PROPOSAL
Hyporheic biogeochemical processes in Arctic tundra streams:
A comparison of two streams of differing geomorphologies
Morgan Johnston
Project Advisor: Breck Bowden
Degree: M.S., Natural Resources,
Aquatic Ecology and Watershed Science
BACKGROUND
Although studies over the past fifteen years have contributed greatly to the growing knowledge base of surface-subsurface exchange and transient storage characteristics of streams, many of the streams that have been studied are temperate or desert (Valett et al. 1990, Valett et al. 1994) ecosystems. Information regarding hyporheic processes in other systems is lacking. This research will focus on the need for more complete information regarding hyporheic processes in Arctic stream ecosystems.
Initially, it was assumed that due to the presence of permafrost in the arctic tundra, hyporheic exchange was not an important process to the functioning of the tundra stream ecosystem. A study by Edwardson et al. (2003) was the first to consider hyporheic processes in arctic streams and through a series of conservative tracer additions and concurrent hyporheic and channel sampling, they found that the hyporheic zone was indeed important to stream ecosystem functioning.
Hyporheic zones in the streams in this study served as sites for the mineralization of organic matter, followed in most cases by rapid nitrification. Ammonium, phosphate and carbon dioxide were all produced in the hyporheic zone and upwelling from the hyporheic zone to the stream channel supplied the stream with nitrate, ammonium and carbon dioxide. In fact, they determined that the equivalent of 14-162% of the nitrogen demands of benthic organisms was supplied to the stream by the hyporheic zone. In this study, none of the hyporheic samples indicated anaerobic conditions (Edwardson et al. 2003).
The study proposed in this paper will follow up on the work done by Edwardson et al. (2003). This particular project is part of a larger NSF-funded arctic hyporheic study which is investigating the potential responses of arctic tundra streams to climate change. The premise of the study is that changes in the geomorphology of a stream can lead to changes in the hydrological flowpaths of a stream and surface-subsurface exchange characteristics, which can have significant implications for the microbial metabolic activity and biogeochemical cycling in the hyporheic and parafluvial zones (Holmes et al. 1998). Therefore, changes in stream geomorphology due to climate change may considerably alter the hydrologic and biogeochemical functions of the hyporheic zone in the context of the stream ecosystem.
It has been predicted that the increase in air temperature associated with global climate change will affect the sensitive arctic ecosystem by promoting the melting of permafrost. As a result of the melting permafrost, erosion and slumping are probable, which will increase sediment loading into streams and lakes (Rouse et al. 1997). Along with the sediments, the associated stored nutrients will be released into streams, with the potential of changing aquatic ecosystem structure and dynamics from the bottom up. Of major concern in these systems is increased phosphorus loading (Hobbie et al. 1999).
Specifically, in the context of arctic stream hyporheic functioning, the major concerns regarding climate change are (1) the changes in temperature, precipitation and runoff patterns that could have an effect on stream geomorphology and (2) changes in the shape and extent of the thaw bulb, which is the area of thawed permafrost around the perimeter of the streambed that develops during the warm season. If the thaw bulb increases, the extent of the hyporheic zone, which is a subset of the thaw bulb, may increase as well, possibly changing the nature and magnitude of influence that the hyporheic zone has on stream ecosystem functioning.
The arctic hyporheic research team is taking an interdisciplinary approach to studying the complex processes of arctic tundra stream hyporheic zones. The following three disciplines are represented by specialists in each area: hyporheic hydrology [Michael Gooseff (Colorado School of Mines) andJay Zarnetske, MS student (UtahStateUniversity), stream geomorphology and ground penetrating radar techniques [John Bradford and Jim McNamara (BoiseStateUniversity)]and biogeochemistry [Breck Bowden and Morgan Johnston, MS student (University of Vermont)].
The goal of my portion of this research project is to identify how transient storage and hyporheic biogeochemical processes differ in twostream reaches of differing geomorphology on the Arctic tundra landscape. The two reach types are an alluvial, high-gradient, cobble bottom reach and ameandering, low-gradient, peat-bottomed reach. In order to determine the effect of stream gradient, substrate and morphology on transient storage, the following three questions will be addressed and compared for the two stream types: (1)What portion of total transient storage is hyporheic storage? (2) What portion of the thaw bulb volume is occupied by active hyporheic processing? (3) What are the net changes in nutrient concentrations of water in transit through the hyporheic zone?
STUDY SITE
As shown in Figure 5, ToolikLake (68ºN, 149ºW) is located on the North Slope of the Brooks Range, Alaska. The lake is located about 180 km south of the Arctic Ocean (Hobbie et al. 1999). The two clear water streams in this study are a part of a series of lakes and connecting streams that flow into ToolikLake and originate from springs in the foothills of the Brooks Range (Figure 5). Although in very close proximity to each other on the landscape, the two streams have very different physical characteristics. The I-8 Inlet is an alluvial, high-gradient, cobble-bottom stream with a slope of 2%, while the Peat Inlet is a meandering, beaded, low-gradient, peat-bottom reach with a slope of 0.07%.
A.
B.
ToolikLake N
↑
Figure 5. (A) Location of Toolik Field Station, Alaska and (B) Infrared image of Toolik Lake and the Inlet series of smaller lakes and streams flowing northward into it.
(Images: Andrew W. Balser)
Research carried out by the Arctic Hyporheic Team in the 2004 field season suggests that these two streams exhibit quite different patterns of transient storage. The I-8 Inlet has been shown to have developed a larger thaw bulb during the warm season than the Peat Inlet. This is likely due to the insulation provided by the peat substrate of the stream. Additionally, stream tracer additions and corresponding models carried out in the 2004 season indicate that the solute transport patterns in I-8 Inlet are best described by a model with a power-law residence time distribution, while the patterns in the Peat Inlet are best described by a model with an exponential residence time distribution (Figure 6).
Figure 6. Results of 2004 stream tracer additions to I-8 Inlet (top panels) and the Peat Inlet (bottom panels) model with an exponential residence time distribution (left panels) and with a power-law residence time distribution (right panels) (Zarnetske et al. 2004).
JUSTIFICATION OF KEY RESEARCH QUESTIONS
Q1. What portion of total transient storage is hyporheic storage? Transient storage in a typical stream reach has an open-channel and a hyporheic component. Traditional field methods and models used to assess transient storage typically do not distinguish between these two components. To accurately assess the potential for nutrient processing and regeneration from the hyporheic zone it is necessary to distinguish between that component of the residence time distribution that is attributable to true hyporheic storage from the component that is simply open-channel storage in eddies and backwaters.
Q2. What portion of the thaw bulb volume is occupied by active hyporheic processing?
The hyporheic zone can only develop in thawed sediments. However, the volume of thawed sediments (thaw bulb) may be greater than the volume of the hyporheic zone that actively exchanges with the open-channel flow. To better understand the nature of hyporheic exchange in Arctic streams it is important to know what portion of the thaw bulb is active hyporheic zone.
Q3. What are the net changes in nutrient concentrations of water in transit through the hyporheic zone?
A portion of open channel flow typically enters the hyporheic zone at riffle heads, traverses riffles and upwells at riffle tails. The flow velocity along this path is slow relevant to the flow velocity of average open channel flow (including storage in eddies and backwaters) and extensive contact with biofilms on hyporheic sediments provides an excellent opportunity for biogeochemical transformations in water in transit through the hyporheic zone. Thus, the hyporheic zone can be a significant source for some nutrients or sink for others.
Q4. What effects do stream substrate and gradient have on hyporheic zone structure and processes?
There are a variety of different stream reach types on the North Slope of Arctic Alaska. High-gradient, alluvial, cobble-bottom reaches and low-gradient, meandering, peat-bottomed streams are two fundamentally different, contrasting stream reach types that are common. General physical hydraulic principles suggest that hyporheic processing should be deeper and more extensive in the high-gradient reaches than in the low-gradient reaches. This hypothesis is supported by early results from our hyporheic project and further suggests that net nutrient exchange is likely to differ in these two contrasting stream types.
GENERAL APPROACH
This research project has two main interrelated components. The first of these is hydrology and the second is biogeochemistry. These two components will be addressed in two field campaigns during the 2005 field season. Because the biogeochemical cycling in the hyporheic zone is so dependent on the hydrology of the system, the hydrologic aspects of the system will be addressed first in the June field campaign.
The focus of the August campaign will be on biogeochemical dynamics. Hyporheic nutrient sampling in the August campaign will focus on only those regions of the thaw bulb that are found to be actively exchanging with the stream in the June campaign. Our previous work suggests that the thaw bulb develops quite quickly (prior to June) and so waiting until August to do the more intensive biogeochemical work should not be a problem.
METHODS
Physical characterization of transient storage and hyporheic exchange
The first two questions will be addressed througha combination of solute tracerinjections andinverse solute transport modeling administered by Gooseff and Zarnetske (as described by the Stream Solute Workshop 1990) and concurrent physical sampling of hyporheic and channel transient storage zones. The conservative tracer used in these injections will be rhodamine WT (RWT), an organic, red dye whose concentrations in water samples can be determined by fluorescence at a wide range of concentrations. RWT is particularly useful as a tracer because its presence can be detected at very low concentrations (as low as 0.1µg L-1) (Sutton et al. 2001).
Based on methods used by Gücker and Boëchat (2004), I will use stream velocity data to identify in-stream transient storage zones. Prior to the study, I will survey each of the stream reaches, recording a channel cross-sectional depth profile at transects spaced ten meters apart. At each of these transects, the relative cross-sectional areas of free-flowing water (A’) and lentic water (ASS) zones will be estimated based on whether or not velocity is detectable with a velocity meter. With this information, I will be able to estimate the in-stream transient storage area relative to stream area (ASS/A’) and compare that value to the total transient storage to stream area ratio (AS/A) in order to help distinguish hyporheic and in-stream transient storage.
Additionally, the reach will also be qualitatively assessed to determine likely transient storage zones in the channel. Approximately ten to fifteen representative zones will be selected to be monitored during the solute injection. These will likely include eddies and the edge and bottom waters of pools. The pre-selected in-stream storage zones will be characterized in more detail by noting their approximate shape and volume and by qualitatively describing the patterns of water flow within each zone.
During the conservative tracer additions administered by Gooseff and Zarnetske, the concentration of tracer in the pre-selected zones will be monitored by taking multiple grab samples from the stream bank. The zones will be sampled frequently on the rising and falling limbs of the solute tracer curve and less intensively during the tracer plateau. In addition, hyporheic samples will be taken throughout the tracer plateau and falling limb. Both the in-stream transient storage and hyporheic samples will be returned to the laboratory where rhodamine WT concentration will be determined using a fluorometer (Turner Designs, 10-AU). By knowing the residence time of tracer and approximate volume of various types of transient storage zones, the residence time distribution can be compared to the transient storage parameters obtained from the transport model.
Hyporheic water samples will be obtained through the implementation of inexpensive, lightweight minipiezometers inserted into the streambed to known depths. At the beginning of the June campaign, these hyporheic samplers will be installed in one riffle-pool sequence in each study reach. Depending on ease of installation, more than one riffle per reach may be outfitted. Samplers will be installed at each of the following locations in each reach: riffle head (downwelling zone), riffle tail (upwelling zone) and pool. The samplers will be installed in nested bundles of four minipiezometers arranged at staggered depths using a steel insertion tool with a one inch diameter. The design of this insertion tool is very similar to that described in Baxter et al. (2003).
Each individual minipiezometer will consist of a rigid delrin rod (1/4”OD, 1/8”ID) in which a series of holes have been pre-drilled in the bottom 10 cm. The bottom of the rod will be left open and the drilled region will be covered with a geotextile sleeve to prevent sediments from entering the piezometer. The four piezometers will be arranged at the desired depth intervals and affixed together with electrical tape before being installed. This arrangement will depend on the depth of resistance at each point of installation.
The insertion tool will be pounded into the streambed to a depth of resistance at each of the three locations within the reach using a fence post driver. The inner steel rod will then be removed from the insertion tool and the minipiezometer bundle will be dropped into the outer casing. Before they are inserted, the piezometer bundles will be arranged so that the deepest piezometer will be installed at the deepest accessible point, the shallowest piezometer will be at approximately 30 cm depth, and the two piezometers at intermediate depths will be spaced at intervals more closely weighted toward the shallow depths. The outer casing will then be removed from the streambed using a jack and thesediments will collapse around the minipiezometer bundle, securing it in place at the desired depth.
Hyporheic samples will be drawn from the installed piezometers by applying suction via 60-ml BD syringe that is connected to the delrin rod with a 3-way stopcock. The hyporheic zone will be sampled periodically in order to note the travel time of the tracer from the point of solute addition to the riffle-head piezometer location. The piezometers at the riffle-tail position will be monitored to determine travel time through the sediments from the riffle-head position. All tracer samples will be taken to the laboratory and analyzed for rhodamine WT concentration using a fluorometer. The presence or absence of tracer at each of the depths will be noted and those data will be used to address Question 2.
Nutrient Sampling and Characterization of Biogeochemical Dynamics
The August campaign of the field season will be devoted to studying the biogeochemical dynamics of the streams’ hyporheic zones. Using the tracer data obtained in the June campaign, I will have determined the extent of the hyporheic zone and the relative contributions of in-stream and hyporheic storage to overall transient storage. Based on this information, hyporheic piezometers will be installed more densely in only the areas of the reach exhibiting active hyporheic exchange.
The piezometers installed at the beginning of the August campaign will be constructed of larger diameter (3/4” in) PVC pipe. Rather than gathering samples from various distinct depths, these piezometers will enable me to collect one composite sample from each sampling location that includes the range of depths at which hyporheic exchange is active. Individual piezometers will be installed more densely at each riffle-pool sequence and additional riffle-pool sequences within the reach will be outfitted.
Hyporheic samples will be taken for analysis of phosphate, ammonium, nitrate, dissolved oxygen (DO) and dissolved organic carbon (DOC) concentrations. Water samples will be collected in the field, filtered through a 25-mm diameter 0.45µm syringe filter and stored on ice until they are brought back to the laboratory for analysis. DO will be measured immediately in the field by connecting Tygon tubing from the three-way stopcock into a glass jar with a stopper. Another piece of Tygon tubing will extend from the glass jar to a hand pump and a third hole in the stopper will accommodate a DO probe. Pressure will be applied via the hand pump and water will be drawn into the jar without coming in contact with the atmosphere. DO will be measured with the probe. Ammonium and phosphate analyses will be carried out at the Toolik Field Station laboratory. Nitrate and DOC samples will be frozen and shipped to the University of Vermont for analysis.
The ammonium analysis will be performed using the OPA (orthophthaldialdehyde) method. In this method, fluorescence is produced when ammonium is combined with OPA. Using the OPA method, ammonium samples are combined with a working reagent containing OPA (active ingredient), sodium sulfite and a borate buffer and the fluorescence of the sample is then measured with a fluorometer. Along with the field samples, a series of ammonium standards is also prepared and measured in order to develop a standard curve from which to determine the ammonium concentrations of field samples based on the fluorometer readings. This method can be used with relatively small-volume samples (2.5 ml) and a wide range of ammonium concentrations. Holmes et al. (1999) document the details of this methodology.