Enhancing flowering plant species richness at restoration sites can increase bee diversity and abundance (Fischer et al., 2016; Hanula et al., 2016; Purvis et al., 2020; Lane et al., 2022; Rubio et al., 2022; Beneduci et al., 2023) and bee visitation rates (Denning and Foster, 2018). A meta-analysis of observational studies by Kral-O’Brien et al. (2021) found that plant species richness was the strongest predictor of bee species richness. Other studies have reported comparable findings, indicating that vegetation type may significantly influence bee community assembly (Brooks, 2020; Novotny and Goodell, 2020). In addition, including high densities of flowering species through the implementation of seed mixes has also been found to increase the chances of pollination and reproductive success for some outcrossing plants (Cane et al., 2012), creating positive feedback loops between associated plant and bee species.

Nevertheless, the reintroduction or supplementation of appropriate combinations of native plant species at restoration sites may be difficult for several reasons. Seed mixes that represent local combinations of sympatric species are often unavailable, due either to their high cost in creating them, the difficulty of sourcing locally adapted genotypes, or challenges in producing seed mixes quickly enough (Nevill et al., 2018; Erickson and Halford, 2020). Nevertheless, carefully designed seed mixes that include seeds sourced from established “seed zones” (seeds from regions with similar environments; these seeds are considered the same in the context of locally adapted seed mixes; Erickson and Halford, 2020) can enhance bee diversity (Harmon-Threatt and Hendrix, 2015; Galea et al., 2016; Lybbert et al., 2022). Despite many benefits, seed zones are not defined for numerous important species in restoration (Johnson et al., 2023). Some research indicates that admixture seed sourcing (sourcing seeds from many different locations) can alter plant-arthropod interactions when flowering species richness is low (Hulting et al., 2024). However, there have been no studies examining the impact of admixture seed sourcing on pollination success or bee diversity.

Empirical tools can be beneficial for selecting plant species for bee habitat restoration (M’Gonigle et al., 2017; Esque et al., 2021; Purvis et al., 2021). M’Gonigle’s genetic algorithm, which uses phylogenetic relatedness, bee visitation rates, and bee diversity, is an effective tool for designing seed mixes (M’Gonigle et al., 2017) and has been empirically tested and used in multiple restoration efforts (Williams and Lonsdorf, 2018; Campbell et al., 2019; Bruninga-Socolar et al., 2023). Continued testing of empirical tools designed to facilitate the selection of plants that support generalist and specialist bee communities in different environments is needed.

Although most bee-centric conservation plans focus on floral resources, the availability of nesting habitats should not be overlooked (Öckinger et al., 2018; Requier and Leonhardt, 2020). Bee nesting biology has been recently reviewed by Orr et al. (2022), and anthropogenic threats to bee nesting in wild bee communities have been reviewed by Harmon-Threatt (2020).

The majority of bee species, including both eusocial and solitary species, nest underground (Danforth et al., 2019; Liczner and Colla, 2019); for a review of ground-nesting bee biology, see Antoine and Forrest (2020). Different species create distinct nest architectures and prefer different microhabitat conditions (Danforth et al., 2019; Antoine and Forrest, 2020). In a study by Buckles and Harmon-Threatt (2019) in tall grass prairies, bee nesting was positively influenced by increasing floral resource abundance as well as increasing the availability of bare ground, low soil moisture, and warmer soil temperature (Purvis et al., 2020). However, some bee species (e.g., bumble bees) prefer increased litter over bare ground (Williams et al., 2019; Smith DiCarlo et al., 2020).

Studies of ground-nesting bees have assessed the effect of ground cover, temperature, texture, space, slope, soil compaction, and soil moisture on nest site selection (Cane, 1991; Xie et al., 2013; Sardiñas et al., 2016; Tsiolis et al., 2022). A multitude of studies have revealed that landscapes undergoing early successional stages, such as habitat restoration efforts, often provide nesting habitats that support diverse and specialized bee species (Rutgers-Kelly and Richards, 2013; Řehounková et al., 2016; Banaszak and Twerd, 2018; Seitz et al., 2019; Mola et al., 2020; Simanonok and Burkle, 2020; van der Heyde et al., 2022). Biotic factors, including plants, pathogens, parasites, predators, and conspecifics, can also influence the nesting density and nesting location at which bees choose their nesting sites (Potts and Willmer, 1997; Michener, 2000; Requier and Leonhardt, 2020). For example, in Hawaii, the nesting sites of Hylaeus anthracinus Smith, 1853 experience lower reproductive success due to invasive ants (Plentovich et al., 2021). Limited research has explored the biotic factors influencing bee nesting, such as soil microbial diversity.

Another group of bees is comprised of native cavity-nesting bee species, which require live or dead biotic material in which to nest. Studies have shown that areas with simplified vegetative structures have low cavity-nesting rates (Flores et al., 2018; de Araújo et al., 2019, de Araújo et al., 2021; Felderhoff et al., 2022). Many bee species are opportunistic nesters, choosing to nest in existing holes, stems, or downed woody debris (Galbraith et al., 2019; Davis et al., 2020; Foote et al., 2020; Glenny et al., 2023; Rappa et al., 2023). In addition, specific habitat types may be preferred as overwintering sites, such as forest habitats for many Bombus sp (Mola et al., 2021a). In sum, whether or not there is sufficient availability of nesting sites for the bees at a given locality will depend on the type of biotic material available, the complexity of the vegetation structure, the presence of existing holes and debris, and the prevailing habitat type, depending on the bee species’ preferences.

The role of non-native plants in bee conservation is highly debated. Parra-Tabla and Arceo-Gómez (2021) provide an extensive review of the influence of invasive plants on plant-pollinator networks, although bees were not a focus. While native plants are recognized for supporting a wide array of bee species (Discua and Longing, 2022), several studies have found that non-native plants can also promote bee abundance and bolster pollination networks (Severns and Moldenke, 2010; Gaiarsa and Bascompte, 2022; Kovács-Hostyánszki et al., 2022). However, most studies of non-native plants and bees focus on the floral resources that non-natives provide (typically in urban environments) and do not account for competition in bee visitation rates between non-native and native plant species (Aizen and Morales, 2020) or disruptions in ecosystem function (Parra-Tabla and Arceo-Gómez, 2021; Tallamy et al., 2021). Hanula et al. (2016) note that when non-native plants outcompete native plant species, this typically negatively impacts pollinator communities, including bees. Additionally, Mathiasson and Rehan (2020) observed that the decline of native bees (particularly specialists) was associated with the proliferation of non-native plant taxa in northern New England due to the loss of their associated host plants. Moreover, non-native plants may affect other aspects of bee biology and ecology, including reproductive success (Hanula et al., 2016), the availability of ground-nesting sites, the abundance of native plant species (Barron and Beston, 2022), as well as floral visitor communities (Denning and Foster, 2017). In some habitats, the removal of non-native plants during restoration increased bee abundance and species richness (Fiedler et al., 2012; Tonietto and Larkin, 2018; Ulyshen et al., 2020). Forb diversity is often negatively associated with non-native grass abundance (Drobney et al., 2020; Molinari and D’Antonio, 2020), high levels of which may cause declines in bee abundance, the simplification of bee communities (Pei et al., 2023), and the alteration of entire insect communities (Luong et al., 2019). Furthermore, non-native grasses create dense litter layers that may block nesting sites for ground-nesting bees (Pei et al., 2023). For example, Pei et al. (2023) observed a decrease in ground-nesting bee abundance at sites occupied by increased leaf litter and high densities of non-native grass Poa pratensis L. in the Northern Great Plains. Other studies have hypothesized that non-native grasses are responsible for a decline in both forb and pollinator diversity and abundance (Lybbert et al., 2022).

It is important to recognize that not all non-native plants are invasive, and some non-native species can provide floral resources to support bee abundance (Carson et al., 2016; Frankie et al., 2019; Gibson et al., 2019; Niemuth et al., 2021; Ulyshen et al., 2022). This is especially the case at the beginning and end of the flowering season, when non-natives may be less likely to disrupt native plant-pollinator networks (Staab et al., 2020). During restoration efforts, non-native plants may serve as temporary food sources for bees while native plants become established (Lybbert et al., 2022; Thapa-Magar et al., 2023). While non-native plant species may support generalist bee abundance in some habitats, prioritizing native plants is recommended as they provide habitat for a broader range of native insects and contribute to ecosystem function (Tallamy et al., 2021).

Heavy machinery used to move soil, water, or vegetation during restoration can affect soil structure (Schäffer et al., 2007; Nawaz et al., 2013). Many species of ground-nesting bees require specific soil characteristics to build their nests (Antoine and Forrest, 2020). Christmann et al. (2022) suggest that heavy machinery could threaten existing nesting sites for ground-nesting bees; however, no studies to date have looked directly at the effect of soil movement or compaction from heavy machinery on bee nesting success.

Mechanical thinning has been found to have several effects on bee communities. When Ealy et al. (2023) compared pollinator communities in old-growth forests with logged early seral forests, they detected negative long-term effects on the bee communities in clear-cut sites, including a decrease in specialist bees. However, when comparing clear-cut vs. young forests with dense understories, clear-cut forests have higher bee diversity, likely due to decreased canopy cover (Ealy et al., 2023). Odanaka et al. (2020) conducted an experiment measuring the effects of mechanical thinning on bee diversity and abundance in longleaf pine savannas compared to untouched remnant plots. They found that bee diversity and abundance were positively correlated with thinning and negatively correlated with canopy cover. Lettow et al. (2018) found thinning coupled with controlled burns significantly increased pollinator richness and abundance in oak savannas relative to unmanaged controls. Other studies observed similar results (Hanula et al., 2015, Hanula et al., 2016; Abella et al., 2017; Milam et al., 2018; Rivers et al., 2018; Glenny et al., 2022; Davies et al., 2023). These findings highlight that mechanical thinning can increase bee abundance and diversity in certain vegetation types such as forests.

Hand thinning offers the advantage of leaving woody debris on the landscape which can support cavity, stem, and opportunistic nesting bees (Rappa et al., 2023). The maintenance of structural heterogeneity to provide nesting sites for diverse bee species should be supported during and following bee habitat restoration (Antoine and Forrest, 2020; Image et al., 2022).

Some plant and animal communities in North America have adaptations that enable them to thrive when exposed to periodic wildfires (Simmons and Bossart, 2020). Prescribed fire (designed to mimic conditions of periodic low-intensity wildfires) can result in bare ground which provides nesting habitat for ground-nesting bees (Hanula et al., 2016; Sitters et al., 2016; Decker and Harmon-Threatt, 2019; Bruninga-Socolar et al., 2022; Brokaw et al., 2023). Ulyshen et al. (2021) examined the effects of frequent prescribed fires on bee abundance and species richness in southeastern U.S. forests. They found that bee abundances significantly increased in burned plots compared to unburned plots, although bee species differed in their tolerance to burn frequencies. Similar results were obtained following controlled burns of tallgrass prairies (Harmon-Threatt and Chin, 2016) and mixed conifer forests.

However, not all species benefit from controlled burns, particularly cavity and stem nesters such as Bombus, Ceratina, and Osmia (Galbraith et al., 2019). Bruninga-Socolar et al. (2022) determined that ground-nesting bee abundance and diversity responded positively to fire, while cavity-nesting bee abundance and diversity increased in the absence of fire, highlighting the importance of heterogeneity in fire regimes. Moreover, plant-pollinator interactions can be disrupted in certain habitats after fire due to the elimination of floral host plants (Love and Cane, 2019). Despite this, Cole et al. (2019) found that burn scars, which contribute to environmental heterogeneity in riparian environments, were positively correlated with bee diversity. Other studies have observed similar results in different habitats (Gelles et al., 2022). In addition, controlled burns have been found to reduce non-native grasses (Ditomaso et al., 2006; Weidlich et al., 2020) and to increase annual wildflower diversity (Peterson and Reich, 2008; Davies and Sheley, 2011; Decker and Harmon-Threatt, 2019; Lybbert et al., 2022; Gelles et al., 2023), which may lead to increases in bee abundance and diversity (Smith DiCarlo et al., 2019). In general, pyrodiversity (the variability in burn size, frequency, duration, and severity across a landscape), including some exposure to high-severity wildfires, has been found to increase bee species richness in fire-adapted regions (Galbraith et al., 2019). Creating a mosaic containing different burn histories will likely provide habitat and resources for the greatest diversity of bee species (Ponisio et al., 2016; Rodríguez and Kouki, 2017; Galbraith et al., 2019).

Mowing is often used to manage weed and grass growth in restored habitats, especially during spring. Mowing has been found to promote forb diversity (Lybbert et al., 2022) but can have the opposite effect if done too frequently (Smith et al., 2018). Additionally, increased mowing frequency has been found to be negatively associated with bee species richness and abundance (Audet et al., 2021; Serret et al., 2022). The “No Mow May’’ movement has spread, in which residents are urged to reduce mowing during peak pollinator flight time (Andrews, 2023). This practice has been found to promote bee abundance on the US East Coast and elsewhere (Lerman et al., 2018), but it may require alterations when applied to other major geographic regions such as the western US, where peak pollinator foraging and flight times occur later in the season. Another method, such as the reintroduction of grazing animals such as wild horses, has been found to enhance forb diversity and boost bee abundance in habitats that historically evolved under herbivory from large ungulates (Garrido et al., 2019). Similarly, Bruninga-Socolar et al. (2022) found that the heterogeneity in vegetation cover caused by cattle grazing and controlled burns benefited ground-nesting bees by providing more bare ground, but implementation of specific grazing regimes is necessary to minimize soil compaction as well as providing habitat for stem and hole-nesting bees. In contrast, Stein et al. (2020) detected that grazing in grassland communities in the upper Midwestern United States led to a reduction in native flowering plant species abundance. In this study, body mass and lipid stores were also measured to assess nutritional health indicators in three sweat bees (Agapostemon spp.). It was found that in ungrazed sites, Agapostemon virescens (Fabricius, 1775) showed greater body mass compared to individuals sampled in grazed areas. Beckett et al (Beckett et al., 2022). determined that deer presence in British Columbia negatively affected bumble bee abundance indirectly by depleting floral resources, indicating a potential decline in colony success. For a review of the known roles of mowing and grazing in restoration as of 2016, see Tälle et al. (2016).

The impacts of pesticide use on native bee health are poorly understood. Experimental studies on honey bees are frequently used to infer the effects of pesticides on all bee species (Franklin and Raine, 2019; Lehmann and Camp, 2021). The U.S. Environmental Protection Agency’s Policy Mitigating Acute Risk to Bees from Pesticide Products states that protecting managed bees will “also protect native solitary and eusocial bees that are also in and around treatment areas” (EPA, 2015). However, honey bee sensitivity to pesticides may differ from the responses of native bees (Chan et al., 2019; Franklin and Raine, 2019). Because some solitary bee species are more vulnerable than honey bees to pesticide exposure, it is crucial to avoid relying solely on honey bees as the risk assessment model when observing the toxic effects of pesticides (Franklin and Raine, 2019). This increased susceptibility can be attributed to solitary bee consumption of fresher pollen and nectar, as well as increased exposure to pesticides through their nesting sites (Goulson, 2013; Chan et al., 2019; Franklin and Raine, 2019; Lehmann and Camp, 2021). In addition, solitary bees have a smaller body size than honey bees (Chole et al., 2019); thus, a dosage calibrated to honey bees could pose a significant risk to most wild bee species. This is particularly concerning as body size is one of the primary predictors of bee species’ vulnerability to pesticides (Schmolke et al., 2021). Furthermore, honey bees are eusocial and thrive in large colonies, whereas native bees are typically solitary and relatively scarce across the landscape. This trait makes them particularly vulnerable to population declines if negatively impacted by pesticides (Straub et al., 2015; Sgolastra et al., 2019).

Herbicides can be useful for removing invasive plants during restoration, but their costs and benefits should be considered before implementation (Bennion et al., 2020). Whenever feasible, employing biological controls can be highly effective and bypass the hazards associated with herbicides (Auld, 1998; Peterson et al., 2020). However, biological control agents are not available for all plant species (Singh et al., 2020). In a review of 372 published articles, Weidlich et al. (2020) reported that 42.3% of the restoration projects used chemicals to eradicate invasive plants. Of these, 40% used glyphosate, an active ingredient in most herbicides (Weidlich et al., 2020). Glyphosate, marketed as Roundup™, can be harmful and sometimes lethal to non-target pollinators including honey bees, bumble bees, and solitary bee species (Abraham et al., 2018; Battisti et al., 2021; Straw et al., 2021). When cavity nests were sprayed with glyphosate, solitary bee reproductive success declined due to reduced brood cell production (Graffigna et al., 2021). In an acute exposure experiment conducted under realistic field conditions, glyphosate exposure impaired fine-scale color recognition and long-term memory in bumble bees, which may disrupt their foraging behavior and lead to overall declines in colony success (Helander et al., 2023). Glyphosate use is restricted or banned in several European countries due to human and environmental concerns, including its negative effect on bee development, behavior, and survival (Kudsk and Mathiassen, 2020; Battisti et al., 2021). Further research is required to investigate the sublethal effects of glyphosate on native bee species observed in field settings, as noted by Battisti et al. (2021). While the effects of herbicides on bee health are considered in agricultural practices, these impacts have not been evaluated in habitat restoration efforts.

Although not commonly used in restoration, insecticides can occasionally be used to protect rare plants that are vulnerable to insect herbivores (Bevill et al., 1999; Flower et al., 2018). However, more commonly insecticides leach into native landscapes from neighboring agricultural fields or watersheds. Neonicotinoids, a class of widely used systemic neuro-acting insecticides absorbed by plants and spread throughout their tissues, are extremely harmful to bees (Alkassab and Kirchner, 2017). Transferred through pollen and nectar consumption, neonicotinoids cause bee mortality or have sub-lethal effects by altering bee communication, foraging behavior, or navigation (Fischer et al., 2014; Alkassab and Kirchner, 2017). Only two studies have investigated neonicotinoid exposure through soil contamination for ground-nesting bees, employing differing experimental designs and yielding conflicting results (Willis Chan et al., 2019; Tetlie and Harmon-Threatt, 2024).

While research on the effects of insecticides on native bee species has increased in recent years, more research is needed, especially on the sub-lethal effects of these chemicals (Dirilgen et al., 2023; Tetlie and Harmon-Threatt, 2024). Further research is needed to assess the impacts of pesticides on bee-centric restoration and to identify or discover practices that minimize negative outcomes. For instance, it has been shown that nighttime spraying is effective in reducing exposure to honey bees (Decourtye et al., 2023). Such studies will provide insights for practitioners to develop more informed, bee-friendly conservation and restoration strategies.