Laboratory Studies of Flatfish Reactivity and Herding Behavior: Potential
Implications for Trawl Capture Efficiency
Figure 1. DIDSON sled ready for deployment. Pictured are the aluminum benthic sled with the forward mounted DIDSON (black box), sweeps and doors. The sled is equipped with a light meter as well as a 'pinger' to aid in recovery should the sled become hung on the bottom.
Trawls are inherently behavioral devices, harnessing the innate avoidance behaviors of fish to facilitate their capture with doors, bridles, and sweeps, herding flatfish into the path of the approaching net. While auditory and tactile senses may warn fish of an approaching trawl, thereby heightening vigilance, it is currently believed that visual signals provide the directional information to stimulate and guide herding behavior. Flatfish have been observed responding to trawl ground-gear, yet few studies have purposefully attempted to elucidate factors controlling their behavior. Several studies have examined secondary behavioral responses, such as swimming endurance and gait and net entry. However, the initial behavioral response, when flatfish first react to disturbance by the trawl ground-gear, has received little attention, even though it likely determines the course of subsequent behavior.
Laboratory experimentation was undertaken, using a simulated footrope in a flume tank, to examine the influence of two extrinsic factors, ambient illumination and temperature, upon the initial behavioral response of subadult Pacific halibut Hippoglossus stenolepis, and northern rock sole Lepidopsetta polyxystra (hereafter rock sole). In using subadults, the assumption is made that their behavior is comparable to those of larger individuals that constitute fishable stocks. Perhaps the most unexpected finding of this study was that 40% of fish did not react to the approaching footrope, that is, they were not flushed from the sand by the footrope. In so far as the apparatus used here simulates actual ground gear, the implication is that trawl ground-gear may frequently fail to stimulate any form of flushing response in flatfish, which means many fish may passively pass beneath the ground-gear and may not be adequately accounted for by current means of estimating trawl efficiency/catchability characteristics.
Light level had a pervasive influence upon the initial behavioral reaction of flatfish to approaching trawl ground-gear. In the light, fish were apt to respond with a 'run' response, where fish swam away from the gear while staying close to the bottom, (i.e., herding). In contrast, in the dark, 'rising' and 'hopping' off the bottom was more common, initiated by a startle response, which, elicited in the absence of visual input, resulted in the observed upward trajectory off the bottom. It has been widely observed that during the day, as fish herd in front of a trawl's footrope, many flatfish opportunistically escape through gaps between the ground-gear and the bottom, or are rolled over by the footrope.
If herding ceases in the night/darkness, with flatfish rising off the bottom in response to ground-gear disturbance, more of these fish would be captured, potentially explaining the widely observed increase in flatfish catches from survey trawls in shallow (<100 m) North Atlantic waters at night. However, where this ambient illumination effect upon ground-gear function may have its greatest impact, is not on the trawl foot rope, but on trawl sweeps. Unlike survey trawls, specialized flatfish trawls often utilize sweeps up to 200 m or more in length between the doors and the wings of the net, to herd fish into the path of the net.
Depending upon configuration, the herding behavior initiated and maintained by sweeps may be responsible for a large proportion of the catch. When flatfish cannot see the approaching sweep, their initial behavioral response may be to rise or hop off the sediment, thereby leaving the sweep's zone of influence, allowing the sweep to pass without initiation of herding. Therefore, while the footrope and net may become more efficient in the darkness, for the reason discussed above, the sweeps may actually become less effective at herding fish into the path of the net, decreasing the overall efficiency of the gear. If sweep efficiency decreases dramatically under low ambient, say at night, it might be plausible that catch would actually be greater during the day, than at night, the opposite of the pattern frequently seen in survey trawls.
Figure 2. Diagram showing the "DIDSON eye view" of sweeps and seafloor. Actual DIDSON imagery is video, allowing detailed observation and behavioral scoring of flatfish responses to initial disturbance by the sweep.
Although small compared to the effect of light, temperature was also demonstrated to influence flatfish behavior in this study. Temperature is known to influence the sustained swimming speed and endurance of fish. It has been demonstrated that American plaice swimming endurance, as measured through failure time analysis, decreased with decreasing temperature. As a consequence, low temperatures would also be expected to diminish the ability of flatfish to engage in herding, both down a sweep, as well as in front of a footrope. The 'hop' response increased in frequency among rock sole at low temperature, particularly in the dark. While there is danger in overinterpreting this result, the consequence of low temperature for flatfish might be a greater likelihood of 'startling' in response to trawl ground-gear disturbance, as opposed to initiation of an ordered behavioral response (herding).
Another study is currently under way, with the goal of testing these hypotheses in the field. This study makes use of a new tool being used at the Alaska Fisheries Science Center: the DIDSON or Dual-frequency IDentification SONar. The DIDSON is a high frequency acoustic camera that provides multiple high-resolution images per second across a 29° fan-shaped sector with a range out to 10 m or more. In this project the DIDSON is mounted on a benthic sled (Fig. 1), which is in turn towed by otter doors and abbreviated (5 m) sections of trawl sweep (3-inch cookies). The DIDSON is aimed along the length of one of the sweeps, providing imagery of the seafloor and fish encountered by the sweep as it is dragged along the bottom (Fig. 2). In a parallel study, the sled is also being positioned on the sweeps of commercial trawls utilized in summer flatfish fisheries out of Kodiak, Alaska.
Importantly, because the imagery is acoustic and therefore independent of ambient illumination, this provides a means of viewing flatfish behavior under differing light conditions, without the potential artifacts associated with artificial lighting. These studies should 'shed light' upon the role of ambient illumination in the herding behavior of flatfishes, as well as potential diel- and depth-related changes in the capture efficiency of both survey and commercial trawl gear.