THE FATE OF NUTRIENTS IN STREAMS
This article also appears in the Oak Ridge National Laboratory Review
(Vol. 26, No. 1), a quarterly research and development magazine. If
you'd like more information about the research discussed in the article
or about the Review, or if you have any helpful comments, drop us a line. Thanks for reading the Review.
X had marked time in the limestone ledge since the Paleozoic seas
covered the land. The break came when a bur-oak root nosed down a crack
and began prying and sucking. In the flash of a century the rock
decayed, and X was pulled out and up into the world of living
things. . . . Between each of his excursions through the biota, X lay in
the soil and was carried by the rains, inch by inch, downhill. . . . One
year, while X lay in a cottonwood by the river, he was eaten by a
beaver. The beaver starved when his pond dried up during a bitter frost.
X rode the carcass down the spring freshet, losing more altitude each
hour than heretofore in a century. He ended up in the silt of a
backwater bayou, where he fed a crayfish, a coon, and then an Indian,
who laid him down to his last sleep in a mound on the riverbank. One
spring an oxbow caved the bank, and after one short week of freshet X
lay again in his ancient prison, the sea.
--Aldo
LEOPOLD, A SAND COUNTY ALMANAC
Traditionally, many people, including the eminent naturalist Aldo
Leopold, have thought of streams and rivers as little more than pipes
channeling excess water and eroded materials from the land to the sea.
This simplistic view of our waterways has been greatly expanded by
scientists in the Environmental Sciences Division (ESD) of Oak Ridge
National Laboratory, who view streams and rivers as dynamic ecosystems,
living laboratories for the study of how organisms interact with and
modify their environment. In addition to important basic questions of
ecological science, studies of streams and rivers at ORNL contribute to
our understanding of how to manage these systems to preserve or enhance
their value as drinking water sources, recreation areas, and wildlife
habitats.
A factor that affects both ecological and human health is the
concentration of various nutrients in streams. If the concentration of
a nutrient in a stream is too high, for example, it can become a matter
of public concern. Too much nitrate in stream water used for drinking
can be potentially toxic to infants. Excess phosphorus can cause algae
to grow rapidly in streams and remove dissolved oxygen needed by fish
and other organisms; as a result, manufacturers have reduced the amount
of phosphates in detergents to lower the amount of phosphorus that is
washed into streams.
So how do streams and rivers work? Why can't they be considered just
pipes transporting runoff from the land to the sea? These questions have
served as the basis for a long history of research in ESD on the
cycling--use and reuse--of nutrients in streams. This effort spanning 25
years has involved a number of current and former ORNL staff members and
guests, beginning with the studies of the late Dan Nelson and Jerry
Elwood in the 1960s and early 1970s on the uptake of phosphorus by algae
in Walker Branch, a small stream on the Oak Ridge Reservation.
Much of the focus of research over the years has been on the cycling of
phosphorus because this element is a critical biological nutrient that
often controls the productivity of plants and microbes in aquatic
ecosystems. When phosphorus availability is very low, plants and
microbes cannot grow as rapidly, restricting the productivity of
organisms higher in the food web, such as fish. In contrast, when
phosphorus availability is very high, plants and microbes may grow
excessively, resulting in the depletion of dissolved oxygen in the water
when they die and decompose.
What sets the ORNL work apart from nutrient cycling research conducted
at universities and other research institutions is the technique of
injecting short-lived radioactive isotopes of phosphorus into streams
and tracking their concentrations in the water, algae, and other stream
organisms over time and distance by measuring their radioactivity. ORNL
researchers have found that, in field experiments, radiotracers, notably
phosphorus-32 (32P), phosphorus-33 (33P), and tritium (3H), provide
powerful tools for measuring phosphorus cycling processes under natural
conditions. Because of the limitations on public access to Walker Branch
and the expertise of ESD staff in the safe use of radiotracers in
ecological studies, ORNL offers a unique environment for radiotracer
studies in natural streams.
MATERIAL SPIRALING
Probably the critical event in the development of stream nutrient
cycling research at ORNL was receiving a three-year National Science
Foundation (NSF) grant in 1978. The NSF-funded project was developed
around a new concept in stream ecology--material spiraling. This concept
was first proposed several years earlier by scientists at the University
of Georgia, but ESD scientists Jerry Elwood, Denis Newbold, and Bob
O'Neill developed its mathematical framework and applied the concept to
a wide variety of issues in stream ecology, particularly those involving
nutrient cycling.
The fundamental premise on which the concept is based is that
biologically required materials (i.e., nutrients) in stream water are
alternately taken up by organisms, most of which are attached to the
stream bottom, and released back to water many times as they are
transported downstream. In this way the downstream velocity of nutrients
is reduced relative to the flow of water and total nutrient uptake
within a given length of stream is increased (see drawing above). The
processes of biological uptake and hydrologic transport are treated
simultaneously, and the nutrient cycling characteristics of streams can
be quantified by an index, known as spiraling length, the average
distance traveled by a nutrient atom in completing one cycle--from water
to organisms and back to water. The shorter the spiraling length of a
given nutrient, the more efficiently that nutrient is used by stream
organisms, and in the case of a limiting nutrient, the higher the
productivity of the stream ecosystem.
During the first three-year NSF grant and in a three-year renewal grant,
ESD scientists determined experimentally that phosphorus does indeed
spiral (i.e., phosphorus is taken up by organisms, released back to
water, and taken up again a relatively short distance downstream). Then
they developed a field radiotracer method for measuring the spiraling
length of phosphorus. In several experiments, the ESD scientists
continuously added several millicuries of radioactive phosphate (32PO4
or 33PO4) to a stream over a 1- to 2-hour period and measured
radiotracer concentrations in the water and in organic material on the
stream bottom (see photo-graph at opening of article). Phosphorus
spiraling lengths were determined to be on the order of 100 meters in
Walker Branch. In other words, on average one atom of phosphorus would
complete its cycle every 100 meters.
At ORNL numerous studies of the roles of different types of organisms
and the effects of different environmental conditions were conducted
during the six years of NSF support for this research. Uptake of
phosphorus from water by stream organisms, particularly by microbes
attached to the surfaces of nonliving organic matter called detritus,
was found to be much more important than purely chemical processes of
phosphorus removal from stream water. Experiments also showed that
phosphorus was more tightly cycled--that is, it had shorter spiraling
lengths--in late fall and winter than in summer in Walker Branch. During
the fall, large amounts of leaves fall into the stream and are
subsequently colonized by aquatic bacteria and fungi. The leaves provide
the carbon necessary for growth of these organisms, but most of the
phosphorus to support that growth must be supplied from the stream
water, resulting in lower stream water concentrations of this nutrient
during this time of year. Other studies demonstrated that stream
organisms also altered the chemical form of phosphorus in transport.
Although inorganic forms of phosphorus are taken up, organic forms make
up a portion of the phosphorus released back to the water.
EXPERIMENTS IN LABORATORY STREAMS CONDUCTED
ORNL also received NSF support for construction of a network of four
laboratory streams used to conduct large-scale experiments under
controlled conditions. These streams, each 40 meters long and 0.3 meter
wide and made of fiberglass, were housed in a greenhouse and supplied
with water from a spring near the west end of ORNL. This type of
research facility is available at only a few other institutions in the
United States. We have used these streams for studies of the effects of
primary consumer organisms--those that eat algae or detritus--on
nutrient cycling in streams. Contrary to our initial hypothesis, we
found that primary consumers reduced the efficiency of phosphorus
cycling, increasing spiraling lengths in streams. The implication of
these results is that stream ecosystems do not develop so as to maximize
use of scarce resources, such as phosphorus. However, because of the
continuous flow of water, inefficiencies in upstream communities become
the inputs to, and resources for, communities downstream. In other
words, nutrients not used by upstream organisms are available to
downstream organisms.
In 1986 a new phase of nutrient cycling research at ORNL was initiated
with a third grant from the NSF, in part as a result of the success of
earlier work. This new project diverged somewhat from previous studies,
focusing on the relationship between nutrient cycling and the rates of
recovery of stream ecosystems from disturbances such as adding chlorine
or blocking out light. The project was conducted by a team of ESD
scientists--Don DeAngelis, Jerry Elwood, Bruce Kimmel, Tony Palumbo, and
me--as well as Alan Steinman, a University of Tennessee researcher,
Anita Parker; the project technician, and several graduate and
undergraduate students from the University of Tennessee, University of
Louisville, and Earlham College. Computer models relating nutrient
cycling and ecosystem recovery were developed, and hypotheses generated
from these models were tested experimentally in the laboratory streams.
The computer modeling work, headed by Don DeAngelis, suggested that high
rates of nutrient cycling would result in ecosystems that recovered more
slowly from disturbances because nutrients would be less available to
stream organisms when cycling was disrupted. However, in streams in
which plant-eating animals prevented the accumulation of plant biomass,
the relationship between nutrient cycling and recovery from disturbance
would be weaker because nutrient cycling should be less of a factor in
these systems.
To achieve greater replication and environmental control, the laboratory
streams were moved inside ORNL's Aquatic Ecology Laboratory and
reconfigured as eight 20-meter-long channels with overhead lights and
water reservoirs (see photograph on p. 11). Pumps to recirculate water
and heat exchangers to control water temperature were also installed. By
recirculating different fractions of the flow independently in each
stream, we could vary the incoming flux of new nutrients to each stream
while maintaining identical conditions of total flow, light, and
temperature in all streams. This degree of flexibility in operation of
the streams is unique among laboratory stream facilities.
In our first experiments we found that nutrient input had little effect
on the amount of algal biomass and productivity in the streams, but
rates of nutrient cycling were much higher in low-input streams (streams
with 90% of the flow recirculated) than in high-input streams. Our
studies suggested that the biological community could compensate for
lower input of nutrients by increasing the efficiency with which
nutrients were cycled but that this response depended on the
accumulation of sufficient biomass to promote cycling. We did not
observe greater nutrient cycling in low-input streams in which the
biomass of algae was held at low levels by the addition of herbivorous
snails.
To test the hypothesis that the rate of recovery from disturbance is
inversely related to nutrient retention and cycling, a variety of
experimental disturbances were imposed on the laboratory streams and
their effects monitored. Disturbances included a 3-hour scour,
simulating the biomass-removing effects of heavy rainfall on a stream;
a 3-month elimination of light (a "nuclear winter" scenario); chlorine
addition; and a drying up of the stream. The immediate impact of these
disturbances was related strongly to disturbance type and the amount of
algae present but only minimally to nutrient cycling. Impacts of
chemical disturbances (e.g., chlorine) were lower in streams with high
biomass (no snails), but impacts of physical disturbances (e.g.,
elimination of light) were lower in streams with low biomass (with
snails).
In terms of the rate of ecosystem recovery, the effect of nutrient
input, retention, and cycling was not consistent. In fact, only in the
case of elimination of light did the ecosystem recover more rapidly when
nutrient inputs were high and retention and cycling were low, as
hypothesized from the model results. The combination of intense
consumption of algae by snails and low nutrient input as a result of
high water recirculation resulted in the slowest rates of recovery of
the algae from most disturbances, although this was not predicted from
the model. Perhaps most importantly, these studies underscored the need
to empirically test predictions made from model simulations to evaluate
and refine those models before using them to make real-world
predictions.
Research at ORNL on nutrient cycling in streams is continuing with the
initiation of a new project in 1991, again with NSF support, to evaluate
how stream hydrodynamic features, such as variations in water velocity
and exchange rates at different points in the stream, determine the
importance of nutrient cycling and response to disturbance. Erich
Marzolf and Susan Hendricks, Oak Ridge Institute for Science and
Education postdoctoral fellows; Ramie Wilkerson, our new technician; and
several new graduate and undergraduate students have joined the project.
Our computer models have demonstrated that hydrologic storage zones
(zones in which water is generally not flowing, such as boundary layers)
strongly influence the amount of living biomass that can be supported in
streams by increasing nutrient cycling and retention (see schematic
diagrams above). A series of experimental studies has been initiated in
the laboratory streams and in several streams on the Oak Ridge
Reservation to determine the extent of these hydrologic storage zones
and their influence on nutrient cycling.
For these studies, nonreactive tracers (chloride or tritium) are
injected into stream water over several hours and a stream hydrodynamic
model is applied to the tracer data to obtain the average water velocity
and the volume of storage zones. Radioactive phosphate (33PO4) is
injected into stream water as in previous work to obtain phosphorus
uptake rates. The researchers also measure whole-stream rates of
metabolism based on changes in dissolved oxygen concentration recorded
at two stream locations over a 48-hour period. For the metabolism
measurements, propane is experimentally injected into stream water to
determine the air-water exchange rate of dissolved gases and account for
this exchange in the metabolism measurements. This work is distinct from
most other studies in stream ecology today because of its attempt to
measure characteristics and processes over an entire 50-meter-long
stream segment rather than over a square meter or so.
WALKER BRANCH WATERSHED NUTRIENTS STUDIED
Although much of our research on nutrient cycling in streams has been
supported by the NSF, DOE's Ecological Research Division (now the
Environmental Sciences Division) in the Office of Health and
Environmental Research has for many years supported studies of nutrient
cycling and transport on Walker Branch Watershed (see Michael Huston's
article in the Review, Vol. 25, No. 1, 1992, pp. 3-9). Some of this
research has focused on the mechanisms controlling the transport and
loss of nutrients from the watershed via the stream. This work has
demonstrated the importance of water pathways through the soil and
bedrock to the stream in determining rates of nutrient transport and
loss. As a result of the long history of weathering and biological
soil-forming processes uninterrupted by glaciation, soils in Walker
Branch are generally deep, with dramatic differences in geochemical,
biological, and hydrological characteristics with depth. Water moving
through upper soil layers is rather acidic (pH 4.5 to 5.5), low in
calcium (Ca2+) and magnesium (Mg2+) ion concentrations, and high in
sulfate (SO42-) ion concentrations relative to water moving through
lower soil layers or through cracks and channels in the dolomitic
bedrock. We have used these differences in flow-path chemistry, as well
as naturally occurring radon (222Rn) concentrations, to show that the
dominant water pathway through the watershed changes dramatically with
hydrologic conditions. During low-flow periods the dominant flow path to
the stream is deep, primarily coming from groundwater flowing through
bedrock cracks and cavities, and the concentrations of nitrogen and
phosphorus in this water are moderately high. However, during high-flow
periods that follow large rain events, the dominant water pathway to the
stream is shallow through the upper soil and the concentrations of
nitrogen and phosphorus in this water are low, primarily as a result of
very efficient biological removal processes by plant roots and microbes
in the upper soil layers.
Future research on nutrient cycling in Walker Branch will focus on water
pathways and nitrogen transformations in the near-stream forest, often
termed the riparian zone. This work is designed to determine whether the
microbial reduction of nitrate to gaseous forms of nitrogen is an
important mechanism for reducing transport of nitrate in groundwater to
the stream. Nitrate in streams and rivers is of particular concern
because of its toxicity, particularly to infants, at high concentrations
in drinking water.
Research on stream nutrient cycles at ORNL has demonstrated the value of
an approach that meshes computer modeling with empirical experimentation
at spatial scales ranging from indoor, laboratory streams to forested
watersheds. This work has contributed significantly to the understanding
and appreciation of streams as dynamic, biologically active ecosystems
capable of altering the amount and chemical form of nutrients and other
materials lost from watersheds. Clearly we now know that streams are
much more than passive pipes. They are an active, dynamic component in
the ecology of the landscape.
BIOGRAPHICAL SKETCH
Patrick J. Mulholland is a research staff member in the Biogeochemical
Cycling Group of ORNL's Environmental Sciences Division. He is also a
principal investigator of the NSF-supported project on nutrient cycling
in streams. A native of Elyria, Ohio, he received a Ph.D. degree in
environmental biology from the University of North Carolina at Chapel
Hill in 1979. He joined ORNL's Environmental Sciences Division in 1979.
Mulholland's 1981 paper published in Ecology was one of the first
comprehensive studies of carbon flow in a swamp ecosystem. He received
the Environmental Sciences Division's 1991 Scientific Achievement Award
for his research on the ecology of streams. Mulholland serves on the
Board of Editors for the journals Ecology and Ecological Monographs and
has recently served on the NSF's Special Review Panel for the Long-Term
Ecological Research Program. He also holds an adjunct faculty position
in the Ecology Program at the University of Tennessee.
Patrick Mulholland
(keywords: streams, rivers, nutrient cycling)
------------------------------------------------------------------------
Please send us your comments.
Date Posted: 1/11/94 (ktb)