Page 7 - Aboud_etal2011

170 Aboud, Pasquale, Sinacori, Massi, Conte, Alonzo
One sample (H7) with water content over 20% was found, that
exceeds the limit established by the European Community and the
International norms (Anonymous, 2001; Codex Alimentarius
Commission, 2001). Five samples (H8, H18, H19, H20, and H21)
have values more than 18% that might lead to fermentation
(Schweitzer, 1998). Also, these high values could be related to a
premature honey harvesting.
A general standard for a minimum content of the sum of
fructose and glucose is 60 g/100 g for all blossom honeys, and 45 g/
100 g for all honeydew honeys and blends of these. The data
analysed for the sugar content were significantly in agreement with
Bogdanov and Baumann (1988) and with the international norms,
although 3 samples (H1, H5 and H19) showed reducing sugar
content lower than 60 g/100 g (Table 2), which was below the
minimum required by law. The low level of the sugar content for
honey sample H1 is normal, because this type of honey contain
lower amount of monosaccharides (glucose, fructose) than for
blossom honeys (Diez et al., 2004), it is a honeydew honey. For the
honey sample H5 and H19, which are chestnut and Sulla
respectively, the low content is probably due to the presence of
some honeydew components or to their immaturity. In general, the
sucrose content does not exceed 5 g/100 g, for some unifloral
honeys not more than 10 g/100 g and for lavender honey not more
than 15 g/100 g (Council Directive of the European Union, 2002).
The sucrose level is useful for the determination of the botanical
origin of honey (Oddo et al., 1995). All samples showed sensitively
low saccharose content values ranging between 0.1- 3 (g /100 g)
(Table 2). Sulla honey from Termini Imerese (H4) showed the
highest value 3 g/100 g. All the honey samples are in agreement
with the limit permitted by the international and the European
community regulations.
The diastatic activity in honey is considered a quality factor. It samples (H2, H23 and H26) were clustered in one group, except one
decreases during storage, heat treatment and feeding of honeybees
during honey flow; thus, it is an indicator of honey ageing,
adulteration and overheating. In general, the diastase number
should not be less than 8, (except for baker’s honey which is an
artificial honey used in commercial baking e.g. sweets and bread). It
should be noted that the use of this parameter in honey as an
indicator of its freshness is often criticised, since the initial enzyme
activity may vary with honey type (Oddo et al., 1999; 1990; Oddo
and Piro, 2004; White, 1994). The honey samples analysed in the
present work show a range of values, between 5.9 and 27.3 Schade
units. Five samples (H14, H16, H19, H21 and H22) show values
below 8 Schade units (Table 2). The explanation for the low content
of diastatic activity found in these 5 samples could be accounted for
an inadequate processing or storage conditions.
The colour and flavor of the honey produced by bees depends
on the flower to which the bees harvested their nectar. The natural
colour of honey presents many tonalities from straw yellow to
amber, from dark amber to almost black with a hint of red. The
samples possessed colour values, varying from 9 to 132 mm Pfund,
with an average value of 50.54 (Table 2). Only four samples were
extra white; extra light amber to dark amber colour samples were
predominant (67.74%). The highest colour value was 132 mm Pfund
for Erica honey from Marettimo (sample 14) (Table 2), judged as old
or overheated on the basis of its low diastase activity (7.1 Schade
units) and its darkness. Unfortunately the colour of honey tends to
darken with age or change according to the method of conservation
or production used by beekeeper (contact with metal, the
temperature of conservation or exposure to light), it may therefore
be only appropriate for the classification of fresh honeys. A strong
interference of polyfloral honeys with unifloral honeys is also to be
expected (Gonzales et al., 1999). This parameter has been reported
to be useful for the classification of the botanical origin of honey by
the Department of Agriculture in the United States, and is commonly
used in the international commerce (Sabatini et al., 2007).
Classification of honey samples by Hierarchical
Cluster Analysis (HCA)
To classify the 29 honey samples according to their aroma profile,
Hierarchical Cluster Analysis (HCA) was applied. The basis of this
approach is to combine objects that are similar to one another into
separate groups; the degree of association between groups is
maximal if they share similar properties, and minimal if share
different properties. The vertical lines in the dendrogram obtained
(Fig. 1) represent the samples studied and the horizontal lines the
similarity values between pairs of samples, a sample and a group of
samples and sample groups. In our samples, it was possible to
distinguish four principal groups, which show the relation of the
botanical and the geographical origin of the honeys with their aroma
profile (Fig. 1). In addition, these groups were divided in subgroups,
which were connected with the botanical origin of the honey
samples. It can be observed from the Figure 1 that all thistle honey
sample (H3) which falls within another group. This sample (H3) was
collected in a different area (Favignana), while the other three came
from Termini Imerese. This suggests that geographical origin affects
honey characteristics. Polyfloral honeys (H8, H10, H11 and H24)
were scattered in different groups because they came from different
botanical origins or maybe they were collected from different places.
The group containing four honey samples (H19, H21, H20 and H22)
demonstrated some defects such as low diastase activity or high
electrical conductivity and water content defects. Erica honey
samples (H14, H15, H18) were all assigned to the same cluster (Fig.
1), thus demonstrated their similarity. Also, it can be observed that
all eucalyptus honeys (Fig. 1) were perfectly differentiated from the
rest, that is they were the closest in the sense of having the lowest
dissimilarity. The similarity manifested among honeydew honey (H1)
and eucalyptus honeys, specially the sample (H17), was due of the
presence of honeydew elements in this sample.