|
Nutrient
Spiraling and Stream Metabolism
Nitrogen
(N), Phosphorous (P), and Dissolved Organic Carbon (DOC) Spiraling
Purpose and Significance
- Phosphorus and nitrogen entering the streams that feed the
NYC reservoirs are likely to be taken up and recycled at least
once and probably several times within the stream ecosystem prior
to reaching the reservoirs. Because this cycling occurs simultaneously
with downstream transport it is sometimes referred to as spiraling.
The spiraling length represents the distance over which the average
nutrient atom travels as it completes one cycle of utilization
from a dissolved available form, passes through one or more metabolic
transformations and is returned to a dissolved available form.
Quantitatively, it is the ratio of the downstream flux of nutrient
to the uptake of nutrient per unit length of stream. More intense
utilization of the nutrient, along with more effective retention
of particulate forms, shortens the spiraling length so that an
individual nutrient atom completes more
cycles in its passage through a stream-river network. Dissolved
organic carbon (DOC) undergoes similar spiraling except that its
utilization eventually results in oxidation, and the spiraling
length in this case refers to the distance traveled until oxidation.
The significance
of spiraling to the NYC watersheds relates both to the function
of the stream ecosystem itself, as well as to implications to
downstream ecosystems (the reservoirs) and the resulting water
quality. For the stream ecosystem, spiraling reflects the degree
of metabolic activity within the system, the ability of the system
to retain nutrients, and the relative utilization rates (hence
degree of nutrient limitation) among different nutrients. Spiraling
length also describes the scale on which upstream processes are
linked to downstream processes. Thus spiraling represents a fundamental
measure of stream ecosystem function. Ecosystem impairment is
likely to increase spiraling length (reduce the cycling intensity),
through reduced uptake, excessive loading, or decreased retentive
ability of the ecosystem. An exception to this rule would occur
when the increased loading of a single nutrient stimulates uptake
of a second nutrient, whose spiraling length would shorten.
The processing or
spiraling of nutrients may have a variety of implications to downstream
ecosystems. Uptake may sequester nutrients for long periods resulting
in seasonal alterations of downstream nutrient loads. Processing
may also alter the partitioning of the nutrient forms (inorganic/organic,
dissolved/particulate) with attendant implications to the availability
of the nutrient reaching the downstream system. In the case of
nitrogen, significant in-stream removal may occur through denitrification.
In the case of DOC, more intense utilization within the stream
ecosystem directly reduces the downstream
loading.
The
measurement of uptake length will provide a first step in addressing
the role of spiraling as an indicator of ecosystem function and
as a potential influence on downstream water quality. A complete
evaluation of spiraling length requires use of isotopic tracers,
but previous studies have shown that the uptake that can be observed
by incrementing the background nutrient concentrations by small
amounts provides a reasonable first approximation to the uptake
length (or distance traveled in the available form). Past studies
have also shown that uptake length is normally >90% of the
total spiraling length (Newbold 1992), a result that can be evaluated
from the fractions of dissolved and particulate nutrient in the
water column.
This section reports the uptake length of inorganic nitrogen (NH4+),
inorganic phosphorus (PO4-3), and organic carbon (glucose and
arabinose) in streams derived from whole-stream injections of
standard solutions of N, P, and C along with a conservative tracer
(bromide). Peak concentrations of the added nutrients were in
the µM range, and carbohydrates were in the nM range. Concentrations
of each constituent were measured at five stations downstream
and the longitudinal rate of nutrient uptake was estimated by
nonlinear regression analysis. The injections were performed once
annually between June and October at each of the 10 integrative
stations.
Methods - Uptake lengths for dissolved phosphate, ammonium,
glucose, and arabinose were determined by whole-stream solute
injections, following the general approach described by the Stream
Solute Workshop (1990). One
injection was made at or near each of the 10 integrative sites
(see Introduction). Each
injection involved simultaneous addition of a conservative tracer
(sodium bromide), PO4-3, NH4+,
glucose, and arabinose, at rates designed to achieve approximate
maximum concentration-elevations (after mixing) in the stream
of 30 µg/L PO4-3, 30 µg/L NH4+,
and 0.20 µM for the carbohydrates. Amendments were metered
in at a constant rate, using a peristaltic pump for time periods
ranging from 75 to 155 minutes, depending on flow and channel
characteristics. These injections were conducted simultaneously
with propane injections made for the purpose of assessing gas
exchange rates.
On
the day prior to an injection, preliminary measurements were made
of streamflow and travel times. Streamflow was measured by wading
cross sections (or from a bridge), and measuring depth and velocity
(Swoffer current meter) at 10-20 intervals. Time of travel was
estimated from the introduction of a pulse of rhodamine WT dye,
which was tracked visually. From these measurements the following
design parameters for the actual injection were determined: (i)
quantity of conservative tracer, phosphorus, nitrogen, and carbon
to be injected; (ii) duration of injection; (iii) concentrations
and metering rates for the injection solutions; (iv) longitudinal
locations of five sampling stations downstream from injections;
(v) schedule for collection of water samples from each station.
The design objective was to achieve target concentrations (with
thorough lateral mixing) of all constituents by the upstream most
sampling station, to minimize longitudinal variation in the peak
concentration of the conservative tracer, and to observe approximately
60% uptake of each nutrient within the study reach. We used an
estimated uptake mass transfer coefficient of 5x10-5 m/s to determine
the length of the study reach (i.e., distance to the downstream-most
station). A spreadsheet-based model was used to
calculate the design parameters. Where rhodamine WT dye was not
used for time of travel, all design parameters were calculated
from a time of travel prediction model derived from Year-1 parameters.
Immediately prior to an injection, background water samples for
nutrient concentrations were taken at each downstream sampling
station. Subsequent samples were taken according to the sampling
schedule relative to initiation of the injection. Five water samples
for assay of N, P, glucose, and arabinose, in addition to the
conservative tracer, were taken from each station within the period
of plateau concentration, or in the period of maximal concentrations.
Samples for N and P assay were field-filtered through a rinsed,
Whatman® 0.45 µm cellulose nitrate membrane filter and
frozen within 24 h of collection for analysis within 60 days.
Samples for glucose and arabinose assay were sterile-filtered
(0.2 µm HT Tuffryn Acrodisc®) and frozen within 24 h
for analysis within 2 months. At the upstreammost and downstream-most
sampling stations additional water samples for the conservative
tracer were collected to describe the complete passage of the
injection pulse. Additionally, samples were collected at the injection
site and at the downstream-most sampling station before, during,
and after the injection for supplemental water chemistry analyses
(NO3-N, NH4+-N, SKN, TKN, TDP,
and TP) to monitor changes in other N and P species throughout
the injections. Stream width (water surface) was measured at each
of 20 transects throughout the reach.
For a complete description of all methodologies,
data analyses, results, literature cited, and interpretations
see Chapter
8 in the Phase
I Report (5.3MB PDF).
|
|
Stream
Metabolism - Go
to the Injection/Metabolism Slide Show
Purpose and Significance
At the integrative study sites,
metabolism measurements provide data on two fundamental ecosystem
processes, primary productivity and community respiration. Primary
productivity is the rate of synthesis of plant biomass, and respiration
is an index of the utilization of reduced chemical energy, including
the metabolic costs of photosynthesis. In these study streams,
algal productivity predominates primary productivity. The goals
of these studies were to rank study sites according to the intensity
of these metabolic processes and to relate the processes to environmental
correlates. It was expected that these ecosystem functions would
be related principally to the biomasses of algae, heterotrophic
microorganisms, and (to a lesser extent) macrophytes and macroinvertebrates
and also to environmental variables including light, temperature,
and dissolved and particulate nutrients. Some of those environmental
variables, in turn, are related to watershed uses and sources
of contaminants. Changes in process rates or in the balance of
these functions over time would indicate changes in watershed
activities and signal that investigative work on upstream tributaries
and the watershed for causative factors is needed. These measures
of ecosystem function add an additional dimension beyond descriptive
variables (e.g., nutrient concentrations, invertebrate densities)
to this research program.
Methods
Field procedures - Community
metabolism was determined using open-system measurements of dissolved
O2 change. Pairs of sondes (YSI model 600XLM, Yellow
Springs Inc., Yellow Springs, OH) were deployed at the upstream
and downstream ends of each study reach to measure dissolved O2
and temperature, usually for three-day periods. The EPA-approved
600XLM sonde coupled with a rapid pulse dissolved O2
probe has a manufacturer-certified precision of 0.01 mg/L and
an accuracy of ± 0.2 mg/L. Temperature was monitored with
a manufacturer-specified precision of 0.01°C and an accuracy
of ± 0.15°C. The study reaches were those referred
to in the nutrient spiraling studies (Chapter 8) and were delimited
by a top (injection) substation, an upstream sonde substation
approximately mid-way through the reach, and a downstream sonde
substation. Conditions affecting reaeration were similar above
the upstream and the downstream sondes. For Year 3 studies, distances
between upstream and downstream sondes ranged from 150 m (Muscoot
River) to 2.2 km (Bushkill) with a mean distance of 0.80 ±
0.78 km ( x ± SD, n=10). The mean reach length was similar
to that for Year 2 (0.7 km) and shorter than mean reach length
in 2000, reflecting another low-flow year.
At the field site, the sondes
were placed in water-saturated Turkish towels (M. Lizzote, YSI,
personal communication) and calibrated according to the manufacturer’s
instructions. The sondes then were placed at a single location
in the thalweg of the study stream for an overnight (7–12
h) period prior to deployment. Differences between sondes were
used when analyzing data to calculate offsets that were applied
to the upstream-downstream approach. Two sondes each then were
transferred to the upstream and downstream substations, with pairings
based on the similarities of dissolved O2 concentrations toward
the end of the field calibration period and probe characteristics
(e.g., DO
charge and voltage). Dissolved O2 concentrations and water temperature
were measured and logged internally at 15-min. intervals. Daily
QA/QC checks were made. The QA/QC sonde was secured to the stake
holding the data sondes and after a 0.5 h equilibration period
instantaneous readings of dissolved O2, % saturation, temperature,
conductivity, and DO charge for each data sonde were checked against
the QA/QC sonde using a YSI 650MDS meter.
Above-water photosynthetically
active radiation (PAR) at each site was measured at 15 sec intervals
using two quantum sensors (Li-Cor, Lincoln, NE) secured to the
stakes holding the sondes. Each 15-min. average was logged on
a Li-COR 1400 data logger.
Reaeration coefficients were
determined empirically from a propane evasion experiment (after
Marzolf et al. 1994, Young and Huryn 1998, Marzolf et al. 1998)
performed once during each measurement period. On the day prior
to the experiment, the time of travel of water through the study
reach was estimated
using rhodamine WT. Data were used to assign sampling times for
the propane evasion experiment. For that experiment, conducted
simultaneously with the nutrient spiraling studies, propane was
bubbled from two 5-gallon tanks into the stream using gas diffuser
tubes at the injection site (a point ~ 50 – 175 m upstream
from the uppermost sampling substation). A bromide conservative
tracer solution was injected simultaneously a few cm upstream
of the propane using a peristaltic pump. Sources were mixed by
the bubbling propane and turbulence during transit from the injection
point to the upstream sampling substation. Five sampling substations
were set over the length of the study reach. The entire injection
was monitored for bromide at the first substation and at either
the fourth or fifth downstream substation (with 5 propane samples
taken at 2–10-min. intervals when concentrations were at
the plateau). Only plateau samples were collected for both propane
and bromide at the remaining substations. Field blanks (FB) were
collected at each substation prior to the start of the injection.
For each field injection, a standard curve of propane concentration
was prepared by diluting water from the
plateau (maximum propane concentration) at the uppermost sampling
substation to three lower percentages (50%, 10%, 1%) in site water
collected prior to the injection. Conservative tracer samples
were collected into 125-mL plastic bottles. Propane samples were
collected in heavy-walled 75-mL glass serum bottles that were
rubber-stoppered and crimp-sealed in the field. In Years 2 and
3, water samples were collected by immersing a bucket into the
flow in an upstream direction and then filling the sample bottles
from the bucket. This approach reduced turbulence during sampling.
Propane bottles were completely filled (no head space) and were
stored under refrigeration.
Open-system metabolism measures
include the whole system, both benthic and water column activity.
Water column metabolism was measured separately at each site as
follows. Ten BOD bottles (six light and four dark) were filled
with stream water. Initial dissolved O2 concentration, temperature,
and percent saturation were measured in each bottle using a YSI
Model 58 DO meter and probe with stirrer
suitable for use with BOD bottles. Water used for incubation in
the bottles was sparged with N2 gas to lower the percent saturation
to ~70% if initial saturation values were greater than 85%. The
bottles were then incubated in the stream for a 4 – 6 h
period. PAR was monitored during the incubation. At the end of
the incubation period the dissolved O2 concentrations were again
determined.
Streambed substrata and periphyton
in each reach were characterized by a mapping effort. Twenty transects
were set at intervals between the beginning and end of the reach.
At each transect river width was measured and 10 – 12 equidistant
lateral sampling points were set. At each point, river depth was
measured and the appearance of both the bed (substratum) and biomass
attached to the substratum (referred to as “cover type”)
were characterized using a viewing bucket. Substrata categories
roughly followed those of Hynes (1970), except that our pebble
category included material he classified as gravel.
Benthic samples for chlorophyll
a and organic mass analyses were collected from substrata constituting
10% or more of the cover types encountered during the mapping
effort. Samples of soft substrata were collected by inserting
a plastic tube (100 cm2 inner diameter) into the riverbed to isolate
a portion of sediment and removing surface sediments with a meat
baster. Samples of periphyton on rocks were scraped, brushed,
and washed into a jar. Samples were held on ice until return to
the laboratory. The planar surface area of the rock was traced
onto a piece of paper for area quantification using image analysis
techniques.
For a complete description of
all methodologies, data analyses, results, literature cited, and
interpretations please see Chapter 9 in the Phase
I Report (5.3MB PDF).
|
