Published online January 17, 2025
https://doi.org/10.5141/jee.24.071
Journal of Ecology and Environment (2025) 49:01
Ehsan Rahimi1 and Chuleui Jung1,2*
1Agricultural Science and Technology Institute, Andong National University, Andong 36729, Republic of Korea
2Department of Plant Medical, Andong National University, Andong 36729, Republic of Korea
Correspondence to:Chuleui Jung
E-mail cjung@andong.ac.kr
This article is licensed under a Creative Commons Attribution (CC BY) 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ The publisher of this article is The Ecological Society of Korea in collaboration with The Korean Society of Limnology
Background: Among the 22 species in the Vespa genus, five have successfully established populations outside their native ranges, while four others have been recorded either in natural habitats or during border inspections in various countries. This study aims to assess the potential threat posed by 12 hornet species—Vespa crabro, Vespa mandarinia, Vespa simillima, Vespa velutina, Vespa affinis, Vespa analis, Vespa basalis, Vespa bicolor, Vespa ducalis, Vespa dybowskii, Vespa soror, and Vespa tropica—within the geographical and ecological context of Iran, an arid Middle Eastern country. Using ecological niche modeling, we analyzed species occurrence data alongside climatic variables with minimal correlation to predict the potential distribution of these hornets across Iran. The mobility-oriented parity method was applied to identify areas where strict extrapolation is relevant for these species. Additionally, we generated a habitat suitability map for Apis mellifera (honey bee) using ecological niche modeling and compared the spatial overlap between the predicted risk maps for the hornets and the honey bee habitat suitability map, employing Schoener’s D metric.
Results: The results revealed two key findings. First, a significant portion of Iran exhibits climatic dissimilarity compared to the native habitats of certain hornet species. Second, the spatial overlap analysis showed varying degrees of overlap between A. mellifera habitats and the potential distributions of different hornet species. Notably, V. mandarinia and V. crabro demonstrated the highest overlap values (D = 0.68), suggesting that these hornets could share substantial habitat preferences or ecological roles with honey bees in Iran.
Conclusions: Although most regions of Iran appear less suitable for hornet invasions, caution is warranted in the northern areas, where trade and exchanges could serve as pathways for Vespa hornet introductions. These findings highlight the importance of targeted monitoring and preventative measures in these high-risk regions.
Keywords: ecological niche modeling, hornets, invasive species, Iran, mobility-oriented parity
The economic value of animal pollination as an ecosystem service in global agriculture ranges from $195 billion to $387 billion (Porto et al. 2020). Additionally, it is estimated that pollinators play a direct role in supplying up to 40% of the essential nutrients in the human diet (Eilers et al. 2011). Insect pollination alone contributes to 9.5% of the total economic worth of agricultural products that are directly consumed by humans (Gallai et al. 2009). Research conducted by Rodger et al. (2021) has demonstrated that a reduction in pollinators can result in a 50% decrease in seed-based reproduction for approximately one-third of flowering plants, indicating that a substantial number of flowering plants rely entirely on pollinators for seed production.
Pollinating insects encompass various groups such as moths, butterflies, bumblebees, honey bees, solitary bees, and hoverflies. Among these, bees hold particular importance as they are accountable for pollinating approximately 35% of the world’s food production (Elias et al. 2017). Bees, categorized as ectothermic creatures, exhibit a significant reliance on the temperature of their habitat for their functioning. Honeybees and bumblebees are known to visit over 90% of global food crops (Doyle et al. 2020). Nevertheless, hornets can pose significant threats to these bee species, especially honeybees. Hornets belonging to the
One of the most notorious invasive insect species is the hornet, such as
While invasive species are usually constrained by their ability to disperse and find suitable new habitats, vespid hornets stand out due to their remarkable invasive achievements and impressive dispersal capabilities (Otis et al. 2023). The genus Vespa comprises a total of 22 hornet species, all of which are found in various regions of Asia (Archer 2012; Rahimi and Jung 2024a). Only two species,
The tendency of species to retain aspects of their fundamental niche over time is called niche conservatism (NC) (Wiens and Graham 2005). If species’ fundamental niches are conserved, species will only be able to invade regions that have a climate similar to that of their native range (Tirozzi et al. 2022). Given climatic NC, the distribution of species in their native ranges may predict where they can successfully invade and subsequently spread (Bock et al. 2017; Wiens et al. 2010). For biogeographic hypotheses, a key idea is that climatically unsuitable conditions can limit geographic ranges when there is NC, and such conditions can potentially be identified and tested using species distribution models (SDMs) (Strubbe et al. 2013). For example, a hypothesis of climatic NC predicts that invasive species will spread primarily in regions that are climatically similar to their native range (Wiens et al. 2010).
Ecological niche models (ENM) have been employed to forecast the potential range of other Vespa species (Alaniz et al. 2021; Barbet-Massin et al. 2013; Barbet-Massin et al. 2018; Bessa et al. 2016; Herrera et al. 2023; Keeling et al. 2017; Moo-Llanes 2021; Nuñez-Penichet et al. 2021; Zhu et al. 2020). For example, Nuñez-Penichet et al. (2021) found that regions highly suitable for
Iran has been identified as home to a variety of pollinating bees across different regions. Among these,
Iran’s ecological landscape is categorized into three distinct phytogeographical regions (Fig. 1) as identified by Talebi et al. (2014): the Euxino-Hyrcanian region, the Saharo-Sindian region, and the Irano-Turanian region. Furthermore, ecologists have classified Iran’s forests into three ecological zones, namely the Caspian or Hyrcanian zone, the Khalijo-Omanian zone, and the Iranian-Turanian zone. These are further subdivided into the Zagros mountainous zone and the central plateau zone. Here, we provide a brief overview of these five ecological regions in Iran, as depicted in Figure 1.
Table 1 presents an overview of the extent and distribution of Iran’s natural resources. As per the data in this table, approximately 81% of Iran’s land area is occupied by various natural features, including forests, deserts, rangeland, and bushes. Notably, rangelands account for nearly half of the country’s land area, with a significant portion characterized as poor quality. Deserts encompass roughly 20% of Iran’s territory, predominantly situated in the central, eastern, and southeastern regions, primarily within the Irano-Turanian ecological region.
Table 1 . Area and proportion of natural resources in Iran (Talebi et al. 2014).
Natural resources | Area (ha) | Proportion (%) |
---|---|---|
Natural forest | 13,364,010 | 8.10 |
Man-made forest | 946,546 | 0.57 |
Bush and woodland | 2,723,756 | 1.65 |
Rangeland | 84,960,321 | 51.60 |
Desert | 32,863,972 | 19.94 |
Total | 134,884,365 | 81.85 |
Two species of hornets are present in Iran:
The species selected for niche overlap analysis in this study include
Bioclimatic layers as predictor variables also were downloaded from the WorldClim database (www.worldclim.org). The Bioclimatic data in the WorldClim database includes 11 temperature variables and 8 precipitation variables, which have a spatial resolution of approximately 4 km. These 19 variables often have a high correlation with each other and therefore it is not recommended to use all these variables in species distribution modeling. For this purpose, we used the usdm (Naimi 2023) package to exclude the highly correlated variables from the set through a stepwise procedure based on variance inflation factor. The remaining variables include Mean Diurnal Range (Bio 2), Temperature Seasonality (Bio 4), Mean Temperature of Wettest Quarter (Bio 8), Mean Temperature of Driest Quarter (Bio 9), Precipitation Seasonality (Bio15), and Precipitation of Coldest Quarter (Bio 19).
In this study, we employed the SDM package (Naimi and Araujo 2018) within the R software environment to model the distribution of
Williams et al. (2007) pointed out that the availability of data can hinder the extrapolation to unfamiliar environments. This limitation arises from the fact that the species’ niche may not be entirely captured or represented in the available data. Depending on the direction of environmental changes, certain aspects of the species’ niche that have not been observed in the existing data may become relevant or newly significant. The advancement of techniques for handling uncertainty has not kept pace with the widespread use of SDMs (Mesgaran et al. 2014). Recently, several valuable approaches have been proposed to detect and visually represent new environmental conditions (Elith et al. 2010; Mesgaran et al. 2014; Owens et al. 2013). New environmental conditions can be classified into two categories: 1. For a specific individual variable, the values may fall outside the range covered during training, which is referred to as univariate or strict extrapolation. 2. Certain areas in the environmental space may lie within the range of individual variables but constitute new combinations of predictors, known as multivariate or combinational extrapolation (Zurell et al. 2012).
To manage this risk and identify analogous environments, we can employ three different approaches by comparing native and invaded ranges: a) Analysis of multivariate environmental similarity surface (MESS) (Elith et al. 2010), which provides a measure of how environmentally similar each location is to the median of the most dissimilar variable. b) Mobility-oriented parity (MOP) (Owens et al. 2013), a method that pinpoints areas of strict extrapolation and quantifies the environmental similarity between the calibrated and projected regions. c) Extrapolation detection (ExDet) (Mesgaran et al. 2014), a technique that detects similarities or novel environmental conditions between native and invaded areas.
In this study, we employed MOP analysis because the MESS tool identifies extrapolation or ‘dissimilar’ points based solely on the ranges of individual (univariate) predictors. It does not consider the correlation structure, which means it doesn’t account for new multivariate combinations of the various covariates that might be included in the model.
Strict extrapolation occurs when the environmental conditions in the area of interest (new areas) fall completely outside the range of conditions observed in the reference or calibration area (original area). A MOP value of 0 for a specific location (or grid cell) in the area of interest indicates that at least one environmental variable at that location lies entirely outside the range observed in the reference area. This suggests the presence of novel environmental conditions that the model has not encountered before. As a result, predictions for these areas are highly uncertain and potentially unreliable due to the absence of prior data on these conditions. On the other hand, a MOP value of 1 indicates that the environmental conditions at a specific location in the area of interest are identical to those found in the reference area. In such cases, the model has already been trained on these conditions, allowing it to make predictions with a higher degree of confidence. Therefore, predictions in these regions are more dependable as they are based on familiar environmental conditions (Qiao et al. 2019; Rahimi et al. 2024; Velazco et al. 2024).
To perform this task, we will need to use the NicheToolBox (ntbox) R package (Osorio‐Olvera et al. 2020). The mop function in this package requires two types of raster stacks: 1. ‘M_stack’: that is a RasterStack (bioclimatic variables) containing variables that represent the calibration area (hornets range). In the context of ENM, this typically means the region where we have species occurrence data and environmental variables that are used to train the model. 2. ‘G_stack’: This is another RasterStack (bioclimatic variables) containing variables that represent the areas or scenarios to which our ENM models will be transferred (Iran). To establish the calibration area, we must account for a zone surrounding the hornets’ presence locations. As an illustration, these hornets are known to inhabit areas up to 100 km (Moo-Llanes 2021) from their occurrence points, but in this case, we’ve opted for a 50 km buffer around each point and masked the bioclimatic variables in ArcGIS software. Our analysis encompasses all recorded instances of hornet presence, both within their natural habitats and regions outside their native range.
We employed Schoener’s
Which
To evaluate the performance of the results obtained from the MaxEnt model for
Table 2 shows the results of model validation metrics for different families based on statistics of AUC and TSS. To measure the values of these statistics, the output of all studied species in different families was evaluated, and for each family, the average AUC and TSS values of all species were reported. According to this table, the value of AUC for all families is between 0.91 and 0.95, which indicates that the results of the models are excellent. According to the TSS test, our results also fall within the perfect prediction range.
Table 2 . Model validation metrics including TSS and AUC for
Name | AUC | TSS |
---|---|---|
0.97 | 0.82 |
TSS: true skill statistic; AUC: area under the ROC curve.
As noted earlier, MOP analysis enables the comparison of environmental conditions between a reference set (South Asia) and an area of interest (Iran). The primary goal of MOP analysis is to identify non-analogous conditions in the area of interest relative to the reference set and to quantitatively evaluate the degree of dissimilarity in these conditions. This approach helps identify potential habitats where hornets could establish successfully based on their original habitat preferences. Predictions for areas with high values of MOP are more reliable due to strong data support from the model. However, low MOP values do not necessarily indicate that an area is unsuitable; instead, they highlight increased uncertainty, making invasion outcomes less predictable.
Figure 2 presents the maps illustrating the risk of hornet invasion in Iran. In these maps, the blue regions with values near zero indicate areas with low climatic suitability for hornets. Conversely, the brighter colors represent higher values, signifying a more favorable climate for these species. A brief examination of these maps reveals two notable observations. Firstly, for certain hornet species, a significant portion of Iran exhibits a dissimilar climate compared to their native habitats. Secondly, none of the cells in the maps have values exceeding 0.7 for any of the species. Specifically, the distribution of
The spatial overlap analysis between
Based on the spatial overlap matrix (Fig. 3), the hornet species that exhibit the highest ecological overlap with each other are
Figure 4 presents both the composite risk map for hornets’ invasion and the climatic suitability map for
Our findings indicate that hornets have limited potential to inhabit specific areas in both the northern and southern regions of Iran. It’s important to note that even in these regions, the climatic suitability for hornets is not exceptionally high (close to 1). Iran is home to two hornet species, namely
Nevertheless, some studies have focused on species distribution modeling of hornets, such as
Others have conducted niche overlap analyses comparing native and non-native ranges of hornets to determine if this species exhibit NC. For example, Barbet-Massin et al. (2018) tried to determine whether
Lioy et al. (2023) also assessed the extent of niche overlap between
Multiple records of nine Vespa species occurring well beyond their native ranges (Otis et al. 2023) provide compelling evidence that hornets have been widely introduced to non-native regions due to human activities. Most hornets are generalist predators with a broad range of habitats, displaying significant adaptability in their behavior as eusocial insects. Additionally, their large size and potent stings make them formidable competitors in various environments. Predictive modeling of potential habitat for hornet species has been performed specifically for
Not applicable.
NC: Niche conservatism
SDM: Species distribution model
ENM: Ecological niche model
MESS: Multivariate environmental similarity surface
MOP: Mobility-oriented parity
TSS: True skill statistic
AUC: Area under the ROC curve
ER has written the paper and has done the modeling part of the analysis. CJ has reviewed the paper and interpreted the results and final edition. All authors have read and agreed to the published version of the manuscript.
This work was supported by a Research Grant of Andong National University (2023–2024).
The datasets used and/or analyzed in the current study are available from the corresponding author on reasonable request.
The authors declare that they have no conflict of interest.
Not applicable.
Not applicable.
The authors declare that they have no competing interests.
![]() |
![]() |
![]() |
![]() |
![]() |
![]() |
Research 2024-12-12 49:47
Assessing the invasive risk of Procambarus virginalis (marbled crayfish) in South KoreaHyungsoon Jeong* and Ju Hui Choi
Research 2024-02-07 49:07
Terrestrial pest gastropod diversity and spatiotemporal variations in highland agricultural lands of Sri LankaDinelka Thilakarathne1,2, Nadeela Hirimuthugoda3, Kithsiri Ranawana3 and Shalika Kumburegama3*
Research 2021-08-14 49:14
Estimating potential range shift of some wild bees in response to climate change scenarios in northwestern regions of IranEhsan Rahimi1, Shahindokht Barghjelveh1 and Pinliang Dong2