In 1997 a preliminary survey of manure management practices and manure quality was carried out in Kiambu and Murang'a Districts, Central Kenya. The results of this survey have been published by Lekasi et al (1998). In order to obtain further information on manure management and quality from a larger sample of farms, a further survey was conducted between the first week of February and the second week of March 1999. This second survey was conducted in Kariti Location, Kandara Division, Maragua District, Central Kenya to within approximately 5 km radius from a small town. Three hundred farmers were interviewed. The survey sought information on practices likely to influence manure quality, namely (1) cattle management - including all aspects of animal housing such as type of animal enclosure, roofing, floor type, drainage, bedding and use of concentrate feeding; and (2) manure management - including the way the manure was handled and stored prior to utilisation including the role of urine and organic materials additions.
From each farm approximately 1 kg of manure was obtained at about 45 cm from the surface of the heap and stored in a plastic bag in a cool hut. The manure was collected weekly and taken to the Muguga laboratories of the Kenya Agricultural Research Institute where it was air dried, ground to pass through a 2 mm sieve and analysed for nutrient content according to methods described by Anderson & Ingram (1996). Only a sub-sample of approximately 50% of the samples was analysed for nitrate-N and ammonium-N.
Only 17 out of 299 farmers did not add both urine and at least some organic matter to the manure during composting. Including urine in the manure did not significantly affect nutrient concentration as measured in heaps reported by the farmer to be ready for application to the field. This observation provided an early indication that addition of urine to composting manure may not necessarily increase nutrient concentration in the finished product.
Some farmers reported adding organic materials directly to the manure heap rather than using it as bedding first. One major source of these materials added directly to the manure heap is rejected fodder. Maize stover was the most frequently added organic material with 87% of farmers using it. It was followed by banana residues (71%), then by napier grass (47%), roadside grass (42%), Grevillea leaves (34%) and other materials (16%) that included avocado leaves, coffee leaves, Lantana prunings, mango leaves and sweet potato vines, reeds, sawdust and weeds. Although these materials were also being used for bedding it is interesting that farmers distinguish between the same material used as feed or as bedding. It might be assumed that bedding consisted of all rejected feed. This is evidently not always the case, and organic materials appear to be selected for particular usage. An example of this is maize leaves being used as fodder whilst maize stems are used as bedding _ both are classified as `maize stover'. Including organic matter directly in the heap affected only the P concentration of the manure, reducing it from 0.32% to 0.30% (p = 0.032).
Considerable differences were observed in the nutrient concentrations of the manures, excluding those where urine and/or organic matter had not been added to the heaps (Table 1).
Housing and roof type: The three categories of housing structure found in the study were traditional boma (kraal), improved boma and zero-grazing (a model design developed by the National Dairy Development Project incorporating roof, separate lying and feeding areas, water trough, feeding trough and impervious flooring), which represented 6, 84 and 9% respectively, of the total number of farms surveyed. Of the total housing types surveyed 16% had no roof, 69% had partial roof and 15% had a full roof.
Floor type: Ninety-six percent of farms had soil floors and 4% had concrete or stone floors in the boma/zero grazing unit.
Drainage: Farms had a variety of different drainage systems that either allowed drainage of urine away (well drained) or retained much of it in the animal sheds (poorly drained). Poor drainage was encountered on 58% of farms surveyed.
Use of bedding: Sixty-nine percent of all farms surveyed used some form of bedding material. The main type of materials used (expressed as proportions of farms that used at least some bedding) were: maize stover (92%), banana residues (51%), grass (45%), napier grass (34%), Grevillea prunings (13%). Other organic materials include reeds (15%), weeds (6%), avocado leaves (4%), mango leaves (4%), bean trash (3%), sawdust (3%), coffee leaves (2%), jacaranda (1%), and Lantana (1%). Most of the farms used more than one type of organic material for bedding.
Concentrate feeding: Seventy percent of farms fed their animals some sort of purchased concentrate.
Manure management practices: Most farmers preferred to store their manure in a heap or pit (67%) rather than by deep littering (33%), and 90% did not cover the manure. Forty-six percent of farmers kept the manure under some sort of shade. Farmers who did not turn, infrequently turned and frequently turned the manure during storage represented 45, 51 and 4%, respectively. The reported age of the manure heaps at sampling time ranged from 1 to 8 months, with 5 months being the most common age (5 months (42%), 4 months (13%), 6 months (13%), 3 months (12%), 2 months (10%), 1 month (5%), 8 months (4%) and 7 months (1%).
Multi-factor Analysis of Variance was carried out using General Linear Model on Minitab to determine the effect of a range of livestock and manure management practices on manure/compost quality, excluding manures where urine and/or organic matter had not been added to the heaps (Table 2). All values were log transformed prior to analysis to normalise the data. Only single factor effects and two-factor interactions were analysed, and rank deficiency and/or colinearity restricted the number of interactions that could be examined with the statistical model used. Very few two-factor interactions were significant and these did not appear to have much biological relevance.
Relatively few of the management practices could, as single factors, be shown to significantly affect the nutrient content of the manure. Percentage P was higher in zero grazing units (0.42%) than in improved bomas (0.30%) or traditional bomas (0.24%); higher with a full roof (0.34%) than with a partial roof (0.31%) or no roof (0.25%); higher when concentrates were fed (0.31%) than when not (0.28%); and higher when manure was stored in a pit or heap (0.31%) than when stored as deep litter (0.28%).
The inclusion of bedding significantly decreased the mineral-N concentration (420 mg/kg compared with 804 mg/kg without bedding) and significantly increased the C:N ratio (23.9 compared with 21.1 without bedding). Turning the heaps significantly increased the mineral-N concentration (667 mg/kg compared with 362 mg/kg without turning) and decreased the C:N ratio (21.5 compared with 24.9 without turning).
Results suggest that modification of traditional livestock housing (boma) to
the zero-grazing system may have beneficial effects on some aspects of manure
quality. It is important to note that these beneficial effects may arise as
an interaction between a number of livestock- and manure-management-factors
and that the analysis of main factors only, presented above, may have overlooked
these. However, in defence of this analytical approach, the aim of this study
was to identify simple management factors that have significant influence on
manure quality. Interacting factors may indeed influence quality but expressions
of these interrelationships
lend themselves to complex extension messages.
Similarities between the current and earlier surveys confirm that management factors have the greatest positive influence upon P content. Both surveys point to P content increasing as the result of feeding concentrates. This is an important finding given that P is considered the primary limiting nutrient in Kenya highland soils. No clear agreement was found between the two surveys regarding best practice for producing manures with high N concentration. The present results suggest that inclusion of bedding and turning affect the C:N ratio and N-mineralisation of the manures and this could have an impact on compost maturity and synchronisation of nutrient release with crop growth.
Although nutrient concentrations are valuable indicators of manure quality, these measurements do not reflect the total amount of nutrients that could be potentially available in the farms. It is quite possible that manures with low nutrient concentration could also have high heap mass, resulting in potentially higher nutrient cycling capability. The full impact of livestock and manure management practices on nutrient cycling can only be determined if mass balances are recorded. Section 3 reports on-station trials to assess the impact of management practices on total nutrient content.
