Soil sand content is a driving force in structuring bee communities

Soil sand content is a driving force in structuring bee communities | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 6 August 2025 V1 Latest version Share on Soil sand content is a driving force in structuring bee communities Authors : Marissa Chase 0000-0003-4878-6658 [email protected] , Nicole Gerjets , Ian G. Lane , Jessica D. Petersen , Zachary Portman , and Daniel Cariveau Authors Info & Affiliations https://doi.org/10.22541/au.175449157.79398130/v1 Published Insect Conservation and Diversity Version of record Peer review timeline 226 views 103 downloads Contents Abstract Abstract 1 Introduction 2 Methods 3 Results 4 Discussion 5 Conclusion References Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract \received DD MMMM YYYY \acceptedDD MMMM YYYY Floral and nesting resources are two major components of habitat specialization that drive patterns of bee distribution. However, nesting resources are largely understudied in comparison to floral resources. For nesting, sand preference is seen as a community-wide trait for bee species and numerous species are considered sand specialists. With little empirical evidence to date, we aimed to test how soil sand content and floral resource availability interact to affect patterns of bee distribution. We also designed the study to evaluate whether sand specialist bees predominantly occur in sandy habitats. We set up a controlled field experiment in Minnesota, USA across three classes of soil sand content (N = 28 plots) using a focal prairie plant species (Dalea purpurea) that attracts a wide range of bee species and can grow in different soil types. We identified four key results: 1) soil sand content, not floral resource availability, affected patterns of bee distribution; 2) contrary to expected, sandier sites did not host the highest diversity of bee species; 3) there was clear evidence of sand specialization for select bee species, but sand specialists were associated with both moderate and high amounts of sand rather than extremely sandy areas; and 4) the proportion of abundance of sand specialists increased as the soil sand content within discrete pockets increased, even in areas with average values of low sand content. Our findings highlight that soil properties, specifically soil sand content, can be more important than floral resources in driving patterns of bee distribution, specifically sand specialists. Therefore, not factoring soil properties into bee conservation and restoration decisions may limit our effectiveness and ability to support habitat specialists and rarer bee communities. Sandy areas and places with high soil heterogeneity should be prioritized for conservation. Soil sand content is a driving force in structuring bee communities Running title: Sand structures bee communities Abstract Floral and nesting resources are two major components of habitat specialization that drive patterns of bee distribution. However, nesting resources are largely understudied in comparison to floral resources. For nesting, sand preference is seen as a community-wide trait for bee species and numerous species are considered sand specialists. With little empirical evidence to date, we aimed to test how soil sand content and floral resource availability interact to affect patterns of bee distribution. We also designed the study to evaluate whether sand specialist bees predominantly occur in sandy habitats. We set up a controlled field experiment in Minnesota, USA across three classes of soil sand content (N = 28 plots) using a focal prairie plant species ( Dalea purpurea ) that attracts a wide range of bee species and can grow in different soil types. We identified four key results: 1) soil sand content, not floral resource availability, affected patterns of bee distribution; 2) contrary to expected, sandier sites did not host the highest diversity of bee species; 3) there was clear evidence of sand specialization for select bee species, but sand specialists were associated with both moderate and high amounts of sand rather than extremely sandy areas; and 4) the proportion of abundance of sand specialists increased as the soil sand content within discrete pockets increased, even in areas with average values of low sand content. Our findings highlight that soil properties, specifically soil sand content, can be more important than floral resources in driving patterns of bee distribution, specifically sand specialists. Therefore, not factoring soil properties into bee conservation and restoration decisions may limit our effectiveness and ability to support habitat specialists and rarer bee communities. Sandy areas and places with high soil heterogeneity should be prioritized for conservation. Keywords: bees, specialization, nesting, sand, soil texture, focal plant species 1 Introduction Understanding the factors that shape bee distribution is critical to the conservation and restoration of native bee communities. One ecological facet that drives distribution is habitat specialization, which for bees is a function of floral and nesting specialization (Ricketts et al. 2008). While bees require both nesting and floral resources for survival, studies have largely focused on how floral resource availability, use, and specialization structures bee communities (Potts et al. 2003; Cane and Sipes 2006; Fründ et al. 2010). Findings emphasize that floral diversity is a major driver of bee composition across ecosystems, and that floral specialization leads to bee species being rarer and more vulnerable to decline (Potts et al. 2003; Biesmeijer et al. 2006; Weiner et al. 2014; Bogusch et al. 2020). Bee nesting requirements, in comparison to floral resources, are largely understudied as a component of habitat specialization. It remains unclear whether nesting resources solely limit bee distribution, or whether nesting requirements interact with specialized floral requirements to influence bee distribution (Cane 1991). Given that many floral specialists have smaller ranges than their host plants (Michener 2007; Buckner and Danforth 2022), it seems plausible that additional constraints of nesting requirements may further restrict patterns of bee distribution. Understanding the link between nesting requirements and bee distribution patterns may help targeted monitoring and conservation efforts, but there are key knowledge gaps when it comes to bee nesting ecology (Harmon-Threatt 2020; Antoine and Forrest 2021). While bees occupy many nesting guilds across species, perhaps the most important is sand-nesting specialization (henceforth sand specialization). Sand specialization refers to bee species that rely on sandy soils to build and excavate their nests. In addition to strict sand specialization by individual species, at a community level, it is thought that bees prefer to nest in soils with high sand content (Cane 1991; Leone et al. 2022). Soil sand content refers to the percentage of sand-sized particles in a soil profile, with sand particles being the largest in comparison to silt and clay particles. Benefits to nesting in high sandy areas include better drainage, less soil compaction, and decreased moisture (Skiba and Ball 2002; Harmon-Threatt 2020), but it is unclear what degree of sandiness and how much of it is required to make suitable nesting conditions. Among bee biologists and ecologists, it is understood that some bee species rely on sand to construct nests and therefore are considered sand specialists (Wolf and Ascher 2008; Leonard and Harmon-Threatt 2019; Harmon-Threatt 2020). However, this has not been well tested and supported with empirical data. It is also worth noting that our perception of sand preference may be the result of detection biases because it can be easier to find nests at bare sandy sites in comparison to low-sand sites covered in vegetation (Harmon-Threatt 2020). It can be difficult to ascertain which resource—nesting or floral—is driving patterns of bee composition. One reason is because bees are central-place foragers and nest near preferred floral resources (Michener 2007; Danforth et al. 2019). This is especially true for floral specialists (Minckley et al. 1994; Cane et al. 2008; Ritchie et al. 2016). Additionally, soil type plays a significant role in dictating plant community composition. One solution is to focus on a core plant species that attracts a wide range of specialist and generalist bee species and occurs across many soil types (e.g., Dalea spp. (Harmon-Threatt and Hendrix 2015). This may allow us to tease apart if nesting, floral preferences, or a combination of the two are driving patterns of bee distribution. Although few studies have quantified the impact of limited nesting availability on bee communities, or even what suitable nesting habitat is (Antoine and Forrest 2021), nesting resources may have a disproportionate impact on bee species compared with floral resources (Roulston and Goodell 2011). The combined effect of nesting and floral specialization might further structure bee community composition but has been largely understudied relative to the singular focus on floral communities. We used a landscape-scale field experiment to examine the factors that drive bee community composition by focusing on the interplay between floral and sand nesting specialization. We arranged the study along a gradient of soil sand content and controlled for differences in the floral community by focusing sampling on the focal plant Dalea purpurea and its foraging bee community. This system is ideal for studying these relationships because 1) D. purpurea can grow in a range of soil conditions and textures (i.e., in both sandy and non-sandy areas - Flora of North America Editorial Committee 2021); and 2) D. purpurea hosts a diverse bee community, including both floral specialists and generalists (Harmon-Threatt and Hendrix 2015). More broadly, Dalea spp. occur throughout North America and many host a specialized set of insect pollinators. Dalea purpurea has the largest range within the genus (though D. candida has a similar range), spreading from the Western Great Plains, through the Midwest, and down to the Southeast of the United States (Flora of North America Editorial Committee 2021) . Additionally, D. purpurea is commonly used in grassland restorations across its range. Focusing on this species and its relationship to bee distribution will have critical implications for bee conservation alongside grassland restoration efforts. We sought to understand how nesting resources structure bee communities by empirically testing whether bees thought to be sand specialists are restricted to sandy areas and if bee communities are more diverse in sandier areas. We had three specific questions: 1) Does soil sand content within and surrounding a plot affect bee abundance, diversity, and composition; 2) Does the soil sand content within and surrounding a plot affect the proportion and diversity of sand specialists?; and 3) Does the interaction of soil sand content, D. purpurea abundance, and overall floral richness affect the proportion of sand specialists and floral specialists? Overall, we predicted that bee composition would vary with soil sand content, and bee abundance and diversity would be greatest in high sand areas. We also predicted that we would find presumed sand specialists in high sand areas. Additionally, we predicted that, while the floral community might moderate the effects of soil sand content, soil properties would be more important in driving patterns of bee distribution. 2 Methods 2.1 Study system and design Our study occurred across the Prairie Parkland and Eastern Broadleaf Forest ecological provinces of Minnesota, USA in tallgrass prairie habitat. We established 50m2 plots that were > 400 m from each other with a known Dalea purpurea record at the center of each plot. After plot selection (see below for exact details), we had 28 plots distributed across three soil sand content categories: nine low sand, nine medium sand, and 10 high sand (Fig. 1). Plots were dispersed across Scientific Natural Areas (SNAs) and Wildlife Management Areas (WMAs) managed by the Minnesota Department of Natural Resources, as well as on The Nature Conservancy (TNC) and US Fish and Wildlife Service (USFWS) lands. Plots were mostly in remnant tallgrass prairie habitat, but five sites had areas that were seeded with native sedge, grass, and forb species. For seeded sites, most seeding occurred >20 years ago without clear documentation and were designed to reconstruct prairie habitats on formerly cultivated areas ( personal communication with site land managers ). We characterized the average soil sand content (i.e., the percentage of sand-sized particles) within the landscape by averaging sand content within 300 m of the plot center. We chose 300 m to accommodate the likely maximum foraging distance of small bees and average distance for medium and larger bees (Wolf and Moritz 2008; Hofmann et al. 2020; Kendall et al. 2022). First, we accessed the Gridded Soil Survey Geographic (gSSURGO) Database for Minnesota (USDA NRCS) via ArcMap (ESRI). Within the database there are 12 broad soil texture classes designated by the United States Department of Agriculture (USDA). These 12 texture classes are further divided into subclasses that correspond to a specific percentage of sand, silt, and clay. With this soil data, we generated a 300 m buffer around each plot center and counted the number of raster pixels of each soil texture class (Fig. 1). We calculated the average sand content of each buffer by multiplying the proportion of pixels for each soil texture class and the corresponding sand content value across the 300 m buffer to yield a sand content value proportional to the amount of area that texture class. With an average sand content value for each buffer, representing a landscape-level percentage of sand, we then used the range of sand content across all buffers to assign a low (8-32%), medium (38-64%), and high sand (68-92%) category value to each plot. We chose to use sand categories as opposed to a continuous measure of sand because there were discrete breaks in the data and not a full range of sand content values from 0-100%. We also determined the dominant soil texture class within each buffer (one of the 12 USDA soil texture classes), the dominant soil texture class within each plot, and the highest amount of sand (the highest sand content value) within a discrete pocket within each buffer. To ensure that plots with similar amounts of sand were not spatially clumped, we calculated Moran’s I statistic to and selected 30 spatially independent plots (P > 0.05; N = 10 low, medium, and high sand content). All 30 plots were ground-truthed in March-April 2024 to verify the presence of D. purpurea and later again in June to confirm that D. purpurea was actively growing in each plot. After ground-truthing, two plots were dropped because we could not find evidence of growing D. purpurea . We sampled each plot once during the bloom time of Dalea purpurea in July-August 2024. All 28 plots were checked two or more times before peak bloom to ensure each plot was sampled as close to the peak bloom of D. purpurea as possible. Bee and plant sampling occurred on the same day. All sampling took place on sunny days with minimal cloud coverage when temperatures were at least 15.5°C and wind speeds were below 4 m/s. To meet these weather conditions, sampling always occurred between 8:30 am and 4:00 pm. Upon arrival, plots were divided into four quadrants by a single 50 m north-south transect and another 50 m east-west transect (Fig. 1). Within each quadrant, the total number of D. purpurea flowers were counted (one floral head equals one flower). Once the total number of D. purpurea flowers were recorded, bee sampling began. Bee sampling occurred over a 60-minute active sampling period. Within each quadrant, we employed targeted hand netting exclusively off of D. purpurea flowers. Time spent netting in each quadrant was proportional to the amount of D. purpurea within each quadrant. For example, if 75% of the total plot’s D. purpurea was in one quadrant, then 45 minutes of active netting would occur in that same quadrant. To sample the flower community, a 1-m2 quadrat was placed every 5 m along each 25 m transect that divided the plot in each cardinal direction (Fig. 1). This resulted in 16 total quadrats per plot. Within each quadrat, we counted the total number of blooms for each flowering plant species, as well as the percentage of bare ground present. Once all flowering species were identified and counted within all 16 quadrats, we found all blooming species in the plot that fell outside of the quadrats through meandering walks. This gave us a measure of floral richness for the plot, even for blooming species that fell outside quadrats. 2.4 Bee identification and specialization determination All bees were identified to species or morphospecies by XXX using published keys and revisions. Keys and revisions include: Rowe 2017; Oram 2018; Gardner and Gibbs 2020; 2023; Robinson 2023; Portman et al. 2022; 2024; Onuferko and Rightmyer 2024. For a full detailed list of taxonomic resources used, refer to Table S1. All specimens are deposited in either the XXX synoptic collection or the XXX Insect Collection (XXX). Prior to analysis, bee species were assigned as floral and sand specialists using literature and expert opinion (Table S1). We recorded whether a given species is particularly a Dalea spp. specialist. We considered a bee species a Dalea specialist if there was literature to suggest they are specifically a Dalea spp. specialist or, alternatively, if they are a Fabaceae specialist but have a documented strong preference for Dalea spp. To further verify presumed sand specialization, we also independently had two Midwestern bee taxonomists and authorities (Mike Arduser and Zachary Portman) look at the final species list and assign whether each species was a presumed sand specialist or not. Both experts independently came to the same conclusions of which species were sand specialists and which were not. All statistical analyses were conducted in R version 4.3.1 (R Core Team 2023). To get plot-level bee abundance and bee diversity measures, we first pooled specimen data across plot quadrants. We then tallied the number of bee specimens at the plot-level to calculate bee abundance. We calculated bee Shannon’s diversity (hereafter, bee diversity) through the vegan package. We used Shannon’s diversity index because it is sensitive to rare species and is appropriate when both abundant and rare species are being considered (Morris et al. 2014). Before testing hypotheses, we initially evaluated the correlation between sand content and D. purpurea abundance using Spearman’s rank correlation, finding the correlation to be not significant (r = -0.11, P = 0.56). We also tested whether D. purpurea abundance and plot-level floral richness significantly differed across sand content categories using ANOVA and verifying model assumptions. We evaluated the effect of sand content on the overall bee community. To assess how soil sand content affects bee abundance, we fit a generalized linear model with sand content category as a fixed effect. We used a Quasipoisson error distribution due to overdispersion. Additionally, we looked at how floral richness affected bee abundance using the same error distribution. For bee diversity, we fit a linear effect model with a Gaussian error distribution after ensuring normality. We fit both models using the lme4 package (Bates et al. 2015) and ran Tukey post-hoc pairwise comparisons with the emmeans() and contrast() in the emmeans package (Lenth 2023). We tested differences in bee community composition between the three sand content categories using PERMANOVA (adonis2(); Okansen et al. 2022) and visualized with nonmetric multidimensional scaling (NMDS) using a Bray-Curtis dissimilarity index via the vegan package. We then assessed whether host plant availability (i.e., the abundance of D. purpurea ) affected sand specialists, as well as Dalea spp. specialists. We first analyzed the interaction of D. purpurea and sand content on the proportion and diversity of sand specialists. We calculated the proportion of sand specialists by dividing the abundance of sand specialists by the total abundance of bees present at a given plot. We calculated the diversity of sand specialists using Shannon’s diversity through the vegan package. Both models were fit with a Gamma error distribution by adding 0.001 to all proportion and diversity estimates. We used the emmeans package (Bates et al. 2015) to run post-hoc pairwise comparisons. We also explored how floral richness affected the proportion of abundance of sand specialists using the same Gamma error distribution. We additionally checked whether D. purpurea alone affected the abundance and diversity of Dalea spp. specialists using a Poisson and Gamma error distribution, respectively. Dalea spp. specialist abundance and diversity was calculated exactly as described above for sand specialists. Once the interaction term of D. purpurea abundance and sand content was dropped, we analyzed how low, medium, and high sand content affected the proportion of abundance of sand specialists, as well as the diversity of sand specialists. We fit two separate generalized linear models with the lme4 package (Bates et al. 2015) using the same error distributions for the proportion of abundance of sand specialists and sand specialist diversity as described above. We again used the emmeans package (Bates et al. 2015) to run post-hoc pairwise comparisons. We added one post-hoc analysis after initial analyses to understand the effects that high sand presence might have on sand specialists, even if the soil sand content across a buffer was not high. To do this, we used a generalized linear model with a Gamma distribution to model the effects of the highest amount of sand present within a buffer on the proportion of bees that are sand specialists. 3 Results With 28 hours of active netting off Dalea purpurea across 28 plots, we collected 959 individual bees, comprising 84 species or morphospecies. Of these 84 unique taxa, eight (9.5%) were designated as sand specialists and 5 (6%) were Dalea spp. specialists. See Table S2 for a complete list of bee species with associated floral and sand specialization information, as well as their occurrence in each sand content category. Table S1 includes a complete species list with identification resources. There was no significant relationship between sand content and D. purpurea abundance (P = 0.74) or between sand content and blooming floral richness (P = 0.34), which showed that the floral community did not vary because of the amount of sand (Fig. S1). Total bee abundance was highest in plots with a medium amount of sand; this was significant compared to high sand plots but not low sand plots (P = 0.027; Fig. 2A). Unlike bee abundance, bee diversity did not significantly differ across sand content categories (P = 0.31; F = 1.2; Fig. 2B). Bee community composition, however, was marginally significantly dissimilar among sand content categories (P = 0.07; R2 = 0.1). PERMANOVA and subsequent pairwise comparisons revealed that low and medium plots (P = 0.008) were compositionally dissimilar from each other, while low and high (P = 0.21) and high and medium (P = 0.9) were not compositionally distinct (Fig. 2C; stress level = 0.2 with two dimensions). 3.1 Sand specialization Across all plots, we collected 197 individuals of 8 species of sand specialist bees (Table S2). High sand plots had the greatest diversity of sand specialists and low sand plots had the lowest diversity of sand specialists, but sand specialist diversity was not significantly different between medium and high sand plots (P = 0.92; T = -0.4). Low and medium sand plots (P = 0.04; T = 2.6), as well as low and high sand plots (P = 0.04; T = -2.6) did significantly differ in sand specialist diversity. Sand content similarly affected the proportion of abundance of sand specialists (P = 0.002), indicating high sand plots had the greatest proportion of sand specialists, while medium sand plots had the greatest raw abundance of sand specialists (Fig. 3). Counter to our hypothesis, sand specialists occurred across both medium and high sand plots, which was also evident through NMDS visualization of bee composition across sand content categories (Fig. 2C). This prompted us to investigate whether this pattern was driven by the presence of high sand patches rather than the average amount of sand on the landscape. When evaluating sand content by the highest percentage of sand available within a discrete patch within a plot buffer (hereafter maximum sandiest patch), rather than average overall sandiness, there was a significant relationship between the maximum sandiest patch and the proportion of sand specialists (P = 0.04; T = -2.123; Fig. 4). As the sand content in the maximum sandiest patch increased, so did the proportion of sand specialists, with results showing that medium sand plots all had discrete patches of soil with >80% sand (Fig. 4). In other words, although the average amount of sand in medium plots was moderate, high-sand areas were still present in discrete patches (Fig. 4). 3.2 Influence of Dalea purpurea abundance and floral richness In total, we collected 96 individuals of 5 species of Dalea specialists (Table S2). In comparison to soil sand content, the floral community had little effect on the bee community. The interaction of sand content and D. purpurea was not significant for both the proportion of abundance of sand specialists (P = 0.75) and overall bee abundance (P = 0.95). The same was true when evaluating the effect of D. purpurea on the proportion of abundance of sand specialists as a single fixed effect (P > 0.5; Fig. 5A). However, D. purpurea abundance as a single fixed effect had a significantly positive relationship with the abundance of Dalea specialist bees (est = 0.0003, P = 0.002), but not diversity (est = -0.002, P = 0.2). Results were consistently not significant when evaluating the effects of blooming floral richness on the bee community. Floral richness did not significantly affect overall bee abundance (P = 0.2) or the proportion of sand specialists (P = 0.75; Fig. 5B). 4 Discussion Floral and nesting resources are two critical habitat components that structure bee communities, yet nesting resources are often neglected as a strong determinant of bee distributions. We controlled for the effects of floral resource availability and focused on a dominant, core prairie plant species ( Dalea purpurea ). In doing so our findings highlight that soil properties, specifically soil sand content, are more important than floral resource availability in driving the distribution of many bee species, specifically sand specialists. This is one of few studies that challenges the commonly held belief that bee communities are more diverse in sandier areas, and one of the first studies to empirically support that sand specialist bees occur in sandier habitats. Contrary to expectations, we found that sand specialists were not limited to high sand areas. Rather, medium sand areas supported the most sand specialists, while host plant ( D. purpurea ) abundance and the number of flowering species present were not significant in structuring bee communities. Though there was a clear association between medium and high sand areas and sand specialists, further examination revealed that pockets of high sand areas were present at all medium sand plots. This suggests that bees can easily seek out high sand areas if they are present at a site or, alternatively, there may be a threshold amount of sand that is required for some sand specialist bee species to occur. Regardless, a fundamental finding of this work is how important sand, and broadly nesting resources, is in driving patterns of bee distribution. Our findings suggest that sand content, not floral resource availability, plays a primary role in structuring bee communities. This confirms that only focusing on floral resources, and not other components of habitat needs, limits our understanding of patterns of bee distribution (Potts et al. 2005; Grundel et al. 2010; Lazarina et al. 2016). Floral richness is not always a reliable indicator of bee community metrics, given a subset of plant species may play an outsized role at attracting both generalists and specialists, such as D. purpurea (Harmon-Threatt and Hendrix 2015). Despite controlling for D. purpurea presence and varying abundance with sand amount, its abundance did not structure bee composition. Contrary to expectations, bee diversity was not highest in sandier areas, but sand content did have an effect on bee community composition. Sand preference is seen as a community-wide trait and we therefore expected high sand areas to support the greatest bee diversity (Cane 1991; Harmon-Threatt 2020). Cane (1991) measured the percentage of sand at the nests of 32 bee species across North America. He recorded a strong connection between bees and sandy/sandy loam soils and exclusively found nests in sandy soils with sand percentages between 33-94%. This is why we expected a strong positive relationship between bee diversity and sand, but our results are more similar to Fortel et al. (2016) that found no relationship between soil texture and either bee richness or abundance. However, our findings suggest that bee composition varies with soil sand content, and as revealed through NMDS visualization, there are sand specialists that are associated specifically with both medium and high sand areas. This means that nesting specialization, like floral specialization, is a facet of habitat specialization and must be prioritized when considering habitat for bees (Roulston and Goodell 2011). Understanding how sand specialization drives bee occurrence and distribution is challenging given the paucity of research on this topic. In a literature review on nesting characteristics of wild bees, there were only 36 instances where authors wrote about broad associations between bees and sandy or sandy-loam soils (Harmon-Threatt 2020). We expected a strong relationship between sand specialization and high sand areas, but this hypothesis was informed by relatively few studies with little evidence for what qualifies sandy soil for bees. While there is a clear association between sand specialists and sandy soils, there may be some flexibility in this relationship. The vast majority (95%) of our designated sand specialists visited D. purpurea in prairies where soils were at least 44% sand, but they were not restricted to high sand areas. Further, not all sand specialists were exclusively associated with medium and high sand areas. For example, Colletes susannae Swenk, 1925, a designated sand specialist, infrequently occurred at two plots with very low sand (8 and 24%, respectively). This might be because soil sand content needs vary by sand specialist species. Alternatively, there could also be a lower threshold of sand that is required for sand specialists to construct nests than previously thought (Heneberg et al. 2018). A study conducted in permanent and ephemeral Czech wetlands documented the occurrence of sand specialists in non-sandy areas (Heneberg et al. 2018), though two problems exist with their methodology that make results incomparable: 1) they did not control for floral resource availability and therefore could not fully isolate the effect of soil texture; and 2) passive trapping (Moericke pan traps; Heneberg and Bogusch 2014) was used to collect bees, meaning it is difficult to show collected sand specialists were actually using and foraging from these wetland areas. Sand specialists may also be able to find high sand pockets even if the majority of the landscape has lower total sand availability. Our results suggest that discrete pockets of high sand across the landscape may be important in driving patterns of sand specialist bee distribution. All medium sand plots within our study had areas with at least 80% sand, and the two low sand plots with the sand specialist Colletes susannae had patches of at least 40% sand. Unfortunately, even though a clear relationship between sand specialists and the maximum sandiest patch available was evident in our results, this study was not designed to meaningfully understand how soil texture heterogeneity might influence bee distribution, nor did we search for nests in these high sand patches. One limitation of our work is that we looked at nesting indirectly. However, by limiting the extent of the study, controlling for floral resource availability, and not using passive trapping methods, our study is a much needed first step in assessing nesting specialization in bees. There may be value in directing monitoring efforts to focus on sampling bees from focal prairie plants rather than using passive methods that overcollect and diminish our ability to ask targeted questions. A worthwhile future research direction should be determining sand specialization thresholds through nest detection, especially given how limited our knowledge of nest selection is for wild and solitary bees (Harmon-Threatt 2020; Antoine and Forrest 2021). 5 Conclusion Although soil nesting resources may seem unlimited in comparison to floral resources (Roulston and Goodell 2011), our findings show that soil sand content can limit the distribution of bee species. We found that bee composition – not diversity – varies with soil sand content, and found direct evidence that supports sand specialization for a subset of species collected in this study. Nesting resource availability (i.e., sand availability) had a larger effect on bee distribution in comparison to floral resource availability. Most conservation or restoration efforts aimed at bees focus on increasing plant abundance and diversity without consideration of how to promote nesting habitat (Roulston and Goodell 2011). While it would not necessarily be possible to increase sand in restored areas (though see Mattson 2024), it may be advantageous to bee conservation to consider soil texture and heterogeneity into where to prioritize the conservation and restoration of habitats (Leone et al. 2022). Sandy areas should be prioritized because they contain both specialist and distinct bee communities. This may also be beneficial to conservation efforts given sandy areas are less arable and subsequently more affordable for conservation organizations to acquire (White 1999). Because Dalea purpurea is so widely used within grassland restorations, this study provides the context for how both nesting and floral resources may be leveraged in conservation and restoration decisions. Figure 1. Geographic distribution of the 28 plots across Minnesota, USA. Plots were designated as low, medium, or high sand by the soil sand content within a 300 m buffer surrounding each plot. The gray panel shows an example of a gSUURGO soil spatial data layer for a plot and surrounding buffer, with the corresponding USDA soil texture classes and sand content values. To obtain sand content values, we accessed the tabular gSSURGO data and performed four table joins to append the representative sand content value (“sandtotal1”) to each soil texture class key (“MUKEY”). We later joined the sand content value to the gSSURGO spatial data layer, using the “MUKEY” field. Figure 2. A) Bee abundance was highest in medium sand plots, with significant differences between medium and high sand plots (P = 0.03). The graph represents bee abundance in each sand content category and 95% confidence intervals. B) Shannon’s bee diversity did not significantly differ between sand content categories (P = 0.31), with the graph showing similar Shannon’s bee diversity values in each sand content category and 95% confidence intervals. C) Bee community composition differed across sand content categories (P = 0.07), with NMDS representing bee community dissimilarity in multivariate space across two dimensions. Sand specialists were only present in medium and high sand plots. \received DD MMMM YYYY \acceptedDD MMMM YYYY Figure 3 . The proportion of abundance of sand specialists significantly differed across sand content categories (P = 0.002), with the stacked bar graph showing medium sand plots had the highest abundance of sand specialists and high sand plots had the greatest proportion. Of the bees collected in each sand content category, low plots had 3.4% sand specialists, medium sand plots had 26%, and high sand plots had 34% sand specialists. Figure 4. Gamma regression of the maximum available sand at a given plot and the proportion of sand specialists (abundance) collected. Maximum available sand had a significantly positive relationship with the proportion of sand specialists (P = 0.04). Points represent each plot and are colored by sand content category, showing that all medium and high sand plots had pockets of soil with > 80% sand. Gray ribbon represents the 95% confidence interval. Figure 5. A) Gamma regression of the Dalea purpurea abundance at each plot and the proportion of sand specialists collected. The proportion of sand specialists was not significantly affected by Dalea purpurea abundance (P = 0.75). B) Gamma regression of blooming floral richness at each plot and the proportion of sand specialists collected. The proportion of sand specialists was not significantly affected by floral richness (P = 0.75). Gray ribbons represent 95% confidence intervals. References 1. Antoine, C. M., & Forrest, J. R. K. (2021). Nesting habitat of ground-nesting bees: a review. Ecological Entomology , 46 (2), 143–159. https://doi.org/10.1111/EEN.12986 Bartoń, K. (2023). MuMIn: Multi-Model Inference. R package version 1.47.5, Bates, D., Maechler, M., Bolker, B., Walker, S. (2015). Fitting linear mixed-effects models using lme4. Biesmeijer, J. C., Roberts, S. P. M., Reemer, M., Ohlemüller, R., Edwards, M., Peeters, T., Schaffers, A. P., Potts, S. G., Kleukers, R., Thomas, C. D., Settele, J., & Kunin, W. E. (2006). Parallel declines in pollinators and insect-pollinated plants in Britain and the Netherlands. Science , 313 (5785), 351–354. https://doi.org/10.1126/SCIENCE.1127863 Bogusch, P., Bláhová, E., & Horák, J. (2020). Pollen specialists are more endangered than non-specialised bees even though they collect pollen on flowers of non-endangered plants. Arthropod-Plant Interactions , 14 (6), 759–769. https://doi.org/10.1007/S11829-020-09789-Y/FIGURES/6 Brockmann, H. J. (1979). Nest-site selection in the great golden digger wasp, Sphex ichneumoneus L. (Sphecidae). Ecological Entomology , 4 (3), 211–224. https://doi.org/10.1111/J.1365-2311.1979.TB00578.XBuckner, M. A., & Danforth, B. N. (2022). Climate-driven range shifts of a rare specialist bee, Macropis nuda (Melittidae), and its host plant, Lysimachia ciliata (Primulaceae). Global Ecology and Conservation , 37 , e02180. https://doi.org/10.1016/J.GECCO.2022.E02180 Cane, J. (1991). Soils of ground-nesting bees (Hymenoptera: Apoidea): Texture, moisture, cell depth and climate. In Article in Journal of the Kansas Entomological Society . http://www.jstor.org/page/ Cane, J. H. (2008). A native ground-nesting bee (Nomia melanderi) sustainably managed to pollinate alfalfa across an intensively agricultural landscape. Apidologie , 39 (3), 315–323. https://doi.org/10.1051/APIDO:2008013 Cane, J.H. & Sipes, S. (2006). Floral specialization by bees: analytical methods and a revised lexicon for oligolecty. In NM Waser and J. Ollerton, editors. Plant-pollinator interactions: from specialization to generalization. Univ. Chicago, Chicago, IL, USA. p.99-122. Danforth, Bryan N., Minckley, Robert L. and Neff, John L. The Solitary Bees: Biology, Evolution, Conservation , Princeton: Princeton University Press, 2019. https://doi.org/10.1515/9780691189321 Environmental Systems Research Institute (ESRI). 2024. ArcMap Version 10.8.2 . Environmental Systems Research Institute (ESRI). Fortel, L., Henry, M., Guilbaud, L., Mouret, H., & Vaissière, B. E. (2016). Use of human-made nesting structures by wild bees in an urban environment. Journal of Insect Conservation , 20 (2), 239–253. https://doi.org/10.1007/S10841-016-9857-Y/FIGURES/5 Flora of North America Editorial Committee. (2021). North of Mexico Volume 11: Magnoliophyta: Fabaceae . Oxford University Press, Incorporated. Fründ, J., Linsenmair, K. E., Blüthgen, N., Fründ, J., Linsenmair, K. E., & Blüthgen, N. (2010). Pollinator diversity and specialization in relation to flower diversity. Oikos , 119 (10), 1581–1590. https://doi.org/10.1111/J.1600-0706.2010.18450.X Gardner, J., & Gibbs, J. (2020). The ‘red-tailed’ Lasioglossum ( Dialictus ) (Hymenoptera: Halictidae) of the western Nearctic. European Journal of Taxonomy, 725 , 1–242. https://doi.org/10.5852/ejt.2020.725.1167 Gardner, J., & Gibbs, J. (2023). Revision of the Nearctic species of the Lasioglossum ( Dialictus ) gemmatum species complex (Hymenoptera: Halictidae). European Journal of Taxonomy , 858 , 1–222. https://doi.org/10.5852/ejt.2023.858.2041 Gridded Soil Survey Geographic (gSSURGO) Database for Minnesota. United States Department of Agriculture, Natural Resources Conservation Service. (USDA NRCS) Available online at https://gisdata.mn.gov/en/dataset/geos-ssurgo Grundel, R., Jean, R. P., Frohnapple, K. J., Glowacki, G. A., Scott, P. E., & Pavlovic, N. B. (2010). Floral and nesting resources, habitat structure, and fire influence bee distribution across an open-forest gradient. Ecological Applications , 20 (6), 1678–1692. https://doi.org/10.1890/08-1792.1 Harmon-Threatt, A. N., & Hendrix, S. D. (2015). Prairie restorations and bees: The potential ability of seed mixes to foster native bee communities. Basic and Applied Ecology , 16 (1), 64–72. https://doi.org/10.1016/J.BAAE.2014.11.001 Harmon-Threatt, A. (2020). Influence of nesting characteristics on health of wild bee communities. Annual Review of Entomology , 65 , 39–56. https://doi.org/10.1146/ANNUREV-ENTO-011019-024955, Heneberg, P., Bogusch, P., & Řezáč, M. (2018). Numerous drift sand “specialists” among bees and wasps (Hymenoptera: Aculeata) nest in wetlands that spontaneously form de novo in arable fields. Ecological Engineering , 117 , 133–139. https://doi.org/10.1016/J.ECOLENG.2018.04.007 Hofmann, M. M., Fleischmann, A., Renner, S.S. (2020). Foraging distances in six species of solitary bees with body lengths of 6 to 15 mm, inferred from individual tagging, suggest 150 m-rule-of-thumb for flower strip. Journal of Hymenoptera Research, 77 , 105–117. https://doi.org/10.3897/jhr.77.51182 iNaturalist. Available from https://www.inaturalist.org. Accessed 2024. Kendall, L. K., Mola, J. M., Portman, Z. M., Cariveau, D. P., Smith, H. G., & Bartomeus, I. (2022). The potential and realized foraging movements of bees are differentially determined by body size and sociality. Ecology , 103 (11), e3809. https://doi.org/10.1002/ECY.3809 Lane, I. G., Portman, Z. M., Herron-Sweet, C. H., Pardee, G. L., & Cariveau, D. P. (2022). Differences in bee community composition between restored and remnant prairies are more strongly linked to forb community differences than landscape differences. Journal of Applied Ecology , 59 (1), 129–140. https://doi.org/10.1111/1365-2664.14035 Lazarina, M., Sgardelis, S. P., Tscheulin, T., Kallimanis, A. S., Devalez, J., & Petanidou, T. (2016). Bee response to fire regimes in Mediterranean pine forests: The role of nesting preference, trophic specialization, and body size. Basic and Applied Ecology , 17 (4), 308–320. https://doi.org/10.1016/J.BAAE.2016.02.001 Lenth, R. (2023). emmeans: estimated marginal means, aka least-squares means. R package version 1.9.0, . Leonard, R. J., & Harmon-Threatt, A. N. (2019). Methods for rearing ground-nesting bees under laboratory conditions. Apidologie , 50 (5), 689–703. https://doi.org/10.1007/S13592-019-00679-8/FIGURES/1 Leone, J. B., Pennarola, N. P., Larson, J. L., Oberhauser, K., & Larson, D. L. (2022). Divergent responses of butterflies and bees to burning and grazing management in tallgrass prairies. Ecology and Evolution , 12 (12), e9532. https://doi.org/10.1002/ECE3.9532;SUBPAGE:STRING:FULL Mattson, B. (2024). An attempt to promote ground-nesting bees in an urban landscape. Investigating the use of sand beds as a conservation effort. University of Gothenburg, MS Thesis. Michener, C. D. (2007). The bees of the world . The John Hopkins University Press. Minckley, R. L., Wcislo, W. T., Yanega, D., & Buchmann, S. L. (1994). Behavior and Phenology of a Specialist Bee (Dieunomia) and Sunflower (Helianthus) Pollen Availability. Ecology , 75 (5), 1406–1419. https://doi.org/10.2307/1937464 Morris, E. K., Caruso, T., Buscot, F., Fischer, M., Hancock, C., Maier, T. S., Meiners, T., Müller, C., Obermaier, E., Prati, D., Socher, S. A., Sonnemann, I., Wäschke, N., Wubet, T., Wurst, S., & Rillig, M. C. (2014). Choosing and using diversity indices: insights for ecological applications from the German Biodiversity Exploratories. Ecology and Evolution , 4 (18), 3514. https://doi.org/10.1002/ECE3.1155 Oksanen J., Simpson, G., Blanchet, F., Kindt, R., Legendre, P., Minchin, P., O’Hara, R., et al. (2022). vegan: Community Ecology Package_ R package version 2.6-4, Onuferko, T. M., & Rightmyer, M. G. (2024). A revision of the simplex species group of the cleptoparasitic bee genus Triepeolus Robertson, 1901 (Hymenoptera: Apidae). European Journal of Taxonomy, 950 , 1–106. https://doi.org/10.5852/ejt.2024.950.2643 Oram, R. J. (2018). Revision of the genus Hylaeus Fabricius (Hymenoptera: Colletidae) in Canada. Master’s thesis. University of Regina Available from: https://ourspace.uregina.ca/handle/10294/8896 Portman, Z. M., Arduser, M., Lane, I. G., Cariveau, D. P. (2022). A review of the Augochloropsis (Hymenoptera, Halictidae) and keys to the shiny green Halictinae of the midwestern United States. ZooKeys, 1130 , 103–152. https://doi.org/10.3897/zookeys.1130.86413 Portman, Z. M., Arduser, M., Powley, M. E., Cariveau, D. P. (2024). Taxonomy of Agapostemon angelicus and the A. texanus species complex (Hymenoptera, Halictidae) in the United States. European Journal of Taxonomy , 958 : 203–241. https://doi.org/10.5852/ejt.2024.958.2671 Potts, S. G., Vulliamy, B., Dafni, A., Ne’eman, G., & Willmer, P. (2003). Linking bees and flowers: How do floral communities structure pollinator communities? Ecology , 84 (10), 2628–2642. https://doi.org/10.1890/02-0136 Potts, S. G., Vulliamy, B., Roberts, S., O’Toole, C., Dafni, A., Ne’eman, G., & Willmer, P. (2005). Role of nesting resources in organising diverse bee communities in a Mediterranean landscape. Ecological Entomology , 30 (1), 78–85. https://doi.org/10.1111/j.0307-6946.2005.00662.x R Core Team. (2023). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.r-project.org/ Ricketts, T. H., Regetz, J., Steffan-Dewenter, I., Cunningham, S. A., Kremen, C., Bogdanski, A., Gemmill-Herren, B., Greenleaf, S. S., Klein, A. M., Mayfield, M. M., Morandin, L. A., Ochieng’, A., & Viana, B. F. (2008). Landscape effects on crop pollination services: Are there general patterns? Ecology Letters , 11 (5), 499–515. https://doi.org/10.1111/j.1461-0248.2008.01157.x Ritchie, A. D., Ruppel, R., & Jha, S. (2016). Generalist behavior describes pollen foraging for perceived oligolectic and polylectic Bees. Environmental Entomology , 45 (4), 909–919. https://doi.org/10.1093/EE/NVW032 Robinson, S. (2023). A taxonomic revision of the subgenus Protandrena ( Pterosarus ) (Hymenoptera: Master’s thesis. University of Manitoba Rowe, G. (2017). A taxonomic revision of the Canadian non- Osmia Osmiini (Hymenoptera: Megachilidae). Master’s thesis. York University Roulston, T. H., & Goodell, K. (2011). The role of resources and risks in regulating wild bee populations. Annual Review of Entomology , 56 , 293–312. https://doi.org/10.1146/ANNUREV-ENTO-120709-144802/CITE/REFWORKS Skiba, U., & Ball, B. (2002). The effect of soil texture and soil drainage on emissions of nitric oxide and nitrous oxide. Soil Use and Management , 18 (1), 56–60. https://doi.org/10.1111/J.1475-2743.2002.TB00050.X Weiner, C. N., Werner, M., Linsenmair, K. E., & Blüthgen, N. (2014). Land-use impacts on plant–pollinator networks: interaction strength and specialization predict pollinator declines. Ecology , 95 (2), 466–474. https://doi.org/10.1890/13-0436.1 White, J. (1999). Kankakee River Area Assessment, Volume 5: Early accounts of the ecology of the Kankakee River Area. Illinois Department of Natural Resources. Springfield, Illinois 62701. 444 pp. Wolf, A.T., & Ascher, J. S. (2008). Bees of Wisconsin (Hymenoptera: Apoidea: Anthophila). The Great Lakes Entomologist , 41 , 129-168. Wolf, S., & Moritz, R. F. A. (2008). Foraging distance in Bombus terrestris L. (Hymenoptera: Apidae). Apidologie , 39 (4), 419–427. https://doi.org/10.1051/apido:2008020 Data archiving statement: Upon acceptance, all data files and associated R scripts will be published through Dryad Digital Repository. Until acceptance, a private copy of the data is available on Dryad while this manuscript is in review: https://datadryad.org/share/bsG47O2gsiyxTvQnyG2lh7B7fOep822OLqtWsPecOl8 Crossref Google Scholar Information & Authors Information Version history V1 Version 1 06 August 2025 Peer review timeline Published Insect Conservation and Diversity Version of Record 22 Dec 2025 Published Copyright This work is licensed under a Non Exclusive No Reuse License. Keywords bees focal plant species nesting sand soil texture specialization Authors Affiliations Marissa Chase 0000-0003-4878-6658 [email protected] University of Minnesota Twin Cities View all articles by this author Nicole Gerjets Minnesota Department of Natural Resources View all articles by this author Ian G. Lane U.S. Fish and Wildlife Service - National Wildlife Refuge Program View all articles by this author Jessica D. Petersen Minnesota Department of Natural Resources View all articles by this author Zachary Portman University of Minnesota Twin Cities View all articles by this author Daniel Cariveau University of Minnesota Twin Cities View all articles by this author Metrics & Citations Metrics Article Usage 226 views 103 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Marissa Chase, Nicole Gerjets, Ian G. Lane, et al. Soil sand content is a driving force in structuring bee communities. Authorea . 06 August 2025. DOI: https://doi.org/10.22541/au.175449157.79398130/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.175449157.79398130/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ffde9331f71df94',t:'MTc3OTQ3NDgyNA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();