In an earlier survey (Lekasi et al, 1998) farmers were able to suggest many ways in which management of livestock and of manures might improve manure quality as opposed to quantity. Suggestions covered aspects such as better feed, capturing urine, mixing manures from different species, composting, storing in a covered pit, adding ash and inorganic fertiliser, adding green biomass, and roofing the cattle pen. In both the earlier survey and the survey described above it was difficult to ascribe manure quality differences to individual management practices. Furthermore, in both surveys, the associated manure analysis included only nutrient concentration. No data was collected on crop response to the different manures and no analyses of chemical parameters that might influence manure quality, other than C:N ratio were undertaken. There was thus a need to investigate the effect of manure management on manure quality under more controlled conditions and with the inclusion of crop response in field experiments designed specifically to compare manure qualities.
A maize response field trial was conducted at Kariti in Maragua District. Kariti soil is classified as a humic nitisol (FAO, 1997) with top soil (0-20 cm) composed of 31% sand, 56% clay and 13% silt, and described as having a clay soil texture. Soil pH (1:2.5 0.01M CaCl2) 6.02; total OC 0.71 %; total nitrogen 0.1%; available phosphorus by Bray P2 36.3 mg/kg ads; exchangeable bases, potassium 20 mg/100g soil, calcium 120 mg/ 100g soil, magnesium 28 mg/100g soil.
The soil contains inherently low total carbon hence organic matter and nitrogen according to guidelines given by Tekalign et al (1991). With 36 mg of available P/kg, the soil seems to contain sufficient P to maintain plant growth. Okalebo (1987) and Okalebo et al (1991), working on a similar type of soils, observed that 15 mg P/kg of Bray No.2 extractable P is the critical level below which responses are expected to occur. The site receives mean annual rainfall of 1300-1600 mm with mean annual temperature of 19.7-28.0 oC.
The experiment had eight treatments including the five experimentally-constituted manures described in Section 3.2. Two additional manures were also used for comparison purposes, Maasai manure, obtained from a Maasai boma (kraal) in Kajiado District and manure obtained from the farmer on whose farm the experiment was conducted. Maasai manure is of economic importance because it is widely used in central Kenya and yet is purchased at high prices (approximately Kshs 2000/t, 1UK£=100 Kshs) from as far as 150 km away in the Rift Valley Districts. Extensive use of Maasai manure demonstrates the extent of nutrient transfer from the drier lowland regions of the country most suited for ranching, to the arable land of the Central Highlands.
Farmer manure (FM) was provided by the owner of the farm from an 8-month old manure compost heap made from cattle manure that he had prepared to use in his crops that season. The cattle enclosure was partially roofed and poorly drained. The manure looked dark and well composted but with visible soil contamination, which may have occurred at the heaping stage during storage as it was being scraped from a deep litter soil floor. The storage heap was not covered and neither was it shaded.
The experimental design was a randomised complete block design with four replicates. With the exception of the farmer's manure, all manures were applied at a rate equivalent to 75 kg N/ha, evenly broadcast in the plots and then incorporated into the soil. Incorporation was done by digging the manure into the soil using a forked hoe and burying it as much as was possible. The farmer's manure was applied at the same rate as the Maasai manure, 13.7 t (fresh weight)/ha, but analysed after application, when N application rate was calculated to have been 121 kg N/ha. Maize (Zea mays) variety Pioneer 3452 was used as the test crop. Unfertilised plots were included as controls. The plot size was 4 x 6 m and maize was planted at a spacing of 30 cm (intra-row spacing) x 75 cm (inter-row spacing) giving a population of 43,000 plants/ha. Maize was planted in the first week of April 1998 and harvested in the third week of September 1998. Two seeds were planted per hill and thinned to one plant per hill 4 weeks after planting. Routine agronomic practices, such as weeding and pest control, were carried out according to the recommendations of extension staff.
At maturity, an area of 9 m2 in each plot comprising four middle rows was harvested and cobs and stover weighed. Sub-samples of ten randomly selected cobs and six plants were taken for moisture by oven drying at 65 oC for 72 h. Nutrient analyses were carried out using methods described by Anderson & Ingram (1996).
A second season trial was planted in the second week of November 1998 in order to study the residual effect of the manures with all agronomic practices being the same as in the first season except that no fresh manure was applied.
The rates of manure applied in the different treatments are shown in Table 6 together with the N and P application rates. The wide range of application rates required to provide 75 kg N/ha is a result mainly of the different moisture contents arising from the different manure management strategies.
Table 7 shows the maize grain and stover yields in the field trial. All of the manures except the farmer manure significantly improved grain yield in the first season compared with the unfertilised control, despite the higher N application rate with the farmer manure. Of the experimental manures, the greatest yield (F+FR, 4336 kg/ha) was significantly higher than the lowest yield (F+U, 2916 kg/ha). Of the experimental manures, all except F+U gave significantly higher grain yields than the farmer manure. Similarly, all of the manures except the farmer's own significantly improved stover yield compared with the unfertilised control. Stover yield with the best experimental manure (F+U+FR, 3805 kg/ha) was significantly higher than the lowest yield (F+U, 2648 kg/ha).
In the second season, due to prolonged drought, the crop struggled to reach maturity. Only Maasai manure, F+U+FR and F+FR showed significantly higher grain yields than the control, while Maasai manure, F+U+FR and F significantly increased stover yield. In the second season, there was no significant difference in grain or stover yields among the five experimental manures. The two-season overall grain yields were significantly higher than the control for all manures except the farmer's own, while the two-season overall stover yield was higher than the control for all manures except the farmer's own and F+U. With the two-season data F+U gave the lowest yield of grain and stover among the experimental manures. Yields were significantly higher with F+FR and F+U+FR respectively. There was no significant difference in harvest index between the unfertilised and fertilised crop nor among the manure types in either season. Harvest index ranged from 0.47 to 0.52 in the first season and 0.40 to 0.50 in the second season.
The five manures composted on-station had a known history of source and chemical composition of the constituents from which the manures were derived, but considerable variation in chemical properties of the finished product was observed, which could be attributed to the different manure management strategies during composting. Differences in manure quality, derived from the different collection practices, influenced crop response over two seasons, even when the manure was applied at the same rate of total N. A variety of parameters may have influenced the efficacy of the different manures, other than total N applied, which was controlled, and total P applied which was measured but did not correlate with yield performance. These factors could include, among others, the relative supply of other, unmeasured, macro- or micro-nutrients, effects on the soil physical chemical or biological properties, and chemical properties of the manures that influence nutrient mineralisation. In order to investigate possible factors, the mineralisation of N from the experimental manures was investigated and correlations were tested between some chemical characteristics of the manures and crop yield response.
Net nitrogen mineralisation, has been considered as a measure of nitrogen availability of organically bound N in soils. The amount of N mineralised or immobilised from manure and compost depends on soil mineralogy (Beckwith & Parsons, 1980), organic material chemical and physical characteristics (Castellanos & Pratt, 1981; Janssen, 1996) and environmental conditions (Adriano et al, 1974; Virgil & Kissel, 1995). A good manure should synchronise mineral N (Min-N) release and plant demand such that the peak Min-N release coincides with peak plant biomass development and hence peak N requirements (Myers et al, 1994).
The five experimental manures and Maasai manure were studied to determine the rate of net N mineralisation. Topsoil (0-20 cm) was obtained from the farm used for the maize field trial, described in Section 4.1, from a site with no prior history of fertiliser usage. The soils were air dried in a greenhouse and ground to pass through a sieve of 2 mm mesh openings. Fifty grams of soil were weighed into 200 ml plastic bottles with four replicates. Manure was applied at the rate of 10 mg N/50 g soil. Replicate samples were included with no manures added as controls. Water holding capacity of the soil had been predetermined by the method described by Anderson & Ingram (1996).
The bottles with soil and manure were stoppered and shaken on an end-to-end shaker for 15 minutes to ensure a homogenous mixture. They were removed and allowed to stand for one hour for the dust to settle before gently applying distilled water to 60% water holding capacity. The bottles were closed loosely so as to allow gaseous exchanged and yet maintain the same level of moisture. The bottles were incubated at 25 oC and the moisture was checked and adjusted weekly. At the end of 1, 2, 4, 8, 12 and 16 weeks of incubation, duplicates of each treatment and the controls were withdrawn for mineral N (NH3-N+NH4-N and NO3-N+NO2-N) analysis. 10 g of the moist soil was extracted for mineral N in 50 ml 0.5 M K2SO4 solution. Mineral N analysis was done by the colorimetric methods described by Anderson & Ingram (1996).
Min-N was obtained by summation of the cumulative (NO3+NO2-N) and NH4-N. All the manures showed net NH4-N release at 2 and 4 weeks after incubation, followed by a net decline up to week 16, with the exception of Maasai manure, S and F+U which showed a further net release at week 12.
Net cumulative NO3-N decline was observed for F alone and F+U throughout the study period and S manure only showed a net NO3-N release at week 12. Net cumulative NO3-N release was observed for Maasai manure and F+U+FR (week 4) and thereafter a net decline at weeks 8-16. F+FR showed a net decline for the first 2 weeks and a net N release at weeks 4, 8 and 12 followed by a net decline at week 16.
There was no net total mineral N (Min-N) release from any of the manures in the first week of incubation. Net Min-N release, was maintained by F+U+FR (weeks 2 and 4) F+FR (at weeks 4, 8 and 12), Maasai manure (week 4) and S (week 12). F and F+U showed no net Min-N release at all during the 16 weeks of the experiment. During storage these two types of manure formed a crust due to desiccation, covering the whole area exposed to the atmosphere. This crusting may have created anaerobic conditions in the manure heaps that reduced the composting process. Similar results have been reported by Castellanos & Pratt (1981), who observed immobilisation of N in soils treated with anaerobically composted dairy cattle and beef feedlot manures.
During sampling of manures F+U and F at the end of composting period, it was also observed that their physical characteristics were similar to freshly voided faeces. Compared with manures S, F+U+FR and F+FR, these two manure types also finished with the lowest and similar N concentrations of 1.59 and 1.6% after composting. This observation suggests that mixing of urine with faeces does not necessarily result in forms of organic N that are easily mineralised to available mineral N.
Thus, broadly speaking, the two manures, F and F+U, that performed badly in the field trials also showed no net Min-N release in aerobic laboratory incubation trials, while the other manures mineralised N at different rates and over different periods, but did all show net N mineralisation at some stage during the experiment. However, only a limited amount of information can be deduced from laboratory aerobic incubation studies and these alone were not enough to account for the influence of manure quality on crop yield when Maasai manure as well as the experimental manures was considered.
The manures were analysed to examine whether the manure management strategies affected a number of chemical characteristics that might influence the fertiliser quality of the final product, and whether these parameters were correlated with crop response. One chemical characteristic that is commonly used to define the quality of organic soil amendments is the C:N ratio because of its influence on organic N mineralisation. Other parameters that could be used to describe the quality of organic materials include lignin, polyphenols and NDF-N (Mellilo et al, 1982; Tian et al, 1992; Lekasi et al, 1999) as these compounds normally impose an effect on the rate of nutrient mineralisation, especially that of N.
Neutral detergent fibre nitrogen (NDF-N) was determined by the ANKOM method. Using neutral detergent solution comprising a mixture of sodium lauryl sulphate, ethylenediaminetetraacetic disodium salt, sodium teteraborate decahydrate, sodium phosphate dibasic and triethylene glycol, the neutral fibre was extracted, then dried in the oven at 65 oC and analysed for N using the Kjeldhal method. This N is referred to as the NDF-N. The NDF-N is always lower than the total N since during the extraction of the fibre part of the total N is lost.
The polyphenolics were analysed by the Folin-Denis method and included hydrolysable tannins and condensed tannins as well as non-tannin polyphenolics. This method is an adaptation from King & Heath (1967) and Allen et al (1974) and is fully described in Anderson & Ingram (1996). Lignin was analysed via the acid detergent fibre by boiling the manures with sulphuric acid solution of cetyltrimethyl ammonium bromide (CTAB) under controlled condition (Van Soest, 1963). The CTAB dissolves nearly all the nitrogenous constituents and the acid hydrolyses the starch to leave a residue containing lignin, cellulose and ash. Cellulose is destroyed by 72% sulphuric acid; lignin is then determined by weight-loss upon washing.
The major quality characteristics that were found to be significantly different (Table 8) were soluble C and N, total kjeldhal nitrogen (N%), C:N ratio, NDF-N, polyphenols and lignin. These observations suggest that the nature of the materials added to the compost heap does affect the quality of the cattle manure compost. For instance, F and F+U, which did not have organic materials added produced composted manures with lower N, NDFN and soluble C content compared with manures that had organic materials added to them, that is S, F+U+FR and F+FR. A similar trend can also be observed where the C:N ratio is considered as a measure of manure quality. F and F+U resulted in lower quality manures with higher C:N ratio compared with the lower C:N ratio observed with manures that had feed refusals added to them.
The importance of these parameters as indicators of manure quality has been shown further in the field trial described in Section 4.1. Hence it was observed that the manures with high N and NDF-N, for example, have lower C:N ratio and also resulted in higher maize yield.
Correlation coefficients were calculated for all parameters and field yield data for the five experimental manures. Table 9 shows only significant correlation coefficients between manure quality measurements and maize grain, stover and aboveground biomass dry matter production over the two seasons. The importance of the C:N ratio as an indicator of manure quality is suggested even in the second season where a significant negative correlation of total C:N ratio with grain yield was observed as in the first season.
Significant positive correlation was observed between NDF-N (grain) or lignin (grain, stover, total) and yield in the first season
and some of these correlations remained significant when combined data for seasons 1 and 2 were considered. As with C:N
ratio, C:NDF-N ratio can also be used to describe the quality of an organic material. Table 9 shows a significant negative
correlation indicating that maize grain yield increased with decrease in initial manure C:NDF-N ratio.
However, when the Maasai manure was included in the correlation tests somewhat different results were obtained which contradict the idea that C:N ratio is a key predictor for crop yield. The Maasai manure gave the highest grain and stover yield, despite appearing to be the manure of lowest quality in terms of N content and C:N ratio. With the low fresh weight application rates required because of the high dry matter content of this material, and the excellent crop response, it is no wonder that Maasai manure is a valued and sought after commodity for the small intensive farms. With Maasai manure included (Table 10), significant positive correlations were found between lignin concentration and grain, stover and total yield in the first season, and grain and total yield in the second season. Significant correlations were also found between lignin + polyphenols and lignin:N ratio for some yield parameters, while negative correlations were found between polyphenol concentration and the second season grain and total yield. C:N ratio was not correlated with yield.
The parameters lignin, polyphenol and lignin:N ratio were subject to multiple regression to examine their ability to predict crop yield in the field trial when both experimental and Maasai manures were included. The best multiple regression equations were:
1. First season grain yield (kg/ha) =
120 L - 866 PP - 54.2 L:N + 2550
nitrogen and other nutrients in such a manner that the synchrony with maize crop demand was achieved to the end of the cropping season.
Overall, the results suggest that manure lignin or NDF-N could be manipulated by varying the concentration in manure before application so as to synchronise N release with plant nutrient demands. Manures derived from forages containing high N bound in the form of lignin and NDF-N, that are able to maintain these high levels after composting are more likely to result in greater crop yields, not only in the immediate application season, but also for subsequent crops by controlled gradual release of nutrients, especially nitrogen. It is increasingly evident that N released in a slow manner may fit more closely to the requirements of growing plants than that from highly available sources (Brinton, 1985, Myers et al, 1994). The field results suggests that high lignin content of organic soil amendments, such as manure, should not always be viewed negatively as undesirable because it is associated with reduced nitrogen mineralisation in incubation studies (Palm & Sanchez, 1990; Myers et al, 1994).
It is vital to understand what lignin levels of manures should be termed as
undesirable and which organic material additions would lead to these undesirable
levels and therefore should be avoided when making manure composts. This means
that, considering only one parameter, for instance C:N ratio, as the sole manure
quality predictor is not advisable, especially when working at the farm level.
Consideration of several parameters singly or in combination would be a better option. However, the combinations and the
critical levels of these parameters to predict manure quality can only be achieved by realistic field experimentation in order to develop
user-friendly models for the desired crop types.
Of course the occurrence of significant correlation does establish a cause and effect relationship between lignin content and
crop performance. Maasai manure is also physically very different from the experimental manures and may possess other,
unmeasured chemical and physical attributes that influence its value. For example, Maasai manure is normally obtained dry and composed
of particles most of which can pass through 10 mm screen openings. This ensures that a high surface area comes into contact with
the soil for microbial activity compared with manures of bigger clods. This then leads to enhanced nutrient mineralisation.
