Soil versus rock-dominated landscapes

Introduction

When you look at a hilly, gently sloped landscape do you ever wonder why it's covered with soil? Similarly, when you're on your favorite hike through a steep, mountainous landscape do you ever wonder why the ridges and peaks are rocky? If you have wondered about this difference between the hills where deer and antelope like to roam and the mountains where peaks are hard to climb and few living creatures live, then you're not alone. Figuring this out is one of the oldest problems that geomorphologists have struggled with. There are so many potential reasons to explain this very simple difference between landscapes that a whole text book could be devoted to them – that's because understanding the difference gets to the core of understanding the processes and rates of soil production. These processes are complicated and can depend on many external forces, such as the tectonic forces acting upon the Earth's surface, the climate that is helping to set erosion rates, as well as impacts on the landscape by human activity. In this vignette we're going to focus on one simple explanation for why a landscape may be "soily" or rocky: when the erosion rate is greater than the soil production rate the soil is stripped, leaving a rocky landscape behind.

Surprisingly enough, this relatively simple explanation is very difficult to quantify. To do so requires knowing what the erosion rates are across a landscape, as well as what the soil production rates are. If we knew what these rates are, then we could compare the rates to our observations of "soily" and rocky landscapes and confirm whether our intuitions about a landscape could be described quantitatively. To get to this level of quantification means that we must understand that soil is produced from the rock below and that it is eroded by different processes as discussed in the text. We move forward assuming this understanding is clear.

In this vignette we look at the controls exerted on landscape form by erosion and soil production rates. We also show that the transition from soil-mantled to rocky landscapes is set by the maximum rate of soil production. Although never previously documented in the field, it is widely accepted that this threshold rate should correspond to the transition from soil-mantled to rocky slopes, expected where erosion rate exceeds the maximum near-surface soil production rate [Heimsath et al., 1997]. Recent work from other places supports this suggestion. For example, Binnie et al. [2007] show a roughly linear increase in average erosion rates as average slopes increase towards a threshold until about 30 degrees. Similarly, Ouimet et al. [2009] show that average slopes do reach a threshold at about the angle of repose with increasing erosion rates. Above this threshold slopes don't increase very much even though erosion rate continues to increase. Both studies suggest that landscapes become rocky above this threshold. We go a bit further to show that as erosion rates increase past this threshold, the soil production rates do indeed mark the transition from soil to bedrock dominated landscapes.

Field Site and Methods

We focus this vignette on the San Gabriel Mountains because so much is known about the tectonic setting, long-term erosion rate patterns [Blythe et al., 2000; Spotila et al., 2002], short-term erosion rates [Lavé and Burbank, 2004], as well as the climatology and hydrology. The rugged mountains are part of an active tectonic zone resulting from a bend in the famous San Andreas Fault and its interaction with the San Jacinto Fault near Los Angeles, California (Fig. 1). Imagine one huge earthquake fault crossing paths with another and then put a small bit of land in between to get an idea of how the Earth's crust might be getting pushed and pulled into hills and mountains.

Both the average elevation and the relief increase from west to east, consistent with short- and long-term erosion rate patterns and interpretations of relative tectonic activity. We use this west-east change in topography and erosion rates to get a better understanding of the rocky versus "soily" problem introduced above. To do this we studied 32 different watersheds. Average rain or snow in these catchments is mostly from winter Pacific storms and varies just a little with elevation from about 500 to 900 mm/yr. The dominant hillslope erosion processes change from mostly soil creep and dry ravel in the west to debris flows, rock fall, and bedrock landslides in the east. Hillslope form changes accordingly, from soil-mantled in the west (Fig. 2) to bedrock-dominated in the east (Fig. 3).

Soil to Rock Dominated Landscapes

Fifty measurements of cosmogenic ¹⁰Be concentrations from river sands collected from catchments across the whole field area show a clear increase in the average slope of a watershed as the average erosion rates increase [DiBiase et al., 2010](Fig. 4). Slopes increase rapidly from the gentle catchments in the west (about 10 degrees) to the steepest soil mantled catchments near the middle of the range (about 30 degrees). As average erosion rates increase to the east, average slopes reach a threshold and do not get steeper than about 35 degrees. Our field observations show that soil-mantled landscapes typically correspond to slowly eroding landscapes, while rough, rocky, jagged landscapes are associated with high erosion rates. We observe that the transition from soil to rock-dominated landscapes occurs at the inflection point marked by the vertical dashed line on Figure 4. Theoretically, this point should correspond to the maximum rate of soil production.

To test this theory, we measured soil production rates from samples in 46 soil pits from across different soil-mantled ridges spanning the study area (Fig. 5). These soil production rates show a great deal of scatter, but do define a clear exponential decline of soil production with increasing soil thickness. Such an exponential soil production function is discussed at length in papers by Heimsath and others and, importantly, shows that the maximum rate of soil production occurs under very thin to no soil cover. Examining this soil production function in comparison with average erosion rate data shown in Figure 4 reveals a remarkable confirmation of theory. Namely that the peak soil production rate, approximately 200 m/Ma at zero soil depth, corresponds to the dashed line drawn between soil and rock-dominated landscapes.

• Binnie, S. A., et al. (2007), Tectonic uplift, threshold hillslopes, and denudation rates in a developing mountain range, Geology, 35(8), 743-746.
• Blythe, A. E., et al. (2000), Structural and topographic evolution of the central Transverse Ranges, California, from apatite fission-track, (U-Th)/He and digital elevation model analyses, Basin Research, 12, 97-114.
• DiBiase, R. A., et al. (2009), Landscape form and millenial erosion rates in the San Gabriel Mountains, CA, Earth & Planetary Science Letters, in press.
• Heimsath, A. M., et al. (1997), The soil production function and landscape equilibrium, Nature, 388, 358-361.
• Heimsath, A. M., et al. (2009), The 'humped' soil production function: Eroding Arnhem Land, Australia, Earth Surface Processes & Landforms, 34, 1674-1684.
• Lavé, J., and D. W. Burbank (2004), Denudation processes and rates in the Transverse Ranges, southern California: Erosional response of a transitional landscape to external and anthropogenic forcing, Journal of Geophysical Research, 109.
• Ouimet, W. B., et al. (2009), Beyond threshold hillslopes: Channel adjustment to baselevel fall in tectonically active mountain ranges, Geology, In Press.
• Spotila, J. A., et al. (2002), Controls on the erosion and geomorphic evolution of the San Bernardino and San Gabriel Mountains, southern California, Geological Society of America Special Paper 365, 205-230.