STUDIES ON THE DEGRADATION OF TESTA OF 
COCOA (tHEOBROMA CACAO L .) BY 
ASPERGILLUS, SPECIES ISOLATED 
FROM MOULDY COCOA BEANS
A Thesis presented by 
MICHAEI ABUGRE ABITEY B .Sc. (HONS) AGRIC 
In partia l fu lfilm ent o f the requirements 
fo r  the M.Sc. Degree o f  the 
University o f Ghana 
JULY, 1982
From: The Department o f  Botany
University o f  Ghana 
LEG ON.
University of Ghana                              http://ugspace.ug.edu.gh
G  " m m
/  (X. D«rw \  •
University of Ghana                              http://ugspace.ug.edu.gh
I hereby declare that the work presented in th is th e s is i
STUDIES ON THE DEGRADATION OP TESTA OF COCOA 
(THEOBRGM CACAO L„) BY ASPERGILLUS SPECIES 
ISOLATED FRom MOULDY COCOA BEANS''
was done en tire ly  by me in  the Botany Department, University o f 
Ghana, Legon,
This work has never been presented either in part or 
completely, fo r  any degree of th is University or elsewhere.
MIC HONS) AGRIC
LEGON.
DR. C. E. AMOaKO-NUAMA 
SUPERVISOR
University of Ghana                              http://ugspace.ug.edu.gh
ABSTRACT
Cocoa beans at the Toma Warehouses were found to be contaw 
minated with a to ta l number o f  21 fungus species , the predominant 
species belonging to the genus A spergillus. The annual average 
contamination was 4,9 per centi In fection  in  the main crop was lower 
(3»9 per cent) than in the minor crop (6 ,4  per cen t). The higher 
the bean moisture content the greater the fungal in fe ct ion .
Extracts o f  testa  of fermenting cocoa beans depressed vegeta­
tive  growth in Aspergi l lu s  f l a vus» Aspergillus fumjgatus, Aspergillus 
n iger, Aspergil lu s  t amari i  and Sclerotium r o l f s i i*  iS, r o l f s i i  was 
more inhibited than the AspergiTlua species* Extracts o f testa  o f 
beans already fermented and being dried had no e f fe c t  on growth o f  
mycelium of the Aspergillus species but stimulated that o f S. r o l f s i i .
A. fumjgatus grew best in Sabouraud's broth medium while A. tamarii 
grew best in Czapek Dox broth medium*
Cotyledons and testa  o f  cocoa beans possessed phenolic 
compounds; free  phenols, orthodihydric phenols, flavonols and 
anthooyanins, The quantities o f these compounds changed in  d iffe ren t 
ways during curing of the beansc Free phenols content in  both 
cotyledons and testa  decreased during fermentation and drying o f 
the beans, Orthodihydric phenol leve ls  in both cotyledons and testa , 
on the other hand, increased during fermentation and decreased 
during drying o f the beans„ Flavonol content in  the cotyledons 
increased during fermentation and decreased during drying, while 
i t  decreased in  the testa  during both fermentation and drying.
The anthocyanin content o f the cotyledons decreased with curing 
but increased s ign ifican tly  in  the testa .
Some o f the phenolic compounds inhibited germination o f 
conidia and germ tube growth in A. fla vu s . li. fumigairus. and 
A« tamarii and germination o f s c le ro tia  o f  _§» r o l f s i i  to  varying 
degrees. Others, however, stimulated germination,. The compounds 
in either case were not iden tified*
Fungal contamination o f cocoa beans was observed in unfermented 
fresh cocoa beans and in  beans at various stages o f fermentation 
and dryingt Invsuling fungi occupied mainly the micropylar and 
sty lar regions o f the beans, The composition o f the flora- changed 
as fermentation and drying of the beans progressed.
University of Ghana                              http://ugspace.ug.edu.gh
COHTEOTS
Page
I .  INTRODUCTION ................................................  1
I I .  LITERATURE REVIEW     7
I I I .  MATERIALS AHH GENERAL METHODS . . . . .  23
i )  Materials .........  . . . . .  . . . . .  23
i i )  General Methods . . . . .  . . . . .  23
a) Maintenance o f Stock Culture 23
b) Methods o f inoculation   24
c) Methods o f s te r il is a t io n  . . . . .  24
d) Assessment o f percentage internal' mouldiness
in commercial beans . . . . .  25
e) Assessment o f mycoflora o f cocoa beans 25
f )  Culture o f te s t  fungi on nutrient media 26
g) Assessment o f growth o f cultures on media 26
£) S o lid  medium . . . . . .  26
i i )  Liquid medium . . . . .  . . . . .  26
h) Assessment o f growth on cocoa bean testa  
extracts rvrti 27
i )  Preparation o f  cocoa bean testa  extracts 27
i i )  Spore germination methods . . .  27
i )  Assessment o f spore germination 28
j )  Investigation  of Stage during curing when
contamination was observed in  cotyledons 28
k) Chemical analyses . **« .  29
1. Preparation o f fa t - fr e e  powders fo r
analyses .........    29
i i
University of Ghana                              http://ugspace.ug.edu.gh
Page
i t  Phonolib compounds Hi .I t  29
3, Estimation of Total free  phenols 30
4 , Estimation o f Orthodihydric phenols 31
5 , Flavonol determination by the v an illin
method . . . . .  . . . . .  32
6 , Ln’ihccyanin determination . . . . .  33
l )  Culture Media . .........  35
m) Chemicals «•••« . .  •. • 36
-n) Ch romatographic analyses . . . . .  36
1.. Paper Chromatography . . . . .  36
20 . Thin Layer Chromatography . . . . .  37
p) Experimentall precautions . . . . .  38
q) S ta t is t ica l methods . . . . .  . 3 8
IV,  R E S U L T S      39
As Level of fungal contamination o f cocoa beans
at Tema > « .»e .• • *• .»»«& 39
B* Fungi iso lated  from stored cocoa beans 42
Cs E ffe ct  o f cocoa bean testa  extract on growth
of selected  contaminant fungi . . . . .  45
D. Growth o f the fungi in nutrient media 54
EC; Analysis o f phenolic constituents o f the oocoa
bean testa  . . . . .  . . . . .  . . . .  & 60
i )  Free phenols content . . . . . .  61
i i )  Orthodihydric phenols content 64
i i i )  Flavonol content . . . . .  ■>.».. 67
iv ) Anthocyanin content . . . . .  69
i i i
University of Ghana                              http://ugspace.ug.edu.gh
Page
P. Chromatographic separation and id en tifica t ion
o f phenolic compounds o f cocoa bean testa  72
Gr. E ffect o f testa  components o f  eluted spots o f
th in -layer p lates cn spore germination in  88
selected  cocoa bean contaminant fungi
1, Germination o f conidia and s c le ro t ia  in  
" flo od  water” o f Potato Dextrose Agar 88
2. Germination o f con idia  and s c le ro t ia  in 
th in -layer eluted spots. . . . .  92
H« Fungal invasion o f cotyledon o f  cocoa beans
during curing. ..«•• . . . .  100
V, GBHEEii DISCUSSIONS .............. 104
VI. S U M M A R X  ........................................... 120
V II. ACKNOWLEDGEMENTS . . . . .  . . . . .  127
VIII, LITERATURE CITED.................................. ............  128
IX. APPENDIG3S . . . . .  .............. 157
APPENDIX A. Standard Calibration curve fo r  free
phenols .-.«»# 157
APPENDIX B. Standard Calibration curve fo r  
Orthodihydric phenols . . . . .
APPENDIX C. Standard Calibration  curve fo r  flavonols 1 9^
APPENDIX D. Standard Calibration curve fo r
anthocyanins . . . . .  . . . . .  140
iv
University of Ghana                              http://ugspace.ug.edu.gh
I .  IHTRODTJCTION
Cocoa produced in Ghana, i s  accepted a l l  over the world as 
being o f  very high quality* Writing on the cocoa bean characteris­
t ic s  that make fo r  good quality  cocoa, Wood (1979) stated , . . .  " i t  
is  unusual fo r  any cocoa to exceed the normal figures fo r  Ghana 
cocoa, so, on this bas is , as on others, Ghana is  the premium on 
cocoa". He l is te d  the unacceptable defects o f merchantable cocoa 
beans as mouldiness, s la tin ess , le v e l o f in se c tic id a l chemical 
residues, in sect damage, foreign  odours (smoke) and foreign  
materials (stones, pieces o f wood, e t c ) .  I t  is  recommended that 
commercial cocoa beans be dried to a moisture le v e l below 8 per 
cent so as to  reduce nycoflora l development (Dade, 1929). Of these 
defects , mouldiness is  the most ob jectionable and has been described 
as by far the ;rorst defect o f commercial coooa (Wood, 1965).
Guidelines set out by the Working Party on Cocoa Grading o f 
the Codex Alimentarius Commission* give acceptable standards that 
commercial cocoa has to meet (Anon, 1969). In Ghana, relevant regu­
lations delim it three grades o f cocoa beans fo r  export. Grade I 
cocoa beans should not have more than 3 per cent internal mouldiness, 
and beans with 4 - 8  per cent mouldiness are classed as Grade I I  
(Anon, 1968), Exportable cocoa beans with more than 8 per cent 
internal mouldiness is  graded as sub-standard and so ld  as in fe r io r  
cocoa, Thorold ( 19T5) put i t  most succinctly  when he stated that 
’’ the ob jective of increased productiv ity  is  not sa t is fa c to r ily
University of Ghana                              http://ugspace.ug.edu.gh
achieved unless the produce is' marketable."
Mouldiness in stored food products is  known to pose serious 
public health problems. Some o f the commonly occurring fungi are 
known to produce mycotoxins (l-layne* Bennett, Tallant* 1971). One 
notable example is  the death of 100,000 turk<37poults in  England in 
1960 which was traced back to  groundnuts in fected  with Asper.srf.llus 
flavus and fed to the poults (Spensley, 1963). Although Wood (1979) 
has assured the chocolate consuming public that none o f  the fungi 
found on cocoa beans has been found to produce mycotoxins in the 
beans, the presence o f the fungi has other unfavourable consequences.
Of great concern to chocolate manufacturers is  the flavour 
o f cocoa beans a fter roasting. The flavour is  markedly a ffected  by 
mouldiness. In the event o f heavy fungal contamination, the beans 
become "o f f"  flavour which can only be partia lly  corrected  through 
expensive deodorising processes. In common with stored seeds o f 
other plant epeciec (Milner and Geddes, 1946; A ltschul, 1948) 
in fection  o f cocoa beans by fungi causes an increase in  the free  
fa tty  acid content o f cocoa beans (Maclean, 1953). This is  undesi­
rable since i t  reduces the amount o f valuable fa t  o f the beans.
In order to minimis0 the incidence o f moulds on the beans in  
storage, i t  is  recommended that cgcoa beans be dried to 8 per cent 
moisture content. The cocoa bean is  hygroscopic and absorbs or 
looaos moisture to get in to equilibrium with ambient re la tive  
humidity. In ssorption isotherms, Soott (1931) showed the rela ­
tionship oetvreen re la tive  humidity and moisture content o f cocoa
University of Ghana                              http://ugspace.ug.edu.gh
beans. At Z6°G and. 84$ R,H,, the equilibrium moisture leve l in  
cocoa is  8 ,4  per cent and at 82$  R .H ,, the moisture le v e l i s  8.0 
per cent.Onfee basi? these derivations, Dade (1929) and la ter  Knapp 
(1937) recommended that at 26°C, the re la tive  humidity conditions 
fo r  storage o f prepared cocoa beans should be at most 82$ R.H. 
Aspergillus glaucus. one o f the fungus species most frequently 
found in mouldy beans and most tolerant o f  low moisture leve ls  
cannot survive below these conditions o f temperature and re la tive  
humidity. The recommended 8 per cent moisture content is  regarded 
as an essential lower lim it , fo r ,  below th is  threshold, the bears 
became b r it t le  '\nci are ea s ily  broken and invaded by in sect pests •>
Bunting (1929) studied the mycoflora o f cocoa beans in  Ghana. 
The most important species he found belonged to  the genera 
Aspergillus. Rhizopus and Hue o r . Some o f  the storage fungi invade 
the heap of fermenting cocoa beans, encouraged by the sugary muci^ 
lage of the beans, Many o f the in it ia l  invaders w ill  be eventually 
k illed  by the high temperature (about 50°Cj o f  the fermentation 
process (Rohan, 1963; Oyeniran and A den iji, 1974), but the thermo­
ph ilic  species , such as Aspergillus fumigatus and Mucor buntingii  
survive. The development o f these thermophilic species is  encouraged 
by prolonged fermentation beyond the six-day period which is 
normally used (Rohan, op. c i t ) .  Poorly fermented beans often  begin 
to  germinate and the crack caused in the testa  by the emerging 
rad icle  creates an entrance for fungi growing on the surface o f  the 
\ 1 suggested that the four-day method o f
3
University of Ghana                              http://ugspace.ug.edu.gh
4fermentation could be used, with adequate mixing o f the beans to 
produce cocoa beans with acceptable flavour when roasted. The
!: I '
observation by Dade (1929) that unformented beans are mostly free  
from internal moulds would suppbrt the proposition  that internal 
mould development mainly starts during fermentations In fection  
of bot.ns in the pod i s ,  however, p ossib le , Ekundayo (1970), for  
example, observed that beans in  the pods were in fected  by Botryodi- 
plodia theobromae^  Many fnngi invade the cocoa beans during 
drying a fter fermentation* The usual humid conditions (85 -  100^ 
R.Ho) o f the cocoa farm where drying is  carried out could prolong 
the drying period to 1 2 - 1 4  days and encourage fungal contamina­
tion , There is  a lower occurrence in  the incidence o f internal 
mouldiness o f the main-crop cocoa harvested during the drier 
(Harmattan) months o f January to March when d iying is  more rapid 
than the ligh t-crop  cocoa harvested during the humid months o f 
Say/ Juno/July when drying takes r, longer tr’ me (Dade, 1927).
Scott (1931) observed that a drying period o f 10 -  12 days caused 
by rains and cloudy conditions.increased fungal in fe ct ion  o- the 
cocoa beans by In Malaysia, quick a r t i f i c ia l  drying has been
suggested as tbs best approach to  reduce contamination (Anselmij 
Liu and Mu* 1974)* I f  a r t i f i c ia l  drying is  to be recommended so as 
to cut down the time taken to dry fermented beans, the cpcoa w ill  
require a "maturation" period o f Z days between the fejpp.fwjtn.tici
University of Ghana                              http://ugspace.ug.edu.gh
5and. drying processes to fu l ly  develop the desired chocolate flavour
precursors (Liau, 1976),
Analyses o f the cocoa bean has shown that i t  has c. varied 
tannin/polyphenol content (Forsyth, 1952b, 1954j G r iffith , 1958; 
Swain, 196C)S It  has been shown that quantitative and qualitative 
changes in  the chemical constituents o f the bean take place during 
fermentation. These changes lead to a varied array o f  hydrolysis 
and polymerisation products o f the phenolic component (Huang,
1955* Roelofsen, 1958; Rohan, 1963) .  ' Some ahenical
constituents o f plants such as a lka loids , v o la t ile  acids, indole 
acetic  acid and polyphenols have been suggested as being preformed 
toxicants against in fection  agents. Polyphenols, however, have 
been d irectly  implicated in plant disease resistance mechanisms 
(Pridhan, 1960; Kuc, 1964; Kosuge, 1969). Since unfernented cocoa 
beans are inmune :~to internal moulds, i t  is  o f  in terest to  find  
out i f  the tannins and pclyphenols in  cocoa play any ro le  in  disease 
prevention„
This thesis contains work carried  out on the chemical changes 
that occur in  the phenolic compoaent o f testa  during cocoa curing. 
The e ffe c ts  o f individual phenolic coiRpouncfe ■ on the growth 
characteristics o f conidia of Aspergillus species iso la ted  from
University of Ghana                              http://ugspace.ug.edu.gh
6mouldy beans was investigated* The resu lts  of th is  study are 
necessary because they wo\ild show sp e c if ic  preformed chemical 
inh ib itors, which occur in cocoa bean testa , and the e f fe c t  of 
curing on these compounds.
University of Ghana                              http://ugspace.ug.edu.gh
I I .  LITERATURE REVIEW
The cocoa bean consists o f two cotyledons forming about 87 
por cent o f the seed^ These cotyledons are enclosed in  the testa  
or seed coat which forms twelve per cent o f the dry weight o f  the 
seed while the gem  or embryo forms one per cent o f  the seed (Knapp 
and Coward, 1934). The seed. soat or testa , sometimes ca lled  the 
skin, shell or husk is  made up o f two .M'vte'guments fused together* 
The outer one contains some vascular bundles and. longitudinal rows 
o f b ig  c e l ls  with very thick nucilagenous c e l l  w a lls ,
in  anatomical examination of the cotyledon by a process of 
d iffe ren tia l centrifugation  (Brown, 1954), showed that i t  is  made 
up of two types o f c e l ls ,  Large polyphenol storage c e l ls  consuiiiti(^q 
10 13$ o f the tissue,usually  in  groups, and scattered among
parenchynaious c e l ls  which contain enzymes, proteins and l ip id  
droplets. They contain the purine bases and have high amounts of 
anthocyanins, , •
The oxidation o f these phenolic compounds by an enzyme complex 
leads tq the formation o f products which inpart flavour to  panufac-- 
tured chocolate and chocolate products (Wood, 1975). This oxidation 
forms part o f the sp e c ific  processes o f curing ( ie  fermentation and 
drying) and roasting which produce, the characteristic  chocolate 
mom. (Swai$, I960)» Curing o f the cocoa bean is  in it ia ted  by
7
University of Ghana                              http://ugspace.ug.edu.gh
8n icrob ic l fercontj.tion  by yeasts followed by la c t ic  acid  and acetic 
acid bacteria (Roelofsen, 1958) ,  Duxing the process o f fermentation, 
the bean is  k il le d  l.y the conbinod action  o f a ce tic  acid and ethanol 
and the high temperature o f 51°C within the fermentation nass. 
(Roelofsen and Geisberger, 1947; Oyeniran and A den iji, 1974). 
Consequently, the c e l ls  .lose th e ir  sen i-perneab ility , and phenolic 
compounds diffus-i into the parenchymatous c e l ls  where enzymatic 
oxidation in it ia ted  by polyphenoyoxidase begins. This oxidation 
continues at a slow rato due to the largely  anaerobic conditions
created by the breakdown o f pulp tissue and the exclusion o f  oxygen
fron the nass. As the sweatings drain out o f  the fom entation  heap, 
and also during drying, prevailing aerobic conditions accelerate the 
oxidation process (Forsyth and Quesnel, 1957).
At the beginning o f fom entation , the nass o f seeds i s  conposed 
o f pulp and beans. Pulp is  nade up o f 80 -  90% water and 8 — 13$
sugars nade up o f glucose (8 - 10 S^) sucrose (0„4 -  1$  as well as
ra ffinose  and stachyoso (Thaler., 1954. Forsyth and Quesnel, 1963). 
Sucrose levels reduce during fernentation due to hydrolysis and 
d iss in ila tion  as well as migration into the testa  (Howat, 1957).
Howat (o p .c it )  reported that cocoa beans contain 0 .2  to 0,4
(w/v) o f n on -vola tile  acids lik e  c i t r i c  acid , oxa lic a c id ,/a ce t ic  i
acid ’ and that these contribute to the a c id ity  o f  the pulp
(pH 3o6 -  3-7) tissue,, In contrast to the pulp, the bean is
University of Ghana                              http://ugspace.ug.edu.gh
9ocmp^ sea of wages' (33^ )# «Oo«?l »*- £■** (ijSQt <#*$■
sugars, purine hases (theobromine and ca ffe in e ) , enaymes, amino 
aoids and phenolio compounds (3 j$ ) (Forsyth and Rombouts, 1951? 
Roelofsen, 1958; Rohan, 19^3)• The phenolio compounds f a l l  under 
three main groups} oateohins (3T fo ) t  anthooyanins (4$) and the 
leuooanthocyanins (58$)• The main oateohins are epioateohins, 
gallooateohins, and spigallocateohina* The anthooyanins are jH^-L 
arabinosidyl oyanidina and 3- 6fc*E galacrfcoeidyl oyanidins (Forsyth,
1952b; G r if f ith , 1958? Roelofsen, 1958? Rohm, 1963)* Leuooantho- 
oyaains in  oocoa oocur as the agluoon leuocanthocyanin and bo do not 
oontain sugar residues (Robinson & Robinson, 1933)• S p ec ific  
oanponents o f  the bean undergo d iffe ren t ohanical changes during 
fermentation*
The water content o f  the bean ixior^ases during fermentation 
from 33^ moisture to  40$ (Howat, 1957)* This suggests among other 
things a movement o f  water from pulp in to  the bean but Howat (op *c it) 
showed that th is  was un like ly . Because the sugar content o f  the pulp 
i s  greater than that o f  the bean, a higher osmotio pressure i s  or^at«<! 
in  the pulp and water should move from the cotyledons to  the pulp thus
dehydrating the former* Howat (1957)» th erefore , suggested that the 
chsnical breakdown o f  some constituents o f  the ootyledons with 
the produotion o f  carbon dioxide and water could be the cause o f
University of Ghana                              http://ugspace.ug.edu.gh
10
the increase in  water content of the beans.
During fermentation^ the fa t  droplets do not d iffu se  out o f 
the cotyledons. Consequently, there is  no change in  the quantity 
or chemical composition o f the cocoa but tor^ Also at th is stage 
hydrolysis o f sucrose occur3 resu lting  in  a decrease in  amounts o f  
non-reducing sugars. The products o f hydrolysis d iffu se  from th 3 
cotyledons in to the sh e ll. Whilst there i s  no change in  the 
content o f starch , soluble amino acids are lost through d iffu s ion  
in to  the testa . Amino acids that have been id en tified  as occurring 
in cocoa beans are aspartic a c id , glutamic acid , alanine, cysteine, 
leucine, serine, threonine, va lin e , proline and o<l-anino butyric 
acid (Maly* 1955; Forsyth and Quesnel, 1963).
The vitamin D concentration in  the sh e ll increases with 
fermentation and drying (Knapp and Coward 1934, 1935)* These 
workers found the sh e ll o^. sun-dried fermented cocoa from Ghana 
to  contain 2S International Units ( l .U .)  of vitamin D per gram dry 
weight o f testa  tissue*
Plants react to  invading parasites, throu^n resistance 
mechanisms 'viJood, 1967). These mechanisms nay consist o f  struc­
tural or physical barriers to the invading process so tiiat 
pathogens are excluded or k il le d  lo ca lly  to arrest further advance. 
Preformed inh ib itors are widely recognised to be involved in  
disease resistance in some plant species , 0-quinone, a tox ic
University of Ghana                              http://ugspace.ug.edu.gh
11
fermentation product, of phenol oxidation in te a , makes the ro lled  
tea uninhabitable to  microorganisms (Lamb and Sreerangachar, 1940). 
Since the chemical constituents of cocoa beans are altered quanti­
ta tive ly  and qua lita tive ly  during curing (Forsyth and Ronibouts,
1951j Forsyth, 1952b), the natural protective action o f cocoa bean 
constituents could be adversely a ffe cted . Laycock's (1931) obser­
vation that unfermented beans were immune to  mould damage appears 
to  r e fle c t  th is  view. While the a b il ity  o f the n on -vo la tile  acids 
as disease inh ib itors has not been investigated , i t  is  established 
that the purine bases, particu larly  theobromine has no inh ib iting  
e ffe c t  on fungal growth. It  i s ,  in fa c t ,  used as a prefered source 
o f nitrogen by micro-organisms. There is  an observed s ign ifican t 
f a l l  in the le v e l of the theobromine in the cotyledons with mould 
in fection  (Maclean, 1953; Roelofsen, 1958). In an eight-day fe r ­
mentation t r i a l ,  Humphries (1944) noticed a decrease in theobromine 
content of 20 ■» 24%. S im ilarly , Knapp and Wadsworth (1924) fe r ­
mented beans fo r  10 days and showed that there was a movement of 
theobromine from the cotyledons in to the te sta . The longer the fe r ­
mentation period, the greater the quantity o f Theobromine that moved 
out of the t-.jsta. I t  was Roelofsen (1958), who described the d iffu sion  
o f theobronnjie into the testa  and i t s  accumulation as crysta ls on 
the in terior and exterior of the te sta . These accumalations could 
be sign ifican t in increasing mouldiness by providing a ready source 
o f nitrogen for  invading fungi (Wadsworth, 1922).
University of Ghana                              http://ugspace.ug.edu.gh
12
In contrast I phenolic compounds have been implicated in  host 
defence nechanisns (Toniyana, Sakai, Sakuna, Ishizaka. 1964; Kosuge, 
1969)« For ,1 proper understanding o f the ro le  o f phenolic conpounds 
in cocoa beans i t  is  necessaiy to obtain information on the changes 
in amounts o f other chemical constituents such as protein , carbohy­
drates and organic acids so as to  be able to ascertain  whether 
decreasing amounts o f these chemicals cause' any noticeable e ffe c ts  
(Kirkham, 1957), The potentia l o f phenolic conpounds and their 
oxidation products in host defence mechanisms have been amply 
demonstrated by Angell, lralker, Link (1930), Flood and Kirkham 
(1959) and Kiraly and Earkas ( 1961) ,  Scevers and Daly (1970) 
however, fa ile d  to correlate  resistance with phenolic conpounds in  
the wheat stem rust and concluded thus; " I t  is  becoming increasingly 
c lear, however, that there is  no single case where resistance and 
biosynthesis o f aromatic conpounds can be unambiguously correlated , 
le t  alone proven"- They pointed out that resistance nay be expressed 
through more subtle metabolic processes than by growth inh ib ition  
due to synthesised tox ican ts .
The phenolic compounds have a basic aromatic structure as 
the building un its . The products o f  th e ir  enzymatic oxidation 
have been well investigated because of their  ro le  in chocolate 
flavour development (Forsyth, 1952b; Forsyth and Quesnel, 1957), 
and in other foods (Bate-Snith, 1954),
University of Ghana                              http://ugspace.ug.edu.gh
13
The bulk o f the tannins and polyphenolic substances in  the 
cotyledon are batechins .and leuooanthocyanins which form 5lf- and 
58rfo o f the to ta l  tannin content, respective ly i Of the soluble 
cat echini! S2/i is  (-jep ica tech in  (Forsyth* i 954) * Apart fron  the 
catechins and leuooanthocyanins, the flavanoids occur in  plants 
as glycosides in which aone o f the phenolic hydroxyl groups are 
conbined with sugar, residues. The sugar-free molecule is  the 
aglycon, Hallas and T-right (1939) had reported that the percentage 
soluble tannins - fe ll  fron  19/i to  8^ within 6 days o f fermentation. 
Forsyth and Rombouts (1951) la ter  found that the decrease was 
through exudation in to  the testa  since the oxidases present were 
active only at the drying state when conditions were more aerob ic. 
This idea had ea r lie r  been proposed by Humphries (1944) and la ter  
supported by Forsyth (1952). Rohan (195T) confirmed the presence 
o f polyphenols in  West African amelonado and determined their 
quantitative changes during fermentation. He found no changes in 
any of the phenolic compounds prior to death o f the beans and the 
subsequent movement of the phenolic compounds from the storage c e l l s .  
Following the movement of the compounds from the storage c e l ls ,  
however, that is  about 12 hours a fter  the beginning o f  fermentation, 
there is  a rapid destruction o f anthpeyanins and si loss o f the other 
polyphenolic compounds but at a puch slower ra te , The oxidation o f 
the phenolic components of a mixture takes place in a sequence
University of Ghana                              http://ugspace.ug.edu.gh
14
(Roberto and Wood, 1950), As with tea fermentation, the sequence 
depends on the oxidation-reduction potential o f the confounds.
Thus in mixture o f compounds, those with a lower oxidation-reduction 
potential are se le c t iv e ly  ox id ized . This observation was supported 
in the case of tea fermentation, by Kefford, (1959), and “fi ll ia n , 
(1957). The order o f  oxidation, in  a mixture, was suggested by 
these workers as catechol, pyroga llo l, ga llocatech in , then catechins, 
ca ffe ic  acid and chlorogenic a c id . Quinones, formed from chlorogenic 
acid , appear to have a high oxidation-reduction potentia l and 
therefore tend not to be oxid ized .
Again, as in  the case of tea fermentation, Swain (1953) 
suggested that the various polyphenols are decomposed to the flavan - 
3: 4 -d io l form and then polymerised into insoluble fractions,* 
Bate-Smith (1954) tr ied  to correlate  the flavanoid tannins with 
the basic flavan -3 :4 ;d io l as an intermediate of the hydrolyses of 
phenolic compounds, Roberts and ¥ood (1950) recognised the* 
existence o f th is intermediate stage as the basic structure o f a l l  
the flavanoids but they did not characterise i t .  Forsyth and 
Rombouts ( l9 5 l)  did thorough investigations and showed that the 
decomposition products o f the major phenolic components are cate­
chins, anthooyanins and leucoanthocyanins. They observed that t ie  
amount of to ta l phenolic compounds remain unchanged, A degree oi' 
condensation occurred. The simpler polyphenols are destroyed anc 
insoluble tannins forhed. These workers also found that the
University of Ghana                              http://ugspace.ug.edu.gh
flavan-3 :4“ d io i was an inevitable intermediate in the decomposition 
and condensation o f phenolic products.
fla van -3 :4 -d io l
Some leucoanthocyanins are known to have flavan -3s4 -d io l units so 
that the structural s im ila rity  to the catechins became obvious.
Anthocyanins are hydrolysed by glycosidoses present in  the 
cotyledons o f the beans (Forsyth and Quesnel, 1957). Some Aspergi­
llu s  spp, produce the enzyme anthocyanase which hydrolyses anthocy­
anins to  anthocyanidins (Nordstrom, 1956 j Huang, 1955). The major 
part of the anth.5eyaa.in hydrolysis occurs during drying as free 
access to a ir  leads to oxidation and browning (Forsyth and Quesnel, 
1957). The r-.si is  lo s t  in storage with time (Fenten, 1965). It 
has been suggested that the action  o f glycosidases during the 
aerobic phase liberates a precursor o f chocolate flavour (Forsyth 
and Quesnel, 1963),
The more complex leucoanthocyanidins are hydrolysed into 
(-)ep icatech ins and cyanidins'(Bate-Smith, 1954; Forsyth and 
Roberts, 1960).
In general, Roberts (195Y) ,  Baruah and Swain (1959) have shown 
in  the case of potato polyphenols, that flavonol g lycosides, unlike 
their corresponding aglycons are not readily oxidized in the
University of Ghana                              http://ugspace.ug.edu.gh
16
presence o f potato phenolase except in  coupled oxidations with 
substrates such as chlorogenic a c id . It  would appear from the 
claim of Roberts (1957), Baruah and Swain (1959) that anthocyanin 
glycosidases would be resistant to d irect attack by phenolases. 
Indeed, Rohan (1957) detected residual anthooyanins in  cocoa during 
the period o f  drying even though the phenolases were s t i l l  presumably 
a ctive . Some plants convert these phenolic compounds into compounds 
more tox ic  to  the invading fungi (Schaal and Johnson, 1955). The 
a b ility  of phenolic compounds to inh ib it fungal a c t iv ity  can be 
reduced by the presence o f proteins and certain amino acids, such 
as those containing sulphur, (eg . cysteine and glutathione which 
act as chelating agents (Roberts, 1957). Sporulation in  Venturia 
inaequalis.the causal agent of the apple scab diseases is  prevented 
by chlorogenic acid and iso-ch lorogen ic acid produced by the host 
plant (Flood and Kirkham, 1959). The onion rot o f the yellow pigmen­
ted onion caused by Botrytis sp, and Collectrotrichum cire irans is  
prevented by the production o f catechol and protocatechuic acid 
(Link, Angell, and Walker, 1929a; Angell, Walker, Link, 1930; Link 
and Walker, 1933). Other workers investigating resistance mechanism
of the Cox's pippin apple, observed that anthooyanins, leucoantho­
cyanins and flagans prevented i t s  in fect ion  by Geosporium nerennans 
(Byrde, Fielding and Williams, 1959; Hulme and Edney, 1959). The 
protective influence was found to be e ffe c t iv e  while the fru it  was 
nn +.kc. ■Ht’or. a-i- +v,n+ ^ e re  £s a high phenolic leve l o f 250 mg.
University of Ghana                              http://ugspace.ug.edu.gh
17
per 100g, dry weight of peal; Once the f r u it  is  harvested* the 
concentration o f  phenolib compounds f a l ls  rapidly  and at 20mgj 
per 100g, dry weight o f peel, the fru its  become more susceptible to  
the disease. F ield  investigations by G r iffith  (1958) on the e f fe c t  
o f  phenolic compounds in Theobroma cacao showed evidence o f an 
intense metabolic a c t iv ity  involving the host polyphenols a fte r  
invasion by Marasmius permiciosus. A lso , while the fungus 
Fhvtophthora palmjvora is  able to in fe c t  and invade the en tire  pod 
within 14 days (Asare Nyako, 1972), i t s  c lose  co-member o f the 
family Sterculiaceae and o f the same environment, Cola n it id a . is  
not a ffe cted . This would suggest in tr in s ic  resistance mechanism 
in £ , n it id a .
At the beginning of the fermentation period, Roelofsen (1958) 
id en tified  eight species of yeasts in the fermentation heap. These 
were Saccharomvces cerevisiae  var e llip so id e s . other Sac charomvo e s 
species, Candida krusei. Kloeclcera ap icu lata, P ichia fermentans. 
Hansenula anomala and Schizepacchromvces pombe. A fter the f i r s t  
36 hours, those yeasts requiring l i t t l e  oxygen, such as the 
Saccharomvces species, dominate. As the macerated pulp drains out, 
the conditipns become more aerobic. Alcohol and acetic  acid  then 
become the main sources o f carbon. At th is  stage, the aeroph ilic 
and thermotolerant yeast Candida krusei becomes the most important.
University of Ghana                              http://ugspace.ug.edu.gh
18
The in it ia l  concentration of c i t r i c  acid in the pulp is  high; 
thus producing a condition g f pH 3*6 (Howat, 1957) which i s  favour­
able to Betabacteridm species as the most s ign ifican t la c t ic  acid 
bacteria* In addition to  Betabacterium xvlinum. Betabacterium 
xvlinoides and Betabacterium ascendens. Acetobacter rancena end 
A, melanogenum have been id en tified  as the main a cetic  acid 
bacteria  in fermentation heaps (Krtepp and Coward, 1935;
Roelofsen, 1958)* A* melanogenum is  known to  be capable o f 
removing pulp together with the outer epidermis of the seed coat, 
possibly by metabolic degradation of mucilage c e l l s .  Candida 
krusei has been observed in  the cotyledons o f cocoa from the Bahia 
region o f Brazil and described there in  visual examinations as 
internal moulds (Maravalhas, 1966)«
Bunting's ( 1929a) investigation  o f internal mouldiness in 
commercially dry cocoa in  Ghana revealed 5 fungal species as causal 
agents though their re la tive  frequencies o f occurrence wore not 
stated . Members of the Aspergillus glaucus group, A, fumigatus 
and A, ustus and Botrvodiplodla theobromae were id en tifie d . Since 
then not much work has been done in  th is  area o f  cocoa quality 
control in Ghana,
The most recent investigations o f internal moulds o f Nigerian 
cocoa beans put the level at 1,6 per cent (Oyeniran and Adeniji, 
1974), These workers iden tified  31 fungus species , belonging to
University of Ghana                              http://ugspace.ug.edu.gh
i iw ia  genera, as the causal agents* These included twelve Aspergi­
l lu s  species and four Penicillium  speciesi They grouped the fungi 
into major and minor pathogens depending on frequency o f occurrence. 
The maj or pathogens were Aspergillus ch eva lie r i. Aspergillus flavus 
Aspergillus fumigatus. Aspergillus n iger. Aspergillus p em ic i l lo id e s . 
Aspergillus ruber. Aspergillus tamarii. Botryodiplodia theobromae. 
Mucor pu s illu s . Paecilomyces v a r io t i . P en ic illiu n  citrinum. Hhizopus 
arrhizus and S.vncephalastrum racemosum. The minor pathogens were 
l is te d  as Absidia corvmbifera. Aspergillus aculeatus. Aspergillus 
nidulans. Aspergillus ochraceus. Aspergillus pseudoglauous. 
Aspergillus re str ic tu s . Cbrvularla lunata. Cylindrocarpon tonkinense. 
Fusarium oxysporium. Fusnrium so lan i. Geotrichum candidun. Hacrophoma 
SP.» Penicillium  deoumbens. Penicillium  s t e c k i i . and Penicillium  
variab le . The occurrence o f J., flavu s . which produces the mycoto- 
xin , a fla tox in , among the l i s t  o f fungi causing internal moulds o f 
Nigerian cocoa, is  o f particu lar in terest since Bunting ( 1929a) and 
Maravalhas ( 1966) did not fin d  th is species in  their observations 
in Ghana and Brazil, respective ly .
Reese and Downing (1951) and O lutiola (1976) have shown that 
Aspergillus terveus. Aspergillus fla v ip e s . Aspergillus fumigatus, 
Aspergillus f is h e r i . Aspergillus ustua. Aspergillus flavus and
.giganteus have c o l lu lo ly t ic  enzymes. With the notable 
exception 01 Asperpil^us .ni^er var luchuensis. the above observation
University of Ghana                              http://ugspace.ug.edu.gh
agrees with the suggestion by Reese and. Downing (1951) that no 
black Aspergillus attacks c e llu lo se . Therefore, in  the in fection  
process, the fungi that can penetrate the cocoa bean could have 
eco log ica l advantages and would pave the way for the entry o f 
those that do not have' o o llu lc ljr t ic  onzymcr. and therefore cannot 
degrade ce llu lose  walls (Reese and Levinson, 1925; Heese and Downing, 
1951).
The method s by which fungi gain entry in to  the cotyledons 
of the bean are not fu lly  known. Laycock (1931) suggested that 
fungi enter the bean through the micropyle end; the in tact bean is  
invariably invaded at a point in  the testa  opposite the . rad ic le  
although there is  no apparent structural d ifferen ce  between that 
part and other areas o f the te sta , Laycock (o p .c it )  showed that 
!&,.• o f invasions of the cotyledons was through the rad ic le  and 
24fr through other means. This report of the fungal penetration of 
the testa  has been supported by Reese and Levinson (1925), de Witt 
( 1952a) and Olutiola (1977). Renaud (1954) demonstrated that the 
point in the testa opposite the rad icle  is  weakened during incip ien t 
germination, a condition encouraged by fau lty  fermentation procedure. 
Knapp (1937), on the other hand, noted that the testa  is  macerated 
by pectinases produced by yeasts during fermentation. "Jerking in 
Bahia, Maravalhas (1966), however, observed that the raphe, which 
is  also a scar tissue, rathor than the micropyle, is  the entrance
University of Ghana                              http://ugspace.ug.edu.gh
21'
point fo r  invading fungi. It  is  the most porous part o f the bean 
and is  also the point through which pigments ejtude from the coty le ­
dons and the entrance fo r  a cetic  a cid .
Phenolic compounds have been described as the an tisep tics  o f  
plants by Pridham (i960 ), In the ligh t o f th is  claim* the a b il ity  
of fungi to invade the cotyledons o f  cocoa beans, which contain 
phenolic compounds, generates a lo t  o f in te rest. Various in ves ti­
gations have been undertaken to  prove the potential o f  phenolic 
compounds as protectants against in fe ct ion  agents (Parkas and 
K iraly, 1962; K iraly and Parkas, 1962; Kuc, 1964).
The cocoa bean has a complex o f enzymes responsible fo r  
browning, oxidation o f phenolic compounds, and proteolysis  (Huang, 
1955; Porsyth, Quesnel and Roberts, 1958; Roelofsen, 1958; Holden, 
1959). The other ensjsmes which have been found are B -fructosidase, 
B -glycosidase, B~galaclosidase, X-amylase, proteinase, asparaginase, 
•phenoloxidase, peroxidase and ca ta lase . However, an ea rlie r  report 
by Bavendamm (1928) claimed that no fungi occurring on cocoa produce 
phenoloxidases. Swain, (1953) observed, in the case o f tea fermen­
tation , that some products o f enzymatic oxidation are able to preci­
p itate protein and therefore enzymes. The more polymerised the 
products, the more they precip itate  the enzymes. The process of 
phenol oxidation and subsequent polymerization o f flavonols leads 
to a situation  in which various enzyme systems can be precipitated
University of Ghana                              http://ugspace.ug.edu.gh
out o f action in  the? fprp of chelates and complexes (Chandler and 
Swain| 1959)t Thus onsymatio reactions leading to flavour develop 
ment could be hampered i f  the enzymes involved were precip itously  
inactivated (Swain. 1960; Jackson and Wood, 1959)o
Besides the broad observation by Biodin and Prevost (1957) 
that some fungi need and indeed use tannins o f cocoa beans in 
their growth processes, no work involving the action  o f s p e c if ic  
phenolic compounds on the predominant moulding agents, the 
Aspergillus species, have been reported?
University of Ghana                              http://ugspace.ug.edu.gh
23
I I I ,  MA'TERTALS AMD GENERAL METHODS
i )  MATERIALS
Cocoa beans used in  assessment studies were obtained from 
parcels of com erc ia l cocoa beans in  the warehouses o f the Ghana 
Cocoa Marketing Board at the port o f Tema  ^ Random samples of Q&tyKtsf 
2 kg; weight were taken o f the main season 's crop and the l ig h t -  
season's cr<jp of beans arriving at the warehouse from transit 
sheds, Ten samples were taken in  a l l5 The samples wore stored 
at 4°C un til required. Cocoa beans o f the Amelonado type usod in  
ehemical analyses and fo r  anatomical studies were taken from experi­
mental fermenting heaps o f the Cocoa Research Institu te at Tafo tn 
March. 1979.
With the exception o f  Sclerotium r o l f s i i . iso la tes  o f a l l  
fungal species used in these studies, were prepared from in fected  
beans co lle cted  from the Ghana Cocoa Marketing Board warehouses.
The isolate ’ of S# r o l f s i i  was obtained from the stock culture 
co lle c t ion  of the Botany Department of the University o f Ghana,
Legonc This iso la te  was or ig in a lly  prepared from naturally in fected  
tomato fru it  (Lycopersicum esculentum) from Mekola market in Accra,
i i ) GENERAL METHODS
a) Maintenance of Stock Culture
Stock cultures of Asperg illus alaucu s . A, n iger. A, fumigatus 
ik* A* nidulans nrjd, 3 , r o l f s i i  were maintained on Sabour.’ud1 s
agar medium (40 g maltose; 10g peptone; 15g agar; 1000ml d is t i l le d
University of Ghana                              http://ugspace.ug.edu.gh
24
water) in McCartney tubes, They were maintained at 25°C and sub­
cultured at two-weekly intervalso
b) Method of Inoculation
For both liqu id  and plate cu ltures, a 3mm agar d isc obtained 
from the growing edge o f the culture was used fo r  the inocu lations,
c ) Methods o f ■Sterilization.
McCartney tubes, p ipettes, nutrient media and d is t i l le d  water 
were s te r ilis ed  by nutoclaving fo r  15 minutes at 1 .1  kg/cm steam 
pressure and a temperature of 120°C. Cotton wool plugs, f i l t e r  
paper and pipettes were temporarily covered with grease paper to 
prevent the penetration of any condensed water during autoclaving.
Petri dishes (5cm and 9cm. diameter) were s te r i l is e d  by heat­
ing at 165^0 fo r  6 hours in an e le c t r ic a l ly  heated oven.
Slides and coverslips were thoroughly washed with detergent, 
rinsed under running tap and in  several changes o f d is t i l le d  water 
and stored in  90 per cent ethanol and flam e-ster ilised  just before 
use. Inoculating needles and loops were flamed to red-heat and 
cooled immediately before use.
Cocoa beans were thoroughly washed under running tap then 
surface s te r iliz ed  by dipping them in  0,1 mercuric chloride fo r  one 
minute and rinsing thoroughly with several changes o f  s te r ile  d is t i l le d
T.ro 4r/?w» -
University of Ghana                              http://ugspace.ug.edu.gh
Cocoa testa  extracts were s te r i l is e d  by f i lt r a t io n  through a 
n ill ip o re  f iI te ra t ion  unit (F ilte r  Holder Part 4* XX1004703) and a 
f i l t e r  of pore size  0.22 trn under suction from a vaceum pump (Type
B.S. 2408 B56),
The inoculating room was always sprayed with d etto l solu tion  
(Dettol 5ml; water 95 ml) 10 minutes Before u se ,
d) Assessment o f percentage internal mouldiness in  Commercial 
beans
One hundred cocoa beans from samples obtained by quartering 
were examined ty p  internal mouldiness. Beans were surface s t e r i l i ­
sed and then drained on s te r ile  f i l t e r  paper (Whatman's Ho, l ) .
The washed beans were asep tica lly  halved to expose as much o f the 
cotyledons as possib le , The cotyledons were incubated at 
25 + 1°C on moist s te r i le  f i l t e r  paper (¥natman!s No, 2) fo r  7 
days. Assessment of growth was done on the 7th day. Three r e p li­
cates o f each tfecutmertt were examined 0
The moisture content of the sample cocoa beans was determined 
by the oven drying method,
0) Assessment  of mvcoflora o f cocoa beans
Infected beans from, an ea rlie r  investigation  were transferred 
to Sabouraud's, Czapek Dox, and Patato Dextrose agar media and the 
mtm  out o f the cotyledons were iso la ted  into
University of Ghana                              http://ugspace.ug.edu.gh
pure cultures* A ll fungal species were Subcultured on Czapok Doz 
agar medium according to the method o f  Than and Raper ( 1960) fo r
complete iden tifica tion s Fungi wore id en tified  under the low~
fo r  10 days before id en tifica tion  was done,
f ) Culture o f te s t  fungi on nutrient media
Sabouraud's agar (15 ml) in a 9 em diameter P etri dish  was 
inoculated with a 3mm culture disc removed with a No* 3 cork borer 
from the growing edge o f a 3 days old culture ra ised  on Sabouraud*s
agar. Tho inoculum was placed at the in tersection  o f two diameters
There were three rep lica tes fo r  each treatment.
For the liqu id  medium, part of the mycelium from a 3 day old 
1. 2 %  water agar was inoculated in to 30 ml broth in  Erlenmeyer
Growth rate on agar medium was assessed by measuring the 
diameter of the cultures along two diameters at right angles to 
each other at 24 hourr in terva ls .
To assess the amount o f mycelium produced by the test  fungi,
power o f a microscope and^oil immersion, A ll p lates were incubated
fla sk s ,
g) Assessment o f  growth o f culturss on media
i )  Solid medium
<3 mm d i  a iDo+.fiT n l the growing edge o f 1 , 29^ water agar
University of Ghana                              http://ugspace.ug.edu.gh
27
culture was used per fla sk i Mycelium was harvested onto a 
previously weighed Whatman’ s N042 f i l t e r  paper, and dried to 
constant weight at 60°Ci The f i l t e r  paper containing the dry 
mycelium was then weighed a fter  i t  had been allowed to  co o l in a 
desLG.cator0
h) Assessment o f groxvth on cocoa bean testa  extracts
i )  Preparation o f cocoa bean testa  extracts
For assessing the e f fe c t  o f bean testa  during curing on the 
production of mycelia, 25g of bean testa  separated from the co ty le ­
dons were ground in a blender (Sunbean, Model XBL) fo r  4 minutes, 
strained with cheese c loth , made up to 500 ml and s te r il is e d  using 
m illipore f i l t r a t io n  ( f i l t e r  pore s iz e , 0 .2 ft® )in  vacuo. Erlenmeyer 
flasks (250 ml capacity) oach containing 20 ml media were used fo r  
a l l  the cultures in extracts,
i i )  Spore germination methods
Spores wore obtained by ca re fu lly  removing spores from agar 
plates o f 3 dayt old cultures,. For each gem ination te s t  the 
density o f spore suspension was s t r ic t ly  standardised to 400,000 
per m ill i l i t r e  o f suspension with the aid o f haenocytometer.
Spores were suspended in the appropriate extract and incubated at 
25 C + 1°C fo r  varying periods of time.
University of Ghana                              http://ugspace.ug.edu.gh
28
i )  Assessment o f Spore germination
At the end of the desired incubation period, spore germina­
tion  was determined by recording; (a) percentage to ta l spore 
germination and (b) length o f germ tubes. Germination assessments 
were done fo r  testa  extracts and the con tro l, A nutrient medium 
o f  Potato Dextrose Agar "floodwater" was used as con tro l, A fresh  
uninoculated PDA plate was flooded with 10 ml, s t e r i le  d is t i l le d  
water and allowed to stand fo r  f iv e  minutes. The water was poured 
o f f  and used in the preparation of spore suspension.
j )  Investigat io n o f otage during curing when contamination was 
observed in cotyledons
Beans used fo r  studies on the stage in  cocoa bean curing when 
contamination was observed were removed from a 155 kg. fermenting 
heap o f Amelonado type cocoa beans at the fermentsry o f  the Cocoa 
Research Institu te at Tafo, F ifty  beans were removed randomly at 
various stages o f curing v iz , fermentation and drying (0 , 2 , 6 and 
12 days). The 6-day method of fermentation was used,, The beans 
were immediately su r fa co -ster ilised  in 0 . 1^ mercuric ch loride 
so lu tion , rinsed at least three times in s te r i le  d is t i l le d  water 
cjid out transversely into 5 equal sections, The sections were 
placed on a moist f i l t e r  paper (Whatmahs No. 2) in s te r i le  potri 
dishes in an a ir b lrm  s te r i le  chamber (Pathfinder, Havant,
England) at 25 C, Beans were incubated fo r  7 days nnd then
University of Ghana                              http://ugspace.ug.edu.gh
29
examined fo r  evidence of fungal growth.
Hyphae were cultured fo r  complete id en tifica t ion .
k) Chemical Analyses
1• Preparation o f fa t -fr e e  powders fo r  analyses
One kilogram o f the cocoa bean testa  separated from the coty ­
ledons o f the sample cocoa beans was dried in  a ventilated  oven. 
(Gallenkanp, Model 1H -  100) at 40°C fo r  48 hours, The dry testa  
was ground into powder. About ten grams o f the ground testa  wore 
defatted by shaking in 100 ml petroleum ether (b .p„ 40 to  60°C) fo i  
3 hours. The solvent was decanted and discarded. The .-pf'tec-IjSitote 
Was-extracted with fresh solvent, thoroughly drained and the powder 
stored in the dark under vacuum in  a desiccator.
2» Phenolic compounds
Three grams o f the defatted testa  powder were extracted with 
50 ml, absolute methanol fo r  2 hours at 22°C in the dark with 
occasional shaking. The supernatant was decanted and the p e lle t  
was suspended in  a fresh  volume o f methanol and the extraction  
repeated tw ice. After the th ird  extraction the supernatants were 
combined, f i l t e r e d  through Whatman's No, 1 f i l t e r  paper and evapo­
rated to dryness at 60°C in a Buchi rotary evaporator (Rotavapor-
u l Glasapparatefabils, F law il). This dried sample contained the 
froo soltiblo phenols.
University of Ghana                              http://ugspace.ug.edu.gh
The amino acids and sugars were removed, by resuspending the 
p e lle t  ill £0 mil o f methanol (50^ 5 v /v ) fo r  1 hour and f i lt e r e d
* j  , ’
through Whatman*s Hoi 1 f i l t e r  paper to remove insoluble material^ 
The f i lt r a te  was discarded! The residual powders wore extracted 
with 70 ml$ methanol ( 5 0  v /v ) under re flux  with b o ilin g  water for1 
15 minutes to  hydrolyse the complex phenols! A drop o f  toluene, 
was added to prevent fungal growths
3 it Estimation o f Total fre e  phenols
The free  phenol;. content was determined using the Folia*- 
Cioloateau phenol reagent with ca ffe ic  acid  (Kock-light Laboratory 
ltd , 4064) as the standard,, A ca lib ra tion  curve was prepared with 
s e r ia lly  d iluted  pure ca ffe ic  acid solutions as shown in  Appendix 
A| To 0 ,5  mlj o f  the absolute methanol fra ction  (d ilu ted  10 times) 
was added 10 ml, o f 2 ,8  N sodium carbonate solu tion  (29.7 g in  200 
a l»  water) and 3*0 ml, phenol reagent with shaking at in terva ls fo r  
2 minutes. The op tica l density o f  the te st  so lu tion  was measured 
at 650 k  on a spectrophotometer (Pye Unican 600 Series 2)
against con trols made up o f the sodium carbonate and the phenol 
reagent and the sodium carbonate and d is t i l le d  water*
University of Ghana                              http://ugspace.ug.edu.gh
31
TABLE 1
Reaction mixture fo r  the determination o f fre e  phenols 
Reactants Quantities Used
Experimental " Controls
1 - 2.+
Test Solution '0 ,5  ml 0.5 ml -
H„0 4.5 ml 14.5 ml 5.0 ml
2.8N Na„C0, Soln. 10.0 ml -  10.0 ml
Phenol reagent-1 + 3 .0 ml 3 .0  ml 3 .0  ml
18.0 ml 18.0 ml 1§ .0  ml
^Control 1 is  fo r  zeroing the photometer since there is  no reaction  
in  the absence of sodium carbonate,
+Control 2 is  the e ffe c t  o f  reagents in  tho absorption o f the 
photometer. It i s  substracted from tho readings o f the experimental, 
++Fo±in~Ciolcate-„u Phenol reagent: d eta ils  o f composition (Spies, 
1965).
4« Estimation of Ortho&ihydrio.phenols
Orthodihydric phenol content o f the testa  was determined with 
A,mow's reagent using chlorogenic acid  as the standard (Johnson 
and Schaal, 1952). The reaction  mixture was made up of 0,2 ml 
extract (diluted 5 tim es), 0 .2  ml, 0.5N HCl, 0.2 ml. o f Arnow's 
ros.gent and 0.4 ml. of H NaOH in a to ta l  volume of 3 .0  ml. The 
reaction  was started by adding the Arnow’ s reagent and within
University of Ghana                              http://ugspace.ug.edu.gh
32
30 seconds the op tica l density (OD) was at 520 nm against a
blank in which wator replaced the Arnow's reagent.
A ca libration  curve/prepared xtfith s e r ia lly  d iluted  pure /wa.5 
chlorogerdc acid (Koch-Light, Laboratory Ltd, 40550) from a stock 
solution  of 200 jug/ml GUndused as indicated in Appendix B.
TABLiij-i
Sp e c trophotometr ic  estimation o fO rth odihydr l c  phenols
Reactants Experimental (ml) Control (ml)
Test extract 0.2 0„2
HgO 2„0 2.2
0.5N HCI 0.,2 0.2
1N HaOTI 0,4 0o4
■-Arnow's reagent 0 ,2 -
3,0 5.0
^Reactions were started by the addition o f the Arnow's reagent 
(lOg IlaNOg, 10g Na^ I'^ gO,, in 100 ml HgC)»
5, F lavonol determination by the vani l l i n  motiiod
The v an illin  method o f flavonol determination was used with 
(-)epicatechin(Koch-Light Laboratory, Ltd 40648) as standard (Swain 
and HilliSj, 1959)* One ml* of the extract was d ilu ted  20 times and
2,0 ml, o f t.ae diluted extract used fo r  the determinations.
University of Ghana                              http://ugspace.ug.edu.gh
Vanillin  reagent (Hopkins £, Williams Ltd. Batch 77314) was used, to 
start the reactions«. A ca lib ra tion  curve was prepared with s e r ia lly  
diluted (-)ep ica tech in  solutions as shown in  Appendix C, Optical 
density readings were taken at 500 nm»
TABLE i i i
Vanillin  Method fo r  the determination of flavanols*
Reactants Reactant Quantities (ml) Blanks (ml)
k B £ D
Extract 0 ,2 0.2 - -
Water 1*8 1 „8 2.0 2.0
70£ (v /v ) H2S04 - 4.0 - 4.0
+ Vanillin  reagent 4.0 _JZ_
6.0 6.0 6f0 6 *0
For the deteTvainat:!on -3D readings produced by the reaction  of the 
flavanol component;, the OE readings o f C and B were substraeted 
from A„ D is  used to zero the spectrophotometer,
+ Reactions were started by the addition o f Vanillin  reagent.
(1 g- v an illin  in 100 mlft 70$ (v /v ) conc, H?S0^)
6 a Anthocyanin determination
The determination of anthooyanins and leucoanthocyanin content 
was made using cyanidin ch loride (Koch-light Laboratory, 40204 as 
the standard (Swain & H ill is , 1959)» The cyanidin chloride was
University of Ghana                              http://ugspace.ug.edu.gh
34
dissolved in 0 . 5N methanolic acid and a ca lib ra tion  curve prepared 
with se r ia lly  d iluted  cyanidin ch loride solutions as shown in 
Appendix D.
The stock oxtract (d ilu ted  20 tin es) was used fo r  the 
determinations and op tica l density readings measured at 525 nra 
against the blanks
TABLE iv
Determination o f anthocyanin concentration
Reactants Experimental (ml) Blank (n l)
Extract 1,0 1.0
^Aqueous methanolic acid 3.0 3 .0
+Methanolic HG1  -  1.0
aHydrogen peroxide 1.0 -
5.0 5.0
*Aqueous methanolic acid (0„5N HClj 855° v /v  absolute methanol). 
+Methanolic hydrochloric acid (5:1 v /v  aqueous methanol, 3N HCl) 
«The reaction  was started by the addition  o f hydrogen peroxide
( 1.0  ml fresh ly  prepared 30$S hydrogen peroxide in 9 nl* methanolic 
a c id ).
University of Ghana                              http://ugspace.ug.edu.gh
55
1) Culture media,
Erlenmeyer flasks (250 ml capacity) each containing 30 ml
of medium were used for  a l l  the liqu id  cu ltures. Each o f the 9 cm
diameter, p etr i dishes contained 20 ml, o f  agar medium.
1. Potato Dextrose Agar: 200g peeled Ir ish  potato were boiled  in
500 ml, d is t i l le d  water, strained and 
made up to 1000 ml; 20g dextrose; 20g 
agar,
2. Sabouraud’s Agar; 40 g maltose; 10 g peptones 15 g agar
in  1000 ml, d is t i l le d  water,
3 . Czapek Pox Agar: 2,0 g Sodium n itrate  (NaNO^); 1.0 g
potassium phosphate (K HP0„); 0 .5g
c. 4-
potassium ch loride (KCl); 0 .5 g magnesium 
sulphate (MgSO^.THgO); 0 ,01g ferrous 
sulphate (PeSO^)? 30g sucrose; 15,0 g
agar; and 1000 ml ml d is t i l le d  water.
4 . Sabouraud's broth medium: 40g maltose; 10 g peptone; and 1000
ml. d is t i l le d  water,
5. Czapek Pox broth medium: 2 ,0  g Sodium n itra te  (NaNO^)j 1.0 g
potassium phosphate (iCjHPO^); 0 . 5g pota-. 
ssium ch loride  (KCl); 0 .5g  magnesium 
sulphate (MgSO^.THgO); 0,01 g ferrous 
sulphate (PeSO^); 30g sucrose; and 1000 
ml, d is t i l le d  water.
University of Ghana                              http://ugspace.ug.edu.gh
36
m) Chor.icals
A ll chemicals used in the preparation o f  media were either o f 
the "Analar" or the "BOH" (B ritish  Drug House) grade.
n) Chromatographic analyses 
1• Paper Chromatography
For the separation, detection  and id en tifica tion  o f phenolic 
compounds, the non-hydrolysed (absolute methanol) and the hydrolysed 
(505® methanol) fraction s o f the cocoa bean testa , were used in  a 
one dimensional ascending chromatogram in  a Shandon "UniKit" tank 
at 25° + 1°C. The chromatograms were developed in  two separate 
solvent mixturest ( i )  BL\! (n -lu tan o l-g la c ia l acetic  a c id , water, 
4 :1 :5 , top layer) (G r iffith  1958); ( i i )  Forestal (a ce tic  acid  -  
conc. HC1-water, 10:3:3 0 ), 50ft HoAc (50^ aqueous a cetic  a c id ),
BA¥, PhOH (Phenol saturated with water), and water (Harbcme, 1973).
Application o f test Solutions and th e ir  detection
Extracts o f (0 . 1ml) were spotted on chromatographic paper 
(T;fhatman No.l) by repeated application  and dried in  a cold  stream 
o f  a ir  from an e le c tr ic  h a ir -d r ie r , Bach chromatogram was run fo r  
10 hours. At the end o f th is  period, the paper was ca re fu lly  
removed from the tank, the margin o f the solvent front marked, and 
the chromatogram dried in a ir .  Standard markers o f known iden tity  
were always run with the unknowns.
University of Ghana                              http://ugspace.ug.edu.gh
37
The dried chromatograms were detected by th e ir  flucreacencc 
in short wave length (253 nm) and long wave length (365 nm) o f 
u ltra  v io le t  ligh t from an Ultra V io le t  lamp (Chromato-Vue, Model 
cc 20) according to the recommendations o f G r iffith  (1958)» Further 
detection  o f the catechins and leucooyanidin was achieved by the 
use o f chromogenic spray reagent diazotised  p -n itroan ilin e  (Swain, 
1953).
2. Thin Laver Chromatography
For further study o f the phenolic constituents o f cocoa bean 
testa , the hydrolysed and non-hydrolysed extracts were separated by 
th in -layer chromatography (TLC) on plates (20 x 20 cm) o f activated 
s i l i c a  gel (a (Type 60) containing fluorescent indicator and 
ce llu lose  MN 300 (1:1)  with a thickness o f 0,25 mm. The extracts 
were spotted on p lates, steamed and developed at room tenperature 
25° + 1°C in  the solvent system o f ch loroform -acetic acid-water 
(4 :1 :5 ) (CAW) as ties crib  ed by van Sumere et al (196 5) *
Id en tifica tion  of the spots was achieved by comparing the
Rf values and colour reactions obtained by the use o f  a chromogenic 
spray reagent d iazotised p -n itroan iline  as described by Swain,
(1953) with those o f  authentic phenolic Barkers, The spots were 
scraped individually  and the phenolic compounds eluted from the 
s i l ic a  gel by washing with ether. The extracts were centrifuged
University of Ghana                              http://ugspace.ug.edu.gh
and the supernatant evaporated to dryness. The crysta ls  foreod 
were stored in a deSiic.cator fo r  subsequent experiments.
Confirmation o f phenolic id en tifica tion s  by U.V. spectral 
comparisons was however, not done.
p) Experimental -precautions
1. Glassware was kept scrupulously clean* Glassware which had 
already been cleaned with water and detergent was rinsed several 
times with tap water and three times with d is t i l le d  water and 
allowed to drain before use.
2 . Oven dried f i l t e r  paper usually lo s t  some weight because o f 
heating. F ilte r  paper used throughout these investigations was, 
therefore, heated at 75°C fo r  24 hours prior to  use.
3 . F ilte r  paper with loads o f dried mycelium were always conveyed 
to the balance room in a closed desS&cator to  avoid absorption o f 
moisture,
4 . P etri dishes with beans on agar plate media fo r  mycoflora 
determinations were stored in platforms with legs  standing in  o i l  
to prevent invasion by in sects and other organisms,
5. 'Conidia o f the same age were used in a l l  experiments.
q.) S ta tis t ica l methods
Results were s ta t is t ic a l ly  analysed where appropriate.
University of Ghana                              http://ugspace.ug.edu.gh
39
IV. RESULTS
A. LEVEL OF FUNGAL CONTAMINATION OF COCOA BEANS AT TEMA
'  In 1930, cocoa from Ghana arriving at United Kingdom ports 
were found to contain 3.1 per cent fungal contamination or internal 
moulds (Passmore, 1932). Because o f  the absence o f any more 
recent records, the leve l of contamination o f cocoa beans, stored 
in  the warehouse o f the port o f Tema, by fungi was investigated .
Cocoa bean samples were removed at monthly in terva ls from 
October 1978 to  July 1979 as they arrived at the warehouse. The 
beans had been stored for  varying periods at the various purchasing 
stations in the cocoa growing areas before arriving at Tema. There 
were neither records of the exact periods o f  storage at the d i f fe ­
rent purchasing centres nor the conditions under which they were 
kept. The beans at the Ten®, warehouses a lso  contained batches o f 
both main crop, harvested in August to December in 1978, and the 
minor crop, harvested in May and June in  1979. Samples from the 
main and minor crops were kept separate.
Beans fo r  the mycoflora studies were obtained by removing a 
handful o f beans from the top , and then m id-section and bottom o f 
every hundredth bag. The three lo ts  from a l l  the bags sampled 
were pooled together. Two kilograms o f  beans were f in a lly  taken 
on each sampling day and stored in sealed k ilner jars in a freezer
University of Ghana                              http://ugspace.ug.edu.gh
40
at 4°C un til required fo r  examination. The moisture content o f 
the beans was in the meantime determined by drying samples from 
the remaining beans at 105°C in  a ventilated  oven fo r  5 hours.
One hundred beans were randomly selected  from those stored 
in the freezer , su r fa ce -s te r ilis ed  with 0 , 1$  mercuric eh loride 
solution  and incubated at 25°C in humid chambers fo r  10 days. The 
percentages o f beans containing fungi in  the various testa  are 
presented in Table 1, Each value in  the Table represents the mean 
o f  three rep lica tes  o f 100 beans each.
The resu lts show an annual average fungal contamination o f 
4,9 per cento The main crop had a lower average percentage o f 3 ,9  
per cent than the minor crop with an average contamination o f  6»4 
per cent.
There was, as should be expected, a re lationsh ip  between 
moisture content o f the beans and the leve ls  o f fungal contemina** 
tion* Percentage contamination rose with increasing bean moisture 
content and f e l l  with reduction in bean moisture content* (F ig , l ) »
University of Ghana                              http://ugspace.ug.edu.gh
41
TABLE 1
MOISTURE CONTENT AND FUNGAL COMMUTATION OF COCOA 
BEANS IN WAREHOUSES AT TEB1A
CROP
YEAR AND MONTH 
OF SAMPLING 
(DATE OF ARRIVAL 
AT WAREHOUSE)
AVERAGE
PERCENTAGE
FUNGAL
CONTAMINATION 
+ S *D °
AVERAGE 
PERCENTAGE 
MOISTURE 
CONTENT 
+ S .D .
1978  NOVEMBER 6 .0  +  2.5 6 ,5 ±  1.9
DECEMBER 2,3 + 0.2 5.4 ±  1.6
M IN 1979 JANUARY 3.7 + 0,3 4,9 +  2.0
FEBRB.j.. _ 2,7 + 1 A 5.7 +  1.1
MARCH 4.0  +  2.5 6 .3 + 0,.7
APRIL 4.7 ±  1.9 6.0 +  0.8
1979 MAY 4,6 1.9 7.5 ±  1 .2
JUNE 7,0 ±  2.2 7.3 + 0.9
MINOR JULY 8 . 0  +  1 ,5 7.6 + 2.1
AUGUST 6 . 0  _+ 1 .4 6 . 3  +  2 . 0
University of Ghana                              http://ugspace.ug.edu.gh
PE
RC
EN
TA
G
E 
M
O
IS
TU
RE
 
C
O
N
T
E
N
T
• 9 -0  
8 0  
■70 
■60 
5 0
3 °
2-0
1978 1979
FIG.l  GRAPH  SHOWING  RELATIONSHIP BETWEEN MO ISTURE  CONTENT O F  STORED COCOA BEANS ( •  
AND THE LEVEL OF FUNGAL CONTAMINATION ( X ---------------- X )
PE
RC
EN
TA
G
E 
FU
N
G
A
L 
C
O
N
T
A
M
IN
A
T
IO
N
University of Ghana                              http://ugspace.ug.edu.gh
43
B. FUNSI ISOLATED FROM STORED COCOA BEAMS
Dade (1929)' and. Bunting (1928, 1931) id en tified  Aspergillus 
fumigatus and Mucor buntingil  as regu larly  occurring in  fermenta­
tion  heaps, Aspergi l lu s  flavu s , Aspergillus tamari i  end Asper g il lu s 
ochraceus were iso la ted  from dry cocoa beans by Bunting (1929a).
In these reports the frequency o f occurrence o f these fungal conta­
minants was not shown. In Nigeria,, as many as 51 fungus species 
have been id en tified  as occurring in commercial cocoa beans 
(Oyeniran and Adeniji, 1974). I t  appears that the problem has not 
been fu l ly  examined in  Ghana, Fungi, occuring in  cocoa beans in 
the warehouses at Tema were therefore c r i t i c a l ly  studied. Infected 
beans were plated on Potato Dextrose Agar, Sabouraud's Agar and 
Czapek Dox Agar media and fungi growing from them were recorded 
a fter  10 days.
Twenty-two fungi l is te d  in Table 2 were iso la ted  from the 
beans. These included three species ; Asperg illu s phoenlc i s . 
Rhizopus oryzae and Neurospora crassa being recorded fo r  the f i r s t  
time on cocoa beans. The dominant contaminants were Aspergillus 
chevali e r i .  Aspergillu s tamarii, Neurospora crassa, and Penicillium  
citrium . Asperg illu s  ochraceus which was iso lated  by Bunting (1931) 
was missing from the f lo ra  on beans in  the Tema warehouse?
University of Ghana                              http://ugspace.ug.edu.gh
44
TABLE 2
FUNGI ISOLATED FROM STORED COCOA BEAU ON SABOURAUEfe AGAR.
FJNGUS SPECIES FREQUENCY OF OCCURRENCE
1°
Aspergillus ch eva lieri (Mang,) Thom & Church 17.6
A. flavus Link 5.3
A, fumigatus Fres 7 .0
A. nidulans (Eidem) ¥iiit 7 ,0
A. niger van Tieghem 2.6
A, pheonicis (Corda) Thom 2.6
A. restrictu s  Smith 7.0
A* ruber (Konig. Spieclc & Brem) Thom & Church 1.0
A. sydowi (Bain and Sart) Thom & Church 0.3
At, tamarii Kit a 13.0
A, ustus (Bainier) Thom & Church 0.3
A. wentii Wehmer 0.6
Botryodiplodia theobromae Pat. 3.6
Fusarium solan i (Mart.) Sacc. 1.0
Mucor lm sillus Lindt 8.6
Hycelia s t e r i l ia 2.6
Neurospora crasaa Shear and Dod^s 12.0
Penicillium  citrium Thom 14,3
Penicilj.ium sp , No, 1 4.3
Penicilliuu^ sp. Nnr .? 1.6
Rhizopus arrhizus FisnViPr- 2.0
.2£ZSae. Went and Prinsen-Geerlings 8.3
University of Ghana                              http://ugspace.ug.edu.gh
C-„ EFFECT OF COCOA J  '-.AN TESTA EXTRACT QV 6RQT'!TH OP SETECTED
COBTAI-ffilAICT FtMGT
The influence of the host plant in host~paras.ite in teraction  
is  well known (eg, Tomiyama, Sakai, Sakuma, IV "'.izaka, 1564? Kosuge, 
1969), It  was therefore considered necessary to  find  the ro le  o f 
the testa  in the invasion o f the cocoa "beans by fungi during 
storage. The influence o f  the bean on the invading fungi cannot 
be predicted in view o f the existence o f both stimulatory compounds 
particu larly* amino acids, carbohydrates and growth fa ctors  (Rohan, 
■'963;), and to x i;  compounds such as phenolic compounds and organic 
acids (Humphries, 1944“ Roelofsen,. 1958? Swain, 1960? Rohan, 1963) » 
The growth c f  f iv e  Aspergillus species, Ao flavus. A« f umigatus.
Ac n ^ u ^ a n s ,  A . n ig e r  and A , t a m a r i i .  i s o l a t e d  from  con tam ina ted  
c o c o a  b ea n s , and one n on -con tam in an t S c le r o t iu m  jr c iy f s i i  * i n  c o c o a  
bean  t s s v a  e x t r a c t s  Was s t e c ’
The extracts were prepared with testa  taken frcni boans at 
varying stages of curing, and the dry weights o f nyeelia produced 
in these media a fter  7 days at 25°C are presents i ia  Table 3a -  f ,
The resu lts  ''how an in teresting  response by fungi to  the 
extracts,, Growth was high in extract .of testa  o f uncured beans it. 
a l l  oases* Extracts from 2- and 6-- day cured beans were inh ibitory  
depressing growth of the fungi to varying degrees, While the 
extracts from beans cured fo r  8 and 12 days remained inh ibitory  to  
.£@46§Sy§, and. A, tama r i i , they supported approximately the same
University of Ghana                              http://ugspace.ug.edu.gh
46
growth as extracts of testa  o f tho uncured boansf These e ffe c ts  
arc d e a r ly  depicted in  Fig. 2 which brings the data together fo r  
comparative purposes.
University of Ghana                              http://ugspace.ug.edu.gh
47
TABLE 5a
GROWTH OF ASPERGILLPS FLAVUS IN EXTRACT OF COCOA BEAN AT
25°C FOR 7 DAYS
Time o f Curing Mean dry weight
o f  Bean 1-- fc-ng of mycelium
extract (mg6 + S.D .)
(Days)
0 45.6 ±  5.1
2 27.7 + 6.2
6 50.4 ±  7.5
8 47.9 + 1.9
12 44.7 + 5 .8
University of Ghana                              http://ugspace.ug.edu.gh
48
TABIE3b
GROWTH OP ASPERGILLUS PUMIGATUS IN EXTRACT CP COCOA
BEAN AT 25 °0 I'iS 7 DAYS
Time o f  curing o f 
Been providing 
extract
(Days)
Mean dry weight 
o f mycelium
(mffr + S„D,)
0 64,3 ±  4c4
2 23,4 + 7 ,2
6 25.0 _+ 5c9
8 35,6 a. 2 ,3
12 43<3 ±  2.7
University of Ghana                              http://ugspace.ug.edu.gh
49
TABLE,, go
GROWTH OF ASPERGILLUS NIDULANS IN EXTRACT OF COCOA 
BBAH AT 25°C FOR 7 DAYS
Time of curing o f Mean dry weight
Bean providing o f mycelium
extract
(Days) (m g .+ S .D .)
0
2
6
8
12
46.9 + 9.2
29.0 ±  6.4
31.5 ±  5.8
49.5 + 2.3
48.1 + 5.0
University of Ghana                              http://ugspace.ug.edu.gh
5o
TABLE .3d
(SOUTH OF ^JPinQILLIJS NIGSR IN EXTRACT. Qg COCOA BEM. 
AT 25°C FOR 7 DAYS
Mean dry weight 
Boan providing o f mycelium
extract 
(Days) (mg* £  S.D*)
0 4 3 .9 + 3 .8
2 2 5 .7 + 6 .1
6 26.8 + 7.3
8 45.4 + 2^7
12 41.3 + 3.8
University of Ghana                              http://ugspace.ug.edu.gh
51
TAELS 39
GROWTH OF ASPERGILLUS - TAMARII IN EXTRACT OF COCOA BEAN
AT 25°C FOR 7 DAYS
Time o f curing Mean dry weight
o f bean providing o f  mycelium
extract
(Days) (mg. + S.D .)
0 69.2 + 4.5
2 34»0 + 4 ,0
6 28.6 + 3.7
8 47.6 + 2.7
University of Ghana                              http://ugspace.ug.edu.gh
52
TABUS i f
GROWTH OF SGEBEOTITO .HOLESn, IfrECTRAGT. OF COCOA BEAN 
AT 25°G FOR 7 DAIS
Time o f curing o f  Mean dry weight
bean providing o f mycelium
extract
(Days) ■ (® g i± S .D i)
0 4-3,2 + 6 ,6
2 40*9 ± 3 .9
6 15*1 ±  4.8
8 5 6 ,2 + 4 ,3
12 47.5 + 2*4
University of Ghana                              http://ugspace.ug.edu.gh
53
7°6
DAYS OF  CURING  OF BEAN PROV ID ING  E X T R A C T
F IG  2 GROWTH  OF FUNGI  IN 2 0 M L  C O COA  BEAN TESTA 
E X T R A C T S  AT 25°C . FOR 7 DAYS .
■ — ■  ~~^~ ^A ' L f  U M I G A T U S ■ O  O .  \ .^ T A M ^ i T  
A  A A ■ N I D U L A N 5 ; D  — D , 5 C L ERQ T |UM  ROL FS I I
University of Ghana                              http://ugspace.ug.edu.gh
54
D. GROWTH OF THE FUNGI IN NUTRIENT MEDIA
A curious feature of the resu lts of the experiments described 
in the previous chapter was the uniform pattern o f response to  the 
extracts o f the cocoa testa  by a l l  the fungi tested . Besides, dry 
weights of mycelium produced in extracts o f 8-day cured beans were 
p ra ctica lly  the sane. Two fa c ts  need to be established . F irst, 
are these fungi normally sim ilar in th e ir  growth habits or the 
observations were the e ffe c ts  o f the extracts. Secondly, do the 
observations indicate depressed or enhanced growth? The fungi were 
grown, in a subsequent experiment, in nutrient media in  an attempt 
to  provide answers to  these questions.
The fungi were f i r s t  grown in Czapek Dox Broth at 25°C and 
samples were drawn a fte r  3, 7, 11 find 14 days, respective ly , fo r  
dry weight determination o f the mycelium produced. The resu lts 
obtained are shown in Table 4 , Each value is  a mean o f 4 rep lica tes .
The resu lts c lea r ly  show that the fungi varied in their  growth 
habits. D ifferences were observed in (a) rate o f  growth, (b) time 
of attaining maximum growth and (c )  the amount o f mycelium produced. 
The last two features are shown very c lea rly  in the graph in Fig. 3. 
Among the Aspergillus species, which came from cocoa beans, A, tamarii 
grew best, producing a maximum mean dry weight o f 287.2 mg while 
A. nidjJl ans provided the smallest mean maximum dry weight o f 126.2mg,
University of Ghana                              http://ugspace.ug.edu.gh
55
A, fumggtus grew fastest in the medium, reaching maximum point o f 
growth at e ither 3 days or between 3 and 7 days, followed by 
niger  and A. _tamarii which grew the greatest dry weights at 7 
days, while maximum growth was not attained in A, -Ha-giia and A. nidu- 
lam  un til a fte r  14 days.
Sclerotium r o l f s l l* s  pattern o f growth was sim ilar to  those 
o f  Ai nift-er and Afi tc.marii and was also next to these in productiv ity .
The experiment was repeated using Sabouraud's agar and broth 
media, There were fou? rep lica tes  fo r  each species in each t e s t . 
Tables 5 and 6 contain the resu lts obtained in  these experiments.
The resu lts can be summarised as fo llow s: (a) there was variation
in  the amount o f ayee'ivm produced in Sabouraud's broth (b) comparing 
performance in the Czapek Box broth and Sabouraud's broth a fte r  7 
days incubation, A, niger and A, tag.gr i l  grow less  well in  Sabouraud's 
medium,, while A. nidulans was higMy stimulated, (c ) Growth rate 
varied on the agar medium, find (d) the poorest growth o f A, nidulans 
in Cr-r.pek Pox broth among the species was confirmed by growth te s t  
on the a;-a? medium at both 25° and 30°Ct>
University of Ghana                              http://ugspace.ug.edu.gh
56
TABLE A
GROWTH OF ASPSItGILLTJS SPBCTF.S ISOLATED MOM SfOREp. COCOA 
BEANS. AMD SCLERQTIp  ROLFSII IN CZAPBK POX BROTH AT 25°C
_  „w rT „  MSAlf DRY WEIGHT OF MTCELIUM (mg + S.D .) AFTER
°  or'uUi- ,J? INDICATED DAYS OF INCUBATION
5 7 11 14
Aspergillus
flavus 167,2 + 4,7 168,2 + 108 145.0 + 4,0 186,2 + 0 ,9me*
A, funigatus 226,3 ±  8 ,4 199,5 + 3.1 199,0 4  7,6 158,4 ±  4 ,4
A, nidulaias 73,2 ,t 9,7 83,6 ± 3,4 107,1 + 7 o2 126,2 + 5,7
&.? n ig?r 170,9 & 10«4 279.2 ^ 5,1 214,4 + 6,1 218,9 ±  1.9
A 5 tnmaril. 154,2 ±  2,1 287,2 + 4,0 247,3 + 3,2 258,2 +10,7
Seleroto^^
r o l f s i i 95.4 + 0,8 277,2 + 4,9 174.2 + 8,3 196,5 +14,5
University of Ghana                              http://ugspace.ug.edu.gh
57
GROWTH OF ASPERGILLUS SPECIES ISOLATED FROM STORED COCOA BEANS
TABLE 5
AND SCLEROTIUM ROLFSII IN SABOURAUD1S BROTH AT 25°C FOR 7 DAYS
Fungus species Mean dry weight o f mycelium 
(mg. + S.D.)
Aspergillus flavus 126,5 + 9.1
A» funisatus 230.2 + 7.9
A„ nidulans 172,1 + 4.6
A, niger 129.3 + 3«6
A. tanarii 124.0 + 6 ,0
Sclerotium r o l f s i i 182,1 + 5,0
University of Ghana                              http://ugspace.ug.edu.gh
58
GR0>ITH OF ASE5RGILLU1 SP3CIES ISOLATE!) FROM STOPJ3D COCOA BEAMS 
JMD SCLEROTIUM ROLFSII OH SA3QURAIJD!_S AGAR MEDIUM! AT 25° AMD 30°C
TABLE 6
Temperature Mean diameter o f colony (nn .)
of incubation Fungus species a fte r  indicated days o f incubation
2 4 6 8
^sjoergillus £lenms 11,0 25.0 37.0 51.5
A, fjAnxgatug 13.0 18.5 42.0 54.0
A. nidulans 3,0 9.5 13.5. 18.0,
A, niger 10.0 23.0 32.5 44.0
A„ tanarii 9,0 26.0 31 rO 41.5
Sclerotium r o l f s i i 13.0 26,0 39.0 51,5
Aspergillus f la w s 16,0 30.0 43.0 64.5
Ae fumgatus 13.0 25.0 39.0 61.0
A* _nidul,ans 0,0 10.0 13.5 19.0
A. n^ -^ .er 19.0 34,0 44,0. 59.0
A, tamarii 16,5 29.0 39-0 54.5
JScljrqtim  r o l f s i i 11*5 22.5 35.0 50.5
University of Ghana                              http://ugspace.ug.edu.gh
. O A. TAMARt l
-X A. NIGER
S.ROLFS!!
x » A -f lavus
/
* A.FUMfGAT.U-S
..OANtQUtAHS
University of Ghana                              http://ugspace.ug.edu.gh
60
S. ANALYSIS _0ff PHENOLIC CONSTITUENTS OF THE COCOA BEAN TESTA
There is  much published work on the chemical constituents o f 
tho cocoa bean (Forsyth, 1952b; Howat, 1957; G r iffith , 1958;
Swain, 1950; Bohan, 1963)* Phenols are considered to be some of the 
important components of the cotyledons (Knapp, 1937; Forsyth, 1954; 
Forsyth and Quesnel, 1957; William, 1957; Forsyth and Rombouts, 195tj 
Roelofsen, 1958). In contrast, there is  l i t t l e  published on the 
phenolic component of the testa  even though Cadman (1959) considered 
the phenolic compounds o f the cocoa bean testa  to  play a very s ign i­
fican t ro le  in  the in fect ion  of the bean by pathogens <, Reports 
available mostly contain qualitative rather than quantitative 
assessments o f phenolic compounds o f  the te sta . Studies of Forsyth 
and Rombouts, (1951) on quantitative changes showed that the to ta l 
concentration o f phenols in the bean remained unchanged during 
curing, whilst the simpler mono-phenols were destroyed and condensed 
in to insoluble tannins and more complex phenols. In h is analyses 
on tho d istribu tion  of phenolic compounds in  cocoa plants, G r iffith  
(195Q) observed that the testa  contained leuco-anthooyanins not 
found in other parts o f the p lant, Leuco-anthocyanins are formed 
from the cordonsation o f simpler phenols in  the cotyledons which 
are then exuded into the testa . Analysis o f the non-hydrolysed 
extracts of the testa  showed that the only free  soluble phenolic 
component o f the testa  was g on tis ic  acid .
University of Ghana                              http://ugspace.ug.edu.gh
61
Luring' tho during of cocoa beans, anthocyanins are destroyed 
a fter  the f i r s t  day of fermentation (Forsyth and Rombouts, 1951)• 
Theobromine a lso leaches out o f the cotyledons in to  the testa .
Forsyth ( 1952b) demonstrated that as much as 20 -  30 per cent o f 
catechins leached in to tho testa  during fermentation?
The amount o f free  phenols, orthodihydric phenolsj flavonols , 
and anthocyanins in  both the cocoa bean cotyledon, and testa  at 
d ifferen t stages of curing wore measured. There were tw o.rep licates 
for  each test,. Determinations made are presented respective ly  in 
Tables 7.. 8, 9 , 10 and in  Fig*
i ) Fgco Phenols content
A wide variety  of simple and complex compounds possessing 
phenolic hydroxyl groups occur in the cocoa plant tissue (Swain,
1950; G r iffith . 1958) sujh as the simple ca ffe ic  acid  and the complex 
flavonol*
Caffeic acid flavonol (quercetin)
Tho simple or free phenols usually c.orabine in polymerisation reactions 
to form complex insoluble tannins (Forsyth and Qucsnel, 1957)•
2z,llas and F:.Ght (1939) showed that the concentration of soluble 
tannins f e l l  during fermontation from a leve l o f 19$ o f the to ta l
University of Ghana                              http://ugspace.ug.edu.gh
62
phenols to  less than 80 at the end o f fermentation. The resu lts in 
Table 7 show that the in it ia l  concentration o f free phenols was 
higher in tho cotyledons than in  the testa . The free  soluble phenols 
broke down rapidly during fermentation and on the la s t  day o f  fermen­
tation  only 0»73iag/g o f the free soluble phenols was detected, tilth 
drying, the concentration reduced s ligh t ly  to  0*70 ng /g . In the 
testa , the to ta l free  phenols concentration Was lower than that o f 
the cotyledons v iz , 0,41 Bg/g dry weight* This amount also f e l l  
with curing so that at the e:ad o f fermentation i t  had been reduced 
to 0.J2 mg/g and to 0.22 mg/g on the 6th day of drying.
University of Ghana                              http://ugspace.ug.edu.gh
FREE PHENOLS IN COCOA. BEANS AT VARIOUS STAGES Off FERMENTATION
MD DURING DRYING
6r
TABLE 7
Treatment 
o f be an
Mean concentration o f Free 
Phenols (ra g /g+S .D .)
COTYLEDON TESTA
FERMENTATION 
0 DAYS 
2 DAYS
6 DAYS
4.00 + 1.11 
1,15 + 0.82 
0.73 + 0 .1 6
0,41 + 0.09 
0.45 + 0.08 
0.32 + 0,11
DRYING
6 DAYS 0.70 + 0.20 0.22 + 0.06
University of Ghana                              http://ugspace.ug.edu.gh
64
The investigations o f Adomako (1975) on the e ffe c t  o f orth cd i- 
hydric phono?..**' on virus diseases o f the cocoa tree gave one o f the 
f e f  reports on the relationsh ip  between orthodihydric phenols and 
host-parasite in teraction . Forsyth and Rombouts (1951) in  discussing 
flavour development during fermentation, observed that there was some 
condensation resu lting  in the destruction o f simple polyphenols and 
the formation of insoluble tannins, The elucidation  of the flavan— 
3 :4 -d io l as an inevitable  intermediate in the hydrolysis of leu co - 
cyanidins supported th is  claim (Robinson and Robinson, 1933). The 
rapid reduction in  the amounts o f free soluble phenols as observed 
in the previous experiments suggested the formation of insoluble  
tannins as products o f condensation and polymerisation reactions. 
Changes in  the concentration o f orthodihydric phenols in  the cocoa 
bean during curing were determined spectrophotometrically at 520 nn.
As shown in Table 8, the in it ia l  concentration o f orthodihy­
dric phenols in the testa  o f unfermented cocoa beans of 2,§3 mg/g 
was far higher than the concentration o f 0,33 mg/g in  the cotyledons4 
However, with fermentation, the concentration o f orthodihydric 
phenols in the cotyledons increased rapidly  to approximately ten 
times the in it ia l  r'H'unt at the end of the curing process. This 
suggested that some polymerisation o f the free soluble phenols had 
occurred during fermentation o f the cotyledons, On the other hand,
ii )  Orthodihydric Phenols content
University of Ghana                              http://ugspace.ug.edu.gh
the concentration of orthodihydric phenols in  the testa  decreased 
s ligh tly  iti the firs-t two days o f curing before r is in g  again to a 
lev e l higher than that o f  the uncuied beani The concentrations 
o f the orthodihydric phenols in both the cotyledons and testa  
decreased during drying o f the fermented bean to approximately the 
same le v e l.
University of Ghana                              http://ugspace.ug.edu.gh
TAI3LB 8
ORTHODIHYDRIC PHENOLS IK COCOA BEAMS AT VARIOUS. STAGES .OF 
FERMENTATION AND DURING PRYING
Treatment 
of 'bean
Mean concentration o f orthodi­
hydric phenols (mg/g + 8#D.)
COTYLEDON TESTA
FERMSHPATION
0 DAY 0.33 ±  0,18 2.43 + 0.22
2 DAYS 2.18 + 0,47 2.19 + 0.72
6 DAYS 3 i41 ±  0,09 3.15 ±  0.81
DRYING
6 DAYS 2.42 ±  0.88 2.46 + 0.95
University of Ghana                              http://ugspace.ug.edu.gh
m  •
Polymerisation •of free pfoewvols «nd ph scale «o*4d ■Pftsui'C
in  more complex flavonoid sampounds such as flavonols being formed 
(Bate-Smith, 1954)* These have the basic flavan-3i4—d io l structure , 
Sehranf£stater (1948) repoartesd that morin, a derivative  of flavonol
30
flavan -3 :
showed a n ild  bacteriostatics e ffect on Staphylococcus aureus.
The concentration of flavonols in  the cotyledons and the  teats, during 
curing were determined using the v a n il l in  method (Swain and H i l l i s ,  1959)* 
The resu lting  yellowish colour was measured at 500nn (the wavelength at 
which ahsorhence was highest) with ( - )  epieateehin as standard*
There was a higher concentration of flavonols, according to the data 
in  Table 9, in  the unfexmented testa (0,51mg/g dry weight) than in  the 
cotyledons (0f21mg/g). During fermentation, while the flavonol concentra­
tion  in  the cotyledons increased to more than three times the in i t ia l  amount 
a t the end of the curing period, i t  decreased in  the testa . The concentra­
tion  then f e l l  in  both cotyledon and testa during the 6 days of drying*
i i i )  KLavonol oontent
University of Ghana                              http://ugspace.ug.edu.gh
68
FLAVONOLS IN COCOA B3AHS AT VARIOUS STAGES OF FERMENTATION 
AMD DURING DRYING
TABLE 9
Treatment of 
Bean
Mean concentration 
(mg/g + S .D .)
o f flavonols
COTYLEDON TESTA
FERMENTATION
0 DAY 0.21 + 0.25 0.51 ±  0.15
2 DAYS 0,39 ±  0.16 0.33 + 0.14
6 DAYS Ot>72 0.09 0.43 + 0.10
DRYING
6 DAYS 0,49 + 0.11 0,37 + 0.04
University of Ghana                              http://ugspace.ug.edu.gh
Forsyth and Rombouts (1951) observed that anthocyanins were 
rapidly destroyed a fter  the f i r s t  day of fermentation* These 
compounds which form 4 per cent of to ta l phenols in the unfermented 
bean (Forsyth and Quesnel; 1963) were responsible, together with
leuooanthocyanins, fo r  the browning o f cocoa beans during fermenta­
tion*
Extracts o f the cotyledon and testa  were assayed spectropho— 
tom etrically  at 525’nm a fter  precip ita tion  o f other compounds which 
absorb at same wavelength (Swain and Hi11i s , 1959). Pure cyanidin 
chloride dissolved in 0.5N methanolic acid was used as a standard. 
The resu lts of anthocyanin concentrations in cocoa bean during 
curing (Table 10) indicated a higher concentration in the co ty le ­
dons than in the testa  o f the unfermented beans* The amount in
the cotyledons decreased at the onset o f fermentation before 
rising  again to a leve l above the in it ia l  concentration at the 
end of fermentation. The concentration decreased during drying.
It  was found that in  the testa* the anthoeyanin concentration 
increased throughout the period o f fermentation and during drying*.
iv ) Ant--,oovanir. bontent
University of Ghana                              http://ugspace.ug.edu.gh
70
ANTHOCYANIN IN COCOA BEANS AT VARIOUS STAGES OF FERMENTATION
AND DURING DRYING
TABLE 10
Treatment 
o f bean
Mean concentration o f anthocyanin 
(mg/e + S .D .)
COTYLEDON TESTA
IEBKENTAS ION
0 DAY 0.16 + 0.02 0.07 + 0.02
2 DAYS 0.13 + 0,02 0.07 ±  0.01
6 DAYS 0,20 + .0,05 0.45 ±  0.04
DRYING
6 DAYS 0.14 + 0,01 0.24 + 0.05
University of Ghana                              http://ugspace.ug.edu.gh
71
FlG. 4 PHENOLIC COMPOUNDS IN TESTA OF COCOA SEANS DUPING
CUR iNG .
University of Ghana                              http://ugspace.ug.edu.gh
S’, CHROMATOGRAPHIC. SEPARATION.,AMD IDENTIFICATION OF PHENOLIC 
' COMPOUNDS . OF C.OC QA BEAM TESTA
G riffith  (1958) id en tified  by paper chromatography gentinle 
aci&, cvanidin hydrochloride, epicatechins and anthocyanins in  the 
testa  o f fresh  cocoa beans. He a lso  observed that phenolic compounds 
and other tannins especia lly  s a l ic y c l ic  acid and cinnamic acid 
derivatives were germination inh ib itors o f sp ices and legume* seeds, 
Caffeic a c id ? p-coumaric and protocatechuic acid , iso la ted  from 
other plants, were found to be mild inh ib itors (Mason, 1955? Varga 
and Koves, 1958; de Roubaix and Lazaar, 1960). These workers noted 
that no single  phenolic compound ought to  be considered a sp e c if ic  
germination inh ib ito r  since d iffe ren t combinations o f a l l  these 
substances occurred in d ifferen t f r u it s ,  Forsyth ( 1952b) had 
ea rlier  id en tified  three anthocyanins in forestero  cocoa; cyanidin 
monoglucose, cyanidin arabinoglucose and cyanidin d ig lucoside.
While Yeboah (1965) achieved good separation o f phenolic compounds 
Citru s aurantium on paper chromatograms, Harbojme (1973) reported 
that chromatographic separation o f  phenolic compounds was better 
on thin layer plates,. The technique o f thin layer chromatography 
(TLC) had a lso been used su ccessfu lly  by other workers (Stahl, 1969; 
Seikel, 1962, van Sumere at a l,1965 ). Cocoa bean testa  extracts 
were analysed by seixr.ation methods involving paper chromatography 
and thin layer chromatography, A fter extraction , the samples were 
spotted on Whatman's No, 1 chromatographic paper, the spots developed,
University of Ghana                              http://ugspace.ug.edu.gh
and id en tified  by their Rf values and colour reactions* Paper 
chromatograms o f testa  extracts (non-hydrolysed and. hydrolysed) 
developed in two solvent mixtures ( a) sing le  solvent system, and 
(h) multi-*solvent system are presented in Figs^ 5a *  d* Colour 
reactions of spots to short -  and longwavelengths o f IJV lig h t  are 
shown cn PI at a a 1.a -  a.
The p re -id en tified  spots on the thin layer plates using 
S ilica  Gel C- (Type 60) as support and development in  chloroform** 
acetic acid-water (4—1—5) were co llea ted  separately and the phenolic 
components recovered by eluting in  ether. Solutions eluted from 
these spots were centrifuged to  remove the s i l i c a  gel material and 
the supernatant was evaporated to dryness. The resu lting crysta ls 
dissolved in d is t i l le d  water were designated as NO^  with Rf value 
o f 0 ,42, and with Rf values 0,31 and 0j54 respectively#
Crystals from spots o f hydrolysed extracts o f cocoa beans in the 
2nd and 6th days of curing were recovered and designated 2)1 r^u^  
H(g)^ with Rf values 0,23 and 0,23 resp ective ly . Drawn chromato­
grams o f ■or.hjflrolycad and hydrolysed testa  extracts separated by TI£ 
are presented in  Figs 5a -  b and colour reactions shown in Tables 
1 1 s rad 1 1 f 0
From oharts o f Rf values and colour reactions, only tentative 
id cn tifiea tion s could be made since spectra l studies were not carried 
o u t .  There was better separation in the multi-solvent system
University of Ghana                              http://ugspace.ug.edu.gh
(Forestall 5($> Acetic! a cid5 BAW.jj saturated phenol; and water) than 
in BAV (n-butahol'-acetic acid-water, 4 J1 s5) alone (see Plato 1a and 
1 Id ), Separation of components was better on pa.per than on thin 
layer p lates, Rf values o f compounds ranged from 0,35 to 0,99«
There were more spots observed in unferasntad testa  erfcracts 
than in beans that had undergone some amount o f curing (fermenta­
tion  and drying) as shown in i l ' . t c  la , The number o f spots decreased 
with fermentation so that fermented beans in  the 6th day o f drying 
showed no observable spots, While in  the unfermented beans 4 spots 
were observed in separation methods involving the sing le -so lven t 
system, thera were 3 spots 011 the second day o f  curing, one on the 
sixth  day, and none on the sixth  day of drying (see Plivfcq 1a),
Spots 011 the chromatograms o f non-hydrolysed unfermented testa  
extracts were designated LU ,^ NTJ0, NIT^  and NTJ^  with increasing
Rf values. The corresponding Rf values fo r  the spots were 0 .35 , 
0„48? 0*60 and 0 ,70, respective ly , in  n-butancl~acetic acid-water 
(BA¥) and 0,58 , 0,-66 , 0,88 , 0,97 and 0,99 in  the multi-solvent 
system,-, Spot was pink in v is ib le  ligh t and brown in XN l ig h t . 
These colours suggested that HU,. could be an anthocyanin compound. 
However, the Rf value for  the standard marker was 0„46 and so 
spectra,! studies would be necessary to confirm the chemical nature
o f the'cyanidin compound. The Rf values o f HU_ and NSe, at 0,603 3
in n-butanol-acetic acid-water (BAW) were c lose  enough to  the 
standard marker (epicatechin) value o f 0,66 to  suggest that those 
spots were epicatechin derivatives.
University of Ghana                              http://ugspace.ug.edu.gh
wWhen testa  extracts weru hydrolysed fewer spots were observed.
While 4 spots were observed in non-hydrolysed unfermented beans, 
there were only 2 spots produced by hydrolysed unfermented beans 
when the chromatograms were developed in  n -butanol-acetic acid-water 
solvent system, Thore was a sim ilar trend with testa  extracts in 
la ter  stages of curing., The spots o f  hydrolysed unfermented beans 
were labelled  HTT^ , HU2, HU-j with Rf values o f 0.35, and 0.57
respectively  in  n-butanol-acetic g,cid-water (BAW) anal HU^HU3..w'dh Rf values o j  0-?5r
OAS  ^0 ,6 8  when developed in  the multi-solvent system. Wo anthoeyanin
was observed in  the chromatograms o f  extracts o f  testa  from beans
in the la ter  stages of curing. From the Rf values of 0,35 and
0,57 fo r  spots 171, and HU  ^ fo r  hydrolysed testa  extracts of
unfermented beans, i t  could be in ferred  that the aglyconc could
be malvidin which has an Rf value of 0 .5 8 -in n -butanol-acetic
acid-water (BLW), Spots HSi:  ^ o f  testa  extracts o f beans in the
sixth  day o f curing and spot HSe o f extracts o f beans in the
second day of curing possessed the same Rf value o f 0.56 in
n -butanol-acetic acid-water (M 'j) and 0 .44- in the multi-solvent
system and could be the same compound»
University of Ghana                              http://ugspace.ug.edu.gh
I f c
COLOUR REACTION OF CHROMATOGRAM SPOTS Of MOM-HYDROLYSED
COCOA BEAK TESTA BZPRACI DIFiYE LOPED ICT SmrGLB-£JOIff2!MP SYSTEM
DAY OF 
CURING
CHROMATOGRAM
SPOT
Colour P-©action
Rf
V is lb le
Light
Ultra V io le t L i^it [ D iazotisf
Sliort
xravel ongth
Long
wavelength;
lin e
IERMEMTATION
0- DAY M 1 0.35 V V DULL BR LY
Mg 0.48 LV DULL BR LY Y
mr*3 0,60 Cl, .VL Iffl CL LY
MJ.4 0.71 CL CL LY CL
2 DAYS NSe,1 0.35 LY V DULL BR
*ir
HSe2 0 .4 3 OL CL .’JJl LY
KSe,
3
0.60 CL ,?L Iffl CL Y
6 DAYS HSi1 0.69 CL BR LY
DRYING 
6 DAYS
(
-  —— -J
- - -
KEY
V = 
LV B 
CL =
V iolet'
Faint V iolet 
Colouj?! ess
Y
LY
BR
Yellow 
Pale Yellor 
Brotm
DTJLL BR = Dull B:?own
FL WH = Fluoresceiro 
If/hit e
University of Ghana                              http://ugspace.ug.edu.gh
r f
m m j h ,
OF NON-HIDROLTgSBD.
COCOA 1BAN TESTA ECTRACT DEVEIQEBD IN IirLgl-SOLVEiCT SYSTEM
DAY OF CEE0MAT0GR4M
Colour Reaction
CURING- 3I-0X Rf I Ultra V io le t  Light D ia zo ti"rd
V isib le
Light Shortwave- Longwave- length length
pwiiitroan in­
lin e
KSRMSUTAf-
TION
0  DAY NU 0 .5 8 V BR m LY
WTg 0,66 OY BR DULL BR Y
mj3 0.86 CL DEL FL BL-WH LY
nu4 0.97 LY FL BL FL BMH CL
ro5 0.99 CL LBL LY CL
2 DAYS *.kJ -
i
0,61 LY BR LY Y
NSe2 0,71 CL BL FL TO LY
NSa_ 0,85 CL Ft. BL FL BL-WH CL
WSe4 0,91 SL LBL m CL
6 DAYS r* ««• - **
DRYING
6 DAYS *sr ## axS
KEY
V = V iolet BR = Brown LBL = Light Blue
CL = Oolotcrless 
Y = Yellow
DULL BE x> Dull Brown FLBL
LY = Faint Yellow 
BL = Blue
FlT.:,-;-escent blue
FL BL—WH Fli’.o res cent 
bluish white
FL I© = F ;. 'esoent white
University of Ghana                              http://ugspace.ug.edu.gh
TABLE 11ic
COLOUR REACTIONS OF CHROMATOGRAM SPOTS ' “  HYDROLYSED
GOGOL BEjJJ TECTA EXTRACT DEVELOPED IN SINGLE-SOLVENT SYSTEM
CHROMA­
COLOUR REACTION
D-iY OF 
CURING
TOGRAM
SPOT
Rf VISIBLE Ultra V io le t  Light Diazotisod.p-nitroan±>
lin eLIGHT Short
wavelength
Long
wavelength
FERM3N- 
T AT ION
0 DAY HU1 0.35 CL FL BL FL WH LY
HOg 0.57 CI, LBL FL BL CL
2 DAYS HSo1
| 0’56
CL LBL LY CL
6 DAIS KSi1 0.56 CL LET, LI CL
DRYING 
6 DAIS i -t1*.)— .— ,—
- - - ~ -
KEY
G1 = Colourless 
LBL = Light Blue 
LY = Faint Yellow
FL BL = Fluorescent Blue 
FL TO = Fluorescent White
University of Ghana                              http://ugspace.ug.edu.gh
79
TABLE 11 d
COLOUR REACTION OF CHROMATOGRAM _SPOTS OF HYDROLYSED COCOA
BEAN TESTA EXTRACT BE i/ELOPED IN TIE MULTI - SOLVENT SYSTEM
DAY OF 
CURING
2HROMA-
TOGRi.M
SPOT
Rf
COLOUR REACTION
VISIBLE
LIGHT
Ultra V io le t  ligh t D ia zctis ’
p-m tros;
lin e
. . . .
Short L'mg 
wavelength wavelength
FERMEN­
TATION
0 DAY HU.1 0.25 CL FL BL FL WH Lr
HU2 0.45 CL LBL FL BL
eu3 0.68 CL f l  m  f  a r a T---,u„
2 DAYS HSe1 0.44 CL LBL II, IL r ,
6 DAYS KSi 1 0,44 CL LBL i  h  BL CL
DRYING
-nr -r
6 DAYS . — —
— ------------ ----j-------------_ _ _
CL “  Colcurles3 FL BL = Fluorescent Blue
LY -  Faint Yellow FL WH = Fluoroscent White
LBL a Light BIuo
University of Ghana                              http://ugspace.ug.edu.gh
TABLE l ie
COLOUR REACTION OF SPOTS OH THIN LAYER PLATES OF NON- 
HYDROLYSED COCOA BEAN TESTA EXTRACT DEVELOPED IN 
CHLOROFORM-ACETIC ACID-WATER
' — *“ 1
DAY OF 
CURING
THIN-LAYER
SPOT Rf
COLOUR REACTION
VISIBLE
LIGHT
Ultra V io le t Light D iazotised
Shori Long 
wavelength wavelength
' v rOaZlx"
lin e
FERMEN­
TATION
0 DAY N(e)1 0.71 V LY DUIiLBR Y
2 DAYS N(2 )1 0.68 LY DULLBR BR LY
N(2)2 0.90 CL FL BL FL WH OR
6 DAYS - - - -
DRYING
6 DAYS - - «■ -  “
KEY
V = V iolet BR = Brown
Y = Yellow DULLBR = Dull Brown
LY = Pale Yellow FL BL = Fluorescent Blue
CCi = Colourless OR = Orange
University of Ghana                              http://ugspace.ug.edu.gh
81
COLOUR REACTION OF SPOTS OH THIN LAYER PLATES OF HYDROLYSED 
COCOA BEAN TESTA EXTRACT DEVELOPED IN CHLOROFORM-ACET IC AC ID—¥ATER
TABLE 11f
KEY
CL = Colourless
LY = Pale Yellow
FLBL = Fluorescent Blue
University of Ghana                              http://ugspace.ug.edu.gh
82.
< * )
0  n
a  °
O  o
O  < 3
0
CJFFE(& 0  i  B ’lCATTSCWlN fc |o C^WHH
ACIJ) CHtwyML
PLATE 1a; Photograph o f chroiaatogran o f phenol:.?- ocnp-'ro-its of 
nonhydrolysed (cocoa bean) testa  extract on 
days o f curing, developed in s ing le-so lven t r.ystenu
University of Ghana                              http://ugspace.ug.edu.gh
83
CYAKIJilM o ti. ENCtffeCHW t &
CHlfiRiDE *  *
PLATE 1b; Photograph o f chromatogram o f phenolic components o f 
nonhydrolysed (cocoa bean) testa  extract on d ifferen t 
days o f curing, developed in multi-solvent system.
University of Ghana                              http://ugspace.ug.edu.gh
12. Qix'nimnCHUW.I6E
Photograph of chromatogram of phenolic components o f 
hydrolysed (cocoa bean) testa  extract on d ifferen t 
days o f curing, developed in single-solvent system,,
University of Ghana                              http://ugspace.ug.edu.gh
85
PIAgE 1d ; Photograph o f chromatogram o f  phenolic components o f
hydrolysed (cocoa bean) testa  extract on d iffe ren t days 
o f curing, developed in  multi-solvent system.
University of Ghana                              http://ugspace.ug.edu.gh
00
0
F I G .  5 a  DRAW ING  OF  THIN LAYER PLATE OF  PHENOLIC C 
N EN T S  OF  NON - H Y DROL YS ED  C O C O A  BEAN  TESTA  ON  DIF 
DAYS OF  CURING,'
University of Ghana                              http://ugspace.ug.edu.gh
mFIG. 5b DRAWINGS OF T H I N - L A Y E R  PLATE OF PHENOLIC  C OM PO ­
NENTS  OF H YDRO L Y S ED  COCOA  BEAN  T ES TA  OF DIFFERENT  DAYS 
OF CURI  NG .
University of Ghana                              http://ugspace.ug.edu.gh
G. EFFECT OF TESTA COMPOUNDS OF ELUTED SPOTS OF TLC PLATES ON 
HPORFl BBRMTNATIOM IN SELECTED COCOA BAN CONTAMINAHT FOTSI
1» Germination in "flood, water*1 o f Potato Dextrose Agar
The potential fo r  in fect ion  of a pathogen depends on fa ctors  
including the inoculum s iz e  and the germination characteristics  o f 
the spore. Spores of fungi which germinate readily  in d is t i l le d  
water have su ff ic ie n t  nutrient reserve to  support the in it ia l  germi­
nation process* There are e co log ica l advantages i f  the endogenous 
nutrients occur in  su ff ic ie n t  quantities to support in it ia l  growth 
of the germ-tubes„ I t  is  a lso  advantageous to  have an in -bu ilt  
system of discontinuous germination so that part of the inoculum 
can survive adverse conditions that may be letha l to the germinated 
spores<
Manners (1g66) suggested that in  assessing spore germination 
the measurement of three parameters best re fle cted  the gerr~jiation 
potential of fungal spores. These parameters are novr y-xry widely 
used in germination te s t s . The parameters are (a) the percentage 
to ta l spore germination achieved within a certain  time period..
(b) the speed of germination or latent period o f germination ( i e ,  
the time taken to produce 50 per cent spore jrermination) and (c )  
the amount o f germ-tube growth or the y ie ld  which i s  a re fle ct ion  
of the amount o f endogenous nutrient reserves*
University of Ghana                              http://ugspace.ug.edu.gh
In th is investigation , percentage germination and germ-tube growth 
were studied in conidia of the four Aspergillus species used in 
previous experiments and the s c le ro t ia  of Sclerotium r o l f s i i  using 
the methods o f Manners (1966) and Dix (1972), Conidia o f  Aspergillus spp- 
do not normally germinate or do so very poorly in water (L il ly  and 
Bamettj 1951). It would therefore be d i f f i c u l t  to  evaluate the 
e ffe c t  o f inh ib itory  compounds in subsequent studies using d is t i l le d  
water as con tro l, I'hus in it ia l  germination te st  was carried out
with a d ilu te nutrient medium. A fresh  uninoculated PDA plate was
r
flooded with 10 ml o f s te r i le  d is t i l le d  water and allowed to  stand 
fo r  f iv e  minutes. The water was poured o f f  and used in  the prepa­
ration  of spore suspension. It was considered necessary to  use 
this d ilu te nutrient medium which as the results in Table 12- show, 
xras stimulatory enough to support germination, A very high nutrient 
concentration would overshadow the influence o f weak inh ibitory  
compounds or unduly reduce the e ffe c ts  o f strong inh ib itory  compounds, 
Conidia and sc le ro tia  were inoculated at 25° + 1°C»
ji .  fumigatus germinated best among the Aspergillus species, 
with the remaining three showing percentage germination between 27 
and 33 perroent; in  contrast to 71 per cent by A. fumigatus« Forty- 
one per cent o f the sc le ro tia  o f J3. r o l f s i i  germinated in this 
medium, ffierm tube growth was related to percentage germination.
Among the Aspergillus species, A. fumigatus which germinated best 
had the longest germ tubes (mean o f 7 ,110  u^m).
University of Ghana                              http://ugspace.ug.edu.gh
Was positive corre lation  between f in a l percentage 
germination and germ-tube length at r  = 0,20065 and P /  0»05.
University of Ghana                              http://ugspace.ug.edu.gh
91
TABLE jjg
GERMINATION OP CONIDIA OP ASPERGTLLUS SIECIES AMD SCLROTIA 
OP SCLEROTIUM ROLFSII IN EXUDATE OF PDA IN 50 HOURS INCUBATION
AT 25°C
Fungus Species Percentage Germination Mean Germ Tube length (tun + S.D .)
A, flavus 27 3,125 + 10.5
A. fumieatus 71 7 ,1 10+  7 .8
A. niger 33 4,120 + 14.6
A. tamarii 30 1 ,950+  9.9
S. r o l f s i i 41 10,250 + 10.1
University of Ghana                              http://ugspace.ug.edu.gh
92
2„ noT-m-ir'a-hi nn of conidia and, s c le ro t ia  in TLC eluted, spots
Although phenolic compounds have been associated with disease 
resistance (eg , Hawley, Fleck and Richards, 1924), th e ir  in fluence 
on spore germination has not been extensively studied. Experiments 
were, therefore, carried out using A, flavus. A, fumiaatus. A. niger 
A. tamarii and r o l f s i i  to  provide pertinent information. The 
spores o f A spergllte /and the s c le ro t ia  o f IS, r o l f s i i  were germinated 
in solu tion  o f  p re -iden tified  spots on Thin Layer chromatographic 
p lates in  PDA "flood -w ator"« Solution o f eluted portions o f the 
s i l ic a  gel support without any compounds was used as a control,,
The samples tested were obtained from the follow ing  f iv e  chromato- 
grain spots:
(a) F irst spot o f unhydrolysed testa  o f unfermented cocoa bean
N(o ) l ‘
(b) F irst and second spots o f  unhydrolysed testa  o f beans 
fermented fo r  2 days ®(2) l  831(1 ^(2)2®
(c ) F irst spot o f hydrolysed extracts o f beans fermented fo r  2 
days
(d) F irst spot o f hydrolysed testa  extracts o f beans fermented 
fo r  6 days Results o f these investigations and data
cn control te s ts  carried out with solutions o f the s i l i c a  gel 
support are given in Tables 13A -  B.
University of Ghana                              http://ugspace.ug.edu.gh
The resu lts show that both percentage gem ination  and growth 
o f  germ tubes were a ffected  by the compounds o f the spots o f the 
thin layer plates,, The resu lts are, b r ie f ly , as fo llow s: 
i* Compounds eluted from ^ 0 )1 31111 ®(2)1
depressed percentage germination of the species , 
i i .  With the exception o f  A, n iger and J3. r o l f s i i  the germ tubes 
grew better in compounds o f than in  the control (Table
ISA). The depression in germ tube growth o f  those two was, 
however, very s ligh t ; 
i i i o  S im ilarly, germ tube growth was varyngly a ffected  by compounds 
eluted from ^ (2)i (^able 13-B). Growth o f germ tubes of 
A, flavu s . A. niger and jB, r o l f s i i  was b e tte r , and that o f 
i i ’ fumigatus rjid jfc, tamar i i  was worse in  compounds eluted from 
II( 2)j than in  the control medium,
iv . Compounds eluted from 1^2)2* H(2 )l £ai<i H(6)1 s 'fcin'a--a'te<i germi~ 
nation in some species and depressed germination in others 
(Tables 13C -  E)„
v . Percentage germination o f A, n iger was consisten tly  stimulated 
by a l l  the three compounds,
v i .  Other instances o f stimulation were, higher germination o f
A. flavus. A. fumigatus and A, tamarii in compounds eluted from 
^(2)1 (TalDle 13C), and S_, r o l f s i i  in  compounds eluted from ^ (2)1 
(Table 13D) and (Table 1SE).
v ix . Germ tubes generally grow better in  media which stimulated 
high percentage germination in the test medium and poorer in
University of Ghana                              http://ugspace.ug.edu.gh
94
the inh ib itory  media. The exceptions were A, flavus in 
compounds eluted from (Table 13'C) and (Table 13E)
and A. fumigatus in compounds eluted from (Table 1?B),
University of Ghana                              http://ugspace.ug.edu.gh
Germination o f conidia o f  Aspergillus species and sc le ro t ia  of 
Sc r o l f s i i  at 25 °C for 50 hours in  solutions o f compounds eluted
from TLC spots l ' ( o )  of non-hydrolysed extracts o f testa  of
1
unfermented cocoa beans®
TABLE ,15A
Fungus
Species
Germination in solution  
o f compounds o f eluted 
spots
i . . .  . . .
Germination o f solu­
tion  o f s i l i c a  gel 
eluted spots
Percentage
Germination
Moan germ 
tube length 
(jm )
Percentage
Germination
Mean germ 
tube length 
(fm )
A. flavus 18 140 35 60
A. fumigatus 18 525 27 450
A. niger 39 200 49 250
A, tamarii 20 480 43 400
S. r o l f s i i 6*5 930 14 950
University of Ghana                              http://ugspace.ug.edu.gh
TABLE ,1 IB
Germination o f conidia o f .^spergilius species arid s c le ro t ia  of 
.§• r o l f s i i  at 25°C fo r  50 hours in solution  o f conpounds eluted 
from TLC spot N(2)^ o f non-hydrolysed extract o f testa  o f cocoa 
fermented fo r  2 days,
Fungus
................................. ...............f
Gemination in  solution  
o f compounds o f eluted 
spots
Germination in  solu tion  
o f s i lica !, gel eluted 
spots
Species
Percentage Mean germ 
Germination tube length 
(jm )
Percentage Mean germ 
Germination tube length 
(jm )
A. flavus 27.5 475 32.5 225
A. fumigatus 20.0 300 27.0 600
A. niger 35.0 225 52.0 175
A. tamarii 24.0 350 47.0 500
S. r o l f s i i 8.0 1980 12.0 1950
University of Ghana                              http://ugspace.ug.edu.gh
Germination o f  conidia o f Asp erg illu s species and s c le ro t ia  o f  
r o l f s i i  at 25°C for  50 hours in solution  of compounds eluted 
from TLC spot N(2)2 of non-hydrolysed extract of testa  of cocoa 
beans fermented for  2 days.
• '9'7
TABLE mC
Fungus
Species
Germination in solution  
o f compounds of eluted 
spots
Germination in solu tion  o f 
s i l i c a  gel eluted spots
Percentage
Germination
Mean germ 
tube length 
(pm)
Percentage
Germination
Mean germ 
tube length
(p i)
A, flavus 31 180 29 850
A. fumigatus 31 620 27 480
A, niger 23 160 21 180
A. tamarii 52 500 37 450
S. r o l f s i i 5 400 13 950
University of Ghana                              http://ugspace.ug.edu.gh
Germination of conidia o f  Asperg i llu s apscios and sc le ro tia  o f 
S . r o l fs i i .  at 25°0 fo r  50 hours in solution  o f compounds eluted 
from TLC spot H(2)^ o f Hydrolysed extract o f testa  o f cocoa 
bean fermented fo r  2 days.
9C
TABLE 1&D
Fungus
Species
Germination 
o f compounds 
spots
in solution  
o f eluted
Gemination in  solu tion  
o f s i l i c a  gel eluted 
spots
Perc entago 
Germination
Mean germ 
tube length 
(jm )
Percentage
Germination
Mean germ 
tube length 
(pm)
■y.. flavus 9 250 38 800
A. fumigatus 1 150 13 280
A, . niger 33 250 17 140
k t  tamarii 1 100 22 250
S* I 'o l f s i i 15 340 5 60
University of Ghana                              http://ugspace.ug.edu.gh
99'
Germination o f conidia o f A spergillus species and sclerotia . of 
S, r o l f s i i  at 25°C fo r  50 hours in  solution  o f compounds eluted 
from TLC spot H(6)^ o f  Hydrolysed extract of te sta  o f  cocoa bean 
fermented fo r  6 days.
TABLE 15E
Fungus
Species
I ..................
Germination 
o f compounds 
spots
in solution  
of eluted
Germination in solution  o f
s i l i c a  ge l elutod
spots
Percentage
Germination
Mean germ 
tube length 
(pm)
Percentage
Germination
Mean germ 
tube length 
(fim)
A. flavus 26 980 39 770
A. fumigatus 2*5 60 16 50
A. niger 35 410 17 140
A. tamarii 0 0 20 350
S, r o l f s i i 34 275 5 120
University of Ghana                              http://ugspace.ug.edu.gh
Fungi have been observed in coooa beans at d iffe ren t stages 
of curing and storage. Dade (1929), Laycock (1931)» de Witt (1961)? 
Roelofsen (195S)~ Broadbent and Oyeniran (1967) and Wood (1967? 1979) 
have reported that fungi invade the beans at d ifferen t stages o f the 
fermentation process.
The h istory  o f the beans could sometimes influence their  
su scep tib ility  to  contaminant fungi, Dade (1927) demonstrated how 
pod diseases such as, blackpod caused by Phytophthora palmivora 
and mealypod caused by Traohysphaera fructigena. predisposed the 
beans to fungal invasion during fermentation and drying. He 
obtained no contamination when he fermented and dried beans from 
100 healthy cocoa pods, but observed 30 per cent contamination 
when ho fermented a mixture o f  beans removed from 50 healthy pods 
and 50 diseased pods. The pod pathogens might have caused s u f f i ­
cien t damage to the testa  to render i t  more penetrable by other 
fungi.
Very few studies have been devoted to the path o f entry of 
the contaminant fung i, Laycock (1931) showed that 76 per cent o f 
fungal contamination o f the beans occurred through tho micropylar 
end, and tho rest through other points.
In th is  investigation , the occurrence o f fungi in cocoa beans 
at different, stages o f fermentation and drying and the d istribu tion
\ r r -
H. FUNGAL INVASION OF COTYLEDON OF COCOA BEANS DURING C £ 0 5
University of Ghana                              http://ugspace.ug.edu.gh
o f the fungi in the cotyledons were studied. F ifty  beans were 
removed from the batch on each sampling date, they were surface- 
s te r ilis ed  and a scep tica lly  cut transversely into 5 equal sections.
The sections were incubated on moist s te r i le  f i l t e r  paper at ' j  
25° + 2°C fo r  10 days, At the end o f the incubation period, the 
fungi growing on the sections were studied and identified ,, The 
fungi id en tified  and their  d istribu tion  in  the cotyledon are tabu­
lated  in Tables '4  and 15„
The fungal population increased with increase in  tine of 
curing. Regions of the bean nearest to  the nicropyle showed 
evidence o f greater fungal invasion than other parts. Thus sections 
4 and 5 contained greater to ta l number of fungi than section  1, 2 
and 3 , Aspergillus was by fa r  the dominant genus among the contaminants.
■101
University of Ghana                              http://ugspace.ug.edu.gh
Occurrence of fungi in beans o f  cocoa during curing
TABIH 14-
i
Days of 
Curing
Percentage fungal invasion o f bean sections
1
(Stylar
end)
2 3 4 5
(micropyjar
end)
5SRMEMATIDI
0 4 0 0 0 4
2 16 8 12 16 0
6 16 0 32 30 48
DRYING
6 12 68 0 80 90
University of Ghana                              http://ugspace.ug.edu.gh
105
Distribution  o f  contaminant fungi in cotyledon o f cocoa beans 
at varying stages of curing process.
TABLE 15
Days of Fungus Species in  bean Sections
Curing 1 2 3 4 5
Fermentation
0 Aspergillus
sp.
Fusarium
d ec emc e l lu l  ar e
2 Aspergillus Aspergillus Aspergillus 
restrictu s  sp fumigatus
Mucor sp.
Mucor sp.
6 restr ictu s  Hue or sp„
A„ fumigatus
Mucor pusillus A0fumigatus
Rhizoms A. fjmigatr," 
arrhizus
A. fumigatus
Drying
6 A, cheva lieri Mucor sp, 
Penicillium  s p .
SP»
A,, flavus Mucor sp,
R. arrhizus Aspergillus 
sp„
University of Ghana                              http://ugspace.ug.edu.gh
V i  GENERAL DISCUSSION
A gpod chocolate aroma is  o f  fundamental importance in  the 
production o f chocolate and chocolate related  products* Anything 
which adversely a ffe c ts  th is property is  undesirable; Fungal conta­
mination or internal mouldiness is  the major cause of unpleasant 
odours and contaminations as low as 4 per cent could be detected in  
the taste of the chocolate products. Surprisingly, apart from the 
works of Dade (1929) and Bunting (1929), no record o f the leve l o f 
fungal contamination in  Ghana cocoa i s  available even though fungal 
contamination o f cocoa beans provides the main basis  fo r  grading o f 
commercial cocoa beans.
The main aim of th is study was to re-examine the mycoflora o f 
cocoa beans* 50 years a fter  the reports of Dade (o p .c it )  and Bunting 
(o p .c i t ) .  The percentage o f cocoa beans at the port o f Tema conta~ 
minated by fungi was investigated fo r  one cocoa season. October 1978 
to July 1979c The average/4 .9 per cent contamination recorded during 
this period f e l l  within the world average range o f 0 -  6 per cent 
(Dlectanan,, 1962) but considerably higher than the 1.6 per cent o f 
Nigerian cocoa (Oyeniran and Adon iji, 1974)* and s ligh tly  greater 
than 3*1 per cent recorded fo r  Ghana cocoa arriving at the United 
Kingdom ports in 1960 (passmore, 1932)» Seasonal breakdown showed 
that the main crop cocoa contained an average 3-9 per cent (2 ,3  -  
6.0 per cent) and the minor crop 6 .4  per cent (4 .7  -  8»0 per cent) 
fungal contamination (see Table 1 and Fig, l ) „
104
University of Ghana                              http://ugspace.ug.edu.gh
Tho averago moisture cpntcht was 6,5 per cent. This was 
lower than the minimum 7^5 per cent permissible in  mer®hantable 
cocoa beans as recommended by the Working Party on Cocoa Gradings 
Percentage contamination was d ire c t ly  related to  cocoa bean moisture 
content (see Table l )§  Fungal contamination was higher during the 
rainy season and lower during the drier months o f December to  
February.
Dade (1929) and Bunting (1929) iso la ted  and id en tified  fiv e  
fungi belonging to the ^ ^ r g i l l u s  groupf while Oyeniran and 
A don iji (1974)- working at the Stored Products Research In stitu te  
in Nigeria la ter  found th irty-one spscies as occurring on Nigerian 
cocoa. Results o f the present study showed that twenty-one fungal 
species belonging to nine genera rGEe capable o f in fect in g  cocos, 
beans in  Ghana. The l i s t  of the fungi observed and th e ir  frequency 
of occurrence is  given in  Table 2„ S ign ifican tly , three fungus 
species Aspergillua  phconlcis . fthizopuq o:«rsgs and Neurospora erassa 
wMch had not been previously isolated  anywhere on cocoa, were 
found* The resu lts Confirmed the suggestion by Reese and Downing 
(1951) that with tho exception o f .^ j^ r ^ l lu s  jEiieer var^ luchuensis  
no black A sp erg illi in fested  cocoa beanst and the observation by 
Bimting (1929) that the Aspergillus ^laucus group (A, ch ev a lic r i,
4t* anc* A* resty ictu s) forms the most important contaminant
o f commercial cocoa beans.
■ \  1 5
University of Ghana                              http://ugspace.ug.edu.gh
A sp e rg illii  constituted the dominant mycoflora o f stored cocoa 
beans in both th is study and in  the work by Oyeniran and Adeniji 
(1974) on Nigerian cocoa. Consistently, Aspergillus chcva l ie r i  was 
the most frequently observe;! spccies in both ca3es0 Also o f parti­
cular in terest is  the occurrence o f  A, f lavus (5»3 per cent) in 
th is study and as a major contaminant in Nigeria although Bunting 
(1929) in  Ghana and Maravalhas (1966) in Brazil did not find th is 
species in th e ir  stud ies. In the reverse, Asperg illus, aculeatua 
which was found in the beans by Bunting (1931) did not occur on 
beans examined in th is study* I t  is  not surprising that d ifferences 
ex ist between the cocoa ’yean mycoflora o f Ghana o:id N igeria. 
PaecSllomycSes v a r io t i and Syncephalastrun raccmosum which wore 
major contaminants o f Nigeria cocoa as well as the minor species 
Absid ia  oorvnb ifera . A3p.3rfip.l lus .pjso^qglaucus, Cajvnlariax lunataT 
Cvlindrooarpon tonklnense, Fusariim oxvsporiura, Gcptrichum candiduia, 
and Macrophoma species were absent in  Ghana cocoa. On the other 
hand, Nigerian cocoa bean were free  from orassa, Rhizopus
or^zae, Aspergillus ustus, A, sydowij, and A, w entii occurring in 
the beans in  Ghanas
Notwithstanding the assurance by Wood (1979) that no a fla to - 
xins had been detected in chocolate and chocolato products, the 
iso la tion  o f A„ f la vus (5 .3  per cent) in mouldy cocoa beans made i t  
necessary that further study of potential mycotoxin producing fungi 
and the characteristics o f  fungi associated with contaminated cocoa
i 'j  c
University of Ghana                              http://ugspace.ug.edu.gh
be more thoroughly investigated| ?-rhaps the absentee o f a fla tozin
> , ’ ' ' I
in  cocoa iq due to the unsu itab ility  o f cocoa bean as substrate 
fo r  a fla tox in  synthesisk
Temperatures as high as 51°C were reached on the th ird day 
o f  fermentation, Such high temperatures could k i l l  most fungi 
except th f thermophilic and th®rmotplej?ant species such as
rqragatus and .Mucor pusiUas * Any frogus which entered th<j bean
prior  to  fom entation  would thus be k il le d  during the process of 
fermentation so that most contaminants might have invaded the beans 
at the end of fernentatioa, ‘ r:. > drying and in  gtorago* luring
invasion* c s l lu lo ly t ic  species notably As f-tnigat--;;. Boti^ .odiBlodia, 
theobronao and Xucor ju s i^ lu s  would be expected to play the r o le  o f  
pioneers creating access ways fo r  tho entry o f non-.-aellulolytic 
species,} Forsyth (1954) reported that cocoa beans with, acceptable 
aroma could be obtained a fter  only two days fermentation# I f  indeed 
most in fect ion  takes place a fter  fermentation,, the shortening o f the 
fermentation process should have no influence on fungal a c t iv ity  in 
the beans in storage-, Shortening of the fermentation period should 
be accompanied by measures to  control fungal in fect ion  during storage 
and shipping o f the benns5
?he Results o f previous experiments s^ved  that the 
species fort'.£d 64e,3 pei? ©ent of fungi contaminating stored cocoa 
bean, Vive A s p jr g i^ ^  species (4 * f^ v u s , A* fun^atus, A, n jg e^
University of Ghana                              http://ugspace.ug.edu.gh
108
A, iiidulans. and «.» tqgagi^) iso lated  in th is study from the conta­
minated beans were use', fo r  further studies, in comparison with 
Sclerotium r o l f s i i . a so il-in h ab itin g  fa cu lta tive  parasite which is  
not n*.t:.vn to  cocoa beans.
The i is p o rg illi  showed high mycelium production in testa  
extracts o f  unfermented cocoa boans. The greatest amount o f growth 
was recorded in A. tamariS, and A. fumigatus cultures which produced 
69*2 mg and 64.2 mg dry weight, resp ective ly , a fte r  7 days of 
incubation (see Table 3b and 3o)„ The unfermented testa  extract 
which contained much pulp is  known to  be high in carbohydrates 
especia lly  the manosos (Forsyth and Quesnel, 1953;. £3. .r o lfs i^
produced the least amount of mycelium o f  43.2 mg dry weight (see 
Table 3 f ) .  Extracts from 2- and 6-days cured beans depressed growth 
in  a l l  the fungi tested* The AsiDergiUus species were more inhibited, 
than S, r o l f s i j  by the ox ’;racts of 2-day citrod beans,. On the other 
hand, £S. r o l f s i i  was most severely inhibited by the extracts of 
6-day cured beans (see Table 3f and Fig* 2 ), The respective percen­
tage reduction in A. flavus , A, fumigatus,, A. nidulans, A. n iser,
—• tamarii  and 3 . ro.1',,. i i  was 33 .4 , 63o6f 31 °9, 39=8, 57.1 and 65*1 
per cent. A lcohol, which is  a major component o f pulp at the begin­
ning of fermentation as yeast a c t iv ity  increases, would bo a sign i­
fican t constituent in the testa  extract o f cocoa beans (Eoolofson, 
1958). The usual end-products o f fermentation, acetic acid and
University of Ghana                              http://ugspace.ug.edu.gh
l e g
la o t ic  acid  might he contributory fa ctors to^ ie  observed depression 
pf gro'.'th. Testa extracts of cocoa beans dried fo r  2 and 6 days 
a fter  fermentation supported almost the same amounts o f groxrth as 
the extracts of the unfermonted beans. Growth w&S stimulated in 
S,* r o lfsS; in extracts o f beans dried fo r  2 days. Perhaps , oxidation 
o f  acetic acid , accumulated in the fermented beans, to  carbon 
dioxide and water markedly reduced the leve l o f acid content o f the 
testa  during drying*
The rate o f growth in the cxtract media should be related to  
the in tr in s ic  growth habit of the fungi. Dry matter produced by 
the various test fungi in  Czapek Dox broth medium showed that the 
fungi varied in th e ir  growth habits. A, fumigatus grew fa stes t, 
attain ing maximum growth on or about the third day o f .incubation 
in Czapek Dox broth mediui, at 25°C9 In comparison i t  took 7 days 
in Ac tamarii» A, niger and 3 , r o l f s i i  (see Table 4 and P ig, 3)„
A. nidulans grew the slowest in Czapok Dox broth medium attaining 
maximum dry weight on the 14th day o f incubation* I t  could be 
in ferred that the amount o f growth obtained in the previous studies 
using extract media re fle cted  maximum growth in  A* n iger. A, tarnnrii.■ 
and £>* r o l f s i i  and not in  the remaining species.
On Sabouraucfe broth medium. A, n ige r , and A, tamarii grew less  
well* At 25 C A, niger, produced 129„3 mg dry Hatter on Sabouraud?P 
broth compared with 279*2 mg dry weight a fter  7 days incubation in
University of Ghana                              http://ugspace.ug.edu.gh
Czapek Dox broth. Correspondingly, A, tamarii produced 124*0 mg 
dry weight o f mycelium in Sabouraudk broth and 287 mg in Czapek
Dox broth! A| fjamigatus  ^ grow better in Sabourauds broth (230,2 mg
dry weight) than in Czapek Dox broth (l99e5 mg dry weight)* The 
poorest growth o f 4» nidulans among the species tested in Czapek 
Dox broth was confirmed by growth te sts  in Sabourauds agar medium 
at 25°C and 30°C (See Table 4 and 6, and Pig, 3 ) . This should not 
be surprising since the cultures were harvested as long as 7 days 
before reaching maximum growth, _S, r o l f s i i  which was next to  
A* nidulans and A. niger in  productivity in Czapek Dox broth medium,
grew better than both in Sabourauds broth.„
Hall (1975) reported the presence o f a germination inh ib itor 
in the testa* and beans with testa  removed germinated better than 
those with in tact testa^ He, however, did not characterise the 
chemical involved, It i s  possible that phenolic compounds present 
in  the testa  might be involved since these compounds have been 
id en tified  as fungal growth inh ib itors in otEher studios (^jrkham, 1957? 
Parkas and Kiraly, 1962  ^ Kuc, 1964f- Kosuge, 1969)* In the present 
study, i t  was observed that there was a s ligh t in it ia l  decrease in 
tho to ta l phenol concentration in the testa of cocoa bean with 
time during fermentation from 4.4  mg per gram to  3,1 tag per gram 
but rose thereafter during drying of the beans, Forsyth and 
Romboi4ft. (1951) observed a general decrease in the phenolic
University of Ghana                              http://ugspace.ug.edu.gh
compounds o f the cotyledons during curing a$d suggested the f a l l  
fo  be either due to  theii? destruction by enzymes or exudation into 
the te sta j Results o f th is  investigation  tended to support the 
la t te r  view,
The concentration o f anthocyanin compounds in the testa  
increased sharply as soon as fermentation started from 0*073 mg 
per gram in  the unfermentcd bean testa  to 0*074 mg per gram,
0*15 mg per f^ ram, and 0*24 mg per gram  ^ in  the 2nd fth  and 12th 
day extracts resp ective ly , Kenten (1965)* however* reported that 
in the case of Amelonado cocoa beans* the anthocyanin content 
decreased with time in  storage* I t  is  suggested that apart from 
the in it ia l  amount of anthooyanin compounds in the testa , more 
leached from the cotyledons into the testa .
Soluble phenols o f the testa  increased in it ia l ly  from 0.41 
mg. per gram to 0*45 mg; per gram in  the f i r s t  2 days of curing. 
Iiater the concentration f e l l  to 0,32 mg per gram on the 6th day 
and then 0#22 mgr per gram on the 12th day o f curing# Q iffe r i  
(1931) suggested that s in ce  there was a large amount of oxidases 
present in  the cocoa bean* an oxidation process might lead to  the 
degradation o f  phenolic compounds resu lting  in  their  breakdown* 
Forsyth and Rombou&s (1951), however, observed that by a process o f 
condensation, the simpler soluble polyphenols were destroyed and 
insoluble tannins f o -u d .
University of Ghana                              http://ugspace.ug.edu.gh
The orthodihydric phenols and the flavonols followed a 
sim ilar pattern of an in it ia l  decrease in  the amounts o f these 
constituents at the beginning o f fermentation and an increase during 
the la ter part o f curing (see Table 7 and Fig, 4)»  The unfermented 
bean testa  contained 2.43 mg per gram of orthodihydric phenols.
On the second day  ^ the concentration f e l l  to  2 * T9 mg per gram o f 
dry testa  t is su e f Forsyth and Rombouts (1951) suggested that th is 
was e ither due to leaching or polymerisation of simpler leucocyani- 
dins to  more complex leucocyanidins. S im ilarly, flavonol compounds 
in the testa  f e l l  in  amount with fermentation from 0.51 mg per gram 
in the unfermented bean to 0.33 mg per gram in two days o f fermen­
ta tion . Concentrations in the testa  bu ilt  up to  0,43 mg . per gram 
at the end of fermentation. Again during drying, there was a f a l l  
in the amount o f flavonols to  0„37 mg per gram of testa  tissu e .
I t  was suspected that there would be a relationsh ip  between 
these changes and the chemical fa ctors involved in. resistance to  
fungal invasion o f plants and plant products<> Further study of the 
chemical constituents on paper and thin layer chromatograms was 
therefore done* The paper chromatograms revealed more spots than 
thin layer plates» On the paper chromatograms f iv e  spots were 
observed fo r  unhydrolysed unfermented testa  extracts while only on© 
developed on this layer plates (sse Pig. 5a and 6a ) . Separation was 
better in the mult insolvent system of Forestal, 50$ a cetic  acid,.
University of Ghana                              http://ugspace.ug.edu.gh
n-butanol-acstic acid-water* saturated phenol and water, than in 
the chromatograms developed in n-butanol-acetic acid-water (4 :1 :5 ) 
alone (see P ig,5a and Figi 5b ), When developed in the multi-solvent 
system^ the unfermented non-hydrolysed testa  extracts revealed f iv e  
spots as compared to  three in  the hydrolysed extracts (see Pig* 5b 
and 5d)i Roberts (1957) explained that hydrolysis resulted in 
decarboxylation o f phenolic compounds into the basic aglycones and 
the methyl and carboxyl groups. Also, non-hydrolysed testa  of 
beans in 2 days of fermentation had three spots when developed in  
ii-butanol-acetic acid-water while the hydrolysed extracts had one 
spot. Spots o f unfermented non-hydrolysed testa  extracts designated 
NTJ.J, UU2, NU ,^ NU4 and had Rf values o f 0 ,58, 0 .66, 0 ,88, 0.97 
and 0,99 respectively  in the multi-solvent system. Spot KU^  was 
pink in v is ib le  ligh t  and brjwn in  the short wavelength band o f TJV 
ligh t  and could be ia e r t if ie d  as an anthocyanin compound. Forsyth 
( 1952b) had id en tified  three anthocyanins in  Forestero cocoa : 
cyanidin monoglucose, cyanidin arabinocyanins and cyanidin d ig lyco­
s id e . Two o f these, cyanidin arabinoc yninc and cyanidin d ig lycoside 
were la ter  confirmed by Kenten (1965)•
Hydrolysed extracts o f unfermented beans developed three 
spots OT1, HU2, HU with Rf values o f 0*25, 0,43 and 0.68 respectively  
in the multi-solvent system. In the n -butanol-acetic acid-water 
solvent system., HU,, and HU2 had Rf values o f 0.35 and 0 .57 .
University of Ghana                              http://ugspace.ug.edu.gh
The Rf value of HUg was regarded c lose  enough to  the Rf value o f  
0,58 fo r  malvidin in ii-bu tanol-aceiic §.cid~water to  suggest that 
HUg could he a malvidin g lycoside . Results of these investigations 
were not conclusive since spectra l studies were not done and fu lle r  
investigation  is  n~fkssary in any future relevant stud ies.
Spots of phenolic compounds on the thin layer plates were 
eluted anil used in  spore germination te s ts . Since conidia. of the 
A sp erg illi  do not normally germinate in or do so poorly in d is t i l le d  
water (L il ly  and Barnettj 1951), i t  would he d i f f i c u lt  to  iden tify  
inh ib ition  by the solutions i f  d is t i l le d  water was used as con tro l,
A preliminary gem ination test  of the fungi under study using a 
d ilu te  nutrient medium o f  Potato Dextrose Agar "flood-water" was 
found stimulatory enough to  support germination (see Table 14),
A, fumigatus germinated best producing 71 per cent germination 
a fter  50 hours^incubation at 25°Ce A, nig e r . A, tamarii and A, flavus 
wore next with 33* 30 and 27 percen t germination respectively  
(see Table 12), Forty-one per cent o f the s c le ro t ia  o f S. r o l f s i i  
germinated in the medium.
As expected, the s c le ro t ia  had the greatest amount of 
nutrient reserves and S.. ro l f s i i  had the longest hyphae (mean 
10,250 jum). Among the Aspergi l lu s  species , A. fumiaatus which 
germinated best had the longest hyphae (mean 7,110 jm )  and A, tamarii. 
the shortest (1,950 ^im), Even though only 27 per cent o f the conidia
University of Ghana                              http://ugspace.ug.edu.gh
0f  niger had germinated a fter  50 hours incubation, the to ta l 
length o f germ tubes produced was higher than fo r  A. tamarii»
The e ffe c ts  o f solu tions o f  the spots eluted from thin layer 
plates on the conidia of the A sp erg illi and s c le ro t ia  of £3. r o l f s i i  
showed varied responses. Compounds eluted from spot (Table I3A) 
which had been ten tatively  id en tified  as an anthocyanin compound 
and the solu tion  o f spot NSe^  (Table 153) generally depressed germi­
nation. With the exception o f A, niger and S,. r o l f s i i . germ tubes 
o f the others grew better in compounds HU^ than in  the control 
(see Table 12A). In the NSe^  so lu tion , A, flavu s . A. n iger. ard 
A. fumigatus and A. tama r i i  worse than the con tro l. Compounds 
eluted from NSe  ^ stimulated percentage germination in  the Aspergillus 
species but depressed germination in £!. r o l f s i i  (Table 15^). With 
the exception o f A. fumigatus and A. tamarii germ tube growth was 
depressed.
Compounds o f spots from hydrolysed extracts (HSe^) and ®!Si^) 
had the same e ffe c t  on percentage germination. Their e ffe c ts  
stimulated the germination o f A. niger and the s c le ro tia  o f  £!, r o l f s i i  
but depressed a l l  others; the severest e ffe c t  being recorded on 
A. tamarii (99 -  100 per cent) and A. fumigatus (97.5 -  99 per cen t). 
While HSe^  tended to  depress germ tube growth among the Aspergillus 
species, with the notable exception of A. n iger„ HSi^  uniformly 
stimulated growth of the germ tubes and hyphae o f S_. r o l f s i i .
University of Ghana                              http://ugspace.ug.edu.gh
These observations point out tha% uniform e ffe c ts  o f the 
phenolic compounds in  the testa  of cocoa beans on the contaminant 
fungi should not be expected and the response o f each fungus to  
the compounds must be s p e c if ic a l ly  studied. The resu lts have a lso  
shown that the extracts o f the bean could stimulate or inh ib it both 
contaminant and non-contami~ant species. A0 niger tended to  react 
to  the compounds and extracts lik e  the non-contaminant jS, r o l f s i i .
The observation by Laycock (1931) that 76 per cent o f fungal 
contamination of the bean cotyledon occurred through the micropylar 
end and the rest through other points was confirmed in  th is study. 
Other paths suggested are the raphe by Maravalhas (196S) ,  and cracks 
in the testa  <us boanc swell during fermentation. An experiment was 
conducted to find out the s';age in cocoa bean curing when invading 
hyphae could be observed in tho cotyledons. This would show whether 
invasion is  associated with sp e c ific  chemical changes in  the bean.
It was observed that in the unfermented beanr 4 per cent of 
the besns contained hyphae in  the micropylar region while another 
4 per cent showed hyphae in the sty lar region (see Table . Tho 
fungi isolated  wo,;© Botryodiploddg. theobromae and Fqsari.ua decern— 
ce llu la re . Spectrophotometric analyses had shown that the un fer­
mented cotyledon was high in anthocyanin (0 j6m g /g ) and flavonol 
content (4.0mg/g) while the testa  had these in lower quantities,
0.07 and 0.41 mg/g, respectively  (see Table 7 ' ,
116
University of Ghana                              http://ugspace.ug.edu.gh
wCompounds o f extracts of  the unfermented bean eluted from 
thin layer chromatographic plates had inh ib itory  e ffe c ts  on spore 
germination. Spot l(c)^depressed spore germination in A. fla vu s .
Ac fumigatus, A, n iger, and A, tamarii and s c le ro t ia  o f S_„ r o l f s i i  
(see Table 13A). It  is  indeed surprising that despite the presence 
of these compounds, B, theobromae and F, decemcellu lare could invade 
the beans. It  is possible that phenolic compounds alone do not 
confer resistance to  fungal invasion in  the beans, Seevers and Daly 
(1970) arrived at the same conclusion while working on wheat stem 
rust. They stated that the relationsh ip  between phenolic compounds 
and disease, resistance had not been proven. Invasion., thereafter, 
went on stead ily  in the beans throughout fermentation and drying.
There was considerably higher in fection , particu larly  by 
A. furaigatua and Hueor species o f beans that had been fermented fo r  
2 days with hyphae occurring in  a l l  parts of the bean except- in  the 
micropylar region ,
Cn the last day of fermentation? that i s ,  6th day of curing, 
there was considerably higher fungal invasion of the beanst The 
micropylar and stylar regions showed greater in fection  than the 
mid-region and the principal invading fungi were Mucor pusi l lu s ,
R. arrhizus and the Ap^crgi.Lr.is species , A» fumigatus> A, rostr ig tu s 
and A* tamari i . The s im ila rity  o f  d istribu tion  pattern with the 
unfermented bean was strik ing , suggesting that invasion commences 
at both sty lar and micropylar ends*
University of Ghana                              http://ugspace.ug.edu.gh
' F inally , beans dried fo r  6 days a fter  fermentation * that is ,  
on the 12th day of curing revealed high contamination which was 
particu larly  severe at the micropylar and sty lar ends, The major 
species being Mucor  sp* j Penicillium  sp, and three A sp erg illi,
A, ch ava lie r i. A. f lavus and A, tamarii (see Table 14)>
The occurrence o f  the phenolic compoimds during these events 
did not portrny any particular pattern, neither.was the frequency 
o f mycoflora occurrence associated with s p e c if ic  leve ls  o f phenolic 
compounds* Thus as shown in  Table 7, free  phenols remained alkost 
at tho same lev e l on the 2nd and 6th days of fermentation while 
fungal invasion increased. A lso , contamination was very high, on 
the 6th day of curing when the content of orthodihydric phenols had 
risen  tremendously from 0.33 mg/g to 2,18 mg/g on the 2nd day and 
then to  3*41 mg/g on the 6th day (see Tables 8 and 9 ),
Fungus species associated with contamination o f cocoa boanr; 
included those with e e llu lo lrb lc  and those without known c e l lu lc ly t i  
properties; eg, Peni c i l l ium species , A„ f l avus .  R, arrhizus and 
A., taijarij,,, Fungi without known c e l lu lo ly t ic  properties could have 
entered the cotylodons through paths croated by those with c c l lu -  
lo ly t ic  properties. During entry, the fumgi must have withstood 
the inhibitory .f le e t s  of the phenol compounds, Pridham (i960) 
suggested that fungi employed two methods fo r  detoxifying chemical 
barriers during the in fection  process. Such compounds wore either
University of Ghana                              http://ugspace.ug.edu.gh
glycosylated by the addition r f  sugar groups thus changing thoir 
fchemical mature or tlae tox ic  compounds werq so lu b ilized  so that 
compounds insoluble a| physiological pH values; would then b© 
dissolved i>e allow fo r  their translocation*
Some o f the fa ctors that could predispose cocoa beans to 
fungal in fection  have been studied)* The disorganisation of Ihe 
selerotiSed layer of bean testa  during fermentation (Roelofsen, 1!)58)> 
and the breakdown o f ce llu la r  structure around the dome o f the 
mieropylap region (Dade, 1929) were not investigated in th is study, 
These ea rlie r  reports could be confirmed in future studies®
Since some o f  the phenolic compounds have been found to 
inh ib it fungal spore germination* further studios should identify  
potent compounds fo r  possib le future use in the con tro l o f cocoa 
bean fungal contaminants? Another recommendation that could be made 
is  that ffrmentatipn time should be shortened to four days, using 
smaller heaps and the drying time also reduced possibly  through the 
u^e pf a r t i f i c ia l  drying methods in order to. shorten the favourable 
peylod fpr invasion ajid thus minimise fungal contamination*
University of Ghana                              http://ugspace.ug.edu.gh
SUMMARY
The average annual fungal contamination o f cocoa beans at 
Tema from October 1978 to  July 1978 was 4.9 per cen t.
The main crop cocoa beans had lower fungal contamination o f  
3 .9  per cent and the minor crop beans a higher percentage o f 
6 ,4  per cent.
The averag3 moisture content o f coeoa beans at Tema warehouse 
was 6 ,4  per cent.
High incidence of fungal contamination was associated with 
high bean mois-tsure content.
Twenty-one species of fungi belonging to  nine genera were 
Isolated  from contaminated cocoa beans in  addition to the 
presence o f some ster:.le  mycelia0
The most important fungal genera were A spergillus, Penicillium  ^
Mucor and Rhizopus with the A sp erg illi  constitu ting  the 
dominant f lo ra , particu larly  those of the Aspergillus glaucus 
group; A. ch eya lie r i. A, ustus and A. ruber.
4* ch eva lier l was the predominant contaminant o f  stored coeoa 
beans.
Aspergillus phoenicis, Rhizopus orvzae. and Keurospora crassa 
were being recorded on stored cocoa beans fo r  the f i r s t  time. 
Testa extracts of unfermented beans supported vegetative 
growth in A, fla vu s . A. fumigatus. A. n iger. A, tamarii and 
.Solerotimr) r o l fs ^ i ,^
University of Ghana                              http://ugspace.ug.edu.gh
A ll Asperg illu s species grerr better than S., r o l f s i i  in  the 
testa  extracts of the unfermented beans* with A.i tamarii 
producing the greatest amount o f mycelium*
Extracts o f testa  of fermenting beans inh ib ited  growth o f 
the f iv e  fungi, r o l f s i i  was the most severely inhibited 
fungus.
Extracts o f  testa  o f beans being dried a fte r  fermentation 
a lso depressed vegetative growth in the f iv e  species but to  
a le ssor  extent than extracts o f fermenting beans.
In Czapek Dox broth at 25°0 , tamarii grew best, producing 
mean mycelium dry weight of 2 8 7 mg and A, nidulans grew 
poorest producing 126,2 mg dry weight of mycelium.
In th is medium, A, fumigatus grew fa s te s t , reaching maximum 
growth in p days of incubation and A. nidulans slowest* 
attain ing maximum growth a fter  14 days o f incubation.
In Sabouraud’s broth at the same temperature, however, A. funi~ 
gat us grew best and JL, tamr Li the poorest, But on Sabourauds 
agar medium nidulans grew slowest as measured by colony 
diameter; at both 25° and 300C,
iho unfermented cocoa bean cotyledons and testa  hav$ phenolic 
compounds in varying concentrations* The amounts of these 
substuices changed during fermentation and dryingo
University of Ghana                              http://ugspace.ug.edu.gh
The quantities recorded could be summarised as follows 
(a) .Free -phenols, in coty ledons
Uhfernented cocoa bean s 4*00 + m i mg/ g
Bean fermented fo r  2 days : 1 *13 ±  0*82
Bean fermented fo r  6 days : 0.73 ±  0*16 II
Bean dried fo r  6 days : 0.70 + 0.20 it
Free phenols in  testa
Unforuented cocoa bean s CTiO,»o+1>—o mg/g
Bean fermented for  2 days : 0.45 + 0 .08 tl
Bean fermented fo r  6 days : 0.32 + 0.11 tt
Bean dried for  6 days : 0.22 + 0,06 II
Orthodihydric phenols in cotyledon
Unfermented cocoa bean j 0,33 + 0.18 mg/g
Bean fermented fo r  2 days s 2.18 + 0,47 1
Bean fermented fo r  6 days j 3,41 + 0.09 !l
Bean dried fo r  6 days : 2,42 + 0,88 II
Orthodihydric phenols in  testa
Unfermented cocoa bean : 2.43 + 0.22 ng/g
Bean fermented f o r  2 days : 2.19 ±  0.72 !!
Bean fermented fo r  6 days : 3.15 ±  0.81 tl
Bean dried fo r  6 days t 2.46 ±  0.95 tt
University of Ghana                              http://ugspace.ug.edu.gh
(e ) Flavonals Ifa cotyledons
125
Unfermented cocoa bean ; 0.21 CM•
o
mg/g
Bean fermented fo r  2 days t 0.39 + 0.16 a g / g
Bean fermented fo r  6 days ; 0,72 + 0,09 11
Be$n dried fo r  6 days : 0.49 + 0.11 ft
( f ) r ^ i - ^ Z a  .'in •: ocJ.M ■____
Unfermented cocoa bean : 0.51 + 0.15 mg/g
Bean fermented fo r  2 days : 0.33 + 0 ,1 4  "
Bean fermented fo r  6 days ; 0 .4 3 + 0 .1 0  ”
Bean dried fo r  6 days : 0,37 + 0.04 "
(g) Anthocyanin in cotyledon
Unfermented cocoa bean s 0.16 £  0 .02  mg/g
Bean fermented fo r  2 days : 0.13 ±  0.02 H
Bean fermented fo r  6 days : .0.20 + 0,05 "
Bean dried fo r  6 days i ©►14 + 0.01 "
(h) Anthocyanin in testa
Unfermented cocoa bean : 0.07 + 0.02 mg/g
Bean fermented fo r  2 days i 0.07 + 0.01 "
Bean fermented fo r  6 days ; 0.15 ±  0.05 "
Bean dried fo r  6 days : 0 .24 + 0.05
17. Testa extracts of unfermented beans showed four chromatographic 
spots o f phenolic compounds o f  Rf values of 0 .35, 0.48, 0.60 and 
0.70 when solvent n -butanol-acetic acid-water was used.
University of Ghana                              http://ugspace.ug.edu.gh
18. The number of chromatographic spots obtained from extracts o f 
unhydrolysed testa  o f beans oured fo r  0 , 2, 6 and 12 days was 
4, 3j 1 and 0 respective ly  with the respective  Rf values o f  
0.35, 0 ,48, 0 o60 and 0 .70 fo r  unfermented beans and 0 .35 , 0.48 
and 0,62 for  beans in  2 days o f fermentation. The spots fo r  
extract of beans in  6 days o f  fermentation had an Rf value of 
0.59c
19. The number o f chromatographic spots obtained from extracts o f
hydrolysed testa  of beans cured fo r  0, 2, 6 , and 12 days was 
2 , 1 , 1  and 0 respective ly  with the respective Rf value of
0*35 and 0,57 fo r  unfermented beans. The single  chromatographic 
spots observed fo r  extracts of bean cured fo r  2 and 6 days was 
0.57.
20* The percentage o f conidia o f  A. flavus. A* fumigatus. A. niger
and A, t amar i i  and sc le ro tia  o f £3. r o l f s i i  which germinated in
EDA "flood  water" at 25°C a fter 50 hours was 27, 71, 33, 30 
and 41 per cent respective ly ,
21, The germinating s c le ro t ia  o f S_. r o l fs i  i produced many hyphae.
Among the A sp erg illi, A* fumigatus had the longest germ tubes, 
a mean o f 7,110 um, and A. tamarip. th® shortest -  a mean o f
1,950 um,
22. Compounds eluted from chromatogram spots had varying e ffe c ts  an 
the conidia o f the A sp erg illi and sc le ro tia  o f S., r o l f s i i
124
University of Ghana                              http://ugspace.ug.edu.gh
23. Compound from spot F (0 ), of Rf 0.42 o f  the extract of non- 
hydrolysed testa  o f unfermented beans inhibited germination o f 
conidia of a l l  the A sp erg illi  and sc le ro tia  o f  S,. r o l f s i i . and 
reduced growth o f germ tubes and hyphae.
24. Compounds eluted from the spot H(2)^ 0^0.31 o f the extract of 
unhydrolysed testa  from beans fermented fo r  2 days also inh ib ited  
ger!i;jaation o f the conidia o f  a l l  the A sp erg illi  and sc le ro tia
of _S. r o l f s i i . I t ,  however, depressed germ tube growth in 
A. fumigatus and A, tamari i  while i t  stimulated that o f A. flavus 
A. niger and S, r o l f s i i .
25. Compound eluted from chromatographic spot H(2)^ of Rf value
0.23 o f extract o f hydrolysed testa  o f beans on second day of 
fermentation severely depressed germ tube growth in A. flavu s .
A„ fumigatus and _A, tamarii but improved germ tube growth in 
A* niger and S., r o l f s i i .
26. Compound eluted from chromatographic spot H(2)„ o f Rf 0 .54, of 
extract of hydrolysed te3ta o f beans fermented fo r  2 days reduced 
con id ia l germination in A. flavus. A. fumigatus and A, tamarii 
but stimulated i t  in A, niger and improved germination o f
jS. r o l f s i i  s c le ro t ia .
27. Compound from chromatographic spot H (6)1 o f Rf 0.23 o f  exttact 
o f hydrolysed testa  o f beans fermented for  6 days a lso  depressed 
spore germination in A. flavus . A, fumigatus and A. tamarii but
jl 25
University of Ghana                              http://ugspace.ug.edu.gh
stimulated germination in A„ niger  and ,3, r o l f s i i . I t  however, 
e'n.VianoecL germ tube growth in a l l  the species except A. tamarii
28<, Unfemontod cocca beans* bta.y\S be-ing fsrncntcd ana "those boing dried 
a fte r  fermentation a l l  had fungal Qontaminants, but in fection  
was greater in beans being fermented or being dried .
29. Asperg illu s and Mucor species were the major f lo r a s
30. The fungi wore not evenly, distributed in the in fected  cotyledon. 
They occurred in greater quantities in the micropylar and sty lar 
regions of the bean than in  the m id-regions.
126
University of Ghana                              http://ugspace.ug.edu.gh
ACKNOWLEDGEMENT.
I wish to  express my sincere gratitude to  my supervisor 
Dr. C.E. Amoako-Nuama who guided me during the course o f these 
investigations and fo r  her construetiTO critic ism  and suggestions 
during tho preparation of the manuscript.
I am very grateful to Professor G„C. Clerk, fo r  his in it ia l  
help with this study befora he l e f t  fo r  Sabbatical Leave and fo r  
h is general in terest and advice.
My thanks and deep appreciation go to the D irector o f the 
Cocoa Research In stitu te , Tafo- and especia lly  Dr<. Daniel Adcmako, 
Mr, 0“A, Danquah, Mr„ P. Akuley ctruiMcBossuan, o f the Physiology 
Department of the Institu te  fo r  their  help with the analyses of 
phenolic compounds* I a lso wish to record my sincere thanks to  ' 
Mr. K, Asiakwam fo r  his help with s ta t is t ic a l  analyses and Mr.
E. K„ Anyomi who typed the th e s is c
I wish to  acknowledge -.rith thanks, the major financia l support 
of the Ghana Cocoa Marketing Board, and a study leave, which enabled 
me to  carry out th is  work,,
F inally , my sincere gratefu l thanks go to  my w ife , Apini, fo r
her patience, encouragement and support during the course of th is
work; to my n iece , Ayama, fo r  her dedication ; and to  my son, Abu
fo r  his refreshing innocence; and fo r  bearing my long absence from 
home.
127
University of Ghana                              http://ugspace.ug.edu.gh
128
VIII. T.ITBHATPRE CUED
ADOMAKO, Dr ( 1975) A biochemical study o f the mechanism o f  swollen 
shoot virus resistance in  cocoa: the ro le  o f phenolic
compounds. PROCEEDINGS OP V INTER. COCOA RES. CONF. 
IBADAN, NIGERIA, pp 344 -  348. '
ALTSCHUL, A.M. (194©) B io log ica l processes o f the cotton  seed.
COTTONSEED AND COTTONSEED PRODUCTS (BAILEY, A.E. ED) 
IOTERSCIENCE, NEW YORK, pp 157 -  212.
ANGELL, H.R., J. C. WALKER and K.P. LINK (1930) The relationship
o f  protocatechuic acid to disease resistance in  onion, 
PHYTOPATHOLOGY 20, 432 -  35.
ANON, (1968) Procedures fo r  Grading and Sealing- o f Cocoa and other 
Producef INDUSTRY AGREEMENT REGULATIONS PRODUCE 
INSPECTION DIVISION (GCMB) pp 1 -  2.
ANON, (1969c) Committee on commodity problem Study Group on cocoa, 
Working Party on Cocoa grading. THIRD SESSION, PARIS 
27-31, MAY 1969.
ANSELMI, J .A .j S-N,. LIU & C.H. MU (1974) The Barico Cocoa Drier. 
PLANTER, KUALA. LUMPUR j?0, 144.
ASARE-NYAKO, A. (l972) A comparison o f fungicides for  the control 
o f blackpod disease o f  cocoa caused by Phytophthora 
palmivora (Butl) Butl. in  Ghana, REP. 3RD SESSION OP 
THE F.A.O. TECH. PARTY ON COCOA PRODUCTION pp 20.
BARUAH, P. & T. SWAIN (1959) The action  o f  potato phenolase on
flavonoid compounds, J. SCI. POOD AGRIC. _1_0, 125—129
BATE-SMITH, E.C. (1954) Plavonoid compounds in Poods.
ADV. PD. RES. 5., 261-295.
BA7ENDAMM, W. (1.928) Uber das Vorkommen und den Nachweis von
Oxydase be i Holz aertorenden Pilzen . Z. PPLANZENKRANKH 
U. PFLANZENSCHUTZ 3§., 257 -  258.
BIODIN, J & j ,  pREVOST (1957) Les Aspergillus rencontres en tanneries 
Leur action., BULL. ASS.PR. CHEM. IND. GUIR, JS> 47 -  72.
BROADBENT, J.A. & j , o ,  OYENIRAN (1967,/ The penetration of cocoa 
beans -by filamentous mould during fermentation and 
drying„ ANN. REP. NIGERIAN STORED PROD. RES. INST.
1967, TECH. REP, 6, 59 -  65, PEDERAL MINISTRY OP 
INFORMATION, PRINTING DIVISION, LAGOS.
T lO n lJT lT  T l  "B f - l n t r / i N  — .
 atxon o f  pigment c e l ls  o f cacao.
—  <■-*» 492
University of Ghana                              http://ugspace.ug.edu.gh
129
BUNTING, RjHi ■ (1988) Fungi occurring in cacao beah^lDEPT* QF 
AGRIC; GOLD COAST Y^AR BOOK 1928, BULLS 16, 37-4-3,
BUNTING, R.H, (1929a) Defective cocoa .' DEFT* OF AGRIC; GOLD COAST 
YEAR BOOK. 1928, BULL 16, 37 -  43*
BUNTING, R.H, (1931) Deterioration o f cocoa beans by internal
mouldsi BULL. OFFI, IHT* FABRj CHOC4 COCOA. , f jt 295-304,
BYRDE, R; J.W»• A;H, FIELDING & A,Hi WILLIAMS (1959) The role  o f 
oxidized, polyphenols in  the varieta l resistance o f 
apples to brown ro t . "PHENOLICS IN PLANTS IN HEALTH 
AND DISEASE" J.B, PRIDHAM (ED*) pp 95 -  98i
CAD/E.N? C„H, (1959) Some properties of an inh ib itor  o f Vifus 
in fect ion  from leaves o f Raspberry,
Ji GEN. MICROBIOL. 20, 113 -  128.
CH&NDL3R, B.V. & T. SWAIN (1959) Separation of anthooyanins from 
plant extracts, NATURE 183. 989
CIFFERI, R. (1931) Studies on cocoa. I ,  Actinomycetes on cocoa 
beans. J. DEPT. AGRIC. PUERTO RICO .15, 223 -  286
DADE, H.A. (1927) The economic sign ificance  o f cocoa pod diseases 
and fa ctors determining their incidence and con tro l, 
DEPT. OF AGRIC. GOLD COAST, BULL. NO. 6.
DADE, H.A, (1929) Internal moulding o f prepared cocoa. DEPT. OF 
AGRICULTURE, GOLD COAST YEAR BOOK, 1928, BULL.
No, 16-, 74 -  100.
de ROUBAIX, J & 0 LAZAR (i960) The Inhibitory substances contained 
in sugar beet glomerules. "PHENOLICS IN PLANTS IN 
HEALTH AND DISEASE;. PRIDHAM J.B. (ED), pp 35 - 4 1 .
de WITT, K.W, (1952a) Studies on small-scale cocoa fermentation in 
the laboratory. REPT COCOA RES. TRINIDAD. 1954 -  51,
pp 110,
DIECKMN, J ,  (1962) Cocoa bean tests  1961/62. GORDI&N PUBLISHING 
HOUSE; HUMBURG. pp 15 -  20.
DIX, N,J, (1972) E ffect o f s o i l  fungistasis on spore germination 
and germ tube growth in Penicillium  species. TRANS.
BR. ffiTCOL. SOC. ^8 (1 ), 59 66.
University of Ghana                              http://ugspace.ug.edu.gh
130
SKDHDAYO, .
FARKAS, G.l
FLOOD, A.E,
FORSYTH, W. 
FORSYTH, ¥,
FORSYTH, If,
FORSYTH, W.
FORSYTH, W.
FORSYTH, W.
FORSYTH, W. 
GRIFFITHS, 
HALL, T.H.E
JiA, (1970) Pycnidium production by B. theobromae. I l l  
Germination o f pycnidiospores. CAN. J. BOT. 48, 67-70,
j .  & Z. KIRALY (1962) Role o f phenolic compounds in the 
physiology o f plant diseases and disease resistance, 
PHYTOPATHOLOGY 44, 105 -  150.
, & D.S. KERKHAM (1959) The e ffe c t  o f some phenolic
compounds on the growth and sporulation o f  two Venturis 
species , "PHENOLICS IN PLANTS IN HEALTH AND DISEASE." 
PRIDHAM J.B . ED), pp 81.
,G.C. (1952) Cocoa Polyphenolic substances. I I  Changes 
during fermentation, BIOCHEM. J. _5J.f 516 -  520
,G.C. (1954) Cocoa Polyphenolic substances. I l l  
Separation and Estimation o f Paper chromatograms, 
BIOCHEM. J. 60. 108 -  111.
,G.C. & J.E, ROBERTS (i960) Cacao polyphenolic substancesc 
5. The structure o f  cacao. - 1" -
"leucocyanidins 1" BIOCHEM. J. 2 1 t  374 -  378.
G.C. & J.E. ROMBOUTS (1951) Our approach to the Study 
of Cocoa fermentation. RPT. COCOA CONF. LOTION, COCOA, 
CHOC AND CONFECT. ALLIANCE pp 73 - 8 1 .
G.C. & V.C. QUESNEL (1957) Cocoa glycosidase and colour 
changes during fermentation. J. SCI. FD. AGRIC,
8, 505 -  509.
G.C. & V.C. QUESNEL (1963) The mechanism o f cocoa 
curing. ADV. ENZYMOL. AND REL. SUBJ. OF BIOCHEM.,
(F.F. NORD ED.) 25, 452 -  91 INTERSCIENCE PUBLISHERS; 
LONDON.
G.C.; V.C. QUESNEL & J.B. ROBERTS (1958) The interaction  
o f polyphenols and proteins during cacao curing.
J. SCI. FD. AGRIC. j), 181 -  190.
L.A. (1958) Phenolic acids and flavanoids o f -. Cacao 
L. Separation and id en tifica tion  by Paper Chromato­
graphy. BIOCHEM. J. 10, 1 2 0 -1 2 5
-• (1975) The culture o f Theobroma cacao L. in v itro  as 
a possible means of vegetative propagation.
Ph.D. THESIS, BOT. DEPT., UNIV. OF LIVERPOOL, pg. 50.
University of Ghana                              http://ugspace.ug.edu.gh
131!
HALLAS, C.
HARBOR*, 
HAWLEY, L.
HOLDEN, M«
HOWAT, G.R
HUANG, H.T 
HULME, A.C
HUMPHRIES,
JACKSOH, G
KEFFORD, J 
KENTEN, R.l 
KIHALY, Z i
KIRKHAM, D
A. & T.A. WIGHT (1939) The tannins and related substances 
in  cacao. BRIT. ASSOC. EES. COCOA, CHOC., SUGAR 
CONFECT. JAM TRADES, RES, BEC. NO. 70.
J.B. (1973) Phytochemical Methods, CHAPMA.N AND HALL,
LONDON.
F.'j L.C. FLECK & C.A. RICHARDS (1924-) The re la tion  
between durability  and chemical composition in  wood,,
IND. ENG, CHEM. J6., 699 -  700.
(1959) Processing of Raw Cocoa, I I I  Enzymic aspects 
o f Cocoa Fermentation, J. SCI. OF FD. AGRIC. .
JO, 691 -  700,
. (1957) Fermentation and drying -  Research in  the f ie ld ,  
REPT. COCOA CONFERENCE (1957) LONDON pp 190, THE 
COCOA, CHOCOLATE AND CONFECTIONARY ALLIANCE.
• (1955) The decolorisation  o f anthooyanins by fungal 
enzymes, J„ AGRI, POOD. CHEM. 3,» 141 -  146,.
& K.L, EDNEY (1959) Phenolic substances in  the Peel 
o f Cox’ s Orange pippin apples with reference to 
in fect ion  by Geotrichum perennaas, "PHENOLICS IN 
PLANTS IN HEALTH AND DISEASE, 1 J.B, PRIDHAM (Ed.) 
pp 87 -  94,
E.C. (19410 Studies on tannin compounds.. I Changes 
during autolysis of minced cacao bean.
BIOCHEM. J. 38, 182 -  187.
.A.D. & R.B. WOOD (1959) Presence in  Rose Hips of
Substances Inh ibiting  the Oxidation o f  Ascorbic Acid, 
NATURE 184, 902 -  903.
,F. (1959) The chemical constituents o f citrus fru it s .
ADV. IN FD. RES. £ ,  286 -  355.
I. (1965) Loss o f  anthooyanins during the storage o f 
dry fermented cocoa . GHANA J. SCI. 5., 240 -  242.
fc G.L. FARKAS (1961) Relation between phenol metabolism 
and Stem Rust Resistance in Wheat.
PHYTOPATHOLOGY J52, 657 -  663,
,s * (1957) Studies o f  the s ign ifican ce  o f Polyphenolic 
host metabolites in the nu trition  o f Venturia 
j naequalis and V, p ir ina , J. GEN. MICROBIOL. 17, 120-134.
University of Ghana                              http://ugspace.ug.edu.gh
132:
KNAPP, A|¥i (1937) "Cocoa Fermentation", JOHN BALE, SONS AND 
CURNOW LTD., LONDON, 1937, pp 7 8 - 9 2 ,
KNAPP, A«W» & K.H. COWARD (1934) V it . D. in cacao sh e ll,
THE ANALYST 59, 474 *- 478.
KNAPP, A,U, & K.H. COWARD (1935) The V it D a c t iv ity  o f cocoa sh e ll , 
BIOCHEM. J. 2g, 2728 -  2735i
KNAPP* A.Wo & R,V. WADSWORTH (1924) The d istribu tion  o f theobromine 
during the fermentation o f cacao. J , SOC. CHEM.
IND., 42 ; 124.
KOSUGE, T. (1969) The ro le  o f  Phenolics in Host Response to 
in fe ct ion , ANNe REV, PHYTO, 1 , 195 -  222.
KUC, J. (1964) Phenolic compounds and disease resistance in
plants. PROC. SIMP, PLANT PHENOLIC GROUP N. AM. 63 
(RUNECKLES, V.C. Ed, PUNT PHENOLICS GROUP, N. AM., 
1964).
LAMB, J. & H.B, SREERANGACHAR ( 1940) Studies on the "fermentation" 
o f Ceylon tea . 1, The Nature o f the enzyme system
2. Oxidizing enzymes, BIOCHEM. J. 1472 -  1492.
LAYCOCK, T, (1931) An investigation  in  fermentation: i i  An experi­
ment on the fermentation and moulding o f cocoa,
9TH ANN. BULL, DEFT. OF AGRIC; NIGERIA, pp 5 -  26.
(AS PRINTED) GOV'T PRINTER; LAGOS.
LIAU, H.T.L. (1976) Raw cocoa processing, AGRIG, RES. CENTRE,
SABAH, MALAYSIA TECH. BULL -  DEPT. OF AGRIC. (MALAYSIA) 
No. 2, pp 37.
LILLY, V.G. & H.L. BARNETT (1951) Physiology o f the fungi. 
MbGRAW-HILL, NEW YORK.
LINK, P .; H.R. ANGELL; & J.C. WALKER ( 1929a) The iso la tion  o f
protocatechuic acid from pigmented onion scales and 
i t s  sign ifican ce  in re la tion  to disease resistance 
in  onions. J. BIOL. CHEM. 81_, 369 375..
LINK, P. <5: J ,0. WALKER (1933) Tho iso la tion  of catechol.
J. BIOL. CHEM* 100. 379 -  382.
MACLEAN, J.A.R. (1953) Some chemical aspects o f "Blackpod" disease
in W. African amelonado cacao.
EMPIRE J. EXP. AGRIC, 21 , 340 -  349.
University of Ghana                              http://ugspace.ug.edu.gh
133
A, (1955) Cocoa-boon protoin , CHEM, ABST. 8522 -  24
P-NNERS, J,Gi (1966) Assessment of germination* "The Fungus. Spore," 
Madeline M.K. (Ed.) BUTTERWOETH, LONDON, pp 165 -  174I
MilRAVALflii.3» N| (1966) Myeojogical deterioration  o f cocoa beans
during fermentation and storage in Bahia, INI. CHOC* 
REV* 2 i (8)* 375 -  378,
MA.SON, JUS. (1955) Comparative biochemistry o f  the Phenolase 
complex, ADV, IN ENZYMOLOGY j6 ,  105 -  173,
MAYNE, R .J ,} Ji¥ . BENHETT'j & J . TALLANT (1971) In stab ility  o f  
an a flatoxin  producing stra in  o f Aspergillus 
parasiticus. MYCOLOGIA. £3, 644 -  648,
NORDSTROM, C.G, (1956) Biochemical methods in polyphenols,
ACTA, CHEMICAf SCAND, JO, 1491 -  1498,
OLUTIOIA, P.O. (1976) Enzyme systems’ o f Aspergillus tamarii 
PHYSIOLOGIC PIAKT  ^ ’37 (4 ), 313 ~ 316.
OLUTIOLA, P.O. & 0 .0 , COLE (1977) C e llu lo ly tic  enzyme system in 
Culture F iltra tes  of Aspergillus sydowi 
PHYSIOLO, PLANT. 22,(3), 243 -  247.
OYENIRAN, J ,0 . & M.O, ADENIJI (1974) Mould deterioration  o f cocoa 
during processing and storage in Nigeria, NIGERIAN 
JOUR. OP PLANT PROTECTION J.(No. l )  48 -  51
PASSMORE, F,R. (1932) A survey o f damag0L_b^_insects and moulds of
¥, African cocoa before storage in Europe: season 
1930 -  31 . BULL. IMP, INST. 30, 296 -  305.
PRIDHAM, J.Bv (i960) Phenolics in Plants in Health and Disease. 
PERGAMON PRESS, LONDON. (Ed.)
REESE, E,T,, & H.S. LEVINSON (1925) Comparative study o f  the 
breakdown o f ce llu lase  by micro-organisms,
PHYSIOL, PLANT jj, 345 -  366.
REESE, E.T, & M.H. DOWNING (1951) A ctiv ity  o f the A sp erg illi on
ce llu lo se , ce llo lu se  derivatives and wool.
M3TC0L0GIA 12, 16 -  27,
RENAT®, R, (1954) La qualite/  du cacao : le s  moisissures des feves 
fermentees, AGRON. TROP, (NOGEBT-SUR-MiiRNE)
J ,  563 -  566.
University of Ghana                              http://ugspace.ug.edu.gh
134
ROBERTS* KiA.H* (1957) Economic importance o f  flavonoid
substances: Tea fermentation* " CHEMISTRY OF FLAVO-
N0ID CCiTOUKDS1' , (GEISSMN* T.A; Ed|) pp 4l8 & 472j
ROBERTS; E,A,H.- <Sj Ds J| WOOD (1950) The fermentation process in tea 
Manufacture; 10i The condensation o f catechins and 
it s  re la tion  to the chemical changes in  fermentation* 
BIOCHEM; Js 4^*538 -  542.
ROBINSON» G,Mf & R* ROBINSON (1933) A survey o f anthocyanins* I I I  
Notes on the d istr ibu tion  of leuco-anthocyanins,
BIOCHEM» J| 22Jj 206 -  212„
ROELOFSEN, PwL| (1958) Fermentation, drying and storage o f 
cacao beans, ADV, FD, RES. 8 , 225 -  96,
ROELOFSEN, P.A, &. G. GEISBBRGSR (1947) Onderzoekingen over cacao** 
bereiding, ARCH. KOFFEDCULT. (NETH. IND.)
ENGLISH SUMM/iRX 16, 1,mm '■» •
ROHAN, T.A. (1957) Observations on the fermentation o f West African 
Amelonado cocoa* 1957 COCOA CONFERENCE, pp. 203 -  206,
ROHAN, T.A, (1963) Processing o f raw cocoa beans for  the market. 
F*A.O. AGRIC« STUD. 60, 40 -  45,
SCHAAL, L. & G. JOHNSON (1955) The inh ib ito iy  e ffe c t  of phonolic 
compounds on the growth o f Streptomyces scabies as re 
related to tho mechanism of scab resistance. 
PHYTOPATHOLOGY 4£, 626 -  628.
SCHRAUFFSTATTER, E. (1948) The bacter iosta tic  action o f chalcone, 
flavanone, flavone and flavonol. EXP^ RIMEHTIA 40. 
4 8 4 - 9 0 ,
SCOTT, J.L . (1931) Pr@limin.ary investigations on the Moisture
Content and Hygroscepicity o f coeoa beans. DEPT. OF 
AGRIC. GOLD COAST YEAR BOOK 1930, BULL. 2 1 0 7  -  110
SEVERS, P.M, & J.M, DALY (1970) Studies on wheat stem rust
resistance controlled  at the Srb locus. 1, The role  
o f phenolic compounds. PHYTOPATHOLOGY 60, 1322-1328.
SEIiCEL, Ms.K* (1962) Chromatographic methods o f separation, iso la tion  
and id en tifica tion  o f flavonoid compounds,
"THE CHEMISTRY OF FLAVONOID COMPOUNDS", GEISSMAN. T,A. 
m , }  PP 34 -  69.   ' ! ’ ' v 5 1
University of Ghana                              http://ugspace.ug.edu.gh
135
SPBNSIEY, p tCi (1963) A flatoxin j the active princip le  in Turkey 
‘X’ d isease; ENDEAVOR 22 , 75 -  79*
SPIES, E. (1965) Analysing polyphenolic compounds; METHODS IF 
ENZYMOLOGY 467 *. 471
STAHL, E. Ed (1969) Thin layer chromatography; a laboratory 
handbook; SPRINGER-VT5RLAG, BERLIN,.
SWAIN, T, (1953) The id en tifica tion  o f counarins and Related 
compoitr.ds by F ilt  31 -paper chromatography,
BIOCHEM* J, jjg,, 200 -  2084
SWAIN, T. (1960) Economic importance o f flavonoid compounds 
Foodstuff "CHEMISTRY OP FLAVONOID COMPOUNDS", 
GEISSMAN, T,A. (ed) pg, 513.
SWAIN, T & W.E, HILLIS (1959) The phenolic constituents o f  Prumus.
domesticao t The quantitative analysis o f phenolic 
constituents* J, SCI. FD, AGRIC. JO. 63 -  68,
THALER, H. (1954) Die los lich en  kohbenhydrate der kakaobohne, 
NATURWISSENSCHAFTEN, 41., 432 -  35
THOM, C» & K.B, RAPER (1945) A manual of the A sperg illir
THE WILLIAMS AND WILKINS COMPANY, BALTIMORE pp 222,
THOROLD, C.A, (1975) Diseases o f  Cocoa, OXFORD, U.K. CLAVEDON 
PRESS, APPENDIX 368 „
TOMIYAMA, K ,; R, SAKAI; T* SAKUMA; & N. ISHIZAKA (1964) 0 Toxins
o f Host and Paras i t  c-, in  the In fection  Process pp 165. 
'THE DYNAMIC ROLE OF MOLECULAR CONSTITHENTS IN PLANTS- 
PAHASITE INTERACTION"» MIRGCHA C. Je AND I ,  URITANI 
(Ed),
VAN SUMEHE, C.F, WOLF; H, TEUCHY /JSTD A. KINT (1965), A new th in - 
layer method o f phenol'c sr-r.betances and coumarins,
J, CHROMA, 20, 48 -  55,
VARGA, M. & E, KOVES (1958) Phenolic acids as growth and germina­
tion  inh ib itors in  dry fr u it s ,  NATURE 183, 401,
WADSWQRTH, R,V, (1952) The theobromine content o f  cocoa beans 
and Cocoa, THE ANALYST J55, 65 -  74
WADSWORTH, R,V, (1955) The quality o f raw cocoa as i t  a ffe c ts  the 
manufacture!.1, TROP. AGRIC, TRIN, 2 g  ( l ) j ,  1 -  9.
University of Ghana                              http://ugspace.ug.edu.gh
136
WILLIAMj A 
WOODi G;A, 
WOOD, G.A.:
WOOD, G.A„:
WOOD, H.E. 
YEBOAH, S.:
.H, (1957) The Simpler phenolic substances o f plants*
J. SCI. ED, & AGRIC. 8.* 385 -  389.
Rii (1965) Storage o f Raw Cocoa. COCOA GROWERS BULL.
I ,  21 -  24,
Ro (1975) Processing and U tiliza tion  o f Cocoa.
PROCEEDINGS OF 5TH INTER. COCOA RES. CONF., IBADAN, 
NIGERIA. SEPT. 1975> pp 51 -  53.
R. (1979) Manufacturers'need with respect to Quality. 
COCOA GROWERS BULLETIN. NO. 28 pp 4. TRAINING SEMINAR 
ON COCOA GRADING LAGOS, 1972. F.A.O., TA 3195.
(1967) Physiological Plant Pathology. BLACKWELL 
SCIENTIFIC PUBLICATIONS, OXFORD & EDINBURGH, pp 50.
K» (1965) The flavonoid compounds o f Citrus aurantium 
M.Sc. THESIS CHEMISTRY DEPT. UNIV. OF GHANA, 1965.
University of Ghana                              http://ugspace.ug.edu.gh
O
S
93
AWRSDXX A
137
S T A N D A R D  c a l i b r a t i o n  c u r v e  f o r  f r e e  P H E N O L S
University of Ghana                              http://ugspace.ug.edu.gh
Es
ao
I P
CONCEN TRA T IO N  OF  CHLOROGENIC  ACID /Jg/ml.
STANDARD CAL IBRA T ION  CURVE FOR O R TH O D IH Y D R IC  PHENOLS
University of Ghana                              http://ugspace.ug.edu.gh
C O N C E N T R A T I O N  O F  ( - )  E P I C A T E C H I N  yug/ml 
S T A N D A R D  C A L I B R A T I O N  C U R V E  FOR  F L A V O N O L S
University of Ghana                              http://ugspace.ug.edu.gh
CONC.OF CYANIDIN CHLORIDE /ug/ml
s t a n d a r d  c a l i b r a t i o n  CURVE f o r  a n t h o c y n i n
University of Ghana                              http://ugspace.ug.edu.gh