Selecting wilderness areas to conserve Utah's biological diversity

24 Pages·1996·13.9 MB·English

by  D W Davidson

The Great Basin Naturalist
Published at Proxo, Utah, by
Brigham Young University
ISSN 0017-3614
Volume 56 30 April 1996 No. 2
GreatBasin Naturalist56(2), © 1996,pp.95-118
SELECTING WILDERNESS AREAS TO CONSERVE
UTAH'S BIOLOGICAL DIVERSITY
DianeW. Davidsonl, William D. Newmark-,JackW. Sites,Jr.'^ Dennis K. Shiozawa'^,
EricA. Rickart-, KimballT. Harper"', and Robert R. Keiter^
—
Abstr.\c:t. Congress is currently evaluating the wilderness status ofBureau of Land Management (BLM) public
lands in Utah. Wilderness areas playmany importantroles, andonecritical role is the consei'vation ofbiological diver-
sity. Wepropose thatobjectives forconser\'ingbiodiversity on BLM lands in Utahbeto(1) ensure thelong-term popu-
lation viability ofnative animal and plant species, (2) maintain the critical ecological and evolutionaiy processes upon
whichthesespeciesdepend,and(3)preservethefullrangeofcommimities, successionalstages,andenvironmentalgra-
dients.Toachievetheseobjectives,wildernessareasshouldbeselectedsoastoprotectlarge,contiguousareas,augment
existing protected areas, bufferwilderness areas with multiple-use public lands, interconnect existing protected areas
with dispersalandmovementcorridors, conserveentirewatershedsandelevational gradients, protectnativecommuni-
ties from invasions ofe.xotic species, protect sites ofmaximum species diversity, protect sites with rare and endemic
species, and protect habitats ofthreatened and endangered species. We use a few comparatively well-studied ta.xa as
examplestohighlighttheimportanceofparticularBLM lands.
Key words: wilderness, biodiversity, conservation. Utah, Bureau ofLand Management, endemic species, exotic
species, cryptobioticsoils,plants,bees, vertebrates.
The Wilderness Actand Biodiversity historical value" (16 U.S. Code, § 1131 [c][4]).
Ecological concerns have also figured promi-
In the Wilderness Act of 1964, Congress nently in several congressional wilderness
endorsed the presentation offederal land in its bills for Bureau of Land Management (BLM)
natural state (16 U.S. Code, Sections 1131-36). public lands. Both theAlaska National Interest
Congress plainly anticipated that ecological Lands Conservation Act, 16 U.S. Code, § 3101
considerations w^ere an important dimension (b), and the California Desert Protection Act,
of the wilderness concept, since the act pro- 103 Public Law433 Section 2 (b) (1) (B) (1994),
vides that wilderness may contain "ecological" expresslyacknowledge thatwilderness designa-
features of "scientific, educational, scenic, or tion is intended to protect important ecological
115fpartiiu-iitotBinloKx',Unh'ersih-ofUtah,SaltLakeGit)'.UTS4112.
^UtahMuseumofNaturalHistory,Universit\-ofUt;di,SaltLakeCity,UT84112.
'DepartmentofZoology,BrighamYoungUniversity,Provo,UT84602.
"^DepartmentofBotan\',BrighamYoungUniversity,Provo.UT84602.
^CollegeofLaw,Uni%ersit\ofUtah,SaltLakeCit\',UT84112.
95
96 Great Basin Naturalist [Volume 56
values. Among the significant ecological func-
tions ofwilderness areas is their role in con-
sei^vingbiological diversity (biodiversity).
In Utah, undeveloped public lands admin-
BLM
istered by the (Fig. 1) can potentially
play a key role in conserving the state's natural
heritage. The BLM is now pursuing an ecosys-
tem management policy designed to ensure
sustainable ecological processes and biological
diversityonlands underitsjurisdiction (Depart-
ment of the Interior 1994). By using these
same criteria to designate wilderness areas.
Congress could not only advance the BLM's
ecosystem management goals but also reduce
conflict over the agency's multiple-use lands
(e.g., by diminishing the risk offuture endan-
gered species listings and the accompanying
regulatoiy limitations). Over the long tenn, it is
both cheaper and easier to protect species in
aggregate in their intact, functioning ecosys-
tems than to conserve them individually in
fragmented and decimated populations under
the Endangered Species Act. Fig. 1. MapofthestateofUtah showing(inblack)loca-
In short, the use ofbiological and ecologi- tions of all existing roadless areas proposed for BLM
BLM wilderness status. The BLM formally studied a suliset of
cal criteria to designate wilderness areas these areas and recommended a portion ot studied lands
in Utah is consistent with the legal concept of forwildernessstatus. DataarefromaDepartmentofInte-
wilderness and would help to avoid future riormapofBLM Wilderness StudyAreas, BLM Proposed
conflicts overresource management. Wilderness, and the Utah Wilderness Coalition's BLM
Wilderness Proposal. County boundaries also are shown.
Isolated moiuitain ranges in Utah's western deserts are
BioDivERSiTi' Defined identifiedasfollows:a = DeepCreek;b = Fish Springs;c
— — House range, and d = Newfoundland range (not for-
givBeinolaorgeiac—alidnicvleurdseitsythrtehee vhaireireatrychoifcallifecoinm-a mthaellCyoplorroapdoosePdlaotreasut,udeie=dtfhoerwHielndne'rnMeossundteasiingsna.tion). On
ponents: genetic diversity, species diversity, and
ecosystem diversity (e.g.. National Research sity consists ofthe xariety ofmajor ecological
Council 1978, Wilson 1988, Reid and Miller communities within areas that are heteroge-
1989, Raven 1992). Cenetic diversity refers to neous in their physical attributes, for example,
the variety ofgenes within species. Depletion in elevation or soil type.
of genetic diversity during population bottle- Genetic, species, and ecosystem di\'ersity
necks, or because of inbreeding within frag- all result from both interactions bet\\'een organ-
mented and isolated populations, can threaten isms and their environments, and interactions
a species' sundval by reducing the capacity of of organisms with one another. The physical
organisms to adapt to changing environments environment sets limits on wliich species can
(Soule and Wilcox 1980, Frankel and Soule inhabit an area, and interactions among those
1981). Species diversity, or the number of species determine which are most abundant.
species within a region (species richness), can Strategies for preserving biodixcrsitx' must
be divided into three major components therefore take note of all li\ing things in the
(Whittaker 1972): alpha diversity {a), the num- landscape, and the linkages among them.
ber ofspecies in a homogeneous habitat; beta Finally, since different species specialize on
diversity (/3), the rate of species-turnover different stages of natural disturbance cycles,
across habitats; and gamma diversity (y), the it is important to presei-ve a range ofcommu-
total number of species observed in all habi- nities and ecosystems representing all stages
tats within a region. Finally, ecosystem diver- in the disturbance cvcle.
.
1996] WlLDKRNESS SELECTION FOR BlODINKHSITY 97
Objectives have several ofthese traits. On BLM lands in
Utah, examples of such organisms are river
The success ofconserving biological diver- otter {Lutra (•anadi'iisi.s) and both Bald and
sit) within a s\stem ot" protected areas can Golden Eagles {Haliaeefus lencoc('})lialu.s and
only be assessed in relationship to a series of AqiiiJa chrijsaetos). Risk-prone plants include
selected objectives. We propose that the con- Holmgren locoweed {Astragalus hobngrenio-
sei-vation ofUtah's biological diversity depends nim) and Jones cycladenia {Cijclaclenia huinilis
on (1) ensuring die long-term viability ofnative var. jonesii), which have highly specific sub-
plant and animal populations, (2) maintaining strate recjuirements
the criticd ecological andevolutionaiyprocesses Viability of populations depends on both
upon which these species depend, and (3) pro- the level ofriskone is willingto accept, and the
tecting the full range of communities, succes- time frame overwhich one wishes to consene
sional stages, and environmental gradients (e.g., the population (Shaffer 1981, Schonewald-Cox
lUCN 1978, MacKinnonetal. 1986, Noss 1992). 1983, Soule 1987). In general, both survival
Both the size of the network of protected time and the likelihood of population persis-
areas and the selectionofindividualwilderness tence increase with population size. A level of
areas should be guided by these 3 goals. risk and persistence that is commonly pro-
Although it is possible to presence a small sub- posed as a management goal is a 99% chance
set ofspecies and genotypes in zoological and ofsui-vival for 1000 years (e.g., Belovsky 1987,
botanical gardens, communities and species Armbrusterand Lande 1993).
interactions must be consei-ved in situ. Large For large carnivores, the minimum viable
areas with minimal human intrusion, and with population necessary to ensure a 99% chance
natural processes reasonabK' intact, are critical of survival for 1000 vears is estimated to be
elements of an in situ conservation strategy; approximately 10,000-100,000 individuals (Be-
tlieyprovide protection forfiagilehabitats, such lovskv 1987). In habitat area, this is equivalent
as easily eroded soils, and preserve habitat for to 100,000-1,000,000 km2, or 2.5-25 million
reclusive species. Moreover, wilderness areas acres. Although this area requirement may
offer natural ecosystems some protection from seem remarkably large, documented losses of
the biological invasions that have devastated mammalian species from among the largest of
many communities, especially plant communi- North American national parks (e.g., the
ties, across Utah. 10,328-km^ Yellowstone-Grand Teton park
Here we describe a strategy, based upon assemblage) during the last 90 years make
widely accepted principles of conservation clear the importance ofprotecting large areas
biology (see e.g., Primack 1993, Meffe and (Newmark 1987, 1995).
Carroll 1994), for both selecting critical sites
Maintenance ofEcological
for wilderness designation and determining
and Evolutionary Processes
the amount of habitat that should be pre-
served as wilderness (see also Babbitt 1995). In selecting wilderness areas, one must
take care to ensure the maintenance of the
Criteria for Selection ecological and evolutionary processes upon
which all plant and animal species depend
Viable Populations
(Pickett and Thompson 1978, Kushlan 1979).
Utah contains approximately 3000 indige- Among the most important ofthese processes
nous plant species and varieties and about 584 are natural disturbance and recovery cycles.
vertebrate species. Viable populations formost Ideally, criteria for the selection ofwilderness
ofthese plants and animals can be ensured by areas should include information on fre-
focusing, within ecological communities, on quency, size, and longevity of natural distur-
species for which the risk of extinction is bances. Protected areas should be large
greatest. Risk-prone species typically include enough to contain minimum critical areas of
those with small populations, large home the entire range of recovery stages for each
range requirements, low reproductive poten- community type (Pickett and Thompson
tial, restricted geographic ranges, or large 1978). In western North America, natural dis-
temporal xariation in population size (Brown turbance regimes can encompass tens ofthou-
1971, Willis 1974, Terborgh and Winter 1980, sands to millions ofacres, as witnessed by the
Diamond 1984, Pimm et al. 1988, Belovsky et recent and extensive wildfires in Yellowstone
al. 1994, Newmark 1995). Many top predators National Park (Christensen etal. 1989).
98 Great Basin Natueulist [Volume 56
Two other critical ecological processes are designating wilderness areas, high priority
migration and dispersal of terrestrial organ- should be given to lands whose selection
isms across landscapes, and ofaquatic species would enlarge and connect existing protected
within watersheds. The selection of wilder- areas (e.g., national parks, wildlife refuges, and
ness areas requires that attention be given to Forest Service wilderness areas) and thus
ensuring that migratoiT pathways are open to enhance the viability ofanimal and plant pop-
organisms migrating seasonally along eleva- ulations (Newmark 1985, Salwasser et al. 1987,
tional gradients. Of particular importance is Noss 1992, Grumbine 1994). By themselves,
the need to maintain winter ranges and migra- BLM wilderness areas in Utah clearly cannot
toryroutes oflarge mammals such as mule deer satisfy the huge area requirements noted above
{Odocoileus hemioniis), elk {Cervus elaphiis), as requisite for maintaining viable populations
and moose {Alces dices). of large carnivores. However, when linked to
Interactions among competitors, and be- other public lands (e.g., Utahs national parks,
tween predators and prey, are integral aspects and wilderness areas in other states), BLM
of natural ecosystems and should be pre- wilderness in Utah can be a key component of
served. For example, in the southwestern strategies forlong-term presei"vation ofbiolog-
deserts of the United States, the direct and
ical diversity.
indirect effects of seed predation on plant Other high-priority areas are those which,
community structure have been documented alone or together with other protected areas,
in long-term experiments manipulating densi- encompass entire watersheds. In addition to
ties ofrodent and ant grani\'ores (Daxidson et affording direct benefits to humans, watershed
al. 19S4, Samson et al. 1992). These effects protection is the most effective means ofcon-
include transformation ol a shrubland into a sei"ving the aquatic and riparian communities
grassland biome (Brown and Heske 1990). that account for a disproportionate fraction of
Special care must be taken to consei^ve popu- both species diversity and endangered and
lations of predators with large area require- threatened species in arid western North
ments, because extinctions of these species America (Miller 1961, Minckley and Deacon
can alter whole communities (e.g., by leading 1968, 1990, Holden et al. 1974, Johnson et al.
to outbreak densities ofprey, which then over- 1977, Cross 1985, Knopf 1985, Moyle and
exploit their plant resources). Some of the
Williams 1990). Moreover, since populations
strongest e\'idence for such "trophic cascades"
of riparian species are usually isolated from
comes from the Greater Yellowstone Ecosys-
similarcommunities in otherdrainage systems,
tem, where intensive browsing by elk has
species losses from these environments are
greatly altered many riparian zones by the re-
not easily remediedb>' natural recolonization.
moval ofwillows (genus Salix), and has elimi- A 3rd priority in selecting wilderness sites
nated aspen seedlings {Popiihis fremiiloides)
recruiting from seeds and rhizomes shortly is land that fomis or helps to complete the pro-
tection of entire elevational gradients, for
after the extensive 1988 fires. Huge contem-
~ example, in isolated mountain ranges of the
porary elk herds, numbering 40,000 individ-
Great Basin. Scant attention paid to consemng
uals in the park, and 20,000 in the northern
herd alone, are likely the result of reductions these gradients in the past is evident in the
restriction of most national parks and wilder-
in the full complement oflarge predators (Kav
1990, Wagner et al. 1995). Gonsidcrable evi'- ness areas in western North America to higher
dence also suggests that deer and elk herds in ecloemvpaatriaotnivseilteys.loDweseilegvnaattiioonnBoLfMwilladnedrsnewsosulidn
Utah average significant!)' larger at present
afford protection to regions ofgreatest species
than during any extended period in the histor-
ical past (Durrant 1950, Julander 1962, Haiper richness for man\' organisms (e.g., mammals,
birds, amphibians, insects, and trees) whose
1986).
diversity generally declines with elevation
throughout much of western North America
Strategies for Selecting
Wilderness Areas (Harris 1984, Ste^'ens 1992).
Landscape-wide Priorities Optimal Design Goals 1
Given the large area requirements ofmany If BLM wilderness areas are to contribute
extinction-prone Utah species, it is important substantialK' to the preser\'ation of biodiver-
to protect large, contiguous land blocks. In sitv in Utah, then site selection must take into
1996] Wilderness Selection for BiODiVERsm- 99
H
account tlie 3 general goals outlined above. ^'-^ wilderness
iIdnetaelrlyc,onBnLecMtewdilcdoerrenezssonleanodfs srhooadulledssforlmanadns f<ij&if%;: F^orest. Soervice W..,il,d.erness
H
when combined with otlier federal wilderness National Park service
areas, national and state parks, and wildlife
refuges (Fig. 2). Special attention should be Public Lands
given to linking roadless lands so as to pre-
clude further fragmentation ofnatural habitat.
Inagmentation, or the transformation of an
unbroken block ofnatural habitat into a num-
ber of smaller patches separated by altered
habitats, reduces population sizes, increases
their isolation, and threatens their long-term
viability. It is one ofthe greatest threats to bio- Buffer Zone
logical dixersit)' worldwide (Wilcox and Mur-
phy 1985, Wilcove et al. 1986, Saunders et al. <
1991). Across diversehabitats, there are numer-
ous examples of species extinctions precipi- 'KWi;;-^-;.—core zone
tated by both natural and human-induced
habitat fragmentation (e.g., Brown 1971, Ter-
borgh and Winter 1980, Diamond 1984, Fig. 2. An example of a preferred arrangement of
Heaney 1984, Patterson 1984, Newmark 1987, wilderness and multiple-use federal and state lands to
1991, 1995, Case and Codv 1988, Soule et al. conserve biological diversity. Wilderness areas adminis-
1ir9v8oA8od,\jnBaocle1gnetr emtLulalt^.i1pi9lne9m-1\us. e' lands can 1bufrfrer t,mN]ea^ito^^ei-,doi,nbiavlfioth^rdeamrBkiaurccbeoeanrutviiocgfeu,LouaasnndcdjoMraFue-insazihgoenamenedniint^W,w^i/hkFiliolurciehflsetthSSceeerrvmvioicscet,e
human impacts on biological diversity within extinction-prone species in Utahcanbeprotected. Multi-
wilderness areas. Such lands can be expected pie-use lands can effectively buffer this core zone and
topro\'ide marginal habitatfortlie manvspecies P'^^'deadditionalmarginalhabitattospecies thatarepri-
t.1liat. are rest1.ri•cited1 prn• nanl-1y t1.o more prisI.t-me marilvrestrictedtoroadlessareas.
wilderness regions. Thus, proposed wilder-
ness areas surrounded bv public lands should
receive high priority forprotection. their genes move about only through the pro-
cesses of seed dispersal and pollen transport.
E.XAMPLES OF Rare and Therefore, it is not surprising that many plants
Endemic Species have narrowly restricted ranges, are locally
The design advocated above is based adapted to conditions within those ranges, and
are isolated, often by great distances, from
largely on conservation strategies for preserv-
other sites where similar conditions prevail.
ing wide-ranging vertebrate species. Although
Although locally endemic plants can often be
such strategies can help to ensure the long-
relatively abundant inside their ranges, their
term viability of most species within a given
region, exclusive reliance on such approaches populations are easily jeopardized by habitat
may well overlook and endanger many locally alteration (e.g., by all-terrain vehicles) within
isolated, rare, and endemic plants and animals. their narrow distributions. Of Utah's approxi-
We cannot give a comprehensive treatment of mately 2600 plant species and 400 named
this subject here, but we discuss 3 ta.xonomic varieties (Albee et al. 1988, Welsh et al. 1993),
groups oforganisms for which especially high about 180 (or 7% ofspecies) are currently clas-
rates of endemism or existing threats to iso- sified by federal or state agencies as endan-
lated populations present particular manage- gered, threatened, or sensitive. A majority of
ment dilemmas that should be taken into these (133, or —74%) definitely or probably
account in wilderness decisions. In most cases, occur on BLM lands (Atwood et al. 1991), and
specific habitats must be protected to assure a substantial subset of the classified species
are narrowendemics.
thepresei'vation ofthese species.
Shultz (1993) provides a useful summary of
Plants ofSpecial Concern endemism in the Utah flora. Approximately
Unlike the wide-ranging animals discussed 240 species, or 10% ofall Utahplantspecies, are
above, plants occupy fixed positions; they and endemic to the state. This rate of endemism.
100 Great Basin Natufl\list [Volume 56
the percentage ofthe flora considered for hst- plant species are endemic to the region in and
ing as threatened or endangered, and the per- around the proposed wilderness area (PWA)
centage ofrare species in the flora are among near the White River south of Vernal (UWC
the highest in the continental United States. 1990), and most of these are confined to the
Thevastmajority(86%) ofUtah endemics reside Parachute and Evacuation Creek members of
in arid and semiarid regions of the state, and the Green River Shale formation. Another
90% are edaphicalK' restricted to fine-textured dozen endemics occur in a diversity ofhabi-
and/orhigh pH substrates (limestone, clay, silt, tats in and around the San Rafael Swell. Here
mudstone, and shale) that magnify drought the most important habitat is a beige (rather
stress. Plant distributions generally appear to than red) Moenkopi formation, spatially iso-
respond more to edaphic, topographic, and lated from other Moenkopi outcrops and un-
geologic features of the environment when usual in its soil chemistiy. A few endemics also
drought is a factor (Stebbins 1952). Because occur on the younger Carmel and Summer-
most endemics live in close proximity to mor- ville formations surrounding the core of the
phologically similar species (Albee et al. 1988), swell, especially between Muddy Creek and
these species appear to be mainly neoendemics Crack Canyon (S. Welsh personal communica-
that have evolved since the last glacial maxi- tion). Wilderness designation in these 2 regions
mum (18,000yi's BP), orin the Bonneville basin (the San Rafael PWA and the White River
PWA UWC
during the past 10,000 >ts. of the Uinta Basin [Fig. 3]; see
Geographically, endemism ofUtah plants is 1990) couldafford significantprotection to some
highest in the Canyonlands Phytogeographic of Utah's endemic plants. South and east of
Section of the Colorado Plateau Division of the San Rafael, in the Dirty Dexil PWA (UWC
the Intermountain Region (Cronquist et al. 1990), are the distinctixe flora of the Orange
1972, Fig. 3 modified from Shultz et al. 1987). Cliffs region (Fig. 3) and some additional nar-
An unusual diversity ofsubstrates occurs here, rowendemics deservingprotection in the Main
andthese substrates are moreapttobeexposed, and South forks of Happv Canvon (Shultz et
rather than coxered with alhnium as in other al. 1987).
areas ofsemiarid Utah (Welsh et al. 1993). Thus, The Moenkopi formation is also important
fully 50% of Utah's 240 rare and endemic as a substrate for endemics elsewhere in semi-
plant species occur on the Colorado Plateau, arid Utah. Two federalK' listed endangered
whereas just 15% occur in the Great Basin, species, Arctomecon limnilis (the dwarf bear-
11% in the Mojave Desert, and 10% in the claw poppy) and Pediocactiis sileri (a cactus),
Uinta Desert (Welsh 1978, Shultz 1993). and several other species are endemic to par-
About halfofUtah's endemics belong tojust 5 ticular Moenkopi outcrops in southwestern
genera that are both common and physiologi- Utah. Wherever possible, the boundaries of
cally adapted to aridity (total Utah species and wilderness areas and other protected areas
percent endemics, in parentheses): Astragalus, should encompass these specialized habitats.
Fabaceae (114, 36.8%), Penstetnon, Scrophulari-
aceae (106, 26.4%), Cnjptantha, Boraginaceae
Bees and Wasps in the
(61, 36.1%), Eriogoniim, Polygonaceae (60,
San Rafael Desert
23.3%), and Erigeron, Asteraceae (54, 24.1%;
Welsh et al. 1975, Welsh 1978, Shultz 1993). Because oftheircapacit) for directed mo\'e-
Because most ofthe state's endemic plants ments, animals are less likely than plants to
are restricted to particular geologic formations, exhibit high rates of endemism. Nexertheless,
and because multiple endemics often occur on since insects often tend to be host- or habitat-
the same formation, groups of endemics gen- specific (e.g., in pollinators, herbixores, or sub-
erally can be protected simultaneously by safe- strate-specific ground nesters), endemism can
guarding those soil formations and surround- often be high in insect taxa. Bees and wasps
ing areas. Two regions where large nimibers of (order H\menoptera) are examples of such
endemics stand to benefit from wilderness insects. Here, as elsewhere, bees and preda-
protection of BLM lands are the Uinta Basin tor)'wasps are especialh'di\'erse in arid regions
and the San Rafael Swell and surrounding San (Michener 1979). The state supports a mini-
Rafael Desert (Fig. 3, Table 1; M. Windham mum of 950 species of native bees (roughly
personal communication). No fewer than 15 25% of the total number of species known
1996] Wilderness Selection eor Biodin i.Ksin 101
SweU
San RafaelDesert
^Orange CUffe
Fig. 3. Satellite imageofUtah showingthe positions ofthe San Rafael Swell, the San Rafael Desert, and tlicOrange
Cliffs, all within the Canyonlands Phytogeographic Section, ontlined in bold. The arrow in the Uinta Basin shows the
approximatepositionoftheWhite RiverPWA(UtahWildernessCoalition 1990).
from America north of Mexico), and 50 ofthe (Neffand Simpson 1993), and manyoftheareas
Utah species are currently inidescribed (T. currently under consideration for wilderness
Griswold, K Parker, and V. Tepedino personal designation in Utah are centers of endemism
communication). Many areas, especially in the forboth groups. Althoughwe lackextensive in-
southern part of the state, have not been formation on bees ofthe Canyonlands Section
explored intensively and undoubtedly harbor (Fig. 3), where endemism is highest for plants
many additional undescribed species. (see above), intensive collecting in that small
Bees and plants often show comparable geo- part known as the San Rafael Desert has
graphic patterns in diversity and endemism yielded a total of 316 species of bees, 42 of
102 Great Basin Natur.\list [Volume 56
Table 1. Plantsendeinictothe2areaswith thehigliesteiuleinisiiion Utah BLM lands.
Endemicsoftliesontheni Uinta Basi Endei the San Haiael Swel
Aqtiilegiabarnebtji Miniz(Ranunciilaceae) AstragalusrafaclciisisJones(Fabaceae)
Asfragdiiiseqiiisolensis N'eese6cWelsh (Fahaceae) CnjpfaiithacrciitzfclcliiWelsh(Boraginaceae)
A. hainiltoniiC. Porter C.Johnstoiui Higgins
A. hitosii.sJones C.joiu'siaiui (Pa\son) Pa\son
A. saiiriniLs Barnehx Erigeroii inaquirciCronquist(Asteraceae)
CirsiiiinhanwhijiJohnst. (^Asteraceae) Loinatiiimjiinceiiiii Banieh\ 6c N. Holmgren(Apiaceae)
CryptcmtlidhantchijiJohnst. (Boraginaceae) LijgoclcsmiaentradaWelsh 6c(iooilrieh(Asteraceae)
CgrdliainiiJohnst. PcdiocactiisdcspaiiiiiWelsh6cC.ot)drieh(Cactaceae)
CtjDioptcri.s(liiclu'siicn.sisJones(Apiaceae) Pcitsteinoit inarnisii(Keck) \. Holmgren(Scropluilariaceae)
Pensteinoiifloiccifiii Neese&\M'lsh ScliOi'iicraiidH'harncbt/i(Welsh6c.\tA\ood) Rollins(Brassicaceae)
(Scrophulariaceae) 'ndinuiii dtninpsoiiiiAtwood6cWelsh(Portiilacaceae)
P.goodhcliii .\. Holmgren Tt>uiisciidi(iapricaWelsh 6c Re\eal(Asteraceae)
P.grahainii Keck
SchoencniinlH'argilhicea(\\'elsh &Atwood)
Rollins(Brassicaceae)
S. .suff'ruti'.sceus(Rollins)Welsh6cChatterK'
Sclerocactiisglaitciis(K. Schnm.) L. Benson
which are presentK undeseribed (T. Griswold, in semiarid Utah assuredh depends on their
E Paikei; and \^ Tepedino personal conniuuii- relationships with bees. Eor example, a rare
cation). Thus, 33% of the state's total species species of Pcrdita, found in Utah only at the
count, and 84% of Utahs undeseribed (but BeeHi\e Dome site southeast of St. George,
catalogued) species, are endemic to a region pollinates the rare and endangered dwarfbear-
comprising just 2.0% ot the states land area. claw popp\' (^. Tepedino personal commimica-
Fin-thermore, a signiticant portion of this tauna tion). Bees that ha\e specialized b\ collecting
(24%) occin\s onI\' on the Colorado Plateau. pollen onl\ from flowers of a particular plant
The remainder of the Cainonlands Ph\togeo- family, or exen from a single genus within a
graphic Section, in which the San Rafael Desert famih; are termed oligoleges. Such bees tend
is embedded, is likeK" to be equalK di\erse to be most common in arid regions (Neft and
and to ha\e as man> new species. Simpson 1993) and generalK" are regarded as
Other hymenopteran groups, such as the being closely adapted to the phenolog>' and
aculeate \\asps, also are highly di\erse in the floral traits of the plants on \\hich the)' spe-
San Rafael Desert (T. Gris\\old, E Parker, and cialize. Such adaptations tend to make them
V Tepedino personal connnimication). For ex- superior pollinators. Scjuash bees and squash
ample, with a total of22 species there, the cir- flowers are examples ofsuch a co-adapted pair
cinnglobal genus Fhildntluia is more di\"erse in in the Americas (Tepedino 1981). Some oligo-
the San Rafael Desert than an\A\'here else in leges ma> one da\' proxe to be useful as crop
North America, andprobably the world. These pollinators. The legume specialist Osniia san-
predatoiy "digger wasps" nest in the soil and rafachic. a nati\e ot the San Rafael Desert, has
ma\ ha\e di\"ersified in response to the \aried been inxestigated as a potential pollinator of
substrates present in this desert. ClearK, des- alfalfa {Mcdicdgo sativa L.), an important for-
ignation ofwilderness in the San Rafael region age crop (Parker 1985, 1986). Man> of the
UWC
(see 1990) could afford significant pro- species ofthe San Rafael Desert appear to be
tection to an area of\er>- high endemism and oligoleges. A brief list of some of the unde-
di\ersity for the order H\nienoptera. seribed and recentK" described bee species
Bees and wasps are among the most benefi- and their host plants is pro\ided in Table 2.
cial insects. Predaton' and parasitic wasps help These entries were chosen only to illustrate
to control populations of pest species (e.g., the \ariety of plant taxa upon which nati\e
grasshoppers, aphids, etc.) below outbreak bees specialize.
densities. An estimated 67% offlowering plants
Nati\e and Endemic Fishes
depend on insects (primariK' bees) for pollen
transfer and sexual reproduction (Axlerod Freshwater ecosxstems are natinal habitat
1960), and the welfare of nian\ plant species "islands ; as sutli. thcii- long-tcnn isolation b\
1996] WiiJ^EKNEss Selection for Biodivkksitv 103
Tahi.K2. Pollenpiffereiicesforreprescnlativfoli.ujolL'ctifliccs in llicSan lialacl Dl'S(m1(dataCioni'I'.Crisvvold, F Farkt-r
andV.Tepedinc)personalconnnunication).
Plantfamily Plantgenus/species Beespecies
Asteraceae Hcliiiiitliiis(inoinolu.s Perdita nr.laticincta*
Hi'.siH'rapissp.*
W'l/ctliidKc<il)ra Perdilaholiartoruin
Boraginaceae Coldenia Perdita(Heteroperditu) sp.*
Slaiilci/a Perditanr.zehrata*
Eiipliorbiaceae Euphorbiaparnji Perditani'.laln'rgei*
Fabaceae Astrci<!,alus Ashineadii'lla nr.mieheneri*
Loasaceae Meiitzeliaiutillifli)ni Perditauudtiflorae
Onagraceae CcDiiissonia Diijoureasp.*
Papaveraceae Argcinonc Perditaute
Polenioniaeeae Cilia Perditanr.giliae*
Perditaeloiigaticeps
Scn)})liu!ariaceae Pcmtonon Anthocopasp.*
*UiuU'stiil>cilspecie
intei'vening terrestrial liabitats, orby unsuitable Bonneville cutthroat ti'out {Oncorhtjnclms clarki
aquatie habitats, often promotes loeal speeial- iitali), once thought to be extinct (Behnke
ization, evoliitionaiydiversification, andendem- 1992), survives in populations in Trout Creek
ism in aquatic organisms. Seven centers of and Birch Creek within the Deep Creek PWA
endemism are recognized for fishes ofwestern (UWC 1990).
North America (Miller 1959), and Utah includes Where protection of whole watersheds is
substantial portions of 2 of these centers, the notpossible, wilderness that includes keyhabi-
Bonneville Basin and the Colorado River tats may help to stabilize decliningpopulations
Basin. Collectively, 28 fish species are native ofnative fishes, preclude newlistings and draft-
to thesebasins (Smith 1978), and 27 are extant. ings ofrecovery plans, and promote recoveries
Because of their limited distributions, en- and delistings. This should be the case most
demic species are easily endangered by both often for fishes living in headwater streams
habitat alterations and introductions ofnonna- protected by natural and artificial downstream
tive competitors and predators. Seven species barriers from unintended invasions of alien
and subspecies from the Bonneville and Col- cold-water species. For example, habitat in the
orado basins are now federally listed as endan- upper Book Cliffs-Desolation Canyon PWA
gered (U.S. Fish and Wildlife Ser-vice 1993). A may support the Colorado River cutthroat trout
further 11 species and subspecies are consid- {Oncorhyncluis clarki plenriticu.s), considered
ered by fisheiy specialists to be endangered, the rarest of the cutthroat taxa (Behnke and
threatened, or ofspecial concern in Utah (War- Zani 1976) and federally listed as a categon^ 2
ren and Burr 1994). The decline of native species (Kerchner 1995). Although the region
fishes has been associated with both water- has not been surveyed for this subspecies,
shed development (e.g., reservoirs, irrigation native populations occur in streams entering
diversions, channelization, floodplain drainage) the Duschesne Riverfrom the north (Shiozawa
and the introduction ofalien species. and Evans 1994) and have recentlybeen found
Conservation of endemic fish populations in streams ofthe western Book Cliffs, closer to
has been especially successful when much of Price and Soldier Summit (Shiozawaand Evans
the watershed has been protected (Williams unpublished data). Given these obsei-vations, it
1991), but adherence to strict legal definitions is likely that streams flowing into the Book
of wilderness often precludes such wide- Cliffs-Desolation Canyon PWA will also con-
spread protection. In Utah, opportunities for tain this subspecies.
protecting entire watersheds are limited to In relatively large downstream systems
relatively small drainage systems extending (secondary and tertiaiy streams), key habitats
from stream headwaters in mountain ranges of include floodplain wetlands, among the first
the Bonneville Basin to diy or saline lake beds habitats to be lost due to human activities.
at lower elevations. A particularly important Although wetlands have been viewed tradi-
case is in the Deep Creek Range, where the tionally either as breeding sources for insect
104 Great Basin Naturalist [Volume 56
pests or as waterfowl production sites, periodic argentissimus), and the Virgin River chub {Gila
or continuous connection to rivers renders robusta seminuda) are endemic to this system.
them important appendages to lotic systems. Twoadditional species, the flannelmouth sucker
Densities of aquatic invertebrates are signifi- and the desert sucker {Catostomus clarki),
cantly higher in wetlands than in main river have evolved very slender caudal peduncles,
channels, over 100-fold in some cases (Wolz possibly as a response to occasional high flows
and Shiozawa 1995, Mabey and Shiozawa in the Virgin River (Smith 1966).
unpublished data). Floodplain wetlands can The health ofthis unique fish fauna already
therefore serve as important nurseiy grounds is cause for concern. Two ofthe endemics, the
forlaival and immature native fishes. woundfin and the Virgin Riverchub, are feder-
The loss of wetlands may be a significant ally listed as endangered. Although the desert
factor endangering sexeral native fishes in the sucker occurs in Arizona, Nevada, and New
Colorado River (Tyus and Karp 1989). Fishes Mexico, this species merits special concern in
native to the larger streams and rivers of the Utah (Utah Division of Wildlife Resources
Colorado River Basin are predominantly min- [UDWR] 1992), where it is limited to the Virgin
nows (Cyprinidae) and suckers (Catostomidae) Ri\'er drainage. Loss of either this species or
that have evolved in isolation, are adapted to the flannelmouth sucker from the Virgin River
unique local conditions of this drainage (e.g., system would eliminate only a subset oftheir
heav\' silt loads and wide fluctuations in dis- existing populations and is unlikely to move
charge and temperature), and are the most either species to endangered status. However,
moiphologicalK' distinct fishes in NorthAmer- the uniqueness of these populations (Smith
ica (Hubbs 1940, 1941, Deacon and Minckley 1966) may warrant their designation as sepa-
1974, Minckley et al. 1986). Four of these rate subspecies. This, togeflierwifli the concern
native species, the Colorado squawfish {Pty- now e\'idenced for the flannelmouth sucker
chocheilus lucius), the humpback chub {Gila throughout its range, could easily translate into
cijpha), the bonytail chub {Gila elegans), and candidacy forlisting ifexistingpopulations are
the razorback sucker {Xyrauchen texamis), are notprotected.
now federally listed as endangered. The decline Concern for native fishes ofthe Virgin River
ofboth the bluehead sucker {Catostomus [Pan- drainage has already constrained water devel-
tosfeii.s]discobolus) andthe flannelmouth sucker opmentillWashington Count); Utali. An>'actions
{Catostomus latipinnis) within the main stems thatwould help presene the integrit}' ofripar-
of the Colorado and Green rivers may result in ian habitat and stream channels would also
their listings as threatened, especially ifpopu- reduce stress for these fishes. Since the integ-
lations in tributaiy streams are not stabilized. rit\' ofriparian habitats is best maintained over
Several of these species occur in areas under large areas, wilderness designation in PWAs of
consideration for wilderness status. Both the the Beaver Dam slope and the greater Zion
Price River, in the Book Cliffs-Desolation Can- areawould sei"ve this purpose.
yon PWA, and the San Rafael River, in the San Finally, protection of Utahs rare and en-
Rafael PWA, have populations of roundtail dangered fishes would likeK also afford signif-
chub, flannelmoudi sucker, andbluehead sucker icant protection to other aquatic organisms,
Bluehead sucker are also known from the for example, Utah's diverse communities of
Dirty Devil and Muddy Creek drainages (Smith aquatic insects. Reciprocally, the maintenance
1966), andboth flannelmouth suckerand round- of high species diversity in stream insect com-
tail chub are likely to occur there. Wilderness munities is critical to assuring a continuous
designation couldbroaden theprotectedranges food supply to fishes in rivers with wide sea-
of several of these species by stabilizing wet- sonal and annual fluctuations in flow rates.
land habitats in the Dirty Devil, San Rafael, Mayflies (Ephemeroptera) are among the best-
and Book Cliffs-Desolation Canyon PWAs. studied stream insects in Utah, and 16-18
Although the Virgin River drainage is also genera (22-24 species) are known from warm
part of the Colorado River Basin, it has a water tributaries ofthe Colorado Rixer sxsteni
unique fish fauna that appears to have evoKed (G. Edmunds personal commimication). Con-
in isolation from populations in other parts of struction of reservoirs on these rivers has
the basin. The Virgin River spinedace {Lepi- iilreadyinundated many river miles and altered
domeda mollispinus). the woundfin {Plagoptenis flow rates, sediment loads, and downstream