Crosswalking National

Wetland Inventory attributes

to hydrogeomorphic functions

and vegetation communities:

a pilot study in the Gallatin

Valley, Montana

Prepared For:

The Montana Department of Environmental Quality

and
The United Slates Environmental Protection Agency

By:
Linda Vance, Gregory M. Kudray and Stephen V. Cooper

Montana Natural Heritage Program

Natural Resource Information System

Montana Slate Library

MONTANA

Natural Heritage
Program

Crosswalking National Wetland Inventory

attributes to hydrogeomorphic functions and

vegetation communities: a pilot study in the

Gallatin Valley, Montana

Prepared For:

The Muni ana Department of Environmental Quality

and
The United Slutcs Environmental Protection Agency

Agreement Number
205003

By:

Linda Vance, Gregory M. Kudrny and Stephen V* Cooper

I OMA N A

Natural Heritage
Program

w T 5ltate *V Natural Resource

y Lihrar

irary V VDr InfL,rmaliun System

© 2006 Montana Natural Heritage Program

P.O. Box 20 1 800 • 15 15 East Sixth Avenue • Hclciu, MT 59620-1800 • 406-444-5354

This document should be cited as follows:

Vance. Linda. Kudray. Gregory M. and Stephen V. Cooper 2006. Crosswalking National
Wetland Inventory attributes to hydrogcomorphic functions and vegetation communities: a pilot
study in the Gallatin Valley. Montana. Kcpm to The Montana Department of Environmental
Quality and The United Slates Environmental Prelection Agency* Montana Natural Heritage
Program. Helena. Montana 37 pp. plus appendices.

Executive Sum m a r y

The Wcllund anil Riparian Mapping Center of the
Montana Natural Heritage Program has recently
begun new National Wetland Inventory (NWIj
mapping fur Montana. The present study was a
pilot project lo investigate ways to enhance (he
new mapping through additional classification? and
modifiers. Specifically, our goals were to:

• Evaluate CIS-based approaches
lo Hydrogcomorphic (HGM)
classification and profiling:

• Conduct an HCM functional
assessment of the Gallatin Valley study
area based on these classification and
profiles:

• Establish a crosswalk between NWI
classifications, HCM classification,
and vegetation associations:

• Determine whether these additional
classifications and associations could
be efficiently linked with the National
Wetland Inventory in future mapping.

Using existing NWI and gcospatial data for
soils, elevation, and hydrology, we developed a
dichotomous key that could assign HGM classes
to approximately 854 of the wetland polygon* in
the study area with 904 accuracy, as determined
by visual irapectionof the wetlands on t-meler
color imngeiy. The remaining wclland polygons
were classified manually. We used a total of 32
HGM classifications. 17 for unaltered wetlands
and 15 for altered wctlanoS. Nearly 834 of the
wetland acreage in the study area is classified as
li.. Lotic wetland acreage is concentrated in
the Gallatin River-Gallalin Gateway (474) and
Lower East Gallatin <254 ) watersheds, which
contain the main Gallatin River and East Gallatin
River channels. Smaller proportions are found
in the Smith Creek (10^) and Hyalite Circle
(94 i witcishcdv Terrene wetland acreage Ls
most plentiful in the Lower Easi Gallatin (414)
and Smith Creek (27%) watersheds. In the study
area as a whole. NWI mapping from Ihe l9SOi
shows less than 84 of the total wetland acreage
as being direclly altered. However when wetland
numbers are considered, almost 1 7^ of wetlands
wcte altered in some way. Although 854 of all the

wetland acreage in Ihe study area is Lotic. 724 of
Ihe altered wetland acreage is Terrene.

We used Principal Components Analysts to
separule the 5th code HUCs into impact categories,
evaluating road density, land cover, population
density.scplic density, average parcel si/e.
percentage of private land owneiship. percentage
of public ownership and casemenls. and percentage
of public land survey sections wilh one or more
noxious weeds. The watersheds with the highest
impact ranking were the Lower East Gallatin.
Hyalite Creek, and the Upper East Gallatin.
Together, ihcsc walcishcoS hold 87*1- of Teircnc
Inlerfluvc Basin wcllanoV 714 of Terrene
Inlerlluve flat wellanoS. and 334 of Terrene Basin
wetlands. Because these types of wetlands have
high function ratings for nutricnl cycling, sediment
trapping, aquatic and terrestrial habilal. and
wetland biodiversity, we consider those functions
lo be ai risk both within those watershed and
across the study urea as a whole. Similarly, those
three watershcoS contain almost 404 of Lotic
River Lower Gradient Floodplain wellanos. 364 of
Lolic River Lower Gradient Fringe Wetlands, and
39"]+ of Lotic Stream Lower Gradient Hoodplatn
wetlands. Because these weilund types exist
in other, somewhat less impacted watcisheck
(notably the Gallatin River-Gall atin Gateway and
LowerCallatin). their specific functions (surface
ivatcr detention, stiram maintenance, and plant
community maintenance! are at least moderately
at risk in the study area. Furthermore, because
Ihcsc Lotic types, like the Terrene types, also have
high function rating for sediment trapping and
aqualic/teirestrio) habitat. Ihosc functions may be
especially compromised in these three watershcoS.

Prom our lield-hased surveys and Ihe GIS, we
linked Nalional Vegetation Classification .System
vegetation associations, Montana riparian types,
and HGM-derivcd ecological functions to NWI
types in the Gallatin Valley, so tacts could relate
NWf c 1 ass i lie a lions lo detailed information tfieful
for planning and manugement. These associations
air compter A specific NWI type typically
encompasses several vegetation associations, but

(he relationship is usually understandable if past CIS approach to adding HGM attributes K> N\V]

disturbances and regional environmental conditions types, we have ml evaluated ii beyond the study

air considered- NWI associations with HGM area, and therefore cannoi say whether it offers a

modifiers and ecological functions ire similarly cost and lime savings over individually clarify in

complex with a one-lo-many relationship common, each polygon by hand. However, we recommend

linking functions to NWI types continues in

We recommend that mapping in new areas some form Wetlands are valued (and rcgulatcdi

continues to associate vegetation types with because of their function and associated values;

USFW5 types, since the comprehensive, readily connecting mapping with functions will aid

available information about these vegetation types wetland mitigation, restoration, conservation, and

will help those seeking to better undetstand or management,
manage wetlands. Despite our success with a

D

iv

Acknowledgements

Tins project would not have been possible
without the dynamic leadership of Lyndi Saul.
MT Department of Environmental Quality, in
promoting wetland science and conservation. We
also acknowledge Montana's Wetland Council
in ireogni/Jng the need for a new National
Wctlincbi Invcntoiy in Montana. Tammy Crone
of ihcCallutin Valley Local Water Quility
District wis very helpful in contacting private

landowner for access. Erin Farm, an inlern at the
MT Department of Environmental Quality, also
assisted in landowner cont act and other project
support. Cob urn Currier, or the Montana Natural
Heritage Program, took on the task of fomtatiing
and publishing the report. We extend our thanks
to all these individuals, hut remind the reader any
errors or omissions in the report arc entirely the
responsibility of the authors.

Table of Contents

Introduction and Background I

Classification and Assessment I

Crosswalk between NWI Classifications. HGM Classifications, and Vegetation

Associations 2

The Study Area 2

Methods 7

HGM- based Classification and Assessment 7

National Wetlands Inventory Enhancement 1 1

Results 12

HGM Classification of Wetlands 12

HGM Assessment of Wetlands 12

Evaluation of the Classification and Assessment Methodology 20

Crosswalking between NWI, HGM, and NVCS 2 I

Guide to National Wetlands Invcntofy Types and Functions 22

Conclusions and Recommendations 35

Literature Cited 36

Appendix A* Die ho to mo us key for CIS-based Assignment of HGM Classifications to NWI

Wetlands in the Gallatin Valley
Appendix B, NWI and HGM Codes and Descriptions Used in This Report

Lint or Figlies

Figure I. Location of study area 3

Figure 2. Mapped wetlands, Gallatin Valley 5

Figure 3* Public and private owncrship^Callatin Valley 6

Figure 4* Geology of study area IS

Figure 5* Elevation of study area 16

Figured. Hydrology of study area 17

List or Tables

Table I. GISdata sources 7

Table 2. HGM terms and descriptions used in this study S

Table 3. Unaltered wetland acreage in the study area, by HGM class 13

Table 4. Allcrcd wetland acreage in the study area, by HGM class 14

Table 5, Impacts by 5th-codc HUCs in study area 19

Table 6. Correspondence Tabic, NWI loHGM 23

Table 7. Typical HGM types associated with PAB/PUB wetlands 25

Table 8. Functions associated with PAB/PUB wetlands 25

Table 9, Typical HGM types associated with PEM wetlands 27

Table 10, Func lions associated with PEM wetlands 27

Table 1 1, Typical HGM types associated with PSS wetlands 30

vi

List or Tiiles (cuntimed)

Tabic 12. Functions associated with PSS wetlands 30

Tabic 13. Typical HGM types associated with PFO wcllands 32

Tabic 14, Func lions associated with PFO wcllands 32

Tabic 15. Typical HGM types associated with R2US wcllands 34

Tiblc 16. FunLliuiu associated with R2US wetlands 34

vu

Introduction and Background

Wetlands provide numcroiE habitat and
cconDtnic benefits, und arc key factors in flood
reduction, groundwater recharge, and biodivciNity
maintenance (Brinsonct al. 1981. Keddy 2D00).
The scale and importance of wetland values
arc disproportionate Id the physical extent of
wetlands on the landscape, especially in a semi-
arid western slate such as Montana (Finch and
Ruggiero 1993. Fallen 1998). Yet despite their
importance to humans and wildlife, a significant
portion oT Monlana*s wetlands have been lnsi
or degraded (Dahl 1990, Scott ct al. 2003).
Furthermore, assessment of wetland extent am
condition has been impeded by the lack oF wetland
maps. Although the pholo interpretation has been
completed for Ihe National Wetlands Inventory
|NW1) in Montana, digital maps have not been
produced for much of the slate. Because the NW1
is now over 3) ycais old, Montana's Wetland
Council has encouraged new h NW] mapping rulher
than simply digitizing the old maps. This mapping
has recently been undertaken by the Wetland and
Riparian Mapping Centerof ihe Montana Natural
Heritage Program.

The present study was a pilot project to
investigate ways to enhance the new mapping
ilin tugh additional classification and modifiers.
Specifically, our goals were to:

* Evaluate G 1 5- based approaches io
Mydrogcomoiphic (HGM) classification
and profiling:

* Conduct an HGM functional assessment
of the Gallatin Valley study area based on
ihcse classifications and profiles:

* Establish a crosswalk between \W|
classifications. HGM classifications, and
vegetation associations;

* Determine whether these additional
classifications and associations could be
efficiently linked with National Wetland
Inventory in fulure mapping.

Classification and Assessment

TheNWI classification system (Cowaalinct al.
1979) is hierarchical. with an increasing level

of detail There arc five broad Systems, three of
which exist in the interior United States: Pahtftrinc.
Lacustrine, and Riverine. Systems are further
divided into Subsystems* for example. Lacustrine
Systems may be Limnetic or Littoral: Riverine
Syslenn may be Upper Perennial. Lower Perennial.
Intermittent, etc. No Subsystemsoccurinthe
Pulustrinc System. Systems and Subsystems
contain Classes describing the substrate, or. in the
case of Palifttrine wellanuV the vegetation life form
(e g emergent. scrub-shrub, forested, etc), and
hydrologic modifieis describe Ihe water regime.

Most Montana wcllanus arc PaliRlhne. This
System generally includes any wetlands not within
a stream or river channel (Riverine System) or
within bodies of water which arc either larger
than 20 acre$ORtfccpcrthan6.6 feet (Lacustrine
System). Paluslrinc systems can be found in a
broad range of landscape positions and across a
variety of ecological classifications, encompassing
such diverse wetland types as mountain fens and
floodplain oxbows. Consequently, it is difficult
lo determine Ihe specific habitat orecological
functions and bene Ills of a given Paluslrinc wetland
simply from its Class.

The Hydrogconxjrphic (HGM) approach
(Brinson cl al. 1993) is also a nationwide wetland
classification system, but emphasizes wetland
function, recognizing some wetlands perform
certain hydrologic. habitat, or biochemical
functions more effectively than others. Our
objective was to determine if we could develop
a GIS-biscd method to link HGM types lo NW]
types in the study area, and to assess whether this
could be done in other watersheds or on a lurger
scale. We hoped ageospatiu) HGM classification
key would provide a cost-eflcclive. streamlined
methodology for associating wetland function*
with NWI lypcs. so evaluations of function and
condition could be accomplished w ith a GIS. We
recognize, of couiscGlS-bascd assessment* are
noi asuhslilute for full HGM evaluation, which
are based on regional reference conditions and
require substantia) field investigations. However,
recent studies in M an land (Tincr ct al. 2000*

aixl Coloraub (Johnson 2005) hive illustrated the
value o\ this methodology for initial, co use-scale
charactcri/aiion and assessment.

We were especially interested in the appmach
described by Johnson (2005) is Hydrogeomorphic
Wetland Profiling (HGM WP). In non-impocted
sellings wetland abundance and diversity of
wetland types ire Jinked 1o such lantfccapc-lcvcl
faclois as geo morphology, basin hydrology,
and regionil or local climate (Johnson 2005).
Consequently, similar lamlscipcs can be expected
lo have similar wetland "profiles.*" When
landscapes are altered by humin activity, it follows
that these profiles, loo. will change. The goal of
HGM WPis lo determine if differences in wetland
abundance and diversity across a region are
attributable Id differences in the physical setting
or rather to disproportionate impacts caused by
human activities. Johnson (2005) was able to use
HGM WP with some success in Summit County.
Colorado. Although we recognized the extreme
physical gradients and concentrated human
impacts characteristic of Summit County differed
significantly 1mm gradient and impacts in our
relatively homogcneoiA study area, we wanted to
lest out the appmach lo determine its utility for
future use.

Crosswalk he/ween NWI
Classifications, HGM
Classification s y and Vegetation
Associations

While NWI and HGM classifications aie
information rich, we believed there could be
significant added value in our ongoing mapping
efforts if even mote classification systems could

be associated with the mapped NWI types. The
National Vegetation Classification System (NVC5.
Gmssmanclal. 1998) has described specific
vegetation associations occurring in Montana and
nationwide. Hansen el oL (1995) also developed
descriptions of riparian types in Montana, which
ore widely toed. Both NVCS and Hansen et al.
(1995) descriptions include detailed accounts of
associations and their components . and provide
interpretative information on ecological and

m an age menl characteristics. If NWI and HGM
classification could be linked with NVCS andfar
Hansen ct al. (1995) associations, either across
the boaid or in specific environmental contexts,
wetland maps would be greatly enhanced.
By attempting a cmssw alk in this relatively
homcogenous study area, we hoped to gauge
whether this would be an efficient way lo enhance
NWI maps.

The Study Area

The present wxirkis an extension of ecological
surveys and evaluations of wetland and riparian
habitats, carried out by a group of univeixily, state,
local and MTNHP researches in the 4IG\000-
acrc lower basin of the Gallotin River waleishcd.
commonly referred to as the Gallatin Valley {Figure
I). Il isessentiolly Ihcsamesludy area as was
used by the Gallatin Local WaterQualily District
(2004) in their assessment of wetland and riparian
resources in the valley, mums the Bo/eman Creek
wateisbed. The study urea includes all or part of
nine 5lh-codc USOS Hydrologic Units (HUCs):
BridgerCreeL Camp Creek- Dry Creek. Gallatin
River-GollatinGatcwoy. Hyalite Creek. Lower
East Gallatin. Lower Gallatin, Smith Creek, and
Upper East Gallatin. The study area boundary was
initially drawn by thcGallalin Local WaterQualily
District lo encompass the laige concentration of
wellands in the valley. We continued lo use it for
consistency, and became this is the only pan of the
lower Gallatin wateisbed for which NWI digital
maps ore available. However. NWI mapping is not
complete even within these boundaries. We note
loo the study area does not include most o[ the
uplands surrounding the valley, where additional
(unmappedj wetlands may exist.

The sludy area is within the Gallatin River
subbasin. ilself pari ol the Upper Missouri Kixcr
Basin, The subbasin covets almost 1.2 million
acres, extending from the Yellowstone Plateau to
ihe Gallalin Valley (Figure I >. Upper portions of
the subbasin receive over 40 inches of rain per
year. The Gallatin Valley, however, is sub-arid,
with a typical precipitation average of 1 2 inches
per year (Western Regional Climate Center 2006).

Fifitu* I. Ijtcatitirttfxlttdvatra*

Two main river systems dominate Galium Valley
hydrogcomorphology. The Wcsl Gallatin River
(typically refciTed tn simply as the Gallatin
Riven Hows for over 100 miles from its origin in
Ycllowstune Nalional Park to its confluence with
Ihe M adison and Jefferson Rivers at Ihe M issouri
HcadwalenL The East Gallatin River originates in
Ihe Bridgcr Mountain** east of Ihe study area, and
joins Ihe main Gallatin River 4 J miles later near
Manhattan. Several perennial streams flow from
Ihe mountains south and east of the Valley to join
Ihe East Gallatin and Gallatin Rivers. These rivers
and peirnnial streams have left extensive alluvial
deposits in Ihe Valley, and as a result, Ihe Valley
overlies a thick aquifer system, as much as 600 feet
deep at Belgrade (Bridger Outdoor Science School,
no dale). In many pans cif the Valley, deplh to
groundwater is \c*& than 3 feeL

Abundant water and fertile soils support agriculture
throughout Ihe Valley, but especially in its northern
and eastern range, where wheat, barley, alfalfa and
hay fickfc arc common The high gmundwalcr
luble also suppons an extensive complex of
wetland vegetation in swales and ditches, along
riven* and streams, and in low-lying depressions
(Figure 2). Wetland vegetation ranges fmm the
eo Hon woods, ofcgwood, chnkcchcrry, willows

and alder in seasonally Hooded riparian forests lo
Ihe recifc, sedges, and rushes found inswalcs. En
Ihe study area as a whole, the National Weilands
Inventory has identified nearly 5,830 acres of
wetlands.' The northern and northeastern portions
of the study area (Lower East Gallatin and Lower
Gallatin watersheds), which are Ihe general
discharge areas for the aquifer, have the highest
concentration of non-riparian wetlands. The
watersheds associated with the Gallatin and East
Gallatin River (Gallatin River-Gall at in Galeway,
Hyalite Creek and Lower East Gallatin) have
extensive tloodplain and riparian welland systems.
The southeast, nonhcast, and western portions of
Ihe study area (Bridgcr Cieck, Camp Creek. Upper
Easi Gallatin) have relatively few wetland features.

The Gallatin Valley is one of Montana** fastest-
growing areas, having gained 34*i in population
between I W0 and 2000 (Census 2000). Gallatin
County estimates an addition 21 f i population
increase by 2010. Most of Ihe Valley is privately

owned, while uplands have a high concentration of
public ownership (Figure 3\. With farmland rapidly
being converted to residential land, and increased
pressure on water resources from both extraction
and discharge, wetland and riparian resources are at
considerable risL

TheNutionul Wetbndslavestory did not map all njurian forctfs as wet bftds. excluding those ih;i are ooiai least
srasotally flooded. Asa result* this acreage does noi adequately reflect ike cxiensive rijxihan areas uloag the Callu-
i in and Has Galium Rivets aad ibe LinierHirums

Gallatin Valley Wetlands
| We lands
I Quads wilh we* and mapping

2.3

ID

Mil-v-

Figure 2. \I*/fq*rU nrfhuuix* Gutiittht VW/rt

Gallatin Valley ownership

Private Land

Montana Slate Tm&l Lands

Montana Rsh. Wildife. and Parks

Montana Univetdfy Systa*

City Govetnmeni

US Pored Service

US Dcplot Agnntftaa

2 J

10

Miles

Fifflv* X Puttie and {Hywie rnrnrM/V/i GulUilttt Vaifcv

Methods

HGM~based Classification and
Assessment

Clarification

Wc used ArcCIS 9.1 fESRI 3005) ind several
diu sources iTabtc t \ to create a Geographic
InlbniMlton System (GI5) with Uye& representing
elevation, soils, stream?*, wetlinoX land cover,
parcel ownership, conservation holdings, mads,
and potential water impacts in the study aira. We

Tabic /. GIS iAi/jj sunn ?s~

underlaid true color 1-melcrNatinnal Agricultural
Imagery Program (NAIP| aerial photos from 2005
oniheGIS (o allow verification of features, since
land caver and wetland layers weir marc than ten
yearn old. Using dichotomous keys developed
byTiner(2D03al and Johnson (2005). wc mode
several passes over the data to assign HGM
classes to wetland identified by the NWI. basing
assignments cm Utufccapc position, relationship to
stream and river features, and soils, and visually
inspecting the outcome on the NAIP imagery. The

L*y*f Nam«

Oota Source

Oaficripvon

Mapping scila

FJaliaul fir laiito frjerrarf

U Ei Flth.TrrJ VMtltl? Gtr.to:

fJ.llnilJ . I.J ' '■ 1 r = : * .

Dtadal dati taifd un 1 .1*4000 arnrt phakis

torn i«8on

1J4.000

Honiara &ti -cod* wafcitfwd&

ri.urr* nr: ourcr Coraerraltin

Servie*

u S Geological Survey BaGr«iops. IMG

l K-i I'll

rtairaal land Coirr Gatnri

U S Genlofycal Surety B[io£r*-i
H- i:;irrr DftMlon

lOn ptirl lancl&al tnaot'f lUOi?

1*0 1H»

Goil Surveys

r^UJHnnlKf Cnlisrri-alivi
SrfvlCl

SSURGO and MiN-s G

124 DUO

! t 1 i ( H . ■ 1 ■ ■ j: .ij ' : 1 . 1 ■' ' 1

US GnUopcal Surety and EPA

Prtnisinnai non-re mluton Da Ainr. ?co*i

1 »4 DIM

I atdu»r

r.tmhra Ualpai Heritage Pro jam

upratd rasrinrniand adruntainn

ZOQ6

1 »4 DIM

Catiratal Databatt

r.fnnbru D* pafttteniol

Cartattal retard* 2006

f. 1*

Mmlara F4onds

us crttMo aurrau jooo

US Grdoaftai Survey i ma ooamapt

1*100 000

CenjtUI Greupft 3000

U.S. CrmiB Burau 3000

trr^yrar terms data wtiictraus each

f;;-ii

Walei Rwnh Pain* ol

Mantna Drpaitne nl ol tjatrai
Reiuivtci iTwl Cansvrvntiiri Wdrl

Rnuiktc DiJiman 200*

r-ditiakd tian feD>J Eand drtcnpiuni pi

VOuWd uucriinalet

vaitaPte

Grotfidwalri Welt

Crnter

Etimated »o«i nei dtriltiu too*, wafer

ntfib niirxjt Ik Id ftUfveyi

1J4.000

SepicDtntily

U.S. Crtaut Burau 3000

US Gralootcal Surrey Map* andiurtty

data

MOO 000

n; " 4 -i : * s 11 1. *

Srrvu*

HrpofU torn Araty Corpt ol Entprvrii

1J4O00

wastewater PrfirvH

Man Ira DEO. 3000

t'lalfd lainDeo records

1^4 QIW

keys were nrfinxrcl until iheCIS-bascxicJissiiicalion
and visual inspection produced agreement in 9i :
of 100 randomly selected wcllanus (Appendix
A)* Approximately J5ft ofthc wctlimfc in the
study area could not be classified using keys: Ihcsr
weie classified manually. HGM attributes were
then coded onto the NWI allhbuie tables, and
statistical summaries were carried oul 1o evaluate
the distribution of HGM types across the study area
watcisbeds.

HCM terminology followed categories described
by Tiner (2DQ3al in his work with NWI

enhancements in Mainland. This terminology
differs fium other HGM classifications in its use
of such tcims as "Tenrnc*" and "Lotic** instead
of "Dcprcssional" or "Riverine," and by iDi
extension or the lenn "Flat** K» mean any wetland
nol having adisiinct basin form and is located on
relatively level ground. The decision to use Tiner's
terminology was based on a desire to maintain
consistency of NWI products and enhancements.
A description of each oT the terms can be found in
Table 2.

Table 2. N(jftf lenttx tmt tSrM'tiftiitint turd in //aV MiinJy-

HGU LANDSCAPE POSiTIOH

Definition

Lntfe
Lolio River
Lotto Stream
Terrene

Associated witi a lake . reservor or ofaer large standing wafer body.

There were no len&e we lands in t\e siidy area

A&oacioFMj will a IbAing wafer system depfcled as a 2*1™ feature on
a 1^4 ooo U5GS topomap.

Associated witi a Ibwing waer p/stem depicted as a ssKjhHine Fea*pe
on a 1 34 OQO U5GS lopo map

Sunaunded ur afrnosl ajnui»*ded by uplaniH These ATrlanda may

have inlet or tmfel channels, bul do no I ha^e a channel entemg and
eufno. hdides hiroan^nade ponds*

HGU GRADIENT (with Loie londscope position)

Definition

Love r Gradient

Middle Gradient

The do wet -lb vi ing ara ol a rivet ot steam ohaiocterced b/ meandets
and IDodpiain de^-elopmenl. Corresponds to NWI Lo**r Perennial*

Faster* no wing area of a livet or steam <*iti limiled Ibodpluwi
development Conesponds to NWI Upper Perennial.

HGU LAHOFORM

Definition

fiosb

Fbl

Flood plain

Fringe
blond

A depresbonaJ tandform octurnricj in upland mis or on foodpftains (e ^j

Oibow). NWI Aquatic bed sl**s aie aliva|-i dnssified as Badn Can be
pari of Terrene or Lotc landscape position.

A level Ian J or in usually but no I always on a feodplain or inter fcn*e

Typical^ only ideniftted Itom .Imal inspection.

Brnad Hal landforrn rJiaped bf ri'ivrine or oher AivlaJ processes h fhi=

case t*e area del bed b j tie presence ol soils listed as Ibodplawi soils
in ISeNftCSSSURGO database. Can be Further described as Basin.

Rat Or 0>bow Avough qumbinaliun al G G and ,i- ji.i inspecton

Wetland on tie moigin of a steam river, pond, ui lake, hclutfes non*
*rcjetated or sparsely vege toted point bora along a river's edoe.

Alancform ttaiis completely surrounded by a sfream ri»er pond, si
lake

LiMfl CVtnfj

■in

HGM WATER PATM

Deiailba

Inflow

bo toted
Tbroughflow

A Ireland receiving {pound in surface Aaler but **ti no significant
d^thaige. Typicall/ only identfed tn^ugti visual vispecfon-

A lenene we land witi no apparent surface ivaier oilbvi or nuill™

A *eland wtti suriace of orounoVal-i finding tvough H lo a
naw body or another we land. ThrounhAow may be epher»eraJ or
internrilienl

Assessment

Johnson (20051 used geology, hydrology, climate
and elevation to assign Summit County 6th code
HUCs inlo three relatively homogeneous process
domains isettxu Montgomeiy 1999), and then
assigned them into impact categories ("reference"
and "impacted") hascd on land cover and mad
density. He was then able to determine whether
characteristic wetland profiles differed between the
Ihrcc process domains (yes) and between impacted
and reference examples ul individual process
domains (ycsi. Our initial intent was to follow
hts methodology in our study area, but when we
examined our G1S. we determined the mapped
wetlands were almnst entirely confined to a single
geologic layer (alluvium), and a narrow elevation
band 0250m to 1575m) within a 12-14 inch
precipitation zone. Principal components analysis
<MVSP20D3)of the CIS led to thedelenninatkm
that 5th code HUCs would be the best units of
analysis, since they had suflicicnt homogeneity*
on the one hand and enough impact variability
on the other to make comparisons. Furthermore*
5th code HUC boundaries an: well-defined, and
areuscdinNRISsGIS Data Bundler and Digital
Atlas applications, allowing anyone who wishes to
examine lurthcr attributes of the landscape units in
ilits study to tk> so.

We also used principal components analysis to
separate the 5th code HUCs inlo impact categoric*.
We added several impacts lo the ones used by
Johnson |2005|. evaluating mad density, land
cover, population density, septic density, average
parcel size, percentage of private land ownership,
percentage of public ownership and easements.
Because field surveys revealed extensive weed
infestation in the studv area, we also ihlIixLM

percentage of public land survey sections with one
or more noxious wcco>* us on impact. Based on
these impact factor*. 5th code HUCs were ranked
from "Less Impacted" to "More Impacted." (We
did not feel the term "reference" was appropriate
in the study area, even in a relative context.) We
examined a suite of additional factors lo determine
whether they would altcci the analysis (water
diversions. Section 404 Clean Water Act permits,
wastewater discharge permits and number of
groundwater wells relative lo wetland acreage),
bul did not find these changed the outcome of the
analysis.

Once 5th code HUCs were determined as units
of analysis and impact categories assigned, we
(allied areal coverage of each HGM class within
each HL'C and summed HGM classes using the
gcopiutcssing functions of ArcGI S 9.1. These
weir then exported into Microsoft Excel, and
pivot table functions were used to generate
descriptive statistics of HGM classes. Functions
weir identified fn>m a review of HGM literature
(Hnnson 1993. Junkovsky- Jones et al. 1999a and
1999b, Hauerct al. 2002a and 2002b, Sheldon et
id. 2DQ3. Tincrcl aL 2000. Tincr 2003a and 2003b)
and professional opinion.

The study assessed eight wetland functions:
I) surface water detention andslreamllow
maintenance. 2) nutrient cycling. 3)sedimenl
trapping: 4) plant community maintenance. 5)
shoreline stabilization; 6i aquatic habitul; 7)
terrestrial habitat; and 8) conservation of wetland
biodivcisiiy. We assigned functions to wetlands as
a class, rolherthanio Individual wetlands, so si/e
was not considered us a factor We then ranked
each wetland class as having "high", "jmderute"

or "low"* importance vis-a-vis that function. based
on literature reviews, field surveys in the Gallatin
Valley, and pmlc.ssion.il opinion.

1 ) Surface water detention andslreamllow
maintenance. Floodplain wetland^, especially
along I o we r gradient streams and rivets, an:
typically inundated during flood events, and to
a lesserdegrec. during o verb ink (lows. These
wetlands reduce flood peaks by capturing and
storing water, stored water is released later, helping
In maintain baseline flows We ranked Lotic River
and Lotic Stream Lower Gradient Floodplain
wetlands "high" on this calegoiy. Island wellanoS.
fringe wetlands, and Middle Gradient Lolk
wetlands were given a "moefcrate" ranking on this
category, is they typically do not store as much
noDclwalcrfTiner2D03a|. Tenrnc Intcrfluvc
Basin wetlands, while not on lloodplain soils, may
receive and store considerable amounts of snow
and rainfall, and arc often groundwater recharge
onddischargesitcs. Because of their landscape
position (the intcrfluvc area), they contribute

lo both detention andslreamflow maintenance,
so these were also given i "mndcrale~ ranking.
Tencne wellanoV not in Ihe interfluve area and
Tenene Inieriluvc Hat wetlands were ranked as
"low.

2) Nutrient cycling. Forested riparian wetlands.
Ihickly vegcluled upland wetlands, and ponded
wetlands with aquatic bottoms all take up and
recycle nutrients (c>g. nitrogen and pfmsphorus).
Slreamside wctlanch. through Iheir sediment
(rapping functions, funhcr reduce phosphon&
input into flowing waters. We ranked all Terrene
Basin wetlands is high on this function, and oil
Floodplain wetlands is moderale. although we
note thai if we weir assessing individual wetlands
Ihcse rating could change depending on vegciaiion
structure (i.e. large B asin wetlands with subslantial
open water could rank "moderate** and some
heavily vegetated lloodplain wcllamb could rank
"high.""i. Fringe wetlands and flat wetlands were
ranked "low." because vegciaiion is ofien sparse

in the former, and flat wetlands do not generally
contain as much ofganic maltcr is basin wetlands

3) Sediment (rapping. All wetland* perform some
sediment trapping functions, but Ihcsc funclions
air especially significant near walcnxiuntcs

in agricultural areas. Therefore, we gave all
Floadplain and Intcrfluvc Basin wcllanus a ranking
of "high" on this calegoiy. while upland Terrene
Basins weir rated as "moderate** and all Flat
wetlands were rated as "low."

4) Plant community maintenance. Wcllanus wilh
fluctuating water table depths, layered soils, and
oiganic mailer can support Ihc most diverse plant
communities, while wetlands with considerable
open water or wellamh subject to ice scour and
frequent bankful flows cannot. We gave floadplain
wetlands a "high" score in this calegoiy. because
Ihcy typically have the characteristics associated
wilh high plant diversity. Beaver ponds* were
ranked low because of the open water associated
wilh them (the edges of beaver ponds in Ihe
Gallatin air characterized simply as Floodplain
wetlands, and therefore rank as "high"*), as were
Flal wetlands, because Ihcy trap less organic maiter
and have less of the horizontal complexity that
leads io plant diversity. All other wetlands- ireeived
a score of "moderate" on this function.

5) Shoreline stabilization. This is on important
function for wctlaixls in coastal, estuarine, and
lacustrine settings, none of which are found in Ihc
Gallatin. In our study area, the only shoreline exists
along rivets and streams. Floodplain throughtlow
wetlands were all ranked as high an this function.
Fringe wetlantfc often contain deep-rooted
vegetation providing critical bank anchoring* bul
because many of them are associated with unstable
substrates, and have lillle deep moled vegelalion in
Ihe Gallatin, we gave Ihc diss is a whole a ranking
of "moderate." Beaver ponds, basins, and Lotic
Island wetlands were given a rating of "law.*"

6) Aquatic habitat. Fringe, and island wellaiufc
offer cover, nutrient inputs, temperature
moderation, andjuvenilc and rearing habitat
for fish, especially along lower gradient rivers
and streams, so these were all ranked as "high.**
Floadplain wetlands fitter sediments that could
clog spawning gravels, contribute woody debris

m

lo streams, and shelter reptiles and amphibiara, so
these wen: ranked as "moderate. ** unless ihcy were
beaver pomK in which case Ihcy were "high."
Basins arccritical to reptiles and amphibians,
so these too were given a rating of "high." Flit
wetlands were ranked as "low" because ihcy
typically lack open water and have less structural
diversity.

7) Terrestrial habitat. Wbody wetlands ofter
food, shelter access to water, and breeding and
reproduction sites for a wide range of mammals
and blob, so all Hood pi am wetlands received a
"high" ranking in thiscalcgoiy Jntctfluvc basins,
because of their proximity m woodlands and their
ownstnarlural complexity, also have considerable
habitat value for hinls and small mammals, so
they too were ranked as high* Non-inleifluvc
basins and Inlerfluvc Hat wclluntfe wen: given a
score of "mndenrtc," while Lotic Fringe and island
wcilands. which provide more limited terrestrial
habitat, were ranked as "low"

8) Conservation of wetland biodiversity* These
rankings were assigned ex past facto based on the
representation of wetland classes in Ihe study area.
The least common wetlands were given a ranking
of "high/ while the most common and well-
represented wetlands were scored as "low" and
"moderate," respectively.

National Wetlands Inventory
Enhancement

Sites for field sampling were selected by using
existing NW1 mapping to identify localiom we
could access and represented Ihe spectrum of NWI
types. We coordinated with the Gallatin Valley
Water District to gain private land access. Sites
were characterized for vegetation and the correct
NWt type was rerouted if the mapped type was

incorrect. Vegetation communities were coded to
Ihe closest NVCS or Hansen rial. |IW5l type and
a crosswalk table was coratructcd.

We adapted Ihe HGM modifier key developed
by Tiner (200 3a I to lest Ihe application of HGM
modifiers tn NWI delineation in the field Major
NWI types were assessed far their ecological
functions through a review of the HGM literature
cited above and professional opinion, and were
added to Ihe crosswalk.

lb refine Ihe crosswalk for future use in mapping,
we wanted to determine if any or all of the
Palusthnc and Riverine wetland classes would
predictably align with a specific HGM class. If
I his were so, Ihen future mapping efforts could
use a simple "if-then" logic approach to assign
HGM categories to given NWI attributes. Because
we hod observed a onc-tu-many relationship in
our visual inspection of the data, we began by
Icsling Ihe hypothesis NWI Riverine types would
always be HGM Lolic types and NWI Palustrinc
types would always be HGM Terrene types. As
we sifipected. NWt Riverine types had a onc-
lo-onc relationship with HGM Lotic types, both
as a group and one a class-by-cltss basis, but
HGM Lotic types had a onc-tu-many relationship
with NWI Riverine types. All Paluslrine types
had complicated onc-to-many relationships with
HGM classes. We then constructed a pairwisc
comparison/correspondence table, using numerical
scores of 1 to 10 lo represent the percentage of
observed incidences of an exact pairwise mulch
(sec Table 5 under Result, below) However we
urge Ihe reader K> refer to the dichotonuiK key
rather than to the correspondence table to extend
this methodology beyond the Gallatin, This caveat
Is disclosed more fully under the Results section,
below.

U

Rksults

HGM Classification of Wetlands

Tables 3 and 4 show the breakdown of HGM
types amiss the study area, as determined
by our classification keys and accompanying
visual inspections We used a total or 32 HGM
classifications. 17 for unaltered wetlands and 15
Tor altered wellantfc. Nearly 85 - I of the wetland
acreage in the study* area is classified as Lotic.
In the NWI classification scheme, the Riverine
System includes only those wetlands within the
active channel of a flowing water feature depicted
as a 2-line channel (i.c. a channel wide enough to
be represented as an area, rather than a single line)
on a 1 :24.G0D U5C5 tono map. Riverine wetlands
most typically include fnnge wctiintts ilong thr
liver's edge, mid-channel islands and point bam
submerged at bunkful flow, and side channels
completely flooded by the river at least part of
the year. Stable point bais and doodplains above
the active channel are classified as Palustrinc,
even if their main source of water isover-bankfu)
How?;. By contrast. Lolic wcllaixls in the HGM
system include all those wetlands found within
the approximate 100-year floodplain. as that i-
dctcrmincd fmm floodplain mans or soils maps
indicating floodplain soils. Lolic wetlands would
therefore include wetlands along the maigin
of single-line channels (i.c. streams), saturated
backwaters, old meunders. oxbows, and the like.
In this way. the Tcrrene-Liilic HGM classification
dichotomy conveys more information about the
origin and hydrology of wctlanos in a given area
than docs the Riveri ne-Paluslrinc dichotomy. It
is not surprising, therefore, that Lolic wetland
acreage is concentrated in the Gallatin Rivcr-
Gallatin Gateway (47 r 4) and Lower East Gallatin
(25ft) walcishcds. because these watctshcA
contain the main Gollalin River and East Gallatin
River channels. Smaller proportion are found in
the SmithCreck <IQ r +) and HyaliteCreek <9*lj
watersheds. Terrene wetland acreage is most
plentiful in the Lower East Gallatin i4l ,: o urxl
SmithCreck (27 1 *) wale&heds.

Wetland alterations (draining, ditching,
impounding, excavating) were determined solely
by reference to NWI polygon attributes, except

when the polygon in question hud to be visually
inspected for its primary classification or for
accuracy. For example, a PuluHirinc. Aqauiic Bed.
Semipermanently Flooded. Diked/Impounded
(PABFh) polygon entirely surrounded by upland
soils wuuld be classified by the G15 as a Terrene.
Basin. Diked/Impounded HGM type. In the study
area us a whole, less than 84 of the wetland
acreage fell into the altered category, although
when wetland numbers were considered, almost
1 7 l ft of wetlands were altered in some way. This
suggests two possibilities: fiist. smaller wellanoV
ore more susceptible to manipulation or alteration
and second, activities such as draining, ditching,
impounding or excavation typically result in the
creation of small wetlands. Alterations also tend to
be clustered on upland soils. Although SS^tf of all
the wetland acreage in the study area is Lotic. as
noted above, 72 r 4 of the altered wetland acreage is
Terrene.

HGM Assessment of Wetlands

Figures 4. 5 and 6 show the watetshed boundaries
for the 5th code HUCs over maps oT the geology,
elevation, and hydrologic features of the study area.

The Gallatin River-Gallatin Gateway watershed
and the Lower East Gallatin watershed were the
most similarof the landscape units (ix* 5lh code
HUCs) . Each has a major (4th or 5lhonJcn river
(lowing through a broad, alluvial valley. Both
have uplanoSof scdimenlaiy rock and (at higher
elevations) gneiss and schist, and both are in the
1 2-14 inch precipitation /one. The main difference
between them is in elevation; the East Gallatin
River originates in a lower elevation hand than the
Gallatin River. This is reflected in the wetland
profiles: although they have similar proportions of
Lolic and Terrene wetlands, including inlcrfluve
wetlands, the Gallatin Rivcr-Gallatin Gateway
watetshed has aver 5()0 acres of Lolic Stream and
Lolic River Middle Gradient Floodplain wetlands,
while the Lower Fast Gallatin has none

The HyaliteCreek watershed was similar lo these
Iwu in terms of geology and precipitation, but b
higher in elevation, with only its northern end in

12

fcjftfr i L'tttdirmt wetfanct octrafp irt tftt* xltutv a> i-j- bv HGAI ciast.

—

iGM CLASS

5r.dgerCk

DampCk )ryCk

£allatrn River

-ya ne C hi

.ower E. Gallatin

.GMerGaltatin

I-mi-h CI.

Upper 1

; Galbtr

Total

3 <TpevWS to t'-ak-s be**n dia(ned h

039

021

BUB

453

olc Rta*. hfcmHheni Rnoafolain &

032

■:i 13

.olc Rrrcr Loa« GraSeni. Rw^an. D 1

171

1.71

.crfc Rh-er.Lo'Arr Grt*£enl Hw^an. Ei

039

039

.ofc W«c .Middfe Gradenl. Roodptoin. & 1

031

n a

.of c St raiTi Lower GraalenL Roodplain. 0.0

19,19

1*00

2312

0031

n. Steam Lower Gradient Roodplavv 1

DM

1"4

14 4B

8 19

4.19

234

ioi

3267

,o*c Steam. Lower Granien! Roodplain. t

MB

1741

3.14

327

237

20 48

.o(c Stem. Middle Gracteni. Roodploin. D 1

0B1

048

133

Terrene. Basin. DD

12 87

MO

28 TO

30.17

Terrene. Basin. D 1

072

EH

340

? II

230

— -*j

239

129

3 10

31 43

Terrene. Basin. Ci

11.97

1041

0,08

1333

073

020

34 80

Terrene, her ftuve. 0D

328

328

Terrene, hterfluve, &iai DD

1.1*

1-87

774

1073

Terrene, hterliuve, Bcc*i 1

189

909

430

1127

Terrenp. herftjve, Basvi E&

7739

30.70

113 11

17320

J rand To tol (Acres)

0.72

in oa

430

WQM

87 43

irvm

I93J

30 rr

1143 4Mir

* These? wete MW| weland prij*gons
eumwied dunng lie das^icalon and ^afng
pioceso tiai ^eie nor distngijiJiatte OS

we lands on NA1P imager y.

TubJe4+ Altftrd urthstut uf/ra^c itttftc tUkJvarett. i

WflGilrV

/tfff.

UGH CLASS

a^srrCk

:amp Ck

DryCk

"akitin FHvri

-lyalifeCk

jiwpi E jallat n

.owet

Goltotln

SmrthCb

iff*

■£. Gallatin

fatal

*i|3pes** to have fawn drain-d*

008

021

003

093

oic Rfrer kitemaltenr Hoocfcilajn. Ei

032

□ aj

o*c Ritfrf.LoMTf GraSenl- Roodptaln. D 1

1.71

1.71

o*c fcp^r Lowe* GraofenL Roodpfein. &

□ 3S

ao9

.o#c Ffrrr.Mddfe GrarfenL Roodplan. 0.1

03

031

.o*c Steam Lower Gmimi. Roodplnn. DP

1919

1830

23.12

0031

ni:: Sroflrn Lower Gm;5fl-H. Roocfctlmi. D 1

i :-u

104

1440

8.19

4.19

2.94

131

32 67

.o*c Steam Lower Gradfonl. Roodplan. E 1 ■

039

174

3.14

327

237

2049

.oic Steam MkfcJe Grculeni. Roodplan, D'l

aSi

048

130

Feticns.Sasin.DD

U£^

830

28. 7C

30.17

Isnww Ha=in D 1

0.73

238

340

7.10

230

228

339

129

H.lfl

3143

terrene Soon. E*

1197

104

ona

1333

□ .73

n20

%4sa

terrene hter Ajve, D

328

32a

r-nerw tiler ftjve, rfcrai, DO

Iittl

137

7,74

10 '3

terrene bteii;p.&iin Dl

ISO

3 09

4.30

1127

IfOBis, frtlerlljve, Basil. E*

27.38

30.70

113.11

17320

jrand TotaHAcm)

0.72

13 OS

•130

90.19

7743

17331

1934

30.77

1343 4MD7

* These nere NWI fc^landpd)^c<is exam-
ined in*ing tie 0sL-3fr.ii ton and luring pro-
Cetei hal Werenoi dstngushahJe as tve lands

on NAlP Imnjei y

INTNAJJ&

,* ^

J <#

A^V

///

A--

D 23

D

ID

Fi£wr 4, Gmitt&ytf if.'iiV 1 . uini.

15

y r V- i -— ^— n

Jj l iV **<►*■** j^'**^* , *^"™\

Lrrftairn^tr r.d.irr.4-^ 1 h,4rcm I / *J

? ( // 1

^^-^T^J^m i

ELEVATION, JnmtMcre

| 1J19 -1581
] 1,561 -1334
] IBM -2.192
| 2.192 -2.907

23 3 10

1 1 1 1 1 1 1 1 1

Ufea

Fi&wv 5. EtevaShm ^jftliufv ami*

in

fi&or A Hydndagy tJ tfiutv usru*

17

Ihc brand, low-lying valley floor, and although its
northeastern edge cxlcnds into Ihc li as\ Gallatin
River fioodplain (giving it a complement or Lotic
River Fioodplain wctlatKti). il is nol bisected
by a major river Like Ihc Lower East Gallatin
watcishcd. the majority of its terrene wetland arc
in the inlcrfluvc area between the East Gallatin and
Gallalin Rivcis. No other watershed was suitable
for pairing with Hyalite Creek.

The Smith Creek and Dry Creek walcishctK in the
northeastern part of the study area are similar in
most respects, although Smith Creek has a more
substantial set of perennial streams, several of them
extending onlo the East Gallalin fioodplain. Many
of the wetlands in the Smith Creek watctshed were
classified is Lotic Stream rather than Lotic River
because they were physically closer to stream
features than to river features, so the fact thai the
wetland profiles appear different (Dry* Creek has
Lotic River Fioodplain wetlands while Smith Creek
has Lotic Stream Fioodplain wetlands) is probably
more an artifact of the classification than of any
genuine difference.

The portions of thcBridger and Upper East
Gallatin walcisheds lying within the study area
differ from Ihcothcrsevcn primarily in their
elevational profiles and geology, being gencrully
higher They also have a higher percentage or
gneiss, schists, and sedimentary rocks than
alluivium. and share a wcsi-facing aspect Bccat&c
only small portions o[ these watersheds were
included in the NWI mapping, it is impossible
to say whether they truly have similar wetland
profiles, although it appears this is the case, with
only scattered Lotic wetlands.

The LowcrGalliLin and Camp Creek share
similarities in aspect, elevation, precipitation,
and geology, but differ insofar as the lower
Gallatin includes portions of the Gallatin River
Because of this, their wetland profiles appear to be
considerably different, with Camp Creek lacking
any significant Lotic River wcllanus. However,
these watersheds are also not completely mapped,
and their profiles may be more similar than they
appear

Table 5 gives the scores on the metrics wc used
for assessing impacts, as well as additional metrics
(eg. groundwater wells, water diversions) collected
but were not useful in the analysis* We were
unable lo discern any fundamental differences in
Ihc wetland profiles of these
matched watcisheds thai could be attributed lo
impacts. In some cases this was due to the mapping
gap. The Bridgcr Creek/Upper East Gallalin pair
had the greatest difference in impact ranks, with
Bridger being the least impacted of the watershed
and Upper East Gallatin the most

However, there are only 2 mapped wctlanoS in the
BridgcrCrcck walcishcd and _VJ in the Upper East
Gallalin. More than half the wetland acreage in
the Upper East Gallatin has been altered, which b
consistent with its impact ranking, but the same
is true of Bridger Creek* where the larger of the
Iwu wellaiufc is altered. Comparisons between
these two wetlands watcisheos are therefore
meaningless. A similar problem exists with the
CampCreek/LnwerGallatin pain although they are
closer on the impact scale (4 and 6. respectively).
It appeals similar percentages of Terrene wetlands
have been directly altered in the two watershed*,
bul mapping covets only half the watersheds, and
in the case of the Lower Gallatin, docs not even
include all of the Gallalin River.

The Gallatin River-Gallaiin Gateway/Lower East
Gallatin pair, however, did exhibit differences in
line with their impact ranks (5 and 7, respectively),
at least in terms of direct alteration. Less than
4 f * of the wetland acreage in the Gallatin River-
Gallaiin Gateway watetshed has been directly
altered, while alteration have been made to almost
1 1 ' 1 ol Lower East Gallalin wetland acreage,
accounting for 37.68^ of all altered wetland acres

mm

across the entire study area. It may be. therefore,
the difference in wetland profiles between these
two watcisheds lie. the absence of Middle
Gradient wcllamb) has an anthropogenic cat&e
such as residential development along stream
corhdois. It is also worth noting the Hyalite
Creek watcishccL which differed from these two
wateisheds only in having m fourth or fifth oalcr
stream, had a higher impact ranking than either
of them (8). and the highest percentage of altered

IS

*©

Tahtr JT Irttf tacit (*\ Stfactnir HUCs in x!ttd\ atvtt^

kdfaotor

BridgcrCli Ca»pCh DryCh Galbtii ft-GG

UyaliteCfc Lo«erE. Galblia LowerGalfatin

Mac*

Upper E.Garblin

ID* permits*

13 058

0.03

008

oxw

0O3 0-01

2.84

SW w*tb"

09 9J9

047

12

ua

141

089

034

74 8

r/at-rdrveroons*

107 10 33

129

1,39

42

147

114

123

233

Wastewater pGtmita*

c

0JD1

028

tooddenrty"

13 73 10J)3

9.79

1734

29 00

27J

|7| 1024

39.10

3 Opubtfon den&rty"*

0*12

DJ4

o.n

[■ J

184

iuman Loftdcover****

1328 7940

39.19

39.01

93 .-":

03.13

7337 014

74.93

Weed coverage"

3833 942$

8113

87.10

87,04

87-04

100 02.3

83,0

Septic Drnorty**

0i>1

04

3*13

0.72

0.10

2.72J

^I'tiage Parce! S'=e

3008 39-70

104,21

1020

8.39

4.93

144

3748

244

Percent private ownership

71 BO

91 17

84 .OS

90.19

924)4

M4I

8034

OBBI

74.70

^errenl pubic ownership ease merits

3323J 1838

23,91

2334

7J0

20

19B3J 2? 2*

29.34

mpnet ranking (1 denoted least. 9 denotes moot)

1

J

3

8

8

i

7 2

9

otal acre*

3SO0 453 17

19487

tOS44|

393SS

29422 3404 2

28Z73

ffetbnd Acrr*

23 108

2420

334 ioos

304 740 23

* pe* tVef and z*re

H i — ■. per ojiare nrfe of watershed
" **oJatre'G *iti papijakm deceit/ > 103 per

per oquaf e mile

" 'aol aoe-iki residential. agnctJUah iiiljstial

or rantnet dd iters

* ■• of sections iwti one cr rKie noxious *reds
" "» of acres «ti sepic density rating THyfi

wetland acreage (16.374) of my watershed with
100 or more wetland at to The Smilh Creek/Dry
Creek pair, which ranked 2nd and 3rd in Impact*,
differed in their alterulion percentages, with Smith
Creek (the less impacted) having almost 84 on Is
acreage altered and Dry Creek only 3fi. We cannot
say for certain this is not a reflection of incomplete
mapping, hut we helievc it is likely attributable m
(he Smith Creek waleishcd wetlanos being more
concentrated cm the noodplain*

We also iKed impact rankings to determine which
wetland functions wen! most at risk in the study
area us a whole. Johnson (2005) has made a
useful distinction between impacts and effect*,
noting impacts are actions while effects arc Iheir
rcsulti. Draining. ditching, ordirccl alteration of
a wetland is an impact: loss of specific function
is an effect. Although altered wetlands can still
perform wetland functions (e.g. excavated wetlands
can provide water storage, or habitat for ducks and
lish), Ihcy generally do not perform ul as high a
level as unaltered wetlands in intact landscapes*
A future project will examine the loss of wetland
function in the Gallatin Valley over the past twenty-
plus yeus based on an JiGM analysis of past
and present wetland distribution and abundance.
Heir we are primarily concerned with identifying
Ihe functions at risk in watersheds where overall
human impacts —that is, impacts at the wetland and
whole-watetshed level— are highest*

The Gall atin River-Gallatin Gateway, Lower East
Gallatin, and Hyalite Creek watcntheck conlain
7fis of the altered wetlands in the entire study
area, but they also contain 7SS of all the wetland
acreage, so this is not particularly surprising.
What we are concerned with here is the threat to
Ihe integrity of non-altered wetlands that is posed
by more generalized impact: road building,
residential and commercial development, irrigation
and other water withdrawals, the spread of noxious
weeus. and so on. Assuming ihcse sorts of
generalized threat to wetland integrity arc nearest
at hand in those watcishcds with Ihe highest impact
ratings, what are the watershed functions particular
to their wetland profiles?

The watctshech with the highest impact rankings
ojt Ihe Lower East Gallatin, Hyalite Creek, and Ihe
Upper East Gallatin Together, these watersheds
hold 87*< of Terrene Interiluvc Basin wctlancfc.
7! ft of Terrene lntcrfluve Flat wclluiuK and 33M
of all Terrene B asin wetland?. Because these types
of wetlands typically rank high on nutrient cycling,
sediment trapping, aquatic and terrestrial habitat
and wetland biodiveisily. it follows that those
functions are at risk both wilhin those watersheds
and across the study area as a whole. Similarly,
those three watersheds conlain almost 40*ft of Lobe
River Lower Gradient Hoodplain wellaiub. 36 1 * of
Lolic River Lower Gradient Fringe Wetlands, and
39*tf of Lolic Stream Lower Gradient Hoodplain
wetlands. Because these wetlands exist in other,
somewhat less impacted watcishcds (notably
Ihe Gallatin Rivcr-Oallalin Gateway and Lower
Gallatin), their functions arc somewhat less at risk
across Ihe study area as a whole. Nevertheless, we
should consider surface water detention, stream
maintenance, and plant community maintenance
are al least moderately at risk in the study area,
Eurthcrmore, because these Lolic types, like the
Terrene types, also have high function m lings for
sediment trupping and aquatic/terrestrial habitat,
these functions may be especially compromised in
Ihcse three watersheds.

Impact rankings are not sialic, and as rankings,
Ihcy are relative to each other We do not have any
reference standaid watersheds in the study area
againsl which impact scores could be calculated.
The least impacted watershed wilh any significant
wetland acreage. Smith Creek, still has over 51 **
human land cover (based on the best available
but 15-yearold land cover maps), almost 69*(
private ownership, and no&jous weedS in63*!t
of its sections. Wetland function may be less at
risk in this watershed than in the Hyalile Creek
wateished. bul it isslill at risk, and if development
pressures continue h push new* development into
Ihe foothills, impacts and risks will both increase.

Evaluation of the Classification
and Assessment Methodology

GIS-based attribution of HGM descriptors was
simple and efficient m the case of NWI Riverine

20

System wet I amis. Bcciusc these well and* ore
by definition within an active eiutmwl* they were
implicitly pan of the HGM Lalie landscape
position, which includes tny wctlanoS within
the K\WtflutMlplain. USCS topo maps allowed
npid classification into Lotic River and Lotic
Slream types, and the existence nr a good soils
layer allowed us to complete further assignment
of floodpluin wcllanus without difficulty. The
GIS approach began to have limitations when we
attempted to assign Jandfomt. Paluslrine Aquatic
Bed wetlands wen: assumed to be depression)],
and thus were attnbuted as Basin wetlands* We
made the some assumptions about Palustrine
Unconsolidated Bottom wctlanoK but as these weir
all directly altered, they wen: classified as such
(e xi > Terrene, Inlerfluve, Basin. Excavated). The
otthbution of tandform to Palustrine Emergent.
Scrub-shrub, and forested welluixls was more
difficult- Most of these were classified as Lotic
Stream or Lotic River Floodplain wcllaiuh. or
as Terrene Inlerfiuvc wellanos with no further
londform descriptors, hut the ones we inspected
visually as pan of our accuracy assessment wen:
often given additional modifieissuch as "Basin**
or "Basin Oxbow." since Ihesc characteristics were
clearly visible on the aerial photographs. Because
oxbow wetlands have a notably high functional
capacity instrcamllow maintenance and water
detention, the inability to identify these from a
GIS alone should be considered a limitation of
the approach* We aho note the classification of
Polusthne Emeigcnt wetlands that were neither on
a floodplainor an interfluve could only proceed
beyond the "Terrene** classification with visual
inspection. Appioximately 15^ of wetlands could
not be classified at all with the GtS. primarily
because of incomplete geomoiphic information
on the soils maps: it was unclear whether the
underlying soils wen: or wen: not pari of the
floodpluin until the aerial photographs were
ljD.iU-v.V in .1 Mrti >in '■. Mh .'iii.Mi' li ■ -\ :>i\

floodplain maps, this would not have been an issue,
and would have saved the step of drawing out a
putative lloodplain based on the soils maps.

not refine our dichotomous key beyond its initial
iteration. We caution the reader that this is a
limitation of the key in future use.

We believe the assessment methodology shows
great potential for future use. Because the study
atra has neither complete nor current maps, we did
ml attempt a highly detailed functional assessment
that would integrate HGM and Palustrine types
to tease out differences in functional capacity
(e.g. between Lotic River Floodpluin-Paluslrinc
Scrub-Shrub wellanos and Lotic River Floodplain-
Paluslrine Aquatic Bed wellanos). Similarly,
because the National Land Cover Database is
almost 1 5 years old, and does not capture the
demographic change in (he Gallatin Valley, we did
ml assess potential land use impocts to functions
along wetland orstieam corridors. However
w ith updated wetland and riparian maps, and
edited (or newly generated) land caver layers,
the combination of mare detailed functional
assessments and more focused impact assessments
would yield a comprehensive coarse-scale
evaluation of wetland function and condition. This
is the direction we encourage in future work.

Crosswalking between A 7 W7,
HGMandNVCS

National Vegetation Classification System (NVC5)
and other vegetation types (Hansen cl aL 1995)
were readily associated with U5FW5 types after
ourfieldwork in the Gallatin Valley, although the
relationship was typically complex with several
types possible for a USFW5 type, Within NWI
classes we also associated water regime modifiers
with specific vegetation associations when
possible. This result also allows an understanding
of the considerable vegetation (and also the related
habitat) diversity inherent within a type. For
example. Palustrine Emergent wetlands have a high
conservation of wetland biodiversity functional
value in the HGM cmsswolk due to the wide
variety of vegetation associations potentially
present.

Because almost all our wetland* were on the volley
floor or the floodplain terrace of larger streams,
we did not encounter slope wellanos and so did

The relationship between HGM and NWI types
was also complex, especially in the case of the
NWI Palustrine Class*, ami the HGM Lotic Class.

21

Tabic 6 is a correspondence tabic showing the
relationship between wetland classes (Appcndu

B contains the abbreviations used in ihc tabic).
The NWI Palustrine Class has a one-to-many
relationship with MOM types, including both
Terrene and Loiic types, depending on ihcir
location and underlying soils. Similarly, although
NWI Riverine Class wetlands arc always HGM
Lolic wetlands. HGM Loire wetlands can be
Rivcnne or Palustrine. In general, once Ihc
relationships were defined through repeated
iterations, descriptive statistics, probability
analyses and inspection oT aerial photographs, the
CIS-based key worked fairly well. However, it has
not been tested beyond the Gallatin, and therefore
we do not know how well it would work in areas
with more high-gradient slopes and less fioodplarn
development

While the key did not allow us to produce a simple
one-to-one crosswalk, the ability to attribute
individual wetland polygons with HGM classifiers
holds great promise for detailed wetland analyses
that combine the best of both systems. In the field,
we aligned specific HGM wetland functions
to wellantK based on site-specific factors that
could be discerned from observation of wetland
composition, hydrologic relationships, and
overall site condition. A similar kind of detailed
connection is possible when a database includes
both HGM and NWI attributes for specific
polygons. For example. Lolic River Floodplain
wetlands with thick forest coverage lake up
more nutrients than sparsely treed or semi-baic
wetlands, but vegetation cannot be inferred from
an HGM classification atone. When a particular
polygon has both an NWI attribute and an HGM
attribute, there is enough information to begin
characterizing the degree to which a particular
function is carried out by assigning weight values
to modifiers or to attribute combinations. That
type of detailed analysis was beyond the scope of
the current study, but with additional research and
fie Id- testing, enhanced wetland databases would
provide considerable assessment potential. We
are attributing HGM types to NWI and riparian
polygons now inourBillenool Valley mapping, so

Hut would be an ideal starling poinl for advanced
analysis.

Despite the complex relationship between cltuttcs.
we were able to produce a guide lo NWI types,
functions and vegetation, reproduced in Ihc next
section. We recommend that work continues in Ihc
association of USFWS wetland and riparian types
with NVCS vegetation typos as mapping coniinucs.
Vegetation strongly influences habitat and many
other ecological functions. Most land managets
directly manage vegetation. so a linkage to the
information rich NVCS with spatially explicit
wetland/riparian mapping will be beneficial for
many aspects o[ planning and management.

Guide to National Wetlands
Inventory Types and Functions

We examined wctlanoS in the Gallatin Valley and
reference literature lo link the common NWI types
lo National Vegetation Classification Syslem types,
other vegetation types (Hansen el al 1995). and
Hydmgcomoiphie (HGM) (Brinson IW3) wetland
functions. While the HGM Approach is useful due
lo its focus on the functions aT wcilandc, there is
no HGM mapping. Since NWI maps are available
our goal was to associate HGM functions with
NWI types to improve assessment and mitigation
of wetlands in Montana. The association o[
NWI types with HGM types has pmven useful in
watershed- based wetland planning and evaluation
in other areas (Tinerct al. 2000).

The NWI classification system is hierarchical
with an increasing level of detail. Five broad
Systems cusl: Palustrine. Lacustrine, Riverine.
Estuarine. and Marine. Only the first three exist
in Montana. Lacustrine and Riverine Syslcms arc
further divided into Subsystems, for example.
Lacustrine Systems may be Limnetic or Littoral;
Riverine Systems may be Upper Perennial. Lower
Perennial . Intermittent, etc. No Subsystems exist
in the Palustrine Syslem. Below* Systems and
Subsystems are Claries describing the substrate,
and hydmlogic modifiers describing the water
regime. Accuracy of map coding decreases
further down the classification system; hydn>logic
modifiers are the least accurate, especially in
Montana with our periodic droughts.

22

&!»/«• a

Co*rrjfiprt(iffir<*

foW*All7fr>//GM.

-:

m*

tBVl

|RO»t

■ H ♦-!

L ■ W» 1

i K(4

m^i

L *:i ifUl

HD-T

Cr*l

iHiN

ppp

» W»

i * >n

K«.

.11.:-*

Llll^l

1H#

P*r

-**■

-«■

■b

■V

:>*

>*>

■■

,m

THI

IB*

-rw

»

BB)

*twf *M

■■

4

i»

a

J9

*.»

.

4

4

Q

lt|

!•■♦

H

*3

4*Ut

.

a

+■»*

*"-

B

'

H

3

t'

M

n

1

"

■ ■II

i

4

n

■■»

n

.

«..

■

a

T

41M

»*

a*

*K

-

*j

■ r

3

•

at

»<*

a

«■(«

^4

■

o

JT

«c.

■

1

a

a

a

™

i

■

^

a

■W

* U *.

BBBh

.

t

BUM

£j

-^

-*1

B

j

>>

* IH

4

a

;-

pK

a

BBB

-■

1

•

**r

m,

a

".i- ■

a

.»

4

4

n

*— .

)

4

»

a

K

.»

4

*«*

.

a

^.

U

«

r :

B

>...,

«4

Mi.

...

■K.

:-»•-

ft

it«""

*

T

tn«<

BT

Dimj

1*

or

Pi"

#■

**

:aiv

i

■M

I

a

«H

U]t(

be

..;.,.

-, j ,i.

■

■

1

"

■

T*4

"

rt

1

tf

17

'

:,]

n

s

H

1 ■

i

*

•**

^

O 12

a

/

3 IB tZT

'

■

: ui

IBflhlJQJlBpd cdurana ind*aa? racist anrani N'A'I or HGM dtt&esin ;'. -:l . a ea

Most Montana wellaniK ore Paluslrinc. This
System includes any wcllaiufc ncil within a stream
orrivcrchanncl (Riverine System) or within bodies
of water which ire either > 20 acres OR deeper
linn 6.6 feet (Lacustrine System). SecCowardtn
el al. (1979) for complete details on the NWI
classification system* Nine possible Classes in the
Palustrine System exist. Five were common and
sampled in the Gallatin Valley. Only one Riverine
System was described.

The HGM Approach Hi wcUand classification
emphasizes wetland function, recognizing some
wetlands perform certain hydrologic. habitat, or
biochemical furclions more effectively lhanotheis.
We chose eight broad wetland functions, and
evaluated the relative importance nl'eoch function
for the common NWI types wc sampled. The
following crosswalk is specific to our study area
in the Gallatin Valley, with applicability elsewhere
depending on the similarity of environmental
conditions,

Palustrine System Unconsolidated

Bottom Class (PUB h Aquatic Bed

ClamtPABt

Stamtiary

The Unconsolidated Bottom Class has <_1n'i
vegetation cover and a surface with >25 r i of the
panicles smaller lhan stones. The closely related
Aquatic Bed Class has >3D^t vegetation cover of
plants growing on or below the water s surface for
most of the growing season in almost all years.
Most of our wetlands in this cliv* have a mix of
silt. clay, and organic matter us substrate.

Envinmment

Typically found along watercourses in meander
cutoffs (oxbows) and beaver pomh. The low
vegetation cover is ofien due to relatively deep
(>~5m) and turbid water inhibiting a higher cover of
emergent or submerged plants. In these situations,
the hydrologic modifier will be "permanently
flooded" but during years of drought, this may vary
considerably.

Vegetal it m

These shallow potxls wilh silly bolloms usually
have a concentric zoneoremcrgent vegetation
around the drier perimeter. Sec the Falustrlnc
Emergent Vegelalion System for a description of
this vegetation. As the walcrdepth increases to
about 0.5m typical emergents like cattail {Typlta
ftilijnfia)* sedges (Ca/r.vspp*). and bulrush {Scityms
aixl Sclmeiwplectut spp.) become very sparse
and submerged plants become more abundant,
including species of waterweed {Eiodea spp).
watermilfoil tAtyiiop/iyllutn spp.), bladderworis
{Utticalario spp. ). pondwectb iPotatno&eloa spp.)
and Mare's tail {Hip/Hiiis mlgarix}.

Ecolftgical Dynamics and Management
CtMtsideru tit vis

Beavers are responsible for the formation of
many of ihcsc wetlands, which may peisisl in this
stale forsomc period of Lime but will eventually
drain or become filled with silt and transition lo a
different wetland or non-wetlund environment. In
mountainous landscapes, this type and the animals
depending on it for habitat would be extremely
rare without the beavers 1 dum-building activity
Floodplain dynamics create another typical setting
for these ponds.

If river hydrology is altered, or if beaver activity*
ceases, existing types may fill in wilh sediment and
new locations may nol be created. If livestock have
access to this type, they can degrade the habitat
by trampling the vegelalion. pugging the soft soil
nearby, ami exposing soil for weed colonization
sites. As is true with all wetland^, maintaining the
hydrology is key to the welland: changes in the
quulily or quantity of water reaching the weiland
will alter the vegetation and habitat values.

Hydmge€m&rphic Types and Functions
PUB and PAB wetlands correspond to several
HGM types depending on location. IntheNWI
classification system, only those wetlands occurring
within the stream or river channel are classified as
Riverine. Wetlands occurring within Ihe bankful
channel and/or the active fioodplain of a river
or stream have a lotic landscape position: those
surrounded by uplands have a terrene landscape

M

position. Within the Gallatin sampling area, the
PUB wetlands hid a lotic landscape posilinn when
Ihcy were associated with beaver activity, or a
Icrrcnc position when Ihcy were ml on an active
floodplain The R\B wetlands we sampled wcie
also typically beaver ponds, and so were classified
as lube river or stream. These air import ml

habitats due to ihe dependable presence of walcr in
most years, a rarity in much of our arid landscape.
Amphibians and aquatic reptiles are especially
dependent but waterfowl and numerous other
ciratuirs also need these wetland*. Table 7 shows
Ihe HGM classifications Lhatcan be associated with
PUB andPAB wetlands.

Tabir7. Tx^rtJ HGM /

i/vi atutciafrd i

iith PAU/PVB urtlatub

Landstupc
Position

! - » i i ; i - ■ j in

ho Ik Water Size
Modifier

Lu tic Flow
Recline

Water Ftoti
Path

I. Terrene

Depression
Floodplain

Bidirectional
Thro ugh flow
Isolaled
Inflow

Oulllow

2 Lotic

River
Stream

Dammed Reach
Low Gradient

TafJr J\ Fim-titm tustxiultd tart FAH/VLB u rf/f i#*/s

f 11 IK Ll"El

Relative
Importance

Comments

1. Surface and
Groundwater
Storage and
Slrcamflow
Maintenance

High to
Moderate

Larger and deeper depressions arc the most effective. Tine
textured and high organic content soils are better Throughflow
walcr paths and terrene oulllow wetlands an: less effective.
Wetlands on higher order streams are more important

2. Nutrient
Cycling

High to Low

Vegetatcd types have higher importance. Denitrification
is higher with organic substrates. A fluctuating water table
is best. 1 ligh organic content and fine textured substrates
are most effective. Lotic wetlands are more important in
preserving water quality, especially on lowerorder streams.

X Maintain Plant
Community

High to
Moderate

Dammed reaches and other deeper water environments can
have little vegetation* Otherwise, a variety of water depths
create different vegetation /ones for these distinctive plant
communities.

4. Retention
of sediments,
elements, and
compound*

High to Low

Terrene Lhroughllow basins and most lotic wellands other than
Hals are Ihe most cffeciive. Ponds with a small watershed are
of low h importance. Vegetated wctlunoS are more effective.

5. Shoreline
Stabilization

Low

Most of these types are ponds with little shoreline erosion
potential.

6. Terrestrial
Habilai

High

These walcrsourccs are critical for waterfowl and many other
animal species.

25

ZttfeA ftinbwr.-/.

7. Aquatic Habitat

High

Amphibian incl aquatic reptile habitat value is high, especially
with deeper water depths. These types may also provide good
Jish habitat. The king water mention lime is important for
some macro in vertebrates.

8. Conservation
of Well and
Biodiversity

High

These are relatively common types but the absence of beaver
could significantly diminish these occurrences.

Palustrine System Emergent Wetland

Class IPEM)
Summary

This type is defined by the dominance of erect
rooted herbaceous (not woody) wetland plants.
If then: tsvXffc cover of woody plants, then the
type would be Scrub-Shrub or Forested. Then: arc
a variety of water regimes influencing the soil,
vegetation and habitat characteristics.

Effvi/tMtmefil

In the Gallatin Valley these wetlands were
mostly associated with floodplains but were also
occasionally found on sub irrigated terraces. The
floodplains range from seasonally inundated and
perennially saturated at their wettest to merely
seasonally sub irrigated at Ihcirdry extreme
Landscape features include sloughs created by
channel meandcts. shorelines of oxbow lakes
and drier or drained beaver ponds. While this
community often occurs arouixl the edges of
beaver ponds, this fringe is typically too small to
be mapped on NWI maps unless it is a veiy large
pond. The more xcric riparian sites occur higher in
the landscape and are inactive floodplains. terraces
and toes lopes: these sites are primarily seasonally
subinigated and have high water tables receding
below the mating /one by mid to late summer.

Of communities with predominantly native
plants those dominated by beaked sedge (Cwvx
uSrictdala) arc the nwsl common and exhibit
the broadest range of water regime, from
permanently flooded (oxbows and beaver potufco
to saturated and temporarily flooded. The Baltic
rush {Juncia baltiens) vegetation community b
neatly as well represented. It exhibits a brood

range of adaptability to water regimes, but does
not occur under permanently flooded conditions*
The woolly fruit scenic (Carex latut&uwsa).
slender sedge <C Imittcarpa) and bluejoint
reedgrass {Caiamn&m&Us canadensis) vegetation
communities represent a gradient (in wettest to
driest oitfcr) from sen ii- permanently flooded,
to saturated soil conditions tn wet meadows
experiencing mid- to late-summcrdiying below
the rooting /one. On the dncst surfaces, but
still potentially wetland if subinigated, is the
western wheatgrass {t ascopyntm smithii} — grccn
needlegross ^Xasetfa \ritidtda) community. .MI
these communilies are typically dominated by the
named species with only scattered other graminoids
and f»rhs~ In all but the western whcalgross-grcen
needlegross community, the most common forte
are field mint {MnHlta uiwiisis). large-leaved
avens [Geion maemphyllum), water smartweed
{Polygonum amplunhtn\).cux\y dock iRmnex
crispiis)* buckbean iAfeityanthej irifnliata)* western
water hemlock \ Cicala doughtsi), Canada thistle
{Cirsitim aiwi5t*).silvcrwced \Argvntinu amerina
formerly Paientilla anserina) and purple cinquefoi)
{Patenlilla f*uttts!ris\.

Ecological Dynamics and Management
Cons id era I urns

Hood plain dynamics create the typical setting for
these wetlands. Like all wetlands, maintaining
the hydrology^ is key to maintaining the wetland,
changes in the quality or quantity of water reaching
the wetland will alter the vegetation and habitat
values. When beaver activity is responsible for the
formation of these wetlands, sites are quite dynamic
and the community may transition to a different
wetland or a non-wetland environment within fairly
short lime periods. As is tiue with PUB and PAB

26

wetlands, beaver activity is especially impoiunl in
mountainous landscapes where this type, anil the
habitat, unuld otherwise be rare.

Canada thistle and spotted knapweed yCeittaun'ti
Iriebersttinii) are common invadcis on these fresh
Etta.

If livestock have access Id this Class, especially
when the soils arc wet, (hey cm degrade the habitat
by trampling the vegetation, pugging the sofi soil,
and exposing soil for weed colonization sites. Wc
noted sheep found beaked sedge palatable both
early in the growing season and post-gmwing
season, but caused no perceptible compositional
alteration in these perennially saturated silcs. In
the wellest sites with pugging from livestock,
western water hemlock and silvcrweed were
the most pmlilic incre&scrs, but the natives still
maintained dominance. At the drier end of
the moisture gradient, disturbance of western
whcalgrass communities can result in dominance
by introduced pusluie grusses including smooth
bmmc iBn'tnus Olenitis)* Kentucky blucgrass {toa
pmtemixi, timothy {Pldeiim ptvleitxe\ and Canada
blucgrass (/*. vontptvssa). Perhaps the condition
most favorable to exotic introduction is ciratcd
when flooding deposits fresh alluvium and seeds
or other propagulcs ate waler- or wind-borne.

HG\f types and modifiers
The Paltfitiine Emergent Wetland class is found
on active floodplairc and terraces. Hoodplain
PEM well inus in (he Gallatin sampling area were
typically Lotic Riven M kkllc Gradient, Eloodplain
wetlands- (Tables 9 and 10). When these Eloodplain
well anus hid a Throughflow water regime, Baltic
njsh communities wen? most common: when
the water regime was Inflow, beaked sedge and
wixilly fruit sedge were morc characlehstic.
We also saw h PEM wetlands along the edges of
beaver ponds. o\bows. or other dammed stieams:
these corresponded la Lotic Stream, Dammed
Reach. Fringe Wetlands withBidircctional water
regimes but weie often too small to map* Catlail
{Typhi) tati/olia) dominated these types. When
PEM wellanos were not on floodpluins. they
corresponded to Terrene. Slope, Throughflow HGM
types. Both were typically dominated by beaked
sedge, although woallyfroit sedge, slender sedge
and Baltic rush communities were also found.

Table 9 t 7Vf wrtff HGht tvpes associated with PEM wetlcttttk.

Uindsaipc

|\.MlH.]l

[jindform

I. oik Water Sfcc
Modltier

l<otk Flow

Regime

Waler Flow
Path

1 . Terrene

Depression
Eloodplain
Fringe
Slope

Bidirectional

Throughflow

Isolated

Inflow

Outflow

2. Lotic

River
Stream

Dammed Reach
Low Gradient
Moderalc Gradient

Table SO. Futtrtitt

\U iiAItH'i

fafad n?Wi PEM uetfatxfx.

I iMuiinn

Re la the
Importance

CotuineDls

1. Surface and
Ground waler
Storage and
Streamflow
Maintenance

High to
Moderate

Larger and deeper depressions are the most effective but with
greater depth cmejgenl vegetation will disappear. Fine textured
and high organic content soils are belter. Throughflow water
paths and terrene outflow wellanos arc less effective forstorage.
Wetlanbs on higher order streams are more important.

27

Tahte HI CwtinvtfiL

1 Nutrient Cycling

High hi
Moderate

Dc nitride a lion is higher with more organic material in the
substrate. Fine textured substrates ore mote effective than coaisc
substrates. A fluctuating water table is best Lolic wetlands are
more important in preserving water quality, especially on lower
order streams*

3. Maintain Flint
Co mm unity

High to Low

A variety of water depths will create different vegetation zones
for these distinctive plant communities* Can be dominated :■>
nonnalive vegetation.

4. Retention
of sediments,
elements, and
compound*

High to Low

Terrene ttuoughflow basins and lotic wetlands are the most
effective. Depressions with a small watershed are low importance.

5. Shoreline
Stabilization

High in Low

Most important in Lotic positions with higher water gradients and
ttuoughflow water paths.

6- Terrestrial
Habitat

High In
Moderate

These types are often in transition zone between deeper water and
uplantk providing cover and food for many species. Waterfowl
and other birds use this habitat.

7. Aquatic Habitat

High to Low

Amphibian and aquatic reptile habitat value can be high with
deeper walcrdepths. These types may also provide fish habitat.

S. Conservation
of Wetland
Biodivcrsitv

Moderate to
Low

These arc relatively common types.

Pulustrine System Scrub -Shrub
Wetland ClamfPSSt

Summary

This type is defined by >X*t campy cover of
shmhs or small trees <6m (20 ft.) in height. All
water regimes except subtidul and permanently
hooded are potentially included, but Montana
types are. in approximate order of extent: saturated,
seasonally flooded, temporarily Hooded and
semipermanently flooded. The semipermanently
flooded condition is especially common in areas
where beavers have been active. Although all four
subclasses (Broad-leaved Deciduous. Broad-leaved
Evergreen. Needle-leaved Evergreen and Dead)
occur within Montana, the Broad-leaved Deciduous
Subclass was the only sampled subclass in the
Gallatin Valley and is the typical Suhcltss across
the state, even in mountainous terrain.

Environment

These types are most common in floodploin

environments, often where deposition on inside

curves creates sand or gravel bats quickly
colonized bv shrubs and trees. The drier fringes of
beaver ponu\ and oxbows can also have this type,
although these are often loo small in area to be
mapped in NWI mapping.

Vegetation

The most abundant Scrub-Shrub types in the
Gallatin Valley are dominated by sandbar willow
{SuH i ev&Huy These cod \ uj from i dersc bed
of seedlings on recently deposited alluvium, which
probably will be exposed to seasonal Hooding for
a number of years, to a sapling class or a mature
stand approaching 5m (16 ft.) in height in which
other wilkiw \Sa/i\ spp.) and Cottonwood and
aspen [to/mtus spp.i species are well established
with a saturated to temporarily flooded water
regime. Three other common early succession^
willow species i.Si;/f\ spp.). Pacific willow (5*
Utcida ssp* fasiandra)* yellow willow (5. hitea)*
and Bebb willow (5. bef>hiantt). may co-occur with
sandbar willow, or in any mix with one another or
be sole community dominants. Other than sandbar

2K

willow only Pacific willow appcan* capable of
colonizing raw gravel bant.

The undergrowth dominants of Lhcsc stands
range from beaked sedge (Catvx titriatfttla)
on ihc wettcsl of siles* to bluejoint rcedgross
(Calattta&nixtix ewtadenxis) and tufted hairgrora
(Dactltttnpsia CVSfrttma) nns\\cssiXiSTltct\
to (he surface but seldom flooded, to smooth
bmmc {Brwnus otortiux)* Kentucky hlucgrass
{toa pru tenxi x I arcl. in the uncommon instances
of native dominance, western wheatgrass
(Pastwpynim stnitlui) on the driest sites. Reed
canaiygrass {Plmlatix atwulituuiw and fowl
hlucgrass (fott pa/iisffis) have a broad tolerance
range of water regime and degree of shading.
Although both are exotics, (here is no indication
disturbance (beyond what is c\peclcd in riparian
systems) is necessary for their introduction
or maintenance. Kccdcanarygrass is a very
aggressive colonizer capable of forming a virtually
monospecific undergrowth.

Although the western snowbeiry {Sympfioricurpax
oecidenSalis) community* type is quite abundant in
the riparian zone on alluvial lloodplain terraces,
only one wetland site, probably subirrigated or
tempomrily flooded* was identified- This type is
often not considered wetland or even riparian but
occurs more typically in upland swales (Hansen et
at. 1995). Et has the potential to support western
■■ he it -■ i .!>.- . green nccdlegrass {Xaxello \itidula\
and a variety of native herbs, but is more typically
dominated by exotic grasses including quackgross
U\grupywn tvpens)* smooth brome {Bmmtts
inenmsu Kentucky blucgrass ytoit prtttenxixi and
limothy ythtetfm praletuey Wood's rose {Roxa
uuwiixii) also may attain high coverages. Only
the very* wettest margin of this type seems capable
of supporting the otherwise ubiquitous reed
c anaiygrass*

Alder \.\hms iiuwiat communities were often
associated with beaver activity, both old and
new. Beaver activity (as a disturbance) did not
result in the establishment of these stanch; rather
they occupy lloodplain terraces, which may have
experienced ice jam phenomena in the post. Most
stands observed during the fleld season of 2005

are currently sufficiently removed from the main
channel to moke ice scour a remote probability.
The Gallatin Valley occurrences had a strong
component of willows (Bcbb willow and Pacific
willow), and willow-dominated communis types
were Intermingled with mountain aldcrstands.
Narrow-leaf coltnnwood (Fopuhis anpisUJoUa) is
scattered in the stands or occurs nearby and will
probably become the canopy dominant. Wcll-
browscd red-osier dogwood (Conuis svricva)
is often present in the mid-shrub layer* Stands
adjacent io beuvcr ponds and semipermanent
flooding are dominated by beaked sedge in the
herbaceous layer Higher positions* although
continuously saturated* have bluejoint recugrass,
redtop \,\£ivxtix xlolonifera), reed canaiygrass and
drooping woodreed \C\mm tatifntia) as dominant*
of the herbaceous layer, with the most constant forb
being the obligate wetland species western water
hemlock (C tfon&lasii).

Ecological Dmunths and ManaRemenl
CtHisidemik mA

Where the lloodplain is relatively narrow*
succession may not progress beyond a Scnjb-Shnjh
wetland before the watercourse meanders back,
eroding the bank where a shrub- or tree- dominated
stand was developing. Young stanus of cotton wood
would also be in the Scrub-Shrub Wetland Class:
with continued development they would grow
into the Forest Wetland Class. Some willow
communities, especially those dominated by Pacific
willow* also transition Io Forest Wetlands.

Floodplain dynamics create the typical setting for
these wcilands. Like oil wetlands, maintaining
Ihc hydrology is key to maintaining the wetland;
changes in the quality or quantity o[ water reaching
Ihe wetland will alter the vegetation and habitat
values, and new locations may not becreaied*
When beavers are responsible for the formation*
sites are quite dynamic and the community may
transition to a different wetland or a non-wclJand
environment within fairly short time periods.
Beavers are especially important in mountainous
landscapes where this type, and the habitat, would
be extremely rare without their dam-building
activity

29

If livestock have access Id this type, especially
when the suits arc wet, they can degrade the habitat
by trampling the vcgelalion. pugging the soft soil,
and exposing soil for weed colonization sites.
Fcrfiaps the condition most favorable lo exotic
introduction is created when flooding deposits
fresh alluvium, and seeds arothcr pmpagules are
dcliveird waiter or wind-home.

HGM types and modifiers
Most of the Shrub-Scrub wetland* in our sampling
area arc Lotic River or Stream. Low to Middle
Gradient Floodplains (Tables II and 12)* Sandbar
willow communities were most characteristic of
the Lotic River or Stream. Middle Gradient Inflow

wetlands* Lotic River Thioughflow types often
have iklcr as ihe dominant Bcbb willow wis
typical of Low r Gradient I i i \< Stream wetlands*
On some of the alluvial ban* associated with
Lolic Stream. Middle Gradient, wetlands, young
narrowleaf collomvooox were establishing.

Beyond the active Hoodplain, Palustrinc scrub-
shrub well anus were likely to be Tenrne
Ploodplain or Slope Wetlands with Throughflow
water regimes. Early succcssiona) shrubs like
yellow willow with a beaked sedge unde&tory
dominated Ihe Terrene HoodpUin wetland.*. Slope
wetlands had alder. Pacific willow or yellow
willow as dominants.

Tijtf/f //. TifwYiW HG\t l\jtr% uMiKuiJf*/ ut/Jr PSS ntltunUs.

Landscape
Position

! .ml]"] n.

Lotk Water Stcc
Modifier

Lotfc Flow
Rc£lmr

Water Flow
Path

I . Terrene

Depression
Ploodplain
Slope

Bidirectional

Thioughflow

Isolated

Inflow

Outflow

2. Lotic

River
Stream

Dammed Reach
Low Gradient
Moderate Gradient

Tiifrtc 12. I iir*. !wtu itiu<-iu!cti with PSS wttttituii

Function

RcUtVt

IniptirtimtL-

Comments

1 . 5urface and
Ground water
Storage and
Streamflow

Maintenance

High to Low

Fine textured and high organic content soils are belter.
Thiuughflow water paths and tenrne outflow wellanos are less
effective. Well amis on higher order streams are more imponant.
Miny of these types are on coarse soils near watercourses and
will not rale highly for Ihis function.

2. Nut lien! Cycling

High to Low

Deni trifle a lion is higher with more organic material in the
substrate. Tine textured substrates are more elTective than the
coaise substrates that most of these types have. A fluctuating
water luble is best* Lotic wetland^ are more important in
preserving water quality, especially on lower order sire atns.

3. Maintain Plant
Community

High to
Moderate

The undcislory can be dominated by normative vegetation.

30

Tii/rfr tl CitnfrwiL

4. Retention
of sediments,
elements, and
compound*

High lo Low

Tenrne thmuehflow basins and loik wetlands are the most
effective. Depressions with a miuII watershed arc low importance.

5. Shoreline
Stabilization

High Id
Moderate

Most imponanl in Lone positions with higher water gradient and
Ihioughflow water paths. Shrubs typically have erosion resistant
mot systems lhat arc important For bank stabilization.

6- Terrestrial
Hahiial

High

Especially important IbrhinJs but other species also rely on the
dense rover and the food resource some shrub species otter
Important food source for beavets.

7. Aquatic Habhal

Moderate fri
Low

Can provide some stream shading and woody debris in Lotic
positions.

S. Conservation
of Wetland
Biodive&ity

Moderate fri
Low

These on? relatively common types but some have been impacted
by improper grazing regimes.

Palustrine System Forested Wetland

Class <PFO>
Summary

This type has woody vegetation >6m (20 ftj in
height with >30W canopy cover. The Broad-leaved
Deciduous, Needle-leaved Evcigrecn and Dead
are the only Subclasses present in Montana. In the
Gallatin Valley only Broad-leaved Deciduous types
are present. Potentially all water regimes, except
sub I kill and permanently flooded, are included, but
in the Gallalln Valley only the temporarily flooded
and saturated regimes were associated with this
class.

Eftvinwttnettl

The most common landform is Hoodplains where
seedlings* of narmwlcaf Cottonwood (Poptdas
tirtgusliJotUii and/or Pacific willow (S. /uriuV/ ssp.
laxiandru) establish on recently deposited alluvium.

Vcwtalion

Willows typically establish at the some lime as
the iree seedlings and dbminatc the stand until
the slower glowing trees overtop the shrubs.
In the Gallatin Valley we sampled narcowleaf
collonwuod/ red osier dogwood (Cornui
sericea) and nanowleaf cottomvood / herbaceous
vegetation communities. In the Unit community
red-osier efagwood is indicative of the type, but

msc species (ftaraspp.) and western snowbeny
\Sytnphoricurpos otvidentafis) may have greater
canopy cover. The herbaceous component includes
redtop {A&tvsfis sMoniJera). reed canorygrass
{Pluduris anmdinitceti). latge-leaved avens|C7evim
fnacmplty/IuHr) and Canada thistle (Cinittm
arwnse) with low to moderate cover we have
noted Minus where iced canarygrass has become
abundanl f> 25*1 cover) and dominant. The
nanowleaf cottomvood /hert>aceous vegetation
type is a default type (Hansen ct al. 1995) only
defined by having < 25*% shrub cover representing
a severely disturbed secondary succession stage of
other community types. The undcrgmwih in the
Gallatin Valley included smooth brome \Bmmux
inennisj, Kentucky blutgrass {Poa prafemish
timothy (/Vzfrmtt/'ra/titff}. quuekgross iA&tvpynm
repens)^ and western wheatgrass (Pascopynun
simtlui). Reed canarygrass is present in small
patches, suggesting these stands arc at the diy
extreme of this species moisture tolerance. The
nmious weeds leafy spuigc {Eiiphoiina aula).
knapweed* (Cenhturca spp.), uixl Canada thistle
{drsiutn urvmtt*) can be common, especially in
openings.

Eeirfo&icol Dynamics and Management
CtMtsidera! Urns

Floodplain dynamics create the typical selling for
these wetlands. River hydmlogy has been altered

:;i

through dams on miny rivets, and the floods
creating lire regeneration sites may no longer
occur Like all wetlands, maintaining the hvdrology
is key to maintaining (he wetland: changes in the
quality or quantity of water retching the wetland
will alter the vegetation and habitat values.

If livestock have access 1o this type, especially
when the soils arc wet, (hey can degrade the
habitat by trampling the vegetation, pugging the
soft soil, and exposing soil for weed colonization
sites. Livestock are active hmwscts of some or
the shrubs native to these habitats and, along with
trampling, can eliminate or change ihe composition
of ihe shiuh layer

These lloodplain locations arc particularly
vulnerable lo colonization by invasive plant

species due to the regular disturbance of soil and
Ihe deposition of seeds or other pmpagulcs by
floodwalcrs. These stands ohen have a herbaceous
layer almost totally composed of nonnaiivc species.

Stream meander can place older stanoV inclose
pmtimily lo the active channel. There can be active
erosion that may eventually eliminate some or all
of ihe stand.

HGM types and modifiers
In the Gallatin Valley wetlands we surveyed. PPO
wetlands were classified asLotic River Middle
Gradient. Floodplain Wetlands with a Throughflow
water regime (Tables 1 3 and 14).

1M? I J. TvtHcatHGbtlxf/esaMUKiatedwiihPFQ wtltaitcU.

I^indsvapc
Position

I. a (id form

Lotfc Water Sbe
Modifier

Lotk Flow
Regime

Wafer Fm

Pnlh

Lotic

Doodplain

River
Stream

Low Gradient
Moderate Gradient

Throughflow*

Tatitt M. Fwirtitttti ux&it:itilfd frith Pt wftltiikh

1 iiii- r ■■- ii

Relative
importance

Comments

1 . Surface and
Groundwater
Storage and
Slreamflow
Maintenance

Moderate to
Low*

Trees will root deeply and transpire considerable water during Ihe
growing season- Fine lex lured and high oiganic content soils ore
better

2. Nutrient Cycling

High to
Moderate

Denitrification is higher with more oigunic material in Ihe
substrate. Fine textured substrates are more effective than coarse
substrates* A fluctuating water table is best. These Lalic wetlands
ore important in preserving water quality.

3. Maintain Plant
Community

Moderate to
Low*

The undetsloty is often dominated by nonnaiivc vegetation and
the shrub uixlciMoty may be absent or diminished. M any of these
stands are decadent and dying. Regeneration can be limited due to
altered hydrology.

4. Retention
of sediments,
elements, and
compounds

High In
Moderate

These vegetated wetlands in lloodplain locations ate generally
effective for this function*

32

Tuht* 14. Cvttiiittted.

5. Shoreline
Stabilization

High

These trees with large woody mot systems occur in settings with
high erosion potential and an: very effective at stabilizing banks.

6. Terrestrial
Hihilal

High

Especially important for bint* and bats, paniculatiy cavity
ncsters* A variety of other species abo rety on this habilal.
Important food source for be a vets.

7. Aquatic Habitat

Moderate to
Low

Can provide some stream shading, woody debris, terrestrial insect
input as a food source and leaf drop us u nutrient source.

8. Conserve lion
of Wetland
Biodiversity

High

Non-wetland riparian forests are much more common. Both typos
arc in decline due to altered hydrokigic regimes and invasion
from Russian olive and salt cedar. They have also been impacted
by improper grazing regimes so siles with good shrub structure
ore relatively unoimmon.

Riverine System, Lower Perenninl

Unconsulitlated Shore 1R2US1

Summary

This type is contained within a lower perennial
river channel An unconsolidated shore musl have
<75* cover of stones or laigcr rock and <30ft
canopy cover of vegetation other than pioneering
plants. Water regime can be irregularly exposed,
irgularly flooded. iiTegulariy flooded. seasonally
flooded, temporarily flooded, intermittently
flooded, saturated or artificially Hooded. The
Unconsolidated Bottom type is similarly defined
but has a wetter water regime.

EnvinMtiticttt

Riverine Systems occur within a lower perennial
liver channel and have a disturbance regime
limiting the establishment or dense permanent
vegetation. These locations experience periodic
flood and ice scour. Fairly narrow streomsidc
Tringes. point bars and c\irnsive braided gravel
bats ore the typical sites for this type.

Vegetal Urn

By definition, vegetation is sparse. Vcgetotion
can vary annually depending on the species of
propagulcs ir aching and establishing on a site
since these environments arc regularly disturbed,
although some species may be able R> persist like
wcll-rooied woody species or reed canarygrass
{Phalarix anuulinacea^ Although spante
overall, there can be relatively dense clumps

of vegetation, usually willows (5al/xspp.).
narrowleaf collonwood iPo/ndtti tm&itslijatia)
scedlingt/saplings. and reed c an un grass {Plmtarix
antntlinacettt. Water knotweed {Polygonum
amphibium). rccltop {Agwsiis xtolonrjera). and
other herbs may occur. Knapweed (Cenhmmt
spp. muslly spotted knapweed. C bieberxtetnii)*
common mullein lYcrtmsats llutpxnx), leafy
spuige \Enphoiina esula). Canada thistle (Cirsiam
unwise} and other weedy species arc also rapid
cotnnuecs on the drier portions of these open sites.
Most of these sites would be considered to have
"sparse vegetation" (canopy cover < 10*% \ in the
National Vegetation Classification System (hence
mostly unclassified).

Ecological Dynamics and Management
CtMtsUtrra tions

Floodplain dynamics create these wetland sites.
River hydrology has been altered through dams
on many rivcis and the flooth creating these siles
may have changed in intensity' or liming. Like
all wetlands, maintaining the hydrology is key to
maintaining the wetland, changes in the quality
or quantity or water reaching the wetland will
alter the vegetation and habitat values. These
locations experience active change and the shape or
existence of any specific site can vaiy annually.

These floodplain locations are particularly
vulnerable to colonization by invasive plant species
duct* the regular disturbance of the substrate and

33

the deposition of sceth orother propagules by
(lood waters.

HGAI types and modifiers

Wetlands classified us Riverine under the NW1

system anr always Lotic River or Lotic Si trim

HGM types. In iheCallalin. R2U5 wetlands
corresponded lu Lotic River Middle Gradient,
Ploodplain Wetland* with a ThmughnYiw water
regime (Tables 15 and 16|.

Table 15. Tvt*c<JHGStl\v/esat*wafcd\riittR2UStor'lfaiab.

I :ili!-- a pi
1*0 MI- 3!

[ imdforni

Lollr Water SLrc
Modifier

I <»ik Flow
Regime

Water Flm
Path

Lotic

Floodplain

River
Sire am

Low G radical
Moderate Gradient

Throughllow*

Tuhic Iff. i Ji/ft -frivu tkWi^ndfti %\\th R2US \s*lhmii\.

1 unit Ion

Relative
Importance

Comments

1 . Surface and
Groundwater
Storage and
StirimJlnw

Munlc nance

Low

These typically coaise soils drain quickly*

2. Nutrient Cycling

Moderate to
Low

These Lotic wetlands are important in preserving water quality,
bul the coarse soils are not ideal Tor this function.

3, Maintain Plant
Community

Low*

Vcgelatinn is sparse lo absent and mnnalivc species I '.■
establish.

4. Retention
of sediments,
elements, and
compounds

Moderate to
Low*

Sediments can be retained durintz flootb and other erosion events
bul the coitsc suils ore not well suited fur retention of elements ar
compounds,

5. Shoreline
Stabilization

Moderate i >
Low

These lypes are eroded faster than vegetated types but they ore
Ihe liiM line of defense for shore stabilization*

6. Terrestrial
Habitat

Moderate to
Low

Some insccfr and birds use these habitats.

7. Aquatic Habilal

Moderate

Rocks and stones along the shoreline provide colonization
surfaces fnr macro invcjicb riles, undercut banks provide shelter
for fish

8. Conservation
of Wetland
Biodiversity

Low*

Common and non- threatened.

34

Conclusions and Recommendations

Updated USFWS NWI and itpamn mapping is
underway in areas of Mont ina where funding is
available. We recommend mapping in new areas
continues 1o associate vegetation types wilh
USFWS types, since the comprehensive, iradily

available information about these vegetation
types will help those seeking 1o better understand
or manage wetlands. Although this effort wis
confined to theCalluiin Valley, it represents
a Strang beginning in our effort* *» accrue
this information for Montana. We have also
demonstrated that it is possible to use a GIS to
attribute HCM classes lo NWl-cl ossified polygons,
especially when (he majority of wetlands occur on
Hoodplainsoil, and this permits a bmad assessment
of wetland function and potential impairment on a
landscape scale. What we have not yet determined
is the degree to which this approach can be used in

other landscapes, even in Montana. Individually
attributing HGM modifiers at a detailed level
to NWI polygons during the mapping process
represents a significant effon, but if dicholumous
keys have to be ircidiftcd for even' t incise ape
and leave enough uncertainly that 2tt to 30** of
wetlands still have to be individually classified,
there may be little time savings in a CIS-based
appmach. In our new mapping cltorfe, we will
analyze both methods to determine the costs and
benefits or adding an HGM attribute lo every NWI
polygon, but whatever the outcome of that analysts,
we recommend linking functions to NWI types
continues in same form Wcllancb arc valued (and
regulated) because of their functions and associated
values, connecting mapping wilh functions will aid
wetland mitigation, restoration, conservation, and
management.

35

Literature Cited

BridgexOutdnor Science School. No dale. A
reference guide Id wuter resources in the
Gallatin Watershed. Available from the NRCS
til: htln^/wwwmLiiRs iisdi.^ov/lrLhmc i\?a:st
vw to r% teiU/g a Lsmirce hook/index .hi ml

Brinson, M. M. 11993). "A hydrogcomorphic
classification for wetlands," Technical Report
WRP-DE-4 . U.S. Army Engineer Waterways
Experiment Station. Vicksburg. MS. NTIS No.
AD A270053.

Brinson, M.M., B.L.Swift. R. C. Plantico, and
J. S. Barclay. 1981. Riparian ccusystcim: their
ecology and status. U.S. Fish and Wildlife
Service Biological Report HI . U.S. Government
Printing Office, Washington. D. C.

Census MIX). Profile of General Demographic

Charac (eristics. Gallatin County, Montana. V.S.
Census Bureau. Washington, DC.

Cowaidin. L.M., V. Carter. F.C. Golel, and ET.
LaRoc, 1979. Classification of Wetlands and
Dcepwater Habitats of the United States.
i." 5 Department of the Interior, Pish ami
Wildlife Service, Washington, D.C. FWS/OBS-
79/31.

Grossman D.H., Faber-Langcndoen D., Weakley
A.S., Anderson M.. Bourgeron P., Crawford R..
i if..M.!m K . Lindaal 5., Mel/lerK.. PattCW n
K.D., Pync M>. Reid fct, and Sneddon L. 199H.
International classification of ecological
communities: terrestrial vegetation or the
United States, \blume I. The National
Vegetation Classification System: development,
status, and application. The Nature
Conservancy: Arlington. VA.

Hansen, P. L.. R. D. Pfistcr, K. Bogg*, B. J. Cook.
J. Joy and D. K Hinckley. 1995. Classification
and management of ManUnVs riparian and
wetland sites. Miscellaneous Publication No.
54. Missoula, MT: Univeisity of Montana.
School of Forestry, Montana Forest and
Corocrvation Station. 646 pp.

Haucr. F. R., Bradley J. Cook. Michael C.
Gilbert, Ellis J. Clairain, Jr.. and R Daniel
Smith. 2002a. "A Regional Guidebook for
Applying the Hydrogeomorphic Approach
to Assessing Wetland Functions of Riverine
Floodplains in the Northern Rocky Mountains."
ER DC/EL TR-02-2K U.S. Army Engineer
Research and Development Center. Vicksburg,
MS.

Dahl. T. E. 1990. Wetland losses in the United
Slalcs 1780s lo l9JOTs. US. Department of
Interior. Fish and Wildlife Service. Washington,
D.C.

ESRI 2005. ArcGIS 9.1.
RcdlantkCA.

E5R1 Corporation.

Finch, D. M.. and L. F. Ruggtero. 1993. Wildlife
and biological diversity in the Rocky
Mountains and northern Great Plains. Natural
Aims Journal 13:191-203.

Gallatin Local Water Quality District. 2004.
Wetland and Riparian Resource Assessment
of the Gallatin Valley. Available from the
Gallatin Local Waler Quality Resource District

Haucr. F. R., Bradley J. Cook. Michael C.
Gilbert. Ellis J. Clairain, Jr.. and R. Daniel
Smith. 2002b. "A Regional Guidebook for
Applying the Hydrogeomorphic Approach
lo Assessing Wetland Functions of
Inlermontane Prairie Pothole Welluntk in the
Northern Rocky Mountains," ERPC/ELTK-02-
7, U.S. Army Engineer Research and
Development Center, Vicksburg, MS.

Jankovsky-Jones, M., B. Bengc. F. Fink, P.
Guillory. and J. Olson. 1999a. Idaho interim
functional assessment for law-gradient broad
hasin, groundwater Ted, slope wetland* with
spring fed riverine inclusion. Prepared by the
Idaho Wetland Functional Assessment
Committee. 17 pp.

36

Jankovsky-Jones. M.. B. Bcngc, F. Fink- P.

Cuillory. and J. Olson. 1999b. Idaho interim
functional assessment for riverine wetlands on
(he lloodplains of low- to moderate g rodicnt.
2nd or 3id onler streams on fine textured
substrates. Prepared by the Idaho Wetland
Functional Assessing ni Committee. 1 H pp.

Tincr, R.W. 2003a. DicholoirotKKcys and
Mapping Codes for Wetland Landscape
Position. Landlbrm. Water Flow Path, and
Waterbady Type Descriptor U.S. Fish and
Wildlife Service, National Wcilanck Invcntoiy
Program, Nonheosl Region. Hudlcy. MA. 44
pp.

Johnson. J.B. 2005. Hydrogcomofphic wetland
profiling: An approach to landscape and
cumulative impacts analysts. EP.V620/R-
05ODI. U.S. Environmental Protection Agency,

Washington. DC,

Ketkiy. P, A. 2000, Wetland ecology: principles
and conservation. Cambridge Untvcisily Press.
Cambridge. U.K.

Montgomery, D.R. 1999. Process otamuins and the
river continuum. Journal of the American Water
Resources Association 35:397-410.

M VSR 2003. Multivariate Statistics Package.
Kovach Computing Services. Anglesey, Wales.

Patten. D.T. I99S. Riparian ecosystems ofscmi-
orkl North America: diversity and human
impacts. Wetlands 18:498-512.

Scott. M. U, S. K. Skugen, and M. F. Mcrigliano.
2003. Relating gcomorphic change and
grazing to avian communities in riparian
forests. Conservation Biology 17:284-296.

Tincr, R.W. 2003b. Correlating Enhanced

Naiiono) Wetlaixls Inventory Data with Wetland
Functions for Watcishcd Assessments: A
Rationale for Northeastern U.S. Wetlands.
VS. Fish and Wildlife Service. National
Wetland* Inventoiy Program, Region 5.
Hadley,MA, 26 pp.

Western Regional Climate Center. 2006. Data
available online at www.wrcc.dri.edu.

Shckton. D..T. Hmby. P. Johnson. K. Harper. A.
McMillan. 5. Stanley. E. Siockdale. August
2003 Draft. Freshwater Wetland* in
Washington Stale \blumc I : A Synthesis of the
Science. Washington State Department of
Ecology PuhlKaTiPii#[)3-0fr-0lP,

Tincr, R.. M. Starr. H. Bctgquist. and J, S wonts.
2000. Watershed-based Wetland
Characterization for Maryland's Nonttcoke
River and Coastal B ays Watershed;: A
Preliminary Assessment Report. U.S. Ftsh&
Wildlife Service, National Wetlands Inventory
iNWIj Program, Nonheosl Region. Hadley.
MA. Prepared for the Maryland Department
of Natural Resources, Coastal Zone
Management Program (pursuant to National
Oceanic and Atmospheric Administration
award}. NWl technical report.

37

Appendix A. Diciiotomous Key for GIS-baskd Assignment of HGM
Classifications to NWI Wetlands is the Gallatin Valley

Key A: Key lo Wetland Landscape Position

1. Wetland is classified as Riverine in NWI 3

L Wetland Ls not classified as Riverine in NWI 2

2. Wed a nd is completely or almost completely surrounded hy upland sods Terrem (T)

Go to AVv Bjor inland latuiform

2. Wetland is completely or almost completely surrounded by floodplam soils 3

3. Wetland is associated with a stream U linear or single-line watercourse on a 1:24*000 U.S.
Geological Survey topographic map) « . Go to Cgggfej "a " hcltrw

a. Wetland is the source of a stream bul this watercourse Joes not e\icnd through the

wetland Terrene <T)

Go to AVv B lor inland tmtf/bfm
a. Wetland is not the source of a stream or the watercourse extends through the

wetland Lotk Stream (LS)

Go to Couplet "/> M below
3. Wetland is associated with a river (a broad channel mapped as a polygon or 2-lincd water-
course on a 1:24,000 U.S. Geological Survey topographic map) Lotk River (LR)

Go to Couplet "b" bdow
h Stream or river associated with wetland is mapped as intermittent on a 1:24,000 USGS

map. Intermittent Gradient (IG)

Gtt to AVv D toy water How path
fc. Stream is mapped as perennial on a 1:24-000 USGS map .€

t\ Slope at stlc of wetland is greater than 4% Hfeh Gradient (HG)

Gi> to AVv Djor jtgjggjjflg path

c. Slope i\ site of wetland is less than 4'* but more than 2% Middle Gradient (MG)

Go to AVv Dfoy water flow path

i\ Slope at site of wetland is 2** or less Low Gradient (LG)

Go to Kt-\ D tor wvitcr tltnv path
3. Wetland is classified as Lacustrine in the NW ; I Lentlc (LE)

Go to Key C for Luke Type Then Go to AVv B lor inland htmlU *tvt
Key B: Key to Inland Landforms

L Wetland is Terrene and occuis on slope >4% Slope Wetland <SL)

Go to Key D for water How* path

1. Wetland is not Terrene or occurs on a slope <2% 2

2. W f etland is classified as Lacustrine in NWI or is classified as Riverine but is not completely
surrounded by a 2-lincd watcicoutscon a 1:24,000 USGS topo map Fringe Wetland (FR)

Go to Key D for water (tow path
2. Wetland is not classified as Lacustrine in NWI 3

Appendix A- I

3, Wetland is completely surrounded by walcr in a 2-lincd watercourse an a 1:24,000 USGS

lopo map Island Wcllnnri (I)

Gn to Key I) far water fl<w path

3. Wetland not as above .4

4, Wetland is completely or almost completely surrounded hy flood plain

soils „ Flwidpla In Wetland <F)

Gn to ctmplei 6 far further tamllhttn descriptors Then Gtf to Key D fag water fli w pulh

4. Wetland is surrounded by upland soils or a mixture of upland and tloodplain soils S

5. Wetland occurs in a broad intcrfluvc area between two rivers I tii> i lluve Wetland (I )

Gn to Key Dft>r water jltrw pulh

5. Wetland not as above ,6

6. Wetland exists in a distinct depression visible on aerial pholos or a DEM. or is classified as a
Paluslrinc Aquatic Bed or a Paluslrinc Unconsolidated Bed in NW1 Basin Wetland (Hi

6. Wetland is not as above Flat Wetland (FL)

Gn to Kcv D for water ft<nv paih

Key C. Kcv to Lakes

1. Waterbody is not dammed or impounded Natural Lake ( 1 )

1 1 Waterbody is dammed, impounded, or excavated Dammed I . n !-.*■

KevD: Key to Water Flow Paths

L Wetland lies in the flow palhof groundwater or surface water as determined from a DEM or
NHD flowlinc. and that water appcarc 1o pass through the wetland to another wetland or
waterbody at a lower elevation ThniQEhOcnv (TF>

L Water does not pass through this wetland to other wetlands or watcrbodics 2

2. Water flows in and out of the walcrbody through the same channel Bidirectional till)

2. Water flow is not bidirectional —J

3* There is no surface inflow or outflow to the wetland on NHD mafti. and no inflow or outflow

path is apparent on DEMs or aerial photos Isolated (I)

3 4 Wetland is not hydrologically or geographically isolated »„„3

4. Wetland lies in the flow path of groundwater or surface water as determined from a DEM or
NHD flowlinc but no outflow flowline exists and no outflow path is apparent on DEMs or
aerial pbotos Inflow (IF)

4. Wetland does not lie in the flow path of groundwater or surface water as determined from a
DEM or NHD but surface or ground water is dtschaigcd from this wetland to another wet-
land or water body at a lower elevation, as determined by an NHD flowline. a DEM or aerial
photos Oulllow(OU)

Appendix A-2

Appendix B. NWI am* HGM Conns and Descriptions Used in This Report

INWI CODE FULL DESCRIPTION

"*ABF Palustrine. Aquatic Bed, Semipermanently Flooded

PABFn Palustrine, Aquatic Bed, Semipermanently Flooded, Diked/Impounded

PABFX Palustrine. Aquatic Bed, Semipermanently Flooded, Excavated

PABGu Palustrine. Aquatic Bed, Intermittent^ Exposed. Beaver

PABGn Palustrine, Aquatic Bed, Intermittently Exposed. Diked/Impounded

PABKX Palustrine, Aquatic Bed, Artificially Flooded, Excavated

PEN! A Palustrine. Emergent Temporarily flooded

PEMAd Palustrine. Emergent Temporarily flooded. Partially Drained/Ditched

PEMAn Palustrine, Emergent Temporarily flooded, Diked/Impounded

PEMAx P alustnne. Emergent T emporanty flooded, E xcavated

PEMB Palustrine. Emergent Saturated

PEMC Palustrine. Emergent Seasonally Flooded

PEMCd Palustrine, Emergent Seasonally Flooded, Partially Drained/Ditched

EMCn Palustrine. Emergent Seasonally Flooded, Diked/Impounded

EMC)* Palustrine. Emergent Seasonally Flooded, Excavated

EMF Palustrine, Emergent Semipermanently Flooded

EMFb Palustrine, Emergent Semipermanently Flooded, Beaver

EMFn Palustrine. Emergent Semipermanently Flooded, Diked/ Impounded

EMFX Palustrine. Emergent SemipermanentJyFlooded, Excavated

PEMKX Palustrine, Emergent Artificially Flooded, Excavated

PFOA Palustrine. Forested. Temporarily flooded

PFOB Palustrine. Forested. Saturated

PFOC Palustrine. Forested. SeasonalyFlooded

PS 5 A Palustrine, Scrub-Shrub, Temporarily flooded

PSSAd Palustrine. Scrub-Shrub, Temporarily flooded. Partially Drained/Ditched

PSSAx Palustrine, Scrub-Shrub, Temporarily flooded. Excavated

P55B Palustrine. Scrub-Shrub Saturated

P5SC Palustrine. Scrub-Shrub, Seasonally Flooded

PSSCb Palustrine. Scrub-Shrub, Seasonally Flooded. Beaver

PSSCh Palustrine, Scrub-Shrub, SeasonallyFlooded. Diked/Impounded

PUBFx Palustrine, Unconsolidated Bottom. Semipermanently Flooded. E "tavated

PUBGX Palustrine. Unconsolidated Bottom. Intermittently Exposed, E xcavated

PUS A Palustrine, Unconsolidated Shore, Temporarily flooded

PUS AX Palustrine. Unconsolidated Shore. Temporarily flooded. Excavated

PUSCn Palustrine. Unconsolidated Shore. SeasonallyFlooded. Diked/Impounded

PUSCX Palustrine, Unconsolidated Shore. Seasonally Flooded. Excavated

Appendix B - /

INWI CODE FULL DESCRIPTION

R2UBF Riverine. Loner Perennial Unconsolidated Bottom. 5 errdpermanently Flooded

H2UBG Riverine. Lower Perennial, Unconsolidated Bottom. Intermittently Exposed

R2UBH Riverine, Lower Perennial, Unconsolidated Bottom. Permanently Flooded

R2USA Riverine. Loner Perennial, Unconsolidated Shore, Temporarily flooded

R2U5C Rivenne. Loner Perennial, Unconsolidated Shore, Seasonally Flooded

R3UBF Riverine, Upper Perennial, Unconsolidated Bottom. Semipermanently Flooded

R3UBH Riverine, Upper Perennial, Unconsolidated Bottom. Permanently Flooded

K3U5A Riverine, Upper Perennial. Unconsolidated Shore, Temporarily flooded

R3USC Riverine, Upper Perennial, Unconsolidated Shore. Seasonally Flooded

R45BCK Rivenne. Intermittent, Streambed, SeasonallyFlooded. Excavated

HGMCODE FULL DESCRIPTION

ftD Appears to have been drained

LRIFX Lobe River. Intermittent Floodplain, Excavated

LRLGFu Lobe River Lower Gradient Floodplain, Dammed, Beaver

LRLGFu Lobe River Loner Gradient Floodplain, Diked/impounded

LRLGFRtf Lobe River Lower Gradient Fringe. Through (low

LRLGItf Lobe River F Loner Gradient Island. Throughflow

LRMGFdl Lobe River Middle Gradient Floodplain. Diked/impounded

LRMGFRtf Lobe River. Middle Gradient Fringe. Throughflon

LSLGFb Lobe Stream, Loner Gradient Floodplain. Dammed. Beaver

LSLGFu Lobe Stream, Lower Gradient Floodplain. Diked/impounded

LSLGFdr Lobe Stream, Lower Gradient Floodplain. Drained/ ditched

LSLGFlf Lobe Stream, Lower Gradient Floodplain. Inflow

LSLGFtf Lobe Stream, Lower Gradient Floodplain. ThroughPon

tSLGFX Lobe Stream, Lower Gradient Floodplain. Excavated

SMGFu Lobe Stream, Middle Gradient Floodplain. Diked/impounded

_5MGFtf Lobe Stream, Middle Gradient Floodplam.Throughflon

FBu T errene, B asm. Diked/impounded

FBdr Terrene. Basin. Drained 'drlched

rblf T errene. B asin, Inflow

i BIT Terrene, Basin, Throughlovr

FBK T errene. B asm, E xcavated

TFtf T errene. F lat T hroughflow

TIBd Terrene. InterHuve. Basin, Diked/impounded

I I B dr T errene. Interfluve. B asm. Drained/ditched

Appendix B - 2

HGMCODE

FULL DESCRIPTION

flBi

Terrene, Interfluve. Basin. Isolated

flBif

T enene. Interfluve. 8 asm. Inflow

riBtf

Terrene. Interfluve. Basin. Thro ugh iow

riBK

T errene. Interfluve. B asm, E icavated

riFtf

Terrene. Interfluve. Flat Throughfiow

rn

Terrene. Interfluve. Isolated

fiif

Terrene. Interfluve. Inflow

AppcndixB -j