sign in sign up
<<
Home , Cured fish, Smoked fish

Influence of Locally Available Trees or Tree Products (Acacia, Prosopis, Neem and Coconut Husks) on Quality and Post Harvest Losses of Smoked Marine and African In Kipini and Tana Delta

Item Type Report

Authors Odote, P.M.O.

Publisher Western Indian Ocean Marine Science Association

Download date 01/10/2021 17:26:21

Link to Item http://hdl.handle.net/1834/7859

INFLUENCE OF LOCALLY AVAILABLE TREES OR TREE PRODUCTS (ACACIA, PROSOPIS, NEEM AND COCONUT HUSKS) ON QUALITY AND POST HARVEST LOSSES OF SMOKED MARINE AND AFRICAN CATFISH IN KIPINI AND TANA DELTA

BY

Peter Michael Oduor-Odote

REPORT NO: WIOMSA/MARG-I/2009 –02

ii

Contents 1.0 Introduction ...... iv 1.1 The hypotheses...... vi 2.0 Materials and Methods ...... vi 2.1 Construction of ovens ...... vi 2.1.1 Sample collection ...... viii 2.1.2 Preparation of samples for processing ...... viii 2.1.3 Moisture analysis ...... ix 2.1.4 Smoke drying characteristics ...... ix 2.1.5 Sensory evaluation ...... ix 2.1.6 Storage trials ...... x 2.2 Properties of Wood fuel ...... x 2.2.1 Determination of calorific values ...... x 2.2.2 Determination of moisture content and ash content ...... xi 2.2.3 Density ...... xi 2.2.4 Moisture content ...... xi 2.2.5 Ash content ...... xi 2.2.6 Burning Characteristics ...... xii 2.2.7 Fuel wood value Index ...... xii 3.0 Data analysis ...... 12 4.0 Results ...... 12 4.1 Sensory evaluation ...... 12 4.2 Insect attack during storage ...... 15 4.3 Mould attack during storage ...... 17 4.4 Moisture during storage ...... 19 4.5 Burning properties ...... 21 4.6 Wood fuel consumption ...... 25 4.7 Drying characteristics ...... 26 5.0 Conclusions ...... 29 6.0 Acknowledgements ...... 29 7.0 References ...... 30

LIST OF FIGURES

Figure 1 Traditional oven and communities participating in construction of improved ovens ..... vi Figure 2 Completed ovens & The Minister for Dr. Otuoma (Checked shirt) inspecting the project in Moa ...... vii Figure 3 Humidity during storage of ...... 16 Figure 4 Temperature during storage of fish ...... 16 Figure 5 Mould on MCP No mould on FCN ...... 18

iii

Figure 6 % Moisture during storage of freshwater catfish smoked with Acacia, Prosopis, Neem and Coconut husks ...... 20 Figure 7 % Moisture during storage of marine catfish smoked with Acacia, Prosopis, Neem and Coconut husks ...... 21 Figure 8 Rate of water loss for freshwater catfish smoked with Acacia, Prosopis, Neem and Coconut husks ...... 28 Figure 9 Rate of water loss for marine catfish smoked with Acacia, Prosopis, Neem and Coconut husks ...... 29

LIST OF TABLES Table 1 Mean sensory scores of taste, appearance and overall acceptability attributes of freshwater and marine catfish smoked using Acacia, Prosopis, Neem and Coconut husks ...... 13 Table 2 Mean sensory scores of taste, appearance and overall acceptability attributes of marine catfish smoked using Acacia and compared with Prosopis, Neem and Coconut husks .. 13 Table 3 Mean sensory scores of taste, appearance and overall acceptability attributes of freshwater catfish smoked using Acacia and compared with Prosopis, Neem and Coconut husks ...... 14 Table 4 Insect scores and % of fish attacked during storage in open air ...... 17 Table 5 Mould scores and % of fish attacked during storage in open air ...... 17 Table 6 % Moisture during open air storage...... 19 Table 7 Burning properties of wood Acacia, Prosopis, Neem and Coconut husk ...... 22 Table 8 Ranking of burning properties of wood Acacia, Prosopis, Neem and Coconut husk ...... 22 Table 9 Moisture, Ash, volatile matter, fixed carbon and calorific value of the different tree species ...... 23 Table 10 Fuel wood ranking of Acacia, Prosopis, Neem, Coconut husks on the basis of Calorific value, Density, ash, Moisture and Fuel value Index. The ranking ranges from 1 (best) to 4 (last) ...... 24 Table 11 Fuel wood ranking of Acacia, Prosopis, Neem, Coconut husks on the basis of Calorific value, Density and ash. The ranking ranges from 1 (best) to 4 (last) ...... 24 Table 12 Wood fuel used during smoking (Weights in Kg) ...... 25 Table 13 Weight loss, % water (g) in fish after smoking, final weight after smoking, hours of smoking to 90%, 95% and constant weight during smoking of African and Marine Catfish ...... 26 Table 14 Drying rate of marine catfish smoked by Acacia, Prosopis, Neem and Coconut husks. 28

iv

1.0 Introduction Fish and are readily available sources of animal protein in the coastal region of Kenya. The Tana River area is known for catfish landings which are smoked by artisanal fishermen to prevent spoilage. As smoke is produced by artisans, its quality depends upon the skill of the individual fish processor in the village. Because of its significant role in the preservation and consumption of fish, traditional fish smoking technology has been studied in the Tana delta area and improvements made in terms of process efficiency and product quality (Oduor-Odote et al, 2008). The advantages of improved ovens over traditional ones include increased capacity, fuel economy and better quality product (Oduor-Odote et al 2008). In Kenya, reports on the wood quality of various species of trees as fuel energy source for smoking fish are limited. The burning properties of wood used as fuel are important in selecting wood for fish smoking (Lartey et al 1994). Wood smoke has variable composition which depends on various conditions such as procedure and temperature of smoke generation, origin and composition of wood, water content in wood etc (Simko 2005). So far, up to 1100 various chemical compounds have been identified and published in literature (Simko 2005). The smoking treatment itself is based on successive deposition of compounds such as phenol derivatives, carbonyls, organic acids, esters, acetones, lactones, pyrazines, pyrol and furan derivatives on a food surface with their subsequent migration into bulk (Simko, 2005, Martinez et al, 2007, Guillén and Martinez,2002). Many factors influence the quality of products including the properties of the fish flesh, maturity, age, sex, seasonal variation and factors involved in the smoking procedure such as wood type, composition of smoke, temperature, humidity, velocity and density of the smoke (Simko, 2005). Specific volatile compounds in particular phenolic compounds have been related to the different smoking techniques which directly influence the sensory characteristics of smoked fish (Cardinal et al 2006; Jónsdóttir et al 2008). Some of the phenolic compounds are guaiacol and syringol which are very characteristic in smoked fish (Jónsdóttir et al 2008). Apart from the African catfish, there is however the marine catfish which closely resembles it in appearance, comes to brackish waters and is thrown away as discard by prawn trawl fishers. In other places like Kipini, it is considered a low value fish and also discarded or dried and sold at throw away prices (KMFRI Ungwana bay report 2002). In places where it finds its way upstream therefore around the Tana estuary, it can be landed and smoked and provide a chance for its protein to be utilized alongside

v the African catfish. Little work has also been done to assess post processing and general quality deterioration of smoked fish in Tana delta area during storage. Studies in West African countries indicate that traditionally smoke- dried fish are stored in round smoking ovens and covered in polyethylene and jute sacks. Occasional re-smoking is undertaken to maintain dryness and drive off insect pests. Storage conditions under such conditions are unsatisfactory due to frequent insect infestation, microbial decomposition and rodent attack (Cauri et al 1979; Nerquaye-Tetteh 1979, Okraku-Offei, 1970). Insect infestation can cause a loss of 30% to 50 % fish meat (Khan and Khan 2001; Osuji, 1977, Proctor, 1977). In Kenya, exact statistics of storage losses of smoked fish are not available but observations indicate that post processing losses of the smoked fish do occur during storage. Insect infestation especially beetles are common. Efforts to reduce high losses are needed if the beneficial effects of improved processing techniques are to be derived. Any reduction in processing and post processing loss by simple modifications of existing methods will benefit the fishing villages in the Tana delta area. The main interest in using Neem (Azadirachtica) in fish smoking is because of activity of its components as a deterrent for insect activity. It has multiple pesticidal and medicinal properties, smoke from its leaves are used as insect repellant, about 135 different compounds are found in every part of the tree and it has antimicrobial effects (Battacharya et al, 2007; Coventry and Allan, 2001, Modue and Nisbet 2000; Biswas et al 2002). In Kenya, interest started being shown in the Neem tree for commercial and industrial potential in 1980’s and it continues till now (ICIPE, 1995). No work is reported for the use of Neem tree in Kenya for control of insect infestation during smoking and storage of smoked fish. It is therefore with this in mind that we intend in this part of the study to utilize Neem tree firewood in fish smoking and later monitor insect infestation during storage of such smoked fish. Other fuel woods selected for quality effect on fish smoking are Prosopis (Mathenge) an alien species, Coconut husks of Cocos nucifera and Acacia raddiana which is the tree commonly used in the Tana delta region for fish smoking. Prosopis (“mathenge”) was introduced in Kenya (Pasiecznik et al, 2006) and has potential as a wood fuel (Puri et al, 1994). It has however not been tried to smoke fish. If found suitable then its use in fish smoking will ease off pressure on the Acacia tree currently in use in the Tana delta region in fish smoking where it is the tree of choice. Coconut plantations provide numerous products such as coconut meat, milk and husks. Coconut husks can be used for , heating and as the heating base for smoking meat and fish. The use of coconut husks for fish smoking

vi had not been tried at the coast and if found to be effective will reduce the demand on mangrove wood. The purpose of this study was to evaluate the properties and relative effect of 4 tree species or products- Acacia raddiana, Prosopis julifora, Neem (Azadirachta indica) and Coconut husks from Cocos nicifera on smoking the African and marine catfish and to monitor insect infestation in the fish smoked with Neem tree during storage

1.1 The hypotheses

1. The 4 wood species have different quality properties –calorific value, specific gravity, density and burning properties 2. The 4 wood species/types shall have different effects on taste, appearance and shelf life of the 2 selected smoked fish species 3. The Neem tree fuel wood shall have an effect on insect infestation during storage of the 2 selected catfish species.

2.0 Materials and Methods

2.1 Construction of smoking ovens

6 clay-walled improved double door fish smoking ovens were constructed according to Oduor-Odote (2006 Brownell & Lopez, 1985; Sharing innovative experiences, 1999). The area for construction of the ovens was selected after deliberations with the local community to come to agreement. 5 ovens were located at Moa village and one at Manyala village due to high demand for the ovens. The ground was cleared and leveled.

Figure 1 Traditional oven and communities participating in construction of improved ovens

vii

Figure 2 Completed ovens & The Minister for Fisheries Dr. Otuoma (Checked shirt) inspecting the project in Moa

The outline of the oven walls was drawn and a trench 20 cm deep dug for the foundation. The wall frames were then constructed using locally available “fitoes”, logs and mud. The inclusion of the frame of “fitoes” is an improvement introduced for further reinforcement. The combustion chamber which was the base of the smoking oven was rectangular, about twice as long as wide and with two stoke holes in front. The length of the combustion chamber was 225 cm, width 112.5 cm, height 60 cm, wall thickness 12.5cm, width and height of stoke pole 37.5cm and depth of fire pit 15cm for each stoke hole (Figure 1). A dividing wall in the middle was constructed to give added strength to the oven and protect the median cross piece of the tray from burning and to give greater support to the loaded trays and allow for smoking of small quantity of fish over just one chamber, using less wood. The stoke holes are arched and in the middle of the rectangular base. The top of the oven wall was square, level and flat so that the wooden-framed trays fitted flush and no smoke or heat allowed to “escape” through gaps. A space of 5cm was left between the frame wall and inner wall of the base. The smoking trials reported here were done with 1 tray. The tray was 235 cm long including handle and 215.9 cm without handle but including inner dividing wall frame. The internal width was 96cm and internal length without dividing wall frame was 208 cm. The average depth of the tray was 6.25cm. Wire netting of ¾ inch mesh and 20 gauge was used in the inner part of the tray and a 2 inch mesh size wire below it. This modification was necessary to avoid smaller fish falling through into the fire. Fire was kept at least 50 cm from the lowest tray. A plane iron sheet cover measuring 4ft by 8 ft was used as a cover for regulating smoke and heat in each oven (Figure 2).

viii

2.1.1 Sample collection

A total of 500 pieces of marine and freshwater catfish were purchased when fresh during the months of August and September 2008 in Kipini and Tana River areas respectively. It was not possible to get all fish of the same size as they have different sizes in different seasons. The marine catfish were generally smaller in size compared to the freshwater catfish.

2.1.2 Preparation of samples for processing

The fish were gutted then washed thoroughly. The tails were then slit by making a small cut in them, then coiled and by squeezing the gill region between the thumb and forefinger to open the mouth, the tail was fixed in the mouth and with pressure released, the tail remained inside the mouth. (Oduor-Odote et al, 2007). The fish were once washed thoroughly in fresh water with ice. 10 marine and freshwater catfish each with a mean weight of 392.5± 61.7 and 808.7 ±196.3 respectively were kept fresh in ice and not processed to provide baseline data for body composition. The mesh of the trays were oiled prior to the distribution of fish using liquid food vegetable oil so that fish do not stick during smoking. 60 pieces of each species of fish were randomly placed on each of the 4 fish smoking trays. Each tray therefore contained 120 pieces of fish on either side of marine catfish or freshwater catfish. The trays with the fish were transferred under shade and at an angle to let the fish drip dry for 2 hours. Small logs of wood from Acacia raddiana, Prosopis julifora and Azadirachtin indica were cut into chunks using a power saw HUSQVARNA 61-268 272XP. Smaller pieces were cut using hand saws from the chunks to smaller pieces with a length of 50 cm and a fairly uniform thickness of 4.0 to 7.0 cm. The coconut husks were utilized the way they were. The fire was lit 30 minutes before smoking started using grass and let to burn till the flames died off and only smouldering wood remaining. The pieces of wood were weighed each time before being put in the fire pits of the smoking ovens. 6 fish species were weighed in a PETER electronic field balance to give a representative sample of weight at the beginning of the experiment in each of the trays. They were returned to the trays. The starting weights of wood fuel were weighed before the start of the experiment. The starting temperatures of the fish were measured using normal mercury thermometers. The fish were transferred to the smoking chambers already pre-prepared according to Oduor-Odote et al (2007) and Oduor-Odote (2006). Each set of trays with 60 pieces

ix of freshwater catfish on one side and marine catfish on the other were smoked using the 4 tree types all at the same time in each of the 4 double door ovens. The fish on each of the trays were labeled as follows- MCA –(Marine catfish-Acacia); MCP –(Marine catfish Prosopis); MCN- (Marine catfish Neem); MCC- Marine catfish coconut; (FCA- (Freshwater catfish –Acacia); FCP- (Freshwater catfish fish-Prosopis); FCN (Freshwater catfish-Neem), FCC (Freshwater catfish Coconut). The fire from the coconut was difficult to control as it was burning fast and on the third day it was mixed with Acacia then smoking continued. Its results therefore include contribution from Acacia. To control the excessive temperature in the ovens, the intensity of the fire was reduced by intermittent withdrawal of some of the logs from the fire point. At intervals the positioning of the fish was interchanged to effect uniform penetration. Temperature of the smoking ovens as well as drying rate was measured every 2 hours with the latter using 6 randomly chosen fish from each of the trays. At the end of smoking the fish were removed from the kilns and exposed to air to cool for 2 to 3 hours (Akande et al 1996). The fish were packed in labeled plastic open sided milk crates and transported for storage in the open on benches in KMFRI lab Mombasa 350 km away for proximate composition and organolepic analysis.

2.1.3 Moisture analysis Moisture content was determined by the air oven method (AOAC 1990) using 6 fish samples.

2.1.4 Smoke drying characteristics 6 samples each of MCA, MCP, MCN, MCC, FCA, FCP, FCM, and FCN were analyzed for weight or moisture loss during smoking in the ovens every 2 hours till end of smoking. This was done using a SALT PETER electronic field balance.

2.1.5 Sensory evaluation This involved locally trained panelists drawn from KMFRI staff. The organoleptic parameters that were evaluated included appearance, taste and with provision for a score on overall acceptability. A 5 point hedonic scale was used as shown in Annex 1. The tests carried out were in three sets. The first one was MCA vs. FCA, MCP vs. FCP, MCN vs. FCN and MCC vs. FCC. The second set was MCA vs MCP, MCA vs MCN and MCA vs MCC. The third set was FCA vs

x

FCP, FCA vs FCN and FCA vs FCC. Set one was to establish sensory scores between freshwater catfish and marine catfish, set two for sensory scores due to due marine catfish smoked Acacia against Prosopis, Neem and Coconut husks and set three for sensory scores due to due freshwater catfish smoked Acacia against Prosopis, Neem and Coconut husks A score below 2 was considered not acceptable. The fish samples were coded with numbers of 2 digits indicating no information about the samples to avoid bias in preferred treatments. Samples for taste (cooked flavor) were heated in a microwave oven LG INTELLOWAVE without any additives e.g. salt before tasting. Normal consumption temperature for food was used.10 panelists who were neither hungry nor satiated were used at each sampling time. The panelists were advised not to smoke or eat 1 hour before the sensory evaluation, to avoid perfumes or aftershave and avoid being in the panel if sick or suffering from cold. Each panelist was served as similar a part of the fish as possible i.e. a sample from the tail parts or loin parts. The quantities of samples were at least 2-3 bites for each panelist. The panelists received each sample separately. Rinsing the mouth between samples was done using plain unsalted crackers then water.

2.1.6 Storage trials Insect and or mould attack was monitored during the storage period. Attacks by insects or moulds were determined according to the scale indicated in annex 2 (Khan& Khan 2001). Any sign of mould attack was the limit of acceptability while a score of 3 for insects was the limit of acceptability beyond which the fish were rejected. Insect and mould attacks were monitored regularly.

2.2 Properties of Wood fuel

All analyses to describe wood fuel properties were carried out at KEFRI laboratories in Muguga.

2.2.1 Determination of calorific values

This was done according as described in Puri et al (1994). The 4 wood samples-Acacia, Prosopis, Neem and coconut husks were each ground to fine powder in a grinder. 1g each of the samples were weighed in duplicate, wrapped with a weighed tissue paper of a known calorific

xi value then tied with an ignition wire of platinum also of known calorific value. Both ends of the wire were connected to the bomb calorimeter electrodes (+,-) and placed in a bomb and firmly closed. 30kg of oxygen was led into the bomb and the bomb immersed into the cylinder filled with distilled water up to 2,100g. The bomb calorimeter was calibrated with Benzoic acid tablet of a known calorific value.

The following formula was used to calculate the calorific value of the samples:-

CV (Cal/g) = Water equivalent (g) + Water quantity of inner Cylinder (g) x Rise in temperature (°C) – Calorie Correction Quantity of Sample (g)

2.2.2 Determination of moisture content and ash content

This was done according to British standard 3631 (1973). The laboratory crucibles were reconditioned by putting them into a muffle furnace at 900± 25oC for 2 minutes. The crucibles were removed then cooled in a dissector for 10 – 15 minutes and weighed to the nearest 0.1mg.

2.2.3 Density Density was determined according to Puri et al (1994)

2.2.4 Moisture content

1.000g of the samples was weighed into the crucible in duplicate and put into the oven for at least more than 12 hours at 100 ± 3 oC.

Moisture Content = Initial weight (g) – Oven dry weight x % Oven dry weight

2.2.5 Ash content

This was continued from the volatile matter samples. The crucibles and its residue was returned into the muffle furnace at 900 ±25 oC for the hour.

xii

The crucibles were removed from the muffle furnace, placed into the desiccator and cooled for 15 minutes and re-weighed. The formula below was used to determine the ash content of the samples.

Ash Content =Final weight of residue x100 Original weight of sample

2.2.6 Burning Characteristics

Samples were split into small sticks the size of match sticks. 3 sticks of each sample were taken and put into a Japanese stove and lit using a match box.

2.2.7 Fuel wood value Index

This was calculated according to Puri et al (1994) FVI= Calorific value× density Ash content×moisture content

3.0 Data analysis

Analysis of variance (ANOVA) was carried out on the sensory data in the statistical program NCSS 2000 (NCSS, Utah USA). The program calculates multiple comparisons using Duncan’s test to determine if sample groups are different. Significant difference was defined at the 5% level.

4.0 Results

4.1 Sensory evaluation

The sensory scores were as shown in table 1. There was a significant difference (p<0.05) in taste between marine and freshwater catfish smoked with Acacia. There was no significant difference in taste, appearance and overall acceptability between freshwater and marine catfish smoked with Prosopis, Neem and coconut husks (coconut husks + Acacia).

13

Table 1 Mean sensory scores of taste, appearance and overall acceptability attributes of freshwater and marine catfish smoked using Acacia, Prosopis, Neem and coconut husks Groups (FCA vs MCA) FCA MCA Taste 3.7a 4.3ab Appearance 3.2 3.8 Overall acceptability 3.2 3.8 FCP vs MCP Taste 3.3 4.2 Appearance 3.5 2.9 Overall acceptability 3.6 3.8 FCN vs MCN Taste 4.2 4.2 Appearance 3.7 3.9 Overall acceptability 4 4 FCC vs MCC Taste 3.6 4.1 Appearance 3.4 3.6 Overall acceptability 3.5 3.7

There was no significant difference in taste, appearance and overall acceptability between marine catfish smoked with Acacia and marine catfish smoked with Prosopis, Neem and Coconut husks (Table 2) Table 2 Mean sensory scores of taste, appearance and overall acceptability attributes of marine catfish smoked using Acacia and compared with Prosopis, Neem and coconut husks Groups (MCA vs MCP) MCA MCP Taste 4.3 4.2 Appearance 3.8 2.9 Overall acceptability 3.8 3.8 MCA vs MCN Taste 4.3 4.2 Appearance 3.8 3.9 Overall acceptability 3.8 4 MCA vs MCC Taste 4.3 4.1 Appearance 3.8 3.6 Overall acceptability 3.8 3.7

There was no significant difference in taste, appearance and overall acceptability between freshwater catfish smoked with Acacia and freshwater catfish smoked with Prosopis, Neem and Coconut husks (Table 3)

14

Table 3 Mean sensory scores of taste, appearance and overall acceptability attributes of freshwater catfish smoked using Acacia and compared with Prosopis, Neem and coconut husks Groups (FCA vs FCP) FCA FCP Taste 3.7 3.7 Appearance 3.2 3.5 Overall acceptability 3.2 3.6 FCA vs FCN Taste 3.7 3.3 Appearance 3.2 3.7 Overall acceptability 4 3.2 FCA vs FCC Taste 3.7 3.6 Appearance 3.2 3.4 Overall acceptability 3.2 3.5

These results suggested that Prosopis, Neem and coconut husks can also be used to smoke both freshwater and marine catfish without differences in sensory values being noticed. Inspite of Prosopis, Neem, Acacia and Coconut husks being used for smoking fish, the components of their smokes have not been studied in relation to fish smoking. The difference in taste observed between fresh and marine catfish smoked with Acacia could be due to the different compounds in Acacia smoke interacting differently with body components of the marine and freshwater catfish. Different wood sources give wood smokes that have distinctly different sensory properties. Many factors influence the quality of smoked fish products including the properties of fish flesh, maturity, age, sex, seasonal variations and factors involved in the smoking procedure such as wood type, composition of smoke, temperature, humidity, velocity and density of the smoke (Simko, 2005). Phenolic and carbonyl compounds contribute towards taste in smoked fish (Maga and Fopajuwo, 1986; Martinez et al, 2007). Organoleptic properties of smoked foods are decisively influenced by composition of the smoke and nature of wood involved. There is no agreement about which wood or mixture of woods imparts the preferred sensorial properties to smoked woods (Guillén and Manzanos, 2002). There was still a better score for marine catfish meaning that marine catfish competes well with freshwater catfish even when Acacia is used for smoking. The fact that there is no significant difference in sensory attributes between freshwater catfish and marine catfish during smoking in this study indicates that smoked marine fish can be used as a fish product in the market just like the freshwater catfish. There could just be a psychological feeling towards inferiority of marine catfish. This could basically be due to the influence in culinary habits and customs of each region on food

15 acceptability Giullén and Manzanos, 2002). The marine catfish therefore is a protein source being ignored and more work needs to be done towards optimization of processing. Neem, Prosopis and Coconut husks have potential for use in fish smoking can be analyzed further for intrinsic properties in smoking marine and freshwater catfish.

4.2 Insect attack during storage

Insects and mites are often found infesting during and after processing. In this study, no identification was made on beetles or mites and all were lumped as “insects”. Insects were first detected in the marine catfish smoked by Acacia and Prosopis, on day 35 of storage. The score for insects for Acacia and Prosopis for smoked marine catfish was 1 (Table 4). The marine catfish smoked by Neem and Coconut husks had no signs of insect presence on day 35. The freshwater catfish smoked by Acacia, Neem and Coconut had no presence of insects on them. Only the freshwater catfish smoked by Prosopis had the presence of insect on them by this day. The score for insects for Prosopis smoked freshwater catfish was 2. Scores of 2 and below were still considered acceptable though it suggested heavy presence. The marine catfish smoked by Acacia, Prosopis and Neem had insect scores of 2 on day 48 of storage. This was the first time that the Neem smoked marine catfish showed any presence of insects. The marine catfish smoked by Coconut husks also showed the presence of insects with a score of 1. The freshwater catfish smoked by Acacia and Coconut husks had a score of 1, Prosopis a score of 3 and Neem a score of nil on day 48 of storage. This was the first time that Prosopis smoked freshwater catfish reached the threshold for rejection. On day 56, the marine catfish smoked by Acacia and Prosopis had a score of 3 Neem and coconut husks had insect scores of 2 and 1 respectively. The marine catfish smoked by Acacia and Prosopis reached the rejection point on day 56 in respect of insect presence. The insect scores for marine smoked coconut and Neem were still within acceptable levels with scores of 2 and below. On day 56 the freshwater catfish smoked by Acacia and Coconut husks had a score of 2 and 3 respectively. The freshwater smoked catfish using coconut crossed the threshold value of acceptance. The freshwater smoked Neem had a score of 1. This was the first time the Neem smoked freshwater catfish showed any sign of insect presence. The coconut smoked marine catfish also had a score of 1 on the 56th day of storage. Insect development during storage of

16 cured fish in the open is influenced by temperature, relative humidity and moisture content (Haines and Rees, 1989, Daramola et al, 2007). Optimal temperatures for insect development in such cured fish are between 25 to 350C (Haines and Rees, 1989). The mean humidity during the entire 56 day storage period was 81.2% ±1.27 with a high frequency between 80-81% while temperature was 26.330C ±1.73 with a frequency of 26 to 27.5 (Figures 3 and 4).

84

83

82

81

80

79 % Humidity % 78

77

76

75 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 Days during storage Figure 3 Humidity during storage of fish 30

25

20

15

Temperature oC Temperature 10

5

0 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 Days of Storage Figure 4 Temperature during storage of fish

In this study, both storage temperature and humidity favoured development of insects during storage. The Neem tree in this study is a better insect deterrent during storage in both fresh and marine catfish owing to the limited insect attacks observed during storage.

17

Several Neem compounds that prevent insect infestation have been identified in the Neem tree (Bhattacharya et al 2007; Mordue and Nisbet 2000; Biswas et al, 2002, Singh and Singh, 1998). It can be postulated that the Neem smoke still contains bioactive ingredients which when deposited on fish muscle during smoking prevents insect infestation.

Table 4 Insect scores during storage in open air Day 0 Day 14 Day 21 Day 28 Day 35 Day 48 Day 56 Insect score MCA 0 0 0 0 1 2 3 FCA 0 0 0 0 0 1 2 MCP 0 0 0 0 1 2 3 FCP 0 0 0 0 2 3 3 MCN 0 0 0 0 0 2 2 FCN 0 0 0 0 0 0 1 MCC 0 0 0 0 0 1 1 FCC 0 0 0 0 0 2 3

4.3 Mould attack during storage

Reports on the role of moulds in the spoilage of smoked foods and their mycotoxigenecity is scanty (Essien et al 2005). The marine catfish and freshwater catfish were attacked by mould on different days over a period of 56 days (Table 5). The differences in mould attack could be due to differences in active microbial components in the respective smoke sources (Guillen and Manzanos, 2002).

Table 5 Mould scores and % of fish attacked during storage in open air Day 0 Day 14 Day 21 Day 35 Day 48 Day 56 Mould % Mould % Mould % Mould % Mould % Mould % score present score present score present score present score present score present MCA 0 0 0 0 0 0 0 0 1 20 2 60 FCA 0 0 1 10 1 20 1 10 2 50 3 70 MCP 0 0 0 0 0 0 2 50 3 80 3 100 FCP 0 0 1 10 1 10 2 50 3 80 3 100 MCN 0 0 0 0 0 0 0 0 2 50 3 70 FCN 0 0 0 0 0 0 0 0 0 0 1 20 MCC 0 0 0 0 0 0 1 20 0 20 1 20 FCC 0 0 1 20 1 20 0 40 3 80 3 90

Carboxylic acids and phenols have been reported to show the highest antimicrobial activity with carboxylic acids lowering the PH while phenols inhibit (Benjakul and Aroonrueng, 1999; Guillen and Manzanos, 2002). The marine catfish smoked by Acacia and

18

Neem and stored on open benches simulating storage conditions the fish are subjected to in the rural villages had first signs of mould growth on them on day 48 of storage. The mould covered more than 20 and 50% of body surface respectively. Those that were smoked with Prosopis and Coconut husks had mould attack on them by day 28 of storage. The mould covered 20 and 10% of body surface respectively. The freshwater catfish smoked by Acacia, Prosopis and Coconut husks and stored in the open benches had mould attack on them by day 14 of storage. For serious marketing purposes any fish with mould on them should be rejected though at community level marketing they would be re-smoked and sold. The mould covered 10, 10 and 20% of body surface respectively. For purposes of this study any sign of mould attack led to rejection. The freshwater catfish smoked with Neem did not show mould attack during storage till day 56 when 20% of the body surface was attacked. Moulds grow on smoked fish whenever there is enough moisture. Moulds also like humid and damp conditions. Under normal circumstances when moulds occur, they are relatively easy to remove if caught early enough as they attack surface of the fish first and do not immediately penetrate the fish (Ward 2003). This however is unacceptable and proper standards must be met as it is not easy to establish extent of toxin penetration. Among the 33 mould isolated from 20 samples of wood smoked Chlamydoselachus anguincus (shark-fish), 20 isolates were capable of producing toxic metabolites Essien et al (2005).

Figure 5 Mould on MCP No mould on FCN

In this study, mould was seen to attack the fish during storage at different times depending on the tree species used during smoking. The development of mould as a measure of spoilage rate was less pronounced in the fish smoked with Neem whether fresh or marine water catfish. Marine

19 catfish generally stored for a longer time before mould attack than freshwater catfish. When moisture is reduced to 25 % wet basis, contaminating agents cannot survive and autolytic activity is greatly reduced (Bala, 2001). However to prevent mould growth during storage moisture must be reduced to below 15% (Bala, 2001). Daramola et al (2007) indicates that for smoked fish to survive mould attack during storage after a few days, moisture should be below 12%. Other factors however like different arrangements of fish muscles vis-à-vis free/bound waters, and the fat content of fish species could be important factors to storage (Daramola et al, 2007). In this study, after 14 days of storage, moisture was above 15% creating an environment for mould growth. The Neem smoke could have bioactive compounds that prevent mould growth. Coventry & Allan (2001) have reported antifungal activity due to Neem extracts. The Neem smoke probably contains these active antifungal ingredients that are deposited on the fish muscle during storage and subsequently reduces attack by mould.

4.4 Moisture during storage

Storage of smoked fish is accompanied by increase in moisture (Daramola et al 2007). The % moisture in this study was monitored on day 0, 14, 21, 28, 35, 48 till 56 (Table 6). Moisture was plotted only for smoked fish showing no mould attacks during storage (Figure 6 and 7).

Table 6 % Moisture during open air storage Day 0 Day 14 Day 21 Day 35 Day 48 Day 56 MCA 16.1 16.6 16.4 16.3 16.4 FCA 13.3 16.3 MCP 16.1 18.9 16.6 21.7 FCP 16.8 18.2 MCN 16.0 18.8 15.9 20.5 19.0 FCN 16.6 15.6 16.0 18.3 17.7 17.5 MCC 15.8 17.4 16.2 17.9 FCC 15.3 16.6

The moisture content in marine catfish smoked Acacia was 16.1%, 16.6, 16.4, 16.3 and 16.4 % on day 0, 14, 21, 35 and 48 when mould was detected and storage therefore would have been ended. The moisture content of marine catfish smoked with Prosopis was 16.1, 18.9, 16.6 and 21.7 on day 0, 14, 21 and 35 when mould was detected and storage therefore would have been ended. The moisture content in the marine catfish smoked with Neem 16.0, 18.8, 15.9, 20.5 and

20

19.0% on day 0, 14, 21, 35 and 48 when mould was detected and storage therefore would have been ended. The moisture content in the marine catfish smoked with Coconut husks 15.8, 17.4, 16.2, 15.9, 20.5 and 19.0% on day 0, 14, 21, 35 and 48 when mould was detected and storage therefore would have been ended.

25.0

20.0

15.0

% Moisture % 10.0

5.0

0.0 Day 0 Day 14 Day 21 Day 35 Day 48 Days of storage MCA MCP MCN MCC Figure 6 % Moisture during storage of freshwater catfish smoked with Acacia, Prosopis, Neem and Coconut husks

The moisture content in freshwater catfish smoked Acacia was 13.3% and 16.3% on day 0 and 14 when mould was detected and storage therefore would have been ended. The moisture content of freshwater catfish smoked with Prosopis was 16.8 and 18.9% on day 0 and 14 when mould was detected and storage therefore would have been ended. The moisture content in the freshwater catfish smoked with Neem was 16.0, 15.6, 16.0, 18.3, 17.7 and 17.5% on day 0, 14, 21, 35, 48 and 56 when mould was detected and storage therefore ended. The moisture content in the marine catfish smoked with Coconut husks was 15.3 and 16.6, on day 0 and 14 when mould was detected and storage therefore would have been ended. It can therefore be assumed that the fish muscle absorbed moisture during storage. Such increase in levels of moisture in smoked fish have been reported by Ikeme (1996).

21

20.0

18.0

16.0

14.0

12.0

10.0

% Moisture % 8.0

6.0

4.0

2.0

0.0 Day 0 Day 14 Day 21 Day 35 Day 48 Day 56 Days of Storage FCA FCP FCN FCC Figure 7 % Moisture during storage of marine catfish smoked with Acacia, Prosopis, Neem and Coconut husks

Mould attack is generally related to moisture content or moisture (humidity) in the environment. Humidity and temperature also influence moisture content. When the relative humidity is greater than 75%, rehydration of dried fish occurs if product is not wrapped in polythene or any cover and in an atmosphere with more than 60% humidity, the fish will tend to pick up moisture with consequent risk of spoilage (FAO 1981; Khan & Khan 2001; Benjankul and Aroonrueng 1999). This is because fish muscle is normally hygroscopic. The hygroscopic nature of dried fish muscle and the high humidity played a part in increasing moisture thus creating an environment for mould attack. In this study the final moisture content of the marine and freshwater catfish was between 13 to 16%. Daramola et al (2007), stored five different smoked fish with initial moisture contents of 10.4% that increased to 10.7% over 6 weeks storage period and still had mould attack. Increase in moisture content could be attributed to the difference in the moisture of the smoked fish relative to the surroundings. According to Jallow (1995), fish at 10-15% moisture content had a shelf life of 3-9 months when stored properly. The trees used in smoking the fish in this study were suitable because they all attained suitable moisture values recommended for storage during the smoking period (Bala, 2001; Sankat and Mujaffar (2005).

4.5 Burning properties The burning properties and energy composition values of the different trees used in this study are shown in table 7

22

Table 7 Burning properties of wood Acacia, Prosopis, Neem and Coconut husk Species Smoke Flame Ash Acacia raddiana Light-grey smoke Luminous Burns with a lot of soot hot yellow completely into ash.

Cocos nucifera A lot of soot Luminous flame. Burns completely (Naazi) produced and grey- Burns rapidly like into Ash light smoke oily containing Substance

Azadiratchta indica (Neem) Light grey smoke -Luminous yellow Does not burn into with soot complete Ash. Black residue remains.

Prosopis julifora Burns with a lot of Strong Luminous Burns (Mathenge) soot and a light grey flame produced Completely smoke produced. into ash.

Acacia burnt with light grey smoke with a lot of soot, had luminous hot yellow flame and burnt completely to ash. Prosopis burnt with a light grey smoke, a lot of soot a strong luminous flame and burnt completely to ash. Neem burnt with light grey smoke with soot, a luminous yellow flame and did not burn completely to ash with a black residue remaining. Coconut husks burnt with a light grey smoke, a lot of soot, a luminous flame that burnt rapidly like an oily substance and burnt completely to ash. All the four tree species burnt with light grey smoke. Only Neem had less soot. Acacia and Prosopis burnt with a luminous hot yellow flame and strong luminous flame. Only Neem did not burn completely to ash and left some soot. The soot could be due to higher moisture content in Neem (Baker, 1982). The performances are ranked as shown in table 8. Table 8 Ranking of burning properties of wood Acacia, Prosopis, Neem and Coconut husk Species Smoke Flame Ash Total Rank Acacia raddiana 1 1 1 3 1 Prosopis julifora (Mathenge) 1 1 1 3 1 Azadiratchta indica (Neem) 2 2 3 7 3

Coconut nucifera 1 3 1 4 2

23

It would be difficult to compare these burning properties in the laboratory with actual field conditions where the experiments were done because other factors like wind, size of fish and composition of fish come into play. Acacia and Prosopis had better burning properties than Neem and Coconut husks. This part of the study gives Prosopis a chance to be utilized in fish smoking Other properties of the tree analyzed were % moisture, volatile matter, ash, fixed carbon, calorific value and density (Table 9).

Table 9 Moisture, Ash, volatile matter, fixed carbon and calorific value of the different tree species Sample %MC %AC CV Density FVI

Acacia raddiana 7.23±0.42 4.73±0.35 4.90±0.03 0.85 0.12

Coconut nucifera 6.70±0.07 5.53±0.09 5.02±0.02 0.09±0.009 0.01

Azadiratchta indica (Neem) 8.71±0.11 3.68±0.40 4.96±0.01 0.51 0.08

Prosopis julifora (Mathenge) 5.62±0.59 4.60±0.11 4.92±0.04 0.95 0.18

MC=Moisture content; AC= Ash content; CV = Calorific value,

Emphasis was laid on Density, Moisture, Calorific Value and Ash. That is the data reported. Acacia had a moisture content of 7.23±0.42, Prosopis 5.62±0.59, Coconut husks 6.7±0.07 and Neem 8.7±0.11. Prosopis was the most dry of the tree species as at time of collection from field followed by Coconut husks then Acacia then Neem. Calorific value for Acacia was 4900 kcal/Kg, Prosopis 4920 kcal/Kg, Neem 4960 kcal/Kg and Coconut husks 5020 kcal/Kg. Coconut husks had the highest calorific value followed by Neem, Prosopis then Acacia. The density of Acacia was 0.85g/cc, Prosopis 0.95g/cc, Neem 0.51g/cc and Coconut husks 0.09± 0.009g/cc. Prosopis had a higher density than Acacia than Neem. The ash content of Acacia was 4.73% ± 0.35, Prosopis 4.6% ± 0.11, Neem 3.68% ±0.40 and Coconut husks 5.53%±0.09. Ash was highest in Coconut husks followed by Acacia, Prosopis and Neem. Such findings have been reported by Puri et al 1994; Pasieznick et al 2006; Fact Sheet 97; Ndayambaje, 2005, Oduor & Githiomi, 2003; Baker, 1982). Ash content was rather high in all the species analyzed. The tree components included bark. Baker (1982) reported that ash content including bark is about 3%. The fuel wood value index was calculated taking into account the calorific value and density as

24 positive characteristics and high water content and high ash content as negative characteristics (Puri et al, 1994). The FVI for Acacia was 0.12, Prosopis 0.18 Neem 0.08 and 0.01. Prosopis had the highest FVI followed by Acacia then Neem then Coconut husks. A Summary of these findings is shown in table 10 where the tree species are ranked with respect to their fuel wood characteristics. This table includes the parameters studied and shows “winners”. Puri et al (1994) used such rankings for over 10 wood fuel species.

Table 10 Fuel wood ranking of Acacia, Prosopis, Neem, Coconut husks on the basis of Calorific value, Density, ash, Moisture and Fuel value Index. The ranking ranges from 1 (best) to 4 (last) Species Moisture Calorific value Density Ash FVI Total Rank Acacia 3 4 2 3 2 12 2 Prosopis 1 3 1 2 1 7 1 Azadirachta 4 2 3 1 3 14 3 Coconut husks 2 1 4 4 4 15 4

Since calorific value, density and ash content are important parameters in this kind of study, they could be weighted more heavily in the final decision. In this study, the rankings would change as shown in table 11 (Puri et al,1994). There would be no significant change for ranking of Prosopis but Acacia would be below Neem tied with Coconut husks.

Table 11 Fuel wood ranking of Acacia, Prosopis, Neem, Coconut husks on the basis of Calorific value, Density and ash. The ranking ranges from 1 (best) to 4 (last) Species Calorific value Density Ash Total Rank Acacia 4 2 3 9 2 Prosopis 3 1 2 6 1 Azadirachta 2 3 1 6 1 Coconut husks 1 4 4 9 4

For an ideal fuel wood however, high calorific value, high density of wood, low ash content, low N percentage and low water content are the most desirable characteristics (Bhatt and Todaria, 1990). Wood density correlates positively with calorific value (Deol, 1983; Jain, 1990, Singh et al, 1984, Tietema et al, 1991). Most of the variation in calorific values is due to inherent

25 differences in cellulose, lignin, hemicellulose and wax or resin composition of different species (Tillman, 1978) and is unaffected by age of trees between 1 to 4 years (Bewbaker et al, 1984). Fuel wood value index can be calculated to give preference for woods (Puri et al, 1994) and calorific value, density and ash content are the most important parameters in this kind of study and should be weighted more heavily in the final decision (Puri et al, 1994). In this part of the study, Acacia and Prosopis as well as Neem now compete well with Prosopis showing good fuel wood properties. These tree properties do not necessarily transform to any advantage during actual smoking because other quality issues come into play.

4.6 Wood fuel consumption

The quantities of wood fuel are as shown in table 12. The quantity of Acacia used to smoke marine and freshwater catfish was 34.5kg and 35.5kg respectively. While for Prosopis 31.5kg and 30.5kg was used. The quantity of Neem used to smoke marine catfish and freshwater catfish was 32kg and 30kg respectively. For coconut husks, there was a substitution with Acacia after 16 kg had been used. There was not much variation seen in the quantities of wood used during smoking. Similar weights for Acacia had been reported by Oduor Odote el al (2008) during smoking of freshwater catfish though using Acacia. Prosopis and Neem tree species compared well in quantity for use during fish smoking. The difference seen in weight could be due to their burning properties and non uniform starting weights of fish as well as heterogeneity of the fish structure leading to different drying rates and patterns. It is still possible to optimize use of Coconut husks in future studies.

Table 12 Wood fuel used during smoking (Weights in Kg) Starting Weight Unused wood Used wood Weight of ash after smoking MCA 37 2.5 34.5 1.8 FCA 37 1.5 35.5 1.5 MCP 32 0.5 31.5 1.5 FCP 32 1.5 30.5 1.2 MCN 33 1 32 1.7 FCN 33 3 30 1.4 MCC+ *(Acacia) 18 (16) 0 16 0.4 (Coconut only) FCC + *(Acacia) 16 (16) 0 16 0.4 (Coconut only) * On day 3 Acacia was added to Coconut husks because they burnt fast

26

4.7 Drying characteristics

The freshwater and marine catfish are treated separately owing to differences in weight at the start of smoking brought about by difficulty in synchronizing availability of both species. The weight loss of the marine catfish smoked with Acacia was from 383.3g± 28.9 to 116.7± 28.7. This was equivalent to 70% (Table 13). The time taken to smoke the marine catfish to a weight loss of 90% was 68hr, to a weight loss of 95% was 74hr and to constant weight loss then stored was 74hr. Table 13 Weight loss, % water (g) in fish after smoking, final weight after smoking, hours of smoking to 90%, 95% and constant weight during smoking of African and Marine Catfish

MCA MCP MCN MCC FCA FCP FCN FCC Initial 383.3±28. 275±2 425±2 441.6±38. 600±86. 583.3±80. 750±75 566.6±57. fresh 9 5 5 2 6 4 7 weight of fish Final 116.7±28. 110±1 125±2 125±25 225±25 185±62.6 241.6±38. 188.3±37. weight 7 0 5 1 5 after smoking % 16.1 16.1 16.0 15.8 13.3 16.8 16.6 15.3 Moistur e after smoking % 70 60 70.1 71.7 62.5 68.3 67.8 66.7 Weight loss Hours to 68 44 50 44 44 66 66 46 lose 90% weight Hours to 74 48 50 48 50 68 74 72 lose 95% weight Hours to 74 68 66 74 70 70 74 72 constant weight

The weight loss of the marine catfish smoked with Prosopis was from 275g± 25 to 110g± 10. This was equivalent to a 60% weight loss. The time taken to smoke the marine catfish to a weight loss of 90% was 44hr to a weight loss of 95% was 48hr and to constant weight loss then stored was 68hr. The weight loss of the marine catfish smoked with Neem was from 425g± 25 to 125± 25. This was equivalent to 70.1 % weight loss. The time taken to smoke the marine catfish

27 to a weight loss of 90% was 50hr to a weight loss of 95% was 50hr and to constant weight loss then stored was 66hr. The weight loss of the marine catfish smoked with coconut husks was from 441.7g± 38.2 to 125± 25. This was equivalent to 71.7%. The time taken to smoke the marine catfish to a weight loss of 90% was 44hr to a weight loss of 95% was 48hr and to constant weight loss then stored was 74hr. The weight loss of freshwater catfish smoked with Acacia was from 600± 86.6 to 225±25. This was equivalent to 62.5%. The time taken to smoke the freshwater catfish to a weight loss of 90% was 44hr to a weight loss of 95% was 50hr and to constant weight loss then stored was 70hr. The weight loss of the freshwater catfish smoked with Prosopis was from 583.3g±80.4 to 185g±62.5. This was equivalent to 68.3%. The time taken to smoke the freshwater catfish to a weight loss of 90% was 66hr to a weight loss of 95% was 68hr and to constant weight loss then stored was 70hr. The weight loss of the freshwater catfish smoked with Neem was from 750g±75 to 241.6g±38.2 equivalent to 67.8%. The time taken to smoke the freshwater catfish to a weight loss of 90% was 66hr to a weight loss of 95% was 74hr and to constant weight loss then stored was 74hr. The weight loss of the freshwater catfish smoked with coconut was from 566.6g±57.7 to 188.3g±37.5 Equivalent to 66.7%. The time taken to smoke the freshwater catfish to a weight loss of 90% was 46hr to a weight loss of 95% was 72hr and to constant weight loss then stored was 72hr. The tree species used in the study smoke dried the fish to weight losses of 60 to 70% over a period of 66 to 74 hours. Although Coconut husks burnt fiercely at first, its burning was controlled by closing the stoke hole of the oven more often than the for the other tree species and supplement with Acacia. The moisture loss rates were closer when the range of starting weights of fish was not wide (Table 14 Figures 8 and 9). The starting weight of for marine catfish smoked with Neem and Coconut husks was 425g and 441g respectively. Their drying or moisture loss rates were quite close. The same was observed for freshwater catfish smoked with Acacia and Prosopis with weights of 600g and 583g respectively. The wider margins in drying rates were observed between marine catfish smoked Acacia and marine catfish smoked Prosopis with starting weights of 383g and 275g respectively as well as freshwater catfish smoked Neem and Coconut husks with starting weights of 750g and 566g respectively. The time taken to reach constant weight during smoking was between 66hr to 74 hr with most tree species attaining constant weight between 70 to 74hr except for marine catfish smoked with Prosopis and Neem with 68hr to 66hr respectively. The starting weights of marine catfish

28 smoked with Prosopis were slightly smaller and by inference thinner at the start of the experiment. Thinner or smaller fish would normally dry faster than bigger or thicker ones. The surface area to volume ratio of smaller fish is normally higher resulting in faster drying rates (Mujaffar and Sankat, 2005). This did not happen despite the fact that the fish had a greater surface area to volume ratio. This means that other factors come into play during smoking. The heterogeneous nature of smoking, differing quantities of wood fuel, temperatures in smoking ovens, uneven fish surfaces may not allow for easy comparison of water loss rates but the results seem to suggest that all the trees compete well during smoke drying and can therefore be used in smoking of fish. No tree seems to enjoy undue advantage over the other in terms of moisture loss. The rate of water loss (g/hr) in the marine and freshwater fish is as shown in table 14

Table 14 Drying rate of marine catfish smoked by Acacia, Prosopis, Neem and Coconut husks 2 MCAwater loss rate y =14.3ln(x) + 36.05 R = 0.50 2 MCPwater loss rate y=-6.02 ln(x) + 16.11 R = 0.48 2 MCNwater loss rate y=-10.1ln(x) + 27.18 R =0.56 2 MCCwater loss rate y=-10.6 ln(x) + 28.46 R =0.62 2 FCAwater loss rate y = - 15.1 ln(x) + 40.47 R = 0.52 2 FCPwater loss rate y = -15.7ln(x) + 41.06 R =0.54 2 FCNwater loss rate y=-26.6 ln(x) + 66.50 R = 0.57 2 FCCwater loss rate y= -14.1ln(x) +38.58 R = 0.43

70

60

50

40

30

20 Water loss rate (kg/s)rate loss Water 10

0 0 2 17 19 21 23 25 27 29 31 33 35 37 39 53 55 57 59 Drying time (hrs) MCN MCA MCP MCC Figure 8 Rate of water loss for freshwater catfish smoked with Acacia, Prosopis, Neem and Coconut husks

29

140

120

100

80

60

40 Water loss rate (g/hr)rate loss Water

20

0 0 2 17 19 21 23 25 27 29 31 33 35 37 39 53 55 57 59 Drying time (hrs) FCN FCA FCP FCC Figure 9 Rate of water loss for marine catfish smoked with Acacia, Prosopis, Neem and Coconut husks

5.0 Conclusions

Acacia is the tree commonly used for smoking fish in the Tana delta region. The performance of Prosopis, Neem and Coconut husks is commendable especially in terms of organoleptic properties like appearance and taste. There was no difference in overall acceptability of the fish smoked using the different trees. Freshwater catfish is commonly smoked in the North coast of Kenya while the marine catfish is sun dried. There was no significant difference in appearance and overall acceptability between marine and freshwater catfish. The difference in taste between marine and freshwater catfish during smoking using Acacia even showed that marine catfish was preferred to freshwater catfish. The Neem tree smoked marine and freshwater catfish stayed for a longer time before mould or insect attack suggesting presence of antifungal and insect infestation deterrent factors in the Neem smoke. All the tree species had burning and energy properties that are within acceptable range and their fuel wood consumption was not varied. The drying characteristics of the trees used dried fish to acceptable moisture levels for storage.

6.0 Acknowledgements

WIOMSA, Director-KMFRI, Maurice Obiero, Maurice Omega, Cyprian Odoli, Janet Ntabo, Florence Waithera, Dick Odongo, Moa BMU

30

7.0 References

Akande G.R. Oladosu, O.H; Tobor, J.G. (1996). A comparative Technical and economic appraisal of fish smoking: Two traditional ovens and a new improved Magbon-Alade oven. FAO expert consultation on Fish Technology in Africa. Kisumu, Kenya. 27th- 30th August 1996. FAO Fisheries Report No. 574

AOAC (1990). Official Methods Of Analysis: Association Of Official Analytical Chemists International, Airlington, U.S.A.pp 931. 341-342; 449.

Baker A.J. (1982). Wood fuel Properties and Fuel Products from Woods, In: Fuelwood Management and Utilization Seminar: Proceedings. East Lansing, M.I. 1982 November 9-11 MI Michigan State University, 1983: 14-25

Bala, B.K.; Mondol, M.R.A (2001). Experimental Investigation on Solar Drying of Fish Using Solar Tunnel Dryer. Drying Technology 19 (2). 427-436.

Battacharyya N.; Chutia, M and Sarma S (2007). Neem (Azadirachta indica A. Juss), A potent Biopesticide and Medicinal Plant: A Review. Journal of Plant Sciences 2 (3) 251-259

Benjakul S and Aroonrueng N (1999) Effect of Smoke Sources on Quality and Storage Stability of Catfish Fillet (Clarias macrocephatus Gunther). Journal of Food Quality 22 213-224

Bewbaker, J.L.; Beldt R van den and Macdicken, K (1984). Fuelwood Uses and Properties of Nitrogen-Fixing Trees. Pesq AgroPec Brais Brasilia 19: 193-204

Bhatt, B.P and Todoria, N.P. (1990). Fuelwood Characteristics of some Mountain Trees and Shrubs. Biomass 21: 233-238

Biswas K, Chattopadhyay I.; Barnajee R.K and Bandyopadhyay U (2002). Biological Activities and Medicinal Properties of Neem (Azadirachta indica). Current Science Vol 82 No 11 1336-1345

British standard 3631 (1973

Brownell, B.; Lopez, J. (1985). The Chorkor Fish Smoking Method:A Truely Appropriate Technology. Proceedings of The FAO Expert Consultation on Fish Technology in Africa. Lusaka, Zambia. 21st-25th June 1985

Cardinal, M.; Cornet, J.; Serot, T and Baron R (2006). Effects of the Smoking Process on Odour Characteristics of Smoked (Clupea harengus) and Relationships with Phenolic Compound Content. Food Chemistry 96 137-146

Cauri M.; Okoso-Amaa K.; Chichester C.O.; Lee T.C (1979) Artisan Technology: Ghana-A Case Study of West African Fishery. University of Rhode Island, Kingston.

Coventry E and Allan E.J (2001). Microbiological and Chemical Analysis of Neem (Azadirachta indica) Extracts: New Data on Antimicrobial activity. Phytoparasitica 29 (5) 1-10

31

Daramola, J.A.; Fasakin, E.A.; Adeparusi, E.O. (2007). Changes in Physicochemical and Sensory Characteristics of Smoke- Dried Fish Species Stored at Ambient Temperature. African Journal of Food Agriculture Nutrition and Development Vol 7 No 6 Deol, G.S. (1983). Fuelwood Plantations to Alleviate Energy Crisis. Indian Journal of Forestry

Essien J.P.; Ekpo, M.A and Brooks, A.A. (2005). Mycotoxigenic and Proteolytic potential of Moulds Associated with Smoked Shark Fish (Chlamydoselachus anguincus). Journal of Applied Science and Environmental Management Vol 9 (3) 53-57

FACT SHEET-Fact Net (1997) Azadiratcha indica- Neem. A Versatile Tree for the Tropics and Sub tropics-A Quick Guide to Multipurpose Trees from around the World 5pp. A publication of FACT NET, Arkansas, USA.

FAO (1981). The Prevention of Losses in Cured Fish. FAO Fisheries Technical Paper 219. FAO Rome

Giullén M.D and Manzanos, M.J (2002). Study of the Volatile Composition of an Aqueous Oak Smoke Preparation. Food Chemistry 79 283-292

Haines, C.P and Rees D.P (1989) A Field Guide to the type of Insects and Mites Infesting Cured Fish. FAO Fisheries Technical Paper No. 303

ICIPE (1995). An Affordable, Efficient and Environmental-Friendly Source of Pest Control Products. International Center of Insect Physiology and Ecology.

Ikeme , A.I. (1992) Cooperative Research Programme on Fish Technology in Africa: A Review: Proceedings of Symposium on Post-harvest Fish Technology. CIFA FAO CIFA Technical Paper FAO No. 19 Symposium on Post Harvest Technology Cairo, Egypt Oct 21-22 Teutscher F (Ed) Rome pp103-112 Jain, R.K. (1990). Fuel wood Characteristics of Trees Grown on Alkaline Soils of Lucknow. Range Manage and Agroforestry 11:95-98. Jallow, A.M. (1995). Contribution of Improved Chorkor Oven to Artisanal fish Smoking in the Gambia. Workshop on Seeking Improvement of in Fish Technology in West Africa. IDAF Technical Report. No 66.

Jónsdóttir, R.; Olafsdóttir, G.; Chanie, E.; Haugen, J.E. (2008). Volatile Compounds Suitable for Rapid Detection as Quality Indicators of Cold Smoked ( salar). Food Chemistry 109 184-195.

Khan, A. A. M.; Khan, A.S.Y. (2001). Insects Infestation and Preventive Measures in Dry Fish Storage of Chittagong, Bangladesh. Journal of Biological Sciences 1 (10) 963-965

KMFRI Ungwana Bay Report (2002) Current Status of Fisheries in the Ungwana bay- Malindi area. A Report by Kenya Marine and Fisheries Research Institute, P.O.Box 81651, Mombasa. 199p

Lartey, A.; Asiedu, M.; Okeke, E.C. (1994). Assessing the Effects of Different Smoking Ovens and Fuel Wood Types on Fish Nutritional Quality. Progress Report Submitted to

32

Ghana/Netherlands Artisanal and Applied Research Project in Africa. University of Ghana. Legon.

Maga J.A and Fopajuwo O.O. (1986). Aroma Intensities of Various Wood Smoke Fractions. Journal of Sensory Studies Vol 1 Issue 1 9-13 (Abstract only)

Martinez O, Salmeron J,; Guillen M.D.; Casas C (2007) Sensorial and Physicochemical Characteristics of Salmon (Salmo Salar) treated by different Smoking Processes during Storage. Food Science and Technology International 13 (6) 477-484

Martinsdottir,E.; Sveinsdottir, K.; Luten, J.; Schelvis-Smit, R.; Hyldig, G. (2001).Sensory Evaluation of Fish Freshness. Reference Manual for the Fish Sector. QIM EUROFISH. 49pp

Mordue, J.A. and Nisbet J.A (2000). Azadirachtin from the Neem Tree Azadirachta indica: Its Action against Insects. An. Soc. Entomol. Brasil 29 (4) 615-632

Mujaffar, S and Sankat, C.K. (2005). The Air Drying Behaviour of Shark Fillets Canadian Biosystems Engineering Vol. 47. 3.11-3.21

Ndayanbaje J.D (2005). Agroforestry for Wood Energy Production in Rwanda. A paper presented in the Workshop on Alternative Sources of Energy in Rwanda. Organized by IRST 19th -20th May CENTRE IWACU KABUSUNZU

Nerquay-Tetteh G.A. (1979). The Traditional Post-harvest Fish Processing Technology in Ghana. FRI Project Report. Food Research Institute, Accra

Oduor N and Githiomi J (2003) Utilization of Potential of Prosopis julifora grown in Kenya Drylands. In Proceedings of the Workshop on Integrated Management of Prosopis Species in Kenya. SOI Safari Club, Lake Baringo, Kenya. 1st-2nd October 2003 p 41-51

Oduor-Odote P.M (2006). Improved Traditional Fish Processing Methods by Smoking and Solar Drying in Tana River and South Coast areas of Kenya. WIOMSA, MARG 1 Report. 66pp.

Oduor-Odote, P.M; Ohowa, B.O.; Odoli, C.A.; Obiero, M. (2007). Performance of Improved Smoking Kilns Introduced at the Tana-Delta Region of Kenya. A Perfomance Contract Technical Report by KMFRI To Ministry of Livestock and Fisheries, Kenya.

Oduor-Odote, P.M; Ohowa, B.O.; Obiero, M. (2008). Performance of Improved Smoking Kilns Introduced at the Tana-Delta Region of Kenya. Samaki News- A Magazine of the Department of Fisheries, Kenya. Vol V No 1 April p 23-28

Okraku-Offei G.A (1970). Processing and Preservation of Fish in Ghana. FRI Project Report. Food Research Institute, Accra.

Osuji, F.N.C. (1977). The Influence of Traditional Handling Methods on the Quality of Processed Fish in Nigeria. Proceedings of Conference on Handling, Processing and Marketing of Tropical Fish.319-322. London Tropical Products Institute.

33

Pasiecznik, N.M.; Choge, S.K.; Muthike, G.M; Chesang, S.; Fehr, C.; Bakewell-Stone, P.; Wright, J.; Harris, P.J.C. (2006). Putting Knowledge on Prosopis into Use in Kenya.

Puri, S.; Singh, S.; Bhu Shan, B. (1994). Evaluation of Fuelwood Quality of Indeigenous and Exotic Tree Species of India’s Semi-arid Region. Agroforestry Systems 26: 123-130

Proctor, D.I. (1977). The control of Insect Infestation of Fish during Processing and Storage in the Tropics. Proceedings of Conference on Handling, Processing and Marketing of Tropical Fish.307-311. London Tropical Products Institute.

Sharing Innovative Experiences (1999a). “Improved Fish smoking in Ghana”Science and Technology, Chapter 3 Vol. 5. pp. 240-254

Simko, P (2005). Factors affecting elimination of Polycyclic Aromatic Hydrocarbons from Foods and Liquid Smoke Flavourings. Mol. Nutr. Food Res. 2005 49 637-647-A Review

Singh, S and Singh R.P. (1998). Neem (Azadirachta indica) Seed Kernel Extracts and Azadirachtin as Oviposition Deterrents against the Melon Fly (Bactrocera cucurbitae and the Oriental Fruit Fly (Bactrocera dorsalis).

Singh, B.; Khanduja, S.D and Srivastava, G.S. (1984). Qualitative Analysis of some Firewood Shrubs. Biomass 5: 317-320

Tietema, T.; Dithogo, M,; Tibone, C and Mathalza N. (1991). Characteristics of Eight Firewood Species of Botswana. Biomass and Bioenergy 1: 41-46

Tillman, D.A. (1978). Wood as Energy Source. Acedimic Press, Ney York. 252 pp

Ward A (2003) A study of the Trade in Smoked Dried Fish from West Africa to the United Kingdom. FAO Fisheries Circular No 981 FIIU/C981 17pp

34

Annex 1 CONSUMER RESPONSE TO SMOKE DRIED MARINE & FRESHWATER CATFISH (TICK SCORE). CODE FOR SPECIES DATE NAME (OPTIONAL) ATTRIBUTE RESPONSE SCORE TASTE Extremely acceptable 5 Very acceptable 4 Acceptable 3 Fairly acceptable 2 Not acceptable 1 Mean

TEXTURE (FIRM) Extremely acceptable 5 Very acceptable 4 Acceptable 3 Fairly acceptable 2 Not acceptable 1 Mean APPEARANCE Extremely acceptable 5 Very acceptable 4 Acceptable 3 Fairly acceptable 2 Not acceptable 1 Mean OVERALL ACCEPTABILITY Extremely acceptable 5 Very acceptable 4 Acceptable 3 Fairly acceptable 2 Not acceptable 1 Mean

35

Annex 2 INSECT /MOULD INFESTATION OF STORED MARINE & FRESHWATER CATFISH IMPROVED OVEN SMOKED MARINE OR FRESHWATER CATFISH (Tick) TREE TYPE FISH SPECIES NAME NAME OF INSECT SPECIES NAME OF INSECT SPECIES NAME OF INSECT SPECIES NAME OF INSECT SPECIES DATE DAYS OF STORAGE SCORE SCORE FOR NUMERAL CODE FOR INSECT CODE SCORES OBSERVATION MOULDS S No sign of infestation/mould O 0 damage Occasional infestation/mould L 1,2,3 damage M 4,5,6 Noticeable consistent infestation Heavy insect/mould infestation H 7,8,9 covering whole fish