VZJ sign up for etocs
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 20 November 2007
Published in Vadose Zone J 6:841-848 (2007)
DOI: 10.2136/vzj2006.0161
© 2007 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Water Content
Right arrow Capillary Fringe Processes

ORIGINAL RESEARCH

Controls on the Spatial Dimensions of Wetted Hydrologic Margins of Two Antarctic Lakes

Michael N. Gooseffa,*, John E. Barrettb, Melissa L. Northcottc, D. Brad Bated, Kenneth R. Hille, Lydia H. Zeglinf, Michael Bobbf and Cristina D. Takacs-Vesbachf

a Dep. of Civil & Environmental Engineering, Pennsylvania State Univ., University Park, PA 16802
b Dep. of Biological Sciences, Virginia Polytechnic Institute and State Univ., Blacksburg, VA 24061
c Dep. of Geology & Geological Engineering, Colorado School of Mines, Golden, CO 80401
d Environmental Studies Program, Dartmouth College, Hanover, NH 03755
e Dep. of Geography, Univ. of Colorado, Boulder, CO 80322
f Dep. of Biology, Univ. of New Mexico, Albuquerque, NM 87131
* Corresponding author (mgooseff{at}engr.psu.edu).

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


Received 3 November 2006.



   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
Polar deserts are characterized by limited water availability across much of the terrestrial landscape. In the McMurdo Dry Valleys, Antarctica, soil active layers are wetted only briefly after infrequent snowfall (<10 cm yr–1). Adjacent to glacial meltwater streams and lakes, however, sediments and soils appear visibly wetted throughout the summer. We hypothesized that this apparent wicking of water from aquatic environments is controlled by capillary action determined by the physical properties of near-shore sediments including slope, particle size distribution, and depth to permafrost. We performed synoptic sampling of near-shore sediments around two lakes in Taylor Valley, East Antarctica. We measured the water content, active layer depth, particle size distribution, and topography to determine the relative importance of shore slope, pore size distribution, and depth of the active layer on the spatial dimension of wetted hydrologic margins. The horizontal dimension (distance from the lake edge) of wetted margins (HW) and slopes ranged from 1.1 to 32.9 m and 1 to 61% and active layer depth ranged from 5.5 to >120 cm. Our results indicate that particle size distributions of near-shore sediments are homogeneous; slope had a greater effect on the dimensions of wetted margins, explaining 73% of the variance in HW. These findings show that the spatial dimensions of wetted zones are controlled by a suite of physical parameters that may be easily measured using remotely sensed data and a limited number of samples. Such information is useful for constraining landscape-scale models of nutrient redistribution because intermittently wetted soils are zones of enhanced biogeochemical transformation and transport in desert ecosystems.

Abbreviations: ELB, east lobe of Lake Bonney • LF, Lake Fryxell


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Site Description
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
There has been much attention focused on the spatial distribution and temporal dynamics of subsurface water in catchments (Niedzialek and Ogden, 2004; Seeger et al., 2004; Western et al., 2004), largely driven by the need to identify sources of water driving in situ biological activity and stream flow at broader scales. In most catchments, sediments in the riparian zones of streams and lakes are generally found to be wetter than upland soils due to the proximity of these environments to both surface water and shallow groundwater. In dry environments, broad-scale variation in the distribution of soil water controls the potential for soil microbial activity (Austin et al., 2004; Berkowitz et al., 2004; Fernandez et al., 2006) and the composition and productivity of terrestrial vegetation (Sala et al., 1988, 1992; Lauenroth et al., 1994). Thus, near-shore environments where soils and sediments are preferentially wetted are likely to be biological "hot spots" in desert environments, which may control fluxes of nutrients between terrestrial and aquatic ecosystems.

Whereas most investigations of soil water have occurred in temperate regions, polar deserts are also environments structured by the scarcity of liquid water (Kennedy, 1993). For example, in the McMurdo Dry Valleys of Antarctica, liquid water is only available to most of the landscape for a brief duration as meltwater following infrequent snowfall events of low accumulation, <10 cm yr–1 (Campbell, 2003; Witherow et al., 2006). The life history strategies of soil biota in the dry valleys appear well adapted to tolerate such conditions (Treonis et al., 2000; Weicht and Moorhead, 2004). Adjacent to glacial meltwater streams and closed-basin lakes, however, sediments and soils appear visibly wetted up to 10 m from open water throughout the austral summer. These "hydrologic margins" host distinct biological communities relative to dry upland soils (Treonis et al., 1999; Ayres et al., 2007) and exhibit enhanced rates of nutrient cycling (Barrett et al., 2002; Gooseff et al., 2003a).

We performed a synoptic surveying and sampling of wetted soils and sediments in the hydrologic margins around Lake Fryxell (LF) and the east lobe of Lake Bonney (ELB) in Taylor Valley, East Antarctica. Our goal was to characterize the controls over the dimensions of hydrologic margins around these lakes. Understanding the factors controlling the spatial extent of hydrologic margins is essential for making predictions about the potential for hydrologic and biogeochemical exchanges between terrestrial and aquatic ecosystems. We hypothesized that the apparent wicking of liquid water from lakes and streams observed during the austral summer months (November–February) is controlled by the physical properties of near-shore sediments and soils including (i) local topography or slope, (ii) particle size distribution, and (iii) depth to impermeable ice-cemented permafrost, which bounds the extent of the water table beneath near-shore sediments. We predicted that large, well-developed hydrologic margins would be associated with gentle slopes, fine grain sizes, and deeper active layers (depth to ice cement or the 2-yr mean depth of the 0°C isotherm in dry permafrost), whereas the dimensions of hydrologic margins would be limited by steep slope, coarse tills, and shallow active layers. Furthermore, we expected that prediction of hydrologic margins around lakes and streams, based at least partly on shoreline topography, might be useful for predicting margin extent and wetted subsurface habitat, perhaps when coupled with remote sensing products that provide high resolution of topographic information (i.e., airborne laser swath mapping) for remote or extraterrestrial landscapes.


   Site Description
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
The McMurdo Dry Valleys (–77.35 to –77.45 lat., 162.20 to 163.20 long.) are generally ice free, though there are a number of alpine and piedmont glaciers that surround the valleys. The climate in the dry valley region (on the western edge of the Ross Sea, East Antarctica) is a polar desert with annual air temperatures of about –20°C and <10 cm of precipitation annually (Doran et al., 2002; Witherow et al., 2006). The McMurdo Dry Valleys provide an excellent natural laboratory for studying hydrologic processes because the movement of water in these systems is not confounded by multiple, contributing factors such as extensive subsurface flow, rainfall infiltration, or redistribution of near-surface soil water by vascular vegetation. Precipitation falls during short, infrequent snowstorms, after which ablation occurs within hours, as a function of both melt and sublimation (Campbell, 2003; Gooseff et al., 2003a). Taylor Valley, the site of the McMurdo Long-Term Ecological Research Project, is bounded to the west by the Taylor Glacier, an extension of the East Antarctic Ice Sheet, and opens to McMurdo Sound to the east. There are three major closed-basin lakes in the valley (west to east): Lake Bonney (into which Taylor Glacier terminates), Lake Hoare, and Lake Fryxell (Fig. 1 ). Each of these lakes maintains a perennial ice cover, but during the austral summer a liquid moat develops around the edge of each lake. The moats range from several meters wide in deep locations to tens of meters wide in shallow locations (Lawson et al., 2004). Glacial meltwater streams flow from glaciers to lakes along well-established channels for 10 to 12 wk during the austral summer (Conovitz et al., 1998).


Figure 1
View larger version (93K):
[in this window]
[in a new window]

  		FIG. 1. Location map of the McMurdo Dry Valleys, LANDSAT image of Taylor Valley, and maps of surveyed lakes: (A) east lobe of Lake Bonney, and (B) Lake Fryxell. Circles denote transects at which all measurements were made and samples were collected, and gaps in Panel A represent surface bedrock at the lakeshore. Transect numbering is sequential around each lake.

 
Taylor Valley ecosystems occur on a heterogeneous mix of tills resulting from local glacial dynamics and exotics from the Ross Sea incursion of the West Antarctic Ice Sheet (Bockheim, 2002). The soils and sediments in the area of our study sites were inundated by glacial Lake Washburn within 5000 ybp (Doran et al., 1994). Soils and sediments are composed of unconsolidated sand; silt- and clay-sized fractions typically account for <10% of the soil by weight (Campbell and Claridge, 1987). Organic matter content is quite low, typically <0.1% organic C by weight (Burkins et al., 2001; Barrett et al., 2006). Salt content of soils and sediments varies over several orders of magnitude, while salt composition varies with proximity to the Ross Sea and till age (Bockheim, 1997). Active layer depths are generally shallow, <1 m (Conovitz et al., 2006). Because the McMurdo Dry Valleys landscape is devoid of vascular vegetation, exposed soils are conspicuously dry, except adjacent to streams and lakes, where intermittent to prolonged wetted conditions are often observed (Fig. 2A ). These visually conspicuous lakeside hydrologic margins can extend up to 10 m from shorelines, and have been observed to be stable from year to year, although the mechanisms controlling their extent remain uncharacterized.


Figure 2
View larger version (58K):
[in this window]
[in a new window]

  		FIG. 2. (A) Conceptual model of observed horizontal (HW), vertical (ZW), and lengthwise (LW) hydrologic margins adjacent to a lake and the horizontal dimension of the water table in the active layer (HWT), computed from measurements to point of refusal, (B) plan view of sampling transects, and (C) image of the hydrologic margin along the north shore of Lake Joyce, as viewed from the shoreline. Hydrologic margin is approximately 5 m in length from the shoreline.

 

   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
Results
 Discussion and Conclusions
 REFERENCES
 
Field Data Collection
We surveyed the spatial dimensions of wetted zones and the topography of near-shore environments at LF and the ELB in Taylor Valley. Soil samples were collected at regular intervals to characterize the influence of particle size distribution and salinity on hydrology. Topographic surveys of the wetted margins around LF and ELB were conducted using a roving Trimble 5700 Real Time Kinematic Global Positioning System (GPS) receiver mounted on a backpack with an antenna fixed on a survey rod. This unit was used to manually trace the edge of each lake and the edge of the hydrologic margin of both lakes. A base station transmitting unit was established on a USGS benchmark near the main camp at each lake, which continuously acquired data from the GPS. Rover data (>6000 observations per lake) were corrected by the base station data, and therefore had an estimated precision of ~1.5 cm horizontal and ~3 cm vertical for both surveys. Sampling transects were marked at approximately 100-m intervals around each lake (as indicated by real-time position read from the roving GPS unit) and surveyed for topographic position (easting, northing, and elevation) and soil moisture and active layer depth (Fig. 2B and 2C). The in-transect sites were chosen in the field to characterize a moisture gradient from Site 1, always ~20 cm from the shoreline, to Sites 2 and 3, approximately equally spaced between Site 1 and the wetted edge, to Site 4, which was approximately 1 m beyond the wetted edge of the hydrologic margin.

At each of the sites within a transect (Fig. 2B), we measured the liquid water content of the surface (10-cm) soils and sediment using a Delta T Theta Probe and meter (Dynamax, Inc., Houston, TX) and depth to refusal, using a small-diameter steel T rod, 120 cm in length, pushed in manually. We interpreted the depth to refusal to be the depth of the active layer within these hydrologic margins. Depth of refusal measurements were consistent with the depths of ice-cement layers determined by excavation of soil pits in the Lake Fryxell basin (R2 = 0.95, Fig. 3 ). Although many of the soils in the McMurdo Dry Valleys are characterized by dry permafrost (i.e., frozen unsaturated soil), the soils in the vicinity of the lakes typically have sufficient soil moisture to maintain ice cement (i.e., frozen soil near saturation) in the permafrost (Bockheim, 2002). This sampling design resulted in measurements along each of 144 transects around LF, and 59 transects around ELB. At every fifth transect on the LF survey ( about every 500 m, Fig. 1B) and every seventh transect on the ELB survey (about every 700 m, Fig. 1A), we also collected surface soil samples (to a depth of 10 cm) at each site for gravimetric water content and particle size analysis. This sampling design resulted in a total of 120 soil samples collected from the LF survey (four sites in each of 30 transects), and 36 samples from the ELB survey (four sites in each of nine transects).


Figure 3
View larger version (18K):
[in this window]
[in a new window]

  		FIG. 3. Measurements of depth to ice cement in six pits dug near Lake Fryxell correlated to adjacent depths of refusal, measured with an active layer t-bar probe.

 
Soil moisture is reported here as gravimetric soil moisture percentage. At sites where soil samples were collected, we determined gravimetric soil moisture by taking the difference of sample weights before and after drying in a laboratory oven at 105°C for >24 h, normalized to the post-drying mass. At all sites, however, we measured moisture in situ with the Delta T Theta Probe. We developed a fifth-order polynomial regression relationship between the probe measurement and the gravimetric values from the samples collected to translate all probe measurements to gravimetric soil moisture. Dried soil samples were sieved (4.5, 2, 0.425, and 0.075 mm) using a mechanical shaker for 5 min. We estimated the soil salinity of bulk soil samples (0–10-cm depth) collected from a subset of the survey plots as the electrical conductivity of a 1:5 solution of the <2-mm size fraction in deionized water using a YSI 3100 conductivity meter (Yellow Springs Instruments, Yellow Springs, OH) calibrated with a 0.01 mol L–1 KCl solution.

Data Analysis
To test our hypothesis that hydrologic margin dimensions are a function of shore slope, soil texture, and active layer size, we performed a number of regression analyses of the data. We expected that dimensions would increase (that is, there would be more extensive hydrologic margins) with lower slope, finer soil texture, and deeper active layers. From the GPS survey data, we computed the relative coordinates of the following parameters for each site and transect: hydrologic margin elevation (ZW), lateral hydrologic margin distance (HW), hydrologic margin distance along the slope (LW), elevation of the active layer at a site (as ZW minus the penetration depth at that point), lateral dimension of the water table (HWT) (Fig. 2A), which integrates the depth and slope of the active layer, and the slope of the shore transect (S). Analyses of LW trends vs. those of HW were not significantly different, so we present only those of HW here. To test whether shoreline slope influences the hydrologic margin extent, we regressed S with HW for each transect. We examined the influence of multiple factors (lake basin, slope of shoreline, and soil particle size) on HW using the stepwise multiple regression procedure in the JMP statistical software package (SAS Institute, Cary, NC).


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
Discussion and Conclusions
 REFERENCES
 
Survey Results
Hydrologic margins extended farther and were relatively more variable around LF than the ELB (Table 1). Around LF, hydrologic margin HW ranged from 1.92 to 32.91 m with a mean of 8.90 m (Fig. 4A ), and shore slopes ranged from 0.01 to 0.32 m m–1 with a mean of 0.08 m m–1 (Fig. 4B). Around the ELB, HW ranged from 1.01 to 9.17 m with a mean of 3.44 m (Fig. 5A ), and shore slopes ranged from 0.05 to 0.61 m m–1 with a mean of 0.29 m m–1 (Fig. 5B). Hydrologic margin ZW above the lake level ranged from 0.13 to 0.94 m with a mean of 0.54 m around LF, and from 0.35 to 1.96 m with a mean of 0.89 m around the ELB. Moisture gradients were, on average, –1.01 and –4.71% gravimetric moisture content per meter horizontal distance from the shoreline for LF and the ELB, respectively.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Data summary from circumnavigation surveys of Lake Fryxell (144 transects) and the east lobe of Lake Bonney (59 transects). All data represented as means ± standard deviations of hydrologic margin dimensions, where HW is the lateral distance from shore, LW is the distance along the slope, and ZW is the elevation above the shoreline.

 

Figure 4
View larger version (29K):
[in this window]
[in a new window]

  		FIG. 4. Calculated (A) hydrologic margin horizontal lengths (HW) and (B) transect slopes (S) from circumnavigation survey of Lake Fryxell. Small plots within each panel are the normalized frequency distributions for each data set. Order of survey transect numbering noted between panels (see Fig. 1B).

 

Figure 5
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 5. Calculated (A) hydrologic margin horizontal lengths (HW) and (B) transect slopes (S) from circumnavigation survey of the east lobe of Lake Bonney. Small plots within each panel are the normalized frequency distributions for each data set. Order of survey transect numbering noted between panels (see Fig. 1A).

 
Particle size distributions showed no significant variation with distance from the lake edge (i.e., no significant influence of transect position) for either lake (Fig. 6 ); however, there were significant differences in particle size distribution between samples collected from LF and the ELB. Samples from LF were enriched in the <75-µm fraction relative to samples from the ELB (P = 0.002, two-tailed t-test assuming unequal variance), particularly in transect positions closest to the lake edge (Fig. 7 ), while samples from the ELB had a significantly greater coarse sand fraction relative to LF (P < 0.001, two-tailed t-test assuming unequal variance). In a multiple regression analysis, the proportion of coarse size fraction of the samples and slope explained 88% of the variability of observed HW values around the ELB (Table 2). No clear influence of soil texture on HW was apparent for LF.


Figure 6
View larger version (12K):
[in this window]
[in a new window]

  		FIG. 6. Mean sediment size distribution for all samples, grouped by site within each transect for (A) Lake Fryxell, and (B) the east lobe of Lake Bonney.

 

Figure 7
View larger version (29K):
[in this window]
[in a new window]

  		FIG. 7. Horizontal position as a function of sediment size proportions for (A and B) >4.5 mm, (C and D) <4.5 but >2 mm, (E and F) <2 mm but >425 µm, (G and H) <425 but >75 µm, and (I and J) <75 µm.

 

View this table:
[in this window]
[in a new window]

  		TABLE 2. Stepwise multiple regression for the horizontal dimension of the hydrologic margin from the lake edge (HW) at the east lobe of Lake Bonney. The best-fit model accounts for 88% of the variability in HW as a function of slope and the coarse sand fraction (sum of squared error of full model = 1.94, MSE = 0.32).

 
The HW values were found to be well correlated with transect slope (R2 = 0.73) as a power-law function for the combined data sets from the two lakes, representing a broad range of shore slopes (Fig. 8 ). Because there is no clear relationship between ZW and soil texture, we regressed ZW vs. HW (Fig. 9 ), which suggests little to no relationship around LF but a moderate relationship around the ELB, with HW explaining 53% of the variability in ZW.


Figure 8
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 8. Regression analysis of lateral hydrologic margin dimension (HW) as a function of shoreline slope (S) for transects measured around Lake Fryxell (LF) and the east lobe of Lake Bonney (ELB). Best-fit regressions based on highest R2 for LF data alone is HW = 0.55S–1.02, R2 = 0.73; and for ELB data alone is HW = 6.48exp(–2.70S), R2 = 0.31.

 

Figure 9
View larger version (20K):
[in this window]
[in a new window]

  		FIG. 9. Relationships between the lateral dimension (HW) and vertical extent (ZW) of hydrologic margins around A) Lake Fryxell, and B) the east lobe of Lake Bonney. See Fig. 2 for a diagram of dimensions.

 
The active layer depths at each position within a transect were greater and more variable at the ELB than at LF (Table 1) and slopes around the ELB were greater than those around LF. Water tables under these hydrologic margins provide a nearby reservoir to the extended soils, from which moisture can be wicked. Thus, deeper active layers, which can provide longer HWT dimensions, may serve as an important control on the observed surface soil moisture distribution. Active layer depths range from a mean of 0.42 m near the shore to 0.28 m in the dry soils around LF, and from 0.43 to 0.33 m around the ELB (Table 1). Active layer slopes, computed from measured depths below surveyed transect points, range from 0.01 to 0.55 with an average of 0.11 around LF, and from 0.07 to 0.63 with an average of 0.31 around the ELB. Water table extents, computed from active layer and slope geometries, range from 0.17 to 29.00 m and average 6.79 m around LF, and from 0.22 to 4.71 m with a mean of 1.60 m around the ELB (Fig. 10 ). Around LF, HWT values were strongly correlated with HW observations (R2 = 0.90, Fig. 11A ), whereas, around the ELB, the smaller range of HWT values was less strongly correlated with the smaller range of HW observations (R2 = 0.39, Fig. 11B). In most cases, HW distances are longer than HWT around both lakes, as most points plot above the 1:1 lines in Fig. 11A and 11B.


Figure 10
View larger version (23K):
[in this window]
[in a new window]

  		FIG. 10. Average horizontal and vertical locations of sampling sites (points), and corresponding active layer depth ranges for hydrologic margins of (A) Lake Fryxell and (B) the east lobe of Lake Bonney.

 

Figure 11
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 11. Relationship between the lateral extent of the water table (HWT) estimated from the geometry of measurements, and the lateral extent of the hydrologic margin (HW) at each transect for (A) Lake Fryxell and (B) the east lobe of Lake Bonney. Note that at most transects, HW is greater than HWT, indicating a lateral component to the wicking of lake water into the shore sediments.

 
Based on HW and the depth of the active layers, we estimated the area and volume of the hydrologic margins adjacent to LF and the ELB. The surveyed perimeter of the LF shoreline (~90% of total) was 18.36 km, whereas the hydrologic margin perimeter around LF was 18.94 km. The total hydrologic margin area was 0.16 km2. Unfortunately, we could not completely circumnavigate LF because of the Canada Glacier, bounding LF to the west, and on the northwest corner of the lake, an Antarctic Specially Protected Area (ASPA no. 131) has been established (http://cep.ats.aq/cep/apa/index.html [verified 8 Oct. 2007]), for which we did not have a permit to access. These missing shore distances are ~2 km. The surveyed perimeter of the ELB shoreline was 11.23 km, and the ELB hydrologic margin perimeter was 11.88 km. The total hydrologic margin area measured was 0.05 km2 around the ELB. Calculated volumes of the hydrologic margin zones (product of hydrologic margin area and average depth to permafrost) are 58,925 m3 around LF and 21,473 m3 around the ELB.


   Discussion and Conclusions
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion and Conclusions
REFERENCES
 
Intermittently saturated soils and sediments are likely to be important locations of biogeochemical cycling and microbial biodiversity in polar deserts. Previous work has remarked on the distinct biotic communities occurring in soils and sediments adjacent to streams and lakes (Treonis et al., 1999; Bobb, 2006; Ayres et al., 2007) while others have demonstrated enhanced potential of nutrient cycling and mobilization. Gooseff et al. (2003b) noted that subsurface water adjacent to a Taylor Valley stream had much higher nutrient and major ion concentrations than stream water. Barrett et al. (2002) found that resin-exchangeable soil NO3 was strongly correlated with soil moisture in near-stream soils, and Bate et al. (2007) reported depleted levels of total and available P in dry valley stream channels, indicating weathering and mobilization of essential nutrients. A growing body of literature illustrates the importance of these hydrologic margins, and terrestrial–aquatic connectivity in general as a major control over biogeochemical cycling in the McMurdo Dry Valleys (Barrett et al., 2007). Thus an analysis of the landscape controls over the spatial distribution of soil and sediment water content in near-shore environments is an important step toward understanding the role of hydrologic margins in facilitating biogeochemical transformations and transport of nutrient elements.

The dynamics of these hydrologic margins has not been documented in the past, though they have been empirically assessed to be stable in time. Lake levels moderate from year to year, rising or falling up to several centimeters per year, during the course of a few months. We expect that because of these gradual changes in boundary conditions, the wicking extent up the shoreline may lag some, but is generally likely to keep up with lake level changes. Over millennial time scales, the lake rise and fall may greatly change the boundary conditions for hydrologic exchange between terrestrial and aquatic domains, especially if lakes rise and become more dilute. Thus, for approximately the current conditions, we expect that these findings will apply.

Our results indicate that the particle size distribution of near-shore sediments and soils are fairly homogeneous within hydrologic margins, with only minor differences between lake basins; slope had a greater control on the spatial extent of the hydrologic margin (HW), explaining 73% of the variance in the dimensions of the hydrologic zones around Taylor Valley lakes. Around the ELB, ZW was moderately correlated with HW, further supporting the minor role that the larger size fractions analyzed here play in controlling soil wicking of water. It is worth noting, however, that this moderate correlation and the lack of correlation at LF may be due, in part, to the very mild lake shore topography in LF compared with the ELB.

Dimensions of the water table extent (HWT) were correlated with HW around LF, and to a lesser extent around the ELB. The fact that most of the points in Fig. 11A and 11B plot above the 1:1 lines indicates that there is a lateral component to the wicking of water up these slopes. Given the slope of the linear regressions of these relationships, there is approximately a 10% extension of the surface hydrologic margin beyond the estimated water table distance around LF, and 17% extension around the ELB. In the ELB hydrologic margins, greater vertical extension of the wetted edge was observed, in part because the ELB shorelines were generally steeper than those of LF; however, the average LW of the ELB transects was 3.56 m, whereas the average LW of the LF transects was 8.93 m. The relationships between hydrologic margin dimensions and slope were best correlated around LF, while around the steeper shore of the ELB weaker relationships were found. One potential cause of this variability is the difference in soil salinity between the two lake basins and its potential for influencing water movement in the subsurface.

The differences in the spatial dimensions of the hydrologic margins between the ELB and LF may be due to differences in basinwide soil salinity and the hygroscopic properties of salt minerals that occur in the two lake locations (Campbell and Claridge, 1987; Bockheim, 1997; Barrett et al., 2004). Shoreline soil salinities are generally greater than those found in the adjacent surface waters near shore, supporting the notion that these soil waters are wicked from the adjacent water bodies. The ELB has generally high soil salinities (1000 µS cm–1 in upland soils is typical), whereas LF generally has much lower soil salinities (500 µS cm–1 typical). In both basins, soil salinities typically increase near the evaporative zones adjacent to streams and lakes (Treonis et al., 1999; Barrett et al., 2002; Ayres et al., 2007). Thus, pore water in the ELB shores is more likely to be enriched in dissolved salts than pore waters around LF. This may suggest that more water is wicked through the ELB shoreline soils than those around LF; however, the soils of the Lake Bonney basin are generally higher in salinity overall, so we cannot conclude that the difference in surface soil salinity is due solely to differences in flux. The differences do suggest that pore water movement and extent across the ELB hydrologic margins are likely to be different than across the LF hydrologic margins. Salinity differences may also, of course, affect biological activity differences between these basins and across these hydrologic margins. While water is a potentially important pore-scale environmental mediator and transport medium that may enhance biological or biogeochemical activity, it is also possible that salinity may counteract some of this potential.

On average, neither lakeshore displayed >10% proportion of the finest size fraction of soil. Thus, it is not surprising that soil texture is not a more important control on HW; however, the very fine size fractions (<10 µm) that most strongly influence capillary flow are probably present in very low concentrations given that the entire <75-µm fraction represented only 4% of the soil by weight. Regressions between soil texture and HW (not shown) were not significant, with the exception of a relationship between the coarse fraction and HW, which improved the multiple regression analysis around the ELB, where soils were typically coarser than soils from LF (Fig. 7). This, too, could be due to influences of salts, which often coat the undersides of gravel- and cobble-sized particles and have been shown to have a disproportionate effect on soil moisture in very dry soils (Campbell, 2003).

The strong regression of HW from the combined data with shore slope (Fig. 8) suggests that one could, with reasonable information about local slope (e.g., digital elevation model, topographic maps, etc.), predict the dimensions of the hydrologic margin surrounding lakes in polar deserts for which such detailed data is not available. This is a potentially important predictive tool because these zones of enhanced moisture in very dry settings are likely to be locations of enhanced biological and biogeochemical activity (Bobb, 2006; Treonis et al., 1999). When combined with the findings of Fig. 9, however, it seems that a single model of HW as a function of S is not really appropriate. We propose that the explanation for this is that HWT is the most important control on HW. That is, if LF had steeper slopes, that is, more sediment overlying the developed water table (HWT), the observed ZW would probably be greater, similar to those observed around the ELB (Fig. 10). The difference in slopes of the two lakes is related because the steeper slopes of the ELB provide more overlying sediment in which hydrologic margins can develop, with deeper ZW values observed. Thus, shorelines are limited by the extent to which HWT develops, which is dependent on the surface energy balance and heat exchange within the shore sediments.

Our findings show that the spatial dimensions of hydrologic margins around polar deserts are controlled by a suite of physical parameters that may be easily obtained from remotely sensed data and limited sample collection. Prediction of soil moisture from remote sensing products is a burgeoning area of both hydrology and remote sensing (Schmugge et al., 1974; Njoku and Entekhabi, 1996; Kerr, 2007). Similarly, the products of remotely sensed surface topography from aerial platforms have also become common tools in hydrology (Ritchie, 1996; Tenenbaum et al., 2006). While we did not use remote sensing products, our findings may be useful to other projects that may wish to incorporate such tools or resources. Furthermore, our findings are useful for constraining landscape-scale (tens of square kilometers or greater) models of nutrient redistribution since intermittently saturated sediments along lake margins are zones of enhanced biogeochemical transformation and transport in this polar desert ecosystem.


   ACKNOWLEDGMENTS
 
We gratefully acknowledge Raytheon Polar Services Corp., Petroleum Helicopters, Inc., and UNAVCO for logistical and field support, and the McMurdo Long-Term Ecological Research Project. This research was funded by the National Science Foundation under collaborative research grants OPP 03-38267, 03-36970, and 03-38174. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Site Description
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Gooseff, M. N.
Right arrow Articles by Takacs-Vesbach, C. D.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Water Content
Right arrow Capillary Fringe Processes

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome