Acoustic comparison among soniferous sand gobies
The sounds of seven soniferous sand gobies, Knipowitschia panizzae Verga, 1841, Ninnigobius canestrinii (Ninni 1883), Orsinigobius punctatissimus (Canestrini 1864),Pomatoschistus marmoratus (Risso 1810), P. pictus (Malm, 1865), P. microps (Krøyer, 1838) and P. minutus(Pallas 1770), were previously recorded and characterised (Lugli et al., 1995; Lugli & Torricelli, 1999; Malavasi et al., 2008; Amorim et al., 2013, 2018; Bolgan et al., 2013; Pedroso et al., 2013). However, these acoustic data were never combined into a single phylogenetic dataset and analysed interspecifically. Therefore, we studied their interspecific acoustic variability (P. marmoratus was separated geographically into two populations, Italian and Portuguese). Briefly, the species were sampled either from brackish habitats in north Adriatic Sea (K. panizzae , P. marmoratus and N. canestrinii ), from freshwaters of north-west part of Reggio Emilia Romagna, Italy (O. punctatissimus ) (Lugli et al., 1995, 1997; Lugli & Torricelli, 1999; Lindström & Lugli, 2000), from Portuguese marine/brackish waters (Amorim et al., 2013, 2018; Bolgan et al., 2013) or the west coast of Sweden (Pedroso et al., 2013). Sound recordings gathered from the previously conducted laboratory experiments were re-analysed to allow for interspecific comparison with a minimal measurement experimental error. All investigated sand gobies produced pulsatile sounds, thus enabling acoustic interspecific comparisons. The dataset was composed of 36 individuals of eight soniferous sand gobies including O. croaticus (min – max: 3 – 5 individuals, except for a single individual of P. microps ), with at least three sounds recorded per individual. In total we calculated the means for five acoustic variables (temporal: DUR in ms, NP, PRR in Hz; spectral: PF and FM, both in Hz) for each individual. Since gobies included in the current study were recorded at different water temperatures (range: 15.8 – 22.6°C) and it is well known that the ambient water temperature affects fish acoustic signals (Vicente et al., 2015; Ladich, 2018), we conducted two separate multivariate analyses: the first involving the complete dataset (all five acoustic features for each species), and the second excluding the temporal features (DUR and PRR) known to be influenced by water temperature (Lugli et al., 1996; Vicente et al., 2015).