Developmental experience with anthropogenic noise hinders adult mate location in an acoustically signalling invertebrate

Abstract

Phenotypic plasticity facilitates survival and reproduction in rapidly changing and novel environments. Traffic noise spectrally overlaps with (i.e. masks) the sounds used by many acoustically signalling organisms to locate and secure mates. To determine if pre-reproductive exposure to noise improves adult performance in noisy environments, we reared field crickets (Teleogryllus oceanicus) in one of three noise environments: masking traffic noise, traffic noise from which frequencies that spectrally overlap with the crickets' song were removed (non-masking), or silence. At reproductive maturity, we tested female mate location ability under one of the same three acoustic conditions. We found that exposure to noise during rearing hindered female location of mates, regardless of the acoustic environment at testing. Females reared in masking noise took 80% longer than females reared in silence to locate a simulated singing male who was less than 1 m away. Impaired mate location ability can be added to a growing list of fitness costs associated with anthropogenic noise, alongside reductions in pairing success, nesting success and offspring survival.

1. Introduction

Adult traits, including behaviour, are shaped by ecological and social environments experienced during development and beyond [1], when adaptive, behavioural plasticity can reduce negative impacts of environmental change on individual fitness and enhance population persistence. Mating preferences and decisions are particularly plastic, varying, for instance, with risk encountered [2] and social experience [3]. Developmental experience alters adult mating behaviour in ways that likely reshape evolutionary trajectories (e.g. [4–5]). Here we ask how developmental experience with anthropogenic noise impacts reproductive components of fitness at adulthood [6], because noise transforms the mating environment [7].

Anthropogenic noise is a major and expanding human-induced global pollutant that can have dramatic physiological (reviewed in [8–10]) and behavioural (reviewed in [7,8,10,11]) impacts on animals. Noise could influence reproductive success through effects on signals and signalling strategies (e.g. [6,12]), contest behaviour, location of mates and mate preferences (e.g. [13]), nesting or pairing success (e.g. [14,15]) and parental investment (e.g. [16]). Much research has focused on whether signallers can improve detection in noisy environments (reviewed in [17]), but less attention has been paid to effects of noise on receivers [18–20]. Anthropogenic noise may impede receivers’ ability to locate signallers if it impacts hearing development, distracts mate searchers, masks acoustic cues or induces stress responses. Given this, we might expect receiver behaviour to depend on, and perhaps compensate for, experience with anthropogenic noise.

The taxonomic focus on vertebrate study systems in noise research [11,17,21] limits our understanding of the effects of anthropogenic noise on reproductive success (but see [14–16,18]); this is likely because of logistical difficulties measuring their fitness. Yet, mating behaviour is key to the evolution of male signals, and to fitness more generally. Switching the focus to invertebrates offers advantages: invertebrates compose most of the biodiversity on earth, are often small, have short generations, and can be laboratory-maintained under experimental conditions [11]. We use a field cricket study system to ask (i) does pre-reproductive exposure to anthropogenic noise impact adults' ability to locate mates, and, if so, (ii) does developing in noise improve performance in noisy environments?

Teleogryllus oceanicus lives in habitats ranging from urban lots in Australia to undisturbed fields on sparsely populated Pacific Islands. Traffic noise overlaps with the frequency of the calling and courtship songs males use to attract mates from afar and to entice them to mate once in close proximity (4–6 kHz). Females are locomotory and search for stationary calling males less than 1 m to more than 20 m away in a matrix of grass and rocks. We manipulated pre-reproductive experience with traffic noise, rearing female T. oceanicus under masking noise (traffic noise that overlaps spectrally and temporally with male calling song; electronic supplementary material, figure S1), non-masking noise (traffic noise that does not overlap spectrally with male calling song), or silence, and then tested adult female location of mates under the same three acoustic environments in a fully factorial design.

2. Material and methods

To produce masking and non-masking traffic noises, we recorded traffic noise at five Denver, CO, USA locations using a Marantz (PMD620MKII) digital recorder and Shure SM58 microphone. Locations captured varied vehicular types, volumes and speeds. We compiled two representative 30 s clips from each of the five locations into a single continuous 5 min track (electronic supplementary material, figure S1A). We produced a non-masking traffic noise by filtering out frequencies from 3 to 6 kHz using the ‘filter’ command in RavenPro14 (electronic supplementary material, figure S1B).

We pulled females from our laboratory stock (established in 2014 from Mo'orea, French Polynesia) when sex could be reliably identified and the hearing organs were apparent, and randomly assigned them to one of three acoustic rearing environments: masking (n = 44), non-masking (n = 43) or silence (n = 42). We broadcast the masking or non-masking noise inside the incubators (Percival I36VLC8) crickets were reared in for 14 h a day (1 h pre-dawn to 1 h post-dusk, mimicking traffic patterns) from EcoXBT wireless speakers. We rotated treatments among incubators every two weeks and rotated container positions within incubators during cleanings. Because incubators produce background noise (76–92 dBA), we kept them off during the entire experiment, but maintained a photo-reversed 12 h light–dark cycle. The temperature fluctuated between light and dark phases (21.2°C−30.5°C at the light source) but did not exceed those experienced in nature. We reared females in 64 oz Tupperware containers until sexual maturity with rabbit food ad libitum egg carton shelters and fresh water [3]. Females spent 15.5 ± 0.7 days in their rearing treatment prior to eclosion, regardless of treatment (F_2,126 = 2.58, p = 0.08, electronic supplementary material, figure S2).

We conducted phonotaxis (mate location) trials in a randomly assigned acoustic environment (masking, non-masking or silent) when females were 7 days post-eclosion. Phonotaxis trials took place inside of a square arena 1.45 m² in size, with a 10 cm grid on its floor, located within a 2.3 × 2 m room with acoustic foam-lined walls. We conducted phonotaxis trials 0–7.5 h post-dusk (mean = 2.9 ± 0.2 h). Time of testing did not differ among rearing (F_2,126 = 1.81, p = 0.17) or phonotaxis environments (F_2,126 = 1.50, p = 0.23). In each trial we placed the focal female at the centre of the arena under an inverted plastic cup for 2 min, after which we simultaneously released the female and projected (i) a strongly preferred T. oceanicus calling song (electronic supplementary material, figure S3) from a speaker in one randomly assigned corner and (ii) the noise treatment from a speaker suspended 141 cm above the arena. Both the song and noise treatment were broadcast at realistic volumes (70 dBA from the female's starting point) using EcoXBT wireless speakers. We measured the time to first movement, whether or not a female contacted the speaker broadcasting song, contact time (the difference between start of trial and touching the speaker), search time (the difference between time to first movement and contact time) and the number of grid lines females crossed (as a measure of search path). Trials lasted 5 min. Females who did not contact the speaker were assigned the maximum contact time.

We tested if experience with noise alters location of mates and whether developing in noise prepares females for mate searching in noisy environments using two-way analyses of covariance (ANCOVAs) in JMP Pro 13.0. Rearing environment, phonotaxis environment, and their interaction were main effects, and female pronotum width (size) was a covariate. Size did not differ across rearing environments (p = 0.76). We also considered whether females reared under noise shifted their mate searching behaviour temporally using ANCOVAs that included rearing environment, phonotaxis testing time (time post-dusk) and their interaction as main effects, and size as a covariate. Continuous outcome variables were natural log transformed to meet assumptions of normality. We ran logistic regressions (size = covariate) to address whether rearing or phonotaxis environments affected likelihood of contacting the speaker because the parameter estimates in the full model were unstable.

3. Results

Rearing environment was the most important predictor of adult female mate location behaviour (table 1). Differences in contact time (figure 1a) were due to the time it took females to initially move (figure 1b), rather than the search time or search path (number of grids crossed) (table 1). Females reared in masking noise took 209% longer to begin searching (figure 1b), and 81% longer to reach the signalling male than females reared without traffic noise (figure 1a). Surprisingly, the acoustic environment experienced during phonotaxis never influenced mate location behaviour (table 1; electronic supplementary material, figure S4), and we found no interactions between rearing environment and phonotaxis environment (table 1). Females who were larger were slower to begin moving and crossed fewer grids during the search (table 1; electronic supplementary material, figure S5). Female mate location behaviour did not depend on phonotaxis testing times (time post-dusk; all p > 0.39), or on the interaction between rearing environment and phonotaxis testing times (all p > 0.11) (electronic supplementary material, table S1). Of the 129 females, 120 contacted the speaker. Whether or not females contacted the speaker did not depend on rearing environment (χ²₂ = 5.77, p = 0.12) or phonotaxis environment (χ²₂ = 3.21, p = 0.36). All data have been deposited in Dryad [22].

Table 1.

ANCOVAs testing effects of rearing and phonotaxis noise environments on female location of mates. Significant p-values are indicated in italics.

outcome variable	effect	F	d.f.	p
time to first movement	rearing environment phonotaxis environment rearing × phonotaxis pronotum width	3.38 0.52 0.61 5.30	2,126 2,126 4,120 1,128	0.038 0.596 0.658 0.023
time to contact speaker	rearing environment phonotaxis environment rearing × phonotaxis pronotum width	3.18 1.42 1.12 0.60	2,126 2,126 4,120 1,128	0.045 0.247 0.353 0.808
search time	rearing environment phonotaxis environment rearing × phonotaxis pronotum width	1.50 0.81 1.12 3.21	2,126 2,126 4,120 1,128	0.226 0.445 0.351 0.076
no. grids crossed (search path)	rearing environment phonotaxis environment rearing × phonotaxis pronotum width	2.49 1.22 0.60 4.26	2,126 2,126 4,120 1,128	0.087 0.299 0.662 0.041

Figure 1.

Adult female mate location responses by rearing environment. (a) Time to first movement and (b) total time to contact the speaker from the start of the trial. Non-transformed means and standard errors are shown for ease of interpretation. Letters indicate statistically significant differences according to a Tukey's test (p < 0.05).

4. Discussion

Anthropogenic noise experienced prior to sexual maturity hindered adult mate location behaviour, regardless of the acoustic environment encountered at the time of searching. Females reared in masking noise took more than 200% as long to move and more than 80% longer to contact a simulated calling male than females reared in silence. Effects of previous and sub-adult exposure to noise may be underappreciated because studies often test for immediate behavioural responses (i.e. vigilance or foraging) to projected noise or make comparisons across habitats that are regularly exposed to more or less anthropogenic noise ([21], but see [23]). While certainly valuable, such studies can miss effects of prior exposure altogether or confound previous and current experience. In general, we expect masking noise to affect both signals and receiver responses [24], but organisms like singing insects that cannot alter their signal frequency plastically [25] or those that cannot quickly leave undesired areas [26] may suffer greater costs of noise, unless receiver behaviour can compensate.

Similar to our results (figure 1; see also electronic supplementary material), noise that masks a focal signal often elicits greater plastic and evolutionary change in signals and signalling behaviour than non-masking noise [17,27]. The mechanism underlying reduced mate location ability of females reared in masking noise is currently unknown, but we are testing several possibilities. There are strong effects of juvenile experience with sexual signals (or lack thereof) on adult mating decisions in this species [3], and masking noise might decrease signal detection during development, limiting learning opportunities, for instance. Alternative explanations for our results include generalized stress responses, physiological damage or impaired hearing development stemming from juvenile experience with masking noise [9].

We were surprised to find no effect of phonotaxis environment on female location of mates, though there is precedence for this in the literature (e.g. [25]). Nearly all females eventually located the speaker broadcasting song, which lends support to the hypothesis that juvenile exposure to masking noise produced a more generalized effect on physiology or learning that hindered location of mates, but that the broadcast noise did not completely eliminate females' ability to localize song.

With repeated adult exposure to noise, females may become tolerant, reducing the costs of developing in noise. However, our experimental design minimized factors other than noise that might impede female mate location. Animals searched for a highly preferred simulated mate who was less than 1 m away and experienced noise during their inactive period (daylight hours) for roughly two weeks prior to sexual maturity. Costs of developing in noise might be magnified by longer-term exposure, a search environment that includes males of varying attractiveness at more realistic (longer) search distances, and/or unpredictable onset and cessation of noise disturbances.

Data accessibility

Data can be accessed at https://doi.org/10.5061/dryad.53qb3 [22].

Authors' contributions

G.A.G.-S. and R.M.T. conceived of the study, and contributed to experimental design, data interpretation, and writing. G.A.G.-S. collected and analysed the data. Both authors approve the final content and are accountable for all aspects of the work.

Competing interests

We declare we have no competing interests.

Funding

Research was supported by grants from Sigma Xi and the Orthopterists' Society to G.A.G.-S.