Prof. Gachene Charles K.K.

Nitrogen (N) fixation was estimated for three Acacia senegal (L.) (A. senegal) Willd. varieties (A. senegal var. senegal, kerensis, and leiorhachis) growing naturally in different sites in the dryland areas of Kenya. The quantities of N2 fixed were estimated by the 15N natural abundance method, using leaves as the sampling material. Balanites aegyptiaca (B. aegyptiaca) was selected as the reference species growing in the same area. Soil samples were also collected under A. senegal trees for nodule assessment. Leaf 15N natural abundance values (δ15N) were significantly different between A. senegal and B. aegyptiaca. These values averaged 6.35, 4.67, and 3.03% for A. senegal var. kerensis, leiorhachis, and senegal, respectively, and were lower than those of the adjacent reference species. There were also significant differences in the amount of N2 fixed (%Ndfa) among the varieties. A. senegal var. senegal showed the highest levels of N2 fixation with a mean of 36% while A. senegal var. kerensis and leiorhachis had equal estimates of 25%. However, no nodules were observed in the collected soil samples. Leaf N values were significantly different among the varieties with a mean of 2.73, 2.46, and 4.03% for A. senegal var. kerensis, leiorhachis, and senegal, respectively. This study shows that the three varieties of A. senegal are able to fix N2 in their natural ecosystems and the differences could probably be due to soil properties and nutrient availability under the different environments. The species can hence be utilized as plantations in agriculture and land rehabilitation programs.

Water relations are among the most important physical phenomena that affect the use of soils for agricultural, ecological, environmental, and engineering purposes. To formulate soil-water relationships, soil hydraulic properties are required as essential inputs. The most important hydraulic properties are the soil-water retention curve and the hydraulic conductivity. The objective of this study was to determine the soil hydraulic properties of a Nitisol, at Kabete Campus Field Station. Use of an internal drainage procedure to characterize the hydraulic properties and soil and water retention curves allowed for the establishment of the moisture and matric potential at field capacity and permanent wilting point. The Bt2 (84 -115) and Bt3 (115 - 143 cm) had the highest clay contents of 619 compared to Ap, AB and Bt1 horizons. The PWP was attained at soil moisture contents of 0.223, 0.284, 0277, 0.307 and 0.314 m3m-3 in the Ap, AB, Bt1, Bt2, and Bt3 horizons, respectively. Horizontal saturated hydraulic conductivity (Ksat) was high at 6.0 cm hr-1 in Ap horizon and decreased to 0.4 cm hr-1 in the subsurface horizon (Bt3). Ksat in the vertical direction was higher than horizontal and ranged from 8.3 cm hr-1 in surface layer to 0.6 cm hr-1 in Bt3 horizon, with exception of Bt1 and Bt2 where horizontal Ksat was greater than vertical. The Ap horizon also had the highest crop extractable water. Though the AB and Bt1 had the same water content at low matric suction, the variation was very wide as the SWRC approached saturation point. Bt1 and Bt2 also had similar water contents at suction range of – 7kPa after which Bt1, tended towards Bt3. Bt3 had the narrowest range of crop extractable water and thus was attributed to texture. The Bt3 retained the most amount of water at 0.314 m3m-3concluding that θPWP increased with depth. The total available water capacity between FC and PWP in the profile was 79.2 mm m-1. The study observed that the field capacity, crop available water contents and hydraulic conductivities were influenced positively by soil organic matter. The Van Genuchten parameters of air entry value (α) and pore size distribution (n) indicated that pore size distribution was not even in the AP and AB horizons. The field capacity was attained at higher matric potential at -5kPa for Bt1 while Bt2 and AP, AB, Bt2 and Bt3 was at -10kPa.The functional relationship, K(θ) = aθb that deals with water redistribution as a result of soil hydraulic properties and evaporative demand of the atmosphere was highly correlated to soil moisture content and texture with R2 values > 0.85.

A two phase study was carried out from 2002 to 2005 in the semi-arid rangelands of Kajiado District, Kenya to determine the effect of introduced forage legumes on soil fertility improvement of natural pastures. During legume evaluation phase, Neonotonia wightii (Glycine), Macroptilium atropurpureum (Siratro), Lablab purpureus cv. Rongai (Dolichos), Mucuna pruriens (Velvet bean) and Stylosanthes scabra var. Seca (Stylo) were screened for adaptability and growth performance under the semi-arid conditions for two years. Results of soil analysis showed there were significant increases in soil pH (4.92 to 5.36), organic carbon (1.17 to 2.57%) , nitrogen (0.17 to 0.22%) and potassium (1.23 to 1.68 me%) probably due to the large amounts of organic residues produced by the legumes (particularly Glycine, Siratro and Stylo which are perennials). The calcium content decreased significantly from 7.97 to 4.50 me% (which was attributed to plant uptake) while the decrease of phosphorus was not significant. During the second phase of study for 1½ years Glycine, Siratro and Stylo were integrated into natural pastures. The results showed that only the soil pH significantly increased from 5.23 to 5.31 while all the other nutrients decreased results, which were attributed to production of less organic residues by the legumes compared to the residues produced during the legume evaluation phase. The study concluded that Glycine, Siratro and Stylo were capable of improving the soil fertility of semi-arid natural pastures only if the respective dry matter production was 10.31, 7.81 and 3.52 tha-1, amounts which were able to produce large amounts of organic residues.

Acacia senegal is a drought-tolerant, multi-purpose tree species, highly valued for gum arabic production and increasingly being used in agro-forestry in Sub-Saharan Africa. Despite its long history of use, there has not been exhaustive genetic evaluation of the extant genetic resource base of A. senegal in Kenya for genetic improvement of the species. Inter-simple sequence repeat (ISSR) and chloroplast microsatellite (cpSSR) markers were used to study genetic diversity among seven Kenyan populations of A. senegal embracing three putative varieties: kerensis, leiorhachis and senegal. The two marker types detected similar levels of Nei’s gene diversity (HISSR = 0.211, HcpSSR = 0.212) among the A. senegal populations. Acacia senegal var. kerensis exhibited the highest diversity using ISSR markers (HISSR = 0.248), followed by varieties leiorhachis (HISSR = 0.218) and senegal (HISSR = 0.151). Analysis of molecular variance (AMOVA) detected significant genetic variations within and among populations (P < 0.001 and P < 0.01 for ISSR and cpSSR, respectively). Based on the unweighted pair group method with arithmetic mean (UPGMA) dendrogram of the seven populations, two regions were differentiated (north and south). Both markers demonstrated their potential for delineating population structure at local and regional levels, and infra-specific relations within the species, hence their potential as tools for conservation, improvement programmes and sustainable use of the species. This study provides baseline genetic information for the domestication of A. senegal varieties in Kenya.

The current pace of rangeland degradation imparted by appalling land use and management systems is greatly limiting the potential of the soil resource to support pasture production in semi - arid rangeland s of Uganda. Our objectives were to determine the effects of land cover change and production systems on pasture biomass yield and to identify the critical soil factors affecting pasture production in Nakasongola. The area was stratified into three product ion systems and three land cover types from which six pasture and soil samples were collected following a Modified - Whittaker sampling method. Pasture biomass was significantly high (p < 0.0001) under herbaceous cover (2019 kg/ha) compared to woody (1302 kg /ha) and bare which had no pasture biomass. The settled production system had a significantly (p = 0.013) high pasture biomass (1266 kg/ha) compared to non settled (1102 kg/ha) and semi settled systems (953 kg/ha). Biomass yield was more associated with hi gh levels of organic matter (r = 0.91), calcium (r = 0.91), magnesium (0.83), nitrogen (r = 0.77) and base saturation (r = 0.88). It can be concluded that maintaining native vegetation cover of the rangelands and increasing levels of limiting nutrients are the major strategies for increasing pasture production in semi - arid rangelands of Nakasongola

Acacia senegal is a multipurpose drought-tolerant tree or shrub legume and is commonly used in agroforestry systems in sub-Saharan Africa for gum arabic production and soil fertility improvement. Despite its wide distribution in Kenya, there has not been exhaustive evaluation on the effects of the extant varieties (kerensis, leiorhachis and senegal) on soil properties under their canopies for sustainable utilization of the species. Three sites in the drylands of Kenya representing the three varieties were selected for assessment. Soil samples were collected under tree canopies at a depth of 0 to 25 cm and were compared with the soils from the open canopies. There were significant differences in soil physicochemical properties among the three varieties (P<0.05 and P< 0.01). Soil nutrients under the canopies were higher than in the open canopies mainly due to effects of litter accumulation. The three varieties have beneficial effects on soil nutrient status in their natural ecosystems and would most likely improve crop productivity in agroforestry systems as well as enhance herbage productivity in the rangelands. The varieties growing under different soil types may have an effect on their gum Arabic production and quality. Key words: Acacia senegal varieties, soil nutrients accumulation, sustainable utilization.

This study analyzed changes in natural resources in Machakos District of Kenya using participatory approaches. The results show that natural resources have decreased since the ranch became a settlement scheme. Natural forests decreased, vast land was cleared, rivers dried-up while soil erosion, drought, temperatures and land degradation increased. Land productivity declined and most farmers abandoned the traditional crops for modern high value crops. However, farmers adopted various coping strategies. Drought resistant crops, early maturing crops and water harvesting were some of the strategies adopted by farmers. The results also show that resource base management at the community level was still a challenge and a lot of investment needed to be made in this area. As part of the study, farmers knowledge of changes in natural resource use was assessed in a resettlement area. Over a period of about 50 years, significant land degradation has occurred as a result of increased population pressure, poor natural resource management and climate change effects. This was reflected in poor/low land productivity and reduced availability of water. Farmers responded by moving away from growing indigenous crops to growing short-duration crop

Low crop yields in the semi arid areas of Kenya have been attributed to, among other factors, low soilfertility, low farm inputs, labour constraints and inappropriate tillage practices that lead to pulverized soils. The aim of this study was to determine the effects of legume cover crops (LCC) on soil propertiesand maize growth in the semi arid area of MachakosmDistrict, Kenya. The study was undertaken in farmers’ fields. The field experiments were carried out in a randomized complete block design with four treatments each replicated four times during the 2008/2009 long (LR) and short rain (SR) seasons. The treatments were T1 = maize + dolichos (Lablab purpureus) + subsoiling; T2 = maize + dolichos + nosubsoiling; T3 = maize alone + no subsoiling; T4 =maize alone with subsoiling). Results from the field experiments showed that rainfall amount and its distribution affected the growth and yield of dolichos and maize. There were significant differences in ground cover between the treatments at P ≤ 0.05 in all the different weeks after planting when measurements were taken. The penetration resistance in all the plots ranged from 3.83 - 4.18 kg cm-2 with treatment T4 having the highest and treatment T1 lowest penetration resistance. There were also siginificant changes in soil N in plots which were under dolichos compared to plots without dolichos. The results obtained in this study also indicated that subsoiling in combination with dolichos had the greatest potential of improving soil properties and crop yields in semi arid environments of Kenya.

Maize grain yield is constrained by inadequate supply of nitrogen caused by insufficient application of fertilizers that are found to be costly and unaffordable in smallholder farming. Therefore there is need to search for locally available and potentially low-cost N sources. Beneficial residual effect from mucuna biomass application during subsequent cropping seasons of maize, hence fertilizer saving, is considered one of the possible solutions. However, there are residual aspects of herbaceous legume green manure application that are not clear and warrant further investigation: The application rate of green manure biomass required to make substantial residual effect on yield of maize produced in subsequent cropping seasons is unknown. The effect of mucuna green manure application rate on seasonal persistence of the residual influence is not adequately described. Consequently, on-farm research was carried out in southwest Kenya in the period 2002-2005. The objective was to determine residual effect of mucuna green manure application rate on maize grain yield during the subsequent first and second cropping seasons after incorporation. Treatments investigated were: mucuna green manure applied at rates of 0, 30, 60, 120, 240 and 480 kg N haG1; and inorganic fertilizer-urea at 30, 60 and 120 kg N haG1. At tissue N concentration of 1.6 to 2 % for mucuna, the application rates worked to 1.5, 3, 6, 12 and 24 t DM haG1 equivalent of its green manure biomass. Randomized complete block design with three replications was used and maize variety H614 planted. Data was collected on maize grain yield at harvest. Genstat discovery edition 2 was used in data analysis. Results obtained showed that mucuna green manure and inorganic fertilizer-urea N do not have residual effect on maize grain yield during the first and second subsequent planting seasons, regardless of the rate applied. These being wide range of mucuna application rates, 1.5 to 24 t DM haG1, would suggest that change in tissue N concentration of the herbaceous legume, while other factors remain similar, would have little if any dramatic alteration on the observed residual response trend.

Gender mainstreaming is crucial in soil and water conservation initiatives. The existing technologies though, as designed, are expected to be gender neutral, lead to gender differences at the adoption stage. This was confirmed during a study conducte d in Kathekakai settlement scheme, Machakos district where despite both men and women participating in soil and water conservation initiatives, women’s efforts to adopt the recommended technologies were hampered by their limited access to authoritative infor ma- tion and lack of control over land. Women though playing major roles as farmers (64.6%), could not make key decisions on land u se. Previous reports indicate that the women in Machakos district contributed significantly to soil and water conservation efforts in the mid 1980’s leading to terracing of over 70% of the district. Gender mainstreaming efforts need to be enhanced for achievement o f sustainable and effective soil and water conservation for improved agricultural production and livelihoods

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

An experiment including planted Tephrosia vogelii and Tithonia diversifolia fallow species and natural fallow was conducted at Maseno, Kenya, for assessing whether these fallows grown on a nutrient depleted land could produce sufficient green manure in six month period, whether their biomass retained on the same plots or transferred to continuously cropped plots with or without added P fertiliser could increase yield of consecutive maize crops and whether it is useful to regularly repeat these fallows on same plots. First fallow was established in randomized complete blocks with three replicates. At harvesting, biomass was recorded, then either incorporated in situ or transferred to continuous cropped plots split with and without added P fertiliser and monitored for the effect in improving consecutive maize crops. The second fallow was managed on this split plot design. The two-planted shrubs fallows produced more than 9 Mg total dry biomass and accumulated 154 to 234 kg N.ha-1, which were significantly higher compared to the production in the natural fallow. The shrubs were also superior to natural fallow for P accumulation (5-22 kg versus 2 kg.ha-1). The aboveground dry biomass harvested from planted T. vogelii and T. diversifolia and either incorporated in situ or transferred into continuously cropped plots increased maize yields by 2.5 folds compared to the unmanured crop, the control. Supplementing the organic materials with an additional 20 kg P inorganic fertilizer increased the 1st maize yield by about 40%. Productivity in the plots with T. vogelii or T. diversifolia aboveground biomass removal was low for the subsequent fallow and maize crops when compared to the performance in plots where biomass was incorporated. To achieve sustained yields of maize in depleted soils requires regular improved fallowing at least one season alternating with one season maize, and additional P inputs

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

Effect of liming and application of phosphatic fertilizer on growth and root nodulation of common bean in an old tea land was investigated. A field trial was conducted where tea had been uprooted after 70 years of monoculture under high inorganic fertilizer input. Lime was applied at 0, 4, 8 and 12 tonnes ha-1 and triple superphosphate at 0, 5, 10 and 15 g per planting hole. The treatments were combined in a factorial design and investigated for effects on the soil pH, nodule formation, and some plant growth parameters. Lime application raised the soil pH and also significantly (P= 0.05) increased dry matter production and nodule quantity. Phosphorus also increased dry matter production and nodule quantity significantly (P = 0.05). These observations are discussed in regard to ways of enabling bean crop grow on old tea lands.

Maize is an important crop in the high and medium rainfall areas of Kenya and thus, there is a need for additional information on the effect of tillage and soil management practices on water conservation and yield of maize. The objective of this study was to determine the effects of seven soil management practices: bare with conventional tillage (BC), bare with minimum tillage (BM), incorporated mulch with conventional tillage (IMuC), manure with conventional tillage (MaC), manure with minimum tillage (MaM), surface mulch with conventional tillage (SMuC), and surface mulch with minimum tillage (SMuM) on water conservation and yield of maize. Soil water content was greater in minimum tilled plots than in the conventionally tilled (hand hoeing) plots during the study period. This was demonstrated by the manure with minimum tilled treatment, which had the highest soil moisture (7.6% and 8.3%) both at 0-7 cm and 7-23 cm depth. However, when organic matter additions were analyzed separately, it was observed that manure irrespective of tillage had the highest soil moisture in both depths (7.6% and 8.1%). The differences within the treatments occurred when soil water contents were highest shortly after the rains. When the amount of soil water for each management practice was averaged for all the seasons, a significant difference was found at P = 0.0001. Manure (Ma) and surface mulched (Smu) treatments had significantly higher soil moisture content than the other treatments. The higher moisture content found under these treatments was attributed to higher infiltration rates, better cover that reduced the rainfall kinetic energy, and to better structural stability contributed to by higher organic carbon. Compared with the traditional form of hand hoeing, minimum tillage increased available soil water content drastically and crop performance was improved. Grain yields were greatest in manure and lowest in bare treatments. There were significant differences in grain yield in all treatments.

In order to meet the challenges of increasing food production, many agricultural production technologies are being developed and recommended for use by the small- scale fanners who are also the main agricultural producers in sub-Saharan Africa. Whereas some ofthe technologies seem promising in the short-term, their long-term effects on soil productive properties are largely unknown. Soil organic carbon which is closely related to soil productivity has been used to measure the impact caused by the application of external inputs to soils since it is known to respond to such changes. However changes in soil organic carbon caused by the applied inputs are slow, often taking several decades to emerge. Therefore simulation models have been used to predict the likely long¬tenn changes in soil organic carbon, hence soil productivity, as a result to continuous application of external inputs to soils. The work reported here evaluates the applicability of ROTH C-26.3, the current version of the Rothamsted carbon model, to simulate the turnover of soil organic matter in soil under tropical conditions using soil organic carbon concentrations from the Kabete long-term trial measured 22 yrs after the start of the experiments. Measured and Long-term field experiments offer the best modeIled soil organic carbon were closely practical means of monitoring and correlated showing the applicability of the understanding the changes caused by model to simulate soil organic matter different agricultural systems, since these turnover in the tropical soil. The results show changes are slow, often taking many years that even the continuous application of to appear (Powlson et al., 1987; Poulton, inorganic fertilizer alone failed to stop the 1996). The Kabete long-term experiment in decline in soil C. When crop residues were Kenya has now run for long enough to yield returned, decline was slowed. However the useful information on the effects of inorganic only treatments able to stop the decline involved large additions of animal manure. The model predicts a lengthy and difficult amelioration process for degraded soils that have received no organic inputs particularly in the smallholder farms where only small amounts of organic resources are available to restore soil organic matter back to agronomically viable levels. Howeverthe use of Farmyard manure alone or in combination with inorganic fertilizer wi111ead to higher equilibrium C concentrations than the initial amount of soil organic C at the start of the trials.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

The Legume Research Network Project (LRNP) was started in 1994 (by then known as the Legume Screening Network) to evaluate suitable legume species for different agro-ecological environments and to subsequently incorporate the “best bets” into the existing farming systems. Initial Network activities included the screening of about 40 legume species, among them, green manuring species, food legumes and forage species. The screening trials were conducted in 11 sites across the country especially where soil infertility had been identified as a major constraint to crop production. The Network extended its activities to include research on legume residue management, integrated nutrient management, livestock feeding and cowpea screening trials. Each site had the task of bulking seeds of promising legume species. The Network members are from KARI, University of Nairobi (UoN), Environmental Action Team (EAT, an NGO based in Kitale) and Community Mobilisation Against Desertification (C-MAD, an NGO based in Rongo near Kisii). The main collaborators are the Ministry of Agriculture, and Rural Development staff, and the farmers from different regions of Kenya. The following are the major highlights of phase 1 activities: ♦ Promising green manure (GM) legume species: The most outstanding green manure legume species across Network sites based mainly on biomass accumulation are Mucuna pruriens, Lablab purpureus, Crotalaria ochroleuca, and Canavalia ensiformis. ♦ Inoculation of best-bet legume species: The rhizobia inoculation study concluded that inoculation of best-bet legumes in the study sites was not necessary but further systematic studies to characterise the native rhizobia and to determine their levels in the soil should be undertaken. ♦ Response of legume species to phosphorus: Three Network sites participated in this trial, namely Kakamega, Kisii and Gatanga. In Kakamega and Kisii, legumes did not respond to application of P. In Gatanga they responded to application of P at the rate of 20 kg ha-1 but did not respond substantially to application beyond this rate. ♦ Potential benefits of GM legume technologies for improved maize yields: Incorporating mucuna biomass (4 - 11 t DM ha-1) into the soil for maize production increased maize yields by 120%. The additional labor required for digging mucuna into the soil was compensated by increased maize yields. Returns to labour were higher in mucuna (US$ 11.50) than in maize only plots (US$ 8.00). Besides, farmers in Gatanga and Kisii Network sites reported that additional labour required for incorporation of legume biomass was minimal because incorporation and land preparation for the companion crop were done simultaneously. ♦ Potential for soil moisture conservation: In a semi-arid site Machakos,mucuna on the surface as mulch gave better yields than incorporating it in the soil probably because of the moisture conservation effect. A farmer in Embu reported that soil moisture was retained for a longer time in plots where mucuna was grown than in plots without mucuna. ♦ Potential for soil improvement: Integrated Nutrient Management (INM) at three sites, Kakamega, Embu and Mtwapa, mucuna and crotalaria were evaluated in field studies that involved the combinations of green manure, FYM and inorganic fertilisers. Higher maize yields were obtained by combining green manure legume with FYM and inorganic N. ♦ Potential for feeding livestock: Livestock feeding studies at, Mtwapa and Katumani showed that performance of cattle and goats improved when fed on legume forage. In Mtwapa, dairy cows fed on mucuna and lablab forage had a daily DM intake of about 9.2 kg cow-1, which was similar to cows fed on Gliricidia sepium forage, a proven fodder tree for the coastal Kenya. Milk yield (6.5 kg day-1) was only 8% less than that produced by cows fed on gliricidia forage. In Katumani, goats supplemented with Neonotonia wightii gained on average 16.37 g while those on basal diet alone lost 23.81g daily.

Gully control and reclamation activities using low-cost measures were carried out in early March 2001 at Gatanga division, Kenya. The study area was selected on the basis of previous work carried out in farmers fields by the Legume Research Network Project (LRNP). The project’s main objective is to introduce green manure legume species that perform well in different agro ecological zones of Kenya mainly for the purpose of soil fertility improvement and erosion control in smallhold farms. Area studied is characterized by a mean annual rainfall of about 1100 mm with a bimodal distribution, deep red soils, steep slopes and intensive landuse. Field activities were carried out in one of the farms which had literally been abandoned due to gully erosion. The length of the gully was 130 m with an average width and depth of 1.62 and 1.4 m, respectively. Work involved planting of grasses (mainly Brachira humidocola) and mucuna (Mucuna pruriens) on the floor and sides of the gully. In addition ‘macro-contour lines’ were constructed in the farm which involved planting lines of mucuna, sesbania (Sesbania sesban) and napier grass (Pennisetum purpureum) along the terrace embankments. Through photographs taken over a 3 year period, evidence is given to show that the gully has completely healed and that the farm has been brought back to productivity.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

Plant residues from trees, crops or weeds play an important role in determining the soil fertility status in smallholder farmers of the central highlands of Kenya region. A study was conducted in Embu district of Kenya, located in the densely populated windward side of the south-eastern slopes of Mount Kenya, to document farmers knowledge and practices in the prevalence and utilisation ofthese residues. A total of 134 small-scale . farmers cutting across five major agro ecological zones of the 30 kIn transect were interviewed. Farmers indicated that trees, Degraded soils are a major constraint to crops or weed residues found in their farms agricultural production and food security in were associated with generation of plant the central highland of Kenya region residues which either improve or impoverish (Smaling, 1997; Hudgens 1996). The the farm niches where they are found. Plant traditional residues that were associated with soil Long term experiments provide some fertility enhancement included: Zea mays, insights in the consequences of land Physeolus vulgaris, Grevillea robusta, management strategies that cannot be Combretum molle, Cordia africana, Ficus obtained through other means. For instance, sycomorous, and Camellina sinensis. On the trials established at Kenya Agricultural other hand, the residues of Eucalyptus Research Institute land at Kabete near saligna, Macadamia integrifolia, Cupressus Nairobi have shown that the soil organic lusitanica, Croton megalocarpus, Acacia matter declined from 2.1 % (in the original meamsii, Sorghum bicolour and Mangifera coffee estate) to 1.2% C after planting a indica continued to impoverish the soil in the maize-beans rotation over a period of twenty farm fields where they occur. The farmers years (Kapkiyai et al., 1999). Swift and attributed the good or poor crop Woomer (1993) state that it is not the loss performance in farm sections with certain of the soil carbon which poses the threat to types of plant residues to the decomposability the smallholder livelihood, but rather the assocated decline in crop yields resulting from degraded soils. Th euse of plant residues play a crucial role in the long-term improvement of soil physical and chemical properties.

Studies were conducted in a semi arid area of Kenya to determine the effects of tillage methods and soil management practices on crust strength and thickness during the 1998/1999 cropping seasons. Conventional and minimum tillage methods, surface mulch, incorporated mulch and goat manure were studied. Crust strength and thickness were measured at sowing, mid season and harvest. The results obtained indicate that there were large seasonal changes in the field measurement of crust strength and thickness over the two seasons. There were significant differences in crust strength and thickness between sampling times within the treatments. There was a good relationship between crust strength and thickness at R2 = 0.46. All the soil management practices and tillage methods resulted into a decrease of crust strength and thickness with time. Minimum tillage however had stronger crusts than conventional tillage at. Crust thickness was significantly influenced by tillage methods and the soil management practices.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

Aerial photograph interpretation on erosion trends from 1961 to 1998 showed that by 1998, approximately 17 and 50% of agricultural land in Ndome and Ghazi respectively had been permanently lost due to the combined effect of rill, inter-rill, and gully erosion, and sand deposition. Although abnormally heavy rains received between 1996–98 caused much of the present land damage, the severity of the damage was enhanced by inappropriate and persistent human perturbations in the land-system notably through vegetative degradation and destruction of soil structure through inappropriate tillage practices. With soil organic matter contents of 1.6 and 1.9%, and clay ratios of 10.1 and 10.6 for Ndome and Ghazi, respectively, the areas inherent erodibility was considered as high thus pointing to the need for careful use and management of the soil resource. That farmers appreciated only land quality indicators that were visible to the naked eye and that directly affected their subsistence, revealed a knowledge gap that saw damages from intrinsic processes like rill and inter-rill erosion proceed unnoticed. This paper argues that the spread of erosion damage in rural agro-ecosystems is survival-driven. And as a remedy to this problem, there is need for the diversification of livelihood endeavours to alternative off-farm income sources to reduce pressure on the already fragile land resource. Further, being the ultimate implementers of conservation technologies and by virtue of the multiplicity and inter-relation of rural household needs, adoption of an integrated erosion management approach with food security as paramount presents the most practical entry point for sustainable land management in such rural agro-ecosystems

The study involved two experiments; the first experiment was at Machakos Farmers Training Centre (semi-arid) while the second was at Kabete Campus Field Station (sub-humid). The aim of the first experiment was to assess the effect of different residue management practices using green manure cover crop (GMCC) on maize yield. The treatments, which were repeated during the subsequent seasons were: maize without fertiliser, maize with fertiliser, maize intercropped with Lima, bean (Phaseolus lunatus), maize+Sunhemp (Crotalaria ochroleuca) and maize+Mucuna (Mucuna pruriens). The biomass obtained was either incorporated or left as surface mulch or removed during the subsequent seasons. The second experiment aimed at assessing the effect of short-duration fallows of green manure cover crops on maize yield. The species planted during the fallow period were Mucuna pruriens, Purple vetch (Vicia benghalensis) and Crotalaria ochroleuca. Two additional plots were planted with a pure stand of maize, with or without fertiliser. The biomass obtained after the fallow period was either incorporated in the soil, removed or left as surface mulch during the subsequent season when all the plots were planted with maize. Biomass accumulation during the first season of the first experiment ranged from 2.8 to 5.9, 5.6 to 18.5 and 11.5 to 20.9 t DM ha-I for P. lunatus, C. ochroleuca and M. pruriens respectively. Maize grain yields were generally higher in the incorporation followed by mulching. The removal treatment had the lowest yield. However, in seasons of low rainfall, the trend especially for mucuna changed, the mulching treatment gave the highest grain yield (4.01 t ha-I) followed by the incorporation treatment (1.90 t ha-I) and the removal treatment gave the lowest yield (1.51 t ha-I). The increase in yields in the mulched plots was attributed to soil moisture conservation. Biomass accumulation for the short duration fallows ranged from 2.0 to 15.6, 11.2 to 18.0 and 15.6 to 20.0 t DM ha-l for' V. benghalensis, C. ochroleuca and M. pruriens, respectively. Maize grain yield after residue incorporation was higher than from maize+fertiliser plots although no fertiliser had been applied in the former plots. Incorporating V. benghalensis and C. ochroleuca more than doubled the yields when compared with the control. Yield increments of 16 to 58% compensated loss of yields during green manuring.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

The effect of different soil management practices on crust strength and thickness, soil water conservation and crop performance was investigated on a ferric lixisol in a semi-arid environment of eastem Kenya. The studv proved that manure and mulching with minimum tillage have a greater effect on the water balance of crusted soils and maize emergence. There was increase in steady infiltration rates, amount of soil water stored in the soil and better drainage. The physical effect of mulch was less important in the rehabilitation of crusted soils in the study site when it was incorporated into the soil. Manure and surface mulch with minimum tillage should therefore be taken into account in land management and water conservation in the semi-arid areas of Kenya. The response of crops to the improved water availability due to manure with minimum and with conventional tillage and surface mulch was very clear. These management practices should be recommended when considering the effectiveness of soil and water management techniques in the study area. Keywords: Crust strength and thickness; Soil management: Tillage methods: Steady infiltration rates and available soil moisture

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

The Arpolo gullying control project in the West Pokot Region of Kenya is described. This is an example of conservation work done to protect the Arpolo dispensary which cost >KES350,000 to put up and serves about 3,000 people and is threatened by a fast-growing gully. The initial proposal for a concrete wall in the first phase of implementation was not feasible since it would be undermined during the rains. The alternative was to plan, design and apply water-harvesting methods and implement biological measures that would reduce the erosive energy of the surface runoff. The design of the water-harvesting structures is shown. The success of the project shows that land and water management requires skillful planning and design, especially where soils are vulnerable to erosion. The physical and chemical properties of the soils in the area are described. Given proper hydraulic designs, water-harvesting measures are still a reliable means of soil conservation.

Observations on crop response to soil water content were made over a 2-year period from 1991 to 1992 on 3m by 10m plots at Kabete, Kel1ya. The total anmial rainfall was 948 and 1125mm in 1991 aind 1992 respectively with 101 and 90 rainy days respectively. Two plots, each replicat¬ed two times were cropped to continu~us maize (Zea mays L.) for four seasons ahd were either fertilized or non-fertilized' Due to the more vigorous crop growth] in the fertilized plots, a higher rate of soil water depletion was noted than in non-fer¬tilized plots at both 20cm and 50cm depth. Crop water stress index values also confirmed this observation. The CWSI taken on 3.4.92 were 630 and 813 for fertilized and non-fertilized crop respectively. The difference in soil water content between fertilized and non-fertilized crops was more pronounced at 50cm depth indicating that the former crop had a well developed root system in the subsoil. In particular, most of the available water was left unuti¬lized during the growing period of 1992 short-rains in the non-fertilized plots. During the four seasons, the fertilized crop generally reached tasselling two to two and a half weeks earlier than the non-fertilized crop. Average maximum crop height, LAI and maize grain and above-ground dry matter yields during the four seasons were 245cm, 5.08, 5555.6 kg and 12,110.7 respec¬tively for fertilized crop. The correspond¬ing figures for non-fertilized crop were 191 em, 3.44, 3305.6 kg ha-l and 6,623.4 kg ha-l.

This paper outlines the development of a new methodology to modify and improve the indigenous practices of soil fertility management in Eastern and Southern Africa and suggests requirements for involvement by collaborating institutions. It provides a brief history of the Tropical Soil Biology and Fertility (TBSF) Programme in Africa, ongoing resource integration work in Zimbabwe, proposed work in Kenya and team development in Zimbabwe.

The effects of three different micro-catchment sizes on survival and growth of plant species was assessed and success in establishment and growth of Croton megalocarpus and Cassia spectabilis was compared in a semi-arid region of Kitui District, Kenya. Both species were planted in 25 x 25 cm, 45 x 45 cm, and 65 x 65 cm spherical micro-catchments. Height and diameter were measured and a survival count was taken. Results show that then micro-catchment size influenced (P>0.05) lateral growth of both species. Lateral growth of C. spectabilis in the smaller two micro-catchments (3.66 and 4.60 cm, resp.) was not significantly different (P>0.05), but was less than in the largest micro-catchment(5,31 cm). These results indicate that the two species are suitable for afforestation in these areas and that their survival is not limited by provision of a catchment in the area.

Stover placement (surface mulch, incorporated or a mixture of mulch and incorporation) was compared with stover removal in the presence and absence of 50kg N fertilizer/ha in trials over two successive seasons in continuous maize cropping systems in two regions in Kenya. Stover applied at 4 tonnes DM/ha was found to have highly variable effects on maize growth and yield according to site., method of stover placement, N application and season. Relative to controls without stover, stover incorporation reduced yield by 39% in the first season, followed by yield increases of 15% in the second season. In the first season there was little or no response to N in the presence of stover. Low N uptake and N use efficiency suggested N immobilization in the incorporation treatment. Yield responses and large N uptake in the following season suggested mineralization of N immobilized in the previous season. Surface mulching at the Kabete site increased grain yields in the first season by 39% and 6% compared to stover removal without and with fertilizer, resp. In the second season, surface mulching markedly reduced yields possibly due to a combination of reduced phytotoxicity and N immobilization. At the Katumani site, stover amendments increased yields compared to removal in both seasons with incorporation results being superior to surface mulch. At this site, in both seasons, application of N reduced the effect of stover mulching and incorporation. You must log in to CAB Direct in order to view search results. If you have forgotten your log

Long-term use of compost (4, 8, 12 or 125 kg/ha for maize and 10, 20, 30 or 200 kg/ha for beans) on improving crop yields and soil physical characteristics was studied at the steep-land research site, Kabete campus, Kenya. Preliminary results showed that maize yields under compost were lower than under diammonium phosphate (DAP). Compost increased the maize yield by 15% compared to an increase of 50% caused by DAP. Bean yield did not show any response to any rates of compost or DAP. The use of compost improved soil physical conditions, mainly bulk density and infiltration rates.

The effectiveness of different grass spp. in reducing runoff and soil loss was studied at Kabete campus, University of Nairobi, Kenya. Half metre wide strips of Napier grass (Pennisetum purpureum), Nandi setaria (Setaria anceps) and tall Signal grass (Brachiaria ruziziensis) were established on 11m x 2m runoff plots, and runoff and soil loss were monitored for each rainfall event during the long and short rains of 1990. In terms of runoff control, there was no significant difference between treatments during early establishment; however, runoff form plot with filter strips was always lower than controls. B. ruziziensis was most effective at runoff and soil loss reduction and this was attributed to growth habit and slow rate of establishment. The capability of the strips to impede runoff improved with time.

The effects of P supply (0, 20, 40, 80, 160 and 320 kg/ha) on growth and nodulation of cowpeas (Vigna unguiculata) cv. Vita 4 and Ife Brown grown in a podzolic soil (Haplustult) was studied in a greenhouse trial. The seeds were inoculated with Bradyrhyzobium CB 756. Plants were harvested at the beginning of flowering. The number of trifoliate leaves and DM yield in both cv. increased with increasing P application. The yield of the tops of Vita 4 was maximum at 160 kg P/ha and that of Ife Brown at 320 kg P/Ha. Extractable P related linearly to the rate of applied P, and DM yield increased with increase in extractable P. Plants grown without P had fewer and smaller nodules. An increase in P from 1 to 160 kg/ha increased the number of nodules/plant from 16 to 113 in Vita 4 and from 14 to 70 in Ife Brown. Similarly, increasing P increased nodule dry weight. The P conc. of the youngest fully expanded leaf (YFEL) and in the whole plant top increased with increasing P supply in both cv. and was generally higher in the YFEL than in the whole top. A critical P conc. of 0.3 and 0.25% was found in the YFEL and the whole top, resp., for cv. Vita 4.

This paper examines the impact of sand harvesting from rivers on the environment, land use and social life in the Masinga Division, Machakos District, Kenya. Sand acts as a safe aquifer for water flowing below and through it. Removal of sand results in destruction of underground aquifers and loss of safe water. sand scooping adversely affects surface water quality and quantity and damages the aquatic ecosystem. Haulage of sand by heavy trucks causes environmental degradation by accelerating soil erosion and affecting soil stability. Storage of sand causes destruction of surface areas through clearing of vegetation and uses land that could be used for agriculture. Related social and health problems include prostitution and high school drop-out rate leading to serious social and health problems. The beneficial effects of sand harvesting include local employment; however, the share of monetary benefits to locals is minimal. The results show that the local community gains the least from sand harvesting, but stands to suffer the most if the degradation of the river system continues. Suggestions are made for safe and sustainable methods of managing sand harvesting, in which greater local involvement and stricter enforcement of regulations to protect the environment are vital.

The possible environmental impacts of water development projects and land use practices such as deforestation and over grazing on water supply in Kenya are reviewed. Measures that can be taken in conjunction with properly conducted environmental impact assessments to mitigate adverse impacts are suggested

This paper describes earthflow features in Nyairoko sub-location of Ol Joro Orok Division, Kenya. In 1989 more than 900 earthflows were counted in Nyandarua District, most of them in Ol Joro Orok and Ol Kalou Divisions. Individual flows affected areas between a few square metres and two hectares. More than 50% of the farms in the area were affected by earthflows. Such earthflows occurred only on grazing land and on convex slopes and were usually the result of extremely slow processes. Soils in the area are mainly moderately well drained, dark reddish brown Luvisols and well-drained, red to reddish-brown Nitisols. The first visible signs were usually shallow scars along the contour, breaking up the grass vegetation. Subsequent soil movement was very gradual. According to old residents of the area some of the earthflow features had been observed as long as 60 years ago and could not be attributed to land use change. One earthflow studied in detail revealed that soil stratification impeded drainage, leading to oversaturation of the topsoil layers and that seasonal subsurface flow generally seemed to be responsible for the soil movement. No obvious triggering could be identified.

Forty controlled gullies distributed in Kiambu, Murang'a and Nyeri Districts of Central Province, Kenya, were randomly selected to evaluate their control measures. A recording schedule was developed to assist in collecting pertinent information. Results indicate that the failure of gully control in Central Province can be attributed to lack of technically skilled personnel and poor maintenance of gully control structures. It is concluded that there is a great need to impart technical knowledge to technical staff, and that it is essential to intensify research on gully control techniques and that there is regular and proper maintenance of all gully control structures after rainstorms.

This paper looks at factors which influence land allocation and crop production patterns in Botswana, across traditional and commercial farmers. It found that, among smallholder traditional farmers, less land is used (on average 5.6 ha) and is subdivided into the growing of three major crops: sorghum, maize and beans, with more land allocated to sorghum, the main staple and more adaptable to the climate. The small amount of cultivated land is attributed to lack of capital, transport and labour. Although large areas of land are owned by commercial farmers a high percentage is left idle due to constraints such as poor marketing systems, poor storage facilities and shortage of labour.

Runoff and soil loss was studied on steep terraced slopes containing various combinations of lines of Grevillea robusta, hedgerows of Calliandra calothyrsus and strips of Pennisetum purpureum in Kabale, southwest Uganda. A factorial design, replicated three times, included three levels of intra-row spacing (0, 3 and 5m) of Grevillea spp. and three understorey types (none, Pennisetum and Calliandra spp). The understorey and Grevillea were interplanted along a contour in the middle of the terrace parallel to the two adjacent risers. Results indicate that when contour lines of trees are on fallow ground there is little or possibly a negative effect in controlling soil and water loss. When contour lines are on steep cultivated terraces they have a significant effect in reducing losses. The understoreys appear to filter out much of the sediment from the runoff, holding it in or above the contour line of the vegetation. The potential for using contour lines of Pennisetum and Calliandra for controlling soil and water loss is discussed.

The implications of global climatic change on water resource development and management in Kenya was examined. The extent to which the water resources of the various regions of Kenya will be affected and the effects on streamflow regimes, rainfall amount, frequency and distribution patterns and the overall effects of these on socio-economic development are discussed. It is suggested that increased drought could occur in low-potential regions of Eastern, Rift valley and North Eastern Provinces, whilst regions that already receive a favourable amount of rainfall could have even higher amounts. Changes in rainfall patterns and stream discharge may require translocation of current water development projects with possible high costs and dire consequences for those dependent upon such projects. The increases aridity in the arid and semi-arid lands may lead to increased salinity of groundwater, limiting it's use. Recommendations include water conservation strategies, groundwater basin recharge, inter-basin water transfer from high to low potential zones, cloud seeding above arid and semi-arid areas, better soil and water conservation strategies, and controlling the use of products contributing to global warming.

Trends in land use practices in the Lake Nakuru catchment area and, in particular, the causes and impacts of deforestation, which threaten the lake's unique ecology. It addresses the importance of natural forests for the successful management of the lake's watershed area. Findings are all discussed in the broader, national context and some suggestions for programmes of action to alleviate problems caused by deforestation are presented.

The constraints on land use and water resources for sustainable agricultural production in Kenya were assessed during May to September 1992 using questionnaires and personal interviews along with existing secondary data using districts representative of the country as a whole. It is concluded that land use planning in Kenya is disastrous. A large proportion of good land is under commercial crops that are less important than food crops, from social and political standpoints. Soil degradative processes, other than accelerated erosion, should be given due attention. Erosion control measures should be evaluated for technical feasibility, economic benefit, social acceptability and environmental suitability. Soil erosion classification is required. Drainage and irrigation potentials are under-exploited, whilst fertilizer use on food crops is minimal. Due to maldistribution and mismanagement of land and water resources, low crop production per unit of land is common. It is recommended that techniques be developed to ensure sustained high productivity of land to satisfy changing needs without unnecessary opening up of new land

Preliminary findings on the effects of land use in the Masinga Dam catchment, Kenya, on the storage capacity of the reservoir are presented. Remote sensing and GIS techniques, supplemented with ground reports, were used to determine areas most susceptible to erosion. A representative catchment was then chosen for rainy season monitoring of soil loss, river suspended sediments and discharge response to rainfall. In addition, Gerlach-type traps were used to evaluate erosion rates under different crop covers and slope gradients. A sample of 200 households was interviewed about their perceptions of erosion problems on their farms. Preliminary results suggest that the major sediment-contributing areas are the densely populated and intensively cultivated foothills of the Aberdares, rather than the semi-arid lowlands directly bordering the reservoir. Poorly drained sealed and murram roads, together with footpaths, cattle tracks and gullies, act as extensions of the drainage network during storms, channelling sediment-laden runoff from bare areas around homesteads and schools directly into the river. Subsistence crops, particularly mono-cropped maize, appear to provide poor cover for a major part of the wet season.

This report reviews work on legume cover crops in eastern Africa and provides a summary of recommendations of the appropriate germplasm, management and posible niches for these plants in farming systems. It briefly identifies various reseasons for the lack of adoption of LCC in farming sysytems. The focus is primarily on herbaceous legume cover plants that attain maximum biomass in less than 12 months.

A review of available literature on the hydrologic impacts of drainage is presented and management practices that reduce such impacts are identified. Factors that influence the impact of drainage include land use, soil type, drainage density, type and condition of the drains, size and duration of the storm, and extent and location of the drainage. Suggested options for minimising impacts include: appropriate land use (fish ponds, duck farming, harvesting natural products, dry season grazing or cultivation); controlled drainage through use of shallow drains, drainage for growing water-loving plants, land preparation (hilling, cambered beds), and blocking drains at the end of the rains; and water storage (dams along the river or farm ponds).

During the last 40 years, water development in the southern Marsabit District of Kenya concentrated mainly on drilling boreholes and constructing large dams and pans which are difficult to maintain without financial aid. In order to make the nomads independent of outside aid, the Marsabit Development Programme has introduced animal traction for dam and pan construction and promotes the management of shallow wells. This paper reports the experiences encountered so far in the integration of water development within the natural resource improvement programme and discusses lessons learnt during the 1991/2 drought in the area in terms of water development and nomadic lifestyle

This paper briefly presents a methodology for planning development in arid and semi-arid areas. It highlights the issues of problem identification and assessment of the natural and socio-economic environment, and describes the database contained in the Range Management Handbook. The handbook makes available baseline data for planning development in arid and semi-arid areas following a successful Farm Management Handbook that covers the high- and medium-potential areas of Kenya. In three parts, the Range Management Handbook covers: the status, principles and applications in Kenya; texts and maps relating to climate, landforms and soils, vegetation types, water sources, range unit inventory, livestock marketing and human ecology; special reports (guide to tolerant plants, pictorial key for goat stocking rates, large scale remote monitoring of vegetation, and a survey method for classification of range conditions) relevant to land planning and use in arid and semi-arid areas

This paper presents results from a trial in Swara Farm, Njoro, Kenya, in which a field was sown with Sorghum almum on a volcanic ash soil. Immediately after sowing, 14 strips measuring 156 m x 9 m were marked out and alternate strips were oversown with inoculated seed of lucerne cv. Hairy Peruvian. During the first year of the trial no differences were noticed in the growth of S. almum, but during the second year it grew better when associated with lucerne. After two years the field was ploughed and sown with wheat. The strips were harvested separately with a combine harvester. Mean yields after S. almum were 2,708 kg/ha and after the S.almum/lucerne mixture were 3,244 kg/ha

This paper reflects on the role of root crops (cocoyam, Colocasia esculenta or C. antiquorum) as a case study in riverbank protection and food security in Kenya. Field performance, water erosion control, ecologically friendly production and food value were studied. It is concluded that the national relevance of cocoyam cultivation lies in its role in its conservation through erosion control and soil fertility maintenance; its low energy input; its potential to supplement per capita calories through high yield per unit area; its perpetuity; its multiple nutritive value, ease of preparation and digestibility; and its industrial potential and competitive market opportunity for various end-products.

This paper presents results of a study aimed at quantifying soil and water management practices and identifying major constraints and implications for future adoption of appropriate technologies. The study was conducted in the Machakos and Makueni Districts of Kenya. Approximately 83% of the farmers interviewed were using manure, while only 4% use fertilizer. 60% had fanya juu terraces on their land, while only 13% were using grass strips. The study concluded that finance is a major constraint limiting farmers' adoption of practices enhancing soil fertility and that lack of conservation practices on grazing land is of great concern. The lack of knowledge of possible benefits of soil and water management practices is noted.

The existing soil erosion damage caused by water drained from minor roads in the Nyeri District of Kenya was assessed and the soil conservation works needed to stabilize the waterways and gullies along the roads are specified. The study evaluated the land affected by road drains/culverts on 25 minor roads with a total length of 149 km. Of the total of 321 culverts identified, 171 (53%) were found to require channel rehabilitation. 68% of the culverts discharged onto steep slopes (>10%). Erosion was found to be more severe in the coffee-growing zones than in tea-growing areas. In the plateau areas, soil erosion from the culvert outlets was minimal. Due to gentle slopes and more perennial vegetation. 20,346 m of channel excavation was needed to provide artificial waterways for the discharge of water drained from the roads. Channel stabilization with grass cover or installation of scour checks was necessary on very gentle slopes. Steep slopes required stone check-dams and single-row post/stone check-dams. Very steep slopes also required lock-and-spill drains and post/stone/wire check-dams. Gabions (57 crossings) were needed mainly for rehabilitation of large gullies along with double-row post/stone check-dams and post/stone/wire check dams. Cut-off drains were necessary in some cases to divert water from the culvert outlets. In areas where vegetation was easily accessible, brushwood check-dams could be used.

This study was carried out with the aim of assessing the effect of accelerated erosion on soil productivity of a humic nitisol at Kabete, Kenya. Twenty runoff plots which had been subjected to varying levels of ero¬sion for four seasons (from 1991 to 1992) were planted with maize during the 1993 long-rains. Ten of the plots were cropped with maize under fertilized conditions while the rest were cropped with maize under non-fertilized conditions. There was very little change in soil physi¬cal properties following erosion. However, there was a decrease in percentage volu¬metric water content after four seasons at pressure heads between - 100cm to - 1500 em. Differences in %C, %N and available P between the least and most eroded plots were significant at the 0.01 probability level. Maize grain and above-ground dry matter yields and crop height were always greater in the least eroded plots compared to the more eroded plots in both fertilized and non-fertilized crops. These crop para¬meters were highly and negatively correlat¬ed with cumulative soil loss. Relative to the least eroded plot, there was a decline in maize grain yields of 214 kg ha-l cm-l of topsoil lost. Plant height was reduced 34cm per cm of soil lost for the first 2.5cm of topsoil.

This manual was developed by the Legume Research Network Project of the Kenya Agricultural Research Institute to provide basic, simple information about soils and their management in order to maintain or improve their fertility. The subjects addressed by the manual are: composition and formation of soils, soil fertility, inorganic ferlitilizers, organic manures, soil erosion and conservation, and cropping systems

This paper examines the anomalies in the sedimentation rating equations in terms of existing notions of statistical analysis and concludes that c=aqb is the preferred relationship. The method of estimating parameters a and b through ordinary least squares and the method of prediction using the log normal probability distribution of the error component z (c=aqbz) is presented using data for the Mathare river at Kabete, Kenya. The need for nonlinear least squares for estimation of parameters a and b is discussed in relation to the additive nature of the error component (c=aqb + z) n the nonlinear form of rating equation. The equation c=Faqb (where F is a correction factor arising due to log normal distribution of the error term z) predicted the sediment yield of the Mathare catchment quite well for 61 days during the long rainy season of 1991.

This paper describes a "near-farmer" applied research project working on 10 field stations scattered through the arid and semi-arid lands of Embu, Meru and Isiolo Districts of Esatern Province, Kenya. The objective of the research was to define better extension messages for resource-poor farmers to enable them to improve their land and water management techniques for improved and sustained yields. Most of the trials related to soil fertility and soil moisture, as well as trials on the use of Vetiver grass for soil conservation, control of the legume root parasite Alectra vogelii, and new introductions such as groundnuts, simsim (sesame), tubers, fodder crops and fruits.

This paper examines some of the reasons for the low success rate in gully control and argues that a more careful examination of each situation and a greater understanding of the processes at work could lead to more successful interventions. It outlines a systematic evaluation of a gully erosion problem and analysis of the options for control or reclamation. Evaluation should involve assessment of the causes of gully formation, gully morphology, gully erosion/sedimentation processes, soil characteristics, land use in the vicinity, and catchment characteristics.

Initial observations of a two-phase trial on the use of Mucuna deeringianum [M. deeringiana] to improve soil productivity are presented. In phase one, 42 seeds were sown at Kabete, Kenya, in October 1990 with 89% germination rate. By September 1991, 4 kg of beans had been harvested. The second phase studied the effects of M. deeringiana on improving crop cover and reducing soil loss at Mbooni and Kabete, resp. Initial observations show that the crop preformed fairly well, providing a cover of up to 25cm uncompressed thickness. Treatments on the runoff plots at Kabete include: bare ground (control), M. deeringiana; maize; the two crops intercropped and sown at the same time; intercropping with M. deeringiana sown one week after the maize; or intercropped with M. deeringiana sown two weeks after the maize. The highest soil loss was found on the bare plot (65 t/ha), followed by the intercropped plot with two weeks between sowing (50 t/ha). Soil loss was lowest in runoff plots with a pure stand of M. deeringiana (11 t/ha). Intercropping and sown at the same time provided highest percentage cover.

The results of soil surveys at the ICIPE research site at Ungoye, South Nyanza, and the Ngori Ngori toposequence in Narok, Kenya, are discussed to provide insight into the applicability of soil survey information to soil and water management. It is concluded that soil survey information in soil and water management programmes can be used through understanding and appreciation of the criteria used in categorizing soils and their management implications. From the examples given it is noted that chemical aspects of the soil should be viewed not just as a means to assess its fertility, but also as a base to predict the behaviour of the soils when subjected to different types of management.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A stream flow model was developed for use in rural catchments in Kenya. In the model the physical environment was divided into three zones: the unsaturated zone (consists of multiple hydrological response units defined by homogenous land use and soil type in which a daily soil moisture balance is maintained), shallow saturated and deep saturated zones which were modelled as regional aquifers. Rainfall and evapotranspiration were distributed according to altitude, and runoff was determined using the curve number method devised by the US Soil Conservation Service: a function of five-day antecedent precipitation. Evapotranspiration was varied according to the soil moisture content, and the weather generator allowed Monte Carol simulations over a long period. The data consisted of daily rainfall, mean daily evaporation for each month, land use/vegetation and soil type. The ARC-INFO GIS package was used to assemble the topographical, hydrological, land use/vegetation type and soil information in different layer. The model was tested on the Naro Moru catchment (172kn2) in Kenya, which has climatic conditions that vary from the glaciated peaks of Mount Kenya (5,200 m) to the semi-arid Laikipia plateau (1,800 m). The model was calibrated over a two-year period and validated over a different two-year period. A comparison of the observed and simulated stream flow showed that minimal calibration was required. The simulated stream flow compared well with the observed values for both the calibration and the validation periods (70-85% for the 10-day period) indicating that the model is appropriate for ungauged catchments.

This paper examine past and present trends in the development of Kenyan water resources, and the status of Kenya's arid and semi-arid lands. It considers the parameters linked with soil erosion and sedimentation processes and suggests means of combating the problem.

A case study was conducted in two reaches in the Naro Moru river basin, which extends from the humid western slopes of Mount Kenya to the semi-arid Laikipia plateau. The study involved measuring the amount of water abstracted from all abstraction points, as well as river discharge, for a period of eight months starting from November 1990. Results showed that the amount of water abstracted from each reach as a percentage of the available river inflow rose from 10% at the end of the wet season to over 60% at the end of the dry season. A survey of water permits revealed that legal abstractions constituted 30% of the amount abstracted during the wet season, but only 8% of the dry season abstractions. Communal water supply systems taking water for domestic use, for watering livestock and for irrigation accounted for over 90% of the water that was abstracted from the river during the period of study.

This paper examines the land use patterns in the upper Tana river basin, Kenya, and the resultant erosion that is responsible for the continued loss of storage capacity of the Masinga reservoir. The main Tana branch of the Masinga reservoir contributes about 85% of the total sediment inflow to the reservoir. The high production rates of sediment are linked to the rivers that feed this branch passing through intensively cultivated slopes of the Aberdares and Mount Kenya. Lack of adequate ground cover and the steep slopes often being cultivated without soil conservation measures result in increased surface runoff and soil loss. The need for effective soil conservation in the catchment area is stressed to prevent the expected lifespan of the reservoir form being reduced dramatically.

This paper describes a completely randomized block experiment that was laid out on four terraces at the Kabete campus, Kenya, to test the suitability of seven plant spp. (Tripsacum laxum, Brachiaria decumbens, Panicum trichocladum, Pennisetum purpureum, Setaria anceps, Brachiaria humidicola and Panicum coloratum var. Makarikariense) for riser stabilization and fodder production. A further treatment, stone pitching, was used as a control. The trial was started in the short rains of 1982 and the land between the risers was cropped, with maize, beans or sunflowers, for the next 12 seasons. In three seasons the rains failed and in four seasons the rains were poor. Observations were made of plant cover and persistence and fodder yields were measured. The impact of the fodder on the adjacent crops was assessed by measuring crop yields in the adjacent rows. The results show that of the fodder grasses that were most vigorous and productive, (P. purpureum and T. laxum) caused greatest depression in crop yield. In seasons of low rainfall, the control showed significantly better crop yield mainly due to lack of competition for water. Of the grasses tested, B. decumbens exhibited best cover and persistence. The costs and benefits of bench-type terracing in terms of loss of cropping area and gain in fodder area are briefly discussed.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

Nutrient loss is one of the factors which causes productivity decline in soils when erosion occurs. This paper gives preliminary results on nutrient losses in eroded soil material in relation to original soil material of some commonly occurring soils in Kenya. Results showed that nutrient losses in eroded soil materials are quite high as indicated by the enrichment ratios of 1.0 for all the nine Kenyan soils studied. Soil conservation measures are recommended

Kenya's landscape is of great topographic diversity, rising from sea level to 5,199 m a.s.l. at the summit of Mount Kenya. Climate, geology, and relief have created different agro-climatic zones with highly variable soil developments. These are discussed with reference to the highland areas above 1,500 m. Of the fourteen major soil types, the dominant ones are Ferralsols, Acrisols, Phaeozems, and Nitosols. Deforestation, agricultural use of slopes, and overgrazing have significantly degraded most soils in the highlands, although they have good structure and relative resistance to soil erosion. A number of conservation measures were introduced in the late 1920s and again after 1972. Since 1980, the government has also heavily encouraged afforestation to control erosion.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.

A model for the establishment of a four-dimensional regional geodetic reference datum is presented. Starting from the three-dimensional integrated geodetic network model, formulations for the establishment of a four-dimensional regional datum are developed. Astronomic latitudes, astronomic longitudes, gravity values, gravity potential differences, gravity differences, and GPS-vectors are considered as observables. The estimated parameters defining the datura are point coordinates, deflections of the vertical and geoidai undulations, and velocities and accelerations on the positional coordinates. The network datum is considered observed over several epochs with parameters estimated from previous epochs being introduced into later epochs as stochastic prior information parameters.