Project Title: Development of odour-free mushroom compost by modifying the organic and inorganic nitrogen sources and process technology Project Number: M 3d Project Leader: Ralph Noble Report: Final Report 2001 Previous Reports: Year 1 – Aug 1998, Year 2 - Aug 1999, Year 3 - Aug 2000 Key Workers: Ralph Noble, Phil Hobbs, Alun Morgan, Andreja Dobrovin-Pennington, Tom Misselbrook Location of Project: Horticulture Research International, Wellesbourne, Warwick, CV35 9EF Horticulture LINK Project 180 Institute of Grassland and Environmental Research, North Wyke, Okehampton, Devon, EX20 2SB Project Co-ordinator: Mr Peter Woad Blue Prince Mushrooms Ltd, Poling Arundel, West Sussex, BN18 9PY Date Project Commenced: 1 August 1997 Date Project Completed: 31 July 2001 Keywords: Mushrooms, Compost, Odour, Smell, Poultry Manure, Nitrogen Sources Whilst reports issued under the auspices of the HDC are prepared from the best available information, neither the authors or the HDC can accept any responsibility for inaccuracy or liability for loss, damage or injury from the application of any concept or procedure discussed The contents of this publication are strictly private to the Consortium of Horticulture LINK Project No. 180. No part of this publication may be copied or reproduced in any form or by any means without prior written permission of the Consortium. © 2001 Consortium of Horticulture LINK Project No. 180. Intellectual property rights are invested in the Consortium of Horticulture LINK Project No. 180. Consortium members: Horticulture Research International Institute of Grassland and Environmental Research Department for the Environment, Food and Rural Affairs Horticultural Development Council Middlebrook Mushrooms Ltd Blue Prince Mushrooms Ltd Shepherds Grove Ltd Tunnel Tech Ltd Pond Chase Nurseries Ltd Hensby Composts Ltd J. Rothwell & Son Ltd Shackleford Mushrooms Ltd Chesswood Produce Ltd Bulrush Peat Co Ltd Osmetech plc North Tamar Business Network 1 CONTENTS Page Consortium members 1 PRACTICAL SECTION FOR GROWERS Project objectives and targets Part 1: Alternative nitrogen sources Part 2: Odour quantification techniques Part 3: Microbial and chemical degradation of odours Action points for growers Project Deliverables 4 4 6 7 8 PROJECT MILESTONES 10 SCIENCE SECTION Part 1: Alternative Nitrogen Sources Introduction Materials and methods Bench-scale composting equipment Aerated bulk composting tunnels Windrow composting Compost analysis Odour analysis Mushroom cropping procedure Experiments Results Bench-scale flask composts Aerated tunnel composts Windrow composting Commercial farm tests Role of gypsum in composts Conclusions – Part 1 Figures and Tables – Part1 12 12 12 12 13 14 14 14 14 15 17 19 20 21 22 22 23 Part 2: Odour Quantification Techniques 39 Introduction 39 Materials and methods Odour sample collection 39 Composting yards and composts 40 Olfactometry 40 GC-MS analysis 41 Gas detector tubes 41 Electronic sulphide detectors 41 Aromascan electronic nose 42 Odour and sulphide concentrations on and around composting sites 43 Comparison of real and synthetic mushroom composting odours 43 Results Olfactometric analysis 43 2 GC-MS analysis Gas detector tube analysis Relationship between gas detector tube analysis and odour Compost analysis and type Performance of Aromascan electronic nose Performance of electronic sulphide detectors Odour concentrations on and around composting sites Validation of the odour/sulphide relationship Conclusions – Part 2 Figures and Tables – Part 2 43 44 44 45 45 46 46 47 47 49 Part 3: Microbial and Chemical Degradation of Sulphides in Compost Introduction Biodegradation of odours Strains Testing isolates on composting odours Testing systems for levels of sulphides and odour Effect of Hyphomicrobium strain and Biofilm on H2S production Chemical degradation of odours Conclusions - Part 3 61 61 Overall Conclusions 66 TECHNOLOGY TRANSFER Industrial relevance and plans for future commercial exploitation Plans for future R&D resulting from the project Publications and Presentations resulting from the project References 61 61 61 63 64 66 67 67 68 68 69 3 PRACTICAL SECTION FOR GROWERS Background Odour pollution is a major problem facing mushroom compost production in the UK and several other countries. Conventional composting involves wetting and mixing straw and animal manures in heaps (pre-wetting) and then in long stacks (Phase I composting). During these stages fermentation is uncontrolled, resulting in the evolution of gaseous pollutants, causing environmentally unacceptable odour levels. Broiler poultry manure has largely replaced straw bedding horse manure as an integral part of most mushroom composting due to its low cost, high N content and ease of handling. However, supplies of poultry manure are declining and it also has a serious odour problem both on its own and when incorporated into compost. This project will investigate the use of alternative N sources and straw types other than wheat and their effect on the availability of N, composting odours and subsequent mushroom yield. Project Objectives This project is divided into three objectives, with the overall objective of developing a quantifiable method of producing odour-free mushroom composts. The three objectives were to: 1. Develop alternatives to broiler poultry manure as a nitrogen source in mushroom compost with a view to reducing the reducing the use of poultry manure by at least 50%. 2. Develop objective methods for quantifying mushroom composting odours which relate to odour panel assessments 3. Develop microbial inocula and chemical treatments which degrade the odourants produced during composting compared with uninoculated composts. The project also aimed to integrate experimental odour-free composting processes and methods of odour quantification in a commercial scale composting system (Objective 4). Summary of results Part 1: Alternative Nitrogen Sources Poultry and horse manures are the main sources of sulphur in the generation of odourous sulphides from composting. Experiments had the following objectives: to determine if mixtures of organic and inorganic N sources could be used to replace or substitute poultry manure to determine the effect of replacing wheat straw with other straw types (rape, bean and linseed) on composting odours and compost quality to examine the replacement of poultry manure with alternative N sources in aerated and conventional windrow composting systems to examine the performance of the composts and composting methods on commercial sites. 4 Initial experiments using small-scale aerated flasks for composting identified a number of compost ingredients which could be used in large-scale experiments. These were: spent hop waste, cocoa meal, molasses waste (AminoPro), and rape straw (in place of wheat straw). Degradation using inorganic nitrogen sources (urea or ammonium sulphate) was slower than with poultry manure. However, the nitrogen from urea was readily available. Rape straw had a higher nitrogen content than wheat straw and required a lower inclusion of poultry manure, the main cause of mushroom composting odour. Replacing wheat straw with rape straw resulted in a significant reduction in odour in both windrow and aerated tunnel composts without affecting compost density, but wheat straw produced a higher yield (see Summary Table). Mushroom yield from bean or linseed straw composts were lower than from wheat or rape straw composts. Following farm studies, rape straw is now used at 20% inclusion rate by one of the commercial partners (Tunnel Tech Ltd) (Summary Table). Substituting 50% of poultry manure N with cocoa meal or urea in large-scale experiments reduced mushroom yield, although cocoa meal was better than urea. Hop waste as the sole N sources produced a good mushroom yield (241 kg/tonne) when the initial compost N was less than 2% of dry matter. Using inorganic N sources (urea or ammonium sulphate) resulted in lower compost bulk density. Substituting poultry manure by 50% with organic (spent hop powder, cocoa meal) or inorganic (ammonium sulphate or urea) nitrogen sources resulted in significant reductions in odour and sulphide concentrations. In large-scale tests the release of nitrogen from cocoa meal was more delayed than with poultry manure. This meant that the material needed to be incorporated during the early stages of pre-wetting. Following farm studies, this method is now used successfully at one of the commercial partners (J. Rothwell & Son Ltd). Urea is used during pre-wetting at three commercial sites (Tunnel Tech Ltd, Hensby Composts Ltd and Blue Prince Mushrooms Ltd) (see Summary Table). The economics of using different N sources depend on locality and transport costs: cocoa meal is available in the north of England and Wales, spent hop waste is available in Kent. Odour concentrations from windrow composts were higher than from aerated tunnel composts using similar composting materials (see Summary Table). The omission of gypsum (a major source of sulphur) from compost did not affect the emission of sulphides or odour, but resulted in a large reduction in mushroom yield. The S in gypsum is in an oxidised form (sulphate); the compost microbiota does not appear to be able to utilise this S to produce odourous reduced (sulphide) compounds. 5 Summary Table of results with alternative nitrogen sources (Project Part 1) Treatment Scale % reduction of standard compost* odour mushroom compost yield density 48 - 72 76 - 82 107 - 112 Rape straw + poultry manure Large experiment 20% rape, 80% wheat straw + poultry manure Farm studies 80 100 100 Wheat straw 50% cocoa meal + 50% urea Large experiment 14 73 100 50% cocoa meal or urea + 50% poultry manure Large experiment 37 - 61 73 - 79 100 100% hop waste or 50% hop waste + 50% urea Large experiment 13 - 39 89 100 25% urea (pre-wet) + 75% poultry manure Farm studies 100 100 25% urea, 25% cocoa meal (pre-wet) + 50% poultry m. Farm studies 100 100 Aeration Large expt. and farm st. 10 (prewet) 10 (prewet) 10 10 96 100 95 100 * Percentage compared with standard wheat straw + poultry manure or poultry manure / horse manure composts in conventional pre-wetting and Phase I windrows. Part 2: Odour Quantification Techniques Odour samples obtained from eleven composting sites showed there was a close correlation between the compost odour concentration measured by human panels (OC) of the pre-wet and Phase I compost air samples and the combined hydrogen sulphide and dimethyl sulphide concentrations using gas detector tubes.In order to measure sulphide concentrations of less than 0.1 ppm (100 ppb) eight electronic sulphide detectors were assessed. Only one instrument (a laboratory-based pulsed fluorescence sulphide analyser, manufactured by Thermo Environmental Instruments Ltd) responded to low sulphide concentrations in compost odour samples (see Summary Table below). Air samples were also obtained from five composting sites, at increasing distances downwind from the Phase I composting stacks, to the site boundary and beyond. The pulsed fluorescence analyser was found to be sensitive to sulphides in composting odours at 10 ppb. There was a good correlation between the instrument readings and 6 odour concentration. The sensitivity of the analyser enabled it to detect odour plumes at the boundary sites, about 50 m from the Phase I composting stacks. Synthetic pre-wet and Phase I odours were prepared from sulphides, ammonia and other odour compounds which closely simulated real composting odours when presented to a human odour panel. These results were consistent with the hypothesis that certain S compounds are mainly responsible for mushroom composting odour. Odour concentrations and sulphide concentrations from aerated composting systems were generally lower than those from non-aerated systems. There was no difference in odour or sulphide concentrations between poultry manure composts and compost prepared with horse and poultry manures. No relationships were found between compost analysis and composting odours. A tenfold dilution of the air from composting sites resulted in almost a threefold reduction in odour intensity. Summary Table of odour quantification techniques (Project Part 2) Method Human odour panel (Olfactometry) Advantages Relates to actual odours High sensitivity Disadvantages Not on-site High cost of measurements Partly subjective Gas detector tubes (sulphides) Cheap Not cross sensitive to ammonia or water vapour Only suitable for gases in concentrations > 0.1 ppm Pulsed fluorescence analyser High sensitivity (10 ppb) Not cross sensitive to ammonia or water vapour Can be used on-site High cost (£10,000+) Electronic nose or electronic sulphide detectors Cross sensitive to ammonia and moisture Low sensitivity to sulphides Part 3: Microbial and Chemical Degradation of Composting Odours Ten bacterial isolates were obtained from mushroom compost which were able to remove odourous sulphur containing compounds from compost air. The bacterial isolates belong to the following species: Pseudomonas putida, Pseudomonas fluorescens, Bacillus cereus/thurigiensis and Hyphomicrobium spp. Several compost systems were compared for testing the use of the bacteria in removing hydrogen sulphide and dimethyl sulphide from compost air. Ferric (iron) sulphate solution was more effective in removing sulphides and anaerobic odours from mushroom composts than microbial inocula (see Summary Table oppposite). 7 There was a distinct step in oxygen concentration (5%) below which anaerobic odours and sulphides developed. Above this threshold concentration, strong odours were prevented and no sulphides were detected (see Figure below). Scatter plot of H2S vs O2 levels in mushroom composts contained in incubated flasks 10 02 level (%) 8 6 4 2 0 0 50 100 150 200 H2S level (ppm) Summary Table of microbial and chemical degradation of odours (Project Part 3) Method Biofiltration Hyphomicrobium spp. Advantages Could be used on a biofilter in removing sulphides; Does not require frequent changes Chemical scrubbing Ferric sulphate solution Effective at normal composting temperatures and pH Not affected by compost ammonia Disadvantages Cost of biofilter Microbes sensitive to high composting temperature, ammonia and pH Biofilter requires ammonia pre-washing Requires replacement of ferric sulphate at intervals Action Points for Growers 1. Sulphides appear to be the main cause of mushroom composting odours, and there is a good correlation between odour and sulphide concentration in the air of composting sites. Gas detection (Draeger) tubes can be used to measure the two main sulphides (hydrogen sulphide and dimethyl sulphide) at concentrations above 0.1 ppm. A pulsed fluorescence analyser can be used to measure sulphide concentrations as low as 10 ppb on site boundaries. 8 2. Replacing wheat straw with rape straw reduces odour emissions without affecting mushroom compost bulk density, but mushroom yield is lower if rape straw is used as a 100% replacement for wheat straw. Inclusion of some rape straw (20%) in composts reduces odours by preventing anaerobic pockets developing in the compost, without affecting mushroom yield or compost density or costs. 3. Replacing poultry manure with organic nitrogen sources (spent hop waste or cocoa meal) also reduces compost odours without affecting compost density. Mushroom yield is lower than with 100% poultry composts. Cocoa meal has been shown to a suitable low odour additive during pre-wetting in farm studies since it reduces odours without affecting compost quality. 4. Replacing poultry manure with inorganic nitrogen sources reduces mushroom yield and compost bulk density although urea is better than ammonium sulphate. Urea was successfully used in farm studies during pre-wetting in that odours were reduced but compost quality was not affected. 5. The omission of gypsum (a major source of sulphur) from compost did not affect the emission of sulphides or odour, but resulted in a large reduction in mushroom yield. 6. Hyphomicrobium spp. and 10 other isolates were successfully used as biofilters in removing sulphides in the laboratory. The filter did not require frequent changes. However, on a commercial scale, the cost of the biofilter, its requirement for ammonia pre-washing and sensitivity to high composting temperature, ammonia and pH are disadvantages. 7. Ferric (iron) sulphate solution was more effective than microbial treatment in removing sulphides and anaerobic odours from mushroom composts than microbial inocula. 8. Aeration of pre-wet and Phase I areas, more frequent turning of windrows and reduction in compost moisture all reduced anaerobic compost and emissions of odours. Maintaining a minimum of oxygen concentration of 5% was found to prevent the development of anaerobic odours. Project Deliverables Odours were reduced by over 90% using a combination of aeration and alternative N sources, although completely odour-free composts were not produced (Objective 1). New nitrogen sources which can be used as low-odour, poultry manure alternatives in mushroom compost include: urea, cocoa meal, cotton seed meal, spent hop waste and molasses waste. The first three have been shown to be commercially economic, are now in commercial use, and have the potential to meet public odour concerns. The use of rape straw in mushroom compost is successful in reducing compost odour (less dependency on supplies of wheat straw alone). Rape straw is now used commercially by one of the partners. Objective 2, to develop objective methods for quantifying mushroom composting odours which relate to odour panel assessments, was fully achieved. Identification of a close relationship between mushroom composting odours and sulphides, enables rapid and objective measurement of odour sources. A pulsed fluorescence analyser method for measuring low concentrations of sulphides on site boundaries was developed. 9 Chemical scrubbing of odours with ferric sulphate was found to be more effective than microbial treatment (Objective 3). Composting methods and methods for measuring sulphides were successfully integrated on 5 commercial sites (Objective 4). Unexpected Benefits Rape or bean straw could be used in organic mushroom compost, particularly if there was a shortage of organic wheat straw, or a shortage of conventional straw in some seasons. Cocoa meal and urea were found to be better than a proprietary compost activator, Sporavite, with a 70% cost saving in N sources by one of the project partners (J. Rothwell & Son Ltd). A citric acid by-product was found to be a suitable and cheaper source of gypsum for mushroom compost than agricultural gypsum, and is now used by three of the project partners (Hensby Composts Ltd, Blue Prince Mushrooms Ltd and Gateforth Park - Shepherds Grove Ltd). 10 Project Milestones Task Target Date Milestones 1/1 12 months Methodology to monitor and sample odours developed: Relationship between laboratory-based odour quantification techniques and subjective methods established (HRI, IGER, Aromascan plc) 1/2 1/3 12 months Organic waste N sources analysed and processed into odourfree forms (HRI, Bulrush, Holdsworthy Bioplant) 1/4 12 months Different N sources compared in laboratory-based composting processes; N balance in substrate and its biomass and mushrooms determined (HRI) 12 months Consortium to decide whether odour quantification techniques are sufficiently reliable to use in intermediate and large scale composting experiments or whether further laboratory development is needed. Decision on which N sources should be taken into intermediate scale facilities. 2/1 18 months If Task 1/1 completed on time, odour quantified in intermediate-scale facilities and bulk chamber composts using GC-MS, olfactometry and electronic nose (HRI-W, Aromascan, IGER) 2/2 24 months Composting temperature and aeration regimes developed for bulk chambers (HRI) 2/3 24 months If Tasks 1/1 and 1/2 were completed on time, quantitative methods for measuring smell at commercial sites will have been developed (Industrial partners a – f, HRI, IGER, Aromascan) 2/4 24 months If Task 2/3 was completed on time, odour of conventional composting will have been quantified (Industrial partners a – f, IGER, Aromascan) 2/5 24 months Evaluation of N sources in bulk chamber experiments completed (HRI) 24 months Consortium to decide on which temperature/aeration regime(s) and composting system(s) to be used for larger scale experiments, or whether further intermediate scale experiments are needed to develop the regimes. Decision on which N sources are suitable for larger scale experiments. 11 24 months Decision on whether odour quantification of conventional composting sites is reliable or whether further assessments or development of quantification techniques are needed. 3/1 30 months Influence of microbial inoculation on odour emissions and degradation and additives to enhance water retention determined in laboratory conditions. (HRI) 3/2 30 months Compost odours synthesised in the laboratory to test the parameters of the electronic nose (IGER). 30 months Consortium to assess the performance of microbial inocula in reducing odour levels. Decision on whether larger scale experiments with microbial inocula can commence. 3/3 36 months If Task 2/2 was completed on time, temperature and aeration regimes will have been evaluated at commercial sites (Industrial Partners a – f, HRI). 3/4 36 months Evaluation of N sources in commercial scale experiments completed (HRI, Bulrush, Holdsworthy Bioplant, Industrial Partners a – f). 36 months Consortium to compare conventional and experimental composting odour levels and to decide what levels of improvement are necessary in the experimental and commercial systems. Consortium to decide which further developments in N sources are required. 4/1 42 months If Tasks 3/1 was completed on time, the effect of microbial additives and methods for enhancing compost water retention in bulk chamber experiments will have been determined (HRI). 4/2 42 months Modified N sources further examined in laboratory experiments; methods for reducing N losses from compost determined (HRI). 46 months Consortium to decide on the format of the manual and preparation of remaining publications. 4/3 47 months Combined effect of new composting regimes, N sources, microbial additives and methods of enhancing water retention determined at commercial sites (Ind. Partners a-f; HRI) 4/4 48 months Production of a manual on odour assessment, compost preparation and N sources in compost (HRI, IGER and Industrial Partners) All the milestones have been achieved within the specified timescales. 12 SCIENCE SECTION Part 1: Alternative Nitrogen Sources Introduction Odour pollution is a major problem facing mushroom compost production in the UK and several other countries (Miller and Macauley, 1988; Derikx et al., 1990, Noble and Gaze, 1994). Conventional composting involves wetting and mixing straw and animal manures in heaps (pre-wetting) and then in long stacks (Phase I composting). During these stages fermentation is uncontrolled, resulting in the evolution of gaseous pollutants, causing environmentally unacceptable odour levels. The traditional ingredient of mushroom compost, straw bedding horse manure, provided both a source of carbon (C) and nitrogen (N). Due to difficulties in availability and variability in the material, compost formulations with wheat straw and various N sources were developed. These included other animal manures such as poultry, pig and bullock (Ross, 1968; Grabbe, 1974; Dawson, 1978), other organic wastes such as dried blood, cotton seed meal, brewery wastes, horn, whey powder, molasses and sewage-based products (Riethus, 1962; Delmas and Laborde, 1968; Smith and Spencer, 1997; Gerrits, 1988) and inorganic N sources such as ammonium nitrate and sulphate, calcium nitrate, urea and urea formaldehyde (Reithus, 1962; Bech and Rasmussen, 1968; Delmas and Laborde, 1968; MacCanna, 1968). Due to its low cost, high N content and ease of handling, broiler poultry manure is now an integral part of most mushroom composting in the UK and many other countries (Gerrits, 1988). However, due to subsidised use as non-fossil fuels in power stations, supplies of broiler chicken manure and turkey manure for mushroom composting have recently declined. In addition, poultry manure has a serious odour problem, both on its own and when incorporated into compost. This is mainly due to the sulphur-containing amino acids which are precursors of volatile, odourous sulphur compounds particularly under anaerobic composting conditions (Miller and Macauley, 1988). Wheat straw is preferred to other cereal straws such as rye, barley and oat since it loses its structure less during composting and it is widely available (Flegg & Loughton, 1961; Gerrits, 1988; Noble & Gaze, 1995). However, the use of other straw types such as rape, bean and linseed, which have higher nitrogen contents, may reduce the requirement for poultry manure, as well as increasing available straw supplies for organic production. In these experiments, the effects of different straw types and N sources on the availability of N, composting odours and subsequent mushroom yield were examined. Initial experimental work used a bench-scale flask composting system; subsequent experiments were conducted using large-scale turned windrow and aerated bulk tunnel composting systems. Materials and Methods Bench-scale composting equipment Substrate ingredients were composted in ‘Quickfit’ multiadapter flasks immersed in thermostatically controlled water baths, each holding two 10-litre flasks (Noble et al, 1997). The prepared ingredients (3 kg samples) were placed on a perforated stainless steel platform within each flask and the flasks immersed in the waterbaths such that 13 the water level was above the level of the enclosed substrate. Each flask was connected to ancillary equipment providing independent aeration of the compost. The oxygen concentration in the substrate was controlled regularly by adjusting the airflow through the compost in each flask within the range 0 - 16 litres kg-1 substrate h-1 by means of flow meters. The temperature of the substrate in the flasks was monitored with Squirrel multipoint temperature loggers (Grant Instruments Ltd, Cambridge, UK). For the first 48 h of the composting process, the thermostat of the waterbath was set at 48°C to allow a natural succession and gradual build-up of microorganisms. The substrate temperature was then increased to 72°C for 5 days, after which the substrate was re-mixed and the temperature reduced to 47°C for the remainder of the composting period, which was seven days, or prolonged until the air in the flask was clear of ammonia. An oxygen concentration of 11(± 1.5)% v/v was maintained in the substrate. Aerated bulk composting tunnels Bales of straw were wetted and formed into stacks using a compost turning machine. Further water was added to the straw in a separate turn to achieve a moisture content of 70%. After four days, 50% of the required N source was mixed into the stack; the remaining N source and gypsum at 30 kg tonne-1 fresh compost ingredients were mixed into the stack after a further two days. Water was added in a further three turns after day 4, to achieve a moisture content of 78%. Stack temperatures were monitored with platinum resistance sensors and data logger. The preparation time for the blended compost ingredients, before filling into the tunnels was seven days. Six aerated bulk composting tunnels at HRI Wellesbourne were used for the experiments. Compost was filled on to a slatted base in the tunnels, mounted above an air plenum through which a controlled flow of fresh and/or recirculated air could be blown. Two of the tunnels consisted of modified insulated cargo containers. Both of these tunnels had a vertical partition, which did not extend into the air plenum below, to enable two different composts to be filled into each tunnel (Type A). The other four tunnels consisted of insulated polythene tunnels, inside which were two parallel walls, joined by a wall at one end (Type B). The compost was enclosed by a removable end wall, which fitted across the sidewalls. Details of the tunnels, temperature, oxygen and air flow measurement and control, methods for filling and emptying the tunnel and methods for measuring ammonia concentrations are given in Noble & Gaze (1994 & 1998). The tunnel composting regime consisted of three stages, designated Phases 0, I and II. The tunnels were filled with 4 t batches of blended compost ingredients to a height of 1.5 m. The Type A tunnels were each filled with two 4 t batches of blended compost ingredients separated by a central partition. In the Type B tunnels the airflow was set at 9 m3h-1, unless the oxygen concentration in the compost fell below 6%, in which case the airflow was increased to 13 m3h-1 until the oxygen concentration was above 6%. The respective airflows in the Type A tunnels were 12 and 26 m3h-1. After five days (Phase 0), the compost was emptied from the tunnels, mixed and if necessary re-wetted to achieve a moisture content of 77%, and then refilled. The subsequent 6-day Phase I was similar to the Phase 0 regime. For the Phase II pasteurisation regime, the tunnels were filled with 2.5 t of material from the Phase I stage to a height of 0.9 – 1.1 depending on the ingredients. Following a 20 h equalisation of compost temperature at 45 - 48°C, the composts 14 were pasteurised at 58 - 60°C for 6 h. Compost temperatures were then reduced to 46 - 49°C (conditioning). A minimum oxygen concentration of 13% was maintained during Phase II. Composting was completed when the compost temperature had fallen to within 1°C of the air temperature and ammonia could no longer be detected in the compost. Details of temperature and airflow control during Phase II are given in Noble & Gaze (1998). Windrow composting Straw bales were formed into windrows and wetted on day 0 without a pre-wetting procedure. Nitrogen sources were added on day 2 and water applications and windrow turns were made on alternate days during a 16 day period. Water applications after day 10 were varied according to compost moisture content. Compost moisture content at filling of bulk pasteurisation tunnels was 76%. Both aerated and windrow composts were pasteurised at 58 C for 6 hours and conditioned at 45-48 C. Compost analysis Analyses were conducted on freeze-dried, finely milled samples of the compost ingredients and of the substrates before and after processing in the flask composting equipment or Phase II tunnels. Dry matter (DM), N, ammonium (NH+4) and ash contents and pH were determined as described previously (Noble and Gaze, 1994). Compost sulphur content was measured according to the method in Anon (1986). Odour analysis Odour samples were collected in 20 L Teflon bags as follows: (i) from flasks by evacuating the air in the flask (ii) from aerated tunnel composts, 0.2 m downwind of the compost during the emptying of the Phase 0 tunnel stage. (iii) from windrow composts, 0.2 m downwind of the stack during turning. The odour samples were then transported to IGER North Wyke and analysed within 24 h. Odour concentration (OC) was determined by an odour panel using dilution olfactometry and volatile organic compounds detected by gas chromatography – mass spectrometry (GC – MS). A Dräger Accuro bellows pump with appropriate detector tubes (Drägerwerk, Lübeck, Germany) was used for on-site measurement of ammonia and specific sulphides, in the same way as sampling odours for collection in Teflon bags. Two replicate measurements were made for each sampling. Mushroom cropping procedure (i) Bench-scale flask composts At the end of the composting period, the material in each flask was weighed. After samples were taken for analysis, 2 kg of the residual material was inoculated with mushroom spawn (spawned) at two percent of the fresh weight of compost with Agaricus bisporus spawn (Hauser A15) and filled into plastic pots, 230 mm diameter x 220 mm depth. The pots were placed in polythene bags in an incubator at 25°C and when the substrate was fully colonised with mushroom mycelium, about 15 days after 15 spawning, the containers were cased with a moist mixture of peat and sugar beet lime (900 g). When mushroom mycelium was visible on the surface of the casing, the containers were transferred to a controlled environment chamber with an air temperature of 18°C, relative humidity of 90% and a CO2 concentration of 0.1% to induce fruiting. Mushrooms were harvested daily over a 30 day period (cap diameter 25-30 mm). (ii) Large-scale, aerated tunnel and windrow composts The cropping procedure is outlined in Noble & Gaze (1998). The composts were spawned using the Hauser A15 (Sylvan Spawn Ltd, Peterborough, UK) and 2100 (Amycel-UK Ltd, Burton-on-Trent, UK) strains. Half the compost spawned with each strain was supplemented with the soya meal-based “Betamyl 1000” (Sylvan Spawn Ltd) at a rate of 1% of compost fresh weight. Spawned trays were stacked four high in cropping sheds, with four replicate trays of each spawn and supplement sub-treatment from each of eight compost treatments (128 trays per shed with 16 trays from each of the compost treatments). A split-plot design was used with compost treatments allocated to main plots, which were arranged in a Latin square design, and spawn/supplement treatments allocated to sub-plots within each main plot. Mushrooms were picked as large buttons (diameter 30-40mm) over a 24-days period (three flushes of mushrooms). The yields of each run were analysed separately and the means of each treatment were incorporated into the analysis structure of the tunnel composting stage of the experiment. Percentage dry matter content of each batch (treatment and run) of mushrooms from the first and second flushes was calculated from the fresh weight of 20 mushrooms, and the dry weight after oven drying (Burton & Noble, 1993). The N and ammonium (NH4+) contents of mushrooms from the first and second flush were determined on freeze-dried samples of 20 mushrooms from each batch (Noble & Gaze, 1994). (iii) Commercial farm tests Samples of compost were transported in bulk bags to tray and shelf farms. The compost was spawned with the same strain used on the farm for standard compost, and received the same cultural treatment. The commercial trays contained 200 kg of compost. Bags were also filled with samples of compost, spawned with same strains used on the commercial farms. Batches of 20 bags, each containing 20 kg compost, were sent to the commercial farms. Tests with alternative nitrogen sources were conducted at one of the commercial sites using a 100 tonne aerated composting tunnel. Experiments Bench-scale flask composts Experiment 1: Effect of compost N from poultry manure on mushroom yield Compost N content was varied by altering the ratio of broiler poultry manure to wheat straw, taking into account the N and moisture contents of the poultry manure and straw (Noble and Gaze, 1994). Water was first added to the straw to achieve a moisture content of 70% before the addition of poultry manure. Gypsum was added at 30 g kg-1 fresh compost ingredients. Further water was added over a period of 4 16 days to achieve a moisture content of 78% (±1.5%), before the substrate ingredients were filled into the flasks. Compost ingredients were prepared with six target N contents varying between 1.1 and 2.6% of dry matter. Six composts were prepared with each target N content. A total of 36 substrates were prepared. Quantities of poultry manure used with wheat straw varied between 17 and 55% by weight. Experiment 2: Effect of substrate N from difference N sources on mushroom yield Compost ingredients were prepared from wheat (Triticum aestivum) straw and the following organic, inorganic and proprietary N sources: Broiler poultry manure, a mixture of droppings, feathers and sawdust (control) Brewers' grains (Everards Brewery Ltd, Narborough, Leicester, UK) Hop (spent) waste powder (English Hop Products Ltd, Tonbridge, Kent, UK) Digester waste, animal manure-based (Holdsworthy Bioplant, Devon, UK) Cocoa meal, extracted from bean shells (Cadbury Ltd, Chirk, Wrexham, UK) Cocoa shells, pulverised (Cadbury Ltd) Chipboard sawdust waste, containing urea formaldehyde resin (Pelican Fabrications Ltd, Northfleet, Kent, UK ) Ammonium sulphate Urea Sporavite: molassed fibrous meal with added urea (a proprietary mushroom compost activator produced by Amycel (UK) Ltd, Burton on Trent, UK) AminoPro: molasses waste liquid including inorganic N (marketed as a proprietary animal feed supplement by United Molasses, Burton on Trent, UK). Mycelium by-product ( a waste from citric acid production) Amide waste (from oilseed rape processing) Compost was also prepared from: Oil seed rape (Brassica napus) straw with poultry manure. The proportions of straw and N sources in the formulations were calculated on the basis of their N and dry matter contents using the formula in Noble & Gaze [1994] to achieve blended ingredients with N contents of 1.6, 2.1 and 2.6 % of DM. Analyses of the raw ingredients and the quantities used in the formulations are shown in Table 1. Each of the 12 compost formulations x 3 N content treatments had 3 replicates. Brewers' grains, cocoa shells and animal digester waste were not used at the highest N rate (2.6 % of DM) because the N content was too low (Table 1). Analyses of the composts at spawning are shown in Table 2. Aerated tunnel composting Experiment 3: Use of different compost nitrogen and sources and straw types The substrates were prepared from wheat (Triticum aestivum) and oilseed rape (Brassica napus) straws as the main carbon (C) source. Rape straw was only used with broiler poultry manure. Wheat straw was used with poultry manure (control treatment), cocoa meal, spent hop powder, ammonium sulphate and urea. The analysis of the raw ingredients and the quantities used in the formulations are shown in Table 3. The proportions of C and N sources in the formulations were calculated on the basis of their N and dry matter contents using the formula in Noble & Gaze (1994) to achieve blended ingredients with an N content of 1.8 and 2.3% of dry matter. The moisture content of the composts at filling of the Phase 0 tunnels was 77 - 78%. 17 Analyses of the composts at filling of the Phase II tunnels and at spawning are shown in Table 4. Windrow composting Experiment 4: Composting different straw types and N sources in windrows Substrates were prepared from wheat, oilseed rape, bean and linseed (flax) (Linum usitatissimum) straws. All the straw types were used with broiler poultry manure. Wheat straw was also used with three other organic N sources in combination with urea: cocoa meal, spent hop waste, and molasses waste (AminoPro). In the latter two formulations, the organic N sources and urea each provided 50% of the supplemented N. Linseed straw was used only once. The combinations of straw types, organic and inorganic N sources are shown in Table 7. Analyses of ingredients are shown in Table 1. Compost formulations were prepared with an initial N content of 2 % of DM. Composts were analysed, gaseous emissions measured with detector tubes, and mushroom yields and DM contents determined as previously described. The experiment was conducted as a series of three runs, with six tunnels (two Type A and four Type B). Role of gypsum in compost Since the 1930s, gypsum (calcium sulphate) has been a traditional ingredient in mushroom compost, partly as a flocculating agent to prevent 'greasiness' and partly to reduce the emission of ammonia from ammonium nitrogen. However, it also one of the main sources of sulphur in mushroom compost, but there is no information available regarding its role in the production of odourous S compounds. Experiments were conducted to determine whether: gypsum contributes to the emission of sulphides from mushroom compost? mushroom yield is influenced by the addition or rate of gypsum in compost? there are differences between sources of gypsum in terms of composting and mushroom yield? Anaerobic, straw-poultry manure composts with gypsum added at 0%, 0.5% and 3.5% w/w fresh weight were prepared in flasks. Windrow (6 tonne) stacks of compost were prepared with (1) agricultural gypsum added at 25 kg/tonne compost ingredients (2) citric acid by-product gypsum added at 25 kg dry gypsum (33 kg fresh material) /tonne compost ingredients (3) no gypsum. The experiment was repeated twice. Results Bench-scale flask composts Experiment 1: Effect of compost N from poultry manure on mushroom yield Increasing the N content of the compost ingredients from 1.1 to 2.6% of DM increased the duration to clear NH3 from the compost from 71 hours to 343 hours and increased the maximum NH3 concentration in the flasks from 200 to 840 ppm. This resulted in an increase in compost N losses from 352 to 3142 mg/kg. Final ammonium and dry matter content and pH of the composts were: (mean values) 0.09 (±0.03)% of DM, 27(±2.3)% and 7.4 (± 0.31)%. 18 There was considerable variation in the mushroom yields obtained from the small pots. The optimum compost N content at spawning in terms of mushroom yield was 2-2.5% of DM. This was equivalent to an initial compost N content of 1.5 to 1.8% of DM. Experiment 2: Effect of substrate N from difference N sources on mushroom yield Analysis of materials Analyses of the N sources used are shown in Table 1. The molasses waste (AminoPro), Croda amide waste and mycelium by-product had N contents above 3% of dry matter, although the molasses waste and mycelium by-product had higher moisture contents. The cocoa shells waste and brewers grains had lower N contents than poultry manure, cocoa meal waste or spent hop waste. The N content of hop haulm waste on a fresh weight basis was significantly lower than that of the other alternative N sources. Hemp straw had a similar N content to wheat straw. These latter materials were therefore not tested in the composting experiment. Compost DM at spawning was 27.0 ± 1.1%. Increasing the N rate in the compost increased the maximum NH3 concentration, and NH3 and N losses from compost (P<0.001) (Table 2). Ammonium contents of the composts at spawning increased with initial compost N rate (P<0.001). Maximum NH3 concentration and NH3 and N losses during composting were greater from Sporavite composts than from other compost formulations (Table 2). With urea, Sporavite and AminoPro composts, NH3 was an immediately evolved from the composts in the flasks, but emission whereas with other formulations, there was delay in the main evolution of NH3 until the compost temperature had been increased to 72°C (Figure 1). Maximum NH3 concentrations were higher from poultry manure, brewers' grains, hop waste and urea composts than from the remaining formulations (Table 2; Figure 1). Losses of NH3 and N from digester waste, chipboard waste, and cocoa meal or shells composts were lower than from other formulations. Composts prepared from digester waste, cocoa meal and shells, and chipboard waste had lower ammonium contents at spawning than poultry manure or urea composts; the highest levels were in ammonium sulphate, Sporavite and AminoPro composts (P<0.001). The ash contents at spawning of digester waste composts were higher than those of other formulations, except poultry manure composts. The pH values of cocoa meal or shells and Sporavite composts were higher than those of digester and chipboard wastes, brewers' grains, AminoPro and ammonium sulphate composts, with other formulations intermediate in pH (Table 2). Increasing N rate in the compost from 1.6 to 2.1 or 2.6% of DM resulted in a higher ash content at spawning but did not significantly affect pH (Table 2). Mushroom cropping. There was a significant interaction between the effects of N rate and N source on mushroom yield (P<0.001) (Table 3). Cocoa meal and Sporavite composts produced higher yields at the 1.6% N rate than at the higher rates, whereas hop waste and rape straw composts produced higher yields at the 2.1% rate than at the 1.6 % N rate. The 2.6% N rate produced either a lower yield than the 1.6 and 2.1% N rates or no mushrooms (cocoa shells and Sporavite composts), with the exception of rape straw + poultry manure compost, where there was no significant difference in yield between the 2.1 and 2.6% N rates. Mushroom yield results for the 2.6% N rate are therefore not shown. At the 1.6% N rate, the highest mushroom yields were obtained from cocoa meal and Sporavite composts (Table 3). The cocoa meal compost also produced the 19 highest yield in terms of numbers of mushrooms. At the 2.1 and 2.6% N rates, rape straw + poultry manure compost produced the highest yield, both in terms of weight and numbers of mushrooms. At all three N rates, yields from digester and chipboard wastes, and ammonium sulphate composts were lower than from the wheat straw + poultry manure composts. At the 1.6% N rate, yields from urea, hop waste and rape straw composts were lower than from the wheat straw + poultry manure compost. Cocoa shells compost produced only one mushroom per pot. Yields N-Rich and amide waste composts were also low and these treatments were therefore not repeated. Aerated tunnel composts Experiment 3: Use of different compost nitrogen and sources and straw types Analysis of composts Analyses of the materials used in the experiment are shown in Table 3. The cocoa meal and spent hop powder had similar N contents on a fresh weight basis to the poultry manure. Rape straw had a significantly higher N content than wheat straw. During composting, composts with organic N sources reached higher temperatures than those with inorganic N sources (ammonium sulphate or urea) (Table 4). Initial S contents of poultry manure composts (1.0–1.1% of DM) were intermediate between those of ammonium sulphate composts (1.8% of DM) and the other formulations (0.8–0.9% of DM). Of this S, about 0.8% could be attributed to the inclusion of gypsum. Losses of S during composting followed a similar pattern to the initial S content of treatments (Table 4) with S losses from ammonium sulphate composts greater than from poultry manure composts, which in turn lost more S than the other formulations (P<0.001). There was no significant effect of compost N rate on S losses. Maximum NH3 concentrations, NH3 and N losses were greater from the 2.3% N rate than from the 1.8% N rate composts (P<0.001). Maximum NH3 concentrations and losses were also higher from the poultry manure and urea composts than from the cocoa meal, hop waste, and ammonium sulphate formulations (P<0.01) (Figure 2). This corresponded to higher N contents at spawning in the latter composts (P<0.01). Ammonium contents of all composts at spawning were in the range 0.02–0.10% of DM. At spawning compost DM was not significantly different between treatments (26.9 ± 1.4%), but ash contents of poultry manure compost was higher that of urea compost. The pH of hop waste composts were higher, and those of ammonium sulphate composts lower than those of other composts (P<0.001). There was no significant effect of N rate on compost ash content, pH or bulk density at spawning. However, composts prepared with organic N sources had greater bulk densities at spawning than those prepared with inorganic N sources (urea or ammonium sulphate) (P<0.001). Mushroom cropping. Mushroom yield was higher from the wheat straw + poultry manure composts than from the rape straw + poultry manure and hop waste (1.8% N) composts, which produced higher (P<0.01) yields than the remaining formulations (Table 3). Hop waste compost with an initial N content of 1.8% of DM produced a higher yield than the 2.3% N rate. There were no other significant effects of N rate on mushroom yield. There were no significant differences in mushroom yield between the strains A15 and 2100, or between supplemented and unsupplemented compost. 20 Mushroom DM (mean 8.2 ± 0.4%), N (mean 5.8 ± 1.0% of DM), and ammonium (0.6 ± 0.2% of DM) contents were not significantly affected by the treatments (Table 7). Odour and gaseous emissions. The bag sample OC of wheat straw + poultry manure compost (2.3% N rate) was higher than that of the other composts (Table 5). The other poultry manure composts and hop waste compost (2.3% N rate) had higher OCs than the remaining formulations. With all the N sources, the 1.8% N rate produced a higher OC than the 2.3% N rate. Compounds found in concentrations exceeding their detection thresholds (Devos et al, 1990) were NH3, butanol and those containing S (Table 5). Several other compounds (mainly organic acids) were found at concentrations just above their detection thresholds, but were not significantly different between treatments (footnote to Table 4). The S-containing compound mainly responsible for exceeding its detection threshold in compost air was dimethyl sulphide (DMS), followed by H2S, except in ammonium sulphate composts where methanethiol (MeSH) predominated. Little or no sulphides were detected in the air from cocoa meal, hop waste or urea composts. In the other formulations, increasing the rate of the N source increased the emission of sulphides. The concentrations of NH3 and acetone were also higher from all the 2.3% N rate composts than from the 1.8% N rate composts, and butanol concentration was higher from the 2.3% N rate composts prepared with poultry manure or hop waste. There was a close correlation between the combined on-site concentration of H2S + DMS and the bag sample OC of compost air. There was no correlation between OC and total losses of S during composting (Table 4). Windrow composting Experiment 4: Use of different compost nitrogen and sources and straw types Composting process and compost analysis Composting process and compost analysis. Compost prepared from linseed straw was slow to degrade and maximum compost temperature was only 55°C. Subsequent compost bulk density and mushroom yield were very low (Table 8) and the treatment was therefore not repeated. Maximum temperatures during composting were not significantly different between the other treatments (mean 73 ± 3°C). Maximum NH3 concentrations from poultry manure composts (wheat, rape or bean straws) were higher than from the other wheat straw formulations (P<0.01)(Table 6; Fig. 3). Maximum NH3 concentration during composting was correlated positively with NH3 and N losses (r = 0.74 and 0.71; P<0.05), and negatively with compost N content at spawning (r = 0.68; P<0.05). Composts prepared with cocoa meal had higher N contents at spawning than the other formulations. Calculated N losses during composting of equivalent formulations in windrows (Table 6) and aerated tunnels (Table 4) were similar. However, maximum NH3 concentrations and calculated losses were higher for aerated tunnel composts with equivalent formulations. This was probably due to the dilution of NH3 concentrations during recording of open-air windrow composting. At spawning, all composts in Experiment 4 had ammonium contents in the range 0.03–0.11% of DM. Compost DM content at spawning was 27.1 ± 1.2% and was not significantly different between treatments. The ash content and pH of the AminoPro + urea compost was lower than that of the other composts except for the pH of rape straw 21 compost. The ash contents of windrow composts (Table 6) were higher than those of aerated tunnel composts (Table 4) or flask composts (Table 2) where equivalent formulations were used. This indicates that carbon losses and compost degradation were greater in windrow composting. Compost bulk densities of bean straw, hop waste + urea and AminoPro + urea composts were significantly lower than those of the other treatments. Bulk densities of windrow (Table 6) and aerated tunnel composts (Table 4) were similar. Mushroom cropping. Mushroom yield from wheat straw + poultry manure was significantly greater than that from other compost formulations (Table 8). AminoPro + urea produced a significantly lower mushroom yield than rape straw + poultry manure or hop waste + urea. In both Experiments 2 and 3, there was a correlation between the maximum NH3 concentration and mushroom yield obtained from different compost formulations (r = 0.61 and 0.68; P<0.05), indicating that availability and release of N for compost degradation is a factor in determining subsequent mushroom growth. This is not the only factor since NH3 was released more readily from urea-based composts than from poultry manure-based composts although mushroom yield was poorer. However, it was observed that mushroom mycelial growth on the casing layer over urea-based composts was more vigorous than on the other treatments. There were no significant differences in mushroom yield between supplemented and unsupplemented compost, but strain A15 produced a higher yield than 2100 (mean values 227 and 213 kg t-1. Mushroom dry matter content was not significantly different between treatments or different from that in aerated tunnels, Experiment 3 (mean 8.3 ± 0.5%) (Table 9). Odour and gaseous emissions. The bag sample odour concentrations (OCs) of poultry manure composts were significantly higher than those of compost formulations which did not include poultry manure (Table 6). The H2S and DMS concentrations of the wheat straw + poultry manure compost were significantly higher than those of the rape or bean straw + poultry manure composts, which in turn were higher than the concentrations from other treatments. As in Experiment 3, there was a correlation between the combined on-site concentration of H2S + DMS and the bag sample OC of compost air. Commercial farm tests All the experimental composts produced lower yields than the commercial standard composts. Poultry manure and poultry manure + cocoa meal performed better than cocoa meal + urea (Table 10). Cocoa meal, cotton seed meal and urea were used as pre-wetting materials in place of poultry manure at J. Rothwell & Son Ltd. At Hensby Composts Ltd, Blue Prince Mushrooms Ltd and Tunnel Tech Ltd, 50% of the poultry manure applied during pre-wetting was replaced with urea. In all cases, there was a large (>90%) reduction in odour during pre-wetting, with no effect on mushroom yield, quality or compost density. At Tunnel Tech Ltd, 20% of the wheat straw was replaced with rape straw. This reduced the formation of anaerobic pockets in the compost, and did not affect the quality or cost of the compost. 22 Role of gypsum in compost Since gypsum is a major source of S in mushroom compost, it may have been expected that the presence or absence of gypsum in compost would influence the emission of S compounds. Hydrogen sulphide and dimethyl sulphide concentrations in the anaerobic flasks after 24 and 72h exceeded 1000 ppm and there were no significant differences between treatments (composts with and without gypsum at 0.5 and 3.5 % w/w). There was no significant difference in odour, air sulphide or ammonia concentrations from the windrow composts with different gypsum treatments. There were no clear differences in compost analysis at Phase I or Phase II between treatments, except that the first replicate compost had higher moisture than the second Tables 11 and 12). The absence of gypsum resulted in a large yield reduction (about 90%) in both crops (Table 13). There were no significant differences in mushroom yield or compost analysis between the agricultural and citric acid by-product gypsum sources. Further windrow composting tests showed that there was no significant effect of gypsum inclusion rate in compost (0.5 to 3.5 %) on mushroom yield (Fig.4). The results indicate that gypsum is a necessary ingredient in mushroom compost (particularly if wet) but it has no significant effect on sulphide or odour emissions. The S in gypsum is in an oxidised form (sulphate); the compost microbiota does not appear to be able to utilise this S to produce odourous, reduced (sulphide) compounds. Conclusions - Part 1 1. 2. 3. 4. 5. 6. 7. 8. Replacing wheat straw with rape straw resulted in a significant reduction in odour in both windrow and aerated tunnel composts without affecting compost density, but mushroom yield was lower (see Summary Table below). Rape straw has a higher nitrogen content than wheat straw and required a lower inclusion of poultry manure, the main cause of mushroom composting odour. Mushroom yield and compost density from bean straw and linseed straw composts were lower than from wheat straw or rape straw composts. Substituting poultry manure by 50% with organic (spent hop waste, cocoa meal, or molasses waste) or inorganic (ammonium sulphate or urea) nitrogen sources resulted in significant reductions in odour and sulphide concentrations, but also reduced mushroom yield. However, cocoa meal and cotton seed meal were found to be suitable as low odour N sources during pre-wetting in farm studies (see Summary Table). Spent hop waste as the sole N source produced a good mushroom yield (241 kg/tonne) when the initial compost N was l.8% of dry matter (summary Table). Using inorganic N sources (urea or ammonium sulphate) resulted in lower compost bulk density and mushroom yield than poultry manure compost. Release of ammonia from urea is more rapid than from poultry manure, but from cocoa meal it is more delayed. This material was successfully used for pre-wetting in farm studies (summary Table). Odour concentrations from windrow composts were higher than from aerated tunnel composts using similar composting materials (Summary Table). The omission of gypsum from compost did not significantly affect the emission of sulphides or odours but resulted in a significant yield reduction. 23 9. There was no difference in mushroom yield between two types of gypsum (agricultural or citric acid production by-product). Summary table of results with alternative nitrogen sources (Project Part 1) Treatment Scale % reduction of standard compost* odour mushroom compost yield density 48 - 72 76 - 82 107 - 112 Rape straw + poultry manure Large experiment 20% rape, 80% wheat straw + poultry manure Farm studies 80 100 100 Wheat straw 50% cocoa meal + 50% urea Large experiment 14 73 100 50% cocoa meal or urea + 50% poultry manure Large experiment 37 - 61 73 - 79 100 100% hop waste or 50% hop waste + 50% urea Large experiment 13 - 39 89 100 25% urea (pre-wet) + 75% poultry manure Farm studies 10 (pre-wet) 100 100 25% urea, 25% cocoa meal (pre-wet) + 50% poultry m. Farm studies 10 (pre-wet) 100 100 Aeration Large expt. and farm st. 10 10 96 100 95 100 * Percentage compared with standard wheat straw + poultry manure or poultry manure / horse manure composts in conventional pre-wetting and Phase I windrows. 24 Table 1 Analysis of straw types and N sources used in the experiments and quantities used in the compost formulations in Experiments 1 and 2. Each value is the mean of 5 determinations (± S.D.) Straw typea or N source Rateb % w/w DM % Wheat straw 17-97 Rape straw Price, £c % of DM N NH+4 Ash t-1 kg-1N 88 (±1.4) 0.5 (±0.15) 0.05 (±0.007) 7 (±1.3) 36 8.18 57-73 85 (±2.6) 1.2 (±0.09) 0.04 (±0.008) 6 (±1.8) 30 2.94 Bean straw 72 88 (±2.1) 0.6 (±0.12) 0.01 (±0.006) 4 (±1.2) 35 6.63 Linseed straw 68 87 (±1.3) 0.7 (±0.08) 0.01 (±0.006) 4 (±1.1) 30 4.93 Poultry manure 23-49 69 (±6.8) 5.9 (±0.59) 0.98 (±0.361) 15 (±0.8) 3 0.07 Digester waste 40-73 27 (±1.6) 2.8 (±0.51) 1.03 (±0.110) 22 (±2.7) 0 0 Cocoa meal 29-50 93 (±2.5) 4.2 (±0.23) 0.07 (±0.004) 5 (±1.5) 50 1.28 Cocoa shells 50-75 98 (±0.7) 2.6 (±0.15) 0.12 (±0.002) 9 (±0.6) 25 0.98 Hop waste 38-60 90 (±2.0) 3.3 (±0.52) 0.05(±0.007) 8 (±0.5) 50 1.68 Brewers' grains 69-83 24 (±5.2) 2.8 (±0.30) 0.04 (±0.001) 5 (±0.9) 22 3.27 Chipboard waste 39-67 87 (±4.5) 3.1 (±0.58) 0.26 (±0.149) 1 (±0.2) 0 0 Amm. sulphate 5-12 100 21.2 27.27 - 235 1.11 Urea 3-5 100 46.7 0 - 260 0.56 AminoPro 15-26 61 (±2.2) 6.3 (±0.32) 3.37 (±0.053) 9 (±1.6) 70 1.82 Sporavite 23-44 75 (±1.9) 6.6 (±0.54) 2.30 (±0.099) 16 (±1.9) 150 3.03 a Wheat straw was used in combination with all the N sources, inclusion rates with individual N sources were (100 - N source) %; rape, bean and linseed straws were only used with poultry manure. b Original fresh weight in compost excluding added water and gypsum. c Average UK prices, excluding transport and tax (Farmers Weekly and companies listed in Experiment 1). 25 Table 2 Ammonia and nitrogen losses during flask composting, compost analysis at spawning, and mushroom yield in Experiment 2. Values for compost parameters are the means of three N rates (1.6, 2.1 and 2.6% of DM) and two replicate composts. Straw type and N source in compost During composting Compost at spawning Max. NH3 NH3 loss N loss conc., μL L-1 mg kg-1 mg kg-1 % of DM pH N NH4+ Ash Mushroom yield , g kg a numbers per pot -1 1.6%N 2.1%N 1.6%N 2.1%N Wheat straw + Poultry manure 339 [5.83] b 689 [6.54] b 385 [5.95] b 2.28 0.15 16 8.1 126 138 15 [3.9] c 16 [4.0] Digester waste 30 [3.42] 170 [5.14] 18 [2.93] 1.77 0.05 22 7.8 79 60 16 [4.1] 15 [4.0] 8 [2.08] 135 [4.91] 151 [5.02] 2.36 0.06 12 7.7 55 55 18 [4.2] 17 [4.2] 119 [4.78] 516 [6.25] 148 [5.00] 2.22 0.38 12 7.4 93 97 16 [4.1] 18 [4.3] Cocoa meal 17 [2.87] 67 [4.21] 77 [4.35] 2.42 0.07 10 8.4 199 94 24 [5.0] 14 [3.9] Cocoa shells 10 [2.33] 53 [3.97] 403 [6.00] 2.16 0.06 11 8.5 10 0 1 [1.3] 0 [0.6] Hop waste 103 [4.63] 544 [6.30] 78 [4.36] 2.54 0.10 13 8.0 89 132 13 [3.6] 23 [4.8] Sporavite 927 [6.83] 4538 [8.42] 2490 [7.82] 2.21 0.62 14 8.2 168 0 16 [4.0] 0 [0.6] AminoPro 80 [4.39] 897 [6.80] 698 [6.55] 2.72 0.56 13 7.8 118 97 10 [3.3] 16 [4.0] 309 [5.73] 1740 [7.46] 1315 [7.18] 1.85 0.17 12 7.9 89 108 11 [3.3] 18 [4.2] 78 [4.36] 428 [6.06] 444 [6.10] 2.41 0.50 9 6.9 74 98 14 [3.8] 16 [4.0] Chipboard waste Brewers' grains Urea Amm. sulphate N-Rich Amide waste 1 5 12 2.35 0.03 11 7.2 53 42 13 7 100 412 316 2.29 0.04 11 8.4 50 65 7 12 [7.50] 2.18 0.14 15 8.0 82 177 13 [2.18] 0.26 0.06 4 0.3 46 46 Rape straw + Poultry manure LSD (P < 0.05) a 651 [6.48] [1.18] 1261 [7.14] [1.24] 1812 [3.7] [1.0] 27 [5.0] [1.0] Mushroom yield expressed as g mushrooms per kg of spawned compost, in composts with initial N content of 1.6 and 2.1% of DM. Figures in square parentheses are loge transformations, shown next to back-transformed values. c Figures in square parentheses are square root transformations, shown next to back-transformed values. b 26 Table 3. Analysis of straw and N sources and quantities used in aerated tunnel composts. Analyses are means of three replicate samples. % w/wc N-rate % of DM DM% Ingredient 1.8 2.2 N N Wheat straw 100-N source 87 0.7 0 Rape straw 100-poultry manureb 84 1.2 0 0 Poultry manurea 34.5 40.4 71 4.9 Poultry manureb 31.0 39.1 - - Spent hop powder 36.6 52.7 89 3.8 0 Shell extracted cocoa meal 35.9 44.1 90 4.3 0 Ammonium sulphate 3.2 6.7 100 21.2 2 Urea 2.1 3.6 100 46.7 a rate with wheat straw b rate with rape straw c original fresh weight in compost excluding added water and gypsum, at compost N-rates of 1.8 and 2.1% 27 Table 4 Compost temperatures, ammonia and nitrogen losses during aerated tunnel composting, compost analysis at spawning, and mushroom yield in Experiment 3. Values are the means of two replicate composts. Compost formulation Straw type and N source N rate During composting a % of DM Compost at spawning Max. Max. NH3 NH3 loss N loss S loss % of DM temp. °C conc., μL L-1 mg kg-1 mg kg-1 mg kg-1 N Ash pH Mushroom BD b kg m-3 kg t-1 Wheat straw + 1.8 76.8 358 [5.88] d 721 [6.58] d 1043 [6.95] d 43 2.41 18 7.7 446 268 Poultry manure 2.3 76.8 881 [6.78] 4024 [8.30] 1525 [7.33] 145 2.66 23 7.7 458 273 Rape straw + 1.8 78.3 299 [5.70] 1772 [7.48] 1054 [6.96] 85 2.12 20 7.6 496 207 Poultry manure 2.3 78.3 876 [6.78] 4583 [8.43] 1604 [7.38] 48 2.35 19 7.6 508 206 Wheat straw + 1.8 80.9 14 [2.68] 62 [4.12] 22 [3.11] 21 3.01 16 7.6 445 108 Cocoa meal 2.3 77.9 77 [4.35] 196 [5.28] 133 [4.89] 19 3.15 12 7.6 470 107 Wheat straw + 1.8 76.9 29 [3.36] 308 [5.73] 889 [6.79] 17 2.60 17 8.0 507 241 Hop waste 2.3 81.4 104 [4.65] 578 [6.36] 1353 [7.21] 0 3.20 11 8.2 520 159 Wheat straw + 1.8 70.8 31 [3.45] 446 [6.10] 388 [5.96] 155 2.56 14 6.7 347 135 Amm. sulphate 2.3 67.8 176 [5.18] 706 [6.56] 461 [6.13] 169 2.79 15 6.4 359 85 Wheat straw + 1.8 70.3 191 [5.25] 2368 [7.77] 1686 [7.43] 49 1.61 15 7.8 397 122 Urea 2.3 74.3 322 [5.78] 4915 [8.50] 2276 [7.73] 36 2.20 11 7.6 397 120 [1.75] 38 0.66 9 0.5 81 16 LSD (P < 0.05) 8.1 [1.77] [1.11] a N content of blended compost ingredients before Phase 0 tunnel composting. b Bulk density. c Mushroom yield expressed as kg mushrooms per tonne of spawned compost, mean of strains A15 and 2100, supplemented and unsupplemented compost. d yieldc Figures in square parentheses are loge transformations, shown next to back-transformed values. 29 Table 5 Odour and gas (GC-MS) concentrations of bag samples from different aerated tunnel compost formulations in Experiment 3. Values are the means of two replicate composts and two bag samples per compost. Compost N Odour conc. Concentration, mg m-3 formulation rate a OU m-3 H 2S DMS MeSH b NH3 acetone ethanol butanol propanol c Wheat straw + 1.8 2216 [7.70] b 0.09 0.26 0.07 7 [1.97] d 0.74 [0.03] d 3.05 [1.23] d 2.58 [1.08] d 1.03 Poultry manure 2.3 6342 [8.76] 0.73 2.06 0.46 80 [4.38] 1.38 [0.07] 10.26 [2.36] 13.60 [2.64] 3.84 Rape straw + 1.8 2619 [7.87] 0.07 0.34 0.15 41 [3.73] 0.29 [-0.41] 2.92 [1.19] 1.94 [0.84] 0.28 Poultry manure 2.3 3584 [8.18] 0.49 0.52 0.33 107 [4.68] 1.25 [0.49] 12.65 [2.57] 13.52 [2.63] 0.15 Wheat straw + 1.8 488 [6.19] 0 0 0 9 [2.24] 0.21 [-0.53] 0.86 [0.21] 1.08 [0.38] 0.29 Cocoa meal 2.3 977 [6.89] 0 0 0 23 [3.16] 0.63 [0.01] 0.08 [-0.79] 0.83 [0.19] 0.10 Wheat straw + 1.8 854 [6.70] 0 0 0 2 [0.91] 1.07 [0.37] 2.07 [0.90] 0.23 [-0.51] 0.63 Hop waste 2.3 2970 [8.00] 0.05 0.13 0.09 50 [3.93] 1.60 [0.69] 1.66 [0.71] 8.16 [2.14] 0.45 Wheat straw + 1.8 699 [6.55] 0 0 0.12 12 [2.49] 0.34 [-0.34] 3.28 [1.30] 2.44 [1.04] 0 Amm. sulphate 2.3 1045 [6.95] 0.02 0.12 0.27 20 [3.00] 1.44 [0.59] 6.04 [1.86] 1.89 [0.82] 0.37 Wheat straw + 1.8 670 [6.51] 0 0 0 5 [1.71] 0.65 [0.03] 1.16 [0.43] 1.18 [0.44] 0 Urea 2.3 843 [6.74] 0 0.02 0 41 [3.71] 0.70 [0.07] 0.55 [-0.07] 0.88 [0.23] 0.15 [1.50] 0.11 1.15 0.35 [1.77] 2.47 0.03 0.006 0.002 LSD (P < 0.05) Detection threshold [rf] [3.11] [0.99] 4 34.67 [2.49] 54.95 1.51 6.03 a N content of DM of blended compost ingredients before Phase 0 tunnel composting. b methanethiol. c Other odourants identified with GC-MS which were not significantly different between treatments were (range in concentrations in mg m -3): acetic acid (0.37-1.043), butanoic acid (0.70-2.68), propanoic acid (0.14-0.74), pentanoic acid (0.63-2.48), methyl ethyl ketone (1.76-4.68), iso-propyl-alcohol( 0.28-0.79), dimethyl disulphide (0.10-0.19), methyl propanoic acid (0.33-1.07), phenol (0.29-0.51), 4 methyl phenol (0.14-0.33), methyl butanoic acid (0.66-4.92). Ammonia was measured with detector tubes. d Figures in square parentheses are loge transformations, shown next to back-transformed values. 30 Table 6 Odour concentration of bag samples, on-site sulphide and ammonia concentrations, N and ammonia losses during windrow and aerated Phase II composting, compost analysis at spawning, and mushroom yields and dry matter content from different windrow compost formulations in Experiment 4. Values are the means of three replicate composts, except the linseed straw compost which was not repeated. Compost formulation Inclusion rate (% w/w) Straw During composting Odour conc. a H2S OU m-3 N source DMS mg m-3 Compost at spawning Max. NH3 NH3 loss N loss % of DM conc., μL L-1 mg kg-1 mg kg-1 N Ash pH BDb kg m-3 Wheat (63) Poultry manure (37) 11950 [9.39] d 2.63 4.97 141 [4.95] d 796 [6.68] d 483 [6.18] d 2.85 25 7.8 463 Rape (67) Poultry manure (33) 5729 [8.65] 1.97 2.23 242 [5.49] 1366 [7.22] 1287 [7.16] 2.34 24 7.7 496 Bean (72) Poultry manure (28) 4189 [8.34] 1.06 1.84 141 [4.96] 1703 [7.44] 672 [6.51] 2.77 20 8.0 452 Linseed (68) Poultry manure (32) - - - 94 [4.54] 584 [6.37] 829 [6.72] 2.49 16 7.7 400 Wheat (78) Poultry manure (21) + 4458 [8.40] 0.43 0.07 50 [3.92] 534 [6.28] 75 [4.32] 2.73 24 7.9 488 7320 [8.90] 0.50 0.08 8 [2.16] 380 [5.94] 65 [4.17] 3.53 22 8.0 489 - Urea (1) Wheat (64) Poultry manure (18) + Cocoa meal (18) Wheat (78) Cocoa meal (21) + Urea (1) 1716 [7.45] 0.24 0.23 27 [3.31] 94 [4.54] 233 [5.45] 3.23 22 7.8 480 Wheat (77) Hop waste (22) + Urea (1) 1609 [7.38] 0.18 0.18 23 [3.14] 35 [3.55] 372 [5.92] 2.67 20 8.0 444 Wheat (84) AminoPro (15) + Urea (1) 1406 [7.25] 0 0 13 [2.62] 63 [4.15] 330 [5.80] 2.79 15 7.5 423 [0.94] 0.38 0.94 [2.62] 0.65 4 0.3 39 LSD (P < 0.05) [1.61] [2.95] a Figures in parentheses are percentage w/w inclusion rates, before the addition of water and gypsum. b Bulk density. c Mushroom yield expressed as kg mushrooms per tonne of spawned compost, mean of strains A15 and 2100 and supplemented and unsupplemented compost. d Figures in square parentheses are loge transformations, shown next to back-transformed values. 31 Table 7. Mushroom dry matter and nitrogen contents from aerated tunnel composts, strain A15, mean of flushes 1 and 2. Treatment N rate at Fill, % of DM Mushroom DM % Mushroom N % of DM Wheat straw + poultry manure 1.8 2.2 7.82 7.89 6.08 6.25 Rape straw + poultry manure 1.8 2.2 7.68 7.44 4.76 4.82 Ammonium sulphate 1.8 2.2 7.48 7.72 6.10 7.08 Urea 1.8 2.2 8.72 8.46 6.25 6.35 Cocoa waste 1.8 2.2 7.89 8.03 6.90 6.91 Spent hop powder 1.8 2.2 8.40 7.96 6.32 6.31 32 Table 8. Mushroom yields (kg t-1 spawned compost) from windrow composts with different straw types and N sources. Mean of 3 replicate crops Treatment* Strain Betamyl A15 2100 - + - + Poultry manure (control)* 288 289 277 276 Rape straw + poultry manure 228 241 208 227 Bean straw + poultry manure 222 220 198 200 Linseed straw + poultry manure 83 103 71 105 Cocoa meal + urea 205 204 204 195 Spent hop powder + urea 239 217 227 254 Poultry manure + urea 239 238 211 207 Cocoa meal + poultry manure 230 203 206 182 * Wheat straw was used unless stated 33 Table 9. Mushroom dry matter contents from windrow composts, mushroom strain A15 unsupplemented compost, mean of flushes 1 and 2 and 3 replicate runs. Treatment* Mushroom DM, % Poultry manure (control)* 7.83 Rape straw + poultry manure 8.81 Bean straw + poultry manure 8.24 Linseed straw + poultry manure 7.88 Cocoa meal + urea 8.24 Spent hop powder + urea 8.27 Poultry manure + urea 8.33 Cocoa meal + poultry manure 7.96 AminoPro + urea 7.83 *wheat straw was used unless stated 34 Table 10. Mushroom yields from experimental composts on commercial farms ___________________________________________________________________ Compost treatment Test site Mushroom strain Mushroom yield Experimental Commercial standard __________________________________________________________________ Poultry manure/ wheat straw h 512 15.4 kg/m2 17.9 kg/m2 Cocoa meal + urea h 512 13.6 kg/m2 17.9 kg/m2 Cocoa meal + poultry f 2100 170 kg/tonne 216 kg/tonne __________________________________________________________________ 35 Table 11. Effect of Gypsum Treatments on Phase I Compost Analysis Treatment Rep. No gypsum 1 2 1 2 1 2 Agri. gypsum Citric acid gypsum NH4+ Ash % of dry matter 2.20 0.30 13 2.21 0.15 15 1.93 0.04 15 2.14 0.16 15 1.79 0.13 12 2.04 0.19 11 N pH moisture % 79 78 78 76 78 78 8.2 7.9 7.8 8.7 7.9 8.2 Table 12. Effect of Gypsum Treatments on Phase II Compost Analysis Treatment Rep. No gypsum 1 2 1 2 1 2 Agri. gypsum Citric acid gypsum NH4+ Ash % of dry matter 2.21 0.19 14 2.53 0.07 22 2.00 0.07 14 2.46 0.06 23 1.93 0.19 12 2.55 0.08 24 N pH moisture % 75 69 75 74 75 67 8.0 7.9 8.5 8.0 8.3 7.8 Table 13. Effect of gypsum treatments on mushroom yield Treatment Replicate crop 1 2 No gypsum 31 177 Agricultural gypsum 210 324 Citric acid by-product gypsum 243 340 __________________________________________________________________________ 36 1 0 0 0 p o u lt r y m a n u r e r a p e s t r a w + p o u lt r y m a n u r e d ig e s t e rw a s t e c h ip b o a r d w a s t e LL -1 8 0 0 6 0 0 Ammonia, 4 0 0 2 0 0 0 1 0 0 0 0 1 0 0 2 0 0 4 0 0 5 0 0 h o p w a s t e b r e w e r s 'g r a in s c o c o a m e a l c o c o a s h e lls 8 0 0 LL -1 3 0 0 6 0 0 Ammonia, 4 0 0 2 0 0 0 LL -1 1 0 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 S p o r a v it e u r e a A m in o P r o a m m o n iu m s u lfa t e 8 0 0 6 0 0 Ammonia, 4 0 0 2 0 0 0 0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 c o m p o s tr e m ix e d 7 0 Compostemperature,C 6 0 5 0 0 1 0 0 2 0 0 3 0 0 T im e ,h 4 0 0 Fig.1 Ammonia concentrations and compost temperature in flasks with different compost N sources. All composts were based on wheat straw unless stated. 37 Fig. 2a Ammonia emissions during Phases 0 and l: poultry manure composts, N = 2.2% of DM at filling 1400 Rape straw Wheat straw 1200 Ammonia, ppm 1000 800 600 400 200 0 0 100 200 300 400 500 Fig. 2b Ammomia emissions during Phases 0 and l: inorganic N composts, N = 2.2% of DM at filling 1400 1200 urea ammonium sulphate Ammonia, ppm 1000 800 600 400 200 0 0 100 200 300 400 500 Fig. 2c Ammonia emissions during Phases 0 and l: organic N composts, N = 2.2% ofDM at fillin 1400 Cocoa waste Hop waste 1200 Ammonia, ppm 1000 800 600 400 200 0 0 100 200 38 Time, hours 300 400 500 Fig. 3a Ammonia emissions during windrow composting and Phase II. N = 2.2% of DM at filling of Phase II. 200 Poultry manure composts Wheat straw Rape straw Bean straw Ammonia, ppm 150 100 50 0 0 200 400 600 800 Time, hours Fig. 3b Ammonia emissions during windrow composting and Phase II. N = 2.2% of DM at filling of Phase II. 120 Alternative materials Cocoa waste + urea Poultry manure + urea Hop waste + urea 100 Ammonia, ppm 80 60 40 20 0 0 200 400 Time, hours 39 600 800 Fig.4 Effect of compost gypsum rate on mushroom yield 400 A12 A15 Mushroom yield, kg/tonne 350 300 A 12 A15 250 200 150 2.4 2.6 3.0 2.8 Compost gypsum rate, % dry weight basis 40 3.2 3.4 Part 2: Odour Quantification Techniques Introduction In order to assess the effectiveness of new composting methods in odour reduction, such as forced aeration and changes to the raw ingredients (Perrin and Gaze, 1987; Gerrits, 1989), techniques are needed for quantifying odour levels. Olfactometry is used to measure the concentration of odour in air through the use of a serial diluter, or an olfactometer, to present odourous air with odourless air dilutions to a panel of people. However, olfactometry is costly, time consuming, subject to error and incurs delays between sampling and measurement (Hobbs et al., 1995). Several studies have attempted to identify the individual compounds associated with compost odours using gas chromatography-mass spectrometry (GC-MS) (Miller and Macauley, 1988; Derikx et al., 1990; Duns et al., 1997). Compounds shown to be responsible for compost odour include amines, ammonia, organic acids and most importantly, sulphur containing compounds. However, GC-MS analysis and interpretation is impractical on a day-to-day basis due to the large numbers of compounds involved, cost and the delay between sampling and measurement. However, many of the important odourants can be measured quickly and cheaply using gas detection tubes (Anon, 1997, 1998). There has been research into the development of electronic nose systems for odour detection and measurement. These instruments employ an array of non-specific electronic gas sensors in an attempt to mimic the human olfactory system by utilisation of artificial intelligence (Gardner et al., 1992). Equipment has been developed for a number of applications, mostly in the food industry, but also for agricultural malodour applications (Persaud et al., 1996; Misselbrook et al., 1997). The AromaScan (AromaScan plc, Crewe, UK) instrument uses an array of conducting polymer sensors, which display reversible changes in electrical resistance when volatile molecules adsorb and desorb from their surface. Such instruments have been shown to successfully distinguish between odours from pig and chicken slurry (Hobbs et al., 1995) between odours from slurry from pigs fed with different diets (Byun et al., 1997) and between breath odours from healthy cows and cows with ketosis (Elliott-Martin et al., 1997). Several sulphide detector instruments are available which operate using different principles. The suitability of these instruments for detecting sulphides from mushroom composting or other organic wastes has not be reported. The objective of the present work was to analyse the odours from mushroom composting using the different methods outlined above, and to determine the relationships between these methods. Materials and Methods Odour sample collection Odour samples were collected in 20 L Teflon bags by placing the bag in a pressure vessel with a PTFE tubing sampling line connected and then evacuating the vessel, thus drawing sample air into the bag over a 4 min period (Hobbs et al., 1995). The open end of the sampling line was held 0.2 m downwind from the compost heap or stack, during turning. Replicate samples were collected simultaneously. Background samples were collected 200 m upwind of the composting sites. Samples were transported to IGER, North Wyke for GC-MS and olfactometry analysis. Two replicate samples were collected for each of the analyses, which were conducted 24 h after sampling. 41 Windspeed at the point of odour sampling was measured with a vane anemometer (Type 949079, airflow Developments Ltd, High Wycombe, UK). Composting yards and composts Odour samples were taken from eleven sites: Blue Prince, Chesswoods (Sussex), Gateforth Park, Hensbys, HRI Wellesbourne, Monaghan Middlebrook (Avon and Market Harborough), Pond Chase, Shepherds Grove, and Tunnel-Tech (North and South). Six of the sites produced a windrow Phase I compost, turned at two-daily intervals; four sites had an aerated tunnel Phase I and HRI Wellesbourne produced compost using both types of Phase I. With the exception of HRI, all the sites had separate pre-wet and Phase I areas. At Blue Prince, Chesswoods, HRI, Monaghan (Avon), Shepherds Grove and Tunnel-Tech (South), prewetting was conducted partially or completely in windrows and at Gateforth Park and Pond Chase, pre-wetting was conducted on an aerated pad. Pre-wetting was conducted in flat heaps on the remaining sites. All the sites used proportions of wheat straw, broiler poultry manure and gypsum, although the proportions of these materials and addition of other manures and additives differed between sites. The total durations of pre-wetting and Phase I composting were 7-14 d and 6-7 d. The duration of the combined pre-wet and Phase I windrows at HRI was 16 d. Phase I tunnel composts at HRI were pre-wetted for 7 d. The pre-wet areas were sampled 3-6 d after setting up of the heaps and the Phase I windrows or tunnels were samples 3-7 d after the start. The combined pre-wet and Phase I windrows at HRI were sampled after 12 d. All the composts were analysed for moisture, nitrogen and ammonium contents and pH according the methods in Noble and Gaze (1994). Olfactometry A dynamic dilution olfactometer (type DTM, Project Research, Amsterdam) was used according to recommendations in van den Berg (1992), ie a forced choice type presentation where six panellists were required to choose between two sniffing ports, one containing odourless air, and the other diluted, odourous air. Threshold values, at which 50% of the panel could just detect an odour, were determined and odour concentration (OC) expressed as Odour Units m-3 (OU m-3) air. A range of six dilutions was presented to the panellists in steps of ascending concentrations, each differing from the next by a factor of two and each range being presented twice. OC was calculated according to the Dravneiks and Prokov (1975) method. Measurements of the sensitivity of the odour panellists for each set of OC measurements was performed with 198.2 mg m-3 (60 ppm) butan-1-ol in nitrogen. Samples were also tested with a newer Olfactomat "C" Project Research, Amsterdam instrument, which operated on the same principles as the DTM instrument. The odour panellists also made an assessment of odour intensity of the samples according to Burton et al (1997). A range of dilutions differing by a factor of two each time were presented to each panellist a number of times through one known port of the olfactometer. The highest dilution was equivalent to the 75% odour threshold value. At each dilution, the panellists were required to indicate the subjective strength of the odour, according to the following scale: 0 No odour 1 Very faint odour 2 Faint odour 3 Distinct odour 4 Strong odour 5 Very strong odour 42 6 Extremely strong odour. The relationship between intensity and odour concentration was calculated by linear regression. GC-MS analysis Volatile compounds were preconcentrated from a 600 ml odour samples by adsorbtion onto silica (Orbo 52, Supelco Inc., Supelco Park, Bellefonte, PA, 16823-0048 USA) and carbon (Orbo 32) based adsorbents. The concentrated odourants were then thermally desorbed from the adsorbents into the GC-MS system for identification and quantification. Chromatographic retention time and mass spectral matching were used to confirm odourant identity. Quantification was performed by desorbing 8 μl of a standard odour identified in the preconcentrated headspace, from the adsorbent. A Hewlett Packard (hp) (hp Ltd, Heathside Park Road, Cheadle Heath, Stockport, Cheshire, UK) GC-MS system consisting of a 5890 II Series gas chromatograph and a 5972A mass selective detector (MSD II) was used for analysis. A 25 m fused silica (cross linked methyl siloxane) hp-1 column with an internal diameter (id) of 0.2 mm and a 0.34 μm film with a 1m deactivated fused silica guard column (0.25mm id) were used. The flow rate of the helium the eluting gas was 0.75 ml min-1. The Optic temperature programmable injector (Ai Cambridge Ltd, Pampisford, Cambridge, UK) was used to desorb headspace samples from the adsorbents and was initially at 30˚C and heated to 16˚C s-1 for 1 min. An electronic pressure controller was used to offset peak pressure broadening with increasing GC column temperature. The GC oven conditions were an initial temperature of 40˚C, then to 220˚C at 15˚C min-1 and remaining at 220˚C for 1 min. The GC-MS interface was at 280˚C. The mass spectrometer scanned from 35 to 250 mass units every 0.2 s to give responses in the ng range. Volatile organic compounds (VOCs) detected by the mass spectrometer were identified using a probability based matching algorithm and a NIST mass spectral library. Compounds were declared unknown if their matching probability was less than 80 (100 being a perfect match). Gas detector tubes A Dräger Accuro bellows pump (Drägerwerek, Lübeck, Germany) was used in conjunction with appropriate detector tubes: acetic acid (6722101), amines (8101061), NH3 (CH2051 and CH31901), carbon disulphide, CS2 (8101891), DMS (6728451), H2S (8101991 and 8101831), mercaptan (thiols) (6728981) and phenol (8101641). Detector tubes were used on-site in the same way as sampling odours for collection in Teflon bags, 24h after on-site sampling. In addition to the eleven sites where odour samples were collected, detector tube measurements were also conducted on two further sites: Shackleford Mushrooms and J. Rothwell & Son. Two replicate measurements were made for each sampling. Electronic sulphide detectors The following electrochemical sulphide detectors from the following companies were examined: Analox Sensor Technology Ltd, Stokesley, Cleveland, UK Draeger Ltd, Blyth, Northumberland, UK Gas Measurement Instruments Ltd, Renfrew, UK The remaining four instruments had differing modes of operation: 43 Graseby Dynamics Ltd, Watford, UK: ‘Environmental Vapour Monitor’ utilizing gas chromatography (GC) and ion mobility spectrometry. 'Chemcassette' hydrogen sulphide analyser. Electronic optical measurement of gas sensitive colorimetric paper tape (Type 7100 Zellweger Analytics Ltd, Bishops Stortford, Herts.,UK). PE Photovac, Norwalk, CT, USA: ‘Voyager’ portable gas chromatograph, utilizing GC and a photoionization detector (PID) Pulsed fluorescence sulphur dioxide (SO2) analyser (Type 43C, Thermo Environmental Instruments Ltd, Franklin, MA, USA) coupled to a sulphide to SO2 converter (Type 45C, Thermo Environmental Instruments Ltd). The analyser was supplied by Unicam Chromatography, Cambridge, UK. All except the latter of these instruments were portable, although data from the Graseby Dynamics and PE Photovac instruments needed to be downloaded to PC for observation and analysis. Specifications of the instruments tested are shown in Table 20. Odour samples were drawn into the analysers using an internal electric pump. Each analysis took about 2 minutes to complete. The analysers were also tested with gas samples containing known concentrations of three sulphides at concentrations from 10 to 1400 ppb: hydrogen sulphide, dimethyl sulphide and methanethiol. Aromascan electronic nose A commercially available electronic nose instrument (AromaScan Sample Station, A8S, AromaScan plc, Crewe, UK) was used. The instrument used an array of 32 conducting polypyrrole sensors. A small pump pulled air across the sensor array at approximately 100 ml/min and the change in resistance of individual sensors was recorded as ∆R/R, where ∆R was the change in resistance and R the base resistance of the sensor. The degree of response of each sensor to a given volatile depends upon the type of polymer used, so a pattern of resistance changes across the array was recorded for a particular odour. The instrument incorporated a reference line, which consisted of filtered ambient air of the same relative humidity as the odour sample, which was used to provide the base resistance for sensors. The odour samples were passed across the sensor array for 8s. An initial problem encountered was the high sensitivity of the sensors to ammonia in the odour samples, to the exclusion of a response to other volatiles in the sample. The odour samples were therefore first passed through a Nafion tube gas dryer (Perma Pure Inc., Toms River, New Jersey, USA). Nafion is a copolymer of tetrafluoroethylene (Teflon) and perflouro-3,6-dioxa-4-methyl-7-octane-sulphonic acid. It has selective retention for sulphur compounds, simple hydrocarbons and oxides (CO2, SOX, NOX) but does not retain ammonia, amines, polar organic compounds (alcohols and organic acids) and water vapour. The Nafion tubing (3 mm diameter) was fitted inside a 16 mm diameter PTFE shell tubing. The relative humidity of the air in the shell tubing was controlled to the relative humidity of the odour sample in the Nafion tubing, to avoid changes in the relative humidity of the odour sample. Nafion tube lengths of 0.1, 0.3 and 0.6 m were used. The odour samples were passed over the sensors for 6 min when the Nafion tubing was used. The electronic nose was tested with and without the Nafion tubing with known concentrations of the following gases in air: ammonia (NH3), dimethyl sulphide (DMS) and hydrogen sulphide (H2S). Tests were also conducted on control samples with known concentrations of the above gases with unknown concentrations of other compost odourants, obtained from flask composting equipment. 44 Odour and sulphide concentrations on and around composting sites Air samples were obtained from five composting sites, (Tunnel Tech South, Shepherds Grove, Gateforth Park, HRI Wellesbourne and Monaghan Middlebrook Avon) at increasing distances downwind from the Phase I composting stacks, to the site boundary and beyond. Odour samples were collected in 20 L nalophane bags. Samples were obtained downwind of the compost stacks during turning at distances of about 0.2, 5, 15, 30, 40 and 50 m. Replicate samples were collected simultaneously. Background samples were collected 200 m upwind of the composting sites. Samples were transported to IGER, North Wyke for analysis using gas chromatography - mass spectrometry (GC-MS), pulsed fluorescence sulphide detector and olfactometry as described previously. Two replicate samples were collected for each of the analyses, which were conducted 24 h after sampling. Comparison of real and synthetic mushroom compost odours In order to test the reliability of the chemical analysis of composting odours, synthetic odours of typical pre-wet and Phase I compost odours were prepared. These were prepared from the six compounds which exceeded their detection thresholds by the greatest order of magnitude in real pre-wet and Phase I odour samples. These compounds were dimethyl sulphide, H2S, butanoic acid, methanethiol, butanoic acid (in both the pre-wet and Phase I odours), indole (pre-wet only) and trimethylamine (Phase I only). Since ammonia occurred in all the compost odours, this was also added to both synthetic mixtures. The concentration of each chemical was the average concentration found in real compost odours. Synthetic and real odours were presented to an odour panel in a series of dilutions. Each mixture was presented twice. Synthetic odour mixtures were also prepared without the ammonia, fatty acids and amine removed in order to test the influence of sulphides with and without other odourants in the air sample. Results Olfactometric analysis There was considerable variation in odour concentration (OC), and the relative strengths of the OC of the pre-wet and Phase I areas differed between both sites and sampling dates on the same site. Whilst there was no significant difference in OC between the pre-wet and Phase I composting stages, aeration (sites D, E, G and K) significantly (P<0.001) reduced OC (Table 14). There was no evidence of this aeration effect varying between the two composting stages. GC-MS analysis Odourants detected more than once with GC-MS in the pre-wet and Phase I air bag samples are shown in Table 15. There was considerable variation in odourant concentrations, both between sites and within sites. Acetone, NH3, ethanol and butanol were detected in all the samples in Table 15. Compounds for which the mean concentrations exceeded their published olfactory detection thresholds were volatile fatty acids (VFAs), sulfur containing compounds, NH3, amines, indole and 4-methyl phenol. In decreasing order, H2S, DMS, butanoic acid, methanethiol and trimethylamine were found to exceed their detection thresholds by the greatest order of magnitude (x100 or greater). Trimethylamine and VFAs were generally at a higher concentration in the Phase I air samples than in the pre-wet air 45 samples; alcohols showed the reverse pattern although the mean concentrations were either below, or in the case of butanol close to their detection thresholds. Odourants detected only once in the pre-wet and Phase I air samples were: acetaldehyde, butanone, heptanol, methanol, 2-methyl propanal, 2-methyl propanol, 2-pentanone and p-xylene. Gas detector tubes analysis Aeration significantly reduced both H2S (P<0.001) and DMS (P<0.001) emissions (Table 16 and 17). After adjusting for this aeration effect, Phase I composting areas had significantly (P<0.05) higher H2S and DMS concentrations than the pre-wet areas (Table 16 and 17). Ammonia concentrations were significantly (P<0.01) higher from the Phase I areas than from the pre-wet areas (Table 18) but there was no significant effect of aeration on NH3 concentration. There were no significant interactions between the effects of aeration and stage of composting for H2S, DMS or NH3 concentrations. The concentration of H2S measured in nalophane bags after 24 h was, on average, 52% of that measured on-site. There were no significant reductions in the concentrations of DMS or NH3 after 24 h. Gas detector tube measurements of H2S and DMS in nalophane bags, 24 h after sampling, were closely correlated with GC-MS measurements made at the same time (r = 0.81, P<0.001 and 0.86, P<0.001 respectively, based on 42 d.f.). The values obtained for H2S and DMS with detector tubes were, on average, 13% and 7 % lower than those obtained with GC-MS. Thiols could only be detected with gas detector tubes at site I (0.2 μL L-1 and 0.5 μL -1 L for the pre-wet and Phase I areas) and acetic acid could only be detected in the Phase I area of site I (0.6 μL/L). CS2 and phenol could not be detected with the tubes previously listed on any of the sites. Values obtained from the amines + NH3 test corresponded with the NH3 concentration, ie additional amines were not detected with detector tubes. Relationship between gas detector tube measurements and odour concentration There was a very close correlation between the combined on-site concentrations of H2S + DMS and the bag sample odour concentration (OC) of pre-wet or Phase I odour samples (Fig. 8). The linear regression equation, with loge transformed data, was: loge OC = 7.601 + 0.934 loge (H2S + DMS + 0.375), r = 0.948 P<0.001 [1] where H2S and DMS were on-site concentrations (μL L-1) and OC was measured using bag samples (OU m-3). The correlation was higher than the individual correlations of OC with H2S concentration (r = 0.912; 42 d.f.) or DMS (r = 0.905; 42 d.f.). There was no effect on the fitted relationship of either stage of composting (pre-wet or Phase I) or aeration. Similarly, multiple regression using H2S and DMS as separate independent variables did not improve the goodness-of-fit. Concentrations of NH3 were above the detection threshold in 42 out of 44 odour sources, but were not significantly correlated with OC. The inclusion of NH3 concentration in the multiple regression model described above did not improve the goodness-of-fit. The CO2 concentrations of the bag samples were 403 to 1223 μL L-1 (mean 841 μL L1 ) for the pre-wet areas and 510 to 1489 μL L-1 (mean 786 μL L-1) for the Phase I areas. There were no relationships between the CO2 concentrations and any of the other gas concentrations or OC. Wind speed at the point of sampling ranged from 0 to 4.5 m s-1 (mean 2.4 m s-1), but there were no consistent relationships between wind speed and any of the odour or gas concentrations. 46 During the third year of the project , a new olfactometer was installed at IGER (Olfactomat "C"). Compost odour samples were presented to the panellists through both the original and new olfactometers. Both machines were calibrated with a known concentration of butanol gas. Although there was still a good correlation between sample odour and sulphide concentrations (Fig.9), the new machine gave a significantly higher odour concentration for samples of an equivalent sulphide concentration. This was probably due to the reduced metal pipe work in the new olfactometer, resulting in reduced absorption of sulphides, and hence higher odour concentrations to the panellists. This emphasises the need to specify the type of olfactometer being used to assess odour concentration of air samples containing sulphides. Compost analysis and type Both moisture and N contents of the Phase I composts were higher than those of the pre-wet composts, but the NH4+ contents and pH of the pre-wet and Phase I composts were similar (Table 19). Aeration slightly reduced the pH of compost (significant at P<0.05) but there were no other effects of aeration on compost analysis (Table 19). There was a weak positive correlation between the compost NH4+ content and the NH3 concentration (r = 0.40, 42 d.f., P<0.05). No other correlations between gas concentrations, the proportions of H2S and DMS, or OC, and compost analysis factors were found. There was no significant difference in the compost analyses between poultry manurebased composts and poultry and horse manure-based composts. These manure types had no significant effect on OC or gas concentrations and there were no interactions between the effects of manure type and aeration or composting stage on these concentrations. Performance of the Aromascan (Osmetech) electronic nose The results from the instrument are presented as the change in resistance of the 32 sensors relative to the base resistance in air of the same rh (∆R/R). This response is affected both by the chemical species present and their intensity. The relative response of the 32 sensors can also be presented as a normalised response to produce a ‘fingerprint’ which is characteristic of the chemical species adsorbed (Persaud et al., 1996). In this situation, the response of the sensors is expressed as a proportion of the response of the entire array of 32 sensors so that the intensity components are nearly eliminated. The duration that the odour samples pass over the sensors could be varied. Due to strong response of the sensors to the odour samples in the absence of the Nafion tubing, this period was restricted to 8 s so that ∆R/R did not generally exceed 15%. The average response of each sensor over this period was used. In the presence of the Nafion tubing, the response of the sensors to the odour samples was delayed and much weaker, so that the period of sensing could be extended to 6 min without ∆R/R exceeding 15%. The average response of each sensor over the final 10 s of this period was used. Ammonia response. All the sensors of the Aromascan responded positively to ammonia at concentrations below 2 ppm. The normalised responses of the sensors to 2 and 50 ppm NH3 were similar with sensor numbers 19 and 31 producing the strongest response. There was a positive relationship between the NH3 concentration in compost odour samples and the average sensor response during the first 8 s of sampling. Hydrogen sulphide and DMS responses. All of the Aromascan sensors responded negatively to both hydrogen sulphide and DMS although the responses were weak and more delayed than for NH3. The normalised responses to 50 and 180 ppm H2S or DMS were similar, as 47 were the normalised responses to 20 and 110 ppm DMS. The strongest (negative) responses to both H2S and DMS were generally in sensors 19, 21 and 31. Lower concentrations of H2S or DMS resulted in a loss in the normalised response pattern for the particular gas. Passing the samples through a 0.6 m length Nafion tube did not significantly change the sensor response to H2S or DMS but reduced the sensor response to ammonia. The detection thresholds for H2S and DMS were about 50 ppm and 20 ppm. Compost odour mixture responses. Without Nafion tubing, the NH3 component (50 ppm) dominated, producing strong positive responses in all the sensors. After the gas mixtures were first passed through a 0.6 m length of Nafion tube, the H2S and DMS components dominated (both 60 ppm), producing negative responses in all the sensors. However, if the H2S and DMS components were below their sensor detection thresholds, there was either no significant sensor response or, if NH3 was present, a positive sensor response. Site I, which had high NH3, H2S and DMS concentrations, produced negative sensor responses for both the pre-wet and Phase I samples with the Nafion tube, but positive responses without. Compost odour samples with OCs of 30,000 OU m-3 could not be distinguished from less odourous samples with OCs of less than 10,000 OU m-3 from either the average sensor response or normalised sensor response pattern. There were negative relationships between average sensor response and the concentrations of H2S or DMS in compost odour samples, after passing through a 0.6 m Nafion tubing. These relationships were not present without the use of Nafion tube and all of the odour samples in produced positive sensor responses due to the presence of NH3. The Aromascan electronic nose was not sufficiently sensitive to odourous sulphurcontaining compounds and too sensitive to ammonia to be of use in evaluating composting odours. Performance of electronic sulphide detectors The three electrochemical sensors tested were found to be cross-sensitive to ammonia and, therefore, unsuitable for measuring sulphide levels in compost odour samples. The Draeger instrument was sensitive to water vapour, and H2S could not be detected in compost odour below 20 ppm. The GMI instrument did not give reliable H2S readings below 2 ppm; the Analox analyser could detect less than 1 ppm H2S, but only in the absence of ammonia. The Zellweger instrument was sensitive to hydrogen sulphide and methanethiol at 50 ppb but was insensitive to dimethyl sulphide at over 1000 ppb. There were no relationships between the Zellweger readings and the sample odour concentration or the sulphide readings from the pulsed fluorescence analyser. The pulsed fluorescence analyser was sensitive to all three sulphides tested (hydrogen sulphide, dimethyl sulphide and methanethiol) at a concentration of 10 ppb. There was a good correlation between the instrument readings (up to 500 ppb) and odour concentration (Fig.5). Odour and sulphide concentrations on and around composting sites The odour and sulphide concentrations decreased logarithmically with increasing distance from the main odour source (turned Phase I stacks) (Figures 6 and 7). The pulsed fluorescence instrument was able to detect sulphides at 10 ppb in odour samples from site boundaries, about 50 m downwind of the Phase I stacks. Both the odour and sulphide concentrations on the downwind site boundaries were significantly above the background (upwind) concentrations (Figures 6 and 7). 48 Validation of the compost odour / sulphide relationship Comparison of real and synthetic compost odours The synthetic pre-wet and Phase I odours smelt similar to poultry manure and the real and synthetic odours had similar detection thresholds. When presented as a series of dilutions, odour intensity score v odour concentration was similar for all the real and synthetic samples (Fig.10). This indicates that the six-compound mixtures were good simulations of pre-wet and Phase I odours, and that the measurement of a large number of compounds is unnecessary. Figure 10 shows that a tenfold dilution of the air from composting sites or synthetic composting odours resulted in almost a threefold reduction in odour intensity. Removing the fatty acids and trimethylamine from the synthetic compost odour mixtures had little effect on the odour concentration (Table 21). Most of the odour concentration of the synthetic pre-wet mixture was due to the sulphides (H2S, DMS and methanethiol) since removal of the ammonia also had little effect on the odour concentration. However, removal of the ammonia from the synthetic Phase I mixtures had a significant effect on reducing odour concentration (Table 21). Conclusions - Part 2 1. Sulphur containing compounds in compost odour samples were found to be most important in exceeding detection thresholds, but volatile fatty acids and trimethylamine were also found in odour samples at concentrations exceeding their detection thresholds. 2. There was a close correlation between the compost odour concentration of the pre-wet and Phase I composts samples and the combined hydrogen sulphide + dimethyl sulphide concentration from gas detector tubes. Concentrations of ammonia were above the detection threshold in most of the odour samples, but were not correlated with odour concentration. 3. Out of eight electronic instruments examined, only one was found to be sensitive to sulphide levels less than 60 ppb in compost air samples (see Summary Table below). 4. A pulsed fluorescence analyser was found to be sensitive to sulphides in composting odours at 10 ppb. There was a good correlation between the instrument readings and odour concentration. 5. The sensitivity of the analyser enabled it to detect odour plumes at the boundary sites, about 50 m from the Phase I composting stacks. 6. The Aromascan electronic nose was not sufficiently sensitive to odourous sulphurcontaining compounds and too sensitive to ammonia to be of use in evaluating composting odours (Summary Table). 7. Synthetic pre-wet and Phase I odours were prepared from sulphides, ammonia and other odour compounds which closely simulated real composting odours when presented to an odour panel. These results were consistent with the hypothesis that certain S compounds are mainly responsible for mushroom composting odour. 8. Most of the odour concentration of synthetic pre-wet odours could be attributed to the sulphide (hydrogen sulphide, dimethyl sulphide and methanethiol) content, although ammonia contributed strongly to the odour concentration of synthetic Phase I mixtures. 9. The type of olfactometer used in measuring odour concentration should be specified, since different types of machines can give different values. 10. Odour concentrations and sulphide concentrations from aerated composting systems were generally lower than those from non-aerated systems. 49 11. There was no significant difference in odour concentration between poultry manure based composts and those prepared with horse and poultry manures. No relationships were found between compost analysis and odour concentration. 12. A tenfold dilution of the air from composting sites or synthetic composting odours resulted in almost a threefold reduction in odour intensity. Summary Table of odour quantification techniques (Project Part 2) Method Human odour panel (Olfactometry) Advantages Relates to actual odours High sensitivity Gas detector tubes (sulphides) Cheap Not cross sensitive to ammonia or water vapour High sensitivity (5 ppb) Not cross sensitive to ammonia or water vapour Can be used on-site Pulsed fluorescence analyser Electronic nose or electronic sulphide detectors Disadvantages Not on-site High cost of measurements Partly subjective Only suitable for gases in concentrations > 0.1 ppm High cost (£10,000+) Cross sensitive to ammonia and moisture Low sensitivity to sulphides 50 Table 14. Compost yard Odour concentrations by olfactometry for compost yards in different years Year 1 Odour concentration, OU m-3 air Pre-wet Phase I Year 2 Year 3/4 Year 1 Year 2 Year 3/4 A - - - 17894 71702 4465 B 11796 3849 34958 10316 395533 81379 C 26777 25177 - 3965 2043 - D 6721 1933 - 20139 921 - E 1209 3698 - 919 885 - F 10467 4821 - 31823 16864 - G 6146 538 - 666 804 - H 10144 3430 8061 19607 20213 34459 I 245100 9553 2248 263758 72319 9518 J - 1626 - 13015 1957 - K - 993 - - 2919 - L.S.D. (P = 0.05) = 492 Aerated composts are shown in bold. 51 Table 15. Concentrationsa and odour detection thresholdsb for odourants identified with GC-MS in compost yard odour samples, μg m-3 air. Compounds with mean values exceeding the detection threshold are shown in bold. Odourant Acetic acid Acetone Ammoniac Butanoic acid Butanol 2-Butanone Carbon disulphide Dimethyl sulphide Dimethyl disulphide Dimethyl trisulphide Ethanol 4-Ethyl phenol Hexane Hydrogen sulphide Indole Iso-propyl-alcohol Methanethiol 2-Methyl butanoic acid 3-Methyl butanoic acid Methyl ethyl ketone Methyl phenol Methyl propanoic acid Methyl sulphide Pentanoic acid Phenol Propanoic acid Propanol Toluene Triethylamine Trimethylamine p-Xylene Pre-wet Phase I Detection threshold Characteristic odour 1588(±941) 3007(±5644) 27976(±28840) 481 3003(±6774) 697 trace 1667(±3290) 657(±1626) 83(±170) 53460(±103860) 75 150(±276) 730(±1310) 7 1837 211(±246) 238 206 660 101(±50) 26 156 216 139(±77) 2004(±2323) 5928(±13171) 30 179(±99) trace 113 1952(±1362) 2679(±5114) 75740(±62750) 2603(±3442) 2715(±6628) 287(±457) trace 3528(±6921) 287(±896) 130(±292) 10808(±16281) 79(±90) 86(±112) 49480(±10595) 3(±3) 195(±245) 126(±165) 165(±350) 869(±1531) 5720(±3354) 196(±101) 444(±379) 167 700(±1231) 189(±170) 351(±259) 678(±879) 18 169(±151) trace 25 25,000 770,000 260 1 x 10-3 33,000 18,000 2.6 255 15.3 2.6 x 10-2 93,000 vinegar ethereal dry urine rancid fusel oil foul foul foul vinous 1.1 600 ethereal rotten egg faecal 1.1 rotten cabbage unpleasant 5,000 80,000 220 800 12,000 3 ,000 80,000 140,000 4 96,000 acetone-like disagreeable unpleasant characteristic rancid stupefying ammonical fishy mean and ± standard deviation; compounds without standard deviations were identified in less than five samples a b after Summer (1971), van Gemert and Nettenbreijer (1977) and Overcash et al (1983) c measured with gas detection tubes d after Budavari (1989) 52 Table 16. Hydrogen sulphide concentrations determined with detector tubes on compost yards in different years Compost yard Year 1 Hydrogen sulphide concentration, ppm Pre-wet Phase I Year 2 Year 3/4 Year 1 Year 2 Year 3/4 A - - - 22 2 0.4 B 10 0.2 1.1 4 160 110 C 0.3 6.5 - 0.5 0.2 - D 3 0.6 - 11 0.4 - E 0.3 0.1 - 0.1 0.1 - F 2 0.3 - 7 9 - G 0.2 0 - 0 0 - H 0.8 0.6 0.4 3 2 30 I 57 6 0.1 95 41 3 J - 0.4 - 3 0 - K - 0.1 - - 0 - L - - - - 3 - M - 0.7 - - 0 - L.S.D. (P = 0.05) = 0.6 Aerated composts are shown in bold. 53 Table 17. Compost yard Dimethyl sulphide concentrations determined with detector tubes on compost yards in different years Year 1 Dimethyl sulphide concentration, ppm Pre-wet Phase I Year 2 Year 3/4 Year 1 Year 2 Year 3/4 A - - - 6 4 1 B 2 1 7 4 52 63 C 12 10 - 1 0.6 - D 2 1 - 1 1 - E 0.2 0.5 - 0.2 0.3 - F 0.1 1 - 4 7 - G 1 0.2 - 0.6 0 - H 2 0.8 2 7 3 0 I 24 2 0.5 97 32 5 J - 1 - 4 0 - K - 0.1 - - 0.1 - L - - - - 2 - M - 2 - - 0 - L.S.D. (P = 0.05) = 0.5 Aerated composts are shown in bold. 54 Table 18. Compost yard Ammonia concentrations determined with detector tubes on compost yards in different years Year 1 Ammonia concentration, ppm Pre-wet Phase I Year 2 Year 3/4 Year 1 Year 2 Year 3/4 A - - - 105 120 40 B 5 55 5 40 100 90 C 120 7.5 - 30 26 - D 5 400 - 30 550 - E 8 60 - 35 50 - F 19 10 - 4 30 - G 14 2 - 15 20 - H 22 30 50 110 45 70 I 27 4 22 276 39 8 J - 5 - 180 110 - K - 75 - - 35 - L - - - - 20 - 6 - - 5 - M L.S.D. (P = 0.05) = 4 Aerated composts are shown in bold. 55 Table 19. Compost analysis, mean values from 11 sites and two visits per site. ________________________________________________________________ Moisture % % of dry matter N pH NH4 ________________________________________________________________ Stage of composting Pre-wet 77.0 1.76 0.49 8.20 Phase I 75.7 2.09 0.51 8.13 0.248 0.103 0.196 LSD† (42 d.f.) 1.15 Aeration Aerated 75.9 2.01 0.49 8.01 Unaerated 76.6 1.88 0.50 8.25 0.281 0.109 0.196 LSD†† (42 d.f.) 1.27 ________________________________________________________________ † Least significant difference, calculated at P < 0.05, for comparing stages of composting and †† for comparing aerated and unaerated composts. 56 Table 20. Electronic sulphide analysers and electronic nose Company Analox Aromascan Draeger Analyser Mode of Operation Compounds Measured Specified L.D.L.* 101DZ electrochemical H2S 0.1 ppm Cross Sensitivity ammonia other sulphides Portable Cost £K Yes 3 No 32 A8S conducting polypyrroles polar compounds - ammonia alcohols PacIII electrochemical H2S 2 ppm ammonia other sulphides Yes 2 Yes 2 Personal surveyor electrochemical H2S 1 ppm ammonia other sulphides EVM GC+ion mobility spectrometry ionized molecules - - Yes 9 PE Photovac Voyager GC+ photoionization volatile organic - - Yes 18 Thermo Env. Instru. 43C+45C pulsed fluorescence sulphur containing 0.5 ppb - No 18 7100 colorometric paper H2S 2 ppb other sulphides Yes 5 GMI Graseby Dynamics Zellweger * lower detection limit 57 Table 21. Synthetic odour mixtures used to simulate typical pre-wet and Phase I compost odours Concentration in mixture, µg m-3 Compound All compounds Sulphides + ammonia Pre-wet Phase I Pre-wet Phase I Dimethyl sulphide 1667 3528 1667 Hydrogen sulphide 730 4362 Methanethiol 211 Butanoic acid Sulphides Pre-wet Phase I 3528 1667 3528 730 4362 730 4362 126 211 126 211 126 33 2129 - - - - 3-methyl butanoic-acid - 869 - - - - Trimethylamine - 582 - - - - Indole 7 - - - - - Ammonia 27976 75740 27976 75740 - - Odour conc. OU m-3 1359 3442 3955 11378 1829 2623 9837 14191 1851 252 1881 5778 58 Fig. 5 Relationship between pulsed fluorescence sulphides (Unicam analyser) and odour concentration 10000 8000 OC, OU/m3 6000 4000 2000 r 2 = 0.61 0 0 100 200 300 Unicam sulphides, ppb 59 400 500 Fig.6 Odour plumes from turned composting stacks. Distances down-wind are from the turned windrows; background measurements are 200m up-wind of composting sites. 11 50,000 site and date MM(Nov 2000) TunTech(Nov 2000) 9 MM(Dec 1999) 5,000 8 7 6 500 5 background 4 50 0 10 20 30 40 50 60 Distance down wind, m Fig.7 Plume of sulphides from turned composting stacks. Distances down-wind are from the turned windrows; background measurements are 200m up-wind of composting sites. 12 site and date MM(Nov 2000) TunTech(Nov 2000) MM(Dec 1999) TunTech(Dec 1999) Sulphide conc., ppb 100,000 10 8 10,000 10006 100 4 2 10 background 0 0 0 10 20 60 30 Distance from stacks, m 40 50 60 Odour conc. (OU/m3) loge odour conc. (OU/m3) 10 Odour Concentration, Odour Units m-3 106 12 105 10 104 8 103 6 loge OC = 7.601 + 0.934 loge (H2S + DMS + 0.375) r = 0.948 ( P < 0.001 ) 102 0 5 0.5 50 500 H2S + DMS concentration, ppm Fig. 8 Relationship between the combined on-site hydrogen sulphide and dimethyl sulphide concentrations and the odour concentration of bag odour samples from mushroom composting yards. Each point is the mean of two sample determinations. Solid circles are 1997/98 measrurements; open circles are for 1999/2000 measurements 14 loge OC = 8.774 + 0.838 loge (H2S + DMS) r = 0.883 13 loge OC 12 11 new old 10 9 8 7 -2 -1 0 1 2 3 4 5 6 loge H2S + DMS + 0.375 Fig.9 Relationship between the combined hydrogen sulphide and DMS concentrations and odour concentration, new olfactometer 61 Fig. 10 Relationship between air sample concentration and intensity score (0 to 6 scale) for real compost yard odour samples (a and b) and synthetic compost odour samples produced by mixing the most important odourant compounds in their typical concentrations (c and d). Each value is the mean of six determinations. (b) HRI Wellesbourne, Phase I (a) Gateforth Park, Pre-wet 6.0 y = 1.41x + 1.87 R2 = 0.95 5.0 4.0 intensity score intensity score 6.0 3.0 2.0 1.0 0.0 0.00 0.50 1.00 1.50 2.00 2.50 y = 1.28x + 2.06 R2 = 0.93 5.0 4.0 3.0 2.0 1.0 0.0 0.00 0.50 log 10 concentration 1.00 intensity score intensity score y = 1.27x + 1.89 R2 = 0.95 0.50 1.50 1.50 2.00 2.50 (d) Synthetised Phase I ( c) Synthesised Pre-wet 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0.00 1.00 log 10 concentration 2.00 2.50 3.00 6.0 5.0 4.0 3.0 2.0 1.0 0.0 0.00 y = 1.48x + 1.13 R2 = 1.00 0.50 1.00 1.50 log 10 concentration log 10 concentration 62 2.00 2.50 Part 3: Microbial and Chemical Degradation of Odours Introduction The target odour compounds identified for removal from compost are H2S and dimethyl sulphide (DMS). Bacterial strains known to remove H2S have been isolated from environmental samples and when re-applied to a biofiltration system can remove this compound from emissions. The isolates used in these types of study are normally present in a separate unit where ammonia has been removed, the temperature and other conditions within the system have been controlled. Such biofiltration systems are expensive and normally require specialised ducting, equipment and out-buildings. This part of the project aimed to isolate bacteria capable of removing H2S and DMS from the environment, and re-inoculating them on to the compost to test if they could remove these compounds directly from the stack. The isolates would have to function on the outside of the stack or on a matrix applied over its surface. As well as isolating strains, the development of reproducible compost capable of generating stable levels of the target compounds were needed. Bioremediation of odours Strains As a positive control for testing strains isolated from compost, one strain Hyphomicrobium species (a gift Dr H. Op den Camp, Netherlands) known to remove H2S in pure culture was used in these studies. The strain was tested for its ability to grow on simple nutrient media at a range of temperatures and large scale laboratory growth conditions in Luria Broth/Luria agar. Ten isolates obtained from the outer layer (5cm) of the compost stack were tested for their ability to remove H2S and DMS. Seven of the ten isolates performed as well as the Hyphomicrobium strain, the remaining three isolates significantly reduced the levels of H2S and DMS from samples, however they were approximately 3 times less active than the other isolates. All 11 strains had the following general characters; they grew between 25 and 40 oC, within a pH range of 6.5 and 8.5, and were aerobic. The 16S ribosomal RNA gene of each isolate was amplified using the universal primers 8f and 1654r. A region within the gene of approximately 500bp from each isolate was sequenced. Analysis of the sequence revealed that the isolates fell into the following lineages: Pseudomonas putida-lineage Pseudomonas fluorescens-lineage Bacillus cereus/thuringiensis/etc - lineage Hyphommicrobium- lineage Testing isolates on compost odours The Hyphomicrobium isolate was introduced on an agar plate or as a broth suspension into an air sample collected from above an 8-day-old compost sample. At time=0 the compost air sample was found to contain 30ppm H2S and 5ppm DMS. After 8 hours at 30oC the sample with the Hyphomicrobium strain had no detectable H2S or DMS, while in the untreated sample 20ppm H2S and no DMS was detected. This indicated that the strain effectively removed the H2S from the sample. Similar results were found for the 10 isolates used on 2 separate treatments, although slightly different starting conditions for both experiments were obtained (H2S 15ppm and 25ppm, and DMS 8ppm and 0ppm at the start). The decline in H2S and DMS detected in the control samples indicated that these molecules were unstable during the incubation period. The results also indicate that the microorganisms removed H2S from samples of compost air sample. 63 Testing model systems for levels of H2S and 02. In this experiment a series of model compost systems were set up on three occasions (A, B. C) and monitored for H2S and O2 levels. From a complete compost stack on the mushroom unit (1 week old) 600g (wet wt) material was placed into the model system (2L volume) and closed. Moisture levels were determined to be 75%. These systems were incubated at 55oC in a steam room and subsequently sampled over a 5d period. A Draeger tube placed into the compost headspace was used to measure H2S levels, and 02 levels were measured by destructively sampling the system at the end of the experiment. The results are presented below (Figure 11). These results indicate that a fixed O2 concentration within this model system dictates if it is likely to produce H2S at a level of 200+ppm. This level is around 3.8% O2. It may not be the overall value that is the key to H2S production, but that within samples with this overall level, regions exist where the O2 level is lower and H2S production takes place. Higher O2 levels may prevent these regions developing. The results also demonstrate that there appears Figure 12. Scatter plot of H2S vs O2 levels in mushroom composts contained in incubated flasks 10 02 level (%) 8 6 4 2 0 0 50 100 150 200 H2S level (ppm) to be a cut off at which point H2S production starts, producing very high levels within the system. It is likely that such areas developing within discrete regions within a compost stack result in the H2S and odour. From this point a dilution effect takes place which results in lower H2S levels at the surface or around the large stack. This is probably why we have been unable to develop model systems where 50ppm H2S is present in the headspace, consistently within the model system. If the O2 level is above the threshold we get no H2S production, and below this threshold we get greater than 2000ppm H2S. 64 Effect of Hyphomicrobium strain and Biofilm on H2S production from model compost systems Hyphomicrobium species has been shown to remove H2S when incubated in odourous mushroom air samples collected from above the stack. In this work, 500g of compost was placed in 2L flasks. Model systems were inoculated with the bacterium pre-grown at 30oC for 1 day on a nutrient agar plate, the agar was cut up and added to the top of the flask. A total of 6 inoculated and 6 control flasks were set up and incubated at 55oC in a water bath. Moisture content was 76%. The results can be seen in Table 22. In this experiment the H2S levels in the inoculated samples were higher than that in the control samples. Although the levels of H2S were lower than desired probably due to the use of a water bath and partial immersion. In a repeat of this experiment where the systems were placed at 55oC in a steam room, H2S levels were above 200ppm for all samples, inoculated and controls. In an additional experiment a biofilm sample was taken from a biofilter on a composting yard operating a treatment plant (250g of this material placed on the surface of the model systems). Samples were incubated at 55oC in a water bath. The results are presented in Table 23. In these results again poor H2S levels were detected within samples and if anything higher levels were detected within treatments compared to controls. In both these experiments the compost model system was only partially submerged in a water bath held at 55oC so that the surface of the compost exposed for treatment could be maintained at a lower temperature. This was used since the bacterium can only tolerate 42oC, and the on site measurements of the biofilter indicated it was operating at 20oC. The results indicate that these conditions are not suitable for generating H2S in our model systems. Overall the data obtained does not give any indication that either treatment will work. Table 22. H2S levels detected in model systems treated with Hyphomicrobium. Sample H2S 24h H2S 48h H2S 5d Control 1 0 0 0 Control 2 0 6 0.1 Control 3 0.1 0 0 Control 4 0.1 0.5 0 Control 5 0 0 0 Control 6 0.1 0.1 0.1 Inoc. 7 0.2 0.5 0.1 Inoc. 8 0.5 0.1 0.1 Inoc. 9 3.5 0.5 0 Inoc. 10 10 0.5 4 Inoc. 11 0.5 2.5 2 Inoc. 12 0.2 0.1 0.1 65 Table 23. H2S levels detected in model systems treated with biofilter material from two locations within 1 treatment plant. Control Sample A Sample B H2S 24h H2S 5d 1 0.1 0 2 10 6 3 10 0 4 0.1 0 5 20 0 6 60 2 7 180 0 8 400 5 9 120 2 10 2.0 10 11 120 2 12 180 30 Chemical degradation of odours - Ferric Sulphate treatment The use of ferric sulphate to eliminate sulphide malodours was developed on waste-water by Song at al (2001). To test the use of this material for mushroom composting odours, model systems were established as before. To these systems, ferric sulphate at 200mM and 500mM were added to chopped straw and this material placed on the surface of the compost. All flasks were placed in a 55oC room. The results can be seen in Table 24. The results illustrate that after 24 h 200mM ferric sulphate had removed over 88% of the H2S, and 500mM over 95% of the H2S . After 48h, 500mM ferric sulphate had still removed 80% of the H2S. In a repeat of this experiment a good effect was seen where H2S had fallen from 745ppm to 43ppm (+/-66.5), and DMS levels fallen from 409ppm to 58ppm (+/- 53). Again we were detecting 95% removal of H2S, and 85% removal of DMS. These results are very encouraging, as we have carried out no optimisation of the treatments used. Further work to study this in greater detail is therefore needed. 66 Table 24. The effect of ferric sulphate on H2S production from mushroom compost in incubated flasks. H2S 24h H2S 48h 02 1 800 800 5.2 2 2000 2000 4.5 3 2000 2000 4.9 4 2000 2000 4.2 200mM 5 400 2000 4.4 ferric 6 200 2000 5.4 sulphate 7 20 800 5.1 8 400 2000 4.6 500mM 9 60 10 5.8 ferric 10 70 10 4.8 sulphate 11 120 600 5.5 12 90 100 5.2 Control Figure12. Structure of the model systems. A, laboratory microcosm; B, compost stack experiment Less than 1ppm H2S B A AIR H2S DMS Am o C 8 sq ft. H2S-10ppm DMS-5ppm C 50-55oC 67 Less than 1ppm 1.5 sq ft H2S-<1ppm DMS-<5ppm Conclusions - Part 3 1. 2. 3. 4. Ten bacterial isolates were obtained from mushroom compost which were able to remove odourous sulphur containing compounds from compost air. The bacterial isolates belong to the following species: Pseudomonas putida, Pseudomonas flurescens, Bacillus cereus/thurigiensis and Hyphomicrobium spp. Ferric (iron) sulphate solution was more effective in removing sulphides and anaerobic odours from mushroom composts than microbial inocula. There was a distinct step in oxygen concentration (5%) below which anaerobic odours and sulphides developed. Above this threshold concentration, strong odours were prevented and no sulphides were detected. Conditions in composting stacks are too unfavourable (high temperatures and pH) for the survival of sulphide metabolising bacteria. Summary Table of microbial and chemical degradation of odours (Project Part 3) Method Biofiltration Hyphomicrobium spp. Advantages Could be used on a biofilter in removing sulphides; Does not require frequent changes Chemical scrubbing Ferric sulphate solution Effective at normal composting temperatures and pH Not affected by compost ammonia Disadvantages Cost of biofilter Microbes sensitive to high composting temperature, ammonia and pH Biofilter requires ammonia pre-washing Requires replacement of ferric sulphate at intervals Overall Conclusions 1. Replacing wheat straw with rape straw resulted in a significant reduction in odour in both windrow and aerated tunnel composts without affecting compost density, but mushroom yield was lower. Rape straw has a higher nitrogen content than wheat straw and required a lower inclusion of poultry manure. 2. Substituting poultry manure by 50% with organic (spent hop waste, cocoa meal, or molasses waste) or inorganic (ammonium sulphate or urea) nitrogen sources resulted in significant reductions in odour and sulphide concentrations, but also reduced mushroom yield. However, these materials may be suitable as low odour N sources during prewetting. 3. Release of ammonia from urea is more rapid than from poultry manure, but from cocoa meal it is more delayed. This material should therefore be used in the early stages of prewetting. 4. The omission of gypsum from compost did not significantly affect the emission of sulphides or odours but resulted in a significant yield reduction. There was no difference in mushroom yield between two types of gypsum (agricultural or citric acid production by-product). 5. Sulphur containing compounds in compost odour samples were found to be most important in exceeding detection thresholds, but volatile fatty acids and trimethylamine were also found in odour samples at concentrations exceeding their detection thresholds. 68 6. There was a close correlation between the compost odour concentration of the pre-wet and Phase I composts samples and the combined hydrogen sulphide + dimethyl sulphide concentration from gas detector tubes. Concentrations of ammonia were above the detection threshold in most of the odour samples, but were not correlated with odour concentration. 7. A pulsed fluorescence analyser was found to be sensitive to sulphides in composting odours at 10 ppb. There was a good correlation between the instrument readings and odour concentration. The sensitivity of the analyser enabled it to detect odour plumes at the boundary sites, about 50 m from the Phase I composting stacks. 8. Odour concentrations and sulphide concentrations from aerated composting systems were generally lower than those from non-aerated systems. 9. There was no significant difference in odour concentration between poultry manure based composts and those prepared with horse and poultry manures. No relationships were found between compost analysis and odour concentration. 10. A tenfold dilution of the air from composting sites or synthetic composting odours resulted in almost a threefold reduction in odour intensity. 11. Ferric (iron) sulphate solution was more effective in removing sulphides and anaerobic odours from mushroom composts than microbial inocula. 12. There was a distinct step in oxygen concentration (5%) below which anaerobic odours and sulphides developed. Above this threshold concentration, strong odours were prevented and no sulphides were detected. TECHNOLOGY TRANSFER Industrial relevance and plans for future commercial exploitation Results from the research work have been transferred to commercial sites as described in Objective 4 of the research results. This commercial exploitation on mushroom composting sites includes the following areas: waste, cotton seed meal and urea) at Hensby Composts Ltd, Blue Prince Mushrooms Ltd, Tunnel Tech Ltd, Pond Chase Nurseries Ltd and J. Rothwell & Son Ltd. Sources of the first two of these materials for mushroom composting have been agreed with Cadburys Ltd and English Hop Products Ltd. use of objective methods for measuring odours and sulphides using gas detector tubes and a pulsed fluorescence analyser for site boundary measurements at all the commercial composting sites of 5% v/v in compost in aerated systems at Blue Prince Mushrooms Ltd, J. Rothwell Ltd, Pond Chase Nurseries Ltd, Shepherds Grove (Gateforth Park) and Tunnel Tech Ltd. -product gypsum in mushroom compost mushroom compost at Hensby Composts Ltd, Blue Prince Mushrooms Ltd and Gateforth Park (Shepherds Grove Ltd) cost saving in N sources for one of the project partners (J. Rothwell & Son Ltd) The results of the project can be used in preparing new composting best practice guidelines, which are due to be renewed by the Environment Agency during 2002. Some of the project results have already been published in scientific journals and industry-based publications (HDC News etc) - see section on publications. Further publicity of the project results will continue in publications and posters, after agreement of content with the project consortium 69 (the Project Consortium will continue to meet at six-monthly intervals as part of the project application indicated in the previous section). The exploitation of results from this project in other related areas will be examined (odour measurement and reduction in waste composting and sewage treatment plants). This will form part of the new BBSRC project, which includes a component on waste composting emissions. Plans for future R&D resulting from the project A research project proposal has been submitted to the Biological and Biotechnology Research Council (BBSRC) in October 2001 entitled 'Identification of early indicators to improve the environmental impact of composting'. The science partners are the Institute of Grassland and Environmental Research (North Wyke), Horticulture Research International (Wellesbourne) and the University of Manchester Institute of Science and Technology, and the industrial partners include 7 of the existing project consortium, as well as 5 new industrial partners. The mushroom composting part of the project will cover the following areas: ct of compost nitrogen on composting odours - effects on compost odours and compost quality populations. The project will also cover waste composting emissions, enabling cross-fertilisation of methods from mushroom composting. As part of this project, the partners will continue to meet as a consortium on a regular basis. On-site testing of compost ingredients developed in this LINK project is continuing by individual project consortium members: Tunnel Tech Ltd, Hensby Composts Ltd (at own sites), J. Rothwell Ltd (at Horticulture Research International) Publications and Presentations resulting from the project Publications Noble R., Hobbs P. J., Dobrovin-Pennington A., Misselbrook T.H. & Mead A. (2001) Olfactory response to mushroom composting emissions as a function of chemical concentration. Journal of Environmental Quality 30: 760 - 767. Noble R., Hobbs P.J., Mead A., & Dobrovin-Pennington A. (2001) Influence of straw types and nitrogen sources on mushroom composting emissions and compost productivity. Submitted to Journal of Industrial Microbiology and Biotechnology. Noble R., Gaze R.H., Dobrovin-Pennington A. & Hobbs P. (1999) The development of odour-free mushroom compost. HDC News 56: 10 – 11. Noble R. (1999) Odour-free mushroom composting. Agriculture LINK Newsletter. Jan-Mar, p10. Noble R., Morgan A., Dobrovin-Pennington A. & Hobbs P. (2000) Compost without the smell. HDC News 66: 22,23. Odour-free mushroom composting. HortLink Leaflet. The project featured as a photograph of odour sampling at Middlebrook Mushrooms Ltd on the front cover of the above May issue of the Journal of Environmental Quality. 70 Presentations Mushroom compost research. Presentation by R. Noble at HRI/HDC Mushroom Subject Day, HRI Wellesbourne, June 2000 Measuring mushroom composting odours. Presentation by R. Noble at Pennsylvania Workshop on Mushroom composting odours, November 1999. References Anonymous. 1986 The Analysis of Agricultural Materials. Ministry of Agriculture, Fisheries and Food / Agricultural Development and Advisory Service Reference Book 427. HMSO, London. pp. 219-221. Anonymous 1997. Dräger-tube handbook. Drägerwerk Aktiengesellschaft, Lübeck, Germany. Anonymous 1998. Gastec gas detection system detector tubes list. Gastec Corporation, Kanagawa, Japan. Anonymous 1999. Air quality – determination of odour concentration by dynamic olfactometry. Draft Paper EN 13725, European Committee for Standardization, Central Secretariat, rue de Stassart 36, B-1050, Brussels. Bech, K and C. Rasmussen, 1968. Further investigations in organic and inorganic supplementation of mushroom compost. Mushroom Sci. VII: 329-342. Beyer D.M., Rynk, R., Peccia, J. and Wuest P. 2000. Improving odour management on mushroom farms. Biocycle 41 (7): 60-63. Bidlingmaier, W. 1992. Emissions from composting plants. Acta Horticulturae 302 19-27. Budavari, S. 1989. The Merck index. Merck & Co. Inc. Rahway, NJ, USA. Burton CH, Sneath RW, Misselbrook TH & Pain BF 1998 The effect of farm scale aerobic treatment of piggery slurry on odour concentration, intensity and offensiveness. J. agric. Engng. Res. 71: 203 - 211. Burton KS, Noble R (1993) The influence of flush number, bruising and storage temperature on mushroom quality. Post harvest Biology and Technology 3: 39 – 47. Byun H G, Persaud K D, Khaffaf S M, Hobbs P J, Misselbrook T M 1997 Application of unsupervised clustering methods to the assessment of malodour in agriculture using an array of conducting polymer odour sensors. Computers Electronics Agric 17: 233-247. Castro de Jimenez, G. F., Smith, J. F. and Love, M. E. 1990. Utilization of oil palm waste in the preparation of substrates suitable for mushroom cultivation. Mushroom J. Tropics 10: 93-100. Clarkson, C.R. and T.H. Misselbrook. 1991. Odour emissions from broiler chickens. p. 194 – 202. In V.C. Nielson, J.H. Voorburg and P. L’Hermite (ed.) Ammonia and odour emission from livestock production. Elsevier Applied Science, London. Dawson, W. M. 1978. The use of cattle slurry as a mushroom compost material. Mushroom Sci. X(2): 105-113. Delmas, J. and Laborde, J. 1968. Influence de la forme d’azote utilisee an cours de la fermentation des composts sur l’effect d’une supplementation proteinique aportee au moment du gobertage, et sur le rendement. Mushroom Sci. VII: 309-328. Derikx, P.J.L., H.J.M. Op den Camp, C. Van der Drift, L.J.L.D. Van Griensven, and G.D. Vogels. 1990. Identification and quantification of odourous compounds emitted during the production of mushroom compost. Appl. Environ. Microbiol. 56: 563-567. Derikx , P.J.L., F.H.M. Simons, H.J.M. Op den Camp, C. Van der Drift, L.J.L.D. Van Griensven, and G.D. Vogels. 1991. Evolution of volatile sulfur compounds during laboratory-scale incubations and indoor preparation of compost used as a substrate in mushroom cultivation. Appl. Environ. Microbiol. 57: 563-567. Devos, M., F. Patte, J. Rouault, P. Lafffort, and L.J. Van Gemert. 1990. Standardized human olfactory thresholds. Oxford University Press. 71 Dravneiks A, Prokop, W H 1975 Source emission odour measurement by a dynamic forcedchoice triangle olfactometer. J Air Pollut Cont Assn 25: 1. Duns, G.J., B.D. Ripley, C.A. Kingsmill, and D.L. Rinker. 1997. The analysis of mushroom composting odours. Mushroom World, June: 58-70. Duns, G.J., B.D. Ripley, and D.L. Rinker. 1998a. Monitoring the production of odourous compounds during Phase I mushroom composting. Mushroom World, March: 46 – 61. Duns, G.J., B.D. Ripley, and D.L. Rinker. 1998b. Off-site detection of odourous compounds produced by Phase I composting. Mushroom World, April: 13 – 20. Elliott-Martin R J, Mottram T T, Gardner J W, Hobbs P J, Bartlett P N 1997 Preliminary investigation of breath sampling as a monitor of health in dairy cattle. J agric Engng Res 67:267-275. Flegg, P.B. and Loughton A. 1961. Cereal straws used for mushroom composts. Experimental Horticulture 5: 50-52. Gardner J W, Shurmer H V, Tan T T 1992 Application of an electronic nose to the discrimination of coffee. Sensors and Actuators B 6: 71-75. Gerrits, J. P. G. 1988. Nutrition and compost. In: The Cultivation of Mushrooms (Van Griensven, L. J. L. D. ed), pp29-72, Darlington Mushroom Laboratories, Sussex, UK. Gerrits, J. P. G. 1989. Kan indoor compost worden uitgezweet? De Champignoncultuur 33: 65-71. Gerrits, J.P.G. 1992. Bijvoeden van conventionele en indoor compost. De Champignoncultuur 36: 31-37. Gerrits, J.P.G. 1994. Developing indoor composting: a historical view. Mushroom J. 530: 15-19. Gerrits, J.P.G., Amsing, J.G., Straatsma, G. and van Griensven L.J.L.D. 1995. Phase I process in tunnels for the production of Agaricus bisporus compost with special reference to the importance of water. In: Science and Cultivation of Edible Mushrooms (ed.) T.J. Elliott. Balkema, Rotterdam, Netherlands. pp. 203-211. Grabbe, K. 1974. Versuche zur Biotechnologischen Aufbereitung von Rinderflűsigmist zu Kultursubstraten fűr hőhre Pilze. Mushroom Sci. IX(2): 99-108. Gulliver, A., F.C. Miller, E. Harper, and B.J. Macauley. 1991. Environmentally controlled composting on a commercial scale in Australia. p. 155-164. In M. J. Maher (ed.) Science and cultivation of edible fungi. Balkema, Rotterdam, Netherlands. Hill, D.T. and C.L. Barth. 1976. Quantitative prediction of odour intensity. Trans. Am.Soc.Agric.Engnrs. 19: 939 – 944. Hobbs, P. J., T.H. Misselbrook, and B.F. Pain. 1995. Assessment of odours from livestock wastes by a photoionization detector, an electronic nose, olfactometry and gas chromatography-mass spectrometry. J. agric. Engng. Res. 60: 137-144. Hobbs, P. J., T.H. Misselbrook, and B.F. Pain. 1997. Characterisation of odourous compounds and emissions from slurries produced from weaner pigs fed dry feed and liquid diets. J. Sci. Food Agric. 73: 437-445. Hobbs, P.J., T.H. Misselbrook, and T.R. Cumby. 1999. Production and emission of odours and gases from ageing pig waste. J. Agric. Eng Res. 72: 291-298. Leonardos, G., D. Kendall, and N. Barnard. 1969. Odour threshold determinations of 53 odourant chemicals. J. Air Pollution Control Assn. 19: 91 – 95. MacCanna, C., 1968. Nitrogen supplementation of composts. Mushroom Sci. VII: 294-307. Miasik J J, Hooper A, Tofield B C 1986 Conducting polymer gas sensors. J Chem Soc, Faraday Trans 1 82 1117-1126. Miller, F. C. and B.J. Macauley. 1988. Odours arising from mushroom composting: a review. Austl. J. Exp. Agric. 28: 553-560. Miller, F.C., B.J. Macauley, and E.R. Harper. 1991 Investigation of various gases, pH and redox potential in mushroom composting stacks. Austl. J. Exp. Agric. 31: 415-425. 72 Misselbrook, T.H., P.J. Hobbs, and K.C. Persaud. 1997 Use of an electronic nose to measure odour concentration following application of cattle slurry to grassland. J. Agric. Eng. Res. 66: 213-220. Noble, R. and R.H. Gaze. 1994. Controlled environment composting for mushroom cultivation: substrates based on wheat and barley straw and deep litter poultry manure. J. Agric. Sci. Camb. 123: 71-79. Noble R. & Gaze, R.H. (1996) Preparation of mushroom (Agaricus bisporus) composts in controlled environments: Factors influencing compost bulk density and productivity. International Biodeterioration and Biodegradation 37: 93-100. Noble R. & Gaze R.H. (1998) Composting in aerated tunnels for mushroom cultivation: Influences of process temperature and substrate formulation on compost bulk density and productivity. Acta Horticulturae 469: 417-427. Noble, R., T.R. Fermor, C.E. Evered, and P.T. Atkey. 1997. Bench-scale preparation of mushroom substrates in controlled environments. Compost Sci. Utiliz. 5 (3): 32- 43. Noble R., Gaze R.H. and Willoughby N. (1998) A high yielding substrate for mushroom experiments: Formula 3. Mushroom Journal 587: 27,28. Noble, R. P. J. Hobbs, A. Dobrovin-Pennington, T.H. Misselbrook and A.Mead. 2001. Olfactory Response to Mushroom Composting Emissions as a Function of Chemical Concentration. Journal of Environmental Quality 30: 760 - 767. Oei, P. 1991. Manual of Mushroom Cultivation. Tool Acta, Wageningen, Netherlands. pp? Overcash, M.R., F.J. Humenik, and J.R. Miner. 1983. Livestock waste management. Vol. I, CRC Press Inc. Boca Raton, Florida, USA. 255pp. Patterson, M.Q., J.C. Stevens, W.S. Cain, and J.E. Cometto-Muniz. 1993. Detection thresholds for an olfactory mixture and its constituent compounds. Chemical Senses 18: 723 – 734. Perrin P S, Gaze R H 1987 Controlled environment composting. Mushroom J 174: 195197. Perrin, P. S. and B.J. Macauley. 1995. Positive aeration of conventional (Phase I) mushroom compost stacks for odour abatement and process control. p. 223-232. T.J. Elliott (ed.) Science and cultivation of edible fungi. Balkema, Rotterdam, Netherlands. Persaud K C, Khaffaf S M, Hobbs P J, Sneath R W 1996 Assessment of conducting polymer odour sensors for agricultural malodour measurements. Chem Senses 21: 495505. Poppe, J. 1995. Use of agricultural waste materials in the cultivation of mushrooms. In: Science and Cultivation of Edible Mushrooms (ed.) L.J.L.D. van Griensven. Balkema, Rotterdam, Netherlands. pp. 3- 23. Riethus, H. 1962. Versuchsergebenisse fűr Stickstoffbilanz in Champignonsubstraat. Mushroom Sci V: 10-18. Ross, R. C. 1968. Experiments on the use of farm waste products in mushroom composting. Mushroom Sci. VII: 365-371.Schaefer J. 1977 Sampling, characterization and analysis of malodours. Agric. Environ. 3: 121-128. Smith, J. F. and Spencer, D. M. 1977. The use of high energy carbon sources in rapidly prepared mushroom composts. Scientia Hortic. 19: 65-78. Song Z., Williams C.J., Edyvean R.G.J. 2001 Coagulation and anaerobic digestion of tannery wastewater. Process Safety and Environmental Protection79 (B1): 23-28 Van den Berg M, 1992 Sensory odour measurements using an olfactometer. Nederlands Normalisatie Instituut, Kalfjeslaan 2, Postbus 5059, 2600 GB Delft, Netherlands. Van der Hoek, K. W. and J. Oosthoek. 1985. Composting: odour emission and odour control by biofiltration. p. 271-281. In J.K.R. Gasser (ed.) Composting of agricultural and other wastes. Elsevier Applied Science Publishers, Amsterdam, Netherlands. 73 Verscheuren, K. 1983. Handbook of Environmental Data on Organic Chemicals. Van Nostrand Reinhold, New York Wood, D.A. and Fermor T.R. 1985. Nutrition of Agaricus bisporus. In: The Biology and Technology of the Cultivated Mushroom (Eds) P.B. Flegg, D.M. Spencer and D.A. Wood). John Wiley and Sons, Chichetser, UK. pp. 43-61. 74