«Dimambro ME, Lillywhite RD & Rahn CR Warwick HRI, University of Warwick, Wellesbourne, Warwick, CV35 9EF Corresponding author: ...»
A randomised and blocked experiment was established in spring 2005. There were three replicates of each treatment; 63 plots of 18m * 4 m wide, 4536 m2 in total. The design incorporated three compost treatments: (1) the amount of compost required to supply 250 kg ha-1 N, (2) the amount of compost required to supply 500 kg ha-1 N and (3) the amount of compost required to supply 250 kg ha-1 N plus mineral fertiliser to supply an additional 125 kg ha-1 N as ammonium nitrate (NH4NO3). The actual amounts of compost applied is detailed in table 2.3. Incorporated within the design was a nitrogen fertiliser response trial with six levels of ammonium nitrate: 0, 42, 84, 125, 167 and 209 kg ha-1 N: No compost were applied to these plots. The plots which received no compost and no nitrogen fertiliser are referred to as the control plots. Mineral fertiliser was applied on 11th May 2005.
The composts were distributed evenly over each plot (10th March) and then incorporated (14th March) to a depth of 15-20 cm using a spading machine in a north-south direction. This was followed by a light spring tine used east-west to assist seedbed preparation. Barley (variety Optic) was drilled at a rate of 200 kg ha-1 in an east-west direction on 15th March 2005.
Herbicide (Duplosan KV 2.0 litres ha-1 and Alpha Briotril plus 19/19 0.5 litres ha-1) was applied at the end of April 2005.
2.3.2. Soil sampling Soil samples were taken on three occasions and analysed, as detailed in table 2.4. In each instance, cores to two depths (0-30 and 30-60cm) were taken, and samples were bulked together to provide one 0-30cm sample, and one 30-60cm sample per plot.
Pre-drilling nine samples per depth were taken from each block. Pre-top dressing and postharvest three samples per depth were taken from each plot.
Soil pH, moisture content, conductivity, mineral nitrogen, water soluble nutrients were determined from both the 0-30 and 30-60cm depths, using methods based on MAFF/ADAS (1986). Total carbon and nitrogen were analysed using a LECO® CN-2000. PTEs (aqua regia extractable) were analysed using the PAS 100 methods (BS EN 13650 and ISO/DIS 16772) in the top 0-30cm fraction only, in soils from all compost treatments, plus the zero fertilizer control treatment.
Soil sample analyses undertaken Soil sampling Analysis date Pre-drilling pH, conductivity, total C & N, mineral N 9th March 2005 water soluble nutrients, PTEs Pre-top dressing pH, moisture content, total C & N, mineral N, 5th May 2005 PTEs Post-harvest pH, conductivity, moisture content, total C & N, mineral N, 8th August 2005 water soluble nutrients, PTEs 2.3.3 Plant sampling Plant samples were taken prior to top dressing and at harvest.
Pre-top dressing The first sampling (9-11th May 2005) was undertaken using a 0.5m2 quadrat. Two samples were taken from 11 plots per block, representing one plot of each of the compost treatments.
All above ground biomass within the quadrat was harvested, fresh weight recorded. Following oven drying at 80oC to constant weight (5 days), the dry weight was recorded. The dried samples were analysed for total C and N using a LECO® CN-2000.
Harvest The barley was harvested on 4th August 2005. An area of 30m2 was taken from the centre of each plot. Total grain yield was measured, and sub-samples taken for chemical analysis.
Samples were oven dried at 70oC for 48h to measure moisture content. 1000 grain weight was measured. Grain was analysed for minerals (organic N, P, K, Ca, Mg, Na, Mn) using methods based on MAFF/ADAS (1986). In addition, PTEs were analysed in grain from all compost treatments, plus the zero fertilizer control treatment using PAS 100 methods (BS EN 13650 and ISO/DIS 16772).
Samples of straw were oven dried (80oC for 5 days) and minerals (organic N, P, K, Ca, Mg, Na, Mn) were analysed using methods based on MAFF/ADAS (1986).
2.3.4 Statistical analysis Two way ANOVA was used to compare treatments. Genstat® was used for all statistical analyses. Significant differences are expressed as probability (P) values, where a P value of
0.05 represents a significant difference.
3. RESULTS AND DISCUSSION OF COMPOST ANALYSISThe results of the batch 1 composts are presented and discussed in section 3.1. The results of the batch 2 composts are presented in section 3.2. In each section, the results and discussion
of the physical and chemical analyses are divided into nine sections, as summarised below:
1. pH and conductivity
2. Bulk density and moisture content
3. Particle size distribution
4. Physical contaminants
5. Water soluble carbohydrates, cellulose, lignin, organic matter and ash
6. Carbon and nitrogen content and the C:N ratio Nutrients: NO3-, NH4+, K, Ca, Mg, P, Fe, Zn, Mn, Cu, B, Na, Cl-, S 7.
8. Potentially toxic elements: Cd, Cr, Cu, Pb, Ni, Zn, Hg
9. Pathogens The results are discussed in relation to the BSI PAS 100 (WRAP 2005) and Apex standards (http://www.apexcompost.co.uk/standards.asp). More information regarding UK compost standards can be found in section 1.3.
3.1 Physical and chemical characterisation of batch 1 composts General observations The composts were black or brown in colour, with the exception of compost G, which was grey – light brown (Figure 3.1). Composts A to J had either no smell, or a faint, woody smell.
Compost K had a strong woody smell. Compost L had a very strong odour similar to silage.
Figure 3.1 Composts A, F, G, H, I and J 3.
1.1 pH and conductivity The pH of the composts varied between 5.1 and 8.7 (Table 3.1), with 11 out of 12 within the PAS 100 pH range of 7.0-8.7. pH influences the availability of nutrients to plants, and plants vary in their tolerance to pH. For cereals, the optimum soil pH is between 6.5 and 7.5 (Wibberley 1989). Using these composts as soil improvers would aid in neutralising acidic soils, and could reduce liming costs. Compost L was very acidic, with a pH of 5.07, which is an indication that this compost was not mature.
Electrical conductivity (EC) is a measure of soluble salt content. EC varies according to the number and type of ions in the solution. In this study, EC varied from 0.67 mS cm-1 (compost
C) to 3.32 mS cm-1 (compost L). The EC of composts G, K and L exceeded the Apex EC limit of 2 mS cm-1, potentially due to the high concentration of some water soluble salts, which could be from the food waste in the feedstock of these composts.
In systems where compost is used alone (as a growing medium) high EC can cause reductions in germination and growth. For example, high EC in mixed MSW compost (5.3 mS cm-1) caused growth inhibition in lettuce and cabbage (Brito 2001). EC levels above 9 mS cm-1 were found to be detrimental to tomato seedling emergence and development (Castillo et al.
2004). Thus it is important for composts to have an EC below the recommended limit when used as a growing medium. However, when composts are used as a soil conditioner in agriculture, they will be incorporated into the soil and thus diluted. Therefore, the EC of the soil will have a much greater effect on the crop than the EC of the compost. Indeed, in some saline and sodic soils, the soil EC could be higher than that of the compost.
3.1.2 Bulk density and moisture content
The bulk density of the composts ranged from 410 g/l (compost A) to 812g/l (compost F;
The Apex guidelines specify the range of bulk density to be 450-550g/l. Composts E and F exceeded this limit (779 and 812 g/l respectively). This can be explained by low moisture content and 75% of particles being smaller than 4mm. Composts which have only been screened down to 30 or 40mm, such as composts C and K, have bulk densities below the Apex range. This could be due to the distribution of large and small particles creating air spaces in the compost.
The moisture content of the composts varied considerably, from 20% (compost F) to 62% (compost B). This range is much greater than the Apex specification of 35-45%, and PAS 100 35-55%. Very wet compost can cause odour problems, while dry compost can be dusty and may require wetting before use. Moisture content will vary according to the feedstock, time of year, composting and storage conditions. Interestingly, composts B and F both contained fruit and vegetable waste (50% and 60-70%) and green waste (50% and 30-40%), which shows that although the feedstock can be similar, moisture content can still vary.
Composts with low bulk density and high dry matter content could be difficult to land spread since windy conditions would dissipate the material before incorporation.
3.1.3 Particle size distribution Particle size grading is an important aspect of compost specification, and will depend on the final use of the product. Coarse particles in composts designated as growing media may be unacceptable to growers. WRAP recommends a maximum screen size of 40mm for general use in arable agriculture, although where a fine seedbed is required, 25mm or 15mm is suggested (WRAP 2004).
The particle size distribution of the composts varied considerably (Figure 3.2), since the composts were screened to different sizes by the composters (Table 1.1). The finer composts (D, E, G, I, and J) had no particles greater than 15mm diameter. For horticulture and gardening, where small volumes of compost are used, consistent and small particle sizes are essential. Composts D, I and J are currently produced for use as a soil conditioner. They are used by landscapers and gardeners as well as for application to agricultural land.
In contrast, composts B, H and K had some particles greater than 30mm in diameter, generally woody material. Compost B is sold as a soil conditioner to landscapers and gardeners. Composts H and K are applied to agricultural land only. Whether the compost is purchased by landscapers and gardeners, or used by farmers, an even-sized and consistent product is more desirable to the end user.
3.1.4 Physical contaminants Physical contaminants are normally removed from composts prior to use. Indeed, most compost suppliers employ screening techniques to remove contaminants. However, despite this, all composts tested contained some physical contaminants or non-compostable elements (NCEs).
The presence and type of NCEs was diverse (Figure 3.3), and was greatly influenced by the type of feedstock. The composts produced from 100% and 72% MSW (G and L respectively) contained greater amounts of glass, metal and plastic than the source segregated composts.
Compost L consisted of over 22% contaminants, including 16% glass.
The 10 source segregated composts (A, B, C, D, E, F, H, I, J, K) contained a small proportion of stones ranging from 1.8% (compost A) to 7.3% (compost H). Glass was less than 0.2% and plastic less than 0.9% in all source segregated composts. No metal was found in the source segregated composts.
Physical contaminants, specifically glass and sharps are a major problem in compost quality.
Concerns over public health will severely reduce the use of mixed MSW composts.
Moreover, where over 10% of the compost is contaminants, it is a waste of energy and resources to transport and apply. Many consider that it is not possible to obtain good quality compost from mixed MSW (Schauner 1998). However, progress in the mechanical sorting of mixed MSW to extract the non-organic fraction is constantly improving. A study in France has indicated that composts of a similar quality can be produced from either source segregated BMW or mixed MSW (Morvan 2004).
3.1.5 Water soluble carbohydrates, cellulose, lignin, organic matter and ash
During composting, the biodegradable carbon sources (simple sugars and starch) are broken down by microorganisms; with water, carbon dioxide and heat being produced. Thus, water soluble carbohydrates (WSCs) decrease during the composting process as they are utilised by the microbial flora (Sánchez-Mondero et al. 1999). WSCs are more readily broken down than cellulose and lignin. Therefore, in mature compost, only small amounts of WSCs should be present, with higher concentrations of lignin and cellulose. This trend was apparent for the composts investigated in this study (Table 3.3).
Water soluble carbohydrates In general, WSC concentrations were relatively low, ranging from 0.09% (compost F) to 0.49% (compost H), with the exception of compost L, with a WSC content of 2.45% (Table 3.3). Garcia et al. (1992) compared the WSC concentration of six composts at different stages during composting. They found after 210 days WSC levels were reduced at least eightfold. As the WSC concentration of compost L was over five-fold higher than the other 11 composts, it is likely that this compost is not mature.
Cellulose and lignin The cellulose content was highest in the two mixed waste composts (G and L), which could be due to the high content of paper and cardboard in MSW. Lignin was highest in composts B and J (31% lignin), potentially due to wood in the green waste fraction of the feedstock. The low cellulose and lignin content in composts E and F (Table 3.3) could be due to the high content of fruit and vegetable waste in the feedstock.
Fruit and vegetables contain lower levels of cellulose and lignin than wood. A study comparing ten different vegetables found cellulose concentrations to be 9-23% and lignin concentrations 10-17% (Rahn and Lillywhite 2001). In contrast wood contains 40-50% cellulose and 20-35% lignin (www.paperonweb.com). Thus the proportions of vegetable waste and green waste (containing woody material) in the initial feedstock will influence the cellulose and lignin concentrations in the mature compost.
A UK study (Ward et al. 2005) showed that green waste compost composition varies with time of year. In addition, a study on source segregated composted BMW in the Netherlands found that the composition of BMW containing leaves, branches and grass varied greatly with time, and also location (Veeken & Hamelers 2002). Thus, the time of year that the BMW was collected may have influenced the cellulose and lignin content of the composts in the current study.