SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 28 June 2005
Published in Soil Sci Soc Am J 69:1275-1287 (2005)
DOI: 10.2136/sssaj2004.0204
© 2005 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 ISI 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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
Related Collections
Right arrow Soil Geomorphology and Geography
Right arrow Frozen Soils
Right arrow Pedology
Right arrow Soil Classification and Mapping

Pedology

Soil Topochronosequences at Storbreen, Jotunheimen, Norway

R. G. Darmodya,*, C. E. Allenc and C. E. Thornb

a Dep. of Natural Resources and Environmental Sciences, 1102 S. Goodwin Ave., Univ. of Illinois, Urbana, IL 61801
b Dep. of Geography, 607 S. Matthews St., Univ. of Illinois, Urbana, IL 61801
c Santa Cruz, CA, 95060

* Corresponding author (rdarmody{at}uiuc.edu)


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 CONCLUSIONS
 REFERENCES
 
The rate at which an undifferentiated pile of geologic materials becomes organized into a soil is unknown in most environments. To this end, we examined soil development and chemical weathering trends on four dated Holocene moraines and on adjacent 9000-yr-old sites at Storbreen glacial foreland in Norway. Our objective was to determine the influence of land surface age, elevation, and landscape position on soil properties in this alpine environment. Sampling sites on moraine crests and bases ranged in elevation from 1165 to 1465 m and in age from 70 to 250 yr old. The soils were mostly frigid or isofrigid, coarse textured, and poorly developed Cryorthents. However, there were differences between same-age soils at different elevations, same-elevation soils of differing ages, and soils from adjacent moraine crest–base pairs. Older and higher elevation soils tended to have more silt and clay, possibly due to eolian additions. Extractable Al and Fe, cation exchange capacity (CEC), organic C, and C/N ratio tended to increase with time, while pH, base saturation, and most extractable elements generally decreased with time. Extractable Mg, Al, Fe, K, P, and Cu tended to increase with elevation. Soils at moraine base locations were cooler and had more organic C than adjacent moraine crest soils. The rate of organic N accumulation initially is rapid, but diminished with time from 0.120 to 0.001 g cm–2 yr–1 on the 70- and 9000-yr-old surfaces, respectively. Organic C accumulation rates were greatest on the 250-yr-old surface, 0.31 g cm–2 yr–1, and slowest on the 9000-yr-old surface, 0.02 g cm–2 yr–1. Rates of organic C and N accumulation both increased with elevation. Primary minerals, quartz, mica, feldspar, and amphibole dominated soil mineralogy. Secondary minerals, in particular hydrobiotite, increased with age and elevation, thus revealing weathering trends. Despite generally poor soil development, we found detectable topochronosequence differences in soils and weathering in this young, cold environment.

Abbreviations: CEC, cation exchange capacity • H, high elevation • ICP, inductively coupled plasma • L, low elevation • M, medium elevation • masl, meters above sea level


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 CONCLUSIONS
 REFERENCES
 
GLACIAL FORELANDS PROVIDE convenient sites to study early changes in soil properties (Chandler, 1942; Crocker and Major, 1955; Crocker and Dickson, 1957; Mellor, 1987; Messer, 1988). The land surface uncovered by retreating ice is initially unvegetated, and presumably unweathered, thereby presenting a rare opportunity to follow a chronological sequence of early biological and pedological developments. This requires a presumption that the materials exposed are similar and differ only in time since exposure. However, these assumptions may not be valid. As glaciers (particularly, alpine glaciers) retreat, they expose topographic irregularities and surfaces at increasingly higher elevations. Pedological chronosequence studies are complicated because temperature and precipitation vary by elevation, and minor differences in topographic position influence effective precipitation through variable snow accumulation and thus soil moisture (Darmody and Thorn, 1997).

Early documented changes in soils on glacial forelands follow the invasion of the fresh land surface by vegetation (Matthews, 1992). These include development of soil horizons accompanied by organic C and N accumulation, pH decrease, iron translocation, and CEC increase, as well as limited mineralogical transformation, such as the formation of hydrobiotite (Chandler, 1942; Crocker and Dickson, 1957; Burt and Alexander, 1996; Mellor, 1987; Messer, 1988; Darmody and Thorn, 1997; Kohls et al., 2003). Some changes in soils are rapid on glacial forelands. In southeast Alaska, for example, Alexander and Burt (1996) found albic E horizons in 70-yr-old soils and identified Typic Haplocryods on terrain deglaciated only 240 yr previously. Potential pitfalls of glacial foreland pedogenic chronosequence studies include problems of soil sampling by a set depth or by horizon, lack of replication, potential differences in source materials, and ignoring the effects of elevation and microtopographic variability.

Our objective was to determine soil changes associated with age since exposure after deglaciation and relate them to differences in elevation and landscape position at Storbreen, Norway.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 CONCLUSIONS
 REFERENCES
 
Study Area
Storbreen glacial foreland (8°21' E, 61°35' N, snout elevation {approx} 1400 m) is a chronologically and ecologically well-characterized alpine location that has also been an attractive site for ecological and pedological research (Mellor, 1984; Messer, 1988; Matthews, 1992; Darmody et al., 1987, Darmody and Thorn, 1997) (Fig. 1) . An unusually well-dated set of moraines associated with the retreating Storbreen glacier affords good age control. In addition, an extensive ecological research program has generated much supporting information (Matthews, 1992). Matthews (1979)(1992) provides a detailed description of the general area and Darmody and Thorn (1997) provide details and references about the immediate research locale. The climate is semioceanic, with a mean annual air temperature of approximately –2.0°C, a mean June air temperature of 4.2°C, and a mean December air temperature of –11.3°C (Messer, 1988). Estimates of mean annual precipitation in the region range from 860 to 1200 mm (I.M. Nordin, Norwegian Meteorology Institute, 2004, personal communication).



View larger version (72K):
[in this window]
[in a new window]
  		Fig. 1. Map of general research area in Jotunheimen, Norway.

 
Ultramafic rocks including peridotites, pyroxenes, pyroxene gneiss, amphibolites, mylontes, and quartz feldspar veins dominate bedrock geology of the region (Battey and McRitche, 1973; H.M. Battey, 1975, unpublished data). Biotite weathers to hydrobiotite in the pre-Holocene age glacial deposits at Storbreen (Mellor, 1984), and there is evidence of similar weathering trends in the older Holocene moraines (Darmody et al., 1987). Messer (1988) studied soil chronosequences on the dated moraines at Storbreen. That work showed relationships between increasing moraine age and increasing soil depth, organic C content, and CEC, as well as decreasing soil pH. Our previous reconnaissance work has shown detectable weathering and soil genesis trends on this landscape. The work reported here builds on earlier efforts. The new experimental design is more comprehensive than the previous work (Darmody et al., 1987; Darmody and Thorn, 1997).

Field Methods
We sampled soils at 19 sites arrayed over the Storbreen glacial foreland and the adjacent hillslope outside the foreland in August 1998 (Fig. 2 , Table 1). The experimental design involved sampling soils from three closely spaced (2–3 m) soil pits from selected sites on dated surfaces across a range of elevations. Surfaces dated by lichenometry and direct observation (Matthews, 1992) included the "Little Ice Age" or AD 1750 Storbreen glacier end moraine and recessional moraines dated approximately at AD 1810, 1870, and 1928. Thus, given the recognized uncertainties, the foreland sites were nominally 250, 190, 130, and 70 yr old when sampled. Also included was a {approx}9000-yr-old surface adjacent to, but outside, the Storbreen foreland. The experimental design included three elevations: low (L), at 1165 to 1180 masl (meters above sea level); middle (M), at 1310 to 1330 masl; and high (H), at 1400 to 1465 masl. We sampled the 250-yr-old moraine more intensely than the other moraines, including samples from each of the three elevations on the south side of the foreland on both its crest and adjacent base on its proximal side. Sampling sites on the north side of the foreland included only moraine crests. Samples on the crest of the 190-yr-old moraine came from all elevations on both sides of the foreland, with the exception of the north side high elevation. The 130- and 70-yr-old moraine crest samples included crest positions at the middle elevation on the south side only. We sampled the 9000-yr-old surface across the full elevation range southeast of the foreland. Given the orientation of the foreland on the west slopes of the main valley, Leirdalen, all sample sites had essentially a similar easterly aspect (Fig. 1).



View larger version (42K):
[in this window]
[in a new window]
  		Fig. 2. Map of Storbreen foreland showing sampling site locations. Isochrones show approximate age of dated moraines, based on the year AD 2000 (modified from Matthews, 1992).

 

View this table:
[in this window]
[in a new window]
  		Table 1. Research site locations and features in Storbreen, Norway.

 
We hand-dug pits to describe and sample the soils (Soil Survey Staff, 1993). Typically, there were three samples, from the A, C1, and C2 horizons, per soil pit, with some sites having a fourth O horizon at the surface or a Bw beneath the A or O horizon (Table 2). We sampled the soils to 50 cm deep, but occasionally shallower if we encountered immovable rock. In all, we described, sampled, and saved for later analyses 196 soil horizons from 57 pedons at 19 locations. In addition, to further characterize the soils, we monitored soil temperature at nine selected sites by means of loggers buried at the 50-cm depth that recorded temperatures four times daily from July 1998 to August 2000.


View this table:
[in this window]
[in a new window]
  		Table 2. Soil physical properties from Storbreen, Norway.

 
Laboratory Methods
Soil samples were air dried and sieved through a 2-mm sieve. The <2-mm fraction was saved for further analysis. Particle size was determined by sieving for sands and by hydrometer for silt and clay (Gee and Bauder, 1986). Soil pH came from a glass electrode and a 1:1 soil to distilled water suspension (McLean, 1982). Inductively coupled plasma (ICP) emission spectrometry determined the composition of 1:10 soil/Mehlich III extracts (Mehlich, 1984). Mehlich III extractible cations correlate well with exchangeable cations and Mehlich III extractable Fe and Al correlate well with oxalate extractable Fe and Al (Eckert and Watson, 1996; Fernández-Marcos et al., 1998). All elements except B were determined directly from the Mehlich III-ICP analysis; B required adjustment to hot water extractable levels. Summation of Mehlich III extracted cations gave the CEC, with exchangeable H estimated from SMP (Shoemaker–McLean–Pratt) buffer pH at pH 7.5 to calculate base saturation (McLean, 1982). Total C and N were determined with a Carlo Erba C–N–S analyzer, and are from organic matter because there were no carbonates in the soil.

We determined mineralogy of the <6-µm fraction on selected samples by x-ray diffraction (Hughes et al., 1994). Semiquantitative estimates of mineral abundance came from diffractogram peak areas (Darmody et al., 1987; Klages and Hopper, 1982), with treatments including glycolation and heat to differentiate expandable and heat-liable minerals.


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
SOIL CHEMISTRY
 CONCLUSIONS
 REFERENCES
 
Site Characteristics
Sample site elevations ranged from 1165 to 1465 masl, and slopes ranged from 1 to 72% (Table 1). Soils at all of the sites were stony. Moraine crest surfaces tended to be particularly stone-covered with a deflation pavement of coarse gravel and cobbles. Stone cover ranged from 20 to 98% of the surface, with the lower values on the 9000-yr-old sites and the higher values on the moraine crest positions. Vegetation on the sites was a mosaic of low lichen heath and cryptogam crusts in the more exposed areas (Matthews, 1979). The northern side of the foreland had somewhat better vegetative cover. The parent material for the soils on the foreland was locally derived pyroxene-granulite gneiss till from the Storbreen glacier. The low-elevation 9000-yr-old sites were on a terrace composed of alluvium from the Leirdalen drainage. The other 9000-yr-old sites were on the valley side slope and the soils were formed in colluvium or till from the glacier that once filled Leirdalen (Fig. 1). The soils generally were well drained; the more coarse-textured soils on convex moraine crests were excessively drained.

Soil Physical Properties
The soils were all very coarse textured with only weak horizon development (Table 2). Coarse fragments were common and ranged from gravel to boulders. Most pedons examined were cobbly to very cobbly. Sand was predominantly very coarse, and total sand contents ranged from 31 to 95%. Silt contents ranged from 25 to 64%. Clay content ranged from 0 to 8%. In general, surface horizons tended to be thicker on older sites and at higher elevations. Thin, <3-cm surficial O horizons were found at some locations and were sampled if thick enough. The A horizons ranged in thickness from 1 to 7 cm, except at the 9000-yr-old high site where the A horizon averaged 30 cm thick and at the 9000-yr-old middle elevation site where the A horizon averaged 17 cm thick. Weak Bw horizons were present at some of the older sites. However, no E horizons were found and most sites had simple A over C profiles. Soil colors tended to be dark gray and dull, reflecting the organic matter and dark gray-colored crushed rock parent materials. Other than darkened A horizons, soil colors showed little evidence of horizonation. Chromas were 1 or 2 and values ranged from 2 to 4. Soil hues were typically 2.5Y to 5Y. Some soils on the older sites had 7.5YR to 10YR hues.

Arranging the samples in chronological order revealed some trends in particle size distribution (Table 3). The chronosequence at elevation M included 73 samples and showed that older sites have more silt and clay than younger sites. However, there was one exception; the 70-yr-old site was out of sequence. This unexplained result may help elucidate the anomalous chemistry and mineralogy of that site. A similar pattern emerged when all 182 mineral soil samples were included in the chronosequence analyses.


View this table:
[in this window]
[in a new window]
  		Table 3. Silt and clay distribution by age and elevation at Storbreen, Norway.

 
A toposequence involving the 54 samples from the 250-yr-old moraines revealed that higher sites have more silt and clay (Table 3). When all 182 samples in the data set were included, the pattern is similar with silt, but clay distribution does not follow a topographic trend. Previous work suggested the presence of eolian materials regionally (Darmody et al., 1987). If eolian additions were an important factor on this landscape, the distribution of silts should show an age and depth related pattern. There was no significant depth relationship, but the weak chronosequence trend may indicate some eolian input. In addition, the coarse textured soils may allow rapid redistribution of silts through the soil profile, thus obscuring the eolian evidence.

Soil Climate
Soil temperatures as measured at selected sites at 50 cm deep were frigid (Table 4). Mean summer temperatures ranged from 8.7°C at 250 SCL to 2.6°C at 250 SBH, and mean winter temperatures ranged from –0.4°C at 9000 STL to –6.6°C at 250 SCH. Mean annual temperatures ranged from –0.2°C at 250 SCH to 1.8°C at 9000 STL. Seasonal differences in temperatures ranged from 14.4°C at 250 SCL to 4.6°C at 250 SBH. Cryic temperature regime and frigid temperature class apply to all soils, except sites 250 SBH and 9000 STL, which are isofrigid due to their narrow seasonality (Soil Survey Staff, 1999). Larger temperature seasonality was associated with sites presumed to be more exposed—that is, the convex moraine crest sites—and which therefore carry a shallower wintertime snow cover, thus allowing colder temperatures to penetrate. Conversely, the more deeply snow-covered areas—that is, the moraine base positions and the more vegetated 9000-yr-old surface—tend to be warmer in the winter and cooler in the summer due to the insulating properties of snow. Lower summer temperatures were also associated with higher sites, moist moraine basal locations where late lying snow presumably occurs, and sites shadowed by terrain to the south. Permafrost is predicted to occur regionally only at elevations above 1460 m (Isaksen et al., 2002), an estimate supported by our only subfreezing mean annual soil temperature of –0.2°C at site 250 SCH at 1465 m.


View this table:
[in this window]
[in a new window]
  		Table 4. Soil temperatures at selected research sites in Storbreen, Norway.

 

   SOIL CHEMISTRY
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
CONCLUSIONS
 REFERENCES
 
The soils showed little differentiation by horizon, except for the accumulation of organic C at the surface (Table 5). They tended to be acid, with pH values ranging from 4.1 to 6.7. Consequently, base saturation ranged from 5 to 100%, averaging 51% across all samples analyzed. Both pH and base saturation correlated negatively with organic C content and therefore increased with depth in the soil. As would be expected, the highest pH and base saturation values were associated with the younger sites. Cation exchange capacity was low in the soil samples analyzed, with the exception of the O horizon samples. Overall, CEC values ranged from 2 to 25 cmolc kg–1. Calcium was the most common exchangeable cation, followed by Mg, K, and Na. The north side of the foreland generally supports more vegetation and consequently has higher levels of organic C and extractable Ca, Mg, and K and total N compared with the south side.


View this table:
[in this window]
[in a new window]
  		Table 5. Soil fertility parameters of soils from Storbreen research sites.{dagger}

 
Mean chemistry data revealed chronosequence patterns (Table 6). Cation exchange capacity tended to increase with age, with the 9000-yr-old sites significantly different from the younger foreland sites. Similarly, pH and base saturation decreased with age. Among extractable elements, S and Al increased with age while P, Ca, Mg, K, Na, and Cu decreased with age. Grouping the samples from the 250-yr-old moraine revealed additional patterns. Mg, K, and Fe showed statistically significant increases with elevation, while most of the other parameters also showed weak trends of increase with altitude. Although not particularly noticeable in the field, horizonation, as indicated by the ratio of extractable Fe in the A to C horizon, also increased with time and decreased with elevation. When averaged across all 57 samples, the ratio of Fe in the A horizon to that in the C horizon was 2.3, 1.7, 1.4, 1.2, and 1.2 for the sites in order of decreasing age. The same ratio was 1.4, 1.6, and 2.0 for the high, middle, and low sites, respectively, indicating that Fe is weathering out of the primary minerals preferentially at the lower and older sites, but is not translocating to the lower horizons significantly.


View this table:
[in this window]
[in a new window]
  		Table 6. Soil fertility parameters by toposequence and chronosequence of soils from Storbreen, Norway.{dagger}

 
The most obvious topochronosequences at the research area involved organic C and N accumulation (Table 7). We assumed that there was none in the original parent materials. To make this comparison, we calculated the total content of organic C and N in each pedon by summing the content for each horizon adjusted by horizon thickness and density. This approach gave a trend indicating increasing C, N, and C/N ratio with time. However, the 70-yr-old site was again an outlier with more N than expected in the chronosequence. The general trends also held when the samples included in the statistical analyses were limited to particular elevations. For example, limiting the samples to those from elevation M (map no. 2, 6, 7, 11, 14, 15, 16, 17; n = 24), gave a least square mean total C of 181 g cm–2, N of 13 g cm–2, and C/N ratio of 14 for the 9000-yr-old site. In contrast, samples from elevation M on the 70-yr-old site had only 7 g cm–2 of total C, 8 g cm–2 of total N, and a C/N of 1. Further limiting the samples in the analyses to only those from the south crest sites (map no. 2, 6, 14, 15, 16; n = 15), or elevation L (map no. 3, 8, 18; n = 9), or elevation H (map no. 1, 4, 13; n = 9), gave similar, but generally weaker trends. Overall, trends in N accumulation were not as simple a those with C; the overall relationship between organic C content and soil age on the foreland, that is, excluding the 9000-yr-old sites, was exponential (Fig. 3a) :

[1]
This relationship gave R2 = 0.94. Including the 9000-yr-old sites gave R2 = 0.98 and a logarithmic relationship between age and organic C (Fig. 3b):

[2]
Elevation trends indicated that both C and N typically increased with elevation (Table 7). This tendency was more pronounced on the 9000-yr-old sites (map no. 1, 2, 3), but the differences were not statistically significant. The analyses included grouping together all the 250-yr-old sites (map no. 4, 5, 6, 7, 8, 9, 10, 11, 12), limiting them to moraine crest sites on the north side (map no. 10, 11, 12) or on the south side (map no. 4, 6, 8), or limiting the analyses to the 190-yr-old south crest sites (map no. 13. 14, 18). There was no clear elevational trend with C/N ratio. Crest sites had lower, but not statistically significant, total C, N, and C/N than did the moraine base sites on the 250-yr-old moraine, presumably a response to the cooler, wetter environment caused by snow accumulation at the base of the moraines.


View this table:
[in this window]
[in a new window]
  		Table 7. Total organic C and N by chronosequence and toposequence of soils from Storbreen, Norway.{dagger}

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 3. (a) Soil organic C accumulation across time at the Storbreen glacial foreland sites. (b) Soil organic C accumulation across time at all sites.

 
Dividing the total organic C or N content by the site age gives the mean rate of organic C or N accumulation at each site (Table 8). Using this metric, the rate of organic C accumulation in the foreland soils increased with time, from about 0.09 g cm–2 yr–1 initially to about 0.3 g cm–2 yr–1 at the 250-yr-old sites. The rate of soil organic C accumulation was much lower on the 9000-yr-old surface, about 0.02 g cm–2 yr–1. Total soil organic N accumulation differs from organic C. The initial N accumulation rates are high, about 0.12 g N cm–2 yr–1, presumably because early colonizers are N2 fixing plants. After N builds up in the soil, N fixers are less competitive and N accumulation rates drop to <0.03 g N cm–2 yr–1 by 250 yr, and by 9000 yr it is much lower, 0.001 g N cm–2 yr–1. The overall relationship of soil age and C accumulation rate was exponential and gave an R2 of 0.77:

[3]
The relationship for N accumulation rate gave an R2 of 0.98:

[4]
As indicated by the time series correlations for organic C accumulation, rates of accumulation of total organic C and N also varied by elevation (Table 8), although they were not significant at 5%. The least square mean rate of accumulation for both C and N increased with elevation, indicating that the cooler and presumably moister conditions at higher elevations allow relatively less decomposition of soil organic matter.


View this table:
[in this window]
[in a new window]
  		Table 8. Rate of accumulation of organic N and C by chronosequence and toposequence in soils at Storbreen, Norway.

 
Mineralogy
In these young, poorly developed soils, mineralogy largely reflects the pyroxene-granulite gneiss parent rock (H.M. Battey, 1975, unpublished data). Primary minerals accounted for 40 to 100% of the minerals identified in the samples. In order of abundance, the primary minerals identified in the <6-µm fraction were feldspar, mica, amphibole, and quartz (Table 9). Secondary minerals identified, in order of abundance, included an interstratified mineral identified as hydrobiotite, plus vermiculite, kaolinite, and smectite. Previous research showed that abundance of secondary minerals varied by position on moraines; the more moist moraine base positions had relatively more secondary minerals than the drier moraine crest positions (Darmody et al., 1987). However, samples from moraine base locations were not included in the present x-ray analyses, so the weathering interpretations are conservative. The analyses included the surface A horizons and the first C horizon at 13 sites along the toposequence; that is, 10 sites on the crests of the dated moraines, plus three sites on the 9000-yr-old tundra area outside the foreland. Statistical analysis of the mineralogy was not possible because we included only one site from each location. In general, primary minerals are less common in the surface horizons (73%) than in the C horizons (81%), indicating some slight surficial weathering. The relative abundance of secondary minerals is evidence that there was more advanced weathering on the 9000-yr-old tundra surface as compared with the foreland, as would be expected given the great difference in age between the foreland and the tundra. In addition, on the 250-yr-old moraine, the northern side of the foreland showed somewhat more weathering than the southern side. A better vegetative stand on that side of the foreland may explain the discrepancy; and the chemistry data, lower pH and greater organic C, independently supported this interpretation (Table 5).


View this table:
[in this window]
[in a new window]
  		Table 9. Mineralogy of <6 µ size fraction by site and horizon of soils from Storbreen, Norway.

 
We determined topochronosequence trends by summing the unrounded mineralogical estimates for both horizons at each site (Table 10). The hypothesis here was that the most easily followed weathering sequence is biotite mica weathering to hydrobiotite, which in turn weathers to vermiculite (Darmody et al., 1987). Trends were noticeable on both toposequences tested, the 9000-yr-old surface and the crests of the 250-yr-old moraines. While total mica content increased slightly with elevation in the 250-yr-old sites, there was no trend in the 9000-yr-old sites. This reflects both random variability and increased weathering of the primary minerals on the older sites. Total interstratified minerals (hydrobiotite) increased with elevation, both in absolute abundance and as a fraction of the total biotite mica content, indicating increased production of hydrobiotite at the expense of biotite with elevation. Differences in extractable K and Mg support this observation; there are more of both at higher elevation on the 250-yr-old moraine (Table 5). Total vermiculite also increased with elevation in the 250-yr-old toposequence, but again, there is no trend in the 9000-yr-old samples. Expressing the vermiculite content as a ratio to the total biotite content strengthens the trend in the 250-yr-old sites and reveals a similar trend in the 9000-yr-old sites.


View this table:
[in this window]
[in a new window]
  		Table 10. Topochronosequences of total secondary minerals of <6 µm size fraction of soils from Storbreen, Norway.

 
These findings would support a counterintuitive trend that there is more weathering at higher elevations. However, the vermiculite/interstratified trend runs counter to the interstratified/biotite trend. The vermiculite/interstratified ratio was greater at lower elevation, indicating inhibition of the full weathering sequence of biotite to interstratified to vermiculite at higher elevations. It proceeds more readily at lower elevations, as would be expected on first principles. Indeed, transformation of biotite to interstratified minerals proceeds rather rapidly as shown by short-term bench scale experiments (Hinsinger et al., 1992; Murakami et al., 2003), and by our previous work (Darmody and Thorn, 1997; Darmody et al., 1987). However, the transformation all the way to vermiculite is something that in this cold environment is inhibited at the highest, coldest sites.

Chronosequence trends in mineralogy were also evident (Table 10). Total mica content decreased with age as expected. However, the 70-yr-old site was anomalous in the analogous hydrobiotite (interstratified) and vermiculite trends. Generally, hydrobiotite and vermiculite increased with age, but the 70-yr-old site showed anomalously high amounts of these two secondary minerals. This incongruity at the 70-yr-old site is unexplained but the unusually high amounts of silt and clay found there may contribute (Table 3). Consequently, the discussion will not include the 70-yr-old data set. The ratios of presumed weathering products to their parent minerals show consistent age trends; the older sites had greater interstratified/mica and vermiculite/mica ratios. The vermiculite/interstratified ratio generally decreased with age, indicating that, while both minerals increase with age, the relative increase is greater for hydrobiotite. This finding is consistent with the toposequence analyses, where the intermediate weathering of biotite to interstratified hydrobiotite is relatively rapid, yet the conversion of hydrobiotite to vermiculite is slow. An additional complication is that the original parent rock may contribute some or all of the vermiculite in these soils, as Mellor (1984) believed.

Soil Classification
Because of their poor development, we classified most of the soils of the research area as Typic Cryorthents with the exception of the low-elevation tundra site (9000 STL), which was an Oxyaquic Dystrocryept (Table 11) (Soil Survey Staff, 1999). The soils at 9000 STL exhibited a weak cambic horizon with the greatest amount of oxalate-extractable Fe and Al (data not shown) of any horizon investigated. In addition, as previously noted, there was a general trend in extractable Fe distribution demonstrating incipient podzolization increasing with age and decreasing with elevation. Soils on the 9000-yr-old surface were coarse loamy, not quite so coarse textured as those on the foreland which were in sandy-skeletal and loamy-skeletal particle size families. All soils are in the mixed mineralogy family. Soil temperature class for all soils, as interpolated from the temperature record at selected sites (Table 4), is frigid, with the exception of 250 SBM, 250 SBH, and 9000 STL, which are isofrigid due to microtopographic effects on snowcover moderating the temperature, as previously discussed.


View this table:
[in this window]
[in a new window]
  		Table 11. Classification of soils from Storbreen, Norway.

 

   CONCLUSIONS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 CONCLUSIONS
REFERENCES
 
The soils of Storbreen foreland are poorly developed and do not exhibit obvious differentiation due to elevation or age within the foreland. Outside the foreland, the 9000-yr-old soils show their much greater age by slightly better horizonation and thicker O horizons; however, the soil development there is still surprisingly limited. In general, cold temperatures and coarse-textured parent materials at Storbreen do not allow much plant growth, which would accelerate weathering and pedogenesis. Nevertheless, when closely examined with laboratory analyses, subtle topochronosequence trends emerged, although they were more subdued than those shown on forelands with conditions more favorable for pedogenesis such as southeast Alaska (Alexander and Burt, 1996; Chandler, 1942). Trends included increased weathering of biotite to hydrobiotite then to vermiculite on older surfaces. Biotite does not weather beyond hydrobiotite as readily at higher elevations. Release and redistribution of Fe in the soil profile is detectable but has not progressed to the formation of a spodic horizon, even on the 9000-yr-old sites, something accomplished in just a few centuries in Alaska (Alexander and Burt, 1996). Accumulation of organic C and N is the most obvious evidence of soil age in this landscape. However, differences by elevation within an age cohort are not obvious on the younger moraine crests. This is due to microtopographic effects, such as exposed, wind-swept, droughty convex moraine crests, or more protected, moist, snow-accumulating moraine base positions, that control plant growth, which provides the larger organic C and N content found in moraine base locations as compared with equally aged moraine crest positions at similar elevations.

Because toposequence effects are due typically to drainage differences, it is surprising that any toposequence differences are detectable on these soils that essentially have the same drainage. In Storbreen, it is not so much drainage that differs with topography, but climate. Mean summer soil temperature differs by 6.1°C over the elevation range of 300 m we studied. That, together with an unknown difference in precipitation, was enough to produce the toposequence trends we observed. In addition, microtopographical impacts on snow accumulation and persistence and on wind exposure as on crest vs. moraine base locations, account for some soil variability at this high elevation and latitude. Chronosequences studied elsewhere may involve millennia. While there were obvious differences between the soils on the 9000-yr-old surface and those on the <250 yr old foreland, there were detectable differences within the foreland. This indicates rapid initial rates of soil formation on initial weathering of fresh parent materials, and colonization of new surfaces by pioneer vegetation. While the unfavorable conditions permit an initial high rate of change, they do not favor continuation of high pedogenic rates. This observation explains the generally poorly developed soils in the region, even after 9000 yr of pedogenesis.


   ACKNOWLEDGMENTS
 
The authors thank John Matthews for his courtesy and assistance in the field. This is Jotunheimen Research Expeditions contribution 155. National Geographic Society grant 6237-98 and the University of Illinois Research Board provided support for this work.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 CONCLUSIONS
 REFERENCES
 
A grant from the National Geographic Society partially supported this work.

Received for publication June 21, 2004.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 SOIL CHEMISTRY
 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 ISI 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 ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Darmody, R. G.
Right arrow Articles by Thorn, C. E.
Related Collections
Right arrow Soil Geomorphology and Geography
Right arrow Frozen Soils
Right arrow Pedology
Right arrow Soil Classification and Mapping

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