Introduction
The genus Apis has been separated into 11 distinct
species historically (Engel, 1999) based on morphological,
behavioral, and physiological characteristics, and also on
the geographic distribution. Although 10 Apis species
were recognized, the validity of Apis laboriosa is still
questionable (Arias and Sheppard, 2005). Among these
species, Apis mellifera L. was the most extensively studied
93
Multivariate Morphometric Study on Apis florea Distributed in Iran
Ayça ÖZKAN1, Mohammed M. GHARLEKO2, Berna ÖZDEN3, İrfan KANDEMİR1,*
1
Department of Biology, Faculty of Science, Ankara University, 06100 Tandoğan, Ankara - TURKEY
2
Department of Biology, Zanjan University, Zanjan-IRAN
3Department of Biology, Faculty of Art Sciences, Zonguldak Karaelmas University, 67100 İncivez, Zonguldak - TURKEY
Received: 18.06.2008
Abstract: Multivariate Morphometric and Spatial Autocorrelation Analyses were performed to determine the morphometric variation
in Apis florea colonies representing 13 localities from 4 states on the coastal north-south diagonal in Iran. New morphometric
characters (hind wing length, hind wing width, and hind wing angles) were also measured to determine the usefulness of these
characters for Apis florea. Analysis of variance of new morphometric characters showed that 7 out of the 8 hind wing variables
displayed statistically significant differences among populations (P < 0.05). The scatter plot from Canonical Variate Analysis revealed
that the colonies from north (Ilam, Khuzestan, and Bushehr states) showed an overlapping distribution. The colonies from south
(Hormuzgan state) formed a non-overlapping distant cluster. Based on the spatial autocorrelation analysis, 14 correlograms were
significant for morphometric characters (According to the Bonferroni criterion). Six characters (CuB, C.Ind, HWW, B4, D7, and I13)
showed clinal type correlogram.
Key Words: Apis florea, Iran, Morphometry, Multivariate morphometric analysis, Spatial autocorrelation
İran’da Dağılım Gösteren Apis florea Üzerine Çok Değişkenli Morfometrik Çalışma
Özet: İran’nın kuzey-güney köşegeni üzerinde bulunan 4 eyaletteki 13 lokasyondan alınan Apis florea kolonilerindeki morfometrik
varyasyonu belirlemek için çok değişkenli morfometrik analizler ve uzamsal otokorelasyon analizi yapılmıştır. Aynı zamanda Apis
florea için arka kanat uzunluğu, arka kanat genişliği ve arka kanat açıları gibi yeni karakterlerin uygunluğunu belirlemek için bu
karakterlerde ölçülüp analize dahil edilmiştir. Yeni karakterlerin varyans analizi, 8 arka kanat değişkeninin 7’si populasyonlar arasında
istatistiksel olarak anlamlı farklılıklar bulunduğunu göstermektedir (P < 0.05). Kanonikal Değişken Analizi’nden elde edilen serpilme
grafiği Kuzey’deki kolonilerin (Ilam, Khuzestan ve Bushehr eyaletleri) örtüşen dağılım gösterdiğini ortaya koymaktadır. Güney’deki
koloniler (Hormuzgan eyaleti) diğer eyaletlerdeki kolonilerden daha uzakta örtüşmeyen küme şeklindedir. Uzamsal otokorelasyon
analizinde elde edilen korrelogramlardan 14 korrelogram Bonferroni kriterine göre anlamlıdır. 6 karakter (CuB, C.Ind., HWW, B4,
D7 ve I13) klinal tipte korrelogram göstermektedir.
Anahtar Sözcükler: Apis florea, İran, Morfometri, Çok değişkenli morfometrik analiz, Uzamsal otokorelasyon
* E-mail: ikandemir@gmail.com
Turk J Zool
33 (2009) 93-102
© TÜBİTAK
doi:10.3906/zoo-0806-6
Research Article
honey bee species in the world utilizing their morphology,
biochemical variations and DNA-RFLP, microsatellites and
mtDNA genes, and DNA sequencing (Ruttner, 1988;
Sheppard, 1988; Smith, 1988; Del Lama et al., 1990;
Cornuet and Garnery, 1991; Cornuet et al., 1991; Hall,
1991, 1992; Garnery et al., 1992, 1993, 1998; Crozier
and Crozier, 1993; Estoup et al., 1995; Sheppard et al.,
1996; Franck et al., 2000; Kandemir et al., 2000, 2006).
Although the other species had not been studied
intensively, recently they were the subject of several
investigations (Diniz-Filho et al., 1993; Hepburn et al,
2001; Rinderer et al., 2002; Radloff et al., 2005).
Remaining 9 species (Apis cerana, Apis dorsata, Apis
florea, Apis andreniformis, Apis laboriosa, Apis
koschevnikovi, Apis binghami, Apis nigrocincta, and Apis
nuluensis) were distributed in Southeast Asia and Far
East. A. cerana and A. florea reach to the border of Iran,
Pakistan, and Afghanistan (Otis, 1996). Moreover A.
florea further distributed into Iraq, Saudi Arabia
(Whitcombe, 1984), and Jordan (Haddad et al., 2008). In
this vast geography, A. florea occupies rainforests,
savannas, subtropical steppes, and semi-deserts with
open-air nest (Hepburn et al., 2005). A. florea naturally
does not overlap with the distribution of A. mellifera.
Anthropogenic factors, nowadays, affect the
distributional ranges of Apis species. Although there was
no natural overlapping distribution of A. mellifera with
the other Apis species, currently A. florea, A. cerana, and
A. mellifera are found in the same geographical areas.
Based on the multivariate analysis of morphometric
data, A. florea populations from several countries were
investigated regionally (Ruttner, 1988; Mossadegh,
1993; Rinderer et al., 1995; Ruttner et al., 1995;
Markmoor and Ahmad, 1998; Tahmasebi et al., 2002;
Chaiyawong et al., 2004; Hepburn et al., 2005). The first
morphometric research on A. florea demonstrated that
clinal type of geographical variability exists in A. florae;
larger bees found in the north and smaller ones in the
south (Ruttner, 1988). Besides geography-related
variability, climate-related variation was also reported for
A. florea (Ruttner et al., 1995). In addition, A. florea was
separated into 2 indistinct morphoclusters in Iran
(Tahmasebi et al., 2002) and morphoclusters were
changed clinally with latitude in a much wider geography
extending from Vietnam and southeastern China to Iran
and Oman (~7000 km) (Hepburn et al., 2005).
In recent years, 2 different approaches have been used
to evaluate geographic variation in Apis. The most
common approach is to extend the traditional multivariate
approach, showing their results with Principle Component
Analysis, Discriminant Function Analysis, or Canonical
Variate Analysis (Tahmasebi et al., 2002; Chaiyawong et
al., 2004; Hepburn et al., 2005; Radloff et al., 2005; Adl
et al., 2007). On the other hand, multiple regression
analysis and direct spatial autocorrelation analysis can also
be performed by pairing geographic coordinates with
character means of local populations (Daly et al., 1991;
Diniz-Filho et al., 1993, 2000; Diniz-Filho and Malaspina,
1995; Kandemir et al., 2000). In addition, new size and
angle characters have been used in morphometric studies
of honey bee populations due to deformations in some of
the body parts and usefulness of wing asymmetry (Dedej
and Nazzi, 1994; Clarke and Oldroyd, 1996). The
geographic variation of A. florea has been tried to be
explained with multivariate analysis based on characters
of fore wing, hind leg, abdomen, wing angles, and the
number of hamuli and antenna (Ruttner et al., 1995;
Chaiyawong et al., 2004; Hepburn et al., 2005). The
geographic variability based on spatial autocorrelation
was not studied in A. florea.
The objectives of our research were to investigate the
extent of morphometric variation in A. florea populations
distributed in the coastal states of Iran and to determine
whether differences could be detected within this species
and also to show whether there is a clinal type of variation
with respect to latitude and longitude.
Materials and Methods
Apis florea colonies were collected from Hormuzgan,
Bushehr, and Khuzestan along the cost of Persian Gulf and
from inland state Ilam between 2005 and 2007 (Table 1,
Figure 1). Apis florea workers were collected from each
colony and preserved in 70% ethanol until morphometric
examinations were carried out. From each colony, 10-15
worker bees were selected randomly. Body parts (fore
wing, hind wing, and hind leg) were mounted on
microscope slides. In total 28 morphological characters
were measured and hamuli were counted for each worker
bee (Ruttner, 1988; Nazzi, 1992). These characters
included: fore wing length (FWL), fore wing width (FWW),
cubital A (CuA), cubital B (CuB), numbers of hamuli (HAM),
femur length (FL), tibia length (TL), metatarsus length
(MTL), metatarsus width (MTW), fore wing angles (A4,
B4, D7, E9, G18, J10, J16, K19, L13, N23, O26)
(Ruttner, 1988), fore wing distance D (WDL), fore wing
distance C (WCL) (Nazzi, 1992), hind wing length (HWL),
Multivariate Morphometric Study on Apis florea Distributed in Iran
94
hind wing width (HWW), and hind wing angles (HW1,
HW2, HW3, HW4, HW5). Hind wing characters were used
in the studies for the first time (Figure 2). Size characters
were measured under a stereo-microscope with ocular
micrometer and angles were measured with a goniometer.
Multivariate morphometric analyses were performed on
1667 individual worker bees collected from 115 colonies
from 13 localities.
Multivariate statistical analysis (Analysis of VarianceANOVA, Discriminant Functions analysis-DFA, Canonical
Variates Analysis-CVA) was performed with SPSS (SPSS,
2004) and NTSYS-pc version 2.20 (Rohlf, 2004)
statistical packages. Regression of morphometric
variables on latitude and longitude and the correlation
between morphometric variables were computed using
SPSS (SPSS, 2004). Spatial autocorrelation analysis was
also computed using SAAP (Wartenberg, 1989). Spatial
correlograms were defined by Moran’s I coefficient for 6
distance classes (Sokal and Oden, 1978) and the
significance of correlogram was established using the
Bonferroni criterion (Oden, 1984).
A. ÖZKAN, M. M. GHARLEKO, B. ÖZDEN, İ. KANDEMİR
95
Table 1. The sampling locations of A. florea in Iran (n: number of
colonies).
State Location n Coordinates
Ilam 1. Dehloran 21 32.41N 47.15E
2. Sarollah 3 32.35N 47.22E
3. Musiyan 9 32.32N 47.22E
4. Dasht Abas 17 32.29N 47.47E
5. Ainsole 4 32.10N 47.41E
Khuzestan 6. Dezful 5 32.23N 48.23E
7. Soush 4 32.11N 48.14E
8. Ahvaz 22 31.17N 48.43E
9. Mollasany 3 31.35N 48.53E
Bushehr 10. Bushehr 13 28.59N 50.50E
11. Hkark 2 29.14N 50.20E
Hormuzgan 12. Bandar Lingeh 6 26.34N 54.52E
13. Siri Jazireh 6 26.07N 54.26E
Figure 1. Geographical distribution map of the analyzed populations of A. florea in Iran.
Result
A total of 29 morphometric characters from fore
wing, hind wing, and hind leg were used for multivariate
statistical analysis. Colony sample means and standard
deviations were computed for each morphometric
character from 10-15 bees per colony. The mean values
and standard deviations of characters are shown in Table
2. Apis florea colonies from 4 states on the coastal northsouth diagonal in Iran show high levels of morphometric
variation. Analysis of variance of morphometric
characters showed that 24 out of the 29 morphometric
variables displayed statistically significant differences
among groups (P > 0.05), (Table 3).
Figure 3A-B shows the results of CVA and the colony
groupings on a scatter plot. The clustering of 115 A.
florea colonies based on DFA and CVA showed similar
results. The colonies from Ilam, Khuzestan, and Bushehr
showed an overlapped cluster, while the colonies from
Hormuzgan state formed an apparent non-overlapping
cluster with the other states.
Multivariate Morphometric Study on Apis florea Distributed in Iran
96








Figure 2. Hind wing of A. florea showing the 7 characters examined.
Table 2. Mean and standard deviations of morphometric characters in honey bee populations from Iran (n: number
of colonies, measurements in mm, angle in degrees).
State Ilam Khuzestan Bushehr Hormuzgan
Character (n = 54) (n = 34) (n = 15) (n = 12)
FWL 6.801 ± 0.089 6.896 ± 0.055 6.859 ± 0.062 6.484 ± 0.028
FWW 2.409 ± 0.044 2.446 ± 0.033 2.434 ± 0.043 2.188 ± 0.014
WCL 0.647 ± 0.007 0.644 ± 0.014 0.646 ± 0.013 0.609 ± 0.003
WDL 1.264 ± 0.011 1.267 ± 0.017 1.270 ± 0.015 1.199 ± 0.017
CuA 0.532 ± 0.010 0.517 ± 0.007 0.555 ± 0.017 0.526 ± 0.001
CuB 0.179 ± 0.002 0.153 ± 0.004 0.169 ± 0.003 0.160 ± 0.007
HWL 4.937 ± 0.073 4.839 ± 0.043 4.842 ± 0.000 4.678 ± 0.024
HWW 1.490 ± 0.037 1.683 ± 0.195 1.466 ± 0.009 1.359 ± 0.018
HAM 11.776 ± 0.088 12.175 ± 0.120 11.868 ± 0.160 11.631 ± 0.004
FL 1.849 ± 0.020 1.838 ± 0.347 1.836 ± 0.013 1.799 ± 0.012
TL 2.313 ± 0.030 2.313 ± 0.114 2.307 ± 0.043 2.251 ± 0.002
MTL 1.427 ± 0.017 1.431 ± 0.241 1.421 ± 0.014 1.379 ± 0.001
MTW 0.642 ± 0.007 0.634 ± 0.204 0.641 ± 0.000 0.628 ± 0.000
A4 31.926 ± 0.264 31.329 ± 0.178 31.736 ± 0.055 33.115 ± 0.117
B4 99.817 ± 0.930 101.765 ± 0.182 97.169 ± 0.081 92.045 ± 0.167
D7 86.174 ± 0.546 87.895 ± 0.443 85.331 ± 0.185 81.097 ± 0.570
E9 19.147 ± 0.233 20.100 ± 0.099 20.038 ± 1.400 18.009 ± 0.491
G18 103.051 ± 1.529 103.655 ± 0.285 104.746 ± 2.483 107.792 ± 0.108
J10 40.061 ± 0.254 40.590 ± 0.317 40.319 ± 0.194 39.131 ± 0.668
J16 116.130 ± 0.887 118.610 ± 0.605 117.822 ± 1.261 112.689 ± 0.411
K19 70.118 ± 0.422 69.794 ± 0.967 69.544 ± 2.253 73.523 ± 0.735
L13 15.094 ± 0.201 14.897 ± 0.146 15.334 ± 0.021 16.117 ± 0.267
N23 76.701 ± 1.355 79.503 ± 1.102 75.050 ± 2.112 74.892 ± 0.819
O26 26.342 ± 0.579 26.714 ± 0.554 29.212 ± 0.496 29.755 ± 0.355
HW1 86.808 ± 1.241 88.320 ± 1.054 88.186 ± 0.627 86.287 ± 0.202
HW2 64.507 ± 0.265 64.069 ± 0.295 64.350 ± 0.496 63.528 ± 0.626
HW3 89.424 ± 0.295 90.311 ± 0.102 90.147 ± 0.397 90.407 ± 0.495
HW4 81.567 ± 0.174 80.774 ± 0.308 80.934 ± 0.330 80.075 ± 0.263
HW5 28.909 ± 0.023 29.323 ± 0.105 29.474 ± 0.005 28.951 ± 0.137
A. ÖZKAN, M. M. GHARLEKO, B. ÖZDEN, İ. KANDEMİR
97










!"


     
!" !"




!"
    
#





!"
Figure 3A. A scatter plot of 115 A. florea colonies from 4 states based
on CVA analysis.
Figure 3B. A scatter plot of 103 A. florea colonies (except Hormuzgan
colonies) based on CVA analysis.
Table 3. Test of equality of means of 29 morphometric characters.
Character Wilks’ Lambda F df1 df2 Sig.
FWL 0.642 18.740 3 101 0.000
FWW 0.625 20.165 3 101 0.000
WCL 0.791 8.907 3 101 0.000
WDL 0.766 10.280 3 101 0.000
CuA 0.815 7.624 3 101 0.000
CuB 0.778 9.583 3 101 0.000
HWL 0.792 8.815 3 101 0.000
HWW 0.714 13.456 3 101 0.000
HAM 0.860 5.490 3 101 0.002
FL 0.958 1.474 3 101 0.226
TL 0.949 1.820 3 101 0.148
MTL 0.871 4.991 3 101 0.003
MTW 0.965 1.212 3 101 0.309
A4 0.898 3.837 3 101 0.012
B4 0.680 15.822 3 101 0.000
D7 0.668 16.738 3 101 0.000
E9 0.756 10.871 3 101 0.000
G18 0.886 4.322 3 101 0.007
J10 0.929 2.574 3 101 0.058
J16 0.730 12.480 3 101 0.000
K19 0.923 2.802 3 101 0.044
l13 0.824 7.183 3 101 0.000
N23 0.823 7.245 3 101 0.000
O26 0.846 6.127 3 101 0.001
HW1 0.830 6.878 3 101 0.000
HW2 0.951 1.738 3 101 0.164
HW3 0.735 12.128 3 101 0.000
HW4 0.835 6.639 3 101 0.000
HW5 0.845 6.159 3 101 0.001
According to the results of multivariate statistical
analysis (DFA) the first, the second, and the third axis
explained 53.1%, 39.7%, and 7.2% of the total variation
respectively. One hundred percent of the total variation
was explained by the first 3 canonical variates. FWW, B4,
HWW, and FWL were the variables with the highest
loadings on the first canonical axis whereas, HW3, HWL,
and HW5 were loaded highly on the second canonical axis.
In the third axis, FWW, WCL, and FWL were the variables
contributing to the separation of the groups.
Table 4 shows correct classification results of 115 A.
florea colonies with respect to their original groupings
based on the CVA analysis. All the colonies from
Hormuzgan state were correctly classified to their
assigned groups. Colonies from Khuzestan and Ilam were
classified 94% to their pre-assigned groups, whereas
Bushehr colonies were classified 93% to their preassigned groups. The rate of correct classification of all
colonies to their original groups was 95%.
When multiple regression analysis was applied to the
morphometric variables using latitude and longitude as
independent variables, no correlation with latitude and
longitude was found (data not shown). Not all of the
morphometric variables turned out to be significantly
dependent on latitude and longitude, meaning that there
were no relation with morphometrical variables and the
geographical variables. According to the spatial
autocorrelation analysis, out of 180 Moran’s I coefficients
calculated for dataset (30 characters 6 distance classes),
51 were significant at 5% or less. Out of 30
correlograms, 14 correlogram were significant according
to the Bonferroni criterion (Table 5, Figure 4).
Linear Correlation Analysis between fore wing distance
D (WDL) and both wing lengths (fore and hind) and
between fore wing distance C (WCL) and both wing widths
(fore and hind) were carried out. The associations between
fore wing distance D (WDL) and both wing lengths (fore
and hind) and fore wing C distance and both wing widths
were high and significant (P < 0.05). The associations
between fore wing distance D (WDL) and both wing lengths
were found to be highly significant. The linear correlation
between WDL and FWL were 0.848, while correlation
between WDL and HWL were 0.816 (P < 0.01). Similarly,
the associations between fore wing distance C (WCL) and
both wing widths (fore and hind) were found to be highly
significant. The correlation between WCL and FWW were
0.801, while correlation between WCL and HWW were
0.774 (P < 0.01).
Discussion
In the present study, the extend of morphometric
variation was studied with respect to 29 morphometric
characters using 115 colonies collected from 4 states in
Iran on the coastal north-south diagonal. This study was
the most extensive study in terms of area coverage and
the number of morphometric characters used. New size
and angle characters utilized for the first time for the fore
and hind wings. The morphometric variation among 4
states and usefulness of the news characters were
discussed.
Distribution of Apis florea
Apis florea is another species in the genus Apis
distributed in Iran, in addition to Apis mellifera. Its
distribution starts from Qhasr-e-Shirin in Kermanshah
(Hashemi, 2004) and spreads to Ilam, Lorestan,
Khuzestan, Bushehr, Fars, Hormuzgan, Kerman, Sistan
va Baluchestan, and Boyarahmad va Kohgiluye in Iran
(Mossadegh, 1993; Ruttner et al., 1995). In addition,
Moradi and Kandemir (2005) stated that the distribution
may extend to Kurdistan states in the north. The present
study covers a vast area extending over 8 parallels, from
Hormuzgan (26°) in the south to Ilam (33°) in the north.
The multivariate morphometric analyses showed clear
groupings with respect to geography.
Geographic variation
Geography-related variation was pronounced in Apis
by Ruttner, who states that clinal type of geographical
variability exists in Apis florea, meaning that larger bees
found in the north and smaller ones in the south.
Tahmasebi et al. (2002) also demonstrated that a
western group of larger bees found at a higher latitude
(29°-34°) and an eastern group of smaller bees found at
a lower latitude (<29°) in Principal Components Analysis
in Iran. The finding is in accordance with the Bergman’s
rule about other animals. However, based on the multiple
regression analysis in this study, geography-related
Multivariate Morphometric Study on Apis florea Distributed in Iran
98
Table 4. Classification results [n (%)] of 115 A. florea populations
based on CVA analysis.
State Bushehr Khuzestan Ilam Hormuzgan
Bushehr 14 (93) 0 (0) 1 (7) 0 (0)
Khuzestan 1 (3) 32 (94) 1 (3) 0 (0)
Ilam 2 (4) 1 (2) 51 (94) 0 (0)
Hormuzgan 0 (0) 0 (0) 0 (0) 12 (100)
A. ÖZKAN, M. M. GHARLEKO, B. ÖZDEN, İ. KANDEMİR
99
   #
$
#





#
$


Figure 4. Significant autocorrelograms of A. florea morphometric characters (P < 0.05). Bold indicates the
average autocorrelation between neighboring localities.
Table 5. Moran’s I coefficients for 30 morphometric characters of A. florea.
Distance Classes
Character
1 2 3 4 5 6 P
FWL 0.04 0.13 0.02 0.15 -0.58* -0.26 0.082
FWW 0.16 0.17 0.04 0.19 -0.56* -0.49* 0.094
WCL 0.21 0.29* -0.55** 0.31* -0.04 -0.72** 0.004
WDL 0.10 0.23 -0.17 0.15* -0.31 -0.60** 0.022
CuA -0.08 0.07 -0.34 -0.16 0.02 -0.01 0.549
CuB 0.22 0.48** 0.09 -0.43* -0.17 -0.69** 0.006
C.IND 0.18 0.39* 0.16 -0.71** -0.10 -0.43* 0.004
HWL 0.18 0.09 -0.20 -0.01 0.12 -0.68** 0.005
HWW 0.07 0.18 0.06 0.03 -0.23 -0.61** 0.020
HAM 0.27 -0.34 -0.14 0.06 -0.43 0.07 0.311
FL -0.10 0.16 -0.22 0.11 -0.05 -0.39 0.350
TL -0.17 0.05 -0.10 0.19 -0.20 -0.27 0.514
MTL -0.04 0.15 -0.04 0.12 -0.20 -0.49* 0.111
MTW -0.13 -0.08 -0.09 -0.03 0.08 -0.24 1.000
A4 0.15 0.15 0.03 0.08 -0.59* -0.31 0.065
B4 0.39* 0.32* 0.26* -0.19 -0.47* -0.82** 0.000
D7 0.37* 0.26* 0.18 -0.09 -0.61** -0.61** 0.019
E9 0.20 -0.19 0.00 0.13 -0.62** -0.02 0.040
G18 -0.02 0.06 0.02 -0.26 0.18 -0.43* 0.128
J10 -0.04 0.10 -0.05 0.11 -0.37 -0.25 0.516
J16 0.19 0.05 -0.10 0.14 -0.76** -0.02 0.008
K19 0.22 0.07 -0.36 0.43** -0.27 -0.58** 0.029
I13 0.17 0.21 0.17 -0.07 -0.47* -0.51* 0.086
N23 -0.16 0.14 -0.05 -0.28 -0.06 -0.09 0.856
O26 0.47** 0.44** -0.08 -0.48* 0.12 -0.98** 0.000
HW1 -0.26 -0.31 0.19 -0.03 -0.10 0.02 0.470
HW2 0.11 -0.35 -0.04 -0.37 0.37* -0.23 0.103
HW3 0.40* -0.25 -0.47* 0.31* -0.17 -0.33 0.073
HW4 0.65** -0.66** 0.08 -0.26 0.36* -0.66** 0.004
HW5 0.60** -0.69** 0.07 0.04 -0.88** -0.35* 0.001
*significant at 5% level, **significant at 1% level.
variation was not found in A. florea colonies of Iran. On
the other hand, A. florea colonies from lower latitude
(Hormuzgan state) were morphologically distinguishable
from higher latitudes (Bushehr, Khuzestan, and Ilam
states). Based on morphometric measurements of size
characters (expect CuA, CuB, and TL), bees were smaller
in the Hormuzgan colonies compared with the colonies of
the other states (Table 2). Canonical Variates Analysis also
supported morphological differentiation in A. florea
colonies.
Cubital index was first used for the morphometrical
discrimination of Apis mellifera subspecies (Ruttner,
1988). Thus, Ruttner (1988) also used the same index
for A. florea and found 2.89 in southern Iran. In this
study, cubital index value varied, ranging between 2.82
(Khuzestan) and 3.30 (Hormuzgan) and there was no
relationship between cubital index and the latitude
according to multiple regression analysis (Table 5).
Besides, 3rd leg length of A. florea populations from
Hormuzgan (5.429) was similar to A. florea populations
from southern Iran (5.431) (Table 6).
In order to examine whether the observed variable
at one locality is dependent on the values of the
neighboring localities, we analyzed the geographic
variation. According to spatial autocorrelation analysis,
14 correlograms were significant according to the
Bonferroni criterion (Figure 4). Six characters (CuB,
C.Ind, HWW, B4, D7, and I13) showed clinal type
correlograms indicating positive autocorrelation in
short distance classes and negative autocorrelation in
long distance classes. For the characters CuB, C.Ind,
B4, D7, and I13, a positive autocorrelation existed at
the first 3 distance classes (Table 5, Figure 4), followed
by negative autocorrelations in the last 3 distance
classes. For HWW, a positive autocorrelation exists at
the first 4 distance classes, followed by negative
autocorrelations in the last 2 distance classes. Rest of
the characters showed more or less crazy quilt type
correlograms, meaning that there is no pattern
between neighboring localities.
Usefulness of new characters
Ruttner (1988) reported that Apis florea samples from
different regions vary not only in size characters, but also
in others, especially wing venations. With this approach,
angles of hind wing were first used for multivariate
morphometric study of A. florea colonies. Analysis of
variance of morphometric characters showed that 7 out of
the 8 hind wing variables displayed statistically significant
differences among populations (P < 0.05) (Table 3). In
addition, 3 characters of hind wing had high loadings on
the second canonical axis in multivariate statistical analysis.
This demonstrated that characters of hind wing were
valuable for discrimination of colonies.
Beside new hind wing characters (HWL, HWW,
HW1, HW2, HW3, HW4, and HW5), fore wing
distance D (WDL) and fore wing distance C (WCL) were
used in this study. Dedej and Nazzi (1994) reported
that WDL and WCL were used instead of fore wing
Multivariate Morphometric Study on Apis florea Distributed in Iran
100
Table 6. Some characteristic of A. florea of different origin (N: number of individuals, measurements in
mm).
Location N FWL FWW Cubital Index Hamuli 3rd Leg
Ilam 806 6.801 2.409 2.98 11.78 5.558
Khuzestan 354 6.896 2.446 2.82 12.17 5.581
Bushehr 206 6.859 2.434 3.29 11.87 5.564
Hormuzgan 178 6.484 2.188 3.30 11.63 5.429
*South Iran 60 6.706 2.313 2.89 11.37 5.431
*Oman 60 6.516 2.248 3.08 13.20 5.260
*Pakistan 40 6.598 2.316 2.85 - 5.202
*Sri Lanka 80 6.168 2.125 3.50 11.60 5.118
*South India 20 6.252 2.140 3.32 10.50 4.901
*Thailand 60 6.433 2.201 2.94 - 5.269
*adopted from Ruttner, 1988
length and fore wing width for Apis mellifera
subspecies to estimate the fore wing length and fore
wing width. Highly significant correlations between
WDL and the length, as well as between WCL and the
width, were in agreement with Dedej and Nazzi (1994)
(P < 0.01). Moreover, correlation analysis to
determine the associations between fore wing and hind
wing length and between fore wing and hind wing
width displayed significant associations between fore
wing and hind wing length (0.763), as well as between
fore wing and hind wing width (0.905) (P < 0.01).
These correlations revealed that one can use fore wing
- hind wing length and fore wing - hind wing width
characters interchangeably.
Beside size dependent morphometric characters,
geometric morphometric characters were also applied to
the same data set (unpublished data, manuscript in
preparation) and similar findings were found on the wing
shape of A. florea. However molecular tools should also
be applied along with size-dependent and size–
independent morphometric characters in order to
elucidate the population differentiation in A. florea.
Acknowledgements
This work was supported by research projects from
Zonguldak Karaelmas University grant no: BAP-2007-2-
1306-17.
A. ÖZKAN, M. M. GHARLEKO, B. ÖZDEN, İ. KANDEMİR
101
Adl, M.B.F., Gençer, H.V., Fıratlı, Ç. and Bahreini, R. 2007.
Morphometric characterization of Iranian (Apis mellifera meda),
Central Anatolian (Apis mellifera anatoliaca) and Caucasian (Apis
mellifera caucasica) honey bee populations). Journal of Apicultural
Research and Bee World 46: 225-231.
Arias, M.C., Sheppard, W.S. 2005. Phylogenetic relationships of honey
bees (Hymenoptera:Apinae:Apini) inferred from nuclear and
mitochondrial DNA sequence data. Molecular Phylogenetics and
Evolution 37: 25–35.
Chaiyawong, T., Deowanish, S., Wongsiri, S., Sylvester, H.A., Rinderer,
T.E. and De Guzman, L. 2004. Multivariate morphometric study
of Apis florea in Thailand. Journal of Apicultural Research 43:
123-127.
Clarke, G.M. and Oldroyd, B.P. 1996. The genetic basis of
developmental stability in Apis mellifera 2. Relationships between
character size, asymmetry and single-locus heterozygosity.
Genetica 97: 211-224.
Cornuet, J.M. and Garnery, L. 1991. Genetic diversity in Apis mellifera.
In: Diversity in the genus Apis (ed. D.R. Smith), Westview Press,
Boulder CO, pp. 103-115.
Cornuet, J.M., Garnery, L. and Solignac, M. 1991. Putative origin and
function of the intergenic region between COI and COII of Apis
mellifera L. mitochondrial DNA. Genetics 128: 393-403.
Crozier, R.H. and Crozier, Y.C. 1993. The mitochondrial genome of the
honey bee Apis mellifera: complete sequence and genome
organization. Genetics 133: 97-117.
Daly, H.V., Hoelmer, K. and Gambino, P. 1991. Clinal geographic
variation in feral honey bees in California, USA. Apidologie 22:
591-609.
Dedej, S. and Nazzi, F. 1994. Two distance of forewing venation as
estimates of wing size. Journal of Apicultural Research 33: 59-61.
Del Lama, M.A., Lobo, J.A., Soares, A.E.E. and Del Lama, S.N. 1990.
Genetic differentiation estimated by isozymic analysis of
Africanized honeybee populations from Brazil and from Central
America. Apidologie 21: 271-280.
Diniz-Filho, J.A.F., Malaspina, O. and Pignata, M.I.B. 1993. Geographic
variation in Apis cerana indica F.:a spatial autocorrelation analysis
of morphometric patterns. Journal of Apicultural Research 32:
65–72.
Diniz-Filho, J.A.F. and Malaspina, O. 1995. Evolution and population
structure of Africanized honey bees in Brazil: evidence from
spatial analysis of morphometric data. Evolution 49: 1172–1179.
Diniz-Filho, J.A.F., Hepburn, H.R., Radloff, S. and Fuchs, S. 2000.
Spatial analysis of morphometrical variation in African honeybees
(Apis mellifera L.) on a continental scale. Apidologie 31: 191-204.
Engel, M.S. 1999. The taxonomy of recent and fossil honey bees
(Hymenoptera: Apida); Apis. Journal Hymenopteran Research 8:
165–196.
Estoup, A., Garnery, L., Solignac, M. and Cornuet, J.M. 1995.
Microsatellite variation in honey bee (Apis mellifera L.)
populations: hierarchical genetic structure and test of the infinite
allele and stepwise mutation models. Genetics 140: 679-695.
Franck, P., Garnery, L., Loiseau, A., Solignac, M. and Cornuet, J.M.
2000. Molecular confirmation of a fourth lineage in honeybees
from Middle-East. Apidologie 31: 167-180.
Garnery, L., Cornuet, J.M. and Solignac, M. 1992. Evolutionary history
of the honeybee (Apis mellifera L.) inferred from mitochondrial
DNA analysis. Moleculer Ecology 3: 145-154.
Garnery, L., Solignac, M., Celebrano, G. and Cornuet, J.M. 1993. A
simple test using restricted PCR-amplified mitochondrial DNA to
study the genetic structure of Apis mellifera L. Experientia 49:
1016-1021.
References
Multivariate Morphometric Study on Apis florea Distributed in Iran
102
Garnery, L., Franck, P., Baudry, E. and Vautrin, D. 1998. Genetic
biodiversity of the West European honeybee (Apis mellifera
mellifera and Apis mellifera iberica), I. Mitochondrial DNA. Génét.
Sél. Evol. 30: 31-47.
Haddad, N., De Miranda, J.R. and Bataena, A. 2008. Discovery of Apis
florea in Aqaba, Jordan. Journal of Apicultural Research 47: 173-
174.
Hall, H.G. 1991. Genetics characterization of honeybees through DNA
analysis. In: The African honey bee (ed., M. Spivak, D.J.C. Fletcher
and M.D. Breed), Westview Press, Boulder CO, pp. 45-73.
Hall, H.G. 1992. Further characterization of nuclear DNA RFLP markers
that distinguish African and European honeybees. Arch. Insect
Biochem. Physiol. 19: 163-175.
Hashemi, M. 2004. The complete guide to bee keeping, 2nd Edition,
730p.
Hepburn, H.R., Radloff, S.E., Verma, S.E. and Verma, L.R. 2001.
Morphometric analysis of Apis cerana populations in the southern
Himalayan region. Apidologie 32: 435-447.
Hepburn, H.R., Radloff, S.E., Otis, G.W., Fuchs, S., Verma, L.R., Ken,
T., Chaiyawong, T., Tahmasebi, G., Ebadi, R. and Wongsiri, S.
2005. Apis florea: morphometrics, classification and
biogeography. Apidologie 36: 359-376.
Kandemir, I., Kence, M. and Kence, A. 2000. Genetic and morphometric
variation in honeybee (Apis mellifera L.) populations of Turkey.
Apidologie 31: 343-356.
Kandemir, İ., Kence, M., Sheppard, W.S. and Kence, A. 2006.
Mitochondrial DNA variation in honey bee (Apis mellifera L.)
populations from Turkey. Journal of Apicultural Research 45: 33-
38.
Makhmoor, H.D. and Ahmad, H. 1998. Biometric studies on four
species of honey bees in Jammu India. Indian Bee Journal 60:
141-142.
Moradi, M.G. and Kandemir, I. 2005. Observations on Apis florea “the
Dwarf Honey Bee” in Iran. American Bee Journal 145: 498-502.
Mossadegh, M.S. 1993. New geographical distribution line of Apis florea
in Iran. In: Asian Apiculture (ed. L.J. Conner, T. Rinderer, H.A.
Sylvester, S. Wongsiri), pp. 64-66.
Nazzi, F. 1992. Morphometric analysis of honeybees from an area of
racial hybridization in north-eastern Italy. Apidologie 23: 89-96.
Oden, N.L. 1984. Assessing the significance of a spatial correlogram.
Geographic Analysis 16: 1-16.
Otis, G.W. 1996. Distribution of recently recognized species of
honeybees (Hymenoptera: Apidae: Apis) in Asia. Journal of Kansas
Entomological Society 69: 311-333.
Radloff, S.E., Hepburn, H.R., Hepburn, C., Fuchs, S., Otis, G.W., Sein,
M.M., Aung, H.L., Pham, H.T., Tam, D.Q., Nuru, A.M. and Ken,
T. 2005. Multivariate morphometric analysis of Apis cerana of
southern mainland Asia. Apidologie 36: 127-139.
Rinderer, T.E., Oldroyd, B.P., Wongsiri, S., Sylvester, H.A., De Guzman,
L., Stelzer, J.A. and Riggio, R.M. 1995. A morphological
comparison of the dwarf honey bees of southeastern Thailand and
Palawan, Philippines. Apidologie 26: 387-394.
Rinderer, T.E., Oldroyd, B.P., De Guzman, L., Wattanachaiyingchareon,
W. and Wongsiri, S. 2002. Spatial distribution of the dwarf honey
bees in an agroecosystem in southeastern Thailand. Apidologie 33:
539-543.
Rohlf, J.F. 2004. Numerical Taxonomy and Multivariate Analysis
System, NTSYS pc, Version 2.20, Exeter Publishing, New York.
Ruttner, F. 1988. Biogeography and Taxonomy of Honeybees,
Springer-Verlag, Berlin.
Ruttner, F., Mossadegh, M.S. and Kauhausen-Keller, D. 1995.
Distribution and variation of size of Apis florea F in Iran.
Apidologie 26: 477-486.
Sheppard, W.S. 1988. Comparative study of enzyme polymorphism in
United States and European honey bee (Hymenoptera: Apidae)
populations. Entomological Society of America 81: 886-889.
Sheppard, W.S., Rinderer, T.E., Meixner, M.D., Yoo, H.R., Stelzer, J.A.,
Schiff, N.M., Kamel, S.M. and Krell, R. 1996. HinF1 variation in
mitochondrial DNA of Old World honey bee races. Journal of
Heredity 87: 35-40.
Smith, D.R. 1988. Mitochondrial DNA polymorphism in five Old World
subspecies of honey bees and in New World hybrids. In:
Africanized honey bees and mites (ed. G.R. Needham, R.E. Page,
M.Delfinado-Baker and C.E. Bowman), Ellis Horwood, Chichester
England, pp. 303-312.
Sokal, R.R. and Oden, N.L. 1978. Spatial autocorrelation in biology. 1.
Methodology. Biological Journal of the Linnean Society 10: 199-
228.
SPSS, 2004. SPSS for Windows, Release 13.0 Standard Version, SPSS
Inc., 1989-2004.
Tahmasebi, G., Ebadi, R., Tajabadi, N., Akhondi, N. and Faraji, S. 2002.
The effect of geographical and climatological conditions on the
morphometrical variation and separation of Iranian small
honeybee (Apis florea F.) populations. Journal of Science and
Technology of Agriculture and Natural Resources 6: 176-185.
Wartenberg, D. 1989. SAAP: Spatial Autocorrelation Analysis Program,
Exeter Publishing, New York.
Whitcombe, R.P. 1984. Bedouin bee. Saudi Aramco World 35: 34-40.