Diagnostic indicators of soil quality in productive and non-productive
smallholders’ fields of Kenya’s Central Highlands
Evah W. Murage
a, Nancy K. Karanja
b, Paul C. Smithson
c, Paul L. Woomer
b,∗aKenya Agricultural Research Institute, PO Box 47811, Nairobi, Kenya
bDepartment of Soil Science, University of Nairobi, PO Box 29053, Nairobi, Kenya
cInternational Centre for Research in Agroforestry, PO Box 30677, Nairobi, Kenya
Received 19 February 1999; received in revised form 7 September 1999; accepted 14 October 1999
Abstract
Food security in the East African Highlands is dependent upon the productivity of lands managed by smallholders who face difficult challenges in maintaining the fertility of their soils. A study was conducted to identify indicators of soil fertility status that are consistent with farmers’ perceptions of soil fertility. Physical, chemical and biological properties of soils were measured from paired fields identified as either productive or non-productive by 12 farmers and compared to findings of a household survey on soil fertility management. Special attention was given to the potential of different soil organic matter fractions to serve as diagnostic indicators of soil fertility. Farmers’ criteria for distinguishing soil productivity included crop performance, soil tilth, moisture and colour and presence of weeds and soil invertebrates. All farmers attributed low fertility to inadequate use of organic and inorganic fertilisers (100%) and removal of crop residues (100%). Other causes included continuous cropping (83%), lack of crop rotation (66%) and soil erosion (42%). Productive soils had significantly higher soil pH, effective cation exchange capacity, exchangeable cations, extractable P and total N and P than non-productive soils. Total organic C and several estimates of soil labile C including particulate organic C (POC), three Ludox density separates of POC, KMnO4-oxidizable C and microbial biomass C were significantly greater in productive soils. Soil microbial biomass N, net N mineralisation and soil respiration were also significantly higher in productive soils. Farmers’ perceptions of soil quality were substantiated through soil chemical analyses and soil organic matter fractions provided precise information on these differences. The similarity of soil physical properties in productive and non-productive fields suggests that differences in chemical and biological indicators may have resulted, in part, from smallholders’ management and are not inherent properties of the soils. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: East African Highlands; Humic Nitisol; Soil fertility management; Soil organic carbon
1. Introduction
Soil fertility depletion results from an imbalance between nutrient inputs, harvest removal and other losses and is reaching critical proportions among smallholders in the East African Highlands. Depletion
∗Corresponding author. Tel.:+254-2-631643. E-mail address: biofix@arcc.or.ke (P.L. Woomer)
of soil organic matter (SOM) is a contributing factor in this decline (Kapkiyai et al., 1998) and several dif-ferent approaches toward soil fertility replenishment are being explored (Buresh et al., 1997; Woomer et al., 1997a). Smaling et al. (1993) estimate that 112, 2.5
and 70 kg ha−1per year of N, P and K, respectively
are lost from agricultural soils of Kenya. Input/output approaches to understanding nutrient dynamics in East African smallhold farming systems were further
Table 1
A profile of smallhold farming in the Central Kenyan Highlands based of a survey of 50 households in Kiambu District, Kenya (after Woomer et al., 1998)
Average farm size 2.3 ha
Age of family farm 26 years
Household members 10 (two adult males, two adult females and six children) Principal crops maize, beans, potatoes, banana, vegetables, coffee, tea Reported crop yields maize 1420, beans 675, potatoes 3800 kg ha−1
Farm animals five cattle, three goats, two swine, 15 poultry
Farm management practices
Apply animal manures 96%
Confine livestock 92%
Feed crop residues to livestock 92%
Apply fertilizers 82%
Combine organic and fertilizer inputs 72%
Produce domestic compost 32%
Sell organic residues 14%
advanced by accounting for internal nutrient flows (Van den Bosch et al., 1998; Woomer et al., 1998), farm resource endowment (Shepherd and Soule, 1998) and specific cropping strategies (Wortmann and Kaizzi, 1998; Elias et al., 1998). Knowledge of these systems has advanced sufficiently that candidate land management options may now be assessed within the wider context of Integrated Nutrient Management, a more holistic approach to judicious resource use that better accounts for farmer preference and opportuni-ties (Deugd et al., 1998).
Typically, smallhold farming systems contain three enterprise areas (Woomer et al., 1998); ‘outfields’ cul-tivated in cereal–legume intercrops mainly intended for home consumption, ‘infields’ of market crops, and home sites where livestock are confined, manures and composts accumulated and kitchen gardens cultivated. Crop residues from these ‘outfields’ are harvested, fed to livestock and manures applied to crops intended for market. The nutrient mining of ‘outfield’ soils results in characteristic patches, plots and fields of nutrient-deficient crops. Depletion of soil organic mat-ter exacerbates this condition (Kapkiyai et al., 1998). Smallholders in East Africa recognize special het-erogeneity within fields and adjust land management practices accordingly (Ugen et al., 1993; Ojiem and Odendo, 1997), however, past studies have not dis-tinguished between inherent differences of soils and those that result from past land management. The Kikuyu Red Clay is the most abundant and important agricultural soil in the Central Highlands of Kenya and
its successful management greatly affects the prosper-ity and food securprosper-ity of Kenya (Siderius and Muchena, 1977). Past studies by researchers at the University of Nairobi have focused upon smallholder land-use in the Central Highlands of Kenya (Table 1) and the en-vironmental consequences of their most common land management practices (Woomer et al., 1997b, 1998; Kapkiyai et al., 1998). The objective of this study was to characterise differences between soils from produc-tive and non-producproduc-tive fields that have resulted from smallholders’ management practices in order to iden-tify soil quality indicators consistent with perceptions of land quality with emphasis on SOM fractions (Cam-bardella and Elliot, 1992; Smith et al., 1993; Blair et al., 1997) and soil biological processes (Woomer et al., 1994).
2. Methods
2.1. Site description
Kiambu District is adjacent to Nairobi, with ready ac-cess to its markets. A short (eight questions), formal, open-ended survey was designed that asked farmers to distinguish between productive versus non-productive sites on their farms, which crops are cultivated in productive and non-productive areas, whether or not and how crop residues, animal manure, mineral fer-tilizers and other soil inputs are utilized, which crops and fields are targeted for inputs, and their opinion on causes and solutions for soil fertility decline.
2.2. Soil sampling
Soils were collected in October 1996 from 12
farms in Kiambu District, Kenya (0◦–0.5◦S latitude
and 36◦30′–37◦E longitude, 1350–2400 m asl).
Farm-ers identified productive and non-productive areas after which 20 soil samples were recovered from the 0–20 cm depth in both types of fields with a small, narrow shovel. The samples were spread on a clean polythene sheet, mixed thoroughly and a 7 kg compos-ite sample recovered. A 2 kg sub-sample was removed from the composite sample for C fractionation and biological process measurements and stored under
re-frigeration at 4◦C. The remaining soil was air-dried,
sieved to 2 mm and stored at ambient temperature for use in the various physical and chemical analyses.
2.3. Soil analyses
Soil texture was determined using a Bouyoucos hy-drometer after Gee and Bauder (1986). Soil pH was
determined in H2O (1 : 2.5), exchangeable Ca and Mg
by 1 M KCl extraction, and exchangeable K and
avail-able P by 0.5 M NaHCO3+0.01 M EDTA extraction.
Extractable inorganic N was measured by extraction in 2M KCl, with ammonium determination after An-derson and Ingram (1993). Nitrate was measured by Cd reduction after Dorich and Nelson (1984) followed by colorimetric determination of nitrite (Hilsheimer and Harwig, 1976). Total organic C was determined
by digesting the soil at 130◦C for 30 min with
concen-trated H2SO4and K2Cr2O7, after which C was
deter-mined colorimetrically (Anderson and Ingram, 1993). Total N and P were determined by Kjedhal digestion (Parkinson and Allen, 1975).
Particulate organic matter (POM) was recovered from soils dispersed in pH 10 water (prepared by
dissolving 0.1 g NaCO3 in 1 l of distilled water) by
end-to-end shaking for 16 h and then collected on a
53mm mesh by wet sieving (Okalebo et al., 1993).
Density fractionation of organic matter using Ludox
at densities 1.13 and 1.37 Mg m−3was conducted by
the method of Meijboom et al. (1995). Potassium per-manganate oxidizable C was determined in 333 mM
KMnO4according to Blair et al. (1997), with the
ex-ception that readings were taken at 24 hr instead of 1 hr reaction time. Soil microbial biomass C and N was determined by the chloroform fumigation–extraction method (Anderson and Ingram, 1993) in a closed
des-iccator for 24 h at 25◦C. Potential anaerobic N
miner-alization was determined according to Anderson and Ingram (1993) following incubation of saturated soils
for 7 days at 40◦C. Soil respiration was determined
us-ing the potassium hydroxide trappus-ing method of Fran-zluebbers et al. (1995) over a period of 24 h followed by titration with standardised HCl using phenolph-thalein indicator after addition of barium chloride.
2.4. Data analysis
Data were entered into a spreadsheet software pro-gramme with measurements as columns and the 12 farms as rows. The data were then imported into SYS-TAT (Wilkinson, 1990) and pairwise t-tests were con-ducted between the two soil categories.
3. Results
3.1. Farmers’ diagnostic criteria of land productivity
Farmers’ criteria for distinguishing productive and non-productive fields included crop performance, ease of tillage, soil moisture retention and soil colour (Fig. 1). Indicator organisms included soil macro-fauna and invading plants. The presence of earth-worms and beetle larvae were regarded as positive features. Weed species were also associated with land quality, particularly Commelina benghalensis (L.) in productive fields and Digitaria scalarum (Chiov.) and Rhynchelytrum repens (Willd.) C.E. Hubb in non-productive fields (data not presented).
Fig. 1. Frequency of criteria cited by smallholders in their as-signment of land quality status as productive or non-productive in Kiambu District, Kenya.
L.), potatoes (Solanum tuberosum L.) and green veg-etables in rotation with maize (Zea mays L.) in the fields they considered productive. Maize and bean (Phaseolus vulgaris L.) intercrops were cultivated in both soil categories but more so in the productive ar-eas. Fodder crops such as napier grass (Pennisetum purpureum Schumach.), lucerne (Medicago sativa L.),
Fig. 2. Frequency of cropping activities by smallholders in fields they regard as either productive or non-productive in Kiambu District, Kenya.
and sweet potatoes (Ipomoea batatas L.) were fre-quently established in the less productive areas.
The main causes of productivity decline identified by farmers were inadequate fertilization (100%), re-moval of crop residues (100%), continuous cultivation (83%), monocropping (67%) and soil erosion (42%). Solutions proposed by the farmers to overcome this condition included use of farmyard manure (100%), increased use of inorganic fertilisers (100%), crop ro-tation (25%) and establishment of contour strips for soil conservation (42%) (data not presented). Although most farmers identified continuous cultivation of land as being a major cause of low fertility status, only 25% of the farmers considered crop rotation a feasible solu-tion to restore lost fertility and no farmers mensolu-tioned fallow rotations as feasible, suggesting land scarcity.
3.2. Soil properties in productive and non-productive soils
Soil physical and chemical properties are presented in Table 2. Mean clay and sand contents for the two soil fertility categories were not significantly different but silt content was significantly higher in productive
soils compared to non-productive soils (p<0.001).
Productive soils contained more exchangeable cations and had higher pH than non-productive soils.
Several soil organic matter fractions were sig-nificantly greater in the productive soils than in non-productive soils (Table 3). Carbon fractions in productive versus non-productive soils were signifi-cantly greater in particulate organic C, Ludox heavy,
medium and light C separates and KMnO4oxidizable
C but not in non-labile C. Soil microbial biomass C and N, potential anaerobic mineralizable nitrogen and 24 h soil respiration were significantly higher in pro-ductive soils than in non-propro-ductive soils (Table 4).
4. Discussion
4.1. Soil management decision-making by smallholders
Table 2
Soil physical and chemical properties for a Humic Nitisol from productive and non-productive soils (0–20 cm) in 12 farms in Kiambu District, Kenya
Soil property Farmers’ category of soil quality Probabilitya
Productive Non-productive
Clay (g kg−1) 350 380 nsb
Sand (g kg−1) 320 350 ns
Silt (g kg−1) 330 270 <0.001
Exchangeable K (cmolckg
−1) 1.9 1.1 <0.001
Exchangeable Ca (cmolckg
−1) 13.0 8.3 <0.001
Exchangeable Mg (cmolckg
−1) 3.4 2.4 <0.001
Extractable NO3− (mg kg−1) 27.0 17.2 <0.05
Extractable P (Olsen) (mg kg−1) 55.2 17.1 <0.001
Effective CEC (cmolckg
−1) 18.3 11.9 <0.001
Soil pH (1 : 2.5 H2O) 6.29 5.56 <0.001
aComparison by Tukey t-test. bns=not significant.
Table 3
Mean organic carbon content in the different organic matter pools for soils from productive and non-productive sites in selected farms in Kiambu District, Kenya
Carbon pool Soil fertility status (mg kg−1) Probabilitya
Productive Non-productive
Particulate Organic C (POC, >53mm) 5753 3829 <0.01
Heavy POC separate (>1.37 Mg m−3) 222 103 <0.05
Medium POC separate (1.13–1.37 Mg m−3) 453 252 <0.01
Light POC separate (<1.13 Mg m−3) 560 420 <0.05
Sum of POC separates 1236 801 <0.01
KMnO4-oxidizable C 19594 14848 <0.001
Non-KMnO4-oxidizable Cc 4554 4420 nsb
Total soil organic C 24148 19268 <0.001
aComparison by Tukey t-test. bns=not significant.
cAn estimate of non-labile soil C.
farmers to assess soil fertility status. Irungu et al. (1996) conducted an on-farm survey among 30 house-holds in the semi-arid area of Tharaka Nithi District to assess soil fertility as perceived by farmers and
re-Table 4
Biological properties of a Humic Nitisol from productive and infertile soils in 12 selected farms of Kiambu District, Kenya
Carbon pool Soil fertility status (mg kg−1) Probabilitya
Productive Non-productive
Soil microbial biomass C (mg kg−1) 145 103 <0.01
Soil microbial biomass N (mg kg−1) 64 46 <0.01
Soil microbial biomass P (mg kg−1) 12 10 nsb
N-mineralization (mg N kg−1 per day) 6.1 3.8 <0.01
Respiration (CO2) (mg C kg−1per day) 61.4 43.7 <0.05
aComparison by Tukey t-test. bns=not significant.
semi-arid, midland agroecosystem. Ugen et al. (1993) also found that small scale farmers’ judgement of soil fer-tility status in Mpigi District of Uganda was based on readily observed visible and tactile soil characteristics. Farmers in Kiambu District manage productive and non-productive soils differently. Productive sites were used for production of high value crops, often due to expectations of more favourable soil moisture conditions into the dry season. These crops included tomatoes, bananas, potatoes and green vegetables. Because of their importance to household diets, beans and maize were cultivated as intercrops in both pro-ductive and non-propro-ductive areas of the farm but more so in productive areas. Plants cultivated in poorer soils were primarily intended for livestock feed but most farmers also recognized their importance in the restoration of soil fertility. Lucerne (M. sativa) fixes atmospheric nitrogen, sweet potato (I. batatas) pro-tects soil from erosion and napier grass (P. purpureum) is an indigenous plant used for soil regeneration by native farmers prior to European contact (Boonman, 1993). This survey and that conducted by Ugen et al. (1993) among smallholders in Uganda show that farmers’ preference to grow high value crops on certain portions of their farm is governed by soil fertility status but the selection of those crops also considers social and economic factors.
Our findings reveal that farmers are more likely to allocate their limited organic resources and fertilizers to higher value crops in more productive areas of the farm than to attempt amelioration of fertility-depleted field areas. Farm ‘outfields’ thereby become mined of soil nutrients (Smaling et al., 1993) due to a conscious decision by land managers. Fertilizer application rates by farmers in Kiambu District are considerably less than that recommended by agricultural extension (FURP, 1994; Woomer et al., 1997b). Increased popu-lation density has led to land scarcity and necessitated continuous cultivation of ‘outfields’ by households, greatly restricting opportunity for fallowing and resulting in accelerated nutrient depletion (Woomer et al., 1998) and soil organic matter loss (Kapkiyai et al., 1998).
4.2. Diagnostic criteria of soil quality
Clay and sand contents did not vary between soil categories, suggesting that differences in soil
prop-erties result from past soil management rather than inherent soil properties. Silt, on the other hand, var-ied between soil categories (Table 2) and differences were greatest where farmers identified soil erosion as a cause of soil fertility depletion (data not presented). This observation is consistent with the general princi-ple described by Brady (1984) that silt is usually the first mineral component of soil washed away by soil erosion.
The effective cation exchange capacity (ECEC) for productive soils was significantly greater than that of
non-productive soils (p<0.001). If soil ECEC and pH
are mainly determined by soil organic matter content and the type and amount of clay and the clay content is not different between the soil categories, then differ-ences in soil pH and nutrient holding capacity (Table 2) mainly arose from changes in the soil organic mat-ter content (Table 3). In contrast, Irungu et al. (1996) failed to detect significant differences in soil pH be-tween good and poor soils in Tharaka Nithi District, Kenya, suggesting that soil pH may not be a reliable indicator of soil fertility status in all smallhold, up-land conditions. The similarity of soil physical proper-ties in productive and non-productive fields observed in this study (Table 1) indicates that differences in chemical and biological properties (Tables 2 and 3) have resulted, in part, from land management and not inherent differences in soil.
conditions, but the additional time and expense of con-ducting these procedures compared to analysis of to-tal organic C may not be warranted in assessment of smallholds in the Central Kenyan Highlands.
Measuring differences in soil C fractions is not as useful as understanding how these differences have occurred. In our case, uncertainties of past crop and soil management histories by smallholders and past colonial estates and the paucity of soils continuing under natural vegetation (Woomer et al., 1998) provide insufficient information with which to compare these various C fractions. For example, in this study it was not possible to generate a Carbon Management Index to compare farms and soils as described by Blair et al. (1997) for lack of a suitable control condition.
The KMnO4 oxidizable C was significantly
dif-ferent between soil categories, but the recalcitrant, non-oxidizable C was not. Blair et al. (1997) reported
that labile C as estimated by the KMnO4 oxidation
technique was extremely sensitive to soil management in Australia. Among the different labile C pools
com-pared in this study, the KMnO4technique emerged as
one of the most convenient and reliable but serious disadvantage rests in its destructive nature, not allow-ing for recovery of a fraction for further
characteri-zation. Fractions inferred through KMnO4 oxidation
were consistently much larger than those recovered through physical procedures, suggesting that these es-timates of lability are based upon different soil organic constituents and are not interchangeable.
5. Conclusions
Smallholders in the Central Highlands of Kenya are aware of the root causes of productivity decline but are often unable to devote sufficient resources to lands undergoing degradation. The internal flow of resources within the farm and allocation of scarce fertilizers are focused upon improving returns in areas viewed as having greatest potential productiv-ity, rather than attempting to ameliorate those areas undergoing decline. This situation results in an in-tensification of human-induced spatial heterogeneity characterizable through several physical, chemical and biological process attributes of soil. The recognition of land management difficulties by smallholders and the documentation of consequences by scientists are
important, but preliminary, steps in the development of new solutions appropriate to farmers’ needs and re-sources. Many adjustments by land managers are un-der way, particularly in shifting toward less intensive cultivation of areas undergoing decline but additional products and technologies that specifically address land rehabilitation without withdrawing lands from food production are urgently required. Several simple field indicators and soil measurements offer potential in assessing the impacts of candidate interventions as these are developed for, and adopted by, the smallhold farming sector in the Central Highlands of Kenya, particularly those cultivating the Kikuyu Red Clay, a Humic Nitisol.
Acknowledgements
We are indebted to the 12 smallholders in Ki-ambu District for their cooperation during soil sam-pling and household survey. We are grateful to V. Mbugua, W. Ngului and V. Karari for laboratory assistance and to P. Mugane for help in sampling farmers’ fields. Financial support to this research project was provided by Rockefeller Foundation Fo-rum on Agricultural Resource Husbandry. Additional funding was provided through a graduate scholarship to the first author from the German Exchange Ser-vice (DAAD). This research was conducted by Ms. Murage in partial fulfilment of a M.Sc. programme in the Department of Soil Science, University of Nairobi.
References
Anderson, J.M., Ingram, J.S.I., 1993. Tropical Soil Biology and Fertility: A Handbook of Methods. CAB International, Wallingford, Oxon, England, 221 pp.
Barrios, E., Buresh, R.J., Sprent, J.I., 1996. Organic matter in soil particle size and density fractions from maize and legume cropping systems. Soil Biol. Biochem. 28, 185–193. Blair, G.B., Lefroy, R.D.B., Singh, B.P., Till, A.R., 1997.
Development and use of a carbon management index to monitor changes in soil C pool size and turnover rate. In: Cadisch, G., Giller, K.E. (Eds.), Driven by Nature: Plant Litter Quality and Decomposition. CAB International, Wallingford, UK, pp. 273–281.
Brady, N.C., 1984. The Nature and Properties of Soils. Macmillan, NY, USA, 750 pp.
Buresh, R.J., Sanchez, P.A., Calhoon, F. (Eds.), 1997. Replenishing Soil Fertility in Africa. Special Publication No. 51, Soil Sci. Soc. Am., Madison, WI, 251 pp.
Cambardella, C.A., Elliot, E.T., 1992. Particulate soil organic matter across a grassland cultivation sequence. Soil Sci. Soc. Am. J. 56, 777–783.
Deugd, M., Röling, N., Smaling, E.M.A., 1998. A new praxeology for integrated nutrient management, facilitating innovation with and by farmers. Agric. Ecosyst. Environ. 71, 93–114. Dorich, R.A., Nelson, D.W., 1984. Evaluation of manual cadmium
reduction methods for determination of nitrate in potassium chloride extracts of soils. Soil Sci. Soc. Am. J. 48, 72–75. Elias, E., Morse, S., Belshaw, D.G.R., 1998. Nitrogen and
phosphorus balances in Kindo Keisha farms in southern Ethiopia. Agric. Ecosyst. Environ. 71, 93–114.
Franzluebbers, A.J., Haney, R.L., Hons, F.M., 1995. Soil nitrogen mineralization potential for improved fertilizer recommendations and decreased nitrate contamination of ground water. Technical Report No. 171, Texas Water Resources Institute.
Fertiliser Use Recommendation Project (FURP), 1994. Fertiliser Use Recommendations, Vol. 4. Kiambu District, Kenya Agricultural Research Institute, Nairobi, 14 pp.
Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis, Part 1. Physical and Mineralogical Methods. Am. Soc. Agron. Soil Sci. Soc. Am., Madison, USA, pp. 383–411.
Hilsheimer, R., Harwig, J., 1976. Colorimetric determination of nitrate from meat and other foods: an alternative colour reagent for the carcinogenic 1-naphthylamine and an improved extraction method. Can. Inst. Food Sci. Tecnol. J. 9, 225– 227.
Irungu, J.W., Warren, G.P., Sutherland, A., 1996. Soil fertility status in smallholder farms in the semi-arid areas of Tharaka Nithi District: farmers’ assessment compared to laboratory analysis. Kenya Agric. Res. Inst. 5th Scientific Conf., Nairobi, 30 October–1 November 1996.
Kapkiyai, J.J., Karanja, N.K., Woomer, P.L., Qureshi, J.N., 1998. Soil organic carbon fractions in a long-term experiment and the potential for their use as a diagnostic assay in highland farming systems of Central Kenya. African Crop Sci. J. 6, 19–28. Meijboom, F.W., Hassink, J., van Noordwijk, M., 1995.
Density fractionation of soil macroorganic matter using silica suspensions. Soil Biol. Biochem. 27, 1109–1111.
Ojiem, J.O., Odendo, M.O., 1997. Farmers’ perception of spatial heterogeneity and its influence on soil management in small-scale farms in western Kenya. African Crop Sci. Conf. Proc. 3 (1), 283–287.
Okalebo, J.R., Gathua, K.W., Woomer, P.L., 1993. Laboratory Methods of Soil and Plant Analysis: A Working Manual. TSBF-UNESCO, Nairobi, 88 pp.
Parkinson, J.A., Allen, S.E., 1975. A wet oxidation procedure suitable for determination of nitrogen and mineral nutrients in biological material. Commun. Soil Sci. Plant Anal. 6, 1–11.
Shepherd, K.D., Soule, M.J., 1998. Soil fertility management in western Kenya: dynamic simulation of productivity, profitability and sustainability at different resource endowment levels. Agric. Ecosyst. Environ. 71, 131–146.
Siderius, W., Muchena, F.N., 1977. Soils and environmental conditions of agricultural research stations in Kenya. Miscellaneous Publication No. 5, Kenya Soil Survey/National Agricultural Research Laboratory, Nairobi.
Smaling, E.M.A., Stoorvogel, J.J., Winmeijer, P.N., 1993. Calculating soil nutrient balances in Africa at different scales II. District scale. Fert. Res. 35, 237–250.
Smith, J.L., Papendick, R.I., Bezdicek, D.F., Lynch, J.M., 1993. Soil organic matter dynamics and crop residue management. In: Blaine Jr., M.F. (Ed.), Soil Microbial Ecology: Applications in Agriculture and Environmental Management. Marcel Dekker, NY, pp. 65–94.
Ugen, M.A., Jjemba, P.K., Fischer, M., 1993. Farmer participation in soil management research in Matugga Village (Mpigi District) of Uganda: an alternative approach. In: Wortmann, C.S., Ransom, J.K. (Eds.), Soil Fertility Research for Maize and Bean Production Systems of the Eastern African Highlands: Proc. Working Group Meeting, Thika, Kenya, 1–4 September 1992. Network on Bean Research in Africa, Workshop Series No. 21, CIAT, Dar es Salaam, Tanzania, pp. 3–20.
Van den Bosch, H., Gitari, J.N., Ogaro, V.N., Maobe, S., Vlaming, J., 1998. Monitoring nutrient flows and economic performance in African farming systems (NUTMON) III. Monitoring nutrient flows and balances in three districts of Kenya. Agric. Ecosyst. Environ. 71, 63–80.
White, F., 1983. The Vegetation of Africa. UNESCO, Paris, 356 pp.
Wilkinson, L., 1990. SYSTAT: A System for Statistics. Systat Inc., Evanston, IL.
Woomer, P.L., Martin, A., Albrecht, A., Resck, D.V.S., Scharpenseel, H.W., 1994. The importance and management of soil organic matter in the tropics. In: Woomer, P.L., Swift, M.J. (Eds.), The Biological Management of Tropical Soil Fertility. Wiley, Chichester, UK, pp. 47–80.
Woomer, P.L., Okalebo, J.R., Sanchez, P.A., 1997a. Phosphorus replenishment in Western Kenya: from field experiments to an operational strategy. African Crop Sci. Conf. Proc. 3 (1), 559– 570.
Woomer, P.L., Palm, C.A., Qureshi, J.N., Kotto-Same, J., 1997b. Carbon sequestration and organic resource management in African smallholder agriculture. In: Lal, R., Kimble, J.M., Follet, R.F., Stewart, B.A. (Eds.), Management of Carbon Sequestration in Soil. Advances in Soil Science. CRC Press, Boca Raton, FL, pp. 58-78.
Woomer, P.L., Bekunda, M.A., Karanja, N.K., Moorehouse, T., Okalebo, J.R., 1998. Agricultural resource management by smallhold farmers in East Africa. Nature and Resources 34 (4), 22–33.
