Fish predation by the long-fingered bat (myotis capaccinii)ecological, behavioural and sensorial clues for a trophic shift

Resumen

The long-fingered bat Myotis capaccinii is a European trawling bat reported to feed onfish in several locations in the Mediterranean, and known to be able to gaff fish from the watersurface when fish occur in high densities. However, the ecological circumstances of thisbehaviour, the conditions bats need for fishing, as well as the detection, capture and handlingtechniques used by them in the field have not yet been studied. To elucidate the importanceof fishing in this bat¿s diet and to study the sensorial and motor plasticity in echolocationand flight that allows M. capaccinii to fish I carried out a study in Dénia (Eastern IberianPeninsula). I worked with a nursery colony in Punta de Benimaquia limestone cave locatedin the Montgó Natural Park, which is one of the three colonies for which fish-eating behaviourwas described (Aihartza et al., 2003).First of all, I evaluated the frequency and seasonal variation of fish remains in faecalpellets from M. capaccinii collected in 2008, 2009 and 2010. Fish consumption occurredevenly throughout the year, without any seasonal pattern, and the percentage of consumedfish across the study period increased six-fold. All otoliths found in the excrements wereidentified as belonging to the exotic surface-feeding fish Gambusia holbrooki. Measuringotoliths, I estimated that the mean size of consumed fish (mean=2.48 cm, range= 1.92¿3.55cm, n=86) was significantly smaller than the mean measured for available fish (mean=3.87cm, range= 2.17¿6.21 cm, n=86), suggesting that the long-fingered bat¿s relatively small bodymay constrain handling larger prey. Moreover, after capturing 15 bats and radio-tracking thefour with the highest number of fish remains in their droppings, I identified two huge artificialponds as fishing grounds. Ponds hold very high densities of G. holbrooki, suggesting that theamount of fish at the water surface may be the principal factor triggering fishing.After identifying the fishing areas, I conducted an experiment to identify the echolocationand flight differences between insect hunting and fishing, and I observed that M. capacciniiregulates the temporal components of its terminal phase and modifies the prey capturetechnique according to the target type. All recordings corresponding to insect hunting showeda terminal phase with the typical two parts; buzz I and buzz II, even though some (5.7%)fishing recordings lacked buzz II. When capturing insects, bats used both parts of the terminalphase of the echolocation call to the same extent, and performed short and superficial dipson the water surface. In contrast, when preying on fish, the buzz I was longer and the buzzII shorter, and the bats made longer and deeper dips. These changes suggest that batsobtain advantages from using buzz I pulses rather than buzz II when fishing, maybe due tothe echolocation scanning needs associated with less conspicuous prey such as fish. Batscontinued emitting echolocation calls beyond the theoretical signal-overlap zone in the threeprey capture attempts, suggesting that they might be able to obtain information even whenthey have surpassed that threshold, at least initially. The feeding buzz produced during insectcapture ended at about 6¿7 cm from the prey, presumably when the pulse-echo overlapeffectively limited information gathering. However, fishing bats ended their feeding buzz atfarther distances from the prey, most likely because they were limited by the earlier insertionof their feet into the water.A subsequent fish-stimulus recognition experiment showed that M. capaccinii respondedalmost exclusively to direct stimulus¿fish protruding above water¿, and used echolocation toidentify and localize them, suggesting that capture attack is triggered only by targets arising¿even minimally¿from the surface. However, that discovery did not provide informationabout how bats discern fish from insects, so I performed another experiment to address thatquestion. I hypothesized that if bats depend on the morphology of the prey to discern fish frominsects, fish displayed either stationary or in motion would trigger fishing-like (buzz I-biasedterminal phase, and long and deep dips) attacks, whereas if bats rely on the motion of theprey to discern fish from insects, bats would attack stationary fish using an insect-huntinglikeattack (even length of both components of the terminal phase, and short and superficialdips), while performing fishing-like attacks to temporary fish. Thus, I analysed the last stageof attacks to observe the changes in echolocation and attack-pattern between stationary andtemporary fish-targets. Stationary targets were attacked using insect-hunting-like attacks,while temporary targets triggered fishing-like attacks, which drove me to conclude that batsrely on the motion of the prey to discern fish from insects instead of their morphologicalfeatures. Additionally, to determine the reaction ability of bats when facing temporary targets,I analysed their response to target disappearance at different moments of the hunting act.Bats were able to adjust their attack pattern when the target disappeared during the approachphase and the first part of the terminal phase, but targets that disappeared during the secondpart of the terminal phase were tackled in the same manner as stationary targets. Finally, thebuzz-ratio and dip-length where correlated to the time of target disappearance, suggestingthat bats are able to make extremely sharp adjustments of their attack depending on whenthe target has disappeared.Aiming to address whether the ability to fish is innate in M. capaccinii or the result of alearning process, I repeated the last experiment in a non-fishing colony, obtaining significantdifferences in the responses to different types of targets. Long-fingered bats of both coloniesresponded to the stationary target similarly, performing short and superficial dips and usingsimilar amount of pulses in each part of the feeding buzzes. However, temporary targetswere attacked using a different hunting technique. The sudden disappearance of the targetled to longer and deeper dips as well as to an increased buzz I-buzz II ratio in both cases,but the intensity of that shift was greater in the fishing colony than in the non-fishing colony,suggesting that the basic response is instinctive, but fishing bats have refined the capturetechnique to enhance fishing efficiency.All these information drove me to conclude that fishing ability of M. capaccinii is not anovel achievement, but an imprinted ability that is triggered when some specific conditionsare met. Slight modifications in its hunting trawling technique, honing the echolocation andflight features were enough to capture fish, and the instinctive reaction to temporary targetsshowed by non-fishing bats suggests that they might be able to add fish to their diet aftersharpening their hunting technique. Hence, the appearance of favourable conditions, namelyhigh densities of surface-feeding fish, is probably enough for any M. capaccinii to start fishing,providing that those conditions remain until bats refine their hunting technique.