Peeking is the word for what an approaching object does to an eye. The image of the thing grows, and grows at an accelerated rate, and that rate carries information: how fast the object is coming and approximately when it will arrive. Animals across the tree of life respond to it, from insects to birds to us. In fish, the response is a whole-body movement called a C-start, a curve in the shape of the letter followed by a burst of speed. What has not been clear is where the line is. A reef fish constantly sees predators, most of them do not hunt, and each false alarm costs it time to feed. Working on shallow reefs off the west coast of Curacao, three researchers put a number on it. No fish in their recordings made an escape response until the stimulus expanded like an object approaching at more than two meters per second. Most leaks occurred at three meters per second or more. The comparison that makes the figure interesting came from the reef itself. In a separate filming effort, the team tracked 61 bar cat attacks, filmed the hunt for typically brown chromis, one of the study’s two dominant species, and found that their rapid approach speeds ranged from 2.7 to 9.0 meters per second. In fact, no escape response was recorded when faced with something as slow as a moving predator. The platform in the coral zone The experiment was carried out in Cas Abou and on Kokomo beach, in five locations at about three meters depth, in twelve test sessions at the two locations. In each of them, the team installed a 12.9-inch iPad Pro next to a secluded area of coral that local fish use as shelter. Every ninety seconds, for about an hour, a black disc played on the screen, expanding on a white background. Three GoPros recorded at 240 frames per second, two facing downwards and one sideways. That allowed the team to reconstruct each fish in three dimensions and measure how far it was from the coral and from what angle it was looking at the screen. Fish that were invisible to one of the stereo cameras or left the field of view were removed, and an entire video was discarded when more than a third of the tracked fish became unusable. After filtering, 247 videos survived, 67 of them containing at least one escaping fish and 180 containing none. The disk expansion rate was calibrated to replicate what a predator approaching at a nine-speed would do to a fish’s retina: 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, or 8.0 meters per second, presented in random order. No animals were handled and the work was carried out with a permit from the Curaçao government held by the CARMABI research station. A flat screen displaying a two-dimensional expanding disk is far from a predator. It eliminates form, body movement, water displacement, chemical signals, and being an object rather than an image. The goal of the design is to isolate the approach speed, and it is also the limit of what the result can say about a real attack, something the paper acknowledges in its discussion. The probability of escape jumps above two meters per second. In all analyses, stimulus speed dominated. Response probability did not increase consistently with speed. It remained flat and near zero during slow presentations and then rose sharply above about two meters per second, and formal comparisons confirmed that curved fits describe the data better than a straight line. That shape is what a threshold looks like in the data. It’s also consistent with the circuitry involved: the Mauthner cell, the giant neuron that fires C-start in fish, is on a pathway tuned to approach speed, so slow growth in the visual field can pass through the system without triggering anything. The authors interpret the near-absence of response to slow stimuli as consistent with threat-sensitive predator avoidance, in which prey rate their escape at the threat level and respond weakly to slow-moving or non-attacking predators. Damselfishes of the genus Stegastes, which is the genus of one of the two dominant species in this study, are already known to behave in this way. Curaçao’s work adds a measured speed at which the response is activated. They also imported two variables into the configuration, and both have to do with visibility. Fish closer to the screen and viewing it from a smaller angle responded more frequently, which the authors attribute to the flat screen becoming reflective and the disc becoming more difficult to distinguish from oblique positions. The attack speeds came from a different deployment: stereo cameras at 120 frames per second on the same reef, at other times recording bar cats instead of damselfish. The team tracked head positions across 61 attacks, each tracked from the moment the cat entered the frame or began running until the prey, typically a brown chromis, escaped or was caught. Instantaneous speed is noisy when measured manually from video, so the team used the 90th percentile of the predator’s speed as a proxy for its maximum attack speed, which captures the rapid approach phase while dampening tracking noise. These values ranged between 2.7 and 9.0 meters per second, and the authors consider that the correspondence between these speeds and the response threshold is close. Two different types of speed are placed side by side, and the piece on which the comparison is based is thinner than it appears. One is a programmed number, the approach speed that a disk on a screen was calibrated to imitate. The other is the tracked speed of an actual animal in the water, summarized as a percentile rather than its maximum. They are close, and the closeness is the finding, but the overlap is not a calibration and should not be read as such. No comparable figure has been published for the speed at which a cat swims when not hunting, and the authors state this. The closest number available is that of a relative, the green cat, which cruises at about a meter per second or less at similar body sizes. That speed is below the threshold, in the band where the fish in this study never responded at all. More extensive work on reef fish reports that attacking movements are several times faster than those of routine swimming, so the gap these fish are exploiting is wide. The gap between species survives the model. The two dominant species behaved differently. Bicolor damselfishes, which are territorial and contain small patches of coral, responded less frequently than brown damselfishes, which form site-attached aggregations and feed on plankton in open water above the reef. That difference tracks where they were when the disk appeared. Chromis were found significantly further from the coral than damselfish, and distance from shelter alone predicted a greater probability of fleeing. If both variables are put into a model, the distance is no longer statistically significant, which is what happens when two predictors contain overlapping information. The difference in species does not dissolve in the same way. Adding log-transformed distance to a species model reduced the species coefficient by more than a third and improved the fit. The effect of species remained statistically significant, which the authors interpreted as an intrinsic difference between the two fish and not a matter of exposure alone. The social context did nothing, which the authors point out as contrary to their expectations. Whether a fish had company nearby and how close that company was had no measurable influence on its escape. Sixty-seven of 247 videos contain some leakage, and the leaks are clustered at the top of the speed range, so the threshold is estimated from the thin end of the data. No natural predator attack occurred during any of the presentations that were analyzed, meaning that the alignment between the two data sets is an alignment between separate recordings and not a case of the same fish responding to both. A threshold measured on a reef, in two species, at three meters depth, with daylight clear enough for a screen to work, is a number for these fish here. The rule that fish follow What the study describes is not so much surveillance as the environment. Fish do not weigh each approaching form on its merits. They’re applying a rule that ignores everything below a speed and reacts to everything above it, and the rule falls just below the speeds at which it was filmed attacking the local predator. That’s a cheap way to be right most of the time. It’s also blind in one direction, at least in principle, and the study didn’t test that direction: such a rule would miss something that closed slowly and never accelerated. The document itself narrows that gap. Note that pursuit, ambush, and stalking all end in a rapid terminal approach, which is why approach speed functions as a signal. The study can’t say how the environment got there, whether it was learned on a particular reef or built and inherited. Either way, it’s a number that a fish carries with it, and it was measured in an animal looking at a screen that no predator has ever resembled. About this articleThis article is for general information and reflection. It is not professional advice. For your specific situation, consult a qualified professional.
