Continued source level reduction during attack in the low‐amplitude bat Barbastella barbastellus prevents moth evasive flight

Abstract

Ears evolved independently in many insect taxa due to the selection pressure of echolocating bats. Eared moths perform evasive flight manoeuvres upon hearing approaching bats, thereby substantially increasing survival probability. Accordingly, eared moths constitute only a minor portion of many bats’ diets. In contrast, the barbastelle bat Barbastella barbastellus almost exclusively feeds on eared moths by emitting low‐amplitude stealth echolocation calls that are undetectable by distant moths. While closing in on the prey, however, the calls might become audible. We thus hypothesised that barbastelle bats lower the source level of their calls even further while closing in, such that the level at the moth's ear remains below the moth's hearing threshold and the moth fails to elicit its evasive manoeuvre. We tested these hypotheses by offering tethered moths to barbastelle bats in the wild and in captivity. We reconstructed the bats’ three‐dimensional flight paths based on time‐of‐arrival differences of the echolocation calls at four microphones, measured the received sound levels at the moth's position with an additional miniature microphone, and calculated call source levels. We show that barbastelle bats continuously reduced source level upon detecting a moth, thereby delaying the time and shortening the distance when the moth might detect the bat. The received level at the targeted moth never exceeded the moth's neuronal hearing threshold by more than 10 dB, which is likely sufficiently low to prevent moth evasive flight. Barbastelle bats emit low‐amplitude calls to counter moth hearing. Continued amplitude reduction during an attack extends the time before the prey becomes aware of the predator, if at all. In combination, these two strategies allowed barbastelle bats to access a food resource which is largely unavailable to competing species, thereby probably altering competition between sympatric predators. A http://onlinelibrary.wiley.com/doi/10.1111/1365-2435.13073/suppinfo is available for this article.

Weather conditions determine attenuation and speed of sound: Environmental limitations for monitoring and analyzing bat echolocation

“…To estimate the localization error caused by errors in c , I generated the call sequences that a microphone array would receive from a bat at different spatial positions, and then analyzed these sequences with a different c than used during generation. Specifically, I used cross‐correlation to calculated TOADs for symmetrical planar star‐shaped four‐microphone arrays with 60 cm (e.g., Goerlitz, ter Hofstede et al., ; Hügel et al., ; Lewanzik & Goerlitz, ) and 120 cm intermicrophone distance, a c of 338 m/s, and a grid (2 m spacing) of bat positions filling half a hemisphere above the array up to 20 m distance ( x = 0–20, y = −20 to 20, z = 2–20; Figure ). The half‐hemisphere left of the array ( x = −20 to 0) was omitted as it is identical to the right one ( x = 0–20) due to the array's symmetry.…”

“…The speed of sound (c) is the speed by which sounds propagate through the air (often approximated by 340 m/s). It is the fundamental physical sound parameter by which bats compute object range based on echo delay, as well as the fundamental parameter underlying the acoustic localization of echolocating bats and other soundproducing animals (reviewed in Blumstein et al, 2011), a method that is increasingly used to obtain bats' spatial positions based on the differences in arrival time of the same call on multiple microphones (e.g., Fujioka, Aihara, Sumiya, Aihara, & Hiryu, 2016;Goerlitz, ter Hofstede, Zeale, Jones, & Holderied, 2010;Hügel et al, 2017;Lewanzik & Goerlitz, 2018;Seibert, Koblitz, Denzinger, & Schnitzler, 2013;Surlykke, Pedersen, & Jakobsen, 2009).…”

Neural representation of bat predation risk and evasive flight in moths: a modelling approach

“…While correlative studies show that moths initiate directional flight at 487 sound levels close to A1 cell threshold (Roeder 1962(Roeder , 1967, direct measurements of A1 488 activity in behaving moths would help to confirm the exact relationship between neural 489 activity and behaviour. Likewise, bat reaction thresholds are unknown and were set to 20 dB 490 SPL, an estimate of hearing threshold for many bat species, although behavioural reaction 491 thresholds can be higher (Lewanzik & Goerlitz 2018 Our model results provide support for the hypothesis that the A2 cell is adapted to fire 529 at a fixed distance between a particular bat species and all sympatric moths regardless of 530 moth species and size. If the A2 cell activity triggers last-ditch behaviour, this suggests that 531 last-ditch flight might be most effective when initiated at a distance to impact that is specific 532 to each bat species, not necessarily when the bat first detects the moth (as predicted by the 533 matched onset hypothesis).…”

“…Note that we performed linear reduced major axis regressions of the logarithmized values, which we backtransformed to the here presented potential function of the non-logarithmized data. Pvalues report the results for testing the hypothesis that there is no relationship between the log-transformed data with sequential Bonferroni-correction for seven tests.. We excluded the apparent source level of B. barbastella due to this species' unusual lowamplitude stealth echolocation strategy (Goerlitz et al 2010; Lewanzik & Goerlitz 2018) and the duration of R. ferrumequinum due to this species' use of high duty-cycle echolocation (Schnitzler 1968) as outliers from all regression analyses (open circles in panels a and c). For data sources, see Table S1.…”

Species‐specific strategies increase unpredictability of escape flight in eared moths

“…Although neuronal audiograms of multiple species suggest that our stimulus is above the threshold of the A2 cell of moths (Gordon & ter Hofstede, 2018;Ter Hofstede et al, 2013;Surlykke, 2003;Zha et al, 2009), little is known about how neuronal activity translates into evasive flight. Behavioural thresholds are generally higher than neuronal thresholds, although the exact differences and potential variation between species are mostly unknown (for discussion, see Lewanzik & Goerlitz, 2017) etation could be a potential anti-predator strategy, as close-by background structures impair bats' capture success due to sensory and motor constraints (Siemers & Schnitzler, 2004).…”