Acoustic masking of soniferous species of the St-Lawrence lowlands

Highlights

• Estimated acoustic masking between human- and animal-induced sounds in three soundscape contexts.

• Measured acoustic overlap in the timing, frequencies, amplitudes of signaling birds, anurans and orthopterans.

• Acoustic masking is unlikely for most species, but anurans in an urban context.

• Temporal overlap between anthrophony and biophony increased along the urbanization gradient for birds and anurans.

• Attention must be paid to other factors that covary with the amount of noise produced by human activities.

Abstract

How much of the acoustic resource shared by vocalizing animal species is masked by human-induced noise has become a central question in the emerging field of soundscape ecology. The goal of our study was to evaluate the joint probability that masking could occur for anuran, bird, or stridulating orthopteran species along the spectral, spatial and temporal dimensions of the acoustic space. We evaluated acoustic overlap in the dominant frequencies, absolute amplitudes (emitter-receiver distances), and time-series (temporal match-mismatch) of human- and animal-induced sounds in three landscape settings: urban, peri-urban and rural. Acoustic overlap was evaluated at the day scale during the active vocalizing period of each taxon. Our results suggest that acoustic masking is unlikely for a vast majority of species in the three taxonomic groups; ranging from 5–7% for anurans to 3–4% for birds and less than 1% for orthopterans. However, in urban contexts, acoustic masking could be a selective force when coupled to other factors that covary with the amount of noise produced by human activities, such as habitat loss and degradation.

Introduction

Communication in the animal kingdom is used by individuals to attract females during the reproductive stage, defend territories and signal food supply to conspecifics (Bradbury & Verenchamps, 2011; Gerhardt & Huber, 2002). Communication can be achieved with visual, olfactory or acoustic signals (Endler, 1993; Rekwot, Ogwu, Oyedipe, & Sekoni, 2001). In noisy environments, acoustic communication can be masked, thus forcing individuals to invest energy into compensatory mechanisms, such as shifting call frequencies (Lengagne, 2008, Slabbekoorn and Peet, 2003), vocalizing louder (Brumm, 2004; Cynx, Lewis, Travel, & Tse, 1998; Penna & Hamilton-West, 2006), or moving away from the noise source (Francis, Ortega, & Cruz, 2011; Goutte, Dubois, & Legendre, 2013). Previous studies have reported examples of adaptations to acoustic masking in anurans (Cunnington & Fahrig, 2010), birds (Hu & Cardoso, 2010; Wood and Yezerinac, 2006) and insects (Morley, Jones, & Radford, 2014; Shieh, Liang, & Chiu, 2012). However, it is not clear if a majority of species in these three taxonomic groups are affected by acoustic masking. For example, a recent meta-analysis by Roca et al. (2016) showed that only small-bodied bird species singing at low frequencies (<3 kHz) increased their dominant frequencies when exposed to anthropogenic noise, whereas anurans were less prone to such shifts.

To estimate the global importance of acoustic masking events, it is important to know the probability of spectral, spatial, and temporal overlap between human-induced and animal-induced sounds. Acoustic masking events occur if the communication signals are emitted i) at the same frequency (spectral overlap), ii) close enough to the noise source (spatial overlap), and iii) concurrently to human-induced sounds (temporal overlap) (Brumm and Slabbekoorn, 2005, Patterson and Green, 1978, Slabbekoorn et al., 2010). Human-induced sounds generally have low dominant frequencies and, as such, should not impact every vocalizing organisms equally (Hu & Cardoso, 2010; Slabbekoorn & Peet,2003). Anurans calls could be more susceptible to acoustic masking due to their spectral overlap with human-induced sounds (e.g., Cunnington & Fahrig, 2010). Hu and Cardoso (2009) similarly argued that birds singing at high frequency should predominate in urban ecosystems because their signaling frequency is out of the range of human-induced sounds.

Acoustic masking can be particularly important in urban settings because species did not coevolve with human-induced sounds (Feng & Schul, 2007). Although, data on the intensity level and the frequency spectrum of urban sounds are relatively easy to obtain (see Wood & Yezerinac, 2006; Wei, van Renterghem, De Coensel, & Botteldooren, 2016), little is known of the amount of temporal overlap between human-induced sounds and animal-induced sounds. Studies that explored the temporal patterns of vocal organisms were mostly conducted in natural ecosystems (e.g., Krause, Gage, & Woo, 2011; Rodriguez et al., 2014) or did not explicitly consider the anthrophony component of the soundscape (e.g., Gage & Axel 2014; Gage, Joo, Kasten, Fox, & Biswas, 2015). Hence, bird and anuran individuals exchanging acoustic signals close to a noise source, and at low frequencies, may avoid acoustic masking by vocalizing when temporal overlap is minimized during the day (Cartwright, Taylor, Wilson, & Chow-Fraser, 2014; Warren, Katti, Ermann, & Brazel, 2006).

The aim of the study was to evaluate the joint probability that acoustic masking could occur for anuran, bird, or stridulating orthopteran species along the spectral, spatial, and temporal dimensions of the acoustic space. We evaluated acoustic overlap in the dominant frequencies, absolute amplitudes (i.e., emitter-receiver distances), and acoustic patterns (i.e., temporal match-mismatch) of human- and animal-induced sounds in three soundscape contexts: urban, peri-urban, and agricultural. We paid particular attention to the daily temporal overlap between human-induced sounds (anthrophony) and animal-induced sounds (biophony), for which no estimates currently exist.