No evidence of predator odor avoidance in a North American bird community

1.   Introduction

It was historically thought that birds were anosmic (Caro and Balthazart, 2010; Caro et al., 2015). However, birds have both functional olfactory bulbs (Balthazart and Taziaux, 2009) and diverse sets of olfactory receptor genes (Steiger et al., 2008, 2009a, 2009b, 2010; Driver and Balakrishnan, 2021). The use of odor has recently been established as an important mode of communication, especially in influencing mate choice in the context of species identification, (Bonadonna and Mardon, 2010; Whittaker et al., 2011a; Bonadonna and Sanz-Aguilar, 2012; Zhang et al., 2013; Van Huynh and Rice, 2019, 2021), sex discrimination (Soini et al., 2007; Zhang et al., 2009, 2010; Whittaker et al., 2010; Amo et al., 2011; Grieves et al., 2022; Krause et al., 2023), and assessment of individual quality (Whittaker et al., 2011b, 2013). For example, European Storm Petrels (Hydrobates pelagicus) use olfaction in kin recognition to avoid inbreeding while selecting an appropriate mate (Bonadonna and Sanz-Aguilar, 2012) while Black-capped (Poecile atricapillus) and Carolina chickadees (Poecile carolinensis) may reduce instances of hybridization by using odor cues (Van Huynh and Rice, 2019). Olfaction has also been shown to play a role in recognition of territory (Caspers et al., 2013; Golüke et al., 2016) and kin (Coffin et al., 2011; Krause et al., 2012; Caspers et al., 2015, 2017; Fracasso et al., 2019). Some species have been shown to use odor cues for foraging (Yang et al., 2015; Saavedra and Amo, 2018; Potier et al., 2019; Potier, 2020; Mahr et al., 2022; Rubene et al., 2022), even cueing in on herbivore-induced plant volatiles to find insect prey (Hiltpold and Shriver, 2018; Nguyen et al., 2022). In some species, odor cues are also important for selecting nesting material (Petit et al., 2002; Gwinner and Berger, 2008; Shutler, 2019) Lastly, navigation in birds, notably seabirds, is governed by olfactory cues (Hutchinson and Wenze, 1980; Pollonara et al., 2015; Abolaffio et al., 2018; Zannoni et al., 2020).

Many studies have investigated the ability of birds to respond to predator odors as a potential threat cue. There have been mixed findings in this context across a variety of avian species. Some birds show avoidance of predator odor cues (Amo et al., 2015), especially during nest visitation or nest site selection (Amo et al., 2008, 2011, 2017; Ekner and Tryjanowski, 2008; Eichholz et al., 2012; Stanbury and Briskie, 2015). However, some species do not respond to such predator odor cues in a nesting context (Godard et al., 2007; Johnson et al., 2011; Stanbury and Briskie, 2015; Stanback et al., 2019; Stanback and Rollfinke, 2023). Additionally, this behavior can be sex-specific. Male, but not female House Sparrows (Passer domesticus) have shown discrimination behaviors between predator and control odors at nesting sites (Griggio et al., 2016). Less work has been done on birds' response to predator odor cues in the context of foraging and current results are mixed (Roth et al., 2008; Zidar and Løvlie, 2012).

Here, we aim to measure bird responses towards predator odors in the context of natural foraging. To test if this behavior is appreciable as a general phenomenon above the individual species level, we quantify this potential effect in a community of eastern Pennsylvania songbirds in a natural setting.

2.   Materials and methods

2.1   Field setup

We placed 14 experimental locations around the DeSales University campus in east-central Pennsylvania (40°32′19″ N, 75°22′44″ W). Each location was either in an open field with tall grass (field location) or surrounded by trees (forest location). Each location contained three treatment sites containing one of three odor treatments, predator odor (bobcat urine), non-predator odor control (rabbit urine), or water control. We determined treatment assignments randomly. Treatment sites were spaced by a minimum of 60 m to prevent possible confoundment caused by scent proximity of the other treatments. At each treatment site, we placed a wooden perch (Fig. 1A). Perches were made of 0.95 cm-wide wooden dowels that were hammered into the ground (total above-ground height 0.7 m). A crossbar was secured with a black zip-tie (crossbar length 0.3 m) 2.5 cm from the top of the vertical dowel. On both ends of the crossbar, we placed clay caterpillars. To measure foraging rates, we created artificial caterpillars out of green clay (Sculpey Ⅲ, Leaf Green, Polyform Product Company, USA). The use of artificial caterpillars to reliably quantify bird foraging has been previously demonstrated (Hiltpold and Shriver, 2018; Nguyen et al., 2022). We rolled 2.5 g of clay into ~5 cm length and scrunched the caterpillar to resemble an inchworm-like shape. We pinned these caterpillars using bright orange pins to resemble the head capsule of insects. The caterpillars were pinned 2.5 cm from the end of each crossbar, with the "head" facing inwards (Fig. 1A, C). Urine dispenser vials were secured 20 cm above the ground with white lab tape. Dispenser vials held 2 mL of urine with a capillary tube inserted through the cap to allow the slow release of odor (Fig. 1B). Additionally, 10 mL of treatment solution (water, rabbit urine, bobcat urine) were sprayed at the base of the perch. Like many other felids, bobcats use urine to scent mark territories, marking on average every 0.13 km (MacDonald, 1980). In bobcats, it is thought that urination and olfactory investigation are the predominant method of communicating territory information as opposed to physical markings such as scrapings (Allen et al., 2015). Because good estimates of urine volume used by bobcats during territory marking are unknown, we chose our 10 mL treatment volume to reflect reported volumes of other predatory mammals such as foxes, coyotes, and wolves (Henry, 1977; Rothman and Mech, 1979; Bowen and Cowan, 1980). Bobcat and rabbit urine odors were detectable, at least to humans, after the initial setup and prior to daily urine reapplication (see below).

Figure  1.  Experimental setup illustrating (A) the wooden perch with two mounted artificial caterpillars and dispenser vial, (B) dispenser vial with capillary tube allowing for the slow release of volatile compounds, and (C) an artificial caterpillar showing peck marks.

DownLoad: Full-Size Img PowerPoint

2.2   Data collection

We set up the experiment on 5 June 2023. Experimental sites were checked daily for the following 7 days (6 June 2023 to 12 June 2023). We collected all data between 8:00 and 12:00 h. In order to quantify foraging on our caterpillars, we recorded whether the sites were visited (Y/N), the number of caterpillars pecked (0, 1, or 2), and how many total pecks were present (Fig. 1C). If the caterpillars were pecked or damaged we reformed and smoothed the clay back to its original form. If the caterpillar was irreparable or missing we replaced it with a new one. We refilled the dispenser vial if less than half the liquid volume remained. Additionally, 10 mL of treatment solution (water, rabbit urine, bobcat urine) were reapplied at the base of the perch at each site every day.

2.3   Statistical analyses

We fit mixed-effects repeated measures models for our three response variables—visits, the number of caterpillars pecked, and the total number of pecks, using the R package 'lme4' (Bates et al., 2015). We included the fixed effects of treatment (bobcat, rabbit, or water), day, location type (forest or field), and all interaction effects as well as a random effect of site location. Because of severe Canadian wildfires that took place during the course of our study, and which severely impacted the air quality of our study site, we also included an additional fixed effect of the daily air quality index (AQI) since the presence of smoke may have masked our odor treatments. We analyzed our mixed-effects model with a type-Ⅱ ANOVA using the R-package 'car' (Fox and Weisberg, 2019). All data is publicly available (Van Huynh, 2023).

3.   Results

There was no effect of treatment on any of our dependent variables (Table 1, Fig. 2). Although compared to rabbit and water controls, bobcat urine-treated locations experienced a slightly lower probability of visits, average number of caterpillars pecked, and number of average total pecks (Fig. 2), these differences were not significant (Table 1). We found a significant and near-significant two-way interaction of type by day concerning the number of caterpillars pecked and the total number of pecks respectively (Table 1, Fig. 3). Field sites were less preyed upon than forest sites during the beginning of the data collection period, but experienced a significant increase in foraging rates over the course of the experiment. Forest sites, on the other hand, experienced a significant decrease in foraging rates as the experiment progressed. Even though the air quality fluctuated dramatically during our study, at points even reaching hazardous levels of particulate matter, this did not seem to have a significant effect on foraging rates (Table 1, Fig. 3).

Table  1.  Analysis of variance (type-Ⅱ tests) of predation rates (visits, number of caterpillars pecked, and total number of pecks) by the fixed effects of treatment, type, day, all interaction effects thereof, and air quality index (AQI). Significant effects are highlighted in bold.

	χ2	df	p-value
Visits
Treatment	0.824	2	0.662
Type	0.093	1	0.761
Day	0.027	1	0.869
Treatment × day	0.624	2	0.732
Treatment × type	3.046	2	0.218
Type × day	2.684	1	0.101
Treatment × type × day	0.384	2	0.825
AQI	0.066	1	0.797
Caterpillars pecked
Treatment	1.148	2	0.563
Type	0.501	1	0.479
Day	0.073	1	0.788
Treatment × day	1.414	2	0.493
Treatment × type	1.828	2	0.401
Type × day	4.167	1	0.041
Treatment × type × day	0.712	2	0.700
AQI	0.279	1	0.597
Total pecks
Treatment	0.799	2	0.671
Type	0.062	1	0.803
Day	0.279	1	0.597
Treatment × day	0.310	2	0.857
Treatment × type	1.914	2	0.384
Type × day	3.371	1	0.066
Treatment × type × day	1.413	2	0.493
AQI	0.805	1	0.370

 | Show Table

DownLoad: CSV

Figure  2.  Predation rates for three odor treatment groups of bobcat urine (red), rabbit urine (green), and water (blue) illustrating (A) probability of visitation, (B) the average number of caterpillars pecked, and (C) the average number of total pecks. Error bars indicate standard error. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

Figure  3.  Time-series predation data for (A) caterpillars pecked and (B) total number of pecks across field (yellow) and forest (green) locations. Yellow and green lines indicate a linear model fits with 95% confidence intervals for field and forest locations respectively. The blue line indicates daily AQI. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

DownLoad: Full-Size Img PowerPoint

4.   Discussion

Here we tested the ability of bird species in a natural community in eastern Pennsylvania to avoid predator odor cues in the context of foraging. We did not observe any significant difference in attempted predation on artificial caterpillars between bobcat treatments and control treatments (rabbit and water). Overall, our findings reflect past mixed results of individual species' responses to predator odor cues. Some species do show avoidance behaviors in response to predator odors. For example, Blue Tits (Cyanistes caeruleus), Great Tits (Parus major), House Sparrows, and some dabbling ducks exhibit anti-predator behaviors towards predator odors at nesting and roosting sites (Amo et al., 2008, 2011, 2017; Ekner and Tryjanowski, 2008; Eichholz et al., 2012; Griggio et al., 2016). Other birds such as House Finches (Carpodacus mexicanus) (Roth et al., 2008) and domestic fowl (Zidar and Løvlie, 2012) also exhibit anti-predator behaviors towards predator odor cues during feeding. Conversely, other species including European Starlings (Sturnus vulgaris), House Wrens (Troglodytes aedon), Eastern Bluebirds (Sialia sialis), Tree Swallows (Tachycineta bicolor), and Wedge-tailed Shearwaters (Ardenna pacifica) exhibit no observable behavioral response to predator odor cues in various contexts (Godard et al., 2007; Johnson et al., 2011; Gérard et al., 2015; Blackwell et al., 2018; Stanback et al., 2019; Stanback and Rollfinke, 2023).

We did find a significant time by type of location (forest vs. field) interaction. We saw a decrease in the foraging rates in forest sites and an increase in the foraging rates in field sites over the course of the experiment. Our field locations were all placed in open fields with tall grass. Birds may have initially avoided field locations because of the ability of predators to hide in the tall grass. However, over the week-long duration of the experiment birds may have become less inhibited with the lack of any reinforcement of the presence of an actual predator. The forest experimental locations experienced less foraging as the experiment proceeded. Visiting birds may have become accustomed to the experimental setup, for example learning that the caterpillars were fake. However, further investigation on potential habituation effects is warranted to verify these possibilities. Additionally, there seemed to be no effect of air quality on avian foraging rates. Air quality fluctuated dramatically over the course of our experiment, reaching hazardous levels on several days, resulting in limited visibility and a strong odor of smoke, at least to humans.

The absence of significant differences between odor treatments could be due to a number of factors. Unlike other studies that examined anti-predator behaviors in response to predator odor in individual species (Godard et al., 2007; Amo et al., 2008, 2011a, 2011, 2015, 2017; Roth et al., 2008; Johnson et al., 2011; Eichholz et al., 2012; Zidar and Løvlie, 2012; Griggio et al., 2016; Blackwell et al., 2018; Stanback et al., 2019; Stanback and Rollfinke, 2023), we did not control for or record the specific species that visited our experimental setups. Common insectivorous songbirds present in our geographic study site include Gray Catbirds (Dumetella carolinensis), Tufted Titmice (Baeolophus bicolor), Eastern Bluebirds, chickadees, and Blue Jays (Cyanocitta cristata) among many others. In total, the community of songbirds present at our study site as a whole did not exhibit any significant avoidance of predator odors, although unobserved species-specific effects may have been present. We likewise did not record the sex of visiting birds, and sex-specific responses to predator odor cues have been observed in some species (Johnson et al., 2011; Griggio et al., 2016). Furthermore, we only recorded evidence of attacks and did not record individual behaviors at our experimental locations. Therefore, we do not know if birds displayed anti-predator behaviors such as latency to land or vigilance behaviors at bobcat treatment sites. Lastly, we collected this data in June of 2023, which means that the local songbird community was composed of both adult birds and recently fledged individuals. It is unknown to what extent odor-based behaviors are learned in birds, but these and other experience-based behaviors may be absent or reduced in the fledgling population. Even if predator odor avoidance has innate genetic components, suboptimal variation in this trait in the naive fledgling population may not yet have been purged by natural selection. An example of a similar phenomenon in a complex behavioral trait can be observed in the spatial learning ability of first and second-year Mountain Chickadees (Poecile gambeli), where individuals with lower cognitive abilities are less likely to survive their first winter (Sonnenberg et al., 2019). Future studies may wish to investigate behaviors of birds towards predator odors with respect to seasonality, i.e. fledgling vs. adult populations. Additionally, Canadian wildfires that took place during our data collection led to high volumes of smoke and particulate matter affecting the air quality in our study location. While we did not observe a statistical effect of the changing degree of air quality, this has the potential of changing the perception of the odors and overall behaviors. Lastly, we only assessed behavioral responses in the context of foraging. Previous studies have found a significant response during foraging in the domestic fowl (Zidar and Løvlie, 2012), while another study found only weak effects in House Finches (Roth et al., 2008). Birds may be more apt to respond to predator odor cues in other contexts, such as nesting, where there is threat to offspring and where predation risk may be higher (Amo et al., 2008, 2011b, 2017; Ekner and Tryjanowski, 2008; Eichholz et al., 2012).

Previous studies of predator odor avoidance in birds have shown mixed results. Some studies show evidence for the use of olfaction when avoiding predators (Amo et al., 2008, 2015, 2017; Eichholz et al., 2012) while other studies suggest that birds do not use olfaction during foraging or nesting behaviors (Godard et al., 2007; Amo et al., 2012; Gérard et al., 2015; Stanback et al., 2019; Stanback and Rollfinke, 2023). Our results, though non-significant, nevertheless reinforce the idea that avoidance of predator odors by birds may not be ubiquitous across contexts and species.

Ethics statement

This study was carried out in compliance with the Ornithological Council's Guidelines to the Use of Wild Birds in Research. As this field study involved no invasive procedures, harms, or material alteration of individual birds, under the Animal Welfare Act regulations [9 CFR 2.31 (c)(2) and 9 CFR 2.31(d)] this work is exempt from the site inspection and protocol review procedures.

CRediT authorship contribution statement

Austin Dotta: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Batur Yaman: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Alex Van Huynh: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We thank the DeSales University Biology Department and Women for DeSales for supporting this work. We also thank Wendy Maybruck for her logistical support.