Ant colonies have vast nutritional needs. Foraging worker ants must meet not only their own nutritional needs but also those of their nestmates. Worker ants require primarily carbohydrates for energy, whereas the queen(s) and brood also require proteins for egg production and larval development, respectively (Markin, 1970; Sorensen and Vinson, 1981; Weeks et al., 2006). To meet these nutritional needs, ants engage in complex and diverse foraging activities. Ants obtain sugary honeydew excretions from hemipteran insects, hunt for insect prey, scavenge for deceased insects, feed on floral and extrafloral nectar as well as pollen, collect plant seeds, harvest plant foliage to cultivate mutualistic fungi, and acquire nutrients from animal excreta and carrion (Hölldobler and Wilson, 1990). The ants’ foraging activities may alter biotic and abiotic characteristics of their habitat, including the plant community composition (Halaj et al., 1997; Macmahon et al., 2000; Swanson et al., 2019).

The means by which foraging ants locate food sources have rarely been studied (Knaden and Graham, 2016). Engaging in certain foraging patterns may increase the likelihood of locating food (Dornhaus and Powell, 2010) but sensing and responding to cues from food sources would make foraging more energy-efficient. Visual cues associated with insect prey seem to guide some foraging ants (Baroni Urbani et al., 1994; Beugnon et al., 2001). Olfactory resource cues guide many foraging insects (Cardé and Willis, 2008; Webster and Cardé, 2017) and – as shown in a few studies – also guide ants (e.g., Zhou et al., 2012; Fischer et al., 2015). Some ants learn to associate odors with food sources, and via trophallaxis pass on food odor information to nestmates (Dupuy et al., 2006; Provecho and Josens, 2009; Nelson et al., 2019; Oberhauser et al., 2019). Innate recognition of certain food odors would expedite the process of locating resources that are reliably present, whereas learned odors may help locate and exploit fleeting resources.

Foraging requires energy expenditures not only for the locomotory physical activity but also for the maintenance of those sensory receptors and nervous tissues that inform foraging activities (Niven and Laughlin, 2008; Dornhaus and Powell, 2010; Elgar et al., 2018). As flightless foragers, ants have limited mobility and would accrue significant energy savings for themselves and for the entire colony, if they were able to track the odor plume from all valuable resources and pinpoint their location from a distance. Specific ant taxa are known to respond to odor cues from specific resources, such as deceased insects (Buehlmann et al., 2014; see below), honeydew (Zhou et al., 2012; Fischer et al., 2015; see below), and floral nectar (Schiestl and Glaser, 2012; De Vega et al., 2014; see below) but to date no study has investigated whether conspecific ants are able to respond to odor cues from multiple macronutrient sources including those consisting of mainly carbohydrates and proteins. This ability would be adaptive because foragers must adjust their foraging activities and priorities in accordance with their colony’s needs. When brood is present, they must collect not only more food but also more proteinaceous food (Cornelius and Grace, 1997; Dussutour and Simpson, 2008, 2009).

Protein-rich food sources such as insect prey, carrion, and animal excreta are often ephemeral. Challenged to locate them quickly, scavenging desert ants, Cataglyphis fortis, use olfaction to find deceased insects (Buehlmann et al., 2014), and the ponerine ant Pachycondyla analis and the formicine ant Crematogaster scutellaris exploit prey odor to locate termite and fig wasp prey, respectively (Schatz et al., 2003; Yusuf et al., 2014). As protein sources commonly release indole – which is a breakdown product of tryptophan and is an indicator of essential amino acid presence (Tomberlin et al., 2016) – many insects, including C. fortis, use indole as a generic semiochemical to locate protein sources (Chaudhury et al., 2015; Zito et al., 2015; Brodie et al., 2016, 2018; Cortez et al., 2017; Zhao et al., 2020). Similarly, workers of C. fortis are attracted to linoleic acid, a necromone indicative of deceased insects (Buehlmann et al., 2014).

Carbohydrates are vital to ant colony survival (Cook et al., 2010; Dussutour and Simpson, 2012; Bazazi et al., 2016; Arganda et al., 2017). Carbohydrates sought by ants originate mainly from floral and extrafloral nectar and sugary honeydew. Previously considered deleterious nectar thieves (Willmer et al., 2009), ants are increasingly documented as floral visitors and pollinators (De Vega et al., 2009; Czechowski et al., 2011; Luo et al., 2012; Ibarra-Isassi and Sendoya, 2016; Kuriakose et al., 2018; Del-Claro et al., 2019; Delnevo et al., 2020). Yet, there are still only a few examples of ant attraction to honeydew or floral and leaf semiochemicals (message bearing chemicals). Workers of both the black garden ant, Lasius niger, and the red-imported fire ant, Solenopsis invicta, are attracted to honeydew excreted by aphids (Zhou et al., 2012; Fischer et al., 2015). Similarly, workers of the African weaver ant, Oecophylla longinoda, respond to leaf odors of cashew trees, Anacardium occidentale, and protect these trees from herbivores in exchange for extrafloral nectar rewards (Wanjiku et al., 2014). Some species of ants respond to herbivore-induced plant volatiles (Agrawal and Dublin-Thaler, 1998; Bruna et al., 2008; Schettino et al., 2017), while others are attracted to floral odors of specific plants (Cytinus hypocistis, Chamorchis alpine), serving as their exclusive pollinators (De Vega et al., 2009, 2014; Schiestl and Glaser, 2012).

If worker ants had reduced olfactory acuity, like other flightless insects (Neupert et al., 2020), they would not likely be able to recognize the specific odor profiles of diverse food sources but might still be able to locate them by responding to key odorants shared between these sources. For example, many inflorescences that provide essential carbohydrates to pollinators share linalool and α-pinene as floral attractants (Knudsen et al., 2006; Nicolson, 2011). Whether foraging ants respond to specific or generic carbohydrate semiochemicals has not yet been investigated.

Western carpenter ants, Camponotus modoc, as a model species in our study, are commonly found in coniferous forests along the west coast of North America (Hansen and Klotz, 2005). They forage on aphid honeydew (Tilles and Wood, 1982; Renyard et al., 2021), scavenge arthropod prey (Hansen and Akre, 1985; Tilles and Wood, 1986), and feed on bird excreta, mammal urine, and carrion (AR pers. obvs). Of the many saccharides present in aphid honeydew, worker ants preferentially consume fructose and sucrose (Renyard et al., 2021) which are widely present also in floral nectar (Blüthgen et al., 2004; Woodring et al., 2004). The carpenter ants’ favorite saccharides also occur in the nectar of fireweed, Chamaenerion angustifolium (Antoń et al., 2017), and are likely present in the nectar of thimbleberry, Rubus parviflorus, which are two plant species common in forest clearings. The ants may less likely encounter hardhack, Spiraea douglasii – which thrives in more riparian habitats – but might still respond to its floral odor when presented with it. Whether carpenter ants are attracted to floral resources has not yet been studied but pollinivory by ants, including Camponotus carpenter ants (Czechowski et al., 2011; Cembrowski et al., 2015), is increasingly observed.

Here, we tested the hypotheses that (1) food sources rich in carbohydrates (aphid honeydew, floral nectar) and rich in proteins (bird excreta, house mouse carrion, cow liver infested or not with fly maggots) all prompt long-distance attraction of worker ants, and (2) attraction of worker ants to inflorescences (fireweed, thimbleberry, and hardhack) is mediated by floral semiochemicals that are shared between these plants.

Ant colonies have vast and diverse nutritional needs including carbohydrates and proteins (Porter, 1989; Evans and Pierce, 1995; Feldhaar et al., 2007; Blüthgen and Feldhaar, 2010; Mankowski and Morrell, 2014), but forager ants have limited mobility to meet these needs. Foragers face the challenge of not only finding enough nutrients for themselves but also for all of their nestmates including the developing brood (Csata and Dussutour, 2019). Foragers would likely accrue significant energy savings for themselves and for their entire colony if they were able to sense olfactory cues from both carbohydrate and protein sources, and to engage in long-distance orientation toward them. If forager ants had reduced olfactory acuity, like other flightless insects (Neupert et al., 2020), they might not be able to recognize the specific odor profiles of multiple food sources but might still be able to locate them by responding to key odorants shared between food sources. Working with Western carpenter ants as a model species, we show that foragers are capable of long-distance orientation toward both carbohydrate and protein food sources. Foragers were attracted to two of four carbohydrate sources (thimbleberry and fireweed) and to one of four protein sources (bird excreta) that we tested in bioassays. However, a blend of floral odorants shared between thimbleberry and fireweed was not attractive to ants, indicating that select floral odorants, while common among plants, are not attractive to ants when presented outside typical floral odor context. As ants were selective in their responses to carbohydrate and protein food sources, we conclude that they can discern between specific food odor profiles, and that they seem to have good, rather than poor, olfactory acuity.

All four sources of carbohydrates (aphid-infested conifer branches, blooming fireweed, thimbleberry, and hardhack) that we tested in our study would have provided nutritional value to foraging ants. Thus, it is surprising that aphid-infested branches and hardhack did not attract ants. Honeydew, in particular, is consumed by many species of ants – including C. modoc (Tilles and Wood, 1986; Yamamoto and Del-Claro, 2008; Ness et al., 2010; Renyard et al., 2021) – and may constitute a large proportion of an ant’s diet (Domisch et al., 2009; Pekas et al., 2011). That worker ants of L. niger and S. invicta were attracted to honeydew (Zhou et al., 2012; Fischer et al., 2015), but C. modoc workers were not (Figure 2), has at least three plausible explanations: (1) there simply may not have been sufficient honeydew accumulation on the aphid-infested branches; (2) the aphid honeydew, which at the time of excretion is odorless (person obs.), was not yet extensively colonized by exogenous microbes whose volatile metabolites attract natural enemies to aphid colonies (Leroy et al., 2011), and also accounted for attraction of L. niger workers (Fischer et al., 2015); and (3) Cinara aphids colonize tall conifer trees and their honeydew odors, or alarm pheromone signals, may not consistently reach ground-dwelling carpenter ants, providing little opportunity for ants to associate Cinara honeydew odor or aphid pheromones with the presence of aphids and carbohydrate rewards (Verheggen et al., 2012).

Insufficient overlap between habitats colonized by carpenter ants and hardhack may also explain the non-attractiveness of hardhack flowers. Hardhack thrives in open riparian habitats (Pojar et al., 1994) and is less common in areas frequented by carpenter ants (A.R.; pers. observ.), whereas fireweed and thimbleberry are common plant community members of the forest ecosystem (Pojar et al., 1994) inhabited by carpenter ants. Alternatively, the nectar or pollen rewards of hardhack are not sufficiently appealing to, or accessible by, carpenter ants.

Animal-derived nitrogenous sources such as bird excreta, vertebrate urine, and carrion are ephemeral resources. If foraging ants were to rely on chance encounters of these resources, they might not be able to meet the protein requirements of their colony’s egg-laying queen and developing brood. Expectedly then, worker ants were attracted to fresh bird excreta (Figure 2). While ants are known to forage on bird excreta (Kaspari, 1993; Jaffe et al., 2001; Sainz-Borgo, 2015), their olfactory attraction to bird excreta has not previously been reported. Bird excreta are nutritionally valuable to carpenter ants not only as a protein source, but also as a source of uric acid and urea. The ants’ obligate endosymbiont Blochmannia spp. enzymatically breaks down uric acid and urea, and converts urea to both essential and non-essential amino acids (Sauer et al., 2000; Feldhaar et al., 2007). This metabolic capability improves the nutritional intake of the host ants, enables them to persist on otherwise nutrient-deficient diets, and allows them to occupy nutritional niches off-limits to ant community members lacking these endosymbionts (Davidson et al., 2003; Feldhaar et al., 2007; Russell et al., 2009; Hu et al., 2018). The ants’ endosymbiotic ability to process urea as an amino acid precursor may also explain their lack of attraction to other protein sources, such as mouse carrion (Figure 2). Several species of carpenter ants selectively feed on urea (Shetty, 1982; Feldhaar et al., 2007; Menzel et al., 2012), and worker ants of Camponotus terebrans even sift through sand containing urea to acquire it (Petit et al., 2020).

Although our synthetic blend of bird excreta odorants was very complex, it still failed to attract ants in Y-tube olfactometers, indicating that essential constituents were still missing from the blend. These constituents could have been too polar to properly chromatograph [e.g., (bi)acids] or too low in abundance to be detectable in GC-MS analyses.

Over 154 species of ants, including C. modoc carpenter ants (Shean et al., 1993), have been found on or near carrion (Eubanks et al., 2019) but – surprisingly – neither mouse carrion nor cow liver at various stages of decay attracted carpenter ants in our study. The odor profile of decaying carrion dynamically changes in relation to the stage of decay (Dekeirsschieter et al., 2009), and each stage attracts a different guild of scavengers. For example, very fresh carrion attracts blow flies (Brodie et al., 2016), whereas the dry (bone and hair only) stage is attractive to clothes moths (Takács et al., 2001). The carrion stage that is preferentially sought by scavenging ants has not yet been investigated, and we may have presented a suboptimal stage in our bioassays. Alternatively, scavenging ants may prefer invertebrate to vertebrate protein, and vertebrate carrion protein is a suboptimal food source. This explanation is supported by findings that laboratory colonies of S. invicta and Solenopsis geminata had greater brood production and growth when provisioned with insect protein instead of liver protein (Gavilanez-Slone and Porter, 2013; Porter et al., 2015; Lin et al., 2022). It follows that ant assemblies on vertebrate carrion may be motivated primarily by prospective encounters with insect prey, such as fly maggots developing in carrion (Lin et al., 2022). In our study, the liver odor may have masked the (faint) fly maggot odor.

Our prediction that foraging ants are reliant upon a simple olfactory search ‘image’ for nectar odor cues was not supported by the data. The prediction was inspired by previous reports that flightless insects have poor olfactory acuity (Neupert et al., 2020). With a simple search image, ‘featuring’ only those floral constituents that are shared between plants, foraging ants would be able to locate and exploit multiple and diverse nectar sources. This concept seemed particularly appealing because more than 50% of flowering plant families have floral bouquets with overlapping constituents, including α-pinene, benzaldehyde, linalool, and E-β-caryophyllene (Knudsen et al., 2006). E-β-Caryophyllene and α-humulene are shared between fireweed and thimbleberry but this 2-component blend was not attractive to ants. This finding, coupled with (i) reports that ants have hundreds of olfactory receptors (Saad et al., 2018), and (ii) our data showing that foraging carpenter ants were attracted to complete floral odor bouquets of fireweed and thimbleberry but not of hardhack, support the conclusion that carpenter ants recognize specific floral odor blends and discern between them, and thus have good, rather than poor, olfactory acuity.

Western carpenter ants had no prior contact or experience with the carbohydrate and protein sources we tested, indicating innate responses. However, ants in general can learn to associate odors with food rewards (Dupuy et al., 2006; Provecho and Josens, 2009; Nelson et al., 2019; Oberhauser et al., 2019), and thus may be able to opportunistically adjust their foraging activities in accordance with the resources that are currently available in their habitat and that they have learned about. The ants’ disposition to respond to olfactory cues is likely affected by both resource-specific factors such as an optimally attractive stage (see above) and intrinsic ‘ant’ factors such as caste, stage, or hunger (Morgan et al., 2006; Seid and Traniello, 2006; Muscedere et al., 2012; Gadenne et al., 2016). The disposition to respond to olfactory cues is further modulated by shifting barometric pressure (Pellegrino et al., 2013) which could explain the variable number of non-responding ants in our bioassays. Spatio-temporal overlap of food and ant presence, and nutritional value of food, are obvious requisites for odor-mediated foraging responses by ants.

Foragers of Western carpenter ants are attracted to food sources rich in carbohydrates and proteins. The foragers’ ability to sense and orient toward sources of these two macronutrients greatly improves their foraging efficiency. Foragers are not reliant on chance encounters of these resources but can detect them from a distance and move upwind toward them. This ability likely translates into significant energy savings for the entire colony. A complex, rather than simple, olfactory search image seems to guide the foraging activities of ants. They discriminated between odor profiles of three flowering plants, selecting only two (fireweed and thimbleberry), and four protein sources, selecting only one (bird excreta). Moreover, the simple blend of only those two floral odorants shared between fireweed and thimbleberry had no ‘ant appeal.’ All these data support the conclusion that carpenter ants have significant olfactory acuity. With a keen sense of smell, flightless forager ants can efficiently locate valuable nutrient sources and meet the vast and diverse nutritional needs of all their worker nestmates, queen, and developing brood.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: doi: 10.17632/7kj4s38rvn.

AR and GG: conceptualization. AR, RG, SC, CG, GG, JB, and AM: experimental design, methodology, and writing—review and editing. SC, CG, AR, JB, and AM: data collection. RG: analytical chemistry. AR: statistical analysis and graphics. AR and GG: writing—original draft preparation. All authors contributed to the article and approved the submitted version.

The authors declare that this study received funding from BASF Canada Inc. and Scotts Canada Ltd. as the industrial funders. The funders were not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.