From the December 1990 issue of "Forest Research West"

Lessons from Lysimeters

by J. Louise Mastrantonio

LYSIMETER: A device for measuring water percolation through soil...something like a 'flower pot" that is buried and filled with soil.

It takes four hundred years or more for an argillic horizon to form in soil. True or false? Until very recently, "true" would have been the correct answer. In fact, textbooks still claim it takes 3-4000 years to form an argillic horizon...a layer of clay that has moved down and accumulated in the soil.

Recent research, however, indicates that an argillic horizon may develop rather rapidly--on the order of decades rather than hundreds or thousands of years. Furthermore, its formation may be more a function of overlying vegetation than of time itself.

This new information comes, not from expensive and time-consuming research, but by, quite literally, digging into the past. And it is one of several important findings from a study conducted recently at Tanbark Flats in the San Dimas Experimental Forest, a U.S. Forest Service research site in Angeles National Forest east of Los Angeles. The Scientists are: Hulton B. "Hutch" Wood, a research forester with the Pacific Southwest Station at Riverside, California; Robert C. Graham, Assistant professor of Soil Mineralogy and Genesis at the University of California, Riverside; and Mary A. Lueking, a soil scientist then with Oregon State University and now with ALPKEM in Clackamas, Oregon.


But we're getting ahead of the story. It actually begins more than fifty years ago. In the 1930's, a major research project was initiated at Tanbark Flats to determine how different species of plants affect water yield. Even back then, people were concerned about the availability of water in southern California. Scientists thought maybe some plants would use less water than others and might be used on hillsides to increase water runoff into storage basins.

Researchers, including engineers, hydrologists, plant physiologists, research foresters, and soils experts, spent several years devising an experiment--a study so elaborate and labor intensive there is no way it could be duplicated today. But labor was cheap then, what with a depression going on, and dozens of laborers from work programs such as the Civilian Conservation Corps were put to work--digging holes. They were lysimeters--devices for measuring the movement of water through soil.

First a large trench was dug, the soil removed and stockpiled. Then lysimeter holes were dug, Some were encased in concrete or metal casings("confined lysimeters"). Others were left "unconfined."

Then devices were installed to collect and measure water that flowed through the soil or ran of the surface. The excavated soil was then sieved, mixed, and returned to the lysimeter holes. Thus, each lysimeter contained identical and homogeneous soil samples. In addition, samples of this soil were ''archived"--stored away in glass jars for future reference.

The lysimeters were completed in 1937 and baseline monitoring was begun soon after. Nine years later, five different types of vegetation were planted over the lysimeters: scrub oak(Quercus dumosa), ceanothus (Ceanothus crassifolia), Coulter pine (Pinus coulteri), buckwheat (Fasciculatum eriogonum) and chamise--(Adenostoma fasiculatum). Data collection continued until 1960 when a major fire burned through the Experimental Forest, and provided a convenient stopping point for the research. A report summarizing the research results was published and, for all practical purposes, the study was abandoned.

Today, Tanbark Flats is far quieter. The Experimental Forest is closed to the public because of fire hazard, but the area is visited frequently by scientists working on different research projects--and by forestry professionals as part of field tours.

Comparing soils

Wood, who confesses to being something of a junk collector, became intrigued with the old lysimeter study after transferring from Hawaii to Riverside in 1982. It was the archived soils that fascinated him. They had been stored all these years in a shed downhill from the lysimeter plots. "When I saw all those 'antique' soils in there--uncommitted--it started my wheels spinning, Here was a beautiful treasure trove of soils that had been put away fro fifty years." What Wood proposed to do was to compare the archived soils with present-day soils that have been exposed to high levels of atmospheric pollution. Because the original samples had been saved, the old lysimeter study offered a unique opportunity to study the effects of air pollution on soils in the watershed--something no one had looked at previously.

More specifically, the study would:

  1. Determine the presence of toxic metals (lead, copper, cadmium, arsenic; and mercury) and basic chemical characteristics of lysimeter soils developed under four different plants species;; ceanothus, scrub oak, Coulter pine, and chamise. The buckwheat plots were not used because, over the years, they had been invaded by other species.

  2. Compare sulfate concentrations and sulfate adsorption capacities for archived and lysimeter soils. Sulfates are common in smog. As excess sulfate is leached from the soil it may be accompanied by losses of cation nutrients.

  3. Assess the degree of soil development achieved after fifty years of soil formation.

The study was proposed in 1987 under a system of competitive grants at the Pacific Southwest Research Station in which scientists must compete for discretionary dollars. Research administrators were intrigued and the study was funded.

Wood located the study in the "unconfined" lysimeters because vegetation growth had been inhibited by the containers of the confined lysimeters. Soil pits were dug to a depth of 30 and 100 centimeters and soil samples taken. Observations were made of their morphological properties and later, in the laboratory, the soils were analyzed for soil chemicals, nitrogen, and toxic metals.


Excessive amounts of toxic metals were found, particularly zinc and lead. This was high "but not extraordinarily high" when compared to some industrial sites, according to Wood. Differences were also found in soil nitrogen under the different plant covers. For example, ceanothus helped fix nitrogen in the soil while chamise inhibited nitrification.

But what has researchers most intrigued is the difference in the way soils develop under different plant species. Under scrub oak, earthworm activity promoted thickening of the "A" (or upper) horizon--to about seven centimeters. The leafy matter of scrub oak is apparently very palatable to earthworms. In eating the leaf litter, they also ingest clay particles, and move them up to the surface.

Under pine, however, something totally different happened. All the needles accumulated on the surface. There were no earthworms at all and only a very thin "A" horizon--one centimeter thick. Below this, however, scientist detected a weak argillic (Bt) horizon in the subsoil.

"The interesting thing," according to Graham, "is that the soil processes are completely reversed. In one, clay is being moved to the surface and accumulating. In the other, clay is being leached down in the soil profile and accumulating in the subsoil."

What may be even more startling is that the soil under the pine actually meets requirements for an argillic horizon. Argillic horizons are subsoils that have been enriched with clay that has moved from the surface down into the soil. Their presence is commonly used as an indicator of soil stability. If an argillic horizon is present, builders and engineers consider it safe to locate buildings and other structures--even nuclear power plants. The more well developed the soil, the more stable--long existing--the site is believed to be.

But at Tanbark Flats, an argillic horizon had formed in only about forty years--certainly not a time frame that any builder would be comfortable with. Thus, depending on the conditions under which it was formed, the argillic horizon may not always be a good indicator of site stability.

"The lysimeter study was inadvertently an experiment in soil genesis," Graham says, adding that soil forms as a function of five factors--climate, organisms (dominantly plants), topography, parent material, and time. "Usually when people study soil formation, they try to find situations where all those factors are constant except one," he adds. "In nature, that's hard to do. But here everything was the same--except the vegetation."

It was, inadvertently, a perfect experiment in soil genesis. Researchers who spent so many years designing and carrying out the original study could not possibly have known their work would produce useful results long after the study was abandoned. And who knows what scientists may learn in the future from the old lysimeter study?