Next Article in Journal
Monitoring Aquatic Debris in a Water Environment Using a Remotely Operated Vehicle (ROV): A Comparative Study with Implications of Algal Detection in Lake Como (Northern Italy)
Previous Article in Journal
Identifying Circular City Indicators Based on Advanced Text Analytics: A Multi-Algorithmic Approach
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Environments
Volume 12
Issue 1
10.3390/environments12010002
environments-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Using Species Distribution Models to Assess the Status of the Declining Western Bumble Bee (Hymenoptera: Apidae: Bombus occidentalis) in Wyoming, USA

by
Lusha M. Tronstad
*,
Christine Bell
,
Katrina Cook
and
Michael E. Dillon
Wyoming Natural Diversity Database, Department of Zoology and Physiology, and Program in Ecology and Evolution, University of Wyoming, Laramie, WY 82071, USA
*
Author to whom correspondence should be addressed.
Environments 2025, 12(1), 2; https://doi.org/10.3390/environments12010002
Submission received: 20 November 2024 / Revised: 17 December 2024 / Accepted: 24 December 2024 / Published: 27 December 2024
Browse Figures
Versions Notes

Abstract

:
Monitoring declining species is crucial to inform conservation but is challenging for rare species with limited information. The Western Bumble Bee (Bombus occidentalis) was previously common in the western United States but has drastically declined. Despite documented populations in the Intermountain West, many areas remain under-sampled. Species distribution models (SDM) can guide sampling efforts in large areas by predicting where the highest probability of suitable habitat may occur. We developed a sampling SDM using historical observations (1910–2010) in Wyoming to predict suitable habitat in the past. Using the model, we selected sampling sites that ranged from low to high predicted habitat suitability and we revisited historical locations where B. occidentalis were observed. Using all data (historical and current), we selected the predictors that explained the most variance, and created separate historical and current (2017–2018) SDM using the same variables to assess how predicted habitat suitability changed. We detected B. occidentalis at 30% of the revisited historical sites and 25% of all sites sampled. Areas predicted to be highly suitable for B. occidentalis in Wyoming declined by 5%; a small decrease compared to declines in the western portion of their range. Predicted habitat suitability increased the most in foothill areas. Creating SDM with landscape and climatic variables can bolster models and identify highly contributing variables. Regional SDM complement range-wide SDM by focusing on a portion of their range and assessing how predicted habitat changed.

1. Introduction

Monitoring declining species is crucial for evaluating their conservation status, abundance and distribution; however, sampling can be difficult if the species is presumed rare or basic distribution information is lacking as is true for many insects [1,2]. Insects represent >67% of named and described animals, but they only account for 11.5% of the animals with federal protection in the USA (NatureServe Explorer, 2024; https://explorer.natureserve.org accessed on 1 November 2024). This may reflect a lack of knowledge that can at least partially stem from challenges with sampling small animals due to patchy distributions across large ranges, highly variable phenology and poor historical records [3]. Historical data (i.e., museum specimens) are often limited, and can be biased temporally and spatially [4,5]. Despite the challenges of using historical data, studies have documented the loss of insect biomass and biodiversity [6] which is concerning because they provide vast ecosystem services, including detritivory, herbivory, parasitism (e.g., biological control) and pollination [7,8].
Pollinating insects are crucial because they support ~88% of plant species by transporting pollen among individuals that helps maintain plant genetic diversity [9,10]. In particular, bumble bees (Hymenoptera: Apidae: Bombus sp.) are highly effective generalist pollinators that provide essential ecosystem services in agricultural and natural settings [11]. Several bumble bee species have experienced notable declines; however, the status of these species is often uncertain as large areas of their historical ranges have not been surveyed recently [12,13]. Causes of declines are largely unknown, but effects from pesticides, climate change, habitat degradation and fragmentation, and pathogen spillover from commercial colonies are suspected [14]. Currently, two bumble bee species are federally protected in the USA and an increasing number of petitions requesting protection for additional species have been submitted, including for the Western Bumble Bee (Bombus occidentalis).
The Western Bumble Bee was once a common insect in western North America, found from central California to Alaska, east to South Dakota, and south into New Mexico [12]. The Western Bumble Bee has largely disappeared from its historical range since ~2010, particularly on the west coast [15]. The status of populations in the Intermountain West is largely unknown and we are not aware of any state-wide surveys in the Rocky Mountains conducted since 2009 [13,16]. To assess the status of B. occidentalis, we need current information on its distribution, as the remaining populations may be crucial to the persistence of the species.
Species distribution models (SDM) predict suitable habitat of B. occidentalis, and inform future sampling and conservation efforts. SDM relate locations of observations to environmental and landscape characteristics to predict suitable habitat of a species where they have not been surveyed [17]. These models are often used in conservation to estimate the current status of the species and aid management decisions [18,19,20,21,22,23]. Developing SDM can aid future research for under-studied species by establishing baseline distribution information [24,25]. Augmenting historical collection data with current occurrence data can improve the performance of SDM by presenting a more accurate representation of the current suitable habitat and add insight into potential range shifts when compared with historical distributions [18]. Many techniques for modeling species distributions—such as generalized linear models, generalized additive models and random forest—use presence and absence points within the spatial extent of the model. Absence points are often randomly generated and potentially create false absences, particularly when using historical or museum data, which could lead to erroneous predictions [26]. One method—maximum entropy (Maxent)—uses presence-only data and reliably produces high performing models [23]. Maxent models can be especially useful to predict the distribution of rare or under-sampled organisms using small data sets or in areas where little or no sampling has occurred [4].
Range-wide models for B. occidentalis using historical observations and limited environmental factors indicated that Wyoming, USA had large swaths of suitable habitat [13,27]. Range-wide maps are critical to assess species at larger scales; however, regional- or local-scale models can identify population declines and provide a more comprehensive understanding of the status of a declining species by identifying unique environmental drivers affecting populations in particular parts of their range [5,28]. For example, declines of bumble bees in North America were first detected in regional surveys [15,16,29,30]. Widespread species like B. occidentalis occur in many different ecosystems across their range, thus estimating changes in their predicted habitat in smaller areas can provide additional information about the status of the species.
Our objective was to investigate the status and distribution of B. occidentalis in Wyoming, USA. Wyoming has unique habitat compared to most of B. occidentalis’ range due to the high elevation, semi-arid environment and lower human density. We produced a SDM using historical occurrence data, and a variety of landscape and environmental variables to predict suitable habitat prior to 2011. We used the historical SDM to guide our sampling efforts which included revisiting historical locations (n = 28). Using the same predictors, we predicted suitable habitat with historical and newly collected observations to compare results. Our specific questions were: (1) In what proportion of historical sites did we observe B. occidentalis? (2) To what degree did predicted suitable habitat change between the historical and current SDM? and (3) What variables best predicted suitable habitat for B. occidentalis? Establishing baseline data will elicit new questions and research directions for this declining bee, and provide information that will aid managers in land management decisions.

2. Materials and Methods

2.1. Study Area

We sampled across the state of Wyoming, USA in locations with varying climatic and landscape characteristics (Figure 1). Wyoming is a high elevation state (953 to 4209 m; mean = 2042 m) with a semi-arid climate. The main ecotypes are short grass prairie, sagebrush steppe, conifer forest and tundra. Annual precipitation varies widely from 15 cm in the basins to 230 cm in the mountains [31].

2.2. Species Distribution Models

We created an initial Maxent (version 3.4.1) SDM to inform where to survey for B. occidentalis in 2017 and 2018 (sampling SDM) to create a balanced, representative sampling plan [32]. We used 90 unique historical locations of B. occidentalis between 1910–2010 obtained from the Global Biodiversity Information Facility (GBIF), the University of Wyoming Insect Museum and personal observations. We tested 25 climatic and landscape variables we choose a priori to predict suitable habitat of B. occidentalis based on previous studies and where they were previously observed. The nineteen bioclimatic variables from WorldClim [33] were used and rasters were resampled to a 30 × 30 m resolution to match the resolution of landscape variables. Landscape variables tested included a shrub index, a conifer index, mean deciduous tree cover, mean herbaceous ground cover, percent forest canopy cover and distance to water which were created from the LANDFIRE Existing Vegetation [34], GAP Land Cover [35], and USGS’s National Hydrography datasets [36]. We removed correlated variables (Pearson’s correlation coefficient ≥ 0.7), used ten-fold cross-validation, and reviewed the overall percent contributions and jackknife contributions of each variable to identify the top predictors [28] (Table 1).
We selected areas on public land that varied in predicted suitable habitat and ecoregions to generate new sampling locations using the sampling SDM. We binned the probability levels (p) into three equal intervals and selected sections with low (0.0 < p ≤ 0.33, n = 17), medium (0.33 < p ≤ 0.66, n = 22) and high (0.66 < p ≤ 1.0, n = 36) predicted probabilities of suitable habitat, spread across 5 Level III ecoregions [37] (Figure 1). Additionally, we revisited 28 historical sampling locations for a total of 103 sites.
We created a new SDM for historical and current data using the same predictors to estimate differences in the amount of predicted suitable habitat. We selected variables that explained the most variation when all data (historical and current) were combined. We used the same predictors to create an SDM with only historical or current data for B. occidentalis locations. We used the same methods and evaluated the same initial suite of predictor variables as was done for the sampling SDM. We subtracted the predicted values of the historical model from the current model to produce a map to visualize changes in habitat suitability.

2.3. Field Sampling

We sampled pollinators in western Wyoming (62 sites) May–August 2017 and eastern Wyoming (41 sites) May–September 2018. We used 3 methods to detect the presence of B. occidentalis and compared each method’s performance [38] to maximize the probability of detection across all sites we sampled [32]. We used two types of passive traps (blue vane traps and sets of pan traps) and one active method (target netting). We set out three vane traps and three sets of pan traps (yellow, white and blue) for 24–48 h at 98 sites. Vane traps and pan traps were placed ≥15 m apart to remain independent [39]. We actively target netted bumble bees for 30 min at 97 sites. Seven of the sites were target netted only. We visited 52% of sites twice over the summers. Specimens were brought back to the laboratory where they were processed and identified by an experienced bumble bee taxonomist [12].

3. Results

3.1. Sampling SDM

Seven top variables were identified for the sampling SDM from the 25 variables evaluated. Past (1910–2010) observations of B. occidentalis were mainly in mountainous areas of Wyoming (Figure 2a) potentially explaining why suitable habitat was most influenced by mean forest cover, precipitation of coldest quarter (snowpack), isothermality (temperature evenness), mean herbaceous cover, mean temperature of wettest quarter, precipitation of wettest month and precipitation seasonality (area under the curve; AUC = 0.87). The model predicted a relatively even proportion of Wyoming that had predicted low (34.3%), medium (32.8%) and high (32.9%) habitat suitability.

3.2. Current Results for Bombus Occidentalis Sampling

We captured B. occidentalis at 25% of sampled sites across the state (n = 26) and 30% of revisited historical sites where this bumble bee was previously observed (n = 8; Figure 1). Bombus occidentalis was detected at 15 sites (24%) in western Wyoming in 2017 and 11 sites (27%) in eastern Wyoming in 2018.

3.3. Historical and Current SDM

Thirteen top variables were identified when all data (historical and current) were combined from the 25 variables evaluated. The same predictors selected in the sampling SDM were retained when all the data were combined except for precipitation seasonality. Shrub cover index, conifer cover index, deciduous tree cover, herbaceous cover and mean temperature of the driest quarter were the most influential variables (Table 1). The historical SDM created using these variables (AUC = 0.88; Figure 2b) predicted a large proportion of the state had a low (69%) probability of suitable habitat, and much less of the state was predicted to have medium (16%) and high (15%) habitat suitability. Three percent of sites were in low, 39% in medium and 58% in high probability of suitable habitat as predicted by the 26 sites where B. occidentalis was observed.
The current model (AUC = 0.89 ± 0.064; Figure 2c) was closely associated with deciduous tree cover, conifer cover index and the mean temperature of the warmest quarter (Table 1). The current model predicted more area with low habitat suitability (76%), and less area with medium (14%) and high (10%) predicted habitat suitability for B. occidentalis. More B. occidentalis sites occurred in areas predicted to have high (80%) habitat suitability compared to medium (15%) and low (3%) probability (Figure 2c). Most suitable habitat was predicted in northwestern Wyoming, and some mountain ranges were highlighted in the rest of the state.

3.4. Changes in Predicted Suitable Habitat

Generally, habitat suitability increased in the foothills at the base of mountain ranges, and decreased in northwestern Wyoming in the national parks and the Wyoming Range (Figure 3). The current SDM predicted more area with low habitat suitability (increased by 7%), and less area with medium (decreased by 2%) and high (decreased by 5%) predicted habitat suitability for B. occidentalis compared to the historical SDM.

4. Discussion

Comparing our historical and current SDM predicted that highly suitable habitat for B. occidentalis declined by 5% in Wyoming. These data provide a baseline of information on the distribution of B. occidentalis in Wyoming where we primarily found them in mountain meadows, forest edges and urban environments. Bombus occidentalis was observed at 25% of the sites visited in 2017 and 2018, the same percentage of sites (n = 20) that Cameron et al. [16] reported for Wyoming surveys in 2008 and 2009. Conversely, we had much higher success in Wyoming compared to surveys in eastern Washington where B. occidentalis was found at 9% of sites [5]. Wyoming is unique compared to most other areas inhabited by B. occidentalis due to the high elevation and semi-arid climate [37]. We discovered that the margins of mountain ranges in Wyoming had some of the largest increases in predicted suitable habitat which may indicate elevational shifts as has been observed in other bumble bees [40]. Species distribution models can direct survey efforts so researchers can stratify sites by predicted habitat suitability resulting in balanced surveys in low, medium and high-quality sites. Selecting sites without using a SDM could result in artificially low or high detection rates that could bias the data and be interpreted as a decline or increase in the species.
Previous models for B. occidentalis only used climatic variables (WorldClim BioClim data) [16,27], a common practice when creating SDM [25,41,42]; however, we included landscape variables in addition to climatic variables. Temperature explained the most variation in range-wide models for B. occidentalis [27,43] which is not surprising considering that air temperatures affect insects [44], including bumble bees [45]. In our models, five measures of air temperature contributed nearly 25%; however, landscape variables (57.6%) contributed more to the historical model than temperature (22.7%) or precipitation (10.6%). Similarly, landscape variables (64.8%) contributed more to the current model compared to temperature (24.8%) and precipitation variables (13.3%). Similarly, landscape variables have contributed the most to SDM created for other taxa [19,20,22].
Most SDM are fitted with historical observations, but are often improved with the addition of current data [18]. The comparison of our historical and current models illustrates the value of up-to-date data. We observed a loss in suitable habitat for B. occidentalis and potentially a range shift to lower elevations. The increase in suitable habitat in the foothills may be correlated with floral resources. Our current model predicting suitable habitat of B. occidentalis provides a foundation for monitoring this declining species. Maxent models often over-predict suitable habitat [46,47]; therefore, our results should be considered conservatively. Species distribution models are a useful tool to guide survey efforts and identify potentially limiting variables; however, they only predict areas of suitable habitat and not the range of the species. Pairing SDM results with habitat characteristics at finer scales (e.g., floral surveys and soil characteristics) will help understand ecological processes driving spatial patterns of B. occidentalis occurrences.

5. Conclusions

Models suggest that suitable habitat for B. occidentalis is declining in Wyoming; however, the decrease is small compared to other areas [43]. Bombus occidentalis was locally abundant in several areas, but its distribution was patchy. Comparing this region-specific survey to areas where the species is nearly extirpated could help identify potential causes for population declines. Wyoming provides a novel environment lacking some of the factors thought to drive population declines, including crops that tend to be pollinated by managed bee colonies, extensive pesticide application and widespread urban development; however, climate change could negatively affect this bumble bee [48]. As B. occidentalis is nearly extirpated in large portions of its former range, Wyoming populations may become a stronghold for the species as a whole [43].

Author Contributions

Conceptualization, L.M.T., C.B. and M.E.D.; methodology, L.M.T. and C.B.; formal analysis, C.B. and K.C.; investigation, C.B.; resources, L.M.T. and M.E.D.; writing—original draft preparation, L.M.T., C.B. and K.C.; writing—review and editing, L.M.T., C.B., K.C. and M.E.D.; visualization, C.B. and K.C.; supervision, L.M.T. and M.E.D.; project administration, L.M.T. All authors have read and agreed to the published version of the manuscript.

Funding

This project was funded by the Wyoming Bureau of Land Management (L16AC0389), US Forest Service, Wyoming Landscape Conservation Initiative, Wyoming NASA Space Grant Consortium, and the University of Wyoming Biodiversity Institute.

Data Availability Statement

Data are available from the Wyoming Natural Diversity Database, University of Wyoming (https://wyndd.org/portal/apps/data_explorer/map; 1 July 2024), due to the sensitive nature of the data.

Acknowledgments

We thank Timothy Collier and Zach Wallace for their thoughts and edits. We are very grateful to US National Parks Service (permit numbers GRTE-2017-SCI-0035, YELL-2017-SCI-8002, FOLA-2018-SCI-0005, DETO-2018-SCI-0007), US National Forest Service, Wyoming State Parks, and private landowners for providing access to sites. We thank Madison Crawford and Charles Anderson for help in the field, Matthew Green for his help in the lab and the field, Mark Andersen for his assistance.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Dunn, R. Modern insect extinctions, the neglected majority. Conserv. Biol. 2005, 19, 1030–1036. [Google Scholar] [CrossRef]
  2. Thompson, W. Sampling Rare or Elusive Species: Concepts, Designs, and Techniques for Estimating Population Parameters; Island Press: Washington, DC, USA, 2013. [Google Scholar]
  3. Koch, J.B.; Strange, J.P. Constructing a species database and historic range maps for North American Bumblebees (Bombus sensu stricto Latreille) to inform conservation decisions. Uludağ Arıcılık Derg. 2009, 9, 97–108. [Google Scholar]
  4. Newbold, T. Applications and limitations of museum data for conservation and ecology, with particular attention to species distribution models. Prog. Phys. Geogr. 2010, 34, 3–22. [Google Scholar] [CrossRef]
  5. Rhoades, P.R.; Koch, J.B.; Waits, L.P.; Strange, J.P.; Eigenbrode, S.D. Evidence for Bombus occidentalis (Hymenoptera: Apidae) populations in the Olympic Peninsula, the Palouse Prairie, and forests of northern Idaho. J. Insect Sci. 2016, 16, 20. [Google Scholar] [CrossRef] [PubMed]
  6. Burkle, L.A.; Marlin, J.C.; Knight, T.M. Plant-Pollinator Interactions over 120 Years: Loss of Species, Co-Occurrence, and Function. Science 2013, 339, 1611–1615. [Google Scholar] [CrossRef]
  7. Conrad, K.F.; Warren, M.S.; Fox, R.; Parsons, M.S.; Woiwod, I.P. Rapid declines of common, widespread British moths provide evidence of an insect biodiversity crisis. Biol. Conserv. 2006, 132, 279–291. [Google Scholar] [CrossRef]
  8. Hallmann, C.A.; Sorg, M.; Jongejans, E.; Siepel, H.; Hofland, N.; Schwan, H.; Stenmans, W.; Muller, A.; Sumser, H.; Horren, T.; et al. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLoS ONE 2017, 12, e0185809. [Google Scholar] [CrossRef]
  9. Michener, C.D. The Bees of the World, 2nd ed.; JHU Press: Baltimore, MA, USA, 2007; Volume 1. [Google Scholar]
  10. Ollerton, J. Pollinators and Pollination: Nature and Society; Pelagic Publishing Ltd.: London, UK, 2021. [Google Scholar]
  11. Losey, J.E.; Vaughan, M. The economic value of ecological services provided by insects. Bioscience 2006, 56, 311–323. [Google Scholar] [CrossRef]
  12. Williams, P.H.; Thorp, R.W.; Richardson, L.L.; Colla, S.R. Bumble Bees of North America; Princeton University Press: Princeton, NJ, USA, 2014; p. 208. [Google Scholar]
  13. Graves, T.A.; Janousek, W.M.; Gaulke, S.M.; Nicholas, A.C.; Keinath, D.A.; Bell, C.M.; Cannings, S.; Hatfield, R.G.; Heron, J.M.; Koch, J.B.; et al. Western bumble bee: Declines in the continental United States and range-wide information gaps. Ecosphere 2020, 11, e03141. [Google Scholar] [CrossRef]
  14. Goulson, D.; Nicholls, E.; Botias, C.; Rotheray, E.L. Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science 2015, 347, 1255957. [Google Scholar] [CrossRef]
  15. Thorp, R.; Jepsen, S.; Black, S.H. Status Review of Three Formerly Common Species of Bumble Bee in the Subgenus Bombus. Prep. Xerces Soc. Invertebr. Conservation. 2008. Available online: https://xerces.org/sites/default/files/2019-10/xerces_2008_bombus_status_review_0.pdf (accessed on 1 November 2024).
  16. Cameron, S.A.; Lozier, J.D.; Strange, J.P.; Koch, J.B.; Cordes, N.; Solter, L.F.; Griswold, T.L. Patterns of widespread decline in North American bumble bees. Proc. Natl. Acad. Sci. USA 2011, 108, 662–667. [Google Scholar] [CrossRef] [PubMed]
  17. Suzuki-Ohno, Y.; Ishihama, F.; Yokoyama, J.; Inoue, M.N.; Nakashizuka, T.; Kawata, M. Estimating bee distributions and their functional range to map important areas for protecting bee species and their functions. Sci. Rep. 2024, 14, 12842. [Google Scholar] [CrossRef] [PubMed]
  18. Rodríguez, J.P.; Brotons, L.; Bustamante, J.; Seoane, J. The application of predictive modelling of species distribution to biodiversity conservation. Divers. Distrib. 2007, 13, 243–251. [Google Scholar] [CrossRef]
  19. Yost, A.C.; Petersen, S.L.; Gregg, M.; Miller, R. Predictive modeling and mapping sage grouse (Centrocercus urophasianus) nesting habitat using Maximum Entropy and a long-term dataset from Southern Oregon. Ecol. Inform. 2008, 3, 375–386. [Google Scholar] [CrossRef]
  20. Crawford, P.H.; Hoagland, B.W. Using species distribution models to guide conservation at the state level: The endangered American burying beetle (Nicrophorus americanus) in Oklahoma. J. Insect Conserv. 2010, 14, 511–521. [Google Scholar] [CrossRef]
  21. Davis, S.L.; Cipollini, D. Range, genetic diversity and future of the threatened butterfly, Pieris virginiensis. Insect Conserv. Divers. 2016, 9, 506–516. [Google Scholar] [CrossRef]
  22. Pogue, C.D.; Monfils, M.J.; Cuthrell, D.L.; Heumann, B.W.; Monfils, A.K. Habitat suitability modeling of the federally endangered Poweshiek skipperling in Michigan. J. Fish Wildl. Manag. 2016, 7, 359–368. [Google Scholar] [CrossRef]
  23. Silva, D.P.; Spigoloni, Z.A.; Camargos, L.M.; de Andrade, A.F.A.; De Marco, P.; Engel, M.S. Distributional modeling of Mantophasmatodea (Insecta: Notoptera): A preliminary application and the need for future sampling. Org. Divers. Evol. 2016, 16, 259–268. [Google Scholar] [CrossRef]
  24. Pelletier, T.A.; Crisafulli, C.; Wagner, S.; Zellmer, A.J.; Carstens, B.C. Historical species distribution models predict species limits in western Plethodon salamanders. Syst. Biol. 2015, 64, 909–925. [Google Scholar] [CrossRef] [PubMed]
  25. Rojas-Arias, L.; Gómez-Morales, D.; Stiegel, S.; Ospina-Torres, R. Niche modeling of bumble bee species (Hymenoptera, Apidae, Bombus) in Colombia reveals highly fragmented potential distribution for some species. J. Hymenopt. Res. 2023, 95, 231–244. [Google Scholar] [CrossRef]
  26. Elith, J.; Phillips, S.J.; Hastie, R.; Dudik, M.; Chee, Y.E.; Yates, C.J. A statistical explanation of MaxEnt for ecologists. Divers. Distrib. 2011, 17, 43–57. [Google Scholar] [CrossRef]
  27. Sheffield, C.S.; Richardson, L.; Cannings, S.; Ngo, H.; Heron, J.; Williams, P.H. Biogeography and designatable units of Bombus occidentalis Greene and B. terricola Kirby (Hymenoptera: Apidae) with implications for conservation status assessments. J. Insect Conserv. 2016, 20, 189–199. [Google Scholar] [CrossRef]
  28. Christman, M.E.; Spears, L.R.; Strange, J.P.; Pearse, W.D.; Burchfield, E.K.; Ramirez, R.A. Land cover and climate drive shifts in Bombus assemblage composition. Agric. Ecosyst. Environ. 2022, 339, 108113. [Google Scholar] [CrossRef]
  29. Colla, S.R.; Packer, L. Evidence for decline in eastern North American bumblebees (Hymenoptera: Apidae), with special focus on Bombus affinis Cresson. Biodivers. Conserv. 2008, 17, 1379–1391. [Google Scholar] [CrossRef]
  30. Tripodi, A.D.; Szalanski, A.L. The bumble bees (Hymenoptera: Apidae: Bombus) of Arkansas, fifty years later. J. Melittology 2015, 50, 1–17. [Google Scholar] [CrossRef]
  31. Arguez, A.; Durre, I.; Applequist, S.; Squires, M.; Vose, R.; Yin, X.; Bilotta, R.; NOAA′s U.S. Climate Normals (1981–2010). 2010. Available online: https://www.ncei.noaa.gov/metadata/geoportal/rest/metadata/item/gov.noaa.ncdc:C00820/html (accessed on 1 March 2024).
  32. Yackulic, C.B.; Chandler, R.; Zipkin, E.F.; Royle, J.A.; Nichols, J.D.; Campbell Grant, E.H.; Veran, S. Presence-only modelling using MAXENT: When can we trust the inferences? Methods Ecol. Evol. 2013, 4, 236–243. [Google Scholar] [CrossRef]
  33. Hijmans, R.; Cameron, S.; Parra, J.; Jones, P.; Jarvis, A. Very high resolution interpolated climate surfaces of global land areas. Int. J. Climatol. 2005, 25, 1965–1978. [Google Scholar] [CrossRef]
  34. Comer, P. Ecological Systems of the United States: A Working Classification of U.S. Terrestrial Systems; NatureServe: Arlington, VI, USA, 2003. [Google Scholar]
  35. Davidson, A.; Aycrigg, J.; Grossmann, E.; Kagan, J.; Lennartz, S.; McDonough, S.; Miewald, T. Digital Land Cover Map for the Northwestern United States; Northwest Gap Analysis Project, USGS GAP Analysis Program; Geological Survey: Moscow, ID, USA, 2009. [Google Scholar]
  36. Simley, J.D.; Carswell, W.J., Jr. The National Map-Hydrography; 2327-6932; US Geological Survey: Reston, VA, USA, 2009. [Google Scholar]
  37. Chapman, S.S.; Bryce, S.A.; Omernik, J.M.; Despain, D.G.; ZumBerge, J.; Conrad, M. Ecoregions of Wyoming. Trans. Wis. Acad. Sci. Arts Lett. 2004, 88, 77–103. [Google Scholar]
  38. Bell, C.; Tronstad, L.; Hotaling, S. Tailoring your bee sampling protocol: Comparing three methods reveals the best approaches to capturing bees. Agric. For. Entomol. 2023, 25, 477–488. [Google Scholar] [CrossRef]
  39. Droege, S.; Tepedino, V.J.; Lebuhn, G.; Link, W.; Minckley, R.L.; Chen, Q.; Conrad, C. Spatial patterns of bee captures in North American bowl trapping surveys. Insect Conserv. Divers. 2010, 3, 15–23. [Google Scholar] [CrossRef]
  40. Fourcade, Y.; Åström, S.; Öckinger, E. Climate and land-cover change alter bumblebee species richness and community composition in subalpine areas. Biodivers. Conserv. 2019, 28, 639–653. [Google Scholar] [CrossRef]
  41. Bucklin, D.N.; Basille, M.; Benscoter, A.M.; Brandt, L.A.; Mazzotti, F.J.; Romañach, S.S.; Speroterra, C.; Watling, J.I. Comparing species distribution models constructed with different subsets of environmental predictors. Divers. Distrib. 2015, 21, 23–35. [Google Scholar] [CrossRef]
  42. Nascimento, A.C.; Montalva, J.; Ascher, J.S.; Engel, M.S.; Silva, D.P. Current and future distributions of a native Andean bumble bee. J. Insect Conserv. 2022, 26, 559–569. [Google Scholar] [CrossRef]
  43. Janousek, W.; Douglas, M.; Cannings, S.; Clement, M.; Dephia, C.; Everett, J.; Hatfield, R.; Keinath, D.; Koch, J.; McCabe, L.; et al. Recent and future declines of a historically widespread pollinator linked to climate, land cover and pesticides. Proc. Acad. Nat. Sci. USA 2023, 120, e2211223120. [Google Scholar] [CrossRef] [PubMed]
  44. Kingsolver, J.G.; Arthur Woods, H.; Buckley, L.B.; Potter, K.A.; MacLean, H.J.; Higgins, J.K. Complex life cycles and the responses of insects to climate change. Integr. Comp. Biol. 2011, 51, 719–732. [Google Scholar] [CrossRef] [PubMed]
  45. Kerr, J.T.; Pindar, A.; Galpern, P.; Packer, L.; Potts, S.G.; Roberts, S.M.; Rasmont, P.; Schweiger, O.; Colla, S.R.; Richardson, L.L. Climate change impacts on bumblebees converge across continents. Science 2015, 349, 177–180. [Google Scholar] [CrossRef] [PubMed]
  46. Marcer, A.; Sáez, L.; Molowny-Horas, R.; Pons, X.; Pino, J. Using species distribution modelling to disentangle realised versus potential distributions for rare species conservation. Biol. Conserv. 2013, 166, 221–230. [Google Scholar] [CrossRef]
  47. Amirkhiz, R.G.; Frey, J.K.; Cain, J.W., III; Breck, S.W.; Bergman, D.L. Predicting spatial factors associated with cattle depredations by the Mexican wolf (Canis lupus baileyi) with recommendations for depredation risk modeling. Biol. Conserv. 2018, 224, 327–335. [Google Scholar] [CrossRef]
  48. Morales, C.L.; Montalva, J.; Arbetman, M.P.; Aizen, M.A.; Martins, A.C.; Silva, D.P. Does climate change influence the current and future projected distribution of an endangered species? The case of the southernmost bumblebee in the world. J. Insect Conserv. 2022, 26, 257–269. [Google Scholar] [CrossRef]
Environments 12 00002 g001
Figure 1. Sampling sites for bumble bees in Wyoming, USA separated by county boundaries (black lines) and colored by ecoregions. We re-visited historical sites where Bombus occidentalis were observed from 1910 to 2010 (squares) and we sampled new sites (triangles) in 2017–2018. Blue shapes denote where we observed B. occidentalis and yellow shapes denote where we failed to detect B. occidentalis.
Figure 1. Sampling sites for bumble bees in Wyoming, USA separated by county boundaries (black lines) and colored by ecoregions. We re-visited historical sites where Bombus occidentalis were observed from 1910 to 2010 (squares) and we sampled new sites (triangles) in 2017–2018. Blue shapes denote where we observed B. occidentalis and yellow shapes denote where we failed to detect B. occidentalis.
Environments 12 00002 g001
Environments 12 00002 g002
Figure 2. (a) Sampling species distribution model (SDM) of Wyoming, USA used to guide site selection created from observations of Bombus occidentalis between 1910 and 2010 using Maxent. (b) Historical and (c) current species distribution models for B. occidentalis. We choose climate and landscape variables by combining all data, and created separate SDM using the same predictor variables. Points represent historic (1910–2010; yellow circles) and current (2017–2018; blue triangles) observations of B. occidentalis. The darker green shading predicts more suitable habitat for the bumble bee.
Figure 2. (a) Sampling species distribution model (SDM) of Wyoming, USA used to guide site selection created from observations of Bombus occidentalis between 1910 and 2010 using Maxent. (b) Historical and (c) current species distribution models for B. occidentalis. We choose climate and landscape variables by combining all data, and created separate SDM using the same predictor variables. Points represent historic (1910–2010; yellow circles) and current (2017–2018; blue triangles) observations of B. occidentalis. The darker green shading predicts more suitable habitat for the bumble bee.
Environments 12 00002 g002
Environments 12 00002 g003
Figure 3. The change in suitable habitat calculated as the difference between habitat suitability predicted by the historical (Figure 2b) minus the current (Figure 2c) species distribution models for Bombus occidentalis in Wyoming, USA. Orange indicates a loss in suitable habitat over time and purple indicates a gain in suitable habitat.
Figure 3. The change in suitable habitat calculated as the difference between habitat suitability predicted by the historical (Figure 2b) minus the current (Figure 2c) species distribution models for Bombus occidentalis in Wyoming, USA. Orange indicates a loss in suitable habitat over time and purple indicates a gain in suitable habitat.
Environments 12 00002 g003
Table 1. Environmental variables used for the sampling species distribution model (SDM), and the historical and current SDM after variable selection for the Western Bumble Bee (Bombus occidentalis; see methods). Variables contributed a different percentage for the sampling (1910–2010), historical (1910–2010) and current (2017–2018) models. Variables are in order of the largest contribution to the historical model. The sampling model was created with past observations only. The historical and current models were made when variables were selected using the complete dataset (historical and current observations) and separate SDM were made for each time period to compare the amount of predicted suitable habitat.
Table 1. Environmental variables used for the sampling species distribution model (SDM), and the historical and current SDM after variable selection for the Western Bumble Bee (Bombus occidentalis; see methods). Variables contributed a different percentage for the sampling (1910–2010), historical (1910–2010) and current (2017–2018) models. Variables are in order of the largest contribution to the historical model. The sampling model was created with past observations only. The historical and current models were made when variables were selected using the complete dataset (historical and current observations) and separate SDM were made for each time period to compare the amount of predicted suitable habitat.
VariableVariable Description% Contribution
SamplingHistoricalCurrent
shrubShrub index 15.42.7
coniferConifer index 13.715.7
decidMean deciduous tree cover 1236.8
herbMean herbaceous cover8.811.52
bioclim9Mean temperature of the driest quarter 11.50.4
bioclim19Precipitation of the coldest quarter (snowpack)25.780.7
bioclim10Mean temperature of the warmest quarter 6.312.4
forest ccPercent forest canopy cover26.657.6
bioclim1Annual mean temperature 4.96.8
bioclim3Isothermality (temperature evenness)103.54.2
bioclim8Mean temperature of the wettest quarter8.53.21
bioclim13Precipitation of the wettest month5.82.66.2
d2wDistance to water 2.53.6
Bioclim15Precipitation seasonality5.2
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Tronstad, L.M.; Bell, C.; Cook, K.; Dillon, M.E. Using Species Distribution Models to Assess the Status of the Declining Western Bumble Bee (Hymenoptera: Apidae: Bombus occidentalis) in Wyoming, USA. Environments 2025, 12, 2. https://doi.org/10.3390/environments12010002

AMA Style

Tronstad LM, Bell C, Cook K, Dillon ME. Using Species Distribution Models to Assess the Status of the Declining Western Bumble Bee (Hymenoptera: Apidae: Bombus occidentalis) in Wyoming, USA. Environments. 2025; 12(1):2. https://doi.org/10.3390/environments12010002

Chicago/Turabian Style

Tronstad, Lusha M., Christine Bell, Katrina Cook, and Michael E. Dillon. 2025. "Using Species Distribution Models to Assess the Status of the Declining Western Bumble Bee (Hymenoptera: Apidae: Bombus occidentalis) in Wyoming, USA" Environments 12, no. 1: 2. https://doi.org/10.3390/environments12010002

APA Style

Tronstad, L. M., Bell, C., Cook, K., & Dillon, M. E. (2025). Using Species Distribution Models to Assess the Status of the Declining Western Bumble Bee (Hymenoptera: Apidae: Bombus occidentalis) in Wyoming, USA. Environments, 12(1), 2. https://doi.org/10.3390/environments12010002

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop