of Beaver, is accessible by
Forest Service road 123, which crosses the
crest of the range at an elevation of 3505 m.

Timberline coincides with the upper limit of
continuous forest and reaches a maximum ele-
vation of 3438 m on a minor ridgecrest on the
western (windward) side of the range. Timber-
line occurs as low as 3341 m on lower ridges
and is edaphically depressed even lower on
some talus slopes. Engelmann spruce and lim-
ber pine {Pinus flexilis) are the only arboreal
species found at timberline. Subalpine fir {Alnes
lasiocarpa) and aspen {Popidus tremidoides)
approach timberline with maximum known ele-
vations of 3365 and 3292 m, respectively. The
krummholz limit, consisting of Engelmann
spruce, occurs at about 3566 m on the steep,
south-facing slope of Mount Baldy; this slope
is protected from prevailing winds by a ridge
extending southwesterly from the summit.

Methods

Voucher specimens were collected from
1984 through 1993 from throughout the range
in preparation of a checklist of the vascular
plants of the Tushar Mountains. Collected
specimens were deposited in the herbarium of
Brigham Young University and a search was
made in this herbarium for other pertinent
specimens. The total known flora for the range
consists of 971 taxa representing 924 species,
381 genera, and 89 families (Taye 1994). The
alpine region was visited during die same period
except for the years 1986 and 1989. Only those
species found above local timberline are
included in this study.

Species nomenclature and life form classifi-
cation follow Welsh et al. (1993).

Sorensen's Index of Similarity (Mueller-
Dombois and Ellenberg 1974) was used to
compare the alpine flora with neighboring alpine
floras to determine possible migrational path-
ways to the Tushars. Differences in nomencla-
ture among the floras were largely resolved
with the references of Dorn (1988), Weber and
Wittmann (1992), and Welsh et al. (1993).
Subspecific taxa were not used in statistical
comparisons.

Alpine Vegetation

Eight types of alpine plant communities
were recognized based on qualitative obsewa-
tions; future intensive study of the vegetation
will likely expand this classification. As noted for
alpine communities in the Uinta Mountains
(Lewis 1970), boundaries between plant com-
munities are usually diffused.

Cushion plant. — Low-growing species
dominate the windswept ridgecrests where soils
are shallow and outcrops of bedrock and rub-
ble formed in place by frost-heaving are com-
mon. Dominant species include Carex ehj-
noides, Cerastiiim beeringianum, Erigeron
compositus, Festuca ovina, Geiim rossii. Phlox
pidvinata, and SUene acaidis.

Dry meadow. — The warmer and drier south-
and west-facing slopes are characterized by a
plant cover in which bare soil is generally pre-
sent between individual plants; rock cover is
frequently high and soil disturbance from
pocket gophers is common. Spruce krummholz
is common at lower elevations. Common species
in this extensive community type include
Achdiea mdlefolinm, Astragalus miser, Carex
elynoides, CastUleja parvula, Cymopterus lem-
monii, Elymus trachycaulus, Haplopappus
macronema, Heleniiim hoopsii. Phlox pidvina-
ta, Potentilla glandidosa, P. gracilis, Poa seciin-
da, and Ribes montigenum. Geum rossii occurs
occasionally in usually mesic microhabitats.
Alpine populations of Gentiana parry i, Jiiniperas
communis, Sambucus racemosa, Thalictrum
fendleri, and Viola nuttallii occur onl\^ in this
community type at low elevations.

Mesic meadow. — Plant cover is generally
higher on suitable (nontalus or bedrock) north-
and east-facing slopes and near drainage bot-
toms and is occasionally cai-pet-like where suf-
ficient soil development has occurred. Alpine
avens {Geum rossii) is perhaps the most com-
mon species in this community. Komarkova

228

Great Basin Natukalist

[Volume 55

(1979) found tliis species to he most abundant
on slopes with deep soil profiles and moderately
proloniied snow cover Other common species
in this extensive community type include
Arenaria obtusiloha, Artemisia scopulorum,
Carex heteroneura, Cerastiinn heerinf^ianwn,
Erigeron simplex, Luziila spicata, Pcdiculans
parnji, PJilox puhimita, Poa reflexa, Pohj^^onwn
bistortuides, Salix arctica, Saxifraga rhotn-
boidea, and Silene acaidis.

Wet meadow. — A few wet meadows occur
adjacent to rivulets and below long-lasting
snowdrifts. Common species include Geiim
rossii. Polygonum bistortoides, and Primula
parriji. Caltha leptosepala, Potentilla diversifo-
lia, Seduin rhodanthum, and Stellaria longipes
reach their upper ele\'ational limit of 3627 m
in this community type on the north-facing
slope of Delano Peak.

Rivulet. — Alpine rivulets from melting
snowfields are mostly transitoiy and occur only
in the southern (Bullion Canyon Volcanics) por-
tion of the alpine region. Cardamine cordifo-
lia, Deschampsia cespitosa. Delphinium occi-
dentale van barbeyi, and Mertensia arizonica
occur at lower elevations while Caltha lep-
tosepala, Pedicularis parryi, and Salix arctica
occur more commonly at higher elevations.
Polygonum bistortoides and Primula parryi are
common throughout this community type.
Two rivulets on the flanks of Delano Peak (to
about 3444 m in elevation) persist throughout
the summer; Epilobium halleanum, Juncus
drummondii, Mimtdus tilingii. and Saxifraga
odontoloma occur at their upper elevational
limit at these continually moist and marginally
alpine sites.

Gravelly barren. — This is perhaps the
most distinctive alpine community type in the
Tushar Mountains. It occurs on saddles of ridge-
crests and on man\' of the higher tributar\'
drainages between Lake Peak and Mount Bel-
knap where snow accumulations are long last-
ing; plant cover is only 0 to about 20% on
largely unaltered, gravelly, grayish parent mate-
rial. Some of the relatively few species that
occur here are Calyptridium umbellatum,
Elymus scribneri, Ivesia gordonii. Phlox pul-
vinata, Poletnonium viscosum, Senecio amplec-
tens, and S. canus. The endemic Draba sobo-
lifera frequently flowers in gravel at the edge of
receding snowbanks. Isolated 'hanging' patches
of Geum rossii turf are sometimes present,
indicating possible recent erosion of surround-

ing material. GravclK barrens usualK inter-
grade into dry meadow or talus/scree commu-
nities over relatively short distances.

Bedrock. — Plant growth on exposures of
bedrock is restricted to rock crevices and ledges
where pockets of soil have accumulated. Species
present include many of those present on sim-
ilar aspects in surrounding dry and mesic
meadow communities. A unique assemblage
of species that occasionally occurs on north-
facing exposures includes A)ien\isia scopulonmi,
Carex heteroneura, Cystopteris jragilis, Erigeron
compositus, Geum rossii, Oxyria digyna,
Saxifraga caespitosa, S. debilis, and Stellaria
longipes.

Talus/scree. — Colluvial deposits are most
common in the area composed of the Mount
Belknap Volcanics and along the glaciated por-
tions of the main ridgecrest. Arenaria nuttallii,
Cerastium beeringianum, Erigeron compositus,
and Polemonium viscosum are commonly pres-
ent on all exposures while Artemisia scopulonmi,
Geum rossii, and Primula parryi are more com-
mon on more mesic north- and east-facing
slopes within this community type.

The Flora

The alpine flora of the Tushar Mountains
consists of 171 species from 102 genera and 34
families. The largest families are Asteraceae
(29 species), Poaceae (20), Brassicaceae (13),
Rosaceae (12), Cyperaceae (11), Caryoph\'lla-
ceae (10), Fabaceae (8), Ranunculaceae (7),
and Scrophulariaceae (7). The largest genera
are Carex, Poa, and Potentilla with 11, 8, and 7
species, respectively, while Saxifraga and
Senecio are each represented by five species.
Bromus inermis and Taraxacum ojficinale are
the only introduced species occurring above
timberline. The species list is presented near
the end of this paper

Thirteen taxa appear to be restricted to the
alpine area: Astragalus australis van glabrius-
culus, Carex elynoides, C haydeniana, C. nar-
dina, Claytonia inegarJiiza, Hyincnoxys grandi-
flora. Lychnis apetala van kingii, Poa patter-
sonii, Potentilla concinna, Salix arctica,
Saxifraga caespitosa, Townsendia condensata,
and Valeriana acutdoba.

Three taxa (1.8% of the alpine flora) are en-
demic to high elevations in the Tushar Moun-
tains. Draba sobolifera and Senecio castoreus
are most common in gravelly barren and talus/

1995]

TusHAR Mountains Alpine Flora

229

scree communities above timberline while
Cirsium eatonii var. harrisonii is most common
on subalpine talus/scree slopes. Other Utah
endemics found in the alpine are Agoseris
glauca var. cronquistii. Astragalus perianus,
CastiUeja parvula var parviila, Gilia fridactyhL
and Lesquerella icarclii.

The perennial herb life form accounts for
86.4% of the indigenous alpine flora. This fig-
ure includes 143 species of angiosperms (110
dicots and 33 monocots), one spikemoss, and
two ferns. Ten species of shrubs (5.9% of the
flora) are present (two of which are gymno-
sperms). The remaining 13 indigenous taxa are
considered to be annual or biennial to short-
lived perennials. Only 1.8% of the flora {Cheno-
podiiun atrovirens, Gentianella tenella, and
Polygonum douglasii) is classified as strictly
annual though Spira (1987) reports Gentianella
tenella to be strictly biennial in the alpine of
the White Mountains, CA. Perennial herbs
increase in importance at higher elevations
and comprise 94.6% of the species (53 of 56
taxa — with exceptions being Androsace scpten-
trionalis, Draba crassifolia, and Salix arctica)
known to occur in the area of 0.6 km^ above
the elevation of 3596 m on Delano Peak. A
similar life form composition is reported for
the alpine flora of the Teton Range (Spence
and Shaw 1981).

Species richness and habitat diversity are
greatest in the vicinity of Delano Peak be-
cause of this peak's geologic substrate, glacial
history, and elevation. Erosion of the Bullion
Canyon Volcanics has produced a mostly
plateau-like topography conducive to soil for-
mation and associated meadow communities.
The northern and eastern slopes of Delano
Peak, though glaciated, are relatively gentle as
compared to the cliff-like glacial headwalls
present along much of the main ridgecrest; per-
sistent snowdrifts (sometimes lasting through-
out the summer), which are necessary for the
growth of some alpine species (Billings 1978)
and which provide moisture to lower eleva-
tions, are thus able to form on these less-inso-
lated, high-elevation, leeward slopes. All eight
types of plant communities and a minimum of
101 species (59.8% of the indigenous alpine
flora) are known to occur within a radius of 1.0
km of the summit within an area of 3.14 km^
(16.3% of the alpine area).

The northern portion of the alpine region
composed of the weathering-resistant Mount

Belknap Volcanics is floristically poor despite
the presence of the second and third highest
peaks; no vascular plants were obsei'ved above
the elevation of 3536 m on Mount Belknap. A
depauperate alpine flora of about 65 species
occurs on the ridgecrest cushion plant com-
munities, block slopes, and in the talus/scree
and gravelly barren communities and small
patches of mesic meadow that occur on the
ridges and flanks of these summits; Crypto-
gramma crispa and Poa pattersonii apparently
occur in the alpine only on this fonnation, how-
ever Soil formation and plant growth on this
substrate may be hindered by unfavorable
nutrient availability as occurs locally in hydro-
thermally altered, highly acidic exposures at
the base of the range (Salisbuiy 1964).

Plant Geography

The Tushar Mountains are located on the
western margin of a floristically similar high-
land region known as the Southern Rocky
Mountains. This area, which includes most of
Colorado and parts of adjacent states, contains
the greatest concentration of alpine tundra in
the United States outside of Alaska (Weber
1965). One hundred fifty-five of the 169 indige-
nous alpine species of the Tushar Mountains are
also reported by Weber and Wittmann (1992)
for the flora of Colorado.

Statistical comparison with 14 neighboring
alpine floras shows the Tushar alpine flora to
be most similar to the adjacent northerly floras
of the Wasatch and Uinta ranges of Utah and
the Teton Range, Wyoming, with Sorensen's
similarity indices of 52.8, 50.2, and 48.8%,
respectively (Table 1). The relatively continu-
ous "Teton-Wasatch-High Plateau mainland
mountain system" (Harper et al. 1978), which
is perhaps best illustrated as an elevated (2000
m and higher elevation) corridor in Figure 19
in Reveal (1979) over which direct migration
of alpine species may have occurred during
glacial times (Billings 1978) and which has
previously been noted to be a migration route
for Utah's boreal species (Harper et al. 1978,
Welsh 1978, Reveal 1979, and Welsh 1993),
has thus likely been a primary source area for
development of the alpine flora of the Tushar
Mountains. In particular, Calyptridium wnhel-
latum, Cymopterus hendersonii, Synfhris pin-
natifida, and Townsendia condensata appear to
have migrated to the Tushars via this north-to-

230

Great Basin Natuiulist

[Volume 55

Table I. Moristic siinilarit\ indices hi'twecii tlu' alpine flora oi tlic Tnshar Nlonntains, UT, and representative nei^li-
boring alpine floras. The index of similarity used is that of Sorensen (Miieller-Donil)ois and Ellenherg 1974). Mainland
area floras arc part of a relatively continuous mountain s\'stem such as the Teton-Wasatch-High Plateau system in contrast
to the more isolated mountain floras of the Great Basin and portions of the Colorado Plateau (Harper et al. 1978).

Number of

Alpine-to-alpine

indigenous

dist

ance from

Percent

Flora^

alpii

lie species

Tnshar Mts

(km)

similarity

M.MNLA.ND ARKAS

1. Wasatch Mountains, UT

202

157

52.8

2. Uinta Mountains, UT

257

269

50.2

3. San Juan .Mountains, CO

250

410

44.9

4. Sawatch Range, CO

285

507

45.4

5. Teton Range, VVT

216

573

48.8

6. Indian Peaks area, CO

249

596

42.1

7. Pioneer Mountains, ID

130

600

36.8

S. Sangre de Cristo Mountains, NM

157

627

40.5

9. Beartooth Plateau, WT-MT

185

750

36.2

Mountain Islands

10. Henr\' Mountains, UT

136

32.4

11. Snake Range, NV

171

25.5

12. Deep Creek Mountains, UT

198

51.2

13. San Francisco Peaks, AZ

332

44.6

14. Ruby Mountains, NV

150

345

42.0

^Alpine floras are from tlie following sources; (1) ,\rnow ct al. (1980) and \ouclier specimens from Allred (1975) and Collins (1980); (2) Lewis (1970), Goodrich and
Neese (1986), and Goodrich (1994); (3) Webber et al. (1976) and Hartman and Rottman (1985); (4) Hartman and Rottman (1988); (5) Spence and Shaw (1981); (6)
Komarkova (1979); (7) Moseley and Bematas (1992); (8) Baker (1983); (9) Johnson and Billings (1962); (10) Neese (1981); (11) Lewis (1973); (12) McMillan (1948);
(13) Schaack (1983) and Schaack and Morefield (1985); (14) Loope (1969) and Lewis (1971).

south route, inasmuch as they occur in west-
ern Wyoming (Dorn 1988) but are unreported
from Colorado (Weber and Wittmann 1992). A
total of 158 of the indigenous Tushar alpine
species are reported by Dorn (1988) for the
flora of Wyoming. The alpine flora of the
Tushar Mountains is more similar to that of
the Teton Range, Wyoming, than to any of the
compared Colorado alpine floras despite a
greater distance of up to 160 km (Table 1).
Pro.ximity along the same migrational pathway
thus appears to be an important factor in
floristic similarity.

The apparent effectiveness of the Teton-
Wasatch-High Plateau migration route is fur-
ther illustrated by 13 boreal species occuning in
the alpine of the Tushars which are apparently
at a southern margin of distribution within the
longitudes of Utah: Astrafialus ati.stralis, Carex
nardina. Lychnis apetala, and Salix arctica are
arctic species (Polunin 1959) not known to
occur further south in Utah (Albee et al. 1988)
or in adjacent Arizona (Lehr 1978). A total of
44 alpine species from the Tushars (26.0% of
the indigenous alpine flora) are reported by
Polunin (1959) as also occurring in the Arctic.
Other alpine species at an apparent southern
margin of distribution are Antcwiaria cdpinci

var media. Arenaria mittalUi, Ccdijptridiwn um-
beUatinih ClunnaerJwdos erccta, Claytonia >neg-
orhiza, Hymenoxys grandiflora, Poa pattersonii,
Saxifraga adscendens, and Townsendio con-
demata. Nonalpine boreal species at a southern
margin of distribution in this range include
Arnica diversifolia. Aster engelmannii, Carex
deweyana, C. hoodii, C. hizidina, Draba lance-
olata, Hieracium graciJe, Leucopoa kingii,
Microseris nutans, and Mitella pentandra.

Though migration of high-ele\'ation species
has occurred between the Colorado Rockies
and the La Sal Mountains of southeastern Utah
(Holmgren 1972, Welsh 1993), significant migra-
tion of alpine species fijrther west to the Tushar
Mountains has perhaps been limited by an
area of relatively low elevation termed the
"Colorado Plateau migrational barrier" by
Hadley (1987). The isolated Hemy Mountains,
located midway between the Tushar and La
Sal ranges (Fig. 1), have a meager alpine flora
of 47 species; absent there are common alpine
species such as Geum rossii, Oxyria digyna.
Polygonum bistortoides, and Silcne acaulis
(Neese 1981). These and other alpine species
may have been eliminated from the Henry
Mountains by the warmer post-glacial hyp-
sithermal chmate (Neese 1981), however, thus

1995]

TusHAR Mountains Alpine Flora

231

masking the true effectiveness of the Colorado
Plateau as a migrational barrier to high-eleva-
tion species.

Species richness, which is strongly corre-
lated with area on mountains (Harper et al.
1978, Hadley 1987), also appears to affect flor-
istic similarity as the Tushar alpine flora gen-
erally has higher indices of similarity with the
larger and generally more distant mainland
floras (Table 1); Harper et al. (1978) note that
the isolated mountain floras of the Inter-
mountain West have fewer species per unit area
than adjacent mainlands and also an uneven
stocking of species as a result of greater ran-
domness of colonization and/or extinction. The
isolated alpine floras of the east central Great
Basin to the west of the Tushars (Loope 1969),
the San Francisco Peaks to the south (Moore
1965), and the Henry Mountains to the east
(Neese 1981) are slightly to extremely depau-
perate examples of the Rocky Mountains alpine
flora.

The Tushar alpine flora is also slightly de-
pauperate in comparison with most other
neighboring mainland area floras (Table 1);
this is likely due to the limited alpine area (in
comparison, the Uinta Mountains have an
alpine area of about 1000 km^ [Lewis 1970]),
scarcity of wet meadows and rivulets, and
presence of the talus-forming Mount Belknap
Formation. The smaller Tushar alpine flora
may be a factor in the relatively low maximum
similarity index of 52.8% with the Wasatch
Mountains; Hartmann and Rottman (1988)
report a similarity index range of 72.5-73.3%
between the larger alpine floras in Colorado.

The alpine flora and vegetation of the Tushar
Mountains are remarkably diverse given the
relatively small alpine area. Interesting, too,
are the number of endemic taxa and species
that reach a southern limit of distribution here.
The wide-ranging alpine species Claytonia
megarhiza, Poa pattersonii, and Saxifraga
adscendens are disjunct here with other in-
state distributions only in the Uinta and La Sal
ranges, while Townsendia condensato occurs
nowhere else in the state (Albee et al. 1988,
Welsh et al. 1993).

Alpine environments are in general fragile
and easily susceptible to disturbance (Billings
1973). This fragility is locally compounded by
poor soil-forming characteristics of some igne-
ous members and by the questionable intro-
duction of Rocky Mountain goats to the range

in 1986. There is evidence these animals feed
on the endemic CastiUeja parvula, and they
endanger the species diversity of the alpine
area by grazing at scarce alpine wet sites.

Acknowledgments

I am grateful to Dn Stanley L. Welsh, curator
of the herbarium at Brigham Young University,
for his assistance and encouragement with this
study initiated as part of a graduate program.
Kaye Thorne, assistant curator of the herbari-
um, provided sustained aid in herbarium
research. Dr. Wesley B. Niles gave helpful
comments on an earlier version of the manu-
script, and Dr. Kimball T Harper graciously
loaned me pertinent references from his per-
sonal library. Ray Wilson of the Soil Conser-
vation Service, Salt Lake City, provided data
on climate for the area, and Mike Smith of the
U.S. Forest Service office in Richfield, UT,
provided information on soils. Information on
plant specimens from the Tushars was provid-
ed by the following individuals: Linda Allen,
assistant curator of the Intermountain Herb-
arium at Utah State University; Dr. Patricia K.
Holmgren, director of the herbarium at The
New York Botanical Garden; Ann Kelsey, cura-
torial assistant at the Garrett Herbarium,
University of Utah; and Tim Ross, senior cura-
torial assistant at the Rancho Santa Ana
Botanic Garden. This checklist would be less
complete without the efforts of early botanists
such as Marcus E. Jones and Drs. Walter R
Cottam, Bertrand F. Harrison, and Bassett
Maguire as well as the more recent prolific
collecting by Dr. Stanley L. Welsh, Dr. N.
Duane Atwood, Mont E. Lewis, and Joel Tuhy

Annotated List of Vascular Plants

The following list of families, genera, and
species is arranged alphabetically within the
divisions of Cronquist et al. (1972). Community
type(s) and maximum elevation noted for each
taxon are based on field notes and herbarium
specimen label information. The following
abbreviations are used for community types:
cushion plant (CP), dry meadow (DM), mesic
meadow (MM), wet meadow (WM), rivulet
(RI), gravelly barren (GB), bedrock (BR), and
talus/scree (TS). Frequency of occurrence for
most taxa is estimated using the following
scale from Thorne (1967): rare, 1-3 collections
or observation stations; infrequent, 4-7 sta-

232

Great Basin Naturalist

[Volume 55

tions; frequent, 8-12 stations; connnon, 13 +
stations. Life form is listed as a = annual, ab
= annual or biennial, ap = annual to short-
lived perennial, bp = biennial to perennial, p =
perennial herb, and s = shrub. Speeies that
also oeeur in the Arctic (Polunin 1959) are fol-
lowed by an asterisk (*).

I collected Botrychiuin lunaria, Junctis
mertensianiis, Pediciilaris groenlandica, and
Salix planifolia at a seep below local timber-
line at an elevation of 3389 m, and Draba
lanccolata has been collected at timberline
(Welsh et al. 14015). These and other taxa may
eventually be discovered from the alpine area.
Erigeron humilis and Taraxacum ceratophorum
have recently been reported for the Tushars
(Cronquist 1994), but I have seen no speci-
mens.

Division Lycopodiophyta
Selaginellaceae

Selaginella watsonii Underw.; rock crevices in CP,
DM. MM, BR. and TS to 3658 m; common; p.

Division Polypodiophyta
Polypodiaceae

Crijptogramma crispa (L.) R. Br. var. acrostichoides
(R. Br.) C. B. Clarke; TS to 3304 m; rare; p.

Cystopteris fragilis (L.) Bernh.; rock crevices in DM,
MM, HI, BR, and TS to 3.505 m; frequent; p.*

Division Pinophyta
Cupressaceae

Juniperiis communis L. var. depressa Pursh; DM at

3444 m in slielter of houlder on south-facing slope; rare;
s.*

Pinaceae

Picea engelmannii Parry; DM, MM, and TS to 3566 m;
common; s.

Division Magnoliophyta

Class Magnoliopsida

Apiaceae

CymopteruH hendersonii (Coult. & Rose) Cronq.; CR
15 K, and TS to 3627 m; frequent; p.

CymopteruH lemmonii (Coult. & Rose) Dorn [Pseudo-
cymopterus montanus (Gray) Coult. & Rose]; CR DM,

MM, RI, and TS to 3700 m; common; p.

Asteraceae

Achillea millefolium L. ssp. hmulosa (Nutt.) Piper;
DM, MM, and RI to .3548 m; connnon; p.*

Agoseris aurantiaca (Hook.) Greene var. purpurea
(Gray) Cronq.; MM(?) to ca 3505 m; rare; p.

Agoseris glauca (Pursh) Raf. var. cronquistii Welsh;
DM to 3353 m; infretjuent; p.

Agoseris glauca (Pursh) Raf. var. dasycephala (T. &

G.) Jepson; (^P to 3414 m; rare; p.

Anteiinaria alpina (L.) Gaertner var media (Greene)
Jepson [A. media Greene]; CR MM, and BR to 3487 m;
connnon; p.

Antennaria microphylla Rydb.; DM to 3536 m; rare; p.

Arnica mollis Hook.; BR/MM to 3444 ni; rare; p.

Artemisia frigida VVilld.; CP and DM to 3505 m; rare;
s.*

Artemisia ludoviciana VVilld. var. incompta (Nutt.)
Cronq.; (JR DM, MNL and RI to 3475 m; common; p.

Artemisia scopulorum Gray; MM, WM, RI, BR, and
TS to 3703 ni; common; p.

Cirsium eatonii (Gray) Robins, var. harrisonii Welsh;
TS to 3444 m; rare; p.

Crepis nana Richards.; CR GB, and TS to 3475 m; fre-
quent; p.*

Erigeron compositus Pursh var. glahratus Macoun;
CR DM, GB, BR, and TS to 3706 m; common; p.*

Erigeron simplex Greene; MM to 3700 m; fiequent; p.

Erigeron speciosits (Lindl.) DC. var. uintahensis
(Cronq.) Welsh [£. uintahensis Cronq.]; DM to 3414 m;
rare; p.

Erigeron ursinus D. C. Eaton; CR MM, and RI to 3536
ni; common; p.

Haplopappus dementis (Rydb.) Blake; MM and GB to
3578 m; common; p.

Haplopappus macronema Gray; CR DM, GB, and TS
to 3.536 ni; connnon; s.

Helenium hoopesii Gray [Dugaldia hoopesii (Gray)
Rydb.]; DM and MM to 3566 m; common; p.

Hymenopappus filifolius Hook. var. nudipes (Maguire)
Turner; DM and GB to 3561 m; infrecjuent; p.

Hymenoxys grandiflora (T. & G.) Parker; "grassy tun-
dra above timberline at 350.5 m; rare; p. The only record
from the range is K. E Parker et al. 6354 at the Rancho
Santa Ana Botanic Garden.

Senecio amplectens Gray var. hohnii (Greene) Har-
rington; MM, GB, and BR to 3700 m; common; p.

Setiecio canus Hook.; DM and GB to 3609 m; com-
mon; p.

Senecio castoreus Welsh; CR GB, and TS to 3536 m;
infrequent; p.

Senecio eremophilus Richards, var. kingii (Rydb.)
Greenman; DM and MM to 3536 m; infrequent; p.

Senecio werneriaefolius (Gray) Gray; TS to 3505 m;
frequent; p.

Solidago multiradiata Ait.; DM, MM, GB, and BR to
3700 ni; common; p.*

Solidago parryi (Gray) Greene [Haplopappus parryi
Gray]; MM to 3505 m; infrequent; p.

Taraxacum officinale Weber; DM and TS to 3536 m;
infrequent; introduced p.

Townsendia condensata D. C. Eaton; CR and GB at
3.50.5-3609 m; infrequent; p.

Boraginaceae

Mertensia arizonica Greene; DM, MM, and RI to

3505 m; common; p.

Brassicaceae

Arahis drummondii Gray; DM to 3414 m; infrequent;
I,p.

Arahis lemmonii Wats.; MM and BR to 3402 m; rare; p.

Cardamine cordifolia Gray; RI to 3444 m; infrequent
(localh common); p.

1995]

TusHAR Mountains Alpine Flora

233

Descurainia richardsonii (Sweet) Schuiz van brevipes
(Nutt.) Welsh & Reveal; RI and TS to 3475 m; infre-
quent; ah.

Draba aiirea Vahl; MM and BR to 3688 m; rare; p.*

Draba crassifolia Graham; MM and RI to 3700 m; fre-
quent; ap.*

Draba sobolifera Rydb.; MM, GB, BR, and TS to 3688
111; common; p.

Draba stenoloba Ledeb.; MM to 3505 m; rare; ap.*

Erysimum asperum (Nutt.) DC.; DM to 3441 m; rare;
bp.

Lesquerella wardii Wats.; DM and GB to 3609 m; fre-
quent; p.

Physaria chambersii Rollins van chambersii; GB to
3414 ni; rare; p.

Smelowskia cahjcina C. A. Mey. van americana (Regel
& Herder) Drury & Rollins; CH DM, MM, BR, and TS
to 3703 111; common; p.*

Thlaspi montaniitn L. van montanum; Cf! MM, and TS
to 3475 m; common; p.

Caprifoliaceae

Sambucus racemosa L. van microbotrys (Rydb.)
Kearney & Peebles; DM and TS to 3444 m; infrequent; s.

Car)'ophyllaceae

Arenaria mittallii Pax; CP MM, GB, and TS to 3505 m;
common; p.

Arenaria obtiisiloba (Rydb.) Fern.; MM and WM to
3676 m; common; p.*

Arenaria rubella (Wahl.) J. E. Sm.; CR DM, MM, and
RI to 3688 m; frequent; p.*

Cerastium beeringianum C. & S.; CR DM, MM, WM,
BR, and TS to 3700 m; common; p.*

Lychnis apetala L. van kingii (Wats.) Welsh [L. kingii
Wats.]; CR DM, and MM at 3536-3688 m; frequent; p.*

Lychnis drummondii (Hook.) Wats.; DM, MM, and
BR to 3487 m; frequent; p.

Sagina saginoides (L.) Britt.; MM and RI to 3414 m;
rare; hp.*

Silene acaulis L. van subacaulescens (F. Williams)
Fern. & St. John; CR MM, WM, BR, and TS to 3676 m;
common; p.*

Stellaria longipes Goldie; DM, MM, WM, BR, and TS
to 3627 111; common; p.*

Stellaria umbellata Turcz.; MM, RI, and TS to 3615 m;
frequent; p.

Chenopodiaceae

Chenopodiiim atrovirens Rydb.; DM in disturbed soil
(pocket gophers?) at 3548 m; rare; a.

Crassulaceae

Sedum rhodanthum Gray; MM and WM to 3627 m;
infrequent; p.

Fabaceae

Astragalus australis Fisch. van glabriuscultis (Hook.)
Isely [A. aboriginum Richards.]; CF and GB at
3505-3609 m; infrequent; p.*

Astragalus miser Dougl. van oblongifolius (Rydb.)
Cronq.; DM, MM, and GB to 3706 m; common; p.

Astragalus periamis Barneby; DM and GB to 3566 m;
infrequent; p.

Lupinus argenteus Pursh van rubricaulis (Greene)
Welsh; DM to 3463 m; rare; p.

Lupinus lepidus Dougl. van utahensis (Wats.) C. L.
Hitchc. [L. caespitosus Nutt. van utahensis (Wats.) B.
Cox]; DM, MM, and GB to 3572 m; fiequent; p.

Oxytropis oreophila Gray van oreophila; CR DM,
MM, and GB to 3706 ni; common; p.

Oxytropis parryi Gray; DM and MM to 3633 m; infre-
quent; p.

Trifolium longipes Nutt. var. rusbyi (Greene)
Harrington; MM to 3597 m; frequent; p.

Gentianaceae

Gentiana parryi Engelni.; DM to 3389 m; rare; p.

Gentianella amarella (L.) Borner; DM and MM to
3535 m; fretiuent; ah.*

Gentianella tenella (Rottb.) Borner; MM and WM to
3566 m; rare; a (h?).*

Swertia radiata (Kellogg) Kuntze [Frasera speciosa
Dougl.]; MM and TS to 3475 m; rare; p.

Grossulariaceae

Ribes cereum Dougl.; DM and BR to 3536 m; fre-
quent; s.

Ribes inerme Rydb.; DM (among rocks) and TS to
3438 m; rare; s.

Ribes montigenum McClatchie; DM, MM, RI, and TS
to 3627 m; common; s.

Hydrophyllaceae

Phacelia hastata Dougl.; DM in gravelly soil to 3444
m; rare; p.

Phacelia sericea (Graham) Gray var. ciliosa Rydb.;

DM to 3475 m; rare; p.

Lamiaceae
Monardella odoratissima Benth.; TS to 3475 m; rare; p.
Linaceae

Linum perenne L. ssp. lewisii (Pursh) Hulten; MM in

gravelly soil at 3536 m; rare; p.*

Onagraceae

Epilobium angustifolium L.; TS to 3414 m; rare; p.*
Epilobium halleaniun Hausskn.; RI to 3444 m; rare

(locally common); p.

Epilobium saximontanum Hausskn.; RI to 3487 m;

rare (locally common); p.

Polemoniaceae

Gilia tridactyla Rydb.; CR and TS to 3414 m; rare; p.

Phlox pulvinata (Wherry) Cronq.; CR DM, MM, GB,
BR, and TS to 3706 in; common; p.

Polemonium pulcherrimum Hook. var. delicatum
(Rydb.) Cronq.; DM and MM to 3444 m; infrequent; p.*

Polemonium viscosum Nutt.; DM, MM, GB, BR, and
TS to 3633 m; common; p.

Polygonaceae

Eriogontim umbellatum Toit. var. porteri (Small) Stokes;

DM, MM, and BR to 3566 m; frequent; p.

Oxyria digyna (L.) Hill; MM, GB, BR, and TS to 3658
m; common; p.*

234

Grkat Basin Natufi\list

[Volume 55

Polygonum histortoides Pursh; MM, \\M, and HI to
3676 111; commoii: p.

Polygonum doiighisii Greene var. douglasii., DM to
3444 111; rare; a.

Rumex salicifolius Weinm. ssp. triangulivalvis Danser;
DM, MM, RI, and BR to 3499 in; frequent; p.

Portulacaceae

Calyptridium umhellatum (Torr.) Greene var. caudi-
cifera Griiy, MM and i'Ai to 3536 ni; infre(iiient; ap.

Claytonia megarhiza (Gray) Parry; BR and TS at 3475
to 3615 m; rare; p.

Lewisia pygmaea (Gray) Robins.; MM and Rl to 3597
ni; frefjiient; p.

Primulaceae

Androsace septentrionalis L.; DM, MM, RI, and TS to
3700 m; eoinmon; ab.*

Primula parryi Gray; MM, WM, RI, BR, and TS to

3658 m; common; p.

Ranunculaceae

Anemone multifida Poir.; CP and MM to 3487 m; rare; p.

Aquilegia scopuJorum Tidestr.; TS to 3438 m; infre-
quent; p. A.S noted in \\elsh et al. (1993), some specimens
are completeK' transitional with A. caendea James.

Caltha leptosepala DC. var. leptosepaloj MM, WM,
and RI to 3627 m; frecjuent; p.

Delphinium occidentale (Wats.) Wats. var. barbeyi
(Huth) Welsh [D. barbeyi (Huth) Huth]; DM, RI, and TS
to 3475 m; comnion; p.

Ranunculus eschscholtzii Schlect.; TS to ca 3490 m;
rare; p.

Ranunculus inamoenus Greene; DM and RI to 3597
m; common; p.

Thalictrum fendleri Engelm.; DM (in shelter n( Ril)es
niontigciUDiij to 3414 m; rare; p.

Rosaceae

Chamaerhodos erecta Bunge var. parviflora (Nutt.)
C. L. Hitchc.; CP and DM to 3505 m; rare; lip.

Geum rossii (R. Br.) Sen var. turbinatum (Rydb.) C. L.
Hitchc; CR DM, MM, WM, RI, GB, TS, and BR to 3700
m; common; p.*

Ivesia gordonii (Hook.) T. & G.; DM and GB to 3609
m; infrequent; p.

Potentilla concinna Richards, var. proxima (Rydb.)
Welsh & Johnston; DM and TS at 3353 to 3536 in; infre-
quent; p.

Potentilla diversifolia Lehm. var. diversifolia; WM,
RI, and TS to 3627 m; frecjuent; p.

Potentilla glandulosa Lindl. var. intermedia (Rydb.)
C. L. Hitchc; DM, MM, and TS to 3487 in; common; p.

Potentilla gracilis Dougl. var. pulcherrima (Lehm.)
Fern.; DM to 3463 m; frequent; p.

Potentilla hippiana Lehm.; DM to 3414 m; p.

Potentilla ovina Macoun var. decurrens (Wats.) Welsh
& Johnston; (>P and DM to 3475 ni; infreciuent; p.

Potentilla pensylvanica L. var. pensylvanica; CP DM,
MM, and TS to 3700 ni; common; p.*

Ritbus idaeus L. ssp. melanolasius (Dieck) Focke.; TS
to 3414 in; rare; s.

Sibbaldia procutnbens L.; MM, RI, and BR to 3627 m;
common; p.*

Salicaceae

Salix arctica Pallas var. petraea Anderss.; MM, WM,

and RI at 3444 to 3676 m; frecjuent (locally common); s.*

Sa.xifiagaceae

Heuchera rubescens Torr. var. rubescens; BR to 3444
m; rare; p.

Saxifraga adscendens L. var. oregonensis (Raf.)
Breitung; MM (ainony; rocks) to 3676 m; rare; p.

Saxifraga caespitosa L. var. minima Blake; MM, WM,
and BR at 3566 to 3676 in; infrequent; p.*

Saxifraga debilis Engelm.; MM and BR to 3658 m;
common; p.

Saxifraga odontoloma Piper; RI to 3444 m; rare; p.

Saxifraga rhomboidea Greene; DM, MM, WM, and
RI to 3700 m; common; p.

Scrophulariaceae

Castilleja miniata Dougl.; DM to 3535 m; infrequent; p.

Castilleja parvida Rydb. var. parvtda; DM and MM to
3688 111; common; p.

Mimulus tilingii Regel; RI to 3414 m; rare; p.

Pedicularis parryi Gray var. parryi; MM, WM, and RI
to 3627 m; common; p.

Penstemon whippleanus Gray; MM and BR to 3450 m;
frecjuent; p.

Synthris pinnatiftda Wats. var. laciniata Gray; DM,
MM, WM, RI, and BR to 3627 m; common; p.

Veronica ivormskjoldii R. & S.; MM and RI to 3487 m;
rare; p.*

Valerianaceae

Valeriana acutiloba Rydb.; DM and MM at 3414 to
3567 m; infrecjuent; p.

Valeriana edulis Nutt.; CR DM, and MM to 3599 m;
infrecjuent; p.

Valeriana occidentalis Heller; DM to 3353 m; rare; p.

Violaceae

Viola canadensis L.; BR and TS to 3444 m; rare; p.
Viola nuttallii Pursh; DM to 3414 m; rare; p.

Class Liliopsida
Cyperaceae

Carex albonigra Mack.; CR and MM to ca 3658 m;
infrecjuent; p.

Carex ebenea Rydb.; RI to 3444 m; rare; p.

Carex egglestonii Mack.; DM to 3414 m; rare; p.

Carex elynoides H. T. Holm; CR DM, MM, and TS at
3353 to 3706 m; common; p.

Carex haydeniana Olney; MM, RI, GB, and BR at
3414 to 3566 in; common; p.

Carex heteroneura W. Boott var. chalciolepis (H. T.
Holm) F. Hermann; the intergrading var epapillosa F.
Hermann also occurs in the range though perhaps not in
the alpine; MM and BR to ca 3658 m; common; p.

Carex tnicroptera Mack.; DM (?) to 3414 m; rare; p.

Carex nardina Fries; MM at 3505 m; rare; p.*

Carex nova Baile> ; unknown community at ca 3505 m;
rare; p.

Carex phaeocephala Piper; CR DM, MM, and GB to
3566 m; common; p.

Carex rossii F. Boott; DM (?) to ca 3353 m; rare; p.

1995]

TusHAR Mountains Alpine Flora

235

Juncaceae

Jimcus drttmmondii E. Mey.; RI to 3444 ni; rare; p.
Luzula spicata (L.) DC; MM, WM, and BR to 3627
m; coinnion; p.*

Liliaceae

Zigademis elegans Pursh; MM, WM, and RI to 3536
m; infrequent; p.*

Poaceae

Agrostis variabilis Rydb.; MM to 3383 m; rare; p.

Bromus ciliatus L.; MM to 3414 ni; rare; p.

Bromiis inermis Leysser; roadside adjacent to MM at
3487 m; rare; introduced p.

Calamagrostis piirpiirascens R. Br.; TS to 3414 m; rare;
p.*

Deschampsia cespitosa (L.) Beauv.; MM and RI to
3499 ni; infrequent; p.*

Elymus ehjmoides (Raf.) Swezey [Sitanion hystrix
(Nutt.) J. G. Sm.]; DM to ca 3505 m; rare; p.

Elymus scribneri (Vasey) Jones [Agropyron scribneri
Vasey]; DM, GB, and TS to 3578 m; common; p.

Elymus trachycaulus (Link) Gould [Agropyron tra-
chycauhim (Link) Malte]; DM and NLM to 3566 m; com-
mon; p.*

Festuca ovina L. van brevifolia (R. Br.) Wats.; CR DM,
MM. GB, BR, and TS to 3706 m; common; p.*

Phleum alpimim L.; MM and RI to 3487 m; frequent; p.*

Poa arctica R. Br.; CP, MM, WM, BR, and TS to 3700
m; frecjuent; p.*

Poa fendleriana (Steudel) Vasey; DM to 3383 m; fie-
quent; p.

Poa glauca Vahl [P. glauca ssp. rupicola (Nash) W. A.
Weber; P interior Rydb.]; CE DM, MM, GB, and TS to
3536 m; common; p.*

Poa nervosa (Hook.) Vasey; TS to 3414 m; infrequent; p.

Poa pattersonii Vasey; TS at 3505 m; rare; p.

Poa pratensis L.; MM to 3444 m; rare?; possibly intro-
duced p.*

Poa reflexa Vasey & Scribn.; MM and RI to 3536 m;
common; p.

Poa secunda Presl [P. sandbergii Vasey]; CP DM, and
TS to 3475 m; frequent; p.

Stipa lettermanii Vasey; DM and MM to 3475 m; fre-
quent; p.

Trisetum spicatum (L.) Richter; CP, MM, BR, and TS
to 3700 m; common; p.*

Literature Cited

Albee, B. J., L. M. Shultz, and S. Goodrich. 1988. Atlas
of the vascular plants of Utah. Utah Museum of
Natural Histon; Salt Lake Cit>'. 670 pp.

Allred, K. W 1975. Timpanogos flora. Unpublished tlie-
sis, Brigham Young University, Provo, UT. 178 pp.

Arnow, L., B. Albee, and A. Wyckofe 1980. Flora of the
central Wasatch Front, Utah. 2d edition revised.
Universit>' of Utah Printing Sendee, Salt Lake Gib,'.
663 pp.

Baker, W L. 1983. Alpine vegetation of Wheeler Peak,
New Me.xico, U.S.A.: gradient analysis, classification,
and biogeography. Arctic and Alpine Research 15:
223-240.

Billings, W D. 1973. Arctic and alpine vegetations: simi-
larities, differences, and susceptibility- to disturbance.
Bio Science 23: 697-704.

. 1978. Alpine phytogeography across the Great

Basin. Great Basin Naturalist Memoirs 2: 105-117.
1988. Alpine vegetation. Pages 391-420 in M. G.

Barbour and W. D. Billings, editors. North American
terrestrial vegetation. Cambridge University Press,
Cambridge.

Bliss, L. C. 1985. Alpine. Pages 44-65 in B. F Chabot and
H. A. Mooney, editors. Physiological ecology of North
American plant communities. Chapman and Hall,
New York.

Callaghan, E. 1973. Mineral resource potential of Piute
County, Utah, and adjoining area. Utah Geological
and Mineralogical Survey Bulletin 102. Utah Geo-
logical Suwey, Salt Lake Cit\'. 135 pp.

Collins, P D. 1980. Comparative life histoiy and floral
characteristics of desert and montane plant commu-
nities in Utah. Unpublished thesis, Brigham Young
University, Provo, UT. 168 pp.

Cronquist, a. 1994. Intermountain flora. Volume 5:
Asterales. The New York Botanical Garden, Bron.x.
496 pp.

Cronquist, A., A. H. Holmgren, N. H. Hol.mgren, and
J. L. Reveal. 1972. Intermountain flora. Volume 1.
Hafner Publishing Co., New York. 270 pp.

Cunningham, C. G., and T. A. Steven. 1979. Mount
Belknap and Red Hills calderas and associated rocks,
Marysvale volcanic field, west-central Utah. U.S.
Geological Suney Bulletin 1468. 34 pp.

Cunningham, C. G., T. A. Steven, E D. Rowley, L. B.
Gl,\ssgold, and J. J. Anderson. 1983. Geologic map
of the Tushar Mountains and adjoining areas, Marys-
vale volcanic field. U.S. Geological Survey miscella-
neous investigations series map I-1430-A. U.S. Geo-
logical Sui-vey, Denver, CO.

DORN, R. D. 1988. Vascular plants of Wyoming. VIountain
West Publishing, Cheyenne, WY. 340 pp.

Goodrich, S. 1994. Written communication. Ashley
National Forest, 355 North Venial Ave., Vernal, UT
84078.

Goodrich, S., and E. Neese. 1986. Uinta Basin flora. U.S.
Department of Agriculture, Forest Service — Inter-
mountain Region, Ogden, UT. 320 pp.

Hadley, K. S. 1987. Vascular alpine plant distributions with-
in the central and southern Rock-y Mountains, U.S.A.
Arctic and Alpine Research 19: 242-251.

Harper, K. T, D. C. Freeman, W K. Ostler, and L. G.
Klikofe 1978. The flora of Great Basin mountain
ranges: diversity, sources, and dispersal ecology.
Great Basin Naturalist Memoirs 2: 81-103.

Hart.man, E. L., and M. L. Rottm.a.n. 1985. The alpine
vascular flora of three cirque basins in the San Juan
Mountains, Colorado. Madrono 32: 253-272.

. 1988. The vegetation and alpine vascular flora of

the Sawatch Range, Colorado. Madrono 35: 202-225.

Holmgren, N. H. 1972. Plant geograhy of the Inter-
mountain region. Pages 77-161 in A. Cronquist, A. H.
Holmgren, \. H. Holmgren, and J. L. Reveal, editors,
Intermountain flora. Volume 1. Hafner Publishing
Co., New York.

Hunt, C. B. 1987. Physiography of western Utah. Pages
1-29 in R. S. Kopp and R. E. Cohenour, editors,
Cenozoic geology of western Utah — sites for pre-
cious metal and hydrocarbon accumulations. Utah
Geological Association Publication 16. Utah Geo-
logical Association, Salt Lake Cit>'.

Johnson, P L., and W D. Billings. 1962. The alpine
vegetation of the Beartooth Plateau in relation to

236

Ghkat Basin Natur.\list

[Volume 55

cryopedo^fiiic ])r()ct'sses and patterns. Iscolouical
Monographs 32: 105-135.

KOMARKOVA, V. 1979. Alpine vegetation of the Inchan Peaks
area. Front Range, Colorado Rocky Monntains.
Strauss & Cramer, Hirschberg II. Cerniany. 650 pp.

Leiik, J. H. 1978. A catalogue of the flora of Arizona.
Desert Botanical Garden, Phoeni.x, AZ. 203 pp.

Lewi.s, M. E. 1970. Alpine rangelands of the Uinta Moun-
tains, Ashley and Wasatch National Forests, Region 4.
U.S. Department of Agriculture, Forest Service,
Ogden, UT 75 pp.

. 1971. Flora and major plant communities of the

liiib\-East Ilumholdt Mountains with special empha-
sis on Lamoille Canyon. Report to Humboldt National
Forest, Elko, NV 62 pp.

. 1973. Wheeler Peak area species list. Report to

Intermountain Region, U.S. Forest Service, Ogden,

UT. 17 pp.
LooPE, L. L. 1969. Subalpine and alpine vegetation of

northeastern Nevada. Unpublished doctoral disser-
tation, Duke University, Durham, NC. 287 pp.
McMlLLA.N, C. 1948. A taxonomic and ecological study of

the flora of the Deep Creek Mountains of central

western Utiih. Unpublished thesis, Universit\' of Utah,

Salt Lake City. 99 pp.
McjORE, T. C. 1965. Origin and disjunction of the alpine

tundra flora on San Francisco Mountain, Arizona.

Ecology 46: 860-864.
MOSELEY, R. K., AND S. Bernatas. 1992. Vascular flora of

Kane Lake circjue. Pioneer Mountains, Idaho. Great

Basin Naturalist 52: 335-343.
MuELLER-DoMBOis, D., AND H. Ellenberg. 1974. Aims

and methods of vegetation ecology. John Wiley &

Sons, New York. 547 pp.
Neese, E. J. 1981. A vascular flora of the Heni-y Mountains,

Utah. Unpublished dissertation, Brigham Young Uni-
versity, Provo, UT. 370 pp.
POLUNIN, N. 1959. Circumpolar Arctic flora. O.xford

University Press, London. 514 pp.
Re\ EAL, J. L. 1979. Biogeography of the Intermountain

region: a speculative appraisal. Mentzelia 4: 1-92.
Salisbury, F B. 1964. Soil formation and vegetation on

hydrothermally altered rock material in Utah. Ecol-
ogy 45: 1-9.
S(:h.\.'\ck, C. G. 1983. The alpine vascular flora of Arizona.

Madroiio 30: 79-88.
Scii,\ack, C. G., and J. D. Morefield. 1985. Noteworthy

collections: Arizona. Madroiio 32: 121-122.
Smouse, F a., and K. D. Gurgel. 1981. ElevaHon. Page 18

in D. C. Greer, K. D. Gurgel, W. L. Wahlquist, H. A.

Christy, and G. B. Peterson, editors. Atlas of Utah.

Weber State College, Odgen, UT. [Printed at Brigham

Young University Press, Provo, UT]

S(ML CoN.SKRX.vnoN Skrvige. 1993. SNOTEL. Snow Sur-
vey Section, Salt l^ake City, UT.

Spence, J. R., and R. J. Sh.wv. 1981. A checklist of the
alpine vascular flora of the Teton Range, Wyoming,
with notes on biology and habitat preferences. Great
Basin Naturalist 41; 232-242.

Spia^, T P 1987. .Alpine annual plant species in the White
Mountains of eastern California. Madrono 34:
315-323.

Steven, T A., P D. Rowley, and C. G. Cunnlngham.
1984. Calderas of the Marysvale volcanic field, west
central Utah. Join^nal of Geophvsical Research 89,
No. BIO: 8751-8764.

T.\ye, a. C. 1994. Annotated checklist of the vascular
plants of the Tushar Mountains, Utah. Unpublished
data. Herbarium, M. L. Bean Life Science Museum,
Brigham Young University', Provo, UT 84602.

Thorne, R. F 1967. A flora of Santa Catalina Island,
California. Aliso 6(3): 1-77.

U.S. Forest Service. 1993. Tushar-Pa\ant-Canyon soil sur-
vey report (preliminaiy unpublished data). Fish Lake
National Forest, Richfield, UT.

Washburn, A. L. 1979. Geocryology, a survey of peri-
glacial processes and environments. Wiley, New York.
406 pp.

Webber, P J., J. C. Emerick, D. C. Ebert May, and V.
KoviARKON A. 1976. The impact of increased snowfall
on alpine vegetation. Pages 201-264 in H. W. Steinhoff
and J. D. Ives, editors. Ecological impacts of snow-
pack augmentation in the San Juan Mountains of
Colorado. Colorado State University, Fort Collins.

Weber, W. A. 1965. Plant geography in the southern
Rocky Mountains. Pages 453-468 in H. E. Wright
and D. G. Frey, editors. The Quaternary in the United
States. Princeton University Press, Princeton, NJ.

Weber, W A., and R. C. Wittmann. 1992. Catalog of the
Colorado flora: a biodiversity baseline. University
Press of Colorado, Niwot. 215 pp.

Welsh, S. L. 1978. Problems in plant endemism on the
Colorado Plateau. Great Basin Naturalist Memoirs 2:
191-195.

. 1993. Description of the environment. Pages 6-10

in S. L. Welsh, N. D. Atwood, L. C. Higgins, and S.
Goodrich, editors, A Utah flora. Print Services,
Brigham Young University, Provo, UT.

Welsh, S. L., N. D. Atwood, S. Goodrich, and L. C.
Higgins. 1993. A Utah flora. 2d edition revised. Print
Senices, Brigham Young Universit>', Provo, UT. 986
pp.

Received 1 September 1994
Accepted 7 November 1994

ECOLOGY OF CELTIS RETICULATA IN IDAHO

Ann Marie DeBoltl and Bruce McCune-

Abstract. — The small deciduous tree Celtis reticulata (netleaf hackberry) reaches its northern limit in Idaho, where,
contraiy to most of its western range, it often occurs as an overstorv' dominant. Two hundred fifty stands of this tree were
sampled throughout Idaho. Celtis is slow-growing, averaging 4 m tall at 50 \'r, and long-lived (to 300-400 yr). It occurs in
a variety of habitats, from riparian to rocky uplands. Trees grow best where topographically sheltered, such as in draws
and narrow canyons, and where soils are loamy. Although plants grow more slowly as surface rock cover increases,
stands are often associated with rock, with a mean surface cover of 39% rock. Differences in growth rates were unrelat-
ed to parent material and aspect. Most stands are reproducing, in spite of habitat degradation caused by overgrazing,
alien plant invasion, and increasing fire frequencies. Stands are typically represented by one dominant cohort; however,
young, even-aged stands are rare and are generally found along watenvays on stream terraces or at the high-water line.
Although slow-growing, C. reticulata shows promise for land managers interested in site enhancement. This native
species is long-lived, produces fruit used by wildlife, and provides structural diversity in a semiarid landscape (with a
maximum height of 12 m) in areas that are becoming increasingly dominated by e.xotic plant species.

Key words: Celtis reticulata, netleaf hackbernj, ecology, Idaho, growth, longevity, stand structure, recruitment, site
characteristics, livestock grazing, rehabilitation.

Celtis reticulata Torr. (netleaf hackberry,
western hackberry) is a deciduous shrub to
small tree in the elm family (Ulmaceae), wide-
ly distributed in semiarid regions of the west-
ern United States (Fig. 1). It occurs in a diver-
sity of habitats, including deciduous riparian
woodlands, mountain shrub, wash scrub, and
live oak-mixed shrub communities, in rocky
canyons, and as scattered individuals in semi-
desert grasslands, pinyon-juniper and Joshua
tree woodlands (Glinski 1977, Plummer 1977,
Brown 1982, Albee et al. 1988). Its elevational
range is from 200 to 2000 m (Elias 1980).
Populations are often small or highly localized
(Daubenmire 1970, Dooley and Collins 1984),
particularly at the northerly latitudes in the
states of Oregon, Washington, and Idaho (Eliot
1938). Despite its broad distribution, little is
known about the plant's ecology, presumably
due to its position as a minor component in
many of its habitats, and its fragmented occur-
rence (Peattie 1953, Lanner 1983).

While C. reticulata is sparsely distributed
in Idaho, near its northern limit (Fig. 2), it
appears to exhibit wide ecological tolerances.
However, it tends toward the warmest portions
of canyons, especially southerly aspects (Tisdale
1986). It is a member of both riparian and
upland communities in Idaho, where it can

occur as a locally abundant, overstory dominant
(Huschle 1975, Johnson and Simon 1987).
Along the Wiley Reach of the middle Snake
River, it forms narrow, but extensive, gallery
forests of nearly monospecific stands (Bowler
1981). On steep shoreline escarpments of the
lower reaches of the Snake River, in the
"Douglas" hackberry vegetation type described
by Huschle (1975), it forms a dense, nearly
closed canopy. On the gentle shoreline slopes,
alluvial fans, and colluvial cones of the lower
Snake River, it grows in an open savanna
(Daubenmire 1970, Huschle 1975). "Open sa-
vanna" is perhaps the best way to describe the
appearance of a typical Celtis community on an
upland site in Idaho, where individuals occur
in a random or clumped pattern with exten-
sive areas of grassland between.

Plants produce a small, fleshy drupe in the
fall, favored by a variety of birds and mammals
(Hayward 1948, Lanner 1983, C. A. Johnson
1990, personal communication). With as many
as 41 species of birds associated with Celtis
communities in Idaho, the tree's importance for
wildlife cannot be overemphasized (Asherin
and Claar 1976). It provides cover for a variety
of big game species, including mule deer and
bighorn sheep (Asherin and Claar 1976), as
well as much-sought-after shade for domestic

'Bureau of Land Management, 3948 Development Avenue, Boise, ID 8.370.5.

^Department of Botany and Plant Pathology, Oregon State University, Conallis, OR 97331-2902.

237

238

Great Basin Naturalist

[Volume 55

»" INTERMITTENT DISTRIBUTION
— CONTINUOUS DISTRIBUTION

Fig. 1. Global distiilnition of Celtis reticulata (revised
from Little 1976).

livestock along the Snake River (Daubenmire
1970).

Due to an apparent tolerance of haish, water-
stressed growing conditions, a strong potential
to resprout following disturbances such as fire
and herbivory, and its high wildlife values,
public land managers are interested in using
C. reticulata to rehabilitate disturbed habitats.
However, we must know more of the growth
rate, longevity, stand structure, and ecological
tolerances of the species to properly evaluate
its potential in site enhancement or rehabilita-
tion projects.

This study sought to answer the following
questions: (1) What are the growth rates and
longevities of C. reticulata, and do they differ
with aspect, parent material, soil texture, per-
cent surface rock cover, topographic position,
topographic shelter, and grazing intensity of a
stand? (2) How does the size class structure of
C reticulata stands differ with the environ-
mental parameters listed above? Is the species
reproducing in Idaho, and does recruitment
diftei- under different environmental conditions?
(3) Are environmental conditions related to
differences in growth form of the plant (i.e.,
the formation of single vs. multiple stems)?

Fig. 2. Idaho distribution of Celtis reticulata.

Methods
Field Methods

Two hundred thirty stands spread over much
of tlie Idaho range of C. reticulata were sampled
in 1990 and 1991. Approximately 20 stands on
the west side of the Snake River, in adjacent
Oregon and Washington, were also sampled
(total N = 250). Stands were selected based on
within-site homogeneity of apparent history,
topography, and parent material, and a mini-
mum population size of six individuals (many
more individuals were usually present). With
these constraints for homogeneity, the sam-
pling areas were t\'pically irregularly shaped
and small, usually less than 0.25 ha. Stands
were chosen to represent a range of sites and
disturbance histories.

Stands were assigned to topographic posi-
tions (Table 1) that included river tenace, high-
water line, draw, rocky draw, bench, toe slope,
lower slope, broken lower slope, mid-slope,
upper slope, and talus. The 11 categories were
narrowly defined on the assumption that

1995]

Celtis reticulata in Idaho

239

Table 1. Definitions of topographic positions in which
Celtis reticulata was sampled.

River terrace Relatively flat horizontal sintace cut or

built b\' river or stream action
High-water line Transition line between flood-tolerant

and -intolerant plant species
Draw Shallow incision in a slope, with <30%

total smface rock cover
Rocky draw Shallow incision in a slope, with

>30% total surface rock cover
Bench Nearly level surface usually well above

active floodplains and terraces
Toe slope Gently inclined, basal part of a slope

continuum that grades to the valley,

usually <14° slope
Lower slope Lower 1/3 of a hillside (above the toe

slope, when present); if steep (> 14°)

and toe slope absent, the basal part of

the slope that meets the valley floor
Broken lower Similar to lower slope but with
slope extensive smface cover of large

boulders and outcrops
Mid-slope Middle 1/3 of a hillside, relative to the

surrounding landscape
Upper slope Upper 1/3 of a hillside, relative to the

surrounding landscape
Talus slope Coarse, angular rock fragments derived

from and King at the base of cliffs or

rock slopes; slopes typically >25°

combining them at a later time, if needed, would
be possible. Based on field observation and
reconnaissance, the number of stands sampled
within each topographic position was approxi-
mately proportionate to how frequently those
topographic positions were occupied by the
species. Stand-level data recorded, in addition
to topographic position, included elevation;
latitude; longitude; aspect; slope; percent sur-
face rock cover; surface soil texture; parent
material; topographic shelter, grazing intensi-
ty, total stand density; density within four
structural classes, including seedling, juvenile,
mature, and decadent individuals; number of
cohort modes; and associated dominant plant
species (explained below).

Surface soil textures were evaluated by moist-
ening in the field according to the Soil Conser-
vation Service "Guide for Textiu-al Classificaton"
(Brady 1974). When soils were unreachable due
to surface rock, the surface rock matrix was
classified instead. For example, stands on talus
slopes had soils categorized as "talus. "

Six categories of parent material were iden-
tified initially, including granite, sandstone,
basalt, river alluvium, rhyolite, and oolitic lime-
stone. However, because of the small sample

size of rhyolite (4) and its chemical similarity
to granite, the two were combined for analy-
sis. A similar situation existed for oolitic lime-
stone, an uncommon and geographically
restricted coarse-grained rock that typically
occurred as a lens within sandstone-dominated
strata. Therefore, the eight stands on oolitic
limestone were combined with sandstone for
analysis.

Each stand was categorized by "topograph-
ic shelter": open (0), intermediate (1), and shel-
tered (2). For example, exposed stands grow-
ing within a valley were classified as "interme-
diate," while stands growing in a side canyon of
the same valley were classified as "sheltered."
"Open" stands were those with unobstructed
exposure to solar radiation. They were typically
not associated with a major, incised drainage;
rather, they faced broad, expansive valleys.

To evaluate recruitment and growth of C.
reticulata under different livestock grazing
pressures, we scored grazing intensity within
a stand as none to moderate (1) or extreme (2).
Stands scored as extreme were recognized by
(1) heavy browsing of trees, with a hedged or
"pasture-tree" growth form; (2) elimination of
vegetation under trees by trampling; (3) tree
roots exposed by soil compaction and erosion;
and (4) dominance of alien plant species.
Thirty-six of the 250 stands were classified as
extreme.

The overall density of Celtis stands was cat-
egorized as (1) widely scattered [mature indi-
viduals more than 10 crown widths apart]; (2)
scattered [mature individuals separated by
gaps of 4-10 individual crown widths]; (3) sub-
continuous [breaks in the total canopy exist but
mature individuals average no more than 3
crown widths apart]; or (4) continuous [little
open space in the canopy; crowns form a con-
tinuous matrix with occasional gaps]. Inter-
mediate sites were recognized with a mid-
point value (e.g., 3.5 for stands approaching a
closed canopy).

To evaluate the composition of C. reticulata
stands, densities in four structural classes were
also estimated in a similar fashion. The four
structural classes were defined as follows: (1)
seedling [individual of the year and < 2 yr
old]; (2) juvenile [individual >2 yr old and
< 1.5 m tall]; (3) mature [>1.5 m tall]; and (4)
decadent [>1.5 m tall and experiencing signif-
icant dieback, i.e., at least one major dead
branch present].

240

Great Basin Naturalist

[Volume 55

Within each stand at least three indixidu-
als, chosen to represent the modal size in the
stand, were measured and aj^cd. Modal size
was defined as typical size of individuals in the
dominant (most abundant) cohort. Measure-
ments recorded for each tree included height,
age, diameter at core height (typically 20 cm
above ground level), number of live and dead
stems, and percent rock cover below the can-
opy as centered over the main trunk. When
two or three modal sizes were present, all
modes were sampled for a minimum total of
either six or nine individuals. When stands
were all-aged with no apparent modal tree
size, at least six individuals of the dominant
canopy cohort were sampled. The number of
modes present, from 1 to 4, with 4 equivalent
to an all-aged stand, was recorded as a stand-
level variable. Most height measurements
were obtained with an 8-m, extendable level
rod. For taller trees, height was determined
with a clinometer.

Increment cores were taken at the same
height the diameter was measured (20 cm).
Cores were transported in plastic straws, glued
onto slotted boards, sanded, and annual growth
rings were counted under a dissecting micro-
scope. When cores did not reach the tree s
center (i.e., because of rot), the number of
missing years was extrapolated by first sub-
tracting the length of the core from the tree's
ladius. This remainder was multiplied by the
number of rings counted in the core's inner
centimeter, which was then added to the num-
ber of rings counted for an estimate of the
total age. When cores were off-center, the miss-
ing radius was estimated by overlaying a clear
transparency with a series of circles of known
radii over the core, matching the ring pattern
in the core with a circle, and multiplying its
radius by the number of rings in the centime-
ter nearest the core's center This amount was
added to the number of counted years to esti-
mate tree age. Small-diameter noncoreable
individuals were cut down and a cross section
was removed, sanded, and the rings counted
as above.

Analytical Methods

Stands were not included in the analysis if
the sample size within a particular topograph-
ic position or parent material was too small, or
if the majority of cores from a stand were illeg-
ible after sanding due to contortions in the

radial growth. Nine stands were dropped, for
a final sample size of 241. SPSS (1988) was used
for all analyses.

A heat load index was generated to account
for differences in heat load from nortlieast- to
southwest-facing slopes (Whittaker 1960, Muir
and Lotan 1985). For each stand, index values
were calculated with the following e(|uation,
where 0 = aspect in radians east of north:
heat load = (1 - cos(0 - 7r/4))/2. Index values
ranged from 0 (NE slopes) to 1 (SW slopes).

To compare C. reticulata growth rate and
stand structure differences under various en-
vironmental conditions, we developed 50-yr
site indices as measures of growth potential
(i.e., site quality), as outlined in Husch et al.
(1972). Site index is based on average heights
of dominant trees at a specified index age
(usually 50 or 100 yr) and is the most widely
used method of evaluating site quality for tree
growth (Husch et al. 1972, Daubenmire 1976).
Site index curves are constructed to allow for
estimation of site index for stands older or
younger than the index age, as index age
stands are seldom encountered (Husch et al.
1972).

The commonly used relationship of tree
height to age formed the basis for one index,
and the relationship of tree diameter to age
formed the basis for the second (DeBolt 1992).
The best linear fit was achieved when log
(height, m) and log (diameter, cm) were
regressed on the log of tree age (fi- = .25, R^
= .54, respectively; N = 939). The resulting
equations were log (height) = 0.428 X log
(age) - 0.135 and log (diameter) = 0.764 X log
(age) - 0.165. Using these two equations, we
obtained the expected (mean) height and
diameter at 50 yr, then back-transformed to
improve inteipretabilit\', yielding an expected
size at 50 yr of 3.9 m tall and 13.6 cm in diam-
eter

For each tree in the data set, the site index
was calculated by first finding its residual from
the regression line, then shifting this residual
to the 5()->'r point on the line, which fields an
estimated height and diameter at 50 >'r Thus,
the equations to calculate site index (SI) for
each tree were;

Log (height SI) = 0.591 +
(LOGheight - ((0.428 x LOGage) -0.135))

Log (diameter SI) = 1.134 +
(LOGdiam - ((0.764 x LOGage) -0.165))

1995]

Celtis reticulata in Idaho

241

To analyze stioictural class differences under
differing environmental conditions, the vari-
able TYPE, representing types of stand struc-
ture, was created. Based on the density of
juvenile, mature, and decadent size classes in
a stand, the five TYPEs were defined as fol-
lows: (1) young (juvenile); (2) mature, nonre-
producing, nondecadent; (3) mature, repro-
ducing, nondecadent; (4) mature, reproducing,
decadent; and (5) mature, nonreproducing,
decadent (Table 2).

Based on field obsei-vations, mortality of C
reticulata seedlings during year one is extreme-
ly high. Because most seedlings were year-
lings, seedlings were not used to define TYPE.
Stands were classified as reproducing when
the juvenile density class was 1 or greater (i.e.,
> 5 individuals).

Celtis reticulata growth rate, expressed by
site indices, was analyzed as the dependent vari-
able in one-way analyses of variance (ANOVA)
against the environmental parameters topo-
graphic position, parent material, soil texture,
grazing intensity, and topographic shelter.
Relationships between site indices and ordered
categorical independent variables were ana-
lyzed by linear regression. With few excep-
tions, height site index was a more sensitive
predictor of growth differences than diameter
site index. Celtis reticulata growth rates and
relationships with topographic position and
other environmental parameters were also
analyzed with analysis of covariance, to com-
bine categorical and continuous factors.
Included in the model was the categorical vari-
able topographic position, with soil texture,
topographic shelter, grazing intensity, and par-
ent material as four covariates. Relationships

between environmental variables and stand
structure (TYPE) and the number of modes
were analyzed bv contingency tables and
ANOVA.

Results
Growth

Log-log regressions best represented the
statistical relationship between height and age
(Fig. 3) and diameter and age of C. reticulata
individuals. An initial impression that regres-
sion lines do not fit the scatter of points at
log(age) <1.2 can be reconciled by recognizing
that the dense central elliptical clouds of points
have controlled the regression results. In both
cases the least-squares fit resulted in a good fit
to the dense cloud of points representing mid-
dle-aged trees, but resulted in almost entirely
negative residuals for trees younger than
10-25 yr. Because these younger trees were
from a small number of sites, many of which
showed battering by floods, distributions of
residuals were judged to be acceptable.

Celtis reticulata diameter and height were
tightly related in a log-log regression {R~ =
.75). Mean height and diameter of dominant
and codominant C. reticulata, regardless of
age, were 5 m and 18 cm, respectively. While
diameter is a better predictor of age than
height (R- = .53 and .25, respectively), height
is more responsive to site characteristics than
is diameter, both in the literature and in this
study. Thus, height was the preferred basis for
the site index.

Fifty-year-old C. reticulata trees in Idaho
averaged 3.9 m tall and 13.6 cm in diameter.
Using height, we constructed site index curves

Table 2. Categorization of the Celtis reticulata stand structure variable TYPE. TYPE represents the five types of stand
structure that were recognized from the density classification. Within each stand, the three size classes of trees (juve-
nile, mature, decadent) were assigned to a density' class based on the following definitions. Mid-point values were used
as needed. Juvenile: (1) wideK' scattered — 5 or fewer juveniles present; (2) scattered — >.5 juveniles present in a nonag-
gregated distribution averaging > 10 canopies apart; (3) subcontinuous — breaks in the total canopy exist but juveniles
average >3 and <10 canopies apart. Mature/Decadent: (1) widely scattered — mature individuals >10 crown widths
apart; (2) scattered — mature individuals separated by gaps of >4 and <10 individual crown widths; (3) subcontinuous —
breaks in the total canopy exist but mature individuals average < 3 crown widths apart; (4) continuous — mature trees
form a continuous matrix with only occasional gaps.

TY'PE Description
1 Yoim

Density cl

ass

Juvenile

Mature

Decadent

< .5

< 2

<.5

>.5

< 2

>.5

HMIH

Nonreproducing, nondecadent
Reproducing, nondecadent
Reproducing, decadent
Nonreproducing, decadent

242

Great Basin Natuiulist

[Vokinie 55

Site index
' Height at 50 years

100 200 300

AGE. years

400

Fig. 3. Nontransformed log-log regression of Celtis
reticulata height (m) on age and site index cui-ves for the
Idaho stands.

for Idaho Celtis stands to allow site classifica-
tion for a stand at any age (Fig. 3). Site quality
of an area can be assessed by determining
average height and age of dominant trees and
locating the position of these coordinates on
the site index graph. The stand's site index is
then read from the closest curve.

Site quality, as expressed by the height-
based site index, differed among the eleven
topographic positions identified {P = .0001, F
= 4.4) (Table 3). However, variation within
topographic positions was large, so that at the
.05 significance level, only draws differed
from any other specific topographic position.
Growth was faster in draws than on talus
slopes, upper slopes, mid-slopes, and stream
terraces.

Although site index means did not differ
statistically between most topographic positions,
a relatively predictable biological ranking of
topographic positions was expressed, with a
general trend of faster growth where sheltered
and mesic to slower growth on more xeric and
exposed sites. For example, site index values
were smallest on talus slopes, followed by upper
slopes, mid-slopes, and stream tenaces (Table 3).

Celtis reticulata occurred infrequently on
north- and east-facing slopes (Fig. 4A). Twenty-
five percent (60) of stands were found on SW
slopes, with a heat load between 0.95 and
1.00, the hottest values of the heat load index;
58% (140) were between 0.74 and 1.00. Only

32 stands (13%) were located on the coolest
sites between 0.00 and 0.20, or between 350°
and 98° east of north. The mean heat load index
was 0.69. No stands were found between 349°
and 9° east of north.

In spite of C. reticulata s affinity for souther-
ly exposures, heat load was not a good predictor
of hackbeny growth characteristics. More often
than not, stands have an affinity for southerly
exposures, but because of topographic shelter-
ing, growing conditions are often not as harsh
or water stressed as they first appear. Of 241
Celtis stands, 168 (70%) had at least an inter-
mediate topographic shelter.

In a stepwise regression analysis from a
pool of six independent variables (soil texture,
rock, grazing intensity, shelter, heat load, and
slope), shelter was the most important predictor
of site index (R2 = .13, p < .001, F = 35.5).
Site index values were largest when shelter
was greatest, with well-sheltered stands differ-
ing from intermediate and open exposures
(Table 4). However, variability in growth rates
within a given class of shelter is large, as
shown by the low R^.

Presence of C. reticulata is correlated with
surface rock or rock outcrops. Of the 241 stands
sampled, 96 (40%) had a surface rock cover of
50% or more (Fig. 4B). Twenty percent of the
stands were extremely rock-y, with rock cover-
ing 75-98% of the ground surface. Average
rock cover was 39%.

A weak, inverse relationship between per-
cent surface rock cover and site index was
found (fi2 = -.28, P = .0001). As rock cover in-
creased, site index tended to decrease slightly.
Rock was a statistically significant variable in a
stepwise multiple regression as well, following
topographic shelter in order of entry.
Including rock in the model increased the R^
value from .13 to .20 (F = 28.9, P < .001). On
sites classified as draws, where topographic
shelter is maximized, surface rock cover is less
important.

Neither parent material nor grazing inten-
sity was a statistically significant predictor of
site index (F = .43 and .14, respectively). How-
ever, site index values differed with soil tex-
ture (F = .023, F = 2.07). As with topographic
position (Table 3), means were ranked by
Fisher's LSD procedure in an intuitively pre-
dictable order. Growth rates were higher on
finer-textured soils (clay or loam) than on
coarse-textured soils (sand). At alpha = .05, the

1995]

Celtis reticulata in Idaho

243

Table 3. Site index values of Celtis reticulata (s — standard desiation) for each topographic position. Mean site index
(SI) vakies have been transformed back into the original scale of measurement to aid interpretation. Topographic posi-
tions with no overlap of similarit\' grouping letters are different from each other at the .05 significance level (Fisher's
LSD).

Topographic
position

Mean SI:
transformed

Mean SI:
back-
transformed

Similarity
grouping

Draw

High-water line
Toe slope
Rock\' draw
Lower slope
Bench

Broken lower slope
Stream terrace
Mid-slope
Upper slope
Talus slope

0.74 (0.16)

5.5

0.65 (0.15)

4.4

0.61 (0.14)

4.1

0.58 (0.17)

3.8

0.57 (0.19)

3.7

0.56 (0.24)

3.6

0.55 (0.12)

3.5

0.51 (0.31)

3.2

0.50 (0.18)

3.2

0.48 (0.20)

3.0

0.47 (0.13)

2.9

only pairs that differed from each other were
talus and loam.

Interactions between soil te.xture and topo-
graphic position were highly significant (Chi-
square, P = .001). When the analysis of site
index and soil texture was restricted to just
upland sites, the effect was slightly more pro-
nounced (F = .014, F = 2.49).

Growth Form

"Shrubbiness" was quantified by counting
the number of live and dead main stems or
trunks of each individual. Regression analysis
of stem number with the variables grazing
intensity, topographic shelter, soil texture, heat
load, slope, average height, average diameter,
and percent surface rock cover produced sev-
eral statistically significant, albeit weak, rela-
tionships. Live and dead stem density per
individual decreased as topographic shelter
increased {R^ = .20 and .30, respectively).
Average height decreased slightly as the num-
ber of live stems increased {R- = .20). In gen-
eral, on sheltered sites C. reticulata has a sin-
gle stem (treelike) rather than multi-stem
(shrublike) growth form.

Differences in plant growtli form were found
among topographic positions and among par-
ent materials. Individuals growing at mid-slope
were generally shrubbier, with a greater num-
ber of live stems (.T = 2.5), than individuals
growing at high-water line {x = 1.4), in draws
{x = 1.6), and in rocky draws {x = 1.5)
(ANOVA, P = .003, F = 2.71). Dead stems
were far less numerous than live stems and

were absent from most individuals. The num-
ber of dead stems at mid-slope (x = 0.6) was
greater than all other topographic positions ex-
cept upper slopes (P = .0001, F = 6.5). Stands
at high-water line, rock-y draw, stream terrace,
draw, and broken lower slope topographic
positions averaged only 0.1 dead stems per
individual. Growth form did not differ with
the number of size modes within a stand.

Individuals on sandstone were more com-
monly multi-stemmed than those on the three
other parent materials, for both living and
dead stems (F < .001, F = 8.5; F < .001, F =
14.7, respectively).

Longevity

The mean age of individuals sampled dur-
ing our study was 66 yr, with a range of 1-374
yr (Fig. 5). Old age and large size are not tight-
ly related. For example, it is common to find
trees 10 m tall but less than 75 yr old.
Diameter was often a better predictor of age
than was height (F- = .54 and .26, respective-
ly, after log-log transformation).

The oldest C. reticulata recorded in this study
(about 374 yr) grew on an exposed talus slope
approximately 300 m above the Salmon River;
it was 4.6 m tall and 48 cm in diameter at 20
cm above ground level. Percent surface rock
cover of the site was 90%, with the small stand
of scattered trees restricted to talus margins
where pockets of soil were exposed. Other
members of the stand ranged in age from 191
yr (3.35 m tall, 28 cm diam) to 320 yr (5.48 m
tall, 46.5 cm diam).

244

Great Basin Naturalist

[Volume 55

-inOi

LIl

-10

-0.2 0.0 0.2 0.4 0.6 0

HEATLOAD

J i_

1.0 1.2

E7i 30 -

20 40 60 80 100 120

PERCENT SURFACE ROCK

Fig. 4. Frequenc\' distrihutions of the niiml:)er of Celtis reticulata .stands by (A) heat load and (B) percent surface rock-
cover.

Stand Structure

Of the 241 Celiis stands, 178 (74%) were re-
producing and only 4 (1.7%) of diese were deca-
dent. Fifty-seven stands (23.5%) were classified
as nonreproducing, 6 (2.5%) of them decadent.
The remaining 6 stands (2.5%) were recently
established (juvenile dominated), with no
mature individuals present.

Structure of C. reticulata stands, in terms of
their relative densities of juvenile and mature
size classes, was unrelated to soil texture (Chi-
square, P = .31). Structural type was weakly
related to topographic position of the stand
(Chi-square, P = .08). Of the 11 topographic
positions, rocky draws had the highest juve-
nile density, or recruitment. In general, juve-
nile densities increased as the percent of sur-
face rock cover increased. Density of C. retic-
ulata juveniles was highest when rock cover
was 50% or greater

Rocky draws consistent])' had the densest
canopies, followed by draws and high-water

line. Rocky draws were never assigned an over-
all density <2.5, where 3 = subcontinuous. In
fact, 75% of rocky draws had closed or nearly
closed canopies (overall density = 3.5 or 4).
Juveniles were often present on the margins of
rocky draws.

The few decadent stands were found higher
on the slope, on steeper slopes, and in less-
sheltered positions than nondecadent stands.
Nonreproducing, decadent stands were more
steeply sloping than young and nondecadent,
nonreproducing stands (ANOVA, P = .003, F
= 4.03) (Table 5). Of the 10 decadent stands,
50% were at mid-slope and 20% were on talus.

While none of the decadent stands were ex-
tremely overgrazed, their distance from water
may have confounded this result. Overgrazed
stands were typically found on fairly gentle
terrain (.v = 14°, S.D. = 8.6) and in close prox-
imity to a water source, where livestock tend
to concentrate, while decadent stands were on
steeper slopes (Table 5) and at higher slope

Table 4. Mean site index (SI) vakies for Celtis reticulata for three lexels of topographic sheher, in both transformed
and back-transformed scales. Topographic shelters with no o\erlap of similarit) grouping letters are different from each
other at the .0.5 significance level (Fisher's LSD).

Topographic
shelter

Sheltered

Intermediate

E.xposed

Mean SI:

transformed

back-

Similarity

(s)

transformed

groupmg

0.69 (0.15)

4.9

0.55 (0.20)

3.6

129

0.49 (0,14)

.3.1

1995]

Celtis reticulata in Idaho

245

100

2
3

UUUU

JUuUUUUUL

0D0== o .

Table 5. Average slopes of decadent, nondecadent, and
yoinig stands of Celtis reticulata, with the variable TYPE
in its original five-categoiy format. TYPE represents the
five types of Celtis reticulata stand structure that were
recognized.

100 200

TREE AGE (years)

Fig. 5. Frequency distribution of the number of Celtis
reticulata trees by tree age.

positions. Less intensively grazed stands aver-
aged 23° (S.D. = 10.1).

Grazing level was related to stand structure
(TYPE; Chi-square, P = .0002). A larger per-
centage of heavily grazed stands (53%) were
nonreproducing than were stands with light or
moderate grazing intensity (18.5%). Even
though sample sizes were veiy different (light
or moderate = 205, extreme = 36), the pattern
confirms field observations of low recruitment
under extreme grazing pressure. However, it
is perhaps even more noteworthy that recruit-
ment on heavily grazed sites is as high as it is,
given how few, if any, other shrub species are
present on such sites.

Among the four parent materials, 37% of C.
reticulata stands growing on sandstone were
nonreproducing, as compared to 21%, 22%,
and 21% of stands growing on granite, basalt,
and river alluvium, respectively (Chi-square,
P = .014; Table 6). A greater number of sand-
stone-associated stands were nonreproducing
than expected (14 and 9, respectively), while
fewer were reproducing than expected (22 and
27, respectively). Expected and observed val-
ues for the three other parent materials were
more similar

Newly established C. reticulata stands are
apparently rare, as few were observed during
the study in spite of efforts to locate them.
Only six young (<33 yr) stands were sampled.
These were typically on rock)' sites with inter-
mediate topographic shelter and gentle slopes

TiTE

Mean slope
(degrees)

S.D.

Young (juvenile)

3.4

Nonreproducing,

nondecadent

9.6

Reproducing,

nondecadent

174

10.3

Reproducing,

decadent

7.7

Nonreproducing,

decadent

5.0

(x = 12°). All had at least 15% surface rock
cover, but most had 75% or greater rock cover
(x = 65%). Five of the six stands were on allu-
vium, including stream terraces, high-water
lines, and benches. All five had sandy soil. The
sixth stand was atypical, occurring near a mid-
slope, sparsely vegetated band of sandstone
with intermediate shelter. All individuals were
shrubby, decadent, and old (18-33 yr) relative
to the average height of 0.7 m (expected age =
8 yr). Soils were sandy loam in texture.

While young stands were only on sites with
intermediate topographic shelter, reproducing
and nonreproducing stands differed little in
the degree of shelter they received (Chi-
square, P = .06). Thirty-three percent of repro-
ducing stands were sheltered, compared to 25%
of nonreproducing stands.

The amount of surface rock differed weakly
across stand structure (TYPE; ANOVA, P =
.038, F = 2.58). Differences were greater when
the variable TYPE was restructured to three
categories (mature reproducing, mature non-
reproducing, young), eliminating decadence
as a factor (ANOVA, P = .015, F = 4.26).
Under the three-level categorization, young
stands were rockier than mature, nonrepro-
ducing stands (x = 32%) but did not differ
from those that were reproducing.

Number of Modes

Celtis reticulata stands typically appeared to
be unimodal (73%), with one dominant cohort.
Stands with two modes were far less common
(11%), but a slightly greater number were all-
aged (16%). Since only two stands had three
modes, they were dropped from analyses;

246

Great Basin Naturalist

[Volume 55

Table 6. Cross tabulation ol tlic miiiibi r oi Ccltis reticulata stands b\ stand structure and parent material. The
hypothesis of independence of stand structure and parent material is rejected with P — .014.

Observed/
expected

Mature

Row total

material

Nonreprod.

Reproil.

Younu

Row %

Granite

obs.

11.0

40.0

0,0

exp.

12.2

37.6

1.3

Sandstone

obs.

14.0

22.0

1.0

Basalt

exp.
obs.

8.8
20.0

27.2
71.0

0.9
0.0

Aihniuni

exp.
obs.

21.5
12.0

66.3
45.0

2.3
5.0

exp.

14.5

44.9

1.5

Column total

obs.

57.0

178.0

6.0

241

100

Column %

24%

74%

thus, the sample size for this portion of the
results is based on 239 stands. Growth form or
number of stems of the individuals was unre-
lated to number of modes.

Although of marginal statistical significance,
all-aged stands were more common on shel-
tered sites (Chi-square, P = .07). For example,
33% of stands in draws, which typically have
at least an intermediate topographic shelter,
were all-aged. The percent of all-aged stands
at other topographic positions ranged from 6%
to 16%.

Livestock grazing intensity may restrict entry
of new cohorts within a C. reticulata stand as
shown by the strong tendency for overgrazed
stands to be unimodal (92%; Chi-square, P =
.0008). In contrast, 70% of light to moderately
grazed stands had only one mode, 11% were
bimodal, and 18% were all-aged.

Size structure of Celtis stands did not differ
with topographic position, parent material,
soil texture, slope, percent surface rock, or
heat load (all P > .2).

Discussion

In our study, trees were typically tallest and
least shrubby when located in draws, on sites
with surface or subsurface moisture, and in areas
where they received maximum topographic
shelter. Similar observations of C. reticulata
have been recorded by others (Eliot 1938, Van
Dersal 1938, Peattie 1953), and for different
species of Celtis as well. For example, Hitch-
cock and Cronquist (1964) noted that Celtis
reticulata is taller in moist areas in the Pacific
Northwest. In Oklahoma, C. laevigata (sugar-
berry) is typically a small tree in open areas,
but in lowland forests it reaches its maximum

development (Schnell et al. 1977). On the
eastern Great Plains, C. tcnuijolia (dwarf hack-
berry) is a gnarled, shrublike tree when grow-
ing on its typical rocky, shallow, calcareous
substrate, but in the bottom of ravines it may
reach heights of 8-10 m (Stephens 1973). In
addition to the influence of an ameliorated
environment, sheltered stands may be less
prone to repeated disturbances such as fire, to
which a vegetative sprouter such as C. reticu-
lata will often respond with a shrubbier growth
form.

In Oklahoma, Celtis occurs almost exclu-
sively on loamy bottomland soils (Dooley and
Gollins 1984), and in west Texas it is best de-
veloped on alluvium (Van Auken et al. 1979).
In the canyon grasslands of Idaho, Tisdiile (1986)
recognized two types of Cc/f /.s-dominated vege-
tation on soils of two principal origins. The C.
reticulata- Agropijron spicatiim habitat type
occurs on lower valley slopes with rocky (50-
60%), weakly developed loam soils derived
from residual and colluvial materials. The sec-
ond vegetation type, unclassified because of
heavy livestock disturbance and alien plant
dominance of the understory, occurs on allu-
vial terraces with deep sandy soils.

Soil texture appears to have a greater influ-
ence on C. reticulata growth on upland sites
than on sites associated with a perennial water
source. While C. reticulata grows on a range
of soil textures in Idaho, we found the tallest
trees on loams, possibly because of their
greater water-holding capacity and nutrient
content. However, 80% of the stands occurred
on soils with some sand component, and 30%
were on sand or coarse sand. The presence of
good drainage may be an imjDortant limiting

1995]

Celtis reticulata in Idaho

247

factor for C. reticulata, as finer-textured soils
of the uplands were nearly always skeletal.
The increased percolation of sandy or skeletal
soils provides greater moisture availability for
deep-rooted shrub and tree species.

In Idaho, C. reticulata is most prevalent on
rocky sites with southeast to westerly aspects,
although heat load was not an important pre-
dictor of growth. The presence of rock, particu-
larly bedrock, may in fact be critical for hack-
beriy's existence on certain sites. It may also
partially explain the fragmented distribution
of the species in Idaho. Other rock-associated
species have been obseived in semiarid regions
as well. In the shrub-steppe region of eastern
Montana, Rumble (1987) found that scoria
rock outcrops provide a unique habitat for
several relatively mesic species. Rhus trilobata
(skunkbush sumac), Prunus virginiana (choke-
cherry), Ribes spp. (currant), and Jiiniperus
spp. (juniper) were found only in association
with rock outcrops in that ecosystem. He con-
cluded that their occunence is probably related
to protection from wind, snowdrift accumula-
tion, shading, and mulch effects of rocks.
Oppenheimer (1964) and Potter and Green
(1964) suggested that the association of mesic
species with rocky substrates is due to tempo-
rary water reservoirs that rock fissures pro-
vide. In Arizona, Johnsen (1962) reported that
Juniperus monosperma (one-seed juniper) is
largely limited to rock outcrops, where he
recorded 2-2.5 times as much available mois-
ture. The theoiy of extra moisture availability
in rock fissures could also hold true for the
deeply rooted C. reticulata, helping explain its
frequent presence on southerly aspects.

Other plausible explanations for the infre-
quency of C. reticulata on northerly aspects
and sites with less surface rock cover include
its sensitivity to late spring frosts (personal
observation) and poor competitive ability with
fast-growing species. In Idaho, C. reticulata is
the last shrub to break dormancy and expand
its leaves in the spring. This strategy, in com-
bination with the tendency to grow on warmer
slopes, generally prevents frost damage from
occurring. The greater effective soil moisture
and dense vegetative cover of north slopes
probably create an environment too competi-
tive for this slow-growing species.

In summary, Celtis reticulata can generally
be described as slow-growing and small-
statured. Fifty-year-old trees averaged 4 m tall

and 13.6 cm in diameter in Idaho, with a mean
tree height and diameter, regardless of age, of
5 m and 18 cm, respectively. Unlike some shnib
and tree species in the Intermountain West,
populations are generally maintaining them-
selves by vegetative sprouting or seedling
recruitment, despite historic and prevailing
large-scale habitat alterations resulting from
overgrazing, exotic plant invasion, and chang-
ing fire frequencies (Tisdale 1986, Whisenant
1990). Hackbeny's general resiliency and abil-
ity to resprout following disturbance or injury
likely play a role in this, as does its positive asso-
ciation with rock. Recruitment, as expressed
by the density of juvenile individuals, in-
creased as surface rock cover increased. How-
ever, under extreme grazing pressure, recruit-
ment was significantly lowered and stands
were nearly all unimodal. All-aged stands
were absent from severely grazed sites. Even
though rock favors Celtis recruitment, its
growth is favored on less-rocky sites, such as
draws.

The most likely explanation for relatively
slow C. reticulata growth on stream terraces,
in spite of the assumed availability of ground-
water, is the extreme annual fluctuation of the
water level and battering by flood debris. These
sites are located below the high-water line.
Above the high-water line the mean site index
is larger and mechanical stresses are less
extreme. While newly established C. reticulata
stands were uncommon, they typically occuired
on these riparian sites, where establishment
conditions occur more frequently than in the
uplands.

Although individuals are often slow-grow-
ing, the variation in site conditions that the
species appears to tolerate and its other posi-
tive attributes (i.e., wildlife food, cover, land-
scape structure, potential large size, tolerance
of southerly aspects), are favorable qualities
for those seeking rehabilitation species. The
species' persistence in heavily degraded
ecosystems may speak to its suitability for
rehabilitation projects as well.

Acknowledgments

This study was funded in part by the Boise
District Office of the Bureau of Land Manage-
ment, witli additional support provided by Idaho
Power Company. Nancy Shaw, Ed Tisdale,
and Steve Monsen provided insight and

248

Great Basin Naturalist

[Volume 55

encouragement during the earliest phase of
the research. Roger Rosentreter assisted in the
field and pro\'ided helpful suggestions and
encouragement throughout the study s dura-
tion. We thank Patricia Muir, Boone Kauffinan,
and Kermit Cromack for their valuable com-
ments on an early version of the manuscript.
Thanks arc also due to Stanle>' D. Smith, Sherel
Goodrich, and an anonymous reviewer tor
their constructive review of the manuscript.

Literature Cited

ALBEE, B. J., L. M. SCHULTZ, AND S. GOODRICH. 1988. Atlas
of the vascular plants of Utah. Utah Museum of
Natural History, Occasional Publication No. 7. Salt
Lake City, UT 670 pp.

Asheri.n, D. a., and J. J. Claar. 1976. Inventoiy of riparian
habitats and associated wildlife along the Columbia
and Snake rivers. Volume 3A. College of Forestry,
Wildlife, and Range Sciences, University of Idaho,
Moscow. 556 pp.

Bowler, P A. 1981. Natural histoiy studies and an evalua-
tion for eligibility of the Wiley Reach of the Snake
Ri\er for National Natural Landmark designation.
Unpublished report, National Park Service, Seattle,
WA. 86 pp.

Brady, N. C. 1974. The nature and property of soils.
Macmillan Publishing Co., Inc., New York. 639 pp.

Brown, D. E. 1982. Great Basin montane scrubland.
Pages 8.3-84 in D. E. Brov\n, editor, Biotic commu-
nities of the American Southwest — United States
and Mexico. Desert plants. Volume 4.

Daubenmire, R. 1970. Steppe vegetation of Washington.
Agriculture E.xperiment Station Technical Bulletin
No. 62. Pullman, WA. 131 pp.

. 1976. The use of vegetation in assessing the pro-
ductivity of forest lands. Botanical Review 42:
11.5-143.'

DeBolt, a. M. 1992. The ecology of Celt is reticulata Torr.
(netleaf hackberi-y) in Idaho. Unpublished master's
thesis, Oregon State Universit>', Coi-vallis. 161 pp.

DooLEY, K. L., AND S. L. COLLINS. 1984. Ordination and
classification of western oak forests in Oklahoma.
American Journal of Botany 71: 1221-1227.

Elias, T. S. 1980. The complete trees of North America.
Times Mirror Magazines, Inc., New York. 948 pp.

Eliot, W A. 1938. Forest trees of the Pacific Coast. G. R
Putnam's Sons, New York. 565 pp.

Glinski, R. L. 1977. Regeneration and distribution of syca-
more and Cottonwood trees along Sonoita Creek,
Santa Cruz County, Arizona. Pages 116-123 in R. R.
Johnson and A. Dale, technical coordinators. Impor-
tance, preservation, and management of riparian
habitat — proceedings. USDA General Technical
Report RM-43. Fort Collins, CO.

Hayward, C. L. 1948. Biotic communities of the Wasatch
chaparral, Utali. Ecological Monographs 18: 473-506.

Hitchcock, C. L., and A. Cronquist. 1964. Vascular
plants of the Pacific Northwest. Volume 2. Univer-
sity of Washington Press, Seattle. .597 pp.

HuscH, B., C. I. Miller, and T. W. Beers. 1972. Forest
mensuration, fohn Wilev & Sons, Inc., New York.
410 pp.

Hlscih.E, Ci. 197.5. Analysis ot the \egetation along the
middle and lower Snake River Unpublished master's
thesis. University of Idaho, Moscow. 271 pp.

JoHXSEN, T N. 1962. One-seed juniper invasion of north-
ern Arizona grasslands. Ecological Monographs 32:
187-207.

Johnson, C. G., Jr., and S. A. Simon. 1987. Plant associa-
tions of the Wallowa-Snake Province, Wallowa-
Whitman National Forest. USDA Pacific Northwest
Region Publication R6-ECOL-TP-255B-86. 272 pp.

Lanner, R. M. 1983. Trees of the Great Basin. University
of Nevada Press, Reno. 215 pp.

Little, E. L., Jr. 1976. Adas of United States trees. Vol-
ume 3-Minor western hardwoods. U.S. Department
of Agriculture Miscellaneous Publication No. 1314.
Washington, DC. 215 pp.

MuiR, P S., AND J. E. LOTAN. 1985. Disturbance histoiy
and serotiny of Finns eontoiia in western Montana.
Ecology' 66:' 1658-1688.

Oppenheimer, H. R. 1964. Adaptation to drought: .xero-
phytism. Pages 10.5-135 in Plant-water relationships
in arid and semi-arid conditions. The Hebrew Uni-
versity, Rehovot, Israel.

Peattie, D. C. 19.53. A natural histoi-y of western trees.
Houghton Mifflin, New York. 751 pp.

Plummer, a. P 1977. Revegetation of disturbed inter-
mountain area sites. Pages 302-337 in J. C. Thames,
editor. Reclamation and use of disturbed lands of the
Southwest. University of Arizona Press, Tucson.

Potter, L. D., and D. L. Green. 1964. Ecology of pon-
derosa pine in western North Dakota. Ecology' 45:
10-23.

Rumble, M. A. 1987. Avian use of scoria rock outcrops.
Great Basin Naturalist 47: 62.5-630.

Schnell, G. D., P G. Risser, and J. F Helsel. 1977.
Factor analvsis of tree distribution patterns in Okla-
homa. Ecology 58: 1345-13.55.

SPSS. 1988. SPSS/PC -h V2.0 base manual. SPSS Inc.,
Chicago.

Stephens, H. A. 1973. Woody plants of the North Central
Plains. University Press of Kansas, LawTence. 530 pp.

Tisdale, E. W 1986. Canyon grasslands and associated
shrublands of west-central Idaho and adjacent areas.
Bulletin No. 40. Forestiy, Wildlife, and Range Exper-
iment Station, University of Idaho, Moscow. 42 pp.

Van Auken, W W, A. L. Ford, and A. Stein. 1979. A com-
parison of some woody upland and ripariiui plant com-
munities of the southern Edwards Plateau. South-
western Naturalist 24: 165-180.

Van Dersal, W. R. 1938. Native woody plants of the
United States, their erosion control and wildlife val-
ues. U.S. Department of Agriculture Miscellaneous
Publication No. 303. Washington, DC. 362 pp.

Whisenant, S. G. 1990. Changing fire frequencies on
Idaho's Snake River Plains: ecological and manage-
ment implications. Pages 4-10 in E. D. McArthur, et
al., compilers, Cheatgrass invasion, shrub die-off,
and other aspects of shrub biolog\' and manage-
ment— proceedings. U.S. Department of Agriculture
General Technical Report INT-276. Ogden, UT

Whittaker, R. H. 1960. Vegetation of the Siski\ou
Mountains, Oregon and California. Ecological Mono-
graphs 30; 279-338.

Received 22 March 1994
Accepted 29 November 1994

MIMULUS EVANESCENS (SCROPHULARIACEAE): A NEW ANNUAL
SPECIES FROM THE NORTHERN GREAT BASIN

Robert J. Meinke^

Abstract. — Recent taxonomic studies in Mhmdiis support the recognition of Mimultts evanescens, a new autoga-
mous species moiphologically allied with M. brevijlorus and M. latidens. Initially known only from herbarium speci-
mens, the most recent from 1958, M. evanescens was relocated in the field in 1990 in northern Lassen Co., CA. A sec-
ond population was found in southern Lake Co., OR, in 1993. Mimidus evanescens is apparendy confined to the Great
Basin and its peripheiy where it has been recorded from 10 localities across Idaho, Oregon, and California. Based on
collection information and visits to the two extant populations, the new species appears to be restricted to vernally moist
sites and fluctuating banks of intermittent streams or pools. Long-term utilization of such sites by livestock may have
contributed to the present-day rarity of M. evanescens. The species should be added to federal and state lists of candi-
date endangered species pending the results of future field studies and surveys.

Key words: Mimulus, Great Basin, faxononiy, Scroplndariaceae, nwnkeyflower, Mimulus breviflorus, Mimulus latidens.

Mimuhis breviflorus is a diminutive, self-
pollinating, annual monkeyflower occurring
primarily east of the Sierra Nevada and Cascade
Mountains in the northwestern United States
and adjacent British Columbia. Little is known
concerning the evolutionaiy or taxonomic rela-
tionships of this or most other taxa in the genus,
which comprises perhaps 100 predominantly
North American species (Thompson 1993). In
the only comprehensive monograph of the
genus, Grant (1924) placed M. breviflorus in
section Paradanthus, an assemblage of small,
problematic species groups that are probably
paraphyletic and considered difficult to align
taxonomically (Argue 1980). Indeed, in a pro-
posed phylogenetic chart Grant (1924) affiliated
the yellow-flowered M. breviflorus with the
M. moschatus alliance, while in the text of her
paper she associated the species with mem-
bers of the M. inconspicuiis group, particularly
the white- to pinkish-flowered M. latidens of
California.

The proposed relationship between Mimulus
breviflorus and M. latidens is largely based on
shared features of the corolla and calyx. Both
species possess short, inconspicuous corollas
and strongly plicate, chartaceous fruiting
calyces that inflate with age. Although inflated
calyces are also described for some members
of the M. moschatus complex (Grant 1924,

Munz 1959, Holmgren 1984), the consistently
reduced, essentially regular flowers of M. brevi-
florus and M. latidens are unlike any species in
that group. The calyx moiphology and texture
of the two species is also different, being sin-
gularly reminiscent of M. inconspicuus and its
proposed relatives (Grant 1924). Moreover, the
general habit of M. brevijlorus and M. latidens
is more comparable to this group than to any
other

Despite the similarities, Mimulus breviflorus
and M. latidens are quite distinct with respect
to geography and habitat. Mimulus breviflorus
is a basin and range species, principally occur-
ring in well-drained, rocky environments near
rain pools, rocky meadows, and ephemeral
streamsides, often at middle and upper eleva-
tions. It has rarely been recorded south of
extreme northeastern California, and only then
above 2000 m. Mimulus latidens occurs mostly
on poorly drained flats or slopes subject to
vernal inundation, primarily below 800 m.
The species is virtually endemic to California,
extending from the Central Valley to northern
Baja California. The apparent uncertainty by
Grant (1924) over the taxonomic placement of
M. breviflorus may have been influenced by
geography, in that the range of the species
overlaps much of the M. moschatus complex
but not M. latidens or the M. inconspicuus

^Restoration Eeolog,' and Plant Conservation Biologv' Cooperative Project. Department of Botany and Plant Pathologx', Oregon State Universit)', Corvallis, OR
97331. (The Restoration Ecolog\' and Plant Consenation Biolog>- Cooperative Project is a collaborati\e reseiucli unit of Oregon State Universit)' and the Oregon
Department of Agriculture.)

249

250

Great Basin Naturalist

[Volume 55

group, which are restricted to cisniontane
Cahfornia.

The present study was prompted b\' several
unusual herbarium collections identified as
Mitiuihis hrcvijloriis, disco\ ered during a taxo-
nomic sune> of the Mimiilus washingtonensis
complex (Meinke in preparation) in which sev-
eral hundred collections (including all rele-
vant types) were examined. Despite the evi-
dently yellow flowers and the fact that the few
collection localities were well within the
known range of M. breviflonis, the plants were
similar to M. latidens in many respects. The
anomalous material originated from several
scattered stations across the upper Great
Basin and its northern periphery, all within
areas belie\'ed historically grazed by livestock.
The most recent of these collections is dated
1958, and tliere was concern that the entity may
have become extinct in die interim. Unexpected
opportunities to obsei've living populations in
the field were presented in 1990 and 1993 dur-
ing chance visits to two reservoirs in Lassen
Co., CA, and Lake Co., OR. The unique and
consistent combination of features noted in
herbarium collections, including bright yellow
corollas, was even more conspicuous in living
plants, prompting a taxonomic reevaluation of
their relationship with M. breviflonis and M.
latidens. After further evaluation, the unusual
populations were considered to represent a new
species which is here described, illustrated,
and contrasted with potentially related taxa.

Description of the Species

Mimulus evanescens Meinke, sp. nov. (Figs.

lA-C).— Type: USA, California, Lassen Co.,
20.5 km east of Adin, north side of Ash Valley
Rd., ca 0.1 km east of the Lassen National
Forest boundary, in broken boulders and heavy
gravel abutting Moll Reservoir, T38N RIOE,
NWl/4 SWl/4, Sect. 25, ca 1500 m, 27 June
1990, Meinke and Kaije 5900 (holotype, OSC;
isotypes, MO, NY, RxM, UC, US, UTC).

Herhae annuae, puberulentes, ± viscido-vil-
losae; caulis tenuis, erectis, (6-) 10-25 cm altis,
internodiis elongatis; /0///5" late ovatis vel
lanceolatis, lamina integerrima vel parce den-
ticulata, acuta, 1.0-3.8 cm longa, 0.7-2.9 cm
lata, 3(-5) nei-vis, base lata, sessili vel subsessili;
pedicel foliis brevioribus, tenuibus, ascenden-
tibus; calyce in statu florifero 3.5-6.5 mm
longo, 1.5-3.5 mm lato, in statu fructifero late

urceolato, 7.0-11.0 mm longo, 5.0-8.5 mm
lato, valde glabro, dentibus ciliati, late triangu-
laribus, ± subaequalibus, acutis; corolla flava,
brevi, 4.0-9.5 mm longa, calyce ca 1.5 plo lon-
giore, tubo incluso, lobis ± aetiualibus, patulis,
erectis; statninibus stylo aequalibus, inclusis,
glabris; stylo glabro, 3.0-7.8 mm longo, labiis
stigmatis laciniatis, subaequalibus; capsula
inclusa, subglobosa, 4.8-9.0 mm longa, sessili
vel substipitata; seininibiis late oblongis, ca
0.3-0.6 mm longis.

Annual herb, ± succulent, glandular-
puberulent throughout (except the calyces),
the hairs short and appearing of even length to
the naked eye, moist or slimy to the touch,
mostly one-celled (excluding the gland); stems
slender, (6-) 10-25 cm tall, erect to slightly
decumbent in robust individuals, simple or
branched from near the base, often sparingly
branched above as well, with elongated inter-
nodes; leaves acute, broadly ovate to some-
what lanceolate, 1.0-3.8 cm long, 0.7-2.9 mm
wide, evenly distributed, not much reduced at
the upper nodes, not forming a basal rosette,
the lower ones abruptly petiolate or subses-
sile, petioles 1-3 mm long, blades broadly ses-
sile above, with 3(-5) primaiy veins, the mar-
gins entire or shallowly denticulate; pedicels
slender, 8-18 mm long, ascending in flower
and fruit, shorter than the leaves in fruit or
longer in depauperate individuals; inflores-
cence racemose, flowers axillaiy; ^o«;ers incon-
spicuous, autogamous; calyx 3.5-6.5 mm long
and 1.5-3.5 mm wide in flower, tubular-cam-
panulate at anthesis, green becoming strami-
neous and anthocyanic along the angles with
age, accrescent and broadly urceolate to oxal
in fruit, 7.0-11.0 mm long and 5.0-8.5 mm
wide, the tube ehartaceous and glabrous, the
orifice narrowing and becoming somewhat
oblique, the angles strongly plicate, the teeth
broadly triangular, acute, 0.8-1.6 mm in fruit,
ciliate on the margins, scarcely unequal, the
uppermost lobe occasionally appearing slightly
longer in some flowers; corolla short and
essentially regular, 4.0-9.5 mm long, clear yel-
low or occasionally with a few tiny brownish
dots in the throat, the inconspicuous petal
lobes rounded or mucronate, the tube includ-
ed or barely exserted, the limb exceeding the
calyx by 2-3 mm, lobes short and subequal,
mostly erect, glabrous externally, bearded in-
ternally with a few ± clavate hairs extending in
a line from the lower palate into the floral tube;

1995]

New Mimulus From Idaho, Oregon and California

251

4 mm

Fig. 1. Mimulus evanescens Meinke: A, habit drawing showing details of leaf moiphology, inflorescence, and relation-
ship between pedicel and leaf blade length; B, close-up of fruiting calyx of M. evanescens; C, calyx of M. evanescens
opened to show sessile capsule insertion; D, calyx of M. latklens opened to show stipitate capsule insertion.

stamens included, about equal with the style,
glabrous, pale; style glabrous, included, 3.0-7.8
mm long, stigma lips equal or subequal, shal-
lowly laciniate-margined; capsule included,
subglobose, 4.8-9.0 mm long, extending to ca
1.0-2.5 mm below the sinuses of the calyx
teeth, sessile, or rarely with an abbreviated
stipe up to ca 0.5 mm long, the placentae
adherent to the apex; seeds ovoid or broadly

oblong, brownish, 0.3-0.6 mm long, dormant
when first ripe, dispersal often delayed or pro-
longed due to the nearly closed, inflated calyx.
Para'HPES. — USA, California, Lassen Co.,
10 mi south of Ravendale, 9 June 1940, Pennell
25763 (P); 4.8 mi south of Madeline, 17 June
1958, Raven and Solbrig 13298 (JEPS); Modoc
Co., along Willow Creek, June 1894, Austin s.n.
(UC). Idaho, Owyhee Co., meadow, 3 mi south

252

Great Basin Naturalist

[Volume 55

of Riddle, 1 July 1949, Hohngren and Holmgren
7973 (CAS, UC, WS, WTU). Oregon, Crook
Co., Grizzly Butte, 18 June 1894, Leiberg 275
(NY, 0RE,'US); Gilliam Co., forks of Cotton-
wood Canyon, 6 June 1894, Leiherg 156 (NY,
ORE, P, US); Grant Co., Ochoco National
Forest, Graylock Butte, 6 July 1912, Ingram
s.n. (RM); Harney Co., dry watercourse near
Frenchglen, 26 June 1942,' Peck 21389 (CAS,
NY, I^ UC, WILLU); Lake Co., moist to muddy
margins of receding water, among rocks, Drews
Reservoir, 15 June 1993, Meinke and Carlson
6401 (BRY, HSC, NY OSC, RM, RSA, SRR
UC, US, UTC, WS, WTU).

Distribution and habitat — Mimulus evan-
escens is distributed widely along the north-
western edge of the Great Basin at elevations
of ca 1200-1700 m, ranging from southwest
Idaho west through eastern Oregon and south
into northeastern California. Mimulus hrevi-
florus is more widespread and considerably
more common. Although inconspicuous even
when in bloom, it has been recorded from
numerous collections located throughout much
of the northwestern United States east of the
Sierra Nevada and Cascade ranges. Outlying
populations are known from southern British
Columbia and south (rarely) to the mountains
near Lake Tahoe. Mimulus breviflorus has a
broader elevational range than M. evanescens,
occurring from roughly 300 to 2900 m.

Mimulus latidens is essentially a California
endemic, distributed below 800 m from the
northern Central Valley south to San Diego.
The most southerly populations are known
from northern Baja California, while four his-
toric collections from extreme southwestern
Oregon, originally identified as M. breviflorus,
represent the northern range limits. A recently
discovered Great Basin population of M. lati-
dens, occurring at ca 1700 m in southern Lake
Co., OR (Shelly 1986), is noteworthy as it is
the single recorded locality in which the range
of this otherwise low-elevation species over-
laps either M. evanescens or M. breviflorus.
The population was persisting over several
acres in a sagebrush-dominated swale as of
1993. This area is along the flyway for various
waterfowl species migrating northeast across
the Great Basin from central California.

The habitat of Mimulus evanescens can be
evaluated only from the two extant localities,
the first adjacent to Moll Reservoir in Lassen
Co., CA, and the second at Drews Resei-voir

in Lake Co., OR, l)()th occurring within sage-
brush-juniper-dominated vegetation zones.
Plants at both sites were scattered among rock
fragments and alongside small boulders, in
moist, heavy gravel that had been inundated
earlier in the spring. The California popula-
tion was discovered in 1990 and visited again
in 1991, while the Oregon population was first
located in 1993. Associated species during these
years (for both locations) included Artemisia
tridentata, Junipenis occidentalis, Mimulus flori-
bundus, M. suksdorfii, Porterella carnosula,
Collinsia grandiflora, C. parviflora, Downingia
sp., Mimetanthe pilosa, Heterocodon rariflorum,
Poa bulbosa, and Bromus spp. The perennials
Machaerocarpus californicus and Marsilea
vestita were common along the shoreline at the
Lassen Co. site. Remaining locations for M.
evanescens are known only through scanty
herbarium labels, with specimens reportedly
taken from rocky stream banks or drying
watercourses. Mimulus breviflorus occurs in
comparable microsites, frequenting wet, rocky
sites that often diy out by late spring or early
summer, as well as lush, gravelly meadows.

Morphological comparisons. — Monkey-
flowers are often phenotypically plastic, and
related annual species in particular may be sub-
ject to overlapping moi-phological variation de-
pending on ecological conditions. In an attempt
to objectively evaluate the phenetic relation-
ships of the new species and its most similar
congeners, a data set was compiled by scoring
18 vegetative and reproductive character
states (Table 1) from 114 Mimulus collections
representing 38 populations. Measurements
were taken from 15 populations each of M.
latidens and M. breviflorus, and 8 of the 10
extant and historical populations of M. evan-
escens. Three plants were measured per collec-
tion to provide population averages for each
quantitative trait. Sample populations of M.
latidens and M. breviflorus were selected from
herbarium collections encompassing the geo-
graphic range for each species. Every effort
was made to choose individuals of the three
species that, based on field experience of the
author, represented normally developed plants
(i.e., not drought-stressed) from approximately
the same life-history stage. Measurements
were made on randomly selected individuals
where possible, insofar as the limited number of
phenologiciilly acceptable collections pemiitted.

1995]

New Mimulus From Idaho, Oregon and California

253

Table 1. List of morphological traits measured from
Mimtihis plants for use in principal components analysis.
Thirty-eight study populations were sampled, including
15 each for M. latidens and M. breviflorus, and 8 for M.
evanescens. An average measurement was derived for
each trait (from 3 samples per population) for use in the
analyses.

(1) Presence or absence of a basal rosette

(2) Length of initial stem leaf

(3) Width of initial stem leaf

(4) Length of upper cauline leaf

(5) Width of upper cauline leaf

(6) Base of upper cauline leaf (sessile versus distincth-
petiolate)

(7) Peduncle length (in fruit)

(8) Caly.x length (in fruit)

(9) Calyx width (in fruit)

(10) Length of caKx teeth (in fi-uit)

(11) Overall corolla length

(12) Corolla color (yellow versus rose, whitish, or
ochroleucous)

(13) Length of lower corolla lip

(14) Width of lower corolla lip

(15) Length of capsule

(16) Width of capsule

(17) Capsule insertion (base sessile versus distinctly
stipitate)

(18) Stem and leaf pubescence (clearly glandular-
puberulent versus glabrous or subglabrous)

The data set was initially used to analyze
moiphological relationships between Mimulus
evanescens, M. latidens, and M. breviflorus
using a principal components analysis (PC A).
Clustering relationships of sample populations
were compared along the first two a.\es of vari-
ation and graphically displayed. As a second
measure of overall dissimilarity, canonical dis-
criminant analysis (DA) was performed on the
same populations using only the quantitative
characters from Table 1 (i.e., eliminating traits
1, 12, 17, and 18). On herbarium specimens
the resolution of certain qualitative traits, such
as flower color and degree of pubescence, may
be open to inteipretation if specimens are poor-
ly preserved or mishandled after collection.
Since many of the available Mimulus collec-
tions were old or otherwise less than optimal
for a moiphometric study, the potential existed
for errors in judgment of qualitative traits to
bias the analysis. As an alternative, DA was
utilized to determine whether the elimination
of diagnostic qualitative traits would result in
a weaker phenetic relationship than that indi-
cated by PCA. The multivariate statistical
package'in STATGRAPHICS (v. 4.0) was used
for the two analyses.

PCA clustered the 38 populations into
three well-defined groups conforming to a
priori determinations of samples as M. evan-
escens, M. latidens, or M. breviflorus (Fig. 2).
The first two principal components accounted
for 88% of total variance (Table 2), indicating
that the PCA scatterplot (Fig. 2) is a good
gauge of overall morphological differences
among the three species. DA resulted in a
comparable pattern, although M. evanescens
clustered somewhat closer to M. latidens when
qualitative characters were excluded (Fig. 2).
In both analyses, M. evanescens is clearly and
consistently intermediate to M. latidens and
M. breviflorus.

Upon first inspection Mimulus evanescens
appears to be merely a robust version of M.
breviflorus. The yellow, nearly regular corollas,
essentially nonstipitate capsules, and short-
puberulent foliage and stems are traits that are
virtually identical in the two species. Since M.
evanescens also develops papery, inflated fruit-
ing calyces, which is the most prominent fea-
ture in most specimens of M. breviflorus, it is
understandable that the identity of the new
species has been obscured. However, the over-
sized habit of M. evanescens is striking, and all
floral and vegetative characteristics average
larger than in M. breviflorus.

The significance of these proportional dif-
ferences was first noted when fresh material of
M. evanescens from the type locality was com-
pared with greenhouse-grown plants of M. bre-
viflorus. Although many individuals of the new
species suffered from insect predation in the
field and others appeared underdeveloped
due to drought, undamaged plants from moist
microsites commonly grew to 2 dm or more,
far exceeding the largest examples of M. brevi-
florus. Conversely, M. breviflorus plants culti-
vated in the greenhouse (originating from three
distinct populations in eastern Oregon) never
exceeded 12 cm in height. Rather than grow-
ing taller wdth age, they tended to branch out
and become unusually floriferous. This obser-
vation was confirmed when plants of M. brevi-
florus and M. evanescens (32 and 27 individu-
als, respectively, from populations in Lake
Co., OR) were grown together from seed in a
common greenhouse environment. Given iden-
tical conditions, all M. evanescens plants grew
to over twice the size of M. breviflorus. In
addition, all quantitative and qualitative differ-
ences for the species originally noted on the

254

Great Basin Natur\list

[Volume 55

Principal Components Analysis

Discriminant Analysis

c
o

E
o
O

"cd
g.
o

c
o
o

Mimulus evanescens

■

• •

Mimulus breviflorus

Mimulus latidens

-5 0 5

First Principal Component (73.9%)

Q
■D

O
O
CU
GO

■

Mimulus evanescens

A A

Mimulus breviflorus

•

A°

•\

Mimulus latidens

8*8

-10 -5 0 5

First Discriminant Function

Fig. 2. Morphometric evaluations (see te.xt for discussion); hvo-dimensional plots depicting principal components (18
qualitative and quantitative characters) and discriminant analyses (14 quantitative characters), contrasting the moipho-
logical relationships of A/, evanescens (triangles), M. latidens (open circles), and M. breviflorus (closed circles).

herbarium specimens were maintained in cul-
ture. Although some herbarium specimens of
M. evanescens are not particularly large, it is
suspected that this is due to moisture limita-
tion rather than genetic potential, based on
obsei'vations of living plants.

In addition to the overall size disparity, other
features readily separate Mimulus evanescens
from M. breviflorus. Most evident are the
leaves, which are ovate to broadly lanceolate
in the new species and rhombic-ovate or nar-
rowly lanceolate in M. breviflorus. Moreover,
only the very lowest leaves of M. evanescens
are petiolate, and these abruptly so (Fig. lA),
while leaf blades of M. breviflorus nanow grad-
ually into slender, evident petioles at all nodes.
In fruit, the pedicels of M. breviflorus general-
ly exceed or at least equal the leaf blades.
Those of the new species are always shorter in
well-developed plants, and in some instances
the leaf blade exceeds the calyx as well. Finally,
the fruiting calyx of M. evanescens is much
more plicate and typically exceeds the length
of the mature capsule by 1.5-3.0 mm (Figs.
1B,C). In M. breviflorus ripe capsules are
approximately the same length as the calyx. The
overall dimensions of Mimulus evanescens, as
well as the strongly angled fruiting calyx and
broad, sessile leaves, are traits that also imply

a relationship with M. latidens. There are sub-
stantial differences between these taxa, however,
including flower color, pedicel length in rela-
tion to leaf length, stem pubescence, capsule
insertion (Figs. 1C,D), and the presence of a
basal rosette in M. latidens. Dissimilarities
among the three species are summarized in
Table 3.

Table 2. Amount of total variance accounted for by
each principal component, in a principal components
analysis of moiphological variation among populations of
Mimulus evanescens. M. breviflorus, and M. latidens.

Component

Percent of

Ciuuulatixe

ninnher

variance

percentage

73.88

14.18

88.06

3.75

91.81

2.41

94.22

2.12

96..34

1.22

97..56

.,57

98.13

.46

98.59

.39

98.98

.29

99.27

.21

99.48

.17

99.65

.12

99.78

.11

99.89

1.5

.06

99.95

.05

100.00

1995] New Mimulus From Idaho, Oregon and California

Table 3. Diagnostic features oi Mimulus evanescens, M. breviflorus, and M. latidens.

255

Character

M. evanescens

M. breviflorus

M. latidens

Plant height

(ft-) 10-25 cm

3_10(-14) cm

10-26 cm

Pubescence

Glandiilar-puberulent

Glandular-puberulent

Subglalirous

Basal rosette

Yes

Leaf base

Petiolate at base,
sessile above

Petiolate throughout

Petiolate at base,
sessile above

Leaf blade shape

Ovate to lanceolate

Elliptic-lanceolate

Broadly ovate

Cauline leaf
length

1.0-3.8 cm

0.4-1.7 cm

0.8-3.2 cm

width

0.7-2.9 cm

0.2-0.5(-0.8) cm

0.5-1.7 cm

Pedicels

0.8-1.8 cm long,
< the blades

0.5-1.9 cm long,
> the blades

1.0-3.3 cm long,
> the blades

Fruiting calyx
length

7-11 mm

4-8 mm

9-12 mm

width

5.0-8.5 mm

3.0-4.5 mm

6-8 mm

Corolla color

Yellow

Whitish, shaded
rose or yellow

Corolla length

4.0-9.5 mm

3.5-5.5 mm

9.0-11.5 mm

Capsule insertion

± sessile

clearly stipitate

Capsule length

4.8-9.0 mm,
clearly inserted

4.5-8.0 mm, about
equaling calyx

6.0-9.0 mm,
clearly inserted

Distribution

Great Basin and
vicinity

Great Basin and
vicinitA'

Cismontane
California

Elevation

-1200-1700 m

300-2900 m

<800 m

Other small-flowered annuals that might be
confused with Mimulus evanescens are primar-
ily members of the M. moschatus complex,
particularly M. floiibundus, M. patulus, and M.
pulsiferae. Of these, only M. floiibundus is
ever characterized as having an inflated fruit-
ing calyx (Grant 1924), which can be distin-
guished from M. evanescens by the multicellu-
lar pubescence throughout and narrow, lance-
olate sepals. These three species are further
differentiated from M. evanescens by distinctly
petiolate upper leaves and bilabiate corollas.
Depauperate annual forms of M. guttatus also
occur in moist sites within the range of M.
evanescens. This common yellow-flowered
species can be separated by petiolate upper
leaves and strongly zygomorphic corollas.
Although the calyces of M. guttatus are also

markedly inflated, they are distinctly irregular
and oriented horizontally in fruit. Mimulus
suksdorfii is the only other annual monkey-
flower in the Pacific Northwest with features
comparable to M. evanescens. Seldom exceeding
6 cm in height, this compact, freely branched
species is easily distinguished by obtuse, linear-
oblong leaves, a cylindrical fruiting calyx, and
flaring, emarginate corolla lobes.

Phylogenetic considerations. — Judging
from moiphology, Mimulus evanescens appears
most closely related to M. breviflorus and M.
latidens, and exhibits characteristics of both
taxa (Fig. 2). Mimulus latidens, in turn, also
seems to have a strong affinity to M. incon-
spicuus, M. graiji, and M. acutidens from Cali-
fornia, based primarily on flower color, stipi-
tate capsules, calyx morphology, leaf shape,

256

Great Basin Natufl\list

[Volume 55

and glabrous habit (Grant 1924, Tlionipson
1993). Aside from general \egetati\e and floial
similarities, the inflated, plieate fruiting eal>'x
is the principal trait linking these six species
together. Whether or not this featui-e implies a
monoplnietie group is open to debate, however,
since inflated calyces have evidently arisen
independently in Mimulus on more than one
occasion. Nonetheless, the shape and texture
of the calyces of these species are distinctive.

The recognition of Mimulus evanescens
allows for a reevaluation of the relationship
between M. hrcviflorus and the rest of the
genus. The morphology of M. evanescens,
transitional between M. breviflorus and M.
latidens, suggests that the new species might
have arisen through hybridization. However,
this hypothesis conflicts with the current geo-
graphical and ecological separation of the
putative parents and the fact that M. breviflorus
is highly autogamous. An alternative scenario
proposes M. evanescens as a descendant of M.
latidens. The smaller-flowered and apparentK'
more successful M. breviflorus (based on the
number of historic collections) may have then
arisen from M. evanescens, perhaps as a result
of a shift to more xeric conditions in what is now
the Great Basin. Mimulus breviflorus is ubiqui-
tous and well represented in herbaria while
M. evanescens is apparently rare and widely
scattered, providing circumstantial support for
this concept. The discoveiy of the disjunct M.
latidens population in Lake Co., OR (Shelly
1986) is intriguing, because it suggests a mech-
anism by which this relationship might have
developed. If genotypes of M. latidens capable
of survival outside of California's relatively
benign Central Valley have been historically
transported to the Great Basin by migrating
ducks or geese, the means and opportunity for
adaptive radiation could have existed.

Conservation. — It is not encouraging that
only 10 extant or historical populations of
Mimulus evanescens are known, with only two
sites recorded since 1958. This contrasts with
hundreds of collections at dozens of localities
for the much less conspicuous M. breviflorus.
As with M. breviflorus, the distribution of M.
evanescens is apparently limited to damp or
wet sites at moderate elevations within open
rangeland. Virtually all such sites in the Great
Basin are associated with a long history of
grazing by domestic livestock. The broad geo-
graphic range and relatively unremarkable

habitat of Miiiuilus evanescens impK that the
compaiativc rarit)' of the species may be the
result of habitat loss or disturbance. However,
the paucity of herbarium records, especially
when contrasted with similar species, suggests
that M. evanescens may have never been com-
mon, even under pristine, pre-grazing condi-
tions. If this is true, the combination of natural
scarcity with contemporary grazing or other
disturbances ma\ now be jeopardizing the
species. As an initial step, M. evanescens should
be added to federal and state lists oi candidate
endangered species. Although confirmed from
Idaho, Oregon, and California, it is expected
that northern Nevada is also within the his-
toric range of the species. Placing Af. evanescens
on candidate lists will bring the species to the
attention of land managers in these states and
will help justify inventory and research, which
may in turn ascertain diat the species is not par-
ticularly rare and has merely been overlooked
by collectors. However, until this is estab-
lished it is piTident to consider the species ex-
tremely vulneral:)le, with ample protection given
to any sites occurring on public lands.

Acknowledgments

The author acknowledges field or green-
house assistance provided by Thomas Kaye,
Matthew Carlson, Steven Gisler, Lisa Lantz,
Crista Chadwick, and Melissa Peterson. Line
drawings were prepared by John Megahan. The
manuscript was reviewed by Kenton Chambers,
Robert Frenkel, Mary Barkworth, Teresa Magee,
and Edward Guerrant. Financial or logistical
support for this study was provided by the
Oregon State University herbaria, the USDA
(Fremont and Winema National Forests), and
the Plant Consei-vation Biology Program of the
Oregon Department of Agriculture. Staff of the
following herbaria graciously lent specimens or
othei'wise provided access to their collections:
BRY, CAS, CU, DS, GH, ID, IDF JEPS, M,
NY, ORE, OSC, P RM, RSA, UC, US, UTC,
WILLU, WS, and VVTU.

Literature Cited

Argue, C. L. 19<S(). PoUt-n morphology in the genus
Mimulus (Scrophnlariaceae) and its taxonomic signif-
icance. American Journal of Botan\' 67: 68-87.

Grant, A. L. 1924. Monograph of the genus Mimulus.
Annals of the Missouri Botanical Garden II: 99-388.

Holmgren, N. 1984. Mimulus. Pages 3.50-364 in A.
Cronquist, A. Holmgren, N. Holmgren, J. Reveal,

1995]

New MiMULus From Idaho, Oregon and California

257

and E Holmgren, editors, Intermountain flora. IV
New York Botanical Garden, The Bronx.

MUNZ, E A. 1959. A California flora. Universit>' of Cali-
fornia Press, Berkeley and London.

Shelly, J. S. 1986. Noteworthy collection o{ Mimiihis lati-
dens. Madrofio 33: 151.

Thompson, D. M. 1993. Mimulus. Pages 1037-1046 in J. C.
Hickman, editor. The Jepson manual: higher plants

of California. University of California Press, Berkeley
and Los Angeles.

Received 2 March 1994
Accepted 5 December 1994

MORPHOLOGICAL AND HOST-SYMBIONT STUDIES OF TRICHODINA

TENUIFORMIS AND APIOSOMA CAMPANULATUM INFESTING MOTTLED

SCULPIN {COTTUS BAIRDI) FROM PROVO RIVER, UTAH

YingQi' and Kicluird A. Heckmannl'2

Abstiuct. — TrichocliiKi teintijonnis Stein, 1979 and Apiosomci cciinpanitlatiiin Tiinoteev, 1962 were found on j^ills of
mottled sculpin {Cotttis bairdi) from two locations in the Provo River, UT. They were studied by light and electron
optics. Dimensions and morphology of the adhesive disc and denticles of T. tenuifonnis were differentiated from other
Trichodma species. A. campanulatum was characterized by its spindle-shaped cell body. Fine features examined by scan-
ning electron microscopy included body shape, pellicle, elements of the adhesive disc, aboral ciliaiy complex, and ado-
ral ciliary spiral. Histopathological studies suggested that the organisms are ectocommensals. Ecological aspects of
organism infestation between two areas were also investigated. This report establishes a new host and distribution
record for these two species in mottled sculpin from the Provo River, UT.

Key words: Trichodina tenuiformis, Apiosoma campanulatum, Cottus bairdi, morphology, host-syinbiont relationship,
ecological aspects. Provo River

High numbers of two ciliated protozoa,
Trichodina and Apiosoma, were encountered
on the gills of mottled sculpin {Cottus bairdi)
during a study of ectoparasites of fishes from
the Provo River

Trichodina is a mobile ciliate belonging to
the subclass Feritrichia, family Trichodinidae
(Lorn and Dykova 1992). This protozoan has
an adhesive disc characterized by very promi-
nent and taxonomically significant denticles
(Van As and Basson 1987). More than 140
species of Trichodina have been reported from
wild, cultured, and laboratoiy fishes in many
parts of the world (Rand 1993).

Sessile peritrich ciliates of the genus Apio-
soma (syn. GlossateUa) belong to the subclass
Feritrichia, family Epistylididae (Lom and
Dykova 1992). They are generally attached to
fish by a scopula (Lom 1973). They have been
largely neglected by fish parasitologists until
recently, when more attention has been given
to this group.

Many species of these two ciliated protozoa
have been investigated (Arthur and Margolis
1984, Cone and Odense 1987, Rand 1993);
however, a detailed study on mottled sculpin
has never been reported. Objectives of this
study were to (1) incoiporate different levels of
microscopy to study ciliate structure, (2) ob-
sei-ve histopathological changes these protozoa

may cause to the host, and (3) evaluate the sea-
sonal infestation rate to provide ecological infor-
mation for the listed ciliates and their host.

Materials and Methods

Studies were earned out in late summer and
fall (August, October 1993), late winter and
spring (March, May 1994). Water temperatures
in the Frovo River ranged from 14 °C to 4°C
and 6°C to 10 °C, respectively. One hundred
si.xty sculpin were collected from two sites:
one in the city of Frovo (Utah County) munici-
pal area, the second in a relatively pristine
region near the Jordanelle Reservoir (Wasatch
County). Sculpin were collected using elec-
trofishing, placed in buckets containing river
water, transported to the laboratory, and exam-
ined within 24 h after capture.

For light microscopy, air-dried smears of
gill filament scrapings were prepared from
freshly killed fish and treated by Klein's diy sil-
ver impregnation technique (Clark and Heck-
mann 1984) to examine components of the
adhesive disc. Other smears were prepared,
fixed, air-dried, and stained with iron hema-
toxylin (Carcia and Bruckner 1988) to obsen'e
the position and structure of the macro- and
micronuclei. Sections of infested gills from the
spring sample were fixed, blocked, cut, and

'Ofpartmcnt ol' Zoology', Brigham Young University, Provo, UT 84602.
^Autlior to whom correspondence should be addressed.

258

1995]

Ciliated Protozoa in Mottled Sculpin

259

stained with hematoxylin-eosin (Garcia and
Bruckner 1988) for histopathological studies.

For scanning electron microscopy, gills of
freshly killed fish were fixed in 2% buffered
glutaraldehyde, followed by repeated washes
in a sodium cacodylate buffer and post-fixed in
a 1% solution of osmium tetroxide. After that
they were washed in the same buffer system.
Specimens were dehydrated through a graded
alcohol series and critical-point-dried and
sputter-coated with gold for examination with
a Joel-840 high-resolution scanning electron
microscope.

For transmission electron microscopy, after
fixation and dehydration, gills were embedded
in Spurr resin and sectioned with a glass knife.
Each section was stained with lead citrate and
examined with a Philip EM400 transmission
electron microscope.

Terminology and methods of measurement
follow those given bv Lom (1958), Lorn and
Dykova (1992), Wellborn (1967), Arthur and
Margolis (1984). Measurements are in micro-
meters (/xm) and are based on 30 specimens
for each species from each of the four sam-
pling periods; range is followed by the mean
and ± standard deviation in parentheses.

Results

Morphology

Trichodina temiifonnis Stein, 1979

Host. — Cottus hairdi (Pisces: Cottidae).

LOCALITY'. — Provo River, Utah and Wasatch
counties, Utah.

Site of infestation. — Gill filaments.

Light microscopy. — Body 39-53 (44.2 ±
4.0) dia (diameter). Adhesive disc 19-30 (26.3
± 2.8) dia, surrounded by a border membrane
2—3 (2.5 ± 0.4) wide, with fine transverse stri-
ae. Various-sized light forms present in center
of adhesive disc when silver-impregnated.
Denticular ring 13.5-20 (17.2 ± 1.8) dia, con-
sisting of 20-26 (23.7 ± 1.3) denticles with
6-10 (7.8 ± 0.8) radial pins per denticle.
Denticle with conical central portions 0.7-1
(0.99 ± 0.06) from which a thorn 2.5-4 (2.9 ±
0.4) extends externally with broadly rounded
lobes, tapered slightly to a blunt tip and blade
2-3 (2.3 ± 0.3) attached to central region,
some with rounded ends (Figs. 1, 2).

Macronucleus horseshoe-shaped 27-48 (39
± 5.7) dia and approximately 10 ^tm thick.

Micronucleus in -Y position (Lom 1958)
observed in six specimens, dimension 3x2
(Fig. 3).

Scanning electron microscopy. — Body
of T. tenuiformis circular in aboral view and
aboral surface relatively flat (Fig. 4). Body
bell- shaped or domed when viewed from the
side (Fig. 5).

The aboral ciliaiy complex consists of three
distinct ciliary bands: the basal ciliary ring,
locomotor ciliaiy wreath, and marginal ciliary
ring. The basal ciliary ring, adjacent to the
border membrane, has a single row of fine,
distally tapering cilia 1-2 /xm long. Separated
from the basal ciliary ring by the basal septum
is the locomotor ciliaiy wreath, which is com-
posed of numerous rows of well-developed,
powerful cilia 2-3 fim long whose primary
function is locomotion. The precise number of
ciliaiy rows composing this wreath could not
be ascertained. It is separated anteriorly from
the marginal ciliary ring by a poorly developed
anterior septum that is evident only when the
aboral ciliary complex is uncovered by the
velum. The marginal ciliaiy ring is difficult to
distinguish from the locomotor ciliary wreath
in T. tenniformis. The velum is a thick, well-
developed structure covering the bases of the
cilia of the aboral ciliaiy complex and separat-
ing this complex from the adoral ciliar)' spiral
(Figs. 5, 6).

The adhesive disc has a smooth pellicular
surface beneath which the outline of the den-
ticles can be clearly seen. The disc is sur-
rounded peripherally by a 2-)U,m-wide border
membrane, which functions to seal the margin
of the disc during adherence and contains fine
vertical striae over its entire surface. These
striae on the internal surface of the border
membrane are the radial pins that give the
membrane rigidity while retaining its ability
to conform to the host's surface (Fig. 7).

The adoral ciliature forms a counterclock-
wise spiral of about 270°. The base of each cil-
ium is inserted into a deep furrow and hidden
from view when SEM is used (Fig. 8).

Deposition of slides. — One slide (HWML
37721) of silver-impregnated specimens and
another slide (HWML 37724) of iron-hema-
toxylin-stained specimens are deposited in
the Harold W. Manter Laboratory, University
of Nebraska State Museum. The senior author
has additional slides in her collection.

260

Great Basin Natur.\list

[Volume 55

Figs. 1-3. Light micrographs ot Triclioclhm tcuuijurmis: 1-2. Silver-impregnated specimens showing body shape and
arrangement of components of the adhesive disc. BM, border membrane; D, denticle; RR radial pins. Bar = 10 /im. 3.
Iron-hemato.xylin-stained specimen showing horseshoe-shaped macronucleus (MA); arrow points to the micronucleus
(MI). Bar = 10/Lim.

Apiosoma campanulatum Timofeev, 1962

Host. — Cottus bairdi (Pisces: Cottidae).

Locality. — Provo River, Utah and W^isatch
counties, UT.

Site of infestation. — Gill filaments.

Light microscopy. — Body campanulate.
Macronucleus round or slightly conical. Size
of stained specimens 3L0-66.0 (47.8 ± 7.2) long
by 25.0-45.0 (35.6 ± 4.2) wide. Macronucleus
lLO-20.0 (15.6 ± 2.4). Micronucleus not ob-
served (Fig. 9).

Scanning electron microscopy. — The
spindle-shaped body is the characteristic fea-

ture of this species. Circular striations of pelli-
cle conspicuous. Pellicle wrinkled into longi-
tudinal furrows. Upper part of body bears the
adoral zone, consisting of a tuft of 1-2-^tm-long
cilia. Most specimens viewed with SEM have
contracted peristomes and contracted peristo-
mial lips (Fig. 10).

Deposition of slides. — A representative
slide of Apiosoma campanulatum (silver stain)
is deposited in the Harold W. Manter Labora-
tory, University of Nebraska State Museum
(HWML 37722). The senior author has addi-
tional slides in her collection.

1995]

Ciliated Protozoa in Mottled Sculpin

261

Figs. -J-7. Scaumug electron iiueiugraplis of the surface of I temiiformis: 4. Aboral view of entire specnnen of I
tenuifonnis. B, bacteria. Bar = 10 fxm. 5. Lateral view of entire specimen of I temiifonnis. ACS, adoral ciliarv spiral; BS,
basal septum; BCR, basil ciliaiy ring; GE, gill epithelium; LCW, locomotor ciliary weath; VEL, velum. Bar = 10 fim. 6.
Higher magnification of Figure 4 showing the structure of aboral ciliaiy complex. ACS, adoral ciliary spiral; BCR, basal
ciliary ring; K, kinetosomes; LCW, locomotor ciliary wreath; MCR, marginal cilian,- ring. Bar = 1 ^tm. 7. Adhesive disc
of T tenuifonnis. D, denticle; PR peripheral pins; RR radial pins. Bar = 1 ^tm.

262

Great Basin Natur.\list

[Volume 55

Fig. 8. Adoral view of I tenuifonnis showing how the adoral cihatiire (ACS) forms a counterclockwise spiral of about
270°. Bar = 10 /xm. Fig. 9. Light micrograph of Apiosoma campanulatum. Note conicle-shaped body. MA, macronucle-
us. Bar = 1 /xm. Fig. 10. Scanning electron micrograph of A. campanulatum attached to the gill epithelium (GE). Note
transverse striations of pellicle and its longitudinal furrows (aiTOw). PL, peristomal lip; PS, peristome. Bar = 5^im. Fig.
11. Light micrographs showing Thchodina (T) and Apiosomu (A) infested gill epithelium. Bar = 20 /xm.

Host-Symbiont Relationships

Light microscopy. — Sections of mottled
sculpin gills had no apparent pathological
damage. The conical body of some A. campan-
ulatum appeared to be attached to host gill
surfaces by the scopula, while others were
freely distributed over the epithelial surflice.
Most T. tenuifonnis glide over the surface;
only a few ciliates adhere to the host epithelial
cells (Fig. 11).

Transmission electron microscopy. —
Sections of the interface between the host
epithelial cell and T. tenuifonnis were pre-
pared. No permanent or temporary structure
could be detected between the adhesive disc,
adoral zone of cilia, and gill epithelial cells
(Fig. 12). However, injury to the epithelium
due to 7^ tenuifonnis can be detected by the
number of mitochondria, which decrease and
disappear in the immediate host cell. Host
necrotic tissue, mucous layers from gill epithe-

lium, and particles dispersed in the water
were on the surface of T! tenuifonnis (Fig. 13).

No ultrastructural damage was observed for
A. campanulatum. Presence of this ciliate in-
flicts no serious damage to the host cell. There
was some change in number of mitochondria,
with cristae showing major changes (Fig. 14).

Ecological Aspects of Infestation

In the Provo River near the Provo residential
area, T. tenuifonnis reached the highest infes-
tation rate in April and May. It was uncommon
during summer and autumn and appeared to
be absent in the winter With the increase of
water temperature in spring, ciliates reinfest-
ed the fish. Apiosoma campanulatum at this
site maintained an average of 35% infestation
rate (no. of infested fish vs. no. of total exam-
ined fish) for all seasons.

In the upper Provo River the tendency of in-
festation of T. tenuifonnis corresponded closely

1995]

Ciliated Protozoa in Mottled Sculpin

263

^^.

Y 4^

^m •••• fife'

Figs. 12-13. Transmission electron micrographs of gill epithelium infested by T. teniiiformis. 12. Host necrotic tissue
(arrows) sloughs off for parasite's food. ACC, aboral cilian' complex. Bar - 1 /j,m. 13. Interface between T. teniiiformis
and mucous layer (ML) of epithelial cells. Note damage to mitochondria (M). C, cilia. Bar — 1 fim.

264

Great Basin Naturalist

[Volume 55

y^^;:^^!^^

''«^' iy

^/fon^t.

=^!-, \

Fig. 14. Transmission electron micrograph of gill epithelium infested hy A. campanulatum. A. campanulatwn (A) caus-
es number of mitochondria (M) to decrease and cristae to disappear. M, mitonchondria; Nu, nucleus of epithelial cell.
Bar = Ifim.

to that of the lower area. The highest infesta-
tion rate occurred in May and then decreased
until the next spring. Percentage of fish infest-
ed by T. tenuiformis in the lower river area
was 20.5% vs. 12.5% in the upper Provo River.
Similar to that of the lower river, A. campanu-
latum at the upper site had an average of 37%
infestation in all four seasons. In general, Apio-
soma did not show measurable fluctuations with
seasons.

Discussion

Taxonomy and host-symbiont studies of
Trichodina and Apiosoma infesting fishes in
the United States have received surprisingly
little attention considering the frequency with
which these organisms have been associated
with fish diseases (Khan et al. 1974, Cone and
Odense 1987, Khan 1991). Wellborn (1967)
described 13 species of Trichodina in south-
eastern United States, but few reports have
been published for this ciliate west of the
Mississippi River (Hechmann et al. 1987). Little
information is available on Apiosoma studies
in this country, which is not the case in the
former Soviet Union (Bauer 1984). Cottus hairdi

represents a new host record for Trichodina
tenuiformis and Apiosoma campanulatum.

Comparative Morphology

At the LM level comparison of the adhesive
disc of r tenuiformis with that of other species
oi Trichodina reveals a few similarities. Tricho-
dina reticulata Hirschmann and Partsch, 1955
described from Carassius auratus has denti-
cles similar to T tenuiformis (Bauer 1984). The
adhesive disc of the former has a central light
zone separated into reticulated structures. But
T. reticulata differs in having larger overall
dimensions (average adhesive disc diameter is
60 /xm vs. 25^tm for our material). T tenuiformis
has a close affinity to T elegans described by
Stein (1979) from fish in Russia. The latter is
characterized by an unbroken light zone in the
adhesive disc. Our specimens have various-
sized light forms in the center of the adhesive
disc. To a lesser extent T tenuiformis is similar
to T puijtoraci Lom, 1962 and T domerguei
Dogel, 1940; however, denticle shape and
structure of the adhesive disc clearly distin-
guish T tenuiformis from these species.

Surface features of the adhesive disc and
arrangement of the aboral ciliaiy complex of T

1995]

Ciliated Protozoa in Mottled Sculpin

265

tenuifonnis seen by SEM were generally simi-
lar to those described for T! truttae, an ecto-
parasite on pacific salmon {Oncorliyncluis spp.)
and steelhead trout {OncoHiynclius inykiss;
Arthur and Margolis 1984), and T. labrisomi,
an ectoparasite on hairy blenny {Labrisomas
nuchipinnis; Rand 1993). However, in T. tenui-
fonnis, aboral cilia length is generally shorter
than in those previously described. Further-
more, comparison of the aboral ciliature of T.
tenuifonnis with these species of Trichodina
showed some differences in the extent of
development of the anterior and basal septa,
in velum structure, and in the degree of evi-
dence of the marginal ciliary ring. The anteri-
or septum is relatively large and the basal one
is small in T. tnittae, whereas in T. tenuifonnis
the basal septvmi is prominent. The velum is
well developed in both T. labrisomi and T.
tenuifonnis, but T. tenuifonnis lacks any protu-
berances (Rand 1993). Similar to T. labrisomi,
the marginal ciliary ring of T. tenuijormis is
poorly developed and cannot be distinguished
from the locomotor ciliaiy ring, whereas in T.
truttae the marginal ciliary ring is well devel-
oped (Arthur and Margolis 1984). Rand (1993)
has suggested these marginal ciliature are sen-
sory structures associated with feeding and
orientation. Unlike T! labrisomi and T. truttae,
T. tenuifonnis has no pellicular pores between
denticles and the pellicular ridges on the oral
surface, which might be a species-specific
characteristic for these two species respective-
ly (Arthur and Margolis 1984, Rand 1993).

Over 50 species of Apiosoma have been
recorded from fishes, the majority of which
have been described by Russian authors (Bauer
1984). Although some are common fish para-
sites in some parts of the world, only one ref-
erence concerning Apiosoma piscicola on Salve-
linus fontinalis was reported in North America
(Cone and Odense 1987). There is a paucity of
data pertaining to Apiosoma over the last two
decades, likely reflecting taxonomic difficul-
ties due to variability in ciliaiy structure and
lack of strict host-specificity.

Apiosoma conica has a body shape similar
to A. campanulatum. But our specimens com-
pared more closely to the original description
of A. campanulatum.

The species identifications were based on
original descriptions from Europe; there is a
possibility that the two species described in
this content are not absolutely identical on
both continents.

Host-Symbiont Relationships

Trichodina tenuifonnis is an ectocommensal
with a tendency to be parasitic in mottled
sculpin. There were no visible pathological
symptoms with light microscopy; however,
electron microscopy disclosed changes in the
organelles of host epithelial cells infested by T.
tenuifonnis. Mitochondria decreased in num-
ber and disappeared, which might indicate
respiratory blockage due to lack of oxygen.
This change in mitochondria was obsei-ved in
Trichophrya infesting other fish (Heckmann
and Carroll 1985). Necrotic host epithelial tis-
sue sloughs off following organelle loss, sup-
plying sustenance for the parasite.

No serious damage to mottled sculpin could
be obsei-ved for A. campanulatum. Lom (1973)
suggested that this simple ectocommensal
relationship could change to parasitism in case
of heavy invasions, although this tendency is
much less pronounced than in trichodinids.

Ecological Aspects of Infestation

This study shows that the infestation of Tri-
chodina has both seasonal and regional fluctu-
ations. The higher infestation rate on fish came
from the Provo residential area during the
spring sampling period. Heavy impact from
the local human population may contribute to
this infestation. After summer, the number of
T. tenuiformis gradually reduces with the
decrease of water temperature and reaches
the highest number the following spring. This
may be related to the ciliate life cycle.

Unlike T. tenuifonnis in this study, A. cam-
panulatum maintained a fairly constant infes-
tation on mottled sculpin from the two sites on
the Provo River in all four seasons.

Acknowledgments

The authors thank the Utah Fish and Game
Department for their cooperation on this study.
Dr. Dennis K. Shiozowa provided help with
fish collections. Technical assistance from Dr
John Gardner and staff members of the Elec-
tron Optics Laboratoiy, Brigham Young Univer-
sity, was greatly appreciated.

Literature Cited

Arthur, J. R., and L. Margolis. 1984. Trichodina truttae
Mueller, 1937 (Ciliophora: Peritriehida), a common
pathogenic ectoparasite of cultured ju\ enile salmonid
fishes in British Columbia: redescription and exami-

266

Great Basin Naturalist

[Volume 55

nation I)y scanning electron microscop\'. C-'aiiadiau
Journal ofZooloj,^ 62: 1842-1848.

Bauer, N., EDITOH. 1984. Parasitic protozoa of freshwater
fislies in the USSR. Volume 1. Academy of Sciences
and Zoological Institute, St. Petersburg. 4.30 pp.

ClAHK G. W., AND R. A. Hf.c;kni.\nn. 1984. An atlas of ani-
mal parasites. Brigliam Young University' Press, Proxo,
UT 218 pp.

Co.\E, D. K., AND R M. Odense. 1987. Occurrence of Het-
eropolaria hvoffi (F"aure-Fremiet, 1943) and Apio-
sorna piscicola Blanchard, 188.5 (ciliata) on Salvelinus
fontinaUs (Mitchill) in Nova Scotia, Canada. Cana-
dian Journal of Zoology 65: 2426-2429.

Garcia, L. S., and D. A. Bruckner. 1988. Diagnostic
medical parasitology. Elsevier Science Publishing
Co., Inc., New York. 500 pp.

Heckm.wn, R. A., and T. Carroll. 1985. Host-parasite
studies of Trichophnja infesting cutthroat trout
{Salmo clarki) and longnose suckers {Catostotniis
catostomtis) from Yellowstone Lake, Wyoming. Great
Basin Naturalist 45: 255-265.

Heck.\l\nn, R. a., a. K. Kimball, and J. A. Short. 1987.
Parasites of mottled sculpin, Cottus bairdi Girard,
from five locations in Utah and Wasatch comities,
Utah. Great Basin Naturalist 47: 1.3-21.

Khan, R. A. 1991. Mortality in Atlantic salmon associated
with trichodinid ciliates. Journal of Wildlife Diseases
27: 153-155.

Khan, R. A., V C. Barber, and S. McCann. 1974. A scan-
ning electron microscopical study of the surface

topograpin of a tricliodinid ciliate. Transactions of
the American Microscopical Society 93: 131-134.

LoM, J. 1958. A contriliution to the systematics and mor-
phology of endoparasitic trichodinids from amphib-
ians, with a proposal of uniform species characteris-
tics. Journal of Protozoology 5: 251-263.

. 1973. The mode of attachment and relation to the

host in Apiosoma piscicola Blanchard and Epistylis
hcojfi Faure-Fremiet, ectocommensals of freshwater
fish. Folia Parasitologia (Prague) 20: 105-112.

LoM, J., AND I. Dykova. 1992. Protozoan parasites of fishes.
Elsevier Science Publishers B. V., New York. 316 pp.

R^ND, T. G. 1993. Light and scanning electron microscopic
studies on Tnclwdina labrisomi n. sp. from Labrisomas
niichipinnis (Osteichthyes: Labrisomidae) from
Mangrove Lake, Bermuda. Canadian Jomnal of Zool-
ogy 71: 1855-1860.

Stein, G. A. 1979. Variability of the ciliates of the famiK
Urceolariidae (Peritricha, Mobilina) in the Baikal
Lake. Journal of Protozoology 26 (No. 3, Part 1,
Programs and Abstracts): 36A-37A (abstract).

Van As, J. G., and I. Basson. 1987. Host specificity of tri-
chodinid ectoparasites of fi^eshwater fish. Parasitology
Today 3: 88-90.

Wellborn, T. L. 1967. Thchodina (Ciliata: Urceolariidae)
of freshwater fishes of the southeastern United
States. Journal of Protozoolog>' 14: 399-412.

Received 13 Janiian/ 1995
Accep1ed'l2 April 1995

EFFECTS OF HORSE GRAZING IN SPRING ON SURVIVAL,
RECRUITMENT, AND WINTER INJURY DAMAGE OF SHRUBS

Dennis D. Austin^ and Philip J. Urness^

Abstract. — The use of domestic grazers to shift the growth advantage toward shrubs is a commonly applied tool on
winter ranges managed primarily for big game. Results from horses grazing in spring indicated grazing also benefits
shrub survival, seedling reciaiitment, and reduced winter injuiy damage on some species of shrubs.

Key words: winter range, range management, mule deer horses, shrubs, browse, Utah, revegetation. n^ountain big
sagebrush, Douglas rabhitbrush, true mountain mahogany.

On winter ranges managed primarily for big
game, the management alternative often
selected to maintain the desired mixture of
shrubs and understory herbage is grazing by
livestock in spring. Numerous studies have re-
ported the benefits of spring livestock grazing
to maintain and improve stands of shrubs on
winter ranges (Christensen and Johnson 1964,
Smith and Doell 1968, Jensen et al. 1972, Hull
and Hull 1974, Reiner and Umess 1982, Austin
et al. 1994, and others). However, information
is limited concerning shrub responses to the
effects of livestock grazing with respect to (1)
survival of individual mature plants, (2) seed-
ling recruitment, and (3) winter injuiy damage.
In this study the responses of shrubs to domestic
horse grazing treatments in spring are reported
for Artemisia thdentota var. vaseyana [Rydb.]
Beetle (mountain big sagebrush), Chrysothcun-
nits viscidiflonis [Hook.] Nutt. (Douglas rabbit-
brush), and Cercocarpus montamis Raf (true
mountain mahogany).

Methods

The studv site, located on the foothills east
of Logan, UX 4r46' N latitude, lir47' W
longitude, at 1600 m elevation, contained three
50 X 50-m adjoining paddocks. Within each
paddock the three browse species were hand-
planted from transplants in spring 1983 in 5 X
5 clusters of 25 plants, with 1 m between plant
centers. Seven clusters were planted in each
paddock, with each cluster separated by a min-
imum of 20 m. Before planting, all vegetation
was removed by root plowing; for two growing

seasons following planting, all seedlings were
removed by hand and rototiller weeding.
Between 1983 and 1987 the three paddocks
received equal use by mule deer {Odocoileus
honionus) in winter and no livestock grazing.
A detailed description of the site is found in
Olsen-Rutz and Urness (1987).

This study was conducted during the six
growing seasons between spring 1987 and fall
1992. In spring 1987 all shrub seedhngs that
had become established from seeds were re-
moved from each paddock by hand pulling to
minimize soil disturbance. The number of sur-
viving, previously transplanted shrubs within
each cluster was counted.

Paddocks were randomly assigned a grazing
treatment by horses as heavy, moderate, or
protected. Three to seven horses were used, de-
pending upon herbage production, to obtain
utilization levels of 35-50% and 65-80% for
moderate and heavy treatments, respectively.
Horses were selected as grazers because of tlieir
high foraging selectivity for grasses and avoid-
ance of shrubs, and the managerial opportunity
to manipulate the herbaceous understory to
improve shrub growing conditions (Reiner and
Umess 1982). Paddocks were grazed yearly be-
tween 1 May and 30 June 1987-1991. In 1992
all paddocks were rested from grazing. In the
moderately and heavily grazed paddocks,
herbage production, comprised almost entire-
ly of annual grasses, and percent utilization
were determined from four paired 1-m^ bas-
keted and unprotected plots, randomly placed
in spaces between clusters. Baskets were con-
structed fiom 1.2-m-high netting wire supported

iRangeland Resources, Utah State Universit>', Logan, UT 84322-5230.

267

268

Great Basin Naturalist

[Volume 55

by steel fence posts. Plots were reestablished
and relocated yearly before grazing.

In fiill 1992 all surviving shrubs were counted
by cluster, all seedlings within 10 m of each
cluster were counted, and percent winter injui")
damage was visually estimated. Winter injury
was defined as the amount of dead stems and
twigs as a percentage of total dead plus live
stems and twigs. Damage was estimated at five-
unit increments from 0 to 95%.

Because we were not able to replicate the
three paddocks established for the previous
study, we considered clusters as experimental
units. We agree with Hurlbert (1984), who
described this experimental design as "simple
pseudoreplication," but because of constraints
of time, space, and costs, this design was the
only option. Consequently, we recognize that
differences between treatments could be caused
by inherent differences between paddocks, but
argue that potential spatial error is low due to
adjoining paddocks, identical use during the
three years preceding our experiment, simple
grazing treatments applied, and lack of differ-
ences in the number of surviving seedlings
among paddocks for each species (P > .10) at
the beginning of the experiment.

T tests of the means were used to deter-
mine differences between grazed and protect-
ed plots within and among paddocks. For plant
survival a split-plot design using repeated mea-
sures (1987 and 1992 data) analysis of variance
determined treatment and year effects. One-
way ANOVAs assessed differences among
treatments for species within years. For seed-
ling reciTjitment, because all seedlings were re-
moved in 1987, one-way ANOVAs were used
for species within years. For winter injury
damage, differences between treatments were
analyzed using chi-square tests. A significance
level of P < .10 was used for all tests.

Results and Discussion

Horse Grazing

Horse use reduced herbage at the end of
the grazing period during all years in both the
moderately and heavily grazed treatments (P
< .10), except in 1987 when neither treatment
was different from protected plots. Mean herb-
age utilization during all years was 46% in the
moderately grazed treatment and 71% in the
heavily grazed treatment. Following grazing,
remaining herbage was different between the

moderately and heavily grazed treatments din-
ing all years except 1987. Herbage production
in protected plots was not different from the
moderately and hea\'ily grazed treatments dur-
ing the first three years. However, the heavily
grazed treatment had lower production during
the last two years, suggesting that heavy graz-
ing by horses reduced production of herbage.

Shrub Sunival

Horse grazing increased suwival oil Artemisia
(P = .01) and Cercocarpus (P = .10) but had
no effect on Chnjsothamnus (Table 1). All
three species declined in numbers between
1987 and 1992 (P = .001).

In 1987 the number of surviving plants
among treatments for Aiionisia, ChrysotJiam-
mis, and Cercocarpus was not different (Table
1). However, in 1992 the number of sui-viving
Artemisia plants among treatments was differ-
ent (P = .005). The protected treatment had
lower survival than both the moderately and
heavily grazed treatments (P = .001), but the
moderately and heavily grazed treatments were
not different. Similarly, for Cercocarpus the
number of sui^viving plants among treatments
was different (P = .03). The protected treat-
ment had lower sui-vixal than both the heavily
(P = .005) and moderately (P = .10) grazed
treatments, but the moderately and heavily
grazed treatments were not different. For
Chrysothammis, no differences were found.

Seedling Recruitment

For Artemisia, seedling recruitment was sig-
nificantly different among treatments (P = .08).
The heavily grazed treatment had more seed-
lings than the protected and moderately grazed
treatments (P = .05). No differences among
treatments were found for Chnjsothamnus,
and no seedlings were counted for Cercocar-
pus (Table 1).

Although the low numbers of seedlings
counted in this study require inteipretive cau-
tion, results are consistent with other studies in
which livestock grazing was reported to in-
crease shiiib densitv (Stewart 1941, Christensen
and Johnson 1964, Hull and Hull 1974).
Furthermore, the results from this study, that
horse grazing in spring resulted in higher sur-
vival of mature plants and increased seedling
establishment for several species of shrubs,
are consistent with reports of increased pro-
duction of shrubs following livestock grazing

1995]

Shrub Responses to Grazing

269

Table 1. Plant sundval (total niiinber/paddock), seedling recruitment (total niimber/paddock), and winter injin-\' dam-
age (mean % per shmb) o{ Artemisia tridentata (ARTR), Chnjsothainniis viscidiflorus (CHVI), and Cercocarpus montanus
(CEMO), as affected by heavy (H), moderate (M), and protected (P) horse grazing treatments'.

Species

Treatment

1987

1992

Shrub suni\al

ARTR

CHVI

CEMO

Seedling recnjitment^

ARTR

CHVI

Winter injur}' damage'^

CEMO

Number/paddock

119

91«

128

93^'

120

42''

125

110

106

101

164

158"

161

140ab

168

119'^

Number/paddock

18"

—

5''

—

Mean

% per shiTib

11"

—

241'

—

41c

^Data with different superscripted, lowercase letters witliin year and species were different at P < .05.

^No seedlings of CEMO were found.

''No winter injury damage on ARTR or CHVI was found.

with horses (Reiner and Urness 1982, Austin
et al. 1994), sheep (Jensen et al. 1972), cattle
(Smith and Doell 1968), or goats (Riggs and
Urness 1989).

Winter Injury

Winter injuiy was not found on either Artem-
isia or Chrysothamntis (Table 1). For Cercocar-
pus, winter injury among treatments was dif-
ferent (F = .001), with highest damage occur-
ring on the protected treatment, medium on the
moderately grazed treatment, and lowest dam-
age on the heavily grazed treatment. All treat-
ments were different fi-om each other (P = .001).

Winter injury has been reported for many
shrub species, including Cercocarpus (Nelson
and Tiernan 1983). However, only one known
report compared winter injury to grazing.
Contrary to our results, Jensen and Urness
(1979) compared heavy (70%) and moderate
(35%) levels of grazing of grasses and forbs by

sheep and reported that injuiy to Purshia tri-
dentata (antelope bitterbrush) was indepen-
dent of grazing intensity or time of use.

Summary

Our results support the use of grazing by
horses of herbaceous understoiy in spring to
maintain and improve stands of browse for
winter use by big game. Herbage production
was reduced by heavy grazing, survival of
mature plants of Artemisia and Cercocarpus
was increased, recruitment oi Artemisia was
increased, and winter injury to Cercocarpus
was decreased. No negative effects on shrubs
from grazing by horses were found.

Acknowledgments

This report is a contribution of the Utah
State Division of Wildlife Resources, Pittman-
Robertson, Federal Aid Project W-105-R.

270

Great Basin Naturalist

[Volume 55

Literature Cited

Austin, D. D., F. J. Urness, and S. L. Di hiiam. 1994.
Impacts of mule deer and horse grazing on trans-
planted shrubs for revcgetation. Journal of Range
Management 47: 8-11.

CHR1STEN.SEN, E. M., AND H. B. JoHNSON. 1964. Presettle-
ment vegetation and vegetational change in three
valleys in central Utah. Brigham Young University
Science Bulletin, Biological Series IV(4): 1-15.

Hull, A. C, Jr., and M. K. Hull. 1974. Presettlement
vegetation of Cache Valley, Utah and Idaho. Journal
of Range Management 27: 27-29.

HURLHERT, S. H. 1984. Pseudoreplication and the design
of ecological field e.xperiments. Ecological Mono-
graphs 54: 187-211.

Jensen, C. H., A. D. Smith, .\nd G. W. Scotter. 1972.
Guidelines for grazing sheep on rangelands used by
big game in winter. Journal of Range Management
25: 346-352.

Jensen, C. H., and P J. Urness. 1979. Winter cold damage
to bitterbnish related to spring sheep grazing. Joimial
of Range Management 32: 214-216.

Nelson, D. L., and C. E Tiernan. 1983. Winter injurv- of
sagebrush and other wildland shiiibs in the western
United States. USDA Intermountain Eorest and
Range E.xperiment Station Research Paper INT-314.

Olsen-Rutz, K. M., and R J. Urness. 1987. Comparability
of foraging behavior and diet selection of tractable
and wild mule deer. Utah Division of Wildlife
Resources Publication No. 88-3.

Reiner, R. J., and P J. Urness. 1982. Effects of grazing
horses managed as manipulators of big game winter
range. JoiuTial of Range Management 35: 567-571.

RiGGS, R. A., and P J. Urness. 1989. Effects of goat brows-
ing on Gambel oak communities in northern Utah.
Journal of Range Management 42: 354—360.

Smith, A. D., and D. Doell. 1968. Guides to allocating
forage between cattle and big game on winter range.
Utah Division of Fish and Game Publication No. 68-
11.

Stewart, G. 1941. Historic records bearing on agriculture
and grazing ecology in Utah. Ecolog>' 39: 362-375.

Received 7 Fehnuinj 1994
Accepted 24 January 1995

NORTH AMERICAN TYPES OF OXYTROPIS DC. (LEGUMINOSAE) AT THE

NATURAL HISTORY MUSEUM AND ROYAL BOTANIC GARDEN, ENGLAND,

WITH NOMENCLATURAL COMMENTS AND A NEW VARIETY

S. L. Welshi

Abstract. — Specimens of Oxytropis in the herbaria of The Natural Histoiy Museum and Royal Botanic Garden
were examined to interpret their role in nomenclature. This is the first attempt at a systematic ovei'view of specimens so
important in understanding the genus as it occurs in North America. The review of specimens at BM and K during the
present research has resulted in realignment of names of some of the ta.\a. Oxytropis cainpestris var. gracilis (A. Nelson)
Barneby is recognized herein as being predated by O. campestris var. spicata Hook., O. sericea var. spicata (Hook.)
Barneby is replaced by O. sericea var. speciosa (Torn & A. Gray) Welsh comb, nov., and O. campestris var. terrae-novae
(Fern.) Barneby is superseded by O. campestris var. minor (Hook.) Welsh comb. nov. One new taxon is proposed;
Oxytropis deflexa (Pall.) DC. var. pidcherrima Welsh & A. Huber, var. nov. Lectotypes are designated for the following
taxa: Astragalus retroflcxus Pall., Oxytropis arctica R. Br., O. arctica var a subumbellata Hook., O. arctica var. [J uniflora
Hook., O. campestris var. P speciosa Torr. & A. Gray, O. campestris var. ^ melanocephala Hook., O. campestris var. 5 spi-
cata Hook., O. multiceps var. minor A. Gray, O. splendens Douglas ex Hook., O. splcndens (3 richardsonii Hook., O.
uralensis (3 subsucculenta Hook., and O. uralensis y minor Hook.

Key words: Legiiminosae, Ox\'tropis, nomenclature. Natural History Museum, Royal Botanic Garden.

Concepts of species within a genus undergo
an evolutionary progression through time as
additional information is obtained. A review of
the history of botanical treatments of the
genus Oxytropis parallels that of other genera
in North America, wherein the early explora-
tions, researches, and publications were under-
taken by explorers and scientists from the Old
World, especially from England. Specimens
arriving from the New World were compared
to those of the same genera from Eurasia.
Specimens of Oxijtropis from Russia and other
regions with arctic, subarctic, or boreal floras
arrived piecemeal at herbaria in Europe,
where important collections accumulated, par-
ticularly at the Royal Botanical Garden at Kew
(K) and the Natural History Museum (BM;
fomierly die British Museum [Natural Histoiy])
in London. These materials formed the basis
for comparison with North American speci-
mens. Some American plants were similar and
were given the same names as some Old
World species. The earliest revision of Oxy-
tropis based on NoiiJi American specimens was
that of William Jackson Hooker (1785-1865),
whose concepts of species, set forth in the
Flora horeali-americana (Hooker 1831), were

to have a profound effect on all later inteipre-
tations of the genus. Annotations of the speci-
mens at K and BM present a history of the use
of epithets by various botanists interested in
this fascinating genus. However, of the revi-
sionaiy workers on North American members
of the genus Oxytropis, only Asa Gray appears
to have systematically studied the historical
collections at the Royal Botanical Garden, Kew,
and no one has examined all materials of the
genus in North America at the Natural History
Museum. Various workers on regional
floras, M. L. Fernald, A. E. Porsild, and N.
Folunin, have annotated part of the specimens,
and R. C. Barneby, whose revision (1952) is a
classic presentation of the genus in North
America, has examined selected material.

The pui-poses of this paper are to clarify the
status of historical specimens, to record their
place of deposit, and to trace their nomencla-
tural histoiy as it affects interpretation of the
genus in North America. Plants at BM and at
K are the center of focus for this treatment.
However, the location of duplicate types in
various herbaria in the United States is also
included where that information is known.
Names and synonyms of the North American

'Life Science Museum and Department of Botany and Life Science, Brigham Young University', Provo, UT 84602.

271

272

Great Basin Naturalist

[Volume 55

taxa were reviewed previously by Welsh
(1991). Abbreviations of the repositories follow
Holmgren et al. (1990).

The synopsis of Oxytropis in Flora horcali-
amevicana (hereinafter Flora) by Hooker is
relevant to an understanding of many of the
names in the following list. This is true even
though Hooker had not seen the materials in
the field, and even though his concepts were
based on limited and often inadequate materi-
als for a definitive understanding of the taxa.
The Flora has no introduction outlining the
scope and rationale for the treatment. It does,
however, contain a ven* detailed title page:

Flora Boreali-AnieriL'ana; or, the Botany of the
Northern Parts of British America: compiled prin-
cipally from the plants collected by Dr Richardson
& Mr Drummond on the late northern expedi-
tions, under command of Captain Sir John
Franklin, R.N. to which are added (by permission
of the Horticultural Society of London,) those of
Mr Douglas, from North-West America, and of
other Naturalists.

That Hooker does not mention the collections
of Captain William Edward Parry and his
associates is not to be considered an oversight;
their materials had been published previously
by Hooker (1825), and their specimens are by
no means neglected in the Flora. It is unfortu-
nate that the specimens on which the names
were based were not routinely so noted by the
authors.

The dedication in the Flora by Hooker
honors both Franklin and Richardson,

under whose auspices, as Commander and
Naturalist of two separate expeditions to The Polar
Seas, a great portion of the more rare and interest-
ing plants that ornament this volume were collect-
ed, under circumstances of singular difficulty,
hardship, and danger.

In this important pioneer work. Hooker recog-
nized only 10 species of Oxytropis but regard-
ed several of them as consisting of variants
designated by Greek letters, some of which
were followed by epithets. The names in order
of their appearance are O. horealis; O.
Uralensis a, ^ siihsucciilenta, jfninor; O. arcti-
ca a, |3 minor, 5 inflata; O. foliolosa; O. argen-
tata; O. lambertii; O. nigrescens; O. carnpestris
a, y siilphiirca, 6 spicata, £ glahrata, and t,
melanocepliala; O. splendens a vestita and (3
richardsonii; and O. deflexa. These names
have occurred in subsequent literature and
have been accounted for by various workers of

the genus, often without citation of type mate-
rial or place of deposition. An attempt is made
here to associate all names noted by Hooker
with their 20th-century equivalents. Hooker's
personal herbarium, containing many of the
Oxytropis types, is at Royal Botanic Garden (K
Hooker); those of Richardson, Douglas,
Drummond, and Parry and his associates are
represented in some part at both K and BM.

Richardson (1823) cited species oi' Oxytropis
based on his owai findings, but he did not name
any as new. The species treated by him include
O. oxyphylla, O. deflexa, O. carnpestris, O.
argentata, and O. uralensis. Of these, only O.
deflexa stands almost in the same sense today.

The treatment of Oxytropis by Torrey and
Gray (1838) followed Hooker's account in
nearly all details, but it added six new species
collected and described by Thomas Nuttall
(1786-1859) on his journey across the conti-
nent with Wyeth in 1834. These were the first
American species described by a botanist who
had seen them in the field. The diversity of
the species in the American West must have
seemed overwhelming even to Nuttall, who
proposed several additional species of Oxy-
tropis not published in Torrey and Gray's
monumental work. Specimens in Nuttall's
personal herbarium, which is deposited at
BM, are evidence of his belief in a greater
number of species than would be published
subsequently. Some of the Nuttall names were
later cited as synonyms, but some were not
mentioned at all. Nuttall was unfamiliar with
most meml)ers of the genus in the Old World,
and some of his proposals reflect that lack of
understanding. The difficulty in comprehend-
ing a genus as large and complex as Oxytropis
is understandable and is not confined to the
19th centuiy

Asa Gray (1810-1888) became the most
important 19th-century North American devo-
tee to the genus, revising it twice, once in
1863 and again in 1884. Concepts in the revi-
sion of Oxytropis by Gray (1884) were influ-
enced by his examination in 1880 of specimens
at Kew, which had formed the basis of the
treatment by Hooker in die Flora, and fiom the
large number of specimens in eastern American
herbaria collected during the inten'ening se\'-
eral decades. Gray's 1884 publication includ-
ed plants from a broader geographic area than
those examined by Hooker and contained
descriptions of 16 species. Specimens collected

1995]

North American Types of Oxytropis

273

by Nuttall, especially, and several other west-
ern American explorers formed the basis of
additional species not included in his and
Hooker's earlier works. Gray accounted for
some proposals, bringing the concepts of the
genus to date as new materials had accrued.
He accounted for some but not all taxa treated
by Hooker, e.g., O. borealis DC. under O. leii-
cantha (Pall.) Pers. The name O. Iciicantha
(Pall.) Pers. was long considered a potential
replacement for viscid members of the genus
in North America. The type was examined by
Welsh (1977) and the name excluded from
inteipretation of North American taxa. It is a
portion of the O. campestris complex in Siberia.
Most of the other names treated by Hooker
were ignored by Gray, cited in synonymy, or
provisionally included in other entities.
Apparently Gray did not see all pertinent his-
torical material in England, particularly not
that at BM. It is evident, likewise, that not all
ambiguities are resolved by the research lead-
ing to this paper. Nevertheless, as noted by
Barneby (1952), "the resulting synopsis of the
genus [by Gray] in 1884 stands as a small but
enduring monument to his genius."

Barneby (1952), in a classic account, recog-
nized 22 species and 21 additional infraspecif-
ic entities as occurring in North America. He
accounted for all names used previously in
North America, with problematical names
being discussed in a list of excluded and
imperfectly known species.

A summary treatment of the genus by
Welsh (1994) for the Flora North America pro-
ject likewise treats 22 species, somewhat
realigned from those of Barneby, but recog-
nizes 35 infraspecific taxa. Many names pro-
posed at infraspecific rank are from Arctic
regions of the continent, areas whose collec-
tions were not well represented in herbaria
prior to 1950.

Students of plant taxonomy must examine
authentic materials of all previously named
taxa, whether currently recognized or not. In
the 19th centuiy David Douglas (1798-1834),
Thomas Drummond (ca 1780-1835), Thomas
Nuttall, and Sir John Richardson (1787-1865)
were the most important contributors of speci-
mens on which North American names in
Oxytropis are based. Repositories for their
specimens, later designated as types, are
Philadelphia Academy of Sciences (PH), New
York Botanical Garden (NY), and Gray

Herbarium (GH) in the United States, and
Royal Botanic Garden at Kew (K) and Natural
History Museum (BM) in England. Later in
the 19th centuiy several other workers gath-
ered specimens that were considered new to
science; these were deposited in numerous
other herbaria in addition to those cited
above.

David Douglas was an intrepid Scottish
botanical explorer of North America whose
contributions to Oxytropis came from his jour-
ney across the continent mainly in 1826-27.
He collected the specimens on which the con-
cept of O. splendens was based. According to
Stafleu and Cowan (1976), the first set of his
North American plants is at K; his own her-
barium is partly at BM and partly at CGE.
Hooker based his treatment of Oxytropis in
the Flora in part on collections by Douglas
(Stafleu and Cowan 1979), as evidenced by
specimens at K.

Sir John Richardson, British (Scottish)
explorer and naturalist with the Royal Navy,
was a medical doctor who accompanied Sir
John Franklin on two expeditions, 1819-22
and 1825-27, and later (1848-49) commanded
an expedition in search of Franklin who was
lost on an ill-fated sea voyage of 1845-47 in
search of the Northwest Passage. Richardson's
herbarium of vascular plants is mainly at BM,
with further material at K and elsewhere
(Stafleu and Cowan 1983). The first expedi-
tion, in 1819-22, was from Great Slave Lake
to the Coppermine River, down which they
traveled to Coronation Gulf, and then cross-
country to the Coppermine in winter of
1821-22. Only 9 of 21 on the expedition sur-
vived the ordeals of hunger, cold, and exposure;
that anyone sumved is a tribute to persistence
of the men and aid of local aborigines who res-
cued them from certain death (Houston 1984).
The expedition is remembered as one of the
most deadly in the histoiy of biological inves-
tigations in North America. Both Richardson
and Franklin barely escaped with their lives.
The second expedition was down the
Mackenzie River to the Polar Sea, with
Franklin exploring westward and Richardson
eastward along the coast to Coronation Gulf
and return. The many plant and animal names
proposed by Richardson, and those named
after him, serve as a tribute to the genius and
perseverance of this remarkable man.

274

Great Basin Naturalist

[Volume 55

Edward Sabine (1788-1883), John Edwards,
James Clark Ross (1800-1862), William Edward
Parry (1790-1855), Alexander Fisher, and
Charles James Beverley collected plants on
the first Parry journey in search of the North-
west Passage (Parry 1821). It is evident from
the introduction to Brown's (1824) treatment
that each of those named, mainly medical doc-
tors with various ships, made their own collec-
tions, which were placed initially in their pri-
vate herbaria. Sabine was astronomer to the
Arctic expeditions led by Pany and collected
plants in Melville Island and Greenland.
Edwards was surgeon to the Hecla on Pany s
voyages of 1819, 1820, 1821-23, and Fisher
was assistant surgeon on the Hecla in the
1819-20 voyage, while Beverley was assistant
surgeon on the Gripper. Ross was in the Royal
Navy with the Parry expeditions in 1819-20,
1821-23, 1824-25, and 1827. Parry was a
British explorer who commanded expeditions
to the Arctic in search of the Northwest
Passage. Specimens from the Parry expedi-
tions are deposited at both K and BM, with
some of the collectors' private herbaria better
represented at BM and some at K.

Peter Simon Pallas (1741-1811), important
for his exploration of Russia and his work on
the Russian flora, named Astragalus deflexus
and Astragalus retroflexus, later included in
Oxijtropis, names whose interpretations bear
on North American species of the genus.
According to Stafleu and Cowan (1983) the
main personal herbarium of Pallas was sold at
a London auction in 1808 to A. B. Lambert,
who subsequently sold one part to Robert
Brown and another to William Robertson.
Both parts presently are at BM and are perti-
nent to this paper; additional parts reside else-
where, but they are not the basis of this treat-
ment except for an important set at LE, which
evidently contains the type specimen of
Astragalus deflexus, which has not been exam-
ined for this treatment.

The author wishes to thank the curators at
BM and K for their cooperation in providing
specimens on loan, and for their hospitality
during a visit to London. Also acknowledged
is Dr. Rupert Barneby, who read critical por-
tions of the manuscript and who provided sug-
gestions and encouragement.

1. Aragallus abhreviatits Greene, Proc. Biol. Soc.
Wash. 18: 12. 1905.

= ()xytr(>i)is liiinhciiii Fiirsh \ar. articulata (Greene)
Harneh}'

Type; "Te.xas, near Dallas, Limestone prairie, Dallas
County, J. Reverchon 603, May 1876"; holotype NDG!;
isotypes NY!, BM!; "Dry calcareous soil near Dallas,
Texas. Curtis 603, April, May"; paratype GH!, NDG!, NY!

2. Ar«^rt//».v aven-nel.sonii Lunell, Bull. Leeds Herb. 2:
6. 1908.

= Oxijtropis lainbeiiii Pursh var. Uimbertii
Type: North Dakota, "Aragallus Aven-Nelsoni
Lunell, n. sp. Butte, Benson Count\', N. Dak., June 14, 21,
July 2,1908," legit J. Lunell; holotype .MIN?; isotypes
BM!, NDA!, MIN!, NY!, US!, WTC! '

3. Arafiallus invenustus Greene, Proc. Biol. Soc. Wash.,
18: 12. 1905.

— Oxijtropis sericea Nutt. var. sericea

Type: "South Dakota, about Fort Meade, Meade
County, W H. Fonvood 96a, 96b, 3 June 1887, 96b, 7 June
1887";'syntypes US!, photo BRY!, K Hooker!

4. Aragallus majusculus Greene, Proc. Biol. Soc.
Wash., 18: 12. 1905.

= Oxijtropis sericea Nutt. var sericea

Type: Utah, Heniy Mts., Gai-field County, Utah, M. E.
Jones 5674, July 1894; holotype US!; isotypes NY!, MO!,
BM!, photo BRY!

5. Aragallus mctcalfei Greene, Proc. Biol. Soc. Wash.,
IS: 12. 1905.

= Oxijtropis lambertii Pursh var. bigelovii A. Gray
Type: New Mexico, Sawyer's Peak, Grant County,

open glade, ca 10,000 ft, O. B. Metcalf 1079, 7 Julv 1904;

holotype US!; isot>'pes NY!. CAS!, GH!, POM, WO^C, BM!

6. Astragalus deflexus Pall., Acta Acad. Sci. Imp.
Petrop. 2: 268. 1779."

= Oxijtropis deflexa (Pall.) DC. var. deflexa

Type: "ad nivalia Dauriae . . . circa Balyra rivum

aliosque Ononem influentibus" [Siberia], P S. Pallas s.n.;

holotype LE.

7. Astragalus retroflexus Pall, Sp. Astragal., 33, tab. 27.
1801.

= Oxijtropis deflexa var. deflexa

Original location: Provenit haec species tantuni in
alpinis transbaicalensibus et circa lacum Baical, praeser-
tim in scaturiginosis frigidus, circa fontes rivulorum
Baltschikan, Carol, Bargusin et Chilik; verosimillime
quoque per omnem alpestrem tractum, Sibiriam a
Sinarum Imperio diffinientem.

Type: Pallas, s.n.; lectotype (here selected) BM
(#45444)!

There are two Pallas collections labeled Astragalus
deflexus at BM, 45443 and 45444; the former (paratype)
with three stems bears juvenile to mature fruit and a label
in Russian script, the latter (lectotype) with a complete
plant in flower and two racemes (one in flower and one in
immature fruit). Appearing on the sheet with the BM
number 45444 on the label are "No. 10" and illegible
script. Both collections simulate what has passed in North
America as O. deflexa var. sericea Torr. & A. Gray. Sheet
45444, here designated as lectot>pe for Astragalus retro-
flexus Pall., is a close match for the illustration in Table 27
of Pallas' Species Astragalorum (1800-1803), except that
the drawing is a mirror image of the actual specimen. The

1995]

North American Types of Oxytropis

275

reversal of the image comes from the use of the copper
phite on which the original drawing was produced.
Flower buds, leaves, and leaflets are the same. Only the
flowers and their proportions are slightly different; those
of the drawing are much too distinct and perhaps too
large. The fruiting branch represented in Table 27 is from
sheet 45443, again in reverse image, and taken from the
branch at the upper left. Pallas (1800-1803) notes that the
plant was growii in a garden at St. Petersburg and flow-
ered and produced fniit the second year It seems that one
specimen or both cited above are from plants grown in
the garden at St. Petersburg.

8. Oxytropis argentata sensu Richardson, Frankl. 1st
Jour., Bot. Append. 745. 1823.

= O. sericea var. speciosa (Torn & A. Gray) Welsh,
pro parte

Authentic specimen; "British North America. Dr
Richardson 1819-22. Astr. argentatus Pallas Astr. Carlton"
(BM #45476!).

9. Oxytropis arctica R. Br., Chloris Melvill. 20. 1823.
T>'pe locality: Canada, "Melville Island, Pariy's First

Vo\age, Sabine, Edwards, Ross, and others, 1819-1820"
(R. Brown I.e.).

Type: "Melville Island, coll. b\' Mr. Beverley"; lecto-
type (here selected) K Hooker!; probable isolectoty-pes S!,
GH!

The reverse of the Beverley sheet contains the nota-
tions, "Winter Harbour, 23 July 1820 [flowering materi-
al?]" and "Winter Harbour, 4th July 1820 [fruiting speci-
men?, the portion here selected as lectotype]." The
Beverley material is the most complete for the species of
any of those collected by the Parry expedition members
and includes both flowering and fruiting material; the
Sabine paratype cited below also has a flowering and a
fruiting branch.

Brown (1823) did not designate any specimens to
support his new species. However, in the introduction to
his list of plants collected in Melville Island, he lists the
herbaria of the officers of the expedition on which the list
was based. Included are "Captain Sabine, Mr. Edwards,
Mr. James Ross, Captain Parry, Mr. Fisher, and Mr.
Beverley, whose names are here given in order of the
extent of their collections."

The following are paratypes of Oxytropis arctica at
BM and K: "36. Oxytropis arctica. [illegible]. Melville Id.
Mr. James Ross" (BM!); "Parry's First Voyage 1819-20.
Melville Island. 11th August 1820, (BM!)"; "Melville Id.
Capt. Pany. 36. O.xytropis arctica" (45446 BM!); "Melville
Isld., Sabine" (K Hooker!); and "O.xytropis arctica. Mr.
Edwards" (BM!).

The literature citation for O. arctica has traditionally
been given as "Parry's First Voy., Append. 9: 278. 1824."
The list of plants by Robert Brown was published twice,
however, once as "Chloris melvilliana," in 1823 and subse-
quently in the appendix to Parry's first voyage. Evidently,
Chloris melvilliana was published a year prior to the
appearance of identical material under different pagina-
tion in the Parry appendix. On an introductory page in
Chloris melvilliana is written tlie following: "The follow-
ing List is printed as Nl. XI of the Appendix to Captain
Parry's Journal of the First Voyage, commencing at page
cclix." It is evident that at least page proofs of the
Appendix were a\'ailable at the time the Chloris was read\'
for printing, and that they formed the basis of the Chloris.

9a. Oxytropis arctica van siihwnbeUata Hook, in W. E.
Pairy, Sec. Voy. 4: 396. 1825.

= O. nigrescens var. nigrescens

Type locality: "Arctic shores and Islands of North
America. Capt. Sir E. Parry; Dr. Richardson; Capt. Sir
John Franklin; Capt. Back, & c." (Hooker 1831: 146).

Type: "Dr Richardson. 1/146. Oxytropis arctica Br.
var sul^umbellata. Coast"; lectotype (here selected) BM!

The lectotype at BM consists of four specimens of
Oxytropis nigrescens var. nigrescens, and they are mount-
ed on one sheet with four specimens of O. nigrescens var.
uniflora bearing the label, "Oxytropis arctica (i, Frankl.
Exp. Dr Richardson. "

9b. Oxytropis arctica var. uniflora Hook, in W E. Pany,
Append. Parry Sec. Voy. 4: 396. 1825.

= O. nigrescens (Pall.) Fischer var uniflora (Hook.)
Barneby

Type: "Barrow River, E coast Melville Peninsula, lat.
67°2rN, on Parrv's Second Voyage, Edwards s.n. 1821-3";
lectotvpe (here selected) BM!; isolectotypes K Hooker!,
GH!, NY!

The Barrow River lectotype at BM, a mere fragment
with three flowers, is mounted with a second much more
complete collection, "Igloolik. Mr. Edwards, Pany's 2nd
Voyage" (a paratype). The collection from Igloolik consists
of several flowering specimens and one with a solitary
fruit. Additionally, there are four almost mature fruits, two
of which have been opened displa\ing the septum. There
are several sheets of this variety at BM taken on various
Parry voyages (e.g., BM 45452, "O.xytropis arctica, Barrow
River, 1822"). The isolectotype at K Hooker is similarly a
mere fragment.

9c. Oxytropis arctica 6 inflata Hook., Fl. Bor-Amer 1:
146. 1831.

= O. podocarpa A. Gray

Type: "Highest summits of the Rocky Mts.
Drummond' ; holotype K Hooker!

The Drummond material at K consists of six plants,
one of which is in young fruit; they are mounted with two
specimens by Beechey from Kotzebue Sound, both of which
appear to be O. nigrescens. Gray cited the Drummond
material with several other specimens when he described
O. podocarpa (q.v.).

9d. Oxytropis arctica, "varietas notabilis," R. Br, Chlor.
Melvill. 51. 1823.

= O. nigrescens (Pall.) Fischer var uniflora (Hook.)
Barneby

The name as noted above was cited by Barneby
(1952) as not validly published; it is here included for con-
sistency of use in North American literature on the genus
Oxytropis. Hooker (1825), in the botanical appendix to
Parr\'"s Journal of a Second Voyage, notes in his discussion
of O. arctica:

This variet\' (|3.) is noticed by Mr. Brown at the end
of his valuable Remarks on the Flora of Melville
Island, as discovered by the gentlemen of the present
expedition, and says of it, "Varietas notabilis, vix
enim distincta videtur species, statura minor, scapo
unifloro passemque umbella biflora, dentibus calycis
respecti tubi paulo longioribus, foliolis saepe 7, quan-
doque 7, \illis persistentibus utrinque argento-seri-
ceis." ... To these remarks I may add, that the plants
are not above half the size of a [subiimbellata], the
stems less woolly, the leaflets fewer, denser, and cov-

276

Great Basin Naturalist

[Volume 55

ered with short, ven white silk\ hairs. Thi- pwlvincle
scarcely rises above the leaves, and each subtended
by small bractea. The corolla is of the most beautiful
deep purplish blue; the calyx and legume black from
the (juautitN of black hairs; but these are mixed with
.several longer white ones. The contrast between the
deep blue of the corolla and the dense, white and sil-
very lea\'es render this a most loveK little plant.

The .specimens arc still silvery white and beautiful after
more than 17 decades.

10a. Oxytropis arctohia Biinge, Mem. Acad. Imp. Sci.
Saint-Petersbourg 22: 114. 1874.

= O. ni<iref>cens (Pall.) Fisch. \ar. uniflora (Hook.)
Barneby

Type: "Habitat in arcticis Americae borealis; v. s. sp.
in herb, olim Fischerano nunc h. bot. Petrop. " (Bunge
I.e.).

This was based b\' Bimge (1874) exactly on a speci-
men of O. arctica (3 uniflora Hook., in the Fischer herbari-
um at LE. The name was utilized by some (Polunin 1940)
for North American specimens.

10b. \ar. hyperarcticd Pohmin, Bot. Canad. E. Arctic,
293. 1940.

= O. nigrescens (Pall.) Fisch. var uniflora (Hook.)
Barneby

Type: "Franklin district, Baffin Island, Arctic Bay, N.
Polunin 2583, 8-11 Sept. 1936"; holotype CAN; isotypes
GH!, BM!, OXF!

11. Oxytropis borealis DC. Prodr. 2: 275. 1925.

It appears that the name was not used by Hooker in
the modern sense as interpreted by Welsh (1990). The
name appears on a sheet of O. maydelliana Trauts'. in the
Hooker herbarium at K. It is, however the earliest name
available at species rank for the viscid-glandular material
that has passed under O. viscicla Nutt., and other taxa (see
various uses elsewhere in this paper).

12a. Oxytropis campestris (L.) DC. var. davisii Welsh,
Leafl. W Bot. 10: 25. 1963.

Tvpe: "British Columbia, mi 403.4, Alaska Hwv', R. J.
Davis 6076, 19 JuK 1962"; holotype BRY!; isotv'pe IDS!

Distribution; SW Alberta, NE British Columbia.

This ta.xon has been represented in herbaria since
early in the 19th century. Specimens by Douglas and
Drummond, almost assuredh' assignable to this taxon, are
present at BM and K. Their collections formed portions of
the concepts O. lamhertii (3. Hook, (c^.v.) and O. uralensis y
minor Hook. (q.v).

This plant is readily distinguished by its colorful
flowers, fasciculate leaflets or tendency to fasciculate
leaflets, and elongate inflorescences. Specimens have
been considered as intermediates between members of
the Oxytropis campestris complex and O. borealis var vis-
cida, or they have been misidentified as O. splendens
because of the fasciculate leaflets. Plants of van davisii are
localK' abimdant on stream gravels and adjacent slopes in
the focjthills mainly of the Alberta Rockies and in north-
eastern British Columbia. Intermediates between var.
davisii and O. sericea var. speciosa occur in northern British
Colinnbia. Another variety with pink-puiple flowers, var
roaldii (Lindstr.) Welsh, occurs in northern Yukon
Territory and adjacent Alaska. The northern plant is of low
growth and has fewer flowers than van davisii. A similar

l)air of low \ersus tall varieties, partialK s\mpatric, exists
in eastern Canada, i.e., van johannensis Fernald and van
minor (Hook.) Welsh [var. terrae-novae (Fernald)
Barneby].

12b. Oxytropis cam))estris e glabrata Hook., Fl. Bor-
Amen 1: 147. 1831.

= O. maydelliana Trautv.

Type locality: "Bear Lake to the Arctic Shores and
Islands, Dn Richardson; Capt. Sir J. Franklin and Capt.
Back; Capt Sir E. Parry, &c." (Hooker I.e.).

Type: Bear Lake to the Arctic Shores and Islands;
neotype GH!

Hooker characterized the ta.xon "foliolis glabriuscu-
lus subsucculentis." No authentic material of van glabrata
was discovered at either K or BM.

Hooker's statement of the locality' infomiation applied
to both vars. glabrata and melanocephala. The specimen
at GH, annotated by Gray indicating that it was based on
Hooker's van glabrata, is here designated as neotype.
Thus, the name is fixed in the sense used by Barneby
(1952).

12c. Oxytropis campestris van johannensis Fernald,
Rhodora 1: 88. 1899.

Type: "Maine, gravelly shores, valley of St. John
River, Fort Kent, Aroostook County, M. L. Femald 2289,
15 June 1898"; holotype GH!; isotypes CAN!, US!, NY!,
BRY!, BM!

Distribution: Newfoundland, New Brunswick, Nova
Scotia, Quebec, Ontario, and Maine.

Plants of this variety from the Farm Riven south of
James Bay, Ontario, have fascicidate leaflets and short
pods. In the latter feature they simulate van chartacea
(Fassett) Banneby, which might best be regarded as only a
disjunct phase of this varietv'.

12d. Oxytropis campestris ^ melanocephala Hook., Fl.
Bon-Amen 1: 147. 1831.

= O. maydelliana Trautw

Synonym: O. maydelliana ssp. melanocephala
(Hook.)' Porsiid

Type locality: "Bear Lake to the Arctic Shores and
Islands, Dn Richardson; Capt. Sir J. Franklin and Capt.
Back, Capt. Sir E. Pany, &c." (Hooker I.e.).

Type: "24 July 2-Augt [?] 1826. O. camp. ^. Dn
Richardson"; lectotype (here selected) BM!; isolectotvpes
GH!, K Hooker! ("^. Arctic Sea. Richardson. Oxytropis
campestris, ").

Paratypes: "Capt. Parn'. 2nd Voy. [and] Parn 's 2nd
Voy.," both at K Hooker!; and "O.xytropis campestris.
Duke of York's Bay. Arctic Regions. Parry's 2nd Voyage,"
BM 45449! Several additional specimens from Pane's sec-
ond voyage are also present at BM.

The Richardson isolectotype at K consists of two
specimens, both with stipules pale and merel\- mottled
with piuple instead of puiplish overall as in most speci-
mens of O. maydelliana. Because they are, however,
somewhat unusual among specimens of the species, they
are not chosen as lectotvpe. Specimens at BM more close-
ly represent the concept of the ta.xon. The isolectotype at
K is mounted on a sheet with two collections by Capt.
Parry from the second voyage, and a third collection by
Simpson from the "Polar Sea." The Party collections are
apparent paratv'pes of t, melanocephala Hook. An extrane-
ous stem of O. borealis van hudsonica, mounted on the
same sheet, appears to belong to one of the Pany collec-

1995]

North American Types of Oxytropis

277

tions. On the lectot\pe specimen at BM the name "Dr.
Richardson" was obviously added later as it is on most
specimens attributed to him, and the date 1826 is coiTect
for the second Franklin expedition. There is a second sheet
at BM!, "1/147. O.xytropis campestris DC. River Rae, '
with the name Dr. Richardson written below the label,
and bearing the notation, "0.\\tropis campestris. DC. P
sordida Lin." Possibly it is a paratope of van rnelanocephala.
Many of the specimens from Parn s second voyage desig-
nated as O. campestris are O. inaydclliana. but some are
O. horealis van hudsonica.

12e. Oxytropis campestris var. minor (Hook.) Welsh,
comb, nov., based on "Oxytropis uralensis y minor Hook.,
Fl. Bor.-Amer 1: 146. 183i.

Synonym: O. terrae-novae Fernald; O. campestris
\ar. terrae-novae (Fernald) Barneby (see O. uralensis for
discussion of the reasons for this combination)

I2f Oxytropis campestris var speciosa Torr. & A. Gray,
Fl. X..\mer 1; 341. 1838.

Based on; O. campestris y sulphurea sensu Hook.,
Fl. Bor Amer. 1: 147. 1831 (see below).

— O. sericea Nutt. var. speciosa (Torr. & A. Grav)
Welsh

Type locality; "British America, west to the Pacific"
(Torrey and Gray I.e.).

Type; "Dr. Hooker"; lectotype (here selected) NY!

Parat\-pes; "O.x. campestris y sulphurea. Dr Hooker's
Fl. B. Am. Rocky Mountains. Drummond," at K Hooker!
and "O.xvtropis campestris y. Frankl. Exp. Dr. Richardson,"
atBM! '

The sheet at NY, communicated by Dr. Hooker,
bears the notation, "Carlton House on the Saskatchewan
to the Rock^ Mountains, Dnimmond," and die initials E. E
S[heldon'r']. The quote is identical to that for varieties a
and 5 in the Flora, but since Hooker did not cite a localitx'
for y sulphurea. the material sent to Gray by Hooker could
ha\e been taken by either Douglas or Richardson and not
necessarily by Drummond. There are authentic specimens
bearing the name "sulphurea" at BM, one by Douglas and
the other by Richardson, and possibly a third by Flichardson
with the simple designation, "Oxytropis campestris y."
Probably the lectotype at NY was taken by one or the
other of the two collectors, and not by Drummond.
However, the specimen at K cited as paratype was collect-
ed by Drummond.

Hooker's brief description of y sulphurea, "foliolis
latioribus, spicis capitatis, floribus majoribus speciosis,"
characterizes this common plant of the western plains and
foothills of the Rockies from far north in Canada south to
Montana, Wyoming, and Idaho.

12g. Oxytropis campestris 5 spicata Hook., Fl. Bor.-Amer
1; 147. 1831.

= O. campestris (L.) DC. var. spicata Hook.

Type locality; "Between Carlton House on the
Saskatchewan to the Rocky Mountains [Alberta], T.
Dnnnmond' (Hooker I.e.).

Type; "O.xytn campestris 8. Carlton House. Fl. Bon
Am. "; lectotype (here selected) K Hooker!

Hooker (1831) characterized 8 spicata as "spicis
elongatis, floribus remotioribus. That description,
although short, matches the lectotvpe. There is only one
sheet of Oxytropis at Hooker's herbarium at K that bears
the designation 8 and the locality information "Carlton
House. ' The fact that Drummond is not indicated as col-

lector on the sheet is apparently of little consequence as
Hooker frequently failed to record collectors on his small
herbariiuii labels or in his annotation of the specimens.
The plants on that sheet clearly belong to what has passed
in recent times imder the name of O. campestris van gra-
cilis (A. Nelson) Barneby. That some plants sent by
Hooker to Gray represented taxa (jther than van spicata is
unfortunate. It appears that Hooker clearly had two enti-
ties in mind when he described van spicata and discussed
van sulphurea. Indeed, Barneby (1952; 279) masterfully
summarized the problems of identification of specimens
of van spicata (as O. campestris van gracilis) and van spe-
ciosa (as O. sericea van spicata). Much of the material
exchanged by Hooker is, indeed, the larger-flowered
phase (here termed van speciosa) with fewer leaflets.
Whether Hooker was responsible for sending wrongly
labeled material that was segregated prior to its distribu-
tion is not known. If Drummond mixed his collection so
as to include both varieties, he was not alone in mixing
the two pale-flowered species of the plains of western
Canada. The van spicata, with its numerous leaflets and
small flowers, grows in the general area occupied by the
larger-flowered phase with fewer leaflets. Many botanists
have made similar misinteipretations. The need to replace
van spicata with van speciosa is an unfortunate but neces-
sary' change mandated as the result of study of the types.

12h. Oxytropis campestris van sulphurea sensu Hook., Fl.
Bon Amen 1; 147. 1831. non DC.

= O. sericea van speciosa (Torn & A. Gray) Welsh
Authentic specimens of O. campestris P speciosa
Torn & A. Gnay; "2. O. campestris. P sulphurea on the ned
deer and eagle hills of the [illegible]. 1827," North
America. D. Douglas, and "Oxytropis campestris y, Frankl.
Exp. Dn Richardson" (both at BM!); and "Ox. campestris y
sulphurea. Dn Hooker s Fl. B. Am. Rocky Mountains.
Drummond" (at K Hooker!).

13. Oxytropis deflexa var. pulcherrima Welsh & A.
Huber, var. nov. O. deflexae var foliolosae (Hook.)
Barneby aspectu similis sed in floribus majoribus, racemis
latioribus et dense pilosis, et a van sericeae Torr. & A. Gray
in racemis compactis et floribus majoribus et purpureis
differt.

Type; USA, Utah, Duchesne County; "T2N, R7W,
S31 NW/NW UB&M, head of Log Hollow, 1.8 mi due
SW of Upper Stillwater Reservoin Uinta Mts., gravelly
subalpine meadow, common, limestone substrate, at ca
3294 m, 12 July 1994, A. Huber 1673"; holotype BRY!,
duplicates to be distributed. Additional materials
(paratvpes); Utah, Duchesne County; Uinta Mountains,
T2N, 'R7W, S31, 11 mi N of Tabiona, 12 July 1972; do,
T2N, R6W, S18, divide between Rock Creek and Brown
Duck Basin, at 3447 m, 28 August 1981, S. Goodrich & D.
Atwood 16163; do, T2N, R7W, S30, 3 July 1978, S.
Goodrich & L. Hart 11705; do, TIN, R8W, SI, head of
Wedge Hollow, 6 July 1990, D. Atwood 13934. Utah,
Summit Count>'; Lost Creek Park, road from Hoop Lake
to Spirit Lake, at 3050 m, 23 August 1977, K. Ostler 932.
Utah, Daggett CountA'; T2N, R17E, Sll, 22.2 km SW of
Manila, at 2815 m, 11 August 1983, S. Goodrich 19661.
Colorado, Chaffee County: 0.7 mi NE of Cumberland
Pass, T12 N, R4E, S12, at 3730 m, 17 August 1982, J.
Peterson et al. 82-58. Colorado, Gunnison County; Virgina
Basin, at 3691 m, 8 July 1946, C. L. Hax'ward 148; do,
Cumberland Pass, at 3874 m, 23 July 1970, L. C. Higgins
3755; do, 9 July 1969, L. C. Higgins 2103; do. North ridge

278

Great Basin Naturalist

[Volume 55

of Mount Belview, at 3750 m, 5 July 198S. L. C. Manin
3105. All specimens at BRY!

Plants of var. piilcherritna may he distinguished
from all other North American materials of the deflexa
complex by their compact racemes (remaining so in fruit),
larger flowers (hence broader racemes), and short, plump
pods that tend to have an abrupt shoulder beyond the
stipe. Their bipartite distribution from alpine sites in the
Uinta Mountains and similar, but higher, areas in the
southwestern Colorado Rockies is unique among Oxtjfro-
pis species. This is material that has been regarded in con-
temporary treatments of the genus in North America as
var. deflexa. Specimens of var. deflexa from Siberia, at
least those examined by me, have elongate racemes,
smaller flowers, and more slender pods.

14. Oxytropis foUolosa Hook., Fl. Bor.-Amer. 1: 146.
1831.

= O. deflexa \'m. foliolosa (Hook.) Barneby

Type locality: "From Carlton-House to the Rocky
Mountains, in lat. 54°" (Hooker I.e.).

Type: "Ox. foliolosa Hook. Fl. Bor. Am. Rocky
Mountains, Drumniond" Alberta, Canada; holotype K
Hooker!; isotype NY!

The holotype consists of two flowering specimens
representing the common acaulescent or subcaulescent
phase of var. foliolosa. The upper specimen displays
immature fruit.

There are collections of O. deflexa var sericea Torr
& A. Gray at K by Nuttall ("Oxytropis defle.xa. R. Mts.")
and by Drummond ("Saskatchewan. Drummond. Ox.
deflexa. DC").

15. Oxytropis hookcriana Nutt. in Torr. & A. Gray, Fl. N.
Amer. 1:340. 1838.

= O. lamhei'tii Pursh var. lainhertii

T\pe: "Oxytropis * Hookeri. O. Lambeiti P Hook. 147.
Platte plains," Nuttall s.n. 1834; holotype BM Nuttall!

The specimen at BM has "Platte plains" written on
the back of the sheet. It seems apparent that xNuttall
hoped to honor Hooker by providing a name for what he
took to be the concept of O. lambertii p, a plant that is a
phase of O. carnpesiris scarcely related to O. lainhertii.

16. Oxytropis ixodes Butters ik Abbe, Rhodora 45: 2,
tab. 745, fig. 1-6. 1943.

= O. borealis DC. var viscida (Nutt.) Welsh
Type: Minnesota, "slate cliffs on north side of a high
hill 1/2 mi. west of the outlet of South Fowl Lake, Cook
County, F K. Butters, E. C. Abbe, & G. W. Bums 611, 27
June 1940"; holotype MIN; isotype GH!, NY!, PH!, US!,
UC!, DAO!, BRY!, BM!

17. Oxytropis lagopiis Nutt., J. Acad. Nat. Sci.
Philadelphia 7: 17. 1834.

Type: Sources of the Missouri, N. B. Wveth s.n.
1833; holotype BM Nuttall!; isotopes NY!. PH!, K!

The specimen at BM has a label of a typical Nuttall
collection, "O.xytropis * lagopus — Sources of the Missouri."
It is mounted on a sheet with a collection from "Rocky
Mts, near the Platte," by Dr. Parry. The Royal Botanic
Garden material consists of a single caudex branch and
two inflorescences.

18. Oxytropis lambertii Pursh, Fl. Amer. Sept., 740.
1813.

Type: "On the ^Missouri, on the liluffs from the Maha
\illage to the Poncars, Louisiana [NE Nebraska or adjacent
South Dakota or Iowa], Bradbury s.n. 1811; lectotype
PH! (Barneby, Proe. Calif Acad. Sci. IV. 27: 285. 1952, as
type); isotype BM!

The sheet at BM is labeled "Louisiana. |. Bradbury
1811-12."

18a. Oxytropis lainheiiii (3. Hook., Fl. Bor -Amen 1: 147.
1831.

— O. campe.sfris (L.) DC. var davisii Welsh

Type locality: "Dry banks on Red River and Saskatch-
awan, (Douglas,) to the Prairies in the vallies of the Rock'v
Mountains. Drummond" (Hooker I.e.).

Type: "O.xytropis lambertii (3. Dr Hooker"; holotype
K! (see discussion below); isotype NY!

Authentic specimen: "O.xytropis Lambertii (3. Frankl.
E.xp. Dr. Richardson" (BM!).

.Neither Douglas s nor Drummond s materials with
the unequivocal designation O. lainhertii (3 have been
seen at either K or BM. There is at K Hooker, however, a
collection by Drummond (cited below as a possible
paratype of O. uralensis y minor Hooker, q.v.) that bears
the penciled notation "Ox. Lamberti??" It appears to be
O. campestris var davisii Welsh. The Richardson collec-
tion at BM (authentic specimen cited above) almost cer-
tainly is that ta.xon also. A second sheet at K Hooker bears
three collections, two small plants by Drummond, two
taller plants presumed to be by Douglas, and a third
extraneous collection by Percival. The first and second are
van davisii, and tliere is a penciled notation, "Ox. Lambertii
Pursh." That Hooker misunderstood that at least some
part of his var minor and his var (3 were conspecific indi-
cates the problem he had in dealing with plants from such
diverse areas as those found in North America, plants he
had not seen in the field. Designation of a lectotype for
the material is probably moot, since the material was not
given more than alphabetical designation. Perhaps these
sheets, as interpreted herein, will lay to rest the name O.
lambertii (3. Had there been an epithet applied, it would
have precluded the use of the name davisii.

18b. Oxytropis lambertii P leucoplnjlla Nutt. in Torr & A.
Gray, Fl. N. Amer. 1: 339. 1838, pro syn.
= O. lambertii e. (see below)

18c. Oxytropis lambertii £. Torn & A. Gray, Fl. N. Amen
1:339.1838.

= O. lagopus van atropiirpiirea (Rydb.) Barneby pro
parte et O. nana Nutt. pro parte

Synonym: O. lambertii P leucoplnjlla Nutt.

Authentic specimen: "Oxytropis * leucophylla. R.
Mts.," Nuttall s.n. 1834, BM Nuttall!

The authentic specimen has the information "Rocky
Mts. Nuttall's Herb." written on the reverse side, and the
following notation on the front: "Oxytropis leucophylla
Nuttall! [Oxytropis] Lamberti Pursh van glabrata Torn &
Gn Fl. N. America 1: p. 339." The name was published as
a synonym by Ton-ey & A. Gray (I.e.), based on a manu-
script provided b\' Nuttall. The>- characterize the plant by
the following description, but evidenth' did not see the
material: "e. very dwarf, canescently woolly; the leaflets
shorter and about 5 pairs; scape scarceh' longer than the
leaves; flowers capitate or nearly so; calyx densely woolly;
bracts small and short; wings emarginate. — O. Lamberti P
leucophylla, Nutt. mss," from "Plains of the Platte."
Barneby (1952: 304) noted that there "seem to be no cor-

1995]

North American Types of Oxytropis

279

responding specimens either at Philadelphia or in the
herbaria of Gray and Torrey." He therefore surmises that
the plant in question belongs to O. lagopus van atroptir-
piirea. There is a sheet at BM bearing Nuttall s character-
istic label with the name "Oxytropis * lettcophylla" in
Nuttalls handwriting. The two plants on the sheet appar-
ently belong to two different taxa, the smaller one to O.
lagopus var atropurpurea as surmised by Barneby, and the
second larger one to O. nana of Nuttall (or perhaps, but
unlikely, a dwarf specimen of O. sericea, my first impres-
sion).

19. Oxytropis mollis Nutt. ex A. Gray, Proc. Amer Acad.
Arts 6: 235. 1864. pro syn.

— O. borealis var. viscida (Nutt.) Welsh

Authentic specimen: "Oxytropis mollis. O. Ochro-

leuca Led. Altai proxima . . . R. Mts. Oregon," Nuttall s.n.

1834 (BM Nuttall!).

20. Oxytropis multiceps Nutt. in Ton; & A. Gray, Fl. N.
Amen 1:341. 1838.

Type locality: "Summit of lofty hills in the Rocky
Mountain range, towards Lewis's River [S. Wyoming],
Rock-y Mts. Nuttall" (Toney and Gray I.e.).

Type: "O.xytropis (Physocalyx) multiceps. R. Mts." T.
Nuttall s.n. 1834; holotype BM Nuttall!; isotypes NY!,
GH!, K Hooker!

The two specimens at BM are mounted on a sheet
with collections by J. M. Coulter and M. E. Jones. The
Nuttall material bears three labels: "*Physocalyx * multi-
ceps R. Mts. and two odiers. The second label makes com-
parisons with Old World species that the proposed new
genus and species could not be, and the third label con-
tains a brief description, "Gal. inflatus, apice 5-fidus, legu-
mine includens." Nuttall was at least entertaining the idea
that the plant represented a new genus.

Distribution: Colorado, NE Utah, SW Wyoming,
and W Nebraska.

The accrescent calyces, broad bracts, and few flow-
ers are characteristic for the species, which stands alone
in the genus in North America in its morphology.

20a. Oxytropis multiceps vai: minor A. Gray, Proc. Amer
Acad. Arts 20: 2. 1884.

= O. multiceps Ton: & A. Gray

Type: Clear Greek County, Colorado, C. C. Parry
991, 1861; lectotA'pe GH! (designated by Barneby, Proc.
Calif Acad. Sci. IV 27: 220. 1952); isolectotv-pe NY!

Paratvpe: "Rockv Mountain Alpine Flora, Lat.
39°-4r. No. 144. E. Hall & J. P Harbour, Colls. 1862";
BM!

21. Oxytropis nana Nutt. in Torr. &: A. Gray, Fl. N.
Amer 1:340. 1838.

Type: Plains of the Platte in the Rocky Mountain
Range [Wyoming], T. Nuttall s.n. 1834; holotype BM
Nuttall!; isotypes NY!, PH!

Distribution: Drainage of the North Platte and
Cheyenne rivers, westward to the Wind River Mountains,
Albany, Carbon, Converse, Fremont, Natrona, Platte, and
Sweetwater counties, Wyoming; endemic.

This is a beautiflil species of cla\s, shales, and gravelh'
bluffs and ridge tops endemic to Wyoming. Barneby (1952)
postulated that it might have arisen through hybridization
oi Oxytropis sericea and O. multiceps, a likely supposition.
Flower colors are \'ariable in a gi\'en population from pale
pinks through lavender and purple, and white-flowered

populations are known. A contribution from O. lamhertii
is also suggested by the presence of incipiently malpighi-
an hairs in some specimens. The relationship to segre-
gates of O. hesseyi postulated by Isely (1983) seems tenu-
ous at best. The relegation of O. nana to that species
might require a realignment of other taxa as well, includ-
ing combination oi lamhertii, sericea, campestris, and even
multiceps. Such a proposal is, of course, absurd. Taxonomy
must be both practical and reflect biological reality.

22. Oxytropis plattensis Nutt. in Torr. & A. Gray, Fl. N.
Amer 1:340. 1838.

= O. lamhertii \"m: lamhertii

Type: "Oxytropis * Plattensis, R. Mts. Platte,"
Nuttall s.n. 1834; holotype BM Nuttall!; isotxpe NY!

The holotype at BM consists of a single plant cut
fiom another sheet. It is topical of the Great Plains phase
of O. lamhertii.

23. Oxytropis podocarpa A. Gray, Proc. Amer. Acad.
Arts 6: 234. 1864.

Type locality: "Labrador, Arctic regions, and Rocky
Mountains, lat. 49°" (Gray I.e.).

Types: Labrador (Schweinitz), Arctic America
(Richardson?), O. arctica 5 injlata Hook. (Drummond),
and Alberta (Bourgeau). The Schweinitz and Bourgeau
specimens at GH! are cotypical, both having been used by
Gray in characterization of the species. However, the
species was lectotypified by Fernald (Rhodora 30; 154.
1928) on the Schweinitz collection from Laborador The
remaining specimens are considered to be paratopes.

Paratypes: "Highest summits of the Rocky Mts.,
Drummond" s.n, K!, type of O. arctica 6 injlata Hook.,
q.v.; "Oxytropis arctica 8 R. Br. Arctic America. Frankl.
Exp., K Hooker!

Distribution: Rock)' alpine ridges and coastal shores
in Colorado, Wyoming, Montana, Alberta, Northwest Terri-
tories, Ungava Peninsula, Labrador, and Baffin Island.

The bladdery-inflated stipitate pods of Oxytropis
podocarpa are characteristic of this and few other oxy-
tropes. The folded, falcate leaflets are useful in distin-
guishing this from other closely related mat- or mound-
forming species, such as O. nigrescens, in vegetative con-
dition.

24. Oxytropis sericea Nutt. in Torr & A. Gray, Fl. N.
Amer 1: 339. 1838.

Type: Rocky Mountains toward the sources of the
Oregon [S Wvoming], T Nuttall s.n., 1834; lectotype NY!
(Barneby, Proc. Calif Acad. Sci. IV, 27: 272. 1952).

24a. Oxytropis sericea var. speciosa (Torr. & A. Gray)
Welsh, comb, nov., based on "Oxytropis campestris P spe-
ciosa Torr. & A. Gray, Fl. N. Amer 1: 341. 1838, this in
turn based on O. campestris y .mlphurea sensu Hook., Fl.
Bor Amen 1; 147. 1831.

Missapplied name: O. sericea van spicata sensu
Barneby, Leafl. W Bot. 5: 111. 1951.

Distribution: Yukon, British Columbia, Alberta, Sas-
katchewan, Manitoba, Montana, Idaho, and Wyoming.

Members of this variety are characterized by
ochroleucous flowers with immaculate keel. In general
aspect they simulate the partially sympatric Oxytropis
campestris van spicata, from which they may be distin-
guished by fewer leaflets and generally larger flowers.
Alpine phases of O. campestris van cusickii approach van

280

Great Basin Naturalist

[Volume 55

speciosii l)()tli ill llower size and color. MaiiiK var. speciosa
does not occur in hijihlancls inhabited In var. cusickil l)ut the
similarities of the two varieties should not l)e discounted.
Apparent hybrids are known between this and O. cam-
pestris var. davLsii in northeastern British (Columbia.

25. Oxijtropis spiendens Douglas e.\ Hook., Fl. Bor.-
Amer. 1: 147. 1831.

Type locality: "On limestone rocks ol the Red River,
and south toward Pembina [S Manitoba]. Douglas" (I.e.).

Type: "On Limestone rocks oi the Red River and on
the south towards Pembina, 1827, a Ox. spiendens.
Douglas"; lectotype (here selected) K Hooker!; isolecto-
types OXF! (photo BRY!), BM! ("O. spiendens. Dry soils
on the plains of Red River. 1827. Douglas s.n."; 2 sheets).
The two sheets at BM are both by Douglas and represent
a vestita, the tvpical phase of the species.

The lectotype at K bears a label with almost the
exact information as the published type locality. It is
mounted on a sheet with a specimen designated "(3," and
with the label information "Rocky Mts. Richardson." A
better choice for lectotype of var. |3 is on a second sheet
(see below).

25a. Oxytropis spiendens a vestita Hook., Fl. Bor.-Amer.
1: 148. 1831

= O. spiendens Douglas

Type locality and type: As for the species.

25b. Oxytropis spiendens (3 richardsonii Hook., Fl. Bor.-
Amer. 1: 148. 1831.

= O. spiendens Douglas

Type locality: "From Cumberland-House on the
Saskatchewan, north to Fort Franklin and the Bear Lake,
and West to the dry prairies of the Rocky Mountains. Dr.
Richardson; Drummond" (Hooker I.e.).

Type: "278. O.xytropis o.wphylla. Dr. Richardson" s.n.
in 1821; lectotype (here designated) K Hooker!; isolecto-
types NY!, O. oxyphylla of Richardson, GH!

Paratype: "Fort Franklin to the Rock-y Mts. Drum-
mond. p. Ox. Spiendens. Dougl. Hook. Fl. B. Am.," K
Hooker!

The lectotype has three specimens, each designated
"P." The label "278. Oxytropis o.xyphylla" is affixed across
the base of the middle specimen; adjacent to the left one
is "Bear Lake," and below the specimen at the right is the
collector's name, "Dr. Richardson." The name "O. oxy-
phylla," in the sense utilized by Richardson in the botani-
cal appendix to Franklin's first journey (1823), is clearK
the basis for Hooker's p. richardsonii.

The Drummond syntype consists of a beautiful pkuit
with several flowering stems and numerous leaves, and a
fniiting raceme and peduncle.

Bameby (1952) notes:

Hooker recognized from the first a typical a vestita,
"valde hirsuto-sericea, bracteis hirsutissiniis calyce
multo longiorihus. ' described from Douglas s Red
River plants, and a p richardsonii. "minus hirsuta,
bracteis vi.\ longitudine caixcis, collected between
the Saskatchewan River and the Rocky Moimtains.

Plants of the two types pass into each other by degree and
have not been recognized at taxonomic rank in recent
times.

26. Oxytropis uralensis sensu American authors, non
(L.)DC."

North American specimens at BM and K bearing
this name are a mixture of (a) O. arctica R. Brown var. arc-
tica (eglandular, with large, pink-purple flowers); (b) O.
horealis var. hudsonica (Greene) Welsh (glandular, the
calyx teeth short and purjilish flowers (see "7/1834.7 0.xy-
tropis Uralensis. British North America. Dr. Richardson
1819-22, BM!," and "Repulse Bay. Parry's 2nd Voyage,
BM!"); (c) "Arctic Regions. Oxytropis Uralensis. Repulse
Bay. Parry's 2nd Voyage"); (d) O. maydelliana Trautv.
(ochroleucous flowers, with stipules castaneous); and (e)
O. canipestris var. minor (Hook.) Welsh (including var. ter-
rae-novae, flowers pink puiple, eglandular).

26a. Oxytropis uralensis a in Hook., Fl. Bor.-Amer. 1:
146. 1831.

= O. arctica R. Brown var. arctica

Locality: "Arctic regions and islands. Dr. Richardson;
Capt. Pariy &c." (Hooker I.e.).

Authentic specimen: "O. uralensis a Frankl. Exp.
Dr. Richardson, "BM!

Hooker s (1831) use of Oxytropis uralensis a in his
discussion of habitat merely indicated acknowledgment of
the taxon in the sense of Old World materials. However,
an authentic specimen at BM with that label information
is O. arctica sens. str. (see above).

26b. Oxytropis uralensis var. subsucculenta Hook., Fl.
Bor.-Amer. 1: 146. 1831.

= O. horealis DC. var. horealis

Type locality: "Arctic seashore, to the east of the
Mackenzie River" (Hooker I.e.).

Tyi^e: "O. uralensis p. 126. H. Sea Coast. Dr. Richard-
son"; lectotype (here designated) BM!

The lectotype at BM is mounted with O. uralensis a,
i.e., O. arctica R. Br. var. arctica. A possible syntvpe of P
suhsucculenta Hook, is also present at BM, with the label
"British North America. Dr. Richardson 1819-22. " Above
the label is a pencil notation, "cut from sheet of Oxytropis
canipestris."

26c. Oxijtropis uralensis var. arctica (R. Br.) Ledebour, Fl.
Ross. 1: 594. 1842.

Basionym: Oxytropis arctica R. Br.

= O. arctica R. Br. var. arctica

26d. Oxytropis uralensis y minor Hook., Fl. Bor.-Amer. 1:
146. 1831,

= O. campestris var. minor (Hook.) Welsh (see 12e
above)

Type locality: "Dr\ hills and prairies of the Rock->
Mountains. Mr. Drummond. Labrador. Mr. Morrison'
(Hooker I.e.).

Type: "Labrador. O. uralensis y. Momson"; lectotype
(here designated) K Hooker!

The International Code of Botanical Nomenclature
allows recognition of a taxon based on discordant material
where the name can be applied to at least one of its parts.
Hence, var. minor is not to be rejected simply because the
specimens on which it was based represent more than one
taxon. Evidence to support the assignment of the name to
the Labrador material is unequivocal, while its application
to materials from western Canada is problematical.

The lectotype at K consists of two specimens, one
flowering and the other in fi-uit. They were mounted pre-
viously with plants of another species, which have been
removed by cutting the sheet. The specimens both bear

1995]

North American Types of Oxytropis

281

the notation "y, and the sheet contains the annotation "O.
uralensis y Fl. Bor. Am. sed certi friictu ab Uralensi diver-
sum," the author unknown. Barneby (1952) was unable to
resolve the application of the name but noted: "The prob-
lem is nomenclaturally important in that var. minor could
prove to be the earliest name in its category for either O.
viscida var. hudsonica or O. campestris van terrae-novae."
The Labrador specimens cited with the original descrip-
tion are O. campestris var terrae-novae in a modern sense
and are here selected as the lectotype for the ta.xon.

Two small specimens at K Hooker! bearing the label
"Saskatchewan. Dnmimond" appear to be var. davisii, but
one cannot be certain of their provenance or that they
represent the material designated by Hooker as var.
minor They are mounted on a sheet with two additional
specimens, apparently var davisii also, but probably col-
lected by Douglas. Possibly all four specimens fonned the
basis for still another of Hooker's plants, O. lambertii P
(q.v.). A more convincing collection possibly included by
Hooker within var. minor, at K, is labeled "Astragalus
uralensis. Dr\' mountain prairies & low hills. Drummond. "
It is a possible syntype of var minor and appears to be var
davisii Welsh. There is no certainty, however, that the
specimen is part of what Hooker indicated as var minor

References

Bar.neby, R. C. 1952. A revision of the North American
species of Oxytropis DC. Proceedings of the Cali-
fornia Academy of Sciences IV, 27: 177-312.

Brow.n, R. 1823. Chloris melvilliana. A list of plants col-
lected in Melville Island (latitude 74°-75° N. longi-
tude 110° -112° W.) in the year 1820; by the officers
of the voygage of discovery under the orders of
Captain Parry. With characters and descriptions of
the new genera and species by Robert Brown. W
Clowes, London.

. 1824. A supplement to the appendi.x of Captain

Parry's first voyage for the discover}- of a North-West
passage, in the years 1819-20. Containing an account
of the subjects of natural histoiy. IX. Botan\'. A list of
plants collected in Melville Island, by the officers of
the expedition; with characters and descriptions of
the new species. Pages 261-309 in Journal of a voy-
age for the discovery of a North-West Passage fi"om
the Atlantic to the Pacific; performed in the year
1819-20, in His Majesty's ships Hecla and Gripper
under the orders of Captain William Edward Parry,
R.N., ER.S., and commander of the expedition. John
Murray, London.

BUNGE, A. 1874. Species generis Oxytropis, DC. Memoirs
of the Academy of Imperial Sciences Saint-Peters-
bourg, VII, 22: 1-166.

Gray, A. 1863. A revision and arrangement (mainly by the
fiaiit) of the Noilh American species of Astragalus and
Oxytropis. Proceedings of the American Academ\- of
Arts and Sciences 6: 188-236.

. 1884. A revision of the North American species of

the genus Oxytropis, DC. Proceedings of the American
Academy of Arts and Sciences 20: 1-7.

Holmgren, P K., N. H. Holmgren, and L. C. Barnett.
1990. Index herbariorum. Part I: The herbaria of the
world. Regnum Vegetabile 120: 1-693.

Hooker, W J. 1825. Botanical appendix. Pages 381-430
in Appendix to Captain Parry's journal of a second
voyage for the discovery of a North-West Passage
from the Atlantic to the Pacific performed in his
Majesty's ships Fury and Hecla, in the years
1821-22-23. John Murray, London.

. 1831. Flora boreali-americana; or, the botany of

the northern parts of British America: compiled
principally from the plants collected by Dr.
Richardson & VI r Dnmimond on the late northern
expeditions, under command of captain Sir John
Franklin, R.N., to which are added (by permission of
the Horticultural Society of London) those of Mr.
Douglas, from North-West America, and other natu-
ralists. Treuttel & Wurz, London. [Published as fas-
cicles, 1829-1840.]

Houston, C. S. 1984. Arctic ordeal. The journal of John
Richardson surgeon-naturalist with Franklin
1820-1822. McGill-Queen's University Press,
Montreal, Quebec. Canada.

IsELY, D. 1983. New combinations and two new varieties
in Astragalus, Orophaca, and Oxytropis (Legumi-
nosae). Systematic Botany 8: 420-426.

Pallas, P S. 1800-1803. Species Astragalomm descriptae
et iconibus coloratis illustratae. Godofredi Martini,
Lipsiae.

Parry, W E. 1821. Journal of a voyage for the discoveiy of
a North-West Passage fi-om the Atlantic to the Pacific:
perfomied in the years 1819-20, in his Majesty's ships
Hecla and Gripper, under orders of William Edward
Parr>-, R.N., ERS., and commander of the expedition.
John Murray, London.

POLUNIN, N. 1940. Botany of the Canadian Eastern Arctic.
Part I. Pteridophyta and Spermatophyta. National
Museum of Canada Bulletin 92: 1^08.

RiGHARDSON, J. 1823. Botanical appendix. Pages 729-768
in J. Franklin, Narrative of a journey to the shores of
the Polar Sea. W. Clowes, London.

Staflei', E a., and R. S. Cowan. 1976. Taxonomic litera-
ture. Volume I: A-G. Regnum Vegetabile 94: 1-1136.

. 1979. Ta.xonomic literature. Volume II: H-Le.

Regnum Vegetabile 98: 1-991.
. 1983. Taxonomic literature. Volume IV: P-Sak.

Regnum Vegetabile 110. 1-1214.
TORREY, J., and a. GR/\y. 1838. A flora of North America.

Volume 1. Wiley & Putnam, New York.
Welsh, S. L. 1977. On the typification of Oxytropis leu-

cantha (Pallas) Pers. Taxon 21: 155-157.
. 1990. On the txpification of Oxytropis horcalis DC.

Great Basin Naturalist 50: 355-360.
. 1991. Oxytropis DC. — names, basion\ms, types,

and synonyms — Flora North America Project. Great

Basin Naturalist 51: 377-396.
. 1994. Oxytropis de Candolle. Flora North .\jnerica.

In press.

Received 20 September 1994
Accepted 2 December 1994

SALTATION IN SNAKES WITH A NOTE ON ESCAPE SALTATION
IN A CROTALUS SCUTULATUS

Breck Bartholomew^ and Robert D. Nohavec^
Key words: Crotalus scutulatus, escape saltation, behavior.

Escape saltation and aggressive saltation
have been reported in relatively few snakes
(Gans and Mendelssohn 1971, Klauber 1972,
Gans 1974, Armstrong and Muiph\' 1979, Gasc
1994). These reports range fi-om the incredulous
to the well documented. Gasc (1994) relates
an unbelievable case of jumping in Atropoides
{=Porthiclimn) nwnmifer in which individuals
"tend to jump, either when they hit a prey or
to clear a height of up to 3 ft (1 m), starting
from a low point [emphasis added]." Certainly
A. nummifer may jump; however, in their
decades of experience with hundreds of these
snakes in both the wild and captivity, W. Lamar,
L. Porras, and A. Solorzano have never seen
nor heard of this behavior (personal communi-
cation). It is possible that A. nwnmifer may
appear to jump as they strike from an arboreal
perch (i.e., a log) and fall to the ground (L.
Porras personal communication).

The best reports of ophidian saltation are
those of Gans and Mendelssohn (1971) and
Gans (1974). These authors analyzed Bitis cau-
dalis jumping behavior in terms of stimulus
and biomechanics. They determined that B.
caudalis weighing less than 23.5 g, with a body
temperature between 31 °C and 37 °C, were
able to jump using sidewinding locomotion.
This type of saltation is energetically expen-
sive, and jumping snakes tire quickly.

Believable reports of rattlesnake saltation
are relatively few. Klauber's (1972) reports con-
sist primarily of exaggerated accounts of rattle-
snakes jumping while striking at either prey or
man. However, one of his reports cannot be
ignored:

Dr. R. B. Cowles told me that he was always skep-
tical of stories of rattlesnakes leaving the ground in
the course of a strike, until he saw this done two or
three times by an angry southwestern speckled

rattler {Crotalus mitchellii pyrrhus). The snake was
on pavement and struck for more than its full
length.

The only other report of rattlesnakes jumping is
of C sciitiilatm salvini which "struck so violent-
ly that their entire body appeared to be momen-
tarily air borne" (Armstrong and Muiphy 1979).
Neither of these reports discusses the biome-
chanics of how these snakes jumped. Since
both accounts are of aggressive saltation, and
neither of the species typically utilizes side-
winding locomotion (Cowles 1956, Klauber
1972), the biomechanics involved in rattle-
snake saltation is likely different fiom that in B.
caudalis. Here we report an instance of escape
saltation in a wild C. s. scutuhitus. Although our
observations are anecdotal, we believe they
offer important insight into the biomechanics
of rattlesnake jumping.

On 4 September 1993 we observed an un-
usual flight behavior by a wild C. s. scutulatus
in the Hualapai Mountains, Mojave County, AZ.
When approached, the snake lunged forward
using its tail as the origin of force. This lunge
was powerful enough to cause the snake's en-
tire body to lift off the ground (Fig. 1). Actual
forward movement from this "jump" was mini-
mal, and the snake recoiled into a series of
tight S-cun'es and jumped again. This type of
saltation was observed a total of four times.
None of the four jumps were directed toward
a person, and the snake s mouth appeared to
remain closed.

Of the four types of snake locomotion, this
jumping behavior could only be accomplished
using concertina, in which the tail is the main
point of force during foward movement. Klau-
ber (1972) noted that rattlesnakes use conceiUna
movement for slow progression in open areas
and where restraints are involved (i.e., smooth

'195 Wf.st 200 Ncirtli, Logan, UTS4321-.390.5.

^Venoni Researcli Laboratoir, Veterans Administration Mc-dical Center. Salt Lake Cit\. UT S414S-151H.

282

1995]

Notes

283

:jt<i.

■^'^:;^:<^>~-

•'J^r^^m-^

Fig. 1. Crotalus s. sciitulatus exhibiting escape saltation. Photograph taken just before the tail left the ground.

surface or narrow channel). Neither of these cir-
cumstances was apphcable to this particular
situation. The use of concertina locomotion
rather than sidewinding as a basis for jumping
in this snake is understandable as an anti-
predator response. Crotalus s. sciitidotiis typi-
cally utilize quick seipentine locomotion during
flight; Klauber (1972) noted they are rather
clumsy sidewinders. By reducing the number
of pressure points to one (e.g., the tail), die snake
changed from serpentine to concertina loco-
motion. Whether this change is an effective
use of energy remains to be tested. However,
given the short distance the snake traveled, it
would appear the relative energy cost would
be high.

Acknowledgments

We thank Louis Porras, William Lamar, and
Alejandro Solorzano for the information they
provided about Atropoides niimmifer. James
Glenn offered the financial assistance that
made these observations possible.

Literature Cited

Armstrong, B. L., and J. B. Murphy. 1979. The natural
histoiy of Mexican rattlesnakes. University' of Kansas
Museum of Natural Histor\' Special Publications 5:
1-88.

CowLES, R. B. 1956. Sidewinding locomotion in snakes.
Copeia 1956: 211-214.

Cans, C. 1974. Biomechanics; an approach to vertebrate
biology. University of Michigan Press, Ann Arbor
261 pp.

Cans, C, and H. Mendelssohn. 1971. Sidewinding and
jumping progression of vipers. Pages 17-38 in A. de
Vries and E. Kochva, editors. Toxins of animal and
plant origin. Gordon and Breach, New York.

Gasc, J.-P 1994. Locomotion. Pages 60-75 in R. Bauchot,
editor, Snakes: a natural histoiy Sterling Publishing,
New York.

Klauber, L. M. 1972. Rattlesnakes: their habits, life histo-
ries, and influence on mankind. University' of Cali-
fornia Press, Berkeley. 1536 pp.

Received 14 September 1994
Accepted 29 November 1994

A TRAP FOR BLUE GROUSE

Eric C. Pelreii' and Jolm A. Crawford'
Key tconls: Blue Grouse, Dendrauai^iis ()l)sciiriis, interception trap, Ore<ion. triij)piii<^ success, ualk-in trap

Numerous techniques have been devel-
oped to capture Blue Grouse {Dendragapiis
ohscurus) for scientific purposes, but efficacy
differs among methods and is influenced by
terrain, habitat conditions, and wariness of
birds. Zwickel and Bendell (1967) captured
over 1200 Blue Grouse in British Columbia
between 1950 and 1967 with a "noosing pole.'
Zwickel estimates that 10,000-15,000 Blue
Grouse have been captured with nooses in
parts of their range from Alaska to California
and from Alberta to Colorado (personal com-
munication). However, noosing poles and other
approach-based techniques may not be effective
for capturing Blue Grouse in areas where birds
do not permit close approach by humans,
where baiting is ineffective, or where topo-
graphic or habitat conditions preclude use of
drive traps. Consequently, additional methods
are needed to supplement existing techniques.

Henderson (1960), Standing (1960), and
Bauer (1962) used walk-in traps to capture over
520 Blue Grouse in north central Washington
fi-om 1956 to 1961. We developed a similar trap
for spring and summer capture of adult and
juvenile Blue Grouse in northeastern Oregon
where the waiy nature of the birds made noos-
ing poles and other approach-based capture
techniques ineffective. Our objective is to
describe the design and success rate of a walk-
in trap for Blue Grouse and to present this trap
as a viable alternative when approach-based
capture methods are ineffective.

We used a walk-in style trapset designed to
intercept the movement of birds through the
trapping area and guide these individuals along
wire fences into funnels connected to trap
boxes (Gullion 1966). Our trapsets usually con-
sisted of three open-bottomed boxes (range 1-7)
connected by fences (Fig. 1) and required 1-3 h
for placement. Trap boxes were constructed in

approximately 45 min from plastic or metal
wire or netting. A primary concern in choos-
ing trap material is that birds may injure them-
selves on the traps. Fish netting and/or vegeta-
tion on top of the box reduces the possibility
of injury (Henderson 1960, Standing 1960,
Bauer 1962). Size and placement of entrance
funnels is critical. We fitted 5-cm hexagonal-
weave chicken wire funnels to the entrance
holes; funnels had inner-opening dimensions
of 6 cm width X 10 cm height if the trap was
intended primarily for immatures and 8 cm
width X 12 cm height if the trap was intended
primarily for adult birds.

Successful traps were constructed where
birds had been located repeatedly within a 5-
to 10-m^ area. Fences between boxes were
shorter where grouse were more highly local-
ized to increase the concentration of trap boxes
in those areas.

We captured 140 grouse in 2327 trap(set)-
days during summers 1991, 1992, and 1993 at
rates of 23, 19, and 12 trap-days/capture, re-
spectively (Table 1). Increased trapping effi-
ciency in 1992 and 1993 was likeK' a product
of refined trapping techniques, experience of
trappers, and knowledge of bird movements.
Henderson (1960) achieved capture rates of 7
and 8 trap-days/captme in 1958 and 1959, re-
spectively. Variations in capture success rates
may be influenced by grouse population density
and distribution. Habitats in which grouse are
evenK' distributed are more difficult to trap than
areas in which grouse congregate in dense
clumps.

Many variations in design, material, con-
struction, and placement of walk-in traps exist.
Flexibility, creativity, and an understanding of
the system in which trapping efforts take place
are critical for successful application of this
trap to specific conditions.

'Departiiu-i.t ot Fislu-ries and Wildlife, Oregon State University, 104 Nash Hall, Corvallis, OR 97331-3803.

284

1995]

Notes

285

135 cm

35 cm

18 cm

hog rings

wooden stakes '

75 cm v/''^ sticks, brush, rocks

o^sOK ^^ ^^ or wire fence

fence height
cm

\| fj- ^ ^ lence nei

Fig. 1. Interception trap design incoiporating V-shaped
and linear arrangement for capturing Blue Grouse.

The noosing pole is an effective method for
capturing Blue Grouse in much of its range.
This technique requires significantly less time,
manpower, and money than traps; however, in
areas where birds are too wary to be approached
and captured by noosing pole, interception
traps may provide a viable alternative capture
technique for Blue Grouse.

Acknowledgments

This study was funded by the U.S. Forest
Service and Oregon Department of Fish and

Table 1. Blue Grouse interception trap capture statistics,
June through August 1991, 1992. and 1993, in Wallowa
Gountv, OR.

^ J

Year

rebar stakes

1991

1992

1993

Total

Number of birds

captured

140

Trap -days

618

941

768

2327

Trap - days/capture

Wildlife (ODFW). Special thanks go to ODFW
Wallowa District Biologist Vic Coggins and his
assistant Pat Matthews for technical support
and suggestions during the trapping process.
Trapping assistants Mark Porter and Jim Wach
helped to refine our technique, and Fred
Zwickel provided valuable input during the
editing process. This is Oregon Agricultural
Experiment Station Publication No. 10,355.

Literature Cited

Bauer, R. 1962. Ecology of Blue Grouse on summer
range in northcentral Washington. Unpublished the-
sis, Washington State University, Pullman. 81 pp.

GULLION, G. W. 1966. Instructions for placing and operat-
ing traps for Ruffed Grouse. Forest Wildlife
Relations Project. University of Minnesota Forest
Research Center, Cloquet. [Mimeo.] 14 pp.

Henderson, U. B. I960. A stud\ of Blue Grouse on sum-
mer range, northcentral ^Vashington. Unpublished
thesis, Washington State University', Pullman. 96 pp.

Standing, K. M. 1960. Fictors in relation to population
fluctuations in the Blue Grouse. Unpublished disser-
tation, Washington State Universitx; Pullman. 193
pp.

Zwickel, E C., and J. F Bendell. 1967. A snare for cap-
turing Blue Grouse. Journal of Wildlife Management
31: 202-204.

Received 29 August 1994
Accepted 16 November 1994

BOOK REVIEW

Mountains and plains: the ecology of Wyo-
ming landscapes. Dennis H. Knight. Yale
Unixersit)' Press, New Haven and London.
1994. xi, 338 pp., illus. $40 liardbaek.

Many l)ooks describe the landscapes and
ecology of subregions of the United States.
This book provides a detailed description and
ecological interpretation of Wyoming and
adjacent areas. Although this region has long
been a magnet, drawing national and interna-
tional visitors to Yellowstone National Park,
Teton National Park, the Black Hills, Devil's
Tower National Monument, and to the expan-
sive, sparsely populated plains and mountains
that lie between those public attractions, this
book provides more detail and a more holistic
overview of the region than any existing refer-
ence designed for public consumption. Unusu-
ally well illustrated and handsomely designed,
Mountains and Plains will appeal to and enrich
the understanding of any thoughtful visitor to
this land of natural wonders.

The author, Dennis H. Knight, has distilled
for the reader a delightful mixture of land-
scape images and informative facts from his
teaching and research experiences of over 30
years in the natural landscapes of Wyoming.
Knight has written the book with interested
laymen in mind. Roughly 120 careHilly selected,
well-reproduced photographs enrich the text.
Many of the photos are masterfully selected to
illustrate a puzzling or unexpected relation-
ship. The photographs made my first "glance"
at the book a rewarding four-hour experience!
Numerous photos and their brief captions so
inflamed my curiosity that I had to search the
text for the author's well-researched com-
ments before I could continue on. In keeping
with his objective of writing for laymen, Knight
uses only common names for animals and
plants in the text, but well-organized appen-
dices permit interested professionals to quick-
ly link an unambiguous Latin binomial to any

given common name. In the same vein, the
author cites supporting references in such a
way that the flow of the narrative is not bro-
ken, but interested scholars can readily deter-
mine where detailed studies of particular
problems are published. The brief but clear
explanations in the text are startlingly well
supported by a massive body of references
(over 1150 original sources cited).

Knight, a past president of the Ecological
Society of America and a science advisor to
numerous management agencies in the nation,
has obviously poured many months of intense
effort into Mountains and Plains. The entire
book is carefully integrated and clearly written.
His manuscript has been artfully and profes-
sionally assembled into a handsome, durable
hardback volume by Yale University Press.
Photographs are clearly reproduced and taste-
fully integrated into an essentially error-free
text. The background sections of the book
(Chapters 1 and 2) are supported by a variety
of maps that permit the reader to quickly
determine climatic, geologic, edaphic, and
land ownership status at any geographic point
in the area of concern.

Mountains and Plains is a book that anyone
concerned with the natural histoiy or manage-
ment of Wyoming (or nearby) landscapes
should own. Interested laymen will find the
book to be informative and reliable. Managers
of natural resources in the area will find the
volume to be an essential companion in the field
or at the desk. The price is modest for a hard-
back book of such sturdiness. Both Knight and
Yale University' Press are to be commended for
a valuable contribution to our understanding of
the natural environments in that area where the
High Plains encounter the Rocky Mountains.

Kimball T. Haiper

Department of Botany and Range Science

Brigham Young University

Provo, UT 84602

286

INFORMATION FOR AUTHORS

MANUSCRIPT PREPARATION. Consult Vol. 51, No.
2 of this journal for specific instructions on style
and format. These instructions, GUIDELINES FOR
Manuscripts Submitted to the Great Basin
Naturalist, supply greater detail than those pre-
sented here in condensed form. The guidelines are
printed at the back of the issue; additional copies
are available from the editor upon request. Also,
check the most recent issue of the Great Basin
Naturalist for changes and refer to the CBE Style
Manual, 5th edition (Council of Biology Editors,
One Illinois Center, Suite 200, 111 East Wacker
Drive, Chicago, IL 60601-4298 USA, PHONE: (312)
616-0800, FAX: (312) 616-0226; $24).

TITLE PAGE includes an infonnative title no longer
than 15 words, names and addresses of authors, a
running head of fewer than 40 letters and spaces,
footnotes to indicate change of address and author
to whom correspondence should be addressed if
other than the first author.

ABSTRACT states the purpose, methods, results,
and conclusions of the research. It is followed by
6-12 key words, listed in order of decreasing
importance, to be used for indexing.

TEXT has centered main headings printed in all
capital letters; second-level headings are centered
in upper- and lowercase letters; third-level head-
ings begin paragraphs.

ACKNOWLEDGMENTS, under a centered main
heading, include special publication numbers when
appropriate.

LITERATURE CITED, also under a centered main
heading, lists references alphabetically in the fol-
lowing fomiats:

Mack, G. D., and L. D. Flake. 1980. Habitat rela-
tionships of waterfowl broods on South Dakota
stock ponds. Journal of Wildlife Management
44: 695-700.

Sousa, W P 1985. Disturbance and patch dynamics
on rocky intertidal shores. Pages 101-124 in
S. T. A. Pickett and E S. White, eds.. The ecolo-
gy of natural disturbance and patch dynamics.
Academic Press, New York.

Coulson, R. N., and J. A. Witter. 1984. Forest ento-
mology: ecology and management. John Wiley
and Sons, Inc., New York. 669 pp.

TABLES are double spaced on separate sheets and
designed to fit the width of either a single column
or a page. Use lowercase letters to indicate foot-
notes.

PHOTOCOPIES OF FIGURES aie submitted initially
with the manuscript; editors may suggest changes.
Lettering on figures should be large enough to
withstand reduction to one- or two-column width.
Originals must be no larger than 22x28 cm.

NOTES. If the manuscript would be more appro-
priate as a short communication or note, follow the
above instructions but do not include an abstract.

FINAL CHECK:

• Cover letter explaining any duplication of
information and providing phone number(s),
FAX number, and E-mail address

• 3 copies of the manuscript and WordPerfect
diskette

• Conformity with instructions

• Photocopies of illustrations

(ISSN 001 7-361 4)

GREAT BASIN NATURALIST vc 55 no 3 juiy 1995

CONTENTS

Articles

Benthic community structure in two adjacent streams in Yellowstone National

Park five years after the 1988 wildfires G. Wayne Minshall,

Christopher T. Robinson, Todd V. Royer, and Samuel R. Rushforth 1 93

Effects of browsing by native ungulates on the shrubs in big sagebrush commu-
nities in Yellowstone National Park .... Francis J. Singer and Roy A. Renkin 201

Soft sediment benthic macroinvertebrate communities of the Green River at the

Ouray National Wildlife Refuge, Uintah County, Utah Eric R. Wolz

and Dennis K. Shiozawa 213

Alpine vascular flora of the Tushar Mountains, Utah Alan C. Taye 225

Ecology ofCeltis reticulata in Idaho Ann Marie DeBolt

and Bruce McCune 237

Mimulus evanescens (Scrophulariaceae): a new annual species from the northern

Great Basin Robert J. Meinke 249

Morphological and host-symbiont studies of Trichodina tenuiformis and Apiosoma
campanulatwn infesting motded sculpin {Cottus hairdi) from Provo River, Utah
Ying Qi and Richard A. Heckmann 258

Effects of horse grazing in spring on survdval, recruitment, and winter injury

damage of shrubs Dennis D. Austin and Philip J. Urness 267

North American types of Oxytropis DC. (Leguminosae) at The Natural History
Museum and Royal Botanic Garden, England, with nomenclatural comments
and a new variety S. L. Welsh 271

Notes

Saltation in snakes with a note on escape saltation in a Crotalus scutulatus

Breck Bartholomew and Robert D. Nohavec 282

A trap for Blue Grouse Eric C. Pelren and John A. Crawford 284

Book Review

Mountains and plains: the ecology of Wyoming landscapes Dennis H. Knight . .

Kimball T Harper 286

5\o^

H E

GREAT BASIN

NATURALIST

VOLUME 55 N2 4 — OCTOBER 1995

BRIGHAM YOUNG UNIVERSITY

GREAT BASIN NATURALIST

Editor Assistant Editor

Richard W. Baumann Nathan M. Smith

290 MLBM 190 MLBM

PO Box 20200 PO Box 26879

Brigham Young University Brigham Young University

Provo, UT 84602-0200 ' Provo, UT 84602-6879

801-378-5053 801-378-6688
FAX 801-378-3733 E-mail: NMS@HBLL1.BYU.EDU

Associate Editors

Michael A. Bowers Paul C. Marsh

Blandy Experimental Farm, University of Center for Environmental Studies, Arizona

Virginia, Box 175, Boyce, VA 22620 ' State University, Tempe, AZ 85287

J. R. Callahan Stanley D. Smith

Museum of Southwestern Biology, University of Department of Biology

New Mexico, Albuquerque, NM University of Nevada-Las Vegas

Mailing address: Box 3140, Hemet, CA 92546 Las Vegas, NV 89154-4004

Jeffrey J. Johansen Paul T. Tueller

Department of Biology, John Carroll University Department of Environmental Resource Sciences

University Heights, OH 441 18 Universit>' of Nevada-Reno, 1000 Valley Road

„ „ Reno, NV 89512

Boris C. Kondratieff

Department of Entomology, Colorado State Robert C. Whitmore

University, Fort Collins, CO 80523 Division of Forestry, Box 6125, West Virginia

University, Morgantown, WV 26506-6125

Editorial Board. Jerran T. Flinders, Chairman, Botany and Range Science; Duke S. Rogers, Zoology;
Wilford M. Hess, Botany and Range Science; Richard R. Tolman, Zoology. All are at Brigham Young
University. Ex Officio Editorial Board members include Steven L. Taylor, College of Biolog>' and Agriculture;
H. Duane Smith, Director, Monte L. Bean Life Science Museum; Richard W. Baumann, Editor, Great Basin
Naturalist.

The Great Basin Naturalist, founded in 1939, is published quarterly by Brigham Young University.
Unpublished manuscripts that further our biological understanding of the Great Basin and surrounding areas
in western North America are accepted for publication.

Subscriptions. Annual subscriptions to the Great Basin Naturalist for 1995 are $25 for individual sub-
scribers ($30 outside the United States) and $50 for institutions. The price of single issues is $12. All back
issues are in print and available for sale. All matters pertaining to subscriptions, back issues, or other busi-
ness should be directed to the Editor, Great Basin Naturalist, 290 MLBM, PO Box 20200, Brigham Young
University, Provo, UT 84602-0200.

Scholarly Exchanges. Libraries or other organizations interested in obtaining the Great Basin
Naturalist through a continuing exchange of scholarly publications should contact the Exchange Librarian,
6385 HBLL, PO Box 26889, Brigham Young University, Provo, UT 84602-6889.

Editorial Production Staff

JoAnne Abel Technical Editor

Jan Spencer Assistant to the Editor

Official publication date: 31 October 1995 10-95 750 15818

The Great Basin Naturalist

Published at Provo, Utah, by
Brigham Young UwivERsiri'

ISSN 0017-3614

Volume 55 31 October 1995 No. 4

CLASSIFICATION OF THE RIPARIAN VEGETATION ALONG A 6-KM
REACH OF THE ANIMAS RIVER, SOUTHWESTERN COLORADO

Gillian M. WalforcU and William L. Baker^-^

Abstract. — Riparian ecosystems are important components of landscapes, particularly because of their role in biodi-
versitx'. A first step in using a "coarse-filter" approach to riparian biodiversity conservation is to determine the kinds of
riparian ecosystems. These ecosystems vaiy substantialh' in plant species composition along a single river reach, as well
as between rivers, and yet the river-reach scale has received little attention. We sampled the vascular plant composition
of 67 contiguous patches of riparian vegetation along a reach of the Animas River, in southwestern Colorado's San Juan
.Mountains, that is relatively undisturbed by human land uses. Using cluster analysis and detrended correspondence
analysis, we identified eight riparian community' types along the reach. Using a new technique, we combined overstory
size-class data and understoiy cover data to identify community types. The eight community types, which are in part the
products of past floods, are spatially arranged along the reach in relation to variation in valley morphology, tributaiy
location, and geomorphic landforms. These eight community types do not necessarih' represent successional stages of a
single potential vegetation t\'pe. This study at the river-reach scale suggests that sampling and analysis, as well as con-
sei-vation, may need to be tuned to the scale of patchiness produced b> flood disturbances in the riverine landscape,
since xegetation varies significantly at this scale.

Key words: riparian vegetation, Rocky Mountains, Colorado, multivariate analysis.

Riparian vegetation provides several impor- will be preserved (Hunter 1991, O'Connell

tant functions in landscapes, and riparian com- and Noss 1992).

munities have thus been a focus for conserva- Classification of vegetation communities is

tion. Riparian vegetation contributes to water an essential first step in implementing this

quality, stream bank stability, and healthy fish coarse-filter conser\'ation approach, but classi-

habitat (Johnson et al. 1985, Malanson 1993). fication of riparian vegetation in the Rocky

Riparian vegetation also provides cover and Mountains is incomplete. The montane and

forage for wildlife that is particularly important subalpine riparian zones of Colorado's west-

in the arid portions of North America (Knopf ern slope have been classified (Baker 1989).

1985). The idea of a "coarse-filter ' approach to Riparian community type classifications for

biodiversity conservation is that by preserving U.S. Forest Service lands in Utah and parts of

viable communities, associated species also Idaho and Wyoming are available (Mutz and

^Wyoming Natural Diversity Database, The Nature Conservancy, 1604 Grand A\e., Laramie, \VT 82070, USA.
"Department of Geography, University of Wyoming, Laramie, \VT 82071, USA.
"^ Author to whom correspondence should be addressed. ,

287

288

Great Basin Naturalist

[Volume 55

Gnihani 1982, Youngliloocl ct al. 1985a, 19851),
Padgett et al. 1989 ). A riparian-wetland classi-
fication and key were produced for Montana
(Hansen et al. 1991), and other classifications
are available for small parts of the Rocky
Mountains (Mutz and Graham 1982, Cooper
and Cottrell 1990).

Methods of riparian vegetation classification
vary with the researcher and project goals.
The U.S. Forest Service classifies forest eco-
systems into "habitat t\'pes" based on potential
climax vegetation (Daubenmire 1952, Pfistcr
and Arno 1980). Climax vegetation represents
the stable, self-perpetuating community pre-
dicted on the basis of climate, topography, and
soils in the absence of distmbance. Often, how-
ever, riparian vegetation never reaches climax
due to frequent floods (Campbell and Green
1968). Therefore, some researchers classify
vegetation into "community types" according
to existing structure and composition without
reference to successional stage following dis-
turbance. This approach, however, is not con-
cerned with whether vegetation consists pri-
marily of native or e.xotic species or has been
disturbed by human land uses (e.g.. Young-
blood et al. 1985b, Padgett et al. 1989). Another
classification approach (1) recognizes that cli-
max vegetation is seldom reached due to nat-
ural disturbances, but focuses on the more
mature successional stages; and (2) empha-
sizes that classification of "natural vegetation,"
vegetation that is as free as possible of exotic
species and the effects of human land uses,
provides essential information for effective
biodiversity conservation (Baker 1989).
Vegetation types classified using either a habi-
tat type approach or Baker s approach are
referred to as "associations."

Even if there is only one association along a
river reach, there may be several community
types. Riparian community types along a river
reach comprise a complex which Winward and
Padgett (1989) name on the basis of the most
prominent community type plus geographical
features describing where it occurs. This spa-
tially complex mosaic of community types cre-
ates difficulties for classification, but the
diversity of communities is an important com-
ponent of liiodiversity (Hunter 1991).

An additional problem is that classification
may use only overstory species, or it may be
based on the entire flora. Classification tradi-
tionally uses one technique or a combination

of tcclmi(|ues including subjective grouping,
evaluating and scjrting of stand tables, cluster
analysis, or ordination (Whittaker 1962, Gauch
1982). However, the dominance of the over-
story in forests may skew mathematical analy-
ses that use the entire flora (Padgett et al.
1989). In northern regions, where the under-
stoiy flora often is more sensitive to environ-
mentiil variation than is tlie overstoiy (Whittaker
1962), quantitative techniques that give equal
weight to the undcrstory and overstory may
not be ideal.

Along the Animas River in southwestern
Colorado s San Juan Mountains, we investigat-
ed variation in plant species composition of
contiguous patches of riparian vegetation. We
classified riparian vegetation along a 6-km
reach using both understory and overstory
vegetation. Goals of this paper are to identify
community types found along the river reach,
to describe the community types in relation to
their environment, and to explain a new
approach to balance the use of both overstory
and understoiy vegetation data in quantitative
classification. This approach, we suggest, may
lead to community classifications more useful
for consei"vation and management.

Study Area

The Animas River starts in the San Juan
Mountains of southwesteiTi Colorado and flows
south to the San Juan River in New Mexico.
The study area is approximately 40 km north-
east of Durango in LaPlata County (Fig. 1),
along a continuous 6-km reach of the Animas
River between 2430 and 2550 m in elevation.
This is one of the least disturbed montane
river reaches in western Colorado (Baker 1990).
A narrow-gauge railroad track and a wilder-
ness access trail occur along the reach, but the
reach has probably never been grazed b\' cat-
tle or sheep. There are some silver and gold
mines upstream.

The Animas is an unregulated gravelbed
river with a mean annual peak flow of 145
m^s"!. Within the stud\' reach the river has a
mean gradient of 0.0193, a mean channel widtli
of 34.3 m, and a mean channel depth of 3.6 m.
The river is entrenched in a deep canyon sur-
rounded by the Needle Mountains. The valley,
varying in width from less than 100 m to about
400 m, is lined with alluvial deposits fomied fi-om
Precambrian granites in the north and south

1995]

Animas Riner Riparian Vegetation

289

Community Types

Populus angustifolia I Alnus incana

Populus angustifolia I Agrostis scabra

Populus angustifolia - Pseudotsuga menziesii I Pyrola asarifoUa

Picea pungens - Populus angustifolia I Antennaria parvifolia

Picea pungens / Alnus incana I Equisehtm arvense

Picea pungens- Populus tremuloides I Mahonia repens
Pseudotsuga menziesii I Acer glabrum

I Pyrola asarifoUa
Populus tremuloides - Pseudotsuga menziesii

I Bromus ciliatus

\:^:y\ Not sampled
Animas River

fi.S

1.0

kilometer

Fig. 1. Animas River study area and its location in Colorado. Patches are shaded according to their comniunit\ t\pe
and numbered for identification.

sections of the study area and from Precambrian tion to floods and climatic fluctuations (Baker
gneiss and schist in the central section of the 1988, 1990). Baker (1988) identified 57 vegeta-
study area (Osterwald 1989). tion patches having distinct boundaries recog-

nizable on aerial photographs and on the
Methods ground. This patchiness is largely the result of

tree regeneration after past floods (Baker 1990).

Previous work on the study reach focused on In the field we refined the boundaries of some

the structure of riparian tree populations in rela- of the 57 patches and identified new patches

290

Great Basin Naturalist

[Volume 55

for a total of 67 patches available for sampling.
Patches were mapped on aerial photographs
in the field, then digitized and rectified using
the GRASS geographic information system
(USA-CERL 1991). A final map of patch loca-
tions and vegetation (Fig. 1) was produced with
ATLAS Draw (Strategic Mapping 1991).

Vegetation Sampling and
Environmental Data

Within each of the 67 patches, one 20 X
50-m plot (0.1 ha) was subjectively placed par-
allel to the river to obtain a representative
sample of herbaceous and woody vegetation.
Methods of herbaceous vegetation sampling
followed Peet (1981). Percent cover of each
vascular plant species (except trees) present
was estimated, during mid-growing season, in
25 contiguous 0.5 X 2-m quadrats along the
50-m center line of each plot. Percent cover of
0-10% was estimated to the nearest 1%; per-
cent cover of 10-100% was estimated to the
nearest 5%. Species located during a sui-vey of
the plot, but not found in the quadrats, were
assigned 0.1% cover. Nomenclature follows
Kartesz and Kartesz (1980).

Baker (1988) collected data on diameter-at-
breast-height (dbh) size classes of tree species
in his original 57 stands. We added to this data
set by tallying trees (>2.5 cm dbh in 10-cm
classes), saplings (<2.5 cm dbh and >1 m
tall), and seedlings (<2.5 cm dbh and <1 m
tall) of each species in plots of the 10 addition-
al patches as Baker had done. Increment cores
were extracted from the bases of 5-15 of the
largest trees in each patch for estimating patch
age. Ages of the largest trees tend to be simi-
lar, reflecting a common origin following
floods (Baker 1990). Each patch was assigned
to a 10-year age class according to the maxi-
mum age of the 5-15 cored trees. Age zero is
A.D. 1990.

A set of environmental variables was mea-
sured in the field in each patch. Patch slope
was measured using an Abney level and sur-
vey rod. Aspect of the patch was measured in
degrees with a compass. We surveyed the dis-
tance to the channel and the height above the
channel using the rod, level, and a distance
meter. Patches were identified as either on
terraces or depositional bars. The depositional
bar is the lowest prominent feature higher than,
but within, the channel bed, while terraces are
older, higher fluvial landforms (Osterkamp

and Hupp 1984). At eveiy 0.5 m along the 50-
m center line of each plot, we measured the
intermediate axis of the surface particle at that
point and assigned it to a size class, in a varia-
tion of the Wolman (1954) technique. Later,
using Rodriguez's (1986) MOMENTS pro-
gram, we calculated mean size, %<1 mm,
%<2 mm, and sorting value for each patch.
Soil samples of the upper 15 cm of the profile
were taken in onK' 20 of the 67 patches, due to
the cost of chemical analyses. These 20 sam-
ples spanned the spectrum of patch ages and
floristic and environmental variation. Samples
were later analyzed for standard fertility
(organic matter, pH, N, F, and electrical con-
ductivity) by the University of Wyoming Soil
Testing Lab.

Quantitative Analyses

We used the SPSS/PC + cluster analysis
program (SPSS 1990) to determine groups of
patches similar in overstory and understory
vegetation composition (Romesburg 1984). After
experimenting with several clustering methods,
we identified the BAVERAGE method (aver-
age linkage between groups) and the cosine
distance measure (angular separation of vec-
tors of variables) as the best clustering combi-
nation. This combination emphasizes relative
abundances within a plot and de-emphasizes
absolute abundance differences between plots
(SPSS 1990).

Species composition data were also ordi-
nated by detrended correspondence analysis
(DCA) using CANOCO (Canonical Community
Ordination), a multivariate statistical program
for applications in community ecology (Ter
Braak 1988). Correspondence analysis pro-
vides a geometrical representation of the rela-
tionships among samples and species in a data
set and identifies the dominant trend of \ aria-
tion in community composition.

Initial ordination and cluster anal> sis of the
combined overstory and understory data set
resulted in groupings primarily reflecting just
the high cover values of overstory tree species
rather than the joint pattern of both overstory
and understor)' species. To counteract this, we
analyzed the overstoiy tree species size-class
data and the understory shrub and herbaceous
species cover data separately, and then merged
the two results. The overstor\ size-class data
of each plot were first clustered; then the per-
cent cover data of understorv shrubbv and

1995]

Animas River Riparian Vegetation

291

herbaceous species in each pk)t were clus-
tered. These understory cover data were also
ordinated using DCA. Final classification
groups were the result of intersections of over-
story cluster groups with understory cluster
groups overlain on the understory DCA ordi-
nation diagram. We calculated the mean value
for several environmental variables in each
community type. Environmental variables we
used are those found to be important to vege-
tational variation along the reach based on a
separate, but related, gradient analysis (Baker
and Walford 1995).

Classification groups referred to here are
"community types" because they represent
existing rather than potential natural vegeta-
tion. Each community type is based on the
entire flora but is named based on tlie dominant
species in the overstoiy and the dominant or
most diagnostic indicator species in the
understory (Mueller-Dombois and Ellenberg
1974). When there are co-dominants in a layei;
both species are included in the name and are
separated by a hyphen.

Results
Classification

The cluster analysis and DCA ordination of
the plot understoiy cover data (grasses, forbs,
and shrubs) suggested four major groups and
one outlier (Fig. 2). The similarity cut level was
kept coarse so that overstory cluster groups
could be incoiporated later. This specific level
was chosen after considering alternative cut
levels at slightly greater or lesser similarity
(Fig. 2). Groups A and B, for example, would
become one group if the cut level were at a
slightly lower similarity, yet these two groups
are quite different (Fig. 2). Ordination of the
same data set is represented by the DCA Axis
1 vs. Axis 2 ordination diagram (Fig. 3a). The
distinctiveness of understoiy groups produced
by cluster analysis is supported by the com-
paratively distinct location of the groups on
this ordination diagram.

Understory groups identified by cluster
analysis and ordination are compositionally
distinct and occur in different environmental
settings. Group A was dominated by Ainus
incana and Eqiiisetum arvense. These patches
were predominantly located on bars. Group B
was located entirely on bars and had the fewest
species of any group. Agrostis scahra was always

Fig. 2. Understory cluster analysis dendrogram based on
percent cover of herbaceous and shrubby species. Plot
numbers correspond with patch numbers on the study
area map (see Fig. 1). The dashed line indicates the simi-
larity level at which miderston- groups were separated.

present in Group B patches. The third group
(C) was dominated by Rosa woodsii and Pijrola
asarifolia. Patches of Group D are almost always
on terraces and generally have the highest
species richness. Mahonia repens is always
present, and Rosa woodsii, Bromus ciliatus,
and Onjzopsis asperifolia are usually well rep-
resented.

Overstory size-class data were clustered by
the same method. Three overstory groups were
identified at approximately 25% similarity
(Fig. 4). Each of these major groups has mem-
bers from at least three different understory
groups. Group I is recognized by a dominance
of Populus angiistifolia seedlings, saplings, and
small trees (Table 1). All understoiy Group B
members are found within diis overstoiy group.
But other members of this overstory group
have the understoiy of Groups A, C, D, or E.
Overstoiy Group II is characterized by Picea
pimgens of all sizes and larger P. angiistifolia

292

Great Basin Natur.\list

[Volume 55

(Table 1). Most of its menilHMs ha\e an uiider-
story of Groups A or D. The third overstory
group tends to have a mixed canopy dominat-
ed by all sizes of Pseudotsuga menziesii and
small Abies voncolor (Table 1). Fopuhis tremu-
loicles and Picea pungens are often present.
Half of the members of understory Groups C
and D have this mixed overstor)' composition.

These overstoiy groups are indicated on the
same DCA ordination diagram (Fig. 3b). Since
this diagram represents the ordination of
shrubby and herbaceous species in plots, and
the understory composition varies within the
overstoiy groups, it is not sui-prising that these
overstory cluster groups are scattered within
the ordination diagram. This suggests that the
understory is to some extent independent of
the overstoiy.

Final classification groups resulted from
the intersection of the understoiy groups and
overstoiy groups overlain on the ordination
diagram (Fig. 3c). This results in eight final
classification groups plus two single-member
groups and one outlier plot. The symbol for
each classification group is a combination of
its overstoiy cluster group (I, II, or III; Fig. 4)
and its understory cluster group (A, B, C, or
D; Fig. 2). Groups are presented in an age
sequence within their overstoiy group, from
youngest (IB) to oldest (HID).

Community Types

The following paragraphs summarize tree
composition and structure, understoiy species
composition, and environment of each of the
eight community types (Tables 1, 2). In prior
analyses (Baker and Walford 1995) the gradi-
ent controlling spatial variation of the shrubby
and herbaceous vegetation mosaic was found
to be age and disturbance related. Variables
most affected by disturbance events are illus-
trated for each community type (Fig. 5). A map
of the patches and their community type is in
Figure 1.

[IB] Populus (ingustifolia / Agrostis scahra. —
The 14 patches constituting this community
type are found on bars close to the channel in
both height and distance (Fig. 5). Of the eight
major types, this type was most recently estab-
lished (mean age = 31 years) and has the largest
mean surficial sediment size. Its soils have low
organic matter. Patches of this type are most
common in the middle parts of the study reach
(Fig. 1). This type is characterized by an abun-

(b)

I , oi ■*
III**

, Of o

Overstory
Cluster
Groups

Fig. 3. Ordination diagrams obtained by detrended cor-
respondence analysis of data on percent cover of shruliby
and herbaceous species in the plots: (a) plots are coded
according to their understory cluster analysis group (see
Fig. 2); (b) plots are coded according to their overstory
cluster analysis group (see Fig. 4); (c) plots are coded
according to their final community type, based on the
intersection of understory cluster groups with overstoiy
cluster groups.

dance of P. angiistifolia seedlings and saplings
(Table 1) and sometimes small to medium-size
trees. Picea pungens seedlings are almost always
present and are sometimes abundant along with
saplings and small trees. Small Pseudotsuga
menziesii and Abies concolor may be present.
Herbaceous vegetation is veiy sparse (Table 2).
Graminoids dominate the understor)' with both
Agrostis scabra and either Trisetem niontanuni
or T. spicatum always present. Epilobiwn lati-
foliiun often occurs in significant amounts.

[lA] Popuhis (ingustifolia I Alnus incana. —
Patches of this type are on average 10 years
older than those of Type IB (Fig. 5). This com-
munity type is found on bars slightly higher
above the channel than those of IB. Surface
sediment sizes are diverse, but soils have little
organic matter. Like Type IB, these patches

1995]

Animas River Riparian Vegetation

293

Fig. 4. Overstoiy cluster analysis dendrogram based on
size-class data of tree species. Plot numbers correspond
with patch numbers on the study area map (see Fig 1).
The dashed line indicates the similarity level at which
overstoiy groups were separated.

are most common in the middle part of the
study reach (Fig. 1). Populus angiistifolia char-
acterizes stands of this txpe; many small and
some medium- size trees are present along with
abundant seedlings and saplings (Table 1).
Picea pungens and Pseudotsiiga menziesii seed-
lings and saplings are usually present. The
understoiy composition distinguishes this type
from IB (Table 2). Shrubs are more common.
Alnus incana is always present and Salix drum-
mondiana is usually present. Agrostis scahra is
occasionalh' present in minor amounts.

[IC] Populus angustifoIia-Pseudotsuga men-
ziesii I Pyrola asarifoUa. — This small commu-
nity type comprises two patches on bars and
one on a terrace, with an average surface par-
ticle size <10 mm and soils with low organic
matter (Fig. 5). Patches in this community type
are scattered along the study reach (Fig. 1). The
largest trees and most abundant seedlings of this
type are P. angustifolia (Table 1). Pseudotsuga
menziesii are always present as seedlings

dirough medium-size trees. Medium-size Picea
pimgens or Abies concolor may also be pres-
ent. Pinus strobiformis seedlings or saplings
are always present in this type. The under-
stoiy of this type is not dense, and all but one
of the understory species have cover values
<0.8% (Table 2). Three shrub species occur in
small amounts. Pyrola asarifoUa is always pres-
ent in the highest amount of any understory
species.

[ID] Piceo pungens-PopuIus angustifolia I
Antennaria parvjifolia. — Three terrace patches
and two patches on bars make up this commu-
nity tyi^e, which occurs on surfaces < 1 m above
the channel that have soils with low organic
matter content (Fig. 5). Patches in this type are
scattered along the study reach (Fig. 1). They
have strong similarities in overstory composition
and weak ones in understoiy composition. Picea
pungens and P. angustifolia are the largest
trees of this type and are always present as
seedlings, saplings, and small trees (Table 1).
Abies concolor and Pseudotsuga menziesii seed-
lings can always be found. The lack of a con-
stant understory is reflected in the low simi-
larity level at which patches 65 and 71 are
joined in the dendrogram (Fig. 2). Six species
are present at 80% constancy, Antennaria
parvifolia having the greatest mean cover in the
type. No single herbaceous or shrubby species
is present in all five patches of this type, but in
general there is much more herbaceous and
shrubby vegetation present than in types lA,
IB, and IC (Table 2).

[IIA] Picea pungens I Alnus incana / Equi-
setum arvense. — Patches of this community
type occur in more persistently moist areas.
They span several age classes, can be found on
bars or terraces, and have developed finer sur-
ficial sediments than might be expected for
their age class (Fig. 5). Their soils typically
contain only a little more organic matter than
soils in patches of Type I. Patches in this type
are scattered along the study reach (Fig. 1).
Picea pungens seedlings, saplings, and small to
medium-size trees as well as P. angustifolia of
various sizes characterize the type (Table 1).
Few other tree species occur, although Pseudo-
tsuga menziesii may be present in small
amounts. Alnus incana and Salix drummondi-
ana are the dominant shrubs of this type, both
occurring in greater amounts here than in any
other types (Table 2). Equisetum arvense is
always present in substantial amounts. Sedges

294

Great BasiiN Naturalist

[Volume 55

i *

§ c^

ID
I-

\n> K

CO o

1 =*t

-r

-r yi

i *

i^ *

i =*t

5 =**:

5 =tfe

10 lO -- O ID ac

JO ;o ^H -f cc oi ;c
-H lo t~^ oi o o o

M Ol Ol Ol

oi — ^ d o o d d

;r) o cr; o p o o
t~^ CO lo -r d d d

I- o o o o o
d d d d d d

t- d d d d d d

-r t^ o o o o

-r '^ di d (6 d> di

2 m j- p c c j:
a— c ^cococo '"-'

^T^plScotMCOrrcO

£ O ^'^ CO CO CO 7'
•~ X i^cO^tNcoA

cc -H CO io -r — ; lo
^ -t c-i —H d -^ d

00 CO CO CO CO CO O
-i d '>\ d d d d-

O CO I^ OJ CO -C

d CO t- d CO ^

CO oi --r cc -r p -f
^ -t t--^ d c^i d d

CT) -* CO

00 oi p cc -r o oi
oi -r ^ in -t CO oi

O CO CO CO t^ o o

i-^ —i -H — * -H d d

t— DO t- O p p p

in t-^ in ^ d d d

05 -H -H p p p p

t^ 00 00 d d d d

§i bt c P S £ p

SiS C ^COCOCO ^

^'^ .^ CO CM c^ "^ CO

S 5j ^ 1 CO CO c^ '

j; 3^ -y! CO -^ ci CO A

35 O O 05 00 05 05

CO CO in CO p

CC Ol o o

d d d d

o o o o o o o
d d d d d d d

o o o -r fM o o
d d d d d d d

t^ O O CO CO o o

^ d c<i i>i d d d

o r^ 1-- i^ t^ o o

o oq t~- c^ CO o o

d d d d d d d

C^ O fM o o o o
— < d d d d d d

£;= '-. '^cococo '-'

s.-a i= CO <M CO tt CO

a 5i &^ CO CO CO -^

s ih ;^co^c<icoA

1995]

Animas River Riparian Vegetation

295

B =tt

i =fe

5 gi

i =tfe

i =tt:

2i =**:

5 =«=

i ^

lo lo 00 in CD oq "*

O O ■-< t^ LO CO ^

00 O CO OD IC ^ o

oi rr CO CO r^ CO lo

O-l f^ 00 CO --

lO lO t-^ CO t--

-r p 00 CD p oq o
o oj t-^ to in c<i (N

00 o -t; p p CD -r
-f CO -t t--^ CO -r -r

CO t-; t-; CO p t^ t^

r-^ cq '-r t^ t-^ (N (N

o r- (M o 00 o o

^ lO lO CD lO '-H O

in 00 o CD ^ o o
r^ ^ c<i

00 iXJ ^ CM 05 CD o-i

00 in o '-i i~~: o CD

CO 00 CO O 00 CO O

05 oi C5 t^ c<i — H ci

C72 ^H 00 CO 00 CZD ^H

o -r 00 CO — H o o

o o o o o o o

(M O O O O O O

d c> cj o o c> o

o o o o o o o

-r o o -* o o o
CO —5 -r -r d d d

O CO o o o o o
c4 cN d d d d d

o o o o o o o

o o o o o o o
di d> d> d> di d> d>

— I c^ o o o o o

-r -H o o o o o
— < d d d d d d

o o CO in o CO in

p p CO 00 00 O CO

'H oi oi iri t--^ ^ ^

(N ^ <M -H

-f ^H 00 05 in CO ^H
t-'^ CD -t d d d d

-r 00 -r CD oi o o
eg eg rf ^ — ; d d

CO CO O O CD O CO

CO -t c-i -r ej d d

p p t-; p p p p
00 ^ t-^ iri CO d d

00 00 t>j ej CO p p

d -* in CO d d d

[£;in^a50ooo
.—.oicg^dddd

s~— corMco-fco

f^ bjo ^ c c c ^
-C::C^?cjcjoc
CO— - '^cococo "^
s"0— coe-ico-fco

S 1; ^ I CO CO CO T

;s,5; Jico^cgcoA

-X t/j ^ C C C r-

C "^ .1^ m oi '^ "T c^

i; n: tir' I ro r^ r^ ^

^ -^ Cr CO ^ r>i CO A

296 Great Basin Naturalist [Volume 55

Table 2. Mean |)c'icfnt cover \aliies lor main species in coiniiiinnt\ t\pes (includes species with at least 19( cover in
an\ plot). Species with double underlined co\er \alues had 100% constancx in that coniniunit) t\pe. Species with single
underlined cover values had >S0% and <100% constancy in that conunnnity type.

C^onnnunitx t\ pe

IB lA IC ID IIA III) IIIC: HID

—

.5.1

—

0.6

1.0

1.2

—

2.4

—

0.4

—

1.6

0.5

—

0.8

—

0.2

0.4

0.5

Shrubs

Acer ^hihniiii Ton: — — — 1.7 — 0.2 3^
Alntis incana (L.) Moench ssp.

temiifolia (Nun.) Breitnng OJ. 0^ — 0.1 5A)

Amelanchier alnifolia (Nutt.) Nutt. — — O.I 0.6 —

Corniis sericea L. — — — — —

Jitiiipenis communis L. — — 0.1 2. .3 —
Lonicera involiicrata (Richars.)

Banks ex Sprang. — — 0.1 0.1 0.1

Mahonia repens (Lindl.) G. Don — — — L3 —

Prumifi vir^inia)i<i h. — — — — —

Ru.sa woodsii Lindl. — — 0^ 2.1 —

Ruhiis parviflorus Nutt — 0.1 — — —

Riihiis sfrigosiis Michx. — — — 0.4 0.1
Salix drummondkma Barratt

e.xHook. 0.1 (18 — — 1L3 _ _ _

Salix lasiamlra Benth. — 0.4 — — — — — —

FORBS
Achillea millefolium L. var.

lamdosa (Nutt.) Piper — 0.2 _ 0.9 (U LO — 0.8
Anaphalis margaritaceae (L.) Benth.

& Hook. f. ex C. B. Clarke _ _ _ _ 0.3 0.1 0.1 —

Antenmiria pan ifolia Nutt — — 0.1 2^ — L2 — 4J)

Apocijmim amlro.saemifolium L. — — — 0.1 — — — 1.8

Arabis divaricarpa A. Neh. O.I — 0.3 — — — — —

Arabis hirsuta (L.) Scop. 0.1 — — — — — — —

Arctostaphylos uva-ursi [L.) Spreng. — — — — 0.1

Arnica cordifolia Hook. — — — — —

Artemisia carruthii Wood ex

Carruthers — — —

Artemisia franserioides Greene — — —

Artemisia ludoviciana Nutt. — — —

Cerastiiim fontanutn Baumg. — — —

Epilobmm angu.stifolitmi L. — — 0.3

Epilobium latifolium L. 0.7 0.1 — — — —

Equisetum arvense L. 0.1 0.1 — — 16.8 —

Erigeron eximius Greene — — — 0.3 — 1.3

Erigeron flagellaris Gray — — — 0.2 — 0.5

Erigeron formosisimus Greene — — — 3.0 — 1.3
Erigeron speciosus (Lindl.) DC.

var. speciosus — — — — — 0. 1

0.3

—

1.4

—

1.8

0.8

—

0.1

—

0.3

0.2

0.3

0.1

0.5

0.9

0.2

0.7

—

0.1

—

0.2

—

0.3

0.2

0.3

0.4

0.1

—

0.5

—

2.4

1995]

Animas River Riparian Vegetation

297

Table 2. Continued.

Fragaria vesca L. ssp. aincricunu

(Porter) Staudt
Fragaria virginiana Duchesne
Galium triflonim Michx.
Geranium richardsonii Fisch. &

Traut\.
Haplopappus parryi Gray
Heracleum lanatwn Michx.
Hwnulus hipulus L. \ar lupuloides

E. Small
Lathijrus hracltyculyx Rvdb.
Ligusticwn poiieh Coiilt. & Rose
Lithospermwn multiflorum Torr.

ex Gray
Mertensia franciscana Heller
Otihilia secunda (L.) House
Osmorhiza chilensis Hook. & Am.
Osmorhiza depauperata Phil.
Oxytropis deflexa (Pallas) DC. van

sericea Torr. & Gray
Platanthera hyperborea (L.) Lindl.

van hyperborea
Fotentilla hippiana Lehm.
Pseudocymoptents inuntanus (Gray)

Coult. & Rose
Pteridium aquilinutn (L.) Kuhn
Pyrola a.sarifolia Michx.
Rudbeckia laciniata L.
Senecio fendleri Gray
Silene menziesii Hook.
Smilacina stellata (L.) Desf.
Solidago canadensis L. van salebrosa

(Piper) M. E. Jones
Thai ictnun fendleri Engelm, ex Gra\'
Thlaspi montanum L.
Urtica dioica L.

Vicia americana Muhl. ex Willd.
Viola canadensis L.
Viola nephrophylla Greene

0.1

0.2
0.1

0.1

0.1
0.1
0.1

0.1

Community t\ pe

0.8
0.3
0.7

2.0
2.4
1.0

0.7
0.2

1.2
1.3

1.0

1.4

0.4

03
0.2
5.1

0.1
1.1

0.2
0.3

IIA

0.1

0.4

0.1
0.3

0.1

IID

1.5
0.6
0.2

2.2
2.1
0.2

04
0.2
0.1

0.1

0.3

0.6

0.6
0.3
02
QA

0.1
5.1

0.5
0.6
0.3

I lie

0.1
2.6

9 9

— 0.7

0.1
1.1

3.2
1.6

0.1

18.3
4.0
0.4
0.3

0.1
0.3

0.1
2.1
0.1

HID

0.9
1.4
0.6

M
L5
3.6

0.1
0.3
0.1

L3
0.9
0.2
0.1
0.5

0.2

0.8

0.4
0.1
0.4
1.6
0.3
0.2
2^

0.2

0.2
10
0.7

Graminoids

Agropyron trachycaulum (Link) Malte

ex H.E Lewis van trachycaulum
Agropyron trachycaulum (Link) Malte

ex H.E Lewis van unilaterale

(Vasey) Malte

0.4

0.1

0.2

0.4

298

Great Basin Naturalist

[Volume 55

Tablk 2. Continued.

Coniinunity t\ pc

ID IIA

III)

I lie

HID

Afirostis scabra VVillcl.

Afixostis stolonijcra L. \ ar. .stdlonijera

—

Blepharamuron trichnlcpis (ToiT.)

Nash

—

Bromti.s (inoiniilu.s Huip. ex Fonrn.

—

Bnniitis ciliatu.s L.

—

Caluinagrostis canadensis (Michx.)

Beauv.

—

Carexfoenea VVilld.

—

Carex kelloggii W. Boott ex S. Wats.

—

Carex pityophila Mackenzie

—

Carex rossii Boott ex Hook.

0.1

Desrhainpsia cespitosa (L.) Beauv.

0.5

Elyinns <j,laucm Buck!.

—

Fcstiica tl}urberi Vasey

—

Juncus balticiis Willd.

—

JunciK saxhnonfanus A. Nels.

—

Juncus tenuis Willd. var. uniflorus

(Farw.) Farw.

—

Koeleria cristata (L.) Pers.

—

Muhlenbcr^ia richardsonis (Trin.)

Rydb.

—

Onjzopsis asperifoUa Michx.

0,1

Poo palustris L.

0.1

Schizachne purpurascens (Ton.)

Swallen

—

Sitanion hijstrix (Nutt.) J. G. Sm.

—

Stipa Columbiana Macoun

—

Trisetum niontamim Vasey

0.4

Trisetwn spicatum (L.) Richter

0.1

LXOTICS
Poa pratensis L.

0.8

Poa trivialis L.

—

Taraxacum officinale Weber

0.1

—

0.9

—

0.1

—

1.2

—

0.4
0.2
M

0.1

1.7

—

0.2

0.1

1.3

0.6

0.7

—

1.1

3.6
0.2

3.3

O.S

2.0

0.3

0.2

0.1

0.3

0.6

0.1
1.1

—

0.3

—

0.1

0.4

0.6
0.3
0.3

—

1.0

—

0.5

1.3

0.1

2.3

0.1

0.6

—

0.1

1.7

0.1

0.2

0.1

—

0.6

0.1

—

0.2

0.1

—

O.S

—

0.1

—

0.1

3.7

—

2.3

—

0.6

0.1

1.4

0.1

(L5

and rushes dominate the graminoid layer, with
Juncus balticus always present.

[IID] Picea pungens-Populus trcnudoides I
Mahonia repens. — Ages of these patches span
the range from 40 to 130 years. Most are found
on terraces, with 89% of their surficial sedi-
ment <1 mm in size (Fig. 5). Soils in this type
contain moderate amounts of organic matter
(Fig. 5). Patches in this community type are
most common in the upper two-thirds of the
study reach (Fig. 1). This type is characterized
by P. tremuloides and P. pungens of all sizes
(Table 1). Populus angustifolia is not regenerat-

ing strongly but is sometimes the largest tree
in a patch. Abies concolor and Pseudofsuga
menziesii may also be present, usualK' in small
sizes. Mahonia repens, Juniperus communis, and
Rosa tvoodsii are dominant shrubs (Table 2);
Thalictrum fendleri and Smdacina stellata are
dominant forbs. The graminoid layer is char-
acterized by Bromus ciliatus, with significant
amounts of Carex foenea often present.

[IIIC] Pseudofsuga menziesii / Acer glabrum
I Pijrola asarifolia. — The four patches of this
community type are all found on terraces with
fine surface sediments, have about the same

1995]

Animas River Riparian Vegetation

299

I A

I c

(n=3)

I D

(r>=S)

II A

(n=S)

II D

III c

4«i

H.T

10 15

.■■■T.l.lf. ■■»..■

.I..f..>i

10 15

.. .lilMII

UuU

4...^

10 15

III D ^^

(r>=n)

-Wt-J

10 15

(a) Age Class
(lOyr)

io_

5 J

f Ipp ■ ¥

5 J

t#-

^ 2 :^ S « S !!! § ^ a

1 oj
5 J

0 .

^ 2 !5 S S S !i? 9 ^ 8

10.
5-

T I I I I I I I I I

"?2£S!(i8;!?§^S

10-
5 J

lOJ
5 J

"?2£S!i38^9«8
°'"2£S!^SSS^

(b) Surface Sediment
Size (mm)

10_
5 J

w ■ yip

lO rN I/) c*) lO ^

O - ^n CM lO O

4^0^

^#-

-^^#-

, ' - ' N ' n '

Q in I lO . I/) I tO

- - M tN n n

0.4 %

I n = 3 I

(n = 0)

0.5 %

( n = 2)

1.0 %
(n = 1 )

6.7 %

(n = 2)

4 4.0 %
(n = 2 )

2 0.3 %
(n = 61

Mean
(d) Organic
Matter

Fig. .5. En\'iionmental attributes within each community type (see Fig.
vertical axis.

1). Histograms (a-c) show the frequenc\' on the

age-class span as patches in Type IID, and have
the largest mean organic matter content of
community types along the reach (Fig. 5). The
overstory is a mixed forest similar to Type
HID, but the understory is not as rich. Patches
in this type are restricted to the lower one-
third of the study reach (Fig. 1). Pseudotsuga

menziesii, Popiilm angustifolia, and usually Abies
concolor are the large trees of these patches
(Table 1). Popiilus angustifolia seedlings or sap-
lings are rare. Regeneration appears strongest
in Abies concolor and P. menziesii (Table 1).
Acer glabrum is the dominant shrub usually
associated with lesser amounts of A/nu,s

300

Ci{i' AT Basin Naturalist

[Volume 55

incaruL Priiiius lir^iiiidiid. and liosa uoodsii
(Table 2). Pyrola asarifolia is the most preva-
lent forb with cover values averaging 18%.
Geranium richardsonii, Smilacina stellata,
Haplop(i])])us parnji, and Arlctiiisia franseri-
oides al\va>'s occur. Onjzop.sis a.spcrifolia is the
dominant graminoid. with Broiniis cilidfus in
lesser amounts.

[IIIDJ Fopidus trcmidoides-Pseudoisuga
nienziesii I Bromus ciliatiis. — Examples of this
community' type are found on terraces an aver-
age of 1.75 m above the channel (Fig. 5). Surface
particles are predominantly <1 mm, while
soils have about 20% organic matter content
(Fig. 5). The 11 patches comprising this type
have the oldest average age and are most com-
mon in the lower one-third and upper one-
third of the study reach (Fig. 1). This group of
patches is a mixed forest type, with the dens-
est underston' of all the types. Popuhis treinu-
loides and P. menziesii tend to be the largest
and the most abundant trees of these patches.
Some patches have very large Populus angusti-
folia as well. Abies concolor seedlings and sap-
lings are iilways present, sometimes in veiy large
numbers. A few Picea pungens of various sizes
usually can be found. Mahonia repens\ Rosa
woodsii, and Jiiniperus conununis are co-domi-
nant shrubs. Five Erigeron species were found
in the 1 1 patches of this type, with one to three
species present in each patch. Antennaria parci-
folia and Geranium richardsonii had high cover
values in most patches. Dominant grasses were
Bromus ciliatus and Orijzopsis asperifolia,
their quantities being distinctive from any
other type.

Discussion

The New Classification Technique

Ecologists working in northern climates
have long recognized that individual stratal
layers (e.g., tree and shrub) or "synusia" within
forest communities may be distributed some-
what independently and may not all have the
same value in distinguishing vegetation units
(Whittaker 1962). Classification approaches of
the northern European schools, such as the
"sociations" of the Uppsala school and "site-
types" of Cajander (Whittaker 1962), empha-
size that lower strata often are more useful in
classification, although the tree layer is of
some importance as well.

Yet, conunon multivariate techniques for
analyzing \'egetation data ignore the synusial
affiliations of the species in a community (e.g.,
Gauch 1982). Since cover values of overstory
tree species are often large relative to cover
values of understoiy species, overstoiy species
may prevent understory species from having
much influence on the outcome of multivariate
analyses (e.g., Padgett et al. 1989). Moreover,
these techni(|ues do not take advantage of the
different kinds of data that are useful in
describing the structure of different synusia.
For example, it is much easier and more useful
to obtain size-class structure data than cover
data for overstory trees in forests, as size-class
data can indicate tree composition and abun-
dance as well as population structure (e.g.,
regeneration status). The value of both kinds
of data in forests has long been recognized
(e.g., Pfister et al. 1977). Yet, size-class data
and cover data are incompatible and cannot
both be used readily in a single multi\ariate
analysis. The technique used here allows an
equal consideration of both the understoiy and
overstoiy data and data of different types from
different synusia. Community types that are
produced have homogeneous tree populations
combined with homogeneous understories.

Community types identified using this tech-
nique can be useful in consei-vation and man-
agement. The eight communitx' t\pes reflect
major variations in vegetation and environment
along the reach. A preserve could be designed,
shorter than the full length of the reach, that
contained all eight connnunity types; it is like-
ly that much of the floristic variation along the
reach would then be protected. Moreover, the
community types should be useful in manage-
ment because they are (1) functionalK' homo-
geneous, in the sense that tree populations
within a type might respond similarK to dis-
turbances, and (2) environmentalh' sensitive,
in the sense that the finer en\'ironinental dis-
crimination of understory synusia common in
northern regions has been incoiporated.

The Community Types in a
Regional Context

Community types identified in this study
have not been widely reported. This is proba-
bly due to the river-reach scale of the study
and the tendency to exclude \'eiy >'oung vege-
tation in developing regional classifications.
This is also one of a very few river systems in die

1995]

Animas River Riparian Vegetation

301

southern Rocky Mountains with a complete
mosaic of riparian vegetation relatively free
from human land uses; thus, there have been
few other opportunities for this kind of studv'.

Our Type II IC is very similar to Baker s
(1989) Abies concolor-Picea pungens-Popidus
angiistifolia / Acer glabrum association, previ-
ously found along the Animas River and the
San Juan River, as well as in northern New
Mexico (DeVelice et al. 1986). Baker collected
data from within the study reach, which ex-
plains the similarity of his association with our
Type IIIC, but his goal was to classify vegeta-
tion regionally based upon the similarity of the
more mature vegetation patches along sepa-
rate rivers. Baker did not sample the mature
stands containing Popiihis tremidokles that are
found in our Type IID and HID, thinking they
were earlier successional stages of our Type
IIIC. However, the age-class data (Fig. 5) sug-
gest that stands within Types IID and HID
are just as old as or older than those in Type
IIIC. Thus, although it may be a necessary
shortcut for regional classification efforts, sam-
pling and classifying only the mature vegeta-
tion may result in errors if the successional se-
quence along a reach is not clearly understood.

Sampling and Classification of
Riparian Vegetation Complexes

In riparian areas, and perhaps anywhere
vegetation classification is being approached,
it is important to sample and classify not only
mature vegetation stands but younger stands
as well. The diverse patch structin-e along rivers
may only reach a homogeneous mature com-
position similar to that in the older patches if
the fluvial disturbances that have produced
the mosaic are controlled. Moreover, younger
stands may not all be leading to the same
mature community; there may instead be
more than one serai sequence.

This spatial and temporal complexity at the
river-reach scale compounds the difficulty of
developing regional classifications. One solu-
tion to this problem is to adjust the scale of
sampling to the scale of patchiness produced
l^y the primaiy ecological processes (e.g., fires,
floods) in a particular landscape. An area such
as this free-flowing river requires fine-scale
sampling as there is a fine-scale mosaic pro-
duced by disturbances and geomorphic varia-
tion. A river with less geomoiphic complexity
or a coarser, more infrequent flood-produced

patchiness may require only a coarse sampling
focused on the more mature vegetation.

Spatial Variation in Vegetation
Along the Reach

The sampling and vegetation analysis sug-
gest that substantial landscape diversity is pro-
duced by floods and geomorphic variation
along this reach of the Animas River (Fig. 1).
The spatial aiTangement of this diversity is con-
trolled in part by location of tributaries and
width of the valley floor, both of which influ-
ence how and where floods create new patches.
Needle Creek flows into the Animas River in
approximately the middle of the study reach
(Fig. 1). Smaller tributaries enter above and
below this point, but none carries as great a
volume of water. The valley floor also widens
approximately 0.5 km below the entry of
Needle Creek.

The first four community types (lA, IB, IC,
and ID) with Populus angiistifolia in their over-
stoiy are found primarily in this section at the
outer river curves or mid-channel where scour-
ing is greatest. The wet environment IIA
patches also are found in this wider section,
often away from the main channel on side
channels that dissect major patches. None of the
largest trees is found in this middle section of
the reach. In contrast, community types with a
more mature overstory are more common in
the lower one-third and upper one-third of the
study reach. These parts are narrower and
have fewer substantial tributaries. Many of the
mature vegetation patches are located on ter-
races quite high above the channel in these
parts of the reach.

Conclusions

This study of riparian vegetation on the
river-reach scale revealed considerable spatial
and temporal complexity. Flood disturbances,
modulated by variation in valley morphology
and tributary location, have created distinct
patchiness in the vegetation. A new technique,
based on both overstoiy and understory species,
offers an improved quantitative method for
identifying community types. If classification
is to be used effectively to aid in consenation,
greater attention to younger, less mature stands
of vegetation may be needed. These young
stands are a major component of the biodiver-
sity on the river-reach scale and can represent

302

Great Basin Naturalist

[Volume 55

serai stages of new vegetation associations
unlike the association represented by present
mature stands. Spatial complexity' along a single
river may make the development of regional
classifications, based on many ri\ers, more dif-
ficult. However, regional classifications can
still be completed, and will be more valuable,
if sampling efforts are tuned to the scale of
patchiness and complexit> along river reaches.

Acknowledgments

This research was completed with funds
from the Ecological Research Division, Office
of Health and Environmental Research, U.S.
Department of Energy (Grant No. DE-FG02-
90ER6()977). This support does not constitute an
endorsement by DOE of the views expressed
in this article. Comments of Sherman Swanson
and an anonymous reviewer improved the
manuscript.

Literature Cited

Baker, W. L. 1988. Size-class structure of contiguous

riparian woodlands along a Rocky Mountain river.

Physical Geography 9: 1-14.
. 1989. Classification of the riparian vegetation ot the

montane and subalpine zones in western Colorado.

Great Basin Naturalist 49; 214-228.
. 1990. Climatic and hydrologic effects on the regen-

eration of Populus angustifolia James along the
Animas River, Colorado. Journal of Biogeography 17:
59-7.3.

Baker, W. L., and G. M. Walford. 199.5. Multiple stable
states and models of riparian vegetation succession
on the Animas River, Colorado. Annals of the Associ-
ation of American Geographers 85; .320-.3.38.

Campbell, C. J., and W. Green. 1968. Perpetual succes-
sion of stream-channel vegetation in a semiarid region,
journal of the Arizona Academy of Science 5; 86-98.

Cooper, D. J., and T. R. Cottrell. 1990. Classification of
riparian vegetation in the northern Colorado Front
Range. Unpublished report. The Nature Consen'ancy,
Colorado Field Office. Boulder, CO. 85 pp.

Daubenmire, R. 1952. Forest vegetation of northern
Idaho and adjacent Washington, and its hearing on
concepts of vegetation classification. Ecological
Monographs 22; .301-330.

DeVelice, R. L., J. A. Ludwig, W. H. Moir, and F Ronco,
Jr. 1986. A classification of forest habitat types of
northern New Mexico and southern Colorado.
US DA Forest Service General Technical Report
RM-131. Rocky Mountain Forest and Range E.xperi-
ment Station, Fort Collins, CO. 59 pp.

Gauch, H. G., Jr. 1982. Multivariate analysis in commu-
nity ecology. Cambridge University Press, Cambridge,
UK. 298 pp.

Hansen, R, K. Boggs, R. Pfister, and J. joy. 1991.
Classification and management of riparian and wet-
land sites in Montana. Draft version 1. Montana

Riparian .Association, Montana Forest and Conserva-
tion E.xperiment Station, School of Forestn; Univer-
sity of Montana, Missoula. 478 pp.
HlNTER, M. L., Jr. 1991. Coping with ignorance; the
coarse-filter strategy for maintaining biodiversity.
Pages 266-281 in K. A. Kohn, editor. Balancing on
the brink of extinction; the Endangered Species Act
and lessons for the future. Island Press, Washington,
DC.

JOFINSON, R. R., C. D. ZlEBELL, D. R. P.ATTON, R F
Ffollkjtt, and R. H. Hamre. 1985. Riparian
ecosystems and their management; reconciling con-
flicting uses. US DA Forest Service General Tech-
nical Report RM-120. Rocky Mountain Forest and
Range Experiment Station, Fort Collins, CO. 523 pp.

K.'\RTESZ, J. T., AND R. K.\RTESZ. 1980. A synonymized
checklist of the vascular flora of the United States,
Canada, and Greenland. University' of North Carolina
Press, Chapel Hill. 498 pp.

Knope F L. 1985. Significance of riparian vegetation to
breeding birds across an altitudinal cline. Pages
105-111 in R. R. Johnson, C. D. Ziebell, D. R. Patton,
P. F Ffolliott, and R. H. Hamre, editors. Riparian
ecosystems and their management: reconciling con-
flicting uses. USDA Forest Service General Tech-
nical Report RM-120, Rock-y Mountain Forest and
Range Experiment Station, Fort Collins, CO.

Malanson, G. P 1993. Riparian landscapes. Cambridge
University Press, Cambridge, UK. 296 pp.

Mueller-Dombois, D., and H. Ellenberg. 1974. Aims
and methods of vegetation ecology. John Wiley and
Sons, New York. 547 pp.

MUTZ, K. M., and R. Graham. 1982. Riparian community-
type classification — Big Piney Ranger District,
Wyoming. Unpublished report, USDA Forest Semce
Region 4, Ogden, UT. 92 pp.

O'CONNELL, M. A., and R. E Noss. 1992. Private land
management for biodiversity consenation. En\'iron-
mental Management 16: 43.5—450.

Osterkamr W. R., and C. R. Hupp. 1984. Geomorphic
and vegetative characteristics along three northern
Virginia streams. Geological Society of America
Bulletin 95: 109.3-1101.

Osterw'ald, D. B. 1989. Cinders and smoke. A mile by mile
guide for the Durango to Silverton narroyv gauge
trip. Westei-n Guideways, Lakewood, CO. 142 pp.

Padgett, W. G., A. P Youngblood, and A. H. Winward.
1989. Riparian community tyise classification of Utali
and southeastern Idaho. Unpublished report R4-
Ecol-89-01, USDA Forest Service Intermountain
Region, Ogden, UT. 191 pp.

Peet, R. K. 1981. Forest \egetation of the Colorado Front
Range: composition and d>namics. Vegetatio 45: 3-75.

Pfister, R. D., and S. F Arno. 1980. Classifying forest
habitat types based on potential clima.x vegetation.
Forest Science 26; 52-70.

Pfister, R. D., B. L. Kovalchik, S. E Arno, and R. C.
Presby. 1977. Forest habitat types of Montana. USDA
Forest Service General Technical Report INT-34,
Intennountain Forest and Range Experiment Station,
Ogden, UT 174 pp.

Rodriguez, J. 1986. User's guide to MOMENTS. Page 21
in H. R. Burger, editor. Personal computer softyvare for
geological education. National Association of Geolog>'
Teachers, USA.

Romesburc;, H. C. 1984. Cluster analysis for researchers.
Lifetime Learning Publishers, Belmont, CA. .334 pp.

1995]

Animas River Riparian Vegetation

303

SPSS. 1990. SPSS/PC + . Statistical package for the social
sciences. SPSS, Inc., Chicago.

Strategic Mapping, Inc. 1991. Atlas Draw user's guide.
Strategic Mapping, Inc., San Jose, CA.

Ter Braak^C. J. K 1988. CANOCO— a FORTRAN pro-
giam for ciinonical comniunit)' ordination by [partial]
[detrended] [canonical] correspondence analysis,
principal components analysis and redundancy
analysis. Technical Report LVVA-88-02, Agricultural
Mathematics Group, Wageningen, The Netherlands.
Microcomputer Power, Ithaca, NY. 95 pp.

USA-CERL. 1991. GRASS 4.0 user's reference manual.
United States Army Construction Engineering
Research Laboraton-, Champaign, IL. 280 pp.

VVhittaker, R. H. 1962. Classification of natural commu-
nities. Botanical Review 28: 1-239.

WiNVVARD, A. H., AND Padgett, W. G. 1989. Special con-
siderations when classifying riparian areas. Pages
176-179 in Proceedings — land classification based

on vegetation: applications for resource management.
US DA Forest Service General Technical Report
INT-257, Intermountain Forest and Range E.xperi-
ment Station, Ogden, UT.

WOLMAN, M. G. 1954. A method of sampling coarse river-
bed material. Transactions of the American Geo-
physical Union 35: 951-956.

YouNGBLOOD, A. P, VV. G. Padgett, and A. H. Winward.
1985a. Riparian community type classification of
northern Utah and adjacent Idaho. Unpublished
report, USDA Forest Sei-vice, Intermountain Region,
Ogden, UT 103 pp.

. 1985b. Riparian community type classification of

eastern Idaho-western Wyoming. Unpublished report
R4-Ecol-85-01, USDA Forest Service, Intermountain
Region, Ogden, UT.

Received 24 June 1994
Accepted 24 January 1995

ADl^n IONS TO KNOWLEDGE OF PALEOCENE MAMMALS FROM
THE NOm II HORN FORMATION, CENTRAL UTAH

Kichard L. C>itclli', Nicliolas J. Czaplewski', and Kenneth D. Rose^

Abstkact. — The distinctive but inadeejuately known Paleocene faunas of central Utah are significant in that they
sample a time interval not well represented l)y secjucnces in other areas. New materials from the Wagon Road (late
Puercan) and Dragon (earl\' Torrejonian) local faunas. North Horn Formation, provide additional information on the
composition of the assemblages and systematics of included mannnal ta.xa. The proteutherian tPropalaeosinopa is
recorded, for the first time, from the Wagon Road fauna, indicating a significant extension for the enigmatic family
Pantolestidae, othei-wise first known from the Torrejonian. Associated premolars oi Aphnmorus sirnpsoni, a pentacodon-
tid proteutherian from the Dragon fauna, indicate that the species is more distinct from its Torrejonian congener, A.
fmudalor, than previousK' suspected. New materials of Desmatuclaenus hermaeus uphold the synonymy of this species
with D. paracreodus and permit more adequate definition of the genus with respect to the arctocyonid Loxolophua and
the phenacodontid Tetraclaenodon; because Desinatoclaeniis appears to share derived morphology with Loxolophu.s, we
refer it to the basal condylarth family Arctocyonidae. The periptychid condylarth Haploconus, represented in the Wagon
Road fauna by the geologically oldest described species of the genus, H. elachistus, is shown to be distinctive in the con-
figuration of lower molars and premolars; H. elachistus appears to be more primitive than species known from the
Torrejonian of New Mexico. Some features oi Haploconus are suggestive of the Conacodontinae, a distinctive clade of
diminutive periptychids.

Kcij words: Paleocene, North Horn Fonnation, Puercan, Torrejonian, Dragon local fauna. Wagon Road local fauna,
Mainnuitiii.

Paleocene mammals were first reported fi-om
the North Honi Formation, Emery and Sanpete
connties, UT, by Gazin (1938). Further field-
work resulted in the recoveiy of additional taxa,
interpreted as representing two faunas, from
two main localities (Gazin 1939, 1941). In sub-
sequent years, additional sites in the region
have yielded finther specimens, including more
taxa and a third faunal assemblage (Spieker
1960, Van Valen 1978, Tomida and Butler 1980,
Tomida 1982, Robison 1986, Archibald, Rigby,
and Robison 1983). Three assemblages are cur-
rently recognized, the Gas Tank, Wagon Road,
and Dragon local faunas (Robison 1986). On
the basis of the latter two, a "Dragonian" land-
mammal age was initially established (Wood et
al. 1941). Later work, including magnetic
stratigraphy and biostratigraphic comparisons,
suggests that the Gas Tank and Wagon Road
faunas are Puercan and the Dragon fauna Torre-
jonian in age (Tomida and Butler 1980, Tomida
1981, Robison 1986). Archibald et al. (1987)
tentatively assigned the Gas Tank to Pu2
{Ectoconus I Taeniolahis taocnsis interval zone).
Wagon Road to Pu3 {Taeniolahis taocnsis I

Periptyclws interval-zone), and Dragon to Tol
{Periptijchus I Tetraclaenodon interval-zone).
Both Pu2 and Pu3 are interpreted to occur
within magnetic polarit>' chron 29N (Butler and
Lindsay 1985); the Dragon faima is considered
to lie within anomaly 27N (Tomida and Butler
1980).

The Paleocene mammals of central Utah are
of special interest in both temporal and geo-
graphic contexts: they fall within a time inter-
val not well represented elsewhere, and they
lie geographically between the classic sequence
of the San Juan Basin, NiM, and faunas from
more northerly parts of the Western Interior (cf.
Archibald et al. 1987: fig. 3.1). Mammals from
the Paleocene of the North Horn Formation
are not, in general, well known. We describe
herein newly collected materials that provide
further details on the moiphology and SNstem-
atics of some of the included taxa.

The approximate locations of the major
mammal sites in the Paleocene part of the
North Horn Formation, taken from data pre-
sented by Gazin (1941) and Robison (1986),
are given in Figure 1. The materials described

OklaliDiiK. Mil

)r\'atiinil llistiin and Drpartnicnt ofZiKiIdRv, University nfOkhihoma. Norman, OK 73019.

^Department (il Cell HiclciKy ami Anatomy, Johns llopknis I niversil\ Seliool of Medicine. 725 North \\c

304

1995]

Paleocene Mammals, Utah

305

— 1

Fig. 1. Approximate locations of mammal-bearing sites
in Paleocene part of North Horn Formation, Emeiy and
Sanpete counties, UT; data from Gazin (1941) and
Robison (1986). Localities, Dragon local fauna: Dragon
Canyon (1). Wagon Road local fauna: Wagon Road (2),
\Vagon Road Ridge (3). Gas Tank local fauna: Gas Tank
Hill (4), Dairy Creek (5), Jason Spring (6), Ferron
Mountain (7; probably equivalent to OMNH V829), Blue
Lake (8), and Sage Flat (9).

herein were collected in 1993-94, through sur-
face prospecting methods. With one excep-
tion, all specimens are from the classic Dragon
Canyon (Dragon local fauna; ?Tol) and Wagon
Road (Wagon Road local fauna; ?Pu3) sites
described by Gazin (1941). The exception is a
specimen assigned to Ectoconus ditrigoniis
(OMNH 28111), collected by Jon Judd of Castle
Dale, UT, at a site south of Ferron Mountain.
The site, OMNH V829, is probably the same
as Robison's (1986) Ferron Mountain locality
(Gas Tank local fauna; ?Pu2).

The following abbreviations are used for in-
stitutions cited in the text: BYU, Brigham Young
University, Provo, UT; OMNH, Oklahoma
Museum of Natural Histoiy, Norman; USNM,
National Museum of Natural Histoiy, Washing-
ton, DC. Measurements, in mm, are as follows:
L, anteroposterior length; W, transverse width;
WTal, transverse width of talonid; WTri, trans-
verse width of trigonid.

Descriptive Accounts

Order Proteutheria

Family Pantolestidae Cope, 1884

?Propalaeosinopa sp.

Figs. 2A-B

Material.— OMNH 27681, fragment of
right dentaiy bearing the talonid of P4 (WTil
= 1.5) and complete M^ (L = 2.8, WTri = 1.8,
WTd = 1.8).

Locality and horizon. — OMNH V800,
"Wagon Road" locality (Gazin 1941, Robison
1986); Wagon Road local fauna, late Puercan
(early Paleocene). Joes Valley Member, North
Horn Formation, Emeiy County, UT.

Description and discussion. — The den-
taiy fragment includes the anterior root of P4
and the anterior root of M2. The anterior root
of P4 is bowed forward as in pentacodontids
and most pantolestids, and its placement indi-
cates that P4 was relatively long, longer than
Mj. The posterior mental foramen is large and
is positioned between the posterior root of P4
and the anterior root of M j. The talonid of P4
includes a large hypoconid and a small entoco-
nid; these two cusps are united by a small, thin
postcristid, forming a small talonid basin. The
apex of the hypoconid is on the midline of the
tooth, at the posterior termination of a cristid
obliqua that angles lingually toward the front;
the postcristid is oriented almost peipendicu-
lar to the cristid obliqua. Posterior to the post-
cristid and separated from it by a tiny trans-
verse basin, a small cuspule (hypoconulid?) is
present; this cuspule is connected to the hypo-
conid b>' a thin ridge. A tiny entoconulid, not
connected to the other cusps, is present at the
lingual base of the talonid basin.

The trigonid and talonid of M ^ are of equal
width; the trigonid is distinctly higher than
the talonid, though the tooth is moderately
worn. The protoconid and metaconid are both
triangular in occlusal outline and of equal
occlusal area; the protoconid is the taller of the
two cusps. The paraconid is small, low, and
transversely oriented. Anterior and posterior
carnassial notches are present in the para-
cristid and protocristid, respectively. Because
of the transverse orientation of the paraconid,
the paracristid forms an obtuse angle, with its
apex at the anterior carnassial notch. A short
anterior cingulum, which disappears at the
anterolingual corner of the tooth, is present.

306

Great Basin Naturalist

[Volume 55

Fig. 2. Proteutheria from tlie North Horn Formation. A, B, P4-M1 of Propalaeosinopa sp. (OMNH 27681) in occlusal
(A) and labial (B) views. C-G, Aplmmonis simpsoni (OMNH 27667); C, D, right P4 in occlusal and labial views, respec-
tively; E, F right P3 in occlusal and labial views, respectively; G, left P"* in occlusal view. Scale bar represents 2 mm;
tooth roots and jaw fragments have been eliminated to improve clarity.

1995]

Paleocene Mammals, Utah

307

The posterior wall of the trigonid is planar; the
cristid obliqua meets the base of the posterior
wall of the trigonid below the posterior carnas-
sial notch. Although it has been mostly obliter-
ated by wear, an entoconulid (or at least an ento-
cristid) appears to have been present anterior
to the entoconid.

Of described species, OMNH 27681 most
resembles the Torrejonian Propalaeosinopa di-
luculi (which we tentatively regard as distinct
from P. albertensis following Rose 1981; see
discussion in Van Valen 1967). However, the
Utah taxon differs in several respects. The pos-
terior mental foramen is more anteriorly located
than in figured specimens of P. diluculi (Simp-
son 1936: fig. 3; Krause and Gingerich 1983:
figs. 8, 9). Ml of OMNH 27681 is long and
narrow relative to the corresponding tooth of
P. diluculi: it slightly exceeds published size
ranges (Simpson 1937a, 1937b, Krause and
Gingerich 1983) in length but not width. In
the Utah taxon the cusps of M^ are somewhat
more robust and the postvallid wall more
obliquely oriented with respect to the long
axis of the tooth; the paracristid is higher, and
the metaconid lower, than in P. diluculi. The
talonid of P4 is broader and more basined than
in P. diluculi (or other species of the genus).
We regard the specimen from the North Horn
Formation as representing a distinct species,
but materials in hand are inadequate to proper-
ly diagnose and circumscribe it. Gazin (1941)
briefly described two morphs, represented by
upper molars, as generically undetermined
pantolestids; both were from die Dragon local
fauna. Of these, he found pantolestid "a" to
compare favorably with Bessoecetor {=Propo-
laeosinopa), differing from "J5. thomsom' {=P.
diluculi) in being slightly larger and in a few
morphological details. It is possible that pan-
tolestid "a" and OMNH 27681 represent the
same species, although we point out that they
derive from different horizons in the North
Horn Formation. Differential representation
precludes direct comparison with OMNH
27681.

If referral of the newly recovered specimen
to Propalaeosinopa is correct, it represents the
oldest record of the genus and of the family
Fantolestidae, a somewhat aberrant group of
enigmatic affinities. The new occurrence is
estimated to be late Puercan (Pu3) in age; the
genus and family are otherwise first known
from the late Torrejonian (To3; Archibald et al.

1987). In this context, we note that several
morphological details show the North Horn
taxon to be distinct, at the species level at
least, from described species; when better
known, it may prove to be generically separable.

Family Pentacodontidae

(Simpson, 1937) Van Valen, 1967

ApJironorus simpsoni Gazin, 1938

Figs. 2C-G

Newly referred material. — OMNH
27667, right dentaiy fragment with P3^ (P3L
= 2.4, W = 1.4; P4L = 4.0, W = 2.5) and asso-
ciated left P^ (L = 3.3, W = 4.1).

Locality and horizon. — OMNH V799,
"Dragon" locality (locality 2 of Gazin 1941: p. 7,
fig. 1), Dragon local fauna, early Torrejonian
(early or middle Paleocene). Joes Valley Mem-
ber, North Horn Formation, Emery Gounty,
UT

Description and discussion. — OMNH
27667 differs from the type of A. simpsoni
(USNM 15539) in minor ways but is clearly
referable to the species. P4 is slightly larger
than in the type and differs in having a weaker
anterior cingulum, which is barely indicated
on the anterolingual part of the tooth and is
completely absent labial to the keel extending
down the anterior face of the protoconid. The
minute ridge that extends down the posterior
wall of the metaconid (to meet with the cristid
obliqua) is lacking; however, the development
of this ridge in the type may be due partly to
the advanced wear in that specimen. P4 of
OMNH 27667 bears a small but distinct ento-
conid; this region of the tooth is broken in
USNM 15539. The anterior end of P4 in OMNH
27667 is slightly more developed downward
than in USNM 15539, vaguely recalling the
more advanced condition seen in Pentacodon
(Simpson 1937a: 124). Unlike either species of
Pentacodon, however, the P4 lacks a basal para-
conid, the protoconid is not as inclined poste-
riorly from base to apex, and the talonid is bet-
ter developed.

P3 has not been previously figured or de-
scribed for Aphronorus simpsoni, though this
tooth is known for A. fraudator (illustrated in
outline by Simpson 1937a, Gazin 1941). P3 of
OMNH 27667 is more anteroposteriorly elon-
gate than in A. fraudator. The tooth is distinct-
ly two-rooted and is much smaller than P4;
maximum width occurs just posterior to the

308

Great Basin Naturalist

[Volume 55

protocoaid. A small talonid basin is developed,
with a minute hypoconid and a "cristid oblicjua
connected to a ridge running down the poste-
rior flank of the protoconid. A small, short ridge
and swelling on the posterolingual (lank of the
protoconid are suggestive of a metaconid. A
faint cingulum is present anterolingually.

No associated upper teeth have been previ-
ously described for Aphronorus siinpsoiii, al-
though a few isolated specimens may belong
to the species (Gazin 1941, Robison 1986). P^*
of OMNH 27667 is broken near the paraconule
and at the lingual edge of the tooth, between
the cingulum at the base of the protocone and
the lingual root; the labial side of the meta-
conid is also damaged. Three roots are pres-
ent. The tooth, although similar to P"* of A.
fraudator, differs in several respects. The para-
style is absent; a small paraconule is present; a
metaconule as such is lacking, although there
is a vague swelling of enamel in this position.
The basal protoconal cingula show no tendency
to develop cuspules, as they do in A. frauda-
tor, and the metacone is much smaller in size,
relative to the paracone, than in that species.
The labial cingulum of p4 in OMNH 27667 is
also less developed than in A. fraudator

Aphronorus situpsoni was diagnosed as dis-
tinct from the comparatively well-known A.
fraudator mainly on the basis of differences in
proportions of F_^ and the lower molars (Gazin
1941). OMNH 27667, which includes teeth pre-
viously unreported for A. simpsoni, shows that
it is further distinct in having a somewhat more
elongate P3; P4 has a narrower, smaller-basined
talonid. P'^ differs from that of A. fraudator in
several respects, including the lack of a meta-
conule and parastyle, and the much lesser de-
velopment of the metacone. Considering the
specializations of the posterior premolars in
pentacodontids (Simpson 1937a) and the pos-
sibility that they represent a relatively archaic
group (Van Valen 1967), it is difficult to judge
which conditions are apomorphous, although
some of the states possessed by A. simpsoni (e.g.,
smaller P4 talonid; P^ with small metacone
and no metaconule) would appear — by com-
paiison to more primitive Eutheria — to be prim-
itive. The Tiffanian species A. orieli, known by
remarkably complete specimens (Gingerich et
al. 1983), appears to be considerably more
advanced, with greatly expanded crushing sur-
faces (particularly the protocone) on P"*.

Order Cond\ larthra

Family PArctocyonidae

(Giebel, 1855) Murray, 1866

Desniatoclaenus hennaeus Gazin, 1941

Fig. 3A

Newly referred material. — OMNH
27682, associated skull and jaw fragments with
broken right and left P^ (right P^L = 6.5), right
Ml-3 (MlL = 7.3, W = 8.6; M^L = 7.3, W =
11.0; M^L = 6.2, W = 8.7), left M2-3 (M^ bro-
ken, L = 7.4; M^L = 6.0, W = 8.8), left M^ (L
= 8.8, WTri = 7.2, WTal = 7.4), talonid of
right Mo (W = 6.4), trigonid of left M2 (W =
6.0), and talonid of right M3 (W = 5.2)."

Locality and horizon. — OMNH V800,
"Wagon Road ' locality (Gazin 1941, Robison
1986); Wagon Road local fauna, late Puercan
(early Paleocene). Joes Valley Member, North
Horn Formation, Emery County, UT.

Description and discussion. — P-* has dis-
tinct conules, with the paraconule being taller
than the metaconule. These cusps have not
previously been noted for P"^ of the species,
perhaps because of wear on the type specimen
(USNM 16202; see Gazin 1941:' fig. 19; West
1976: fig. 2). The upper molars have a labial
cingulum that is continuous. Interruption of the
ectocingulum at the base of the paracone was
cited as a generic character o{ Desmatodaenus .
However, the cingulum is complete in other
specimens, such as BYU 3800 (Robison 1986:
pi. 2, fig. 10), and we regard this as a feature
that is intraspecifically variable. M'^ bears a
small but distinct cingular hypocone, another
character that is apparently variable in the
species (Gazin 1941: figs. 19, 20; Robison 1986).
The only variation worthy of note in the lower
dentition of OMNH 27682 is the hypoconulid
of M3, which apparently projected posteriorly
as a distinct lobe, unlike the condition seen in
USNM 16202 (Gazin 1941: fig. 19).

Gazin (1941) originally described tvvo species
of Desniatoclaenus, D. hennaeus and D. para-
creodus, both fiom the Wagon Road fiiuna. West
(1976) synouNinized the two, a view apparent-
ly shared by Tomida and Butler (1980), but
Robison (1986) recognized them as distinct
and reported additional materials of both
species from other localities. In the original
diagnosis (Gazin 1941), D. paracreodus was
said to be larger than D. hermaeus, with the
lingual portion of upper molars more inflated
and widi a relatively larger M'^, bearing a better-

Paleocene Mammals, Utah

309

5 mm

t 1 ^ ■ ■ I

Fig. 3. ?Arctoc\oniclae and Periptychinae fioni the North Horn Formation. A, right P"*-M^ of Desmatoclaenm her-
maeiis (OMNH 27682) from the North Horn Formation; base of M- restored from contralateral tooth of same specimen,
and maxilla eliminated to improve clarity-; B, left dP^-^ and Ml of Ectoconus ditrigonus (OMNH 28111) in occlusal view;
maxilla eliminated to improve clarit)'.

developed hypocone. As shown by West (1976),
these differences in size and morphology are
both minor and inconsistent. In this context,
we note that M^-^ of OMNH 27682 are rela-
tively small (a supposed character of D. her-
maeus), yet M-^ is proportionately large, with a
well-developed hypocone (characters cited for
D. paracreodiis). We follow West (1976) in
regarding the species as synonymous.

In the original diagnosis and discussion of
Desmatoclaenus, Gazin (1941) compared the

genus with Tetraclaenodon and Frotogonodon,
as the latter taxon was tlien conceived (Matdiew
1937, Simpson 1937a). Van Valen (1978) placed
"?F. " protogonioides (cf. Matthew 1937) —
originally referred (Cope 1882a), in part, to
the genus Mioclaenus — in Desmatoclaenus ,
adding to the genus two additional species, D.
diaiiae and D. mearae; Frotogonodon was syn-
onymized with Loxolophus. We are in agree-
ment with these assignments; D. protogo-
nioides is relatively well represented and adds

310

Great Basin Natur.\list

[Volume 55

significantly to knowledge of the genus. Tlius
recognized, Desniatoclaenus is distinct from
Loxolophus in having stronger protocones on
P*^^; better-developed, more lingually placed
hxpocone on \l'~-, with hypocone occasionally
distinct on M'^; and paraconid of lower molars
placed more posterolingually and closely
appressed to the metaconid. Desmatoclaenus
differs from Tetraclaenodon in having less
molarized premolars (a metacone is lacking on
P'^""^; the trigonid is poorly developed and a
talonid basin is lacking on P4), upper molars
lacking mesostyle and with lesser develop-
ment of the hypocone; and lower molars with
more distinct, anteriorly placed paraconid.

Gazin (1941) considered Desmatoclaenus to
be stmcturally inteimediate between the archa-
ic ungulate "Protof^onodou' (then considered a
creodont) and Tetraclaenodon, a primitive
phenacodontid; the differential comparisons
presented above uphold this view. Subsequent
workers have referred Desmatoclaenus to the
Ai^ctocyonidae on the one hand (Van Valen 1978,
Cifelli 1983) or the Phenacodontidae on the
other (Simpson 1945, West 1976, Robison
1986). The positioning of the upper molar h)TDO-
cone somewhat more lingually in Desmatoclae-
nus than in Loxolophus is vaguely reminiscent
of the presumably derived condition in the
Phenacodontidae; similarly, the low, bunodont
cusps bearing mainly flat, apical wear are sim-
ilar to conditions generally obtained in mem-
bers of that family. Desmatoclaenus may well
be a transitional taxon, but in the absence of
compelling evidence in the form of synapo-
morphies, we here tentatively retain it in the
Arctocyonidae. In this context, we note that the
referred species D. protogonioides apparently
has a reduced anterior dentition, a condition
shared with loxolophine arctocyonids (Cifelli
1983).

Family Periptychidae Cope, 1882
Anisonchus ?oligistus Gazin, 1941

Fig. 4A

Newly referred material. — OMNH
27679, right M'l

Description. — OMNH 27679 is missing tlie

lingual base of the crown and enamel from the
posterior margin of the tooth; its estimated L
is 3.1. This specimen is appropriate in size for
only two of the four species of Anisonchus re-
ported from the North Horn Formation (Gazin
1941, Robison 1986); OMNH 27679 differs
from M'^ referred to A. athelae (including A.
eowijnae; Robison 1986) and is tentatively
referred to A. oligistus, for which M"^ was not
previously known. Although the tooth is incom-
plete and worn, it can be seen that the anter-
ocingulum was relatively weak and lacked a
pericone. Similarly, the hypocone was weak in
comparison to the condition in A. athelae,
being more similar to the larger A. dracus in
this respect. The pattern of wear suggests that
both paraconule and metaconule were present,
placed near the base of paracone and meta-
cone, respectively.

Haploconus elachistus Gazin, 1941
Figs. 4B-F

Newly referred material. — OMNH
27670, fragments of mandible with left M^_2
(MiL = 3.8, WTri = 2.7, WTal = 2.8; M2L =
3.9, WTri = 3.2, WTal = 2.9) and right M2 (L
= 4.0, WTri = 3.1, WTal = 3.0); 27713, frag-
ments of left mandible with P3 (L = 4.5, W =
2.8) and a heavily encrusted molar; OMNH
27680, right P4 (L = 4.5, W = 3.3).

Description and discussion. — Available
lower premolars (OMNH 27680, 27713) are
too worn to detennine whedier a paraconid was
present; Gazin (1941) reported the presence of
this cusp on P3 but not P4 of Haploconus ela-
chistus. The protoconid is a large, inflated cusp,
particularly on P4. A talonid crescent extends
posteriorly from the lingual base of the proto-
conid, curving labially at the posterior margin
of both P3 and P4. The metaconid of lower
molars is nearly as tall as the protoconid and is
transversely aligned with that cusp; a weak
paracristid descends anterolingually from the
protoconid, teniiinating in a small paraconid,
which lies in a median position. As described
b\' Gazin (1941), the pre-entocristid is taller
than the cristid oblicjua. The entoconid forms
a distinct pillar and projects somewhat on the

1995]

Paleocene Mammals, Utah

311

Fig. 4. Anisonchinae fiom the North Horn Formation. A, Anisonclut.s 'fuli^istus (UMNH 27679, right M^ in occkisal
view). B-F Haploconus elachistus: B, left Mi_2 (OMNH 27670) in occlusal view; C, E, left P3 (OMNH 27713) in occlusal
and labial views, respectively; D, F, right P4 (OMNH 27680) in occlusal and labial views, respectively. Scale bar repre-
sents 2 mm; tooth roots and jaw fragments have been eliminated to improve clarit)'.

312

Gril\t Basin Naturalist

[Volume 55

lingual side of the tooth; the in poeonulid
forms a fingerlike projeetiou at the baek of the
tooth and is somewhat lingual in position, an
appearance emphasized in later wear stages.

Two species of Haploconus, H. angustus
and the larger H. coniictilaiiis, are recognized
from the Torrejonian (To2; Archibald et al.
1987) of the San Juan Basin, NM (Matthew
1937). The apparent last record of Haploconm
is represented by a single molar, of uncertain
specific affinities, from Swain Quarry (To3?;
Archibald et al. 1987), WY (Rigby 1980). The
genus is othenvise known only from the North
Horn Formation. Gazin (1939) described H.
inopinatiLS fiom the Dragon fauna, later adding
a second species, ?//. elachistm, fi-om die Wagon
Road (Gazin 1941). More recently, Robison
(1986) has reported specimens of Haploconus
sp. from the Gas Tank local fauna; these mate-
rials are of interest in documenting the first
appearance of the genus, but unfortunately
they are not specificalh' diagnostic. H. inopina-
tiis, of Tol age, is similar in size to the later H.
angiistus but differs from that species in pro-
portions of the upper molars (Gazin 1939). H.
elachistus, the geologically oldest described
species, is smaller than the Tonejonian species
and, as noted by Gazin (1941), differs from
them in a number of respects. In the lower
dentition, P3_4 are less inflated than in H.
angmtus. Similarh, the trigonids of lower molars
in H. elachistus lack the inflated appearance
seen in Torrejonian species; a small paraconid
is still present, whereas in remaining species the
paracristid forms a bladelike surface extending
anteriorly from the protoconid and bears no
cusp. Lower molars of H. elachistus also lack
the crenulated or striated enamel and promi-
nent labial cingulum seen in other species. As
might be expected, the geologically older H.
elachistus appears to be more primitive than
the Torrejonian species for the characters
cited. In this context the apparent presence of
a more derived species in the Gas Tank local
fauna (Robison 1986) is somewhat surprising.

Haploconus is distinctive in the extreme
modification of lower molar trigonids (with
reduction to loss of the paraconid) and in the
unusual configuration of the talonid in posteri-
or lower premolars (with a lingual rather than
labial crescent), characters that are both ex-
pressed in H. elachistus. The affinities of the
genus are puzzling; Gazin (1941), noting the
primitiveness of some features of H. elachistus.

considered the species to be transitional be-
tween Conacodon and more derived species of
Haploconus. In retaining unreduced lower
molar trigonids and relatively unspecialized
lower premolars, species of Conacodon are
primitive with respect to Haploconus. In terms
of characters that are probably derived within
tlie context of Condylarthra, Conacodon, Haplo-
conus, and Oxyacodon have a lingually placed
hypoconulid and hypertrophied postmeta-
cristid on lower molars, lingually placed hypo-
cone on upper molars, loss of protocone on P^,
and, possibly, a columnar, lingually placed
entoconid on lower molars (not clearly seen in
all species of Oxyacodon). However, the exclu-
siveness of these characters and their potential
status as synapomoiphies remain to be estab-
lished. Archibald, Schoch, and Rigby (1983)
have shown that Conacodon and Oxyacodon
represent a distinctive subfamily, Conacodon-
tinae, whose relationship to other periptychids
is unclear; further investigation of the position
of Haploconus with respect to this clade is
clearly warranted.

Ectoconus ditrigonus (Gope, 1882)
Fig. 3B

Newly referred material. — OMNH
28111, fragment of left maxilla with dP'^^ and
Ml (dp3L = 7.5, W = 7.0; dP-^L = 7.5, W =
8.4; MlL = 9.6, W = 13.5).

Locality and horizon. — OMNH V829,
probably the same as Robison's (1986) Ferron
Mountain localit)'; Gas Tank local fauna, middle
Puercan (early Paleocene). Joes Valley Member,
North Horn Formation, Emeiy Gounty, UT.

Description and discussion. — The decid-
uous teeth, dP'^~^, are markedly smaller than
M^; both have conspicuous parast\'lar and meta-
stylar lobes. The third deciduous premolar has
a roughly triangular occlusal profile and is
longer than it is wide. The paracone and meta-
cone are sube(]ual in height; a large parastyle
is present almost directly anterior to the para-
cone. A prominent ridge extends lingually
from the parastyle to the protocone, which is
nearly as tall as the paracone and metacone;
another ridge descends the labial slope of the
parastyle, continuing posteriorly as a weak ecto-
cingulum. Labial to the metacone, the stylar
shelf broadens; a small cusp, serially analo-
gous (if not homologous) to a similar cusp on
upper molars of Ectoconus ditrigonus (Osborn

1995]

Paleocene Mammals, Utah

313

and Earle 1895), is present labial to the meta-
cone. A salient postmetacrista descends pos-
terolabially from the apex of the metacone,
extending to the posterolabial corner of the
tooth. Weak paraconule and metaconule are
present on the pre- and postprotocrista,
respectively. Faint pre- and postcingulae are
present on the lingual slopes of the protocone.
The fourth deciduous premolar is more molar-
iform than dP'^, differing from M^ in having
smaller conules and associated cristae, and in
the lesser development of the protocone
region. The parastyle of dP^ is more labially
placed than on dP'^, and the ectocingulum and
cingular cusp better developed than on that
tooth; a small mesostyle is also present. The
lingual cingulae are strong; pericone and hypo-
cone are present. M^ is typical o{ Ectoconus
and complete description is unnecessaiy The
ectocingulum is strong and bears both a meso-
st)'le and posterior stvlar cusp. The latter is sub-
conical and is connected to the base of the
metacone by a low ridge. Paracone, metacone,
and protocone are subequal in height; conules
are strongly developed and are only slightly
lower than the principal cusps.

Ectoconus ditrigonus, the type species, was
first described on the basis of material from the
San Juan Basin, NM (Cope 1882b). Matthew
(1937) reported a second species from the San
Juan Basin, E. majusculus, considered by Van
Valen (1978) to be synonymous with E. ditri-
gonus. The genus is known from several locali-
ties, including both Pu2 and Pu3 horizons, in
that area (Archibald et al. 1987). Gazin (1941)
described the species E. sijmbolus from the
Wagon Road (?Pu3) fauna. North Horn Forma-
tion. Robison (1986) described additional mate-
rials of E. sijmbolus from localities of the Gas
Tank fauna, thereby extending the range of the
species to ?Pu2, and reported E. ditrigonus
from two Gas Tank localities. OMNH 28111 can
be referred to the latter species on the basis of
size (larger than E. symbohis) and the pres-
ence of a relatively small posterior cusp, con-
nected to the base of the metacone b\' a low
ridge, on the ectocingulum of M^ (Robison
1986).

Deciduous teeth of archaic ungulates have
not been widely described or illustrated, a
notable exception being the deciduous premo-
lars of Phenacodontidae (West 1971). To our
knowledge, deciduous teeth of Periptychidae
have not been previously described, so that

there is no basis for comparison with dP'^~l of
Ectoconus ditrigonus .

Acknowledgments

We are especially grateful to Dale Harber
for the cooperation of the U. S. Forest Service.
We thank Jon Judd, Monte Swasey, and Scott
Madsen for help in the field; Dr Scott Russell,
Noble Electron Microscopy Laboratory, for
access to equipment and facilities; and Estelle
Miller for preparing the SEM photographs.
Drs. David W. Krause, J. David Archibald, and
Jeffrey G. Eaton provided invaluable com-
ments that improved the manuscript. Field-
work was supported by grant number 5021-93
from the National Geographic Society.

Literature Cited

Archibald, J. D.. E D. Gingerich, E. H. Lindsay, W. A.
Clemens, Jr., D. W. Krause, and K. D. Rose. 1987.
First North American land mammal ages of the
Cenozoic Era. Pages 24-76 in M. O. Woodburne,
editor, Cenozoic mammals of North America: geo-
chronolog\' and biostratigraphy. University of Cali-
fornia Press, Berkeley.

Archibald, J. D., J. K. Rigbv, Jr., and S. E Robison. 1983.
Systematic revision of Oxijacodon (Condylarthra,
Peript\'chidae) and a description of O. ferronensis n.
sp. Journal of Paleontology 57: 5.3-72.

Archibald, J. D., R. M. Schoch, and J. K. Rigby, Jr.
1983. A new subfamily, Conacodontinae, and a new
species, Conacodon kohlbergeri, of the Peript)chidae
(Condylarthra, Mammalia). Postilla 191: 1-24.

Butler, R. E, and E. H. Lindsay. 1985. Mineralogy of
magnetic minerals and revised magnetic polarity
stratigraphy of continental sediments, San Juan Basin,
New Mexico. Journal of Geology 94: 53.5-.554.

ClFELLi, R. L. 1983. The origin and affinities of the South
American Condylarthra and earh' Tertiaiy Litoptema
(Mammalia). American Museum Novitates 2772:
1-49.

Cope, E. D. 1882a. Some new forms from the Puerco
Eocene. American Naturalist 16: 83.3-8.34.

. 1882b. Synopsis of the Vertebrata of the Puerco

Eocene epoch. Proceedings of the American Philo-
sophical Society' 20: 461-471.

Gazin, C. L. 1938. A Paleocene mammalian fauna from
central Utah. Journal of the Washington Academv' of
Science 28: 271-277.

. 1939. A further contribution to the Dragon Paleo-
cene fauna of central Utah. Journal of the \^ashington
Academy of Science 29: 273-286.
. 1941. The mammalian faunas of the Paleocene of

central L'tah, with notes on the geolog)'. Proceedings
of the Lhiited States National Museum 91: 1-53.
Gingerich, R D., P Houde, and D. W. Kr.4USE. 1983. A
new earliest Tiffanian (late Paleocene) mammalian
fauna fiom Bangtail Plateau, western Craz\- Mountain
Basin, Montana. Journal of Paleontology' 57: 957-970.

314

Great Basin Naturalist

[Volume 55

Kr\use, D. VV., ano E D. C;i\(;khi(;ii. 1983. Mammalian
fauna from Douglass Quany, earliest Tiffanian (late
Paleocene) of the eastern Crazy Mountain Basin,
Montana. Contributions from the Museum of Paleon-
tology, University of Miehigan 26; 157-196.

M.vnilEW, W. D. 1937. Paleoeene faunas of the San Juan
Basin, New Mexico. Transactions of the American
Philosophical Societ>', new series 30: 1-510.

OsBORN, H. K, AND C. Earle. 1895. Fossil mammals of
the Puerco beds. Collection of 1892. Bulletin of the
American Museum of Natural Ilistoiy 7: 1-70.

RiGBV, J. K., JH. 1980. Swain Quarr\' of the Fort Union
Formation, middle Paleocene (Torrejonian), Carbon
Count)', Wyoming; geologic setting and mammalian
fauna. Evolutionar\' Monograph 3. 178 pp.

RoBiSON, S. F 1986. Paleocene (Puercan-Torrejonian)
mammalian faunas of the North Horn Formation,
central Utah. Brigham Young University Geology
Studies 33; 87-133.

Rose, K. D. 1981. The Clarkforkian land-mammal age and
mammalian faimal composition across the Paleocene-
Eocene boundarv'. University of Michigan Papers on
Paleontology 26; 1-196.

Simpson, G. G. 1936. A new fauna from the Fort Union of
Montana. American Museum Novitates 873; 1-27.

. 1937a. The Fort Union of the Crazy ^Mountain

Field, Montana, and its mammalian faunas. Bulletin
of the United States National Museum 169; 1-2S7.

. 1937b. Additions to the upper Paleocene fauna of

the Crazy Mountain Field. American Museum
Novitates 940; 1-15.

. 1945. The principles of classification and a classi-
fication of mammals. Bulletin of the American
Museum of Natural Histoi-v 85; 1-350.

Spieker, E. M. 1960. The Cretaceous-Tertiaiy boundary
in Utah. 21st International Geological Congress,
Copenhagen 5; 14-24.

ToMiDA, Y. 1981. "Dragonian" fossils from the San Juan
Basin and status of the "Dragonian" land mammal
"age." Pages 222-241 in S. G. Lucas, J. K. Rigby, Jr.,
and B. S. Kues, editors. Advances in San Juan Basin
paleontology. University of New Mexico Press,
Albuquerque.

. 1982. A new genus of picrodontid primate from

the Paleocene of Utah. Folia Primatologica 37; 37—43.

TOMIDA, Y. AND R. F Butler. 1980. Dragonian mammals
and Paleocene magnetic polarit\' stratigraphy of the
North Horn Formation, central Utah. American
Journal of Science 280; 787-811.

Van Valen, L. 1967. New Paleocene insectivores and
insectivore classification. Bulletin of the American
Museum of Natural Histoid 135; 217-284.

. 1978. The beginning of the Age of Mammals.

Evolutionary Theory 4; 45-80.

West, R. M. 1971. Deciduous dentition of the early
Tertian' Phenacodontidae (Condylarthra, Mammalia).
American Museum Novitates 2461; 1-37.

. 1976. The North American Phenacodontidae

(Mammalia, Condylarthra). Contributions to Bio-
logical Geology, Milwaukee Public Museum 6; 1-78.

Wood, H. E., II, R. W. Chaney. J. Cl.\rk, E. H. Colbert,
G. L. Jepsen, J. B. Reeside, Jr., and C. Stock.
1941. Nomenclature and correlation of the North
American continental Tertiaiy Bulletin of the Geolog-
ical Society' of America 52: 1-48.

Received 6 May 1994
Accepted 12 December 1994

SPRINGTIME MOVEMENTS, ROOST USE, AND FORAGING

ACTIVITY OF TOWNSEND'S BIG-EARED BAT {PLECOTUS

TOWNSENDII) IN GENTRAL OREGON

David S. Dolikinl, Ronald D. Gettinger^, and Michael G. Gerdes^

Abstract. — Seasonal movements, roost-site fidelity, and foraging activity patterns are largely unknown for western
populations of Townsend's big-eared hat [Plecotus toicnsendii). We used miniature radiotelemetry units to track spring-
time movements of si.x bats inhabiting forested lava flows in central Oregon, and found that bats moved up to 24 km
from hibemacula to foraging areas. Individual bats returned to the same foraging area on successive nights but shifted to
different areas in presumed response to changes in insect availabilit>-. Both se.xes apparently use a series of interim roost
sites between emergence from hibernation and the time females enter into maternitv' colonies, with little individual
fidelity to these sites. In regions characterized by extensive lava-flow topography, suitable daytime roosts are numerous
and dispersed over a large area, allowing bats to move relatively great distances to locate foraging ranges. Hence, the
actual area of concern for effective management of individual populations can be considerably larger than indicated
solely by locations of hibemacula and maternity caves of this declining species.

Key icords: Toicnsend s big-eared bat. Plecotus townsendii,/orag»ig movements, roost sites, roost fidelity, hibemacula,
caves, central Oregon, radiotelemetry, lavaflous, candidate species.

Townsend's big-eared bat {Plecotus town-
sendii) is distributed over much of western
North America (Hall 1981), although popula-
tions may be widely scattered within its range.
The species appears to be a habitat generalist,
reportedly inhabiting coniferous forests in nortli-
ern New Mexico (Jones 1965), mixed meso-
phytic forests in Kentucky (Adam et al. 1994),
deserts in Arizona (Hoffmeister 1970), native
prairie in Kansas and Oklalioma (Humphrey and
Kunz 1976), riparian communities in north-
eastern Montana (Swenson and Shanks 1979),
Kansas, and Oklahoma (Humphrey and Kunz
1976), and agricultural areas and coastal regions
in California and Washington (Dahlquest 1947,
1948, Pearson et al. 1952). In Oregon the dis-
tribution of Townsend s big-eared bat is dis-
continuous and highly local across forest and
shrubsteppe habitats throughout the state
(Perkins and Levesque 1987).

Two disjunct subspecies occur in eastern
North America, both of which are listed as en-
dangered under the U.S. Endangered Species
Act. Kunz and Martin (1982) suggested that
western populations also are vulnerable, espe-
cially to disturbance in winter hibemacula and
summer maternity caves. Both subspecies

found in the western United States are declin-
ing markedly (Perkins and Levesque 1987,
Pierson et al. 1991), and the species is listed as
endangered, sensitive, or of special concern
by several western states and federal land
management agencies.

Big-eared bats feed almost exclusively on
Lepidoptera (Ross 1967, Whitaker et al. 1977,
1981, Dalton et al. 1986, Sample and Whitmore
1993) and are viewed as moth specialists (Dalton
et al. 1986, Sample and Whitmore 1993).
Probably most limiting to their distribution,
however, is availability of suitable sites for
roosting, hibernation, and reproduction, which
consist primarily of caves and abandoned
mines. These three activities require different
microclimatic conditions (Dahlquest 1947, Pear-
son et al. 1952, Twente 1955, Barbour and Davis
1969, Martin and Hawks 1972, Humphrey and
Kunz 1976, Marcot 1984, Center 1986, Perkins
and Levesque 1987, Pierson 1989, Pierson et
al. 1991, Lacki et al. 1993, Clark et al. 1995).
Any single site generally is unsuitable for more
than one function, although microclimates in
different regions of the same cave sometimes
differ sufficiently to accommodate more than
one activity (e.g., Clark et al. 1995).

'High Desert Ecological Research Institute, 15 S.W. Colorado Avenue, Suite 300, Bend, OR 97702.

^Biologs' Department, Randolph-Macon Woman's College, Lynchburg, VA 24.503.

■'Deschutes National Forest, United States Forest Service, 1645 Highway 20 East, Bend, OR 97701.

315

316

Great Basin Naturalist

[Vokime 55

Big-eaix'd bats are colonial tor most of the
year, but colony dynamics and seasonal move-
ments have not been studied in the Inter-
mountain West. In central Oregon, P. town-
sendii undergoes arousal from hibernation and
movement from hibernacula in April, although
the precise timing of these events appears to
vary with weather conditions and topography
(U.S. Forest Service, Deschutes National Forest,
inipublished data). Females form maternity
colonies in late spring or early sunnner (USFS
unpublished data), but the timing of their
amval at mateniit)' roosts is poorly documented,
and it is unclear whether they move immedi-
ately to maternity roosts upon departure from
hibernacula.

A much better understanding of seasonal
movements among roost sites is necessaiy for
effective management of populations. Although
it is clear that traditional site use (sensu Dobkin
et al. 1986) occurs for specific hibernacula and
matemit)' roosts, the extent of roost site fidelity'
by individual bats is unknown. Recent teleme-
try studies have been conducted for both
endangered subspecies found in eastern North
America (Clark et al. 1993, Adam et al. 1994,
Lacki et A. 1994), lout no telemetiy studies have
examined the movements of western sub-
species. Our primaiy objective was to acquire
information concerning the extent of move-
ments by individual Townsend's big-eared
bats during the period following arousal from
hibernation in an area containing a significant
proportion of Oregon's known population.

Study Area and Methods

Fieldwork was conducted in Deschutes
County on the Fort Rock Ranger District of
the Deschutes National Forest and adjacent
lands administered by the Bureau of Land
Management. The primaiy study area (Fig. 1)
consists of a NW-SE-oriented basin contain-
ing extensive forested lava flows, and the sur-
rounding buttes from 44° 25' to 43° 37' N, and
121° 15' to 120° 48' W. Elevations range from
1400 m on the basin floor to nearly 2000 m
atop Pine Mountain. Forests are open stands
of ponderosa pine {Piniis ponderosa) with bitter-
brush {Purshia tridentata), manzanita {Arcto-
staphijlos spp.), and bunchgrass understories.
Scattered, relatively closed, stands of lodge-
pole pine {P. contorta) also occur throughout.
Areas adjacent to lava flows consist of shrub-

steppe habitat dominated by sagebrush
{Artemisia tridenlata).

The study area lies within the broad zone
of intergradation between the western interior
form {P. t. pallescens) and the coastal Pacific
form {P. t. townsendii) of Townsend's big-eared
bat (Handley 1959). We concur with Handley's
(1959: 199) assessment that "allocation . . .
from much of this area to one race or the other
is largely a matter of personal opinion. "

Based on USFS cave surveys conducted
from 1985 to 1991, two hibernacula (SI and
S2, Fig. 1) but no maternity caves were known
from the southern end of the basin. The north-
em end of the basin contained a series of hiber-
nacula and one maternity cave (N3, which was
gated), as well as one other cave (N2) that
reportedly was used as a maternity roost in the
past. The maternity cave and the northern-
most hibernaculum in the southern portion of
the basin are separated by 30 km, which
prompted the assumption that big-eared bats
in the basin consisted of two separate popula-
tions (J. M. Perkins, unpublished report to
USFS). Subsequent to completion of our field-
work, a previously unknown maternity cave was
discovered beyond the southern end of the
basin, 17 km southeast of SI.

Fieldwork in 1992 commenced on 7 April
and continued through 9 June. Six big-eared
bats (5 females, 1 male) were captured by hand
between 1100 and 1730 h from four different
caves in April and May (Table 1). Each bat was
fitted with a battery-powered (14-21 da>' bat-
tery longevity), miniature radiotransmitter
(0.6-0.7 g; Model BD-2B, Holohil Systems,
Ltd.) affixed to the dorsal, interscapular fur
(Dobkin et al. in press) with eyelash cement.
Transmitter units averaged 6% of bat body
mass (x = 10.6 g. Table 1), which should have
had minimal effect on maneuverability and
energy costs for this species (Davis and Cock-
rum 1964, Aldridge and Brigham 1988).

Bats carrying transmitters were tracked
with portable receivers (Telonics) equipped
with directional antennae (Wilkinson and Brad-
bury 1988). Bats were monitored for nearly
850 observer hours over the 64-day period
through a combination of daytime ground
searches and nighttime triangulations from
fi.xed locations. Two or three observers with
receivers were located on the tops of buttes
widely separated aroimd the basin (Fig. 1) to
provide the directional data necessary for

1995]

Radiotelemetry of Townsend's Big-eared Bats

317

:hina

HAT S2

• •

•
EAST

•

BUTTE

•

QUARTZ

MTN. •

FOX

BUTTE

Fig. 1. Map of the study area in central Oregon showing
locations of the four caves in whicli Townsend's big-eared
bats were captured and fitted with radiotransmitters (SI,
S2, Nl, and N2), and location of the onl\- known maternit\'
cave (N3) in the basin. Telemetered bats were monitored
from atop Pine and Quartz mountains, Coyote, East, Fox,
and China Hat buttes.

determining bat locations. The monitoring
protocol for fixed-point triangulation consisted
of scanning all active frequencies for the initial
five minutes of each quarter hour. If contact
was made, tlie other observers were notified by
radio and the bat was tracked continuously. In
addition, seven flights were made at night by
fixed-wing aircraft carrying a receiver and
wing-mounted antennae and equipped with a
LORAN system. LORAN fixes were integrated
with simultaneous directional information ob-
tained from ground-based receivers.

We conducted ground searches on foot and
from moving vehicles. Efforts were concen-
trated in the vicinity of caves known to be
used by bats, including caves in which teleme-
tered bats originally were captured. These
searches continued for 7-14 days following
attachment of transmitters. Due to rugged top-
ography and the distances between northern

and southern ends of the basin, only southern
caves were checked systematically following
tagging of the first three bats, all of which
were from the southern basin. Likewise, only
northern caves were checked systematically
following tagging of the last three bats, all of
which came from the northern basin. All caves
were checked as opportunity permitted, re-
sulting in essentially complete coverage of all
known cave sites in the basin at least weekly.

Results
Movements and Roost Site Fidelity

All marked females left their caves within
two nights of capture and neither returned to
these caves nor entered the known maternity
cave (N3) during the remainder of transmitter
battery life. Upon emergence from their
hibernacula, all three females from the south-
ern end of the basin moved 11-12.5 km north-
east to the western slopes of Pine Mountain
and did not return to the vicinity of their
hibernacula in the southern end of the basin.
Only female #579 was located subsequently,
again on the western slope of Pine Mountain.
Faint signals were received briefly from one of
these females on 3 May on a precise bearing
toward the then-unknown maternity cave
southeast of the study area.

The most extensive telemetry data were
collected for female #707, which left Nl on
the second night following capture. She was
located again five nights later and was tracked
for the following five nights (including a series
of LORAN fixes made from the air), and then
to a day roost located just east of the crest of
Pine Mountain, ca 20 km from Nl but only
2-4 km from where she had been foraging on
the preceding five nights. Although we do not
know whether she had used this roost previ-
ously, she was not found there subsequently.
This bat went undetected over the next three
nights and was then located for the last time
on the following night. All foraging locations
beyond the immediate vicinity of Nl were on
the western slope of Pine Mountain, 17-24 km
from Nl.

Foraging locations for the fifth female
(#728) were within 2-5 km of N2: southwest
of N2 on one night and northeast of N2 two
nights later This bat dropped her transmitter,
which we recovered 15 days after attachment,
at a location 5 km west of N2 and within 1 km

318

Grkat Basin Naturalist

[Volume 55

Table 1. Suninuin of radiotelemetrv' contacts witli Townsentl's big-eared hats earning transmitter nnits on the Deschutes
National Forest in central Oregon, 1992.

Max

. distance''

Bat#

)cation-'

(km)

Sex

Mass (g)

Dates of contact

558

11.0

17-19 April

568

11.2

17-19 April

579

12.0

19 April-2 May

707

10..3

28 April-10 May

728

10.6

12-25 Mayt

768

9.0

20-26 Ma>'

^Indicates location of cave where bat v\as captured and fitted witli radiotransniitter. To maintain site security, caves are designated by alphanumeric codes; S and N
indicate cave location in southern and northern portions of the stud> area, respectively. SI, S2, and Nl were winter hibernacula; N2 was an interim roost site.
''MiLKinium distance moved from cave of initial capture, as cletiTUiinetl by radio contact with foraging bat.
^Transmitter dropped from bat on 26 or 27 May and recovered on 27 May.

of unnamed caves known to have harbored
big-eared bats occasionally in the past (L.
Becker, unpublished USFS sui-vey data).

The single telemetered male (#768) for-
aged extensively in the immediate vicinitx' of
N2 upon evening emergence, then moved 6-8
km east to forage over Horse Ridge. This bat
was not located again until five nights later,
when he returned to N2, and was recorded over
the next two nights foraging in and around the
sinkhole immediately in front of N2. Although
male #768 returned to roost for two consecu-
tive days in the cave where originally captured,
he then left and did not return again prior to
the end of fieldwork 12 days later.

The habitat used for foraging consisted of
sagebrush shrubsteppe (western slopes of Pine
Mountain and Horse Ridge) and very open
ponderosa pine woodland with extensive bit-
terbrush and interspersed areas (<5 ha) of
sagebrush. Relatively little time appeared to
be spent foraging in more densely forested
areas.

Times of Activity

Big-eared bats emerged from their cave
roosts to forage shortly after sunset, with time
of emergence becoming later as day length
increased in the spring (Fig. 2). Although our
data are ver\' limited, an interpretable pattern
of activity can be seen in the May data. Big-
eared bats foraged in the immediate vicinity of
their cave roosts during the first few hours of
darkness, moved to areas farther from their
roost to forage (perhaps intermittently) from
around midnight to within an hour or two of

sunrise, and then once again returned to for-
age in the vicinity of their day roost.

Discussion

Continuous monitoring of movements and
activity in small, cave-dwelling bats like Town-
send s big-eared bat is constrained by (1) the
need to minimize load mass carried by an ani-
mal, which strongK' limits both strengdi of trans-
mitter signal output and batteiy longevity, and
(2) the difficulty of signal detection in land-
scapes of rugged, rock>' topograph}' and from
witliin caves. Despite diese limitations, a number
of salient points can be deduced fi'om our study.

Our data clearly indicated that female big-
eared bats in central Oregon did not move
directly from their winter hibernacula to mater-
nity caves, but instead utilized a series of interim
roost sites over a period of perhaps as much as
two months. The four females marked in April
were captured in winter hibernacula in the
company of other roosting conspecifics. In con-
trast, the two bats captured in May were the
only big-eared bats roosting in the cave on the
dates of capture; we assumed that neither of
these bats hibernated in N2 during the pre-
ceding winter, although we cannot exclude this
possibility. None of the four caves in which bats
were captured was used as a maternity cave.

We suggest that little fidelity to interim roost
sites occurs because neither of die two females
found in day roosts returned to these roosts on
subsequent days. In addition, male #768 left
his roost cave, returned five days later, re-
mained for two days, and then left again for at

1995]

Radiotelemetry of Townsend's Big-eared Bats

319

Bat I.D.
Date

558

4/19/92

568

4/19/92
579

4/19/92

5/2/92

707

4/28/92

5/3/92
5/4/92

5/5/92
5/6/92
5/7/92

5/10/92

728

5/14/92

5/16/92
5/25/92

768

5/20/92

5/25/92
5/26/92

LEGEND

< 0.2 km from cave
> 1 .0 km from cave

^^■H

V/////////M

Y////////////A

2300 2400

Time of Night (H)

0400

Fig. 2. Temporal distribution (Pacific Daylight Savings Time) of foraging activit)' by Townsend's big-eared bats in rela-
tion to distance from daytime roost sites in forested lava flows of central Oregon.

least the next 12 clays. With such small sample
sizes, we cannot say whether males and
females differ in their use of roosts during this
period or whether both sexes exhibit the same
pattern of periodic use. We believe that the
most reasonable interpretation of the data is
that both sexes opportunistically use interim
roost sites during this period, and that the
choice of roost area is most likely determined
by spatial and temporal variation in prey avail-
ability. Even species that exhibit strong indi-
vidual fidelity to day roosts and repeated use
of the same foraging areas on successive nights
(e.g., Euderma maculatum) shift both roost site
and foraging area seasonally (Wai-Ping and
Fenton 1989).

Big-eared bats in our study moved up to 24
km from hibernacula to foraging areas, al-
though our data suggest that distances moved
from interim day roosts to foraging areas are
typically 2-8 km during the period prior to
entry into maternity colonies. These shorter

moves between roosts and foraging areas are
consistent with research on eastern subspecies
of big-eared bats in which females foraged at
distances of 2-7 km from their roosts (Clark et
al. 1993, Adam et al. 1994). Repeated use of
the same foraging area on successive nights or
alternation among several sites appears to
characterize both eastern subspecies of big-
eared bats (Anonymous 1991, Clark et al. 1993,
Adam et al. 1994), as well as big-eared bats in
central Oregon (e.g., bat #707).

Although big-eared bat diets are composed
primarily of forest Lepidoptera, bats in eastern
Oklahoma foraged preferentially at the inter-
face between forested and open pasture habi-
tats (Clark et al. 1993). Nevertheless, bats ex-
tensively used open, forest, and edge habitats,
and significant shifts in relative habitat use
were recorded by Clark et al. (1993). Similarly,
in central Oregon we found that Townsend's
big-eared bats foraged primarily (but not ex-
clusively) in the more open habitats provided
by shrubsteppe and forest-shiTib ecotones.

320

Great Basin Natufl\li.st

[Volume 55

In our study, activih' patterns of hiu-eared
bats in spring most closely resembled patterns
documented for females of eastern subspecies
during late lactation and prior to parturition
(Clark et al. 1993, Lacki et al. 1994), i.e., por-
tions of the annual cycle when females are less
constrained in the amount of time they can
spend away from tlie maternity cave. Flight initi-
ation inside caves and subsequent emergence
documented by Clark et al. (1993) and by
Lacki et al. (1994) were identical to the pat-
terns exhibited in our study.

Primaiy determinants of habitat suitability
for Ozark big-eared bats are the availal)ilit\ of
an adequate food supply and appropriate roost
sites (Clark et al. 1995). Unlike areas where
big-eared bats are limited by a small number
of suitable roost sites, the extensive forested
lava flows found in the Pacific Northwest offer
numerous potential temporaiy roost sites that
enable individual bats to forage over a consid-
erable area by using a succession of roost sites
during the period following emergence from
their hibernaculum. Bats still are limited sea-
sonally, however, to a very small number of
sites that provide suitable microclimatic con-
ditions for hibernacula and maternity caves.
Such an inteqDretation of potential movement
patterns is consistent with our tracking data
and the loss of contact with telemetered bats
for successive days followed by subsequent con-
tact. Even our seven attempts to locate bats by
aircraft, which should have avoided problems
arising from topographic interference with
transmitter signals, succeeded only once, indi-
cating that bats may well have left the basin
entirely, as was apparently the case for at least
the one bat we detected southeast of the study
area in the vicinity of the previously unknown
maternity cave.

Populations of Townsend's big-eared bats
inhabiting regions with extensive lava flows
likely use many roost sites dispersed over large
areas. The extent of movements that we docu-
mented and the use of the same foraging areas
by bats from both ends of the basin make it
unlikely that bats from southern and northern
hibernacula represent separate populations. A
better understanding of movements among
seasonal and interim roost sites is urgently
needed for successful conservation of dwin-
dling populations. Our data demonstrate that
the actual area of concern for management of
individual populations is considerably greater

than indicated solely by locations of hibernac-
ula and maternity caves.

Acknowledgments

We tliank Bijaya Kattel and Jamie Haskins for
their invaluable field assistance; this stud\' could
not have been completed without their con-
siderable help. Lew Becker of the Deschutes
National Forest and Chris Carey of the Oregon
Department of Fish and Wildlife contributed
in many ways to the success of this project.
Helpful discussions with Brad Sample and
Bruce Wunder and review of earlier versions
of the manuscript by Brenda Clark and William
Clark improved the final manuscript. This
project was carried out in part with funding
provided by the United States Forest Sei-vice
under Contract No. 43-04GG-2-69020.

Literature Cited

Adam, M. D., M. J. Lacki, and T. G. Barnes. 1994. Forag-
ing areas and habitat use of the Virginia big-eared
bat in Kentucky Journal of WikUife Management 58:
462-469.

Aldridge. H. D. J. N., AND R. M. Brigham. 1988. Load
carrying and maneuverabihty in an insectivorous bat:
a test of the 5% "rule" of radio-telenietiy Journal of
Manmialogy 69: 379-382.

Anonymous. 1991. Endangered Species Technical Bulletin
16: 14.

Barbour. R. W., and W. H. Davis. 1969. Bats of America.
Universit)' Press, Lexington, O'.

CL.4RK, B. K., B. S. Clark, D. M. Leslie, Jr., and M. S.
Gregory. 1995. Characteristics of caves used by the
endangered Ozark big-eared bat. Wildlife Society
Bulletin; in press.

Clark, B. S., D. M. Leslie, Jr., and T. S. Carter. 1993.
Foraging activity of adult female Ozark big-eared
bats {Plecutus townsendii ingem) in summer. Journal
of Mammalogy 74: 422—427.

Dahlquest, W. W. 1947. Notes on the natural histoiy of
the bat Corynorhimis rafinesqiiii in California. Joimial
of Mammalogy 28: 17-30.

. 1948. Mamniiils of Washington. Universib.' of Kimsas

Publications of the Museum of Natural History 2:
1-444.

Dalton, V. M., V. Brack, Jr., and E M. McTeer. 1986.
Food habits of the big-eared bat, Plecotm townsendii
virginianii.s. in Virginia. N'irginia Journal of Science
37:248-254.

D.WIS, R., and E. L. Cockrum. 1964. E.xpcrimentally
determined weight lifting capacity' in indi\ iduals of
five species of western bats. Journal of Mammalogy
45: 64.3-644.

DoBKiN, D. S., J. A. Holmes, and B. A. Wilcox. 1986.
Traditional nest-site use by White-throated Swifts.
Condor 88: 252-253.

DoBKiN, D. S., B. K.\TTEL, AND R. D. GETriNGER. Com-
parative retention of radiotransmitters by fur-clipped

1995]

Radiotelemetry of Townsend's Big-eared Bats

321

and undipped Townsend's big-eared hats and pallid
bats. Bat Research News: in press.

Center, D. L. 1986. Wintering bats of the Upper Snake
River Plain, occurrence in lava-tube caves. Great
Basin Naturalist 46; 241-244.

Hall, E. R. 1981. The mammals of North America.
Volume 1. John Wiley & Sons, New York.

Handley, C. O., Jr. 1959. A revision of American bats of
the genera Eitderma and Plecotiis. Proceedings of
the United States National Museum 110: 95-246.

HOFFMEISTER, D. E 1970. The seasonal distribution of
bats in Arizona: a case for improving mammal range
maps. Southwestern Naturalist 15: 11-22.

Humphrey, S. R., and T. H. Kunz. 1976. Ecology of a
Pleistocene relict, the western big-eared bat {Plecotiis
townsendii), in the southern Great Plains. Journal of
Mammalogy 57: 470-494.

Jones, C. 1965. Ecological distribution and activit}' periods
of bats of the Mogollon Mountains area of New
Mexico and adjacent Arizona. Tulane Studies in
Zoology 12: 93-100.

Kunz, T. H., and R. A. Martln. 1982. Plecotus townsendii.
Mammalian Species 175: 1-6.

Lacki, M. J., M. D. Adam, and L. G. Shoemaker. 1993.
Characteristics of feeding roosts of Virginia big-
eared bats in Daniel Boone National Forest. Journal
of Wildlife Management 57: 539-543.

. 1994. Observations on seasonal cycle, population

patterns and roost selection in summer colonies of
Plecotus townsendii virginianiis in Kentucky American
Midland Naturalist 131: 34-42.

Marcot, B. C. 1984. Winter use of some northwestern
California caves by western big-eared bats and long-
eared Mijotis. MuiTclet 65: 46.

Martln, R. A., and B. C. Hawks. 1972. Hibernating bats
of the Black Hills of South Dakota. I. Distribution
and habitat selection. Proceedings of the New Jersey
Academy of Science 17: 24—30.

Pearson, O. R, M. R. Koford, and A. K. Pearson. 1952.
Reproduction of the lump-nosed bat [Conjnorhiniis
rafinesqiiii) in California. Journal of Mammalogy 33:
273-320.

Perkins, J. M., and C. Levesque. 1987. Distribution, sta-
tus, and habitat affinities of Townsend's big-eared
bat {Plecutits townsendii) in Oregon. Oregon Depart-
ment of Fish ik Wildlife Technical Report 86-5-01.

Pierson, E. D. 1989. Help for Townsend's big-eared bats
in California. Bats 7: 5-8.

Pierson, E. D., W. E. Rainey, and D. M. Koontz. 1991.
Bats and mines: experimental mitigation for Town-
send's big-eared bat at the McLaughlin Mine in
California. Pages 31-42 in Proceedings V: Issues and
technology in the management of impacted wildlife.
Thome Ecological Institute, Boulder, CO.

Ross, A. 1967. Ecological aspects of the food habits of in-
sectivorous bats. Proceedings of the Western Founda-
tion of Vertebrate Zoology 1: 205-264.

Sample, B. E., and R. C. Whitmore. 1993. Food habits of
the endangered Virginia big-eared bat in West
Virginia. Journal of Mammalogy 74: 428-435.

SwENSON, J. E., AND G. F Shanks, Jr. 1979. Noteworthy
records of bats from northeastern Montana. Journal
of Mammalogy 60: 650-652.

TwENTE, J. W, Jr. 1955. Some aspects of habitat selection
and other behavior of cavern-dwelling bats. Ecology
36: 706-732.

Wai-Ping, V, AND M. B. Fenton. 1989. Ecology of spot-
ted bat (Eiidenna maculatwn) roosting and foraging
behavior. Journal of Mammalogy 70: 617-622.

Whitaker, J. 6.. Jr., C. Maser, and S. R Cross. 1981.
Food habits of eastern Oregon bats, based on stomach
and scat analyses. Northwest Science 55: 281-292.

Whitaker, J. O., Jr., C. Maser, and L. E. Keller. 1977.
Food habits of bats of western Oregon. Northwest
Science 51: 46- 55.

Wilkinson, C. S., and J. W. Bradbury. 1988. Radio-
telemetiy: techniques and analysis. Pages 105-124 in
T. H. Kunz, editor. Ecological and behavioral methods
for the study of bats. Smithsonian Institution Press,
Washington, DC.

Received 5 January 1995
Accepted 28 April 1995

NAMES AND TYPES IN PERENNIAL ATK/PLEX LINNAEUS

(CHENOPODIACEAE) IN NORTH AMERICA SELECTIVELY

EXCLUSIVE OF MEXICO

Stanley L. VVelshl and Clifford Cromptoii-

Absthact. — Cited are names and eoniliinatioiis within the woody species of Atriplcx as the\' occur in North America.
Tv'pes and tlieir repositories are inchided ibr all ta.\a except those for which that information could not be located. New
nomcnclatural proposals include Atriplcx gardneri var aptera (A. Nelson) Welsh & Crompton, comb, nov.; A. garrettii
van navajoensis (C. A. Hanson) Welsh & C>rompton, comb, nov.; Atriplcx acanthocarpa vai. coahiiilcnsis (Henrickson)
Welsh & Crompton, comb. nov. A lectotvpe is designated for A. breweri S. Watson.

Key words: Chcnopodiaccac, Atriplcx types. North America.

This list of names and synonyms of peren-
nial and woody Atriplex taxa is preliminaiy to
the preparation of a taxonomic treatment for
the woody species of Atriplex as they occur in
North America, both indigenous and intro-
duced species. All names, whether treated as
taxa recognized by me or as mere synonyms,
are included. The taxonomic treatment that
will appear subsequently in the publication of
the Flora North America Project will distin-
guish between the names of taxa per se and
their included synonyms. The relatively large
number of names and synonyms for this small
group of plants is indicative of the changes in
generic concepts, the ever-changing inteip re-
lation of the status of a taxon, and the general
phenotypic plasticity of this amazing group of
shrubs, subshrubs, and perennial herbs, which
hybridize freely among themselves and some-
times with other taxa not apparently closely
allied. They grow on a surprising array of sub-
strates in the American West, from the cold
temperate of northern Alberta to the much
warmer climates of Mexico. Often they are
among the most important shrub species on
saline, fine-textured substrates, and some-
times they are the only shrubby inhabitants.
Their ability to survive and even thrive in
saline sites has placed them in a position of
importance for browsing animals where other
browse is scarce or lacking. They cover huge
areas where geomoiphological processes have
exposed raw, saline strata in vast expanses.

Niobrara Shale, Mancos Shale, Morrison
Formation, and numerous other geological for-
mations support these plants. Saline pans and
other poorly drained lowlands are occupied by
these species. Despite the affinity for saline
areas, where they have little competition
(except from other halophytes), some of the
species thrive where total soluble salts are low.
The four-wing saltbush, Atriplex canescens
(Pursh) Nuttall, is such a plant. It grows from
the edge of saline areas up gradient into far
less saline substrates, often in grasslands or in
shrublands dominated by sagebrush and other
shrubby species.

Hybridization is an important factor con-
tributing to the diversity of woody Atriplex
species. There are at least two main taxa
around which many of the remainder are
placed, and with which most form at least
occasional hybrids, i.e., A. canescens (Pursh)
Nuttall and A. confertifolia (Torrey & Fremont)
S. Watson. Some of the hybrids have received
names and formal taxonomic recognition.
Most of them are of occasional occurrence, and
some of the taxa treated in contemporaneous
taxonomic works are apparently of hybrid
derivation — now more or less stabilized as pop-
ulations, mainly on veiy peculiar substrates.

The following list is thought to be exhaus-
tive for woody and perennial Atriplex names
in North America, especially for those north of
Mexico. A few taxa represented in Mexico are
included where thev roimd out the names for

HMc ScicDcc Miiscuin and Department (i( Botaiu' and Kan);e Science, Brighani Young Universit>', Piovo, UT 84602.
^Biosysteniatics Research Centre, W'ni. Saunders Bnilding. C;. E. F. Ottawa, Ontario KIA PC6, Canada.

322

1995]

North American Perennial Atr/plex Types

323

species complexes largely confined north of
that country. Pertinent types have been received
on loan by the gracious kindness of curators of
the herbaria cited with the specimens. Abbre-
viations for the herbaria are those standard
ones cited in Index Herbariorum, except that
the origin of the collection is indicated by use
of such designations as NY Torrey for historic
specimens. This is thought to be important
because it authenticates the antiquity of the
specimen and might prove important in cer-
tain cases in judging whether a particular
author had access to a given specimen.

The type information is presented below in
dual format for some taxa, with the type locali-
ty or collector information (herein arbitrarily
designated "Type locality") as recorded with
the protologue cited first and with the label
data of the type specimen (herein designated
"Type") cited second where there is a substan-
tial difference in the two accounts.

Atriplex acanthocarpa (Torrey) S. Watson, Proc. Amer.
Acad. Arts 9: 117. 1S74.

Basionym: Obione acanthocarpa Toney

This is a shrub or subshriib, generally less than 1 m
tall, characterized by spong)' fruiting bracteoles S-15 mm
long, borne on slender to stout pedicels 4-20 mm long.
Leaves are variable, but often sinuate-dentate to undu-
late-crisped and with hastately lobed base. The species
occurs from west Texas and southern New Me.xico south
to Mexico. It is represented in the United States by two
varieties, i.e., van acanthocarpa from western Texas west
through southern New Mexico to southeastern Arizona,
and var coahiiih'nsis in southern Texas.

Atriplex acanthocarpa ssp. coahiiilensis Henrickson,

Southw. Naturalist 33: 4.58. 1988.

= A. acanthocarpa var coahuilensis (Henrickson)
Welsh & Crompton (cited below).

Type: Mexico, Coahuila, ca 2 km W of Nadadores in
saline pastured flats near El Porvinir along Hwy. 30, with
Suaeda, Sporobolus, Distichlis, nar 27° 03' N lat,
10r37'W long, .540 m, 6 Dec 1975, J. Henrickson 14784;
holotype TEX; isotypes MEXU, NY!, RSA.

This ta.xon is distinguished by its fruiting bracteoles
bearing radiating processes, and stems with at least the
medial leaf blades hastate-lanceolate, and with mature
fi-uiting inflorescences ver>' long. Its range is from south-
ern Texas to southeast Coahuila and coastal Tamaulipas.

Atriplex acanthocarpa var. coahuilensis (Henrickson)
Welsh & Crompton, comb, now

Basionym: A. acanthocarpa ssp. coahuilensis Henrick-
son, Southwest. Nat. .33: 458. 1988.

Atriplex acanthocarpa var. cuneata (A. Nelson) M. E.
Jones, Contr West. Bot. 11: 20. 1903.

= A. gardneri var cuneata (A. Nelson) Welsh

Atriplex acanthocarpa var. pringlei (Standley) Henrick-
son, Southwest. Nat. 33: 461. 1988

Basionym: A. pringlei Standley

The taxon is endemic to Mexico, from northern
Zacatecas and southern Nuevo Leon south to San Luis
Potosi.

Atriplex acanthocarpa ssp. stewartii (I. M. Johnston)

Henrickson, Southwest. Nat. 33: 457. 1988.

Basionym: A. stewartii 1. M. Johnston

The taxon is endemic to Coahuila, Mexico, and is dis-
tinguished by its 4-winged fruiting bracteoles, although
specimens are transitional to A. acanthocarpa var acan-
thocarpa.

Atriplex amnicola P. G. Wilson, Flora of Australia 4: 322.
1984.

TyiDe: "Yalgoo, W. A." [western Australia], 10 Oct. 1945,
C. A.' Gardner 7751a; holotype PERTH!

Atriplex angustior Cockerell, Proc. Davenport Acad. Nat.
Sci. 9: 7. 1902.

= A. canescens (Pursh) Nuttall

Type: New Mexico, Dona Ana Co., Mesilla Park,
Cockerell in 1900; holotyi^e US!

The US specimen bears the following label data:
"Atriplex angustior, n. sp. Distinguished fi-om A. canescens
by the \'ery nanow (3 mm broad) leaves. Apparently = A.
canescens angustifolia but that name is preoccupied. Sand
Hills Mesilla Park, N. M. 1900. TD.A. Cockerell."

Hall and Clements (1923) cited this as a new name for
A. canescens var angustifolia, but it seems obvious that
while Cockerell recognized the equivalency of the taxa,
he was proposing a new taxon, not merely a new name.

Atriplex aptera A. Nelson, Bot. Gaz. 34: 356. 1902.

= A. gardneri var. aptera (A. Nelson) Welsh &
Crompton

Type locality: Wyoming, Laramie, Sept. 1901, E.
Nelson 738; A. Nelson (1902).

Type: "Atriplex aptera A. Nels. n. sp. Moist saline soil.
Laramie, Albany Co., Wyoming. Sept. 1901. Elias Nelson
No. 738"; holotype RM!;'isotype GH!, UC (frag.)!

Hanson (1962) suggested that his ta.xon was of hybrid
derivation involving A. canescens and A. huxifolia as
parental taxa. Distribution of specimens assignable to the
concept is sporadic, possibly indicating multiple origins,
and it cannot be considered a taxon in the usual sense.
The type specimen of A. aptera has definite wings aligned
in four rows similar to some A. canescens but agrees in
aspect, size, and general features with A. gardneri.

Atriplex berlandieri Moquin-Tandon, Chenop. Enum. 65.
1840.

= A. canescens (Pursh) Nuttall

Type: "In regno Mexicano. Berlandier 1828"; holotype ?

Moquin-Tandon enlarges on the type information in
his treatment in Prodromus (13[2]: 114. 1849), "In regno
Mexicano inter Laverdo et Bejar (Berland.! n. 1450)."

Atriplex bonnevillensis C. A. Hanson, Stud. Syst. Bot.
Brigham Young Univ. 1: 2. 1962.

= A. gardneri var bonnevillensis (C. A. Hanson) Welsh
Type: Utah, Millard Co., "diy lake bed 1.5 miles north-
east of headquarters. Desert Range Experiment Station
(dominant plant)," 12 July 1961, C. A. Hanson 354; holo-
type BRY!; isot>'pes GH!,'mO!, NY!, UTC!

The sheets at GH and NY have the date printed as 13
July 1961, probably representing t\'pographical errors.

324

Great Basin Natuiulist

[Volume 55

Atriplex brcweri S. Watson, Proc. Anicr. Acad. Arts 9: 1 19.
1874.

Type locality: "Fremont; 459 Torrcy; 75 Brewer" (I.e.).

Paratypes: "Fremont's 2nd Expedn. Atriplex Breweri
S. Wats.'"; NY Torre\'!; "No. 459. Santa Barbara County,
California. J. Torrey 1865"; NY! and NY LeRoy!, GH!

Type: "Geological Survey of California, 1863. Coll. H.
Brewer. No. 75. Atriplex Breweri n. sp. Sea Shore — Sta
Monica. 6"|ft] high or more"; lectotype GH!, here desig-
nated; isolectotypes NY!, UC, US!

Since the plant was described by Watson on the basis
of at least three collections, and as there are duplicates of
the Brewer collection, it is proper to designate the materi-
al at GH as lectotype. The sheet at US bears a sketch and
notes by John Torrey: "75. Obione — near the Sea, at Sta
Monica: Probably O. lentiformis (large fruited) in an
abnormal state. The bracts appear to have been changed
by galls.

Recognition of A. breweri at taxonomic level as either
a variety or subspecies of A. lentiformis is not without
merit. Indeed, the plants have typically larger leaves and
fruiting bracteoles that average larger However, there is a
series of intermediates that connect the robust coastal
material with the less robust plants in the interior. Plants
designated as belonging to A. breweri are considered by
me as ta.xonomically negligible.

Atriplex buxifolia Rydberg, Bull. Tone\ Bot. Club 39: 311.
1912.

= A. gardneri van aptera (A. Nelson) Welsh
Type locality: Wyoming, Sheridan Co., Dayton, 1220
m altitude, September 1899, Tweedy 2456; holotype NY!

Type; "F Tweedy 2656 (2456 in publication), Dayton,
4000 ft, Sheridan Co., "Wyoming, September 1899"; lecto-
type NY! (Basset et al. Genus Atriplex in Canada 58. 1983).

Atriplex canescens (Pursh) Nuttall, Genera N. Amer PI. 1:
197. 1818.

Basionym: Calligonum canescens Pursh

Putative or actual hybrids are known between A.
canescens and A. confertifolia or A. gardneri (various vari-
eties). Such hybrids are only occasiontd; tliey do not swamp
the characteristics of the taxa nor persist as populations.
The two e.xceptions to the sporadic nature of the hybrids
involving A. canescens as one of the parental types are A.
gardneri var. bonnevillensis and A. gardneri var aptera.
Neither of these ovei-whelms the parental taxa, but being
long-lived, they persist for long periods of time and occu-
py rather large areas in specific habitats. Bracts with four
wings appear to have arisen independently at several
places within the woody atriplexes. Such a condition is
not necessarily an indication of close genetic affinities.
Indeed, the garrettii and acanthocarpa complexes seem to
be more distantly removed from A. canescens than from
other taxa.

Atriplex canescens var. angtistifolia (Torrey) S. Watson,

Proc. Amer Acad. Arts 9: 121. 1874.
— A. canescens (Pursh) Nuttall

Basionym: Obione occidentale var angiistifolia Torrey
Narrow-leaved shrubs from west Texas are transitional

with broader-leaved materials both there and elsewhere.

They do not seem to constitute a taxon.

Atriplex canescens ssp. aptera (A. Nelson) Hall & Cle-
ments, Publ. Carnegie Inst. Wash. 326: 343, pi. 58. 1923.

Basion\in: A. aptera A. Nelson

= A. gardneri \ ar. a])tera (A. Nelson) Welsh

Atriplex canescens var. aptera (A. Nelson) C. L. Hilchc.,

Vase. Pis. Pacif NW. 2: 186. 1964.
Basionym: A. aptera A. Nelson
= A. gardneri var aptera (A. Nelson) Welsh

Atriplex canescens ssp. garrettii (Rydberg) Hall &
Clements. Publ. Carnegie Inst. Wash. 326: 344. 1923.
Basionym: A. garrettii Rydberg

Atriplex canescens var. garrettii (Rydberg) Benson, Amer
J. Bot. 30: 236. 1943.

Basionym: A. garrettii Rydberg

Atriplex canescens var. gigantea Welsh & Stutz, Great
Basin Nat. 44: 189. 1984.

Type: Utah, Juab Co., Lynndyl sand dunes, T35S,
R4W; 8 Sept. 1965, S. L. Welsh & G. Moore 5126; holo-
type BRY!; isotype NY!

The variety is based on its very broad bracts, stems
that produce roots b\' layering, thus accommodating bur-
ial in dimes, and diploid chromosome number

Atriplex canescens var. laciniata Parish, in Jepson, Fl.
Calif. 442. 1914.

= A. canescens X A. polycarpa? as to possible origin.

Type locality: California, Imperial Co., "Caleb,
Colorado Desert, Parish 8256" (Jepson I.e.).

Type: California, Imperial Co., "Plants of Southern
California, Salton Basin, Caleb. About 200 feet below sea
level. No. 8256. Coll. S. B. Parish. Oct 11. 1911"; holotype
UC J E PS!; isotype GH!

This variety has been suggested as based on speci-
mens intermediate between A. canescens and A. linearis
(C. A. Hanson I.e.), although Stutz (personal communica-
tion 1994) poses quite another possibility, i.e., that a chro-
mosomal race of A. polycarpa forming hybrids with A.
canescens has resulted in at least partially stabilized popu-
lations of var laciniata within the Salton Basin. The type
is characterized by deeply laciniate, 4-lobed bracteoles
within the size range of A. canescens. It has slender
branch lets and narrow leaves approaching those of both
A. linearis and A. canescens var niacilenta. which had a
similar origin from a separate chromosomal race of A.
polycarpa forming hybrids with A. canescens.

Atriplex canescens ssp. linearis (S. Watson) Hall & Cle-
ments, Publ. Carnegie Inst. Wash. 326: 344, pi. 58. 1923.

Basionym: A. linearis S. Watson

= A. linearis S. Watson

Atriplex canescens var. linearis (S. Watson) Munz, Manual
S. Calif Bot. 141. 1935.

Basionym: A. linearis S. Watson

= A. linearis S. Watson

Atriplex canescens ssp. macropoda (Rose & Standley) Hall
& Clements, Ph> log. Meth. 11«on 344. 1923.
Basionym: A. macropoda Rose & Standley
This ta.\on is known from Baja California.

Atriplex canescens var. macilenta Jepson, Fl. Calif 1: 442.
1914.

Tvpe locality: California, Imperial Co., "Holhille, Colo-
rado Desert, Parish 8258" (I.e.).

1995]

North American Perennial Atr/plex Types

325

Type: California, Imperial Co., "Plants of Soutliern
California. Salton Basin. Bluffs of Alamo River, Halhartle.
About 15 feet below Sea Level, S. B. Parish 8258, Oet. 18,
1912"; holotype UC JEPS!; isotypes DS (Xale.vico"),
GH!, POM!

The type has leaves to 4 mm wide, narrowly oblanceo-
late and obtuse apically. Bracts are small, as in A. linearis,
and toothed along the margin of the wings. The toothed
margin of the wings hints at the laciniate nature of bracts
on plants from the nearby Salton Basin and named van
laciniata Parish. Plants called var. macilento approach but
do not e.xactly match the more characteristic specimens of
A. linearis from southern Arizona and northern Mexico.
According to Stutz (personal commimication 1994), the
var. macilenfa is a high polyploid, while A. linearis is a
diploid. The relatively broader, thicker leaves of var. maci-
lenta are apparently diagnostic.

The specimen at DS, Parish 8258, Oct. 1912, is labeled
as having been taken on "Bluffs of the Alamo, Calexico."
It is one of three localities cited under Parish's number
8253, and the specimens other than the one taken at
Holtville are probably best considered as paratypes.

Parish made a series of collections from the Salton
Basin in October 1912. His numbers 8255 and 8256 were
collected on October 11; 8255 is a small-bracteoled, nar-
row-lea\'ed plant assignable to van uiacileuta, the type of
which (8258) was taken on 18 October Parish's number
8256, the type of var. laciniata, is evidently closely placed
geographically within the Salton Basin, which also sup-
ports A. pohjcarpa, which is potentially involved in the
origin of both vars. macilenta and laciniata through
hybridization with different chromosome races of A. pohj-
carpa through hybridization with A. canescens. Number
8255 approaches A. linearis in size of bracts and width of
leaves, and possibly that species is also involved in the
derivation of both vars. laciniata and macilenta.

Atriplex canescens var. occidentale (Torrey & Fremont)
Welsh & Stutz, Great Basin Nat. 44: 188. 1984.

Basion\'m: Pterochiton occidentale Torrey & Fremont
= A. canescens (Pursh) Nuttall var. canescens
This name was resurrected on false supposition that
tlie type of A. canescens sensu stricto differed from the tall
phases of the plant so widely distributed in the American
West. It is an unfoiiunate later synonym.

Atriplex collina Wooton & Standley, Contr. U.S. Natl.

Herb. 16: 119. 1913.

= A. confeiiifolia (Torrey & Fremont) S. Watson
Tyjje: Aiizona, Apache Co., "diy hills near the north end

of the Carrizo Mountains," P C. Standley 7481, 31 July

1911; holotype US!

Atriplex confertifolia (Torrey & Fremont) S. Watson, Proc.
Amer. Acad. Arts 9: 119. 1874.

Basionym: Obione confertifolia Torrey & Fremont, in
Fremont

Atriplex corrugata S. Watson, Bot. Gaz. 16; 341. 1891.

T\pe locality: "Nearly allied to A. nutiallii. Discovered
by Miss Alice Eastwood at Grand Junction, Colorado, in
well formed fruit on 20th May, 1891. Miss Eastwood notes
it as the earliest in fruit of several perennial species of the
genus growing in the same locality (I.e.).

lype: "Atriplex conaigata Watson, n. sp. Grand Junction,
Colorado. Miss Alice Eastwood — May 20/1891 "; holotype

GH!; isotypes UC (fiagments taken from holotspe bv H. M.
Hall)!, K,' MO, US!

The type consists of two fertile branches, one pistillate
and the other staminate. Both have the small, narrow
leaves characteristic of the taxon throughout its rather
small range. The species is almost exclusively restricted to
saline substrates of such fine-textured strata as the members
of the Cretaceous Mancos Shale and Jurassic Morrison
Formation, inter alia, where it often occurs as a monotype.
It forms occasional hybrids with A. confertifolia and A.
gardneri var. cuneata, with whom its ecology is sporadic.
The taxon is probabh' most closely allied to the latter, with
which it shares large land areas, but from which its aute-
cology is restricted. It is regarded herein at species rank
because of the maintenance of morphological integrity
despite occasional contact with the other taxa over much
of its area. Additionally, there are hints in its morphology
of close ties in still another direction, i.e., with A. ohovata.

Atriplex cuneata A. Nelson, Bot. Gaz. 34: 357. 1902.
— A. gardneri var. cuneata (A. Nelson) Welsh
Type locality: "M. E. Jones 5443, Emery, Utah, 1894,"

Nelson (1902).

Type: M. E. Jones 5443, Emeiy, 7000 ft., Emeiy Co.,

Utah, 16 June 1894; holotype RM!; isotypes MO!, NY! (3

sheets), US!

Atriplex cuneata ssp. introgressa C. A. Hanson, Stud. Syst.
Bot. Brigham Young Univ. 1: 4. 1962.

= A. gardneri var. cuneata X var. tridentata
Type: Utah, Carbon Co., "Wellington, ca 0.1 mi S of
Price River, in clay hills along road leading to city dump,"
9 July 1961, Hanson 346; holotype BRY!; isotypes GH!,
POM!

The specimens on which this taxon are based demon-
strate intermediacy between the cuneata and tridentata
phases of A. gardneri. Their recognition at any taxonomic
level is problematical.

Atriplex curvidens T. S. Brandegee, Proc. Calif Acad. Sci.
II, 2: 201. 1889.

= A. pohjcarpa (Torrey) Watson

Type: Baja California, Comondu, four feet high,
rounded April 24, 1889, Brandegee sn; holotype UC!

Atriplex decumbens S. Watson, Proc. Amer. Acad. Arts 12:

275. 1877.

= A. watsonii A. Nelson.

Type locality: California, "Near San Diego; Dr. E.
Palmer, 1875 (n. 334)" (Watson 1877).

Type: "Southern part of San Diego Co., California.
Coll. Edward Palmer, M.D., 1875. No. 334. Atriplex
decumbens, Watson n. sp. San Diego"; holotype GH!; iso-
type NY! (2 sheets).

The type consists of a small and a large branch, both
staminate. The large branch is evidently from a sprawling
herbaceous perennial. Leaves are luostly opposite, becom-
ing subopposite above, elliptic to ovate-lanceolate, obtuse
to roinided apicalK'; the glomerules are 3-5 mm thick and
are arranged in terminal spikes 1—4 cm long.

Atriplex eremicola Osterhout, Bull. Torrey Bot. Club 25:
284. 1898a. nom. no\-.

Basionym: A. fruticulosa Osterhout.

= A. gardneri (Moquin-Tandon) Dietrich var. gardneri

326

Great Basin Naturalist

[Volume 55

Atriplex falcata (M. E. Jones) Standley, N. Anier. Fl. 21;
68. 1916.

Ba.sionvni: A. mittallii \dr. fiilcata M. K. Jones, Coiitr.
W. Bot. ll! 19. 1903.

= A. gardneri vnr. falcata (M. E. Jones) Welsh

Atriplex fruticosa Nuttall ex Moquiii-Tandon, in de
Candolle, Prodr. 13(2): 112. 1849. pro syn.

= A. gardneri var. gardneri

Type: "Atriplex * fruticosa. A. Haliinuni afiinis. R. Mts."
Nuttall; holotypcBM!

The type oi A. fruticosa is mounted with collections
with the notation "British North America. Dr. Richardson
1819-28, " and designated as A. caiu'scens. In Inde.x Kewen-
sis the name fruticosa is noted as a synonym of A. canes-
cens, a supposition possibly based on the identity of the
Richardson material, but more probaljly on the publica-
tion of the name as a synonym of A. canescens by Moquin-
Tandon. The epithets /n/licosa and heterophyUa, both
herbarium names of Nuttall, were published as synonyms
and are not to be regarded in considerations oi priority.

Atriplex fruticulosa Jepson, Pittonia 2: 306. 1892.

Type: California, "Little Oak, Solano Co., Aug. 16,
1892. Willis L. Jepson"; holotyiDe UC!; isotype MO!

This plant functions mostK' as an annual but is appar-
ently capable of a longer life span, extending to become a
short-lived perennial. The name has priority' over the later
homonym, A. fruticulosa Osterhout (1898).

Atriplex fruticulosa Osterhout, Bull. Torrey Bot. Club 2.5;
207. 1898. non A. fruticidosa Jepson 1892.
Basionym for: A. eremicola Osterhout
= A. gardneri (Moquin-Tandon) Dietrich var. gardneri
Type locality: Wyoming, Albany Co., Steamboat Lake,
"The type was collected near a small alkaline lake in South-
ern Wyoming," G. Osterhout s.n. 2 July 1896; holotype
(no. 1324) RM!; isotype NY!, RM! (this second sheet, pre-
sumably an isot\'pe, lacks the collector's number). A col-
lector's number was not cited with the protologue, but the
holobt'pe sheet at RM bears the number 1324. Mateiial on
which this entity was based differs in no particular way
from A. gardneri var. gardneri.

Atriplex gardneri (Moquin-Tandon) Dietrich, Syn. Pi. 5;
537. 1852.

Basionym: Ohione gardneri Mocjuin-Tandon
There is a sheet, possibly identifiable as belonging to
this species and not bearing on the nomenclature of the
species, in the Lewis and Clark herbarium at PH; "A half
shrub from the high plains of Missouri. July 20th 1806." It
is cited here to demonstrate that the species was known
from the earliest collections into the western plains.

Atriplex gardneri var. aptera (A. Nelson) Welsh &
Crompton, comb. nov.

Basionym: Atriplex aptera A. Nelson, Bot. Gaz. 34:
356. 1902. '

This entity was treated by Hanson (1962) as a proba-
ble derivative of hybridization between A. canescens and
A. buxifolia (A. gardneri sens. kit.). It is a low subshrub
most similar to the latter, but with bracteoles winged as in
A. canescens or with tubercles aligned in foiu' rows, and
with yellow staminate flowers. It is likely that the condi-
tion of 4-winged fruits has arisen independently on many
occasions and that the resulting populations are not asso-
ciated genetically as in a typical taxon. Regardless of origin.

however, the resultant plants are readiK recognizable and
are widely distributed from southern Canada south along
the plains to Nebraska and Wyoming. A. canescens also
forms hybrids with other phases of the gardneri complex
(see below).

Atriplex gardneri var. honnevillensis (C. A. Hanson) Welsh,
Great Basin Nat. 44: 190. 1984.

Basionym: A. honnevillensis C. A. Hanson
This \ariety is more or less intermediate between A.
gardneri \ar falcata and A. canescens, but it most nearly
resembles the former in habit. The bracteoles are 5-8 mm
long and 3-9 mm wide, ovoid, with four lateral wings or
rows of flattened tubercules to 3 mm wide, or the wings
rarely absent. The plants are confined to playas and saline
pans in the valleys of western Utah and across Nevada.

Atriplex gardneri var. cuneata (A. Nelson) Welsh, Great
Basin Nat. 44; 191. 1984.

Basionym: A. cuneata A. Nelson

Atriplex gardneri \ar. falcata (M. E. Jones) Welsh, Great
Basin Nat. 44: 191. 1984.

Basionym: A. inittaUii vdi: falcata M. E. Jones

Atriplex gardneri var. tridentata (Kuntze) Macbride, Contr
Gray Herb. 3: 11. 1918.

= A. gardneri var utahensis (M. E. Jones) Dorn

Basionym: A. tridentata Kuntze

Atriplex gardneri var. welshii (C. A. Hanson) Welsh, Great
Basin Nat. 44: 191. 1984.

Basionym: A. welshii C. A. Hanson

Atriplex gardneri var. utahensis (M. E. Jones) Dorn, Vase.
PI. Wyo. 130. 1988.

Basionym: A. nuttallii \m: utahensis M. E. Jones

Atriplex garrettii Rydberg, Bull. Tone)' Bot. Club 39; 312.
1912.

T\pe; Utah, Grand Co., "Vicinit>' of Moab, " JuK* 1-2,
1911, P A. Rvdberg & A. O. Garrett 8465; holot>pe NY!;
isotypesGH!;US!, UT!

Despite earlier treatments in which this taxon was
regarded at infraspecific status within A. canescens, the
nearest allies appear to be in the gardneri complex.
Apparent hybrids are known between A. garrettii and A.
confertifolia (C. A. Hanson 1962), but not witli A. canescens.

Atriplex garrettii var. navajoensis (C. A. Hanson) Welsh &
Crompton, comb. nov.

Basion\m: A. navajoensis C. A. Hanson, Stud. Syst.
Bot. Brigham Young Univ. 1; 3. 1962.

This variet\' differs from the type material in plant
size, length of staminate inflorescences, color of staminate
flowers, and other intangibles. Generalh' the plants are
very similar. The few known localities, from the vicinit>- of
Lee's Ferry to Navajo Bridge in Coconino County, AZ, are
only disjunct by about 100 km from the nearest popula-
tions of \'ar. garrettii.

Atriplex gordoni Hooker, J. Bot. 5; 261. 1853. nom. nov.
pro A. gardneri McMiuin-Tandon.

= A. gardneri (Mocjuin-Tandon) Dietrich var gardneri

Atriplex greggii S. Watson, Proc. Amer. Acad. Arts 9; 118.
1874.

1995]

North American Perennial Atr/plex Types

327

= A. ohovata Moquin-Taiidon

Type locality: "New Mexico to Sonora. Collectors: —
1346 Berlandier; 462 Gregg; Emorv; Thurber; Bigelovv;
572, 1137, 1138 Wright" (Watson I.e.).

Paratypes: "No. 462. Atriple.x obovata Moc]. Perros
Bravos, Coahiiila, Mexico, Dr. J. Gregg, leg. 1S48-49"
(GH Lowell!); "Berlandier, No. 1346. Bae de del Salad.
San Luis Potosi, 1827" (Gil!).

Type: "462. Atriplex. Perros Bravos, north of Saltillo. 1
ft. tall. Abundant. State of Coahiiila, Mexico. Dr. J. Gregg,
leg. Sept. 20, 1848"; lectotype GH! (I. M. Johnston, J.
Arnold Arb. 25(2): 147. 1944); isolectot\'pe GH Lowell!

Atriplex griffithsii Standley, N. Amer Fl. 21: 63. 1916.
= A. lentifonnis \ar. grijfithsii (Standley) L. Benson
Type: Arizona, Cochise Co.: "Wilcox," Griffiths sn.

1895,' Oct. 12, 1900; holotype NY!; isotype US!

This is a distinctive tiixon with silveiy, thick leaves. It

is disjunct fiom die remainder of the species.

Atriplex heterophylla Nuttall ex Moquin-Tandon, in de

Candolle, Prodr. 13(2): 112. 1849. pro syn.

= A. gardneri (Moquin-Tandon) Dietrich var. gardneri
Type: "Atriplex * heterophylla. R. Mts." Nuttall; in-
tended type BM!

This is yet another herbarium name by Nuttall cited as
a synomym of A. canescens by Moquin-Tandon in de
Candolle's Prodromus. It again demonstrates that the
species was well represented in collections prior to the
collection of the type material of A. gardneri.

Atriplex hymeneltjtra (Torrey) S. Watson, Proc. Amer.
Acad. Arts 9: 119. 1874.

Basionym: Obione hijmenelijtra Torrey

Atriplex johnstonii C. B. Wolf, Occas. Pap. Rancho Santa
Ana Bot. Card. 1:3. 1935.

= A. numimilaria Lindl.

Type: California, Los Angeles County, Coastal cliffs,
Playa del Rey. C. B. Wolf 1821, 23 Dec! 1930; isotvpes
CAS!, GH!, NY!

The isotype at GH consists of four woody, leiily branches,
two ot them with fruiting bracts. Leaves are short-petio-
late, with blades 1.2-3.5 cm long and 1-3 cm wide.

Atriplex jonesii Standley, N. Amer. Fl. 21: 65. 1916. nom.
nov. pro A. sahitlosa M. E. Jones.
= A. ohovata Moquin-Tandon

Atriplex lentifonnis (Torrey) S. Watson, Proc. Amer Acad.
Arts 9: 118. 1874.

Basionym: Obione lentifonnis Torrey, in Sitgreaves
This is a wami-desert species, important in saline pans
along drainages at low elevations in the valleys of the
Colorado and Gila rivers and Salton Sink. The species is
distributed from western and southern Arizona, through
southern Nevada and California, and also in Mexico.
Hanson (1962) notes that A. lentifonnis sens. lat. forms
liybrids with A. leiicophijlla (Moquin-Tandon) Dietrich, a
perennial not especially woody species, and possibly even
with an annual species. Such hybridizations might indi-
cate that A. lentifonnis and its near relative A. torreiji have
alliances elsewhere than with the other wood)' species
treated herein.

Atriplex lentiformis ssp. hreueri (S. Watson) Hall &
Clements, Publ. Carnegie Inst. Wash. 326: 335, pi. 54.
1923.

Basionym: A. breiveri S. Watson
= A. lentifonnis sens lat?

Atriplex lentiformis var. breweri (S. Watson) McMinn,

Man. Calif Shrubs 113. 1939.

Basionjm: A. breweri S. Watson
= A. lentifonnis sens lat?

Atriplex lentiformis ssp. griffithsii (Standley) Hall &
Clements, Publ. Carnegie Inst. Wash. 326: 336, pi. 55.
1923.

Basionym: A. grijfithsii Standley

= A. lentifonnis sens, lat?

Atriplex lentiformis var. griffithsii (Standley) Benson,
Amer. J. Bot. 30: 236. 1943.

Basionym: A. griffithsii Standley

— A. lentifonnis sens, lat?

Atriplex lentiformis ssp. torreiji (S. Watson) Hall &
Clements, Publ. Carnegie Inst. Wish. 326: 335. 1923.
Basionym: Obione torreiji S. Watson

Atriplex lentiformis var. torreiji (S. Watson) McMinn,

Man. Calif Shrubs 113. 1939.

Basionym: Obione torreiji S. Watson

Atriplex linearis S. Watson, Proc. Amer Acad. Arts 24: 72.
1889.

T^pe locality'; Mexico, Sonora, alkaline soil about Guav-
mas. Palmer 120, 121, 235; s>'nt>pes GH.

Parat>pes: "Flora of Gua>anas, Mex. Dr. Edward Palmer,
1887. No. 120. Atriplex linearis Watson, n. sp. Garden
fences in alkaline soil. July"; GH! and "Flora of Guaymas,
Mex. Dr. Edward Palmer, 1887. No. 121. Atriplex Linearis,
Watson, n. sp. Garden fences, alkaline soil. July"; GH!

Type: "Flora of Guaymas, Mex. Dr. Edward Palmer,
1887. No. 235. Atriplex linearis Watson, n. sp. Plains in
alkaline soil. Sept."; lectot\pe GH! (G. D. Brown, Amer.
Midi. Nat. 55: 210. 1956).

Paratypes 120 and 121 are immature, the former pistil-
late, the latter staminate. The lectotype sheet #235 has at
least four branches with more or less mature fruiting
bracteoles. The bracteoles are 4-winged, rather deeply
laciniately lobed to merely toothed along the wings, and
are 3-6 mm wide.

Hanson (1962) regarded A. linearis as the most sub-
stantial variant within the canescens complex but recog-
nized that it forms hybrids with A. canescens. The plants
are certainly moiphologically distinct from most phases of
that entity. The slender, short to elongate leaves (seldom
more than 4 mm wide and to 3.8 cm long), fi-uiting brac-
toles seldom over 6 or 7 mm wide, and very slender
branchlets are apparently diagnostic in most instances.

Atriplex macropoda Rose & Standley, N. Amer. Fl. 21: 72.
1916.

= A. linearis S. Watson (sens lat?, but the fruiting
bracteoles are long pedicellate, unlike A. canescens)

T\pe localit}': "T\pe collected on Pinchillinque Island,
Lower California, March 27, 1911, J. N. Rose 16518 (U.S.
Nat. Herb. no. 638567)."

Type: Lower California, Pinchilinque Island, Gulf of
California, J. N. Rose 16518, March 27, 1911; holotype
US!

Atriplex matamorensis A. Nelson, Proc. Biol. Soc. Wash.
17: 99. 1904.

328

Great Basin Naturalist

[Volume 55

Nom. ii()\. pro. A. opixi.sitifolia S. Watson

Atriplex navajoemus C. A. Hanson, Stud. S> st. Hot. Brig-
ham \bung Univ. 1: 3. 1962.

= A. t^airettii var. navajoensis (C. A. Hanson) Welsh 6c
Crompton

Type; "Arizona: Coconino Co., east side of the Navajo
Bridge, July 21, 1961," C. A. Hanson 388; Iiolotype BRY!;
isoh'pe CH!

Atriplex X neomexicana Standley, N. Amer Fl. 21: 67. 1916.

= A. gardneri van cimeata X A. confertifolia

Type locaht)'; "Type collected on dry hills near Rmning-
ton, New Mexico, altitude 1550-1650 m, July 19, 1911,
Paul C. Standley 7066 (U.S. Nat. Herb. no. 686089)."

Tvpe: New Mexico, "Diy hills near Farmington," San
Juan Co., New Mexico, July 19, 1911, E C. Standley 7066;
holot>'pe US!

The name is evidenth' based on plants intermediate
between A. gardncri var. cuneata and A. confertifolia.

Atriplex nummularia Lindley, Mitch. J. Exped. Trop.
Australia 64. 1848.

T\'pe: Australia, "Cultivated in Italy, seed from South
Australia"; holot\'pe not seen.

Atriplex mtttallii S. Watson, Proc. Amer Acad. Arts 9: 116.
1874. nom. nov.

= A. canescens (Pursh) Nuttall sens. str.

It is unfortunate that one must at this late date attempt
to analyze Watson's use of the name mittallii for a portion
of the woody atriplexes in the American West. From its
publication in 1874 the name has been the source of much
confusion, sei-ving to clutter Atriplex nomenclature for all
subsequent time. It seems certain fiom a study of Watsons
proposal, justification for which can only be inferred, that
he was merely presenting a new name for material that he
thought to be misinteriDreted by contemporary' botanists.
The evolution of botanical thought with regard to the
perennial atriplex species parallels that for other newly
discovered ta.xa in the American West and was initiated
when the first of the woody specimens arrived from west-
ern botanical explorers. Few names were available, speci-
mens were few and often fragmentan, literature was diffi-
cult to obtain, and it was easy to misapply concepts and
mix names, a symptomology not of that era alone.

Supposed sensu names cited by Watson (1874) within
the synomymy of A. nuttall ii include Atriplex canescens as
used by Nuttall and an assortment of other historical
authors, Obione canescens of Moquin-Tandon and other
authors, and still another synonym, i.e., "A. gordonii
Hook.," with the citation "Pi. Geyer in Lond. Jour. Bot. 5:
261?," and l)\ implication the type of A. gordonii (i.e., A.
gardneri).

Watson first cited the name A. canescens as published
by Nuttall (1818), the implication being that Calligonwn
canescens Pin'sh, basionym of A. canescens, could not appK'.
Nuttall is indeed author of the combination Atriplex
canescens, and the place of citation is his 1818 publication,
wherein he cites C. canescens as the basionym of his com-
bination; furthermore, Nuttall's description is clearly C.
canescens Pursh, sensu stricto. It is Watson's understand-
ing of Nuttall's use of the epithet that is in error. Thus, A.
canescens of Nuttall is certainly not a mere sensu name,
however one might wish to interpret the application of
the epithet. Both the name and the concept as supplied by
Nuttall are A. canescens, including its basionym. A. mtttallii

of Watson thus includes the type oi Calligonwn canescens,
and the epitiiet nntlallii is illegitimate under stipulations
of the International CJode. Hence, from a nomenclatural
viewjDoint there is no problem. Nuttall based his Atriplex
canescens squarely on CaUigonum canescens Pursh, and
Watson quoted A. canescens Nuttall as the name-bringing
synonym of A. mittallii, which was stillborn. The lectotype
of Caligomiin canescens Pursh is at PH and is therefore
the lectotype of both Obione canescens and A. mittallii,
which cannot be transferred to a different species or
brought to life by sophisticated arguments. Hence, the
proposal for lectotypification by McNeill et al. (1983) is
illegitimate.

Atriplex mittallii var. anomala M. E. Jones, Contr W. Bot.
11: 19. 1903.

= A. gardneri var. falcata (M. E. Jones) Welsh

Type localitv: "The type is my specimens from Dolly
Varden Smelten E. Nevada, July 1894 [1891]."

Type: Nevada, Elko Co., "Marcus E. Jones Herbarium.
Atriplex nuttallii var anomata [sic] Jones n. var Dolly
Varden at the Smelter, VII-24-91. N.W of Ibapah, Utah."
M. E. Jones sn; holotype POM!; isotype UC (frag.)!

Jones was clearly in error in citing the date of the col-
lection as 1894. His itinerai-y cited in Leaflets of Western
Botany (10: 189-236) places him at the Dolly Varden
Smelter on 24 July 1891, not 1894.

Atriplex mtttallii ssp. buxifolia (Rydberg) Hall & Clements,

Phylog. Meth. Taxon. 325. 1923.
Basionym; A. buxifolia Rydberg
= A. gardneri (Moquin-Tandon) Dietrich var gardneri

Atriplex mittallii corrugata (S. Watson) A. Nelson, in
Coulter & Nelson, New. Man. Bot. Rocky Mts. 168. 1909.
= A. corrugata S. Watson

Atriplex mittallii ssp. cuneata (A. Nelson) Hall & Cle-
ments, Publ. Carnegie Inst. Wash. 326; 324, f 45. 1923.
Basionym; A. cuneata A. Nelson
= A. gardneri var cuneata (A. Nelson) Welsh

Atriplex nuttallii ssp. falcata (M. E. Jones) Hall & Cle-
ments, Publ. Carnegie Inst. Wash. 326; 324. f 45. 1923.
Basionym: A. nuttallii \m: falcata M. E. Jones
= A. gardneri vm: falcata (M. E. Jones) Welsh

Atriplex nuttallii vm: falcata M. E. Jones, Contr W Bot.
11; 19. 1903.

= A. gardneri vm: falcata (M. E. Jones) Welsh
Type locality: "Weiser, Idaho, July 1899, Jones" (I.e.).
Type; Idaho, Washington Co., "Flora of Idaho. Type
material. Atriplex nuttallii var falcata Jones n. \ar. Weiser,
Wash. Co. July 7 1899. Alt. 2200 Ft." M. E. Jones sn; holo-
type POM!; i.sotype UC!

Atriplex nuttallii ssp. gardneri (Moquin-Tandon) Hall &
Clements, Publ. Carnegie Inst. Wash. .326; 324. 1923.
= A, gardneri (Mociuin-Tandon) Dietricli var gardneri
Basionym: Obione gardneri Moquin-Txndon

Atriplex nuttallii ssp. tridentata (Kuntze) Hall & Cle-
ments, Publ. Carnegie Inst. Wash. 326: 324. 1923.

= A. gardneri var iitahensis (M. E. Jones) Dorn

Basionym: A. tridentata Kuntze

Atriplex nuttallii van gardneri (Moquin-Tandon) R. J.
Davis, Fl. Idaho. 261. 1952.

1995]

North American Perennial Atr/plea Types

329

= A. gardneri (Moquin-Tandon) Dietrich van gardneri
Btisionym: Obione gardneri Mncjuin-Tandon

Atriplex inittaUii van tridentata (Kuntze) R. J. Davis, Fl.

Idalio 261. 1952.

= A. gardneri var. utaJiensis (M. E. Jones) Dom
Basionym: A. tridentata Kuntze

Atriplex nuttaUii var. titahensis M. E. Jones, Contr. VV. Hot.

11: 19. 1903.

= A. gardneri var. utahensis (M. E. Jones) Dom

T\pe locality-; "This is No. 1760 Jones from Salt Lake

Cit}', and is the more common form in Utah."

Type: Utah, Salt Lake Cit\-, Salt Lake Co., M. E. Jones

1760, 16 June 1894: holot>pe POM?; isot>'pe UC (frag.)!

Atriplex oblanceolata Rydberg, Bull. Toney Bot. Club 31:

403. 1904.

= A. gardneri var. ciineata (A. Nelson) Welsh
Type locality: Colorado, Delta Co., Delta, Cowen 4071
(Rydberg 1904).

Type: "Plants of Colorado. No. 4071. Atriplex oblance-
olata Rydb. Delta, J. H. Cowen. Sept 3, 1897"; holotype
NY!; isotypes GH!, RM! (2 sheets), US!

Atriplex obovata Moquin-Tandon, Chenop. Enum. 61.
1840.

Type locality: "In Peruvia. (v. s. in herb. Mus. Paris) ' (I.e.).

Type: "No. 1346. Bae del Salad, Saint Louis Potosi. Dbre.
1827," and "Herbarium Berlandierianum Te.xano-mexi-
canum. No. 1346. Atriplex obovata, Moq.! O. canescens,
var? Torr. San Luis Potosi; Mexico, State of San Luis
Potosi, Berlandier 1346"; lectotype P? (I. M. Johnston, J.
Arnold Arbor 25[2]: 148. 1944); isolectotype GH!

The isolectotype sheet at GH consists of three leafy
branches, now lacking fruiting bracteoles or staminate
flowers. The material is certainly a match for what has tra-
ditionally passed under the name obovata; hence, there is
no problem with its interpretation.

Atriplex obovata var. tuberata Macbride, Contr Gray
Herb. 3: 11. 1918.

— A. obovata Moquin-Tandon

Type locality: Texas, El Paso Co., Fornillo Creek,
Harberd 103.

Type: "No. 103 (see specimen of male). 1-2° [feet] high
— Foliage & specially fruit different from that of A. acan-
thocarpa. Tornillo Creek. W. Texas. Aug. [1S]S3. V. Havard,
U.S.A."; holotvi^e GH!; isotype US!

The sheet at GH has two branches, one staminate and
one with fruiting bracteoles. The bracteoles are rather
stronglv' tuberculate, a feature not unusual witliin the species
as a whole.

Atriplex occidentalis (Torrey & Fremont) Dietrich, Svn.
PI. 5: 537. 1852.

Basionym: Pterochiton occidentalc Torre\' 6f Fremont

= A. canescens (Pursh) Nuttall

Atriplex odontoptera Rydberg, Bull. Torrey Bot. Club 31:

404. 1904.

— A. canescens X A. gardneri var gardneri

Type: Wyoming, Johnson Co., "3302. Atriplex canescens
(Pursh) James. A. odontoptera Rydb. (Type) Buffalo. Ele-
vation 4000-5000 feet. Frank Tweed>'. September 1900";
holotype NY!; isofype RM!

This is a coarse specimen, very woody and obviously
intermediate between A. canescens and A. gardneri var
gardneri

Atriplex oppositifolia S. Watson, Proc. Amer Acad. Arts 9:
118. 1874, non DC.

= A. matamorensis A. Nelson; Obione oppositifolia (S.
Watson) Ulbrich, in Engler & Prantl

Type locality': "In the Rio Grande Valley on the Mexican
side, collected only by Berlandier (No. 3201, 'Matamoras
to San Fernando ) (Watson I.e.).

Type: "de Matamaras a San Fernando circa Guijano,
Oct. 1830," and "Herbarium Berlandierianum Texano-
Me.xicanum. No. 3201. A. oppositifolia n. sp. S.W! [initials
are Sereno Watson's on sheet at GH]," Berlandier; holo-
type GH!; isotype NY!

The specimen at GH is doubly mounted with Palmer
1160, 1879. It is a portion of a herbaceous perennial with
minute leaves ca 2-3 mm long and 1 mm wide. Bracteoles
are conspicuously veined on the faces and prominently
toothed lateral to the apical tooth.

Atriplex orbicularis S. Watson, Proc. Amer Acad. Arts 17:

377. 1882.

= A. lentifonnis (Torrev) S. Watson (the A. breweri S.
Watson phase)

Type locality: "At Santa Monica, California, on the sea-
shore at the base of the bluffs; S. B. & W. F. Parish,
October, 1881" (Watson 1882).

Type: "Flora of Southern California, S. B. & F W.
Parish, No. 1126, perennial, somewhat woody at base, 3-4
ft high, base of bluffs, sea shore, Sta Monica, Oct. 1881";
holotype GH!; isotypes DS!, NY!, US!

The fruiting bracts are ca 3 mm high and 4 mm wide.
Leaves are elliptical and obtuse, tapering basally to a short
petiole.

Atriplex pabularis A. Nelson, Bull. Torrey Bot. Club 25:

203. 1898.

= A. gardneri var utahensis (M. E. Jones) Dom

Type locality: Wyoming, Sweetwater Co., Point of

Rocks, A. Nelson 4429, Aug. 30, 1897.

Tv'pe: "A. Nelson 4429, Bitter Cr, Point of Rocks, 6500

ft, Sweetwater Co., Wyoming, 30 August 1897"; lectot\pe

at RM! (Hall & Clements, Publ. Carnegie Inst. Wash. 326:

324. 1923); isolectotypes GH! (two sheets, male and

female), NY!, US!

Atriplex pabularis var. eremicola (Osterhout) A. Nelson,
Coulter & Nelson, New Man. Bot. Rocky Mts. 168. 1909.

Basionym: A. eremicola Osterhout

= A. gardneri (Moquin-Tandon) Dietrich var gardneri

Atriplex parrtji S. Watson, Proc. Amer Acad. Arts 17: 378.
1882.

Type locality: "Near Colton [actually at Lancaster
according to Parish in Zoe 5: 113, 1901], California; Dn
C. C. Parry 1881"; holot\pe (Pam- 221) GH!; isotypes NY!,
UC (frag.)!

The type consists of a branched stem, with lateral
spinescent stems to 4 cm long. The leaves are ovate-orbic-
ular The plant is obviously allied to A. conferiifolia, but
distinct.

Atriplex polycarpa (Torrey) S. Watson, Proc. Amer Acad.
Arts 9: 117. 1874.

Basionym: Obione polycarpa Tomey

330

Great Basin Naturalist

[Volume 55

Atriplex pringlei Standley, N. Amcr. Flora 21: 68. 1916.
= A. acanthocarpa s.sp. pringlei (Standley) Henrickson
Type locality: "Type collected on alkaline plains.

Hacienda de Ango.stiiia, San Lui.s Potosi, Mexico, July 15,

1891, C. G. Prinj^le 3775 (U.S. Nat. Herb. no. 48298)."'
Type: "Mexico, San Luis Potosi, alkaline plain, Hacienda

de Angostura, 15 Jul 1891," C. C Pringle 3775; holotype

US!;isotypeC;H!

Atriplex sahidosa M. E.Jones, Contn \\'. Bot. 11: 21. 1903.
non A. sdbulosd \\m\\\ 1890.

Basioin in of: A. jonesii Standley

= A. ohovata Moquin-Tandon

Type locality: Arizona, Navajo Co., "No. 4109 Jones,
Winslow, Ariz., Sept., 1884, distributed as A. Greggii"
(Jones 1903).

Type: "Flora of Arizona. 4109. Atriplex Greggii, Watson.
Winslow, M. E. Jones, September 1, 1884"; holotype US!;
i.sotypesGH!, NY!, POM!

The isotype at GH consists of three branches, two sta-
niinate and one pistillate.

Atriplex spinifera Macbride, Contr. Gray Herb. 53: 11.

1918.

Type locality: California, Kern Co., Maricopa Hills, May
15, 1913, East\vood 3269 (Macbride 1918).

Type: "3269. Flora of California. Atriplex. Mai^copa hills,
Kern Co., Alice Eastwood May 15. 1913"; holotype GH!;
isotype CAS!, US!

The holobt'pe at GH consists of a branched stem bear-
ing lateral spinescent branches to 4.2 cm long; that at US
consists of spinose branchlets and two packets of fruiting
bracteoles.

Atriplex spinosa (Moquin-Tandon) D. Dietrich, S\n. Pi.
5: 536. 1852.

Basionym: Obione spinosa Moquin-Tandon, in de
Candolle

= A. canescens (Pursh) Nuttall

Atriplex stewartii I. M. Johnston, J. Arnold Arbor. 22: 110.
1941.

— A. acanthocarpa ssp. steuartii (I. M. Johnston)
Henrickson

Type locality: Mexico.

Type: "Mexico: western Coahuila. Atriplex stewartii n.
sp. Jour Am. Arb. 22: 110. 1941. Eastern border of the
Llano de Guaje, along road from Tancjue del Aparejo 20
miles southeast of Tanque Armendais. Abundant on flats
margining playa at base of Lomas del Aparego (3 miles
south of Tangue Asparejo). Plant erect, 10-15 inches tall.
I. M. Johnston, C. H. Muller No. 777. Aug. 28, 1940";
holotype GH!

The plant is obviously allied to A. anthocarpa. the \ari-
ably 4-winged fruiting bracteoles having been derived
independently or possibly through introgression from A.
canescens. Henrickson (1988) does not suggest the latter
possibility but does note that the 4-winged condition is
not consistent, that there is a transition from that condi-
tion to those wheie the wings are replaced by radiating
processes.

Atriplex subconferta Rydberg, Fl. Rock-v Mts. 248. 1917

[1918].

= A. confertifolia (Torrey & Fremont) S. Watson
Type locality: Idaho, between Twin and Sliosiione

Falls, Nelson & Macbride 1379; holotvpe NY; isotvpes

POM, UC.

Type: "No. 1379. Atriplex confertifolia (Torr.) Wats.
Dry bench lands, alt. 3700. Twin Falls and Shoshone Falls,
3700 ft., Idaho, July 27, 1911, Aven Nelson & J. F
Macbride"; holotype NY!; isotypes GH!, MO!, PO.M, RM!,
UC, US!

This appears to be a small-leaved phase of A. confcrti-
jolia of little or no taxonomic significance.

Atriplex tetraptera (Bentham) Rydberg, Bull. Torre\ Bot.
Club 39: 311. 1912.

Basionym: Obione tetraptera Bentham

= A. canescens (Pursh) Nuttall

Atriplex torreiji (S. Watson) S. Watson, Proc. Amer. Acad.
Arts 9; 119. 1874.

Basionym; Obione torrcyi S. Watson

Atriplex torreiji var. griffithsii (Standley) G. D. Brown,

Amer. Midi. Nat. 55: 205. 1956.
Basionym: A. grijfithsii Standley
= A. lentifonnis (Torrey) S. Watson

Atriplex tridentata Kuntze, Rev. Gen. Pi. 2: 546. 1891.

= A. gardneri var. utahensis (M. E. Jones) Doni

Type locality: Utah, Box Elder Co., Corrine, Kuntze
,3084^ 1874.

Type: O. Kuntze .3084, "Bei Corinne am Salzsee, 7000
[much too higli] ft, [Box Elder Co.], Utah, September 1874";
holotype NY!; isotype? K!

The specimen at K, labeled "Atriplex tridentata OKze
n. sp. U.S. N. Am. zw. Cheyenne & Corinne. 7000'. Sept. 74.
3084. Herbarium Otto Kuntze, is perhaps best regarded
as a paratype.

Atriplex watsonii A. Nelson, Proc. Biol. Soc. Wash. 17: 99.
1904. nom. nov. pro A. decwnbens.
Basionym: A. decwnbens S. Watson

Atriplex welshii C. A. Hanson, Stud. S) st. Bot. Brigham

Young Univ. 1:1. 1962.

= A. gardneri var. welshii (C. A. Hanson) Welsh
Type: "Utah: Grand Co., 4 mi south of Cisco along state

highway 128, July 5, 1961"; C. A. Hanson .322; holotype

BRY!; isotypes GH!, ISC!

Calligontim canescens Pursh, Fl. Amer. Sept. 2: 370. 1814.

= A. canescens (Pursh) Nuttall

Type locality: Lyman or Buffalo counties. South Dakota,
M. Lewis in 1804.

Type; "Big Bend of the Missouri, Sept. 21, 1804," Lewis
and Clark Herbarium; lectoype PH!, G. D. Brown, Airier.
Midi. Naturalist 55: 209. 1956.

The original description of Calligonum canescens
Pursh is "C. dioicum, pidverulento-fruticulosum; folis
lanceolatis, floribus iixillaribus glomeratis in apice ramulo-
rum subspicatis, fructibus alatis, alis venosis cristato-den-
tatis. In the plains of the Missouri, near the Big bend. H.
July, Aug. v.s. in Herb. Lewis. Flowers exceeding small.
Goats delight to feed upon this shrub."

The sheet at PH contains three branches, the left one
with immature friiit, the middle one sterile, and the one at
right with mature fruiting bracteoles. This latter specimen
was designated specifically as the lectotype by McNeill et
al. (1983); it clearly fits the concept of the species as inter-
preted by contemporary authors, except for Stutz and
Sanderson (1979), who claim that the type belongs to what
was subsequently named A. aptera A. Nelson, based on the

1995]

North American Perennial Atr/plex Types

331

assumption that the Lyman County, South Dakota, type
locahty is not within the current range of A. canescens as
presently accepted but is within the range of A. aptera.
Examination of a great many specimens from throughout
the western plains has failed to yield a plant of A. aptera
with fruiting bracteoles identical to the lectot\'pe, which is
matched many times among the specimens traditionally
passing as A. canescens.

The type sheet bears the designation "Sept. 2L 1804,"
and the site of the Lewis and Clark camp on that date is
adjacent to present Lower Brule, Lyman or Buffalo comi-
ties, a short distance above the confluence of the White
River. That portion of the Missouri River has been inun-
dated by waters behind the Fort Randall Dam, far dovvTi-
stream. Nuttall had traversed the river corridor in 1811,
going upri\'er as far as Fort Mandan. The description and
discussion by Nuttall (1818) of the species is pertinent to
the inteipretation of die Lewis Wpe material. He describes
the plant as about 3 or 4 feet high, with the "cali.x (i.e.,
fruiting bracteoles) 2-partecl, becoming indurated, acute,
with 4 unequal cristated or dentated angles "; the habitat
was designated: "On the denudated saline hills of the
Missouri [possibly a reference to the lower-growing, vari-
able, gc/rc/;j('n'-like A. aptera]; commencing about 15 miles
below the confluence of the White River, and continuing
to the mountains [i.e., to the Mandan, as near as he went
toward the mountains]. Much of the habitat where plants
typical of A. canescens, as traditionally inteipreted, could
ha\e grown is beneath the waters of Fort Randall Dam,
and a valid assumption that bi'pical A. canescens did not
occur there cannot be made. Some plants from areas of
South Dakota adjacent to Lower Brule clearly approach
hpical A. canescens. There is no justification for inteipre-
tation of the name differently from that used in the his-
toric past.

Obione acanthocarpa Torrey, U.S. & Mex. Bound. Bot. 2:
183. 1859.

= A. acantliocarpa (Torrey) S. Watson

Type locality: "Plains between the Burro mountains;
September, Bigeloiv. (in fruit.) On the Rio Grande, below
Presidio del Norte; Parry. Near the Piloncilla, Sonora,
September"; Thurber (No. 1739; Wright. His No. 1737
seems to be a slender form of the same.)

Type: "Rio Grande below Presidio del Norte (El Paso),
Aug." Pan-y s.n.; lectotvpe NY! (Henrickson Southwest. Nat.
33: 454. 1988); isolectotype NY!

Obione berlandieri (Moquin-Tandon) Moquin-Tandon,

in de Candolle, Prodr. 13(2): 114. 1849.

Basionym: A. berlandieri Moquin-Tandon
= A. canescens (Pursli) Nuttall

Obione canescens (Pursh) Moquin-Tandon, Chenop.
Enuni. 74. 1840.

Basionym: Calligonuin canescens Pursh

— A. canescens (Pursh) Nuttall

Obione confertifolia Torrey & Fremont, in Fremont, Rep.

Explor. Exped. Oregon & California 318. 1845.

= Atriplex confertifolia (Torre)' & Fremont) S. Watson
Type localits': "On the borders of the Great Salt Lake"

(I.e.).'

Type: "Obione confertifolia. Torn if Frem. in Freni.

2nd Reprt. (1845). Borders of the Great Salt Lake [near

mouth of Weber River, Weber Co.], Utah. 761. 1843";

Fremont s.n. probably 10 September 1843; holotjpe NY!

This species is noted by Fremont (1845) in his journal
entry for 10 September 1843, on his return trip from
Disappointment [Fremont] Island. The plant was probably
collected on the trip from the water's edge to the camp on
the lower Weber River, in Weber Co., Utah. The holot\pe
consists of a single branch in young fruit. The sheet bears
the notation in Torreys handwriting, "Obione rigida var.
confertifolia n.sp. (crossed out) T. & F" Below the notation
is a drawing of a fioiiting bract, with one side folded back,
and an ovary. This is clearly the specimen on which the
species was based. There is a second sheet at NY!:
"Fremont's 2nd Expedn." with the notation "Grayia or near
it." The specimen has male inflorescence fragments and
clearly is not a portion of the t>'pe collection.

Obione coriacea (Forssk.) Moquin-Tandon, Chenop.

Enum. 71. 1840.

This Egyptian species was compared by Torrey and
Fremont (Fremont 1845) with Obione confertifolia (see
abo\'e). It does not occur in North America.

Obione gardneri Moquin-Tandon, in de Candolle, Prodr.

13(2): 114. 1849.

= A. gardneri (Moquin-Tandon) Dietrich var. gardneri

Type locality: SE Wyoming or W Nebraska, "Ad La
Platte, Gardn. n. 250 " (Moquin-Tandon in de Candolle I.e.).

Tvpe: "Gordon 250. La Platte. Obione Gardneri Moq.
A low female plant, lax spike in fruit," possibly 1843; holo-
t>'pe K?; isot\pe GH!

The fragments at GH consist of a leaf and two imma-
ture fiiiiting bracteoles, probably taken from the t\pe at K
(Hooker herbarium), which we have not seen. Writing on
the fragment enxelope is in ink, but partly illegible. The
name of the collector is subject to inteipretation, but is
presumed to be "Gordon. " Moc]uin-Tandon inteipreted it
as "Gardner, and named the species after the person
assumed by him to be the collector. The epithet was
spelled gardneri on purpose and is not an orthographic
variant. It is legitimate under stipulations of the Inter-
national Code.

Obione hymenelytra Torrey, in Whipple, Pacif R. R. Rep.

4: 129. 1857.

= Atriplex hymenelytra (Torrey) S. Watson
Type locality: "Hills and gravelly places on the William's
River [Bigelow]. This species was found by Dr. Parr)' and
by Colonel Fremont on the Gila" (I.e.).

Type: "Fremont's Expedition to California, 1849.
Obione hymenelytra, n. sp. " (lectotype NY'!, Brown, Anier.
Midi. Nat. 55: 203. 1956.). "Fremont's Expedition to
California, Gila" [1849] (presumed isolectotypes NY Crooke!,
GH!). A third sheet, "Fremont's 2nd Expedition," is at NY!
Except for the sheet designated as lectot>pe, the Fremont
materials fi-oiii 1849 are scant}', consisting mainly of fh^iiting
bracts (presumed isolectotypes NY!, GH!) and a branchlet
of equivocal source (GH!). The lectotype at NY bears all of
the accoutrements of a Torrey type specimen, except for
lack of illustrations, but includes a descriptive label in
Torre\ 's handwaiting and the name Obione hymenelytra,
11. sp., on the label.

Obione lentiformis Torrey, in Sitgreaves Rep. 169. 1854.
= Atriplex lentiformis (Torrey) S. Watson
T\pe localit)': Ciilifoniia, along the Colorado River, S. W.

Woodhouse s.n., 6 November 1851 (Sitgreaves E.xpedition,

November 1851) (I.e.).

332

Great Basin Naturalist

[Volume 55

Type: "Sitgreaves Report hSol. ()l)i()iic Iciitifonnis Toit.
in Sitgreaves ex Torrey. Rio CJoIoracIo, (^alil. — Nev. ex
Torrey "; "Rio Colorado. Nov. 6th 1851. Dr. Woodhoiise ";
lectotype NY! (.selected by E V. Covillc, Contr. U.S. Natl.
Herb. 4: 181. 1894); isolectotype GH!

Obione lentifonnis (i rhomhifolia Ibrrey, Pacific R. R. Rep.
4: 129. 1857.

Type: Arizona, NY?

I have been unable to locate material at NY with this
designation nor make a detennination as to its disposition
tiLxonomically.

Obione leucophylla Moquin-Tandon, in de CandoUe,
Prodr 13(2): 109. 1849.

= Atriplex leucophylla (Mo(iuin-Tandon) D. Dietrich
Type locality: "In California (Chamisso!), San-Fran-
cisco (Barclay!)" (I.e.).

Type: "San Francisco. Barclay ; holotype K!

Obione obovata (Moquin-Tandon) Ulbrich, Natm. Pfl. ed.
2. 16(c): 508. 1934.

= A. obovata Mocjuin-Tandon

Obione occidentalis (Torrey & Fremont) Moquin-Tandon,

in de Candolle, Prodr. 13(2): 112. 1849.

Basionym: Pterochiton occidentale Torrey & Fremont
== A. canescens (Pursh) Nuttall

Obione occidentale van angtistifolia Torrey, in Emoiy, Bot.
Mex. Bound. 2(1); 189. 1848.

= A. canescens (Pursh) Nuttall

Type localit\': Texas, Valley of the Rio Grande, Wright
in 1852.

Type: "Field No. 394. Obione, Sandy ridge on Rio
Grande, 3-5 ft tall, much branching, June 17, 1852. Rio
Grande below El Paso Te.xas. [Wright] 1742 = 394"; holo-
type NY!; isotypes GHl (3 sheets).

All three sheets at GH bear the number 1742 on the
laliel. One of them also has the number 394, which was
evidently the field collection number The number 1742
was subsequently applied. The specimens all have veiy
narrow leaves to ca 4 mm wide and immature fruiting
bracteoles. The sheet at GH with the number 394 is dou-
bly mounted with a second Wright collection (1741 = No.
24), which has 4-winged fruiting bracteoles to 7 mm wide.
Notes appear above both labels on the sheet. That above
24 reads: "24, Chenop. hills near Erontera, 3-4 ft tall,
branching widely, July 19, 1851. El Paso Co., Texas"; above
324 is, "324. Obione, sandy ridges on Rio Grande, 3-5
feet tall, much branching, June 17, 1852, Rio Grande
below El Paso, Texas.' The latter is an isotype. All of the
specimens appear to be A. canescens, sens. lat.

Specimens with narrow leaves occur here and there
throughout the range of the species. Those from western
Texas that fit within the concept of van angustifolia seem
not to represent a taxon worthy of consideration.

Obione oppositifolia (S. Watson) Ulbrich, in Engler &
Prand, Die Natun Pflanzenf Ed. 2. 16c: 508. 1934.
Basionym: Atriplex oppositifolia S. Watson

Obione polycarfm Torrey, in Whipple, Pacific R. R. Rep.
4: 130. 1857.

= A. polycarpa (Torrey) S. Watson

Type: Arizona, Graham Co., "With the preceding," i.e.,
"Hills and gravelly places, on William's River valley of the

Gila River [near base of .\lt. Graham, ca 13 mi SW of
Staiford]," October 28, 1846, Enioiy s.n.; holotype NY!

Obione rigida Torrey & Fremont, in Fremont, Rep. Explon
Exped. Oregon &: C'alifornia 318. 1845 (nom. nud.).

= Atrij)lex confeiiifolia (Torrey & Fremont) S. Watson
Authentic specimen: "Obione rigida T. & F On an
island [Fremont Island] in Great Salt Lake, [Weber Co.,
Utah], Fremont 767, 1843"; Fremont s.n., 9 September
1843 (NY!, ToiTey!).

The name was published without a description and is a
nomen nudum. The specimen was taken on 9 September
1843 when Fremont and his boating party were on
Disappointment [Fremont] Island in the Great Salt Lake.
It seems clear from the notation that Torrey intended, at
least initial!)', to name the species O. rigida, with the spec-
imen taken later on "borders of the Great Salt Lake" as
van confertifolia of that species. Reasons for change of
mind are not apparent, but Torrey abandoned the epithet
rigida in favor oi confertifolia. The application of the same
number in this case 767, to two sheets of the same ta.xon
is in keeping with the practice of Fremont, at least occa-
sionalK', of using the number to indicate a species and not
a collection. The sheet bears drawings of bracts, fruit,
seed, and embiyo, roughly sketched by Dn Torrey.

Obione spinosa Moquin-Tandon, in de Candolle, Prodr
13(2): 108. 1849.

= A. canescens (Pursh) Nuttall

Type locality; "In Columbia (Nutt.!). Phyllocaipa spin-
osa Nutt.! in herb. Hook." (I.e.).

T\pe; "Lophocan'a * Pterocarya (crossed out) * spinosa.
R. Mts of the Colimibia. Pt. canescens. Atriplex canescens?,"
Nuttall s.n.; holotype B\l!

This name has consistently been treated as a synonym
of A. confertifolia. but the Nuttall specimen at BM is A.
canescens.

Obione tetraptera Bentham, Bot. Voyage Sulph. 48. 1844.

= A. canescens (Pursh) Nuttall

Type locality: California, San Diego.

Type; "Ex Herbariae Musei Brittannici Voyage of
H.M.S. Sulphur Capt. E W Beechey 1836-37, Capt. E.
Belchen 1837-41. (Type collection of Obione tetraptera
Benth.) California, San Diego. Sept-Oct. 1839. Straggling
shrub 7-9 ft. Hills San Diego. George W Barcla\' 3060";
holot>'pe BM!; isot>pes GH!, K!, MO!

The isotypes at GH and K each consist of a large
branch witli few leaves and fruiting bracteoles still attached.
The leaves are up to 4 mm wide and the bracts somewhat
laciniate. Specimens approach the "laciniata" phase of A.
canescens and possibly represent intergradation of A. lin-
earis with A. canescens. The specimen at K bears the label
information, "Oliione tetiptera. California. Barkle\'. Hooker
1844."

Obione torreyi S. Watson, Rep. Geol. Explon 40th Parallel
5; 290. 1871.

= A. torreyi (S. Watson) S. Watson

Type locality: Nevada, Humboldt Co., diy valleys bor-
dering the Trukee and Carson rivers, ToiTey 463 (Watson
1871).

Type: "Herbarium of Columbia College, New York, No.
463. Obione torreyi S. Wats. Sterile saline plains, Hum-
boldt Co., Nevada. Collected by J. Toney 1865"; lectotype
GH! (G. D. Brown, Amen Midi. Nat. 55; 205. 1956);
isolectotype NY!

1995]

North American Perennial Atr/plex Types

333

The holot\'pe at GH is doubly mounted with Parry
280, 1881. It is staniinate, with glonierules ca 2 mm thick
aiTanged on short lateral spikes (to ca 1.5 cm long) on lat-
eral branches of a much larger paniculate cluster to 28 cm
long. Branches are longitudinally striate and ridged with
low, acute ridges.

PhyUocarpa spinosa Nuttall ex Moquin-Tandon, in de

Candolle, Prodr, 13(2); 108. 1849. pro syn.
= A. cdiu'scens (Pursh) Nuttall

Pterochiton canescens (Pursh) Nuttall, J. Acad. Nat. Sci.
Philadelphia 1: 184. 1847.

Basionym; Calligonum canescens Pursh

= A. canescens (Pursh) Nuttall

Pterochiton occidentale Torrey & Fieniont, in Fremont,
Rep. E.xplor. E.xped. Oregon & California 318. 1845.

A. occidcntalis (Ton-e>' & Fremont) Dietrich; A. canes-
cens var. occidentalis (Torrey &: Fremont) Welsh & Stutz

= Atriplex canescens (Pursh) Nuttall var. canescens

T>pe locality; "The precise locality- of this plant we
cannot indicate, as the label was illegible; l)ut it was prob-
ably ft'om the borders of the Great Salt lake" (I.e.).

Tvpe; "Pterochiton occidentale Toix & Frem." Fremont,
probably 10 September 1843 [locality data are missing
from the type specimen] (holotype NY!; microfiche BRY'!).

The herbarium sheet bears a folded sheet of paper
with the usual careful and detailed drawings of bracts,
embiyo, and seed, and the designation "Pterochiton. " In
the lower right corner of the sheet is written "Fremont, N.
Gen. Pterocaly.x," and at the bottom center the words
"Pterochiton occidentale, Torr. & Frem." The sheet con-
tains three branches, with the bracts mainly fallen away.
This sheet was designated as lectoype by G. D. Brown.
Amer. Midi. Nat. 55; 209. 1956, but no other specimens
are cited with the protologue and the designation should
be holotype.

References

Brown, G. D. 1956. Taxonomy of American Atriplex.

American Midland Naturalist 55; 199-210.
Davis, R. J. 1952. Flora of Idaho. W. M. C. Brown Com-
pany, Dubuque, lA.
Dietrich, N. F. D. 1852. Synopsis Plantarum. Sec. V.

\'imariae; Frieder. Voigtii.
Hall, H. M., and F E. Clements. 1923. The phylogenetic

method in taxonomy. Publications of the Carnegie

Institute, Washington 326; 1-355.
Hanson, C. A. 1962. New species of perennial Atriplex

fiom the western United States. Studies in Systematic

Botany, Brigham Young Universit\' 1; 1-4.
Henrickson, J. 1988. Revision oi Atriplex acanthocarpa

complex. Southwestern Naturalist 33; 451-463.
Jepson, W. L. 1914. Chenopodiaceae. Flora of California,

1(4); 428-448. Associated Students Store, University

of California, Berkele\-,
J(JHNST()N, I. M. 1941. New phanerogams from Mexico.

IV. Journal of the Arnold Arboretum 22; 110-124.
. 1944. Plants of Coahuila, eastern Chihuahua, and

adjoining Zacatecas and Durango. \. Journal of the

Arnold Arboretum 25; 135-182.
Jones, M. E. 1903. Chenopodiaceae. Contributions to

Western Botany 11: 18-22.

Kuntze, D. E. O. 1891. Revisit) Genera Plantarum 1;
1-374.

Mac:bride, J. F 1918. New or othei-wise interesting plants,
mostly North American Liliaceae and Chenopodia-
ceae. Contril)utions from the Gray Herbarium 3:
1-22.

McMinn, H. E. 19.39. An illustrated manual of California
shrubs. J. W Stacey, Inc., San Franciso.

McNeill, J., I. J. Bassett, C. W. Crompton, and E M.
Taschereau. 1983. Ta.xonomic and nomenclatural
notes on Atriplex L. (Chenopodiaceae). Tiixon 34;
549-556.

Moquin-Tandon, A. 1840. Chenopodeanim Monographica
Enumerato. R J. Loss, Paris.

. 1849. Atriplex L. In: A. P de Candolle, Prodro-

mus Svstematis Naturalis Regni Vegetabilis 13(2):
90-115.

Nelson, A. 1898. New plants from Wyoming, I. Bulletin
of the Torrey Botany Club 25; 202-206.

. 1902. Contributions fi^om die Rocky Mountain Her-
barium, IV. Botanical Gazette 34: 355-371.

. 1909. Atriplex L. In: J. Coulter and A. Nelson, New

manual of botany of the central Rocky Mountains.

American Book Compan\', New York.
Nuttall, T. 1818. The genera of North American plants 1;

1-312. D. Heartt, Philadelphia, PA.
. 1847. Plantes Gambler (?). Journal of the Academy

of Natural Sciences, Philadelphia 1; 184.
Osterhout, G. 1898a. A new Atriplex. Bulletin of the

Toney Botany Club 25; 207.
. 1898b. A correction. Bulletin of the Toney Botany

Club 25: 284.
PuRSll, F T. 1814. Flora Americae Septentrionalis 2:

359-751. White, Cocchrane and Company, London.
Rydberg, R a. 1904. Studies on the Rocky Mountain flora

XI. Bulletin of die Torrey Botanical Club 31; 399-410.
. 1912. Studies on the Rocky Mountain flora XXVI.

Bulletin of the Torre>' Botanical Club 39: 99-113.
. 1917 [1918]. Flora of the Rocky Mountains and

adjacent plains. Published by the author, New York.
. 1932. Flora of the prairies and plains of central

North America. The New York Botanical Garden,

New York.
Standley, P C. 1916. Chenopodiales. North American

Flora 21; 1-93.
. 1917. The Chenopodiaceae of the North American

flora. Bulletin of the Torre\' Botanical Club 44;

411-429.
. 1922. Trees and shrubs of Mexico. Contributions

from the U.S. National Herbarium 23: 1-690.

Stutz, H. C. 1978. Explosive evolution of perennial A//7'/j/fx
in western North America. Pages 161-168 in K. T.
Harper and J. L. Reveal, editors, Intermountain
biography: a symposium. Great Basin Naturalist
Memoirs 2.

Stutz, H. C, and S. C. Sanderson. 1979. The role of
polyploidy in the evolution of Atriplex canescens. In:
J. R. Godin and D. K. Northington, editors. Arid
plant resources. International Center for Arid and
Semi-arid Land Studies, Lubbock, TX.

Torrey, J. C. 1848. Botany In: W Emorv', Notes of a mili-
tary reconnaissance from Fort Leavenworth, in
Missouri, to San Diego, in California. U.S. Government
Printing Office, Washington, DC.

. 1852. Appendix D, Botany. 7/j; H. Stansbuni', Ex-
ploration and survey of the valley of the Great

334

Great Basin Naturalist

[Volume 55

Salt Lake of Utali. Lippencott, (Jiainlio (S: (^o.,
Philack'lpliia.

. 1854. Botany. In: L. Sitgreavcs, lii'port ol an ex-
pedition down the Zuni and Colorado ri\er,s. U.S.
Governinent Printing Office, Washington, DC.

. 1857. Description of the general botanical collec-
tions. In: A. VV. Whipple, E.xplorations and surveys
to ascertain the practical)le and economic route for a
railroad from the Mississippi River to the Pacific
Ocean (Pacific Railroad Report). A. O. R Nicholson
printer, Washington, DC.

1859. Botany of the honndaiy Pages 30-259 in W

Emory, Report of the U.S. and Mexican boundary
survey. Volume II. U.S. Department of the Interior,
Washington, DC.

TORREY, J. C, AND J. C. FREMONT. 1845. Descriptions of
some new genera and species of plants, collected in
Captain J. C. Fremont's exploring expedition to
Oregon and North California, in the years lS43-'44.
Pages 311-319 in J. C. Fremont, A report of the ex-
ploring expedition to Oregon and North California,
in the years 1843-'44. Senate Document 174, 28th
Congress, 2nd Session. Blair & Rivers Printers,
Washington, DC.

Ulbrich, E. 1934. Chenopodiaceae. In: A. Engler and K.
Prantl, Die Naturlichen Pflanzenfamilien. Ed 2. 16c;
500-519.

W.vrsON, S. 1871. Botany. In: C. King, Report of geological
exploration of the fortietli parallel 5: 1-525. U.S.
Government Printing Office, Washington, DC.

. 1874. A revision of the North American Cheno-
podiaceae. Proceedings of the American Academy of
Arts 9: 82-126.

. 1877. Descriptions of new species of plants, with

revisions of certain genera. Proceedings of the Aineri-
can Academy of Arts 12: 246-278.

. 1882. Contributions to American botany. XVII. List

of plants from southwestern Texas and northern
Mexico, collected chiefly by Dr. E. Palmer in 1879-80.
— PoK'petalae. Proceedings of the American Academy
ofArtsl7:316-.378.

. 1891. Atiii)I('x in articles. Botanical Gazette 16:

345-346.
WooTEN, E. O., AND R C. Standley. 1913. Descriptions
of new plants preliminar\' to a report on the flora of
New Mexico. Contributions of the U.S. National
Herbarium 16: 109-196.

Received 15 Fchnmnj 1995
Accepted 25 April 1995

NEW RECORDS OF SCOLYTIDAE FROM WASHINGTON STATE

Malcolm M. Furniss^ and James B. Johnson^

Abstract. — Eighteen species of Scolytidae are reported from Washington state for the first time or raised from
obscurity: Scieriis annectens LeConte, Hijlesiniis californicus (Swaine), Phloeotribus lecontei Schedl, Carphoborus
vandykei Bruck, Polygraphiis rufipennis (Kirby), Cnjpturgus borealis Swaine, Pityogenes knechteli Swaine, Ips rnexicanus
(Hopkins), Ips pertiirbatus (Eichhoff), Ips plastographiis plasfographus (LeConte), Ips woodi Thatcher, Trypodendron
betidae Swaine, Trypophloeus striattdus (Mannerheim), Procryphalus mucronatus (LeConte), Procryphalus iitahensis
Hopkins, Pseudopityophdwrus piibipcnnis (LeConte), Pityophdionis alpinensts G. Hopping, and Pityophdwnis grandis
Blackman. Host tree and collection data are given for these species. A total of 105 scolytid species known from
Washington are listed.

Key words: Scolytidae, fanned list, Washington state.

Washington is a large state with seven physi-
ographic provinces (Franklin and D\niess 1973),
ranging from sea level (Fuget Trough) to over
4450 m on Mount Rainier (southern Washington
Cascades). Under the influence of moisture,
temperature, and substrate, natural vegetation
types range from coniferous forests through
woodland to shrubsteppe. Along Washington's
western edge, the Coast Range and Olympic
Mountains intercept the moisture-laden pre-
vailing winds from the Pacific Ocean, helping
to make the temperate forests of western
Washington (and northern Oregon) the most
dense in the world. They are composed almost
exclusively of conifers and in that respect are
also unique among temperate forests.
Eastward lies the Cascade Range that contains
Mount Rainier and other volcanic peaks. Mixed
conifers prevail in these mountain ranges.
Farther east is the Columbia Rasin, largest and
most arid of the provinces, occupying virtually
the southeast quarter of the state, except for a
bulge of the Rlue Mountains extending north-
ward from Oregon. Trees of this province are
restricted mainly to water courses and urban
areas. North of the Columbia Basin is the
Okanogan Highlands province, bordering on
British Columbia and Idaho, which provides a
vegetational bridge to the more diverse north-
ern Rocky Mountain flora.

The provinces of Washington vary greatly
in their climate, resulting from complex inter-
play between maritime and continental air

masses and the mountain ranges, particularly the
Cascade Range that divides the state into east-
em and western parts. For example, Quinalt on
the Pacific side of the Coast Range receives 337
cm of precipitation annually, whereas Yakima,
in the rain shadow to the east of the Cascade
Range, has only 20 cm. Average January and
July temperatures for Seattle (Puget Trough)
are 4.5°C and 18.7°C, whereas those for
Yakima (Columbia Basin) are -2.5 °C and
21.7°C.

The Scolytidae of Washington are host spe-
cific to vaiying degrees, and the extent of their
diversity is related to the diversity of their
woody host plants. Conifers are hosts of 87
species listed herein. A majority of these (81
species) are restricted to one or a few species
of Pinaceae in the genera Abies, Larix, Picea,
Piniis, Pseudotsuga, and Tsiiga, while six species
infest Cupressaceae {Thuja, Chamaecyparis, and
Juniperus). The remaining 19 species infest
angiosperms {Popidus, Salix, Alnus, etc.). By
their habits, Washington Scolytidae are charac-
terized as true bark beetles, living in phloem
(90 species); ambrosia beetles, living in xylem
where they may feed entirely or partly on
symbiotic fungi that they transmit (13 species),
living in pine cones {Conophthorus ponder-
osae Hopkins), or living in the roots of red clover
{Hijlastiniis obscurus [Marsham]).

Patterson and Hatch (1945) listed 73
species of Washington Scolytidae, adjusted to
present-day synonymy. Wood (1971, 1982) lists

lDi\ision ot Entomolog), Universih- of Idaho. Moscow, ID 83S44-2339.

335

336

Great Basin Naturalist

[Volume 55

Washington in the distribution ol 82 speeies of
Scolytidae; six adchtional speeies are hsted h\
Wood and Bright (1992). We herein update
those pubhcations with 15 new state records
collected by us or found in museum collec-
tions, and three species collected by M. A.
Deyrup (personal communication). Similar
lists have been published for Idaho (Furniss
and Johnson 1987), Montana (Cast et al. 1989),
and Oregon (Furniss et al. 1992).

Additional species of Scolytidae are likely
to be collected in Washington in the future.
The\' may include species known to occur in
adjacent states or British Columbia, hosts of
which occur in contiguous areas of Washington.
Also, commerce from foreign countries enter-
ing Puget Sound and the Columbia River may
bring exotic species accidentally. Species that
infest xylem (ambrosia beetles) are especially
well adapted to such transport. The establish-
ment of ambrosia beetles, which typically are
not very host-specific, is enhanced by the
moderate climate and great diversity of native
and e.xotic flora in the Seattle area. Indeed, it
is probable that such introduced scolytids may
have already gained a foothold there and have
not yet been detected.

The following are abbreviations for reposi-
tories listed for specimens new to Washington:
ABS = Aichbold Biological Station, Lake Placid,
FL; FS-Rl = Forest Sei-vice, USDA, Region
1, Missoula, MT; PNW = Pacific Northwest
Forest and Range Experiment Station, Forest
Service, USDA, Coi-vallis, OR; SLW = S. L.
Wood, Brigham Young University, Provo, UT;
WFBM = W F Ban- Entomological Museum,
University of Idaho, Moscow, ID.

Species New to Washington
Subfamily Hylesininae

Scierus annectens LeConte

Biology. — Monogynous. Infests lower bole
and roots of felled Picea spp., rarely Pinus con-
torta, often by entering a galleiy of Demlroc-
tonus rufipennis (Kirby). The parent galleiy is
3-4 cm long, inclined diagonally across grain.
One generation per year (Stewart 1965).

Distribution and notes. — Canada: Alta.,
B.C., N.B., Newf., Ont., Que., NWT; USA: Alas.,
Ariz., Calif., Colo., Ida., Me., Mont., N.H.,
N.M., Ore., Ut.; Washington: Tieton Ranger
Station, Yakima Co., 17-VIII-1955, Picea engel-
mannii, K. H. Wright (4 PNW, 1 WFBM).

Hylesuuis califoniiciis (Swaine)

Biology. — Monogynous. Infests the bole
and limbs of Fraxinus spp. Egg galleries are
transverse and deepK' engrave the wood. Over-
wintering beetles evidently form feeding timnels
in green bark o( Fraxinus spp. (Wood 1982).

Distribution and notes. — Mexico: Chih.;
USA: Ariz., Cahf., Colo., N.D., N.M., Okla.,
Ore. Tex., Ut.; WASHINGTON: Pack Forest, La
Grande, Pierce Co., lO-V-1941, Fraxinus latifo-
lia (=()reg.ona), R. L. Furniss. Two trap trees,
4" and 7" diameter, felled 4-II1-1941. Pairs of
beetles and eggs present in 2.5-cm galleries
lO-V-1941. Ten km N Adna, Lewis Co., 14-
VII-1991, Fraxinus latifolia, M. M. Furniss and
J. B. Johnson (approx. 100 WFBM, 2 SLW).
Infesting underside of a 12-cm-diameter bro-
ken-off branch on ground. Galleries each with
a female and male parent, eggs present. Adult
progeny reared, some larvae tunneled into
xylem for a depth of four annual growth rings
before transforming to adults.

Phloeotribus lecontei Schedl

Biology. — Monogamous. Male constnicts an
entrance tunnel and the bases of two egg gal-
leries that are then completed by the female.
Egg galleries run obliquely across the grain of
shaded-out branches in merchantable-size liv-
ing trees. Adults and larvae may be present
throughout the year; ovei"wintering adults may
occur in brood galleries, special hibernation or
maturation tunnels, or newly formed parental
galleries (Wood 1982).

Distribution and notes. — Canada: Alta.,
B.C.; USA: Aiiz., Calif, Colo., Ida., Mont, N.M.,
Ore., Ut., Wyo.; WASHINGTON: 7 km S Harts
Pass, Okanogan Co., 5-VII-1988, Picea engel-
mannii, M. M. Furniss (1 WFBM). Collected
from a branch of a 60-cm-diameter wind-
felled tree. Horseshoe Lake, Skamania Co.,
17-VII-1991, Picea engelmannii, M. M. Furniss
and J. B. Johnson (9 WFBM). New attacks in
1-cm-diameter shaded-out branch, without
needles, attached to live tree. Swank Pass,
Blewett, Chelan Co., ll-V-1975, Abies grandis,
M. A. Deyrup (ABS). In a small branch. Same
locality and date, Pseiidotsuga nienziesii, M. A.
Deyrup (ABS). In a shaded-out branch.

Carphoboriis vandykei Bruck

Biology. — Polyg>'nous, unstudied. Members
of the genus infest small, shaded-out branches
of living trees or boles of small, suppressed.

1995]

New Records of Washington Scolytidae

337

unthrifty trees. Most species live in host tissue
that is drier than is typical for bark beetles
(Wood 1982).

Distribution and notes. — Canada: B.C.;
USA: Calif., Ore.; Washington: Heritage
Campground, Olympia, Thurston Co., 14-VII-
1991, Pseudotsuga tnenziesii, M. M. Fumiss and
J. B. Johnson (approx. 200 WFBM). Infesting a
2.3-m-long, 6-cm-diameter, broken-off branch
with red foliage. Also present was Pseudohyle-
sinus nebulosus LeConte. Two to four egg gal-
leries radiated from the central nuptial cham-
ber, deeply etching the sapwood. Egg galleries
each extended 2-5 cm, their length inversely
dependent upon attack density. Eggs present,
laid alternately on opposite sides (not opposite
each other) in deep niches at a rate of 6 per cm
and sealed with a reddish brown coating of
frass. Hatched larvae fed in the phloem, not
etching the wood. Some lai-val mines equaled
or exceeded the length of egg galleries but
most were shorter and very broad, apparently
influenced by brood density. Kept at room
temperature, adult brood pulverized the bark
and deeply scored the xylem before emerging
from veiy diy branch-wood one and one-half
years later. The scored xylem had a powdeiy
white appearance, perhaps due to presence of
associated yeast. Carson, Skamania Co., 18-Vll-
1991, Pseudotsuga menziesii, M. M. Fumiss and
J. B. Johnson (approx. 100 WFBM). Infesting
1-2 V2-cm-diameter branches of a 25-cm-
diameter standing tree that had discolored
foliage (dying). Galleries with parent beetles
and larvae. Phloem very dry. Little Rock,
Thurston Co., 30-IV-1975, Pseudotsuga men-
ziesii, M. A. Deyrup (ABS). In a dead branch.
Tahuya, Mason Co., 21-VI-1975, Pseudotsuga
menziesii, M. A. and N. Deyrup (ABS). In a
small, suppressed tree.

Polygraphus rufipennis (Kirby)

Biology. — Folygynous. Recorded common-
ly from Picea spp., especially P. glauca and P.
engelmannii, rarely from other genera of Pina-
ceae. Occasionally kills small-diameter, sup-
pressed trees, commonly occurs as a secondaiy
species in trunks of felled or dying trees. Two
to five egg galleries radiate from each nuptial
chamber, most commonly two, each made by a
different female. One generation per vear
(Hilton 1968).

Distribution and notes. — Canada: all
provinces; USA: Alas., Ariz., Colo., D.C., Ida.,

Me., Mass., Mich., Minn., Mont., N.H., N.M.,
N.Y., N.C., N.D., Ore., Penn., S.D., Tenn., Ut.,
Ven, W.V., Wise, Wyo.; WASHINGTON: Evans
Creek, King Co.; Nacotta, Pacific Co. (Hilton
1968). Kooskooskie, Walla Walla Co., 28-IX-
1955, Picea engelmamiii, W J. Buckhorn. Lake
Wenatchee, Chelan Co., 22-IX-1955, Picea
engelmannii, P W Orr. Metaline Falls, Pend
Oreille Co., 1929-1931, Picea engelmannii and
Pseudotsuga menziesii, H. J. Rust and W D.
Bedard. Park-way, Pierce Co., 17-V-1934, Pinus
contoi-ta, J. A. Beal. Plain, Chelan Co., 19-IX-
1955, Picea engelmannii, P W Orr. Mt. Rainier
N.P, 29-X-1930, Picea engelmannii, F P Keen
and W J. Buckhorn. Winthrop, Okanogan Co.,
22-X-1935, Picea engehnannii, R. L. Fumiss (all
PNW). Horseshoe Lake, Skamania Co., 17-VII-
1991, Picea engelmannii, M. M. Fumiss and J. B.
Johnson. Infesting shaded-out branches of a
60-cm-diameter, wind-felled tree (3 WFBM).

Comment. — This common beetle is certain
to occur throughout the range of P. engelman-
nii in the Cascade Range and Okanogan High-
lands. The Pacific Co. record is likely to be in
P. sitchensis; if so, it is a new host record.

Subfamily Scolytinae

Crypturgus borealis Swaine

Biology. — Monogamous. This smallest
Washington scolytid enters galleries of other
bark beetles in stems of conifers {Abies, Picea,
Pinus). They then tunnel irregularly into the
phloem. Apparently one generation per year,
ovei-wintering as adults in the brood galleries
(Wood 1982).

Distribution and notes. — Canada: Alta.,
B.C., Man., N.B., NWT, N.S., Ont., Que., Sask.;
USA: Ai-iz., Colo., Ida., Me., Mich., Mo., Mont.,
N.M., N.Y., Ore., Penn., S.D., Ut.; Washington:
Harts Pass, Okanogan Co., 5-VII-1988, Abies
lasiocarpa, M. M. Furniss (6 WFBM). Infest-
ing lower trunk of a 30-cm-diameter standing
tree having orangish red foliage and new
attacks by Pityokteines sp. Seventeen km W
Mazama,' Okanogan Co., 12-VII-1991, Abies
lasiocarpa, M. M. Furniss and J. B. Johnson (3
WFBM). Infesting lower trunk of a 25-cm-
diameter standing tree having red foliage and
abandoned galleries of another scolytid, either
Pityophthonis sp. or Pityokteines sp.

Pityogenes knechteli Swaine

Biology. — Polygynous. The egg gallery is
stellate with 4 to 6 branches radiating from the

338

Great Basin Naturalist

[Volume 55

nuptial chamber. Ovenvintering stages include
larvae, pupae, and adults (Alberta, Canada).
One and a partial second generation occur per
year at that latitude (Reid 1955).

Distribution and notes. — Canada: Alta.,
B.C., Sask.; USA: Aiiz., Calif., Ida., Mont., Ore.,
Ut., Wyo.; Washington: Twisp, Okanogan
Co., 12-Vin-193(), Pimis contorta, E R Keen (2
PNW).

Ips mexicanus (Hopkins)

Biology. — Polygynous. Not studied. Infests
Piniis spp.; egg galleries cune outward from a
central chamber (Wood 1982).

Distribution and notes. — Canada: Alta.,
B.C.; Mexico: Baja Calif, Distrito Federal,
Chiapas, Dgo., Hildago, Mex., Mich., Pue.,
Vera.; GUATEMALA; USA: Alas., Ariz., Calif,
Colo., Ida., Mont., Ore., Ut., Wyo.; Washing-
ton: Tieton Ranger Station, Yakima Co., 18-
VI- 1956, Pinus alhicaulis (new host record),
P W On- (15 PNW, 2 WFBM). Horseshoe Lake,
Skamania Co., 17-VII-1991, Pinus contorfa,
M. M. Furniss and J. B. Johnson (3 WFBM).
Sparse galleries in 30-cm-diameter standing
tree with dead top and mottled (dying) foliage.
Hyhtrgops porosiis (LeConte) also sparse in base.
Umatilla National Forest, 45 km S Pomeroy,
Garfield Co., 19-VII-1991, Pinus contorta, M. M.
Furaiss and J. B. Johnson (2 WFBM). Infesting
a 23-cm-diameter standing tree with red foliage.
Egg galleiy deeply etched xylem, its branches
aligned more or less witli wood grain but cui-v-
ing somewhat and irregular due to several
turning niches. Base with moist, sour bark.
Also present were Trypodendron lineatum
(Olivier), Dendroctonus valens LeConte, Oi-tho-
tomicus caelatus (Eichhoff), and Pityophthonis
confertus Swaine. Bremerton, Kitsap Co., 21-
IV-1974, Pinus contoi-ta, M. A. Deyrup (ABS).
In a standing, dead tree.

Ips pciiurbatus (Eichhoff)

Biology. — Polygynous. Breeds abundantly
in Picea glauca logging slash and in tops of
trees killed by Dendroctonus beetles. Parental
galleries have a tuning fork pattern with mod-
erately long larval mines. One generation
annually but two sets of egg galleries may be
constructed by females in one season (Furniss
and Carolin 1977).

Distribution and notes. — Canada: Alta.,
B.C., Man., N.B., NWT, Ont., Que., Sask.,
Yukon; USA: Alas., Me., Mich., Minn., Mont.;

Washington: Montesano, Grays Harbor Co.,
8-IV-1973, Picea sitchensis, M. A. Deyrup

(ABS).

Ips plastographus plastograpJnis
(LeConte)

Bioloc;y. — Polygynous. Usually infests upper
side of fallen Pinus contoi'ta, rarely Pinus pon-
derosa. Two or three longitudinal egg galleries
radiate from each nuptial chamber. Mature
larvae and young adults may bore 1 cm into
wood prior to emerging (Wood 1982).

Distribution and notes. — Canada: B.C.;
USA: Calif, Ida., Ore., Mont., Wyo.; Washing-
ton: Kettle Falls, Stevens Co., IX-5-1968,
Hopkins U.S. no. 54222, Pinus ponderosa, E W.
Honing and J. E. Dewey (FS-Rl).

Ips woodi Thatcher

Biology. — Polygynous. Infests large limbs
and boles of unthrifty or felled 5-needle Pinus
spp. Egg galleries parallel, resembling a nar-
row tuning fork (Wood 1982).

Distribution and notes. — Canada: Alta.;
USA: Ariz., Ida., Mont., Nev., N.M., Ut., Wyo.;
Washington: Tieton Ranger Station, Yakima
Co., Pinus alhicaulis (new host), 21-IX-55 to
12-VII-1956, P W. Orr (26 PNW, 3 WFBM).

Trypodendron betulae Swaine

Biology. — Monogynous. Tunnels are con-
structed by females radially through bark into
sapwood of Betula spp, rarely Alnus sp. The
main tunnel branches at close intei"vals, left or
right, in the same plane. Eggs are laid in nich-
es oriented above and below the gallery.
Larvae excavate short cradles in which they
develop and feed on ambrosia fungus. Males
are active in keeping the tunnels clean and
aerated (Wood 1982).^

Distribution and notes. — Canada: Alta.,
B.C., Man., N.B., N.S., NWT, Ont, Que.; USA:
Ida., Me., Mass., Minn., Mont, N.H., N.J., N.Y,
S.D., Wise; Washington: Metaline Rills, Pend
Oreille Co., 31-V-1930, Betula occidentalis,
Hopkins no. 19839 (PNW).

Trypophloeus striatulus
(Mannerheim)

Biology. — Monogynous. Unstudied, infests
stems of Salix scouleriana, Salix spp., Alnus
crispa, and A. rugosa.

Distribution and notes. — Canada: Newf ,
N.S., Que., Yukon; USA: Alas., Colo., Ida.,

1995]

New Records of Washington Scolytidae

339

Minn., Ut.; WASHINGTON: King Co., 20-VI-
1976, Populus trichocarpa, M. A. Deyrup
(ABS). In branch.

Procryphahis miicronatus

(LeConte)

Biology. — Monogamous. Infests smooth,
outer bark of stems of larger, dying, standing
Populus tremuloides. Ovei-Avinter as lai"vae and
adults; one and one -half to two generations
per year (Petty 1977).

Distribution and notes. — Canada: Alta.,
B.C.; USA: Alas., Colo., Ida., Mont., Nev.,
N.M., Ore., Ut.; Washington: Kamiak Butte,
Whitman Co., 18-VI-1944, Populus tremuloides,
M. M. Furniss and Jianlin Zhou (4 WFBM).
Infesting a 30-cm-diameter recently dead tree
that had no foliage. The bark was necrotic and
had an almond odor. New attacks at a density
of nine per dm occurred at 10-m-height, 11-
cm-diameter. Galleries contained one to two
parent beetles, eggs and first instar larvae.

Procrijphahis utahensis
Hopkins

Biology. — Monogynous. Unstudied, infests
stems of willows, particularly Salix scouleriana.

Distribution and notes. — Canada: B.C.,
Que.; USA: Alas., Calif, Colo., Ida., Ore., S.D.,
Ut.; Washington: Bremerton, Kitsap Co., 26-
VII-1975, Salix scouleriana, M. A. Deyrup
(ABS).

Pseudopityophthonis pubipennis
(LeConte)

Biology. — Monogynous. Infests bole and
branches of Que reus spp. that are felled or
recently dead. Galleries aligned horizontally
across grain, averaging 5 cm long, closely
spaced. Lai^val mines are mainly hidden in the
phloem and oriented longitudinally.

Distribution and notes. — Canada: Soutli-
ern B.C. (Bright 1976); USA: Calif., Ore.;
Washington: Carson, Skamania Co., 18-VI-
1991, Quereus garrayana, M. M. Furniss and
J. B. Johnson (6 WFBM). Infesting a broken,
20-cm-diameter branch on ground.

Pityophthorus alpinensis
G. Hopping

Biology. — Folygynous. Infests broken
branches and twigs of Larix lyallii, apparently
one generation annually.

Distribution and notes. — Canada: Alta.;

USA: Ida., Mont.; WASHINGTON: Harts Pass,
Okanogan Co., ll-VII-1991, Larix lyallii, M. M.
Furniss and J. B. Johnson (3 WFBM). Cadavers
collected from old galleries in dead branches
0.5-2.5-cm-diameter. Galleries were branched
and variable in shape, each branch containing
few (9-11) egg niches; larval mines short,
broad, restricted to phloem; adult brood had
scored the sapwood as if by feeding.

Pityophthorus grandis
Blackman

Biology. — Polygynous, unstudied. Infests
shaded-out branches and young, standing
Pin us ponderosa (Wood 1982).

Distribution and notes. — Canada: B.C.;
USA: Aiiz., Calif, Colo., Nebr, N.M., S.D., Tex.,
Ut.; Washington: Trout Lake, Klickitat Co.,
17-VII-1991, Pinus ponderosa, M. M. Furniss
and J. B. Johnson (4 WFBM). Infesting 4-cm-
diameter standing tree with straw-color
foliage. Umatilla National Forest, 53 km S
Pomeroy, Garfield Co., 19-VII-1991, Pinus
ponderosa, M. M. Furniss and J. B. Johnson (1
WFBM). Reared from stem of a small, felled
tree.

WASHINGTON SCOLYTIDAE

Hylesininae
Hylastini

Scients annectens LeConte

Scierus puhescens Swaine

Hyhirgops porosiis (LeConte)

Hijhirgops rcticiilatits Wood

Hijliirgops rugipcnnis rugipennis (Mannerheim)

Hyhtrgops suhcostulatus subcostulafiis (Mannerheim)

Hylastes gracilis LeConte

Hylastes longicollis Swaine

Hylastes macer LeConte

Hylastes nigrinus (Mannerheim)

Hylastes ruber Swaine

Hylesinini

Hylastiniis obscuriis (Marsham)
Hylesinus califoniiciis (Swaine)
Alniphagiis aspericollis (LeConte)
Alniphagus hirsutus Schedl

Tomicini

Psetidohylesinus dispar pullatiis Blackman
Pseiidohylesiniis granulatus (LeConte)
Pseiidohylesiniis nebulosiis nebiilosus (LeConte)
Pseiidohylesinus nobilis Swaine
Pseitdohylesinus pint Wood
Pseiidohylesiniis sericeus (Mannerheim)
Pseiidohylesiniis sitchensis Swaine
Pseiidohylesiniis tsiigae Swaine
Xylechiniis nwntaniis Blackman
Dendroctonus brevicomis LeConte

340

Great Basin Naturalist

[Volume 55

Di'udroi tonus ponderosac Hopkins
Dendroctoniis pseudotsu<iae 1 lopkins
Dendroctoniis rtifipcnnis (Kirhy)
Dcndructoiiiis valcns LcConte

Phoeotribini

Phlocotrihus Iccontci Stlicdl

Phloeosinini

Phloeosinus cupressi Hopkins
Phloeosinus keeni Bkickman
Phloeosinus punctatus LeConte
Phloeosinus scopulorum scopulonnn Svvaine
Phloeosinus sequoiae Hopkins
Phloeosinus serratus (LeConte)

Hypoboiini

Chaetophloeus heferodoxus (Casey)

Polygraphini

Carphohorus iutcrnicdius Wood
Carphohorus ponderosac Swaine
Carphohorus vandykei Briick
Pohi<j,rap}uts rufipennis (Kirby)

Sc:oLYTi\AE
Scolytini

Scolytiis
Scoh/tiis
Scolytus
Scotytus
Scolytus
Scolytus
Scolytus
Scolytus
Scolytus
Scolyttis
Scolytus
Scolytus

laricis Blacknian
uwnticolae Swaine
multistriatus (Marshani)
opacus Bkickman
oregoni Blackman
piceac (Svvaine)
praeccps LeConte
rugulosus (Miiller)
subscaber LeConte
tsugae (Swaine)
unispinosus LeConte
ventralis LeConte

Micracini

Hylocurus liirtcllus (LeConte)

Crypturgini

Dolurgus puinilus (Mannerlieim)
Crypturgus borealis Swaine

Dryocoetini

Dryocoetcs affaber (Mannerlieim)
Dryocoetes autographns (Ratzebiirg)
Dryocoetcs betulae (Hopkins)
Dryocoetes confusus Swaine

Ipini

Pityogenes earinulatus (LeConte)
Pityogenes fossifrons (LeConte)
Pityogenes knechteli Swaine
Pityokteines elegans Swaine
Pityokteines lasiocarjn (Swaine)
Pityokteines minutus (Swaine)
Pityokteines mystacinus Wood
Pityokteines ornatus (Swaine)
Oiihotomicus caelatus (Eichlioff)
Ips concinnus (Mannerbeim)
Ips emarginatus (LeConte)
//).v integer (Licbboff)
Ips latidens (LeConte)

Ips niexiianus (Hopkins)

//w niontanus (EicbhoO)

Ips perturbatus (Eicbhoff)

Ips pint (Say)

Ips plastographus plustographus (LeConte)

Ips tridens engelnmnni Swaine

Ips tridens tridens (Mannerbeim)

Ips icoodi Tbatcber

Xyloterini

Trypodcndron betulae Swaine
Trypodendron lineatwn (Olivier)
Trypodcndron retusum (LeConte)
Trypodendron rufitarsis (Kirby)

Xyleborini

Xyleborus dispar (Faliricius)
Xyleborinus saxeseni (Ratzebnrg)

Ciypbabni

Trypophloeus salicis Hopkins
Trypopldoeus striatulus (Mannerbeim)
Procryphahis inucronatus (LeConte)
Procryphalus utahensis Hopkins
Cryphalus pubescens Hopkins
Cryphalus ruficoUis Hopkins

Corthylini

Pseudopityophthorus pubipennis (LeConte)
Conophthorus ponclerosae Hopkins
Pifyophthorus alpinensis G. Hopping
Pifyophfhorus confertus Swaine
Pityophthorus confinis LeConte
Pifyophthorus digestus (LeConte)
Pityophthorus grandis Bkickman
Pityophthorus nmrrayanae Blackman
Pityophthorus nitiduhis (Mannerbeim)
Pityophthorus pseudotsiigae Swaine
Pityophthorus tuberculatus EicbbofF
Gnathotrichus retusus (LeConte)
Gnathotrichus stdcatus (LeConte)
Monarthnim scutellare (LeConte)

Acknowledgments

Curatorial help was provided by F. W.
Merickel, University of Idaho. Dr Donald E.
Bright, Jr., Centre for Land and Biological
Resources Research, Ottawa, Canada, identi-
fied the Pityophthorus species except P.
alpinensis. Records of Washington Scolytidae
contained in the Hopkins U.S. numerical file
of the Forest Service, US DA, were provided
by Dr. M. E. McKnight, Forest Sendee, Wash-
ington, DC (retired). The manuscript was
reviewed by Dr. M. A. Deyrup, Archbold
Biological Station, Lake Placid, FL, and b\' Dr.
S. L. Wood, Brigham Yoimg University, Provo,
UT, both of whom provided new records
unknown to us. This is University of Idaho
Agriculture Experiment Station Research
Paper No. 95721.

1995]

New Records of Washington Scolytidae

341

Literature Cited

Bright, D. E., Jr. 1976. The bark Iieetles of Canada and
Alaska. The insects and arachnids of Canada, part 2.
Biosysteniatics Research Institute, Research Board.
Canada Department of Agriculture Pidilication
1576. 241 pp.

Franklin, J. E, and C. T. Dyrness. 1973. Natural vegeta-
tion of Oregon and Washington. USDA Forest
Service General Technical Report PNW-8. Portland,
OR. 417 pp.

FuRNiss, R. L., AND V. M. Carolin. 1977. Western forest
insects. USDA, Forest Service, Miscellaneous Publi-
cation 1339. 654 pp.

Furniss, M. M., and J. B. Johnson. 1987. List of Idaho
Scolytidae (Coleoptera) and notes on new records.
Great Basin Naturalist 47; 375-382.

Furniss, M. M., J. B. Johnson, R. L. Westcott, and T. R.
TORGERSEN. 1992. List of Oregon Scolytidae (Coleop-
tera) and notes on new records. Great Basin Natiualist
52: 373-378.

Cast, S. J., M. M. Furniss, J. B. Johnson, and M. A. Ivie.
1989. List of Montana Scolytidae (Coleoptera) and
notes on new records. Great Basin Naturalist 49:
381-386.

Hilton, D. E J. 1968. A review of the genus Pohjgraphus.
Kansas University Science Bulletin 48(2): 21-44.

Patterson, G. K., and M. H. Hatch. 1945. An annotated
list of the Scolytoidea of Washington. University of
Washington Publication in Biology 10; 145-156.

Petty, J. L. 1977. Bionomics of two aspen l:)ark iieetles,

Tnjpophloeus popitli and Frocnjphaliis mucronatus

(Coleoptera: Scolvtidae). Great Basin Naturalist 37:

105-127.
Reid, R. W 1955. The bark beetle complex associated

with lodgepole pine slash in Alberta. Canadian

Entomologist 87: 311-323.
Stewart, K. W 1965. Obsei-vations on the life history and

habits of Scierus annectens (Coleoptera: Scolytidae).

Entomological Society of America, Annals 58:

924-927.
Wood, S. L. 1971. Family Scolytidae. Pages 395-428 in

M. H. Hatch, The beedes of the Pacific Northwest.

Part V. 662 pp. University of Washington Publica-
tions in Biology. Volume 16.
. 1982. The bark and ambrosia beetles of North and

Central America (Coleoptera; Scolytidae), a ta.xo-

nomic monograph. Great Basin Naturalist Memoirs

No. 6. 1359 pp.
Wood, S. L., and D. E. Bright, Jr. 1992. A catalog of

Scolytidae and Platypodidae (Coleoptera). Part 2;

taxonomic index. Great Basin Naturalist Memoirs

No. 13. 1553 pp.

Received 30 May 1995
Accepted 23 June 1995

RELATIVE VULNERABILITY TO EXTIRPATION OF MONTANE
BREEDING BIRDS IN THE GREAT BASIN

J. Michael Reedl

Abstiuct. — Seventy-four species of montane breeding birds were evaluated for their vulnerability to extiq^ation in
the Great Basin. Although none of these species are endemic to the Great Basin, the montane island system results in a
unique pattern of species associations. Loss of species from these montane communities could be indicative of region-
wide habitat degradation. 1 ranked susceptibility to extiipation based on seven biological variables: geographic range,
population size, reproductive potential, susceptibilitv to cowbird parasitism, migratory status, and diet specialization.
Each variable was weighted equally in its contriliution to \ulnerability, and scores were the sum of trait scores for each
species. Different suites of life-histor>' traits led to similar vulnerabilities. The following 10 montane bird species were
categorized as most vulnerable to extirpation from the Great Basin, listed as most to least vulnerable: Olive-sided
Flycatcher {Contopiis borealis). Painted Redstart (Mijiohorus pictiis), Hammond's Flycatcher (Empidonax luiiiiinondii),
Veery (Cathanis fuscescens). Whip-poor-will {Capriiniil^iii.s vociferiis), Lincoln's Sparrow {Melospiza lincolnii). Black-
backed Woodpecker {Picoides arcticus). Three-toed Woodpecker {P. thdactylus), Himalayan Snowcock (TetraogaUus
liimalayensis), and Nashville Warbler {yennivora ritficapilla). Species of similar vulnerability scores often were dissimilar
in threats related to their vulnerability. No ta.xonomic patterns in vulnerability were found. This type of analysis should
be used proactively to identify vulnerable species or populations and to set priorities for research and management.

Key words: vulnerability, conservation ]morities, avian diversity. Great Basin, montane islands.

Extinction of species worldwide is occur-
ring at a high rate (Stanley 1985). For the most
part, species disappear following habitat loss
(Ehrlich 1988) or after stochastic events elimi-
nate relatively small or isolated populations
(Mac-Arthur and Wilson 1967, Shaffer 1981,
Gilpin and Soule 1986, Rabinowitz et al. 1986,
Reed 1990). Because time, money, and other
resources for species preservation are in short
supply, it is imperative to identify the relative
susceptibility to extinction, or extirpation,
among species to aid in setting conservation
and management priorities.

Extremely vulnerable species often are easy
to identify because of their scarcity, although
sometimes they might be difficult to verify as
extant (Solow 1993). Slightly more common
species, however, often are difficult to classify
by their relative susceptibility to extirpation
even if it varies greatly among species (Rabino-
uitz 1981, Rabinowitz et al. 1986, Reed 1992).
Methods that discriminate among species' sus-
ceptibility to extirpation would be valuable for
setting management priorities. Such methods
exist for selecting geographic areas for conser-
vation based on the number or variety of species
present (e.g., Kirkpatrick 1983, Margules and

Usher 1984, Miller et al. 1987, Scott et al. 1991),
but these methods are not applicable to priori-
tizing conservation efforts among species.

Economic methods can be used to priori-
tize conservation efforts (Bishop 1978, Hyde
1989), but they do not accommodate non-
monetary appraisals of wildlife conservation
goals (Sagoff 1988). The triage method (Myers
1979), whereby species are divided into three
categories based on likely success of conserva-
tion efforts, might not protect the species that
are biologically or anthropocentrically the
most important. In the present analysis, I used
biological traits to determine the relative sus-
ceptibility among species to extiipation.

I analyzed susceptibility to extirpation
(local extinction) of bird species breeding in tlie
semi-isolated montane habitats of the Great
Basin. This is a classic island-biogeographic
system that has been used to test ideas about
extinction and colonization processes (e.g..
Brown 1971, 1978, Johnson 1975, 1978, Behle
1978, Wilcox et al. 1986, Britton et al. 1994).
Although there are no endemic bird species in
the Great Basin, loss of species from diese mon-
tane communities reduces biodiversity and
could be indicative of region-wide problems.

'Biological Hcsoiirces Research C:enter, and Deparliiient of Enviroiiniental and Resource Sciences, University of Nevada, Ren
NV 89512.

lOOU \alle\ Road, Reno,

342

1995]

Montane Bird Vulnerability

343

Furthermore, the naturally fragmented habitat
of the Great Basin montane forest can act as a
model for human-caused fragmentation occur-
ring throughout the world. The 74 species con-
sidered here differ greatly in their life histo-
ries, abilities to colonize, and susceptibility to
extirpation. My goal was to rank species by bio-
logical characteristics related to their vulner-
ability to extirpation, in the anticipation that
the information would be useful for setting
priorities for research, conservation, and man-
agement.

Assessing susceptibility to extirpation in-
volves some type of decision analysis (sensu
Maguire et al. 1987). There are many methods
available for assessing susceptibility to extiipa-
tion, and they vaiy in complexity from simple
classifications to complex multivariate analyses
(Table 1). More importantly, classification meth-
ods differ in their data requirements. Some sys-
tems, such as the lUCN classification scheme
(Mace and Lande 1991), are data intensive,
while others require far less data (Table 1).
The more data available for decision making,
the more certain the results, but it is impor-
tant to chose a method that makes proper use
of the available data. Biological data are rela-
tively scarce for birds in the Great Basin. In
this analysis, I used a method with intermedi-
ate data needs to look at vulnerability to extir-
pation of 74 montane breeding bird species.

Methods

I combined the methods of Burke and
Humphrey (1987), Millsap et al. (1990), and
Rabinowitz et al. (1986) to develop an analysis
appropriate for the species and available data.
This analysis involved assessment using seven

biological characteristics related to persis-
tence ability. Values for each characteristic
ranged from 0 to 1, with higher values associ-
ated with higher susceptibility to extii-pation.
Values for each character were summed to
arrive at a final score of susceptibility to extir-
pation from the Great Basin. All variables had
the same range so that no single character
contributed disproportionately to the suscepti-
bility score (Given and Norton 1993).
Himalayan Snowcock and Ruffed Grouse (sci-
entific names are given later) are introduced
species in the Great Basin (Alcorn 1988). They
were included in the analysis because they are
established in the Great Basin avifauna.
Variable descriptions used in scoring vulnera-
bility to loss from the Great Basin follow.

Geographic range. — Species distributions
were taken from a subset of 20 montane sites
from the Great Basin (Johnson 1975). The con-
tribution of this variable to the vulnerability
score was calculated as 20 minus the number
of ranges on which the species occurs, divided
by 20. This results in a value ranging from 0 to
1.0, with higher values associated with fewer
ranges occupied by the target species, i.e.,
greater vulnerability. Mountain ranges here
and in Table 2 are numbered the same as in
Johnson (1975): 1-Warner, 2-Pine Forest,
3-Santa Rosa, 4-Jarbidge, 5-Raft River,
6-Desatoya, 7-Toiyabe- Shoshone, 8-Ruby,
9-Spruce-S. Pequop, 10-Deep Cr.-Kern,
11-Snake, 12-White-Inyo, 13-Plametto,
14-Grapevlne, 15-Panamint, 16-Spring,
17-Sheep, 18-Mt. Irish, 19-Quinn Canyon-
Grant, and 20-Highland. Distributional data
were supplemented from Behle (1978),
Herron et al. (1985), Ryser (1985), Alcorn

Table L Methods for assessing susceptibility to extirpation and for scoring conservation priorities.

Data

AnaKsis

Method

intensit\'

complexib,'

Citations

Anthropocentric

low

veiy low

the history of the world

Decision analysis:

contingency

low

Rabinowitz 1981, Rabinowitz et al. 1986,
Kattan 1992, Reed 1992

ordinal

variable

low

Burke and Humphre> 1987,
Millsap etal. 1990,' this study

classical

variable

medium

Maguire et al. 1987

multivariate

variable

high

Given and Norton 1993

Economic

variable

Bishop 1978, Hyde 1989

Viability analysis

high

Kinnaird and OBrien 1991, Boyce 1992

lUCN

\ en- high

high

Mace and Lande 1991

344

Gkeat Basin NAXURyVLisT

[Volume 55

Table 2. Additions to Johnson's (1975) original l)ii(l dis-
tributions. Site numbers are the same as those used b\
Johnson (1975) and are listed in Methods. Scientifie
names are listed in Talile 3.

Species

Sites added

American Wigeon

<s\

Northern Goshawk

3",5", 71', 101^251 '.SOB

Sharp-shinned Hawk

16"

Flamnuilated Owl

3H, 4^;, 7H

Northern Pygnn-owl

16H

Short-eared C)\\ 1

3", 8H, 16", 25"

Northern Saw-wliet Owl

5'5,6", 11",23",31"

Calliope Huinniingbird

Hammond's Flycatcher

8\ lOB

Olive-sided Flycatcher

5B, lOB

Steller's Jay

Red-breasted Nuthatch

3\ 8\ 31A

White-breasted Nuthatch

Golden-crowned Kinglet

Ruby-crowned Kinglet

10^

Water Pipit

Solitary' Vireo

Orange-crowned Warbler

23A

Nashville Warbler

Lincoln's Sparrow

5B,8A

White-crowned Sparrow

lOB

Himalayan Snowcock

■V-Ucom (1988), Bfiehle (1978), '-■S. Dunham (iiiipiihlisliecl data limi. l.Rcdins;
bird .surveys in 1993), "Herron et al. (198.5)

(1988), and S. Dunham (unpublished data
from breeding bird surveys in 1993).

Population somewhere large? — The
other component to the relative rarity of a
species is its local abundance. This variable is
used regularly in rarity studies and generally
is used subjectively, often because detailed
numbers are not available (Rabinowitz et al.
1986). Because of this, I use it subjectively as
well. The definition of "common" varies by
taxa. For example, carnivores typically are less
common than insectivores at a given site
(Brown and Maurer 1987). Therefore, using a
single numeric criterion above which a species
is considered "common" will result in a system-
atic bias in vulnerability scores even if no sys-
tematic bias exists in susceptibility to extiipation.
Therefore, for a given species, if anywhere in
the Great Basin there is a population that is
"common" for its taxon, I have given the species
a score of 0. Data for this assessment came
from Ryser (1985), Alcorn (1988), and breed-
ing bird surveys done during the breeding
season by myself or my students.

Habitat specialization. — Habitat special-
ists (score value = 1) were defined as species
that exist in Nevada only in montane habitat.
Limited use outside montane habitat, such as

mi.xed forest or riparian areas, scored .50, and
relative generalists were given a value of 0.
Data on habitat use came from Ehrlich et al.
(1988).

Susceptibility to cowbird parasitis.vi. —
Although there is extensive literature on the
potential effects of Brown-headed Cowbirds
{Molothnis (iter) on passerine reproduction (e.g.,
Mayfield 1977, Brittingham and Temple 1983),
the problem has been little studied in the Great
Basin. Brown-headed Cowbirds are seen at
high elevations in the Great Basin (Fleischer
and Rothstein 1988, Fleischer personal com-
munication), particularly associated with
riparian habitat and human impact, including
cattle grazing. Also, cowbirds can range far from
foraging areas in order to brood parasitize
(Rothstein et al. 1984). However not all species
are susceptible to parasitism from cowbirds,
and some parasitized species are unaffected
(e.g., precocial species). I scored species based
on my expectation of potential impact ol cow-
bird parasitism. Only passerines that are open-
cup nesters could score above 0; Corvidae
received 0. Species known to reject cowbird
eggs were given a score of 0. Large hosts that
did not eject eggs (larger than a female
Brown-headed Cowbird; >39 g; Dunning
1993) were given a value of 0.5. Species that
are small and not known to reject eggs were
assigned a score of 1. Data came from Fried-
man (1971), Rothstein (1975), Airola (1986),
Marvil and Cruz (1989), and Briskie et al.
(1992).

Migr\tory STATUS. — There is some contro-
versy regarding relative costs of migration ver-
sus residency in birds. However, because
migrants are dependent on habitats in more
than one geographic area, I consider them more
vulnerable than nonmigrants. I scored migra-
tory status as no latitudinal migration = 0
(lowest risk), migrates primarily to U.S. = .25,
migrates primariK' to Middle or South America,
winters in nonforest = .50, winters in sec-
ondan' forest = .75, winters in mature forest
= 1.0.'

Reproductive potential. — I considered
reproductive potential to be the anticipated
ability to recover from a population crash and
based it on the first age of reproduction, clutch
size, and number of broods within a year (data
from Ehrlich et. al. 1988). I classified repro-
ductive potential based on an index. The index
was the mean clutch size times the number of

1995]

Montane Bird Vulnerability

345

broods in a year, divided by the age of first
reproduction. With this index, a species that
breeds repeatedly, at an early age, and with
large clutches will have a low score. When no
data were available for number of broods, one
brood was assumed. Age at first breeding was
assumed to be one for small birds, unless data
from the literature indicated otherwise. The
relationships between the index, reproductive
potential, and risk value were made arbitrarily
and are presented in Table 3. Data and refer-
ences associated with this calculation for each
species can be obtained from the author.

Diet specialization. — Information on diet
breadth came from Ehrlich et al. (1988), and
species were classified as generalists (score =
0), moderate specialists (0.5), or specialists
(1.0) based on diet described there. This
assessment was subjective, based on number
of food types typically in the diet and foraging
method used.

With this system, vulnerability scores could
range from 0 to 7, with 7 being the greatest
probability of extirpation from the Great
Basin. One variable not included in the analy-
sis that is important in biological risk to extir-
pation was local population trends. Local pop-
ulation trends were omitted because they are
generally unknown for nongame birds in the
Great Basin. Local endemism should be con-
sidered in scoring as well, but the Great Basin
has no endemic bird species. Another variable
that has been suggested as a risk to sunaval is
ground nesting. Traditional thought places
ground nesters at higher risk to predation than
off-ground nesters (e.g., Ricklefs 1969,
Slagsvold 1982, Collias and Collias 1984).
However, in a reanalysis of the data, Martin
(1993) found that ground nesters were not dis-
proportionately susceptible to depredation.
Given this important ambiguity, nest location
was omitted from the analysis.

Results and Discussion

There were 41 additions of various mountain
ranges to breeding bird distributions (Tiible 2).
The 74 breeding bird species used in this
analysis, their associated scores for each life-
histoiy trait, and their vulnerability scores are
listed in Table 4. Taxonomy follows the con-
vention of the American Ornithologists' Union
(1983). Vulnerability scores ranged from 0.60
for the American Robin (scientific names are

Table 3. Reproductive potential and its relationship to
risk score. The index is mean clutch size times the numher
of broods in a year, divided by the age of first reproduction.

Index

Reproductive

Risk

value

potential

score

<L5

very low

1.0

1.5-2.9

low

0.75

3-5.9

medium

0.50

6-11.9

medium-high

0.25

>11.9

high

found in Table 4) to 5.70 for the Olive-sided
Flycatcher and Painted Redstart. None of the
variables alone was sufficient to assess vulner-
ability to extirpation. This has been seen by
others (e.g., Burke and Humphrey 1987) and
is due to other life-history factors affecting
susceptibility to extir^Dation (Arita et al. 1990).
Therefore, range and density estimates alone
cannot be used to assess vulnerability to extir-
pation. Another problem with using range and
density as the only criteria for extiipation risk
is that slice-in-time assessments of rarity can
give misleading results due to natural fluctua-
tions in distribution and population size (Hanski
1985). Species ranges expand and contract, and
population densities can undergo large fluctu-
ations annually, even in long-lived species such
as birds. Therefore, being uncommon does
not, de facto, make a species vailnerable to extir-
pation; in contrast, being common does not
assure continued presence (e.g., the Passenger
Pigeon [Ectopistes migratorhis]; Bucher 1992).

Passerines tended to rank as more suscepti-
ble to extirpation than other orders, primarily
because one threat, vulnerability to cowbird
parasitism, did not impact non-passerines.
Unlike some earlier studies of birds (Terborgh
and Winter 1980, Kattan 1992), I found no tax-
onomic pattern in susceptibility to extiq^ation.
The 10 species with the highest vulnerability
score come from seven families in four orders.
There are several likely explanations for this.
The first is that no inherent patterns exist.
Alternatively, a true taxonomic pattern in
extirpation proneness might exist for Great
Basin birds but was missed because of incom-
plete data, because of a subsampling effect
(not enough of tlie Great Basin surveyed), or be-
cause tlie anah'sis considers only cuiTcnt species
(implying that extirpation-prone species are
gone).

Many species with similar or identical vul-
nerabilit\' scores were vulnerable for different

346

Great Basin Naturalist

[Volume 55

Table 4. Data used in analyses and \ulncial)ilit\' scorinj^s; variable definitions given in text. Higher values indieate
higher susceptibility to extiipation from the (Ireat Basin.

Vulner-

(Criteria

Some-

Habitat

Diet

Species

ability

where

special-

Cowbird

,\ligratoi"y

Reproductive

special-

score

Range

large?

ization?

problem?

status

potential

ization

Canada Goose

(Branta canadensis)

L9()

.90

.25

.75

Green-winged Teal

(Anas crecca)

2.90

.90

.25 •

.25

American Wigeon

(A. americana)

2.90

.90

.25

Canvasback

(Aythya valisineria)

2.90

.90

.25

Sharp-shinned Hawk

(Accipiter striatus)

4.00

..50

.75

Northern Goshawk

(A. gentilis)

.3.10

.60

.25

.75

Hiniala\an Snowcock

(TetraogaUits hiinahiycnsis)

4.20

.95

.75

Blue Grouse

{Dendragapus ohscunis)

2.75

..50

.25

Ruffed Grouse

(Bonusa umbellus)

3.20

.95

.25

Mountain Quail

(Oreortyx pictus)

L90

.65

.25

Common Snipe

(GalUnago gallinago)

2.70

.70

.50

Flammulated Owl

{Otus flammeolus)

3.05

.55

.50

Northern Pygmy-owl

[Glaucidium gnoma)

3.30

.80

.50

Short-eared Owl

[Asia flammeus)

2.55

.80

.25

.50

Northern Saw-whet Owl

{Aegolius acadicus)

2.85

.60

.25

.50

Common Nighthawk

{Chordeiles minor)

2.10

.35

.50

.75

Whip-poor-will

(Caprimidgus vociferus)

4.70

.95

.75

Calliope Hummingbird

{Stellula calliope)

3.15

.65

.75

Broad-tailed Hummingbird

(Salaspliorus platycercus)

2..30

.05

..50

.75

Lewis' Woodpecker

{Melanerfjes lewis)

1.90

.90

.25

Yellow-bellied Sapsucker

{Sphyrapicus varius)

2.55

.30

.75

.50

Red-breasted Sapsucker

(S. ruber)

3.15

.85

.75

.50

Williamson's Sapsucker

(S. thyroideus)

3.35

.55

.75

.50

Downy Woodpecker

(Picoides pubescens)

2.10

.60

.50

Hairy Woodpecker

[E villosus)

2.00

.50

White-headed Woodpecker

(P. albolarvatus)

3.45

.95

.50

Black-backed Woodpecker

{P. arcticus)

4.45

.95

.50

Three-toed Woodpecker

{P. tridactylus)

4.45

.95

.50

Olive-sided Flycatcher

(Contopus borealis)

5.70

.45

.75

.50

Hammond's Flycatcher

{Empidonax haintnondii)

5.45

.70

.75

.50

Dusky Flycatcher

(£. oberholseri)

3.30

.05

.75

.50

Western Flycatcher

(£. difficilis)

3.95

.45

.50

Horned Lark

[Eremophih alpestris)

2.60

.85

.25

.50

1995]

Montane Bird Vulnerability

347

Table 4. Continued.

Vulner-

Criteria

Some-

Habitat

Diet

Species

ability

where

special-

Cowbird

Migratoiy

Reproductive

special-

score

Range

large?

ization?

problem?

status

potential

ization

\'iolet-gieen Sw;illo\\

(Taclujcineta thalassina)

3.00

..50

.50

Gray Jay

(Perisoreus canadensis)

2.95

.95

.50

Steller's Jay

(Cyanocitta stelleri)

2.10

.60

..50

Clark's Nutcracker

{Nucifraga cohnnbiana)

1.65

.15

..50

Mountain Chickadee

{Panis gambeli)

1.50

Red-breasted Nuthatch

(Sitta canadensis)

2.15

.40

.25

.50

White-breasted Nuthatch

(S. carolinensis)

2.10

.10

Pygmy Nuthatch

(S. pygmaea)

1.95

.70

.25

Bro\\Ti Creeper

(Ceiihia americana)

1.65

.40

.25

.50

American Dipper

(Cinclus mexicamis)

3.25

.50

.25

Golden-crowned Kinglet

{Regidus satrapa)

2.40

.65

.25

Ruby-crowned Kinglet

{R. calendula)

2.65

.15

.75

.25

Western Bluebird

(Sialia niexicana)

2.55

.80

.25

.50

Mountain Bluebird

(S. curntcoides)

1.60

,10

.25

Townsend's Solitaire

(Myadestes townsendi)

3.00

.25

.50

V'eeiy

(Catharus fuscescens)

4.90

.90

.50

Swainson's Thrush

(C. ustidatus)

3.60

.60

1»

.50

Hemiit Thrush

(C. guttatus)

2.55

.05

.75

.25

American Robin

{Turdtis migratorius)

0.60

.10

.25

Water Pipit

(Anthus spinoletta)

3.65

.90

.25

.50

Solitary' Vireo

(Vireo solitarius)

3.55

.30

.25

Orange-crowned Warbler

{Vennivora celata)

2.60

.35

.75

.50

Nashville Warbler

(V nificapilla)

4.15

.90

.75

.50

N'irginia's Warbler

{V. virginiae)

3.25

.25

.75

.25

Vellow-rumped Warbler

{Dendroica coronata)

2..30

.05

.50

.25

Grace's Warbler

(D. graciae)

4.05

.80

1»

.50

.25

MacGillivray's Warbler

{Oporomis tohniei)

3.35

.35

.50

Wilson's Warbler

(Wihunia pusilla)

3.85

.85

.50

Painted Redstart

(Myioboms pictus)

5.70

.95

1»

.75

.50

Western Tanager

(Piranga hidoviciana)

3.15

.15

.50

Green-tailed Towhee

(Pipilo chlorunis)

1.75

.50

.25

Fox Sparrow

(Passerella iliaca)

1.95

.45

.25

Lincoln's Sparrow

{Melospiza lincolnii)

4.60

.85

.50

.25

348

Great Basin Naturalist

[Volume 55

Table 4. ContiiuR'd.

Vulner-

Criteria

Some-

Habitat

Diet

Species

ability

where

special-

('owbird

.\1 iterator)

Repr

oductive

special-

score

Range

largc-'^

ization?

lirobleni:'

status

teulial

ization

White-crowned Sparrow

{Zonotrichia Icucoph njs )

2.00

..50

.25

Dark-eyed Jiiiico

(Junco hyemalis)

2.05

.05

.25

Gray-crowned Rosy Finch

{Leucosticte tephrocotis)

.3.70

.95

1''

.25

.50

Black Rosy Finch

(L. atrata)

.3.50

.75

.25

.50

Cassin's Finch

(Caq)oilacus (•(i.s.sinii}

2..50

.25

Red Crossbill

(Loxiii cunimstra }

3.15

.40

.25

.50

Pine Siskin

{Carduelis pinus)

2.40

.40

.25

Evening Grosbeak

{Coccothraustes vespcrtUais]

2.35

.85

..5

..5-'

.25

''Assumed to not eject Browii-lieaded Coubird eggs

suites of threats to persistence. That is, some
equal scores were made up of low values for
one or more characteristic and corresponding-
ly high values for other traits, which balanced
in the ranking. This observation is consistent
with Rabinowitz's (1981, Rabinowitz et al.
1986) observations of plant species' rarity in
Great Britain. It should be noted tliat this analy-
sis refers to species loss in the Great Basin and
does not reflect species-wide vulnerability.

This type of analysis is sensitive to the num-
lier of variables included. Adding or deleting
characters from the analysis would change
scores. For example, if ground nesting were
decisively shown to increase vulnerability, it
could be added to tlie analysis and would change
relative scores. Results also would be altered
if the characteristics were weighted differently.
I did not weight any variable as more impor-
tant than another because of the lack of data
that demonstrates the validity of weighting
particular traits over others. Arbitrarily assign-
ing different weights in the absence of inde-
pendent data supporting the weighting would
result in unwarranted bias in the vulnerability
scores.

The results presented are not absolute rank-
ings for susceptibility to extirpation because
data are incomplete and more threats might
become apparent, which would have to be
added to the analysis. Validity of these results
depends entirely on reliability of the data used
and how representative the 20 mountain ranges
are of the rest of the Great Basin. There is a
dearth of distributional and life-historv data

on many Great Basin birds. Therefore, my
results should be taken as a guide for detailed
local studies of species and their surrounding
communities. Results of these studies can
then be used to develop proactive manage-
ment plans.

Vulnerability Ranks and Management

Vulnerability to extirpation and manage-
ment priorities are not equal. Scores based
strictly on biological variables ignore homo-
centric values, such as hunting or local tradi-
tional uses. For example, the top 10 vulnerable
species in this analysis include only one hunt-
ed species (an introduced one at that), though
others were scored. In addition, how a given
rank comes about can affect management pri-
orities. There are four ways a species can have
a high score, and they should be interpreted
differently for management.

(a) High score occurs when the Great Basin
is within the greater bounds of a species' dis-
tribution and local declines have reduced a
species range and population sizes in the Great
Basin. These species are probably declining
because of local problems, and in this analysis
might include Mountain Quail and Northern
Goshawk. Specific management plans should
be enacted to increase population numbers,
sizes, and distributions.

(b) High score occurs when the Great Basin
is within the greater bounds of a species distri-
bution, and the species is declining through-
out its range. Problems could be occurring on
the breeding grounds, wintering grounds, or

1995]

Montane Bird Vulnerability

349

migratory routes. If the cause of decline is
known and can be improved through local
management, then this should be done. If the
cause of the decline is known, but occurs out-
side the Great Basin, then I would recom-
mend monitoring populations but not making
any management efforts. If the cause of the
decline is not known, as for many Neotropical
migrants, gather information to determine
whether or not local management could
improve local or region-wide population con-
ditions. If management efforts are suspected
to work, implement them with proper controls
and follow-up work. If no effect is found, dis-
continue management.

(c) High score occurs partly because the
Great Basin is at the edge of a species distri-
bution, thus limiting its local distribution and
population sizes. Of the top 10 scored species
in this analysis, five have Nevada as part of
their distributional boundaiy This is possibly
tlie trickiest categoiy for management. Species'
ranges fluctuate, and population declines
might be range retractions having nothing to
do with local conditions. These species should
be monitored because range retraction might
be an early indicator of a species-wide decline
(e.g., Laymon and Halterman 1987). However,
it can also indicate local problems that require
local management solutions. These species
need further investigation.

(d) High score occurs when species has de-
clined severely (thus reducing its range and
commonness) but is recovering. Continue exist-
ing management efforts, if any, and monitor
populations to make sure recoveiy continues.
If it does not, these species belong in one of the
other three sub-categories.

In all instances involving management plans,
efforts should be made to set up proper stud-
ies or experiments to ascertain the limiting
factor(s) and the coiTcct method(s) for counter-
acting the problem (MacNab 1983, Gavin 1989,
1991, Muiphy and Noon 1992). This includes
monitoring suitable control sites. Without
using adequate experimental design, it will
not be possible to ascertain the effectiveness
of management efforts. Low-score species
should still be monitored and management
plans developed. Low-score species are those
that are closest to recovery or those not threat-
ened and thus have potential for the quickest
success from management.

Acknowledgments

I thank J. A. R. Alberico, E E Bmssard, D. A.
Delehanty, C. Elphick, N. Johnson, B. Maurer,
and one anonymous reviewer for commenting
on this manuscript, and K. Reed and S. Dunham
for help summarizing the data. I also thank
G. Henon, R. Hamlin, M. Elpers, T. Baron, and
P. Zenone for discussions regarding threat
variables. This work was supported by NSE
grant DEB-9322733, the Biological Resources
Research Center at the University of Nevada,
the U.S. Forest Sei-vice, the Center for Conser-
vation Biology at Stanford University, and a
donation from the Wells Family Foundation.
This is Contribution No. 004 of the Nevada
Biodiversity Initiative.

Literature Cited

AlROLA, D. A. 1986. Brown-headed Covvbird parasitism
and habitat disturbance in tlie Siena Nevada. Journal
of VVildhfe Management 50: 571-575.

Alcorn, J. R. 1988. The birds of Nevada. Fairview West
Pubhshing, Fallon, NV 418 pp.

American Ornithologists' Union. 1983. Check-list of
North American birds. 6th edition. Allen Press,
Lawrence, KS. 877 pp.

Arita, H. T, J. G. Robinson, and K. H. Redford. 1990.
Rarity in Neotropical forest mammals and its ecolog-
ical correlates. Consei^vation Biology 4: 181-192.

Behle, VV. H. 1978. Avian biogeography of the Great
Basin and Intermountain Region. Great Basin
Naturalist Memoirs 2: .55-80.

Bishop, R. C. 1978. Endangered species and uncertainty:
the economics of the safe minimum standard.
American Journal of Agricultural Economics 60:
10-18.

Bovce, M. S. 1992. Population viability analysis. Annual
Review of Ecology and Systematics 23: 481-506.

Briskie, J. v., S. G. Sealy, and K. A. Hobson. 1992.
Behavioral defenses against avian brood parasitism
in sympatric and allopatric host populations. E\'olu-
tion 46: 334-340.

BRirriNGHA.vi, M. C., and S. A. Temple. 1983. Have cow-
birds caused forest songbirds to decline? BioScience
33: 31-35.

Britton, H. B., P E Brussard, D. D. Murph\; .and G. T.
Austin. 1994. Colony isolation and isozyme variabil-
ity of the western seep fritillar>', Speijeria nokomis
apacheana (Nymphalidae) in the western Great Basin.
Great Basin Naturalist .54: 97-105.

Brown, J. H. 1971. Mammals on mountaintops: nonequi-
librium insular biogeography. American Naturalist
105: 467-478.

. 1978. The theory of insular biogeography and the

distribution of boreal birds and mammals. Great
Basin Naturalist Memoirs 2: 209-227.

Brown, J. H., and B. A. Maurer. 1987. Evolution of
species assemblages: effects of energetic constraints
and species dynamics on the diversification of the

350

Great Basin Naturalist

[Volume 55

North American avilaima. Anicrican Naturalist 130:
1-17.

BUCHER, E. H. 1992. The causes oT e.xtinctiou of the
Passenger Pigeon. Current Ornitholog)' 9: 1-36.

Burke, R. L., and S. R. Humphrey. 19(S7. Rarity as a cri-
terion for endangennenf in Florida s fauna. ()r\x 21:
97-102.

CoLMAS, N. E., AND E. C. CoLLUS. 1984. Nest building
and bird behavior Princeton Universit\' Press, Prince-
ton, NJ.

Dunning, J. B., Jr. 1993. CRC: handbook of" avian body
masses. CRC Press, Boca Raton, FL.

Ehrlich, P R. 1988. The loss of diversit)': causes and con-
sequences. Pages 21-27 in E. O. Wilson, editor.
Biodiversity. National Academy Press, Washington,
DC. '

Ehrlich, P R., D. S. Dobkin, and D. Wheye. 1988. The
birder s handbook: a field guide to the natural histo-
ly of North American birds. Simon & Schuster, New
York. 785 pp.

Fleischer, R. C., and S. I. Rothstein. 1988. Known sec-
ondary contact and rapid gene flow among sub-
species and dialects in the Brown-headed Cowbird.
Evolution 42: 1146-1158.

Friedman, H. 1971. Fnrther information on the host rela-
tions of the parasitic cowbirds. Auk 88: 239-255.

Gavin, T. A. 1989. What's wrong with the questions we
ask in wildlife research? Wildlife Society Bulletin
17: 345-350.

. 1991. Why ask "Why": the importance of evolu-

tionaiy biology in wildlife science. Journal of Wild-
life Management 55: 760-766.

Gilpin, M. E., and M. E. Soule. 1986. Minimum viable
populations: processes of species extinction. Pages
19-34 in M. E. Soule, editor, Conservation biology:
the science of scarcib,' and diversity. Sinauer, Sunder-
land, MA.

Given, D. R., and D. A. Norton. 1993. A multivariate
approach to assessing threat and for priority setting
in threatened species consei-vation. Biological Conser-
vation 64: 57-66.

Hanski, I. 1985. Single-species spatial dynamics may con-
tribute to long-term rarity and commonness. Ecology
66: 335-343.

Herron, G. B., C. A. Mortimore, and M. S. Ravvlings.
1985. Nevada raptors: their biology and manage-
ment. Biological Bulletin No. 8. Nevada Department
Wildlife, Reno. 14 pp.

Hyde, W F 1989. Marginal costs of managing endangered
species: the case of the Red-cockaded Woodpecker
Journal of Agricultural and Economic Research 41:
12-19.

Johnson, N. K. 1975. Controls of the number of bird
species on montane islands in the Great Basin. Evolu-
tion 29; 545-567.

. 1978. Patterns of avian biogeography and specia-

tion in the Intermountain Region. Great Basin
Naturalist Memoirs 2: 137-160.

Kattan, G. H. 1992. Rarit>' and \'ulnerability: the birds of
the Codillera Central of Colombia. Conservation
Biology 6: 64-70.

Kinnaird, M. E, and T. G. O'Brien. 1991. Viable popula-
tions for an endangered forest primate, the Tana River
crested mangabey (Cercocehus galeritiis fooler if us).
Consei-vation Biology 5: 203-213.

Kirkpatrick, J. B. 1983. An iterative method for establish-
ing priorities for the selection of nature reserves: an

example from Tasmania. Biological Consenation 25:
127-134.

L.\Y.MON, S. A., and M. D. Halterman. 1987. Can the
western subspecies of the Yellow-billed Cuckoo be
saved from extinction'? Western Birds 18: 19-25.

Mac;Arthur, R. H., and E. O. Wilson. 1967. The theorv'
of island biogeography. Monographs in Population
Biology No. 1. Princeton Universit)' Press, Princeton,
NJ. 203 pp.

M.^CE, G. M., AND R. Lande. 1991. As.sessing extinction
threats: toward a reevaluation of lUCN threatened
species categories. Consei-vation Biology 5: 148-157.

MacNab, J. 1983. Wildlife management as scientific ex-
perimentation. Wildlife Society Bulletin 11: .397-401.

Ma(;uire, L. a., U. S. Seal, and R F Brus,sard. 1987. Man-
aging critically endangered species: the Sumatran
rhino as a case study. Pages 141-158 in M. E. Soule,
editor. Viable populations for conservation. Cam-
bridge Universitv' Press, United Kingdom.

Margules, C. R., and M. B. Usher. 1984. Conservation
evaluation in practice. I. Sites of different habitats in
north-east Yorkshire, Great Britain. Journal of Envi-
ronmental Management 18: 153-168.

Martin, T E. 1993. Nest predation among \egetation lay-
ers and habitat types: revising the dogmas. American
Naturalist 141: 897-913.

Marvil, R. E., and a. Cruz. 1989. Impact of Brown-
headed Cowbird parasitism on the reproductive suc-
cess of the Solitaiy Vireo. Auk 106: 476-480.

Mayfield, H. F 1977. Brown-headed Cowbird: agent of
extermination? American Birds 31: 107-113.

Miller, R. I., S. P Gratton, and P S. White. 1987. A
regional strategy for reserve design and placement
based on an analysis of rare and endangered species'
distribution patterns. Biological Conservation 39:
255-268.

MiLLSAP, B. A., J. A. Gore, D. R. Runde, and S. I.
Cerulean. 1990. Setting priorities for the consei-va-
tion of fish and wildlife species in Florida. Wildlife
Monograph No. 111.

Murphy, D. D., and B. R. Noon. 1992. Integrating scien-
tific methods with habitat conservation planning:
reserve design for Northern Spotted Owls. Ecolog-
ical Applications 2: 3-17.

Myers, N. 1979. The sinking ark. Pergamon Press, Oxford.

Rabinowitz, D. 1981. Seven forms of rarity. Pages
205-217 in H. Synge, editor, The biological aspects
of rare plant consei-vation. Wiley, Chichester

Rabinowitz, D., S. Cairns, and T. Dillon. 1986. Seven
forms of rarity and their frequency in the flora of the
British Isles. Pages 182-204 in M. E. Soule, editor.
Conservation biology: the science of scarcity and
diversity. Sinauer Associates, Sunderland, MA.

Reed, J. M. 1990. The dynamics of Red-cockaded
Woodpecker rarity and conservation. Pages 37-56 in
A. Carlsson and G. Aulen, editors, Conservation and
management of woodpecker populations. Swedish
University of Agricultural Science Report 7, Uppsala.

. 1992. A system for ranking conservation priorities

for Neotropical migrant lairds based on relative sus-
ceptibility to extinction. Pages 524-536 in J. M.
Hagan and D. W. Johnston, editors, Ecology and
conservation of Neotropical migrant landbirds.
Smithsonian Institution Press, Washington, DC.

RicKLEFS, R. E. 1969. An analysis of nesting mortality in
birds. Smithsonian Contributions to Zoolog)' 9: 1—48.

1995]

Montane Bird Vulnerability

351

RoTHSTEIN, S. I. 1975. An experimental and teleonomic
investigation of avian hrood parasitism. Condor 77:
250-27 L

RoTHSTEiN, S. I., J. Verner, .■\.\d E. Stevens. 1984. Radio-
tracking confirms a unique diiunal pattern of spatial
occurrence in the parasitic Brown-headed Cowhird.
EcologN' 65: 77-88.

Ryser, E a., Jr. 1985. Birds of the Great Basin; a natural
histor\'. University of Nevada Press, Reno.

S.\GOFF, M. 1988. Some problems with environmental
economics. Environmental Ethics 10: 55-74.

Scott, J. M., B. Csuti, and S. Caicco. 1991. GAP analy-
sis: assessing protection needs. Pages 15-26 in W.
Hudson, editor. Landscape linkages and biological
diversit\': a strategy' for sundval. Island Press, Covello,
CA.

Shaffer, M. L. 1981. Minimum viable populations for
species consenation. BioScience 31: 131-134.

Slagsvold, T. 1982. Clutch size variation in passerine
birds: the nest predation hvpothesis. Oecologia 54:
159-169.

SoLOW, A. R. 1993. Inferring extinction from sighting
data. Ecology 74: 962-964.

Stanley, S. M. 1985. Extinction as part of the natural evo-
lutionary process: a paleobiological perspective.
Pages 31-46 in R. J. Hoage, editor. Animal extinc-
tions: what everyone should know. Smithsonian
Institution Press, Washington, DC.

Terborgh, J., AND B. Winter. 1980. Some causes of
extinction. Pages 119-133 in M. E. Soule and B. A.
Wilcox, editors. Conservation biolog}': an evolutionary-
ecological perspective. Sinauer, Sunderland, MA.

Wilcox, B. A., D. D. Murphy, P R. Ehrlich, and G. T.
Austin. 1986. Insular biogeography of the montane
butterfly faunas in the Great Basin: comparison with
birds and manmials. Oecologia 69: 188-194.

Received 27 October 1994
Accepted 19 May 1995

GRASSHOPPER DENSITIES ON GRAZED AND UNGRAZED RANGELAND
UNDER DROUGHT CONDITIONS IN SOUTHERN IDAHO

Denni.s |. Melcliii^' and Merlyii A. Brusven'-

AUSTKACT. — Low-dt'nsit\ grasshopper populations were sampled at 15 pairs of rangeland sites in south central
Idaho. One site of each pair had not been grazed by livestock for at least 10 \ ears. Grazed sites were managed under
normal grazing regimes established by the Bureau of Land Management.

Mean grasshopper density was higher on ungrazed sites than on grazed sites. Proportions u{ ML'lano})his san^uinipes
were higher on ungrazed sites than on grazed sites and were higher on annual grasslands than on otiier \egetation types.
Effects of grazing appeared to be independent of vegetation type.

Proportions of Gomphocerinae, a subfamily of grasshoppers that feeds almost exclusively on grasses, were affected
by \'egetation t\pe, but not grazing. Crested wheatgrass seedings supported the highest proportions of Gomphocerinae.
Proj^ortions of Oedipodinae were affected by grazing and vegetation type. Higher proportions of Oedipodinae were
found on grazed sites than on ungrazed sites, and on sagebrush/grass sites than on annual grasslands. Results indicate
that li\ estock grazing during drought conditions tends to reduce grasshopper populations on southern Idaho rangeland.

Keij words: Oiihoptcm, Acrididac. Melanoplus sanguinipes, liccstock grazing, drought, population density, range
management.

Grasshoppers are frequently the most abun-
dant arthropods, in terms of biomass, in the
intermountain sagebrush ecoregion of the
western United States. As primaiy consumers
they may be important in energy and nutrient
cychng, and, at outbreak densities, they compete
with Hvestock and wildhfe for forage. Because
of their ecologic and economic importance,
the potential effects of range management
practices on grasshoppers are a concern to
those interested in the health of rangeland
ecosystems. Several studies have addressed
the role of livestock grazing on grasshopper
populations (Coyner 1938, Smith 1940,
Campbell et al. 1974, Holmes el al. 1979,
Capinera and Sechrist 1982, Jepson-Innes and
Bock 1989, Quinn and Walgenbach 1990,
Miller and Onsager 1991). Onsager (1987) sug-
gested that there is probably geographic varia-
tion in grasshopper responses to grazing
among rangeland types and their constituent
grasshopper communities. To date no studies
have investigated the relationship between
livestock grazing and grasshopper densities on
rangelands in the intermountain region.

This study, conducted during years of below-
normal precipitation and low grasshopper
densities, examined differences in grasshopper
densities between rangeland under normal

livestock grazing regimes administered by the
Bureau of Land Management and rangeland
that had not been grazed for at least 10 years.

Study Area

The study area is located southeast of Sho-
shone, ID, within the Bureau of Land Manage-
ment's (BLM) Shoshone District, between
longitude 114°30' and 114°00' W and latitude
42°37.5' and 43°00' N. This area receives an
average of about 26 cm of precipitation annu-
ally, most of it between October and May.
Average annual temperature is about 9.0 °C.

The intermountain sagebrush ecoregion
was subjected to heavy grazing pressure in the
late 19th and early 20th centuries, frequent
fires, and subsequent invasion by cheatgrass
and other e.xotic plant species (Pickford 1932,
Stewart and Hull 1949, Mack 1981, Yensen
1982). As a result, stands of grazing-intolerant
native grasses were greatly diminished over
much of the region, and cheatgrass has become
the dominant species on more than 40 million
ha of the Intermountain West (Mack 1981,
Fellant and Ilall 1994). The present vegetation
widiin die stud\' area consists primarih' of cheat-
grass, Broniis tectorum L., with sagebrush,
Artemisia tridentata wyomingensis Beetle &

' Dcpartnitiil i>t Ph.iit. Soil ciml Knliiniolomcal ScicncfS, University ol Iclulio. Moscow, ID 83843.
-.\utlior Ic) wliniii coi rfspondrucc sliould lie acklivssed.

352

1995]

Grasshoppers and Livestock Grazing

353

Young and A. tridentata tridentata (Rydb.)
Beetle, where it has not burned reeently. As of
1988, about 23% (ca 40,000 ha) of the study area
consisted of crested wheatgrass, Agropijron
cristatum (L.) Gaertn., plantings (USDI-BLM
1984, 1990).

Materials and Methods

Ungrazed sites were selected on the basis of
grazing history (not grazed for at least 10 years),
size (at least 16 ha), and shape (at least 100 m
across the narrowest dimension). Fifteen
rangeland sites were found within the study
area that met these criteria. Most sites were
isolated tracts fenced to exclude livestock and
to provide habitat for upland game birds.
Grazing by wildlife within the ungrazed tracts
was negligible. Black-tailed jackrabbits were
not abundant during the years in which sam-
pling took place, and populations of prong-
horn antelope, the only other large vertebrate
herbivore present in the summer, are quite low
and widely dispersed across the study area, ca
300 individuals over 180,000 ha (J.' Russell,
USDI-BLM, personal communication).

A grazed site was selected to match each
ungrazed site for a total of 30 sites. In most
cases grazed sites were adjacent to, and
shared a boundary with, ungrazed sites. For
six sites adjacent matched pairs were not pos-
sible; consequently, grazed sites were chosen
within 2 km. All grazed sites matched the
ungrazed sites in soil type, topography, vege-
tation, slope, and aspect. All grazed sites were
located within BLM -administered grazing
allotments. Stocking rates for the grazed sites
varied from 1.9 to 2.8 ha/AUM (USDI-BLM
1990). Not all grazed sites were grazed each
year, as prescribed by rest-rotation grazing
management plans.

Elevation of the sites ranged from 1180 to
1320 m. Five pairs of sites were located on
areas replanted with crested wheatgrass, four
pairs were on annual grassland sites having lit-
tle or no sagebrush, and six pairs were located
on sagebrush-grass sites.

Grasshopper and vegetation sampling. —
Grasshoppers were sampled on 19 July-7
August 1990 (adult stage), 21-28 June 1991
(primarily nymphal stage), and 2-13 August
1991 (adult stage). Corresponding sites of a
grazed and ungrazed pair were always sam-
pled on the same day. Grasshopper density on

a site was estimated by counting the number
of grasshoppers flushed from 50, O.l-m^ rings
5 m apart in a circular transect (Richards and
Waloflf 1954, Onsager and Heniy 1977). Species
composition on a site was determined by a
"flush-capture method (Capinera and Sechrist
1982). Thirty to 100 specimens were captured
and identified at each site on each sampling
date by slowly walking in a circular transect
and, to avoid bias toward more conspicuous
species, counting only those grasshoppers en-
countered directly in the path of the obsei-ver

Vegetation was sampled on the same dates
as the grasshopppers by visually estimating
the percentage ground cover in 5% incre-
ments by plant species in each of 40, 0.1-m^
square quadrats in a circular transect. Plant
species unidentifiable in the field were col-
lected and identified later. The percent cover
of cryptogams, cattle dung, and bare ground
was also estimated. Vegetation data from the
three sampling dates were combined for sub-
sequent analyses.

Data analysis. — We classified the sites into
three vegetation tyi3es based on dominant vege-
tation on a site. Sites that had been seeded to
crested wheatgrass were categorized as re-
planted. Sites with sagebrush as the dominant
plant species were placed in the sagebrush
category, and the remaining sites, dominated
by cheatgrass without significant sagebrush
cover, were categorized as annual grasslands.

Differences in percentage ground cover
among vegetation types were confirmed with
a Kruskal-Wallis non-parametric one-way
analysis of variance (Zar 1984). Non-paramet-
ric statistical tests were used with the ground
cover data because of the large number of zero
values involved. Gomparisons between vege-
tation types were made with a non-parametric
analog of Tukey's test (Dunn 1964, Zar 1984).

Because paired sites were in close proximi-
ty and of similar vegetation, we used Wilcoxon
paired-sample tests to identify differences in
percentage ground cover between grazed and
ungrazed sites.

Grasshopper densities were too low to con-
duct meaningful statistical comparisons sepa-
rately for all species. Accordingly, analyses were
conducted on densities of total grasshoppers, on
proportions of Melanoplus sanguinipes (the major
pest species in the region), and on proportions
of the \hree subfamilies of Acrididae within the
region, Melanoplinae, Gomphocerinae, and

354

Great Basin Naturalist

[Volume 55

Oedipodinae. Grasshopper densities were
transformed by log^,(x + 1) to nornuili/.e the
data. The aresin transformation was apphed to
the proportions (Zar 1984).

Three-way analysis of varianee (PROG
GLM, SAS Inst.) was used to determine the
significance of sampling date, vegetation type,
and grazing treatment effects on grasshopper
density and proportions. For the ANOVA,
sites were not blocked by location; i.e., pairing
was ignored. Gomparisons among vegetation
types were made with least significant differ-
ence mean separation tests (PROG GLM, SAS
Inst.). Paired-sample t tests were used for
comparisons between grazing treatments.

Trends between habitat characteristics and
grasshopper densities/proportions were evalu-
ated by Spearman rank correlations (PROG
GORR, SAS Inst.). Mean values from the three
sampling dates were used for the correlation
analyses.

Results

The replanted vegetation type was domi-
nated by crested wheatgrass and also had the
greatest amount of bare ground (Table 1).
Annual grasslands were dominated by annual
plant species, primarily cheatgrass (Table 1).
Sagebrush sites had the greatest shrub cover,
although other vegetation types had small
amounts of sagebrush (Table 1). Annual grass-
lands and sagebrush sites had little perennial
grass cover, other than Poa sandbergii.

Grazing treatment did not greatly affect
most ground cover variables (Table 2), but sites
grazed by livestock had more bare ground and
cattle dung and less total vegetative cover and
perennial grass cover than the permanently
ungrazed sites.

Twenty-three species of grasshoppers were
indentified from the 30 sites. Melanoplus san-
guinipes was found at all 30 sites and repre-
sented 36% of all grasshoppers on the study
sites. Other common species included Oeda-
leonotus enigma (Scudder), Ageneotettix deonun
(Scudder), Aulocara eUiotti (Thomas), Conozoa
sidcifrons (Scudder), and TracJiyrachys kiowa
(Thomas). No species other than M. san-
guinipes comprised more than 10% of all
grasshoppers from all sites. Total density of
grasshoppers on the three sampling dates
ranged from <().2 to 2.6 per m^.

Table L Median (s.d.j percentage ground cover by veg-
etation t\pe.

Vegetation t) pe

Ground c()\'er
variables

Replanted

(.V = 10)

Sagebrush

(.V = 12)

Annual
grasslands

(A' = 8)

Annual grasses

1.2bi

(3.7)

7.8b
(8.1)

2().0a

(9.0)

Annual Forbs

0.5a

(1.6)

1.2b

(3.9)

3.1b

(2.8)

A^ropynm
ciisfafiDn

16.8a

(7.2)

O.Ob

(0.02)

O.Ob

(0.03)

Poa sandbergii

6.9a

(4.9)

.5.0a

(1-7)

12.0a

(5.4)

Other perennial 0.1a 0.6a 1.1a

grasses (0.4) (2.9) (2.8)

Sagebrush

O.Ob

(2.8)

13.0a

(5.0)

O.Ob

(2.4)

Total vegetation 26.1a 32.9a 41.8b

(8.6) (5.7) (8.9)

Cn'ptoganis 1.8a 7.0b 3. lab

(3.2) (5.9) (4.3)

Bare ground 40.,5a 24.()ab IS.Ob

(13.7) (6.5) (S.3)

'Meiisures within rows followed by different letters are significantly different,
P < .05, non-parametric analog of Tukey's test (Dunn 1964, Zar 1984).

Total density of grasshoppers was affected
by sampling date and grazing treatment (Table
3). No differences in density were detected
among vegetation types (LSD mean separa-
tion, P = .05; Table 3, Fig. 1).

Proportions of M. sangidnipes and all species
within the subfamily Melanoplinae were
affected by sampling date, vegetation type,
and grazing treatment (Table 3). Annual grass-
lands had the highest proportions of A/, san-
guinipes and of all species within the subfami-
ly Melanoplinae (LSD mean separation, F =
.05; Fig. 1). The proportion of grasshoppers
within the subfamily Gomphocerinae was
strongly affected by sampling date and vegeta-
tion type, but not grazing (Table 3). Replanted
(crested wheatgrass) sites had the highest pro-
portions of Gomphocerinae (LSD mean separa-
tion, P = .05; Fig. 1). Proportions of grasshop-
pers within the subfamily Oedipodinae were
significantly affected by sampling date, vegeta-
tion i\npe, and grazing, although F-values were
not as great as for proportions of the other
subfamilies (Table 3). The mean proportion of

1995]

Gr.\sshoppers and Livestock Gfl\zing

355

Table 2. Comparison of median (s.d.) ground cover
between grazed and ungrazed sites.

Grazing tre

atment

Grazed

Ungrazed

Annual grasses

10,0

(S.5)

7.0

(12.3)

Annual forbs

1.0

(1.9)

1.5

(3.8)

AJI perennial grasses*

3.0

(6.4)

4.3

(11.3)

Poa sandhergii

8.7
(5.6)

7.1

(4.9)

Sagebrush

3.0

(7.1)

3.0

(6.8)

Total vegetation*

32.5

(7.6)

39.9

(6.3)

Cattle dungi

0.7

(0.5)

0.0

(0.0)

Cryptogams

4.6

(5.7)

6.2

(6.3)

Bare ground**

31.0

(13.4)

23.0

(10.2)

^Because no cattle dung was recorded on tlie ungrazed sites, no statistical test

of significance was perfomied.

*Measures are signiflcantlv different (F < .05, Wilcoxon paired-sample test,

N = 15),

**Measures are signiflcantK- different (P < .(II, Wilcoxon paired-sample test,

IV = 15).

Total grasshopper density was not conelated
with any ground cover variables. Proportion of
M. sanguinipes was negatively correlated with
percentage bare ground and cover of perenni-
al grasses {r^ = -.59 and -.62, respectively, N
= 30, P < .001), and was positively correlated
with percentage ground cover of cheatgrass
and annual forbs (r^ = .41 and .42, respective-
ly, N = 30, P < .05). Proportion of all
Melanoplinae combined was correlated posi-
tively with cheatgrass (r^ = .52, N = 30, P <
.01) and negatively with perennial grasses and
percentage bare ground (-.70 and -.64,
respectively N = 30, P < .001).

As proportions of Melanoplinae declined
with increasing cover of perennial grasses and
bare ground, proportions of other species in-
creased. Gomphocerinae showed trends oppo-
site those of the Melanoplinae. Proportions of
Gomphocerinae were correlated positively
with perennial grasses and bare ground (.66
and .46, respectively, iV = 30, P < .01) and
negatively with cheatgrass and annual forbs
(-.52 and -.42, respectively N = 30, P < .05).
Proportions of Oedipodinae were not signifi-
cantly correlated (P > .05) with any of the
ground cover variables.

Discussion

Oedipodinae was greater on sagebrush sites
than on annual grassland sites (LSD mean sep-
aration, P = .05; Fig. 1).

The effect of grazing treatment was consis-
tent across vegetation types and sampling dates
for all grasshopper variables (Table 3). Because
no significant interactions between date and
grazing, or vegetation type and grazing, were
detected (Table 3), comparisons of grazing
treatments were made across all vegetation
types and sampling dates.

Overall density of grasshoppers was greater
on ungrazed than on grazed plots (paired-sam-
ple t test, P < .001; Fig. 2). Proportions of Af.
sanguinipes and of all species of Melanoplinae
combined were higher on the ungrazed sites
(paired-sample t test, P < .005; Fig. 2). Oedi-
podinae showed a trend opposite that of the
Melanoplinae, being found in greater propor-
tions on the grazed sites (paired-sample / test,
P < .001; Fig. 2). Proportions of Gomphocerinae
were not affected by grazing (paired- sample t
test, P> 0.10; Fig. 2).

Grazing influenced both total density and
species composition of grasshoppers. Members
of the subfamily Melanoplinae accounted for
most of the increase in total density on un-
grazed sites. Although vegetation did not affect
density in this study, it strongly influenced
species composition. Relative abundance of
Gomphocerinae increased, and Melanoplinae
decreased, with increasing coverage of peren-
nial grasses and bare ground, while total num-
bers of grasshoppers remained the same.
Proportions of Af. sanguinipes, the primaiy pest
species in the region, were negatively associ-
ated with grazing, perennial grasses (primarily
crested wheatgrass), and percentage bare
ground.

Habitat preferences of A/, sanguinipes, or any
organism, represent an integrated response to
many stimuli. Short-term changes in habitat
due to grazing may include reduced quantities
of food, less escape space, increased amounts
of bare ground, altered host plant quality, and
changes in microhabitat temperature and

356

Grkat Basin Natuiulist

[Volume 55

T.-VBLE 3. Summary of Type III F-values (and sijiiiifieaiice knt'ls) from three-way ANOVA for densities of total
grasshoppers, Melanoplus sanguinipes, Melanoplinae other tluin M. saii^iiinipcs, Gomphocerinae, and Oedipodinae.

SoiMce

d.f

CJrasshopper
(lensit\7m-

Percentage
Melanoplus
sanguinipes

Percentage
Melanoplinae

Percentage
Gomphocerinae

Percentage
Oedipodinae

Date (D)

12.9 (<.0 1)

18.7 (<.01)

11.9(<.01j

21.3 (.01)

6.1 (<.01)

Vegetation t>'pe (V)

1.2 (.30)

19.3(<.01)

29.3 (<.01)

22.2 (<.01)

5.6(<.01)

Cirazing (G)

5.6 (.02)

13.4 (<.01)

11.7(<.0])

0.5 (.50)

8.8(<.01)

V X D

2.1 (.09)

2.2 (.07)

2.8 (.03)

1.9 (.13)

1.2 (.34)

G X D

0.6 (.58)

0.3 (.77)

0.1 (.88)

2.1 (.13)

0.7 (.48)

V X G

0.7 (.49)

0.1 (.89)

0.1 (.87)

0.8 (.43)

0.9 (.43)

D X V X G

0.3 (.91)

0.6 (.69)

0.5 (.76)

0.2 (.96)

0.2 (.9.3)

Replanted

Sagebrush

Annuals

Oedipodinae
M. sanguinipes

^^ Gomphocerinae
li-:s»il Other Melanoplinae

Fig. 1. Mean density of Oedipodinae, Gomphocerinae, Melanoplinae other than Melanoplus sanguinipes, and M. san-
guinipes by vegetation type.

humidiW. Differences in plant species compo-
sition between grazing treatments were mini-
mal, indicating that long-term alteration of the
plant community composition was probably
not a factor.

Results of this study are consistent with
some previous studies. Proportions of M. san-
guinipes were negatively coirelated with crested
wheatgrass in this study. Fielding and Brusven
(1992) demonstrated that crested wheatgrass
is not a prefeired host plant for M. sanguinipes in

southern Idaho. Perennial grasses are favored
food plants for Aiiloeara eUiotti (Pfadt 1949,
Fielding and Brusven 1992), the most common
gomphocerine in the study area. Proportions
of M. sanguinipes were also negatively corre-
lated with percentage bare ground. Nerney
and Hamilton (1969) and Kemp and Sanchez
(1987) reported that M. sanguinipes avoids ovi-
position in bare soil, whereas A. elliotti prefers
to oviposit in bare ground (Kemp and Sanchez
1987, Fisher 1992). High percentages of bare

1995]

Gr.\sshoppers and Livestock Gr.\zing

357

Grazed

Ungrazed

Oedipodinae
M. sanguinipes

^^ Gomphocerinae
I I Other Melanoplinae

Fig. 2. Mean density' of Oedipodinae, Gomphocerinae, Melanoplinae other than Melanoplus sanguinipes, and M. san-
guinipes b>' grazing histoiy

ground were associated with both grazing and
crested wheatgrass seedings in the present
study (Tables 1, 2).

Previous studies that examined grazing
effects on grasshoppers reported results similar
in some respects to those reported here. On the
short-grass prairie of Colorado, Capinera and
Sechrist (1982) reported that Oedipodinae were
most abundant on the most heavily grazed pas-
tures, while lightly grazed pastures supported
the highest total grasshopper densities. Quinn
and Walgenbach (1990) found Melanoplinae,
particularly Melanoplus sanguinipes, to be dom-
inant on ungrazed sites on mixed-grass prairies
of South Dakota, even though total grasshop-
per abundance was less than on grazed sites.
However, Miller and Onsager (1991) were
unable to detect any effect of different grazing
regimes on adult grasshopper populations,
including M. sanguinipes, in a crested wheat-
grass pasture in Montana.

Our obsen'ations were made under condi-
tions of low grasshopper density and drought
in southern Idaho. Fielding and Biaisven (1990)
showed that grasshopper population density
in southern Idaho was positively correlated
with precipitation. Results of this study cannot
be extrapolated to predict how grasshopper

populations will respond to livestock grazing
during more favorable years when rangeland
productivity is high and grasshopper popula-
tions are rapidly expanding.

Rangeland grasshoppers have traditionally
been viewed solely as destructive rangeland
pests. However, in an ecosystem context they
may have net beneficial worth during most
years as an important food source for at least a
pait of the life cycle of many species of mam-
mals, birds, and reptiles. Results presented here
may serve as a cautionary note regarding range-
land ecosystem management under drought
conditions. Grazing during years of drought
and low grasshopper populations could con-
ceivably add to the stress experienced by
insectivorous animals by reducing available
food resources (i.e., grasshopper populations),
especially if other arthropods serving as alter-
nate foods are also at low densities.

Acknowledgments

We thank the staff of the ELM Shoshone
District office for their support. J. A. Onsager,
M. A. Quinn, and L. P Kish reviewed earlier
versions of the manuscript. William Price, sta-
tistical research associate, Universitv of Idaho,

358

Great Basin NATuii\LiST

[Volume 55

advised on statistical matters. Research was
flmded in part by Bureau of Land Management
as Cooperative Agreement No. 919-CA7-05
and published with the approval of the director
of the Idaho Agricultural Experiment Station
as Paper No. 92731.

Lite MTU RE Cited

Campuell, J. B., W. H. Arnett, J. D. Lambley, O. K. Jantz,
AND II. Knltson. 1974. Grasshoppers (Acrididae) of
the Flint Hills native tall grass prairie in Kansas.
Kansas State University Agricnltural Experiment
Station Research Paper 19. 147 pp.

Capinera, J. L., AND T. S. Sechrist. 1982. Grasshopper
(Acrididae)-host plant associations: response of
grasshopper popnlations to cattle grazing intensity.
Canadian Entomolgist 114: 1055-1062.

Coyner, W. R. 19.38. A report of the effect of overgrazing
on the Acrididae. Proceedings of the Oklahoma
Academy of Science IS: 83-85.

Dunn, O. J. 1964. Multiple contrasts using rank sums.
Technometrics 6: 241-252.

Fielding, D. J., and M. A. Brusven. 1990. Historical
analysis of grasshopper (Orthoptera: Acrididae)
population responses to climate in southern Idaho,
1950-1980. Environmental Entomology 19:
1786-1791.

. 1992. Food and habitat preferences oiMelanoplus

sanguinipes and Aulocara elliotti (Orthoptera:
Acrididae) on disturbed rangeland in southern
Idaho. Journal of Economic Entomology 85:
783-788.

Fisher, J. R. 1992. Location of egg pods of Aidocara
elliotti (Orthoptera: Acrididae) in a field of crested
wheatgrass in Montana. Journal of Kansas Ento-
mology Societ)' 65: 416-420.

Holmes, N. D., D. S. Smith, and A. Johnston. 1979.
Effect of grazing by cattle on the abundance of
grasshoppers on fescue grassland. Journal of Range
Management 32: 310-311.

Jepson-Innes, K., and C. E. Bock. 1989. Response of
grasshoppers (Orthoptera: Acrididae) to livestock
grazing in southeastern Arizona: differences be-
tween seasons and subfamilies. Oecologia 78: 430-431.

Kemp, W. R, and N. E. Sanchez. 1987. Differences in
post-diapause thermal requirements for eggs of two
rangeland grasshoppers. Canadian Entomologist
119: 653-661.

Mack, R. N. 1981. Invasion of Broinus tectorum L. into
western North America: an ecological chronicle.
Agro-ecosystems 7: 145-165.

Miller, R. H., and J. A. Onsager. 1991. Grasshopper
(Orthoptera: Acrididae) and plant relationships
imder different grazing intensities. Environmental
Entomology 20: 807-814.

Nerney, N. j., and a. G. Hamilton. 1969. Effects of rain-
fall on range forage and populations of grasshoppers,
San Carlos Apache Indian Reservation, Arizona.
Journal of Economic Entomology 62: 329-.333.

Onsager, J. A. 1987. Current tactics for suppression of
grasshoppers on range. Pages 60-66 in J. A. Onsager,
editor. Integrated pest management on rangeland:
state of the art in the sagebrush ecosvstem. USDA-
ARS, ARS-50.

Onsager, J. A., and J. E. Henry. 1977. A method for esti-
mating the density of rangeland grasshoppers
(Orthoptera: Acrididae) in experimental plots. Acrida
6: 231-237.

Pellant, M., and C. Hall. 1994. Distribution of two exotic
grasses on intermountain rangelands: status in 1992.
In: Proceedings — Symposium on ecology, manage-
ment, and restoration of intermountain annual range-
lands. USDA-FS, Intermountain Research Station,
General Technical Report INT-GTR-313.

Pfadt, R. E. 1949. Food-plants, distribution, and abun-
dance of the big-headed grasshopper, Aulocara elliotti
(Thos.). JouiTial of the Kansas Entomological Societ\'
22: 69-74.

PiCKFORD, G. D. 1932. The influence of continued heav>
grazing and of promiscuous burning of spring-fall
ranges in Utah. Ecology 13: 159-171.

QuiNN, M. a., and D. D. Walgenbach. 1990. Influence
of grazing history on the community structure of
grasshoppers of mixed-grass prairie. Environmental
Entomology 19: 1756-1766.

Richards, O. W, and N. Waloff. 1954. Studies on the
biology and population dynamics of British grass-
hoppers. Anti-locust Bulletin 17. London. 182 pp.

Smith, C. C. 1940. The effects of overgrazing and erosion
upon the biota of the mixed grass prairie of Okla-
homa. Ecology' 21: 381-397.

Stewart, G., and A. C. Hull. 1949. Cheatgrass in south-
ern Idaho. Ecologv' 30: 58-74.

USDI-BLM. 1984. Proposed monument resource man-
agement plan and final environmental impact state-
ment. USDI-Bureau of Land Mangement, Wash-
ington, DC.

. 1990. Monument rangeland program smnmaiy

Progress report. USDI-Bureau of Land Manage-
ment Shoshone District Office, Shoshone, ID.

Yensen, D. 1982. A grazing histoiy of southwestern Idaho
with emphasis on the Birds of Prey Study Area.
Research project report. United States Department
of the Interior, Bureau of Land Management, Boise,
ID. 82 pp.

Zar, j. H. 1984. Biostatistical analysis. 2nd edition.
Prentice-Hall, Inc., Englewood Cliffs, NJ.

Received 30 March 1994
Accepted 10 April 1995

PLANT NOVELTIES IN LEPIDIUM (CRUCIFERAE) AND
ARTEMISIA (COMPOSITAE) FROM THE UINTA BASIN, UTAH

Stanley L. Welsh* and Sherel Goodrich^

Abstr.'VCT. — Named as new tiixa are Lepidiwn hiiberi Welsh & Goodrich, sp. nov., and Artemisia nova A. Nels. var.
duchesnicola Welsh & Goodrich, var no\'. The taxa are provided with diagnoses and descriptions, and their relation-
ships, provenance, and hahitats are discussed.

Key icords: Lepidium huberi, Artemisia nova var. duchesnicola, new faxa, Utah, Uinta Basin.

Noted historic pioneer Utah botanist Marcus
Eugene Jones (1852-1934) has been quoted,
perhaps apocryphally, as saying that he felt
sorry for all future generations of botanists
because so few plants remained for them to
describe and name. Whedier die quote is tnae
or not, the generations beyond Jones' time
have not suffered from a shortage of areas of
botanical inquiiy, including the discoveiy and
naming of scores of plants new to science, and
there are indications that future generations of
botanists beyond the 1990s will continue to
find and describe novelties. The flora is not
yet fully understood.

The Uinta Basin harbors numerous narrow
endemics in many genera of plants, due in
some part to the availabilit>' of unique geologi-
cal substrates. Geomorphological processes
have, through time, exposed geological strata
of vaiying and diverse composition around the
periphery of the basin and onto the slopes of
mountains and plateaus that form its borders.
The basin proper is the result of uplift during
and following the Laramide Revolution, which
resulted in a topographically low area south of
the Uinta Mountains and north of the Tavaputs
Plateau. The exposed strata vaiy in age fiom the
present into the remote Precambrian epoch.
Revealed are mud and siltstones, shales, sand-
stones, limestones, and quartzites of enormous
total thickness, each displayed in sequence
like pages from a book. Some of the strata,
especially the shales and mud and siltstones,
weather into fine-textured, salt-laden sub-
strates, others into sand and gravel, and still
others into platy shales. Each of the substrates
presents a different array of texture, salinity.

trace elements, and other features important
to plant growth. Plants have become adapted
to the peculiarities of salt content or its lack,
to textural differences, and to the peculiarities
of water relationships. During the past several
millions of years formations have been exposed
and cut by erosional processes, and during that
same period floras have developed. Varying
attributes of the resulting erosional surfaces
have allowed the evolution of present floras of
the basin. Some Uinta Basin plant endemics
are directly correlated to geological formations
and are aligned along the strike of formations
as though planted mechanically by some gigan-
tic drill. In odiers die coiTclation is more subtle,
but most endemics show some affinity to par-
ticular formations.

The present paper deals with two more
narrowly restricted Uinta Basin endemics. Bodi
of them have been known in collections since
the 1980s.

Lepidium huheri
Welsh & Goodrich, sp. nov.

Similis Lepidio inontano var. ohjssioides in
habitu generali amplitudine sed foliis praecipue
caulinis (foliis basalibus nullis vel evolutis
debiliter) in basim lignosam et in siliculam
amplitudinam difiPert.

Plants subshrubs, the stems woody at the
base, ashy or brownish, 1-2.5 dm long; branches
puberulent throughout, green, 15-53 cm long;
leaves all cauline, the eophylls reduced, prin-
cipal lower leaves 2-3.5 cm long, 8-20 mm
wide, pinnatifid, 5- to 7-lobed, the lobes often
again lobed or dentate, smaller and entire
upwards; panicles 3-10 (14) cm long, branches

^Departnient of Botany and Life Science Museum. Brigliam Young Universih; Pro\o, UT 84602.
^U.S. Forest Service, Ashlev National Forest, .3.5.5 N. Vernal Avenue. Vernal, UT 84078.

359

360

Great Basin Naturalist

[Volume 55

corymboscly arranged; pedicels 2-4.5 nnii
long, puberulent; sepals glabrous, 1.4-1.9 mm
long, oval, green, the margin white; petals
white, 2.3-2.6 mm long, 1.8-2.2 mm wide,
shallowlv incised, the stvlc 0.4-0.8 mm long
(Fig. 1). '

Suffrutices, caules lignei ad basin, cinerei
vel brunneis, 1-2.5 dm long; rami puberulenti
onuiino, virides, 15-53 cm longi; folia totus
caulina, eoph\ His reductis, principalibus infer-
nis foliis 2-3.5 cm longis et 8-20 nun latis,
pinnatifidis, 5- to 7-lobatis, lobi plerumque
lobati vel dentati, parvascens et integra sur-
sum; paniculae 3-10 (14) cm longae, ramis
corymbose dispositis; pedicelli 2-4.5 mm
longi, puberulenti; sepala glabra, 1.4-1.9 mm
longa, ovales, virides, marginibus latis albis;
petala alba 2.3-2.6 mm long, unquibus 1 mm
longis; staminalis filamenti glabra; siliculae
glabrae, ovatae vel ovales, ca 2.3-2.6 mm lon-
gae, 1.8-2.2 mm latae, incisura vadosa, styli
0.4-0.8 mm longi.

Type:— USA: Utah: Uintah County, grow-
ing under ledges of Park City (Phosphoria)
Formation, above Weber Sandstone, T2S,
R21E, S15 NEl/4, Big Brush Creek Gorge,
Uinta Mountains, adjacent to black sage-
brush/grass community, west exposure, at
2179 m elev., A. Huber 2400, 18 August 1994
(Holotype BRY; isotypes to be distributed).
Additional collections: USA: Utah: Uintah
County, 8 km N of Maeser, at base of Taylor
Mountain, S. Goodrich, 1548, 13 August 1973;
do, TIS, R20E, S30, Ashley Creek, drainage N
of Sims Peak, 30 km N of Vernal, at 2959 m, D.
Atwood 9128a, 30 July 1982; do, T2S, R21E,
S34, SW/SW, N side of Red Mts., ca 16 km N
of Vernal, at ca 2320 m, J. Tuhy 2693, 31 July
1986; T2S, R21E, S14 NW/SW, Big Brush
Creek Gorge, A. Huber 858, 13 June 1994; do,
A. Huber 880, 14 June 1994; do, T3S, R21E,
S34, SW/SW, N slope of Red Mountain, ca 6.5
km NW of Steinaker Resei-voir, A. Huber & S.
Goodrich 2390, 18 Aug. 1994; do, T3S, R21E,
S3 NEl/4, Red Mt., ca 6 km NW of Steinaker
Reservoir, A. Huber & S. Goodrich, 2392, 18
Aug. 1994; do, T2S, R21E, S34 SW/SE, Red
Mt., ca 6 km NW of Steinaker Reservoir, A.
Huber & S. Goodrich 2393, 18 August 1994
(all BRY, with numerous duplicates to be dis-
tributed).

There is a collection, apparently of this,
taken from Moffat Countv, CO (R. C. & K. W.
Rollins 8387, off countv roads 13 and 789, S of

Fig. 1. Habit sketch (a) and silick- (b) oi Lcpidiu)n huhch
Welsh & Goodrich.

Hamilton), at BRY. Its main difference is the
merely toothed unlobed leaves. Pinnately
lobed leaves are featured prominently in the
material from the range of the species in
Uintah Count)'.

The following key will serve to distinguish
L. hiiheri from other members of the L. )non-
tanuiu complex.

1995]

Lepidium AND Artemisia, Uinta Basin

361

1. Plants slightK if at all woocK' above the base,

biennial to perennial herbs; silicles 2.8-4.1 mm

long, 2.1-2.5 mm wide

L. inontanum sens. lat.

— Plants wood}' well abo\'e the base, long-lived
perennial siibshrubs; silicles various 2

2(1). Silicles 4.5-7.5 mm long, 5.2-6.5, obovate;
plants of the Moha\'e desert region of SW Utah
and southward L. fronontii

— Silicles 2.3-2.6 mm long, 1.8-2.2 mm wide;
plants montane, in Uintah County, Uttih

L. hiiheri

This taxon, a definite subshrub, differs from
L. montanum Nutt. sens. lat. in about the same
degree and manner that the Mohavaean
desert L. frcinontii Wats, differs from that
species complex, i.e., in degree of woodiness
and in size of the silicles, which in L. freviontii
are on the large size for that complex and in L.
hiiberi are smaller. Members of the montanum
complex are widely distributed in the
American West and occur in an array of mor-
phological races, many of which are geograph-
ically or edaphically correlated. Hitchcock
(1936) treated 13 infraspecific taxa, some of
which are now regarded at specific rank. The
phase of the L. montanum complex that is
apparently most closely allied to montane L.
huheri is the extralimital van ahjssoides (Gray)
Jones, to which early collections of this novel-
ty were assigned. That variety, which ranges
widely from Colorado to New Mexico,
Arizona, and Texas, sometimes has a branching
subligneus caudex, but is seldom if ever sub-
shrubby, and lacks the other morphological
features of L. huheri. The spatially and eleva-
tionally isolated var. spathulatum (Robinson)
C. L. Hitchc, also an ally, is rather common in
Uintah County and elsewhere in eastern Utah.
It is a tall plant, apparently biennial or short-
lived perennial, with a single stem from the
base, the caudex not woody or much branched.
It is most common at low elevations along
drainages, growing with sagebrush. Most
phases within the montanum complex have
been regarded at specific rank in the past, and
there is more than marginal justification for so
treating them in the future. Justification for
regarding L. huheri at specific rank involves
its combination of morphological characters,
i.e., long-lived perennial habit, ligneus base,
deeply lobed lower cauline leaves, and small
silicles.

Lepidium huheri grows in sand or silty
sands derived from formations of various age

from the Shinarump Member of the Chinle,
Park City, and Weber Sandstone, all on the
south-plunging flank of the Uinta Mountains.
It occurs in black sagebrush, mountain brush,
ponderosa pine, lodgepole pine, and spruce-
fir communities at 2225 to 2960 m elevation.

Artemisia nova A. Nels. var. dtichesnicola
Welsh & Goodrich, var nov.

Persimilis Artemisia nova A. Nels. in mag-
nitudine et habitu sed in folius pilis albis
dense non-glanduliferis et in floribus gener-
aliter 5 (raro 4) et bracteis plus numerosis
(10-20, nee 8-12) differt (Fig. 2).

Shrubs, 1-3 (5) dm tall, main branches
spreading, vegetative stems 1-3 dm long
(rarely more); flowering stems mainly 1.5-3 (4)
dm long; leaves dimorphic, 0.3-2 cm long,
those of old stems shallowly to deeply 3- to 5-
lobed or -toothed, lobes or teeth rounded,
cuneate basally, appressed white canescent
and not punctate; inflorescence narrowly pan-
iculate, seldom more than 3 cm wide; involu-
cres 3.1-5.8 mm long, 1.4-3.4 mm wide, cylin-
dric to narrowly campanulate; bracts 10-20,

Fig. 2. Drawing of floral head of Artemisia nova A. Nels.
\'ai-. dtichesnicola Welsh &c Goodrich.

362

Great Basin Naturalist

[Volume 55

white canescent, the margin hyahne; flowers 5
(rare!)' 4), all perfect; receptacle glabrous; ach-
enes glabrous.

Type.— USA: Utah: Uintah County, T5S
R20E S5 NEl/4, 16 km W of Vemal, 1710 m
elevation, desert shrub community, on heavy,
reddish clay of the Duchesne Ri\ (m- Formation,
S. Goodrich 23215, 17 Sept. 1990 (holotype
BRY; isotypes to be distributed).

Additional specimens. — USA: Utah:
Uintah Count>', T5S R19E NW 1/4 S2, along
Hw^ 121, 3 km E of Lapoint, 1740 m eleva-
tion, Neese et al. 11013, 19 Sept. 1981; do,
T5S R19E S2, 3 km NE of Lapoint, along
Hwy 121, 1665 m, on red silty clay of the
Duchesne River Formation, S. Goodrich
22225, 5 Sept. 1986; do, TIN RIE S26 SEl/4
USiM, 0.6 km SE of Tridell, 1720 m, heavy
clay of the Duchesne River Formation, S.
Goodrich 23212, 17 Sept. 1990; do, T5S R19E
S2 NVVl/4, 3 km E of Lapoint, 1720 m eleva-
tion, on heavy, reddish clay of the Duchesne
River Formation, S. Goodrich 23214, 17 Sept.
1990; do, T3S R19E S35 El/2 SLM, about 1.4
km N of Hwy 121 between Lapoint and
Maeser, red clays of Duchesne River Fomiation,
1800 m elevation, S. Goodrich 23255, 27 Sept.
1990 (all BRY, with numerous duplicates to be
distributed).

This taxon differs from typical A. nova A.
Nels. in the densely white pubescent outer in-
volucral bracts and generally denser pubescence
of leaves and flowering stalks, and in the lack
of conspicuous glandular dots on leaves.
Leaves are not the green to lead-gray color
typical of most populations of van nova, most
of which also have glandular dots. There are,
however, a few known populations of var nova
that lack glandular dots, l>ut they possess the
lead-gray to green color. In var. duchesnicola,
the dense white, or silveiy, pubescence of leaves
that lack glands is diagnostic. Additionally,
mature involucres of var. duchesnicola are less
lustrous, the number of involucral bracts is
greater on the average (8-12 in var. nova,
10-20 in var. chichesnicola), and the flower
number is almost uniformly 5 (not 3-8 as in
var. nova). Practically all other features of the
variety proposed herein are similar to var.
nova.

The proposed new variety would key in
Welsh et al. (1993) to A. arhuscula Nutt. From

that species var. duschesnicola can be distin-
guished by its relatively shorter flowering
stems, uniformly three-lobed vegetative
leaves, much larger number of involucral
bracts (10-20, not 4-8), and uniformly 5-flow-
ered heads (not 4-9).

The following key, modified from Welsh et
al. (1993) will aid in identification of this taxon
and its near allies.

1. Inflorescence open-paniculate, commonly more
than 2 cm wide; plants often more than 5 dm
tall A. tridentata var wijo)ningensis

— Inflorescence narrowly paniculate, commonly
less than 2 cm wide; plants usually less than

5 dm tall 2

2(1). Plants commonly 3-5 dm tall (sometimes taller);

involucral bracts 4-8 A. arhuscula

— Plants commonly 3 dm tall or less; involucral
bracts averaging more than 8 3

3(2). Vesture of plants silverx' white; involucral bracts

10-20 A. nova var duchesnicola

— Vesture of plants mainly lead-gi^ay; involucral
bracts 8-12 A. nova var nova

The var duchesnicola is the dominant plant,
often in association with other desert shrubs,
on reddish clay soils of the Duchesne River
Formation, for which the variety is named,
from about 15 km west of Vernal to Tridell. It
occurs from about 1700 to 1800 m elevation
on low clay uplands in a position ecologically
between A. tridentata var wyorningensis (Beetle
& Young) Welsh of desert drainages and A.
nova var. nova, which grows in rocky sub-
strates formed by ancient stream pediments.
Suggested as the origin of this entity is poten-
tial hybridization of A. nova and A. tridentata
var wyomingensis. Although differing only in
minor ways, the plants are continuous and
uniform over rather large expanses of the
Duchesne River Formation, and they are wor-
thy of taxonomic recognition.

References

Hitchcock, C. L. 1936. The genus Lepidhnn in the United

States. Madrono 3: 265-320.
Welsh, S. L., N. D. Atwood, S. Goodrich, and L. C.

HicciNS. 1993. A Utali flora. 2nd edition. Life Science

Museum, Brigham Young Unixersitv; Provo, UT. 986

pp.

Received 22 March 1995
Accepted 26 June 1995

PREY CHOICES AND FORAGING EFFICIENCY OF RECENTLY
FLEDGED CALIFORNIA GULLS AT MONO LAKE, CALIFORNIA

Chris S. Elphick^ and Margaret A. Ruhega^

Abstr.'\ct. — We studied the foraging hiologv' of recenth- fledged California Culls {Lams califoniiciis) at Mono Lake
during August-September 1991. We made behavioral observations to collect information on the relative proportions of
different prey types in the diet of these birds and took plankton tows to determine the relative abundance of each prey
in the water column. These data show that alkali flies {Ephijdra hians) were the primaiy constituent of the diet and that
they were eaten at a much higher rate than one would expect based on their abundance. We also detennined the num-
ber of feeding attempts and successful captures made during each behavioral observation. From these, we calculated the
birds' feeding efficiencies on emergent adult alkali flies and on all other pre>' t\'pes combined. We found that foraging
efficiencies on emergent flies were ver\' high and significantly greater than those obtained on other prey types. These
results suggest that flies were actively sought in preference to the alternative prey type, brine shrimp {Artcmia monica),
presumably because they are easier to capture and of greater nutritional value.

Key words: California Gull, Larus californicus, diet, foraging ejficiency. Mono Lake.

California Gulls {Larus californicus) breed
widely in die arid West, widi the largest con-
centrations at two saline lakes: Great Salt Lake
in Utah and Mono Lake in east central Cali-
fornia (Conover 1983). Various factors may in-
fluence the size and reproductive success of
the California Gull colony at Mono Lake: pre-
dation, food supply, weather, parasitism, nest-
ing habitat, and access to freshwater (Winkler
1983, Winkler cited in Botkin et al. 1988). Of
these, increased risk of predation caused by
the exposure of a "land-bridge" between the
mainland and islands on which the birds
breed has received most attention (Patten et
al. 1987, Botkin et al. 1988).

The role of food abundance has received
relatively little discussion, primarily because in-
fonnation on the diet of California Gulls at Mono
Lake is limited. Brine shrimp {Artemia monica)
and alkali flies {Ephijdra hians) are the main
sources of food available to gulls, although
other items (e.g., cicadas, fish, and garbage)
are occasionally taken (Patten et al. 1987).
Previous reports have focused on the food
brought to chicks at the nest. Some of these
studies show chick diets to be dominated by
brine shrimp (Grinnell and Storer 1924, Winkler
et al. 1977, Jehl and Mahoney 1983), while

others found high proportions of alkali flies
(Nichols 1938, Young 1952, Mason 1967). Diet
data for other age classes of gulls are not wide-
ly available. Young (1952) dissected two indi-
viduals and found their guts to be full of alkali
fly pupae, and Jehl and Mahoney (1983) found
high proportions (>90% by volume) of shrimp
in a sample of free -swimming gulls (18 adults,
20 fledglings). These studies show that both
brine shrimp and alkali flies are used by
California Gulls at Mono Lake under certain
circumstances. The factors that determine
which of the two prey species, or which life
stages of alkali flies, are taken are not known.
Do the patterns simply reflect variation in rel-
ative abundances of prey species, or is one
species preferred but not always available?

During three summers of fieldwork we
noticed that over the latter part of summer Cali-
fornia Gulls feed extensively on alkali flies,
particularly recently emerged adults. Flies of
this age class are immotile and presumably
easier to catch than either brine shrimp or fly
lai-vae (though not necessarily fly pupae). We
therefore hypothesized that they would be a
preferred prey source when available. In this
paper we quantify the incidence of alkali flies
in the diet of recently fledged California Gulls.

'EcologN. E\()lution and Conservation Biologx', Universit>- of Nevada, Reno. 1000 \'alle\- Road, Reno, NV 89.512. ."Vuthor to whom correspondence shonld be
addressed.

^Department of Ecologv- and Evolutionary Biology, Universit>' of California, Inine, CA 92717. Present address: Ecolog\-, Evolution and Conservation
Biology, University- of Nevada, Reno, 1000 Valley Road, jleno, NV 89512.

363

364

Great Basin Naturalist

[Volume 55

We restiicted our study to juvenile gulls because
inexperienced birds are t\'picall\ the least pro-
ficient foragers (Porter and Sealey 1982, Burger
1987, Wunderle 1991) and hence most likely
to benefit from the availability of easily cap-
tured prey. We demonstrate that under certain
circumstances alkali flies (1) constitute a major
proportion of the diet and (2) are not eaten in
direct proportion to their abundance. As a
potential explanation for the birds apparent
preference for alkali flies when available, we
also test the hypothesis that fledgling gulls are
able to achieve greater feeding efficiencies
when eating emergent adult flies than when
foraging on alternative prey.

Methods

Data were collected on five days during
August and September 1991 from waters just
off the northeastern shore of Mono Lake,
where feeding gulls were numerous.

Feeding observations. — We obtained
feeding data by videotaping foraging birds
with a Sony 8 mm HandyCam video recorder
with an 8X zoom lens (n = 50) or by direct
observations {n = 20). In all cases the focal
bird was within 10 m of the observer, and for-
aging behavior was scored over a 1-min feeding
trial. No more than 10 birds were obsei"ved at
any site to reduce the chance of obtaining
repeated samples of the same individual.

Feeding trials were scored for the number
of feeding attempts and successful captures,
which were divided by one minute to give
attempt and success rates. When possible,
prey items were identified. An attempt was
defined as any occasion on which the bird's
bill entered the water or the bird lunged for a
prey item on the water's surface. Attempts
were deemed successful if (1) the gull was
seen "head-throwing" (i.e., inertial feeding;
Gans 1961) and swallowing after the attempt,
(2) the prey item was observed in the bird's
mandibles and not dropped, or (3) the prey item
was visible on the water surface before the
capture attempt and was picked off by the
gull. Filmed trials were scored at half-speed to
improve accuracy. Data from the one day when
both methods were used were compared to
assess the relative accuracies of videotaping
and direct obsen'ation.

Diet. — We used two measures to determine
the incidence of alkali flies in the diet of juve-

nile gulls. First, we used the number of
attempts directed at flies (all life stages), divided
by the total number of attempts, as a measure
of the proportion of foraging effort directed at
alkali flies. Second, we calculated the mini-
miun proportion of the birds diet that consti-
tuted flies:

fl\' captures
attempts on all prey minus known failures.

Prey abundance. — Prey abundance was
determined from horizontal plankton tows
taken at the site of, and immediately after, a
series of feeding trials. Tows were made with a
0.5-^tm mesh plankton net, 1 m in diameter, and
supported at the surface by floats. The tows
sampled approximately 6 ni'^ of water, down to
a maximum depth of about 60 cm. Samples
were sorted and individuals of each alkali fly
life stage counted. Because shrimp were too
numerous to count, their numbers were calcu-
lated from a previously deteniiined wet weight
to number relationship (Rubega unpublished
data):

Weight (g) = 0.002207*Number (r- = .96, n = 10).

Feeding efficiency. — We calculated feed-
ing efficiency of juvenile gulls by dividing the
number of successful prey captures by the
number of attempts for both emergent adult
alkali flies and all other prey types combined.
These values were compared using a paired /
test in which the two efficiency measures for
each individual were paired. Feeding efficiency
could not be calculated individually for other
prey types because, unlike adult flies, they
occurred below the water's surface and often
could not be seen unless they were captured.
Hence, usually we were unable to determine
the object of the foraging attempt unless the
attempt was directed at an adult fl\. All esti-
mates are given in means (± standard error).

Results

Table 1 compares the minimum propor-
tions of the total diet for each prey type with
the relative abundances of each prey in plank-
ton tows. Alkali fly adults and pupae both
were eaten in much higher numbers than
expected if prey were taken in proportion to
their abundance. The mininumi proportion of

1995]

California Gull foraging ecology

365

Table L Mean proportions (± SEM) of different prey
b.'pes in the diet of fledged California Gulls {n = 70) and in
plankton tows taken where birds were feeding (n = 21).
No diet data are axailahle tor fl\' lai"vae or shrimp because
they could not be distinguished in our feeding trials.

Abundance

Abundance in

in diet

plankton tows

Prey type

(% by number)

Alkali fix adults

> 22.59 ± 0.35

0.01 ±0.003

Alkali fl\' pupae

> 18.20 ± 0.39

0.67 ± 0.40

Alkali fl\ larvae

—

0.05 ± 0.0007

Alkali flies

(all life stages)

> 40.79 ± 0.36

0.74 ± 0.04

Brine shrimp

—

99.25 ± 0.04

foraging attempts directed at flies (all life
stages) and the minimum proportion of the
diet comprised of flies were 41.7 ± 3.0% and
40.8 ± 3.0%, respectively {n = 70). In compar-
ison, only 0.7 ± 0.8% (n = 22) of prey items
sampled in the water column were alkali flies;
the remainder were all brine shrimp. These
data indicate that alkali flies were favored over
brine shrimp.

The two sampling methods are compared
in Table 2a. Attempt and success rates for all
prey types combined did not differ significant-
ly between the videotaped feeding trials and
those obtained by direct observation (^33 =
-0.1, P = .933 and ^33 = 1.56, P = .128,
respectively). Proportions of different prey
types recorded did differ, however, with
videotaped trials, underrecording the number
of pupae captured by an average of 79.7% on
the one day for which a comparison was possi-
ble. Similar numbers of adult flies were
detected by the two methods. This discrepancy
was probably because, unlike adult flies, pupae
do not float on top of the water surface and are
difficult to see on film due to reflection. Values
given above for the incidence of alkali flies in
the diet are therefore underestimates.

Mean foraging efficiency for recently fledged
gulls feeding on emergent alkali flies was very
high and significantly greater than mean effi-
ciency on all other prey (Table 2b; paired ^45
= 10.8, P < .0001). In addition, a comparison
of the two measures for each individual
showed that in all but one case a bird's effi-
ciency was greater when feeding on emergent
flies. Although our foraging efficienc)' data for
alkali fly pupae are limited because we did not
always know what prey type an attempt was
directed at, they do indicate that pupae were
caught as easily as adult flies (Table 2a).

Discussion

The large difference between alkali fly use
and abundance strongly suggests that flies were
actively sought in preference to brine shrimp
and that flies were an important component in
the diet of the birds we obsei-ved. It is likely
that our prey sampling regime underestimated
the availability of alkali flies because (1) we
sampled deeper in the water column than gulls
forage and (2) emergent flies are most abun-
dant at the surface. It is unlikely, however, that
this could account for the 60-fold difference
between observed and expected values for fly
abundance in the birds' diet. Two factors may
contribute to the apparent preference for flies
over shrimp. First, we have shown that 27%
higher foraging efficiencies can be attained
when feeding on emergent alkali flies than on
alternative prey types combined. Second,
Herbst (1986) reported that alkali flies are larger
and have a greater nutritional value than the
alternative food, brine shrimp. Both factors
mean that there is an increase in food intake
per unit effort when feeding on emergent
flies. Although we have no quantitati\'e data
for adult gulls, observations made during the
course of this study suggest that they also fed
predominantly on alkali flies. A supply of easi-
ly caught prey, however, would be expected to
benefit juveniles more that adults because the
former lack foraging experience and are more
likely to have difficulty feeding on more
motile prey.

Conclusions that can be drawn from these
results are obviously limited. Our sampling was
restricted to a few dates in one year and one
portion of Mono Lake. Our anecdotal observa-
tions from two additional years and surveys
conducted across the entire lake suggest that
these findings are not atypical for late summer,
when emergent flies and dislodged pupae are
common at the water surface. We have no data
for other time periods; however, chick diet data
collected earlier in the summer suggest that
flies were eaten throughout the post-hatching
period in 1991 (D. Shuford personal commu-
nication). Jehl and Mahoney's (1983) data
clearly show that under some circumstances
brine shrimp make up a major portion of the
diet of fledgling California Gulls. The differ-
ence between their result and ours mirrors the
variation seen in the diet of chicks (Grinnell
and Storer 1924, Nichols 1938, Young 1952,

366

Great Basin Naturalist

[Volume 55

Table 2. Mean feeding pcifonnancc v;ilues (± SEM). Sample sizes given in parentlieses. (a) Coniparati\'e \aliies for tlie two
obsenation methods from tlie one da> on which hoth wiMe nsed. (li) Values for tlie two prey classifications for which accurate
data could be collected from all studv (lavs.

Prey type

Attempt/min

Success/min

Efficiencv (%)

(a) Comparison of observation methods
Alkali tK adults (\idco trials)

Alkali tl\ adults (direct observation)
Alkali fly pupae (video trials)
Alkali fly pupae (direct obsen'ation)
All pre\' (\'ideo trials)
All pre\' (direct observation)

(b) Comparison of prey types

Alkali fly adults

All prey except adult flies

0.53 ±0.06 (15)
0.65 ±0.05 (20)
2.40 ±0.31 (15)*
9.50 ±0.22 (20)*
16.67 ±0.43 (15)
16.85 ±0.31 (20)

8..30±0.15 (70)
17.20 ±0.99 (70)

0.53 ±0.06 (15)
0.65 ±0.05 (20)
1.93 ±0.286 (15)
9..50±0.22 (20)
13.07 ±0.40 (15)
10.15 ±0.25 (20)

7.79 ±0.14 (70)
11.99 ±0.09 (70)

100 ±0 (5)

100 ±0 (8)
81.75 ±0.03 (8)*

100 ±0 (20)*
78.25 ±0.85 (15)
59.00 ±0.90 (20)

95.77 ±1.0 (70)
68.40 ±0.23 (70)

*These data should be viewed with cautinn as attempt rates are

Mason 1967, Winkler et al. 1977, Jehl and
Mahoney 1983). Alkali fly abundance varies
seasonally with an increase during May and
June, peak numbers between July and Sep-
tember, and a gradual decline thereafter
(Herbst 1986). Research by Point Reyes Bird
Observatory shows that the relative propor-
tions of flies and shrimp in food brought to
chicks differ considerably between samples
collected during the day and night, and
between years (D. Shuford personal communi-
ation). These observations not only suggest diat
relative availability of the two prey is quite
variable at daily, seasonal, and annual time
scales but also help explain the discrepancies
between studies. Previous diet studies did not
present data on relative prey abundances in
areas where birds were foraging. In demon-
strating a higher than expected abundance of
alkali flies in the diet of fledgling gulls and the
high foraging efficiencies that can be attained
when feeding on them, our study suggests that
flies are the preferred prey when they are
available.

In light of recent research on Red-necked
Phalaropes {Phalaropus lohatus), which are
physiologically imable to survive on a diet of
pure brine shrimp (Rubega and Inouye 1994),
our data lead us to speculate that brine fly
production may be an important factor in
determining fledgling survival rates (currently
unknown) for the Mono Lake gull colony.
California Gulls clearly eat brine shrimp on a
regular basis and apparently are not as depen-
dent on alkali flies as Red-necked Phalaropes.
However, it is not clear whether the prey supply
is limiting the gull population size. Experiments

needed to address that issue have yet to be per-
formed. In addition, it is possible that gull
predation plays an important role in determin-
ing alkali fly recruitment rates. The extent to
which these issues are important can only be
established through further study of the inter-
actions between flies and gulls, both at Mono
Lake and elsewhere.

Acknowledgments

We thank D. Elphick, D. Dawson, staff at
the Sierra Nevada Aquatic Research Labora-
tory, the High Sierra Shrimp Plant, and W
Hamner for their contributions to the comple-
tion of fieldwork. D. Shuford kindly gave us
access to unpublished data collected by Point
Reyes Bird Obsei-vatoiy. E. Beedy, L. Oring,
M. Reed, D. Shuford, R. Whitmore, and three
anonymous reviewers made useful comments
on earlier versions of this paper. Fieldwork
was supported by a grant fi-om the Universit\' of
California, Irvine Foundation to MAR, made
possible by a donation from Jones & Stokes
Associates, consultants to the Los Angeles
Department of Water and Power and the
California State Water Resources Control
Board. CSE received travel funds fi-om the Uni-
versity of East Anglia's Expedition Committee
and the Sir Phillip Reckitt Educational Trust.

Literature Cited

BoTKiN, D. B., W. S. Broecker, L. G. Everett, J. Shapiro,
AND J. A. WiENS. 1988. The future of Mono Lake:
report of the Community and Organization Research
Institute (CORI) "Blue Ribbon Panel." Universit>' of
California, Water Resources Center Report 68.

1995]

California Gull foraging egology

367

Burger, J. 1987. Foraging efficiency in gulls: a congeneric
comparison of age differences in efficiency and age
of maturity. Studies in Avian Biolog\' 10: 83-90.

CoNOVER, M. R. 1983. Recent changes in Ring-billed and
California Gull populations in the western United
States. Wilson Bulletin 95: 362-.383.

Cans, C. 1961. The feeding mechanism of snakes and its
possible evolution. American Zoologist 1; 217-227.

Grinnell, J., AND T. I. Storer. 1924. Animal life in the
Yosemite. University' of California Press.

Herbst, D. B. 1986. Comparative studies of the popula-
tion ecology and life histon patterns of an alkaline
salt lake insect: Ephydra (Hijdropyrtts) hians (Diptera:
Ephydridae). Unpublished doctoral dissertation,
Oregon State University, Conallis.

Jehl, J. R., Jr., and S. A. Mahoney. 1983. Possible se.xual
differences in foraging patterns in California Culls
and their implications for studies of feeding ecolog\'.
Colonial Waterbirds 6: 218-220.

Mason, D. T. 1967. Limnology of Mono Lake, California.
University' of California Publications in Zoology 83.

Nichols, W. E 1938. Some notes fi-om Negit Island, Mono
Lake, California. Condor 40: 262.

Patten, D. T, et al. 1987. The Mono Basin ecosystem:
effects of changing lake level. National Academy
Press, Washington DC.

Porter, J. M., and S. C. Sealey. 1982. Dynamics of sea-
bird multispecies feeding flocks: age-related feeding
behaviour. Behaviour 81: 91-109.

RuBEGA, M., AND C. Inouye. 1994. Prey switching in
Red-necked Phalaropes {Phalaropus lobatus): feed-
ing limitations, tlie flinctional response and water man-
agement at Mono Lake, California, USA. Biological
Consei-vation 70: 205-210.

Winkler, D. W. 1983. Ecological and behavioral determi-
nants of clutch size: the California Cull (Lams cali-
fornicits) in the Great Basin. Unpublished doctoral
dissertation. University of California, Berkeley.

Winkler, D. W, C. R Weigen, E B. Engstrom, and S. E.
Birc:h. 1977. Ornithology. Pages 88-113 in D. W
Winkler, editor. An ecological study of Mono Lake,
California. Institute of Ecology Publication No. 12.
University of California, Davis.

WUNDERLE, J. M., Jr. 1991. Age-specific foraging profi-
ciency in birds. Current Ornithology 8: 273-324.

Young, R. T 1952. Status of the California Gull colony at
Mono Lake, California. Condor 54: 206-207.

Received 1 7 October 1994
Accepted 20 June 1995

HYBRIDIZATION BETWEEN BUFO WOODHOUSII AND BUFO
PUNCTATUS FROM THE GRAND CANYON REGION OF ARIZONA

Keith Malmos', Holiert Reed', and Bnan Starrett^
Key words: hijhridizatidii, Bulo woodhoiisii, BiiFo punctatus, toads, Anura, dislrihiifion.

Natural hybridization between toads of the
genus Biifo is eommon; most accounts involve
representatives from the same species group
(Sullivan 1986). Species groups within the
genus Biifo are hypodiesized to be monophyletic
groups, based on data that include osteology,
lab hybridization studies, advertisement calls,
and release calls (summarized in Blair 1972a).
Intergroup hybrid adults are expected to be
relatively more rare in nature because of the
low proportion that develop completely (Blair
1972b). Here we report intergroup hybridiza-
tion between Biifo woodhoiisii {ainericanus
group) and Biifo punctatus (jninctatus group;
Blair 1972c). Hybrid B. punctatus x B. wood-
hoiisii previously reported from Colorado near
Grand Junction were described as "sterile
males with atrophied testes" (McCoy et al.
1967). We present evidence that B. wood-
hoiisii and B. punctatus have hybridized at two
new localities in Arizona, Coconino Co., and
that atrophied testes are not universal in these
hybrids. The localities are approximately 3 km
upstream from the Colorado River, near
Powell Canyon in the Little Colorado River
Gorge, and approximately 8 km downstream
of the confluence of the Little Colorado River
and Colorado River where Lava Creek emp-
ties into the Colorado River. We also analyzed
specimens collected by S. W. Aitchison in
1973 from Choal Canyon, Coconino Co.,
approximately 22.5 km NNE of Kaibito; these
specimens include putative hybrid B. puncta-
tus X B. woodhoiisii. Hybrids from this series
are likely the toads that support the comment
by Miller et al. (1982) that hybridization
between B. punctatus and B. woodhoiisii
occurs in Grand Canyon National Paik.

Toads were identified and analyzed mor-
phologically using methods similar to those of
Ferguson and Lowe (1955) and McCoy et al.
(1967). Each toad was dissected to determine
sex and condition of testes of putative hybrids.
Twelve specimens from the Little Colorado
River (LCR) site [3 B. punctatus (ASU28935-
28937), 8 B. woodhoiisii (ASU28939-28946),
and the hybrid (ASU28938)], and 15 of the 17
specimens fi'om Choal Canyon (CC) [8 B. piinc-
tatm (MNA Z6.529-536), 5 B. woodhoiisii (MNA
Z6.522-526), and 2 hybrids (MNA Z6.527-
528)] were analyzed. The tv\'o toads fi-om Choal
Canyon excluded from the analysis were too
small to evaluate reliably since ontogenetic
changes in cranial crest and parotoid gland
morphology occur in some toads (Sullivan
1986). Measurements were taken from pre-
served male toads that were all the size of
reproductively mature individuals. A Helios
vernier caliper precise to 0.05 mm was used.
Body size and parotoid gland variation among
species of toads are diagnostic for many
species. For the toads we examined, B. wood-
hoiisii is larger and has more elongate parotoid
glands than B. punctatus, which is a smaller
toad with small, round parotoid glands. We
measured snout-vent length (SVL) and paro-
toid gland length (PL) and width (PW). A ratio
of parotoid gland dimensions (PL/PW) was
formed to evaluate gland shape.

All toads from the LCR collection have
developed gonads and secondary sexual char-
acteristics. The three B. punctatus and four B.
woodhoiisii males exhibit darkened vocal sacs,
well-developed thumb pads, and testes typical
for the species. The other four B. woodhoiisii
appear to be spent females containing ovaries

^Arizona State Universit>, Department of Zoology, Tenipe. AZ 85281-1501.
2plioeiiix Zoo, 455 North Galviii Park-wa\, Phoenix, AZ 85008.

368

1995]

Notes

369

with undeveloped eggs. The hybrid male pos-
sesses one typical looking testis and one great-
ly enlarged testis, approximately 10 times nor-
mal size.

Morphological analysis supports identifica-
tion of ASU28398 as a hybrid. Values present-
ed are the mean ± SD. The hybrid was 58.80
mm SVL, larger than B. piinctatus (42.17 ±
1.48 mm) but similar in size to B. woodhoiisii
(59.69 ± 6.03). Shape of the parotoid gland,
PL/PW, was intermediate for the hybrid, 1.43,
relative to B. piinctatus (1.017 ± 0.053) and B.
woodhoiisii (2.161 ± 0.330; Fig. 1). Both
PL/PW and SVL are different between the
species with at least 95% confidence because
the means ± 2STD do not overlap.

Although no specimens were retained from
the Lava Creek site (LC), photographs taken
in April 1993 provide clear evidence of hybrid-
ization between B. piinctatus and B. wood-
housii at this second site in Grand Canyon
National Park (Fig. 2). Body size, parotoid
gland moiphology, and coloration of the adult
male hybrid are intermediate. Biifo ivood-
housii is larger, has much more elongate paro-
toid glands, and lacks the spinose red warts
seen in Biifo piinctatus.

We submit this photographic evidence and
morphological analysis of toads as support for
the suggestion by Stevens (1983) that B. piinc-
tatus X B. woodhoiisii hybrids occur in the
Grand Canyon region of Arizona. We also sug-
gest that, based on specimens not from the
Grand Canyon region, but from specimens
collected associated with Glen Canyon, Miller
et al. (1982) reported that B. piinctatus X B.
woodhoiisii hybrids occur in the Grand
Canyon. Whether B. piinctatus x B. wood-
housii hybrids from LCR and LC could repro-
duce would require histological analysis and
additional sampling to determine if hybrids
have viable sperm. We are, however, unaware
of other reports of enlarged testes in hybrid
toads.

Three toads from Choal Canyon, MNA
Z6.527-528 and MNA Z6.496, may be hybrids
based on intermediate values of PL/PW (1.42
± 0.04). As in the LCR series, SVLs of hybrids
(58.01 ± 3.37) are greater than B. punctatus
(48.84 ± 6.17), but similar to B. woodhoiisii
(56.02 ± 7.87; Fig. 1). The means ± 2STD for
SVL and PL/PW overlap for the CC sample;

45 50 55 60 65
snout-vent length (mm)

Fig. 1. Comparison of relative sizes and parotoid gland
dimensions of specimens from the Little Colorado River
locality (closed symbols) and the Choal Canyon locality
(open symbols). Circles are Bufo woodhoiisii, squares are
Biifo punctatus, and triangles are hybrids.

therefore, significant statistical differences do
not exist. A small sample size is likely influen-
tial. Gonadal development in some CC hybrids
is unusual; MNA Z6.496 could not be sexed by
its gonads or secondary sexual characters. The
other two hybrids, MNA Z6.527 and 528, have
darkened thumb pads and vocal sacs. Both
testes of MNA Z6.527 appear normal, but
MNA Z6.528 has one enlarged testis and the
other absent or greatly reduced. Again, whether
hybrid males of this cross are reproductively
functional is unknown.

Field observations suggest that hybridiza-
tion at LCR may be relatively common. When
the LCR collection was obtained, 13-14 May
1993, advertisement calls typical of B. wood-
hoiisii and B. punctatus were both heard at
night, as well as calls that sounded aberrant,
approximately intermediate in duration, pulse
rate, and pitch of each species. No other species
of toads were obsened during spring months

370

Great Basin Naturalist

[Volume 55

Fig. 2. Photographs of toads from the Lava Creek locah-
ty: (a) Biifo woodhoiisii, (b) hybrid, (c) Biifo piincfatiis.

at the LCR site for two years. Advertisement
calls produced by hybrid toads often have
characteristics intermediate to their parental
forms (Blair 1956, Zweifel 1968, Sullivan 1986,
1990). Calls of suspected hybrids were not
heard at Lava Creek, but both species cho-
rused together there in April 1993.

Habitat disturbance and environmental
change associated with Glen Canyon Dam may
contribute to hybridization between these taxa
in die Grand Canyon region. Other hybrid zones
between toads are associated with river regu-
lation projects or human impacted areas (Sulli-
van 1986 and examples cited therein). Altera-

tions to the Colorado River have reduced sea-
sonal peak flows, created large daily fluctua-
tions in flow, and dramatically lowered the
temperature of the water. Tributaries such as
the Little Colorado River and Lava Creek are
relatively less affected. Perhaps departure
from historic conditions contributes to the
likelihood of contact and hybridization
between B. woodhousii and B. punctatus in
the Grand Canyon. Other possible explana-
tions for hybridization include natural pertur-
bations that disrupt ecological separation.
Also, natural cycles in population size and
species range are hypothesized to account for
many hybrid zones (Hewitt 1989).

Acknowledgments

We thank M. E. Douglas at Arizona State
University and M. Morales and D. Hill at the
Museum of Northern Arizona for use of speci-
mens. We thank B. K. Sullivan for suggestions
on the manuscript. We also thank the Navajo
Fish and Wildlife Branch of the Navajo Nation
for providing a collecting permit to RNR
(#930709-058). Aiizona State University-West
provided funds to KBM for some costs associ-
ated with this project.

Literature Cited

Bu\iR, W. F. 1956. The mating calls of h\l)rid toads. Texas
Journal of Science 8: 350-355.

. 1972a. Evolution in the genus Biifo. Universit>' ot

Texas Press, Austin. 459 pp.

. 1972b. Exadence from lixbridization. Pages 196-232

in W. F Blair, editor. Evolution in the genus Biifo.
University of Texas Press, Austin.

. 1972c. Bufo of North and Central America. Pages

93-101 in W. E Blair, editor, Evolution in the genus
Bufo. University of Texas Press, Austin.

FERGU.SON, J. H., AND C. H. LoWE. 1969. The evolution-
ary relationships in the Bufo punctatus group.
American Midland Naturalist 81; 435—466.

Hewitt, G. M. 1989. The subdivision of species by hybrid
zones. Pages 85-110 in D. Otte and J. A. Endler, edi-
tors, Speciation and its consequences. Sinauer
Associates Inc., Sunderland, MA.

McCoy, C. J., H. M. Smith, and J. A. Tihen. 1967.
Natural hybrid toads, Bufo punctatus X Bufo wood-
housei, from Colorado. Southwestern Naturalist 12:
45-54.

Miller, D. M., R. A. Young, T. W. Gatlin, and J. A.
Richardson. 1982. Amphibians and reptiles of the
Grand Canyon National Park. Grand Can\on Natural
Histoiy Association, Monograph No. 4.

1995]

Notes

371

Stevens, L. 1983. The Colorado River in Grand Canyon;

a guide. Red Lake Books, Fkigstaff, AZ. 110 pp.
Sullivan, B. K. 1986. Hybridization between the toad

Bufo microscaphiis and Biifo woodhoiisei in Arizona:

morphological variation. Journal of HerfDetology 20:

11-21.
. 1990. Natural hybrid between the Great Plains toad

{Bufo cognatus) and red-spotted toad [Bufo piinc-

tatiis) from central Arizona. Great Basin Naturalist

50: 371-372.

ZWEIFEL, R. G. 1968. Effects of temperature, body size,
and hybridization on mating calls of toads, Bufo a.
aynericanus and Bufo woodhousii fowleri. Copeia
1968(2): 269-285.

Received 14 October 1994
Accepted 20 March 1995

REPRODUCTION IN THE BANDED SAND SNAKE,
CHILOMENISCUS CINCTUS (COLUBRIDAE), FROM ARIZONA

StepluMi H. Goldhergl
Key words: Chilonicnisciis ciiictus, handed sand snake. CoJuhridae. reprodneiion. Arizona.

The banded sand snake, Chilomeniscus cinc-
tiis Cope, 1861, ranges fi-om central Arizona to
extreme southern Sonora, and throughout all
but the northern part of Baja California
(Stebbins 1985). Anecdotal comments on the
reproduction of this species have been pub-
lished in Stebbins (1954), Wright and Wright
(1957), and Behler and King (1979), and in this
report I provide data on reproduction in C.
cinctus from Arizona.

I examined 38 Chilomeniscus cinctus (24
males, 14 females) from Arizona in the heipe-
tology collections of Arizona State University
(ASU), Tempe; Natural Histoiy Museum of Uos
Angeles County (UACM), Uos Angeles; San
Diego Natural History Museum (SDSNH),
San Diego; and the University of Aiizona (UAZ),
Tucson. Museum numbers of specimens exam-
ined are given in Appendix 1. All Arizona C.
cinctus in the above collections were exam-
ined; however, some had been damaged (road-
kills) or had not been preserved promptly
enough to avoid autolysis. These were not
used and are not in Appendix 1. Counts were
made of oviductal eggs or enlarged follicles.
The left gonad was removed for histological
examination, embedded in paraffin, and cut
into histological sections at 5 /xm. Slides were
stained with Harris' hematoxylin followed by
eosin counterstain. Testes slides were exam-
ined to determine the stage of the male cycle;
ovary slides were examined for the presence
of yolk deposition.

Data on the male C. cinctus seasonal testic-
ular cycle are presented in Table 1. Testicular
histology was similar to that reported in
Goldberg and Parker (1975) for two other
North American colubrid snakes, Masticophis
taeniatus and Pituophis nielanoleucus. In the
regressed testes, seminiferous tubules con-

tained spermatogonia and Sertoli cells. In
recrudescence, there was renewal of spermato-
genic cells characterized by spermatogonial
divisions; primary and secondary spermato-
cytes, and spermatids, may have been present.
In spermiogenesis, metamorphosing sper-
matids and mature sperm were present.

Small sample sizes from all months except
May-Iune (Table 1) prevented a definitive de-
scription of the male cycle. However, since all
10 May males and 5 lune males were under-
going spenniogenesis, it is likeK' that C. cinctus
breeds during these months. Epididymides
from 2 May and 1 lune males contained sperm.
The smallest spermiogenic male (sperm pres-
ent) measured 151 mm in snout-vent length
(SVU).

Data on the C. cinctus seasonal ovarian
cycle are presented in Table 2. I recorded two
clutch sizes: 6 lune, 3 enlarged follicles (3—4
mm diameter), 188 mm in SVU; 4 Inly, 2 ovi-
ductal eggs (6 mm diameter), 192 mm in SVU.
Yolk deposition (vitellogenic granules) was
found on histological examination of ovarian

Table 1. MonthK distribution of conditions in seasonal
testicular cycle oi Chilomeniscus cinctus. Values shown
are the numbers of males exhibiting each of the three con-
ditions.

Recru-

Spermio-

Montli

Regressed

descence

genesis

|anuar\

Pebruar\

March

April

Mav

June

lulv

September

December

^Department of Biology, VVhitticr College, VVIiiltier, CA 90608.

372

1995]

Notes

373

Table 2. Monthly distribution of conditions in seasonal ovarian cycle of Chilo)neni.$cus cinctus. Values shown are the
number of females exhibiting each of the four conditions.

Month

Inactive

Yolk deposition

Enlarged follicles

iductal eggs

Februan

March

April

lime

Julv

August

September

October

November

tissue from two June females (173 mm and
198 mm in SVL). No yolk deposition was seen
in the remainder of the female sample. The
lack of vitellogenesis in some adult females
during the reproductive season may indicate
that not all C. cinctus females breed each year.
Breeding by only part of the adult female pop-
ulation has been reported for other North
American temperate zone snake species (see
Aldridge 1979). The smallest reproductively
active female (yolk deposition in progress)
measured 173 mm in SVL.

The biology of C. cinctus is poorly known.
A few reports on its food habits reveal that it
eats centipedes and insects (Vorhies 1926,
Stebbins 1954, 1985, Behler and King 1979).
According to Lowe et al. (1986), C. cinctus has
grooved rear teeth; it is not known whether it
has toxic gland secretions. The small numbers
of C. cinctus in the two major Arizona her-
petology collections (ASU, UAZ) reflect the
secretive nature of this snake. Intensive study
will be rec^uired before the biology of C. cinc-
tus is known.

Acknowledgments

I thank Charles H. Lowe (University of
Arizona), Robert L. Bezy (Natural History
Museum of Los Angeles County), Michael E.
Douglas (Aiizona State University), and Sally Y.
Shelton (San Diego Natural History Museum)
for permission to examine snakes in the her-
petology collections of their respective institu-
tions. Jorge Martinez assisted with histology.

Literature Cited

Behler, J. L., and E VV. King. 1979. The Audubon Society
field guide to North AiTieric;m reptiles and amphibians.
Alfred A. Knopf, New York. 743 pp.

Goldberg, S. R., and W. S. Parker. 1975. Seasonal tes-
ticular histology of the colubrid snakes, Masticophis
tacniatus and Pituophis melanoleucus. Heipetologica
31:317-322.

Lowe, C. H., C. R. Schvvalbe, and T. B. Johnson. 1986.
The venomous reptiles of Arizona. Arizona Game
and Fish Department, Phoenix. 115 pp.

Stebbins, R. C. 1954. Amphibians and reptiles of western
North America. McGraw-Hill, New \brk. 536 pp.

. 1985. A field guide to western reptiles and am-
phibians. Houghton Mifflin Company, Boston. 366
pp.

Vorhies, G. T. 1926. Notes on some uncommon snakes of
southern Arizona. Copeia 1926:158-160.

Wright, A. H., and A. A. Wright. 1957. Handbook of
snakes of the United States and Canada. Volume I.
Comstock Publishing Associates, Ithaca, NY. 564 pp.

Received 29 November 1994
Accepted 7 February 1995

Appendix 1

Specimens examined by count>' from herpetology col-
lections at Arizona State Universit\' (ASU), Natural History
Museum of Los Angeles County (LACM), San Diego
Natural History Museum (SDSNH), and University of
Arizona (UAZ).

Maricopa: ASU 04669, 09161, 13903, 26367-26368.
LACM 112460. UAZ 24104, 35645, 35795, 35818. Pima:
SDSNH 33383. ASU 01231, 15391, 28401. LACM 34918.
UAZ 24087, 24089, 24092, 24095-24096, 24103, 24107-
24108, 30241, 33815, 34411, 34680-34681, 35166, 36108,
37819, 37821, 42197. Pinal: ASU 15376, 23.573. 26411,
26413. UAZ 24097.

Aldridge, R. D. 1979. Female reproductive cycles of the
snakes Arizona elegans and Crotalus viridis. Heipe-
tologica 35: 256-261.

NO ACOUSTIC BENEFIT TO SUBTERRANEAN CALLING IN THE CICADA
OKANAGANA PALLIDULA DAVIS (HOMOPTERA: TIBICINIDAE)

Allen E Sanborni .^ik-i poUy K. Phillips-
Key words: cicada, acomfic hehavioi; calling, .sound pressure level, predator avoidance, Okanagana pallidiila.

Most male cicadas produce a loud calling
song to attract their mates. Sound pulses are
produced when specialized muscles buckle the
rib-strengthened chitinous membranes, the
timbals, located on the dorsolateral surface of
the first abdominal segment. Sound pulses are
then modified by several body components
(Pringle 1954, Bennet-Clark and Young 1992)
before being radiated through the tympana
(Young 1990).

Male cicadas generally use an accessible
perch fi-om which they advertise their presence
to conspecific females. We came across an ex-
ception to this behavior south of Lone Pine,
Inyo County, CA, on 15 July 1994. We en-
countered the cicada species Okanagana pal-
lidiila Davis singing in a scrub habitat. As we
began collecting, we noticed that one individ-
ual continued to sing as we approached and
was very difficult to locate on the plant. By
circling the plant we found that the sound was
actually coming from the ground near the base
of the plant and not from on the plant itself
After clearing some grass we could see a hole
about 1 cm in diameter from which the sound
emanated. Within the hole we could see the
head of a cicada that was calling from this sub-
terranean site.

We measured intensity levels from males
calling from burrows and from plants to deter-
mine if there is an acoustic benefit for the
cicadas calling in burrows. Peak sound pres-
sure levels (SPL) were recorded with a Briiel
& Kjaer 2235 SPL meter, a Type 4155 1/2"
prepolarized condenser microphone, and a
UA 0237 wind screen. The system had been
calibrated with a Briiel & Kjaer 4230 portable
sound pressure calibrator. The SPL meter was
used in the linear frequency mode. The peak
setting has a time constant of less than 100 ms

and was used to ensure that rapid sound tran-
sients were measured. Measurements were
made peipendicular to the long body axis with
the apparatus oriented medially along the dor-
sal surface of a singing cicada at the thorax-
abdomen junction or directly above the hole
in which a cicada was singing. This procedure
minimized any inconsistencies between readings
due to possible asymmetries in the sound field
produced by cicadas (Aidley 1969, MacNally
and Young 1981). Each intensity measurement
was made at a distance of 50 cm. The distance
was kept constant by placing a 1/4" (6.5 mm)
dowel, attached to the SPL meter, near a call-
ing cicada. If the cicada was disturbed by
placement of the SPL meter, the reading was
made only after the normal calling pattern had
been reestablished. All intensity measure-
ments are relative to 1 X IQ-^^ W/cm^.

Power output was determined using the
following equation:

Q = 47Cr2(I)

where Q = sound power (W), r = distance
from source in cm (= 50 cm), and I = intensi-
ty reading for the individual (dB). Since inten-
sity is measured on a logarithmic scale, all
intensity measurements (dB readings) were
converted to pressure levels (W/cm^) prior to
calculating the statistics. Mean power output
was then used to calculate mean sound inten-
sity at 50 cm for each species.

Intensity measurements are summarized in
Table 1. SPL values recorded for cicadas call-
ing from within a burrow are lower than val-
ues recorded when the animals were calling
from a plant; however, the values are not sig-
nificandy different {t = 1.49, d.f = 3, F =
.1159). A greater number of trials ma\' provide

'Bari-y University, School of Natural and lli'altli Scicnt'i's, 11300 N.E^ Second A\enue, Miami Shores, FL 33101-6695.
2Mianii-Dade Coninuinity College North C;anipns, Biology Department, 11380 N.W. 27th .\venne, Miami, FL .33167-3495.

374

1995]

Notes

375

Table 1. Intensity of Okan(i<j^ana pallidtila calling songs.
Song power was calculated from individual measure-
ments. Mean power was used to determine the mean
sound pressure level for each perch. Measurements are
relative to 1 X lO-lf* VV/cm^.

Location of
calling perch

Sound
power

(mW)

Mean

intensity

(dB)'

Range
(dB)

Exposed calls

1.106 ±.6823

(" = 3)

Subterranean calls 0.340 ± . 1 190

{n = 2)

85.46 83.4-87.8

80.34 79.1-81.3

the statistical significance. We were unable to
find a large number of individuals singing
from burrows. Similarly, the vast majority of
specimens calling from plants did not permit
our approach to a distance necessaiy to collect
a greater number of intensity measurements.

Subterranean calling has been described
for the cicadas O. pallidiila (Davis 1917, 1930,
1944, Beamer and Beamer 1930) and O. von-
diizeei (Distant) (Beamer and Beamer 1930).
Crickets have also been shown to call from
burrows (Bennet-Clark 1970, 1987, Ulagaraj
1976, Forrest 1983, Kavanaugh 1987, Walker
and Figg 1990); however, burrows fashioned
by crickets increase the intensity level of their
song (Bennet-Clark 1987). Subterranean call-
ing in O. pallidido appears to decrease song
intensity rather than increase sound transmis-
sion.

Subterranean calling may function as a
predator avoidance behavior in O. pallidido.
The difficulty we had in extracting a cicada
from its burrow supports this assumption for
vertebrate parasitoids. Burrows appear to be
connected to nymphal development chambers,
which can reach a depth of 2 ft (61 cm; Beamer
and Beamer 1930). They represent a relatively
safe location from which to call and may pro-
vide protection from acoustically orienting
parasitoids. Many predators have been shown
experimentally to orient to the calls of cicadas
(Soper et al. 1976) and other acoustic insects
(Walker 1964, Cade, 1975, Mangold 1978, Bell
1979, Buchler and Childs 1981, Sakaluk and
Belwood 1984, Fowler and Kochalka 1985,
Tutde et al. 1985, Belwood and Morris 1987).
The protective value provided against dipter-
an or other parasitoids to cicadas calling from
burrows remains to be determined.

The majority of the population appears not
to call from a burrow as indicated by the pro-

portion of individuals we found calling from
plants versus those calling from burrows.
Perhaps the benefits of a protected calling site
limit the chances of a male to attract a female
successfully. The males may eventually aban-
don their burrows to increase the chances of
interacting with females. It would be interest-
ing to investigate the ontogeny of subterra-
nean calling in cicadas.

Acknowledgments

We thank Maxine Heath for her taxonomic
assistance. The financial support of Sr. John
Karen Frei is appreciated. Tom Moore and an
anonymous reviewer made suggestions that
improved the manuscript.

Literature Cited

AlDLEY, D. J. 1969. Sound production in a Brazilian cica-
da. Journal of E.xperimental Biology 51: 325-337.

Beamer, L. D., and R. H. Beamer. 1930. Biological notes
on some western cicadas. Journal of the New York
Entomological Society 38: 291-305.

Bell, R D. 1979. Acoustic attraction of herons by crickets.
Journal of the New York Entomological Society 87:
126-127.

Belwood, J. J., and G. K. Morris. 1987. Bat predation
and its influence on calling behavior in Neotropical
katydids. Science 238: 64-67.

Bennet-Clark, H. C. 1970. The mechanism and efficien-
cy of sound production in mole crickets. Journal of
E.xperimental Biolog\' 52: 619-652.

. 1987. The tuned singing burrow of mole crickets.

Journal of E.xperimental Biology 128: 383^09.

Bennet-Clark, H. C, and D. Young. 1992. A model of
the mechanism of sound production in cicadas.
Journal of E.xperimental Biology' 173: 123-153.

Buchler, E. R., \nd S. B. Childs. 1981, Orientation to
distant sounds by foraging big brown bats {Eptesicus
fiiscits). Animal Behaviour 29: 428-432.

Cade, W. 1975. Acoustically orienting parasitoids: fly
phonotaxis to cricket song. Science 190: 1312-1313.

Davis, W. T. 1917. Sonoran cicadas collected by Hany H.
Knight, Dr. Joseph Bequaert and others, with descrip-
tions of new species. Jounid of die New York Entomo-
logical SocietA' 25: 203-215.

. 1930. The distribution of cicadas in the United

States with descriptions of new species. Journal of
the New York Entomological Society 38: 53-72.
. 1944. The remarkal)le distribution of an American

cicada: a new genus, and other cicada notes. Joimial
of the New York Entomological Society 52: 213-222.

Forrest, T. G. 1983. Calling songs and mate choice in
mole crickets. Pages 185-204 in D. T. Cwynne and
G. K. Morris, editors, Orthopteran mating systems:
sexual competition in a diverse group of insects.
Westview Press, Boulder, CO.

Fowler, H. G., and J. N. Kochalka. 1985. New record of
Eiiphasiopteryx depleta (Diptera: Tachanidae) from
Paraguay: attraction to broadcast calls of Scapteriscus

376

Great Basin Naturalist

[Volume 55

acletu.s (Orthoptera: Cnllofali^idaej. Florida Hiitoiii-
ologist 68: 225-226.

KavanaUCH, M. W. 1987. Tlie i^liicit'iic) oi sound piodiic-
tion in two cricket species, CniUotalpa australis and
Teleogrijllus comrnodus ((Jrtlioptera: Giylloidea).
Journal of E.\perimental Biology 130: 107-119.

M.-vcNallv, R. C, and D. Younc;. 1981. Song energetics
of the bladder cicada, Ctjstosorna sauiulcrsii. Journal
ofE.xperinicnlal Biology- 90: 185-196.

Mangold, J. R. 1978. Attraction of Euphasioptenjx ocliro-
cea, Corethrella sp. and giyllids to broadcast songs
of the southern mole cricket. Florida Entomologist
61:57-61.

Pringle, J. VV. S. 1954. A physiological analysis of cicada
song. Journal of Experimental Biology 31: 525-560.

Sakaluk, S. K., and J. J. Belvvood. 1984. Gecko phono-
taxis to cricket calling song: a case of satellite preda-
tion. Animal Behaviour 32: 659-662.

SoPER, R. S., G. E. Shewell, and D. Tyrrell. 1976. Col-
condamtjia auditrix nov. sp. (Diptera: Sarcophagidae),
a parasite which is attracted by the mating song of
its host, Okano'^ana riinosa (Homoptera: Cicadidae).
Canadian Entomologist 108: 61-68.

Tlttle, M. D., M. J. Ryan, and J. J. Belwood. 1985.
Acoustical resource partitioning by two species of
phyllostomid bats {Trachops cirrhosits and Tonatia
sylvicola). Animal Behaviour 33: 1369-1371.

Ul.\GAR.'\J, S. M. 1976. Sound production in mole crickets
(Orthoptera: Gr\'llotalpidae: Scapteriscus). Annals of
the Entomological Societ>' of America 69: 299-.306.

Walker, T. J. 1964. Experimental demonstration of a cat
locating orthopteran prey by the prey s calling song.
Florida Entomologist 47: 163-165.

Walker, T. J., and D. E. Figg. 1990. Song and acoustic
burrow of the prairie mole cricket, Gnjllotalpa major
(Orthoptera: Gryllidae). Journal of the Kansas Ento-
mological Society 63: 237-242.

Young, D. 1990. Do cicadas radiate sound through their
ear-drums? Journal of Experimental Biology 151:
41-56.

Received 24 March 1995
Accepted 28 June 1995

BOOK REVIEW

Natural History of the Colorado Plateau and
Great Basin. K. T. Haiper, L. L. St. Claii;
K. H. Thorne, and W. M. Hess, editors.
University Press of Colorado, Niwot, CO.
1994. 294 pp. $24.95 hardbaek.

Natural History of the Colorado Plateau
and Great Basin, a multi-authored volume, is an
introduction to the spectacular arid and remote
North American landscape known as the Colo-
rado Plateau and the Great Basin. The high,
windswept plateau country is interrupted by
numerous rocky canyons and arid valleys, and
the Great Basin is a huge arid depression with
no external drainages. According to the editors,
this region is within the boundaries of Nevada,
Utah, and Colorado (Fig. 1.2). They indicate
that the intended audience of the volume
includes students and managers of the region's
natural resources. The basic objective of the
major eleven chapters is to provide a "ready
reference to the best of recent studies that are
relevant to the region. " Additionally, the editors
hope this volume will stimulate more research,
especially on the Colorado Plateau, which is
more "biodiverse and perhaps more fragile
ecologically than the Great Basin.

The map in Chapter 1 of the Great Basin
and Colorado Plateau indicates a smaller region
than maps in Chapters 2, 5, and 9, excluding
areas as far noilli as Oregon and as far south as
California, Arizona, and New Mexico. The
boundaries of the Great Basin and Colorado
Plateau therefore appear mildly confusing. A
consensus map or better textual description
(as presented in Chapter 5) could have been
included in Chapter 1.

Chapter 2 presents a rather concise and
useful review of the geologic history of the
Great Basin and Colorado Plateau. Chapter 3
attempts to summarize the complex climatic
weather patterns in the broad context of the
western United States and the globe. Anyone
who has spent time in the Great Basin or
Colorado Plateau knows well the unpredict-
able and often extreme weather patterns that

have helped form the regional geomorphology.
Literature citations of this chapter are very
useful.

Chapter 4 reviews the extinct late Pleisto-
cene mammals of the Great Basin. This region
is rich in late Pleistocene vertebrate fossils,
and the author provides a discussion of the
possible causes of extinction and implications
concerning present faunas. Western Great
Basin archaeology in the context of regional
cultural/environmental models is presented in
Chapter 5. Wilde describes various prehistoric
ebbs and flows of peoples for the past 12,000
years. Chapter 6 touches on the current politi-
cally controversial subject of the changes in
plant communities caused by domestic live-
stock grazing, the most widespread land-man-
agement practice in western North America.
Seventy percent of the western United States
is grazed, and ecological costs have been
great. The author seems to concentrate on
deleterious effects of the introduction of alien
plants species such as Russian thistle and
cheatgrass on the Great Basin plant communi-
ties. He predicts that with continued removal
of cattle, the "predators" of these plants, the
ecosystem structure of the Great Basin may
dramatically change in the near future.

In Chapter 7 Jackson presents an enjoyable
analysis of the unique factors that have influ-
enced modem human development of resources
in the Great Basin. He traces the cultural his-
tory of the region, from the Dominguez and
Escalante expeditions of the 1700s to the
Mormon farmers who shaped the modern
human geography of tlie Great Basin. As Jackson
pointed out, the enduring legacy of the Great
Basin is the "strange juxtaposition of religion
and vice, destruction and recreation.' The
authors in Chapter 8 use macrofossil data from
packrat {Neotoma) middens to reconstruct tlie
evolutionary history of eight modern conifer
species. These conifer species now occupy the
montane islands of the Great Basin, and the
current distribution of these trees is related to
past paleoclimatic changes.

377

378

Great Basin Naturalist

[Volume 55

Sigler and Sigler in Chaper 9 present a ver\'
comprehensive review of the fishes of the Great
Basin and the Colorado Plateau. Excellent dis-
cussions are presented for each species.
However, there appear to be some errors; for
example, the Big Spring spinedace is a native
to the Colorado River Basin, not Lahontan, and
the razorback sucker is a federalh' endangered
species as of 1991. Additionally, if the map pre-
sented in this chapter is inclusixe, then perhaps
several other species could be added: Moapa
dace, Moapa speckled dace. Meadow Valley
speckled dace, Preston speckled dace, White
River sucker, and Sonora sucker Also, I cannot
construe the meaning of the last sentence in
their chapter, "that many of the species, both
native and exotic, have survived in spite of
[human] modifications." The fact is, at least for
the Colorado River Basin, most native fishes
are in serious jeopardy of extinction; they have
survived, but with a veiy precarious hold.

Chapter 10 by Nelson attempts to cover a
daunting subject, the insects of the Great
Basin and Colorado Plateau. An estimated
14,000-26,000 species may occur within these
boundaries. He discusses several of the better
regional known taxonomic groups (stoneflies,
butterflies, robber flies, and ants) to answer
broad questions, such as, "What range patterns
are seen in the Great Basin and Colorado
Plateau?" and "Did these groups evolve in the
Great Basin?" Many of the insects of this region
have a widespread distribution throughout the
West, and the insect fauna of the Colorado
Plateau have strong affinities with the Rocky
Mountains physiographic province. Warren
and Harper in Chapter 11 briefly discuss ele-
vational patterns of insects in the Great Basin

and Colorado Plateau. Most of their examples,
however, are higher elevational patterns of the
Rocky Mountains and elsewhere, and the dis-
cussion is limited to adaptations of insects to
harsh environments. Their literature review is
excellent.

In Chapter 12, Mead and Bell describe the
heipetofauna of the Great Basin and Colorado
Plateau in the late Pleistocene and Holocene
(i.e., during the past two million years, or
Quaternaiy Period). Their comparison of mod-
em fauna with the Pleistocene-Holocene indi-
cates that 61% of the modern fauna is repre-
sented in the fossil record, an interesting
observation considering the climatic and envi-
ronmental change in association with such
events as ice ages.

In Chapter 13 the editors provide recom-
mendations for future directions of research,
emphasizing the need for descriptive work.
They also state evolutionary and ecological
questions about the biodiversity of the Great
Basin and Colorado Plateau that need urgent
attention.

This little book packs in much useful infor-
mation, and with its reasonable price it should
appeal to all students who work or visit the
Intermountain West. The editors have suc-
ceeded in presenting a good introduction to
many important and conspicuous aspects of
the natural history of the Great Basin and
Colorado Plateau.

B. C. Kondratiefif
Colorado State University
Department of Entomology
Fort Collins, CO 80523

H E

GREAT BASIN

NATUKALIST

INDEX

VOLUME 55 — 1995

BRIGHAM YOUNG UNIVERSITY

INDEX

Volume 55—1995

Author Index

Anderson, Loran C, 84
Anstin, Dennis D., 267

Baker, William L., 287
Bartholomew, Breck, 282
Baumann, R. W, 124
Belk, Mark C, 183
Bodie, Walt, 181
Bowlin, W. R., 19
Brusven, Merlyn A., 352

Callahan, J. R., 89
Carter, Bernard, 169
Cates, Rex G., 29
Cieminski, Karen L., 105
Cifelli, Riehard L., 304
Clements, Charlie, 188
Compton, Stephen B., 89
Crawford, John A., 284
Crompton, Clifford, 322
Czaplewski, Nicholas J., 304

DeBolt, Ann Marie, 237
DeWalt, R. Edward, 1
Dobkin, David S., 315

Ehlerin^er, James R., 135
Elphick, Chris S., 363

Fielding, Dennis J., 352
Flake, Lester D., 105
Flinders, Jerran T, 29
Friedman, Jonathan M., 58
Furniss, Malcolm M., 335

Geer, S. M., 19
Gerdes, Michael G., 315
Gettinger, Ronald D., 315
Goldberg, Stephen R., 372
Goodrich, Sherel, 359
Griswold, T. L., 19

Hansen, E. Matthew, 158
Haiper, Kimball T. (rev.), 286
Heckmann, Richard A., 258
Hubert, Wayne A., 169

Ischinger, Lee S., 58

Johnson, James B., 335
Johnson, Jerald B., 183

KondratiefT, B. C. (rev.), 377
Kucera, James R., 92

Lesica, Peter, 142
Longland, William S., 188
Lytle, C. Mel, 164

Malmos, Keith, 368
McArthnr, E. Durant, 151
McCoy, Matthew, 181
McCune, Bruce, 237
Meinke, Robert J., 249
Miller, Richard F, 37
Minshall, G. Wayne, 193
Munger, James C, 74
Muth, Robert T, 95

Nohavec, Robert D., 282

Owen, Wayne R., 117

Pelren, Eric C, 284
Phillips, Polly K., 192, 374

Qi, Ying, 258

Ratliff, Ra\inond D., 46
Reed, J. Michael, 342
Reed, Robert, 368
Renkin, Roy A., 201
Robinson, Christopher T, 193
Rose, JeffeiT A., 37

380

1994]

Index

381

Rose, Kenneth D., 304
Royer, Todd V, 193
Rubega, Margaret A., 363
Rushforth, Samuel R., 193

Sanborn, Allen E, 192, 374
Sandquist, Darren R., 135
Shafroth, Patrick B., 58
Shepard, W. D., 124
Shiozawa, Dennis K., 183, 213
Singer, Francis J., 201
Slichter, Todd A., 74
Smith, Bruce N., 164
Snyder, Darrel E., 95
Starrett, Bryan, 368
Stewart, Kenneth W, 1

Storz, Jay F, 78
Stricklan, Dave, 29

Taye, Alan C, 225
Taylor, El Roy 181
Tepedino, V. J., 19

Urness, Philip J., 267

Vicker>', Robert K., Jr., 174, 177

Walford, Gillian M., 287

Welsh, Stanley L., 66, 271, 322, 359

Winward, Alma H., 151

Wolz, Eric R., 213

Key Word Index

Taxa described as new to science in this volume appear in boldface t>'pe in this index.

acoustic behavior, 374
Acrididae, 352
age, 183

allopolyploid, 151
alpine, 117

vascular flora, 225

vegetation, 225
analysis

elasticity, 142

multivariate, 287
aneuploidy, 174
Anura, 368

Apiosorna campamdatum, 258
aquatic invertebrates, 105
Arizona, 372
Artemisia

arhuscula ssp. longicaulis, 151

nova van duchesnicola, 359
Astragalus, 117, 142
Atriplex types, 322
avian diversity, 342
avoidance

predator, 374

back-waters, 95
banded sand snake, 372
behavior, 282

acoustic, 374

feeding, 192
benthic habitat, 193
benthic macroinvertebrates, 213
benthos, 213

big sagebrush browsing, 210
bioaccumulation

metal, 164
Blue Grouse, 284

browse (lirowsing), 267
big sagebrush, 210

Bufo

woodhousii, 368
punctatus, 368

bumblebees, 177

bundle sheath leakiness, 135

California

bighorn sheep, 181

Gull, 363

Mono Lake, 363

White Mountains, 117
Calileuctra, 124
calling, 374
candidate species, 315
carbon isotope ratio, 135
caves, 315

Celtis reticulata, 237
Ceratopogonidae, 213
Cervus elaphus, 201
chemotaxonomy, 151
Chenopodiaceae, 322
Chilomeniscus cinctus, 372
Chironomidae, 213
cicada, 374
coevolution, 188
Colorado, 287

Green River, 95

Plateau, 225
Colubridae, 372
competition, 117
conservation priorities, 342
Cottus bairdi, 258
Crotahis scutulatus, 282
Cijprinclla lutrensis, 95

382

Great Basin Naturalist

[Volume 54

death canius, 188
defoliators, 158
demography, 142
Dendragapiis ohscunis, 284
density

population, 352
description, 124
desert, 169

ecology, 135
diatoms, 193
diet(s), 95, 363

overlap, 95
distribution, 124, 169, 368
disturbance, 193
diversity

avian, 342
Douglas-fir, 158
Douglas rabbitbrush, 267
Dragon local fauna, 304
drought, 352

ecological aspects, 258
ecolog\', 237
elasticit\' analysis, 142
elk

northern Yellowstone, 201
Erythranthe, 177
escape saltation, 282
evolution, 174

fauna

Dragon local, 304

Wagon Road local, 304
faunal list, 335
fecundity, 117
feeding behavior, 192
fish(es), 169

nonnative, 95
flower color mutations, 177
food source, 192
foraging, 192

efficiency, 363

movements, 315
forest insects, 158

Gila copci, 183
Great Basin, 249, 342
Green River, [Colorado], 95
Green River, [Utah], 95, 213
growth, 183, 237

habitat, 169

benthic, 193
halophyte, 135
herbivory, 142
hibernacula, 315
horses, 267

host-symbiont relationship, 258
hummingbirds, 177

hybrid, 151
h\bridizati()n, 368

Idaho, 181, 237

National Engineering Laboratoiy, 105
industrial wastewater, 105
Insecta. 124
interception trap, 284
invertebrates

aquatic, 105

Lams culijurnicus, 363

lava flows, 315

leatherside chub, 183

Leguminosae, 271

Lepidium huberi, 359

Leuctridae, 124

life histoiy, 183

livestock grazing, 142, 237, 352

longevity, 237

macroinvertebrates, 193

benthic, 213
Mammalia, 304
matrix projection models, 142
Melanophis san<guimpes, 352
metal bioaccumulation, 164
Miinuhis, 174, 177, 249

breviflonis, 249

evanescens, 249

latidens, 249
monkeyflowers, 249
Mono Lake, [California], 363
montane islands, 342
moiphology, 258
mountain big sagebrush, 267
mule deer, 188, 201, 267
multivariate analysis, 287
mutations

flower color, 177

Natural Histoiy Museum, 271

Nematoda, 213

netleaf hackberry, 237

Nevada, 151

new taxa, 359

nomenclature, 271

nonnative fishes, 95

North America, 322

North Horn Formation, [Utah], 304

northern Yellowstone elk, 201

nutritional qualities, 164

Okanagaiui palUdiila, 374
Oligochaeta, 213
Oregon, 284

central, 315
Orthoptera, 352
Ovis canadensis, 181
Oxytropis deflexa van pulcherrima, 271

1994]

Index

383

Paleocene, 304
parturition, 181
phylogeii)', 124
plant(s)

geography, 225

poisonous, 188

rare, 142
Plecoptera, 124
Plecotiis townsendii, 315
pollinators, 177
pohploidy, 174
population density, 352
Potamogefon pcctinatiis, 164
predation, 142
predator avoidance, 374
pronghorn, 201
Provo River, [Utah], 258
Ptychochdus lucius, 95
Puercan, 304

radiotelemetry, 315
range

management, 267, 352

winter, 267
rare plant, 142
recruitment, 237
rehabilitation, 237
relationship

host-symhiont, 258
reproduction, 183, 372
reproductive isolation, 177
revegetation, 267
riparian vegetation, 287
river benthos, 213
Rock Wrens, 192
Rocky Mountains, 287
roost

fidelity, 315

sites, 315
Royal Botanic Garden, 271

sagebrush

browsing, 210

mountain big, 267
sago pondweed, 164
Solpinctes obsoletiis, 192
salt stress, 135
sanitaiy wastewater, 105
Scolyti'dae, 335
Scrophulariaceae, 249
shrubs, 267

site characteristics, 237
sound pressure level, 374

speciation, 174, 177
stand structure, 237
streams, 169, 193
subalpine fir, 158

taxonomy, 151, 249

toads, 368

Torrejonian, 304

Townsend's big-eared bat, 315

trace element cycling, 164

trade-off, 117

trap

interception, 284

walk-in, 284
trapping success, 284
Trichodina tenuifonnis, 258
Tridentatae, 151
Troglodytidae, 192
true mountain mahogany, 267
Tushar Mountains, [Utah], 225
tussock moth, 158

Uinta Basin, [Utah] 359

Utah, 225, 267, 359
forests, 158
Green River, 95, 213
North Horn Formation, 304
Pro\'o River, 258
Tushar Mountains, 225
Uinta Basin, 359

vegetation

riparian, 387
vulnerability, 342

Wagon Road local fauna, 304
walk-in trap, 284
Washington state, 335
wastewater

industrial, 105

sanitaiy, 105
waterfowl, 164
wedand, 213

White Mountains, [Cahfornia], 117
wildfire, 193
winter range, 267
Wyoming, 169

Yellowstone National Park, 193
young- of-year, 95

Zigadenits, 188

384 Great Basin Naturalist [Volume 54

TABLE OF CONTENTS
Volume 55

No. 1— January 1995

Articles

Life histories of stoneflies (Plecoptera) in the Rio Conejos of southern Colorado

R. Edward DeWalt and Kenneth W. Stewart 1

Pollinator sliarint^ 1)\ three s\nipatrie milkvetches, including the endangered species Astragalus

tnontii S. M. Geer, V. J. Tepedino, T. L. Griswold, and W. R. Bowlin 1 9

Factors affecting selection of winter food and roosting resources by porcupines in Utah

Dave Stricklan, Jerran T. Flinders, and Rex G. Gates 29

Historic expansion (A Jiniipcnis occidcntalis (western juniper) in southeastern Oregon

Richard E Miller and Jeffeiy A. Rose 37

Rangeland alpha diversities: Harvey Valley Lassen National Forest, California

Raymond D. Ratliff 46

Effects of salinity on establishment of Populus freinontii (cottonwood) and Tainarix ramosissiina

(saltcedar) in southwestern United States Patrick B. Shafroth, Jonathan M. Friedman,

and Lee S. Ischinger 58

Names and types of Hedysanim L. (Fabaceae) in North America Stanley L. Welsh 66

Whipworm [Trichuris dipodovujs) infection in kangaroo rats {Dipoduinijs spp.): effects on digestive

efficiency James C. Munger and Todd A. Slichter 74

Local distribution and foraging behavior of the spotted bat {Eiidenna inaciiJatum) in northwestern

Colorado and adjacent Utah Jay F Storz 78

The Chrysothamnus-Ericameria connection (Asteraceae) Loran G. Anderson 84

Notes

Reproductive behavior in Merriam s chipmunk {Tamias )nenia)ni) Stephen B. Compton

and J. R. Callahan 89

Additional records of fleas (Siphonaptera) from Utah James R. Kucera 92

No. 2— April 1995
Articles

Diets of young Colorado squaw^ish and other small fish in back'waters of the Green River, Colorado

and Utah Robert T. Muth and Darrel E. Snyder 95

Invertebrate fauna of wastewater ponds in southeastern Idaho Karen L. Cieminski

and Lester D. Flake 1 05

Growth and reproduction in an alpine cushion pldnV. Astragalus koitroplujta var. uuplexus

Wayne R. Owen 117

Calileuctra, a new genus, and two new species of stoneflies from California (Plecoptera: Leuctridae)

WD. Shepard and R. W Baumann 1 24

Carbon isotope discrimination in the G4 shrub Atriplcx confertifolia along a salinity gradient

Darren R. Sandquist and James R. Ehleringer 1 35

Demography oi Astragalus scaphoides and effects of herbivory on population growth . . . Peter Lesica 1 42

Lahontan sagebrush [Artemisia arhuscula ssp. longicaulis): a new taxon Alma H. Winward

and E. Durant McArthur 151

1994] Index 385

Douglas-fir tussock niotli [Orgijia pseiidotsugata McDunnough) on suhalpine fir in northern Utah

E. Matthew Hansen 1 58

Seasonal nutrient c\'cling in Potamogeton pectinatus of the lower Provo River C. Mel Lytle

and Bruce N. Smith 1 64

Factors influencing fish assemblages of a high-elevation desert stream system inWyoming

Bernard Carter and Wayne A. Hubert 1 69

Notes

Speciation by aneuploidy and polyploidy in Mimulus (Scrophulariaceae) .... Robert K. Vicker>', Jr 1 74

Speciation in Mimulus, or. Can a simple flower color mutant lead to species divergence?

Robert K. Vickeiy Jr 1 77

Fall lamb production by a California bighorn sheep Matthew McCoy, Walt Bodie,

and ElRoy Taylor 181

Age, growth, and reproduction of leatherside chub [Gila copei) . . Jerald B. Johnson, Mark C. Belk,

and Dennis K. Shiozawa 1 83

Consumption of a toxic plant [Zlgadenns paniculatus) by mule deer William S. Longland

and Charlie Clements 1 88

Use of an unusual food source by Rock Wrens (Troglodytidae) Polly K. Phillips

and Allen E Sanborn 1 92

No. 3— July 1995
Articles

Benthic community structure in two adjacent streams in Yellowstone National Park five years after

the 1988 wildfires G. Wayne Minshall, Christopher T. Robinson, Todd V. Royer

and Samuel R. Rushforth 1 93

Effects of browsing by native ungulates on the shrubs in big sagebrush communities in Yellow-
stone National Park Francis J. Singer and Roy A. Renkin 201

Soft sediment benthic macroinvertebrate communities of the Green River at the Ouray National

Wildlife Refuge, Uintah County, Utah Eric R. Wolz and Dennis K. Shiozawa 213

Alpine vascular flora of the Tushar Mountains, Utah Alan C. Taye 225

Ecology oiCeltis reticulata in Idaho Ann Marie DeBolt and Bruce McCune 237

Mimulus evanescens (Scrophulariaceae): a new annual species from the northern Great Basin

Robert J. Meinke 249

Moiphological and host-symbiont studies of Trichodina teiiuifonnis and Apiosoma campanulatum

infesting motded sculpin {Cottus bairdi) fi-om Provo River, Utali Ying Qi

and Richard A. Heckmann 258

Effects of horse grazing in spring on sunival, recruitment, and winter injury damage of shrubs . . .

Dennis D. Austin and Philip J. Urness 267

North American types of Oxytropis DC. (Leguminosae) at The Natural Histoiy Museum and Royal

Botanic Garden, England, with nomenclatural comments and a new variety S. L. Welsh 271

Notes

Saltation in snakes with a note on escape saltation in a Crotalus scutulatus Breck Bartholomew

and Robert D. Nohavec 282

A trap for Blue Grouse Eric C. Pelren and John A. Crawford 284

,386 Great Basin Naturalist [Volume 54

Book Review

Mountains and plains: tlie ecology of Wyoniin^ landscapes Dennis H. Knif^ht

Kimhall T. Harper 286

No. 4— October 1995
Articles

classification of the riparian vegetation along a 6-km reach of the Animas River, southwestern

Colorado Gillian M. Walford and Willian L. Baker 287

Additions to knowledge of Paleocene manmials from the North Horn Formation, central Utah ....

Richard L. Cifelli, Nicholas J. Czaplewski, and Kenneth D. Rose 304

Springtime movements, roost use, and foraging activity of Townsend's big-eared bat {Plecotus

toioisendii) in central Oregon David S. Dobkin, Ronald D. Gettinger,

and M ichael G. Gerdes 3 1 5

Names and types in perennial AtripJcx Linnaeus (Chenopodiaceae) in North America selectively

exclusive of Me.xico Stanley L. Welsh and Clifford Crompton 322

New records of ScoKtidae from Washington state Malcolm M. Furniss and James B. Johnson 335

Relative vulnerability to extiipation of montane breeding birds in tlie Great Basin ... J. Michael Reed 342

Grasshopper densities on grazed and ungrazed rangeland under drought conditions in southern

Idaho Dennis J. Fielding and Merlyn A. Brusven 352

Plant novelties in Lepidiwn (Cruciferae) and Arte^nisia (Compositae) from the Uinta Basin, Utah . . .

Stanley L. Welsh and Sherel Goodrich 359

Prey choices and foraging efficiency of recently fledged California Gulls at Mono Lake, California

Chris S. Elphick and Margaret A. Rubega 363

Notes

Hybridization between Bitfo woodhousii and Biifo pitnctatiis from the Grand Canyon region of

Arizona Keith Malmos, Robert Reed, and Bryan Starrett 368

Reproduction in the banded sand snake, Chilonieniscus cinctits (Colubridae), from Arizona

Stephen R. Goldberg 372

No acoustic benefit to subterranean calling in the cicada Okanogana poUiduIa Davis (Homoptera:

Tibicinidae) Allen F Sanborn and Polly K. Phillips 374

Bool< Review

Natural histoiy of the Colorado Plateau and Great Basin K. T. Harper, L. L. St. Clair K. H. Thome,

and W. M. Hess B. C. Kondratieff 377

!i76 t5 6

INFORMATION FOR AUTHORS

SUBMIT MANUSCRIPTS to Richard W. Baumann,
Editor, Great Basin Naturalist, 290 MLBM, PO Box
20200, Brigham Young University, Provo, UT
84602-0200. A cover letter accompanying the man-
uscript must include phone number(s) of the author
submitting the manuscript, and FAX number and
E-mail address when applicable; the letter must
also provide information describing the extent to
which data, text, or illustrations have been used in
other papers or books that are published, in press,
submitted, or soon to be submitted elsewhere.
Authors should adhere to the followdng guidelines;
manuscripts not so prepared may be returned for
revision.

MANUSCRIPT PREPARATION. Consult Vol. 51, No.
2 of this journal for specific instructions on style
and format. These instructions, GUIDELINES for
Manuscripts Submitted to the Great Basin
Naturalist, supply greater detail than those pre-
sented here in condensed form. The guidelines are
printed at the back of the issue; additional copies
are available from the editor upon request. Also,
check the most recent issue of the Great Basin
Naturalist for changes and refer to the CBE Style
Manual, 5th edition (Council of Biology Editors,
One Illinois Center, Suite 200, 111 East Wacker
Drive, Chicago, IL 60601-4298 USA, phone: (312)
616-0800, fax: (312) 616-0226; $24).

ABSTRACT states the purpose, methods, results,
and conclusions of the research. It is followed by
6-12 key words, listed in order of decreasing
importance, to be used for indexing.

TEXT has centered main headings printed in all
capital letters; second-level headings are centered
in upper- and lowercase letters; third-level head-
ings begin paragraphs.

REFERENCES IN THE TEXT are cited bv author and
date: e.g., Martin (1989) or (Martin 1989). Multiple
citations should be separated by commas and listed
in chronological order. Use "et al. " after name of
first author for citations having more than two
authors.

ACKNOWLEDGMENTS, under a centered main
heading, include special publication numbers when
appropriate.

LITERATURE CITED, also under a centered main
heading, lists references alphabetically in the fol-
lowing formats:

Mack, G. D., and L. D. Flake. 1980. Habitat rela-
tionships of waterfowl broods on South Dakota
stock ponds. Journal of Wildlife Management
44: 695-700.

Sousa, W. P 1985. Disturbance and patch dynamics
on rocky intertidal shores. Pages 101-124 in
S. T A. Pickett and P S. White, eds.. The ecolo-
gy of natural disturbance and patch dynamics.
Academic Press, New York.

Coulson, R. N., and J. A. Witter. 1984. Forest ento-
mology: ecology and management. John Wiley
and Sons, Inc., New York. 669 pp.

TABLES are double spaced on separate sheets and
designed to fit the width of either a single column
or a page. Use lowercase letters to indicate foot-
notes.

NOTES. If the manuscript would be more appro-
priate as a short communication or note, follow the
above instructions but do not include an abstract.

FINAL CHECK:

• Cover letter explaining any duplication of
information and providing phone number(s),
FAX number, and E-mail address

• 3 copies of the manuscript and WordPerfect
diskette

• Confonnity with instructions

• Photocopies of illustrations

(ISSN 001 7-3614)

GREAT BASIN NATURALIST Vol 55 no 4 October 1995

CONTENTS

Articles

classification of the riparian vegetation along a 6-km reach of the Animas River,

southwestern Colorado Gillian M. Walford and Willian L. Baker 287

Additions to knowledge of Paleocene mammals from the North Horn Formation,

central Utah Richard L. Cifelli, Nicholas J. Czaplewski,

and Kenneth D. Rose 304

Springtime movements, roost use, and foraging activity of Townsend's big-eared

bat {Plecotus townsendii) in central Oregon David S. Dobkin,

Ronald D. Gettinger, and Michael G. Gerdes 315

Names and types in perennial Atriplex Linnaeus (Chenopodiaceae) in North

America selectively exclusive of Mexico Stanley L. Welsh

and Clifford Crompton 322

New records of Scolytidae from Washington state Malcolm M. Fumiss

and James B. Johnson 335

Relative vulnerability to extirpation of montane breeding birds in the Great

Basin J. Michael Reed 342

Grasshopper densities on grazed and ungrazed rangeland under drought condi-
tions in southern Idaho Dennis J. Fielding and Merlyn A. Brusven 352

Plant novelties in Lepidium (Cruciferae) and Artemisia (Compositae) from the

Uinta Basin, Utah Stanley L. Welsh and Sherel Goodrich 359

Prey choices and foraging efficiency of recently fledged California Gulls at Mono

Lake, California Chris S. Elphick and Margaret A. Rubega 363

Notes

Hybridization between Bufo woodhotisii and Bufo punctatus from the Grand

Canyon region of Arizona Keith Malmos, Robert Reed,

and Bryan Starrett 368

Reproduction in the banded sand snake, Chilomeniscus cinctus (Colubridae), from

Arizona Stephen R. Goldberg 372

No acoustic benefit to subterranean calling in the cicada Okanagana paUidida

Davis (Homoptera: Tibicinidae) Allen E Sanborn and Polly K. Phillips 374

Book Review

Natural history of the Colorado Plateau and Great Basin K. T. Harper, L. L.

St. Clair, K. H. Thome, and W. M. Hess B. C. Kondratieff 377

Index to Volume 55 379

3 2044 072 231 152

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The Great Basin naturalist"

See other formats