Fish sampling, laboratory housing and experimental design
Orsinigobius croaticus was sampled using electrofishing (Hans Grassl, model: EL65 IIGI, power: 13 kW) from a boat during spring 2019. Sampling was performed on the Matica River in Croatia (near the village Vina 43°10’30.33” N, 17°23’12.36” E). In total, we collected 25 individuals (15 males and 10 females) from the main river channel at a depth of 0.5 – 2 m. Fish were transferred to large plastic water containers equipped with aerators and transported to the laboratory. At the laboratory, fish were sexed based on urogenital papilla and body coloration (Miller, 1984) and housed in appropriate community rectangular tanks (120 L). After an acclimatization period of 5 – 8 days, prospective soniferous males were chosen for subsequent laboratory acoustic-visual recordings based on coloration (complete or partially darken body, fins and head; Zanella et al., 2011), aggressiveness and territoriality. Eight males ( ± s.d. = 49.21 ± 0.8 mm total length, LT; range: 48.07 – 50.06 mm; 41.07 ± 1.01 mm standard length, LS; range: 38.96 – 42.29 mm; 1.24 ± 0.11 g weight, W; range: 1.10 – 1.40 g) exhibiting typical reproductive behaviour were chosen for the experiments. Females (N = 5; 36.75 ± 5.10 mm LS; range: 28.75 – 41.27 mm) were chosen for the recording sessions according to yellow belly coloration, luminescent green spot on the first dorsal fin and dark eyes, all indicators of female readiness for spawning (Zanella et al., 2011; Bloom et al., 2016). Measurements were made using digital callipers CD-15APX with a precision of 0.01 mm (Mitutoyo, Japan) and a digital scale (0.1 g precision). As a metric of male body condition, we calculated the condition factor [Fulton’s K , where K = (W/LS3) *105].
We followed the acoustic-visual recording protocol established by previous authors (Amorim & Neves, 2007; Amorim et al., 2013; Pedroso et al., 2013), where experimental tanks were divided into three compartments separated by removable partitions. Each lateral compartment housed one territorial male with a nest (artificial pvc shelter), while the middle section (“arena”) was occupied by a ripe female. The female compartment was not provided with a nest. Briefly, our experimental rectangular tanks were 120 L (60 x 48 x 35 cm) in volume and were divided into three unequal sections by means of two opaque removable acrylic partitions to prevent intersexual (visual) interactions. The tanks had a 5 cm thick layer of fine sand or gravel and each male in each section was provided with a water pump system and aeration. The experiments were performed from mid-April to October, at random times. The reproductive behaviour of resident males was elicited by introducing one ripe female into the “arena”. Before experiments, each female was left 12 – 24 h in the experimental tank for acclimatization. Eight territorial males were kept in the experimental tanks until the end of the recording sessions in October, after which they were returned to male community tanks. Prior to recordings (approx. 15 min), electricity, water pumps and aeration were switched off to minimise ambient noise. Between recording sessions of the same resident male, 15 – 30 min pauses were included. In addition, to reduce substrate-born noise, the experimental tanks were placed onto three cm thick rubber foam shock absorbers. Fish were kept at natural photoperiod and fed daily ad libitum with Daphnia . Water temperature, monitored with a thermometer (AQUATERRA, Garešnica, Croatia), was maintained between 18 – 22°C. The male-female trials lasted approximately 30 min, and began by removing one of the lateral partitions, allowing intersexual interaction. To eliminate the noise from the room light system, recording sessions were performed in the dark, with the experimental tank illuminated by LED light from the side of the aquarium. This procedure had no noticeable effects on fish behaviour.