Louisiana State University LSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1993 Pedogenic Development of Soils at High Altitude in Burundi. Salvator Kaburungu Louisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

Recommended Citation Kaburungu, Salvator, "Pedogenic Development of Soils at High Altitude in Burundi." (1993). LSU Historical Dissertations and Theses. 5646. https://digitalcommons.lsu.edu/gradschool_disstheses/5646

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9419900

Pedogenic development of soils at high altitude in Burundi

Kaburungu, Salvator, Ph.D.

The Louisiana State University and Agricultural and Mechanical Col., 1993

UMI 300 N. Zeeb Rd. Ann Arbor, M I 48106

PEDOGENIC DEVELOPMENT OF SOILS AT HIGH ALTITUDE IN BURUNDI

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mecanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy

in

The Department of Agronomy

by Salvator Kaburungu B.S., University of Burundi, 1981 M.S., Louisiana State University, 1990 December 1993 ACKNOWLEDGMENTS

I would like to express my sincere and deep appreciation to my major professor, Dr. Wayne Hudnall, for his continuous guidance, supervision and encouragement. I feel greatly honored to have worked with him on both my

M.S. and Ph.D. programs. My gratitude is also extensive to the Hudnall’s family, Ms.(Callalya) Hudnall and son Eric for their friendship and support.

I am sincerely grateful to my graduate committee , Dr. S. E. Feagley,

Assoc. Professor, Dept, of Agronomy; Dr. J. S. Hanor, Professor, Dept, of

Geology and Geophysics, Dr. W. H. Patrick, Jr., Boyd Professor, Wetland

Biogeochemistry Institute, Dr. R. P. Gambrell, Professor, Wetland

Biogeochemistry Institute, Dr. W. A. Young, Professor, Dept, of Horticulture,

Mr. B. A. Touchet, SCS - State Soil Scientist; and Dr. F. Rhoton, USDA -

ARS; for their education, guidance and time devoted to review this manuscript.

I wish to express my sincere and deep appreciation to AAI/ATLAS for their financial support on both my graduate degrees without which I could not have undertaken the studies. A deep appreciation is also extended to Ms

Michelle Roberts, my educational adviser, and her assistant, Ms Jackie Allen, for their continuous support.

I would also like to express appreciation to Scott and Michelle Nesbit for their friendship which made my stay at Louisiana State University an enjoyable experience. Special thanks are extended to Xiaowei Zhang who, in spite of her hectic schedule, managed to help me with the typing of this manuscript while taking her Ph. D. general exams. Last but not the least, I am grateful to Mr. A. Gakoko, ISABU, Mr. A. Ndihokubwayo, ISABU and J.P.

Muganza, ISABU, who kindly helped me to describe and collect soil samples. TABLE OF CONTENTS

ACKNOWLEDGMENTS ...... ii

ABSTRACT ...... iv

CHAPTER 1 PEDOGENIC DEVELOPMENT OF SOILS AT HIGH ALTITUDE IN BURUNDI: A LITERATURE REVIEW ...... 1 1.1. Introduction ...... 1 1.2. Rationale ...... 2 1.3. Brief Characterization and Classification of the Soils ...... 3 1.3.1. Soils under Natural Forest...... 3 1.3.2. Soils under Pasture ...... 6 1.3.3. Evolution of Soils under Cultivation ...... 14 1.3.3.1. Effects of Cultivation on Selected Soil Properties ..... 15 1.3.3.2. Effects of Cultivation on Organic Matter Content and Bulk Density ...... 18 1.3.3.3. Effects of Cultivation on Soil Morphology and hydraulic Conductivity ...... 21 1.3.3.4. Effects of Cultivation on Both Surface and Subsurface Horizons ...... 21 1.3.4. Effects of Cultivation on Soils at High Altitude in Burundi ...... 22

2 EFFECTS OF LONG-TERM CULTIVATION AND PASTURE FOLLOWING DEFORESTATION...... 25 1.1. Introduction ...... 25 2.2. The Study A re a ...... 27 2.2.1. Background Information on Burundi ...... 27 2.2.2. Location of the Study A re a ...... 30 2.2.3. Land U se ...... 30 2.2.4. Geological Setting ...... 32 2.2.5. Clim ate ...... 33 2.3. Materials and Methods ...... 33 2.3.1. Experimental Design and Field Methods ...... 33 2.3.2. Laboratory Analysis ...... 37 2.4. Results and Discussions ...... 38 2.4.1. Morphological Comparisons ...... 38 2.4.2. General Chemical and Physical Comparisons ...... 41

3 MINERALOGICAL PROPERTIES AND AGGREGATE STABILITY ...... 57 3.1. Introduction ...... 57

iv 3.2. Materials and Methods ...... 58 3.3. Results ...... 58 3.4. Discussions ...... 62

4 GENESIS AND CLASSIFICATION OF HIGH-ALTITUDE SOILS IN BURUNDI...... 78 4.1. Introduction ...... 78 4.2. Materials and Methods ...... 80 4.2.1. Field Setting ...... 80 4.2.2. Laboratory Analyses ...... 81 4.3. Results ...... 82 4.3.1. Forest S o ils ...... 82 4.3.2. Culture S oils ...... 86 4.3.3. Pasture S o ils ...... 90 4.4. Discussions ...... 91 4.4.1. Genesis of High-Altitude Soils in Burundi ...... 91 4.4.2. Evolution of the Soils Following Deforestation ...... 114 4.4.3. Classification of High Altitude Soils in Burundi ...... 115

5 SUMMARY AND CONCLUSIONS ...... 122 5.1. Summary ...... 122 5.2. Conclusions ...... 125

LITERATURE CITED...... 129

APPENDIX A ...... 141 FIELD DESCRIPTIONS ...... 141

APPENDIX B ...... 160 SELECTED SOIL DATA...... 160

VITA...... 182

v ABSTRACT

To evaluate the effects of long-term cultivation and pasture on

deforested soils at high altitude in Burundi, comparisons of morphological,

physical, chemical and mineralogical properties of soils under natural forest

with their counterparts under cultivation and pasture were performed. A

typical soil under natural forest classifies as fine, oxidic, isomesic, Typic

Haplorthods. Cultivated soils classify as fine, oxidic, isothermic Humic

Hapludoxs. Pastured soils classify as fine, oxidic, isothermic, allic Humic

Hapludoxs. When the sombric horizon is prominent, they classify as

Sombriudoxs. We proposed a Spodic subgroup for the Hapludoxs.

The studied area was highly weathered prior to faulting and uplifting

and the spodic properties formed during the cool and humid parts of the

Pleistocene Epoch. On 32 selected parameters, comparisons showed

treatment effects on 18 of them, namely: OC, Bd, water retention, water

dispersible clay, N, Bray I P, Ca, Mg, K, BaCI2 acidity, extractable acidity,

extractable Al, extractable H, NH4OAc CEC, Sum CEC, NH4OAc BS, Sum

BS, and Al saturation. Seven of these parameters {OC, Bd, water retention,

water dispersible clay, BaCI2 acidity, extractable H and Sum CEC) were

significantly different on both surface and subsurface horizons. Substantial

improvements of pastured soils are expected when the latter are cultivated

with continuous addition of O.M. (30 tons/ha/yr). The studied soils are rich in organic matter, gibbsite and iron oxides.

All three components are well known for their binding effects. The soils exhibit steep slopes and rainfall erosion constitutes a serious limitation to land use. Removal of iron oxides and organic matter resulted in increased clay dispersion. This increase was far from total dispersion. Aggregation at the study area involves Al, besides Fe oxides and organic matter.

Addition of organic matter to pastured soils tremendously improves their productivity though they are already rich in organic matter. The organic matter present in pastured soils does not contribute to the nutrient supply for plants. Characterization of the humic substances should elucidate the mechanisms involved in organo-mineral associations. CHAPTER 1

PEDOGENIC DEVELOPMENT OF SOILS AT HIGH ALTITUDE IN BURUNDI: A LITERATURE REVIEW

1.1. Introduction

The effects of cultivation on deforested soils can be assessed by

comparison of a cultivated soil with its forested counterpart. The influence of

soil management practices on the long-term maintenance of fertility and tilth

has been studied at many locations around the world. General conclusions

can be drawn, but the effects of a given practice may not be identical on all

soil types nor under all climatic conditions.

Most of the soils at high altitude in Burundi have been cleared of their

native forests and used for cropping to support increasing human and animal

populations. These land-use changes have created problems. Since the

forest has been cleared, sheet erosion and gullies made some of the steep

erosive soils practically useless for cropping. Some soils have eroded to the

bedrock, hardpan, or less fertile subsoils. Presently, soil degradation caused

by deforestation and soil misuse is a severe problem at the study area.

Whereas deterioration of cultivated soil without use of either manure or

organic residue from harvesting occurred within a few years, cultivation

supported by adequate amount of organic matter have shown sustainable

agriculture and increase in productivity. Decline in organic matter content

following deforestation increases the soils susceptibility to erosion and

1 2

enhances the processes of soil degradation (Lai, 1984). Mulch-farming

reduces physical degradation by improving soil organic matter content and

enhancing activity of soil fauna. Regular addition of organic matter improves

soil structure by serving as food for soil fauna, which creates the desired

porosity and promotes formation of polysaccharide and other compounds that

stabilize the micro aggregates (Oades, 1984).

Improved morphological, physical and chemical properties brought

about by cultivation have been observed by comparison of a cultivated soil with its counterpart under forest. The soils evolve from Haplorthods to

Hapludox within 20 years of cultivation supported by addition of adequate

amount of organic matter. There is, however, a lack of precise quantitative

published data about these improvements. Many changes in the soil system as a consequence of cultivation are subtle.

1.2. Rationale

The physical, chemical, mineralogical and morphological changes that a soil undergoes from a virgin to cultivated state is of interest to agricultural workers in understanding soil management problems. These problems are concerned not only with fertility, but the physical properties that affect water intake, soil aeration, aggregate stability and others. Changes that occur during a long period of cultivation may be a guide for soil scientists to understand and predict the type of problems that they will face in the future in keeping the soil productive. The main objectives of this study were: 1) to provide morphological descriptions, physical, chemical and mineralogical data of the soils, and therefore ascertain the effects of cultivation on surface soil horizons, and, if any, on lower horizons, 2) to decipher the basic processes involved in soil development, 3) to classify the soils according to Soil Taxonomy to the family level, and 4) to provide statistical validity to selected effects of cultivation. This should help to evaluate the soils and to adjust land uses to the limitations and the potentials of the natural resources and environment, thereby to attempt to limit soil-related failures in land uses.

1.3. Brief Characterization and Classification of the Soils

1.3.1. Soils under Natural Forest

The soils under natural forest are part of a complex ecosystem that concentrates plant and microbial nutrients in the surficial layers. Despite the luxurious biomass covering the soils, they are highly weathered acid soils.

The luxurious vegetation is sustained by an efficient cycling and recycling of organic matter in the soils that results in the overall improvement in soil conditions, the development and maintenance of soil structure, the improvement of physical properties, the decreased susceptibility to erosion, the encouragement of microbial activity and the provision of potentially available plant nutrients (Roose, 1977).

The favorable soil structure is produced by large quantities of roots and by the decomposition of continually renewed and abundant organic matter.

Organic matter influences productivity of soils through the mineralization of nutrients, its high cation exchange and water-holding capacities and its ability to improve soil physical properties. Organic matter is composed of many substances that are often complexed with the fine mineral fraction in soils.

These organic compounds may be transient or rather long-lasting, depending upon their resistance to microbial decomposition. Some organic substances are combined with the clay fraction in aggregates so that they are not accessible to microbial decomposition (Larson and Clapp, 1984).

Soil organic matter represents a particular stage in the continuous exchange of C, H, O, N, P and S between living material and soil minerals.

An equilibrium condition usually exists whereby, as new organic matter is formed, a portion of the older material is mineralized. The two main processes - both microbial in origin - are: 1) the degradation of plant and animal residues with modification of their composition and properties and 2) the synthesis of new microbial cells, which in turn die and are used by other microorganisms. In view of the diversity of both organic materials and microorganisms, it is evident that there is wide variation in the soil organic fraction (Lai, 1984). Soil organic matter distribution in soils can generally be attributed to several factors: 1) climate and vegetation (Jenny, 1941), and 2) human intervention such as manure application (Russell, 1973).

A typical soil profile under this natural forest was reported in the Tenth

International Workshop Proceeding on Soil Taxonomy and Agrotechnology

Transfer held in Burundi and Rwanda (1985). The soil exhibits a litter layer consisting of superposed thin layers of undecomposed, partially decomposed and completely decomposed leaves that have been successively deposited on the soil surface. The organic layer is about 10 cm thick. Beneath the litter occurs a dark horizon about 25 cm thick that exhibits the following properties:

granular structure

sand grains free of finer material

base saturation between 10 and 30%

organic carbon content between 5 and 8%

C/N between 13 and 15

pH of about 4.5 in water.

The mineralogy of the soils is dominated by kaolinite and gibbsite. The high amount of CEC encountered (CEC > 100 cmol(+)kg_1) results from the high content of organic matter. The main processes of soil formation in the study area are podzolization for soils under forest, rainfall erosion and desilication

(ferrallitization and ferritization) following deforestation.

Podzolization is defined as the process by which sesquioxides are translocated in a soil profile (Stobbe and Wright, 1959) and can be explained in part by the solubilities of ferrous and ferric iron. The soluble ferrous iron forms at the sites of eluviation and the insoluble ferric iron forms at the point of illuviation (Buol, 1980). Under natural forest, organic matter is added to the soil surface as litter (Oi horizon) and roots, which are typically very prolific in these soils in the lower O horizons (Oe and/or Oa horizons). The decomposition (transformation) of the organic matter, predominantly by fungi, produces mobile fulvic acids. Cheluviation of iron, aluminum and other

"metals" from upper horizons occurs, together with illuviation, i.e. saturation of fulvic acids by iron and/or aluminum (Fanning, 1989). Because the B horizon illuvial iron maximum usually occurs below the illuvial humus maximum, it appears that there may be some movement of iron within the B horizon by reduction and leaching in the Bh horizon followed by oxidation and precipitation in the Bs horizon (Fanning, 1989).

Podzolization in soils under natural forest at high altitude in Burundi has been recognized since 1985 (SMSS and IBSNAT, 1985). The common end products of podzolic weathering are kaolinite, iron and aluminum sesquioxides and hydrates occurring predominantly as amorphous material. Kaolinite and gibbsite are the common mineral identified. These soils exhibit a thick umbric epipedon and a spodic horizon. They are classified as Typic Haplorthods.

It has been suggested to classify them as Flumic Haplorthods due to their high content of organic matter.

1.3.2. Soils under Pasture

Deforestation is always followed by land degradation. Land degradation means deterioration or decline in the productive capacity of an ecosystem through adverse changes in life supporting processes of its climate, vegetation, soil and/or water resources (Lai and Cummings, 1979).

Deforestation in the study area has been performed by cutting trees manually and clearing the land by use of fire. Until the 1960’s, pasture management had been based on use of fire. Fire is the most ancient agricultural tool and is still used widely in the tropics to clear the forest, manage the residues and improve pasture quality.

Effects of fire on soil and vegetation depend on the fire’s intensity and duration. Depending on the biomass available, the temperature may be as high as 500 - 1000° C (Lai and Cummings, 1979, Goudie, 1981). Soil properties, especially soil moisture retention and water-transmission properties are greatly altered by fire. Excessive burning causes loss of nitrogen and other nutrients by volatilization and makes nutrients available for leaching during the rainy season (Lai, 1984). LeHouerou (1977) estimated that bush fires in the African grasslands burn more than 80 million tons of forage per year. Not only could that amount feed about 20 million cattle every year, but considerable amounts of nutrients released in the ash are washed away or leached from the root zone. Development of savannas in mid-latitudes and shrublands are attributed to repeated cycles of natural or man-caused fires.

Fire alters vegetation through evolution of some fire-tolerant species. Fire helps break seed dormancy and stimulates vegetation regrowth of many species. Repeated cycles of fires are responsible for replacement of fast- growing, light-loving trees and shrubs with fire-tolerant grasses and shrub vegetation. Deep-rooted trees that help in nutrient recycling are eliminated from a fire-prone ecology (Cloudsley-Thompson, 1977). Although measurements of rainfall erosion at the study area are underway and results are unpublished, measurements of soil erosion at similar agroecological zones are available. In East Africa, Pereira et al.

(1961) reported the effects of grazing planted pastures on soil structure, runoff and erosion. They observed that under very intense rainfall, a characteristic of the study area, sealing of the exposed and trampled soil surface resulted in heavy runoff and severe sheet and gully erosion. Rainfall infiltration rates in heavily grazed pastures were much lower than in soil under seasonal crops. Recent studies on grazing at the International Institute of Tropical

Agriculture also have shown more soil compaction, low infiltration rate and higher runoff and erosion than under grain crops (Lai, 1984). Measurements of erosion rates from field runoff have been made for Southern Africa by

Hudson (1971) and by Elwell and Stocking (1982), for Eastern Africa by Rapp et al. (1972), for Western Nigeria by Wilkinson (1975) and Lai (1976), for

Francofone West Africa by Roose (1977). These studies reported that erosion rates of 100 t/ha/year from arable lands are not uncommon. Brown

(1981) estimated that as much as one billion tons of topsoil are lost from the

Ethiopian highlands each year. The average annual soil loss from

Madagascar is estimated to be 25-40 t/ha/year over the whole country (Finn,

1983). Severe erosion also occurs in most parts of sub Sahara Africa.

UNEP (1984) estimated that a total of 87% of the Near East and Africa north of the Equator are in the grip of accelerated erosion. In Zimbabwe, the FAO/UNEP reports that in 1974, only 50 years after most of it was first

opened for cultivation, 41% of the land was already affected by erosion and

of this 12% was moderately to severely damaged.

Another important consequence of deforestation is the biological degradation of the soils. The declining soil organic matter content, the reduced biomass carbon, the decrease in the biological activity and the reduced diversity of soil fauna are the basic factors responsible for soil biological degradation. Because of prevailing optimum soil and air temperatures throughout the year, biological soil degradation is more severe in the tropics than in the temperate zone. The rate at which soil organic matter decomposes is doubled for every 10° C increase in temperature.

Consequently, the organic matter content of soil is usually lower in the tropics than in the temperate regions. There are also differences in soil temperature and soil moisture regimes. Decrease in food availability and diversity are important factors that reduce both the activity and diversity of soil fauna (Lai,

1986). Also, Woodwell et al. (1978) believed that deforestation in the tropics could be a major source of C02 by releasing the vast amount of carbon immobilized in the biomass.

For soil developing on steep slopes (up to 85%), rainfall erosion plays an important role as a process of soil formation on both cultivated and soils kept under pasture. In effect, steep erosive soils at the study area have been cleared of the natural forest and used for cropping where they should have 10 been carefully protected under forest. Some economically poor countries, however, have no choice but to expand their agricultural land by converting tropical rainforest to arable land. Planners in these countries should be provided data on appropriate land use and ecologically compatible soil and crop management practices (Lai. et al., 1987). A marked effort to reforest steep slopes is apparent, but the large human population and the grazing animals make revegetation of the study area very difficult. The A horizons generally contain the most nutrients and best structure. Materials eroded from the upper slope and deposited down slope may actually cover the crop with soil material thus decreasing crop yields and plant growth (Cook, 1936).

When the forest is cleared, the raindrops fall upon the bare soil, and by their impact, loosen the soil particles so that they may be more easily transported by flowing water. Other particles and aggregates plunge into the flowing water and by giving up their kinetic energy in the production of turbulence in the liquid, increase the power of the runoff to scour and transport (Cook, 1936).

Near the top of the slope are found the heads of innumerable small channels that have cut but a fraction of a centimeter into the topsoil. If one follows these "micro-channels" down the slope they are found to coalesce to form larger water courses. These entrench themselves deeper into the soil before they are united with neighboring channels. This process is repeated time after time; rivulets join rills, rill unites with rill and gullies develop. These, 11 in turn, coalesce to form larger and larger channels until at a point below that at which the flow is finally discharged into a so-called "stable" channel, there exists a miniature drainage system of extreme complexity. The rapid increase in the erosiveness of the runoff as it gathers volume and velocity in its descent is made strikingly evident by the increasing dimensions of the erosion channels. Because of the tremendous scouring power generated, the flow capacities of the principal erosion channels often become far greater than necessary to pass the run-off delivered to them. In extreme instances, gullies many meters deep are formed by the runoff from a few acres (Mathieu, 1985).

After deforestation, the litter disappears due to mineralization or to rainfall erosion. On land covered by vegetation, a large portion of the raindrop’s energy is intercepted by the plants and dissipates most of its kinetic energy before reaching the soil surface or the surface of flowing water. It is probable that this diversion of rainfall energy may be responsible in a very large measure for the remarkable reduction of erosion by close-growing vegetation (Cook, 1936).

Until the late 1960’s, management of permanent pasture had been based on the use of fire. One of the reasons was to control diseases in domestic animals. This practice favored soil erosion and disappearance of some grass species. Fortunately, such practice has been banned. Runoff and sedimentation cause damages downslope and on the higher elevations as well. Downstream sediments commonly clog drainage channels and cause 12 increased flooding of crop and urban dwellings. Relatively inexpensive techniques, such as contour cultivation and strip cropping are used to reduce and control soil erosion. Sod crops or pasture are established on steep slopes and trees are planted in areas already seriously eroded. The small farmers and the government cannot afford to build long-lasting terraces or construct the waterways that are required to control erosion for an extended period.

Another important process of soil formation on cultivated and pasture land is desilication. Desilication refers to processes that remove silica from the soil. High temperatures and extreme leaching favor rapid desilication and accumulation of iron (ferritization) immobilized in ferric oxide forms under oxidizing conditions (Buol, 1980). The ultimate product of desilication is laterite, a hard sheet of iron - and/or aluminum - rich duricrust. Iron- or aluminum-rich layers are the result of natural soil-evolutionary processes that remove silica and accumulates sesquioxides. While within the soil body and protected by vegetation cover, such layers are often soft. When exposed and desiccated by deforestation and erosion, they harden into rock-like extensive sheets unfavorable for crop production. Laterized horizons are hard, compact and cannot be cultivated. Once the laterized horizon is exposed to the surface, the soil is irretrievably lost for the purpose of crop production (Lai,

1987). Like the problem of soil erosion, laterization constitutes an important threat to tropical soils. The extent and regional distribution of soils already 13 laterized or those that are in imminent danger of being laterized is not known.

Some have warned that most tropical areas, when cleared of vegetation, will

become worthless brick pavement in a few years (McNeil, 1964). Others

have reiterated that laterites have a limited aerial extent in the tropics (Hardy,

1933). In support of the latter argument, Sanchez and Buol (1975) reported

that there may be only 21 million ha (6% of the total tropical land area) that

may become laterized if the subsoil is exposed.

The climatic factors responsible for laterization are those that cause 1)

intense weathering that leads to removal of silica and accumulation of

sesquioxides, 2) wet/dry climate characterized by an intense rainy season

followed by a prolonged dry season and 3) accelerated soil erosion. The

tropical rainforest is supposedly the original vegetation of the plinthite soil (Lai

and Cummings, 1979).

A typical soil profile under permanent pasture exhibits a surficial

horizon about 25 cm thick. Its bulk density is low. Structural peds are

covered by sand grains free of finer material. The organic carbon content is

still more than 3%. The C/N is between 11 and 15. Base saturation is <10%

and the CEC is relatively high (15 to 35 cmol(+)kg"1). These soils exhibit good physical properties and respond well to fertilizers. They are classified as allic Humic Hapludox or allic Humic Sombriudox, depending on the prominence of the sombric horizon (Mathieu, 1985). 14 1.3.3. Evolution of Soils under Cultivation

Most studies on the effects of cultivation on soil properties have been accomplished by comparison of a virgin soil with its counterpart under cultivation. The influence of soil management practices on the long-term maintenance of fertility and tilth has been studied at many locations in the

United States and elsewhere; Laws and Evans (1949), Bouma and Hole

(1971), Coate and Ramsey (1983), Blank and Fosberg (1989), and many others. From the results of these studies, a large body of information has been accumulated concerning the benefits and detriments of numerous management practices. General conclusions can be drawn, but the effects of a given practice may not be identical on all soil types nor under all climatic conditions.

Several marked changes occur when virgin soil is cultivated. Organic matter content rapidly decreases (Newton et al., 1945; Campbell and Souster,

1982) and eventually reaches an equilibrium depending on cropping system, tillage, and climatic regime (Coate and Ramsey, 1983). Degradation of deforested soils’ properties results primarily from reduction of organic matter content that decreases with cultivation. In effect, organic matter influences physical, chemical and biological properties of soils disproportionately to the small quantities present. 15 1.3.3.1. Effects of Cultivation on Selected Soil Properties

It has been generally believed that additions of organic matter greatly

improve the physical conditions of the soil. Baver (1937), Bradfield (1937), and Russell (1952) have pointed out that the physical properties of soil may be affected differently by additions of manure and plant residues as compared to additions of organic matter by plant roots growing in the soil. Browning

(1937) found that additions of organic matter may or may not affect the physical properties of the soil. He found that the dispersion ratio was decreased, the large-sized aggregates were increased, and that the infiltration and percolation rates were increased by the addition of organic matter to soils having relatively poor structure and containing relatively small amounts of inorganic and organic colloidal material. Conversely, these same physical properties were not greatly affected when the organic matter was added to soils containing relatively large quantities of inorganic and organic colloidal material.

Research conducted on Fargo clay (a Humic clay) at the North Dakota

Agricultural Experimental Station at Fargo from 1913 to 1956 concluded that both total N and organic matter content declined on the soil under cultivation.

The rate of decline was more rapid in check plots than in those that received manure or residues. It appeared that residues and manure were equally effective in the long-term maintenance of these constituents. Plots that received P in addition to manure or residues lost slightly more total N and 16 organic matter than those that received only the manure or residues. It is possible that P stimulated microbiological activity thus resulting in faster decomposition of the original organic compounds (Young, et al., 1960).

During the 40 years cropping period, CEC of the soil declined slightly. Neither manure nor residues exerted significant effect on maintenance of this characteristic. More exchangeable K was found in soils from manure and residue plots than in those from check plots. Liming increased exchangeable

Ca and total exchangeable base content. The content of the individual base did not change appreciably with 40 years of cropping except in soil from plots that received lime. In these, Ca increased. Percent base saturation averaged slightly higher in 1953 samples than in the original samples (1913). Soil pH of non-limed plots has increased 0.3 pH units while those of limed plots increased about 0.9 pH units.

Experiments conducted at the Ohio Agricultural Experimental Station to evaluate the effects of 40 years cropping or physical properties and nitrogen content of Nappanee silty clay loam disclosed that cultivation resulted in a loss of pore space and a corresponding increase in weight per unit volume of the soil (Page and Williard, 1946). Under the original forest with which the soils were covered and in the first years after clearing, the soils were well-aggregated and structured. Cultivation and cropping tremendously decrease the quantity of root and organic matter returned to the soil. This results in a serious decrease in soil organic matter and a corresponding loss 17 of soil structure. At the same time, the larger openings left by decay of tree roots are filled with fine soil washed-in from above. Increasing heaviness of the soil and increasing difficulties of drainage follows, resulting in increased soil erosion (Page and Williard, 1946).

Edwards and Lofty (1975) determined that intensive cultivation suppressed invertebrates, especially earthworm (Lumbricus Sp.) populations.

Cultivation causes deleterious changes in soil physical properties. The pedality and wet-aggregate stability of surface horizons decrease, thereby increasing the risk of soil erosion (Laws and Evans, 1949; Greenland et al.,

1962; Bouma and Hole, 1971). Bulk density of surface horizons increases

(Bauer and Black, 1981; Coate and Ramsey, 1983) and large pores, created by soil aggregation, are destroyed (Laws and Evans, 1949; Bouma and Hole,

1971; Bauer and Black, 1981).

Many changes in the soil system as a consequence of cultivation are subtle and variable among soil types. Although crop-fallow management systems allow significantly greater water movement through the soil profile compared with native soil, information is lacking on whether significant differences in leaching occur between a continuously cropped soil and native sod soil. Bouma et al. (1974), however, suggested that greater leaching occurred through a continuously cropped soil in Wisconsin. Martel and Paul

(1974) found a similar distribution of C and N among organic fractions in cultivated and virgin Black and Brown Chernozems, but, in similar soils, 18 Dormaarc (1978) reported up to a 38% decrease in the humic acid to fulvic acid ratio as a direct result of cultivation. Although the effects of a given practice may not be identical on all soil types nor under all climatic conditions, general conclusions can be drawn concerning the effects of cultivation on some important soil properties.

1.3.3.2. Effect of Cultivation on Organic Matter Content and Bulk Density

Organic matter is thought to be a principal binding agent in models of aggregate formation (Emerson, 1959 and Harris et al., 1965) and its loss would contribute to aggregate destabilization and an increase in bulk density.

Loss of organic binding agents and compaction caused by implement traffic destroys most pores larger than 30 urn (Laws and Evans, 1949). Three mechanisms can explain significantly higher bulk density in the Bt horizons of cultivated pedons. First, heavy implement traffic may cause soil compaction at depth (Sohne, 1958). Moreover, the Bt horizons of cultivated pedons contain noticeably fewer roots in ped interiors than Bt horizons of virgin pedons. Decreased root volume, which both create and stabilize channel voids would lead to increased bulk density. Also, there is a general decrease infaunal populations with cultivation (Edwards and Lofty, 1975), which would be expected to reduce void-space volume and thus increase bulk density.

Hesse (1984) studying the effects of organic matter on soil physical properties reported that the overall effect of organic matter is to improve soil physical properties, most of which are related to soil productivity. Organic 19 manures can counteract the deleterious effect upon bulk density that may be

caused by the continued use of mineral fertilizers. Organic matter can also

increase the proportions ofwater-stable aggregate and increase water-holding

capacity. Generally, organic materials with high C/N ratios like straw and

husks have more effect upon soil physical properties than decomposed or

semidecomposed materials such as compost. Thus in situ decomposition has

been shown to give better results than predecomposition (Hesse, 1984).

Debates on the importance of soil organic matter to crop productivity,

particularly in the tropics, have been going on for some time (Howard, 1972;

Russel, 1973; and Sanchez, 1976). The arguments for organic matter conservation have stemmed partly from a generally held view that tropical soils have less organic matter and thus lower fertility than temperate soils.

Systematic comparison of random samples of 61 profiles from the tropics and

45 profiles from the temperate region has not shown this premise to be valid

(Sanchez et al., 1982). There was no significant difference in the concentration of C or the C/N ratio between the two groups at any depth to

100 cm.

Organic matter is not essential to plant growth if the nutrient elements and water can be provided in the right amount at the right time (Russel 1973).

However, in the field this is a condition not always easily met. In traditional systems of agriculture, the extra inputs of the nutrient elements and water can be minimal. In these situations, soil organic matter can be important. 20

Farmers in traditional systems of agriculture know that long-term productivity of the soil may decline as cropping systems become more intensive unless efforts are made to maintain soil fertility. Organic manures have been the major input used to maintain soil fertility partly because they are readily available in many traditional systems of agriculture. Also, organic manures have the potential to correct several soil problems, at the same time, they may supply organic matter, create favorable air and water regimes around the plant roots and act as carriers of some micronutrient (Russel, 1973).

In most soils the continued use of mineral N fertilizers tends to lower pH values and organic manures have a buffering effect. Aluminum-induced acidity, if present, can be partially remedied by organic matter complexing with the Al. In very acid soils such as acid-sulfate soils, the rate of organic matter decomposition is not sufficient to effectively change the pH values by this process (Hesse, 1984). Another positive effect of organic materials added to soil is an increase in cation exchange capacity; an increase in total C and an increase in potentially available N can be expected. Organic manure can increase the availability of P and it can be a good source of micronutrient

(Russell, 1973; Hesse, 1984). Saline soils inhibit certain phases of organic matter decomposition. Although the precise mechanisms are yet to be satisfactorily explained. Salts depress the solubility of organic matter and it is the soluble fraction that is most readily decomposed (Hesse, 1984). 21

1.3.3.3. Effect of Cultivation on Soil Morphology and Hydraulic Conductivity

Two pedons of the same kind of soil may have contrasting soil moisture regimes as a result of differences in field management or of cultivation of one site and preservation of the soil in the virgin state on another (Bouma, 1969).

Bouma and Hole (1971) in their study on soil morphology and hydraulic conductivity of principal soil horizons of paired virgin and cultivated soil pedons in Wisconsin noted that reductions in hydraulic conductivity were paralleled by increases in bulk density and decreases in porosity and organic matter content in the soil horizons. They concluded that agricultural management had led to changes in soil structure, primarily through compaction of the solum, but also as a result of extraction of soil moisture from subsoil by crops. The changes in morphology were reflected in hydraulic conductivity values.

1.3.3.4. Effects of Cultivation on Both Surface and Subsurface Horizons

Much of the research dealing with the effect of cultivation on the soil system has concentrated on surface soil horizons (Laws and Evans, 1949;

Bouma and Hole, 1971; Low and Stuart, 1974 and Blank and Fosberg, 1989).

Inherent soil spatial variability has made it difficult to provide statistical validity to potential subtle changes in the soil caused by cultivation. Blank and

Fosberg (1989) studied six paired virgin and cultivated pedons of Williams soil

(fine-loamy, mixed Typic Agriboroll) or variants of Williams from north central

South Dakota to evaluate the effects of cultivation based on statistical 22 analysis. The statistical analysis were performed by comparisons of chemical,

physical, mineralogical and micromorphological properties of a virgin soil and

its counterpart under cultivation. When similar horizons where compared,

chemical differences between cultivated and virgin pedons averaged as

follows: 1) organic content was 26% less in Ap horizons, 2) water-soluble Si

was 49, 46 and 21% greater in A, Bt and Bk horizons of cultivated pedons,

respectively; 3) water-soluble Mg, Na and K were 42, 32 and 18% lower

respectively in C horizons of cultivated pedons; 4) oxalate - extractable Fe

was 28 and 56% higher in Ap and Bt horizons of cultivated pedons,

respectively. Physical differences between cultivated and virgin pedons

averaged as follows: 1) bulk density was 18% greater in Ap horizons; 2) Ap

horizons contained 38% more very fine sand and 10% less silt; 3) A and Btk

horizons of virgin pedons possessed greater wet aggregate stability and 4) A

horizons of virgin pedons average 30% more water retained at zero MPa

tension.

1.3.4. Effects of Cultivation on Soils at High Altitude in Burundi

Most of the soils of concern were covered by tropical natural forest until

about 1950. The increasing population has exerted such a tremendous

pressure on the forest that, presently, only some relics of the original forest cover the hilltops. Most of the slopes are very steep (25 to 85%). Rainfall erosion constitutes the most rapid process of soil deterioration after deforestation. These soils are traditionally cultivated. That is, the main tool 23 used is a manual hoe that turns the soil to a depth of about 30 cm. The only inputs are stable manure, compost and organic residue from harvesting.

Mineral fertilizers are almost exclusively used on cash crops such as coffee and tea and at agriculture research stations. The optimum quantity of manure recommended by the Institut des Sciences Agronomiques du Burundi (1975) is 30 tons of manure/ha/year. Whereas deterioration of cultivated soil without use of either manure or organic residue from harvesting occurs within a few years, cultivation supported by adequate amount of organic matter have shown sustainable agriculture and increased productivity. In effect, improved morphological, physical and chemical properties brought about by cultivation have been observed by comparison of a cultivated soil with its counterpart under forest or permanent pasture (Frankart, 1972). Because of the relatively short period of time under which the soils can be compared (5 to 30 years), the properties taken into consideration are those that change rapidly such as soil structure, base saturation, acidity, organic matter content, CEC, translocation and aluminum toxicity. The above properties improve with continued additions of stable manure and/or other organic residues. In most cases, these improvements are noticeable within two years, especially in superficial layers (30 cm). Changes in the horizon immediately below the plow layer become apparent after about five years. Ped surfaces become coated with a dark-colored mixture of organic matter and clay. These accumulate on the side of wormholes and become thick and usually tend to fill them. The organic matter content is about 8%, the C/N ratio is about 10. Most of these horizons are classified as agric horizon, the plow layer being umbric or anthropic. The soils evolve from Haplorthods to Hapludoxs within twenty years. CHAPTER 2

EFFECTS OF LONG-TERM CULTIVATION AND PASTURE FOLLOWING DEFORESTATION

2.1. Introduction

The objective of the study was to evaluate the effects of long-term cultivation and pasture on deforested soils at high altitude in Burundi.

Comparisons of physical, chemical and morphological properties of soils under natural forest and their counterparts under cultivation and pasture were performed. The physical, chemical, mineralogical and morphological changes that a soil undergoes from a virgin to cultivated state is of interest to agricultural workers in understanding soil management problems. Changes that occur during a long period of cultivation may be a guide for soil scientists to understand and predict the type of problems that they will face in the future in keeping the soils productive.

The study area is located in Burundi (Fig. 2.1) in the uplifted zone which lies along the Western Rift of the East African Rift Valleys, within a region comprised between 1,500 and 2,600 meters above sea level. The soils develop in residuum from metamorphic rocks (gneiss, schist, quartzite and their intermediate) and from intrusive igneous rocks (essentially granite) under a humid , warm tropical climate. The soil moisture regime is udic , and the temperature regime is isothermic or isomesic. The main objectives of this chapter are to ascertain the effects of cultivation on surface soil horizon, and,

25 26

Fig. 2.1. Location map of Burundi and Rwanda.

Source: Actes du Dixieme Forum International sur la Soil Taxonomy et les Transferts d’Agrotechnologie, A.G.C.D. Publications Agricoles No. 10, Bruxelles, Belgium, 1987. 27

if any, on lower horizons and to provide statistical validity to selected effects

of cultivation. The effects of cultivation on the soil system were accomplished

by comparison of a soil under long-term cultivation sustained by use of organic manure and organic residues from harvesting with its counterparts under forest and under pasture.

2.2. The Study Area

2.2.1. Background Information on Burundi

Burundi is a central African country bordered by Rwanda on the North, Zaire on the West and Tanzania on the South and East. Burundi is a highland country covering an area of about 27,834 km2. According to the 1979 census, the total population was approximately five million, with a density of about 154 inhabitants per km2, making it the second most densely populated country in Africa behind Rwanda. Ninety-three percent of the population is rural with most people living in the highlands 1500 meters above sea level.

The country is subdivided administratively into 15 provinces. Each province is subdivided into "communes", the U.S. equivalent of counties. Each commune is in turn subdivided in hillsides, which are geographical as well as administrative subdivisions. Farm families live dispersed over the "hillsides" where each farmer strives for his self-sufficiency. That is, the smaller farmer manages multiple enterprises: 1) animal production system for food (meat, milk and eggs), manure and cash; 2) crop production systems for food, cash, compost, and animal feed; 3) silvicultural production system for fire wood, 28 building materials, cash, and maintenance or increase of soil fertility by decreasing soil erosion. The small farmer uses few inputs (manure, improved seeds, insecticides and fungicides) and sometimes none at all. He is still, in most cases, using traditional practices (land clearing and preparation, planting, weeding, harvesting) with traditional tools, (hoe, machete,...). Mixed cropping systems are very common in farmers fields. For the special case of associated growth of leguminous and non-leguminous plants, at least three alternative benefits are expected. 1) efficient agronomic management. This is achieved by seeding mixtures of legume and non-legume in field or pasture to produce a more balanced ration and maximize utilization of the land (i.e. the custom of mixing beans with corn is intended to produce as much as possible on a given land). 2) continuous mixed cropping. This sustains steady production without noticeable detrimental effects by exploitation of the soil volume because leguminous plants concentrate their root system in the superficial layers whereas the non-leguminous plants such as corn utilize a thicker soil layer. 3) Excretion of nitrogen by leguminous plants. In effect, various observers have speculated on the possibility that the advantage of mixed cropping with legumes may be related to the ability of these plants to fix atmospheric nitrogen, part of which is passed on to the non-leguminous companion. Lyon and Bizzel (1911) and especially Lipman (1912) provided evidences which suggested that the non-legume was aided because of excretion of nitrogen by the legume (Wilson, 1937). Production of this traditional agriculture is sustained by use of O.M., which in most cases constitutes the only input known by farmers. Different kinds of organic materials are available. The organic materials most commonly used to improve soil conditions and fertility include stable manure, animal wastes, crop residues (either as such or composted), and, rarely, green manures. The composition of the materials can be extremely variable.

The value and the composition of a compost depend as much upon the composting process as upon the original material (Lai, 1984). The same quantities of the same vegetative or animal matter, for example, will give composts of different value according to the location of decomposition (in a pit or in a heap), the amount of moisture maintained, the degree of aeration and the degree of protection from climate (Hesse, 1984). The excretion of dung and urine from domestic animals such as cattle, sheep, goats, pigs and poultry provides the potential for enormous quantities of plant nutrients. For example, Gaur (1984) reports that 15 t N and 4 t each of P and K are potentially available from 1,000 t of fresh cattle dung. Crop residues are recycled in three ways: 1) after composting, 2) by direct incorporation into the soil, or 3) by mulching. On a very limited scale, legume plants are grown to fix atmospheric N before being incorporated into the soil (this practice is still limited to research centers). It is believed that the latter practice exerts its beneficial effect on the following crop only and that there is little residual or cumulative benefit (Hesse, 1984). 30 2.2.2. Location of the Study Area

The study area covers most of the uplifted zone in Burundi that lies

along the Western Rift of the East. African Rift Valleys, at the altitude ranging

from 1,500 meters to 2,600 meters. Similar agroecological zone extends to

Western Rwanda, Eastern Zaire and Western Uganda.

2.2.3. Land Use

Until 1950, most of the area was covered by natural tropical forest.

The recent increasing population (the population has doubled during the last three decades) has exerted such a tremendous pressure on the forest that

most of the area has been deforested and cultivated. In effect, the forest has

been used either for different kinds of building and/or for domestic energy

(cooking, heating,...). As a result, only the hilltops in most cases, are still covered by natural forest whereas the sides slopes have been deforested and kept either under pasture or cultivation. The pastures are usually overgrazed due to the high density of live stock (cattle, sheep and goat).

Burundi has three agricultural seasons: the first and the second are wet and the third is dry. In the first season (September - January), the principal crops in the study area are corn, beans, potatoes and sweet potatoes.

Cassava and banana are limited to lower altitude because of the low temperatures at high altitude. These crops, however, play an important role in food market because they are available at lower altitude at all seasons. Tea is the main cash crop encouraged in this agropedological zone, due to its 31

tolerance to aluminum. In the second season (January - July), the main crops

are peas, wheat, beans, sorghum and sweet potatoes. Cropping during the

third season (July - September) is limited to wet lowlands where corn, beans

and potatoes are the major crops.

In Burundi, intensive cropping systems have become necessary

because of the limitation of available land. Multiple cropping is widespread throughout the country. The two major patterns of multiple cropping are

sequential cropping and intercropping; each is a form of intensification.

Sequential cropping is intensification in time, cropping intensity increases with the number of crops grown in succession in a field in a year. Intercropping

is intensification in both time and space (two or more crops are grown simultaneously in the same field). In many systems of traditional agriculture,

highly complex forms of mixed cropping involving several crops are often practiced infields of shifting cultivation (Greenland, 1962; Okigho, 1978). In these situations the maintenance of soil fertility is achieved through long periods of fallow. Although intensity may be great at a particular point in time, over the long-run these cropping systems are less intensive than a monoculture. Because Burundian farmers cannot afford long periods of fallow, organic manures have played a most prominent role in maintaining soil fertility in cropping systems. 32

2.2.4. Geological Setting

The study area is located within the system of the rift valleys formed during the Cenozoic Era. This system exhibits three-armed rift along the north east margin of Africa. Two of the arms represent the Red Sea and the

Gulf of Aden (Stanley, 1986).

During the late Cenozoic, much of eastern Africa was arched upward nearly 3000 meters above sea level. Fracturing and faulting occurred across the crest of the arch. The famous rift valleys formed as narrow slivers of crust that slipped downward between great fault blocks. As the faulting continued, volcanos developed along the fault trends. Mount Kenya and Mount

Kilimanjaro are two of the most notable of these volcanoes, but there are many others as well. In addition to the volcanoes, a series of elongate lakes formed within the downfaulted blocks. Today, Lake Nyasa (Lake Malawi) and

Lake Tanganyika are splendid examples of these fault-controlled inland water bodies. This region of lakes, volcanoes, and fault-controlled ranges is well known for its exceptional scenic grandeur (Levin, 1983).

The study area is located on the eastern-horst parallel to Lake

Tanganyika, the second deepest lake in the world. The soils developed in residuum from metamorphic rocks (gneiss, schist, quartzite, and their intermediate), and from intrusive igneous rock (essentially granite). 33 2.2.5. Climate

The study area has a humid, tropical climate. The rainfall is relatively high (Mean Annual Rainfall: 1500 to 1800 mm). The humidity is high (the area is influenced by the moisture from winds moving inland from the Indian

Ocean). The temperatures are moderated by the high altitude: the Mean

Annual Temperature varies from 15 to 20° C. The dry period lasts three to four months (from June to August - September) with a decrease of rainfall in

January. According to Koppen classification, it is classified as CW2-3 (two to three months with Means Monthly Rainfall < 60 mm). The hilltops of this area constitute a divide between two important drainage basins: the Nile basin and the Zaire basin (former Congo basin). According to Soil Taxonomy, the soil moisture regime is udic, and the soil temperature regime is isothermic or isomesic.

2.3. Materials and Methods

2.3.1. Experimental Design and Field Methods

Landforms were examined by field observations supported by topographic maps and aerial photographs. Geologic maps of the study area served as the geological data base. Evolution of land cover was accomplished by comparing available aerial photographs (1959, 1972 and 1984). The present land cover was observed in the field. Landowners were interviewed to check these comparisons. Pedons were located under permanent pasture immediately adjacent to pedons cultivated for at least 25 years. The effects 34 of cultivation on soil properties were contrasted by comparison of a soil under cultivation with its counterpart under natural forest and pasture. An important portion of deforested soils is used as pasture. The effects of ranching on soil properties were contrasted by comparison of a soil under pasture with its counterparts under forest and under cultivation. Three sets of sites making a total of nine soil profiles were used for the study. Each set of soils is composed of a soil under forest, a soil under pasture, and a soil under continuous long-term cultivation sustained by use of organic manure and organic residue from harvesting. To minimize the effects of soil spatial variability, sites within a set were located on soils that were similar prior to deforestation. This statement holds under the assumption that prior to deforestation a spatially limited area under the same climatic environment exhibiting the same land form on a same parent material should form the same type of soil. Thus, the three sites composing each set of soils constitute an experiment on a soil under three different treatments. The nine soil profiles constitute a randomized complete block with three treatments experiment design. Special care was taken to locate all sites on similar positions on the landscape. The distance between two sites within a set was as close as possible. Soil profiles were described and sampled according to

Soil Survey Investigations Reports,No. 1 (USDA-SCS, 1984) to a depth of about two meters. Comparisons of selected soil properties were performed on comparable horizons within the sets of soils. A nested-factorial design was used to achieve a thorough evaluation of different treatment effects on both

surface and subsurface horizons. Though the ecological map showed that all

the soils profiles were located within the same map unit, a detailed

observation of both rainfall and temperature distribution revealed the presence

of slight differences among the three sets of sites. To control systematically

possible variability rising from these observed differences, the three sets of

sites were considered different and they were blocked based on latitude. The

three blocks or locations were designated Bugarama, Teza and Rwegura

locations. Each of the nine soil profiles was divided in six horizons. Horizons

occurring approximately at a same depth and sharing a common designation

in most cases were compared. Because these horizons were assumed

comparable, not necessary identical, they were considered nested within the

soil profiles whereas treatments (forest, culture and pasture) and locations

were crossed. Thus, this design had both nested and crossed factors.

Comparisons between different horizons were performed using Bonferroni

pairwise comparisons with a level of significance p=0.05. On four soil profiles

(Muramvya culture, Muramvya pasture, Bukeye culture and Bukeye pasture) a few horizons were divided in two for analytical purposes. They were

regrouped for comparisons because analytical results did not point out any appreciable differences. During reordering, the regrouped horizons’ parameter values were weighed with respect to soil layer thicknesses. For the nine soil profiles, the horizon layouts are summarized in Table 2.1. Table 2.1. Horizon Layouts.

LOCATION I LOCATION II LOCATION III Depth Horizon Depth Horizon Depth (CM) (CM) Horizon (CM) Bugarama Teza Rwegura Forest Forest Forest 0-20 A 0-10 A 0-21 A 20-44 Bhs1 10-36 AB 21-51 AB 44-88 Bhs2 36-74 Bhs1 51-81 Bhs1 88-121 Bhs3 74-125 Bhs2 81-112 Bhs2 121-152 Bhs4 125-168 Bhs3 112-151 Bhs3 152-196 Bt 168-200 Bt 151-200 Bt Muramvya Bukeye Rwegura Culture Culture Culture 0-30 Ap1 0-21 Ap1 0-14 Ap1 30-50 Ap2 21-46 Ap2 14-33 Ap2 50-76 AB 46-71 Bt1 33-64 Bt1 76-110 Bt1 71-110 Bt2 & 64-96 Bth 110-141 Bt2 & Bt3 96-128 B’t1 Bt/R 110-159 Bt4 128-170 B’t2 141-210 B’t1 & 159-200 Bt5 B’t2 Muramvya Bukeye Rwegura Pasture Pasture Pasture 0-19 A 0-16 A 0-30 A 19-35 AB 16-40 AB 30-46 BA 35-94 Bs & 40-96 Bs1 & 46-80 Bs1 Bhs1 Bs2 80-130 Bs2 94-129 Bhs2 96-138 Bhs 130-180 Bhs1 129-179 B’s 138-171 B’s1 180-210 Bhs2 179-210 Bt 171-200 B’s2 . 37 2.3.2. Laboratory Analysis

The soil samples were air-dried, passed through a 2 mm sieve and stored in air-tight containers for physical, chemical, and mineralogical analysis.

Moisture percentages were determined by oven-drying a triplicate of five grams of the air-dried soil samples at 90° C until constant weight. Water retentions were determined using pressure plate extractions (Richards, 1954).

Aggregates stabilities were measured using vacuum wetting (Kemper, 1965).

Soil reactions were determined using a glass electrode-calomel electrode

(McLean, 1982). Organic carbon was determined by acid dichromate digestion and titration with ferrous sulfate ( Walkley, 1935 and Peech et al., 1947).

Nitrogen was determined by Kjeldahl digestion method (Walkley, 1935 and

Peech etal., 1947). Phosphorus extractions were performed using both weak and strong acids, Bray I and II (Bray and Kurtz, 1945). Dithionite - citrate - bicarbonate (DCB) extractions were performed using Mehra and Jackson method (1960). Ammonium acid oxalate extractions were performed using

McKeague and Day method (1965) and sodium pyrophosphate extractions were performed using the Bascomb method (1968). For each of these extractions, superfloe N-10Q (American Cyanamid Co., Wayne, N.J.) was added to the soil samples prior to centrifugation. High-speed centrifugation was used to clear the extracts from particular matter that otherwise stayed in suspension when low-speed centrifugation was used. Iron and Al extracts were analyzed by ICP. Particle size distribution (PSD) and water dispersible clays were determined by the pipette method (Day, 1965), but due to the poor

clay dispersion, the clay fraction was estimated by 2.5 times the 15 bars

pressure plate water content values (Soil Survey Staff, 1992). Bulk densities

were determined by the clod method (Blake, 1965; Soil Survey Staff, 1984)

and the ring method (Uhland, 1949, 1950) for samples with dry bulk density

value less than one. The exchangeable Al and H was determined by the

Yuan method (1959). The extractable cations and the cation exchange

capacities were determined by Peech et al. method (1947). Extractable

acidity was determined by extraction with BaCI2 triethanolamine pH 8.2

Mehlich method (1938). These methods are from SSI, Report No.1 (1984).

2.4. Results and Discussions

2.4.1. Morphological Comparisons

All the soils studied were described and sampled according to Soil

Survey Manual (Soil Survey Staff, 1984). Under natural forest, a typical soil profile exhibited a litter layer consisting of superposed thin layers of undecomposed, partially decomposed and completely decomposed leaves that had been successively deposited on the soil surface. The whole O.M. layer was about 10 cm thick. A thick, dark umbric epipedon occurred consistently throughout the study area beneath the litter layer. This dark color with low chroma was consistent throughout the major part of the umbric epipedon’s matrix. The umbric epipedon presented following characteristics: both broken and crushed samples had Munsell color value of 3 or darker 39 when moist (the soil profile was moist throughout the year). The thickness ranged from 40 to 80 cm or more. It had massive to weak medium subangular blocky structure; the soil was friable or very friable; common fine and medium roots; common vesicular and tubular fine pores; the soil was either nonsticky or slightly sticky, nonplastic, and weakly cemented. Sand grains were free of finer materials.

A spodic horizon usually underlaid the umbric epipedon.

Morphologically, the spodic horizon was similar to the overlying umbric epipedon. Podzolization was recognized by Fe and Al chemistry analysis rather than by the soil profile morphology. In effect, the Fe and Al depth functions of different extracts (DCB, AAO and SPP) were characteristics of podzolization whereas depletion of organic matter, Fe and Al was difficult to detect due to the masking effect of the dark color. Crypto podzolization has been used to designate the hidden podzolization.

Soils under pasture were different from their forested counterparts.

Following deforestation, the litter disappeared due to natural or artificial mineralization, to rainfall erosion or to a combination of the three processes.

Rainfall erosion truncated an important part of the pedon. Some soils are eroded down to the bedrock, hardpan, or less fertile subsoils. A typical soil profile under permanent pasture was described as follows: a dark reddish brown or brown umbric epipedon. Umbric epipedons with intermediate colors were found throughout the studied area. Their thickness had been subdued, 40 compared to that of their counterparts under forest, to about 25 cm. The typical umbric epipedon had massive to weak fine and medium subangular structure; the soil was friable; common fine and medium roots; common tubular pores; the soil was nonsticky, nonplastic, and weakly cemented. Sand grains were free of fine materials.

An oxic horizon usually underlaid the umbric epipedon. The oxic horizon was described as follows: its color was red or dark brown; oxic horizons with intermediate colors occurred within the study area, but for any given soil profile, the oxic horizon’s color was uniform. The structure was weak fine and medium subangular blocky; the soil was friable; common fine roots; common fine and medium tubular pores; the soil was nonsticky, nonplastic, and weakly cemented. Sand grains were free of fine materials.

The oxic horizon thickness ranged from 40 to 80 cm or more.

Cultivation with addition of adequate amounts of organic manure and organic residue from harvesting improved most of the soil properties. These soils exhibited dark reddish brown to very dark grayish brown umbric epipedons. Their umbric epipedons’ thickness ranged from 30 to 50 cm.

Their structure was massive to weak fine granular, the soil was friable; common fine, medium and coarse roots; common medium and coarse tubular pores; the soil was nonsticky, nonplastic, and weakly cemented.

An oxic horizon with color ranging from dark reddish brown to very dark grayish brown underlaid the umbric epipedon. Its thickness ranged from 40 41 to 80 cm or more. These oxic horizons exhibited structures that ranged from moderate medium granular structure to strong coarse subangular blocky structure; the soil was firm. Cultivation with continuous addition of adequate amounts of organic matter improved both the type and the grade of the soil structure. The most obvious improvement consisted of the development of clay films on most ped’s faces, in pores and in root channels. Changes in the horizon immediately below the plow layers become apparent after five years of cultivation. The ped surfaces become coated with a dark-colored mixture of organic matter and clay. Sand grains were coated with the same clay- organic matter mixture. The clay film’s color was one unit chroma lower than the soil matrix. This structure affects positively both soil aeration and rainfall infiltration. In early years of cultivation, clay films appeared only on vertical faces of soil peds. These coatings increased in intensity with time and accumulated on the side of wormholes, become thick and usually tended to fill them. Penetration of clay films increased with the increasing number of years of cultivation provided enough organic matter was continuously added to the soil. After 25 years of cultivation, they were found up to 100 cm deep in the soil profile.

2.4.2. General Chemical and Physical Comparisons

The significant OC decrease in cultivated pedons compared with forested and pastured pedons corroborates what has been extensively reported in the literature (Newton et al.,1945; Doughty et al., 1954). The OC 42 values of forested pedons were greater than their pastured counterparts.

Both forested and pastured OC values were significantly greater than their cultivated counterparts. O.M. is thought to be a principal binding agent in models of aggregate formation (Harris etal.,1965, Emerson,1959) and its loss would contribute to aggregate destabilization and an increase in bulk density.

The soils of concern exhibit unusually high values of aggregate stability

(pastured pedons averaged 93.1%, forested pedons averaged 92.3% and cultivated pedons averaged 89.5%) and cultivation does not destabilize significantly the aggregates, Bonferroni pairwise comparisons, however, showed significant differences between the surface horizons of pastured and forested pedons. These differences were limited to the top 50 cm. Ninety six percent of the total comparisons among forested, pastured and cultivated pedons’ bulk density (Bd) values were significantly different. There was not enough evidence to conclude significant difference on two comparisons (forest

Bd values compared to culture Bd values, and culture Bd values compared to pasture Bd values) on four horizons located at the bottom of the soil profiles. The cultivated pedons Bd values were greater than their pastured counterparts and the pastured pedons Bd values were greater than their forested counterparts. Also, 96% of similar comparisons were significantly different for both water retention difference and water dispersible clay parameters. Forested pedons showed the lowest bulk density values, the lowest water dispersible clay values, and as one should expect, the greatest 43 water retention. Cultivated pedons showed the highest bulk density values, the highest water dispersible clay values, and they logically showed the lowest water retention. The pastured pedons exhibited intermediate values on all the three parameters, namely, bulk density, water dispersible clay and water retention difference. The effect of cultivation on the three parameters were significant on both surface and subsurface horizons. Nitrogen values comparisons showed a marked treatment effect. Significant decrease in N values in cultivated pedons compared to forested pedons resulted from both a reduction of O.M. supply following the disappearance of the forest and a rapid mineralization of O.M. under cultivation. Forested pedons showed significantly greater N values than cultivated pedons throughout the soil profiles. Forested pedons showed significantly greater N values than pastured pedons for the upper 40 cm at the Teza location. They were significantly greater for the upper 120 cm soil layers at the other two locations.

Pastured pedons’ N values were significantly greater than cultivated N values for the top 30 cm at two locations (Bugarama and Rwegura). There was not enough evidence to support differences at the Teza location. The trends were forested N values greater than the pastured N values, the pastured N values being greater than the cultivated N values. The C/N ratios comparisons showed significant differences for the upper 30 cm soil layers when forested pedons were compared to either cultivated or pastured pedons. The cultivated and pastured pedons comparisons did not show enough evidence 44 to support significant difference, except for the upper 30 cm soil layers that

were different when Bugurama forest was compared to Muramvya Culture.

The trends were pastured C/N ratios values greater than cultivated C/N ratio

values, and the cultivated C/N ratio values were greater than forested C/N

ratio values.

Exchangeable bases comparisons showed a marked improvement on

soil cultivated with continuous addition of O.M. In effect, Ca, Mg and K

concentrations of cultivated pedons were significantly greater than both their forested and pastured counterparts throughout the soil profiles. Forested and pastured exchangeable bases comparisons showed significant differences

limited to the upper 30 cm. The trends were cultivated pedons’ bases values greater than forested pedons’ bases values. The forested pedons’ bases values being greater than those of pastured pedons values.

Comparisons between forested and cultivated pedons showed significant difference to at least 60 cm depth. Forested and pastured pedons comparisons showed significant difference to at least 30 cm. Comparisons between pastured and cultivated pedons showed significant difference to 70 cm depth. Forested values were greater than cultivated values. The cultivated values were greater than the pastured pedons.

Bray I P concentrations comparisons showed strong evidence of significant difference among treatments. There was not enough evidence of treatment effects on Bray II P concentrations. Bray I P concentrations in 45 forested pedons were significantly greater than their cultivated counterparts, the cultivated P values were greater than their pastured counterparts.

Comparisons between forested and cultivated pedons P values were significantly different for the upper 60 cm soil layers. The difference was, however, limited to the top 30 cm when the same forested pedons P values were compared to their pastured counterparts. Cultivated pedons P values were greater than their pastured counterparts to a depth of 70 cm.

Concerning soil acidity comparisons, BaCI2 acidity of forested pedons was greater than its pastured counterpart. The pastured BaCI2 acidity values being greater than their cultivated counterpart. All the comparisons were significantly different throughout the soil profiles. Unlike the BaCI2 acidity that was affected at both surface and subsurface horizons by different treatments, the extractable acidity treatment effect was limited to the soil surface layers.

The trends were forested pedons extractable acidity greater than that of pastured pedons. The pastured pedons were greater than that of cultivated pedons. When forested pedons were compared to their cultivated counterparts, the differences were limited to the upper 70 cm at the Teza location, and to the 100 cm depth at both Bugurama and Rwegura locations.

For forested pedons versus pastured pedons, a significant difference was confined to the upper 40 cm. The difference between cultivated and pastured pedons remained within the upper 30 cm soil layers. 46 Extractable aluminum comparisons showed that forested pedons values were greater than the pastured pedon values. The pastured pedons’ values were greater than the cultivated counterparts. Significant differences persisted throughout the soil profiles when forested pedons were compared to cultivated pedons. They remained within the upper 100 cm soil layers for both forested pedons compared to pastured pedons and cultivated pedons compared to pastured pedons.

Extractable H values were greatest on forested pedons, followed by pastured and cultivated pedons, respectively. Significant differences were observed throughout the soil profiles for all the comparisons at two locations

(Bugurama and Teza). They remained significantly different within the upper

100 cm at the Rwegura location.

For both water pH and KCI pH, statistical comparisons failed to detect any significant treatment effect. Although the cultivated pedons showed consistently higher pH values than both pastured and forested pedons’ pH values throughout the soil profiles. The failure resulted from the logarithmic scale that is used to report extractable H activities, making any statistical difference hard to detect. Likewise, pH NaF in cultivated pedons showed a reduction of this parameter’s values, but statistical comparisons were unable to differentiate different treatment effects.

Significant treatment effects were observed on both NH4OAc and sum

CEC values. There was, however, not enough evidence to support any 47 significant treatment effect on ECEC values. For both NH4OAc and sum

CEC, forested pedons showed the highest values, followed by pastured and

cultivated pedon’s values, respectively. Significant difference between

forested and cultivated pedons persisted throughout the soil profiles. The

difference between forested and pastured pedons remained significant for the

upper 70 cm at Teza location on both NH4OAc and sum CEC. Differences

between cultivated and pastured pedons were significant throughout the soil

profiles for sum CEC, the NH4OAc CEC values were significantly different throughout the soil profiles at Bugarama location. They were limited to the

upper 30 cm at the Teza location and to the upper 50 cm for the Rwegura

location.

Significant treatment effects were also observed on both NH4OAc and sum base saturations. Cultivated pedons base saturation values were significantly greater than their forested counterparts. The forested values were greater than their pastured counterparts. Cultivated pedons’ values remained significantly higher than both their forested and pastured counterparts throughout the soil profiles. There was not enough evidence of any difference below 10 cm depth when forested pedons were compared to their pastured counterparts. Cultivation should improve the base saturation status of pastured soil when they are brought to cultivation sustained by continuous additions of organic matter. 48 Significant decreases of Al saturation by cultivation were observed provided continuous addition of sufficient O.M., especially stable manure.

Significant reductions were observed throughout the soil profiles at Teza location. They were, however, limited to the upper 140 cm soil layers at the

Bugarama location and the upper 60 cm at the Rwegura location. Pastured pedons exhibited the highest Al saturation values, followed by forested and cultivated pedons’ values, respectively.

To appreciate and comprehend the overall comparisons of the soil characteristics among forested pedons and their counterparts under cultivation and pasture, one should bear in mind that the natural forest is a complex ecosystem that concentrates plant and microbial nutrients in the surficial layers. The luxurious vegetation is sustained by an efficient cycling of organic matter in the soils that results in the overall improvement in soil conditions.

The disappearance of the forest litter along with the reduction of O.M. supply that follow deforestation increases soil erodibility. Sheet erosion and gullies made some of the soils almost useless for cropping. Some soils erode to the bed rock, hardpan, or less fertile subsoils. It seems reasonable to suggest that the current surface horizons of both cultivated and pastured soils were part of the subsoil, probably C horizons, when these soils were under natural forest. The cultivated soils’ upper layer improvements should, therefore, be appreciated based on comparisons between cultivated pedons and their pastured counterparts. From the overall physical and chemical comparisons, 49 we concluded that 18 out of 32 compared parameters, namely: OC, Bd, water

retention difference, water dispersible clay, N, Bray I P, Ca, Mg, K, BaCI2 acidity, extractable acidity, extractable Al, extractable H, NH4OAc CEC, Sum

CEC, NH4OAc BS, Sum BS, and Al saturation showed significant treatment effects. Seven of these parameters were significantly different throughout the soil profiles. In other words, they affected both surface and subsurface horizons. They were OC, Bd, water retention, water dispersible clay, BaCI2 acidity, extractable H and Sum CEC (Table 2.2 to Table 2.7). Ten of them, namely: N, C/N ratios, Bray I P, Ca, Mg, K, extractable acidity, extractable Al,

NH4OAc CEC, and Al saturation exerted a significant treatment effect that was limited to the upper horizons of the soil profiles. Some of these parameters remained significantly different within the upper 30 cm of the soil profiles. Others persisted to the 100 cm depth. On fourteen parameters, namely: Bray II P, ECEC, pH water, pH KCI, pH NaF, C/N ratios, aggregate stability, Fed, Ald, Al0, Fe0, Fep, Alp and AIQ+1/2 FeQ, there was not enough evidence to support the existence of any treatment effect. At the end of different treatment effects comparisons, it was concluded that there should be substantial improvement of pastured soils when they are cultivated with continuous addition of enough amounts of O.M. (30 tons/ha/yr). The most morphological improvement brought about by cultivation is the development of a subangular blocky structure that improves soil aeration and water infiltration. Organic matter is not essential to plant growth if the nutrient Table 2.2. Physico-Chemical Properties at Site I.

Bugarama Forest

Depth Horizon Db OC WDC WRD (CM) (gem-3) (%) (%) (gem-3)

it ★ 0-20 A 0.61* 3.18* 0.0 63.3 *

o o * 20-44 Bhs1 0.62* 2.96* 61.7 *

d o * 44-88 Bhs2 0.56* 3.06* 66.2 * o 88-121 Bhs3 0.62* 2.82* d 41.4*

121-152 Bhs4 0.94* 3.08* o b 52.5* 152-196 Bt 0.91* 2.44* 0.0* 25.6*

Muramvya Culture

Depth Horizon Db OC WDC WRD (CM) (gem-3) (%) (%) (gem-3) * 0-30 Ap1 1.23* 2.55* 5.1 25.0* ★ 30-50 Ap2 1.47* 1.67* 5.2 22.4* 50-76 AB 1.46* 1.22* 15.5* 21.6* 76-110 Bt1 1.50* 0.66* 10.2* 22.3* 110-141 Bt2 & 1.62* 0.45* 13.3* 21.5* Bt/R * 141-210 B’t1 & 1.65* 0.27* 8.8 21.0* B’t2

Muramvya Pasture

Depth Horizon Db OC WDC WRD (CM) (gem-3) (%) (%) (gem-3) 0-19 A 0.97* 2.82* 10.4* 42.0* * 19-35 AB 1.21* 2.17 15.5* 34.9* 35-94 Bs & 0.99* 1.78* 10.4* 35.4* Bhs1

★ * 94-129 Bhs2 0.92* 1.43 5.2 31.4* * * 129-179 B’s 1.38 0.66* 5.1 29.2* it it 179-210 Bt 1.35 0.47* 5.1 26.6*

* Significant at the P = 0.05 level of probability. Table 2.3. Physico-Chemical Properties at Site II.

Teza Forest

Depth Horizon Db OC WDC WRD (CM) (gem-3) (%) (%) (gem-3) *

d it 0-10 A 0.62* 3.72* o 90.3 it 10-36 AB 0.69* 2.17* 0.0* 47.7

* *

36-74 Bhs1 0.90 1.49* o b 40.2* 74-125 Bhs2 1.24* 1.14* 0.0* 39.7* 125-168 Bhs3 1.19* 1.48* 0.0* 42.5*

it * o 168-200 Bt 1.37 0.68* b 36.2*

Bukeye Culture

Depth Horizon Db OC WDC WRD (CM) (gcrrf3) (%) (%) (gem-3) * 0-21 Ap1 0.90* 2.41* 5.1 35.4* * 21-46 Ap2 1.06* 2.06* 5.1 34.1* 46-71 Bt1 1.36* 1.05* 10.3* 32.9* 71-110 Bt2 & Bt3 1.32* 1.26* 10.3* 34.6*

110-159 Bt4 1.40* 0.90* 10.3* 30.3* * 159-200 Bt5 1.41* 0.66* 5.1 28.8*

Bukeye Pasture

Depth Horizon Db (gem" OC WDC WRD (CM) 3) (%) (%) (gem-3) * 0-16 A 1.19* 2.82* 5.2 45.2* 16-40 AB 1.23* 2.17* 10.3* 42.6* * 40-96 Bs1 & 1.27* 1.78* 7.3 36.2* Bs2 96-138 Bhs 1.22* 1.43* 5.2* 33.6* it 138-171 B’s1 1,34* 0.66* 5.1 32.9* * 171-200 B’s2 1.29* 0.47* 5.1 33.8*

* Significant at the P = 0.05 level of probability. Table 2.4. Physico-Chemical Properties at Site III.

Rwegura Forest

Depth Horizon Db (gcrrf OC WDC WRD (CM) 3) ...... (%) (%) (gem-3) * 0-21 A _ 0.62* -3,23* 0.0 75.5* * 21-51 AB 0.68* 2.89* o o 68.9* *

51-81 Bhs1 0.82 2.77* o o 56.6* * b o * 81-112 Bhs2 0.85* 2.73* 47.3 * 12-151 Bhs3 1.26* 1.66* 2.7 36.9* * 151-200 Bt 1.40* 0.87* 1.3 36.4*

Rwegura Culture

Depth Horizon Db (gem- OC WDC WRD (CM) 3) (%) (%) (gem-3)

* 0-14 Ap1 1.13* 1.51* 2.6 39.4* 14-33 Ap2 1.24* 2.06* 10.8* 41.6* 33-64 Bt1 1.42* 0.99* 15.6* 36.5* 64-96 Bth 1.54* 0.98* . 18.2* 35.2* 96-128 B’t1 .1.55* 0.63* 12,8* _33,4* 128-170 B’t2 1.52* 0.48* 15.4* 32.2*

Rwegura Pasture

Depth Horizon Db OC WDC WRD (CM) (gem-3) (%) (%) (gem-3) * -0-30 A 0.85* 2.87* 5.5 56.9* 30-46 BA 1.16* 2.03* 18.8* 54.6* 46-80 Bs1 1.15* 1.35* 23.8* 48.6* 80-130 Bs2 .... 1.17* 1.35* 21.1* 42.4* 130-180 Bhs1 0.98* _ 2.67* 11.1* 49.1* * 180-210 Bhs2 1.20* 1.35* 7.8 40.0*

* Significant at the P = 0.05 level of probability. 53 Table 2.5. Chemical Properties at Site I.

Bugarama Forest

Depth Horizon BaCU Extr. H Sum CEC (CM) cmol(+) kg'1 cmol(+) Kg"1 cmol(+) Kg'1 0-20 A 39.7* 12.4* 39.8* * 20-44 Bhs1 42.2* 3.8 42.3* it 44-88 Bhs2 43.5* 3.9 43.6* * 88-121 Bhs3 40.3* 2..1 40.4* * 121-152 Bhs4 31.4* 1.4 31.5* it 152-196 Bt 21.1* 1.0 21.2*

Muramvya Culture

Depth Horizon BaCI2 Extr. H Sum CEC (CM) cmol(+) Kg'1 cmol(+) Kg'1 cmol(+) Kg"1

it 0-30 Ad1 16.6* 0.7 20.0* it 30-50 Ap2 13.4* 0.8 17.1* it 50-76 AB 10.9* 0.7 13.4* * 76-110 Bt1 10.9* 0.7 12.4*

it it * 110-141 Bt2 & 7.8 1.6 8.5 Bt/R

* it it 141-210 B’t1 & 7.5 2.9 7.9 B’t2

Muramvya Pasture

Depth Horizon BaCU Extr. H Sum CEC (CM) cmol(+) Kg'1 cmol(+) Kg"1 cmol(+) Kg’ 1

it 0-19 A 24.3* 4.5 24.4* * 19-35 AB 22.4* 7.3 22.4* i t it 35-94 Bs & 24.6* 3.5 24.6 Bhs1

it 94-129 Bhs2 21.1* 4.7 21.1* it it 129-179 B’s 12.8 5.3 12.8* it 179-210 Bt 12.2* 4.1 12.2*

+ Significant at the P = 0.05 level of probability. 54 Table 2.6. Chemical Properties at Site II.

Teza Forest

Depth Horizon BaCL Extr. H Sum CEC (CM) cmol(+) Kg"1 cmol(+) Kg"1 cmol(+) Kg"1 * 0-10 A 49.3* 5.3 50.4* * 10-36 AB 24.3* 1.5 24.4* * 36-74 Bhs1 23.7* 3.3 23.7* * 74-125 Bhs2 19.8* 3.1 19.8* * 125-168 Bhs3 24.3* 5.5 24.3* * * 168-200 Bt 14.7 3.6 14.7*

Bukeye Culture

Depth Horizon BaCI2 Extr. H Sum CEC (CM) cmol(+) Kg'1 cmol(+) Kg'1 cmol(+) Kg"1 * 0-21 Ad1 17.3* 2.5 20.3* * 21-46 Ad2 21.8* 4.4 22.4* * T- CO CD * 46-71 Bt1 1.2 18.4* * 71-110 Bt2 & Bt3 25.9* 0.4 26.4*

* 110-159 Bt4 18.6* 0.6 18.9* * 159-200 Bt5 16.0* 0.6 16.2*

Bukeye Pasture

Depth Horizon BaCI2 Extr. H Sum CEC (CM) cmol(+) Kg"1 cmol(+) Kg"1 cmol(+) Kg"1 * 0-16 A 23.0* 3.4 23.2* * 16-40 AB 21.1* 0.6 21.2* * 40-96 Bs1 & 19.5* 1.1 19.5* Bs2

* 96-138 Bhs 19.8* 2.4 19.8* * 138-171 B’s1 14.1* 1.8 14.1* * 171-200 B’s2 13.4* 1.8 13.4*

+ Significant at the P = 0.05 level of probability. Table 2.7. Chemical Properties at Site III.

Rwegura Forest

Depth Horizon BaCI2 Extr. H Sum CEC (CM) cmol(+) Kg’ 1 cmol(+) Kg"1 cmol(+) Kg'1 ★ 0-21 A -40.3* 3,6_ 42.0* *• 21-51 AB 37.8* 4.4 37.9* * CO 51-81 Bhs1 33.3* o 33.4* 81-112 Bhs2 33.9* 1.2* 34.0* 12-151 Bhs3 .27.5* 0.0* 27.6* * 151-200 Bt 18.6* 0.3 18.7*

Rwegura Culture

Depth Horizon BaCI2 Extr. H Sum CEC (CM) cmol(+) Kg"1 cmol(+) Kg'1 cmol(+) Kg"1 *

0-14 Ad1 19.8* o o 22.0* * 14-33 Ad2 26.9* 0.0 27.9* * * 33-64 Bt1 16.6* 0.0 17.3 * 64-96 Bth 18.6* 0.0 19.2* * 96-128 B’t1 13.4* 0.0 13.9* 128-170 B’t2 12.2* 0.0* 12.7*

Rwegura Pasture

Depth Horizon BaCU Extr. H Sum CEC (CM) cmol(+) Kg"1 cmol(+) Kg'1 cmol(+) Kg"1 * 0-30 A 26.9* . 0,5...... 27.0* * 30-46 BA 24.3* 0.7 -24.3* * 46-80 Bs1 20.5* 0.5 _20.5* * 80-130 . . Bs2 21.1* 0.4 21.1* * 130-180 Bhs1 32.6* 0.2 32.6* * 180-210 Bhs2 23.0* 0.1 23.0*

+ Significant at the P = 0.05 level of probability. 56 elements and water can be provided in the right amount at the right time

(Russel, 1973). This condition is not always easy to meet in the field. In

traditional systems of agriculture, the extra inputs of the nutrient elements and

water can be minimal. When these can not be supplied, soil organic matter

can be important. In addition to nutrient supply brought about by organic

mineralization, the addition of organic matter supplies a new stock of micro­

organisms that decompose the accumulated inactive organic matter, which

renovates the soil biological cycle. On these highly acid soils, O.M. has a pH

buffering effect. The aluminum acidity is partially remedied by O.M.

complexing with the Al. Though N and P concentrations remained below that

of forested soils, the difference is due to a literal continuous addition of forest

leaves that keeps these concentrations relatively higher than those under

cultivation. There were enough evidence to conclude that cultivation

supported by addition of O.M. improved physical soil properties, increased the

plant nutrient status of the soil and reduced soil acidity. Even though these

improvements are not by any mean impressive, they constitute both a good start and a requirement to a meaningful use of chemical fertilizers. Evaluation of this conclusion should take into account the socio-economic situation of the farmers and the government, along with the world financial and food supply availability and management. CHAPTER 3

MINERALOGICAL PROPERTIES AND AGGREGATE STABILITY

3.1. Introduction

To evaluate the mineralogical properties, especially the mineral

components of the studied soils, one should consider that the studied area was highly weathered prior to faulting and uplifting and the spodic properties displayed by the soils formed during the cool and humid parts of the

Pleistocene Epoch. The luxurious forest that emerged under high rainfall constituted an ideal ecosystem for podzolization processes. The rift valleys that erupted during the late Cenozoic Era, probably during Pliocene Epoch, fractured Precambrian-aged rocks (Cryptozoic Eon) that had most likely weathered to oxidic materials. Climate is not constant with time. There is much geological and botanical evidence showing that climate change may be important within the history of many of the soils we presently observe on the earth’s surface (Buol et al., 1973). An excellent example is the occurrence of

Oxisols in Queensland, Australia, where the present environment does not favor the development of these soils. Whitehouse (1940) indicated that the

Oxisols there were the products of two humid periods in the Pliocene. Wilke and Schwertmann (1977) postulated that the German occurrence of gibbsite in Dystrochrepts was a relict (probably pre-Pleistocene). It did not form under current pedogenic conditions as indicated by a decrease upward in the soil profile. Spodosols at the study area are relict (Pleistocene). They did not

57 58 form under current pedogenic conditions. The soil minerals were weathered to oxidic materials prior to Spodosols’ development. Chesworth residua model (1973) and haplosoil model (1980) seem to apply at the study area.

Chesworth (1980) suggested that the humid tropics, due to a combination of high temperatures, adequate moisture and long term stability of many land forms, provide the best opportunity for attaining equilibrium and that common mineral assemblages found in the humid tropics are consistent with the haplosoil system. According to his residua model (1973), the four-component system consisting of Si02-Al2 0 3-Fe20 3-H20 constitutes a chemical sink or residua system toward which mineral composition tends during weathering.

3.2. Materials and Methods

Soil samples were rich in iron and aluminum oxides and organic matter.

Iron oxides were removed using DCB (Mehra and Jackson, 1960). Organic matter was removed using NaOCI that had been adjusted to pH 9.5 with HCI just before treatment. The mineralogical composition was determined on both total and fine clays using a computerized diffractometer system coupled with

NEWMOD and generated standard diffractograms. Aggregates stabilities were measured using vacuum wetting (Kemper, 1965).

3.3. Results

Kaolinite was found in both total and fine clays in a remarkably uniform distribution with depth (Fig.3.1; Fig.3.2; Fig.3.3). The amounts of kaolinite were estimated between 75 and 85% of the total minerals. 59

sags K* 5S(

r» 4881 o K* 3011

3888 CO to CM

K* Air i 7

1888

•dry 8.8 T " "T 38 48 [*2 0 ]

>0* C 5888

I 3 K* ! 30" C m C3

3888 to to Glycol to

2888

rdry

Airdry 8.8 38 (*2 91

Fig. 3.1. X-ray diffraction patterns of Rwegura forest Bhs1 and Bhs2 horizons showing kaolinite (d = 0.714 and d = 0.357), gibbsite (d = 0.238) and quartz (d = 0.334). The d spacings are in nanometers. 60

530* C

o 300* C co

2188 K* A i dry

t irdry 8.8 28 38 48 1*201 58

p>

CO CO CO CO CM

3888

j [ K* Ai rdry, ^ Mi

8.8 18 28 38 48 (*201 58

Fig. 3.2. X-ray diffraction patterns of Bukeye culture Bt1 and Bt3 horizons showing kaolinite (d = 0.714 and d = 0.357), gibbsite (d = 0.238) and quartz (d = 0.334). The d spacings are in nanometers. 61

K* 550* C

B888

7888 K* 30Q to

CO Mg** Gycol

3880 K* Aird

2888

1888 Mg** Ai ’dry

28 38

18888• [conat*] ■ K* 5 5 0 *C

9188'

8888 o Iv o d k* aoo*c 7888

8888

5888 L Mg** (ilycol 4888

3888 K* Aii dry 2888

1888 Mg1* Airdry * 8.8 JL T “ i— ---fl— 18 28 38 48 1*201 58

Fig. 3.3. X-ray diffraction patterns of Bukeye pasture Bs2 and Bhs horizons showing kaolinite (d = 0.714 and d = 0.357), gibbsite (d = 0.238) and quartz (d = 0.334). The d spacings are in nanometers. 62

Gibbsite depth functions were inconsistent. Some soil profiles displayed the highest amount of gibbsite in the surface horizons whereas in other soil profiles the maximum amount of gibbsite occurred in the lowest horizons. Depth trends of gibbsite were not found to correlate with the land use. The amounts of gibbsite were estimated between 12 and 18%.

Micas were found in extremely low amounts (traces in most cases) in total clays. Most fine clays were devoid of micas. The general trend showed a consistent decrease of micas with depth. Quartz occurred in both total and fine clays with a uniform distribution throughout the soil profiles.

3.4. Discussions

During soil development, weathering preferentially removes mobile components and concentrates the relatively immobile components.

Qualitative and quantitative determinations of the soil mineral composition disclosed the predominance of kaolinite (about 80% or more) throughout the soil profiles. The remaining clay minerals were identified and estimated as gibbsite > mica > quartz.

Kaolinite is an extremely widespread soil mineral. Its formation is most pronounced where weathering is intense, as in the humid tropics and humid temperate zones. Hence, it is dominant in the clay fraction of most Oxisols and many Ultisols. Kaolinite formation is favored in the humid tropics by alternate wet and dry seasons in higher parts of the landscape where drainage is unimpeded (Tardy et al., 1973). Soils described as "lateritic" are 63 mostly dominated by kaolinite. Deeply weathered rock (saprolite) underlying

Oxisols and Ultisols is commonly dominated by kaolinite. Examples of such occurrences have been reported by Melfi et al. (1983) in Brazil, Cady (1951),

Rebertus and Buol (1985), and Calvent et al. (1980) in North Carolina; and

Gilkes et al. (1973) and Anand et al. (1985) in Australia.

At the studied area, soils under cultivation and pasture were classified as Oxisols, and most of the soils under natural forest were classified as

Spodosols. In both soil orders, the dominant alumino-silicate clays were kaolinite. Kaolinite distribution with depth was remarkably uniform. Unlike most of the Spodosols described elsewhere, wherein kaolinite rarely makes up more than a small fraction of clays (Brown and Jackson, 1958, Ross and

Mortland, 1966; Kodama and Brydon, 1968, Gjems, 1970, Kodama, 1979),

Spodosols at high altitude in Burundi were dominated by kaolinite. The unusual abundance of kaolinite in these Spodosols corroborates the suggested pedogenesis processes that took place at the study area. The rift valleys that erupted during the late Cenozoic Era fractured Precambrian-aged rocks (Cryptozoic Eon) that had weathered to oxidic materials. The luxurious forest that emerged under high rainfall constituted an ideal ecosystem for podzolization processes. Also, Norrish and Pichering (1983) reported kaolinite as a major component of some Australian Spodosols.

Many geological episodes have taken place on the earth’s crust in the past. Some of these episodes may have involved local variations of 64 temperature or pressure or both which affected the nature of chemical

weathering and the genesis of soil minerals. There is much geological and

botanical evidence showing that climate change may be important within the

history of many of the soils we presently observe on the earth’s surface (Buol

et al., 1973). While common geological studies reported kaolinite to have a

surface area of about 15 cm2g-1, kaolinite from oxic horizons may have

surface area three to four times higher. Bigham et al. (1978), however, found

that the surface area developed by iron oxides in Oxisols were lower than

those observed in Ultisols.

The amounts of gibbsite in the studied soils were estimated between

12 and 18 percent for all the soil profiles analyzed. Gibbsite AI(OH)3 is the

most common Al hydroxide mineral in soils. It is associated with the latter

stages of weathering when leaching of silica has progressed to such an extent

that phyllosilicate minerals no longer form. Water soluble Si in the studied

soils were very low. The water soluble Si concentrations ranged from 5 ppm to 26 ppm for all the profiles (Table 3.1, Table 3.2 and Table 3.3).

Gibbsite is very common in highly weathered Oxisols of tropical

regions. It has a low cation exchange capacity (CEC) and contributes to the

low native fertility of most Oxisols. It is found at the weathering interface

between igneous rocks and saprolite where Si is leached and Al precipitates.

Depth trends of gibbsite in Oxisols are inconsistent. Often the analysis of the soil and underlying saprolite has not been sufficiently detailed to draw definite 65 Table 3.1. Water Soluble Silicon at Site I.

Bugarama Forest

Depth Horizon Si (CM) (ppm) 0-20 A 9.3 20-44 Bhs1 10.3 44-88 Bhs2 14.3 88-121 Bhs3 14.5 121-152 Bhs4 17.1 152-196 Bt 12.6

Muramvya Culture

Depth Horizon Si (CM) (ppm) 0-30 Ap1 9.9 30-50 Ap2 12.5 50-76 AB 26.0 76-110 Bt1 23.2 110-141 Bt2 & Bt/R 18.0 141-210 B’t1 & B’t2 16.2

Muramvya Pasture

Depth Horizon Si (CM) (ppm) 0-19 A 8.0 19-35 AB 8.0 35-94 Bs & Bhs1 10.2 94-129 Bhs2 9.7 129-179 B’s 7.4 179-210 Bt 6.4 66 Table 3.2. Water Soluble Silicon at Site II.

Teza Forest

Depth Horizon Si (CM) (ppm) 0-10 A 14.3 10-36 AB 5.6 36-74 Bhs1 6.3 74-125 Bhs2 7.6 125-168 Bhs3 9.5 168-200 Bt 8.3

Bukeye Culture

Depth Horizon Si (CM) (ppm) 0-21 Ap1 7.4 21-46 Ap2 9.1 46-71 Bt1 17.1 71-110 Bt2 & Bt3 15.3 110-159 Bt4 15.6 159-200 Bt5 19.3

Bukeye Pasture

Depth Horizon Si (CM) (ppm) 0-16 A 7.4 16-40 AB 7.6 40-96 Bs1 & Bs2 9.2 96-138 Bhs 10.6 138-171 B’s1 7.5 171-200 B’s2 25.0 67 Table 3.3. Water Soluble Silicon at Site III.

Rwegura Forest

Depth Horizon Si (CM) (ppm) 0-21 A 19.4 21-51 AB 16.3 51-81 Bhs1 16.3 81-112 Bhs2 12.2 112-151 Bhs3 10.1 151-200 Bt 11.4

Rwegura Culture

Depth Horizon Si (CM) (ppm) 0-14 Ap1 11.3 14-33 Ap2 25.8 33-64 Bt1 15.9 64-96 Bth 13.3 96-128 B’t1 8.6 128-170 B’t2 21.0

Rwegura Pasture

Depth Horizon Si (CM) (ppm) 0-30 A 7.7 30-46 BA 7.8 46-80 Bs1 8.0 80-130 Bs2 8.3 130-180 Bhs1 9.7 180-210 Bhs2 7.0 68 conclusion. Eswaran and Wong (1978) found a significant increase in gibbsite content from the pallid zone upward in an Orthox. Conversely, Gilkes et al.,

(1973) reported a general decrease toward the surface in lateritic profiles in western Australia. The usual pathway proposed for gibbsite formation is by the desilication of kaolinite. Micromorphological evidence from numerous studies indicated that gibbsite often forms pseudomorphically after feldspars.

Moreover, gibbsite formation after biotite has been reported (Wilson, 1966;

Sidhu and Gilkes, 1977; Bisdom et al., 1982). Reasearch on Ultisols in the southeastern USA showed gibbsite contents generally decrease upward in the profile (Bryant and Dixon, 1963; Cook,1973; Calvert et al., 1980, Rebertus and Buol, 1985). Resilication to halloysite, subsequently transformed into kaolinite higher in the profile, was proposed by Calvert et al.(1980) to explain the trends observed in the soils they studied on the North Carolina Piedmont.

No distinct trends were evident from the results reported by Karathanasis et al. (1983). During particle size analysis, dispersion of the soils was extremely difficult. The soils are rich in organic matter, gibbsite and iron oxides. As one should expect the soils showed unusually high aggregate stability values

(Table 3.4, Table 3.5 and Table 3.6). All three components are well known for their binding effects. The soils studied exhibiting steep slopes that favor soil rainfall erosion, aggregation mechanism should be investigated.

The mechanism of aggregation in the presence of Al hydroxides has not been clearly presented in the literature, but two types of effects are conceivable 69 Table 3.4. Aggregate Stability at Site I.

Bugarama Forest

Depth Horizon Ag. St. (CM) (%) 0-20 A 93.1 20-44 Bhs1 92.0 44-88 Bhs2 91.1 88-121 Bhs3 91.5 121-152 Bhs4 91.6 152-196 Bt 92.5

Muramvya Culture

Depth Horizon Ag. St. (CM) (%) 0-30 Ap1 94.8 30-50 Ap2 94.2 50-76 AB 95.0 76-110 Bt1 82.7 110-141 Bt2 & Bt/R 91.4 141-210 B’t1 & B’t2 94.3

Muramvya Pasture

Depth Horizon Ag. St. (CM) (%) 0-19 A 96.0 19-35 AB 94.9 35-94 Bs & Bhs1 94.1 94-129 Bhs2 93.3 129-179 B’s 86.6 179-210 Bt 91.4 70 Table 3.5. Aggregate Stability at Site II.

Teza Forest

Depth Horizon Ag. St. (CM) (%) 0-10 A 89.2 10-36 AB 86.0 36-74 Bhs1 95.7 74-125 Bhs2 95.9 125-168 Bhs3 95.1 168-200 Bt 96.7

Bukeye Culture

Depth Horizon Ag. St. (CM) (%) 0-21 Ap1 91.6 21-46 Ap2 93.6 46-71 Bt1 96.3 71-110 Bt2 & Bt3 93.9 110-159 Bt4 93.3 159-200 Bt5 84.6

Bukeye Pasture

Depth Horizon Ag. St. (CM) (%) 0-16 A 94.2 16-40 AB 96.7 40-96 Bs1 & Bs2 94.9 96-138 Bhs 94.1 138-171 B’s1 91.5 171-200 B’s2 92.6 Table 3.6. Aggregate Stability at Site III.

Rwegura Forest

Depth Horizon Ag. St. (CM) (%) 0-21 A 88.9 21-51 AB 89.5 51-81 Bhs1 90.7 81-112 Bhs2 89.7 112-151 Bhs3 92.3 151-200 Bt 91.2

Rwegura Culture

Depth Horizon Ag. St. (CM) (%) 0-14 Ap1 89.6 14-33 Ap2 91.8 33-64 Bt1 88.2 64-96 Bth 85.3 96-128 B’t1 78.0 128-170 B’t2 73.6

Rwegura Pasture

Depth Horizon Ag. St. (CM) (%) 0-30 A 93.4 30-46 BA 94.5 46-80 Bs1 93.6 80-130 Bs2 94.0 130-180 Bhs1 91.1 180-210 Bhs2 89.6 72

from reasoning. Small OH-AI polymers, which can be considered fragments

of solid AI(OH)g, can be held more tightly than ordinary exchangeable cations

in the interlayer space of expansible clay minerals (Barnhisel and Bertsch,

1989). Their presence may lead to a reduction in the swelling and the

expansion of clay particles by bonding adjacent silica sheets together and by

displacing interlayer cations of high hydration power, such as Na+, thereby

promoting aggregation.

Aluminum hydroxides and OH-AI polymers also may react with clay

particles on their external surfaces and cement them together. The positive

charges of A!(OH)3 are distributed at the edge. Because of the spatial

restriction, it is impossible to have all the positive charges of one OH-AI

polymer or AI(OH)3 unit entirely satisfied by one clay particle. Instead, when

one AI(OH)3 unit is attached to one clay particle, it tends to link another.

Using a high resolution electron microscope, Jones and Uehara(1973) were

able to show the presence of amorphous gel-hull linkage between clay

particles. Such amorphous gel hull must include noncrystalline AI(OH)3. The

presence of AI(OH)3 as a thin layer around illite particles also has been

reported by El Swaify and Emerson (1975). Because AI(OH)3 decreases its

positive charge with increasing pH or particle size, its cementing effectiveness

must follow the same trend. El Swaify and Emerson (1975) showed that

AI(OH)3 exerted little effect on aggregation at its point of zero charge. In

principle, well-crystallized AI(OH)3 also may be able to act as a cementing 73 agent in acid media, but its magnitude may be negligible as compared with

noncrystalline material. Iron oxides have a stronger tendency than AI(OH)3

to crystallize. Therefore, iron oxides are generally less effective than AI(OH)3

in its cementing effectiveness except in frequently alternating oxidized and

reduced soils (Frenkel & Shainburg, 1980). Because the AI(OH)3 and Fe

oxides/oxyhydroxides in soils vary widely in crystallinity and particle size, a

poor correlation between their contents and aggregation does not necessarily

indicate that they do not exert any effect on aggregation.

Organic matter also has been known to be closely associated with soil

aggregation (Greenland, 1971; Tisdall and Oades, 1982). Nevertheless, much of the soil O.M. is negatively charged and is thus not likely to react directly with clay particles. It has been suggested that O.M. promotes soil aggregation through the following linkage: clay - (Al, Fe) - organic matter - (Al,

Fe) - clay (Edwards and Bremner, 1967). Therefore, the organic matter in soil samples should also be considered in assessing the effectiveness of AI(OH)3 or Fe oxides/oxyhydroxides in aggregation.

Theng and Scharpenseel (1975) and Theng (1976) suggested that Al and Fe should be the most efficient cations in linking acidic organic materials to clay surfaces. The soils of concern are rich in organic matter, Al, and Fe oxides. Tisdall and Oades (1982) suggested to consider micro- and macroaggregates to be defined as less than and greater than 250 urn diameter. The micro-aggregates are stabilized against disruption by rapid 74 wetting and mechanical disturbance including cultivation, by several mechanisms in which organo-mineral complexes play a dominant role.

Polysaccharide are also involved. The binding of micro aggregates appeared to be relatively permanent and was not influenced by changes in the organic matter content of the soil caused by different management. On the other hand, the water-stability of macroaggregates depends largely on roots and hyphae, and thus on growing root systems. Numbers of stable macroaggregates decline with organic matter content decreases as the roots and hyphae are decomposed and are not replaced.

Due to the high organic matter content and the low pH values, the iron oxides found in the soils were poorly crystallized. XRD were devoid of iron mineral peaks. Norrish and Taylor, 1961; Juo et al., 1974; Fey and Le Proux,

1977 indicated that most iron oxides in soils of the tropics were highly aluminum substituted. Bigham et al.(1978) showed that the iron oxides were finely divided but crystalline. Due to their low crystallinity, their small particle size and their high degree of substitution, these clays may develop a relatively high specific surface with associated pH dependent positive and negative charges.

Although generally accepted, the aggregating effect of Fe oxides remains somewhat controversial and seems to vary with different soils. This is probably due to variations in the amount and nature of their Fe oxides, but also to the nature and environment of their pedogenic formation. McIntyre, 1956; Kemper, 1966, and Area and Weed, 1966 demonstrated a significant

correlation between the percent of water stable aggregates or related

structural properties and the content of Fe oxides. McNeal et al. (1968)

demonstrated the dispersion of aggregated soils after removal of their Fe

oxides with a reducing agent. Schahabi and Schwertmann, 1970, and

Blackmore, 1973 demonstrated the aggregating effect of added synthetic Fe oxides. In soils, these high-surface-area Fe oxides may comprise only a very small fraction of the total Fe oxides, explaining why Cambier and Prost (1981)

had to remove only a very small proportion of the total Fe oxides in order to disperse a ferralitic soil from Senegal. Other experiments on this subject indicate, however, a rather weak aggregating effect of Fe oxides. For example, in a number of soils from Denmark and Tanzania, the removal of Fe oxides did not change the degree of aggregation (Borggaard, 1983). Also, effective dispersion of a Red Brown Earth was obtained after treatment with an acidified CaCI2 solution (pH 1.5), which extracted little Fe but considerable amounts of Al (Deshpande et al., 1964, 1968). One of the reasons for the negative effect can be deduced from transmission electron microscopy observations: In kaolinitic soils, many Fe-oxide crystals appear to be aggregated among themselves rather than bonded to the kaolinite flakes, which show rather clean surfaces (Greenland et al.,1968; Jones et al., 1982;

Schwertmann and Kampf,1985). In analyzing single microaggregates (<2 urn) by a microprobe, Fordham and Norrish (1979, 1983) found a large range of 76 Fe concentrations. This indicates that microaggregates may consist essentially of Fe oxides, or essentially of kaolinite, or of mixtures of the two.

On the soil studied, they are geographically close to the Tanzanian soils. Removal of Fe oxides resulted in increased clay dispersion. This increase, however, was far from total dispersion.

Aggregation at the study area must, therefore, involve Al, besides Fe oxides and organic matter. Dispersion should be improved after Al oxides removal. It is interesting to investigate the effect of gibbsite, humic and fulvic acids of these soils on the aggregation phenomenon. The soils developed in micaschist underlain by granite. Muscovite occurring in books was observed along deep road entrenchments near most of the sites studied. Since micas in most soils originate mainly from soil parent materials and tend to weather to other minerals with time, they are generally more prevalent in the clay mineralogy of young soils and less prevalent in more weathered soils

(Jackson et al., 1948, 1952). Soils developed from materials rich in coarse micas, the maximum micas content on a whole-soil basis should occur in the

C horizon because of greater micas weathering in the A and B horizons. This was the case in the soils studied. Micas contents increased with depth for the silt and sand fractions. They were limited in the surface horizons for the total clays. They were absent in the fine clay fractions.

Quartz was found in very low amounts in all the horizons of the soil studied. Quartz is a major constituent of most soils because 1) it is the 77 second most common mineral, after feldspars, in the earth’s crust, and 2) it is highly resistant to weathering. Quartz may be subjected to dissolution under intense weathering when the grains are very small (Senkayi et al.,

1985). Sand and silt fractions were mainly composed of quartz. The remaining components being muscovite, phytolite and rarely tourmaline.

Phytolite was relatively abundant in the spodic horizons.

Soil clays are usually less well ordered and smaller in size than the pure minerals. They often overlap neighboring particles of sheets and interleafings and interstratifications of various layer silicates are common. The mineralogy of soil clays is rarely simple or uniform. Coating of iron and aluminum oxides and organic matter on most layer silicates further complicate the mineralogy of soil clays. Such coatings can alter mineral properties by decreasing CEC and surface area values and by restricting the swelling and collapsing of expansible minerals. Oxide coatings magnify anion exchange and other properties associated with positively charged surfaces (Bohn etal.,

1979). Mineralogical comparisons among soils under natural forest and their counterparts under pasture and cultivation did not detect any difference. CHAPTER 4

GENESIS AND CLASSIFICATION OF HIGH-ALTITUDE SOILS IN BURUNDI

4.1. Introduction

The study area is located in the uplifted zone of Burundi that lies along the Western Rift of the East African Rift Valleys. The altitude ranges from

1,500 meters to 2,600 meters. During the late Cenozoic Era, much of eastern

Africa was arched upward nearly 3,000 meters above sea level. Fracturing and faulting occurred probably during the Pliocene Epoch across the crest of the arch. The famous rift valleys formed as narrow slivers of crust that had slipped downward between great fault blocks. As the faulting continued, volcanos developed along the fault trends. Mount Kenya and Mount

Kilimanjaro are two of the most notable of these volcanoes, but there are many others as well (Levin, 1983).

The soils of concern develop on the eastern - horst parallel to Lake

Tanganyika, the second deepest lake in the world. The hilltops of this area constitute a divide between two important drainage basins: the Nile basin and the Zaire basin. According to Soil Taxonomy (Soil Survey Staff, 1992) the soil moisture regime is udic and the soil temperature regime is isothermic or isomesic. The soils develop in residuum from metamorphic rocks ( gneiss, schist, quartzite and their intermediates) and from intrusive igneous rock

(essentially granite). Most of the soils of concern were vegetated by tropical natural forest until about 1950. The soils under natural forest are part of a

78 79 complex ecosystem that concentrates plant and microbial nutrients in the

surficial layers. Despite the luxurious biomass covering the soils, they are

highly weathered acid soils. The luxurious vegetation is sustained by the

efficient cycling and recycling of organic matter in the soils that results in the

overall improvement in soil conditions. These include improvement of physical

properties, decreased susceptibility to erosion, encouragement of microbial

activity and provision of potentially available plant nutrients (Roose, 1977).

Deforestation in the study area has been performed by cutting trees

manually and clearing the land by fire. When the forest is cleared, the

raindrops fall upon the bare soil and their impact loosen the soil particles so

that they may be more easily transported by flowing water. Sheet erosion and

gullies make some of the soils almost useless for cropping. Some soils erode

to the bed rock, hardpan, or less fertile subsoils.

The study area, uplifted prior to Pleistocene Epoch, exhibits soils that

present numerous challenges to soil scientists. Most of the encountered soils

are composed of spodic-like materials underlain by oxidic materials. To

accomplish a sound genesis study of these soils, one must consider the following:

1) the rift valleys that formed during the late Cenozoic Era, probably during Pliocene Epoch, fractured Precambrian-aged rocks (Cryptozoic Eon) that has most likely weathered to oxidic materials, 80

2) the changes of climate that occurred during the Pleistocene Epoch

(from a warm and humid environment to a cool and humid one) might have

favored podzolization processes,

3) the parent material diversity has been nullified by the extremely high

degree of weathering (Chesworth, 1973).

The most probable scenario conducive to such a situation where

spodic-like materials superimpose oxidic materials is that the study area was

highly weathered prior to faulting and uplifting, and the spodic properties formed during the cool and humid parts of the Pleistocene Epoch. The

luxurious forest that emerged under high rainfall constituted an ideal

ecosystem for podzolization processes. The objectives were to evaluate the

effects of deforestation on the genesis and classification of the soils.

4.2. Materials and Methods

4.2.1. Field Setting

The study area is located at the uplifted zone in Burundi that lies along the Western Rift at the altitude ranging from 1,500 to 2,600m. Three sets of soil profiles were selected along three nearly parallel bands that cross a succession of natural forest, cultivated and pastured lands. Along each band, the soil profiles were selected in such a way that one soil was located under natural forest, the other two under cultivated and pastured lands. Overall, nine soil profiles, three under natural forest, three under cultivated and three under pastured lands were studied. 81 4.2.2. Laboratory Analyses

Iron and Al were extracted using several methods. Dithionite - citrate -

bicarbonate (DCB) extractions were performed using Mehra and Jackson

method (1960). Ammonium acid oxalate extractions were performed using

McKeague and Day method (1966) and sodium pyrophosphate extractions

were performed using the Bascomb method (1968). Superfloe N-100

(American Cyanmid Co., Wayne, N.J.) was added to the soil samples prior to

centrifugation. High-speed centrifugation was used to clear the extracts from

particulate matter that otherwise stayed in suspension when low-speed

centrifugation was used. Iron and Al in the extracts were analyzed by ICP.

Particle size distribution (PSD) was determined by the pipette method (Day,

1965), but due to the poor clay dispersion, the clay fraction was estimated by

2.5 times the 15 bars pressure plate water content values (Soil Survey Staff,

1992). Bulk densities were determined by the clod method (Blake, 1965; Soil

Survey Staff, 1984) and the ring method (Uhland, 1949, 1950). The exchangeable Al and H were determined by the Yuan method (1959). The extractable cations and the cation exchange capacities were determined by

Peech etal. method (1947). Extractable acidity was determined by extraction with BaCI2 triethanolamine pH 8.2 Mehlich method (1938). Organic carbon was determined by acid dichromate digestion (Wakley, 1935 and Peech etal.,

1947). Clay mineralogy was determined by XRD. Results were interpreted by analysis of depth functions of different data. 82 4.3. Results

4.3.1. Forest Soils

Depth functions of DCB, ammonium acid oxalate and sodium pyrophosphate extracts have been plotted for different Fe and Al forms.

These different fractions are referred as Fed and Ald for DCB extracts, FeQ and AL for ammonium acid oxalate extracts and Fen and AL for sodium u p p pyrophosphate extracts. The same denominations are used for all the soil profiles’ data.

For the Bugarama forest soil, Fed values ranged from 1.9 to 3.0%; Fe0 values ranged from 0.7 to 1.7%; Fep values ranged from 1.5 to 1.9% (Fig.

4.1). All three Fe fractions exhibited minima at the lowest part of the umbric epipedon (44 to 88 cm, Bhs2 horizon) and maxima within the spodic horizon

(121 to 152 cm, Bhs4 horizon). Aluminum dithionite values ranged from 0.9 to 2.3%; Al0 values ranged from 1.0 to 2.0%; Fep values ranged from 0.9 to

2.0% (Fig. 4.2). Both AIQ and Ald values exhibited pronounced minima and maxima that coincided with those of the Fe fractions. The shapes of the depth functions for the three Fe fractions and those for AIQ and Ald fractions were almost identical and displayed the classical eluvial-illuvial process pattern. The Alp and, to some extent the Fep fraction, seemed less affected by the processes.

For Teza forest soil, Fed values ranged from 1.3 to 2.5%; Fe0 values ranged from 0.5 to 1.2%; Fep values ranged from 0.7 to 1.4% (Fig. 4.3). i. .. Depth distribution fractions I.Fesite of at Fig. 4.1. DEPTH (CM) 152-195 121-152 44-88 20-44 BUGARAMA FOREST BUGARAMA 10 . 30 4.0 3.0 2.0 1.0 0 Fo ract %) (% ts c a tr x e e F Fe 110-129 & Q 76-110 s 167-210 141-167 129-141 50-76 30-50 0-30 MURAMVYA CULTURE MURAMVYA 10 . 30 4.0 3.0 2.0 1.0 0 ract %) (% ts c a tr x e e F F©

DEPTH (CM) 179-210 129-179 94-129 65-94 35-65 19-35 0-19 MURAMVYA PASTURE MURAMVYA 10 . 30 4.0 3.0 2.0 1.0 0 ract %) (% ts c a tr x e e F Fe Fe Fed CO 00 i. .. et distributions Depth offractionsAl I. site at Fig. 4.2. DEPTH (CM) 121-152 88-121 20-44 BUGARAMA FOREST MURAMVYA CULTURE MURAMVYA FOREST BUGARAMA . 10 . 20 2.5 2.0 1.5 1.0 0.5 Al extracts(%) Al Al Al, I zc 76-110 g u u o_ -- 167-210 141-167 50-76 30-50 . 10 . 20 2.5 2.0 1.5 1.0 0.5 l xrcs (%) extracts Al Al| crs LXJ 65-94Q_ ^ cz> 179-210 129-179 94-129 35-65 19-35 UAYA PASTURE MURAMYVA 0-19 letat (%) extracts Al . 15 . 2.5 2.0 1.5 1.0 Al 00 - p >. i. .. et distributions Depth fractions II.siteFe at of Fig. 4.3. DEPTH (CM) 168-200 74-125 36-74 10-36 0-10 168 TEZA FOREST FOREST TEZA e xrcs (%) extracts Fe . 20 . 4.0 3.0 2.0 1.0 Fe re LU Q_ i— 159-200 110-159 " 85-110 71-85 46-71 21-46 BUKEYE CULTURE BUKEYE eetat (%) extracts Fe . 20 . 4.0 3.0 2.0 1.0 Fe Fe cn u_i o> _ n 171-200 138-171 96-138 63-96 40-63 16-40 0-16 10 . 30 4.0 3.0 2.0 1.0 0 UEE PASTURE BUKEYE eetat (%) extracts Fe U1 CD 86 Except for the surface horizon, ail three Fe depth functions exhibited minima

at the lowest part of the umbric epipedon (74 to 125 cm, Bhs2), and the Fed and Fe0 maxima were in the spodic horizon (125 to 168 cm, Bhs3).

Aluminum dithionite values ranged from 0.5 to 0.9%; Al values ranged from r 0.3 to 0.8% (Fig. 4.4). All three Al depth functions exhibited depletion at the lowest part of the umbric epipedon and addition within the spodic horizon.

The maxima occurred between 125 to 168 cm (Bhs3).

For the Rwegura forest soil, Fed values ranged from 1.8 to 2.6%; FeQ values ranged from 1.2 to 2.3%; Fe values ranged from 1.3 to 2.1% (Fig. r 4.5). The three Fe depth functions were constant throughout the soil profiles control section. Aluminum dithionite values ranged from 0.7 to 1.4%. The AIQ values ranged from 0.4 to 1.4%. The Alp values ranged from 0.5 to 1.5%

(Fig. 4.6). The depth distributions of all three forms of Al exhibited depletion at the lowest part of the umbric epipedon (51 to 81 cm, Bhs1), and maxima were within the spodic horizon (81 to 112 cm, Bhs2).

4.3.2. Culture Soils

The Fed values for the Muramvya culture soil ranged from 1.8 to 2.3%;

Fe0 values ranged from 0.1 to 0.8%; Fep values ranged from 0.2 to 1.3% (Fig.

4.1). For all three Fe forms, the maxima occured within the umbric epipedon.

Aluminum dithionite values ranged from 0.3 to 0.7%; AIQ values ranged from

0.1 to 0.4%; Al values ranged from 0.1 to 0.6% (Fig. 4.2). r i. .. et distributions Depth offractions II. Alsite at 4.4.Fig. DEPTH (CM) 168-200 125-168 74-125 36-74 10-36 letat (%) extracts Al EA FOREST TEZA . 10 1.5 1.0 0.5 l, A l, A D C LU Q_ I— Z H . D C

159-200 110-159 85-110 71-85 46-71 21-46 0-21 UEE CULTURE BUKEYE letat (%) extracts Al . 10 1.5 1.0 0.5 CD D C LU CL. I— Z H

171-200 138-171 96-138 63-96 40-63 16-40 05 . 1.5 1.0 0.5 0 UEE PASTURE BUKEYE Ietat (%) extracts AI -vj 00 i. .. et distributions Depth fractions III.siteFe at of Fig. 4.5. DEPTH (CM) 151-200 112-151 21-51 0-21 ■112 ■81 RWEGURA FOREST RWEGURA eetat (%) extracts Fe . 20 . 3.0 2.5 2.0 1.5 64-96 Q Q_ 33-64 s 128-170 96-128 14-33 0-14 RWEGURA CULTURE RWEGURA . 10 . 20 . 3.0 2.5 2.0 1.5 1.00.5 eetat (%) extracts Fe Q 5:80-130 . 180-210 130-180 46-80 30-46 0-30 WGR PASTURE RWEGURA 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 Fe extracts Fe Fe 00 00 i. .. et distributions Depth offractions Alsite at Fig. 4.6. DEPTH (CM) 151-200 51-81 21-51 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 RW EGURA FOREST FOREST EGURA RW letat (%) extracts Al 128-170 96-128 33-64 14-33 0-14 RW EGURA CULTURE CULTURE EGURA RW 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 letat (%) extracts Al Al Al 80-130 Q Q_ <_3. 180-210 130-180 46-80 30-46 0-30 WGR PASTURE RWEGURA 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 letat (%) extracts Al Al, CO 00 90 For the Bukeye culture soil, Fed values ranged from 2.3 to 3.3%; FeQ values ranged from 0.3 to 1.2%; Fep values ranged from 0.8 to 1.7% (Fig.

4.3). Aluminum dithionite values ranged from 0.6 to 1.0%; AIQ values ranged from 0.2 to 0.4%; Alp values ranged from 0.4 to 0.8% (Fig. 4.4).

For the Rwegura culture soil, Fed values ranged from 2.0 to 2.1%; Fe0 values ranged from 0.4 to 1.3%; Fep values ranged from 0.8 to 1.7% (Fig.

4.5). Aluminum dithionite values ranged from 0.5 to 1.0%; AIQ values ranged from 0.2 to 0.8%; Alp values ranged from 0.3 to 1.5% (Fig. 4.6). All three Al depth functions exhibited maxima between 14 to 33 cm (Ap2). Morphological properties along with physico-chemical properties suggested that this horizon was a sombric horizon and the profile had been truncated.

4.3.3. Pasture Soils

For the Muramvya pasture soil, Fed values ranged from 2.8 to 3.6%;

Fe0 values ranged from 0.2 to 1.3%; Fep values ranged from 0.6 to 1.8% (Fig.

4.1). Aluminum dithionite values ranged from 0.7 to 1.3%; AIQ values ranged from 0.2 to 1.5%; Alp values ranged from 0.2 to 1.1% (Fig. 4.2). With the exception of Al0 which exhibited a maximum between 19 to 35 cm (AB horizon), Fe and Al depth functions exhibited maxima within the sombric horizon (65 to 94 cm).

For the Bukeye pasture soil, Fed values ranged from 2.2 to 2.9%; Fe0 values ranged from 0.2 to 1.0%; Fep values ranged from 0.9 to 1.3% (Fig.

4.3). Aluminum dithionite values ranged from 0.7 to 0.8%; AIQ values ranged 91 from 0.2 to 0.4%; Alp values ranged from 0.3 to 0.6% (Fig. 4.4). All six depth functions exhibited high values within the Bhs horizon (96 to 138 cm).

For the Rwegura pasture soil, Fed values ranged from 1.4 to 1.9%; Fe0 values ranged from 1.0 to 1.9%; Fep values ranged from 0.7 to 1.7% (Fig.

4.5). Aluminum dithionite values ranged from 0.6 to 1.1%; Al0 values ranged from 0.5 to 1.3%; Alp values ranged from 0.6 to 1.7% (Fig. 4.6). All six depth functions exhibited maxima within the oxic horizon. The Fe0/Fed values also exhibited their maxima within the oxic horizon.

4.4. Discussions

4.4.1. Genesis of High-Altitude Soils in Burundi

The elements Fe, Al and Mn are greatly affected by the processes of soil profile genesis. The distribution of these pedogenic oxides and hydroxides can be used to describe the type, direction and extent of pedogenic processes (Schlichting and Blume, 1966). Many pedological processes such as podzolization, gleization and laterization involve mobilization and redistribution of Fe and Al. To evaluate the depth functions of the pedogenic oxides, chemical processes are available to dissolve or extract some or all of the continuum, from crystalline oxides through materials with shortrange order such as ferrihydrite and organic complexes with Al3+ and Fe3+.

To study the forms of Fe and Al present in the soils, we used three well known selective reagents: dithionite-citrate-bicarbonate (DCB), ammonium 92 acid oxalate (AAO), and sodium pyrophosphate (SPP). Dithionite-citrate- bicarbonate extracts crystalline, amorphous and organically bound Fe and Al.

Ammonium acid oxalate extracts amorphous and organically bound Fe and

Al (Campbell and Schwertmann, 1984) and SPP extracts organically bound

Fe and Al (McKeague et al., 1983). In theory, DCB should extract more Fe and Al than AAO, and the latter should extract more Fe and Al than SPP.

Contrary to those expected trends, our results disclose that Fe0% <

Fep% < Fed% for most of the horizons in all the soil profiles except for

Rwegura forest where the trend is Fep% < Fe0% < Fed%. The general trend is Fep% > Fe0% in most soil profile horizons except for the spodic horizons.

Similar trends have been reported by Andriesse (1979), Loi (1980),

Mutwewingabo (1984) and SMSS-IBSNAT (1986). For most of the tropical soils analyzed, Andriesse (1979) speculated that the Fe-organic complexes in tropical soils could be different in nature from those in temperate region soils, or that SPP could preferentially dissolve aluminous goethite. Loi (1980) suggested that suspended colloidal iron oxide could also give high Fe r values. Silicates enriched with geothite remained suspended in SPP extracts after low-speed centrifugation has been shown by Bascomb (1968).

According to Sheldrich and McKeague (1975), either high-speed or low-speed centrifugation with addition of superfloe N-100 can be used to flocculate particulate matter in SPP extracts. 93 We added the superfloe to the soil samples prior to high-speed

centrifugation to clear the extracts from possible suspended solids. We used

the same procedure for DCB, AAO and SPP to obtain comparable results.

Our results show Fep > Fe0 and Alp > Al0 in most cases. This supports

Andriesse (1979) suspicion that the Fe-organic complexes in tropical soils

might be different in nature from those in temperate region soils.

From these results, it seems obvious that the AAO does not extract all

the organic Fe and Al. Oades (1981) reported that in the surface horizons of

leached soils, which usually are loams to sands with pH values of 6 or less,

the organic matter (O.M.) cycle is predominantly controlled by oxidative

degradation to a C 02 end product. Insoluble plant components become large

aromatic humic acids soluble in NaOH but not in Na4P20 7. Infrared

spectroscopy has shown that these humic acids retain many characteristics

of lignin (Posner et al., 1968), which are lost as the humic acids become more

acidic and extractable by Na4P20 7. Although these results seem to challenge

the specificity of SPP and AAO, these trends may elucidate the nature of the

humic substances present in these soils. The poor efficiency of the AAO in

extracting Fe and Al associated with O.M. may be attributed to the nature of the O.M. Humic acids are soluble in alkali but are insoluble in acid.

Conversely, fulvic acid are soluble in both alkali and acid. In a case where

humic acids constitute an important component of the humic substances of concern, 0.1M SPP at pH 10 should extract more Fe and Al associated with 94 O.M. than AAO at pH 3. The SPP’s efficiency results from both its alkalinity

and complexing ability that keep both humic and fulvic acids in solution.

Humic acids and their associated Fe and Al precipitate at pH 3 when

ammonium oxalate is used. It seems logical to obtain lower Fe and Al

concentration with AAO compared to those obtained with 0.1M SSP at pH 10.

We, therefore, suggest that the humic substances present in these soils are

mainly composed of humic acids rich in small aromatic carboxyl functional

groups. Aluminum and Fe cations bridge clay and humic acids that result in

very stable organo-mineral associations. This is supported by the unusually

high aggregate stability (90% and more) and the unusually low water

dispersible clay (traces for soils under forest). Otherwise, it is difficult to

explain why the addition of O.M. to pastured soils tremendously improves their

productivity, though they are already rich in O.M. The most sound

explanation is that the O.M. present in these soils is highly resistant to

mineralization, water insoluble and consequently does not contribute to the

nutrient supply for plants. Addition of O.M. also supplies a new stock of microorganisms that reactivate O.M. decomposition to lower molecular forms that are more water soluble. Research conducted on Fargo clay (a Humic

Clay) at the North Dakota Agricultural Experimental Station at Fargo from

1913 to 1956 (Young et al., 1960) reported that plots receiving P in addition to manure or residues lost slightly more total N and O.M. than those receiving only the manure or residues. They concluded that it might be possible for P 95 to stimulate microbiological activity, thus resulting in faster decomposition of the original organic compounds. Most of the O.M. in soils under pasture, especially in the oxic horizon, are probably precipitates of humic substances with Fe and Al that are quite resistant to decomposition. These soils are P deficient and the nature of the clay mineral along with low pH that favor P retrogradation aggravate P availability to soil microfaune. According to Theng and Scharpenseel (1975) and Theng (1976), Al and Fe should be the most efficient cations in linking acid organic materials to the clay surfaces. The subject is complicated, however, because Al and Fe hydrolyze to form a range of polyhydrospecies or polycations, hydroxides, oxyhydroxides and oxides, which may or may not be associated with clay surfaces (Oades,

1984).

Under forest, our results, along with those reported by previous investigators (Mutwewingabo, 1984, SMSS-IBSNAT 1986, along with numerous unpublished data) show that for Fe forms, all the soil profiles follow the FeQ% < Fep% < Fed% trend. But the Rwegura soil exhibits a Fep% <

FeQ% < Fed% trend. For aluminum forms, the trend is Al0% < Ald% < Alp% for most of the soil profile horizons. This is well pronounced in the lower portion of the umbric epipedon. The trend in the spodic and/or oxic horizons is rather ill defined. An overall podzolization picture of the soil profiles is most obvious on the Fe0 and Al0 depth functions figures. The morphology of these soil profiles does not display any clue relative to the podzolization process, 96 the high amount of O.M. present in the soils masks the eluvial features. The

Fe and Al chemistry discloses the presence of a depleted horizon in Fe and

Al oxides that overlies an incipient enriched horizon. We designate this

process as crypto podzolization (hidden podzolization).

The overlap of the six depth functions (three Fe and three Al functions) together with Fed/clay depth functions (Fig. 4.7, Fig. 4.8, and Fig. 4.9) suggest co-migration of Fe and Al. The Fe and Al illuvial fractions correlate with the amount of illuvial O.M. Blume and Schwertmann (1969) reported that the Ald maxima that occurs in the spodic horizon is characteristic of crypto podzolization. This process is best illustrated by the Al0, Ald and AIQ+1/2 Fe0 depth functions that clearly show the depleted horizons underlain by the incipient enriched spodic horizon (Fig. 4.10, Fig. 4.11 and Fig. 4.12). This trend is moderate under the culture soils (Fig. 4.10, Fig. 4.11 and Fig. 4.12).

It is still obvious under pasture, especially when the sombric horizon is well developed (Fig. 4.10, Fig. 4.11 and Fig. 4.12).

Crystallization of iron oxides in soils is a major process of soil genesis.

The FeQ/Fed ratio ("activity ratio") is used as a relative measure of the degree of aging or crystallinity of free iron oxides (Schwertmann and Kodama, 1986).

High Fe0 coincide with high FeQ/Fed ratios, indicating a low degree of aging

(Fig. 4.7, Fig. 4.8 and Fig. 4.9). This corroborates the XRD results which do not display any iron mineral. This low degree of aging results from unfavorable conditions for crystallization. This is equally valid for soils under i. .. et distributions Depth ratiossite I. Fe0/FedatFed/clay and of Fig. 4.7. HORIZONS Bhs4 Bhs2 BUGARAMA FOREST BUGARAMA 05 . 15 2.0 1.5 1.0 0.5 0 10 (Fejj/Clay) Ratios, Ratios, (Fejj/Clay) 10 e/e Ratios Fec/Fed 1 O(Fed/Clay) g BI2 g I Bt/R Ap2 B'(2 Apl B'tl Btl AB MURAMVYA CULTURE MURAMVYA 10 (Fed/Clay) Ratios, Ratios, (Fed/Clay) 10 05 . 15 2.0 1.5 1.0 0.5 0 e/e Ratios Feo/Fed 10 (Fed/C!ay)10 O z ac r-vi O O C Bhs2 Bhsl B's AB Bs Bt A MURAMVYA PASTURE MURAMVYA 10 (Fed/Clay) Ratios, Ratios, (Fed/Clay) 10 05 . 15 2.C 1.5 1.0 0.5 0 e/e Ratios Fe0/Fed 1 O(Fed/Clay) i. .. et distributions Depth ratios II.site Fe0/Fed at Fed/clay and of Fig. 4.8. HORIZONS Bhs2 Bhsl Bhs3 AB Bt A 05 . 15 2.0 1.5 1.0 0.5 0 10 (Fed/Clay) Ratios, Ratios, (Fed/Clay) 10 EA FOREST TEZA Fe0/FedRatios 1 O(Fed/Clay) O □c r^j O 2 GO : Ap2 Apl B15 Bt4 BI3 BI2 Btl 10 (Fed/Clay) Ratios, Ratios, (Fed/Clay) 10 05 . 15 2.0 1.5 1.0 0.5 0 UEECLUEBKY PASTURE BUKEYE CULTURE BUKEYE e/e Ratios Fe0/Fed 1 O(Fed/Clay) o oc r^j O O C B'sl B's2 Bs2 8hs Bsl AB A 10 (Fed/Clay) Ratios, Ratios, (Fed/Clay) 10 05 . 15 2.0 1.5 1.0 0.5 0 / !■ - - - 1 1 - II A e/e Ratios Fe0/Fed 1 l< L -< " ' 1 ' "1 T 1.— o ^ > ---- 1 F©0/F©d O(Fed/Olay) ------i. .. Depth distributions ratios III.site Fe0/FedFed/clay at of and Fig. 4.9. HORIZONS Bhsl Bhs3 AB Bt A 10 (Fe

cn J 110-129U g •S 76-110 ~ Aid, Al0 and Al0 + 1/2 Fe0 Fe0 1/2 + Al0 and Al0 Aid, 167-210 129-141 50-76 30-50 MURAMVYA CULTURE MURAMVYA 0-30 Extracts in % in Extracts . 15 . 2.5 2.0 3.01.0 1.5 .AI - ai q + *1/2+ Fe Q CL. 179-210 129-179 94-129 Aid, Alo and Alo + 1/2 Feo Feo 1/2 + Alo and Alo Aid, 65-94 35-65 19-35 0-19 MURAMVYA PASTURE MURAMVYA 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 Extracts in % in Extracts AI q - h / Fe 1/2 Al, i. .1 Depth distributions fractions of Fe0 Ald, II. site AIQ and Al0 1/2 atFig. 4.11. + DEPTH (CM) 168-200 125-168 74-125 36-74 10-36 0-10 i,AoadAo+ / Fo Al n 0+ / Fe0 1/2 + l0 A and l0 A , ^ A Feo 1/2 + Alo and Alo Aid, 05 . 15 2.0 1.5 1.0 0.5 0 TEZA FOREST TEZA Extracts in % in Extracts Fe o U L Q 71-85Q_ ^ , o 159-200 110-159 46-71 21-46 BUKEYE CULTURE BUKEYE Extracts in % in Extracts . 10 . 2.0 1.5 1.0 0.5 ,Alo-<-1/2 Fe Q U L Q_ 63-96 ^ CJ> 171-200 138-171 - Al(j, Al(j, 96-138 40-63 16-40 UEE PASTURE BUKEYE 05 . 15 2.0 1.5 1.0 0.5 0 AI q E and and xtracts in % in xtracts l + / Fe 1/2 + Al0 AI q + / F© 1/2 + Al, q i. .2 Dph distributions Depth fractions Fe0 of III. Ald,site 1/2 at AIQ Fig. 4.12. + Al0 and DEPTH (CM) 151-200 112-151 81-112 51-81 21-51 Aid. Al0 and Al0 + 1/2 Fe0 Fe0 1/2 + Al0 and Al0 Aid. 0-21

WGR FOREST RWEGURA 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 Al xrcs n % in Extracts Al 33-64 . s o i 128-170 64-96 14-33 0-14 Aid, Alo and Alo + 1/2 Feo Feo 1/2 + Alo and Alo Aid, ■128

RWEGURA CULTURE RWEGURA 05 . 15 . 25 3.0 2.5 2.0 1.5 1.0 0.5 0 oi - xrcs n % in Extracts l 12 Fe. 1/2 + Alo 80-130 Q Q_ Aid, Alo and Alo + 1/2 Feo Feo 1/2 + Alo and Alo Aid, 180-210 130-180 46-80 30-46 0-30 RWEGURA PASTURE RWEGURA 05 . 15 . 25 3.0 2,5 2.0 1.5 1.0 0.5 0 xrcs n % in Extracts iq ♦ " F©, *♦* "1 2 / Al 103 culture (Fig. 4.7, Fig. 4.8 and Fig. 4.9) and those under pasture (Fig. 4.7, Fig.

4.8 and Fig. 4.9). The retarding effect of organic compounds in the

crystallization of iron oxides has been demonstrated by Schwertmann (1966)

and Schwertmann et al. (1968). Hence, high Fe0/Fed ratios in surface soils

and in sombric horizons are due to O.M. Also, the low pH values retard the

aging process. Since no subsoil clay maxima are found in these soil profiles,

clay enrichment can not be responsible for the Al0 or Fe0 maxima in the

subsoil (Table 4.1, Table 4.2 and Table 4.3). The Al0 maximum, and other

maxima as well, is explained by a downward migration of Al (and Fe). In

general, these well-drained soils exhibit nearly constant Fed/clay ratio depth

functions throughout the soil profile.

For soils under culture, the depth functions of Fe and Al disclose slight

podzolization features. The Fed/clay depth functions are in most cases

straight lines parallel to the depth axis. This is characteristic of oxidic

materials. The effective cation exchangeable capacity (ECEC) cmol+ Kg'1 clay values are less than 12 throughout the soil profiles (Fig. 4.13, Fig. 4.14 and

Fig. 4.15).

For soils under pasture, all the surface horizons exhibit Al accumulations. This absolute addition results from lateral flow. High amounts of O.M., Fe and Al are also found in the sombric horizons. The shape of these depth functions suggests slight podzolization. The ECEC values show the oxidic properties of the soils (Fig. 4.13, Fig. 4.14 and Fig. 4.15). 104

Table 4.1. Particle Size Distribution at Site I.

Bugarama Forest

Depth (CM) Horizon Sand silt clay (%) (%) (%) 0-20 A 31.8 5.5 62.7 20-44 Bhsl 29.4 8.1 62.5 44-88 Bhs2 32.4 4.1 63.5 88-121 Bhs3 31.6 3.7 64.7 121-152 Bhs4 32.8 3.2 64.0 152-196 Bt 34.6 3.9 61.5

Muramvya Culture

Depth (CM) Horizon Sand Silt Clay (%) (%) (%).. 0-30 Ap1 40.9 0.4 58.7 30-50 Ap2 35.9 7.9 56.2 50-76 AB 34.7 8.1 57.2 76-110 Bt1 35.5 9.5 55.0 110-129 Bt2 37.3 10.2 52.5 129-141 Bt/R 42.1 9.7 48.2 141-167 B’t1 46.6 6.7 46.7 167-210 B’t2 45.1 8.7 46.2

Muramvya Pasture

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-19 A 33.9 2.4 63.7 19-35 AB 30.1 7.4 62.5 35-65 Bs 30.3 7.2 62.5 65-94 Bhsl 32.7 7.0 60.3 94-129 Bhs2 33.0 6.7 60.3 129-179 B’s 33.8 6.9 59.3 179-210 Bt 32.8 6.5 60.7 105 Table 4.2. Particle Size Distribution at Site II.

Teza Forest

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-10 A 29.9 7.6 62.5 10-36 AB 38.1 5.4 56.5 36-74 Bhsl 37.9 7.4 54.7 74-125 Bhs2 40.0 6.5 53.5 125-168 Bhs3 39.3 5.2 55.5 168-200 Bt 44.9 6.4 48.7

Bukeye Culture

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-21 Ad1 34.6 1.7 63.7 21-46 Ad2 32.6 4.4 63.0 46-71 Bt1 28.0 9.3 62.7 71-85 Bt2 26.3 10.0 63.7 85-110 Bt3 27.1 11.4 61.5 110-159 Bt4 30.1 7.9 62.0 159-200 Bt5 29.9 5.6 64.5

Bukeye Pasture

Depth (CM) Horizon Sand Silt Clay ...(%) (%) (%) 0-16 A 37.6 1.2 61.2 16-40 AB 32.8 5.5 61.7 40-63 Bs1 33.6 6.7 59.7 63-96 Bs2 30.3 8.2 61.5 96-138 Bhs 30.1 8.7 61.2 138-171 B’s1 31.5 7.8 60.7 171-200 B’s2 31.0 8.8 60.2 106 Table 4.3. Particle Size Distribution at Site III.

Rwegura Forest

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-21 A 28.3 4.7 67.0 21-51 AB 27.9 6.4 65.7 51-81 Bhsl 28.5 9.5 62.0 81-112 Bhs2 28.1 10.6 61.3 112-151 Bhs3 26.6 11.1 62.3 151-200 Bt 30.8 6.9 62.3

Rwegura Culture

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-14 Ao1 37.7 12.3 50.0 14-33 Ad2 31.8 12.7 55.5 33-64 Bt1 32.5 12.2 55.3 64-96 Bth 36.1 10.9 53.0 96-128 B’t1 37.7 11.0 51.3 128-170 B’t2 34.9 11.4 53.7

Rwegura Pasture

Depth (CM) Horizon Sand Silt Clay (%) (%) (%) 0-30 A 34.9 2.4 62.7 30-46 BA 31.3 6.0 62.7 46-80 Bs1 32.2 9.1 58.7 80-130 Bs2 32.2 3.3 64.5 130-180 Bhsl 30.5 3.8 65.7 180-210 Bhs2 32.1 4.2 63.7 i. .3 Dph distributions Depth I.site at ofeffective CEC Fig. 4.13. HORIZONS Bhs4 Bhs3 Bhs2 sl h B Bt A CCco() g1 Clay Kg"1 cmol(+) ECEC BUGARAMA FOREST BUGARAMA

5 0 5 20 15 10 5 0 UAVACLUE UAVA PASTURE MURAMVYA CULTURE MURAMVYA CCco() g Ca EE ml+ Kg' Clay '1 g K cmol(+) ECEC Clay 1 Kg’ cmol(+) ECEC 0 5 1 1 20 15 10 5 0 0 2 15 10 5 0 i. .4 Depth distributions II.site at ofeffective CEC 4.14.Fig. HORIZONS Bhs2 Bhs3 hsl B AB Bl A CCco() g1 Clay Kg-1 cmol(+) ECEC 5 0 5 20 15 10 5 0 TEZA FOREST TEZA Ap2 Ap1 CCco() g1 Clay Kg-1 cmol(+) ECEC Bt4 B!3 B11 BUKEYE CULTURE BUKEYE 5 0 5 20 15 10 5 0 Bs2 ^ O GO O - r r Bhs Bsl A CCco() g1 Clay Kg-1 cmol(+) ECEC UEE PASTURE BUKEYE 5 0 5 20 15 10 5 0 1 I 4 4 4 ► ► ► i. .5 Depth distributions III. site at ofeffective CEC Fig. 4.15. HORIZONS Bhs3 Bhs2 Bhsl AB Bl A CCco() g1 Clay Kg'1 cmol(+) ECEC WGR FOREST RWEGURA 5 0 5 20 15 10 5 0 r^j O DC O CO pl A p - Ap2 B’!2 - B'll B11 RWEGURA CULTURE RWEGURA CCco() g1 Clay Kg-1 cmol(+) ECEC 5 0 5 20 15 10 5 0 2 O C D rvj o O C Bhs2 Bhsl Bs2 sl B BA A CCco() g1 Clay Kg-1 cmol(+) ECEC PASTURE RWEGURA 5 0 5 20 15 10 5 0 110

All the soils studied exhibit podzolization features, the latter being most apparent under natural forest. All the results disclose that the sombric horizon results from podzolization processes. Some of the sombric horizons meet the spodic horizon requirements.

In a typical Spodosol profile, development begins at the top of the profile and, as is common with those of most other soils, surface horizons develop parallel to the air/solid interface and not necessarily perpendicular to the force of gravity or the direction of leaching. Comparisons of a virgin soil

(soil under natural forest) with its counterpart under pasture disclose that the main difference is the organic matter content, which is a result of differences in eluviation-illuviation processes. It is important to understand how the materials are mobilized, transported and deposited to produce Bh and Bhs horizons above the wetting front. Eluviation may be thought as resulting from the following:

1) Simple solubility product - controlled solution and precipitation apply.

If soil solutions are undersaturated with respect to the solubility product of minerals composing the soil, the minerals tend to go into solution. The dissolved ions can then move with the soil solution when it is moved by leaching. Dissolved ions or molecules (e.g. silicic acid) may also undergo diffusion, driven by concentration gradient.

2) Reduction-induced eluviation due to the presence of organic matter.

Schnitzer and DeLong (1955), in their investigation on the mobilization and 111

transport of Fe in forested soils, concluded that both the forest canopy and the

forest floor contribute solutions capable of the mobilization and transport of

iron under favorable conditions. They may transport Al and calcium (Ca) also.

It has been found th a t: a) an environment dominated by sodium (Na) ion is

more favorable than one in which Ca-ion predominates; b) an environment

dominated by Ca-ion is much more favorable than one in which H-ion is

predominant; c) the iron-retaining capacities of these solutions vary directly

as the pH of the environment within the pH range commonly encountered in

soils; and that d) the complexes formed by interaction of such solutions with

Fe separate from solution at a critical Fe content, a content which is dependent on the characteristics of the environment.

The Oa horizon of a typical Spodosol is made up of very dark-colored, greasy, humified organic matter with pH near 3, or even lower. In a Spodosol, the E horizon has been largely depleted of Al and Fe by leaching, and, as a consequence, it is enriched with Si. Any organic matter occurring in the E horizon is quite labile because of the lack of sesquioxides-based positive charge in this horizon. Phenolic acids, tannin, and other organic substances becoming soluble and mobile in the O horizons and, carrying bound Fe and

Al, readily pass through the E horizon as leaching occurs. Low in molecular weight compared with humic acids, these substances assume the characteristics of the fulvic and fraction of the spodic horizon as they accumulate in the B horizon (Blazer, et al., 1984). 112 Although these mobile organics contain some complexed biocycled Fe

and Al, most of the H-containing functional groups, largely carboxyls, are free of metals and therefore contribute to an extremely acid environment, especially in the Oa horizon. The lowered pe associated with the reactive organic surfaces, augmented by the very low pH, brings about reduction of

Fe(lll) to Fe(ll), even though the soil pores are well supplied with oxygen (02).

As the acids move through the mineral material that is in the process of becoming the E horizon, they bind additional Al and Fe by chelation, and some of the Fe is reduced by them. In chelation-induced eluviation

(cheluviation) the solubilities of certain constituents can be greatly increased by soluble organic compounds present in soil solutions beyond that predicted from simple solubility product equations. This kind of eluviation is important in podzolization to produce E horizons. Probably most of the Al and Fe picked up by the organics was already mobilized in colloidal inorganic soils before reaction with the organic groups (Childs, 1983). Protons released from carboxyls, when Al is chelated, may be consumed during Fe reduction.

Protons not consumed may amplify soil leaching by replacing base-forming cations and bring about their mobilization into the soil solution.

How dissoluted materials that are eluviated produce Bh and Bhs horizons far above the depth of wetting front is difficult to explain. Following are so me alternatives that may act individually or in combination. 113 1) llluviation of dissolved ions or molecules can occur when the

solubility products of minerals containing the ions or molecules are exceeded

because minerals should not dissolve beyond the limits imposed by the

solubility product of the particular mineral.

2) Solubility of a mineral that is more soluble than another mineral. The

less soluble mineral would tend to precipitate, lowering solution levels of the

ions so that more of the more soluble mineral could dissolve. In this

way, a kaolinite might replace a feldspar with some excess ions being leached

from the soil.

3) Precipitation of a mineral triggered by desiccation. Desiccation

would cause the ions of molecules in soil solutions to become more

concentrated such that the solubility of minerals would be exceeded.

4) Precipitation of fulvic acids may occur when:

a) Metal cation / fulvic acid (FA) molecule ratio is high. e.g. Fe3+ /FA

ratio = 6:1.

b) Precipitation caused by decomposition of the chelating agent by

microorganisms.

c) Precipitation caused by pH and Eh changes.

For example, oxidation of Fe(ll) held by organic matter lowers solubility and causes precipitation of the Fe organic complex. Increasing saturation of organics with Al occurs simultaneously as the fulvic (phenolic) acids continue to move across mineral surface and the solubilities of the complexes are 114 further lowered as the Al/organic ratios increase. The result is precipitation

and immobilization of organic matter and formation of a Spodic horizon with

the characteristic accumulation of organic C, Al, and Fe (McKeague, et al.,

1971). The thickness of the E horizon varies inversely with the amounts of

Al and/or Fe available from a given amount of parent material. It may range

from thin, in easily weathered materials, to several meters thick in quartzitic

sand (Flach, et al., 1981).

4.4.2. Evolution of the Soils Following Deforestation

The removal of such an important vegetative cover results in marked

changes in soil temperature and soil moisture regimes. The higher temperature accelerates organic matter decomposition and soil water

evaporation. The runoff/infiltration ratio increases tremendously, resulting in

land degradation by rainfall erosion. In effect, the reduction of O.M. content

increases soil erodibility. Soil erodibility is the integrated effect of processes that regulate rainfall acceptance and the resistance of the soil to particle detachment and subsequent transport. These processes are influenced by soil properties, such as particle size distribution, structural stability, O.M. content, nature of clay minerals and chemical constituents. Soil parameters that affect soil structure, slaking, and water transmission characteristics also affect soil erodibility (Lai, 1988). The unusually high aggregate stability and the low water dispersible clay that are due to Al and Fe cations that bridge clays (gibbsite and kaolinite) and humic acids impede E horizon development. 115 The marked reduction of organic chelates following deforestation results in the decrease of the reduction-induced eluviation (cheluviation) of Fe and Al, disrupting the podzolization process that was in steady-state. In fact, ions such as Fe(lll) and Al(lll) that were maintained at high concentration in solution as metal-organic complexes at pH values well above those at which their hydroxides would precipitate in aqueous systems devoid of complexing agents can no longer be maintained at such high concentration. In conclusion, the evolution of Spodosols at high altitude in Burundi following deforestation is marked by the disappearance of litter (due to mineralization) that decreases the podzolization process (due to decrease in chelating agents). The resulting increase in soil erodibility is due to both the disappearance of the land cover and the reduction of O.M. addition.

4.4.3. Classification of High Altitude Soils in Burundi

Soils under natural forest exhibit spodic-like materials underlain by an oxic horizon (Fig. 4.13, Fig. 4.14 and Fig. 4.15). Cultivated and pastured soils have an oxic horizon with spodic characteristics. The accepted genetic processes that lead to Oxisols and Spodosols are incongruent. Soil

Taxonomy does not recognize soils that present chemical, physical and mineralogical data with both oxic and spodic properties. According to the

1992 Keys to Soil Taxonomy, soils with spodic materials show evidence that organic materials and Al, with or without Fe, have been moved from an eluvial to an iiluvial horizon. This is clearly the case for the soils under natural forest. 116

The Fe and Al depth functions of all three extracts (DCB, AAO and SPP) exhibit characteristics of podzolization. The spodic-like materials underlie an umbric epipedon. Though the overlying umbric epipedon present Al + 1/2 Fe percentages (by ammonium oxalate) values that are more than half the corresponding values in the iiluvial horizon, the 0.50 value required to qualify for spodic materials is quadrupled in most cases. We, therefore, waived this criteria because all the other requirements for spodic materials are satisfied.

A typical soil under natural forest classifies as fine, oxidic, isomesic Typic

Haplorthods. On landscape positions so steep that an important part of the forest litter is subjected to removal by either runoff or lateral flow or both, the soils classify as fine, oxidic, isomesic, allic Humic Hapludox (Table 4.4).

The Oxic horizon that occurs consistently underneath the spodic materials is typical of warm and humid tropical environments. In effect, the clay content is about 60 percent and its depth function is almost constant throughout the soil profile. The clay fraction is 75 to 85 percent kaolinitic.

Gibbsite occurs throughout the soil profiles. Its amount is estimated between

12 and 18%. The remainder of the clay fraction is composed of quartz and mica. The iron oxides extractable by DCB reported as Fe percent disclose values that range from 2.0 to 3.6%. The ratios of extractable iron oxides plus gibbsite percentages to the clay percentage are more than 0.20, qualifying these soils materials as oxidic. The CEC values computed on organic mater free clay are less than 16 cmol(+) Kg"1 clay (by 1N NH4OAC pH 7) and the Table 4.4 Classification of Soil Series.

SERIES FAMILY SUBGROUP ORDER Bugarama fine Typic Spodosols Forest oxidic Haplorthods isomesic Teza fine Allic Humic Oxisols Forest oxidic Hapludox isomesic Rwegura fine Typic Spodosols Forest oxidic Haplorthods isomesic Muramvya fine Allic Humic Oxisols Pasture oxidic Sombriudox isothermic Bukeye fine Allic Humic Oxisols Pasture oxidic Hapludox isothermic Rwegura fine Allic Humic Oxisols Pasture oxidic Hapludox isothermic Muramvya fine Humic Oxisols Culture oxidic Hapludox isothermic Bukeye fine Humic Oxisols Culture oxidic Hapludox isothermic Rwegura fine Allic Humic Oxisols Culture oxidic Hapludox isothermic ECEC values are less than 12 cmol(+) Kg-1 clay (sum of bases extracted with

1N NH40AC pH 7, plus 1N KCI extractable Al) (Table 4.5, Table 4.6 and

Table 4.7). The disappearance of the forest litter along with the reduction of

O.M. supply that follow deforestation result in soils that exhibit decreased intensity of the spodic properties. Soil erosion truncates the surface layer rich in organic matter, reducing podzolization rate. The soils, nevertheless, sustain spodic characteristics that are both pedogenetic and important for land management. Cultivated soils classify as fine, oxidic, isothermic Humic

Hapludoxs. When cultivated soils are truncated by rainfall erosion or other anthropic activities, they qualify as fine, oxidic, isothermic, allic Humic

Hapludox. Pastured soils classify as fine, oxidic, isothermic, allic Humic

Hapludoxs. When the sombric horizon is prominent,they classify as

Sombriudoxs. We proposed a Spodic subgroup for the Hapludoxs. Table 4.5. CEC and OC at Site I. 119

Bugarama Forest

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-20 A 3.18 .68.1 24.8 15.5 5.6 20-44 Bhs1 2.96 57.4 35.3 16.5 10.1 44-88 Bhs2 3.06 -61.1 36.9 11.7 7.0 88-121 Bhs3 _... 2.82 49.6 33.1 9.9 6.6 121-152 Bhs4 3.08 15.5 11.8 5.9 3.6 152-196 Bt 2.44 38.4 28.1 7.0 5.12

Muramvya Culture

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-30 Ap1 2.55 30.1 12.1 6.1 2.4 30-50 Ad2 1.67 32.9 16.3 6.9 3.4 50-76 AB 1.22 22.2 12.8 5.9 3.4 76-110 Bt1 0.66 18.0 12.7 5.3 3.7 110-129 Bt2 0.51 15.6 11.7 5.5 4.1 129-141 Bt/R 0.35 14.5 11.6 6.0 4.8 141-167 B’t1 0.41 16.3 12.5 6.6 5.1 167-210 B’t2 0.18 12.3 10.8 6.6 5.9

Muramvya Pasture

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-19 A 2.82 21.3 _8.5 -10.8 4.3 19-35 AB 2.17 17.6 ... 8.0 10.1 4.6 35-65 Bs 1.58 25.4 13.6 7.8 4.2 65-94 Bhs1 1.96 30.2 14.2 7.8 3.7 94-129 Bhs2 1.43 25.4 14.0 7.6 4.2 129-179 B’s 0.66 27.6 20.0 5.4 3.9 179-210 Bt 0.47 14.2 11.19 6.4 5.1

Note: a: CEC cmol(+)Kg'1 clay; b: CEC cmol(+)Kg'1 O.M. free clay; c: ECECcmol(+)Kg-1 clay; d: ECEC cmol(+)Kg"1 O.M. free clay. 120 Table 4.6. CEC and OC at Site II.

Teza Forest

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-10 A 3.72 79.0 25.9 14.7 4.8 10-36 AB 2.17 31.3 13.5 8.0 3.4 36-74 Bhs1 1.49 31.3 16.1 10.0 5.2 74-125 Bhs2 1.14 27.5 15.8 8.0 4.6 125-168 Bhs3 1.48 33.5 17.5 6.3 3.3 168-200 Bt 0.68 24.6 16.6 8.6 5.8

Bukeye Culture

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-21 Ad1 2.41 28.4 12.3 7.2 3.1 21-46 Ad2 2.06 26.2 12.3 7.9 3.7 46-71 Bt1 1.05 21.7 13.8 8.9 5.7 71-85 Bt2 1.21 25.7 15.6 10.7 6.4 85-110 Bt3 1.29 28.9 16.8 11.0 6.4 110-159 Bt4 0.90 13.9 9.2 10.2 6.8 159-200 Bt5 0.66 13.9 10.3 9.1 6.8

Bukeye Pasture

Depth Horizon OC CECa CECb ECEC0 ECECd (CM) (%) 0-16 A 2.37 32.8 14.1 9.3 4.0 16-40 AB 1.84 22.7 11.2 10.4 5.1 40-63 Bs1 1.37 22.1 12.3 10.0 5.6 63-96 Bs2 1.16 23.2 14.1 9.7 5.9 96-138 Bhs 1.09 24.3 15.1 10.0 6.2 138-171 B’s1 0.59 18.1 13.6 8.6 6.4 171-200 B’s2 0.58 16.9 12.7 8.8 6.6

Note: a: CEC cmol(+)Kg"1 clay; b: CEC cmol(+)Kg"1 O.M. free clay; c: ECEC cmol(+)Kg clay; d: ECEC cmol(+)Kg_1 O.M. free clay. 121

Table 4.7. CEC and OC at Site III.

Rwegura Forest C O ^ o Depth Horizon CECa CECb ECECc ECECd (CM) p 0-21 A 3.23 59.5. 22.4 16.9 6.3 21-51 AB 2.89 49.5 19.7 .14.4 5.7 51-81 Bhs1 2.77 44.2 17.4 13.7 5.4 8.1-1.12 _ _ Bhs2 2.73 44.5 17.6 1.1.7 4.6 112-151 Bhs3 1.66 29.7 15.5 9.9 5.2 151-200 Bt 0.87 19.3 13.0 9.9 6.7

Rwegura Culture

Depth Horizon OC CECa CECb ECECc ECECd (CM) (%) 0-14 Ao1 1.51 36.4 17.8 8.8 4.3 < CM 14-33 Q 2.06 42.7 18.7 10.3 4.5 33-64 Bt1 0.99 28.9 17.9 9.2 5.7 64-96 Bth 0.98 24.1 14.8 10.2 6.2 96-128 B’t1 0 . 6.3 _ _ 20.7 14.5 9.2 6.4 128-170 B’t2 0.48 17.1 13.1 7.4 5.7

Rwegura Pasture

Depth 2Horizon OC CECa CECb ECECc ECECd (CM) (%) 0-30 A 2.87 37,0 14.4 .7.5 2.9 30-46 BA 2.03 28.9 13.6 9.9 4.7 46-80 Bs1 1.35 28.1 15.7 . 10.2 5.7 80-130 Bs2 . 1.35 25.3 14.7 8.7 5.0 130-180 Bhs1 2.67 39.4 .16.4 9.1 3.8 180-210 Bhs2 1.35 14.0 5.1 7.7 2.8

Note: a: CEC cmol(+)Kg'1 clay; b: CEC cmol(+)Kg"1 O.M. free clay; c: ECEC cmol(+)Kg'1 clay; d: ECEC cmol(+)Kg O.M. free clay. CHAPTER 5

SUMMARY AND CONCLUSIONS

5.1. Summary

The physical, chemical, mineralogical and morphological changes that a soil undergoes from a virgin to cultivated state is of interest to agricultural workers in understanding soil management problems. Changes that occur during a long period of cultivation may be a guide for soil scientists to understand and predict the type of problems they will face in the future to maintain soil productivity.

The study area is located in Burundi on the uplifted zone which lies along the Western Rift of the East African Rift Valleys, within a region comprised between 1,500 and 2,600 meters above sea level. The soils develop in residuum from metamorphic rocks (gneiss, schist, quartzite and their intermediate), and from intrusive igneous rocks (essentially granite) under a humid, warm tropical climate. The soil moisture regime is udic, and the soil temperature regime is isothermic or isomesic.

The main objectives of this study were: 1) to provide morphological descriptions, physical, chemical and mineralogical data of the soils, and therefore ascertain the effects of cultivation on surface soil horizons, and, if any, on lower horizons, 2) to decipher the basic processes involved in soil development, 3) to classify the soils according to Soil Taxonomy to the family level, and, 4) to provide statistical validity to selected effects of cultivation.

122 123 The effect of cultivation on deforested soils was evaluated by

comparing cultivated soils with their forested counterparts. Soils under natural

forest at high altitude in Burundi are part of a complex ecosystem that

concentrates plant and microbial nutrients in the surficial layers. Despite the

luxurious biomass covering the soils, they are poor, highly weathered acid soils. The luxurious vegetation is sustained by the efficient cycling of organic matter in the soils that results in the overall improvement in soil conditions, the development and maintenance of soil structure, the improvement of physical properties, the decreased susceptibility to erosion, the encouragement of microbial activity and the provision of potentially available plant nutrients. The main process of soil formation under natural forest is podzolization. As a result, Haplorthods dominated by kaolinite and gibbsite are the common soils identified within this ecosystem.

The recent increasing population (the population at the study area has doubled during the last three decades) has exerted such a tremendous pressure on the forest that most of the area has been deforested and cultivated. Most of the steep slopes (up to 85%) have been cleared of the natural forest and used for cropping where they should have been carefully protected under forest. Both excessive human and animal populations (cattle, sheep and goat) contributed to the destruction of the forest lands. After deforestation, the litter disappears due to natural and/or artificial mineralization or to erosion. When the land is covered by vegetation, a large portion of the 124 raindrop is intercepted by the plants and yields most of its kinetic energy

before reaching the soil surface. This diversion of rainfall energy is

responsible in a very large measure for the remarkable reduction of erosion

by close-growing vegetation. When deforested, land degradation has been

observed within a few years. In effect, where the forest is cleared, rainfall

erosion constitutes the most dramatic process of land deterioration. Sheet

erosion and gullies have rendered some of the soils almost useless for

cropping. Some soils erode to the bedrock, hardpan, or less fertile subsoils.

The soils continue to erode as they are overgrazed and overexploited.

Another important process of soil formation under permanent pasture and somewhat under cultivation is desilication. The ultimate product of this process is lateritization, which constitutes an important threat in the study area. Most of the soils under pasture are classified as Hapludox.

Farmers in traditional systems of agriculture know that long-term productivity of the soil may decline as cropping systems become more intensive unless efforts are made to maintain soil fertility. Organic manures have been the major input used to maintain soil fertility, partially because they are readily available in many traditional systems of agriculture. Also, organic manures have the potential to correct several soil problems. They may supply organic matter, create favorable air and water regimes around the plant roots and act as carriers of some micronutrients. Whereas deterioration of cultivated soil without use of either manure or organic residue from harvesting 125 occurs within a few years, cultivation supported by adequate amounts of organic matter (30 t/ha/year) have shown sustainable agriculture and increase in productivity. In effect, improved morphological, physical, and chemical properties brought about by cultivation have been identified by comparison of a cultivated soil with its counterparts under forest and under permanent pasture. Because of the relatively short period of time under which the soils can be compared (five to 30 years), the properties that have thus far been taken into consideration are those that change rapidly. These include soil structure, base saturation, soil acidity, organic matter content, C.E.C., translocation of fine material and aluminum toxicity. The soils evolve from

Haplorthods to Hapludox within 20 years.

5.2. Conclusions

At the studied area, a typical soil under natural forest classifies as fine, oxidic, isomesic, Typic Haplorthods, cultivated soils classify as fine, oxidic, isothermic Humic Hapludoxs. When cultivated soils are truncated by rainfall erosion or other anthropic activities, they qualify as allic (fine, oxidic, isothermic, allic Humic Hapludoxs). Pastured soils classify as fine, oxidic, isothermic, allic Humic Hapludoxs. When the sombric horizon is prominent, they classify as Sombriudoxs. We proposed a Spodic subgroup for the

Hapludoxs.

An overall evaluation of the soils’properties revealed that the studied area was highly weathered prior to faulting and uplifting and the spodic 126 properties displayed by the soils formed during the cool and humid parts of the Pleistocene Epoch. The luxurious forest that emerged under high rainfall constituted an ideal ecosystem for podzolization processes. The rift valley that erupted during the late Cenozoic Era, probably during Pliocene Epoch, fractured Precambrian-aged rocks (Cryptozoic Eon) that had most likely weathered to oxidic materials.

To evaluate the effects of long-term cultivation and pasture on deforested soils at high altitude in Burundi, comparisons of morphological, physical, chemical and mineralogical properties of soils under natural forest with their counterparts under cultivation and pasture were performed, we concluded that 18 out of 32 compared parameters, namely: OC, Bd, water retention difference, water dispersible clay, N, Bray I P, Ca, Mg, K, BaCI2 acidity, extractable acidity, extractable Al, extractable H, NH4OAc CEC, Sum

CEC, NH4OAc BS, Sum BS, and Al saturation showed significant treatment effects. Seven of these parameters were significantly different throughout the soil profiles. In other words, they affected both surface and subsurface horizons. They were OC, Bd, water retention, water dispersible clay, BaCI2 acidity, extractable H and Sum CEC. Ten of them, namely: N, C/N ratios,

Bray I P, Ca, Mg, K, extractable acidity, extractable Al, NH4OAc CEC, and Al saturation exerted a significant treatment effect that was limited to the upper horizons of the soil profiles. Some of these parameters remained significantly different within the upper 30 cm of the soil profiles. Others persisted to the 127 100 cm depth. On fourteen parameters, namely: Bray II P, ECEC, pH water, pH KCI, pH NaF, C/N ratios, aggregate stability, Fed, Ald, AIQ, Fe0, Fepl Alp and Al0+1/2 Fe0, there was not enough evidence to support the existance of any treatment effect. At the end of different treatment effects comparisons, it was concluded that there should be important improvement of pastured soils when they are cultivated with continuous addition of enough amounts of O.M.

(30 tons/ha/yr). The most important morphological improvement brought about by cultivation is the development of a subangular blocky structure that improve soil aeration and water infiltration.

Soil mineral composition disclosed the predominance of kaolinite (about

80% or more) throughout the soil profiles. The remaining clay minerals were identified and estimated as gibbsite > mica > quartz. Mineralogical comparisons among soils under natural forest and their counterparts under pasture and cultivation did not detect any difference.

The studied soils are rich in organic matter, gibbsite and iron oxides.

All three components are well known for their binding effects. The soils exhibit steep slope and rainfall erosion constitutes a serious limitation to land use. Removal of iron oxides and organic matter resulted in increased clay dispersion. This increase, however, was far from total dispersion.

Aggregation at the study area must, therefore, involve Al, besides Fe oxides and organic matter. Future research should investigate the effect of gibbsite, humic and fulvic acids of these soils on the aggregation phenomenon. Addition of organic matter to pastured soils tremendously improves their productivity though the same pastured soils are already rich in organic matter.

The organic matter present in pastured soils is highly resistant to mineralization, water insoluble and consequently does not contribute to the nutrient supply for plants. Characterization of the humic substances should elucidate the mechanisms involved in organo-mineral associations.

Identification of the main organo-mineral components of these soils along with their behavior should improve the soils’ management. The studied area is a part of a wide area that extends to western Uganda and eastern Zaire.

Similar soil properties should be expected within a similar ecological zone and similar land management should be applied. LITERATURE CITED

Andriesse, J.P. 1979. A study into the mobility of iron podzolic Sararrak upland soils by means of selective iron extractions. Neth. J. Agric. Sci. 27:1-12.

Area, M. N., and S. B. Weed. 1966. Soil aggregation and porosity in relation to contents of iron oxide and clay. Soil Sci. 101:164-170.

Barnhisel, R. I., and P. M. Bertsch. 1989. Chlorites and Hydroxy-Interlayered Vermiculite and Smectite, in: J. B. Dixon and S. B. Weed (Editors). Minerals in Soils Environments. Soil Sci. Soc. Am., Madison, Wl pp.729-788.

Bascomb, C. L. 1968. Distribution of pyrophosphate - extractable iron and organic carbon in soils of various groups. J. Soil Sci. 19:251-268.

Bauer, A., and A. L. Black. 1981. Soil carbon, nitrogen, and bulk density comparisons in two cropland tillage systems after 25 years and in virgin grassland. Soil Sci. Soc. Am. J. 45:1166-1170

Baver, L. D. 1937. The relation of vegetation cover to the physical properties of soils, in symposium on "some aspects of vegetative methods of cf erosion control". Annual Meeting, Amer. Soc. Agron., 1937.

Bertramson, B. R., and H. F. Rhoades. 1938. The effects of cropping and manure applications on some physical properties of a heavy soil in eastern Nebraska. Soil Sci. Soc. Am. Proc. 3:32-36.

Bigham, J. M., D. C. Golden, L. H. Bowen, S. W. Buol, and S. B. weed. 1978. Iron oxide mineralogy of well-drained Ultisols and Oxisols. I. Characterization of iron oxides in soil clays by Mossbauer spectroscopy, x-ray diffractometry, and selected chemical techniques. Soil Sci. Soc. Am. J. 42:816-825.

Bigham, J. M., D. C. Golden, L. H. Bowen, S. W. Buol, and S. B. Weed. 1978. Mossbauer and x-ray evidence for the pedogenic transformation of hematite to goetite. Soil Sci. Soc. Am. J. 42:979-981.

Blake, G. R. 1965. Bulk density. jn "Methods of soil analysis," pp. 374-390. Am. Soc. of Agron., Madison, Wisconsin.

Blakemore, A.V. 1973. Aggregation of clay by the products of iron (III) hydrolysis. Aust. J. Soil Res. 11:75-82.

129 130

Blank, R. R., and M. A. Fosberg. 1989. Cultivated and adjacent virgin soils in north central South Dakota. I. Physical and Chemical Comparison. Soil Sci. Soc. Am. J. 53:1484-1490.

Blume, H.P., and U. Schwertmann. 1969. Genetic evaluation of profile distribution of aluminum, iron, and manganese oxides. Soil Sci. Soc. Am. Proc. 33:438-444.

Bohn, H. I., B. L. McNeal, and G. A. O’Connor. 1979. Soil chemistry. John Wiley & Sons, New York.

Borggaard, O.K. 1983. Iron oxides in relation to aggregation of soil particles. Acta Agri. Scand. 23:257-260.

Bouma, J., and F. D. Hole. 1971. Soil structure and hydraulic conductivity of adjacent virgin and cultivated pedons at two sites: A Typic Argiudoll (silt loam) and a Typic Eutrochrept (clay). Soil. Sci. Soc. Am. Proc. 35:316- 318.

Bouma, J., D. J. Van Rooyen, and F. D. Hole. 1974. Estimation of comparative water transmission in two pairs of adjacent virgin and cultivated pedons in Wisconsin. Geoderma 13:73-88.

Bradfield, R. 1937. Soil conservation from the viewpoint of soil physics., Jour. Amer. Soc. Agron. 29:85-92.

Bray, R. H., and L. T. Kurtz. 1945. Determination of total organic, and available forms of phosphorus in soils. Soil Sci. 59:39-45.

Brown, L. R. 1981. World population growth. Soil erosion and food security. Science 214:995-1002.

Browning, G. M. 1937. Changes in the erodibility of soils brought about by the application of organic matter. Soil Sci. Soc. Am. Proc. 11:85-86.

Bryant, J. P., and J. B. Dixon. 1964. Clay mineralogy and weathering of a Red-Yellow Podzolic soil from quartz mica schist in the Alabama Piedmont. Clays Clay Miner. 2:509-519.

Buol, S. W., F. D. Hole, and R. J. McCracken. 1980. Soil genesis and classification. Iowa State University Press. Ames, IA.

Cady, J. G. 1951. Rock weathering and soil formations in the North Carolina Piedmont region. Soil Sci. Soc. Am. Proc. 15:337-342. 131 Calvert, C. S., S. W. Buol, and S. B. Weed. 1980. Mineralogical characteristics and transformations of a vertical rock-saprolite-soil sequence in the North Carolina Piedmont. I. Profile morphology, chemical composition, and mineralogy. Soil Sci. Soc. Am. J. 11:1096- 1103.

Calvert, C. S. S. W. Buol, and S. B. Weed. 1980. Mineralogical characteristics and transformations of a vertical rock-saprolite-soil sequence in the Carolina Piedmont. II. Feldspar alteration products - -Their transformations through the profile. Soil Sci. Soc. Am. J. 44:1104-1112.

Campbell, C. A., and W. Souster. 1982. Loss of organic matter and potentially mineralizable nitrogen from Saskatchewan due to cropping. Can. J. Soil. Sci. 62:651-656.

Chesworth, W. 1973. The residua system of chemical weathering: A model for the chemical breakdown of silicate rocks at the surface of the earth. J. Soil Sci. 24:69-81

Chesworth, W. 1980. The haplosoil system. Am. J. Sci., 280:969- 985.Childs, C.W., R.L. Parfitt, and R. Lee. 1983. Movement of aluminum as an inorganic complex in some podzolized soils, New Zealand, Geoderma. 29:139-155.

Coate, D. R., and J. F. Ramsey. 1983. Quantification of the effects of over 35 years of intensive cultivation on four soils. Can. J. Soil Sci. 63:1-14.

Cook, L. H. 1936. The nature and controlling variables of the water erosion process. Soil Sci. Soc. Proc. 1.

Cook, M. G. 1973. Compositional variations in Typic Hapludults containing mica. Soil Sci. 115:159-169.

Day, P. R. 1965. Particle fractionation and particle-size analysis. Jn C. A. Black (ed.) Methods of soil analysis. Part 1. Agronomy 9:545-577. Am. Soc. Agron., Madison, Wl.

Deshpande, T. L., D.J. Greenland, and J. P. Quirk. 1968. Changes in soil properties associated with the removal of iron and aluminum oxides. J. Soil Sci. 19:108-122. 132

Dormaar, J. F. 1979. Organic matter characteristics of undisturbed and cultivated chernozemic and solonetzic A horizons. Can. J. Soil. Sci. 59:349-356.

Doughty, J.L., F. D. Cook, and F. G. Warder. 1954. Effects of cultivation on the organic matter and nitrogen of Brown soils. Can. J. Soil Sci. 34:406-411.

Edwards, A. P., and J. M. Bremner. 1967. Microaggregates in soils. J. Soil Sci. 18:64-73.

Edwards, C. A., and J. F. Lofty. 1975. The influence of cultivation on animal populations, p. 399-408. ]n J. Vanek (ed.) Progress in soil zoology. Academia Publ., Prague. Czechoslovakia.

El Swaify, S. A., and W. W. Emmerson. 1975. Changes in the physical properties of soil clays due to precipitated aluminum and iron hydroxides: I. Swelling and aggregate stability after drying. Soil Sci. Soc. Am. Proc. 39:1056-1063.

Elwell, H. A. and M. A. Stocking. 1982. Developing a simple yet practical method of soil loss estimation. Trop. Agric. 59:43-48.

Emerson, W. W. 1959. The structure of soil crumbs. J. Soil Sci. 10:235-244.

Eswaran, H., and Wong Chaw Bin. 1978. A study of a deep weathering profile on granite in peninsular Malaysia. I. Physico-chemical and micromorphological properties. Soil Sci. Soc. Am. J. 42:144-149.

Fanning, D. S. and M. C. B. Fanning. 1989. Soil Morphology, genesis, and classification. John Wiley and Sons. New York, N.Y.

Finn, D. 1983. Land use and abuse in the East African region. Ambio 12:296-301.

Fordham, A. W., and K. Norrish. 1979. Electron microprobe and electron- microscope studies of soil clay particles. Aust. J. Soil Res. 17:283- 306.

Fordham, A.W., and K. Norrish. 1983. The nature of soil particles particularly those reacting with arsenate in a series of chemically treated samples. Aust. J. Soil Res. 21:455-477. 133

Frenkel, H., and I. Shainberg. 1980. The effect of hydroxy-AI and hydroxy-Fe polymers on montmorillonite particle size. Soil Sci. Soc. am. J. 44:626-629.

Gaur, A. C., 1984. Response of rice to organic matter: the Indian experience, p. 503-514. in organic matter and rice. International Rice Research Institute. Manilla, Philippines.

Gilkes, R. J., G.Scholz, and G. M. Dimmock. 1973. Lateritic deep weathering of granite. J. Soil Sci. 24:523-536.

Gjems, O. 1970. Mineralogical composition and pedogenic weathering ofthe clay fraction in Podzol soil profiles in Zalesine, Yugoslavia. Soil Sci. Sci. 110:237-243.

Goudie, A. 1981. The human impact: Man’s role in environmental change. MIT Press. Cambridge, Massachusetts.

Greenland, D. J., G. R. Lindstrom, and J. P. Quirk. 1962. Organic materials which stabilize natural soil aggregates. Soil Sci. Soc. Am. Proc. 26:366-371.

Greenland, D. J., J. M. Oades, and T. W. Sherwin. 1968. Electron- microscope observations of iron oxides in some red soils. J. Soil Sci. 19:123-126.

Greenland, D. J. 1971. Interaction between humic and fulvic acids and clays. Soil Sci. Sci. 111:34-41.

Greenland , D. J., and C. J. B. Mott. 1978. Surfaces of soil particles, p. 321- 354. jn D. J. Greenland and M. H. B. Hayes (ed.) The Chemistry of Soil Constituents. John Wiley & sons. Ltd Chichester, UK.

Grossman, R. B., B. R. Brasher, D. P. Franzmeier, and J. L. Walker. 1968. Linear extensibility as calculated from natural clod bulk density measurements. Soil Sci. Soc. Am. Proc. 32:570-573.

Hardy, F. 1933. Cultivation properties of tropical red soils. Emp. J. Exp. Agric. 1:103-112.

Harris, S. 1971. Index of structure: Evaluation of a modified method of determining aggregate stability. Geoderma 6:155-162. 134 Harris, R. F., C.R. Chester, and O. N. Allen. 1965. Dynamics of soil aggregation. Adv. Agron. 18:107-160.

Hesse, P. R. 1984. Potential of organic materials for soil improvement, p.35- 43. In Organic Matter and Rice. International Rice Research Institute. Manila, Philippines.

Jeanroy, E., and B. Guillet. 1981. The occurrence of suspended ferruginous particles in pyrophosphate extracts of soil. Geoderma. 26:95-105.

Jenny, H. 1941. Factors of soil formation. McGraw-HillBook Co.,Inc. New York, N.Y. 218 pp.

Jones, R. C., W. H. Hudnall, and W. S. Sakai. 1982. Some highly weathered soils of Puerto Rico. 2. Mineralogy. Geoderma 27:75-137.

Karathanasis, A. D., F. Adams, and B. F. Hajek. 1983. Stability relationships in kaolinite , gibbsite, and Al-hydroxy interlayered vermiculite soil systems. Soil Sci. Soc. Am. J. 47:1247-1251.

Kemper, W.D.,and W. S. Chepil. 1965. Size distribution of aggregates, in C.A. Black et al. (ed.) Methods of soil analysis, Part I. Agronomy 9:499-510.

Kodama, H., and J. E. Brydon. 1968. A study of the clay minerals in podzol soils in New Brunswich, Eastern Canada. Clay Miner. 7:295-309.

Kodama, H. 1979. Clay minerals in Canadian soils: Their origin, distribution, and alteration. Can. J. Soil Sci. 59:37-58.

Lai, R. 1976. Soil erosion problems on an Alfisol in Western Nigeria and their control. Monogr. No. 1 IITA. Ibadan, Nigeria.

Lai, R. 1984. Soil erosion from tropical arable lands and its control. Adv. Agron. 37:183-248.

Lai, R., and D. J. Cummings. 1979. Clearing a tropical forest. Effects on soil and microclimate. Field Crops Res. 2:91-107.

Lai, R. 1988. Soil erosion research methods. Library of Congress Catalog No.88-11381. 244pp. 135

Larson, W. E., and C. E. Clapp. 1984. Effects of organic matter on soil physical properties, p.363-385. ]n Organic matter and rice. International Rice Research Institute. Manilla,Philippines.

Laws, W. D., and D. D. Evans. 1949. The effects of long-time cultivation on some physical and chemical properties of two Rendzina soils. Soil Sci. Soc. Am. Proc. 14:15-19.

Levin, H. R. 1983. The earth through time. CBS College Publishing. Philadelphia, PA.

Lipman, J. G. 1912. The associative growth of legumes and non-legumes. N.J. Expt. Sta. Bull. 253. Rutgers Univ. New Brunswick, NJ.

Loi, K.S. 1980. Genesis and classification of selected soils intermediate between the red-yellow podzolic and lateritic group in Sarawak Malaysia. M. Sc. thesis, Land Resource Science, University of Guelph, Guelph, Ontario. Canada.

Lyon, T. L., and J. A. Bizzell. 1911. A heretofore unnoted benefit from the growth of legumes, N.Y. Agr. Expt. Sta. Bull. 294. Cornell Univ. Ithaca, NY.

Martel, Y. A., and E. A. Paul. 1974. Effects of cultivation on the organic matter of grassland soils as determined by fractionation and radiocarbon dating. Can. J. Soil Sci. 54:419-426.

Mathieu, C. 1985. Personal communication. Univ. of Burundi. Bujumbura, Burundi.

McIntyre, D. S. 1956. The effect of free ferric oxide on the structure of some Terra Rossa and Rendzina soils. J.Soil Sci. 7:302-306.

McKeague, J. A., and J. H. Day. 1966. Dithionite and oxalate extractable Fe and Al as aids in differentiating various classes of soil. Can. J. Soil Sci. 46:13-22.

McKeague, J.A., J.E. Brydon, and N.M. Miles. 1971. Differentiation of forms of extractable iron and aluminum in soils. Soil Sci. Am. Proc., 35:33- 38.

Mehlich, A. 1938. Use of triethanolamine acetate-barium hydroxide buffer for the determination of some base exchange properties and lime requirement of soil. Soil Sci. Soc. Am. Proc. 3:162-166. 136

Mehra, O. P., and M. L. Jackson. 1960. Iron oxide removal from soils and clays by a difhionite-citrate system buffered with sodium bicarbonate, p. 317-342. in Ada Swinefored (ed) Clays Clay Miner., Proc. Conf. (7th, Washington, D. C., 1958). Pergamon Press, Elmsford, NY.

Melfi, A.J., C. C. Cerri, B. I. Kronberg, W. S. Fyle, and B. McKinnon. 1983. Granitic weathering: A Brazilian study. J. Soil Sci. 34:841-851.

Mutwewingabo, B. 1990. Genese caracteristiques et contraintes de I’amenagement des soils acides a horizon sombre de profondeur de la region de haute altitude du Rwanda. Soils Africains-African Soils. Vol xx111.10-42.

Newton, J. D., F. A. Wyatt, and A. L. Brown. 1945. Effects of cultivation and cropping on the chemical composition of some western Canada prairie province soils. Part III. Sci. Agric. (Ottawa) 25:718-737.

Norrish, K., and R. M. Taylor. 1961. The isomorphous replacement of iron by aluminum in soil goethites. J. Soil Sci. 12:294-306.

Norrish, K., and J. G. Pickering. 1983. Clay minerals, p. 281-308. jn. Soils: An Australian viewpoint, Division of Soils, CSIRO. Melbourne. Academic Press Ltd., London, UK.

Oades, J.M. 1981. Organic matter in the Urrbrae soil. p.63-82. In J.M. Oades, D.G. Lewis, and K. Norrish (ed.) Red-brown earths of Australia. Waite Agric. Res. Inst. Univ. Adelaide and Div. Soils, CSIRO, Adelaide, Australia.

Oades, J. M. 1984. Soil organic matter and structural stability, mechanisms and implications for management. Plant Soil 76:319-337.

Okigbo, B. N. 1978. "Cropping Systems and related research in Africa" (Occasional Publications series. OT-1, Association for the Advancement of Agricultural Sciences in Africa, Addis Ababa, Ethiopia.

Page, J. B., and C. J. Willard. 1946. Cropping Systems and soil properties. Soil Sci. Soc. Am. Proc. 2:81-88.

Peech, M., L. T. Alexander, L. A. Dean, and J. F. Reed. 1947. Methods of soil analysis for soil fertility. USDA Circ. 757. U. S. Govt. Print. Office, Washington, DC. 137

Pereira, H. C., P. H. Hosegood, and D. B. Thomas. 1961. The productivity of tropical semiarid thorn-scrub country under intensive management. Emp. J. Exp. Agric. 29:269-286.

Rapp, A. L., L. Berry, and P. Temple. 1972. Erosion and sedimentation in Fanzanna. Geogr. Annu. 54A. Nos. 3 and 4.

Rebertus, R. A., S. B. Weed, and S. W. Buol. 1986. Transformations of biotite to kaolinite during saprolite weathering. Soil Sci. Soc. Am. J. 50:810-819.

Richards, L. A. 1954. Diagnosis and improvement of saline and alkali soils. U. S. Salinity Laboratory, U. S. Dept. Agr. Hbk. 60, 160pp.

Roose, E. J. 1977. Adaptation of soil conservation techniques to the ecological and socioeconomic conditions of West Africa. Agron. Trop. 32:132-140.

Ross, G. J., and M. M. Mortland. 1966. A soil beidellite. Soil Soc. Am. Proc. 30:337-343.

Russel, M. B., A. Klute, and W. C. Jacob. 1952. Further studies on the effect of long-time organic matter additions on the physical properties of sassafras silt loam. Soil Sci. Soc. Am. Proc. 16:156-159.

Russel, E. W. 1973. Soil conditions and plant growth. 10th ed. Longman, London, 849 p.

Sanchez, P. A., 1976. Properties and Management of soils in the Tropics. Wiley, New York, NY.

Sanchez, P. A., and S. W. Buol. 1975. Soils of the tropics and the world food crisis. Science 188:598-603.

Sanchez, P. A., D. A. Bandy, J. H. Villachila, and J. J. Nicholaides. 1982. Amazon Basin soils: management for continuous crop production. Science 216:821-827.

Schlichting, E., and H.P. Blume. 1966. Bodenkundliches Praktikum. Verlag P. Parey Hamburg, Germany.

Schnitzer, M., and W.A. DeLong. 1955. Investigations on the mobilization and transport of iron in forested soils: II. The nature of the reaction of leaf extracts and leachates with iron. Soil Sci. Soc. Am. Proc. 19:363-368. 138 Schnitzer, M., and S. T. M. Skinner. 1963. Organic-metallic interactions in soils, 2. Reactions between different forms of iron and aluminum and the organic matter of a podzol Bh horizon. Soil Sci. 96:187-189.

Schnitzer, M. 1978. Humic substances, chemistry and reactions, p. 1-64. In M. Schnitzer and S. U. Khan (Editors), Soil Organic Matter. Elsevier, Amsterdam, Netherlands, pp. 1-64.

Senkayi, A. L., J. B. Dixon, L. R. Hossner, B. P. K. Yerma, and L. P. Wilding. 1985. Replacement of quartz by opaline silica during weathering of petrified wood. Clays Miner. 33:525-531.

Sidhu, P. S., and R. J. Gilkes. 1977. Mineralogy of soils developed on alluvium in the Indogangetic Plain (India). Soil Sci. soc. Am. J. 41:1194-1201.

Schwertmann, U. 1966. Inhibitory effect of soil organic matter on the crystallization of amorphous ferric hydroxide. Nature 212:645-646.

Schwertmann, U., H. Kodama, and W.R. Fischer. 1968. The influence of organic compounds on the formation of iron oxides. Int. Congr. Soil Sci. Trans. 9th (Adelaide, Aust.) Vol 1:645-655.

Schwertmann, U., and N. Kampf. 1985. Properties of goethite and hematite in kaolinitic soils of Southern Brazil. Soil Sci. 139:344-350.

Schwertmann,U., H. Kodama, and W.R. Fischer. 1986. Mutual interactions between organic and iron oxides, p. 223-250. in P.M. Huang and M. Schintzer (ed.) Interactions of soil minerals with natural organics and microbes. Spec. Pub. 17. SSSA. Madison, Wl.

Sheldrick, B.H. and J.A. McKeague. 1975. A comparison of extractable Fe and Al data using methods followed in the U.S.A. and Canada. Can. J. Soil Sci. 55:77-78.

Soil Survey Staff. 1972. Soil survey laboratory methods and procedures for collecting soil samples. U. S. Dept. Agr. Soil Survey Investigation Rep. No. 1. U. S. Govt. Printing Office, Washington, DC.

Soil Survey Staff. 1984. Procedures for collecting soil samples and methods of analysis for soil survey. Soil Survey Investigations Report No. 1 (revised) U.S. Dept. Agric., USDA- SCS, U.S.Govt. Printing Office, Washington, D.C. 139 Soil Survey Staff. 1990. Keys to Soil Taxonomy, 4th edition. SMSS technical monograph no. 6. Blacksburg, VA.

Soil Survey Staff. 1992. Keys to Soil Taxonomy, 5th edition. SMSS technical monograph No. 19. Blacksburg, Virginia: Pocahontas Press, Inc. 556 pages.

Sohne, W. H. 1958. Fundamentals of pressure distribution and soil compaction under tractor tires. Agric. Eng. 39:276-281.

Stanley, S. M. 1986. Earth and life through time. W. H. Freeman and Company, New York, N.Y.

Stobbe, P. C., and J. R. Wright. 1959. Modern concepts of the genesis of Podzols. Soil Sci. Soc. Am. Proc. 23:161-164.

Tardy, Y., G. Bocquier, H. Paquet, and G. Millot. 1973. Formation of clay from granite and its distribution in relation to climate and topography. Geoderma 10:271-284.

Theng, B.K.G. 1976. Interactions between montmorillonite and fulvic acid. Geoderma 15:243-251.

Theng, B.K.G. and H.W. Scharpenseel. 1975. The adsorption of R e ­ labelled humic acid by montmorillonite. Proc. Int. Clay Conf. 1975. pp 643-653.

Uhland, R.E. 1949. Physical properties of soils as modified by crops and management. Soil Sci. Soc. Proc. 14:361-366.

Uhland, R.E. 1950. Rapid method for determining soil moisture. Soil Sci. Soc. Proc. 15:391-393.

UNEP. 1984. General Assessment of Progress in the implementation of the plan of action to combat desertification. 1978-84. Mimeo. Governing Council. 12th session. Nairobi, Kenya.

Walkley, A. 1935. An examination of methods for determining organic carbon and nitrogen in soils. J. Agr. Sci. 25:598-609.

Wilke, B. M., and U. Schwertmann. 1977. Gibbsite and halloysite decomposition in strongly acid podzolic soils developed from granitic saprolite of the Bayerischer Wald. Geoderma 19:51-61. 140 Wilkinson, G. R. 1975. Effects of grass fallow rotations in Northen Nigeria. Trap. Agric. 59:97-103.

Wilson, P. W., and O. Wyss. 1937. Mixed cropping and the excretion of nitrogen by leguminous plants. Soil Sci. Soc. Am. Proc. 2:289-296.

Young, R.A., J. C. Zubriski, and E. B. Norum. 1960. Influence of long-time fertility management practices on chemical and physical properties of a Fargo clay. Soil Sci. Soc. Am. Proc. 24-128.

Yuan, T.L. 1959. Determination of exchangeable hydrogen in soils by titration method. Soil Sci. 88:164-167. APPENDIX A

FIELD DESCRIPTIONS 142 Bugarama Forest

Location: South-west of Teza, Burundi. 3.3 Km from Bugarama roads’ junction on National Road # 1; 1.2 Km on the old road toward Teza, 80 m straight on pathway and 10 m norh side of the pathway. X: 29° 33’ E; Y: 3° 15’ 34"; altitude: 2242 m. Muramvya SA - 35 - XXIV - 2c topographic map, IGEBU, 1983. 1:50,000.

Physiographic position: Extremely dissected mountains of the east horst of the East African rift valleys.

Parent material: Mica schist underlain by granite rock.

Topography: 18 percent south-east facing convex slope on upper tier of backslope position.

Vegetation: Mountainous secondary subtropical forest.

O -- 10 to 0 centimeters; primarily made of fibrist material (leaves, branches and roots); black 60% (10YR 2/1) and dark brown 40% (7.5YR 4/4) moist; many fine and few coarse roots throughtout the horizon; nonsticky, nonplastic, weakly cemented; extremely acid; abrupt wavy boundary.

A -- 0 to 20 centimeters; black (10YR 2/1) silty clay; weak fine subangular blocky structure parting to weak fine granular structure; very friable; few fine and medium tubular pores; many fine, few medium and coarse roots throughtout the horizon; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bhs1 -- 20 to 44 centimeters; black (10YR 2/1) silty clay; massive structure and weak medium subangular blocky structure; very friable; common vesicular and tubular fine parting to medium pores; common fine parting to medium roots throughtout the horizon; slightly sticky, plastic,weakly cemented; extremely acid; gradual wavy boundary.

Bhs2 -- 44 to 88 centimeters; black (2.5YR 2.5/0) silt clay; sombric horizon; massive structure and weak medium granular structure; friable; few to common vesicular and tubular fine parting to medium pores; common medium and coarse roots; slightly sticky,nonplastic, weakly cemented; few dark reddish brown mottles (2.5YR 3/4) possibly resulting from decomposed roots; extremely acid; gradual wavy boundary. 143 Bhs3 - 88 to 121 centimeters; transitional between Bsh2 and Bsh4; black (5YR 2.5/1) silty clay; massive structure and weak fine and medium subangular blocky structure; friable; common medium tubular pores; common fine and medium roots; slightly sticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bhs4 - 121 to 152 centimeters; dark reddish brown (5YR 3/2) silt clay; massive structure and weak fine to medium subangular blocky structure; friable; few to common fine vesicular and tubular pores; few medium and coarse roots; slightly sticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bt -- 152 to 196 centimeters; reddish brown (5YR 4/3) silty clay; moderate medium to coarse subangular blocky structure; firm; common fine to coarse continuous tubular pores; few fine and coarse roots; slightly sticky, plastic, weakly cemented; presence of some foliaceous micas in the saprolitic red material (2.5YR 4/6); extremely acid; abrupt wavy boundary.

C - 196 centimeters and more; saprolitic red material (2.5YR 4/6) resulting from strong weathering of mica schist and quartzite. The geologic map discloses that granite underlies the residual acid material. This horizon has not been sampled.

Note: The soil temperature is 13 Degrees Centigrades. 144 Muramvya Culture

Location: Muramvya, Burundi. 1Km north of the Roman Catholic church. Approximately 800 m from the junction of roads, south-east of the church, five meters west of the road. X: 29° 37’ 8"; Y: 3° 14’ 52”; altitude: 2009 m. Bukeye SA - 35 - XXIV - 2a topographic map, IGEBU, 1983; 1:50,000.

Physiographic position: dissected mountains.

Topography: 2 percent north-east facing convex slope on a wavy summit position.

Parent material: Mica schist underlain by granite rock.

Vegetation: banana plantation that has been receiving high amounts of domestic organic residues and manure.

Ap1 -- 0 to 30 centimeters; dark reddish brown (5YR 2.75/2) clay (dry color); weak fine granular structure; friable; common medium and coarse tubular pores; common fine, medium and coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse smooth boundary.

Ap2 - 30 to 50 centimeters; dark reddish brown (5YR 3/2) clay (dry color); moderate fine and medium subangular blocky structure; firm; common fine and medium tubular pores; common fine, medium and coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

AB -- 50 to 76 centimeters; dark reddish brown (5YR 3.5/4) clay; moderate fine and medium subangular blocky structure; firm; common medium and coarse tubular pores; common fine and medium roots, few coarse roots; common distinct continuous dark reddish brown (5YR 3/3) clay films on most faces of peds, in pores and in root channels; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

Bt1 -- 76 to 110 centimeters; reddish brown (5YR 4/4) clay; moderate medium subangular blocky structure parting to moderate coarse subangular blocky structure; firm; common medium and coarse tubular pores; common fine, medium and coarse roots; common distinct continuous reddish brown (5YR 4/3) clay films on most faces of peds; nonsticky, nonplastic, weakly cemented; strongly acid; diffuse wavy boundary. 145 Bt2 --110 to 129 centimeters; reddish brown (5YR 4/5) clay; moderate medium subangular blocky structure parting to moderate coarse subangular blocky structure; firm; many fine and medium tubular pores; few fine and medium roots; few faint discontinuous reddish brown (5YR 4/4) clay films on vertical faces of peds; nonsticky, nonplastic, weakly cemented; strong termite activity; very strongly acid; diffuse smooth boundary.

Bt/R -- 129 to 141 centimeters; dark red (2.5YR 3/6) clay; stoneline; 30% of subangular medium and coarse fragments of quartzite (2.5 to 5 cm of diameter); moderate fine and medium subangular blocky structure; friable; common medium and coarse tubular pores; few fine and coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

B’t1 --141 to 167 centimeters; red (2.5YR 4/6) clay; moderate medium subangular blocky structure; firm; common medium and coarse tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; strong termite activity; very strongly acid; diffuse smooth boundary.

B’t2 -- 167 to 210 centimeters; red (2.5YR 4/6) clay; less than 10% of yellowish red (5YR 5/7) weatherable minerals; moderate medium subangular blocky structure; firm; common medium and coarse tubular pores; few fine roots; noticeable termite’s channels; nonsticky, nonplastic, very strongly acid; weakly cemented. Limit of description.

Note: The soil temperature is 20 Degrees Centigrades. 146

Muramvya Pasture

Location: Muramvya, Burundi. Approximately 800 m north of the Roman Catholic church, 600 m from the junction of roads, south-east of the church, 250 m west of the road. X: 29° 37’ 7" E; Y: 3° 15’ 9"; altitude: 1945 m. Muramvya SA -35 - XXIV - 2c topographic map, IGEBU, 1983; 1:50,000.

Physiographic position: dissected mountains.

Topography: 9 percent north-east facing convex slope, on lower tier of backslope position.

Parent material: Mica schist underlain by granite rock.

Vegetation: unimproved pasture dominated by Eragrostis olivacea and Hyparrhenia newtonii.

A -- 0 to 19 centimeters; dark reddish brown (5YR 3/2) clay; massive structure; friable; common medium tubular pores; common fine and medium roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

AB --19 to 35 centimeters; dark reddish brown (5YR 3/3) clay; massive structure; friable; common medium and coarse tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bs - 35 to 65 centimeters; reddish brown (3.75YR 4/4) clay; weak fine and medium subangular blocky structure; friable; common coarse tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear wavy boundary.

Bhs1 -- 65 to 94 centimeters; dark reddish brown (5YR 2.5/2) clay; sombric horizon; weak fine and medium subangular blocky structure; friable; common medium tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bhs2 -- 94 to 129 centimeters; dark reddish brown (80% 2.5YR 2.5/2 and 20% 3.75YR 3/4) clay; sombric horizon; moderate medium subangular blocky structure; friable; common medium tubular pores; common fine and 147 medium roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

B’s -- 129 to 179 centimeters; dark reddish brown (2.5YR 3/4) clay; moderate medium subangular blocky structure; friable; common medium tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

Bt -- 179 to 210 centimeters; red (2.5YR 4/6) clay; moderate medium subangular blocky structure; friable; common medium tubular pores; nonsticky, nonplastic, weakly cemented; extremely acid; limit of description.

Note: The soil temperature is 19 Degrees Centigrades. 148 Teza Forest

Location: Teza, Burundi. North-east of Teza tea factory 50 meters north-east of Nyabudida camping area. Four meters east from the pathway. X: 29° 33’ E; Y: 3° 11’ 37"; altitude: 2255 m. Bukeye SA - 35 - XXIV - 2a topographic map , IGEBU, 1983, 1:50,000.

Physiographic position: extremely dissected mountains of the east horst of the East African rift valleys.

Topography: 6 percent south-east facing convex slope on a summit position.

Parent material: Mica schist underlain by granite rock.

Vegetation: mountainous secondary subtropical forest.

O -- 5 to 0 centimeters; primarily made of fibrist materials (leaves, branches and roots); very dark brown (10YR 2/2); many fine and common coarse roots throughout the horizon; nonsticky, nonplastic, weakly cemented; extremely acid; abrupt wavy boundary.

A -- 0 to 10 centimeters; very dark brown (10YR 2/2) silty clay parting to clay; massive structure; very friable; few fine and medium tubular pores; many medium and few coarse roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear smooth boundary.

AB --10 to 36 centimeters; dark reddish brown (80% 5YR 3/3 and 20% 5YR 3/4) clay; massive structure; friable; few medium tubular pores; many fine, many medium and few coarse roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear smooth boundary.

Bhs1 -- 36 to 74 centimeters; dark reddish brown (5YR 3/3.5) clay; weak fine and medium subangular blocky structure; friable; many medium and coarse tubular pores; many fine and medium roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear smooth boundary.

Bhs2 -- 74 to 125 centimeters; dark brown (7.5YR 3.5/4) clay; weak fine and medium subangular blocky structure; friable; few coarse tubular pores; common fine and medium roots; nonstiky, nonplastic, weakly cemented; extremely acid; clear smooth boundary.

Bhs3 -- 125 to 168 centimeters; dark reddish brown (5YR 3/2.5) clay; sombric horizon; weak fine and medium subangular blocky structure; firm; few 149 medium tubular pores; few medium and coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear smooth boundary.

Bt — 168 to 200 centimeters; dark brown (7.5YR 3.5/4) clay; weak fine and medium subangular blocky structure; firm; many coarse tubular pores; nonssticky, nonplastic, weakly cemented; very strongly acid; 200 centimeters limit of description.

Note: The soil temperature is 12 Degrees Centigrades 150 Bukeye Culture

Location: Bukeye, Burundi. Approximately 1 Km from Bukeye High school, 300 meters east from the National Road # 1 across from Kuwimpfizi. X: 29° 37’ 44"; Y: 3° 12’ 1"; altitude: 1840 m. Bukeye SA - 35 - XXIV - 2a topographic map, IGEBU, 1983, 1:50,000.

Physiographic position: dissected mountains.

Topography: 11 percent north-east facing convex slope, on upper tier of the backslope.

Parent material: Mica schist underlain by granite rock.

Vegetation: banana plantation that received high amounts of domestic residues and manure.

Ap1 -- 0 to 21 centimeters; dark reddish brown (5YR 3/2.5) clay; massive structure; friable; common medium vesicular and tubular pores; common fine and medium roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse smooth boundary.

Ap2 --21 to 46 centimeters; dark reddish brown (5YR 3/2) clay; massive structure; friable; common fine and medium vesicular and tubular pores; many medium and coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear smooth boundary.

Bt1 -- 46 to 71 centimeters; reddish brown (5YR 4/4) clay; moderate medium subangular blocky structure; firm; common fine and medium tubular pores; common medium and coarse roots; common distinct continuous reddish brown (5YR 4/3) clay films on most faces of peds in pores and in roots channels; very strongly acid; diffuse wavy boundary.

Bt2 - 71 to 85 centimeters; reddish brown (5YR 4/3) clay; moderate medium subangular blocky structure; firm; common fine and medium tubular pores; common fine and medium roots; common distinct continuous dark reddish gray (5YR 4/2) clay films on most faces of peds, in pores and in roots channels; extremely acid; diffuse wavy boundary.

Bt3 - 85 to 110 centimeters; dark reddish brown (5YR 3/2) clay; sombric horizon; moderate medium subangular blocky structure; firm; common fine and medium vesicular and tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear irregular boundary. 151

Bt4 -- 110 to 159 centimeters; dark reddish brown (5YR 3/4) clay; moderate medium subangular blocky structure; firm; common fine and medium tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; clear wavy boundary.

Bt5 - 159 to 200 centimeters; dark red (2.5YR 3/6) clay; moderate medium subangular blocky structure; firm; common fine and medium tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; extremely acid; 200 centimeters limit of description.

Note: The soil temperature is 16 Degrees Centigrades. 152

Bukeye Pasture

Location; Bukeye, Burundi. Approximately 1 Km from Bukeye High School, 500 meters east from the National Road # 1 across from Kuwimpfizi. X: 29° 37’ 53", Y: 3° 12’ 4"; altitude: 1805 m. Bukeye SA - 35 - XXIV - 2a topographic map, IGEBU, 1983, 1:50,000.

Physiographic position: dissected mountains.

Topography: 12 percent north-east facing convex slope, on lower tier of the backslope position.

Parent material: Mica schist underlain by granite rock.

Vegetation: unimproved pasture dominated by Eragrostis olivacea and Hyparrhenia newtonii.

A -- 0 to 16 centimeters; dark reddish brown (5YR 2.5/2) clay; massive structure; friable; few medium tubular pores; common fine and medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear irregular boundary.

AB --16 to 40 centimeters; dark reddish brown (5YR 3/2) clay; massive structure; friable; few medium tubular pores; many fine roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear irregular boundary.

Bs1 -- 40 to 63 centimeters; reddish brown (5YR 4/4) clay; massive structure; friable; common fine and medium tubular pores; many fine roots; nonsticky, nonplastic, weakly cemented; strongly acid; diffuse wavy boundary.

Bs2 - 63 to 96 centimeters; intermediate between red and reddish brown (3.75YR 4/4) clay; weak fine and medium subangular blocky structure; friable; common fine and medium tubular pores; common fine and medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary.

Bhs --96 to 138 centimeters; dark reddish brown (5YR 3/4) clay; sombric horizon; weak fine granular structure; friable; common medium tubular pores; few fine and medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

B’s1 - 138 to 171 centimeters; intermediate between reddish brown and red (2.5YR 4/5) clay; weak medium subangular blocky structure; friable; 153 common fine and medium tubular pores, few coarse tubular pores; few medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse smooth boundary.

B’s2 — 171 to 200 centimeters; red (2.5YR 4/6) clay; weak medium subangular blocky structure; friable; common fine and medium tubular pores; nonsticky, nonplastic, weakly cemented; very strongly acid; limit of description.

Note: The soil temperature is 18 Degrees Centigrades. 154 Rwegura Forest

Location: Rwegura, Burundi. Approximately 400 m north-east of the Provincial Road #4 at Mubuga mountain, five meters west of the road. X: 29° 27’ 32"; Y: 2° 53’ 50"; altitude: 2245 m. Ndora SA - 35 - XVIII - 3d topographic map. IGEBU, 1983, 1:50,000.

Physiographic position: extremely dissected mountains of the east horst of the East African rift valleys.

Topography: 2 percent north-east facing flat slope on a summit position.

Vegetation: mountainous secondary subtropical forest.

O -- 9 to 0 centimeters; primarily made of fibrist materials (leaves, branches and roots); 60% dark brown (7.5YR 4/4) and 40% very dark brown (10YR 2/2); many fine and common coarse roots throughtout the horizon; nonsticky, nonplastic, weakly cemented; extremely acid; abrupt wavy boundary.

A - 0 to 21 centimeters; black (10YR 2/1) silty clay; massive structure; friable; few fine tubular pores; many fine and coarse roots; nonsticky, nonplastic, weakly cemented; extremely acid; diffuse wavy boundary.

AB -- 21 to 51 centimeters; very dark brown (10YR 2/2) silty clay; massive structure; friable; few fine and medium tubular pores; common fine, medium and coarse roots; nonsticky, nonplastic, weakly cemented; strongly acid; diffuse wavy boundary.

Bhs1 -- 51 to 81 centimeters; very dark grayish brown (10YR 3/2) clay; massive structure and weak fine subangular blocky structure; friable; few fine and medium vesicular and tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary.

Bhs2 -- 81 to 112 centimeters; black (7.5YR 2/0) clay; sombric horizon; weak fine subangular blocky structure; friable; few fine and medium tubular pores; few medium and coarse roots; slightly sticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary.

Bhs3 --112 to 151 centimeters; dark brown (7.5YR 3/2.5) clay; massive structure and weak fine and medium subangular blocky structure; friable; few fine and medium tubular pores; few medium roots; slightly sticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary. 155 Bt - 151 to 200 centimeters; dark brown (7.5YR 4/4) clay; weak medium subangular blocky structure; friable; few fine and medium tubular pores; slightly sticky, nonplastic, weakly cemented; presence of big holes; very strongly acid; limit of description.

Note: The soil temperature is 12 Degrees Centigrades. 156 Rwegura Culture

Location: Kibati, Burundi. Approximately 400 m east of Kibati elementary school. X: 29° 29’ 42"; Y: 2° 49’ 21"; altitude: 2210 m. Ndora SA - 35 - XVIII - 3d topographic map, IGEBU, 1983, 1:50,000.

Physiographic position: extremely dissected mountains.

Topography: 2 percent south-east facing slope, on a wavy summit position.

Parent material: Mica schist underlain by granite rock.

Vegetation: corn associated with beans fallow that has been receiving high amount of domestic organic residues and manure.

Ap1 - 0 to 14 centimeters; very dark grayish brown (10YR 3/2) clay; massive structure; friable; common fine and medium tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; strongly acid; diffuse wavy boundary.

Ap2 -- 14 to 33 centimeters; very dark gray (10YR 3/1) clay; massive structure; friable; common fine and medium tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; strongly acid; diffuse irregular boundary.

Bt1 -- 33 to 64 centimeters; dark brown (7.5YR 3/3) clay, few brown (7.5YR 4/2) mottles; moderate medium granular structure and strong medium subangular blocky structure parting to strong coarse subangular blocky structure; firm; common fine and medium tubular pores; few fine roots; common distinct continuous dark brown (7.5YR 3/2) clay films on most faces of peds, in pores and in roots channels; very strongly acid; clear wavy boundary.

Bth -- 64 to 96 centimeters; dark reddish brown (5YR 2.5/2) clay, few reddish brown (5YR 4/3) mottles; sombric horizon; strong medium subangular blocky structure parting to strong coarse subangular blocky structure; firm; common fine tubular pores; few fine roots; common distinct continuous black (5YR 2.5/1) clay films on most faces of peds; very strongly acid; clear wavy boundary.

B’t1 -- 96 to 128 centimeters; strong brown (7.5YR 4/4) clay, few dark brown (7.5YR 3/2) mottles; strong medium subangular blocky structure parting to strong coarse subangular blocky structure; firm; common fine tubular pores 157 and few medium tubular pores; few faint discontinuous dark brown (7.5YR 4/3) clay films on vertical faces of peds; moderate termites’ activity; very strongly acid; diffuse smooth boundary.

B’t2 --128 to 170 centimeters; strong brown (7.5YR 4/4) clay; few dark brown (7.5YR 3/2) mottles; strong medium subangular blocky structure; firm; common fine tubular pores; few faint discontinuous dark brown (7.5YR 4/3) clay films on vertical faces of peds; noticeable termites’ activity; very strongly acid; abrupt irregular boundary.

R --170 centimeters and more; sandstone.

Note: The soil temperature is 19 Degrees Centigrades. 158 Rwegura Pasture

Location: Kibati, Burundi. Approximately 600 meters east of Kibati elementary school. X: 29° 29’ 44"; Y: 2° 49’ 18"; altitude: 2190 m. Ndora SA - 35 - XVIII - 3d topographic map, IGEBU, 1983, 1:50,000.

Physiographic position: extremely dissected mountains.

Topography: 11 percent east-facing slope, on a shoulder position.

Parent material: Mica schist underlain by granite rock.

Vegetation: unimproved pasture dominated by Eragrostis olivacea and Hyparrhenia newtonii.

A -- 0 to 30 centimeters; black (10YR 2/1) clay loam; masive structure; common fine tubular pores and few medium and coarse tubular pores; common fine and medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

BA - 30 to 46 centimeters; 70% brown (7.5YR 4/4) and 30% black (10YR 2/1) silty clay; weak fine and medium subangular blocky structure; friable; common fine and medium tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear wavy boundary.

Bs1 -- 46 to 80 centimeters; dark brown (7.5YR 4/4) clay; weak fine and medium subangular blocky structure; friable; common fine and medium tubular pores; common fine roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary.

Bs2 -- 80 to 130 centimeters; reddish brown (5YR 4/3) clay; weak fine and medium subangular blocky structure; friable; common fine and medium tubular pores; few fine roots; nonsticky, nonplastic, weakly cemented; very strongly acid; diffuse wavy boundary.

Bhs1 - 130 to 180 centimeters; black (10YR 2/1) clay; sombric horizon; weak fine and medium subangular blocky structure; friable; common fine and medium tubular pores; few coarse roots; nonsticky, nonplastic, weakly cemented; very strongly acid; clear irregular boundary. 159 Bhs2 -- 180 to 210 centimeters; dark reddish brown (8.75 4/2) clay; transitional between Bsh1 and B’s1; weak fine and medium subangular blocky structure; friable; few fine tubular pores; few medium roots; nonsticky, nonplastic, weakly cemented; very strongly acid; limit of description.

Note: The soil temperature is 16 Degrees Centigrades. APPENDIX B

SELECTED SOIL DATA 161

Table B.1. Physical Properties at Site I.

Bugarama Forest

Depth Horizon Db Ag. St. WDC WRD (CM) (gem-3) (%) (%) (gem"3) 0-20 A 0.61 93.1 0.0 63.3 20-44 BJisI 0.62 92.0 0.0 61.7 44-88 Bhs2 0.56 91.1 0.0 66.2 88-121 Bhs3 0.62 91.5 0.0 41.4 121-152 Bhs4 0.94 91.6 0.0 52.5 152-196 Bt 0.91 92.5 0.0 25.6

Muramvya Culture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem-3) (%) (%) (gem-3) 0-30 Ap1 1.23 94.8 5.1 25.0 30-50 Ap2 1.47 94.2 5.2 22.4 50-76 AB 1.46 95.0 15.5 21.6 76-110 Bt1 1.50 82.7 10.2 22.3 110-141 Bt2 & 1.62 91.4 13.3 21.5 Bt/R 141-210 B’t1 & 1.65 94.3 8.8 21.0 B’t2

Muramvya Pasture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem"3) (%) (%) (gem"3) 0-19 A 0.97 96.0 10.4 42.0 19-35 AB 1.21 94.9 15.5 34.9 35-94 Bs & 0.99 94.1 10.4 35.4 Bhs1 94-129 Bhs2 0.92 93.3 5.2 31.4 129-179 B’s 1.38 86.6 5.1 29.2 179-210 Bt 1.35 91.4 5.1 26.6 162

Table B.2. Physical Properties at Site II.

Teza Forest

Depth Horizon Db Ag. St. WDC WRD (CM) (gem'3) (%) (%) (gem-3) 0-10 A 0.62 89.2 0.0 90.3 10-36 AB 0.69 96.0 0.0 47.7 36-74 -Bhs1 0.90 95,7 0.0 40.2 74-125 Bhs2 1.24 95.9 0.0 39.7 125-168 _Bhs3_ 1.19 95,1 0.0 42.5 168-200 Bt 1.37 96.7 0.0 36.2

Bukeye Culture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem'3) (%) (%) (gem-3) 0-21 Ap1 0.90 91.6..... 5.1 35.4 21-46 Ap2 1.06 93.6 5.1 34.1 46-71 Bt1 1.36 96.3 10.3 32.9 71-110 Bt2 & Bt3 1.32 93.9 10.3 34.6

110-159 Bt4 1.40 93.3 10.3 30.3 159-200 Bt5 1.41 84.6 5.1 28.8

Bukeye Pasture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem'3) (%) (%) (gem-3) 0-16 A 1.19 94.2 5.2 45.2 16-40 AB 1.23 96.7 10.3 42.6 40-96 Bs1 & 1.27 94.9 7.3 36.2 Bs2 96-138 Bhs 1.22 94.1 5.2 33.6 138-171 B’s1 1.34 91.5 5.1 32.9 171-200 B’s2 1.29 92.6 5.1 33.8 163

Table B.3. Physical Properties at Site III.

Rwegura Forest

Depth Horizon Db Ag. St. WDC WRD (CM) (gem-3) (%) (%) (gem'3) 0-21 A 0.62 88.9 0.0 75.5 21-51 AB 0.68 89.5 0.0 68.9 51-81 Bhs1 0.82 90.7 0.0 56.6 81-112 Bhs2 0.85 89.7 0.0 47.3 12-151 Bhs3 1.26 92.3 2.7 36.9 151-200 Bt 1.40 91.2 1.3 36.4

Rwegura Culture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem-3) (%) (%) (gem'3)

0-14 Ad1 1.13 89.6 2.6 39.4 14-33 Ad2 1.24 91.8 10.8 41.6 33-64 Bt1 1.42 88.2 15.6 36.5 64-96 Bth 1.54 85.3 18.2 35.2 96-128 B’t1 1.55 78.0 12.8 33.4 128-170 B’t2 1.52 73.6 15.4 32.2

Rwegura Pasture

Depth Horizon Db Ag. St. WDC WRD (CM) (gem"3) (%) (%) (gem'3) 0-30 A 0.85 93.4 5.5 56.9 30-46 BA 1.16 94.5 18.8 54.6 46-80 Bs1 1.15 93.6 23.8 48.6 80-130 Bs2 1.17 94.0 21.1 42.4 130-180 Bhs1 0.98 91.1 11.1 49.1 180-210 Bhs2 1.20 89.6 7.8 40.0 164

Table B.4. Chemical Properties at Site I, Part One.

Bugarama Forest

Depth Horizon PHW ph kci PHNaF Extr. H (CM) cmol(+)Kg"1 0-20 A 4.0 3.5 8.6 12.4 20-44 Bhs1 4.4 3.7 10.3 3.8 44-88 Bhs2 4.4 3.9 11.4 3.9 88-121 Bhs3 4.4 4.1 11.3 2.1 121-152 Bhs4 4.3 4.3 11.3 1.4 152-196 Bt 4.3 4.1 11.1 1.0

Muramvya Culture

Depth Horizon pHy\/ ph kci PHNaF Extr. H (CM) cmol(+)Kg-1 0-30 Ap1 _4,9 4.3 9.2 0.7 30-50 Ap2 4.9 4.5 9.3 0.8 50-76 AB 4.9 4.4 9.4 0.7 76-110 Bt1 5.1 4.1 9.1 0.7 110-141 Bt2 & 4.8 3.8 9.0 1.6 Bt/R 141-210 B’t1 & 4.7 3.7 9.0 2.9 B’t2

Muramvya Pasture

Depth Horizon PHW ph kci PHNaF Extr. H (CM) cmol(+)Kg‘1 0-19 A 4.1 3.7 9.3 4.5 19-35 AB 4.3 3.8 9.6 7.3 35-94 Bs & 4.3 3.8 10.0 3.5 Bhs1 94-129 Bhs2 4.4 3.9 10.3 4.7 129-179 B’s 4.4 3.8 9.7 5.3 179-210 Bt 4.4 3.8 9.3 4.1 165

Table B.5. Chemical Properties at Site II, Part One.

Teza Forest

Depth Horizon PHW ph kci PHNaF Extr. H (CM) cmol(+)Kg'1 0-10 A 3.2 3.2 _6.8 5.3 10-36 AB 4.4 4.0 9.8 1.5 36-74 Bhs1 4.2 4.0 _10.1 3.3 74-125 Bhs2 4.2. 4.1 11.0 3.1 125-168 Bhs3 4.9 4.2 10.5 5.5 168-200 Bt 4.7 4.2 9.7 3.6

Bukeye Culture

Depth Horizon P^W ph kci PHNaF Extr. H (CM) cmol(+)Kg-1 0-21 Ap1 4.4 4.1 9.1 2.5 21-46 Ap2 4.6 3.7 9.2 4.4 46-71 Bt1 4.6 3.7 .9.3 1.2 71-110 Bt2 & Bt3 4.4 4.0 9.5 0.4

110-159 Bt4 4.2 3.8 9.3 0.6 159-200 Bt5 4.2 3.7 9.1 0.6

Bukeye Pasture

Depth Horizon P^W ph kci PHNaF Extr. H (CM) cmol(+)Kg"1 0-16 A 4.7 3.6 9.1 3.4 16-40 AB 5.0 3.8. 9.9 0.6 40-96 Bs1 & 5.1 3.8 9.5 1.1 Bs2 96-138 Bhs 4.7 3,8 9.7 2.4 138-171 B’s1 4.9 4.0 9.3 1.8 ...... 171-200 B’s2 4.7 3.9 9.3 1.8 166

Table B.6. Chemical Properties at Site III, Part One.

Rwegura Forest

Depth Horizon P^W ph kci PHNaF Extr. H (CM) cmol(+)Kg'1 0-21 A 4.0 3.8 8.6 3.6 21-51 AB 4.3 3.8 9.7 4.4 51-81 Bhs1 5.1 3.9 10.3 3.0 81-112 Bhs2 4.8 4.0 11.3 1.2 12-151 Bhs3 4.8 4.0 10.6 0.0 151-200 Bt 4.7 3.9 9.9 0.3

Rwegura Culture

Depth Horizon P H\j\i ph kci PHNaF Extr. H (CM) cmol(+)Kg-1

0-14 Ad1 5.1 4.4 9.6 0.0 14-33 Ad2 5.1 4.2 10.7 0.0 33-64 Bt1 4.8 4.0 9.6 0.0 64-96 Bth 4.7 4.2 9.6 0.0 96-128 B’t1 4.8 3.9 9.4 0.0 128-170 B’t2 4.7 4.1 9.2 0.0

Rwegura Pasture

Depth Horizon P^W ph kci PHNaF Extr. H (CM) cmol(+)Kg'1 0-30 A 5.0 4.1 10.6 0.5 30-46 BA 4.9 4.1 10.0 0.7 46-80 Bs1 4.8 4.3 9.9 0.5 80-130 Bs2 4.8 4.2 9.8 0.4 130-180 Bhs1 4.7 4.1 11.1 0.2 180-210 Bhs2 4.8 4.0 10.3 0.1 167

Table B.7. Chemical Properties at Site I, Part Two.

Bugarama Forest

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-20 A 3.18 0.68 4.7 46.5 62.7 20-44 Bhs1 2.96 0.48 6.2 33.7 55.5 44-88 Bhs2 3.06 0.51 6.0 22.7 68.8 88-121 Bhs3 2.82 0.38 7.4 24.0 67.7 121-152 Bhs4 3.08 0.33 9.3 20.5 83.3 152-196 Bt 2.44 0.21 11.6 25.2 58.5

Muramvya Culture

Depth Horizon OC N C/N P I Pll (CM) (%) (%) ratio ppm ppm 0-30 Ap1 2.55 0.22 11.6 26.8 37.9 30-50 Ap2 1.67 0.15 11.1 17.2 30.5 50-76 AB 1.22 0.10 12.2 13.8 32.8 76-110 Bt1 0.66 0.07 9.4 12.8 24.9 110-141 Bt2 & 0.45 0.05 9.0 10.7 21.7 Bt/R 141-210 B’t1 & 0.27 0.02 13.5 11.9 19.2 B’t2

Muramvya Pasture

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-19 A 2.82 0.26 10.8 17.9 25.6 19-35 AB 2.17 0.16 13.6 14.2 27.2 35-94 Bs & 1.78 0.14 12.7 15.1 31.7 Bhs1 94-129 Bhs2 1.43 0.11 13.0 14.9 30.2 129-179 B’s 0.66 0.05 13.2 13.0 23.0 179-210 Bt 0.47 0.03 15.7 12.2 21.1 168

Table. B.8. Chemical Properties at Site II, Part Two.

Teza Forest

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm PPm 0-10 A 3.72 1.03 3.6 58.8 59.4 10-36 AB 2.17 0.25 8.7 18.4 33.6 36-74 Bhs1 1.49 0.13 11.5 16.4 35.6 74-125 Bhs2 1.14 0.10 11.4 16.4 35.3 125-168 Bhs3 1.48 0.13 11.4 20.9 50.3 168-200 Bt 0.68 0.05 13.6 13.0 35.3

Bukeye Culture

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-21 Ap1 2.41 0.23 10.5 42.1 53.6 21-46 Ap2 2.06 0.16 12.9 26.7 36.2 46-71 Bt1 1.05 0.13 8.1 13.6 21.9 71-110 Bt2 & Bt3 1.26 0.07 18.0 14.2 33.7

110-159 Bt4 0.90 0.08 11.3 13.3 31.8 159-200 Bt5 0.66 0.05 13.2 13.3 25.7

Bukeye Pasture

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-16 A 2.82 0.23 10.3 17.6 28.8 16-40 AB 2.17 0.15 12.3 14.9 27.9 40-96 Bs1 & 1.78 0.11 11.4 12.9 28.3 Bs2 96-138 Bhs 1.43 0.09 12.1 14.2 33.9 138-171 B’s1 0.66 0.07 8.4 14.3 25.7 171-200 B’s2 0.47 0.06 9.7 15.9 27.2 169

Table B.9. Chemical Properties at Site III, Part Two.

Rwegura Forest

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-21 A 3.23 0.87 3.7 56.6 72.1 21-51 AB 2.89 0.47 6.1 24.7 45.9 51-81 Bhs1 2.77 -0.27 10.3 16,2 41.0 81-112 Bhs2 2.73 .0.31 8.8 20.7 69.0 12-151 Bhs3 1.66 0.16 10.4 15.2 45.6 151-200 Bt 0.87 0.10 8.7 10.0 40.0

Rwegura Culture

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-14 Ap1 1.51 JL15 10.1 14.9 41.0 14-33 Ap2 2.06 0.21 9.8 38.0 86.7 33-64 Bt1 0.99 0.10 9.9 13.5 45.1 64-96 Bth .... 0.98 0.10 9,8 20.3 55.0 96-128 B’t1 0.63 0.07 9.0 12,9 47.8 128-170 B’t2 0.48 0.06 8.0 10.2 37.7

Rwegura Pasture

Depth Horizon OC N C/N P I P II (CM) (%) (%) ratio ppm ppm 0-30 A ...... 2.87 0.41 7.0 18.9 78.3 30-46 BA 2.03 0.19 10.7 6.1 34.2 46-80 Bs1 1.35 0.13 10.4 3.7 37.3 8.0-130 _Bs2 1.35 0.13 10.4 1.9 38.8 130-180 Bhs1 2.67 0,19 14.0 9.3 65.7 180-210 Bhs2 1.35 0.11 12.3 9.5 51.6 Table B.10. Chemical Properties at Site I, Part Three. 170

Bugarama Forest

Depth Horizon Ca Mg K NH.O Sum (CM) ppm ppm ppm Ac BS BS (%) (%) 0-20 A 1.21 5.76 45.22 0.23 0.25 20-44 Bhs1 0.19 2.32 26.89 0.28 0.24 44-88 Bhs2 0.73 0.59 19.84 0.26 0.23 88-121 Bhs3 0.95 0.58 18.73 0.32 0.25 121-152 Bhs4 0.28 0.25 20.15 0.80 0.32 152-196 Bt 0.50 0.22 20.42 0.42 0.48

Muramvya Culture

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%) 0-30 Ap1 928.26 219.80 108.10 19.77 16.87 30-50 Ap2 999.64 263.10 49.99 20.52 . 21.62 50-76 AB 650.23 197.72 40.72 20.22 18.53 76-110 Bt1 327.50 136.85 28.46 15.21 11.62 110-141 Bt2 & 106.40 59.20 22.39 8.42 7.68 Bt/R 141-210 B’t1 & 87.08 47.19 21.20 6.28 5.07 B’t2

Muramvya Pasture

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%) 0-19 A 10.52 6.22 34.28 0.76 0.40 19-35 AB 3.21 2.93 18.94 0.71 0.41 35-94 Bs & 1.32 1.01 11.28 0.00 0.00 Bhs1 94-129 Bhs2 0.86 0.47 10.08 0.00 0.00 129-179 B’s 1.57 0.51 7.86 0.00 0.00 179-210 Bt 1.99 0.19 5.87 0.00 0.00 171 Table B.11. Chemical Properties at Site II, Part Three.

Teza Forest

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%) 0-10 A 311.80 25.31 66.98 2.26 2.18 10-36 AB 3.94 1.37 18.27 0.57 0.41 36-74 Bhs1 0.68 0.39 13.52 0.00 0.00 74-125 Bhs2 0.81 0.29 12.34 0.00 0.00 125-168 Bhs3 1.06 0.22 14.35 0.00 0.00 168-200 Bt 1.07 0.47 13.02 0.00 0.00

Bukeye Culture

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%)

0-21 Ad1 661.50 151.40 297.20 16.55 14.76 21-46 Ad2 127.90 30.32 94.45 3.65 2.75 46-71 Bt1 179.40 64.72 57.88 5.88 4.28 71-110 Bt2 & 110.30 29.90 44.00 2.86 1.92 Bt3 110-159 Bt4 44.58 15.39 45.42 3.52 1.61 159-200 Bt5 30.45 11.11 27.94 2.25 1.25

Bukeye Pasture

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%) 0-16 A 2.67 6.52 61.91 1.02 0.86 16-40 AB 1.21 2.36 20.72 0.71 0.47 40-96 Bs1 & 1.42 1.41 14.22 0.00 0.00 Bs2 96-138 Bhs 1.05 1.08 12.58 0.00 0.00 138-171 B’s1 1.09 0.56 10.21 0.00 0.00 171-200 B’s2 0.84 0.60 9.48 0.00 0.00 172 Table B.12. Chemical Properties at Site III, Part Three.

Rwegura Forest

Depth Horizon Ca Mg K NH.O Sum (CM) ppm ppm ppm Ac BS BS (%) (%)

0-21 A 488.20 62.90 93.82 4.39 4.09 21-51 AB 13.52 6.89 47.88 0.31 0.26 51-81 Bhs1 3.52 2.36 42.86 0.37 0.30 81-112 Bhs2 0.43 0.25 21.98 0.36 0.29 12-151 Bhs3 4.08 0.63 24.56 0.54 0.36 151-200 Bt 6.90 10.57 21.42 0.83 0.53

Rwegura Culture

Depth Horizon Ca Mg K n h 4o Sum (CM) ppm ppm ppm Ac BS BS (%) (%)

0-14 Ad1 392.40 223.80 132.40 12.72 9.64 14-33 Ad2 114.30 96.40 130.90 4.21 3.71 33-64 Bt1 58.52 48.38 159.30 4.38 4.19 64-96 Bth 35.42 29.67 140.40 4.67 3.22 96-128 B’t1 26.38 23.86 123.80 4.68 3.72 128-170 B’t2 24.86 22.84 109.90 5.42 2.17

Rwegura Pasture

Depth Horizon Ca Mg K ppm n h 4o Sum (CM) ppm ppm Ac BS BS (%) (%) 0-30 A -20.52 3.03 27.50 0.44 0.37 30-46 BA 4.05 1.54 15.99 0.00 0.00 46-80 Bs1 -2.14 1.41 14.21 0.00 0.00 __8Q_-13Q Bs2 0.64 1.36 12.64 0.00 0.00 __130-180 Bhsl 0.68 0.29 15.23 0.00 0.00 180-210 Bhs2 0.46 0.25 11.01 0.00 0.00 173

Table B.13. Chemical Properties at Site I, Part Four.

Bugarama Forest

Depth Horizon ECEC NH4OAc c e c Sum CEC (CM) cmol(+)kg"1 cmol(+)Kg'1 cmol(+)Kg_1 0-20 A 9.7 42.7 39.8 20-44 Bhs1 10.3 35.9 42.3 44-88 Bhs2 7.4 38.8 43.6 88-121 Bhs3 6.4 32.1 40.4 121-152 Bhs4 3.8 12.5 31.5 152-196 Bt 4.3 23.6 21.2

Muramvya Culture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg"1 cmol(+)Kg_1 cmol(+)Kg"1 0-30 Ap1 3.6 17.7 .20.0 30-50 Ap2 3.9 18.5 17.1 50-76 AB 3.4 12.7 13.4 76-110 Bt1 2.9 9.9 12.4 110-141 Bt2 & 2.9 7.7 8.5 Bt/R 141-210 B’t1 & 3.1 6.4 7.9 B’t2

Muramvya Pasture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg“1 cmol(+)Kg"1 cmol(+)Kg”1 0-19 A 6.9 13.6 24.4 19-35 AB 6.3 11.0 22.4 35-94 Bs & 4.8 17.0 24.6 Bhs1 94-129 Bhs2 4.6 15.3 21.1 129-179 B’s 3.2 16.4 12.8 179-210 Bt 3.9 8.6 12.2 174

Table B.14. Chemical Properties at Site II, Part Four.

Teza Forest

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg-1 cmol(+)Kg“1 cmol(+)Kg"1 0-10 A 9.2 49.4 50.4 10-36 AB 4.5 17.7 24.4 36-74 Bhs1 5.5 17.1 23.7 74-125 Bhs2 4.3 14.7 19.8 125-168 Bhs3 3.5 18.6 24.3 168-200 Bt 4.2 12.0 14.7

Bukeye Culture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg"1 cmol(+)Kg'1 cmol(+)Kg"1 0-21 Ap1 4.6 18.1 20.3 21-46 Ap2 5.0 16,5 22.4 46-71 Bt1 5.6 13.6 18,4 71-110 Bt2 & Bt3 6.8 17.3 26.4

110-159 Bt4 6.3 8.6 18.9 159-200 Bt5 5.9 9.0 16.2

Bukeye Pasture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg'1 cmol(+)Kg-1 cmol(+)Kg"1 0-16 A 5.7 20.1 23.2 16-40 AB 6.4 14.0 21.2 40-96 Bs1 & 6.0 13.8 19.5 Bs2 96-138 Bhs 6.1 14.9 19.8 138-171 B’s1 5.2 11.0 14.1 171-200 B’s2 5.3 10.2 13.4 175 Table B.15. Chemical Properties at Site III, Part Four.

Rwegura Forest

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg-1 cmol(+)Kg'1 cmol(+)Kg"1 0-21 A 11.3 39.9 42.0 21-51 AB 9.5 32.5 37.9 51-81 Bhs1 8.5 27.4 33.4 81-112 Bhs2 7.2 27.3 34.0 12-151 Bhs3 6.2 18.5 27.6 151-200 Bt 6.2 12.0 18.7

Rwegura Culture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg~1 cmol(+)Kg"1 cmol(+)Kg-1 0-14 Ap1 4.4 18.2 22.0 14-33 Ad2 5.7 23.7 27.9 33-64 Bt1 5.1 16.0 17.3 64-96 Bth 5.4 12.8 19.2 96-128 B’t1 4.7 10.6 13.9 128-170 B’t2 4.0 9.2 12.7

Rwegura Pasture

Depth Horizon ECEC NH4OAc CEC Sum CEC (CM) cmol(+)Kg'1 cmol(+)Kg'1 cmol(+)Kg"1 0-30 A 4.7 23.2 27.0 30-46 BA 6.2 18.1 24.3 46-80 Bs1 6.0 16.5 20.5 80-130 Bs2 5.6 16.3 21.1 130-180 Bhs1 6.0 25.9 32.6 180-210 Bhs2 4.9 8.9 23.0 176 Table B.16. Chemical Properties at Site I, Part Five.

Bugarama Forest

Horizon Extr. Extr. Depth BaCI2 * •k •k Al Sat. (CM) Acidity Acidity Al (%) 0-20 A 39.7 22.0 9.6 43.4 20-44 Bhs1 42.2 14.0 10.2 72.3 44-88 Bhs2 43.5 11.2 7.3 64.6 88-121 Bhs3 40.3 8.4 6.3 74.1 121-152 Bhs4 31.4 5.1 3.7 71.1 152-196 Bt 21.1 5.2 4.2 79.2

Muramvya Culture

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity+ Acidity* Al+ (%) 0-30 Ao1 16.6 0.9 0.2 4.6 30-50 Ad2 13.4 1.0 0.2 4.2 50-76 AB 10.9 1.6 0.9 21.9 76-110 Bt1 10.9 2.1 1.4 38.9 110-141 Bt2 & 7.8 3.8 2.2 49.3 Bt/R 141-210 B’t1 & 7.5 5.6 2.6 43.6 B’t2

Muramvya Pasture

Depth Horizon BaC!2 Extr. Extr. Al Sat. (CM) Acidity"1" Acidity"1" Al+ (%) 0-19 A 24.3 11.3 6.8 59.6 19-35 AB 22.4 13.6 6.3 59.4 35-94 Bs & 24.6 8.3 4.8 58.0 Bhs1 94-129 Bhs2 21.1 9.3 4.6 49.5 129-179 B’s 12.8 8.5 3.2 37.6 179-210 Bt 12.2 8.0 3.9 48.7

+cmol(+)Kg“1 177 Table B.17. Chemical Properties at Site II, Part Five.

Teza Forest

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity+ Acidity"1" Al+ (%) 0-10 A 49.3 13.4 8.1 55.9 10-36 AB 24.3 5.9 4.4 73.3 36-74 Bhs1 23.7 8.8 5.5 62.5 74-125 Bhs2 19.8 7.4 4.3 58.1 125-168 Bhs3 24.3 9.0 3.5 38.9 168-200 Bt 14.7 7.8 4.2 53.8

Bukeye Culture

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity+ Acidity"1" Al+ ..(%) 0-21 Ad1 17.3 4.1 1.6 22.5 21-46 Ad2 21.8 8.8 4.4 46.8 46-71 Bt1 18.6 6.0 4.8 70.6 71-110 Bt2 & Bt3 25.9 6.7 6.2 87.3

110-159 Bt4 18.6 6.6 6.0 86.9 159-200 Bt5 16.0 6.3 5.7 87.7

Bukeye Pasture

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity"1" Acidity"1" Al"1* (%) 0-16 A 23.0 8.9 5.5 60.4 16-40 AB 21.1 6.9 6.3 90.0 40-96 Bs1 & 19.5 7.1 6.1 85.0 Bs2 96-138 Bhs 19.8 8.5 6.1 71.8 138-171 B’s1 14.1 7.0 5.2 74.3 171-200 B’s2 13.4 7.1 5.3 74.6

+cmol(+)Kg"1 178 Table B.18. Chemical Properties at Site III, Part Five.

Rwegura Forest

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity+ Acidity+ Al+ (%) 0-21 A 40.3 13.2 9.6 64.4 21-51 AB 37.8 13.8 9.4 67.6 51-81 Bhs1 33.3 11.4 8.4 73.0 81-112 Bhs2 33.9 8.3 7.1 84.5 12-151 Bhs3 27.5 6.1 6.1 98.4 151-200 Bt 18.6 6.4 6.1 93.8

Rwegura Culture

Depth Horizon BaCI2 Extr. Al Sat. “ +& (CM) Acidity+ Acidity+ > m (%) 0-14 Ap1 19.8 2.2 2.2 50.0 14-33 Ad2 26.9 4.7 4.7 82.4 33-64 Bt1 16.6 4.4 4.4 86.3 64-96 Bth 18.6 4.8 4.8 88.9 96-128 B’t1 13.4 4.2 4.2 89.4 128-170 B’t2 12.2 3.5 3.5 87.5

Rwegura Pasture

Depth Horizon BaCI2 Extr. Extr. Al Sat. (CM) Acidity* Acidity+ Al+ (%Q 0-30 A 26.9 5.1 4.6 88.5 30-46 BA 24.3 6.9 6.2 89.8 46-80 Bs1 20.5 6.5 6.0 92.3 80-130 Bs2 21.1 6.0 5.6 93.3 130-180 Bhs1 32.6 6.2 6.0 96.8 180-210 Bhs2 23.0 5.0 4.9 98.0

+cmol(+)Kg"1 179 Table B.19. Chemical Properties at Site I, Part Six.

Bugarama Forest

Depth Horizon Fe0 Fed A'o F0P Al|p (CM) (%) f a (%) (%) (%) (%) 0-20 A 2.0 1.1 1.2 0.8 1.5 0.9 20-44 Bhs1 2.0 1.1 1.3 1.0 1.8 1.6 44-88 Bhs2 1.9 0.9 0.7 0.4 1.5 1.6 88-121 Bhs3 2.2 1.5 1.4 1.4 1.9 2.0 121-152 Bhs4 3.0 2.3 1.6 2.0 1.9 1.9 152-196 Bt 2.4 1.3 1.7 1.3 1.7 1.3

Muramvya Culture

Depth Horizon Fed Ald F e o A 'o A lp (CM) (%) (%) (%) (%) f a (%) 0-30 Ap1 2.0 0.6 0.8 0.4 0.7 0.4 30-50 Ap2 2.3 0.7 0.8 0.4 1.2 0.6 50-76 AB 2.3 0.7 0.7 0.3 1.3 0.6 76-110 Bt1 2.2 0.5 0.3 0.2 0.9 0.3 110-141 Bt2 & 2.0 0.4 0.2 0.3 0.5 0.2 Bt/R 141-210 B’t1 & 2.0 0.4 0.1 0.1 0.3 0.1 B’t2

Muramvya Pasture

Depth Horizon Fed Ald F e o A 'o AIP (CM) (%) (%) (%) f a (<) 0-19 A 2.8 0.9 0.7 0.4 1.2 0.6 19-35 AB 3.2 0.9 1.3 1.5 1.6 0.8 35-94 Bs & 3.5 1.1 1.0 0.6 1.7 0.9 Bhs1 94-129 Bhs2 3.5 1.1 1.0 0.5 1.7 0.8 129-179 B’s 3.6 0.8 0.4 0.2 1.3 0.5 179-210 Bt 3.5 0.7 0.2 0.2 0.6 0.2 180 Table B. 20. Chemical Properties at Site II, Part Six.

Teza Forest

Depth Horizon CD AL Fe Feo Alo AIP —'Q- ' (CM) Q 31 (%) (%) (%) ( ° 4 (%) 0-10 A 1.3 0.5 0.7 0.3 0.9 0.5 10-36 AB 2.1 0.7 1.0 0.4 1.4 0.8 36-74 Bhs1 2.1 0.7 1.0 0.5 1.4 0.8 74-125 Bhs2 2.0 0.6 0.9 0.4 1.1 0.5 125-168 Bhs3 2.5 0.9 1.2 0.7 1.2 0.7 168-200 Bt 1.8 0.5 0.5 0.3 0.7 0.3

Bukeye Culture

Depth Horizon Fed Ald Feo Alo Fe A'n (CM) (%) (%) (%) (%) <*? (%) 0-21 Ad1 2.3 0.6 0.5 0.3 0.8 0.4 21-46 Ad2 2.4 0.7 0.6 0.3 0.8 0.4 46-71 Bt1 3.2 0.8 0.7 0.3 1.7 0.7 71-110 Bt2 & Bt3 3.3 1.0 1.1 0.4 1.7 0.7

110-159 Bt4 3.1 0.8 0.7 0.3 1.6 0.7 159-200 Bt5 3.3 0.8 0.3 0.2 1.0 0.4

Bukeye Pasture

Depth Horizon Fed A'd Alo Fe s* s* CD AIP (CM) (%) (%) 0 (%) (%f (%) 0-16 A 2.2 0.8 0.4 0.3 1.1 0.6 16-40 AB 2.4 0.7 0.5 0.3 1.3 0.6 40-96 Bs1 & 2.7 0.8 0.8 0.4 1.3 0.5 Bs2 96-138 Bhs 2.8 0.8 1.0 0.4 1.3 0.6 138-171 B’s1 2.9 0.8 0.3 0.2 1.2 0.4 171-200 B’s2 2.8 0.7 0.2 0.2 0.9 0.3 181 Table B.21. Chemical Properties at Site III, Part Six.

Rwegura Forest

Depth Horizon Fed Ald Feo Alo A l p

(CM) (%) (%) (%) ^ CD (°/S U - n (%) 0-21 A 1.8 0.9 1.6 0.7 1.6 0.9 21-51 AB 2.3 1.0 2.3 0.9 2.1 1.2 51-81 Bhs1 2.4 1.0 2.2 0.9 2.0 1.0 81-112 Bhs2 2.6 1.4 2.2 1.4 2.0 1.5 12-151 Bhs3 2.5 0.9 2.1 0.8 1.8 1.0 151-200 Bt 2.4 0.7 1.2 0.4 1.3 0.5

Rwegura Culture

Depth Horizon Ald Fed F e o A 'o FeP A 'p (CM) (%) (%) (%) (%) (%) (%) 0-14 Ap1 2.0 0.6 1.2 0.4 1.5 0.9 14-33 Ao2 2.1 1.0 1.3 0.8 1.7 1.5 33-64 Bt1 2.1 0.6 1.0 0.3 1.4 0.7 64-96 Bth 2.1 0.6 0.9 0.3 1.3 0.7 96-128 B’t1 2.0 0.5 0.8 0.2 1.1 0.5 128-170 B’t2 2.1 0.5 0.4 0.2 0.8 0.3

Rwegura Pasture

Depth Horizon AL Fed F®o A'o FSP A'p (CM) (%) (%) (%) (%) (%) W 0-30 A 1.4 1.1 1.0 1.2 0.7 0.7 30-46 BA 1.9 0.7 1.1 0.5 1.3 0.8 46-80 Bs1 1.9 0.7 1.3 0.5 1.4 0.8 80-130 Bs2 1.9 0.6 1.2 0.5 1.3 0.6 130-180 Bhs1 1.9 1.1 1.9 1.3 1.7 1.7 180-210 Bhs2 1.9 0.7 1.2 0.6 1.2 0.9 VITA

Salvator Kaburungu was born on July 26, 1956 at Gitsinda, Burundi.

He graduated from Don Bosco College (High School) in 1975. He obtained his Ingenieur Agronome degree from the Faculty of Agronomy at University of Burundi, Bujumbura, Burundi in 1981.

He began his professional career in 1982 working as an Instructor at the Faculty of Agronomy, University of Burundi. In 1985, he was promoted to Maitre-Assistant at the same University.

In 1987, he received a scholarship from AFGRAD/ATLAS for his M.S. and Ph.D. program, including English training program. He obtained his M.S. at Louisiana State University in 1990.

182 DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate: Salvator Kaburungu

Major Field: Agronomy

Title of Dissertation: Pedogenic Development of Soils at High Altitude in Burundi

Approved:

EXAMINING COMMITTEE:

/ /'

7 / O O '' I

lii AM*'-

Date of Examination:

October 28, 1993