The evidence that farmers can accurately assess the quality of organic fertilisers and use this knowledge strategically is building. Garforth & Gregory (1997) document evidence for astute indigenous soil management knowledge from across the world. The complexity of assessing compost biomaturity in smallholder farming systems is attributed to the fact that the different farmers conceive this aspect differently. For instance, some farmers believe that a completely composted manure heap with a characteristic fungal smell is the best manure. Others may not judge the quality until the results are seen in the final crop yield obtained after application of such manures. It has been reported by Motavalli et al (1994), from a survey conducted in the semi-arid tropics of India, that farmers conceptualise farmyard manure quality in diverse ways. They judge the manure from the physical composition, which determines its workability and its effect on crop development, and edaphic and biotic factors. Mugwira & Murwira (1997), in a review of the use of cattle manure to improve soil fertility in Zimbabwe, report on the quality in terms of the nitrogen content, an aspect that farmers may not comprehend by mere visual observation of the physical appearance of the manure heaps.
In an earlier survey (Lekasi et al, 1998), farmers were asked how they knew what a good manure looked like. Thirty percent, 20% and 50% of farmers in the large, medium and small farm categories, respectively, said a good manure is one that is "fully decomposed". The remainder (i.e. the majority of farmers) said that the quality of the manure could only be known by applying it to a crop. It was concluded that whilst farmers are aware of the `ingredients' and methods involved in making good manures they did not display competence in assessing the quality of purchased manures or appreciating when a home-produced manure is ready for application.
Simple indices of manure quality are required that will enable farmers to combine manure more effectively with strategic quantities and placements of inorganic fertilisers and so more precisely meet the nutritional needs of crops. In this survey, links between manure quality in terms of nutrient composition and C:N ratio, and the manure texture, colour, smell and biological activity observed visually were investigated. This study sought to investigate the extent to which simple physical parameters could be used as indicators of manure nutrient concentration or quality. If suitable indicators are identified that are reliable, reproducible and applicable with minimum training, they could provide farmers with a simple decision tool to determine the quality (or maturity) of compost and to assess the approximate fertiliser value of their manure-compost. Such an evaluation would aid decision making on application rates and choice of type and quantity of inorganic fertiliser to use as a supplement to manure. The scope for using a decision tool to determine application time would be less, as this is constrained by the crop cycle rather than by manure maturity. The following parameters were assessed:
Manure texture. The hypothesis for this parameter was that undecomposed manure containing animal faeces and possibly a range of other organic additions would have a coarse texture. The texture should become finer as the decomposition process progresses, resulting, at maturity in the fine loamy material, which is recognised as the mature product from all types of organic composting.
Manure colour. The hypothesis for this parameter was that undecomposed material consisting of a heterogeneous mixture of animal faeces and other organic materials, differing in colour, would have a mottled appearance. As the decomposition process progresses, such material would be expected to become more homogeneous, appearing a uniform dark brown or black at maturity.
Manure smell. The hypothesis for this parameter was that fresh animal manure has a strong smell of ammonia and other organic matter also gives of strong smell of putrefaction during the early stages of decomposition. Later, ammonia is lost by volatilisation and ammonium salts are converted to odourless compounds, and the organic decomposition products generally have little smell. Mature compost is expected to have only a slight `earthy' and inoffensive smell.
Manure biological activity. This parameter was included in the survey speculatively with little qualification. It could thus be interpreted as the activity of macrofauna, such as earthworms and other detritivores, or as visible signs of decomposing microflora such as fungi. The fauna and flora of compost heaps changes with time, both increasing and decreasing with maturity depending on the group of organisms. For example earthworm activity might increase to a maximum and then decline towards maturity, while other soil fauna and fungi might well show peak activity at other times. It is thus not surprising that none of the chemical characteristics differed significantly with level of biological activity (Table 11). It can be concluded that this parameter is of little value in describing compost maturity without considerable further qualification.
Multi-factor analysis of variance was carried out using General Linear Model in Minitab to examine the relationship between the above simple manure characteristics and some of the nutrient characteristics reported in Section 2. Only total N, P and Min-N, and C:N ratio were tested. Tables 11 and 12 show the significance of these relationships. In addition, the relationships between the continuous variable, age, and the nutrient characteristics were examined by regression. Only the regression between age and C:N ratio was significant (p = 0.021 negative correlation, R2 = 0.02; Y (log10C:N ratio) = 6.037 - 1.23x (age in months)), confirming a decline in C:N ratio with manure age.
An attempt was made to utilise the above data to develop a decision tool to allow the assessment of manure quality from
simple physical parameters. The percentage nutrient concentration of the manure-compost is important in determining the total amount
of nutrients applied to a crop. However, the results in Section 4 indicate clearly that quality parameters rather than total
nutrient application rate may be the key factors in determining the crop response to applied manure. This is shown by the strong
negative correlation between C:N ratio and crop yield at iso-N applications for the experimental manures (Section 4). For this reason,
an attempt was made to develop a decision tool for only C:N ratio, a measure of manure quality and maturity, and Min-N, to
some extent a measure of compost maturity and immediate fertiliser value. Unfortunately the chemical analysis of the large sample
of farmers' manures did not include the factors such as lignin and polyphenol concentration, which proved significant in predicting
the field performance of manures when Maasai manure was included (Section 4).
This decision tool took the form of a dichotomous key or `decision tree'. C:N ratio and Min-N differed significantly with texture and age as single factors and these factors were used in the decision tool.
Figures 2 and 3 show examples of the keys produced. In Figure 2, the mean C:N ratio for 288 samples was 23.1. Separating the samples into two texture classes separated the mean values significantly into 21.9 and 24.4. No further division of the coarse to fairly coarse category was possible with any other physical characteristic, but the medium to very fine category was further sub-divided by age, with the mean values being 20.2 and 25.2 for the age classes ³ 5 months and £ 4 months. Figure 3 shows a similar key for predicting the Min-N (mg/kg) concentration of manures.
Figure 2. Dichotomous key for determining manure C:N ratio from simple physical characteristics
Figure 3. Dichotomous key for determining manure total mineral N (Min-N mg/kg) from simple physical characteristics
Thus (Figure 2), it is possible to separate manures by combined characteristics of age and consistency into mature (mean C:N = 20.2) and immature (mean C:N = 24.4 or 25.2). While this is not particularly impressive, it should be noted that the difference in field trials between iso-N applications of manure with C:N = 25 and those with C:N = 19 represented maize grain yield improvements of 113 and 216%, respectively, above the plots without manure in the first season after application, and 18 and 61%, respectively, in the second season after application.
Similarly, the prediction of mean Min-N of 750 mg/kg for the old, fine manure is 84 and 97% higher than for the old coarse and young manures, respectively, and could have a significant impact on crop response and/or on the decisions of application rate of manure and of possible mineral fertiliser supplementation (Figure 3). This suggests that even this level of differentiation may be a useful guide to manure quality.
Thus, some simple physical parameters, which are easily discernible, do show significant relationships to some manure chemical characteristics. There appears to be scope for the development of decision tools for manure-compost quality. However, some of the categories of the physical characteristics chosen appear not to be useful (e.g. general biological activity) or act to confound the analysis (e.g. uniform colour of materials at start and end of composting).
Although the results show some significant differences in mean manure nutrient concentrations and quality, these mean values mask an enormous range of values, making significant differences in means difficult to detect. For example, the C:N ratio of the 288 samples of manure ranged from 5.3 to 81.3, representing materials with C:N ratios similar to soil at the one extreme through to a material with a C:N ratio twice that of cereal straw or 40% that of sawdust at the other extreme. Thus, the significantly different mean C:N values of 20.2 and 25.2 in of Figure 2 have ranges of 6.1-62.1 and 5.3-52.5, respectively. It seems likely that preliminary sub-division of manure-composts into types by some means before assessment of the physical parameters tried so far will be necessary to refine the decision tool. Correlation of physical characteristics with other important chemical characteristics reported in Section 4 might also be productive.
