Regionally Based Clean Water Activities: Work Plan and Proposal

dinnerworkableΠολεοδομικά Έργα

16 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

238 εμφανίσεις

Regionally Based Clean Water
Activities: Work Plan and Proposal

A Proposal Submitted to

U.S. Environmental Protection Agency

July 2005

Submitted by

Desert Research Institute

University and Community College System of Nevada


















Task A. Periphyton Distributions, Dynamics, and Environmental Controls in Nevada



Task A.1: Periphyton Distributions and Dynamics



Task A.2: Temperature Response of Photosynthesis, Respiration, and Growth



Task A.3: Reach
scale Groundwater
Surface Water Exchange as a R
egulator of
Periphyton Dynamics



Task B. Benthic Macroinvertebrate and Periphyton Communities Related to Sediment
Loading in the Lower Truckee River



Task C. High
resolution LIDAR and Hyperspectral Remote Sensing of Rivers in Western



Task C.1: Mapping Streamside Vegetation



Task C.2: Aquatic Vegetation



Task D. Simulation Modeling Studies in Support of Management to Protect Beneficial
Uses and Nutrient Criteria Development



Task E. Public Outreach



Task F. Quality Assurance Project Plan











Error! Bookmark not defined.







Conceptual model of biomass and primary producer dyn




Truckee River

August 2002. a) Biomass of periphyton and b) dissolved inorganic
nitrogen (DIN) and orthophosphorous.




Periphyton biomass on cobble substrates (expressed as chlorophyll a; Chla) in the
LTR from August 2000 to July 2001.




Periphyton biomass on cobble substrates (expressed
as chlorophyll a; Chla) in the
LTR from October 2001 to August 2002.




Dissolved oxygen concentrations at the monitoring site (Tracy) closest to the
location of the

periphyton peak shown in June 2002.




Regions of abundant periphyton growth in meandering rivers.




Carson River Basin.




Nodal structure of DSSAMt with tributaries and withdrawals for irrigation and
municipal industrial uses (Brock
et al
., 2004)




pool sequence in a sinuous channel showing definition of features.




A typical hydraulic reach in a river water quality model composed of multiple
pool units.





Numerical water quality models for dissolved oxygen and temperature that have been
applied to the Truckee and Carson river




Generalized nutrient regimes for the Truckee River.




Relation of proposed research to ongoing DRI projects.





State and local agencies in Nevada are currently under intense pressure to meet
conditions of the Clean Water Act (CWA); particularly those related to nonpoint source
pollution (Section 319[h]), impaired wate
rs (Section 303[d]) and associated total maximum
daily loads (TMDLs). Among the challenges facing the state are sparse data, inadequate
scientific basis for existing water quality standards, a general lack of decision
making tools
such as models and spatia
l analysis software, and insufficient financial resources to support
house technical staff. Discussions with state and local stakeholders (e.g., Nevada Division
of Environmental Protection, or NDEP; Pyramid Lake Paiute Tribe; and Washoe County)
along wi
th staff from U.S. Environmental Protection Agency’s (EPA) Region IX have helped
identify and prioritize a suite of water quality
related activities that address some of the
aforementioned water quality challenges. The geographic focus of these activities
three western Nevada river basins (the Truckee, Carson, and Humboldt rivers). The scientific
focus will involve a suite of laboratory and field
scale activities designed to better understand
the effects of natural and human factors on ecological f
unction in western river basins. A

unifying element for data derived from this research will be application to one or more
numerical water quality models, which will lead to improved capability to simulate future
conditions under varying management scenari


Dissolved Oxygen Dynamics in Western Nevada Rivers

At their lower elevations, rivers in the Great Basin ecoregion of western Nevada tend
to be shallow, with minimal shading from riparian vegetation. Depending on local conditions

turbidity, solar radiation can reach the bottom in riffle sections, and, provided other
conditions are favorable (temperature, nutrients), attached algae (periphyton) or rooted
vascular plants can serve as the dominant primary producers. In these riverine

most metabolic activity is associated with benthic processes because the rivers are not
sufficiently deep, nor residence time long enough, to support development of phytoplankton
communities. Significant biomass of aquatic vegetation and assoc
iated detritus can
accumulate in reaches where nutrients are sufficient and conditions such as irradiance and
bed material (substratum) are favorable. These endogenous accumulations of organic
material can lead to substandard oxygen conditions, especially
during periods of low flow
and elevated temperatures in summer.

A common trait of many water quality models that have been developed for use in
western Nevada (see Table 1) is that they simulate in
stream dissolved oxygen (DO) as a
function of periphyton
biomass dynamics (photosynthesis and respiration). A substantial
portion of the uncertainty associated with DO simulation and prediction relates to our
inability to fully characterize the complex relationships between periphyton biomass
production and exte
rnal inputs such as nutrients, temperature, hydraulics, light, grazing, and
substrate conditions.

The DO regime is a primary characteristic that defines water quality in rivers and is
determined by the magnitude of oxygen
demanding and oxygen
producing pr
ocesses and
substances that impact a parcel of water. These factors can be physical (e.g., reaeration
across the air
water interface), as well as chemical, and biological. The relative contribution


of DO
controlling factors varies among rivers based on th
eir size, channel characteristics,
and the nature of their inputs. In shallow rivers with sufficient nutrients, abundant in
growth of primary producers can lead to accumulations of organic matter and low DO
conditions, provided other factors are con

Table 1.

Numerical water quality models for dissolved oxygen and temperature that have been
applied to the Truckee and Carson rivers.


Time Period


Nevada River




1980 to 1985

in, 1987



1986 to present

et al.,





1995 to present

et al
., 1999




1994 to present

Horvath, 1996

et al
., 1997




1995 to 2002

Berris, 1996;

Taylor, 1998


e River

2003 to present

Limnotech, Inc., AquaTerra,
UNR, 2003




2003 to present

Chen and Weintraub, 2002



Our conceptual model for the balance between photosynthesis (i.e., primary
productivity) and respiration
removal in a
river segment is determined by a suite of factors
(also known as drivers) that interact in a complex fashion (see Figure 1). Photosynthesis and
removal rates are determined by drivers that are physical (e.g., irradiance,
temperature, scour, tur
bulence, and available substrate), chemical (e.g., nutrients, pH), and
















Response Feedback



Conceptual model of biomass and primary producer dynamics (RESP = Respiration and
Temp = Temperature).


biological (her
bivory and community dynamics). The net balance between productivity and
removal processes results in a standing crop of primary producer biomass that can
be a critical determinant of DO levels in a river. The primary drivers affecting primary
producer biomass in rivers are typically thought to be physical and chemical (left side of
Figure 1). However, in some systems during specific seasons, top
down control may exert a
significant effect on primary producer biomass through herbivory (grazing).

Substratum for primary producers varies with the nature of the riverine ecosystem,
and in general terms can be comprised of either bed material or biological features such as
emergent vascular plants that provide a surface for the development of epiphyti
c algae. When
substratum is dominated by bed material, its suitability as a growth medium for benthic algae
varies as a function of particle size (e.g., silt, sand, cobble) and associated mobility of the

Extensive attempts over the past few deca
des to predict primary producer dynamics
based on simple relationships among these variables have met with limited success (e.g., Bott
et al
., 1985; Dodds
et al
., 2002). However, there tends to be some general constraining
factors that affect biomass dynam
ics on a coarse resolution basis. Examples include the

Under conditions where irradiance or essential nutrients are lacking, algae will not
accumulate beyond low biomass levels. In a river system with an oligotrophic natural
lake or impoundment
(e.g., the Upper Truckee River downstream from Lake Tahoe;
the Kootenai River in northern Idaho below Libby Dam), one observes low
phosphorous conditions and low biomass of attached algae.

Rivers with elevated suspended sediment concentrations (e.g., thos
e draining
mountainous glaciated regions) or dark color (e.g., blackwater rivers) may have low
primary productivity due to the highly attenuated irradiance.

Rivers traveling through urban areas often have large inputs of dissolved chemicals
that lead to
the development of large spatial gradients in some constituents (Figures 2a and
2b). These inputs are largely due to anthropogenic inputs but can also be the result of
changes in geologic features along the river. Irrespective of the cause, these large inp
uts can
promote biostimulation in downstream reaches of the river. Elevated nutrients lead to an
increased standing crop of algae and higher trophic levels such as benthic
macroinvertebrates. Figure 2b depicts the profile of dissolved inorganic nitrogen (D
IN) and
orthophosphorous in the Truckee River during August 2002. The spike in nutrient
concentration at about 100 km is associated with loads from agricultural return drains, urban
runoff, and treated wastewater from the Reno
Sparks metropolitan area. The

typical nutrient
regime during base
flow conditions in the Truckee River is characterized in Table 2. The
upper 40
km reach (zone i) has orthophosphorous concentrations below those thought to
saturate growth of attached algae in rivers (~0.030 mg/L; Bothw
ell, 1989). Conversely, the
lower section of the river within zone iv has DIN concentrations (0.019 mg/L; Biggs, 2000)
below those levels at which biomass of non
nitrogen fixing algae may be controlled below
“excessive” levels (Welch
et al
., 1988).

The dow
nstream trend in standing crop of attached algae in the Truckee River for
August 2002 suggests an apparent biostimulatory response to nutrient loading near km 100
(Figure 2a). In this nutrient zone iii, both inorganic nitrogen and orthophosphorous are in


mple supply. Under conditions when a nutrient is limited (zones i and iv), the variability in
biomass tends to be lower than where nutrients are in ample supply (zone iii). The response
of periphyton to nutrient concentration observed on the Truckee suppor
ts Liebig’s law of the
minimum, which is fundamental to the algorithms used to numerically simulate primary
productivity in mechanistic models. According to Liebig’s law, t
he total yield, or biomass, of
an organism will be determined by the nutrient presen
t in the lowest (
concentration in relation to the requirements of that organism. In areas where nutrients are
ample (zones ii and iii), factors other than nutrients (e.g., temperature, flow, scour,
turbulence, herbivory) will limit periphyton grow
th. These d
rivers tend to exhibit a range of
conditions based on microhabitats determined by geomorphologic conditions of the fluvial

and AFDM of Attached Algae
Truckee River - 7-20 Aug 2002
Distance from Lake Tahoe (km)
Ash Free Dry Mass
Chlorophyll a
Dissolved Inorganic Nitrogen and ortho-Phosphorus
Truckee River - 6-8 Aug 2002
Distance from Lake Tahoe (km)
DIN (mg N/L)
ortho-P (mg P/L)


Truckee River

August 2002. a) Biomass of periphyton and b) dissolved ino
nitrogen (DIN) and orthophosphorous. Error bars represent

1 SE (n = 11 to 20). The
cities of Reno and Sparks as well as agricultural areas are located in the Truckee Meadows
between km 90 to 105. Nutrient zones are represented by dashed arrows (see

Table 2).


Table 2.

Generalized nutrient regimes for the Truckee River. Concentrations shown in bold are
generally considered limiting of algal growth.



Total Inorganic Nitrogen


< 0.002











< 0.019


Generally, there is a positive correlation between nutrient concentration in the water
column and benthic algal biomass. A recent comprehensive study by Tank and Dodds (2003)
illustrated the complexity of periphyto
n dynamics in streams. They reported results of
controlled experiments with nutrient
diffusing substrates simulating 10 streams with eight
different biomes representing a range of rivers from the tropics to the arctic.

They observed
threshold values of nut
rients below which nutrient limitation may be observed. However,
factors other than nutrients tend to exert a strong influence on the accumulation and
distribution of algal biomass. Nonequilibrium conditions and habitat heterogeneity in
temperate streams c
an produce environmental noise that results in a statistical variance in
nutrient relationships that is greater in flowing aquatic ecosystems compared with
lakes. In an evaluation of a large number (n = 620) of stream locations, Dodds
et al
. (2002)
ound that nutrients accounted for less that half of the variance in benthic algal biomass.
Factors such as hydraulic conditions, flow, light availability, and grazing were thought by
et al
. (2002

to be responsible for the remaining variability in be
nthic algal biomass.
Improved predictive ability was achieved in detailed studies of periphyton biomass that
accounted for nutrient concentration as well as hydrologic parameters (especially length of
time since the last flood), land use, and geology (Lohm
et al
., 1992; Biggs, 1995).

The U.S. Environmental Protection Agency (1998) initiated the process of developing
nutrient criteria for water bodies that would serve as the basis for setting total maximum
daily loads (TMDLs) for nutrients. It then develop
ed national nutrient criteria
recommendations based on ecoregions, but encouraged states and tribes to critically evaluate
and refine these recommendations at the regional level. California, Arizona, and Nevada
formulated a Regional Technical Advisory Grou
p (RTAG) for developing criteria for EPA
Region IX. One of the initial activities of the RTAG was a pilot project to evaluate regional
reference conditions for streams and rivers in aggregated Ecoregion II (Western Forested
Mountains). The results of the p
ilot project underscored the importance of refining nutrient
criteria on a regional basis, because application of the national criteria in the pilot project
resulted in significant misclassification of reference streams. (A large number of minimally
ed sites were classified as impacted using the national criteria.) In Region IX, there is
a wide range in nutrient levels found in minimally impacted aquatic systems (Tetra Tech,
2000). Development of appropriate nutrient criteria to limit algal biomass to

levels requires better understanding of the interplay between nutrient overenrichment and the
other factors that contribute to reducing algal growth and losses as illustrated by our
conceptual model (Figure 1).

Applications of ecological water

quality models to predictions of DO and periphyton
have demonstrated significant divergence when simulated results are analyzed against


observed data. The ability to model periphyton and DO in rivers has been hampered by our
lack of understanding of relat
ionships among the variables shown in Figure 1. The
uncertainties in our understanding of periphyton dynamics serve as the underlying theme in
this work plan for clean water activities in western Nevada.


With the above as background, the obj
ectives of the work to be completed under the
cooperative agreement are as follows:


to develop the physical and biological basis for improving existing models and other
management tools for use in western Nevada watersheds though a variety of
workplan task
s that address algal (periphyton) kinetics, benthic macro invertebrates;
hydrologic processes, and image analysis;


to integrate newly acquired experimental and field results using the numerical
simulation model(s) as the unifying platform;


to operate the a
mended models under a variety of input (anthropogenic and natural)


to improve the information base for evaluating and applying existing water quality
models; and


to improve the capacity of water managers to identify water quality issues related

algal growth in western streams.


The workplan elements, or tasks, described below are intended to address many of
the complex relationships noted earlier, and will lead to an improved understanding of those
factors influencing w
ater quality in western watersheds. A common theme among tasks A.1
through A.3 is that of periphyton biomass, its impact on in
stream dissolved oxygen, and
improving our understanding of physical and chemical processes that impact primary
productivity and
our associated ability to model and predict water quality under varying land

and water
use scenarios. Task B focuses on the relationship between benthic macro
invertebrate communities and suspended solids. Task C focuses on the acquisition and
analysis of

remotely sensed data within the Carson River basin. Task D describes the
integration of data and results generated under Tasks A
C through the direct application of
numerical simulation models to beneficial use and nutrient criteria issues in western Neva
Task E describes the public outreach and data sharing aspects of the project.

Currently, DRI researchers are directly involved in several applied and basic research
or monitoring programs within the proposed study areas. Table 3 provides a brief descr
of each, including potential linkages between these ongoing projects and the activities
proposed in this work plan.


Table 3.

Relation of proposed research to ongoing DRI projects.

Project Title

Linkages between Proposed
Research and Other Project

unding Agency

Baseline Monitoring for
Truckee River

Nutrient flux, primary producer
dynamics, nutrient assimilative

NDEP and Cities of Reno and Sparks

(DRI PI’s: McKay, Brock, Fritsen)

Application of
Ecosystem Function
Model to Truck
ee River

Geomorphic habitat characteristics
of channel and response of

U.S. Army Corps of Engineers

(Relevant PI’s Brock, Warwick)

Source Assessment and
Preliminary Modeling of
Thermal Loading in the
Carson River Basin

Image analysis; on
field activities in same river
reaches; acquisition of critical
temperature data

U.S EPA (NCER/STAR Grant Program)

(DRI PI’s: McKay, McGwire, Brock)

Assessment of Dissolved
Oxygen Dynamics in the
Carson River Basin

Water quality modeling;

field activities

NDEP [through 319(h) funding]

(DRI PI’s: Fritsen, Warwick)

term Water Quality
Monitoring, Truckee
River Basin

Critical long
term data sets;
compatible and overlapping field

State of Nevada

(DRI PI: McKay)

Task A. Periph
yton Distributions, Dynamics, and Environmental Controls in Nevada


During the initial development of water quality models for rivers of western Nevada
(e.g., the Lower Truckee River [LTR] and the Carson River [Nowlin, 1987; Brock
et al
1991; W
et al
., 1999]) quantitative information on periphyton biomass and dynamics
has generally been lacking. Therefore, evaluating the reciprocal interactions between water
quality and algal biomass could only be addressed indirectly (e.g., through an ana
lysis of the
river’s oxygen dynamics or nutrient budgets). In an effort to provide more direct quantitative
information on the Truckee River’s periphyton dynamics, a synoptic periphyton biomass
monitoring program was conducted during 2000 to 2001 and 2001
to 2002. This program
documented periphyton biomass and composition at 11 sites throughout the LTR on a
monthly basis and was conducted to help evaluate water quality models for the LTR that
could be used to provide TMDL evaluations. Results from the perip
hyton biomass
monitoring documented a seasonal dynamic of the periphyton biomass that included minima
of biomass standing stocks in both early spring as well as in late summer. Conversely, two
biomass maxima occurred during early to mid
summer (June to Jul
y) as well as during mid
winter (December to January) (Figures 3 and 4). The biomass maxima in early summer reached
peaks exceeding 100

g Chl


while the winter maxima were 80 to 100

g Chl


(note: biomass maxima in excess of 20

g Chl


is often considered a eutrophic system
et al
., 1989; Dodds
et al
., 1998]). The general spatial trend was for peaks to occur
ximately 10 km downstream of the confluence of Steamboat Creek and the Truckee
River (the discharge from TMWRF is into Steamboat Creek which occurs at about 110 km
downstream from Lake Tahoe). Secondary peaks in biomass occasionally were detected
m of the main peak in biomass (Figures 3 and 4).


Figure 3.

Periphyton biomass on cobble substrates (expressed as chlorophyll a; Chla) in the LTR
from August 2000 to July 2001.

Figure 4.

Periphyton biomass on cobble substrates (expressed as chlorophy
ll a; Chla) in the LTR
from October 2001 to August 2002. Distance is relative to Lake Tahoe. Reno is located at
about 100 km and Pyramid Lake is located at 180 km downstream.

The spatial distribution of the periphyton biomass in the LTR was somewhat expec
based on prior modeling results, and expectations that nutrient loadings from TMWRF,
agriculture, and groundwater would create localized growths of algae. However, the seasonal
dynamics (that included a summer minima and a second winter maxima) were no
t expected,
as they were not predicted in previous water quality simulations (McKay
et al
., 2003), nor
were they consistent with the general expectation that algal blooms are more restricted to the
spring and early summer months.

The documentation of wint
er biomass maxima (observed in both years of monitoring)
has prompted new evaluations of management plans that could allow increased loading of
nutrients in the Lower Truckee River during winter months (TMWRF, personal
communication). The rationale for suc
h a proposal has been that increased discharges and
nutrient loadings during the winter months may not deleteriously impact water quality
because periphyton may not be as physiologically capable of utilizing the released nutrients

















g cm

Aug 00

Sept 00

Oct 00

Nov 00

Dec 00





Mar 01

Apr 01

May 01

Jun 01

Jul 01




DO (mg/L)
at low temperatures. In a
ddition, the released nutrients would perhaps be transported
downstream more efficiently at colder temperatures. Moreover, oxygen solubility during the
cold temperatures of winter (2 to 5
C) is much higher than during the summer, and the river
could perha
ps handle an increase in the algal productivity during the winter even if increased
releases fueled more algal growth. These management options and their implications are
currently being evaluated by river managers and regulators. It is apparent that bette
understanding of the winter ecology and dynamics of periphyton under cold temperatures in
Nevada’s streams would be beneficial for the evaluation of suitable management strategies.

The biomass monitoring program for LTR also documented a summer peak and

a decrease in periphyton biomass during summer 2002. This decrease occurred in July and
coincided with the time when seasonal temperatures in the LTR were at their maximum.
summer die
offs and sloughing events are notorious in rivers where

occur (Whitton, 1970; Graham
et al
., 1982; Muller, 1983; Feminella and Resh, 1991; Dodds
and Gudder, 1992), and
has been a dominant algae forming biomass peaks in the
LTR (Memmott
et al.,

2002). Factors contributing to these mid
er die
offs include
inhibitory temperatures (25 to 30
C; Graham
et al
., 1982), nutrient limitation (Muller, 1983),
and increased seasonal grazing (Feminella and Resh, 1991) although

is generally
considered a poor, nonpreferred food source for fr
eshwater grazers (Bronmark
et al
., 1991;
Dodds and Gudder, 1992). Mid
summer die
offs of periphyton are of particular concern in
Nevada’s rivers because these events often lead to extremely high rates of oxygen
consumption as the biomass decays. During the

warm summer months, such rates can lead to
extremely low DO concentrations and detrimental “oxygen slumps.” Dissolved oxygen
concentrations did decrease to levels below 5 mg l

at the time of the mid
summer peak and
decline of biomass in the LTR during 2
002 (Figure 5). Because oxygen concentrations below
5 mg l

are threatening to aquatic life, further evaluation of primary producer biomass
dynamics is necessary to understand and identify potential threats to aquatic life in Nevada’s

Figure 5.

issolved oxygen concentrations at the monitoring site (Tracy) closest to the location of
the periphyton peak shown in June 2002.


Algal biomass monitoring and DO monitoring programs throughout the state of
Nevada are often not as comprehensive as the stud
ies and monitoring programs on the
Truckee River. However, DRI has recently conducted preliminary nutrient, biomass and DO
monitoring on the Carson River system. These studies have initially focused on summer
processes and have documented periphyton growth

and DO dynamics that are of concern for
Carson River water managers. Specifically, biomass of periphyton exceeding 40

g Chla cm

and DO minima well below 5 mg l
have been measured during early summer when
periphyton biomass appeared to decrease at sel
ect sites and when water temperatures
increased above 25
C. These results are consistent with the concern that the Carson River’s
numeric water quality standards being exceeded for phosphorous could lead to impacts on
other aspects of water quality (e.g.,
DO). These observations are not nearly as extensive and
robust as those conducted on the LTR. However, they further document periphyton and DO
dynamics that are of concern to Carson River stakeholders.

Task A.1: Periphyton Distributions and Dynamics

To im
prove our knowledge base of periphyton distributions, dynamics, and
composition in rivers of Nevada, we will further assess the seasonal cycle of periphyton
biomass and composition in the Carson River and conduct an initial reconnaissance of the

spatial distributions in the Humboldt River during spring and summer. Initially,
we will target periphyton composition and biomass at Riverview Park (a site located in a
reach where we have previously documented high periphyton biomass and low DO
ations during early summer) and Cradlebaugh Bridge on the Carson River (a site
with relatively high nutrient concentrations, low biomass accumulation, and relatively high
levels of DO).

We will visit these sites seasonally, and collect samples using a str
atified random
sampling design. Stratified sampling will be accomplished using riffles and pools as distinct
units whereby random sampling will occur. At each sampling location, periphyton will be
collected by appropriate methods (e.g., template for episam
mic samples or cobble scrubbing
for rocks; Porter
et al
. 1993). Water depth and velocity will be recorded at each sampling
location. For each sample collected, biomass will be determined as Ash Free Dry
pigment content, and bio
volume (determined mi
croscopically). Pigments will be determined
via HPLC, which will be calibrated with known standards. When standards are not available
for unidentified pigment peaks, absorption spectra, retention times, and molecular absorption
coefficients will be used to

identify and quantify the pigments. Microscopic examinations
will be performed on samples fixed with gluteraldehyde (0.5 to 1 percent) using an Olympus
60 microscope equipped with DIC, epifluorescence, and digital imaging capabilities.

In addition to
the monthly sampling at Riverview Park and Cradlebaugh Bridge, we
will sample at two to three upstream sites and in areas where ground
truthing of
hyperspectral remote sensing may be required (see hyperspectral remote sensing section).

Sampling on the Hum
boldt River will initially focus
on sites that NDEP has used for
water quality sampling (NDEP
Bureau of Water Quality Planning;
bwqp/monitor.htm). Specifically, sites downstream of Winnemucca and those upstream of
Battle Mountain, Carl
in, and Jiggs will be targeted in spring and summer 2005. Biomass
monitoring will follow the sampling and analytical plan as outlined above for the Carson


River monitoring. Thus, initial taxonomic information and biomass information will become
available f
or the Humboldt River system.

Task A.2: Temperature Response of Photosynthesis, Respiration, and Growth

To better determine the factors that allow winter algal blooms to develop in the LTR
(and possibly in other systems) and to document the growth regulati
on at elevated
temperatures, we will assess periphyton response to temperature variations during mid

late summer and winter.

This task will feature growth experiments conducted
in situ

on both the Carson and
Truckee rivers as well as chamber experimen
ts in the laboratory. For photosynthesis and
growth experiments in the laboratory, we will collect artificial substrates (e.g., bricks)
deployed in the study areas over the course of a time sequence. Pigment analysis from these
bricks will allow calculatio
n of
in situ

net specific growth rates for specific algal taxa (e.g.,
Brotas and Plante
Cuny, 1998; Li
et al
., 2002). Substrates will be brought to the laboratory
and monitored in recirculating test chambers, which will provide metabolic rate
(gross and net photosynthesis, and respiration) as a function of temperature
et al
., 1997). At the end of each experiment periphyton material incubated in the
chambers will be analyzed for biomass (pigments, carbon, nitrogen, and phosphorous) and
mined microscopically for taxonomic analysis.

In addition to chamber measurements, we will further evaluate the use of
C uptake
into marker algal pigments, which may allow an assessment of the specific growth rates of
differing algal taxa within the bul
k communities. This will be accomplished by incubating
stones or scraping from stones in the presence of
labeled bicarbonate in chambers and
following the incorporation of
C into specific pigments (as detected through HPLC
fractionation and liquid sc
intillation counting).

These laboratory
based studies will be coordinated with other program tasks that are
monitoring nutrients in the water sheds and those studies that are also using recirculating
growth chambers to assess periphyton metabolism under v
arying nutrient and hydraulic
constraints (see hyporheic exchange section). Chamber studies will be conducted in the
controlled environment of DRI’s Great Basin Environmental Research Laboratory (GBERL).

At different times during the year different periphy
ton communities may be dominant
at different locations along the river. Therefore, our assessments of temperature response of
production rates and growth will initially target the three dominant forms found in these
systems, namely filamentous forms, diato
m felts, and cyanobacterial mats (Fritsen, DRI,
personal observation and result of Truckee River biomass monitoring program). By obtaining
temperature response curves for differing algal taxa and communities, we will be better
suited to evaluate the underl
ying maximum rate of growth formulations that provide the
fundamental basis for algal growth formulations used in water quality models (e.g., Brock
., 1991; Chapra, 1997). To date, these formulations have been based on temperature
response of oceanic
phytoplankton (Epply, 1972), and their relevance to Nevada’s arid
stream algal taxa needs to be evaluated. Improving our knowledge of
in situ

capabilities of
periphyton taxa in semi
arid systems should not only improve water quality modeling for
Nevada’s r
iver but also capabilities for other semi
arid systems throughout the western U.S.


Task A.3: Reach
scale Groundwater
Surface Water Exchange as a Regulator of Periphyton


A defining characteristic of stream communities is high variabil
ity in the abundance
and community composition of periphyton. Periphyton dynamics have been linked in some
meandering rivers to influences by local factors including groundwater (
Triska et al
., 1989;
Williams, 1989; Brunke and Gonser, 1997; Huggenberger

., 1998; Dent
et al
., 2000;
Harvey and Wagner, 2000). Concentrated patches of green algae have been observed
downstream from zones of outwelling (discharging) groundwater located downstream from
point bars (Figure 6) (Fisher
et al
., 1998). It is postula
ted that these areas of elevated algal
biomass and productivity are linked to elevated nitrate concentration on the downgradient
side of the gravel bar.

Figure 6.

Regions of abundant periphyton growth in meandering rivers.

A series of thermal and chemical transformations occur on the scale of a riffle
meander unit sequence (Valett
et al
. 1994, 1997; Brunke and Gonser, 1997). Detrital
accumulations lead to high rates of coupled mineralization
nitrification at the head of

point bar, where there tends to be a net recharge of hyporheic exchange

(Coleman and Dahm,
1990; Boulton, 1993; Dent and Henry, 1999; Franken
et al
., 2001). Under this conceptual
model, the water exiting the point bar will tend to be cooler and elevat
ed in nitrate
concentrations relative to the water entering the bar on its leading edge. In addition to
temperature and nutrients, other factors may influence algal productivity and community
composition, including hydraulics (i.e., velocity), irradiance a
nd herbivory (grazing).


What effects do temperature and nutrients (primarily orthophosphate and nitrate)
have on periphyton growth in river meanders?

Although water velocity would seem the most likely control, in most rivers
ce is great enough that no single “velocity” occurs at any location. We suggest that
some water can “short circuit” the meander bend, acquire dissolved nutrients from beneath

Flow direction

Dominant areas

green algae growth

Zone of


the flood plain, and flow back into the river downstream into the next meander be
nd. This
overall surface water and groundwater exchange process of channel discharge and recharge
at a meander bank is termed hyporheic flow
through. Hyporheic exchange may be an
important factor shaping periphyton communities due to factors that include n
utrients and

This task is divided into four phases: (1) reconnaissance over several potential
reaches of the Truckee River to identify areas of local groundwater exchange; (2) field
measurements of groundwater

surface water exchange at sele
cted locations; (3) riffle
unit measurements of periphyton; and, (4) systematic laboratory experiments to determine
the sensitivity of temperature, water velocity, and nutrient concentration on periphyton

(1) Reconnaissance Methods

Boat Surv


A longitudinal in
stream survey will be carried out on the Truckee
River to identify groundwater discharge locations on a scale of tenths of meters (cobbles,
debris) to several tens of meters (bars, islands). We will use visual cues to identify areas
possible groundwater influence. The visual presence of macophyte beds, periphyton, and ice
free zones may suggest zones of groundwater upwelling. Temperature anomalies in rivers are
excellent tracers and have been used to infer local areas of dischargin
g groundwater
Constantz, 1998; Silliman and Booth, 1993; Woessner, 2000
). The distribution and
characteristics of aquatic vegetation can also be used to identify potential sources of surface
groundwater exchange (
White and
Hendricks, 2000).

The sur
vey will be conducted along the reach including each bank using suitable
watercraft (e.g., kayak, canoe, small catamaran) that can be easily portaged over shallow
sections. A water quality multiprobe (YSI Sonde) will be positioned at the front of the boat
record continuous conditions including temperature and specific conductivity.
Approximately every 100 m a water chemistry sample will be collected for further analysis,
if conditions warrant based on temperature and conductivity. Analysis will include n
concentrations as well as trace metal concentrations on the AAS and/or ICP
MS. This will be
completed on both river banks during the month of February to capture warmer subsurface
discharges and the month of August to capture the cooler subsurface
discharge extremes.

(2) Riffle
Pool Unit Studies of Groundwater
Surface Water Exchange

Based on the survey method described above, areas will be identified in the Truckee
River for continuous measurement of heat flux, head distributions, velocity, peri
phyton and
discrete sampling for water chemistry. At each of approximately three locations, seven
longitudinal (thalweg) and seven lateral (equipotential lines) nested piezometers will be
installed. The shallow, nested piezometers will be installed to dete
rmine local hydraulic
gradients; these data will be used to determine flowpaths within the meanders in three
directions. The piezometers will be constructed of 1.2 cm ID schedule 40 PVC to depths
ranging from 1 cm up to 2 m. Installation will be completed
utilizing an air hammer with a
land based air compressor. A drive couple will be machined for the impact hammer. A 1.2
cm solid stainless steel rod will be inserted into the PVC casing with a screw on coupling for
a stainless steel drive point. Once the de
sired depth is obtained, the inner rod will be
disengaged from the drive point and pulled out.


Measurement of temperature gradients also provides a means to examine groundwater
flow in streams (Constantz
et al.,

2003). Subsurface temperature datalogger str
ings will be
installed within the same meanders utilizing methods similar to those described for the
piezometers. The strings will be constructed of several thermocouple or thermistor sensors,
soldered and sealed, attached to a wooden or plastic rod. The t
emperature probes will be
installed at variable depths below the subsurface with increasing distance between
measurements at greater depths. Once constructed, the temperature strings will be inserted
into the casing and sand will be tremmied into the casin
g to seal the temperature string.
Between each of the thermocouples a bentonite or silicone seal will be emplaced to prevent
vertical flowpaths. Once the temperature strings are installed, the casings can either be left in
place or removed. The thermocoupl
e wires will be run through a ½” conduit pipe from the
streambed to the river bank, into a buried datalogger casing to minimize radiant temperature
variability. The temperature strings will enable us to determine groundwater velocity and
hydraulic conducti
vity through heat variability quantified in a numerical model simulation
such as VS2DH (Healy and Ronan, 1996). This process will also allow us to assess the error
between the measurement methods.

Due to the inherent difficulties of characterizing the flo
w field of groundwater in a
river’s hyporheic and parafluvial zones, our research will adopt a multiple lines of evidence
approach. Small
scale aquifer pumping tests and several spatially and temporally variable
solute injections will be conducted to estim
ate groundwater velocity, surface water flow
velocity, flow patterns, groundwater age, specific storage, porosity, hydraulic conductivity,
and dispersivity; this will allow us to estimate specific discharge through the meander bends.
The aquifer pumping te
sts will be conducted with a low stepped discharge rate with one of
the piezometers as the pumping well and the immediate piezometers as the observation wells.
The solute tracer injections will utilize either a dye (Rhodamine) or salt (NaBr) as the tracer.

The injections will be initiated at a range of locations including the middle (laterally and
vertically) of the river as well as each of the lateral sides of the river. They will also be
initiated approximately 0.5 m below the streambed. Depending on the
tracer, water samples
will be collected from various downgradient piezometers, or the piezometers will be
instrumented and continuously analyzed.

(3) Riffle
Pool Unit Measurements of Periphyton Biomass and Community Composition

Our conceptual model (Figur
e 6) suggests that the composition and biomass of
periphyton communities will develop in response to local environmental influences of
temperature and nutrients that are associated with groundwater
surface water exchange on
the scale of the riffle
pool uni
t. To evaluate these relationships, we will establish a sample
grid on the same riffle
pool units selected for the groundwater
surface water exchange
studies described above. Periphyton will be sampled using standard procedures (Porter
et al.

1993) from lo
cations for which habitat characteristics will also be determined (e.g.,
substratum type, current velocity, and depth). Our goal will be to use a sufficiently large to
be determined sample size (e.g., n = 20
40) to adequately characterize patch dynamics on

scale of a meter within the riffle
pool unit.

Resources permitting, the accrual of periphyton biomass over time will be monitored
in each of the areas and will be correlated to temperature fluctuations, velocity, and nutrient
(orthophosphate and nitr
ate) concentrations. Periphyton accrual will be measured using


procedures involving monitoring of the time
course of colonization of biomass on artificial

(4) Laboratory Experiments

Because factors that are difficult to account for (e.g., graz
ing and/or scour) may
complicate the field studies described above, systematic steady
state laboratory experiments
will be conducted to measure growth of periphyton as a function of periphyton taxonomy,
stream temperature, velocity, and nutrients. This pha
se is conducted to (1) determine the
controls on growth, and (2) determine functional relationships between growth rate and its
controls. Experiments will be conducted in controlled chambers (Dodds and Brock, 1998)
and will be repeated while varying only a

single parameter (temperature, velocity, substrate,
algal assemblage, nutrient concentration). To minimize chamber effects including growth on
chamber walls, the duration of each experiment will be between one and two days.
Experiments to determine growth

rate will be conducted in the same manner; however, in
these experiments, DO is used as a measurement of periphyton growth. Here, all secondary
factors affecting growth are optimized for maximum growth. Because sample size can
exceed available resources,
we will carefully review experimental design before proceeding
with this portion of the research plan.

The laboratory experiments will be designed to scale geometrically and dynamically
to the Truckee River. Dimensional analysis will be used to reveal the
dimensionless groups
controlling hyporheic exchange and, therefore, periphyton growth (Barenblatt, 1996). If the
laboratory results are self
similar, the functional relationships determined in the laboratory
should be true in the river as well.

Task B. B
enthic Macroinvertebrate and Periphyton Communities Related to Sediment
Loading in the Lower Truckee River


State agencies are required to develop regulatory standards for sediment loading
pursuant to section 303d of the Clean Water Act (CWA), in

terms of beneficial uses, lists of
impaired water bodies, and Total Maximum Daily Load (TMDL). Total Maximum Daily
Load standards may require periodic reevaluation to ensure that they accurately identify
healthy and impaired waters (Pahl, 2003). These sta
ndards may be developed using physical
measurements such as load duration curves, but such purely physical methods are frequently
insufficient to determine actual impacts to aquatic life, which is frequently identified as a
beneficial use. Improved methods

are needed to assess the relationship between sediment
quantity and the biotic integrity of streams and rivers. This integration can link physical
process to impairment in terms of beneficial uses and allow calibration of TMDL standards
to a biological re

The Truckee River in Nevada is listed as an impaired waterbody for turbidity,
temperature, and nutrients (NDEP, 2002). Turbidity is related to total suspended solids
(TSS), and to sediment load (Dana
et al
., 2004). Beneficial uses potentially aff
ected by
sediment on the Truckee River include municipal consumption (e.g., Chalk Bluff plant) and
measures of biotic integrity, such as the maintenance of fisheries. Although Nevada has no
beneficial use criteria for invertebrates, these organisms are a c
ritical food source for fish


species that are listed as beneficial uses for the Truckee River, and they may provide a more
robust measure of biotic integrity than fish communities (Karr and Chu, 1999).


Sediment is a ubiquitous pollutant in lotic ec
osystems, and it strongly affects benthic
macroinvertebrate (BMI) distribution in the lower Truckee River (Sada
et al.,

2005). Benthic
macroinvertebrates are frequently used to evaluate in
stream biotic integrity, as they integrate
effects of multiple and/
or cumulative stressors over time, exhibit a wide range of
susceptibilities to those stressors, and they are critical components of the food web (Karr,
1999). We propose developing BMI metrics to quantify threshold values for excessive
amounts of sediment
in the lower Truckee River. In addition, we propose examining BMI
“drift” as an easily measured behavioral endpoint of sediment impairment. This information
is needed because tolerance to sediment by BMIs is not well studied, despite the fact that

is the primary non
point source pollutant in North America (Kuhnle
et al.,

The response of BMIs to sediment may not be correlated to the commonly used biotic index
tolerance values for organic enrichment (Zweig and Rabeni, 2001). Information from b
these studies will provide the background needed to quantify a regional sediment tolerance
index for the lower Truckee River. This will a critical step toward assessing the biological
consequences of sediment in the region and facilitate development of s
cientifically based
TMDL standards. Although we plan to study BMIs on the Truckee River, knowledge gained
from this study may also be applied to other TMDL regulated rivers in our ecoregion, such as
the Carson and Walker rivers.


We will develop gu
idance for standards that are based on the tolerance of periphyton
and individual BMI species to sediment. Our studies will be based on other work showing
that each species and community type has a preference for specific habitats that can be
quantified fo
r a wide variety of parameters (e.g., sediment, dissolved oxygen, water
temperature, nutrient concentration, etc.). These preferences can be quantified to develop an
index that that is based on these requirements (Yuan and Pollard, 2005).

Broad assessment

of habitat preferences will be determined by correlating measures
of water quality and the range of physical habitats (such as substrate size, sediment
deposition, water velocity, substrate composition) to the distribution of individual periphyton
and BMI
s. This will provide guidance to accumulate additional, more quantified information
to correlate the tolerance of individual species with sediment deposition. This work should
permit the identification of indicator taxa that may be signals of factors that
community composition. This work will also assess the biological effects of very high
sediment levels on BMIs.

Using the following methods, we will examine BMIs and periphyton, and relate
sediment deposition and TSS to community composition and B
MI drift.

Field Methods

BMI Scrub Samples

Thirty BMI scrub samples will be collected in each of two seasons; autumn (baseflow
hydrograph), and early spring (pre
peak hydrograph). These samples will bracket the


seasonal biotic, hydrographic, and sediment t
ransport variability of eastern Sierra Nevada
lotic systems. Although peak discharge (May) is the time of maximum sediment transport,
BMI sampling during peak discharge is generally not possible; for this reason, we chose pre
peak hydrograph, which general
ly is also a high sediment transport season. Autumn
(baseflow) was chosen because both TSS and sediment transport are generally low at that
time (Dana
et al
., 2003). Scrub sampling will be equally allocated among riffle, glide, and
pool samples (n= 10 each

habitat x 2 seasons), to bracket the lowest to highest sediment
deposition across two seasons. Each scrub sample will consist of collecting BMIs from a

quadrat using a Hess sampler.
Habitat will be sampled using five rapid point
measures of substra
te, including B
axis size, embeddedness, and vegetative, algal mat, and
detrital depths at each site. During sampling, we will avoid disturbing substrates immediately
upstream of the Hess sampler, to facilitate periphyton and sediment core sampling. BMI
ganisms will be elutriated, preserved, and returned to the lab, using standard operating
procedures (SOPs) of the DRI macroinvertebrate laboratory.

Epilithic Algae Samples

Where possible, epilithic samples will be collected from cobbles taken in
ion with the BMI samples and will be placed in a plastic tub for processing. Whole
cobble scrubbing and surface area determinations will be repeated three times for each. The
epilithic sample will include the three rinses from each of the three cobbles alo
ng with any
periphyton adhering to the brush upon completion of the composite. Samples will be kept on
ice and in the dark until lab processing.

Subsambling for biomass

Cobble washes will be sub
sampled to determination Ash Free Dry Weight and

. These measures provide an indication of organic mass in the periphyton
assemblage. The ratio of the two measures is an indicator of the amount of organic matter
attributed to algae.

Sediment Core Sampling

The amount of sediment will be determined usin
g a stovepipe sampler and hand
pump to collect samples immediately upstream of each scrub site. After placing the stovepipe
sampler, sediments will be vigorously disturbed by hand to suspend them in the water
column. A hand pump will be used to pump 3 lite
rs of this water and suspended sediment
into a receiving bucket. Samples will then be settled in an Imhoff cone, where percentage of
fine and coarse sediments will be measured. Coarse and fine portions will be separated and
returned to the lab for analysis

BMI Drift Samples

Macroinvertebrates sometimes “drift” downstream to relocate. Drift is an indicator of
BMI stress, and has been shown to increase because of elevated sediment load and increased
discharge (e.g., Doeg and Milledge, 1991; Bond and Downes,

2003). Drift will be evaluated
for its utility to identify sediment impairment by quantifying relationships between TSS,
water quality, and sediment quantity and the abundance and habitat preference of drift
organisms. This will be accomplished by samplin
g drift organisms in representative reaches


to assess disturbance related drift, taking care to avoid sampling during times of human
activity in the stream. Drift samples will be collected for a set period of time at representative
sampling locations using

a 250 micron mesh net, and current velocity at each of these sites
will be measured to estimate BMI drift “catch per unit effort”. In addition to information
gathered during scrub samples, the amount of sediment and drift samples will be collected in
e major habitat types (riffles, glides, pools) to associate physical parameters of sediment
deposition, TSS, and discharge to taxonomic composition of drift. Drift BMIs will be
preserved and returned to the lab for taxonomic analysis, and sediment samples
will be
returned to the lab and analyzed as discussed below.

Larval Rearings and Adult Capture

Benthic macroinvertebrate tolerance to stressors is most precise when taxonomic
analysis is highly resolved. This is because species within the same genus may s
differences in tolerance to a particular stressor and identifying taxa only to genus may
provide misleading information. Most BMI larvae cannot be accurately identified to species,
as many taxonomic keys are based on adult stages. For this reason, we w
ill use a combination
of adult capture and limited pupal rearing (Diptera, Trichoptera) to determine species when
possible. Sweep nets will be used in riparian areas during scrub sampling to associate larval
and adult instars present. Laboratory rearing of

pupae collected at field sites will also be
conducted to assist with taxonomic resolution. Knowing species that are present is important
to avoid imprecise generalizations regarding sediment tolerance that is determined at the
generic level.

Water Qualit
y Sampling

Datasonde multi
water quality loggers and USGS gages will be used at upstream and
downstream ends of the study reach to bracket the turbidity, temperature, and conductivity
present during the study, and relate those parameters to TSS, using meth
ods of Dana
et al

Laboratory Methods

BMI samples

All BMI samples will be keyed to the lowest possible level of resolution (minimum
level of genus, and to species whenever possible, for all insects). Species identification will
be facilitated by
collection of adults and pupal rearing in the laboratory, which are discussed

Sediment Samples

Sediment samples will be returned to the laboratory, dried at 60

C, weighed, ashed at
C, and reweighed
for ash free dry mass and organic content an
alysis. From this series of
measurements, we will obtain a number of quantitative variables, including: percent fine and
coarse sediments, sediment dry weight density, ash free dry mass, sand per unit area, silt/clay
per unit area, and organic and inorgani
c (mineral) dry weight densities from each sediment
sampling location.


Periphyton Subsambling/Preservation for Microscopy

An aliquot of the epilithic and water quality samples will be preserved with 0.5% v/v
glutaraldehyde for microscopy. Twenty millilite
rs of the homogenized samples will then be
placed in a twenty milliliter borosilicate glass scintillation vials. After fixing with
glutaraldehyde, capped vials will be sealed with parafilm and placed in a refrigerator until
microscopic analysis. Differenti
al interference contrast (DIC) microscopy using an Olympus
60 will be used for enumeration and identification of genera. Counting methods will
follow those utilized by PhycoTech, Inc. In short, a minimum of 400 natural unit counts will
be made from a 10
0 uL subsample observed under a 25mm x 25mm cover
slip and viewed at
400x. The minimum count will be accomplished by random fields along 15 mm transects.
The side margins of the cover
slip will be avoided due to possible edge affects. For larger
taxa (> 20
0um) an additional slide will be completely enumerated at 100x. The large taxa
counts will be estimated for the area observed at 400x to allow calculation of the 400x and
100x counts. For taxa determination to the generic level, 10 ml of subsample will be
washed (HNO
) and mounted in Naphrax

for viewing at 1000x.

Periphyton taxonomic metrics will be evaluated that are most pertinent to assessing
ecological conditions. Specifically, taxa richness, diversity, the siltation index, the
eutrophication inde
x, the pollution tolerance index and the fraction of the diatom
communities comprised of
Achnanthes minutissima

(an early succession/colonizer indicator
organism) are among those that are pertinent to the assessment of disturbance and
sedimentation (Steven
son and Bahls, 1999; Hill
et al
., 2000) and are applicable to the
Truckee River (Davis and Fritsen, in review).


Field sampling will begin in September 2005, and will continue through August 2006.
Final reporting and associated journal manusc
ripts will include multivariate (gradient)
analysis of sediment data and taxonomy data (BMI and periphyton), which will delineate
community types based upon physical habitat, including sediment deposition and TSS. In
addition, a Truckee basin regional sedi
ment tolerance index for BMIs will be included, using
methods similar to those of Relyea and Minshall (2000).

Task C. High
resolution LIDAR and Hyperspectral Remote Sensing of Rivers in
Western Nevada

In June 2004, a high
resolution survey of the Carson

River in Western Nevada was
performed using Light Detection and Ranging (LIDAR) and hyperspectral imagery. These
data were collected by BAE systems on behalf of a group of stakeholders, including the
Carson Valley Conservation District (CVCD). The effort
proposed here would fund scientists
at DRI to work with these datasets and the local stakeholders to provide a set of value
analyses that will improve our ability to monitor, model, and manage such river systems in
the western U.S. LIDAR and hyperspe
ctral remote sensing represent leading edge
technologies with an unprecedented capability for precise environmental characterization.


The objectives of this workplan element are to develop and test analysis methods for
small river systems of th
e western U.S., including:



Mapping of streamside vegetation using hyperspectral imagery in conjunction with
LIDAR data; and


Using high
resolution hyperspectral remote sensing to quantify aquatic vegetation and
selected habitat parameters.

Study Area

e Carson River Basin is located in eastern California and western Nevada (Figure
7). With headwaters in the Sierra Nevada Mountains of eastern California, the Carson River
Basin is an endorheic (i.e., closed) system that terminates in the Carson Sink of th
e Great
Basin. Mean annual flow, taken from a 50
year U.S. Geological Survey (USGS) record
downstream of Carson City, is approximately 400 ft
/sec (~11 m
/sec). However, temporal
and spatial fluctuations can be extreme,
owing primarily to a combination of
climatic variability and anthropogenic
activities that include withdrawals for
agricultural and municipal and industrial
uses. One by
product of these
fluctuations can be seasonally
low flows
in the 30 ft
/sec range.

The NDEP is
currently working on a numb
er of
studies associated with water quality
standards for the designated beneficial
uses, including recreational,
agricultural, and cold water fishery uses.
These would include standards for
nutrients (N and P); total suspended
solids (TSS); dissolved oxyg
en (DO);
and temperature (T).

Figure 7.

Carson River Basin.

Presently, all portions of the Carson River in Nevada are listed as impaired due to
exceedence in one or more of these water quality standards. In response, the NDEP has
designated the Carson as

a “Focus Watershed” within its 2003 Nonpoint Source Management
Plan. An outgrowth of this priority status has been the direction of programmatic resources
towards addressing one or more of the aforementioned water quality issues. In particular,
recently c
ompleted, ongoing, or pending studies relating to phosphorous source assessment;
DO dynamics; and TSS reflect NDEP’s commitment to bringing the Carson River into
compliance with Section 303(d) of the Clean Water Act.

Stakeholders in the Carson Valley were
able to coordinate funds to contract for
LIDAR and hyperspectral image coverage for much of the Carson River floodplain from near
the state line to Lahontan Reservoir (Figure 1). After processing, this massive data
acquisition will result in approximately
500 gigabytes of data for analysis.



Light Detection and Ranging works on a principle similar to radar, in which a
coherent light beam is pulsed over the land surface and the time delay until the light is
received back at the sensor indicates the dis
tance to target. The laser from a LIDAR may
bounce off multiple objects on the land surface, such as treetops and soil surface, so the
dataset will often contain elevations for the first and last returns.

A drawback that has been identified in existing st
udies for LIDAR mapping of river
systems in the eastern U.S. is the presence of dense riverside tree and shrub canopy. For
many stretches of western rivers, this would be much less of an issue. Another potential
drawback is that the LIDAR lasers that are t
ypically used for terrestrial mapping (including
the Carson Valley survey) are in near
infrared wavelengths that are absorbed by water, so
over water bodies there are no data. Some LIDAR systems have been developed for shallow
water bathymetry, such as the

military’s SHOALS (Scanning Hydrographic Operational
Airborne LIDAR Survey) system.

Bowen and Waltermire (2002) describe the application of LIDAR to measuring river
corridor topography on a reach of the Green River in Utah. In their study, the LIDAR syst
was generally able to penetrate vegetation in relatively flat terrain to provide an accurate
surface representation, however, dense vegetation along stream banks posed problems. The
authors also attribute some error to uncertainty in the georeferencing
of the LIDAR data
relative to ground
level GPS (Global Positioning System) control points. These two issues
might be mitigated by spatial filtering methods and incorporation of image data.

Hyperspectral Imaging

Hyperspectral remote sensing is the latest g
eneration of remote sensing technology in
which the spectral resolution is so fine (narrow bandwidths covering a broad range of
wavelengths) that the actual spectral reflectance curve of the land surface is captured with
great fidelity. Most hyperspectral
remote sensing systems are airborne systems, such as the
National Aeronautics and Space Administration’s (NASA’s) AVIRIS or Earth Search
Sciences International (ESSI) Probe
1. The only (unclassified) hyperspectral satellite
imaging system to date is NASA’s

Earth Observer
1. In addition to providing greater fidelity
than standard multispectral sensors like Landsat Thematic Mapper, the high dimensionality
of the hyperspectral datasets allows new methods for mapping such as spectral mixture
modeling to be used
. Spectral mixture modeling mathematically decomposes each pixel from
an image into a proportion of its constituent components.

Research has been developing on the applications of hyperspectral remote sensing to
aquatic systems. Jakubauskas
et al
. (2000)
found that ground
based hyperspectral radiometry
was capable of measuring cover of spatterdock, an aquatic macrophyte, in Wyoming with an
r2 value as high as 0.95. However, that study did not address total biomass or other
submerged vegetation types. Legle
iter (2003) and Marcus
et al
. (2003) report findings from
hyperspectral imagery collected over the Lamar River in Wyoming. These efforts document
some success in mapping in stream habitat types (e.g., glides, pools, riffles, woody debris) in
order r
eaches, though there was a great deal of spectral heterogeneity that was not
directly attributable to the desired habitat characteristics. Williams
et al
. (2003) used
hyperspectral imagery to map two species of submerged aquatic vegetation in the Potomac
iver. While no quantitative accuracy assessment was provided with their study, Williams


. (2003) do report good visual agreement with known distributions of the two species in the

Task C.1: Mapping Streamside Vegetation

Terrestrial vegetation
in the riparian zone has a number of significant effects on
hydrologic systems, affecting bank stability, evapotranspiration rates, solar heating, and
habitat characteristics. Also, invasive plant species such as tall whitetop (
Lepidium lentiflora
and tam
arisk (
Tamarix ramossissima
) are causing significant environmental degradation
along western rivers. For objective 1, we will quantify the ability of high
hyperspectral imagery to map vegetation in the riparian corridor. Recently, a number of
udies have been exploiting the multiple returns of LIDAR data to provide information on
vegetation canopy structure, including possible classification of certain vegetation types
directly from LIDAR. We will develop methods for joint analysis of hyperspect
datasets in mapping riparian vegetation and quantify the benefits of simultaneous use of this
novel pairing of data sources. A statistically valid map accuracy assessment will be
performed with randomly sampled locations on the map product being
field checked using

Field Sampling

A field survey will be performed to develop training and testing areas for
classification of dominant streamside vegetation assemblages. Carson Valley Conservation
District staff has identified 35 plant types of int
erest in the floodplain. The field survey will
determine the dominant plant types on an approximately 20
km stretch of the Carson River,
from the confluence of the east and west forks to Mexican Dam. Tree cover in these riparian
areas is typically dominate
d by any of several species of willow (

spp.), and in well
developed riparian areas, gallery forests of Fremont's cottonwood (
Populus fremontii
sometimes occur. Shrub cover may include wildrose (

spp.), western chokecherry
Prunus virginiana
), bl
ue elderberry (
Sambucus cerulea
), and/or buffalo
berry (

Access to the Carson River by land is greatly curtailed by private land ownership, and
will be obtained through a float of the river channel. Desert Research Institute staff has

performed floats on this stretch of the river without difficulty in previous studies. The
locations of representative vegetative types along this reach will be recorded with a GPS
receiver. In addition, digital photography will be acquired for the trainin
g/testing locations
along the river. The location of photographic stations will be recorded on the GPS unit, and
timestamps on the picture files will also ensure unambiguous georeferencing of photographic
stations relative to the continuously logged GPS ro

Data Processing

The high sampling rate of hyperspectral imagery results in a high degree of cross
correlation between image bands. Also, the dimensionality of the image data is quite high
compared to the number of training pixels. Thus, classificatio
n methods must be selected that
provide effective ways for reducing the hyperspectral data volume down to a smaller number
of more informative dimensions. Three approaches to data reduction and image classification
will be tested. For the first two approac
hes, the image data will be transformed using the
minimum noise fraction (MNF) transformation (Green
et al
., 1988) that is commonly used


with hyperspectral imagery. The MNF transformation is similar to a principal components
analysis, identifying linear co
mbinations of spectral bands that provide the greatest
information content which are uncorrelated with each other. The MNF transform also has the
beneficial effect of reducing image noise in the subsequent classification. Two image
classification methods w
ill be used on the MNF
transformed data: supervised maximum
likelihood and unsupervised clustering. Image analysis will be performed using the ENVI
image analysis software from RSI Inc.

Since the MNF transform is driven by dominant patterns of variance in

the image
data, it is possible that relatively subtle spectral features may not be captured in the dominant
MNF output bands. Also, because the MNF transformation is driven by the content of a
particular image, it is not clear that it can be translated to

work well with new locations or
hyperspectral sensors. The third classification method that will be tested will address these
issues by working directly with the spectral bands of the original imagery. This classification
will be performed using a stepwis
e discriminant function analysis (DFA). The DFA will be
calculated from training samples using the SPlus statistical software package, and the DFA
will then be applied to the image dataset.

LIDAR data contains information on canopy height based on the dif
ferencing first and last
laser returns to the sensor. Original LIDAR data will be processed to extract canopy height
for the study area. Canopy height will be used in the classifiers as an additional independent


A statistically valid
map accuracy assessment will be performed, using the standard
confusion matrix

technique (Congalton, 1991). The number of test samples needed to
validate map accuracy is derived from the binomial distribution. To calculate the number of
samples required fo
r a particular confidence interval, an
a priori

expectation of map accuracy
is required. The historical standard typically used (e.g., Anderson
et al
., 1976) is 85 percent.
As taken from Hord and Brooner (1976), given an expected map accuracy of 85 percent
, 100
test samples will provide a 95 percent confidence interval of 77 to 91 percent.

Task C.2: Aquatic Vegetation

There are strong feedbacks between many water quality indicators and primary
production in aquatic vegetation. Growth rates are affected by
nutrient concentrations, stream
flow, and temperature, which in turn affect DO. For this task, we will assess the ability of
resolution hyperspectral reflectance measurements to quantify periphyton communities
in the Carson River. Periphyton are commu
nities of algae and heterotrophic microbes that are
attached to firm substrates of the river channel. In addition to being of interest for their
feedbacks on a number of important water quality indicators, these communities have been
identified as useful w
ater quality indicators themselves because they can be easily sampled
and the species composition may react quickly and in predictable ways to a number of
specific stressors. In DRI studies of the Truckee River, immediately north of the Carson
basin, t
plant material is known to be the primary agent in determining the oxygen levels
in some reaches of the river. Thus, excessive growth of periphyton can directly affect water
quality and in turn compromise the river's beneficial uses.


This objective will b
e pursued as a feasibility study using handheld spectral
measurement devices to determine if the type/amount of periphytic and macrophytic
vegetation has predictable relationships to hyperspectral reflectance. Since these aquatic
communities are dynamic an
d may have changed significantly since June 2004, it will not be
possible to use the older BAE hyperspectral imagery in a rigorous analytical manner.
However, the CVCD does have information on selected locations of where aquatic plant
communities were pres
ent at the time of the overflight.

Field Sampling

For this task, we will perform in
field reflectance measurements of periphyte and
macrophyte communities in the Carson River. Seven types of aquatic periphyte and
macrophyte communities will be sampled for

this effort. These include:

stalked or short filamentous greens

long filamentous greens

diatom felts

submerged macrophyte beds

emergent macrophyte beds

litoral macrophtyes

edgewater/backwater macrophytes (azola, duckweed)

Field sites will be sampled once
in early summer and once in mid
late summer to
capture changing flow, temperature, and turbidity conditions. If one of these communities is
not well represented on accessible stretches of the Carson River, we may find suitable sites in
the Truckee basin. S
ampling locations will be randomized in the field within each aquatic
plant community type at selected sites where there is river access. Reconnaissance of all sites
will be performed prior to fieldwork. For aquatic communities that are present at multiple

field sites, the number of samples for each community will be distributed in a manner
approximating the expected area
weighted presence of the community at each site. Once
randomized locations within each community at a field site have been selected, four

proximate samples will be taken within a radius of 5 to 10 m. These four samples will be
selected to represent the combinations of higher/lower vegetative density in shallower/deeper
water. Note that in shallow western rivers like the Truckee and Carson,
much of the river is
accessible on foot with waders during the summer.

The following measurements will be collected for each field sample:

corrected GPS locations

spectral reflectance for wavelengths from 350 to 2,500 nanometers

depth to up
permost level of biomass accumulation

depth of channel


substrate type and average cobble size

periphyte/macrophyte community type

dominant genera (or species if identifiable)

relative coverage of epiphytic algal coverage on dominant periphyt

free dry



content of chlorophyll and other pigments

The measurement routine will be codified on waterproof data entry sheets to ensure
consistency. A brief site description will be collected, including comments on atmospheri
conditions that may influence spectral reflectance measurements. Sample locations will be
recorded to submeter positional accuracy using a GPS unit with real
time satellite differential
correction (in
house). Spectral reflectance will be measured with a
FieldSpec Pro,
manufactured by Analytical Spectral Devices (Colorado). The FieldSpec measures
reflectance from 350 to 2,500 nanometers with a full
width half
maximum bandwidth of

nm and is an accepted industry standard for high
resolution field spectros
copy. Ten
spectra will be collected over each target, any obvious outliers will be removed, and the
remaining spectra for each target will be averaged. Data from the FieldSpec are time
and this can be converted to information on variations in sun a
ngle. Reflectance for the area
averaged vegetation type/density will be measured through the water column prior to
sampling, and then reflectance measurements of specific samples of vegetation will be made
as they are retrieved. In addition, samples repres
enting exposed substrates at each field site
(if present) will be brought to the water surface (minimizing disturbance) for measurement of
spectral reflectance.

Water turbidity will be measured using a YSI 600R sonde