Color profile: Generic CMYK printer profile
Composite Default screen
201
The vegetation and ecological gradients of
calcareous mires in the South Park valley,
Colorado
J. Bradley Johnson and David A. Steingraeber
Abstract: The vegetation, environment, and ecological gradients present on three calcareous mires in the South Park
valley, Park County, Colorado, were investigated. Vegetation was classified into four habitat classes, nine subclasses,
and twelve species associations using two-way species indicator analysis (TWINSPAN). Detrended correspondence
analysis (DCA) was used to ordinate vegetation samples along two axes representing the three predominant ecological
gradients: water table height, miremargin to expanse, and region. Canonical correspondence analysis (CCA) was used
to directly relate local environmental conditions to vegetation. Water table depth, microtopographical development, soil
and water pH and nutrient level, soil organic matter, and hydraulic head were significantly correlated with vegetation
gradients. The mire soils consist of intermixed areas of organic and mineral soils. Mire soils and water are highly alkaline and nutrient-rich. Mean pore water calcium concentration on these mires is 115 mg/L, electrical conductivity averages 575 µS, and mean pH is 7.4. Owing to these conditions, the fen floras include a number of highly minerophilic
species. Based on water chemistry and species criteria, each site was classified as rich to extremely rich fen, with the
two fen types mixing in complex patterns according to local environmental conditions. The species Trichophorum
pumilum, Salix candida, Salix myrtillifolia, Carex microglochin, Carex viridula, Carex scirpoidea, Eriophorum gracile,
Triglochin maritimum, Triglochin palustris, Kobresia myosuroides, Kobresia simpliciuscula, Thalictrum alpinum,
Scorpidium scorpioides, Scorpidium turgescens, and Calliergon trifarium were determined to be indicative of extremely
rich fen conditions in the southern Rocky Mountains.
Key words: Colorado, canonical correspondence analysis, detrended correspondence analysis, extremely rich fen, gradient analysis, mire.
Résumé : Les auteurs ont étudié la végétation, l’environnement et les gradients écologiques présents sur trois tourbières calcaires de la vallée South Park, comté de Park, au Colorado. En utilisant l’analyse bi-directionnelle indicatrice
d’espèces (TWINSPAN), ils ont classifié la végétation en quatre classes d’habitats, neuf sous-classes, et douze associations d’espèces. Ils ont également utilisé l’analyse par correspondances hors tendances pour l’ordination des échantillons de végétation, le long de deux axes représentant les trois gradients écologiques prédominants : hauteur de la nappe
phréatique, bordure de la tourbière à pleine expansion, et la région. L’analyse par correspondances canoniques a été
utilisée pour relier directement les conditions environnementales locales à la végétation. La profondeur de la nappe
phréatique, le développement micro-topographique, le sol, le pH et les nutriments, ainsi que la réserve hydraulique,
montrent des corrélations significatives avec les gradients de végétation. Le sol des tourbières est constitué d’une mosaïque de sols organiques et de sols minéraux. Les sols et l’eau de ces tourbières sont fortement alcalins et riches en
nutriments. Dans ces tourbières, la teneur en calcium de l’eau retenue dans les pores est de 115 mg/L, la conductivité
électrique de 575 µS, et le pH moyen de 7,4. À cause de ces conditions, les flores des tourbières basses incluent plusieurs espèces hautement minérophiles. Sur la base de la chimie de l’eau et le critère des espèces, l’auteur a classifié
chaque site comme tourbière basse riche à tourbière basse extrêmement riche, ces deux types de tourbières basses
s’entremêlant selon des patrons complexes, dépendant des conditions du milieu. Les espèces Trichophorum pumilum,
Salix candida, Salix myrtillifolia, Carex microglochin, Carex viridula, Carex scirpoides, Eriophorum gracile, Trichoglin
maritimum, Trichoglin palustris, Kobresia myosuroides, Kobresia simpliciuscula, Thalictrum alpinum, Scorpidium scorpioides, Scorpidium turgescens et Calliergon trifarium sont retenues comme indicatrices des conditions de tourbières
extrêmement riches, pour le sud des Montagnes Rocheuses.
Mots clés : Colorado, analyse par correspondances canoniques, analyse par correspondances hors tendances, tourbières
extrêmement riches, analyse de gradients, tourbière.
[Traduit par la Rédaction]
Johnson and Steingraeber
219
Received 8 February 2002. Published on the NRC Research Press Web site at http://canjbot.nrc.ca on 21 March 2003.
J.B. Johnson1 and D.A. Steingraeber. Department of Biology, Colorado State University, Fort Collins, CO 80523, U.S.A.
1
Corresponding author (e-mail: [email protected]).
Can. J. Bot. 81: 201–219 (2003)
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:19 AM
doi: 10.1139/B03-017
© 2003 NRC Canada
Color profile: Generic CMYK printer profile
Composite Default screen
202
Introduction
Fens are a common, although relatively minor feature of
the subalpine zone of the Rocky Mountains. In Colorado,
most fens are located in cirques and glacial valleys above
about 2650 m, although fens can occasionally be found at
somewhat lower elevations because of favorable local conditions. Owing to the predominance of crystalline geology,
southern Rocky Mountain fens are most commonly found in
granitic basins and valleys. Such fens are usually described
as transitional or moderately-rich fens based on water quality and floristic criteria (Bierly 1972; Cooper 1986, 1990;
Cooper and Andrus 1994; Johnson 1996).
Recently, fens in South Park, a large, subalpine
inter-mountain valley, were found to be distinct from typical
subalpine fens and contain a large number of regionally rare
calciphilous species such as Salix candida, Trichophorum
pumilum, and Scorpidium scorpioides (Cooper 1996). Many
of the calciphilous species had never, or only historically,
been reported in Colorado (Weber 1990). The calcareous
environment of these fens is due to large limestone and dolomite deposits found in the Mosquito Range to the west. Such
deposits are uncommon in these mountains and calcareous
fens have only been identified at two other locations in the
region: Pine Butte Fen in northwestern Montana (Lesica
1986) and the Swamp Lake Peatlands in northern Wyoming
(Fertig and Jones 1992). Additional calcareous fens may
also be located in the Laramie Mountains in northern Colorado (J. McKee, U.S. Fish and Wildlife Service, personal
communication).
Elsewhere in North America, calcareous fens are likewise
rare, having only been described in small regions within
California (Major and Taylor 1988), southern Minnesota
(Glaser et al. 1990; Almendinger and Leete 1998), Michigan
(Weitzman 1983), southern Wisconsin (Curtis 1959;
Carpenter 1990; Eggers and Reed 1987), northwest Iowa
(Holte and Thorne 1962; Holte 1966; van der Valk 1975),
Alberta (Slack et al. 1980), Ontario (Sjörs 1961a, 1963), and
Alaska (Racine and Walters 1994). Eggers and Reed (1987)
state that calcareous fens are the rarest wetland type in Minnesota and Wisconsin, and probably one of the rarest types
in the United States. While we generally agree with this
statement, it is cautioned that the actual distribution of fens
in general, much less calcareous fens, is not well known in
the United States.
One reason that fens have commonly been overlooked in
wetland surveys is the varied and inconsistent vocabulary
used to classify or describe them. Throughout much of the
United States, including Colorado, fens have erroneously
been referred to as bogs, marshes, or swamps. Semantic inconsistencies are also common within the peatland literature
itself. Following the well-accepted terminology of Canadian
(e.g., Warner and Rubec 1997) and European (e.g., Moore
and Bellamy 1974; Sjors 1961a) peatland scientists, we consider fens to be minerotrophic peatlands. In Rocky Mountain
fens, minerotrophic conditions arise through the presence of
groundwater discharge to, or near, the surface of the fen
(Johnson 2000).
While fen soils are predominantly organic, areas of mineral soil, tufa, or marl usually occur within the wetlands.
Such mineral soil inclusions form an integral part of the fen
Can. J. Bot. Vol. 81, 2003
mosaic, and thus it is artificial and difficult to remove them
from fen investigations. For clarity, when considering
minerotrophic wetlands that contain significant areas of both
organic and mineral soils, we will use the more inclusive
terms “mire” (sensu Sjörs 1961a; Pakarinen 1995) or “fen
complex” to highlight the heterogeneous, yet interconnected,
nature of such systems.
The unusual environment provided by calcareous mires in
the South Park valley provides habitat for some 15 state-rare
or endemic plant species as well as least 10 rare invertebrate
species (Sanderson and March 1996). Because of their biological significance, in most areas, calcareous fens have
been given special conservation consideration, through either
regulatory means (e.g., U.S. Fish and Wildlife Service 1999;
Almendinger and Leete 1998; Pearson and Leoschke 1992)
or purchase by conservation groups such as The Nature Conservancy. In spite of their biological significance, calcareous
fens are commonly threatened, or have been historically impacted, by activities such as peat mining, water diversion,
and groundwater pumping.
While the location and gross floristic composition of
South Park’s calcareous fens have been reported (e.g.,
Sanderson and March 1996; Cooper 1996; Johnson and
Gerhardt 2002), little detailed investigation into their ecological and environmental characteristics has been undertaken.
This study was designed to build upon these earlier works
and broaden our knowledge of South Park’s fens by describing their vegetational and environmental characteristics and
how these two attributes relate to one another. The specific
goals of this study were to (i) describe vegetation assemblages and gradients found within representative fens
throughout the South Park valley, (ii) describe the environmental conditions found at these fens, and (iii) directly relate
fen vegetation to local environmental conditions to assess
the factors influencing species composition.
Materials and methods
Geologic and hydrogeomorphic setting
The South Park valley, hereafter known simply as South
Park, is one of the four major inter-mountain basins in Colorado formed during the Larimide Orogeny (Lozano 1967).
The valley is surrounded on all sides by three mountain
ranges, with peaks reaching elevations over 4200 m above
sea level. Sitting at between 2590 and 3050 m of elevation,
South Park has modest physiographic relief compared with
the surrounding mountains and consists of plains, rolling
hills, and broad alluvial valleys, which are punctuated by
a few north-northwest trending bedrock ridges. Roughly
speaking, South Park is 80 km long and 56 km wide at its
widest points and encompasses 2330 km2 (Stark et al. 1949).
The valley is formed by a complexly faulted, asymmetrical, north-northwest trending syncline filled with Tertiary
sediments and Quaternary glacial outwash (Lozano 1967;
De Voto 1971). Elsewhere in Colorado, inter-mountain parks
are underlain by outwash primarily composed of granitic or
volcanic material. In contrast, the till blanketing South Park
has a high proportion of calcareous and dolomitic material.
South Park mires receive groundwater from both shallow
outwash and alluvial aquifers as well as deep bedrock aquifers (Gard et al. 2000). The relative contribution of these
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:19 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
water sources is not well known and probably varies among
sites. The calcareous and dolomitic composition of the
glacial outwash and deep bedrock formations such as the
Maroon, Belden, Coffman, and Leadville Limestone causes
groundwater flowing through them to become alkaline, often
strongly so, and mineral-rich. Appel (1995), for instance,
reported groundwater pHs greater than 11 during his geochemical study of High Creek Fen.
Climate
South Park is a cool and semi-arid region located in the
rain shadow of the Mosquito Range. Mean annual temperature measured near the southern end of the park is 1.9°C
(Antero Reservoir Station), although temperatures are generally cooler in the north and at the mouth of valleys subject to
cold air drainage (Spahr 1981). The region also exhibits a
north–south gradient in precipitation, with northern areas receiving greater amounts of annual precipitation. At Antero
Reservoir, in the south, average yearly precipitation is
25.8 cm, compared with 40.2 cm at Fairplay near the middle
of the valley. In spite of these annual differences, Spahr
(1981) reported that growing season precipitation is relatively uniform across the valley, since central and northern
areas receive the majority of their precipitation as early
spring snows, while Antero Reservoir receives the greater
part of its precipitation in August during convective thunder
showers.
Growing season evapotranspiration is high in South Park
owing to its warm daytime temperatures, high winds, and
large insolation load. All data available for South Park show
that a moisture deficit exists throughout the growing season,
and in some months, up to a 19-cm moisture deficit has been
measured (Spahr 1981; Walter et al. 1990).
Vegetation
The floor of South Park is dominated by short-grass
steppe reminiscent of Colorado’s eastern plains. This vegetation cover type is uncharacteristic of the mainly conifer-dominated subalpine zone (Marr 1961). Hills and ridges
in the park are typically covered by mixed stands of aspen
(Populus tremuloides), lodgepole pine (Pinus contorta), subalpine fir (Abies lasiocarpa), and Engelmann spruce (Picea
engelmannii). Stands of bristle-cone pines (Pinus aristata)
occur occasionally on exposed rocky ridges.
Study sites
Three fens were chosen for this study: High Creek Fen,
Crooked Creek Fen, and Fremont’s Fen (Fig. 1). The sites
were selected since they span the majority of South Park
from north to south and encompass a range of sizes, landscape positions, and disturbance levels. Each of the predominantly organic soil fens studied here are embedded within
more expansive mire systems. Although the focus of this
study was on the organic soil portion of each wetland, surrounding and imbedded mineral soil areas were also examined to characterize the ecotone between the two wetland
types. Our concentration on organic soil wetland areas is not
intended to detract from ecological importance of the surrounding mire system but rather to provide a detailed investigation of the distinctive fen environment.
203
High Creek Fen is located at an elevation of 2830 m. The
main fen with predominantly organic soils covers about
150 ha, while the entire mire encompasses more than
300 ha. High Creek Fen possesses a shallow topographical
gradient, sloping 1% to the southeast. High Creek Fen is
part of a The Nature Conservancy preserve purchased in
1991. Prior to The Nature Conservancy purchase, the fen
had been grazed since the 1860’s (Appel 1995), and during
the 1970’s and 1980’s, portions of the fen were mined for
peat.
Crooked Creek Fen is located in the Pike National Forest
on the periphery of South Park at an elevation of 3080 m.
The fen covers 24 ha on a fan-shaped slope at the foot of an
abandoned beaver pond complex. The fen slopes relatively
steeply to the south at an average grade of 4 %, but the slope
is non-uniform, being punctuated by a series of steep steps
and flat aprons. Most of Crooked Creek Fen has not been
altered by human activities, except for a ditch that transverses the fen near its foot.
Fremont’s Fen lies at the foot of the Mosquito Range at
2930 m of elevation and forms an important part of a
1625-ha mire complex that spans the valley from west to
east. Fremont’s Fen covers 97 ha. The fen begins at a colluvial fan and slopes shallowly to the east at a grade of 1%. A
series of drainage ditches crisscross much of the fen and an
abandoned peat mine lies in the area where the deepest peat
deposits presumably were.
Environmental sampling
Study sites were equipped with a matrix of sampling stations, each consisting of a shallow groundwater well and one
or more piezometers. High Creek Fen had 28 stations,
Crooked Creek 23, and Fremont’s Fen 16. Stations were
subjectively located in the major geomorphic, hydrologic,
and vegetative zones occurring on the fens. A more objective
sample placement was not practicable owing to the large
number of samples that would be required to encompass the
diversity of wetland habitats present and the consequent
over-sampling of common situations.
Groundwater wells were constructed of 2.54 cm inside diameter polyvinylchloride pipe, approximately 1.5 m long.
Wells bottoms were capped and the bottom 30.5 cm of the
wells was perforated. Wells were driven 1.3 m into the peat
or until they reached an impermeable mineral layer. Once installed, wells were allowed to recharge and then were
drained several times to ensure proper functioning.
Piezometers were made of unperforated 2.54 cm inside diameter polyvinylchloride pipe. Piezometers were installed to
a uniform depth of 75 cm at High Creek Fen and variable
depths at the other fens depending on peat thickness.
Well and piezometer water depths were measured every
10 to 14 days from the beginning of June through September
or early October, which roughly corresponds to the local
growing season length. Measurements were taken from 1995
to 1998 on High Creek fen and from 1996 to 1998 on
Crooked Creek and Fremont’s Fens.
Soil and water characterization
Peat depth was determined using a 5-cm diameter piston
corer. Soil characteristics including peat decomposition and
stratigraphy were evaluated from the extracted cores. To
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:19 AM
Color profile: Generic CMYK printer profile
Composite Default screen
204
Can. J. Bot. Vol. 81, 2003
Fig. 1. Map showing the location of Park County (shaded) and South Park (outlined) within Colorado, U.S.A. The portion of Park
County that includes South Park has been enlarged to the right. Study site locations and the city of Denver are starred and labeled.
Selected landmarks are also labelled.
assess physical and chemical soil properties, an additional
set of samples was obtained from the upper 20 cm of the soil
at each station using a hand trowel. Soil samples were obtained in 1995, 1996, and 1997 at High Creek Fen and in
1996 and 1997 at the other two sites.
Pore water samples from each sampling station were collected in September 1997. An additional set of samples were
obtained at High Creek Fen in August 1995. All samples
were filtered through 0.45-µm nitrocellulose filters using
a vacuum aspirator, placed in polyethylene bottles, and
preserved with nitric acid. All water and soil samples were
analyzed at the Colorado State University Water and Soil
Testing Lab using inductively coupled ion spectrophotometry.
Vegetation sampling
Vegetation composition was determined at High Creek
Fen in July 1995 and at the other two sites during July and
August 1997. At each sampling station, a 5 m diameter plot
was constructed around the groundwater well. Within plots,
the cover of each plant species was visually estimated. Species names are according to Weber (1990), but nomenclature
not consistent with Kartesz (1994) is noted parenthetically in
Table 1. Vegetation plots were positioned at each environmental sampling station so that vegetation composition
could be directly related to measured environmental conditions. The percent cover of hummocks within each plot was
visually estimated, and the heights of 12 subjectively chosen
hummocks, if present, were measured.
Data analysis
Three related multivariate statistical techniques were used
to analyze study data: two-way indicator species analysis
(TWINSPAN), detrended correspondence analysis (DCA),
and canonical correspondence analysis (CCA). Each approach provides a somewhat different view of data structure
and when employed together the techniques can be used to
complement, supplement, and evaluate the other analyses
(Gauch 1982; Økland 1996).
Before analysis, species data were log transformed to
reduce the influence of very abundant species (Jongman et
al. 1995). Mosses were made passive in all analyses since
presence–absence data were collected on these species, and
few bryophyte data were collected at High Creek Fen.
TWINSPAN (Hill 1979), as contained in PC-ORD (McCune
and Mefford 1997), was used to classify samples based on
species composition. Default settings were used during the
TWINSPAN, except that cut levels were 0.0, 0.3, 0.7, 1.0,
1.3, i.e., the log of the default cut values.
Gradients in vegetation and the environment were reconstructed using the DCA and CCA algorithms of Canoco 4
(Ter Braak and Smilauer 1998). DCA is an indirect gradient
analysis, or ordination technique, while CCA is a method of
multivariate direct gradient analysis. In CCA, initial sample
and species placement is based on the correspondence analysis algorithm, but sample and species scores are secondarily
constrained to be linear combinations of environmental variables through multiple regression (Ter Braak 1986).
In the DCA, detrending by segments was in force using
the default 26 segments. For the CCA, the “inter-sample
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:20 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
205
Table 1. Species composition of vegetation associations.
Class
Meadow and dry mire
Subclass
Tall-hummock
fen
Achillea millefolium
Agrostis scabra
Agrostis stolonifera
Antennaria sp.
Alopecurus borealis
Argentina anserina
Artemisia frigida
Astragalus agrestis
Aster occidentalis
Betula glandulosa
Bistorta bistortoides
(Polygonum bistortoides)
Bistorta viviparum
(Polygonum viviparum)
Calamagrostis canadensis
Calamagrostis stricta
Campanula parryi
Carex aquatilis
Carex aurea
Carex capillaris
Carex limosa
Carex livida
Carex microptera
Carex microglochin
Carex parryana subsp. hallii
(C. hallii)
Carex scirpoidea
Carex simulata
Carex utriculata
(C. rostrata)
Chondrophylla aquatica
(Gentiana fremontii)
Cirsium arvense
Cirsium coloradense
Clementsia rhodantha
(Sedum rhodanthum)
Conioselinum scopulorum
Crepis runcinata
Critesion brachyantherum
(Hordeum brachyantherum)
Critesion jubatum
(Hordeum jubatum)
Deschampsia cespitosa
Distegia involucrata
(Lonicera involucrata)
Dodecatheon pulchellum
Eleocharis quinqueflora
Elymus trachycaulus
Epilobium lactiflorum
Epilobium spp.
Equisetum arvense
Erigeron lonchophyllus
Erigeron speciosus
Eriophorum angustifolium
Meadow
1
0
1
1
1
1
1
Willow carr
Dry
mire
Tall
willow carr
Short willow
carr
1
1
+
6
1
Hummocky
fen lawn
1
2
1
+
1
1
4
2
1
1
+
+
1
1
+
+
+
2
+
3
+
3
+
+
+
+
+
+
3
2
2
0
3
0
0
0
+
+
+
0
2
2
+
1
2
+
2
3
1
2
0
1
1
2
2
3
+
3
+
+
2
+
+
+
+
1
+
1
1
+
1
+
2
3
1
+
2
10
1
1
+
2
1
7
1
+
+
+
1
Quagmire
+
+
2
1
Water
track
+
0
1
Fen
lawn
+
+
1
+
2
1
Water track and
quagmire
Fen expanse
+
1
1
2
+
+
2
+
+
1
3
+
+
3
+
1
1
+
+
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:21 AM
Color profile: Generic CMYK printer profile
Composite Default screen
206
Can. J. Bot. Vol. 81, 2003
Table 1 (continued).
Class
Subclass
Eriophorum gracile
Galium boreale
Gentian spp.
Hierochloë hirta
Heterotheca pumila
Juncus albescens
Juncus alpino-articulatus
Juncus arcticus ssp. ater
(J. balticus)
Juncus longistylis
Kobresia myosuroides
Kobresia simpliciuscula
Limnorchis hyperborea
(Habenaria hyperborea)
Maianthemum amplexicaule
(Smilacina amplexicaulis)
Moss sp.
Muhlenbergia richardsonis
Parnassia parviflora
Pascopyrum smithii
Pedicularis crenulata
Pedicularis groenlandica
Pentaphylloides floribunda
(Potentilla fruticosa)
Plantago eriopoda
Poa compressa
Poa glaucifolia
Poa pratensis
Polemonium foliosissimum
Potamogeton pectinatus
Potentilla plattensis
Potentilla subjuga
Primula egaliksensis
Primula incana
Psychrophila leptosepala
(Caltha leptosepala)
Ptilagrostis porteri
Ranunculus cymbalaria
Rumex spp.
Salix brachycarpa
Salix candida
Salix monticola
Salix myrtillifolia
Salix planifolia
Stellaria longipes
Swertia perennis
Sisyrinchium pallidum
Taraxacum officinale
Thalictrum alpinum
Trichophorum pumilum
(Scirpus pumilus)
Triglochin maritimum
Triglochin palustris
Utricularia ochroleuca
Meadow and dry mire
Tall-hummock
fen
Meadow
1
2
1
Willow carr
Dry
mire
Tall
willow carr
Water track and
quagmire
Fen expanse
Short willow
carr
1
Hummocky
fen lawn
Fen
lawn
Water
track
+
1
1
+
+
+
+
+
1
+
1
3
+
1
2
+
3
2
2
1
0
+
+
+
2
+
+
1
1
+
1
Quagmire
+
1
+
1
1
2
1
2
2
3
2
1
+
+
1
+
10
2
1
1
+
+
1
3
1
1
0
1
2
1
6
0
0
1
1
+
1
1
+
+
1
+
+
+
1
+
+
0
+
+
+
+
+
+
+
+
+
+
0
+
+
2
+
1
1
+
1
2
1
2
3
2
+
2
+
2
1
+
+
1
1
+
+
+
1
1
+
+
3
+
1
1
1
+
1
1
1
1
1
1
1
1
1
2
1
1
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:21 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
207
Table 1 (concluded).
Class
Subclass
Valeriana edulis
Viola adunca
Zygadenus elegans
Bare ground
Hummocks
Bryophytes
Brachythecium nelsonii
Calliergon giganteum
Calliergon stramineum
Calliergon trifarium
Campylium stellatum
Distichum capillare
Drepanocladus aduncus
Drepanocladus revolvens
Philonotis fontana
Mniobryum albicans
Rhizomnium pseudopunctatum
Scorpidium turgescens
Scorpidium scorpiodes
Warnstorfia exannulata
Meadow and dry mire
Tall-hummock
fen
Willow carr
Meadow
1
Dry
mire
+
1
4
1
3
Tall
willow carr
Water track and
quagmire
Fen expanse
Short willow
carr
Hummocky
fen lawn
Fen
lawn
Water
track
Quagmire
4
2
+
3
+
1
2
+
2
1
2
1
P
P
P
P
P
P
P
P
P
P
P
2
1
3
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
Note: Values are percent species coverage converted into Braun-Blanquet cover classes. +, <1%; 1, 1– 5%; 2, 6–25%; 3, 26–50%; 4, 50–75%; 5,
76–100%; P, species presence in cases where abundance data are not available.
scores” option was chosen to optimally configure plot placement in the diagram (Ter Braak and Smilauer 1998), and
bi-plot scaling was used. In plotting samples, scores that are
linear combinations of environmental factors were used (LC
scores sensu Palmer 1993). Before employing CCA, a
step-wise selection of environmental variables was performed using the Monte Carlo permutation test available in
Canoco (Ter Braak and Smilauer 1998).
To gauge the axis distortion inherent in CCA, Spearman’s
rank correlation was used to compare DCA and CCA scores
(Allen and Peet 1990; Prentice and Cramer 1990; Johnson
1996; Økland 1996). A standard product–moment test of
correlation was used to determine the significance of the
axis correlations (Sokal and Rohlf 1981).
Results
Fen habitats and species associations
TWINSPAN was used to hierarchically classify the vegetation habitats sampled in this study. At the higher levels of
the classification, gross physical characteristics were used
to describe the vegetation units. For the lowest, species-association level, groups were named using a combination of the names of the dominant or diagnostic species
within that vegetation type.
Divisions 3 and 4 of the TWINSPAN (Fig. 2) were used
to create nine physiognomic or habitat subclasses. These
nine subclasses were subsequently divided into 12 species
associations. Table 1 contains the mean species abundances
measured in each subclass. In this classification, subclasses
are frequently represented by only one species association.
That is, only one study plot was located within a subclass’s
habitat. Although this may at first seem to blur the distinction between the subclass and association levels, it engenders useful flexibility into the classification, allowing mire
species associations defined in other studies (e.g., Sanderson
and March 1996; Cooper 1996) to be placed within its
framework. Lastly, subclasses and associations were assembled into four mire habitat classes. Generally, class designation follows the grouping created by division two of the
TWINSPAN (Fig. 2). The exception to this is fen lawn
class, which combines vegetationally distinct subclasses
based on their overall physiognomy, hydrologic characteristics, and interspersion within the mires.
TWINSPAN arranged vegetation assemblages essentially
according to a water-table gradient, with habitats becoming
successively wetter towards the right of the diagram. The
vegetation descriptions contained below focus on the subclass level of the classification.
Meadow and dry mire vegetation
The three subclasses within the meadow and dry mire
class are found in seasonally wet areas dominated by grasses
and sedges. The tall-hummock fen subclass is characterized
by the presence of Deschampsia cespitosa, Elymus
trachycaulus, and Juncus arcticus in addition to characteristic species such as Carex aquatilis, Carex simulata, Carex
scirpoidea, Kobresia myosuroides and Primula incana. The
high production and cespitose growth forms of Carex
scirpoidea and Deschampsia cespitosa help produce the
large hummocks found in these areas. The meadow subclass
contains the least hydric communities examined. Meadow
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:22 AM
Color profile: Generic CMYK printer profile
Composite Default screen
208
Can. J. Bot. Vol. 81, 2003
Fig. 2. Classification diagram based on TWINSPAN of vegetation samples. The diagram follows the TWINSPAN results, except that
some lower-level divisions have been removed for simplicity. Divisions are based on vegetative data but are described in the diagram
boxes using a simple habitat description. Vegetation classes are separated by bold lines and names are given in large type at the diagram bottom. The nine habitat subclass names are boxed and italicized. Species association names are hyphenated and included near
the bottom of the figure. Species abbreviations use the first three letters of the genus name and the first three letters of the specific
epithet.
sites are always located on mineral soil, although a histic
epipedon may be present. Shrubs are uncommon. Juncus
arcticus, Poa pratensis, and Deschampsia cespitosa dominate this vegetation type.
Like the meadow subclass, the dry mire subclass also has
a high coverage of Deschampsia cespitosa and J. arcticus,
but its flora additionally includes hydrophilic sedges, particularly Carex aquatilis. These sites generally form an ecotone
between the more hydric, peat-dominated areas and the mineral soil meadows, possessing either soil condition. At the
point where the TWINSPAN analysis was truncated, only
one association was defined (Fig. 2), but this subclass also
includes the rare Thalictrum alpinum – K. myosuroides association (Cooper and Sanderson 1997; J.B. Johnson, data not
included).
Willow carr vegetation
The willow carr vegetation class includes areas with open
to densely closed canopies of Salix planifolia and Salix
monticola and a lush carpet of mosses. The prevalence of
shrubs physiognomically separates these sites from the other
mire vegetation types sampled. Tall willow carr subclass
vegetation is located along rivulets within the mires and in
areas with quickly flowing surface water, particularly along
the mire margin. The canopy consists of willows up to 3 m
in height and may be somewhat open to closed. The understory is dominated by Carex utriculata and mosses. These
sites have the highest moss species richness and cover of
any of the communities surveyed (Table 1).
Short willow carr subclass vegetation is typically found
along the margins of peat aprons within the fen expanse and
bordering upland rises. These sites have open canopies
of low willows, primarily Salix planifolia and Salix
brachycarpa, and Pentaphylloides floribunda. Carex
aquatilis and Carex utriculata dominate the herbaceous
layer, and the moss coverage and richness is nearly as high
as in the tall willow carr subclass.
Fen expanse vegetation
Fen expanses are open areas within the interiors of fens
(Sjörs 1950b). The fen expanse habitat class contains two
subclasses: hummocky fen lawn and fen lawn. Hummocky
fen lawn vegetation is dominated by sedges such as Carex
aquatilis and Carex simulata. Shrubs are common, but they
are low, scattered, and found almost exclusively on hummocks. Mosses such as Scorpidium scorpiodes and
Scorpidium turgescens usually carpet the hollow bottoms between hummocks, while Campylium stellatum, Warnstorfia
exannulata, and Drepanocladus aduncus are found on the
hummocks themselves. Hummocky fen lawns are perenni© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:23 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
ally wet and are frequently found on land forms that we call
“peat aprons”. Peat aprons are roughly circular to
ameboid-shaped features with slightly domed, quaking surfaces caused by upwelling groundwater.
The hummocky fen lawn subclass includes two associations: Trichophorum pumilum – Kobresia simpliciuscula and
Carex simulata – K. simpliciuscula. Physically, the two associations differ mainly in microtopography development, with
the latter association generally having less well developed
hummocks. The Trichophorum pumilum – K. simpliciuscula
association is considered quite rare in the Rocky Mountains
(Cooper 1996; Sanderson and March 1996).
Habitats included within the fen lawn subclass are found
in grossly similar physical situations as hummocky fen
lawns. The major physiognomic differences between the two
subclasses are that fen lawn communities, lacking the habitat
differentiation caused by hummock development, tend to be
more monotonous and densely vegetated than hummocky
fen lawns. Fen lawns are generally strongly dominated by
Carex simulata and Carex aquatilis, with patches of
Kobresia simpliciuscula and Eleocharis quinqueflora being
common. Species such as Pedicularis groenlandica, Primula
egalikensis, Thalictrum alpinum, Triglochin maritimum, and
Triglochin palustris are common and diagnostic but do not
reach high cover values.
Water tracks and quagmires
The water track and quagmire class contains habitats that
are continually covered by shallow, slowly flowing water,
except where low rises occur. Water tracks tend to be somewhat linear features in which water is conveyed relatively
quickly, and they are often slightly steeper than the surrounding fen. The soils are strongly quaking, but the floating
vegetation mat is relatively robust. Quagmires, on the other
hand, are very low gradient, amorphous expanses of quaking
soils, covered by shallow, almost standing (topogenous), water. Quagmires have lower plant coverage compared with
water tracks (Table 1) and possess large patches of open water. Virtually all surfaces may be covered with marl and the
bottoms are layered with algae and iron-flocculating bacteria
that form soil strata. The soils of quagmire sites are tenuously thin, hardly holding a human’s weight. Water tracks
frequently form links between quagmire areas and together
these features constitute the major drainage pathways within
fens.
The dominant species in both subclasses are Eleocharis
quinqueflora, Triglochin spp., and Utricularia spp., with
mosses such as Scorpidium scorpioides forming expansive
carpets in many areas. In more hummocky areas, somewhat
less hydrophilic species such as Thalictrum alpinum,
Trichophorum pumilum, Salix candida, and K. simpliciuscula
grow in the drier micro-habitats. The quagmires and water
tracks described in this study seem directly comparable to the
mire “mud-bottoms” and “carpets”, respectively, described by
Sjörs (1950b) and Nordqvist (1950) in Swedish mires.
Ordination
Figure 3 shows axes 1 and 2 of a DCA of vegetation samples. The graph symbols correspond to the nine subclasses
described in the previous section. As is evident in Fig. 3,
209
samples grouped by TWINSPAN have a high fidelity for
one another in the DCA diagrams. Table 2 contains ordination diagnostics for the DCA. The first three DCA axes accounted for 21.3% of the total variance in the species data.
Gradient lengths are high for these axes, and complete species turnover occurs across the first axis, while species turnover is nearly complete along the second axis.
Samples from each study site show a tendency to cluster
near one another, with the array of study sites forming a gradient across axis 1. Such arrangement suggests the presence
of site-specific influences on vegetation composition. Also
exhibited in the axis 1 gradient is a trend going from the
most hydric, quagmire plots on the left, to drier, meadow
plots on the right. Axis 2 is dominated for most of its length
by mire margin to expanse gradients. This gradient repeats
itself in the diagram, with the Crooked Creek Fen marginal
samples placed near the top of axis 2 and the High Creek
and Fremont’s Fen marginal samples placed near the bottom.
The two gradients converge to mire expanse vegetation near
the middle of axis 2.
As observed on-site and represented in Fig. 3, the margin
to expanse gradient on Crooked Creek Fen begins in tall willow carr vegetation on the mire margin. The tall willow carr
grades into short, open-canopy willow carr and, finally,
into fen expanse vegetation. In contrast, High Creek and
Fremont’s Fen margin sites have little or no shrub cover.
Both types of marginal sites tend to have well-developed
microtopography and, at their extremes, may or may not be
situated on organic soils.
Fen environment
Hydrology
Environmental data collected at plots were grouped according to TWINSPAN subclasses. Table 3 contains a summary of mire hydrologic measurements. Negative water
table heights indicate the depth of groundwater below the
ground’s surface. Mean water table depths ranged from
–48.2 cm in meadow areas to 1.4 cm in the quagmires.
Meadow and dry mire areas exhibited the greatest spatial
and temporal hydrologic variation, with season water table
deviations of 25.0 and 28.2 cm, respectively. Quagmires had
the least variable water table level, with an average seasonal
deviation of only 2.7 cm.
Water table levels measured in the other mire vegetation
types ranged between these extremes. Fen lawn sites had
water tables approximately at the ground surface and
showed little fluctuation in water table throughout the year.
In hummocky areas (the tall-hummock fen and hummocky
fen lawns), water tables were below the surface on average;
however, they frequently had standing water for periods during the season.
Mean hydraulic head also varied between vegetation
types, generally following the trends seen in water table
height, with the head values increasing with water table
height. The drier vegetation types had negative head values
on average, indicating that these areas are sites of net
groundwater recharge. Tall willow carrs and fen lawns had
the highest head values, suggesting that these areas are sites
of net groundwater discharge.
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:24 AM
Color profile: Generic CMYK printer profile
Composite Default screen
210
Can. J. Bot. Vol. 81, 2003
Fig. 3. Axes 1 and 2 of a DCA. Symbols indicate the TWINSPAN subclasses into which samples were grouped. Alphanumeric code
shows the sample location and number: h, High Creek Fen samples; c, Crooked Creek Fen; f, Fremont’s Fen. See text for additional
details.
Table 2. Ordination diagnostics for the DCA and CCA. See text for explanation.
DCA gradient length
DCA eigenvalues (unconstrained)
CCA eigenvalues (constrained)
Cumulative percentage of species
variance accounted for by CCA
Axis 1
Axis 2
Axis 3
Axis 4
Total inertia
5.29
0.62
0.52
9.7
3.37
0.34
0.42
17.5
2.33
0.19
0.20
21.3
2.48
0.16
0.18
24.7
5.34
2.04
Water and soil characteristics
There was little difference in the mean groundwater pH
measured within the different vegetation types (Table 4).
Soil pH followed the same general pattern as water pH,
except that soils were generally slightly more acidic than
groundwater. Meadow groundwater had the highest electrical
conductivity (EC), with the tall-hummock fen following.
High groundwater EC in these marginal sites is probably due
to cycles of soil wetting and drying and the resultant mineral
accumulation that occurs under such conditions. The other,
more hydric, habitats all possess similar and comparatively
low ECs, probably owing to the dilution and flushing caused
by abundant, flowing groundwater.
Calcium, the dominant cation in all of the samples obtained, ranged from 219.0 mg/L in the tall-hummock fen
sites down to 81.7 mg/L in the tall willow carr (Table 4).
Average calcium content across all sites was 131.7 mg/L.
The concentration of other cations generally follows the pattern displayed by calcium. There was also close correspondence between groundwater EC and the groundwater cation
concentration. The pattern of soil calcium was nearly the inverse of that of water calcium, with the tall-hummock fen
having the lowest soil calcium and the hummocky quagmires
and water-tracks the highest. As with water, calcium was the
dominant cation measured in the soils.
The deepest peat is generally located in the willow carrs
and hummocky fen lawns, whereas meadows possess
mineral soils with well-developed O horizons or histic
epipedons (Table 4). The remaining vegetation subclasses
are located in areas with organic soils, typically about 50 cm
deep. The highest soil organic matter content (66.4%) was
measured in the most hydric habitats, although given the
high sample standard deviations, all of these habitats seem
to possess approximately equivalent amounts of soil organic
matter (Table 4; Johnson 2000). Even in the most organic-rich fen soils examined here, mineral material makes
up a substantial portion of the soil composition and may be
the dominant soil component.
Surficial microtopography on South Park mires is generally in the form of hummocks and hollows. Strings and
flarks (i.e., ridges and troughs) are not present on the studied
fens, although they occasionally develop at other South Park
fen locations (J.B. Johnson, personal observation). Hummock–
hollow topography is best developed in the tall-hummock
fen areas, wherein hummocks averaged 41 cm in height and
covered 24% of the ground surface (Table 4). Micro© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:25 AM
topography is developed to varying degrees in the other vegetation types, except that it is absent in the driest meadow
areas owing to the soil hardness and low level of soil organic matter which comprises the hummocks.
1.2±0.9
2.9
4.1
–1.5
0.2
Note: Water table (WT) variation within subclasses is included as the sample standard deviation. Refer to Johnson (2000) for additional data and statistics.
Fen lawn
0.0±2.5
3.9
3.8
–4.8
4.4
–10.7±10.1
12.8
2.5
–23.3
0.9
–14.8±10.2
17.0
–1.1
–32.4
5.9
–5.1±9.3
7.2
3.1
–10.4
–1.8
–37.6±22.1
28.2
–5.2
–63.0
–3.9
–48.2±26.0
25
–10.2
–98.6
–7.1
–30.5±6.0
19.2
–6.8
–54.8
–12.4
WT height (cm)
seasonal WT deviation (cm)
seasonal WT Minimum depth (cm)
seasonal WT Maximum depth (cm)
seasonal Head (cm)
Mean
Mean
Mean
Mean
Mean
Hummocky
fen lawn
Tall willow
carr
Short
willow carr
Dry mire
Meadow
Tall-hummock
fen
Parameter
Table 3. Summary of hydrological measurements performed from 1995 to 1998.
1.4±0.8
2.7
4.1
–0.6
0.0
211
Water track
Johnson and Steingraeber
Quagmire
Color profile: Generic CMYK printer profile
Composite Default screen
Direct gradient analysis
CCA was used to directly relate the vegetation and environmental data sets (Fig. 4). Graph symbols correspond to
habitat subclasses as in the DCAs above, and the vectors indicate the maximum direction of change in the environmental variables. Vector orientation indicates the direction in
which an environmental factor is increasing in value, while
its relative length symbolizes the strength of the factor’s correlation with vegetation differences. The interaction of hummock height and hummock cover was used as a synthetic
variable to model the relationship between the attributes
(Johnson 1996; Ter Braak and Smilauer 1998).
Table 2 contains the CCA diagnostics for the first four ordination axes (only two are shown in Fig. 4). Unconstrained
eigenvalues are those calculated only using species data and
represent the total variance in the species data set. Typically,
constrained eigenvalues are lower than unconstrained ones
since species data are adjusted to be linear combinations
of the environmental variables. Dividing the sum of the
constrained eigenvalues by the sum of the unconstrained
eigenvalues provides a measure of how much of the change
in species composition has been accounted for by the measured environmental factors and how much is due to unmeasured factors.
Eigenvalue reductions between the DCA and CCA were not
large, indicating that relatively little vegetational information
was lost by the CCA (Table 2). The measured environmental
factors account for 38% of the species variance. Axis 1 is most
highly correlated with differences in water table depth, soil sodium, and pore water pH (Table 4). Axis 2 is most highly correlated with soil Mn, Mg, and EC, and microtopography. The
canonical coefficients show that the factors most highly correlated with the axes also received high weights in the analysis
(Table 5). A Monte Carlo permutation test determined that both
CCA axes are significant (p = 0.005).
The CCA shows that the marginal habitats, including
tall-hummock fen, meadow, and dry mire, are associated
with nutrient-rich soils and waters that are circumneutral to
slightly acidic. Vegetation in these areas is also associated
with relatively deep water tables and well-developed microtopography. The willow carr and hummocky fen lawn sites,
particularly those found at Crooked Creek Fen, tend to posses the most alkaline conditions and soils tend to be rich in
manganese but low in other cations. Such sites are also associated with high water tables and well-developed microtopography. Fen lawns and quagmires are situated in areas
with poorly developed microtopography, high soil organic
matter, pH, and water table, and moderately high soil cation
concentrations (Fig. 4, Table 5).
Discussion
A comparison of South Park’s mires with related
systems
Water chemistry
The majority of southern Rocky Mountain fens are found
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:26 AM
Color profile: Generic CMYK printer profile
Composite Default screen
212
Can. J. Bot. Vol. 81, 2003
Table 4. Summary of soil and water data collected within each of the vegetation types.
Parameter
Groundwater
EC (µS)
pH
Ca
Mg
Na
K
P
Fe
Mn
Soils
pH
EC (µS)
NO3
P
K
Zn
Fe
Mn
Cu
Ca
Mg
Na
Organic content (%)
Hummock height (cm)
Hummock cover (%)
Coverage(%)
Peat depth (cm)
Tall-hummock
fen
Meadow
Dry
mire
Short
willow carr
Tall
willow carr
Hummocky
fen lawn
Fen
lawn
Water
track
Quagmire
847.0
7.4
219.0
65.6
19.5
1.2
0.15
0.2
0.04
864.0
6.4
144.0
30.9
15.9
0.8
0.16
0.8
0.02
433.0
7.0
99.6
22.9
18.2
0.5
0.09
1.8
0.05
544.0
7.8
110.4
20.0
10.7
2.2
0.18
1.0
1.93
494.0
7.7
81.7
10.7
3.8
0.4
0.05
0.1
0.04
556.0
7.7
125.2
18.9
13.1
0.8
0.09
1.7
0.78
519.0
7.5
103.1
30.1
6.9
0.8
0.09
0.9
0.03
527.0
7.5
119.2
35.9
8.4
0.4
0.08
0.1
0.05
605.0
7.4
140.3
48.7
9.4
0.5
0.09
0.2
0.05
7.6
1.5
19.0
5.1
135.6
8.7
642.4
22.5
5.4
8581
1292.0
114.6
36.5
41.0
24.0
7.1
2.5
40.3
11.5
140.1
13.3
784.1
21.6
4.9
11585
1659.3
412.2
50.1
27.9
11.5
6.6
1.1
18.2
6.8
132.5
23.4
1049.4
33.6
5.8
10199
786.8
199.3
61.8
42.5
19.2
7.7
0.7
17.0
5.5
86.0
4.5
639.3
98.8
4.4
9443
378.4
79.9
36.0
20.7
23.4
7.5
0.6
14.5
17.4
171.3
9.0
574.8
77.4
6.2
11521
789.1
112.9
49.5
37.0
28.0
7.4
0.8
13.3
7.8
163.4
9.1
511.0
73.7
6.6
8897
685.2
134.1
46.5
28.9
18.9
7.2
1.8
11.7
8.5
129.5
7.9
395.6
18.5
4.9
9937
1308.9
126.5
59.7
15.5
13.7
7.3
2.0
14.5
14.5
126.7
9.6
378.2
15.9
4.0
12851
1754.2
155.7
66.4
4.8
12.2
7.4
1.7
13.5
9.3
105.6
7.1
197.8
16.7
5.3
10984
1703.2
146.9
54.5
0.7
6.5
46.7
17.1
25.8
81.7
137.4
128.4
50.1
48.0
52.7
Note: All units are parts per million, unless otherwise noted, except pH, which is in standard units. Sample standard deviations have been omitted for
readability, but these may be found in Johnson (2000).
Table 5. Intra-set correlations and canonical coefficients for the CCA.
Intra-set correlations
Canonical coefficients
Parameter
Axis 1
Axis 2
Axis 3
Axis 4
Axis 1
Axis 2
Axis 3
Axis 4
Soil pH
Soil EC
Soil organic matter
Log soil Mn
Log soil Mg
Log soil Na
Water Ca
Water EC
Water pH
Water table depth
Hydraulic head
Hum cover × hum height
–0.42
0.51
–0.08
–0.24
0.22
0.66
0.28
0.42
–0.65
–0.89
–0.50
0.22
–0.31
0.71
0.27
–0.76
0.72
0.40
0.14
0.13
–0.24
–0.02
–0.10
–0.50
–0.40
–0.02
0.10
–0.02
–0.42
0.21
–0.35
–0.46
0.29
0.00
–0.21
–0.15
–0.31
0.05
0.39
–0.05
–0.01
–0.01
0.46
0.51
–0.55
0.19
0.38
0.27
–0.05
–0.09
–0.17
0.00
–0.03
0.31
–0.03
0.08
–0.15
–0.40
–0.04
0.13
–0.12
0.32
0.06
–0.20
0.21
–0.08
0.21
–0.09
0.16
0.14
–0.02
–0.23
–0.17
0.03
0.15
–0.14
–0.52
0.31
0.18
–0.19
0.16
0.04
0.01
–0.12
–0.14
–0.09
0.17
–0.08
–0.20
0.04
0.23
–0.01
–0.21
0.18
0.11
0.06
in granitic basins and have pHs between 5.0 and 6.7, calcium concentrations between 1.4 and 15 mg/L, and ECs
between 13 and 52 µS (Cooper and Andrus 1994; Johnson
1996; Cooper 1996). According to the rich to poor fen classification, which is based on minerotrophy, pH, and species
composition (Du Rietz 1949; Sjörs 1950a, 1961b, 1963,
1983; Gorham and Pearsall 1956; Gorham 1967; Malmer
1986), most Rocky Mountain fens are categorized as moderately rich or transitional.
The fens examined in this study are considerably more
alkaline and nutrient rich than typical Rocky Mountain
wetlands, and these fens are properly classified as extremely
rich fens or mires. In northern mires classified as extremely
rich, water pH ranges between 6.5 and 8.2, Ca2+ between 18
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:27 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
213
Fig. 4. Axes 1 and 2 of a CCA. Symbols indicate the TWINSPAN subclasses into which samples were grouped. Alphanumeric code
shows the sample location and number: h, High Creek Fen samples; c, Crooked Creek Fen;, and f, Fremont’s Fen. See text for additional details.
and 120 mg/L, and EC from 16 to 456 µS (Glaser et al.
1990; Zoltai and Johnson 1987; Karlin and Bliss 1984;
Slack et. al 1980; Sjörs 1961a). The mean and range of pore
water pH (x = 7.4) measured at the South Park mires fit well
within the range reported from other extremely rich mire
studies. The mean EC measured in South Park mires was
greater than the highest EC measured at other extremely rich
mires. Similarly, the mean concentration of calcium in South
Park mires was near the highest values reported from northern extremely rich mires. Other minerals were also at or
above the typical concentrations found elsewhere in such
wetlands. These results underscore the highly minerotrophic
nature of South Park mires.
Previous rich fen studies have reported a negative correlation between groundwater pH and Ca2+ concentration (e.g.,
Malmer et al. 1992). This pattern is borne out in South Park,
as well, where the correlation between pH and Ca2+ was
− 0.36 (p = 0.001). Station 2 at Fremont’s Fen epitomizes
this relationship, having both the lowest pH and the highest
groundwater calcium concentration.
Plant indicators of extremely rich fens
Plant species are frequently employed as synthetic indicators of fen minerotrophy, and the approach has been used
productively since it was first developed by Du Rietz (1949);
however, problems arise in the generic application of boreal
species indicators in novel areas owing to regional floristic
differences. The influence of biogeographical variation on
mire flora has been addressed by previous authors, and
regional floristic trends have frequently been treated as a
separate ecological gradient or direction of variation (Sjörs
1950b; Malmer 1986; Chee and Vitt 1989; Bridgham et al.
1996). Such differences are especially important to consider
when comparing fens located in the mixed southern Rocky
Mountain flora with fens in the boreal floristic province
wherein most investigations of fen ecology have taken place.
South Park’s calcareous fens have relatively few vascular
species in common with Midwestern calcareous fens, and
their flora includes only three Midwestern species indicative
of extremely rich fens, namely Triglochin maritimum,
Triglochin palustre, and Salix candida (e.g., Gates 1942;
Curtis 1959; Holte 1966; van der Valk 1975, 1976;
Weitzman 1983; Pearson and Leoschke 1992; Choesin and
Boerner 2000). South Park fen floras do include many European and Canadian species characteristic of both rich and
extremely rich fens, though, particularly the vascular species
Carex lanuginosa, Carex livida, and Habenaria (Limnorchis)
hyperborea, and the mosses Scorpidium scorpioides,
Scorpidium turgescens, Calliergon trifarium, Calliergon
giganteum, Campylium stellatum, and Tomenthypnum nitens
(Nordqvist 1950; Sjörs 1950a, 1959, 1961a; Slack et al.
1980; Karlin and Bliss 1984; Chee and Vitt 1989; Glaser et
al. 1990; Malmer et al. 1992).
South Park fens also include a number of boreal species
that have a very high affinity specifically for extremely rich
fen conditions. In Canada and Minnesota, Sjörs (1961a,
1963) and Glaser et al. (1990) found Triglochin maritimum,
Triglochin palustris, K. simpliciuscula, J. albescens, Carex
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:28 AM
Color profile: Generic CMYK printer profile
Composite Default screen
214
microglochin, Carex scirpoidea, Pentaphylloides floribunda
(Potentilla fruticosa), and Utricularia intermedia to be generally indicative of extremely rich fens. Salix candida also
frequently occurs in boreal extremely rich fens, but its value
as an indicator species was not evaluated in those studies.
Surprisingly, the common extremely rich fen indicator,
Muhlenbergia glomerata, which has been found on one occasion in Colorado (Weber 1990), has not been identified in
any South Park fens. It seems that Muhlenbergia
richardsonis may replace M. glomerata in these southern
calcareous fens. The fens of South Park have a number of
other species in common with boreal and arctic extremely
rich fens, such as Carex aquatilis, Carex rostrata, Carex
limosa, Salix planifolia, and J. arcticus, but these species
have wide ecological tolerances and thus have little value as
indicators of minerotrophy.
There are problems with using a number of the above
listed extremely rich fen indicator species in Colorado and
probably in the southern Rocky Mountains, in general. For
instance, Triglochin maritimum is commonly found growing
in saline areas not associated with fens or mires.
Pentaphylloides floribunda (Potentilla fruticosa) is a generalist species in this region and may be found growing in
mesic meadows, carrs, and virtually any mire habitat in
Colorado. Likewise, Carex lanuginosa and Habenaria
hyperborea can be found in a variety of wetland conditions,
and K. simpliciuscula most commonly grows in upland,
alpine habitats.
It is important to note that common subalpine transitional
fen species such as Carex aquatilis, Carex rostrata, Salix
planifolia, Salix monticola, Eriophorum angustifolium, and
Pedicularis groenlandica are usually present in, and may
dominate, portions of extremely rich fens in South Park.
Thus, fen classifications can easily err if based solely on a
superficial floristic evaluation. This observation corroborates
Gorham’s (1950) and Wheeler’s (1980) assertions that it is
not the presence of single or dominant species alone that is
important to fen classification but rather the suite of species
present.
Based on this study and previous calcareous fen investigations in South Park, Montana, Wyoming, and California
(Lesica 1986; Major and Taylor 1988; Fertig and Jones
1992; Cooper 1996; Sanderson and March 1996), we suggest a list of extremely rich fen indicator species for the
Rocky Mountains that includes Trichophorum pumilum,
Salix candida, Salix myrtillifolia, Carex microglochin, Carex
viridula, Eriophorum gracile, Scorpidium scorpioides,
Scorpidium turgescens, and Calliergon trifarium. In the
Rocky Mountains north of Colorado, M. glomerata also
seems to be a valuable indicator. If fen conditions are known
to exist at a site, Triglochin maritimum, Triglochin palustris,
Carex scirpoidea, K. myosuroides, K. simpliciuscula, and
Thalictrum alpinum are also key indicators of extremely rich
minerotrophy.
Ecological gradients in calcareous fens
While continental floristic differences, in part, differentiate southern Rocky Mountain extremely rich fens from their
boreal and midwestern counterparts, the major directions of
variation appear the same among the regions. The sections
below provide discussion of the characteristics and factors
Can. J. Bot. Vol. 81, 2003
underlying three major, synthetic gradients present on South
Park mires.
Water
The high and relatively stable water table present in mires
is clearly the predominant driver of floristic composition,
since it precludes the growth of upland and wetland species
that cannot tolerate continually waterlogged soils. But within
the mires themselves, hydrology also exhibits the strongest
control over plant species composition (Figs. 3 and 4). Hydrology has such a strong effect on mire vegetation because
it affects such a broad spectrum the mires’ physical and
chemical characteristics, including soil anoxia, peat accumulation, redox potential, nutrient supply and availability, and
salinity, among other factors (e.g., Moore and Bellamy 1974;
Clymo 1983; Sikora and Keeney 1983; Mitsch and
Gosselink 1993). Further, organic soils interact positively
with hydrology by raising water tables and buffering water
table fluctuations (Moore and Bellamy 1974; Ingram 1983),
both of which favor further accumulation of peat.
The influence of hydrology on South Park’s mires can be
evaluated on two scales: the meso-scale, which considers
water table patterns across the mire, and the micro-scale,
which encompasses differences in relative water table depth
due to the presence of microtopography. As has been widely
noted (e.g., Keddy 2000), at what we call the meso-scale,
there are two primary hydrograph attributes that affect wetland vegetation: water table depth and stability. On South
Park mires, these attributes are positively correlated, with
drier areas being subject to greater seasonal water table fluctuations than wetter areas (Table 3).
A qualitatively similar water table gradient is repeated
across these mires on a much smaller scale, at each hummock to hollow transition. Hummocks create relatively dry
micro-sites that allow the persistence of less hydrophilic species, especially shrubs. Hummocks grade into wet hollows,
in which only the most hydrophilic species grow. Like the
meso-scale, at the micro-scale, species composition is influenced by both microtopographically driven changes in water
relations and also by indirect effects, such as mineral
accumulation caused by wetting and drying cycles. The
environmental heterogeneity caused by hummock–hollow topography creates a fine-scale but profound effect on species
composition. The importance and effect of such micro-sites
have been described in the literature for various peatland
types (Vitt et al. 1975; Slack et al. 1980; Zoltai and Johnson
1985; Johnson 1996; Chimner and Hart 1996), although only
a few studies have experimentally quantified the effects of
such rises (e.g., Boogie 1972; Boogie and Miller 1976,
Vivian-Smith 1997).
Mire margin to expanse gradient
The presence of a mire margin to expanse gradient has
been well documented in the peatland literature (e.g., Sjörs
1948, 1950b; Du Rietz 1949; Moore and Bellamy 1974; van
der Valk 1976; Malmer 1986; Jeglum and He 1995;
Pakarinen 1995; Johnson 1996), and it has been quantitatively shown in this study to be one of the major directions
of variation on South Park mires (Fig. 3). The gradient
seems to be a synthetic, spatial representation of underlying
environmental patterns in hydrology, and water and soil
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:28 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
chemistry. Surprisingly, the gradient was not reconstructed
by the CCA; thus, it is unclear precisely which environmental factors are driving this gradient. There are two likely explanations for the absence of the margin–expanse gradient in
the CCA. Either environmental factors relative to this gradient were not measured, or this within-site gradient was
masked by stronger inter-site differences. The second option
seems more likely in this case, since factors generally
related to this gradient were measured in this study (e.g.,
Jeglum and He 1995), and further, a CCA including only
samples from Crooked Creek Fen (not shown) recreated the
mire margin–expanse gradient as axis 1. That CCA showed
marginal sites being associated with nutrient-rich soils high
in iron, calcium, manganese, and potassium, with expanse
sites having higher groundwater pHs. This pattern suggests
that, as has been noted in other studies (Gorham 1957;
Moore and Bellamy 1974; Johnson 1994), soil and water
mineral concentrations tend to decrease from the mire margin to the expanse owing to plant uptake and adsorption to
the peat. The graminoid-dominated mire margin to expanse
gradient shows a qualitatively similar pattern of mineral attenuation, but a strong decline in water table depth is also
present (e.g., Table 4).
When present as a shift from marginal willow carr to open
carr to fen expanse, this gradient is visually obvious. Less
apparent, however, is the shift from graminoid-dominated
margin to graminoid-dominated fen expanse. This type of
graminoid-dominated margin to expanse gradient differs
from that generally described in the literature, in that
physiognomic differences between vegetation types are subtle. Although visual differentiation of such a gradient may
be challenging, this is nonetheless an important direction of
variation within southern Rocky Mountain mires. Our studies and Lesica’s (1986) description of this gradient on a
Montana calcareous fen are at odds with Cooper’s (1996) assertion that the margin to expanse gradient is unimportant in
Rocky Mountain calcareous fens.
Regional gradients
South Park’s calcareous fens are all floristically related,
possessing large numbers of calciphilous species indicative
of extremely rich fen conditions, yet the floristic composition of each site is somewhat distinct (Figs. 3 and 4). Since
South Park is quite small on the continental scale, there is no
reason to believe that floristic disparities between sites are
the result of biogeographic effects. Instead, we suggest that
inter-mire differences result from the unique hydrogeologic
and geomorphic setting of each site.
In particular, edaphic characteristics vary among the wetlands as a result of differences in geology and geomorphology (Johnson 2000). For instance, High Creek Fen lies on
Quaternary alluvium primarily derived from the Maroon,
Belden, and Coffman Formations, situated over Maroon Formation bedrock (Stark et al. 1949; Singewald 1950; Tweto
1974). These formations are composed of limestones, sandstones, dolomites, gypsum, and shales, which produce mineral-rich, saline, and alkaline soils (Appel 1995).
Fremont’s Fen also lies on Quaternary alluvium, but this
wetland spans the South Park, Laramie, Foxhills, Reinecker
Ridge, Pierre Shale, and Eshe Porphyry formations (U.S.
Geological Survey, Milligan Lakes Geologic Quadrangle).
215
Fewer limestone beds occur in these members, and they
include a range of deposits, primarily sandstone, shale, volcanic, and coal. This strong heterogeneity in underlying
geology likely gives rise to the wide range of soil and
groundwater characteristics found at Fremont’s Fen.
Unlike the other two sites, Crooked Creek Fen lies in a
bedrock trough comprised of red, arkosic Maroon Formation
sandstone, inter-bedded with thick limestone and dolomite
strata (Stark et al. 1949). Crooked Creek Fen also differs
from the other sites in that beaver have had a strong influence on its formation and expansion. As evidenced by lacustrine strata in soil cores, many of the peat aprons and fen
expanse sites were at one time beaver ponds (Johnson 2000).
Abandoned beaver ponds that are undergoing this type of
terrestrialization or Verlandung (Weber 1911) are currently
found at the head of the mire.
These differing hydrogeologic templates (sensu Bedford
1996) create the environmental context for the mires, producing a unique environment at each site, which in turn
influences wetland floristics and vegetational composition.
This type of intra-regional pattern is in contrast to that
described in larger-scale mire comparisons, wherein differences in the regional species pool can overshadow local
influences.
Ecological processes influencing South Park mire
vegetation
A hierarchy of factors act to produce the patterns in vegetation seen on South Park mires. At the highest level, the regional flora and hydrogeologic and geomorphic setting form
the biological and environmental context for the wetlands
(Major 1951; Muller-Dombois and Ellenberg 1974; Bedford
1996; Winter 2001). Constrained by these factors, gradients
of chemistry and hydrology create a matrix of mire habitats
or potential niche spaces (sensu Hutchinson 1957). Within
these habitats, physiological limitations preclude the growth
of unsuitable species, while biotic interactions and disturbance events act to shape vegetational composition
(Mueller-Dombois and Ellenberg 1974; Keddy 2000).
We have focused on describing patterns in the vegetation
and environment of calcareous mires. Although causal processes were not directly investigated, we can speculate on
the role that fundamental mechanisms such as disturbance,
competition, and facilitation have on community composition. These inferences pose a set of questions ripe for future
study.
Natural and anthropogenic disturbance produce a mosaic
of vegetation patches in natural systems (e.g., Pickett and
White 1985). Johnson (2000) considered the effects of
anthropogenic disturbances such as surface mining and
drainage on South Park mires, but the role of natural disturbance has been unexplored. For instance, South Park mires
occasionally burn on the surface (J.B. Johnson, personal observation), but the frequency, severity, and effects of such
fires are virtually unknown. Likewise, the primary and secondary effects of herbivory have not been investigated, in
spite of the fact that grazers, both cattle and wild species,
are known to cause or heighten the development of
hummock–hollow topography, which in turn strongly affects
species composition (Fig. 4).
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:29 AM
Color profile: Generic CMYK printer profile
Composite Default screen
216
Environmental conditions create the habitats that dictate
potential flora, but ultimately biotic interactions shape the
realized patterns of species occurrence. Only a few species
can persist in the severest of South Park mire habitats, such
as the quagmires and marl flats (Cooper 1996; Johnson
2000). Consistent with both theoretical and empirical studies
(e.g., Grime 1973, 1979; Wilson and Keddy 1986a;
Twolan-Strutt and Keddy 1996), we speculate that in these
stressful areas, which have low total plant coverage and production, negative species interactions such as competition
are minimized (but see Tilman 1982). Positive interactions
in such areas may be quite important, though, as suggested
by theoretical models (Callaway 1997; Brooker and
Callaghan 1998) and studies of fens (Johnson 1997), lakes
(Wilson and Keddy 1986a, 1986b), or, more commonly, saline wetlands (Salzman and Parker 1985; Bertness and
Shumway 1993; Bertness and Hacker 1994; Castellanos et
al. 1994; Hacker and Bertness 1995).
In drier areas, usually sited on the mire margin, the environment is more benign. Here, a “generalist” wetland flora
tends to develop, characterized by plants that can survive periods of both inundation and draw down (Svejcar and Riegel
1998) but which cannot persist under either condition indefinitely because of competitive exclusion, physiological limitations, or a combination of both factors. In terms of Grime’s
(1979) C–S–R plant strategy classification, most species
occurring in such areas are competitors (e.g., Salix spp.) or
competitive stress-tolerators (e.g., Deschampsia cespitosa;
Grime 1979; Grime et al. 1988). Using a somewhat different
approach, Boutin and Keddy (1993), Gaudet and Keddy
(1995), and Keddy et al. (2002) also identified common
marginal species such as Calamagrostis canadensis, Carex
utriculata, Elymus trachycaulus, and Potentilla anserina as
being competitive to highly competitive. High, but not extreme, soil nutrient content could play a role in heightening
competitive interactions in marginal areas (e.g., Wilson and
Keddy 1986a), but small-scale habitat heterogeneity created
by microtopography seems to preclude the development of a
species-poor sward (Johnson 1996; Vivian-Smith 1997). Fen
lawns are similarly viewed as areas heavily influenced by
competitive interactions, being typically species poor and
dominated by strongly competitive species such as Carex
aquatilis, Carex utriculata, or Carex simulata. In terms of
Wisheu and Keddy’s (1992) centrifugal model of plant community organization, the lawns and margins can be thought
of as “core” habitat, whereas quagmires, water tracks, and
marl flats are “peripheral”.
Acknowledgements
The authors would like to thank M. Beardsley, M. Edmiston,
and T. Gerhart for their outstanding efforts during fieldwork.
We are also indebted to M. Scott, L. MacDonald, R. Dix, and
M. Gilbert for their help and guidance throughout this study.
The authors sincerely appreciate Dr. William Weber’s identification of moss species. This project was funded by US EPA
104 (b) (3) program grants, Nos. 100280, 100291, 100295, and
100304.
Can. J. Bot. Vol. 81, 2003
References
Allen, R.B., and Peet, R.K. 1990. Gradient analysis of the forests of
the Sangre de Cristo Range, Colorado. Can. J. Bot. 68: 193–201.
Almendinger, J.E., and Leete, J.H. 1998. Regional and local hydrology of calcareous fens in the Minnesota River Basin, USA.
Wetlands, 18: 184–202.
Appel, J.L. 1995. Hydrogeologic framework of the High Creek
Calcareous Fen, South Park, Park County, Colorado. M.Sc. thesis, Department of Earth and Planetary Sciences, University of
New Mexico, Albuquerque, N. Mex.
Bedford, B. 1996. The need to define hydrologic equivalence at the
landscape scale for freshwater wetland mitigation. Ecol. Appl.
6: 57–68.
Bertness, M.D., and Hacker, S.D. 1994. Physical stress and positive associations among marsh plants. Am. Nat. 144: 363–372.
Bertness, M.D., and Shumway, S.W. 1993. Competition and facilitation in marsh plants. Am. Nat. 142: 718–724.
Bierly, K.F. 1972. Meadow and fen vegetation in Big Meadows,
Rocky Mountain National Park. M.Sc. thesis, Department of
Botany, Colorado State University, Fort Collins, Colo.
Boogie, R. 1972. The effect of water-table height on root development of P. contorta on deep peat in Scotland. Okios, 23: 304–312.
Boogie, R., and Miller, H.G. 1976. Growth of Pinus contorta at
different water-table levels in deep blanket peat. Forestry, 49:
123–131.
Boutin, C., and Keddy, P.A. 1993. A functional classification of
wetland plants. J. Veg. Sci. 4: 591–600.
Bridgham, S.D., Pastor, J., Janssens, J.A., Chapin, C., and
Malterer, T.J. 1996. Multiple limiting gradients in peatlands: a
call for a new paradigm. Wetlands, 16: 45–65.
Brooker, R.W., and Callaghan, T.V. 1998. The balance between
positive and negative plant interactions and its relationship to
environmental gradients: a model. Oikos, 81: 196–207.
Callaway, R.M. 1997. Positive interactions in plant communities and
the individualistic-continuum concept. Oecologia, 112: 143–149.
Carpenter, Q. 1990. Hydrology and vegetation on a calcareous peat
mound fen in southeastern Wisconsin. M.Sc. thesis, University
of Wisconsin-Madison, Madison, Wisc.
Castellanos, E.M., Figueroa, M.E., and Davy, A.J. 1994. Nucleation and facilitation in salt marsh succession: interactions between Spartina maritima and Athrocnemum perenne. J. Ecol. 82:
239–248.
Chee, W., and Vitt, D.H. 1989. The vegetation, surface chemistry,
and peat chemistry of moderate-rich fens in Central Alberta,
Canada. Wetlands, 9: 227–261.
Chimner, R., and Hart, J. 1996. Hydrology and microtopography
effects on northern white-cedar regeneration in Michigan’s Upper Penninsula. Can. J. For. Res. 26: 389–393.
Choesin, D., and Boerner, R. 2000. Vegetation and groundwater
alkalinity of Betsch fen, a remnant periglacial fen in
south-central Ohio. Castanea, 65: 196–206.
Clymo, R.S. 1983. Peat. In Ecosystems of the world. Mires:
swamp, bog, fen and moor. Vol. 4A. Edited by A.J.P. Gore.
Elsevier, Amsterdam. pp. 159–224.
Cooper, D.J. 1986. Community structure and classification of
Rocky Mountain wetland ecosystems. In An ecological characterization of Rocky Mountain montane and subalpine wetlands.
Edited by J. Windell. U.S. Fish and Wildlife Service Biological
Report 86(11), Washington, D.C. pp. 66–147.
Cooper, D.J. 1990. Ecology of wetlands in Big Meadows, Rocky
Mountain National Park, Colorado. U.S. Fish and Wildlife Service Biological Report (90)15, Washington, D.C.
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:29 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
Cooper, D.J. 1996. Water and soil chemistry, floristics, and
phytosociology of the extreme rich High Creek fen, in South
Park, Colorado, U.S.A. Can. J. Bot. 74: 1801–1811.
Cooper, D.J., and Andrus, R.E. 1994. Patterns of vegetation and
water chemistry in peatlands of the west-central Wind River
Range, Wyoming, U.S.A. Can. J. Bot. 72: 1586–1597.
Cooper, D.J. and Sanderson, J. 1997. A montane Kobresia
myosuroides fen community type in the southern Rocky Mountains of Colorado, U.S.A. Arct. Antarct. Alp. Res. 29: 300–303.
Curtis, J.T. 1959. The vegetation of Wisconsin. University of Wisconsin Press, Madison, Wisc.
De Voto, R.H. 1971. Geologic history of South Park and geology
of the Antero Reservoir quadrangle, Colorado. Q. Colo. School
Mines, 66: 1–90.
Du Rietz, G.E. 1949. Huvudenheter och huvudgränser i svensk
myrvegetation. Sven. Bot. Tidskr. 43: 274–309.
Eggers, S., and Reed, D. 1987. Wetland plants and plant communities of Minnesota and Wisconsin. U.S. Army Corps of Engineers, St. Paul, Minn.
Fertig, W., and Jones, G. 1992. Plant communities and rare plant
species of the Swamp Lake Botanical Area. Unpublished professional report available from Clark’s Ranger District, Shoshone
National Forest, Wyo.
Gard, M., Yarbrough, S., and Lewis, B. 2000. South Park wetland
evaluation. Available from U.S. Bureau of Land Management,
National Applied Resource Sciences Center, Lakewood, Colo.
Gates, F. 1942. The bogs of lower Michigan. Ecol. Monogr. 12:
214–254.
Gauch, H.G. 1982. Multivariate analysis in community ecology.
Cambridge University Press, Cambridge, U.K.
Gaudet, C.L., and Keddy, P.A. 1995. Competitive performance and
species distribution in shoreline plant communities: a comparative approach. Ecology, 76: 280–291.
Glaser, P.H., Janssens, J.A., and Siegel, D.I. 1990. The response of
vegetation to chemical and hydrological gradients in the Lost
River Peatland, northern Minnesota. J. Ecol. 78: 1021–1048.
Gorham, E. 1950. Variations in some chemical conditions along
the borders of a Carex lasiocarpa fen community. Oikos, 2:
217–240.
Gorham, E. 1957. The development of peatlands. Q. Rev. Biol. 32:
145–166.
Gorham, E. 1967. Some chemical aspects of wetland ecology.
Technical memorandum of the National Research Council of
Canada Association Committee on Geotechnical Research,
Ottawa, Ont.
Gorham, E., and Pearsall, W.H. 1956. Acidity, specific conductivity, and Ca content of some bog and fen waters in Northern Britain. J. Ecol. 44: 129–141.
Grime, P.J. 1973. Competitive exclusion in herbaceous vegetation.
Nature, 242: 344–347.
Grime, J.P. 1979. Plant strategies and vegetation processes. James
Wiley, New York.
Grime, J.P., Hodgson, J.G., and Hunt, R. 1988. Comparative plant
ecology. Unwin Hyman, London.
Hacker, S.D. and Bertness, M.D. 1995. Morphological and physiological consequences of a positive plant interaction. Ecology,
76: 2165–2175.
Hill, M.O. 1979. TWINSPAN – a FORTRAN program for arranging multivariate data in an ordered two-way table by classification of individuals and attributes. Cornell University Press,
Ithaca, N.Y.
Holte, K. 1966. A floristic and ecological analysis of the Excelsior
217
fen complex in northwest Iowa. Ph.D. Thesis, University of
Iowa, Iowa City, Iowa.
Holte, K.E., and Thorne, R.F. 1962. Discovery of a calcareous fen
complex in northwest Iowa. Proc. Iowa Acad. Sci. 69: 54–60.
Hutchinson, G.E. 1957. Concluding remarks. Cold Spring Harbor
Symp. Quant. Biol. 22: 415–427.
Ingram, H.A.P. 1983. Hydrology. In Ecosystems of the world.
Mires: swamp, bog, fen, and moor. Vol 4A. Edited by
A.J.P. Gore. Elsevier, Amsterdam. pp. 67–158.
Jeglum, J., and He, F. 1995. Pattern and vegetation-environment relationships in a boreal forested wetland in northeastern Ontario.
Can. J. Bot. 73: 629–637.
Johnson, J.B. 1994. The vegetational dynamics of a subalpine treed
fen in Rocky Mountain National Park, Colorado. M.Sc. thesis,
Department of Biology, Colorado State University, Fort Collins,
Colo.
Johnson, J.B. 1996. Phytosociology and gradient analysis of a subalpine treed fen in Rocky Mountain National Park, Colorado.
Can. J. Bot. 74: 1203–1218.
Johnson, J.B. 1997. Stand structure and vegetation dynamics of a
subalpine treed fen in Rocky Mountain National Park, Colorado.
J. Veg. Sci. 8: 337–342.
Johnson, J.B. 2000. The ecology of calcareous fens in Park
County, Colorado. Ph.D. thesis, Department of Biology, Colorado State University, Fort Collins, Colo.
Johnson, J.B., and Gerhardt, T.D. 2002. Mapping and characterization of mires and fens in South Park, Park County, Colorado.
Unpublished professional report available from the U.S. Bureau
of Land Management, Canon City, Colo.
Jongman, R., ter Braak, C.J.F., and van Tongeren, O.F.R. 1995.
Data analysis in community and landscape ecology. Cambridge
University Press, Cambridge, U.K.
Karlin, E.F., and Bliss, L.C. 1984. Variation in substrate chemistry
along microtopographical and water-chemistry gradients in
peatlands. Can. J. Bot. 62: 142–153.
Kartesz, J.T. 1994. A synonymized checklist of the vascular flora
of the United States, Canada, and Greenland. Timber Press,
Portland, Oreg.
Keddy, P.A. 2000. Wetland ecology: principles and conservation.
University of Cambridge Press, Cambridge, U.K.
Keddy, P.A., Nielsen, K., Weiher, E., and Lawson, R. 2002. Relative competitive performance of 63 species of terrestrial herbaceous plants. J. Veg. Sci. 13: 5–16.
Lesica, P. 1986. Vegetation and flora of the Pine Butte Fen, Teton
County, Montana. Great Basin Nat. 46: 22–32.
Lozano, E. 1967. Stratigraphy and structure of Tertiary rocks in
central South Park, Park County, Colorado. Mountain Geologist,
4: 119–126.
Major, J. 1951. A functional, factorial approach to plant ecology.
Ecology, 32: 392–412.
Major, J., and Taylor, D.W. 1988. Alpine. In Terrestrial vegetation
of California. Edited by M.G. Barbour and J. Major. California
Native Plant Society Special Publication, 9. Sacramento, Calif.
Malmer, N. 1986. Vegetation gradients in relation to environmental
conditions in northwestern European mires. Can. J. Bot. 64:
375–383.
Malmer, N., Horton, D.G., and Vitt, D.H. 1992. Element concentrations in mosses and surface waters of western Canadian mires
relative to precipitation chemistry and hydrology. Ecography,
15: 114–128.
Marr, J.W. 1961. Ecosystems of the Front Range in Colorado. University of Colorado’s Studies in Biology, No. 8.
McCune, B., and Mefford, M.J. 1997. PC-ORD for Windows.
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:30 AM
Color profile: Generic CMYK printer profile
Composite Default screen
218
Multivariate analysis of ecological data. Version 3.04. MjM
Software, Gleneden Beach, Oreg.
Mitsch, W.J., and Gosselink, J. 1993. Wetlands. Van Nostrand
Reinhold Co., New York.
Moore, P.D., and Bellamy, D.J. 1974. Peatlands. Springer-Verlag,
New York.
Mueller-Dombois, D., and Ellenberg, H. 1974. Aims and methods
of vegetation ecology. John Wiley and Sons, New York.
Nordqvist, Y. 1950. Calcareous fens in Jämtland. In The plant
cover of Sweden. Acta Phytogeogr. Suec. 50: 193–197.
Økland, R.H. 1996. Are ordination and constrained ordination
alternative or complimentary strategies in general ecological
studies? J. Veg. Sci. 7: 289–292.
Pakarinen, P. 1995. Classification of boreal mires in Finland and
Scandinavia: A review. Vegetatio, 118: 29–38.
Palmer, M.W. 1993. Putting things in even better order: the advantages of canonical correspondence analysis. Ecology, 74:
2215–2230.
Pearson, J., and Leoschke, L. 1992. Floristic composition and conservation status of fens in Iowa. J. Iowa Acad. Sci. 99: 41–52.
Prentice, H.C., and Cramer, W. 1990. The plant community as a
niche bioassay: Environmental correlates of local variation in
Gypsophila fastigiata. J. Ecol. 78: 313–325.
Pickett, S.T.A., and White, P.S. (Editors). 1985. The ecology of
natural disturbance and patch dynamics. Academic Press, San
Diego, Calif.
Racine, C.H., and Walters, J.C. 1994. Groundwater-discharge fens
in the Tanana lowlands, interior Alaska, U.S.A. Arct. Alp. Res.
26: 418–426.
Salzman, A.G. and Parker, M.A. 1985. Neighbors ameliorate local
salinity stress for a rhizomatous plant in a heterogeneous environment. Oecologia, 65: 273–277.
Sanderson, J., and March, M. 1996. Extreme rich fens of South
Park, Colorado: Their distribution, identification, and natural
heritage significance. Unpublished professional report available
from the U.S. Environmental Protection Agency, Region VIII,
Denver, Colo.
Sikora, L.J., and Keeney, D.R. 1983. Further aspects of soil chemistry under anaerobic conditions. In Ecosystems of the world.
Mires: swamp, bog, fen, and moor. Vol. 4A. Edited by
A.J.P. Gore. Elsevier, Amsterdam. pp. 247–256.
Singewald, Q.D. 1950. Gold placers and their geologic environment in northwestern Park County, Colorado. U.S. Geological
Survey Bulletin 955.
Sjörs, H. 1948. Myrvegetation i Bergslagen. Acta Phytogeogr.
Suec. 21: 1–299.
Sjörs, H. 1950a. On the relation between vegetation and electrolytes in North Swedish mire waters. Okios, 2: 241–258.
Sjörs, H. 1950b. Regional studies in North Swedish mire vegetation. Bot. Not. 1950: 173–222.
Sjörs, H. 1959. Bogs and fens in the Hudson Bay lowlands. Arctic,
12: 2–19.
Sjörs, H. 1961a. Forest and peatland at Hawley Lake, northern Ontario. Natl. Mus. Can. Bull. Contributions to Botany, 171: 1–31.
Sjörs, H. 1961b. Surface patterns in boreal peatlands. Endeavour,
80: 217–224.
Sjörs, H. 1963. Bogs and fens on Attawapiskat River, northern Ontario. Natl. Mus. Can. Bull., Contributions to Botany, 186: 45–133.
Sjörs, H. 1983. A comparison between mires of southern Alaska
and Fennoscandia. Aquilo Ser. Bot. 21: 89–94.
Slack, N.G., Vitt, D.H., and Horton, D.G. 1980. Vegetation gradients of minerotrophically rich fens in western Alberta. Can. J.
Bot. 58: 330–350.
Can. J. Bot. Vol. 81, 2003
Sokal, R.R., and Rohlf, F.J. 1981. Biometry. W.H. Freeman and
Co., New York.
Spahr, N.E. 1981. Variations in climate characteristics as related to
evapotranspiration in South Park, Central Park County, Colorado. Unpublished professional report available from the U.S.
Geological Survey, Water-Resource Investigation 80-86,
Lakewood, Colo.
Stark, J.T., Johnson, J.H., Behre, C.H., Powers, W.E.,
Howland, A.L., and Gould, D.B. 1949. Geology and origin of
South Park, Colorado. Geol. Soc. Am. Mem. 33: 188.
Svejcar, T., and Riegel, G.M. 1998. Spatial pattern of gas exchange
for montane moist meadow species. J. Veg. Sci. 9: 85–94.
Ter Braak, C.J.F. 1986. Canonical correspondence analysis: A new
eigenvector technique for multi-variate direct gradient analysis.
Ecology, 67: 1167–1179.
Ter Braak, C.J.F., and Smilauer, P. 1998. Canoco reference manual
and user’s guide to Canoco for Windows: software for canonical
community ordination. Version 4. Microcomputer Power, Ithaca,
N.Y.
Tilman, D. 1982. Resource competition and community structure.
Princeton University Press, Princeton, N.J.
Twolan-Strutt, L., and Keddy, P. 1996. Above- and belowground
competition intensity in two contrasting wetland plant communities. Ecology, 77: 259–270.
Tweto, O. 1974. Reconnaissance geologic map of the Fairplay
West, Mount Sherman, South Peak, and Jones Hill 7½ minute
quadrangles, Park, Lake, and Chaffee counties, Colorado. U.S.
Geological Survey Miscellaneous Field Studies Map MF-555,
scale 1 : 62 500.
U.S. Fish and Wildlife Service. 1999. Peatland mitigation policy
considerations, Region 6. Policy document available through the
Ecological Services Center, Colorado Field Office, Lakewood,
Colo.
van der Valk, A.G. 1975. Floristic composition and structure of fen
communities in northwest Iowa. Proc. Iowa Acad. Sci. 82: 113–118.
van der Valk, A.G. 1976. Zonation, competitive displacement and
standing crop of northwest Iowa fen communities. Proc. Iowa
Acad. Sci. 83: 50–53.
Vitt, D.H., Achuff, P., and Andrus, R.E. 1975. The vegetation and
chemical properties of patterned fens in the Swan Hills, north
central Alberta. Can. J. Bot. 53: 2776–2795.
Vivian-Smith, G. 1997. Microtopographic heterogeneity and
floristic diversity in experimental wetland communities. J. Ecol.
85: 71–82.
Walter, I.A., Siemer, E.G., Quinlan, J.P., and Burman, R.D. 1990.
Evapotranspiration and agronomic responses in formerly irrigated mountain meadows, South Park, CO. Unpublished professional report available from the Board of Water Commissioners,
City and County of Denver, Denver, Colo.
Warner, B., and Rubec, C.D.A. 1997. The Canadian wetland classification system. Available from the Wetlands Research Center,
University of Waterloo, Waterloo, Ont. Catalog No.
CW66-156/1997.
Weber, C.A. 1911. Das Moore. Hannoversche Geschichtsblätter,
14: 255–270.
Weber, W.A. 1990. Colorado flora: eastern slope. University of
Colorado Press, Boulder, Colo.
Weitzman, A. 1983. Summerby swamp, an unusual plant community in Mackinac County, Michigan. Mich. Bot. 23: 11–18.
Wheeler, B.D. 1980. Plant communities of rich fen systems in
England and Wales. II. Communities of calcareous mires. J.
Ecol. 68: 405–420.
Wilson, S.D., and Keddy, P.A. 1986a. Species competitive ability
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:30 AM
Color profile: Generic CMYK printer profile
Composite Default screen
Johnson and Steingraeber
and position along a natural stress/disturbance gradient. Ecology, 67: 1236–1242.
Wilson, S.D., and Keddy, P.A. 1986b. Measuring diffuse competition along an environmental gradient: results from a shoreline
plant community. Am. Nat. 127: 862–869.
Winter, T.C., 2001. The concept of hydrologic landscapes. J. Am.
Water Resources Assoc. 37: 335–350.
Wisheu, I., and Keddy, P. 1992. Competition and centrifugal organization of plant communities: theory and tests. J. Veg. Sci. 3:
147–156.
219
Zoltai, S.C., and Johnson, J.D. 1985. Development of a treed bog
island in a minerotrophic fen. Can. J. Bot. 63: 1076–1085.
Zoltai, S., and Johnson, J.D. 1987. Relationships between nutrients
and vegetation in peatlands in the prairie provinces. In Proceedings, Symposium 1987: Wetlands/Peatlands, August 23–27, 1987,
Edmonton, Alta. Edited by C.D.A. Rubec and R.P. Overend.
pp. 535–542.
© 2003 NRC Canada
J:\cjb\cjb8103\B03-017.vp
Wednesday, March 19, 2003 8:55:30 AM