Myoclonus induced by cathode ray tube screens and low‐frequency lighting in the European starling ( Sturnus vulgaris )

light sensor; Crotech Instruments) confirmed that the lights, and all of the stimuli, ran at the frequencies stated. The bird’s responses to a uniform grey field (37 x 28 cm) displayed on a 120 Hz CRT monitor (Trinitron monitor model GDM-F520; Sony), and a grey isoluminant field (29 x 36 cm) displayed on a 110 kHz liquid crystal display (LCD) screen (model LM919; AOC International) were recorded and compared. In both cases the control phases involved leaving each monitor running but occluded by a blanket placed over the screen. The mean luminance (ML) of the stimuli was measured (Minolta Chroma Meter CS-100; Minolta) from a position above the bird’s perch through the viewing window. A device that could measure bird-perceptible UV light was not required, since the viewing window blocked UV light. The luminances of all stimuli were altered to match the CRT monitor, which had the lowest ML when running at maximum brightness (mean [sd] ML 70·62 [0·77] cd/m2 for the CRT monitor, and 71·10 [0·42] cd/m2 for the LCD). The bird only twitched once when the monitors were occluded, but frequently did so when the CRT monitor was uncovered (Table 1, Fig 1a). As magnetic fields and sound would have passed through the blanket, this indicated that the trigger was visual. The bird was more likely to experience myoclonus, and to have muscle jerks of worse severity, when viewing the 120 Hz stimulus than the LCD screen (Fig 1a). The two monitor types differed in their spatial properties as well as in their flicker rate (D’Eath 1998, Fleishman and Endler 2000). Certain spatial frequencies can be visually provocative, at least to human beings (Wilkins 1995) so, to determine specifically whether the flickering light was a trigger, the bird’s responses to either high-frequency (100 kHz) light (Durotest Truelite 18W, ballast; Tridonic) or lowfrequency (100 Hz) fluorescent light (Durotest Truelite 18W, ballast; Fitzgerald Lighting) falling on to a 76 x 52 cm white card were compared. During control phases, the fluorescent lights were turned on, but their light emissions were blocked by a blanket. The mean stimulus radiance was equated to that of the monitor and screen in the previous test, by placing strips of neutral density filter (Filter number 210, 0·6 ND; Lee Filters) over the lamps as required (high frequency: ML 72·28 [24·80] cd/m2, low frequency: ML 75·40 [28·43] cd/m2). Previous measurements of the same lights showed that the emitted wavelength spectrum of the highand low-frequency lights were similar (Greenwood and others 2004) and that the modulation of the type of lamps used was close to 100 per cent (Wilkins and Clark 1990). The bird was more likely to have a muscle jerk when the fluorescent light was not occluded, regardless of flicker rate. However, the probability of myoclonus occurring was much higher upon exposure to 100 Hz than upon exposure to 100 kHz (Table 1, Fig 1b). The Myoclonus induced by cathode ray tube screens and low-frequency lighting in the European starling (Sturnus vulgaris)