Production gradients in smallholder banana (cv. Giant Cavendish) farms in Central Kenya

Banana is an increasingly demanded food and cash crop in sub-Saharan Africa. Reported yields in smallholder farms vary substantially. The importance and spread of yield constraints have not been properly quantified. A study was carried out in Central Kenya to (i) quantify the yield levels, the primary yield constraints, and the spatial production gradients in such systems (ii) explore how soil fertility gradients relate to gradients in soil fertility management, and whether this is a function of farmer resource availability. Data was collected on crop management aspects, pests and diseases, and soil and plant tissue samples analyzed for nutrient contents. Bunch yields were higher near homesteads (29.8 t ha⁻¹ yr⁻¹) than at mid-distance (26.8 t ha⁻¹ yr⁻¹), or far away 20.2 t ha⁻¹ yr⁻¹. Yields were much higher than previously reported (11–14 t ha⁻¹ yr⁻¹) in Kenya. Both soil and tissue K levels were higher near and mid-distance, than far from the homestead. Gradients of soil pH, total N, available P and Organic carbon were found, being higher near the homestead, while Mg and Ca were lowest near the homesteads. K was the most deficient nutrient, with tissue K index (I_K) decreasing when moving away from the homesteads. P and Ca deficiencies were also observed. Resource-poor farmers’ soils were higher in exchangeable K and Mg, pH, and total N, and supported higher mat densities compared to resource-endowed farmers’ farms. Soil quality problems were the biggest yield loss factors and not pests and diseases.     

Interactive effects from combining fertilizer and organic residue inputs on nitrogen transformations

Concerns about sustainability of agroecosystems management options in developed and developing countries warrant improved understanding of N cycling. The Integrated Soil Fertility Management paradigm recognizes the possible interactive benefits of combining organic residues with mineral fertilizer inputs on agroecosystem functioning. However, these beneficial effects may be controlled by residue quality. This study examines the controls of inputs on N cycling across a gradient of (1) input, (2) residue quality, and (3) texture. We hypothesized that combining organic residue and mineral fertilizers would enhance potential N availability relative to either input alone. Residue and fertilizer inputs labeled with ¹⁵N (40–60 atom% ¹⁵N) were incubated with 200 g soil for 545 d in a microcosm experiment. Input treatments consisted of a no-input control, organic residues (3.65 g C kg⁻¹ soil, equivalent to 4 Mg C ha⁻¹), mineral N fertilizer (100 mg N kg⁻¹ soil, equivalent to 120 kg N ha⁻¹), and a combination of both with either the residue or fertilizer ¹⁵N-labeled. Zea mays stover inputs were added to four differently textured soils (sand, sandy loam, clay loam, and clay). Additionally, inputs of three residue quality classes (class I: Tithonia diversifolia, class II: Calliandra calothyrsus, class III: Z. mays stover) were applied to the clay soil. Available N and N₂O emissions were measured as indicators for potential plant N uptake and N losses. Combining residue and fertilizer inputs resulted in a significant (P < 0.05) negative interactive effect on total extractable mineral N in all soils. This interactive effect decreased the mineral N pool, due to an immobilization of fertilizer-derived N and was observed up to 181 d, but generally became non-significant after 545 d. The initial reduction in mineral N might lead to less N₂O losses. However, a texture effect on N₂O fluxes was observed, with a significant interactive effect of combining residue and fertilizer inputs decreasing N₂O losses in the coarse textured soils, but increasing N₂O losses in the fine textured soils. The interactive effect on mineral N of combining fertilizer with residue changed from negative to positive with increasing residue quality. Our results indicate that combining fertilizer with medium quality residue has the potential to change N transformations through a negative interactive effect on mineral N. We conclude that capitalizing on interactions between fertilizer and organic residues allows for the development of sustainable nutrient management practices.     

Leaf quality of selected hedgerow species at two canopy ages in the derived savanna zone of West Africa

Prunings in hedgerow intercropping systems are a potential source of organic matter and their quality is an important characteristic driving decomposition and nutrient release. To determine the potential range of residue characteristics and the impact of canopy age on those, selected characteristics of Leucaena leucocephala, Gliricidia sepium, and Senna siamea hedgerow leaf residues were determined for young (5–10 weeks) and old (39–47 weeks) hedgerow canopies in six sites, representative in terms of soil for the derived savanna zone of West-Africa. The N content of the Leucaena and Senna residues decreased with age (from 4.8 to 3.9% and from 4.1 to 3.0%, respectively, while the N content of the Gliricidia residues remained constant (3.9%). The P content of all species decreased with age (from 0.30 to 0.20%, on average). The lignin content of the Leucaena residues increased, and their polyphenol content decreased with age, while for the Senna residues only an increase in polyphenol content with canopy age was evident. Neither the lignin nor the polyphenol content of the Gliricidia leaf residues was markedly affected by canopy age. For all species, the P and ADF content were positively correlated with the ash and lignin content, respectively. The N content of the Leucaena and Senna residues was positively correlated with their P content while the lignin content of the same species was negatively correlated with the polyphenol content. Because of the low variation in certain residue characteristics between sites, the presented data and equations could potentially be used to reasonably estimate those characteristics if no other information is available. For other characteristics, such as polyphenol contents, measurements are preferable in view of the relatively high variation encountered.This revised version was published online in November 2005 with corrections to the Cover Date.     

Redistribution of particulate organic matter during ultrasonic dispersion of highly weathered soils

The use of ultrasonic energy for the dispersion of aggregates in studies of soil organic matter (SOM) fractionation entails a risk of redistribution of particulate organic matter (POM) to smaller particle‐size fractions. As the mechanical strength of straw also decreases with increasing state of decomposition, it can be expected that not all POM will be redistributed to the same extent during such dispersion. Therefore, we studied the redistribution of POM during ultrasonic dispersion and fractionation as a function of (i) dispersion energy applied and (ii) its state of decomposition. Three soils were dispersed at different ultrasonic energies (750, 1500 and 2250 J g^?1 soil) or with sodium carbonate and were fractionated by particle size. Fraction yields were compared with those obtained with a standard particle‐size analysis. Undecomposed or incubated (for 2, 4 or 6 months) ¹³C‐enriched wheat straw was added to the POM fraction (0.25–2 mm) of one of the soils before dispersion and fractionation. Dispersion with sodium carbonate resulted in the weakest dispersion and affected the chemical properties of the fractions obtained through its high pH and the introduction of carbonate. The mildest ultrasonic dispersion treatment (750 J g^?1) did not result in adequate soil dispersion as too much clay was still recovered in the larger fractions. Ultrasonic dispersion at 1500 J g^?1 soil obtained a nearly complete dispersion down to the clay level (0.002 mm), and it did not have a significant effect on the total amount of carbon and nitrogen in the POM fractions. The 2250 J g^?1 treatment was too destructive for the POM fractions since it redistributed up to 31 and 37%, respectively, of the total amount of carbon and nitrogen in these POM fractions to smaller particle‐size fractions. The amount of ¹³C‐enriched wheat straw that was redistributed to smaller particle‐size fractions during ultrasonic dispersion at 1500 J g^?1 increased with increasing incubation time of this straw. Straw particles incubated for 6 months were completely transferred to smaller particle‐size fractions. Therefore, ultrasonic dispersion resulted in fractionation of POM, leaving only the less decomposed particles in this fraction. The amounts of carbon and nitrogen transferred to the silt and clay fractions were, however, negligible compared with the total amounts of carbon and nitrogen in these fractions. It is concluded that ultrasonic dispersion seriously affects the amount and properties of POM fractions. However, it is still considered as an acceptable and appropriate method for the isolation and study of SOM associated with silt and clay fractions.     

Organic resource quality influences short-term aggregate dynamics and soil organic carbon and nitrogen accumulation

Organic C inputs and their rate of stabilization influence C sequestration and nutrient cycling in soils. This study was undertaken to explore the influence of the combined application of different quality organic resources (ORs) with N fertilizers on the link between aggregate dynamics and soil organic C (SOC) and soil N. A mesocosm experiment was conducted in Embu, central Kenya where 4 Mg C ha⁻¹ of Tithonia diversifolia (high quality), Calliandra calothyrsus (intermediate quality) and Zea mays (maize; low quality) were applied to soil compared to a no-input control. Each treatment was fertilized with 120 kg N ha⁻¹ as urea [(NH₂)₂CO] or not fertilized. The soils used in the mesocosms were obtained from a three-year old-field experiment in which the same treatments as in the mesocosm were applied annually. No crops were grown in both the mesocosms and the thee-year field experiment. Soil samples were collected at zero, two, five and eight months after installation of the mesocosms and separated into four aggregate size fractions by wet sieving. Macroaggregates were further fractionated to isolate the microaggregates-within-macroaggregates; all soils and fractions were analyzed for SOC and N. The addition of ORs increased soil aggregation and whole SOC and soil N compared to the control and sole N fertilizer treatments. There were no differences among different OR qualities for whole SOC or soil N, but maize alone resulted in greater mean weight diameter (MWD), macroaggregate SOC and N than sole added Calliandra. The addition of N fertilizer only influenced SOC and soil N dynamics in combination with maize where SOC, soil N and aggregation were lower with the addition of N fertilizer, indicating an increased decomposition and loss of SOC and soil N due to a faster aggregate turnover after addition of N fertilizer. In conclusion, compared to high quality ORs, low quality ORs result in greater aggregate stability and a short-term accumulation of macroaggregate SOC and N. However, the addition of N fertilizers negates these effects of low quality ORs.     

The prospects of reduced tillage in tef (Eragrostis tef Zucca) in Gare Arera, West Shawa Zone of Oromiya, Ethiopia

Soils in Ethiopia are traditionally ploughed repeatedly with an oxen-drawn plough before sowing. The oxen ploughing system exposes the soil to erosion and is expensive for farmers without oxen. This study was undertaken to assess agronomic and economic impacts of alternative, reduced tillage methods. Field experiments were carried out on a Vertisol and a Nitisol for 2 years to study the effect of zero tillage, minimum tillage, conventional tillage, and broad bed furrows (BBF) on the yield of tef (Eragrostis tef Zucca). No significant differences in tef biomass and grain yields were observed between the treatments on both soils in the first year. In Nitisol in the second year, yield was lower in the zero tillage treatment as compared to the other treatments. No difference in yield was observed between single plough, conventional, and BBF. On Vertisol, the yields were higher in BBF as compared to the other treatments. The yields on Vertisol were 1368, 1520, 1560 and 1768 kg ha⁻¹ for the zero tillage, minimum tillage, conventional tillage and BBF treatments respectively. More than twice as much grass weed was observed on zero tillage treatment as compared to the BBF treatment on both soils. Zero tillage gave the lowest gross margin on both soils whereas BBF gave the highest gross margin. The gross margin on Nitisols for the zero tillage and BBF treatments were −108 and 1504 Birr/ha respectively and corresponding numbers for the Vertisol were 520 and 1924 Birr ha⁻¹. On Vertisol there were no significant difference in gross margin between minimum tillage and conventional tillage. Minimum tillage is an interesting option on Vertisols, particularly for female-headed households as it reduces the tillage cost. It may also improve overall productivity of the farming system because it allows partial replacement of oxen with cows and reduces soil erosion.     

Evaluation of symbiotic dinitrogen inputs of herbaceous legumes into tropical cover-crop systems

 Reliable estimates of symbiotically fixed N₂ in herbaceous legumes are important in order to determine their role in maintaining or improving N levels in tropical low-external-input farming systems. We have studied the effects of different management systems on the suitability of two non-N₂-fixing reference crops, imperata [Imperata cylindrica (L.) Rauescel] and maize (Zea mays L.), for estimating N₂ fixation in mucuna [Mucuna pruriens (L.) DC var. utilis (Wright) Bruck] and lablab [Lablab purpureus (L.) Sweet] in the field. The total-N-difference (TND) method of estimating N₂ fixation was compared to the ¹⁵N-isotope-dilution (ID) technique. The two methods did not differ with respect to estimates of N₂ fixation under in situ mulch (IM) systems. In contrast, under live-mulch (LM) systems the TND method underestimated N₂ fixed in mucuna by 29% and in lablab by 40% compared to estimates made with the ID method. Irrespective of the treatment, estimates of N derived from fixation in both herbaceous legumes were not influenced by either of the reference plants. Using the ID technique, the proportion of N₂ derived from fixation in mucuna and lablab at 12 weeks varied from 52% to 90% depending on whether the treatments were N fertilized, inoculated or uninoculated, cover-crop systems. In view of the nature of cover-crop systems in the derived savanna of West and Central Africa, where imperata is usually present as a weed or maize is grown in IM or LM systems, imperata or maize could be used to estimate N₂ fixation and N contributions of the legumes to soil fertility and subsequent crop improvements. Received: 16 October 1999     

Assessment of soil factors limiting growth and establishment of Mucuna in farmers' fields in the derived savanna of the Benin Republic

Mucuna pruriens was been used to control Imperata cylindrica and improve soil fertility in maize and cassava cropping systems in the derived savanna of the Benin Republic, West Africa. However, field observations showed that Mucuna had poor establishment in some farmers' fields. This could be due in part to the poor symbiotic effectiveness of Mucuna and/or its poor nutrition because of mineral deficiencies in the soil. A short-term survey was carried out in 34 farmers' fields located in four different sites (Zouzouvou, Eglime, Tchi, and Niaouli) in the derived savanna to assess the natural nodulation and mycorrhizal infection of Mucuna. This survey was followed by a nutrient-omission trial conducted in a pot experiment using soil collected from two groups of farmers' fields at Zouzouvou where Mucuna had poor establishment. Mycorrhizal infection ranged from 2 to 31% and correlated positively with nodulation and shoot dry matter production of plants grown only in one site at Zouzouvou. The number of rhizobia ranged between <0.05 (near the detection limit) and 15 cells g^-1 soil depending on the plot history and the fields. Nodulation occurred in 79% of the fields with numbers of nodules ranging from 0 to 135 plant^-1. The nutrient-omission trial showed that when N and P were absent in the complete fertilizer treatment, biomass production decreased significantly, on average by 69% (N) and 33% (P). Mg, S, K and micronutrient deficiencies did not reduce significantly the biomass production in the two groups. However, N fertilizer applied additionally each week to some treatments drastically reduced Mucuna nodulation. Strategies to enhance Mucuna establishment and growth are discussed.     

Impact of rainfall regime on the decomposition of leaf litter with contrasting quality under subhumid tropical conditions

In alley-cropping systems, hedgerow trees are regularly cut back. Losses of N released from the decomposing prunings are minimized when N release is synchronized with crop N demand. In this study, the sensitivity of the decomposition of Leucaena leucocephala, Senna siamea, and Dactyladenia barteri leaf litter to the nature of the rainfall regime is correlated with the residue quality. The litterbag technique was used to measure decomposition. Four periods of 115 days, each starting at a time when hedgerow trees are normally pruned, were selected and the rain that fell during each of these periods in 1986 was simulated on a day-to-day basis by applying irrigation water on the litterbags. The number of rainfall events was better correlated with the percentage dry matter loss than with the total amount of precipitation. The relationship consisted of two lines. The slope of the first line, indicative of the sensitivity of the decomposition to varying numbers of rainfall events, correlated well with the watersoluble fraction (P<0.05), the C:N ratio (P<0.05), and the polyphenol: N ratio (P<0.01) of the residues. The decomposition process was shown to be dominated by microbial catabolism, rather than leaching. Because the decomposition of the higher quality residues is affected by varying rainfall patterns and because rainfall may often be unpredictable in frequency and intensity, synchronization of N released from a significant part of the decomposing residue with crop N demand may require additional management practices.     

Heterogeneity of crop productivity and resource use efficiency within smallholder Kenyan farms: Soil fertility gradients or management intensity gradients?

The decrease in crop yields at increasing distances from the homesteads within smallholder farms of Sub-Saharan Africa (SSA) is normally ascribed to the existence of within-farm soil fertility gradients. Field observations also suggest that a large part of such variability is concomitantly caused by poor agronomy. To understand the interaction between soil fertility (S factors) and management decisions (M factors) affecting crop variability, we combined field research conducted in western Kenya (Vihiga, Kakamega and Teso districts; rainfall: 1600, 1800 and 1200 mm, respectively) with explorations using the simple dynamic crop/soil model for dynamic simulation of nutrient balances, previously tested for the region. Field measurements indicated within-farm differences in average maize grain yields of 48% (2.7 vs. 1.4 t ha⁻¹) in Vihiga and of 60% (1.5 vs. 0.6 t ha⁻¹) in Teso, between fields that were close and far from the homestead, respectively. Extreme values ranged widely, e.g. between 4.9 and 0.3 t ha⁻¹ for all the farms surveyed in Vihiga, where the average farm size was 0.6 ha. Maize grain yields tended to increase with increasing contents of soil C, total N, extractable P and exchangeable bases. However, the negative relationship between S factors and distance from the homestead was not as strong as expected, and yield variability was better explained by multiple regression models considering M factors such as planting date, plant density, resource use and weed infestation (40–60% across sites). Then, we analysed the variation in resource (cash, labour, N) use efficiency within farms of different resource endowments with the aid of the simulation model. N balances at plot scale varied from ca. +20 to −18 kg ha⁻¹, from −9 to −20 kg ha⁻¹ and from −16 to −18 kg ha⁻¹ for the different fields of the high, medium and low resource endowment case-study farms, respectively. Labour productivities ranged between ca. 10 and 38 kg grain man-day⁻¹ across field and farm types. The results indicate the need of considering within farm heterogeneity when designing soil fertility management interventions. Resource use efficiency was strongly affected by soil quality. As farmers invest more effort and resources in the more productive and less risky fields, the interaction between S and M factors leads to farmer-driven resource use efficiency gradients within smallholder farms.