Consumer-Ready Insect Hotels: An Assessment of Arthropod Visitation and Nesting Success

Attracting beneficial arthropods to garden and landscape areas can increase insect biodiversity, promote arthropod-mediated ecosystem services, and promote overall ecological health (Häussler et al. 2017). To further encourage and increase insect visitation in gardens, insect hotels have been developed as a consumer product for homeowners and gardeners to use as a decorative addition to their landscape. Arthropod-mediated ecosystem services include biological control, pollination, and decomposition (Klein et al. 2007). Insect pests affect 37% of agricultural and ornamental crops in the United States annually and more than $30 billion is spent on pest control (Rufus et al. 2009). Biological control or the regulation of pests by beneficial insects is important not only for crop production but also for home gardens and landscape areas. Pollination is also necessary for crop production and wild plant species viability. Insect pollinators provide pollination for 35% of crop production and for 60–80% of wild plant species (National Research Council 2007).

Recent studies have evaluated habitat management to attract beneficial and pollinating insects to landscape and garden areas using flowering ornamental plant species (Davis et al. 2017, Harris et al. 2016, Motzke et al. 2016, Sutter et al. 2017, Tscharntke and Gathmann 1998). Habitat management provides insects with overwintering or nesting sites, alternative hosts, and pollen and nectar resources. The goal of habitat management is to increase pollinator and beneficial arthropod visitation and biodiversity in a garden, which can contribute to improved pollination and biological control, further reducing pesticide use (Gurr et al. 2017, Landis 2017).

Insect hotels can serve as potential habitats and overwintering sites for arthropods including bees, wasps, lady beetles, spiders, and lacewings. One study has reported increases in Polistes paper wasp species when insect boxes were established and monitored in residential locations (Frankie et al. 2005). Solitary bees utilize insect nesting boxes, and solitary bees include bumble, carpenter, leaf-cutter, mason, and sweat bees. The bee genera Hylaeus and Osmia occupy nesting boxes in urban gardens (Gaston et al. 2005). Likewise, other genera such as Xylocopa, Megachile, Centris, and Andrena inhabit nesting boxes (Gathmann and Tshcarntke 2002, Krombien 1967, Roubik and Villanueva-Gutiérrez 2009, Sihag 1993). Additionally, a few studies have reported the impact of nesting box distribution (i.e., location and nesting box placement) on visitation of pollinators and beneficial insects (Boyle and Pitts-Singer 2017, Fortel 2016, MacIvor 2016). Therefore, in the current study, we sought to evaluate the impact of insect hotel placement on arthropod visitation and diversity.

Commercially available insect hotels can be aesthetically pleasing when used in gardens, while acting as a functional nesting site for bees, wasps, and other beneficial arthropods (Rees 2008). Bee hotels have the potential to serve as alternative habitats for beneficial, pollinator, and plant-feeding insects (Le Roux et al. 2016). Although easily and inexpensively constructed from wooden blocks, bamboo stems, etc., a large variety of prefabricated “bee hotels,” “insect homes,” “nesting boxes,” etc. have been marketed in recent years. Various insect hotel designs and materials aimed at creating nesting habitats for different types of pollinators and other beneficial arthropods are often combined in a single design, for example, holes in a wooden block, hollow stems of different plant sources (smaller diameter straw and larger diameter bamboo). A pleasing aesthetic appearance is the goal, while also providing essential habitat features (e.g., metal roof to keep moisture from the box's interior).

However, several studies report that insect hotels may serve as “population sinks” for native bees as they can harbor parasites and pathogens (Moenen 2012, Macivor and Packer 2015). This may potentially be due to their high nesting densities, which are two-dimensional rather than multidimensional as natural nesting locations such as plant stems or decaying wood logs (Wcislo 1996). While insect hotels encourage different insect species to nest in the box, this cohabitation may actually increase parasite visitation (Lee-Mäder et al. 2010, Rosenheim 1990). Brood parasitism occurs when hymenopteran, coleopteran, or dipteran parasites lay their eggs in cells of pollen-collecting bee hosts. For example, parasitic wasps Melittobia acasta (Walker) and Coelopencyrtus spp., and the parasitic fly Cacoxenus indigator Loew attack mason bees (Moenen 2012). Microorganisms such as bacteria and fungi also may develop in the boxes due to condensation and moisture, affecting bee host health and potentially leading to brood mortality (Packer and Knerer 1986).

While ready-made nesting boxes are becoming widely available, their effectiveness to attract desirable insects has not been studied as well. The current study objectives were to (a) document pollinator and other predatory or beneficial insect taxa use of commercially available boxes, (b) quantify and compare arthropod visits and nesting among different taxa, (c) evaluate impact of floral resource proximity on abundance and diversity of arthropods visiting nesting boxes, (d) assess influence of tree placement on insect hotel visitation, and (e) determine potential of nesting boxes as arthropod overwintering sites. With answers to those questions, we hope to provide practical recommendations regarding arthropod visitation to the insect hotels, nesting box placement, and seasonality of insect nesting and visitation.

This study was conducted in the University of Georgia Research and Education Garden on the University of Georgia Griffin Campus (Spalding Co.; 33°24′67″N, 84°26′40″W). Below follows a general description of the garden, experimental design, and treatment layout. Butterfly and conservation theme gardens (hereto referred to as “the Garden”) were established in October 2012, each with 75 commercially available annual and perennial, herbaceous and shrubby, native and exotic plant species (Harris et al. 2016). Plants were chosen based on their attractiveness to pollinating and beneficial arthropods, horticultural attributes, and adaptability to the southeastern United States as well as the overwintering resources provided. Efforts were made to create aesthetically pleasing landscape designs in each garden necessitating the use of various foliage textures and contrasts (e.g., fine vs. coarse), a variety of plant habits (e.g., groundcover vs. upright), extended blooming (spring to fall seasonal interest) and low maintenance. Nonnative exotic species were included due to desirable horticultural characteristics such as colorful foliage, and included taxa such as Colocasia, Hibiscus, Stachys, and gingers. Foeniculum, Petroselinum, Melissa, and Passiflora were included for larval food source. Each Garden plot contained one of each mature tree species: river birch, Betula nigra L. and crape myrtle, Lagerstroemia indica L.

There were four replicate plots (total 1,705.2 m², 426.3 m² per plot) within the Garden; three were contiguous and separated by buffer pathways, while the fourth was located approximately 500 m away. In addition, two naturalized areas located 46 m away from the Garden served as controls. Each naturalized area contained one B. nigra and one L. indica of the same age as in the Garden plots and was surrounded by vegetative cover including annual weeds and volunteer species. Therefore, treatments were nesting box location (proximity to floral resources [within the four Garden plots] and away from floral resources [naturalized area controls]), and nesting box placement on trees (B. nigra or L. indica).

Insect hotels (18.62 cm × 14.36 cm × 32.19 cm) were purchased from a commercial supplier (Esschert Designs Insect Hotel, Gander Mountain, Greenville, NC; Fig. 1A, B). Insect boxes were secured with 2.54-cm brad nails at 1 m height on B. nigra and L. indica in each garden plot and in the naturalized area (control), for a total of 12 boxes. No insecticides, pesticides, or ant traps were used in proximity of the insect boxes.

Visual observations of nesting boxes were timed for optimal arthropod activity and took place at 10-min intervals from 1 p.m. to 3 p.m. Observations occurred twice a week over 3-mo period from June to August in both 2016 and 2017. During each 10-min observation time, arthropods landing on the insect hotel as well as those nesting in bamboo tubes and openings of the boxes were inspected and the number of bamboo stems and wooden drilled tunnels occupied were recorded (Fig. 1C). To assess overwintering arthropod populations, nesting boxes were removed in December 2017. Boxes were then stored in a freezer for preservation over a 3-wk period until arthropods were identified. Insect hotels were disassembled to expose inner tunnels and tubes. Bamboo tubes and wooden drilled tunnels were further examined for nesting and occupancy; determination of arthropod family was made based on presence of adults and/or other identifying features (i.e., leaf or mud material in brood cells) using keys including the online Discover Life Species Guides (Ascher et al. 2017).

Visual observations of arthropod visitors to the insect boxes, as well as the number of enclosed bamboo stems and drilled holes on the boxes were analyzed for each year. Efforts to distinguish returning female bees from new visitors were not made. Insect nesting box location (garden area vs. naturalized area) and nesting box placement (B. nigra vs. L. indica) were analyzed as the main effects. The main and interactions effects of the variables were included in the model. Data were subjected to analysis of variance using the general linear model procedure in SAS® (SAS Institute 2010). Data were transformed prior to analysis using the square root transformation for count data and back-transformed data are reported. Data from 2016 and 2017 were analyzed separately, and no direct comparison between both years was attempted. Main treatment means for visual observations and total enclosed bamboo stems and drilled holes were separated using Tukey's honestly significant difference test (SAS Institute 2010).

We first report on total arthropod visitations by taxa and functional groups followed by treatment comparisons.

Abundance of pollinator and predatory arthropod taxa. Pollinator taxa observed on insect hotels during June through August in 2016 and 2017 consisted of small bees (i.e., Halictidae, Colletidae, Andrenidae, and Megachilidae), potter wasps (Vespidae), thread-waisted wasps (Sphecidae), house flies (Muscidae), soldier flies (Stratiomyidae), midges (Chironomidae), and moths (Lepidoptera) (Table 1). For the purposes of our study, these groups were combined into a single category “All Pollinators,” which may include members of other functional groups as defined by Harris et al. (2017). Only one taxa, thread-waisted wasps (Sphecidae), had significantly higher numbers compared to other arthropod taxa in 2016 (F = 1.69; df = 22, 203; P = 0.024; Table 1). Predatory insect taxa observed at nesting boxes during June through August in 2016 and 2017 included spiders (Araneae), earwigs (Dermaptera), and predatory ants (Formicidae). For the purposes of our study, these groups were combined into a single category “All Predatory Arthropods,” which may include members of other functional groups as defined by Harris et al. (2016). Within that category two groups were highly significant, predatory ants (F = 6.14; df = 22, 203; P < 0.0001) in 2016, and spiders (F = 5.65; df = 24, 227; P < 0.0001) in 2017. These two taxa were primarily responsible for the significance of All Predatory Arthropods in each respective year (F = 6.16; df = 22, 203; P < 0.0001 in 2016 and F = 5.70; df = 24, 227; P < 0.0001 in 2017; Table 1).

Proximity to floral resources. Visitation of potter wasps (Vespidae) (F = 4.75; df = 1, 203; P = 0.031), soldier flies (Stratiomyidae) (F = 4.14; df = 1, 203; P = 0.04), predatory ants (Formicidae) (F = 27.99; df = 1, 203; P < 0.0001), spiders (Araneae) (F = 3.62; df = 1, 203; P = 0.06), and All Predatory Arthropods (F = 28.28; df = 1, 203; P < 0.0001) were significantly influenced by nesting box location in 2016 (proximity to garden areas vs. naturalized insect nesting box location; Table 2). In 2017, insect nesting box location (proximity to garden areas vs. naturalized area) significantly influenced visitation of spiders (Araneae) (F = 3.46; df = 1, 227; P = 0.064), All Pollinators (F = 3.26; df = 1, 227; P = 0.073), and All Predatory Arthropods (F = 4.05; df = 1, 227; P = 0.045; Table 2). Insect nesting boxes placed in close proximity to floral resources were visited more frequently by potter wasps, predatory ants, spiders, and All Predatory Arthropods as compared to controls (naturalized location) in 2016 (Table 3). In 2016, higher visitations of soldier flies were observed in boxes placed in the naturalized area. In 2017, spiders and All Predatory Arthropods were most frequently found visiting insect boxes located within the naturalized area (Table 3). Insect nesting boxes placed in garden areas had higher total visitations of All Pollinators compared to those in the naturalized areas (control) in 2017.

Placement on tree species. In 2016, nesting box placement significantly impacted visitation of thread-waisted wasps (Sphecidae) (F = 2.94; df = 1, 203; P = 0.089), predatory ants (Formicidae) (F = 59.75; df = 1, 203; P < 0.0001), and All Predatory Arthropods (F = 59.59; df = 1, 203; P < 0.0001) (Table 2). Predatory ants and All Predatory Arthropods primarily visited boxes placed on L. indica (Table 4). However, there were higher counts of thread-waisted wasps visiting insect boxes on B. nigra (Table 4). In 2017, nesting box placement was significant for spiders (Araneae) (F = 2.29; df = 1, 227; P = 0.022) and All Predatory Arthropods (F = 5.88; df = 1, 227; P = 0.016; Table 2). Higher counts of spiders and All Predatory Arthropods were found on insect boxes placed on B. nigra (Table 4).

Neither treatment (nesting box location or nesting box placement) significantly influenced total number of bamboo stems and drilled tunnels occupied during 2016 (Table 5). In 2017, however, location (F = 5.26; df = 1, 36; P = 0.023) and placement (F = 37.65; df = 1, 36; P < 0.0001) of nesting boxes significantly impacted arthropod occupancy (Table 5). Boxes placed in the plots with floral resources had higher number of bamboo stems and tunnels filled as compared to controls (naturalized area). When insect boxes were placed on L. indica trees, there were higher numbers of populated bamboo stems and drilled tunnels as compared to those placed on B. nigra trees (Table 5).

Although not subjected to statistical analysis, total counts for occupancy postexperiment (2017) were recorded (Table 6; Fig. 2). Out of 24 overwintering arthropods counted, 15 were found in insect hotels placed in garden areas and identified as small bee brood (n = 8), potter wasp (n = 4) brood, Araneae (n = 2), and Lepidoptera pupa (n = 1). In boxes removed from the naturalized control areas, there was a total of nine overwintering arthropods: small bee brood (n = 3), potter wasp brood (n = 3), and Araneae (n = 3).

Overall, in this 2-yr study, we observed low numbers of desirable pollinators and predatory arthropods visiting and/or nesting in consumer-marketed insect hotels. Highest numbers included predatory ants, which we did not attempt to control. For the average consumer, ants are not likely to be desirable insect visitors. To the best of our knowledge, this study is among the few to assess arthropod visits to consumer-ready insect hotels with respect to proximity to floral resources and nesting box placement on different tree species. Several studies have reported the importance of providing dense plantings as well as other materials such as old wood, stems, or nesting boxes which can serve as overwintering and sheltering sites for pollinators and arthropod predators (Dumroese et al. 2016, Goulson 2003, Kudo 2014, MacLeod et al. 2004). Providing overwintering locations in gardens and landscapes could aid in creating beneficial arthropod habitats and in the conservation of bees, pollinators, and other beneficial insects. Woltz et al. (2012) affirm that most diversity and abundance of arthropods are observed in complex landscapes replete with flowering plants, overwintering refuges, alternative hosts, and prey.

Insect groups observed in this study included members of the taxa Vespidae, Sphecidae, Stratiomyidae, predatory Formicidae (Solenopsis sp.), and Araneae. Our results suggest that insect boxes could serve as overwintering and nesting sites for certain pollinator and predatory arthropod groups including potter (i.e., Euodynerus sp.) and thread-waisted wasps. There was higher drilled tunnel and stem pollinator occupancy when insect hotels were placed in proximity to diverse floral resources (i.e., the Garden) and on L. indica mature trees. Although total overwintering arthropod counts were low, during certain study years, there were increased numbers of predatory and pollinator arthropods in boxes placed in garden areas.

Maclvor and Packer (2015) observed high numbers of solitary wasps (Vespidae) throughout their survey of insect hotels. Similarly, Jenkins and Matthews (2004) found high levels of Vespidae, Sphecidae, Megachilidae, and Apidae in nesting boxes in Georgia and South Carolina. Le Roux et al. (2016) also found that commercially available bee boxes have the potential to serve as alternative habitats for beneficial, pollinator, and plant-feeding insects. Solitary wasps have been known to nest in bee hotels placed in shaded areas (Taki et al. 2004). In the current study, although statistical significance was found for the tree species, results were not consistent from year to year and, in practical terms, the differences between the treatments were small. During summer boxes were shaded and in winter, boxes were in full sun as both tree species were deciduous. Therefore, further studies should be conducted to fully examine the importance of insect hotel placement and orientation on arthropod visits and nesting success.

The present study evaluates commercially available insect hotels specifically marketed for use by homeowners as a decorative addition to their landscape. Although marketing information on the box listed four types of pollinating and beneficial (predatory) insects—lady beetles (Coccinellidae), green lacewings (Chrysopidae), earwigs (Dermaptera), and mason bees (Osmia sp.) (Fig. 1B)—Dermaptera was the only taxon observed in the present study. With the exception of a single Dermaptera, none of these taxa were observed in our study. It is doubtful that the average consumer will be as accepting of wasp and ant visitors as mason bees and lady beetles. It is also pertinent to note that the structural integrity of the boxes was largely lost by the end of the 2-yr period; the red door fell off in over half of the boxes during the first year, necessitating frequent repair.

While there was pollinator and predator visitation and nesting within the commercially available nesting boxes in our study, it is important to note the potential of these insect hotels to negatively impact their targeted arthropods by promoting pathogens and parasites (Maclvor and Packer 2015, Rosenheim 1990). To gain deeper insights and to be able to provide practical guidelines to consumers, future studies should compare nesting success in naturally occurring overwintering sites versus marketed insect boxes in a single study. Through a study such as this, consumer-marketed insect hotel materials and designs may be improved to simulate natural nesting conditions and further boost nesting activity while stimulating brood maturation (Potts et al. 2005, Torné-Noguera et al. 2014).

With increased urbanization and natural habitat dispersion, the current study indicates that implementation of insect hotels in urban garden and in naturalized landscapes could successfully be utilized to enhance beneficial arthropod biodiversity. Further research is needed to increase understanding of insect hotel design and materials as well as insect hotel location and orientation which can further be provided to insect hotel manufacturers in order to minimize negative impacts of nesting boxes while serving as communal nesting sites for pollinator and predatory arthropod taxa.

We thank Martha Sikora, Asa Bradley, and Katie Gwaltney for technical assistance provided for the duration of this study and Jerry Davis for statistical assistance. We thank Scott Horn and Elizabeth McCarty for their internal reviews.