Water shrews detect movement, shape, and smell to find prey underwater

Supporting Information

Files in this Data Supplement:

Movie 1. Three movie clips captured from high-speed video showing a water shrew capturing fish under infrared lighting conditions. Each slow motion clip is followed by a real-time clip. In clip 3, the shrew encounters a live fish and a dead fish simultaneously, and pursues the live fish, supporting the suggestion that movement is an important cue.

Movie 2. Three movie clips captured from high-speed video showing water shrews attacking brief water movements. The arrow indicates the active outlet, from which water occasionally pulses. In clip two the shrew investigates the area of the outlet before it becomes active, but returns to attack the later water movement. Note the brief underwater sniff (expiration and inspiration of air) that occurs at the end of unsuccessful attacks.

Movie 3. Slow motion video of water shrews investigating silicone objects and choosing the silicone fish.

Movie 4. Slow motion video of water shrews reacting to moving objects. Movements were generated by inserting iron into the object and placing a spinning magnet below the aquarium.

Movie 5. Slow motion video of water shrews reacting to moving model fish.

Movie 6. Slow motion video of water shrews reacting to a clay and metal model cricket. Note the underwater sniffs, during which air is exhaled from the nostrils, and some or most of the air is subsequently re-inhaled.

Movie 7. Slow motion video of a water shrew encountering a fish. Note the underwater sniff occurs just prior to the fish's escape response.

SI Text

Fish capture latency trials

For trials examining the latency to fish capture under lighted and infrared conditions, three shrews were tested. Minnows (Pimephales promelas) were presented in a 13 ´ 17-cm rectangular Plexiglas enclosure set within a larger (50 ´ 25 ´ 16-cm) aquarium filled to a depth of 4.5 cm with water. The shrew's home cage was connected to the larger aquarium by a flexible tube with a removable door. Trials were organized in pairs (infrared lighting or lighted conditions) with first condition alternating (e.g. infrared trial followed by visual trial then visual trial followed by infrared trial, etc.). Fish were chosen by first identifying two size-matched fish and capturing them from a separate aquarium (to avoid effects of individual fish speed or size differences in speed). One of the two fish was then randomly chosen for the first trial. Trials were initiated by opening the door, allowing the shrew to enter the larger aquarium with access to the behavior chamber, and the resulting fish capture was recorded on a Redlake HS-3 black and white high-speed camera at 1,000 frames per second with each frame at 1,024 ´ 1,024 pixels. The camera's calibration was checked by filming the screen of an oscilloscope at 20 ms per division and was found to be precisely accurate. The latency to fish capture was determined immediately after each trial by identifying the frame in which the shrew's nose first contacted the water during its entry dive (e.g., see SI Movie 3, clip 1, for example) and ended with the frame in which the shrews teeth closed on the fish (e.g., SI Movie 1).

Model fish retrieval trials

A range of different fish were made by suspending dead Pimephales promelas in a container filled with AeroMarine Silicone RTV rubber (John Greer and Associates). After curing, the mold was split, washed with water and alcohol, and the cavity was lightly coated with petroleum jelly. The mold was reassembled and filled with AeroMarine Silicone using a syringe and cured. Cylinders and rectangles were also cast and cut to different sizes that spanned the size of the model fish. All castings were washed extensively with water and alcohol prior to trials. Model fish were occasionally damaged as the shrews bit and attempted to eat them, and these were replaced by new model fish. The shrew's home cage was connected to the larger aquarium by a flexible tube with a removable door. Prior to the trials, the shrews were fed dead fish in the behavior chamber. The distractors (silicone cylinders and rectangles) and the model, cast fish, were then placed in the chamber. Trials were initiated by opening the door, allowing the shrew to enter the larger aquarium with access to the behavior chamber. The resulting behavior was recorded on a Redlake HS-3 black and white high-speed camera at 250 frames per second. An item was scored as chosen if the shrew grasped it in its jaws (see SI Movie 3). On most occasions, shrews carried the chosen object to the home cage. Shrews did not grasp more than one object per trial. After every two consecutive trials, the shrews was offered a dead fish in the behavior chamber to prevent extinction of responses.

Echolocation tests

To test for the use of audible and/or ultrasonic emissions in either navigation or prey location, five water shrews were tested individually in a 190-liter glass aquarium (94 ´48 ´ 50 cm) with an elevated platform (a 20.3-cm-high half cinder block) positioned 5 cm from one end. Each shrew was tested in both terrestrial and aquatic settings (with the aquarium filled to a depth of 17.5 cm with 20°C tap water), in both the presence and absence of a prey item (a single live earthworm) or obstacle (a 15 cm length of 3.8-cm OD black PVC pipe) placed on the bottom of the aquarium »15 cm from the end opposite the platform on which shrews were introduced to the apparatus. Treatment type was systematically arranged to progress from terrestrial to the aquatic trials, and within each of those settings, progressed from an open field to the addition of the obstacle, and then finally, after removal of the obstacle, to the addition of the prey item over a series of 6 days between 10 and 25 July 2006. Within days, the five shrews were tested in random order, and the apparatus was cleaned thoroughly with a 50% vinegar and water solution between each trial. A single 25-Watt incandescent red light bulb suspended 0.25 m above the aquarium provided illumination during trials, and all trials were videotaped with a tripod mounted SONY DCR-TRV120 camcorder set to "NightShot." In terrestrial trials, airborne ultrasound was detected by an Ultra Sound Advice U30 bat detector (20-200 kHz), with 10 consecutive 19-sec blocks of signal cycled through an Ultra Sound Advice Portable Ultrasound Processor (PUSP; 224-kc/sec sampling rate) and archived (if detected) in expanded form (1:5) on MiniDisc using a SONY MZ-N707S MiniDisc recorder. The audible output of the bat detector recorded to on videotape via the camcorder ultimately allowed an examination of whether predatory behavior or investigation/negotiation of an obstacle was associated with ultrasound production, while all audible signals were captured directly by the camcorder microphone. In aquatic trials, underwater acoustic signals (100 Hz-75 kHz) were detected by a Sonotronics DH-3 hydrophone, with any ultrasound again being cycled in 19-sec blocks through the PUSP and archived to MiniDisc. Audible signals were conveyed via the hydrophone preamplifier to a Honeytone miniamp/speaker, which broadcast to the camcorder microphone, allowing synchronization of sound production with videotaped shrew behavior. For all trials, video recording commenced with the introduction of the subject shrew from a closed PVC tube to the platform and continued until 10- to 19-sec sampling periods had cycled through the PUSP subsequent to the subject's movement off the platform in the case of open field trials or discovery of the obstacle or prey item in object-based trials. Meter sticks adhered to the bottom horizontal and left vertical frame members of the aquarium provided spatial reference for the videotape records and between trials. All trials were videotaped at 250 fps under lighted conditions.

• Early Edition
• Archives
• Online Submission

• Feature Articles

  (Expanded research articles of exceptional breadth)

• Commentaries

  (Expert commentary on leading research)

• Letters

  (Online-only letters to the editor)

• Inaugural Articles

  (Articles by recently elected NAS members)

• PNAS Profiles

  (Biographical profiles of Academy members)

• Sustainability Science

  (Interactions of human and environmental systems)

• Collected Papers

  (Editorials, Perspectives, Reviews, Colloquia, etc.)

• Special Features

  (Solicited articles on exceptional research topics)

• Sackler Colloquia

  (Scientific reports held under NAS auspices)