Morphological Diversity of Moroccan Honey Bees (Apis mellifera L. 1758): Insights from a Geometric Morphometric Study of Wing Venation in Honey Bees from Different Climatic Regions

Abstract

The morphological diversity of Moroccan honey bees (Apis mellifera) was investigated using geometric morphometrics to assess wing venation patterns among three populations representing three climatic zones: desert, semiarid, and Mediterranean. A total of 193 honey bee samples were analyzed and compared to historical reference samples from the Morphometric Bee Data Bank in Oberursel, representing the three subspecies: A. m. intermissa, A. m. sahariensis, and A. m. major. Principal component analysis and linear discriminant analysis revealed significant, yet overlapping morphological differences among the climatic groups. Spatial modeling showed a significant southwest–northeast clinal gradient in wing morphology. Almost all samples were assigned to the African evolutionary lineage, except one, suggesting a dominant African genetic background. Interestingly, all three populations showed greater morphological affinity to A. m. intermissa than to A. m. sahariensis, which could indicate introgression or limitations in the current reference dataset. These discrepancies highlight the necessity of revising subspecies boundaries using updated morphometric and genomic approaches. These findings improve our understanding of honey bee biodiversity in Morocco and provide valuable information for conservation and breeding programs.

1. Introduction

The Moroccan bee fauna exhibits remarkable diversity, ranking the country as the second richest in Africa and the fifth in the Mediterranean region [1]. These vital pollinators play a crucial role in supporting both natural and agricultural ecosystems by pollinating a wide variety of wild plants and crops [2,3]. Their pollination services are of significant economic importance, contributing an estimated value of 1.2 billion USD to the country’s economy [4]. The most well-known bee species is the western honey bee (Apis mellifera), which is native to Africa, Europe, and western Asia [5]. This species has been subdivided into numerous subspecies or geographic races, each of which displays specific genotypic, phenotypic, and behavioral characteristics that have been long shaped by natural selection [5,6,7]. Currently, we count about thirty-three distinct honey bee subspecies that were grouped into seven evolutionary lineages [8,9].

Moroccan honey bee populations belong to the African (A) evolutionary lineage [10,11,12]. Morphometric and genetic studies have grouped the Moroccan honey bee into two main subspecies [5]. The first one, A. m. intermissa, is a specific subspecies of the north African countries that is present in Morocco, Algeria, and Tunisia [13,14]. It is known for its aggressive behavior and ability to swarm [5]. The second one, A. m. sahariensis, is distributed in the Saharan areas to the south of Morocco and Algeria [5,15]. It is known for being less aggressive and for its adaptation to the extreme climatic conditions of the area [5]. In addition to these two subspecies, Ruttner [5] suggested the possibility of a third one, which he called A. m. major. This subspecies is localized exclusively in the Northern Rif region near the Mediterranean coast at approximately 35° latitude [13]. The A. m. major was described as a local adaptation of A. m. intermissa to the unique climatic conditions of a specific area in the Rif region [5]. A. m. major exhibited larger forewing sizes (9.52 mm, exceeding those of A. m. intermissa), longer tongues (6.88–7.13 mm), a smaller tomentum (0.59–0.76 mm), and a mean cubital index of 2.24 [13].

To safeguard the biodiversity within honey bee populations, conservation must begin with their accurate identification. In Morocco, however, honey bee populations remain relatively understudied. Large-scale research aimed at characterizing Moroccan honey bee subspecies dates back more than two decades [10,13], with few instances of recent research [16,17]. The pioneering study of Cornuet et al. [13] identified three distinct subspecies grouped through classical morphometric analysis. To further validate these findings, a mitochondrial DNA analysis focusing on the intergenic COI-COII region was later conducted [10]. Interestingly, the genetic analysis revealed the presence of two main, closely related groups: one corresponding to the southern Saharan subspecies (A. m. sahariensis), the other encompassing the Tellian subspecies (A. m. intermissa). This genetic distribution differed from the three-group structure observed in the earlier morphometric study, highlighting discrepancies between morphological and molecular approaches. Such divergences emphasize the complexity of honey bee population dynamics in Morocco and the need for more comprehensive studies to resolve these taxonomic ambiguities and guide effective conservation efforts.

This study aimed to investigate the morphological diversity of Moroccan honey bee populations across different climatic zones (desert, semiarid, and Mediterranean) using geometric morphometric analysis of wing venation. The main objectives were to (i) assess the extent of morphological differentiation among honey bee populations from distinct environmental contexts, (ii) compare current populations with reference subspecies (A. m. sahariensis, A. m. intermissa, and A. m. major), and (iii) evaluate the spatial structuring of wing shape variation and its potential environmental and anthropogenic drivers.

2. Materials and Methods

2.1. Sampling

Samples were collected from three different climatic zones in Morocco (Figure 1). The climatic groups were selected based on the Köppen–Geiger climate classification [18]. According to this classification, Morocco has five climates: warm desert, warm semiarid, cold desert, cold semiarid, and warm Mediterranean. A large part of the study area belongs to the warm Mediterranean climate in the north and the warm desert climate in the south. Between these two regions, there is a relatively narrow belt of the other three climates, which we treated as one group to obtain a similar number of samples per climate. We refer to the three climatic groups as desert (warm desert climate), semiarid (combined warm semiarid, cold desert, and cold semiarid climates), and Mediterranean (warm Mediterranean climate). The Mediterranean group occurs mainly in the north and along the Atlantic and Mediterranean coasts and in mountainous regions. It is characterized by cool to moderately warm winters and hot, dry summers. The semiarid group is distributed across the interior plateaus and the Atlas Mountains and is characterized by low and variable annual precipitation. The desert group dominates the southern and southeastern regions, characterized by extremely low rainfall and very high summer temperatures.

Sampling was carried out at 121 locations, which represented three climates: desert (68 samples), semiarid (68 samples), and Mediterranean (57 samples) (

Table 1. Sampling data of Moroccan honey bees, as well as reference subspecies used in the study.

Subspecies	Number of Colonies/Sites	Number of Honey Bees	Total Images
Mediterranean	57	750	1641
Semiarid	68	624	900
Desert	68	974	925
A. m. intermissa	19	229	229
A. m. major	1	10	10
A. m. sahariensis	4	40	40
Total	217	2627	3745

). A total of 193 honey bee samples were collected to ensure geographic and ecological coverage across the country. The dataset included older material (110 samples) collected in 2022 and already published by Aglagane et al. [16], as well as newly collected material (83 samples) from 2023 to 2024 that have not yet been published (Table S1). These additional samples were added to extend coverage to more climates and territories in Morocco.

At each location, the worker bees were collected either from individual hives maintained by local beekeepers or directly from flowers using an insect net (Table S1). For comparative purposes, historical reference samples of honey bees were included to represent the three Moroccan subspecies. These consisted of nineteen colonies of A. m. intermissa, four colonies of A. m. sahariensis, and one colony of A. m. major. The reference samples were sourced from Ruttner’s reference collection curated at the Morphometric Bee Data Bank in Oberursel, Germany. These historic samples were collected between 1950 and 1976.

2.2. Geometric Morphometric Analysis

The right and left forewings (when possible) of each worker honey bee were detached at their base using forceps. The wings were then mounted between two microscopic slides and photographed using various equipment (Table S1). Wing images of the recent samples were obtained using an SZX7 Stereomicroscope (Olympus, Tokyo, Japan) (0.8×–5.6×) equipped with a digital camera type X7CAM4K8MP (ToupTek, Hangzhou, China). In all samples, nineteen landmarks were digitized using the software IdentiFly v.1.8.0 following the order displayed by Nawrocka et al. [19]. The raw landmark coordinates were then exported and stored in CSV format for subsequent geometric morphometric analyses. To eliminate potential inconsistencies caused by variations in scale, position, and orientation, the landmark coordinates were subjected to the Generalized Procrustes Analysis (GPA) in the geomorph package (v. 4.0.4) [20,21,22]. The aligned coordinates were averaged within colonies/localities, and the averages were used in the subsequent analysis.

2.3. Statistical Analysis

To explore patterns of morphological variation, principal component analysis (PCA) was performed to identify the main axes of shape variation across samples. Wing shape was described with the first 34 principal components. Group differentiation in wing shape was further assessed using linear discriminant analysis (LDA) with the Morpho package (v. 2.12), allowing clear visualization of shape differences among predefined groups. In the case of A. m. major, the only specimen was projected into the space of the linear discriminants. Mahalanobis distances were calculated to quantify pairwise shape differences between groups, and their statistical significance was tested using multivariate analysis of variance (MANOVA). To assess the spatial structure of shape variation, the association between principal component scores and geographic coordinates (latitude and longitude) was examined using generalized additive models (GAMs) with the mgcv package (v. 1.9.3). Evolutionary lineage assignment was carried out by comparison with a reference dataset from Nawrocka et al. [19], comprising four main evolutionary lineages: A (85 colonies), C (37 colonies), O 49 colonies), and M (colonies). All statistical analyses were conducted in RStudio (version 4.5.0) and are available as Supplementary Document S1.

3. Results

The analysis of wing shape revealed significant morphological differences among the studied climatic groups, as confirmed by multivariate analysis of variance (MANOVA: F = 4.6, p < 0.0001). The principal component analysis indicated that the first two principal components accounted for 25.6% and 12.0% of the total shape variance, respectively (Figure 2a). The scatterplot showed considerable overlap among groups, suggesting a degree of morphological similarity. Pairwise Mahalanobis distance comparisons showed significant differences between all groups (

Table 2. Mahalanobis distances (lower triangle) and significance of pairwise statistical differences (upper triangle) in wing shape of climatic groups of Moroccan honey bees.

	Desert	Semiarid	Mediterranean
Desert	0.00	<0.0001	<0.0001
Semiarid	1.58	0.00	<0.0001
Mediterranean	3.51	3.22	0.00

). The greatest Mahalanobis distance was observed between Mediterranean and desert groups (3.51), while the smallest was found between semiarid and desert groups (1.58; Table 2). Linear discriminant analysis provided an enhanced separation among the groups (Figure 2b). Here, the desert, semiarid, and Mediterranean samples are more distinctly clustered, with reduced overlap. The overall accuracy of the discrimination between climates using cross-validation was 80.31%.

The spatial distribution of the first two principal components was visualized using generalized additive models. For PC1, the model revealed that the variation in wing shape was strongly associated with geographic location (F = 9.08, p < 0.0001). The fitted values of PC1 displayed a clear increasing gradient from southwest to northeast across the sampling range (Figure 3). Interestingly, samples from southeastern Morocco representing the desert climatic group also showed relatively higher values. On the other hand, PC2 exhibited no significant spatial structure across Moroccan locations (F = 0.9, p = 0.47).

Comparing the studied samples to evolutionary lineage references revealed a strong morphological affinity with lineage A (the African lineage), as shown in Figure 4. The vast majority of colonies 192) were classified as belonging to lineage A, while only one colony from the desert climatic group was classified as belonging to lineage O.

When comparing wing shape variation of Moroccan honey bees from climatic groups and the three reference subspecies, differences were statistically significant (MANOVA: F = 4.79, p < 0.0001. The first two principal components explained 22.7% and 13.2% of total variance, respectively, and showed overlapping groups (Figure 5). Pairwise Mahalanobis distance comparisons showed significant differences between all groups (

Table 3. Mahalanobis distances (lower triangle) and significance of pairwise statistical differences (upper triangle) in wing shape of climatic groups in comparison with reference subspecies.

	Desert	Semiarid	Mediterranean	A. m. intermissa	A. m. sahariensis
Desert	-	0.0053	0.0001	0.0001	<0.0001
Semiarid	1.47	-	0.0001	0.0001	<0.0001
Mediterranean	3.48	3.17	-	<0.0001	<0.0001
A. m. intermissa	6.01	6.00	5.56	-	<0.0001
A. m. sahariensis	8.25	8.63	8.25	6.62	-

). Considering only the comparison with reference subspecies, the greatest Mahalanobis distance was observed between the semiarid group and A. m. sahariensis (8.63), while the smallest was found between the Mediterranean group and A. m. intermissa (5.56; Table 3). Nevertheless, the linear discriminant analysis revealed deeper differences, clearly separating groups with minimal overlap (Figure 6). The first linear discriminant clearly shows two clusters of points; the first cluster represents the three climatic groups, and the second cluster groups the three reference subspecies (Figure 6a). The third linear discriminant separates A. m. sahariensis and A. m. major from the rest of the samples (Figure 6b). The overall accuracy of discrimination using cross-validation reached 78.24%. Interestingly, the one colony sample representing A. m. major was classified as A. m. sahariensis with a high probability.

4. Discussion

Our results showed that nearly all of the analyzed honey bee samples belong to lineage A, which is consistent with previous genetic studies of Moroccan populations [11,12,13,17,18]. Only one colony from the desert climate group was assigned to lineage O, which confirms earlier observations reported by Aglagane et al. [16]. This result aligns with the observation of a high degree of wing venation similarity between the two lineages [19]. However, previous studies using single-nucleotide polymorphisms (SNPs) have shown a clear distinction between the O and A lineages [9,23,24]. Additionally, microsatellite analyses conducted on the same Moroccan population revealed no admixture from other lineages, eliminating the possibility that the single colony was introduced by beekeepers from Southwest Asia [17]. Therefore, we assume that the single colony was misclassified as lineage O, which may be due to the inadequate size of the reference samples. Identifying honey bee lineages using wing measurements may become more reliable in the future with larger and more representative reference samples.

A spatial analysis of wing shape revealed a significant southwest–northeast gradient (Figure 3). A similar spatial structuring was observed in a previous study [10], in which the geographic variation of Moroccan honey bees, as determined by mtDNA data, was associated with a southwest–northeast cline. The observed geographical variations in wing morphology can be attributed to environmental selection pressures. This morphological divergence aligns with the well-documented phenomenon of clinal variation in A. mellifera, where populations exhibit gradual morphological changes along environmental gradients [5,15,25]. In the Moroccan context, a pronounced south-to-north clinal gradient has been reported for several morphological traits, including proboscis length, forewing length, hind leg dimensions, and the cubital index [5,13]. These patterns suggest that local environmental conditions play an important role in shaping the morphological architecture of honey bee populations. This is likely an adaptive response to regional ecological pressures [5,26].

Despite the overlap among the three climatic groups in Morocco, honey bee populations sampled across these groups exhibited statistically significant morphological differentiation (Figure 2a,b). This pattern was evident in the relatively low yet statistically significant Mahalanobis distances (Table 3). Since there are no barriers to gene flow between the climates, overlap in wing shape among the three groups is expected. While climate changes mainly from north to south, the spatial pattern of wing shape variation is more complex (Figure 3). This suggests that wing shape variation in Morocco is related not only to climate but also to other factors, including evolutionary history and beekeeping practices.

The wing shape of the honey bee samples investigated differed markedly from the subspecies reference. Similar findings were previously reported in Romania, where historical and contemporary samples significantly differed [27]. Interestingly, our finding that honey bee samples from all three climatic zones were more similar to A. m. intermissa than to A. m. sahariensis is consistent with the genetic evidence reported by Aglagane et al. [17]. They demonstrated extensive introgression between the two subspecies, which was more prominent within the natural range of the Saharan honey bee. Previous studies have shown that, although geographic barriers such as the High Atlas Mountains can restrict gene flow [10,13], intensive beekeeping practices and transhumance in recent decades have accelerated hybridization in Moroccan bee populations [16,17]. Together, these patterns suggest that wing shape variation reflects not only evolutionary history but also ongoing genetic mixing driven by anthropogenic factors [28,29].

Another possible explanation for the discrepancies between the historical and contemporary samples is the limitations of the reference dataset used for subspecies comparison. The reference material for A. m. intermissa and A. m. sahariensis obtained from the Oberursel database does not comprehensively represent the full geographic and ecological variability present within Morocco. Notably, only four reference samples of A. m. sahariensis were available, which is insufficient to capture the intra-subspecific diversity of these subspecies. Similarly, only six out of the 19 A. m. intermissa reference samples were collected from Morocco, while the remaining 13 were collected from neighboring countries like Algeria and Tunisia. These external samples may not fully reflect the morphological traits of Moroccan A. m. intermissa, given regional differentiation within subspecies. As a result, the mismatch between our field samples and reference profiles could partially be attributed to incomplete or geographically biased reference data [5,16,28]. These findings raise critical questions about the adequacy of reference datasets as a baseline reference for honey bee subspecies identification, highlighting the need for updated and locally representative reference collections when conducting morphometric comparisons.

The only available reference sample of A. m. major, collected by Brother Adam in 1976 [19], was more similar to A. m. sahariensis than to A. m. intermissa. This suggests that the sample may have been mislabeled. In general, information about A. m. major is limited, and it does not show how it could be distinguished from other subspecies. Ruttner himself considered A. m. major as a “possible subspecies” [5]. Additionally, an earlier study considered this subspecies a synonym of A. m. intermissa [30]. Its distribution range was restricted to a small area that is not isolated from surrounding populations by any barriers to gene flow. Therefore, it can be concluded that A. m. major should not be considered a valid subspecies. This conclusion is based on only one colony. However, no additional material is available. We collected contemporary samples from locations where A. m. major should be present (Figure 1), but none of them were similar to the historic sample.

The conservation of honey bee subspecies relies heavily on their accurate identification, which is essential for developing and implementing effective management and protection strategies. In this study, we investigated wing venation patterns to assess morphological variation among honey bee populations sampled from regions corresponding to three climatic zones. The results obtained here indicate the need for further investigations combining morphological and genetic approaches to better understand the patterns of Moroccan honey bee diversity. Integrating geometric morphometrics with advanced genomic tools, such as SNPs analysis and whole-genome sequencing, would provide deeper insights into the evolutionary history, population structuring, and hybridization dynamics of these honey bee populations. Additionally, incorporating environmental and ecological data could help disentangle the relative contributions of genetic makeup and environmental influences on wing venation patterns. Future studies should also focus on expanding sampling efforts across broader geographic regions and over longer time scales to assess potential shifts in honey bee morphology in response to climate change, human-mediated introductions, and changing apicultural practices. Such multidisciplinary research is crucial for developing effective conservation and management strategies aimed at preserving the genetic integrity and adaptive potential of Morocco’s native honey bee populations.

5. Conclusions

This study demonstrates that Moroccan honey bee populations vary geographically, especially along the southwest–northeast gradient. At least some of this variation can be attributed to environmental conditions, as significant differences were observed between desert, semiarid, and Mediterranean climates. Contemporary Moroccan honey bees differ from historic reference samples of A. m. sahariensis and A. m. intermissa. This difference is most likely related to migratory beekeeping and the importation of non-native genetic material by beekeepers. These findings highlight the need for effective biodiversity conservation strategies for Morocco’s native honey bee populations.