June 01, 2006

Amazonian Dark Earths

What I've been up to lately, a.k.a., the 'after all the wailing and gnashing of teeth, my professor is NOT going to be the only other person on the planet to read this' post. However, light reading it ain't. If you get lost, skip glibly through the parts with the hairy numbers. ...


Patches of Amazonian Dark Earth (ADE) as large as 350 ha (hectares) but generally between 1-6 ha (German, 2002) and with observed depths ranging from 0.2-1 m, also known as Terra Preta do Indio, are oases of fertility in a region known for its nutrient-poor, heavily weathered soils. Years of research have established the human origin of ADE soils in the Amazon Basin and that conclusion has prompted a growing number of researchers to investigate the properties of these soils, their origins and their context within the modern world. There are hints that they could represent both an effective method of carbon sequestration as well as a guide to sustainable tropical agriculture, though there remains much to discover.


As late as 1996, the origins of ADE soils were still considered highly controversial, particularly in relation to what they implied about the social complexity, size and stability of the Amazonian population before European contact. While it was well supported that most habitation was centered around the rivers, it was often said that communities lived and grew their food almost entirely in the várzea, or lowlying floodplains. Little activity at all was believed to have been present on the higher terra firme land between river channels due to the extreme poverty of the soils, with the exception of low-density, migratory slash and burn agriculture. (Denevan, 1996)

William Denevan noted that regular flooding in these várzea lowlands rendered them unreliable sites for agriculture and settlement, in spite of their often richer, alluvial soils. As he said, “flooding is what floodplains do,” often unpredictably. (Denevan, 1996)

Denevan went on to propose that the bluffs overlooking the Amazonian tributaries and the floodplains they’d carved out were the actual sites of the large, relatively stable settlements reported by the first European visitors to the region. He suggested that in regions where bluffs overlooked river channels that remained navigable all year, settlements were established on the bluff edge. In this model, crops were established both on the floodplain and on the more reliable upland soils adjacent to the settlement, which were then modified to support permanent cultivation. When river channels shifted away from a bluff as they sometimes did, the settlement would generally be abandoned, to be resettled only in the event that the river drew close again. Rivers provided access to travel, commerce, fresh water and a diversified cropping system that could support both subsistence and trade. (Denevan, 1996)

Denevan briefly noted the main compositional features of the ADE found at the sites examined in his research as distinguished from adjacent soils, aside from the thick and omnipresent remains of shattered pottery at these archaeological sites. Dark Earths have increased phosphorus, a severely limited nutrient in the tropics, believed to be from additions of ash, bone, shells, feces and urine. The soils have a high carbon content from charred plant remains, being the main determinant of their dark color, providing the benefits of the slow fraction of soil organic matter. Due to the ash and char, the soil pH is higher and the decreased acidity lowers the amount of free aluminum present. He said that shifting cultivation has not been observed to produce ADE soils in the present, indicating a very different lifestyle for the former inhabitants of these sites. (Denevan, 1996) While the length of occupation is unknown, charcoal and black carbon on ADE sites has been found dating between 390- 6,850 years BP (Falcão, et al., 2003), though widespread generation of ADE first occurred from 450 BC and 950 AD (Erickson, 2003).

Other researchers in the Colombian Amazon outlined the distinction between the two observed types of ADE, the black terra preta, or Terra Preta do Indio, and the dark brown terra mulata in terms of what they believed were their likely uses. (Herrera, et al., 1992) The black soils were developed under the settlement sites and benefited from more concentrated inputs as the result of accumulation of wastes, while the dark brown soils appear to have been developed by the application of similar amendments over wider sites of intense permanent cultivation. The sites’ pollen and carbon dating information indicated that the addition of organic matter and wastes allowed relatively continuous cultivation for between 800-1000 years.

Considering the extent and persistence of native cultivation of soil types that are known for giving out after two to three years of modern or swidden agriculture when cleared from their original forest (Erickson, 2003), Denevan’s argument for a larger pre-colonial Amazonian population, upwards of 5 million as opposed to earlier calculations of less than a fifth that, gains credence (Denevan, 1996).

Charles Mann reviewed the modern study of human activity in the Amazon (Mann, 2000), beginning with Denevan’s 1961 observations of enormous earthworks in the remote Beni region of the Bolivian Amazon that first sparked his interest. The mounds and berms, too linear to be natural, were later determined by Clark Erickson and his team to be a vast series of fish farms that still catch enough rain to run with fish in the dry season, though they haven’t been maintained for an untold number of years.

Mann reports anthropologist Emilio Moran as saying that even if the ADE sites only covered 10% of the Amazonian uplands they would represent a significant resource base, but that this scenario would still be compatible with an overall representation of Amazonian terra firme as a region of poor soils. Mann wrote that geographer William I. Woods estimated that fewer than 1000 Amazonian soil samples had been analyzed as of 2000. (Mann, 2000)

It’s now estimated that ADE plots comprise between 0.1-0.3% of the Amazon Basin, or an area at least 6 million km2 (Sombroek, 2003), or a total area approaching two thirds the area of the United States. However, there has yet to be a systematic soil survey of the region that could provide more exact figures (Erickson, 2003).

Properties of Dark Earths

Among the soils of the Amazon, the black ADE soils are most obviously distinguished by their color and good tilth, or structure. Because their most defining qualities are anthropogenic, or defined by prolonged human input of some type of organic matter, they are classified as Anthrosols. In use, they are noted for high fertility, which benefits both exotic agricultural plants and invasive weeds alike.

The typical Amazon soil is a red or yellow Kaolinite Ferralsol, Acrisol, or more occasionally a sandy Podzol, which are international soil classification types as described in the World Reference Base for Soil Classification. Ferralsols roughly correspond to the USDA classifications of Ultisol and Oxisol (German, 2002), and the Brazilian Latosol designation, for most of the upland amazonian soils referred to in ADE research that would normally be covered by rainforest. For the sake of simplicity in writing and in order to bridge the differing terms used by researchers of various backgrounds, this report uses the term Ferralsol to describe rainforest soils not exhibiting ADE properties.

Ferralsols have a deserved reputation as being of limited use in agriculture, commonly wearing out after 2-3 years. The rapid loss of soil organic matter through decomposition after clearing is the main reason for the loss of fertility, which is accelerated by frequent drying and wetting (Glaser, Guggenberger, Zech, Ruivo, 2003). The tropical heat also contributes to this rapid decay, as once the temperature reaches 25˚C, decay rises with temperature more rapidly than carbon fixation in plants (Brady, Weil, 2002). This reduces the tropical usefulness of techniques for fertilizing soil through organic material additions (Glaser, et al., 2002).

Fig. 1: Colloid properties, adapted from Nature and Properties of Soils 13th Ed, Brady & Weil, 2002.

P. 319, Table 8.1


Surface Area in m2/g

Charge cmolc /kg







Fine mica









Ferralsols generally have pH values between 4 and 5, which is at the low end of the habitable range for plants and plant beneficial soil microorganisms. That degree of acidity mobilizes the toxic aluminum fraction of the soil and immobilizes the necessary plant nutrient phosphorus, which is the main limiting factor of plant growth in the soil types present in the Brazilian Amazon. These soils also tend to be low in calcium and magnesium cations, because the mineral fraction of the soil has a low cation exchange capacity (CEC), which is to say that their chemical makeup offers few negatively charged sites to attract and hold the plant-available ionic forms of calcium and magnesium. The issue of low CEC is worsened in acid conditions and these necessary positive ions, or cations, then tend to be easily lost in runoff or leached down through the soil layers out of the reach of plant roots. (Brady, Weil, 2002)

A distinctive baseline feature of the upland soils of the Amazon region, whatever they are called, is their predominately quartz and kaolinite clay mix. Quartz minerals are slow to weather and yield few nutrients to plants, while kaolin clays have very low cation exchange capacities, therefore retaining little water and few nutrients. (Brady, Weil, 2002)

Dark earth soils exist across a wide range of their parent soil types. In most cases, they are identical to the background soils in terms of texture (Steiner, Teixeira, Zech, 2004), with the notable exception of a site in the Colombian Amazon. This particular Colombian ADE site also showed evidence of labor-intensive, repeated application of river silt over a cultivated ADE plot (Herrera, et al., 1992), however this is unusual in most of the ADE sites described in the literature, where organic inputs are generally solely responsible for the differences between them and surrounding soils. In general, ADE soils are most clearly distinct from their surroundings in terms of structure, appearance and their rich nutrient profile. (Lehmann, Kern, et al., 2003)

Because the parent material of ADE soils is too weathered to have significant nutrients remaining to account for the difference, the activity on soil organic matter exchange sites has been shown to be of far more consequence in sustaining plant growth than the mineral fraction. Additionally, ADE organic matter is proportionally more active than the organic matter of adjacent Ferralsols. (Lehmann, Kern, et al., 2003) Lacking in a high CEC clay colloid, ADE contains an abundance of humus, or decay resistant soil organic matter with numerous sites suitable for bonding with calcium and magnesium ions. In particular, these soils are dominated by humic acid and humin (Lima, et al., 2002), categories of carbon compounds whose effective lifespan in soil is measured in hundreds or thousands of years respectively (Brady, Weil, 2002). The humin in ADE soils is made up largely of charcoal residue, the product of the incomplete combustion of plant material at high temperatures, which contains high concentrations of highly stable aromatic carbon compounds because of the method of formation. The levels of pyrogenic carbon from charcoal inputs to ADE soils may be up to 70 times higher than surrounding Ferralsols (Glaser, Zech, Woods, 2004)

Amazonian Dark Earth soils’ agricultural properties more closely resemble the Mollisols found under natural prairie grassland in the central plains of North America than they do surrounding soils, with an organic-rich layer of ADE topsoil that may extend a full meter deep. The resemblance of ADE and Mollisols may, however, be more than superficial. At least one micromorphology study of ADE has noted that the crumb structure of the topsoil or A horizon is “typical” of a mollic epipedon, the epipedon being the upper mineral layers that are mixed with organic matter (Schaefer, 2004).
Further, analysis of Saskatchewan prairie soils in southern Canada (Ponomarenko, et al. 2001) at the northern reaches of central North American Mollisol distribution indicated that a significant fraction of their soil organic matter was recalcitrant charred plant material high in aromatic carbon compounds. The study of the Canadian Mollisols, or Black Chernozem soils by the Canadian system, turned up between 10-64% char in the humic fraction of the analyzed soils’ A horizons. The authors noted that in contrast to the current models of soil organic matter that present humus accumulation as a matter of slow decomposition mainly carried out by soil biota, fire quickly produces a full range of such substances, as well as promoting polymerization and aromaticity in the remaining carbon compounds. They also said that there was an “increase in aromaticity with darkening of the soil along a Brown to Black sequence of soils.” I’m unaware of whether any studies that have explored whether this last observation may shed light on the distinctions between terra preta and terra mulatta, if in addition to less concentrated inputs the proportion of aromatic carbon in their organic fraction is different, but it seems like a useful question to answer.

The current model assumes that biological oxidation of the aromatic fraction of pyrogenic carbon forms carboxylic acid groups that increase the CEC capacity (Steiner, Teixeira, Zech, 2004), while the exact mechanism for their formation is unclear, aromatic polyacids are among the recognized molecular markers for combustion (Glaser, Guggenberger, Zech, 2003). A biomass pyrolysis technique explored for making a slow-release nitrogen fertilizer relied on the properties of the organic acids and phenols that formed under charring conditions to react with urea or ammonia to form amines (Radlein, et al., 1996). The pyrolysis process was tried with a wide range of biomass inputs and seems to support the formation through charring of the stable and reactive portions of the ADE organic matter. It has also been demonstrated that increased char temperatures increase the aromatic fraction of the pyrogenic carbon (Glaser, 2002), and aromatic compounds are well known for their stability and resistance to decay. These properties combined may be responsible for the higher stability of pyrogenic carbon, as well as the observation that the CEC values of soil organic matter in ADE are proportionally higher than that of neighboring Ferralsols (Lehmann, Kern, et al., 2003).

The behavior of differing nutrients within ADE varies among nutrients as well as within and between sites (Lehmann, Kern, et al., 2003), though the pyrogenic carbon itself isn’t the main source of the nutrients it helps to stabilize in the soil (Steiner, Teixeira, Zech, 2004 and Glaser, Guggenberger, Zech, 2004). Calcium is the ionic nutrient that shows the greatest increase in ADE over surrounding soils compared to other cations, due to its high affinity with cation exchange sites (Lehmann, Kern, et al. 2003, and Kern, et al. 2004) and the likely addition of mussel and mollusk shells (Sombroek, et al., 2003). Exchangeable calcium may be 55-1400 times more elevated on ADE than on Ferralsols, while magnesium shows increases in ADE of 8-130 times (Lima, et al., 2002). Potassium availability in ADE ranges from being equal to or lower than surrounding soils (German, 2004), to increases of up to 10 times that of comparable soil horizons in Ferralsols (Lima, et al., 2002). The sometimes low potassium concentrations can limit the growth of bananas in the Amazon (Clement, et al., 2003), an important tropical and regional crop.

Regarding phosphorus concentrations in ADE, high values of 1,500 ppm have been reported near Santarém (Kern, et al., 2004) and ranging from 136-3,921 ppm in an analysis of the A horizons of three ADE soils sampled near Manaus (Lima, et al., 2002). For comparison, a northwestern United States soil is considered well fertilized with 20-60 ppm phosphorus (Hart, et al., 1990). A probable reason for the wealth of phosphorus lies in the observation that charcoal adsorbs P better than soil clays on a per gram basis (Falcão, et al., 2003).

More recent studies into the main source of phosphorus in ADE have identified apatite fragments from fishbones as the prime addition (Lima, et al., 2002; Lehmann, et al., 2004), along with other animal bones (Schaefer, et al., 2004). The high presence of calcium-phosphorus compounds, which are more in evidence in the inedible parts of fish, as well as the lack of livestock in pre-Colombian agriculture, have been pointed to as evidence against human or animal manure constituting a primary input for the generation of ADE soils (Lehmann, et al., 2004). Over time, biogenic mixing carried out by earthworms favored by the high calcium content of the soil (Lima, et al., 2002) encourages the transformation and immobilization of phosphorus from bone fragments through compounding with aluminum, stabilizing it in acidic conditions and mixing it to soil depths as low as 1.5 m (Schaefer, et al., 2004).

Nitrogen reacts predictably to the high carbon to nitrogen ratio of ADE, even though ADE soils may have higher levels of nitrogen than neighboring Ferralsols, by being mineralized more slowly and thus being less available to plants (Lehmann, da Silva, et al., 2003 and Lehmann, Kern, et al., 2003). Conditions on ADE are also ideal for nitrogen fixing bacteria and legume nodulation, with little to no exchangeable aluminum, lower levels of manganese and high levels of phosphorus and calcium. This may be another effect of charcoal in ADE, as charcoal amendments have been shown to stimulate the formation of the root nodules colonized by microbial nitrogen fixers (Steiner, Teixeira, Lehmann, Zech, 2004). Still, in spite of higher incidences of nitrogen fixing bacteria than adjacent soils, even legumes growing on ADE may be low in nitrogen(Lehmann, Kern, et al., 2003).

More research has begun in the last few years on the microbial populations distinct to charcoal amended soils. Charcoal has been found to stimulate higher overall levels of bacterial respiration and growth, with specifically higher populations of nitrogen fixing bacteria (Steiner, Teixeira, Lehmann, Zech, 2004) and actinomycetes (Clement, 2003), which are important in releasing nutrients from resistant organic compounds through decomposition (Brady, Weil, 2002).
ADE properties also appear to benefit the growth of certain fungi. A personal communication from M. de Lourdes Ruivo of the Museu Paraouse Emílio Goeldi in 2002 described a species of fungi she and her colleaugues found on ADE, but not surrounding soils (Clement, 2003). Charcoal also stimulates the colonization of crop roots by arbuscular mycorrhizal fungi (AMF), believed to result from creation of microhabitats for AMF hyphae aided by a microporous surface area of 400 m2/g (Steiner, Teixeira, Lehmann, Zech, 2004). AMF are known by themselves to increase plants’ access to water and phosphorus beyond what would be expected by other soil conditions (Brady, Weil, 2002), so it would be interesting to separate the magnitude of this effect from the overall nutritional profile of ADE.

On a larger scale of biological interaction, interest in soils often centers around the growth of plants. Plant biomass has been positively correlated with microbial population size and activity, which is aided by stable organic matter and increased nutrients such as exists in ADE soils (Steiner, Teixeira, Lehmann, Zech, 2004). ADE also appears to be preserving plant species that can’t survive elsewhere in the Amazon, one of the factors in their tendency to promote biodiversity (Clement, 2003).

There are cases where the properties of ADE have been seen to be lost and its nutrient-carrying organic material degraded. Unrelieved modern cultivation methods without the addition of fresh organic matter have been observed to lighten ADE, an indication of the loss of soil organic matter, and they must be protected from compaction like other agricultural soils or face a loss in their productivity. (German, 2002)

Yet in general, ADE is known to be resilient, recovering quickly in fallow and freshly cleared ADE is often fertile without additional inputs. Some ADE sites in western Amazonia are even known to have been under continuous cultivation without fertilization for 40 years (Lehmann, da Silva, et al., 2003). The most common reasons for the modern destruction or loss of ADE is mining for use as potting soil or construction fill (Erickson, 2003).

Modern Context

Laura German further investigated the uses to which ADE is being put today in the central Amazonian basin region near Manaus, Brazil in order to add to the picture of how these soils were likely used in pre-colonial times. (German, 2002) Currently, many traditional local farmers still practice rotating swidden agriculture, a form of migratory slash and burn with long fallow periods, which neither generates ADE nor supports large populations. Yet in spite of ADE soils’ greater fertility, indigenous and locally assimilated non-native farmers who were growing crops for subsistence tended to avoid them.

Using interviews and observations, German discovered that ADE was most likely to be used by farmers close to large market towns and for the cultivation of non-native crops. Farmers reported that the important local staples of cassava (manioc) and banana, as well as other fruit trees, produced poorly on ADE. The highly fertile soil yields native crop plants with abundant leaves and shoots, but small or stunted fruits and edible roots. Because of weed ingress, the plots also had to be returned to fallow faster. Though fallows didn’t have to be as long, the increased weeding and clearing labor doesn’t make these soils as attractive for the subsistence growing of the studied farms. (German, 2002)

Yet farmers reported that ADE was better for growing export crops like maize, beans, peppers and melons. Vegetable crops planted over small areas were also reported to produce very fast and thus seemed to balance out the issue of weeding labor to the satisfaction of the farmers. (German, 2002)

Similar observations were reported in a follow-up examination of agriculture and ADE in the blackwater ecosystems of the Rio Negro and Rio Urubú regions surrounding Manaus. (German, 2004) The interviews and crop measurements focused mainly on caboclos, those smallholders with non-indigenous ancestry that have adopted many native practices, along with some indigenous subsistence farmers and the owners of mechanized growing operations near Manaus that made use of ADE.

In pre-colonial times, the ADE soils would have been particularly important to the residents of these blackwater ecosystems who had to farm exclusively on the terra firme uplands. Because the rivers contain such a high proportion of organic acid leachate, their alluvium is acidic and infertile, unlike the rich varzéa floodplains in other Amazonian ecosystems that provide good crop land when they aren’t inundated (German, 2004).

Another study of the area near Manaus in 2005 attempted to quantify the difference in weeds produced between ADE and adjacent Ferralsols. (Major, DiTommaso, Lehmann, Falcão, 2005) Plots were compared under weeded and unweeded maize production, as well as by leaving prepared plots unseeded and free to grow whatever weeds were present in the local seedbank. A further seedbank comparison using soil cores from unprepared land was performed on ADE sites likely abandoned for at least 300-400 years and forest sites that showed no apparent evidence of human disturbance. Non-degraded ADE sites showed an across the board increase in weed diversity, though the maize on these sites generally performed better and responded well to weeding at critical stages or limited fertilizer application.

This indicates the likelihood that best practice management techniques for these sites simply require rediscovery and retransmission to a farming population that has already displayed its adaptability. Smallholders near Manaus have successfully used ADE for homegardens, which are in almost all cases more diverse than the typical swidden plot, sites that tend to be larger and may be kept in monoculture for a significant portion of their active cultivation time. Homegardens in the region are also seen to respond to market pressure by showing greater species dominance in these gardens, with proximity to towns where produce could be sold to larger markets. (Major, Clement, DiTommaso, 2005) In spite of a broken chain of transmission, considerable local knowledge has been amassed about the types of crops and practices that are best suited to ADE, which research has been able to corroborate in some cases (German, 2002, 2004).

Investigations of historical pollen data indicate that the rise of a majority of ADE sites corresponds with the introduction of maize and beans from Central America, higher nutritional value crops with shorter cropping seasons than cassava. (German 2002) Though because the generation of ADE predates non-native crop introduction, the initial technology seems likely to have developed for other reasons. Yet German also notes that though the current native population eats fresh maize, they have no tradition of using dry maize for human consumption and simply feed it to their animals, “suggesting the absence of (or historical disjunction with) a regional tradition of maize consumption.” (German, 2002)

These observations underscore both the difficulty and import of ADE research. People using stone tools, situated on some of the poorest and most fragile soils in the world, managed stable cultivation over spans of centuries, supporting population densities comparable to any of their contemporary societies. Geographer William I. Woods said that the region around the current day Brazilian town of Santarém alone likely supported “200,000 to 400,000 people a few centuries before the Spanish came,” putting it on par with the Aztec city Tenochtitlán, the biggest known city of the time. (Mann, 2002)

The utter collapse of the civilization that created ADE has virtually erased direct knowledge of their cultural practices and made it difficult to know which, if any, of them have continued to be passed down to the modern indigenous inhabitants. And yet their achievements in feeding more people than the region supports today without the benefit of metal tools or Green Revolution farming inputs is truly impressive.

Still, this should in no way downplay the current achievements of the Amazon region’s post-colonial indigenous populations. Studies of the presence of disturbance vegetation and the concentrations of useful food plants in the Xingú River basin of Brazil (Balée, 1990) and the rainforests of Guyana indicate that much of what was once considered virgin, climax or primary forest should be regarded as a cultural artifact. These forests in their current state are products of the shifting cultivation methods still employed by some of the descendants of societies like those who built the Beni mound fisheries and transformed collectively vast swathes of land into what continue to be some of the world’s richest soils.

Darna Dufour reviewed and summarized additional ways in which the rainforest is used today by some Amazonian natives, particularly their methods of swidden cultivation. The average swidden plot is cleared of its vegetation, which is burned and the ash mixed into the soil, then cultivated intensely for about two to three years. They may be planted initially with 2-16 types of root and herb crops, though up to 48 varieties of cassava (Dufour, 1990). Without access to crop insurance, money to purchase outside food, pesticides or reliable weather data, they can provide a varied diet for themselves and reduce their risk of total crop failure. These intercropping techniques act as a deterrent to the spread of disease or high concentrations of specialized pests, while also allowing them to plant varieties of edible plants suited to a wider range of environmental conditions.

As woody shrubs and saplings begin to encroach on the plot and cassava yields drop off, the swidden is managed as a prospective orchard with active selection for, or transplantation of, intermediate shrubs that continue to provide some type of food. Harvest at the orchard stage may continue for many years, with a higher proportion of food, fiber and timber species than would be present in unmanaged forest. A family group may at any one time be harvesting several plots at different successional stages and clearing new primary plots as the cassava yields drop off at the last one. Indigenous farmers also preserve access to useful plants and a ensure a broad genetic reservoir by transplanting useful roots and herbs to trailsides and campsites. (Dufour, 1990)

Other benefits of this managed succession from disturbance back to forest are more rapid biomass addition and high diversity, or low species dominance. Throughout the stages of cultivation, they attract game and edible insects, such as palm grubs, ants and termites. In one example, a certain variety of cassava was planted with the expectation of attracting a species of rodent that could be hunted by the Tukanoan women who tended the plots. (Dufour, 1990)

Susanna Hecht’s study of the Kayapó people of Gorotirê in Brazil’s Xingú River Basin gives more hints of the sophistication of previous techniques, but remain only as a subset of what’s been lost. (Hecht, 2003) The village of Gorotirê is a fairly young settlement, somewhere around a century old, which splintered off from a larger village community that broke up after a series of divisive disease epidemics made large permanent settlements untenable. The Kayapó have their own soil classification system, use different types of composting and mulching with ash or plant debris, cultivate medicinal plants and maintain nursery seedbeds for transplant. They also prefer to grow root crops, especially sweet potatoes and manioc, which have high phosphorus and potassium requirements that they meet through controlled burn strategies. By specifically mulching root crops with burned debris from potassium accumulating palm and banana plants, as well as ant nest ash, they ensure a steady supply of potassium in addition to the expected phosphorus that they can then fertilize crops with.

The Kayapó use of fire in the landscape also includes backyard hearths, cooking fires in the field to ease the handling of root crops and low-intensity fires to clear weeds and underbrush. These native types of landscape management were demonized by the British, who regarded shifting cultivation as a threat to political control and the use of fire as an unproductive technology. Such charges were also made against indigenous peoples in India and Africa. In all cases, disapproval of native means of subsistence agriculture was used as a justification for land appropriation and restricting native people to small subsets of their former lands. (Hecht, 2003)

The same logic and results were applied when British colonists came to North America, where the low-intensity burns started twice yearly by the natives of southern New England were considered a sign of laziness and made a clear case to the colonists that the land’s resources were underused. It did not help that Indian men secured their meat by hunting, which the British regarded as an upper class leisure activity. As in South America, colonists did not regard fire as a land management tool constituting an improvement of the land that would establish a legal claim. Once the Indians were cleared off the land, however, settlers complained that the park-like pastures dotted by trees that had attracted and fed so much game later became nearly impassable. (Cronon, 2003) Pushed out by cultural misunderstandings, displacement and disease epidemics, it seems surprising that any significant native knowledge of maintaining landscapes through fire has survived to the present. The rest must be constructed by study and experiment.

Though the Amazon rainforest is understandably a prime global conservation concern, the evidence indicates that the landscape can support larger populations than previously thought possible. Yet perhaps more importantly, the extent of human intent in shaping such a vibrant and diverse ecosystem suggests the viability of more proactive conservation strategies than mitigation of loss and the possibility that preservation of biodiversity and human habitation can both be pursued without contradiction.

The Amazon stands today as a living testament to a series of native cultures that saw no trade-off between providing for themselves and preserving biodiversity and not, as is commonly presented, an untouched wilds that can’t bear the touch of humanity. It should be seen both as the remarkable gene bank recognized by the conservation perspective and a significant repository of agricultural and agroforestry knowledge, of which ADE represents just a fraction.

Slash and Char: Making New Amazonian Dark Earth?

Researchers are currently exploring the possibility of cost-effectively replicating ADE in tropical agricultural soils. Most of the research so far has focused on charcoal additions and a method under consideration for adding it is the technique called slash and char, a low technology alteration to traditional slash and burn swidden agriculture as practiced in Brazil.

Instead of burning the vegetation cleared from fallow on the ground where it falls, the wood is burned in a kiln under low oxygen, high temperature conditions. The resulting charcoal tends to have a rate of 15% fine material that’s unsuitable for fuel or sale, a rate that increases with the addition of non-woody biomass to the kiln. These fine charcoal portions are suitable for application to the soil and while they may not increase soil fertility immediately, other advantages can be realized in the initial cropping season, such as increased water retention during the dry season. (Steiner, Teixeira, Zech, 2004)

The average carbon mass recovery from woody material burned incompletely through a charcoal process compared to its original carbon mass ranges from 29-50%. In contrast, a typical solid carbon recovery after slash and burn is around 3% (Glaser, et al., 2002). Using published data, it’s been estimated that a 4-20 year forest fallow in the Amazon could yield around 13 Mg (Mg: 106 grams) of carbon as charcoal per hectare (Sombroek, et al., 2003). Still, additions of as little as 1-3 Mg per hectare to the top 30 cm of soil could provide measurable agricultural benefits, with 1 Mg of charcoal requiring 9.45 m3 of wood and 26 days of work for one person using a kiln (Glaser, et al., 2002).

Not only is the potential pool of soil organic matter increased through slash and char, forest health and biodiversity benefits. Because charcoal can be made during the wet season, it can be made throughout the year in the Amazon. Spreading out the burning done for land clearing throughout the year could limit ozone concentrations that can harm plant growth, as well as cut down on localized aerosol accumulations that can block up to 40% of the solar radiation during the dry season and reduce precipitation over millions of kilometers. Without literally scorching the earth, as in typical slash and burn, it’s also possible for some species to regenerate from stumps or for species that lack fire-resistant seed to come back more readily. The saleability of the charcoal end product further encourages longer fallows in slash and char, 8-12 years on average, compared to the typical 5 years of more intensive slash and burn. The extra 3-7 can increase the biomass per hectare by over 50%. (Steiner, Teixeira, Zech, 2004)

It’s also possible that the charcoal addition itself promotes regrowth, as a trial in Zambia showed notable increases in the growth of native woody plants following amendment with charcoal. Yet because of wide variability among the few tests conducted so far, there’s no standard set of application guidelines for varying soil types and ph values. Its effects even vary with respect to water retention, increasing the field capacity of sandy soils while improving drainage in clay soils, which would have to be taken into account when considering the needs of different plant species. (Glaser, et al., 2002)

A trial amending samples of both an ADE and a Ferralsol with various fertilizer and manure treatments, and amending the Ferralsol with charcoal, yielded additional data about nutrient cycling for attempts to build soil organic matter to ADE levels using charcoal (Lehmann, da Silva, et al., 2003). Charcoal additions to the Ferralsol reduced most types of leaching while increasing plant nutrition, nutrient availability and the ratio of nutrient uptake to leaching, leading to improved efficiency of applied fertilizers. However, more nitrogen was immobilized in the charcoal trials due to the high carbon to nitrogen ratio and potassium leaching was increased. The most striking results with the ADE trials, aside from increased plant growth when comparing the unfertilized controls, was its poor response to inorganic fertilizers. While the ADE trials did well with additions of chicken manure, they had very high rates of leaching when inorganic fertilizers were used and the authors suggested that it could be better to stay with organic fertilizers when growing crops on ADE.

Considerations For Future Study And Application

An urgency may underlie research into ADE techniques as humanity faces population increases beyond today’s teeming billions, while already co-opting at least 31-50% of terrestrial primary productivity out of terrestrial ecosystems (Vitousek, et al., 1986 and Vitousek, et al., 1997). Land transformations account for approximately 20% of anthropogenic carbon dioxide emissions (Vitousek, et al., 1997), which have “increased from 290 to 370 ppm during the past century alone” (Brady, Weil, 2002).

Carbon-containing soil organic matter is critical when considering the global carbon cycle and methods of carbon sequestration to reduce the accumulation of greenhouse gases, as soil contains around three times the amount of carbon that’s present in the world’s plant cover. Yet soils don’t just absorb carbon in the form of buried organic matter, the processes of microbial decomposition that take place within soils also release carbon dioxide to the atmosphere. Today, 2 Pg/yr (Pg: 1015 grams) more carbon are released from soils than are added to them, while an additional 5 Pg/yr are released from fossil fuel use. Under typical circumstances, 20-33% of plant material is added to the soil, while only about 20% of that will eventually be transformed into humic substances that are likely to sequester their carbon over long periods of time. (Brady, Weil, 2002)

Under these circumstances, soil management techniques that could stabilize up to 50% of the carbon contained in local woody biomass, while at the same time stimulating plant regrowth that would take more carbon out of the atmosphere, hold significant promise.

In spite of their potential, many questions remain about ADEs and most current knowledge is based on a very limited data set. No one has fully determined how a soil is transformed from a Ferralsol to an Anthrosol with all the properties of the archaeological ADE. Other major gaps in current knowledge include the full nutrient dynamics in these soils, the precise chemical transformations that make pyrogenic carbon reactive with soil nutrients, a quantification of the microbial diversity within ADE and how it’s affected by land use, as well as the best way to sustain its fertility with prolonged use (Madari, et al., 2004). Another neglected area of study concerns the pottery fragments in ADE, which are removed for soil sampling, so nothing is known about any possible effects they may have on in situ soil properties (Erickson, 2003). While exploring ways to recreate ADE, estimates are needed of how much primary productivity ends up as stable soil organic matter and black carbon, as well as finding cost-effective means for small tropical farms to use any techniques discovered (Sombroek, et al., 2003).

Sequestering carbon through charcoal addition to agricultural soils could have applications that go well beyond tropical regions like those that were home to the pre-Colombian Amazonians. However, many local factors would have to be taken into account in determining the suitability of these techniques. Here on the Pacific Northwest coast of the United States there are also areas where high annual rainfall has created leached, acidic soils. Yet the native woody vegetation is predominated by conifers and other acid-loving plants.

If attempts were made to replicate ADE in some fashion in this region, there would be questions to answer about its sustainability using local materials. While hardwood charcoal has been observed to raise CEC, conifer charcoal was seen in one trial to decrease it (Glaser, et al., 2002). Decreasing soil acidity would pose a problem for acid-loving native species, though conifer charcoal doesn’t lower soil pH as much as hardwood charcoal and any such soil treatment would have less of an affect on the pH of higher clay soils (Glaser, et al., 2002) such as those common to parts of the Northwest. Red alder, or Alnus rubra, is a hardwood that grows readily in disturbed regions of the coastal Northwest, but there’s no guarantee that it would replicate the properties of the typically denser, tropical hardwoods. Regarding nutrient availability, nitrogen is chronically lacking in Northwest soils and charcoal may reduce the leaching of nitrogen compounds, but may also immobilize it in forms inaccessible to plant roots (Lehmann, da Silva, et al., 2003), as noted above.

Other properties observed in tests on charcoal suggest uses that go beyond agriculture and carbon sequestration. Charcoal not only holds on to nutrients and water, but is capable of adsorbing polar pesticides, non-nutritive metal ions and hydrophobic polycyclic aromatic hydrocarbons (Glaser, et al., 2002). Considering the pollution hazards posed to waterways by these types of substances, it could be worthwhile to investigate whether charcoal additions would be useful mixed into the soil beneath greenbelts that already act as pollution breaks for runoff between agricultural or industrial zones and open water. In this respect, wood charcoal might not even be the most useful. Data referenced in a communication of R.C. Srivastava of the Industrial Toxicology Research Center in Lucknow, India indicates that activated coconut charcoal is over 2.5 times more absorbent of mercury than wood charcoal (Srivastava, 2002) when used in water filtration.

Returning to the question of promoting sustainable agriculture over the long term, would the additional absorbency of coconut charcoal translate to differences in CEC or soil field capacity if used as a soil amendment? Even if it didn’t, it seems possible that there could be other organic materials that could be put through a charring process and used to increase the soil pool of long-lived organic matter that might be either more effective or more suitable to local resource availability.

While questions about ADE and the properties of its soil organic matter seem pertinent to the global community, they are of significant personal importance to residents of tropical rainforest regions who must find a way to feed their families off the land that is available to them in often precarious wage economies.

Overpopulation has generally been blamed for rainforest destruction, this seems to be an insignificant to nonexistent factor in comparison to the lack of local food security. When large, commercial plantations take up all of the best farmland in an area, local residents are pushed onto increasingly marginal land that they might not have otherwise used for agriculture. When those plantations attract workers from outside their local area and have to shut down because of bust cycles in commodity prices, their stranded workers are left with the same problems of finding suitable land that faced the previously displaced inhabitants. Often, these dislocated people end up following logging trails deeper into pristine rainforest to take advantage of the work done by loggers to clear the forest, with the transformation of land from forest to permanent agriculture often causing more damage than the logging. (Vandermeer, et al., 2005)
With the potential for ADE to point the way to more effective carbon sequestration, as well as to inspire ideas for sustainable agriculture in environmentally sensitive regions like tropical rainforest zones, these Amazonian artifacts certainly deserve more study.


Many sources referenced in this paper were themselves reviews of other research or contained citations of past research that I did not have the time to read and formally reference here. For now I can only offer thanks and apologies to those whose observations have been credited here secondhand.

This report could not have been compiled without the Evergreen State College’s library services, interlibrary loan support and research journal database access.

Literature Cited

Balée, W; Campbell DG. (1990) Evidence for the sucessional status of Liana forest (Xingú River Basin, Amazonian Brazil). Biotropica, vol. 22 1:36-47, by The Association for Tropical Biology and Conservation.

Brady N; Weil R. (2002) The Nature and Properties of Soils: Thirteenth Edition. Copyright by Pearson Education, Inc. Prentice Hall publishers, New Jersey.

Clement CR; McCann JM; Smith NJH. (2003) Agrobiodiversity in Amazonia and its relationship with Dark Earths. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 153-178. Kluwer Academic Publishers, Netherlands.

Cronon W. (2003) Changes in the Land: Indians, Colonists, and the Ecology of New Enngland. Hill and Wang, a division of Farrar, Straus and Giroux, New York, NY.

Denevan W. (1996) A Bluff Model of Riverine Settlement in Prehistoric Amazonia. Annals of the Association of American Geographers, Vol. 86, No. 4, Dec., 1996, 654-681.

Dufour, DL. (1990) Use of tropical rainforests by Native Amazonians. BioScience, Vol. 40 9:652-659.

Erickson, C. (2003) Historical ecology and future explorations. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 455-500. Kluwer Academic Publishers, Netherlands.

Falcão NPDS; Comerford N; Lehmann J. (2003) Determinging nutrient bioavailability of Amazonian Dark Earth soils – methodological challenges. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 255-270. Kluwer Academic Publishers, Netherlands.

German LA. (2002) Historical contingencies in the coevolution of environment and livelihood; contributions to the debate on Amazonian Black Earth. Geoderma, Vol. 111 (3-4), Feb 2002, 307-331.

German LA. (2004) Ecological praxis and blackwater ecosystems; A case study from the Brazilian Amazon. Human Ecology, Vol. 32 (6), Dec. 2004, 653-683.

Glaser B; Lehmann J; Zech W. (2002) Ameliorating physical and chemical properties of highly weathered soils in the tropics with charcoal – a review. Biology and Fertility of Soils, Vol. 35 (4), Jun 2002, 219-230.

Glaser B; Guggenberger G; Zech W. (2003) Organic chemistry studies on Amazonian Dark Earths. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 227-241. Kluwer Academic Publishers, Netherlands.

Glaser B; Guggenberger G; Zech W; Ruivo MDL. (2003) Soil organic matter stability in Amazonian Dark Earths. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 141-158. Kluwer Academic Publishers, Netherlands.

Glaser B; Zech W; Woods WI. (2004) History, current knowledge and future perspectives of geoecological research concerning the origin of Amazonian anthropogenic dark earths (Terra Preta). B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 145-158. Springer-Verlag, Germany.

Glaser B; Guggenberger G; Zech W. (2004) Identifying the Pre-Columbian anthropogenic input on present soil properties of Amazonian Dark Earths (Terra Preta). B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 145-158. Springer-Verlag, Germany.

Hart J; McNeilan R. (1990) Fertilizing Guide: Fertilizing Home Fruit, Vegetable and Ornamental Gardens. Fertilizing Guide 66, Oregon State University Extension Service.

Hecht S. (2003) Indigenous soil management and the creation of Amazonian Dark Earths: implications of Kayapó practices. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 355-372. Kluwer Academic Publishers, Netherlands.

Herrera LF; Cavalier I; Rodriguez C; Mora S. The technical transformation of an agricultural system in the Colombian Amazon. World Archaeology, Vol. 24, 1:98-113.

Kern D; Lima da Costa M; Frazão FJL. (2004) Evolution of the scientific knowledge regarding archaeological Black Earths of Amazonia. B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 19-28. Springer-Verlag, Germany.

Lehmann J; da Silva JP; Steiner C; Nehls T, Zech W, Glaser B. (2003) Nutrient availability and leaching in an archaeological Anthrosol and a Ferralsol of the Central Amazon. Plant and Soil, Vol. 249 (2), Feb. 2003, 343-357.

Lehmann J; Kern D; German L; McCann J; Martins GC; Moreira A. (2003) Soil fertility and production potential. J. Lehmann, et al. (eds.), Dark Earths: Origin, Properties, Management, 105-124. Kluwer Academic Publishers, Netherlands.

Lehmann J; Campos CV; de Macêdo JLV, German LA. (2004) Sequential P fractionation of relict anthropogenic Dark Earths of Amazonia. B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 113-123. Springer-Verlag, Germany.

Lima HN; Schaefer CER; Mello JWV; Gilkes RJ; Ker JC. (2002) Pedogenesis and pre-Colombian land use of ‘Terra Preta Anthrosols’ (‘Indian Black Earth’) of Western Amazonia. Geoderma, Vol. 110 (1-2), Nov 2002, 1-17.

Madari B; Sombroek WG; Woods WI. (2004) Research on anthropogenic Dark Earth soils. Could it be a solution for sustainable agricultural development in the Amazon? B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 170-181. Springer-Verlag, Germany.

Major J; Clement CR; DiTommaso A. (2005) Influence of market orientationon food plant biversity of farms located on Amazonian Dark Earth in the region of Manaus, Amazonas, Brazil. Economic Botany, 59(1) pp. 77-86.

Major J; DiTommaso A; Lehmann J; Falcão NPS. (2005) Weed dynamics on Amazonian Dark Earth and adjacent soils of Brazil. Agriculture, Ecosystems and Environment, 111:1-12.

Mann C. (2000) Earthmovers of the Amazon. Science, Vol. 287 (5454), Feb 2000, 786-789.

Mann C. (2002) The forgotten people of Amazonia. Science, Vol. 297 (5583), Aug 2002, 921.

Ponomarenko EV; Anderson DW. (2001) Importance of charred organic matter in Black Chernozem soils of Saskatchewan. Canadian Journal of Soil Science. 81:285-297.

Radlein D; Piskorz JK; Majerski P (1996) Method of producing slow-release nitrogenous organic fertilizer from biomass. European patent application 0716056 A1

Schaefer CEGR; Lima HN; Gilkes RJ; Mello JWV. (2004) Micromorphology and electron microprobe analysis of phosphorus and potassium forms of an Indian Black Earth. Australian Journal of Soil Research, Vol. 42 (4), 2004, 401-409.

Sombroek W; Ruivo MDL; Fearnside PM; Glaser B; Lehmann J. (2003) Amazonian Dark Earths as carbon stores and sinks. Lehmann J; Kern DC; Glaser B; Woods WI (eds.), Amazonian Dark Earths: Origin, Properties, Management, 125-139. Kluwer Academic Publishers, Netherlands.

Srivastava, RC. (2002) Global Mercury Assessment: Comments and inputs of draft assessment report. (pdf) Personal communication to Dr. James B. Willis, Director, Global Mercury Assessment, UNEP Chemicals, Geneva, Switzerland. Industrial Toxicology Research Center, Lucknow, India. Accessed May, 2006.

Steiner C; Teixeira WG; Lehmann J; Zech W. (2004) Microbial response to charcoal amendments of highly weathered soils and Amazonian Dark Earths in Central Amazonia – preliminary results. B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 195-212. Springer-Verlag, Germany.

Steiner C; Teixeira WG; Zech W. (2004) Slash and Char: An alternative to slash and burn practiced in the Amazon Basin. B. Glaser, Woods WI. (eds.), Amazonian Dark Earths: Explorations in Space and Time, 183-193. Springer-Verlag, Germany.

Vandermeer J; Perfecto I. (2005) Breakfast of Biodiversity: The Political Ecology of Rain Forest Destruction. Food First Books, Oakland, CA.

Vitousek PM, Ehrlich PR; Ehrlich AH; Matson PA. (1986) Human appropriation of the products of photosynthesis. BioScience, 36(6):368-373.

Vitousek PM, Mooney HA. (1997) Human domination of Earth’s ecosystems. Science, 277(5325):494-99.

Posted by natasha at June 1, 2006 06:20 AM | Science | Technorati links |

Had heard about Amazonian Dark Earth, and being a backyard organic grower, was intrigued. Your piece was just the information I needed to make the subject hum.

From your paper, it seems there's a very exciting tradeoff possible between those who desire carbon sequestration and those who could benefit from intense soil-building and cultivation therefrom. This prospect would also address world hunger, thus hitting three hot-button issues:
* hunger abatement
* carbon sequestration
* conservation of remaining intact wildlands

When the techniques you identify have been firmed up, financing could be provided by a carbon tax levied in the industrial north, part of which--perhaps 10%--could be for sequestration.


On a critical note: I think there's something wrong with this:

"It’s now estimated that ADE plots comprise between 0.1-0.3% of the Amazon Basin, or an area at least 6 million km2 (Sombroek, 2003), or a total area approaching two thirds the area of the United States."

I think you mean that the total area of the Amazon Basin is approximately 2/3 that of the US. So, a fraction of one percent of that would be, what?, maybe the Hawaiian islands? Or several times that?

Posted by: Jared Scarborough at June 2, 2006 06:08 PM