Wind Tunnel Experiments Provide Data to Improve Sea Turtle Biotelemetry

In recent experiments, PIFSC researchers in the Fishery Biology and Stock Assessment Division have collaborated with colleagues from the University of British Columbia (UBC), NOAA's Southwest Fisheries Science Center, and Moss Landing Marine Laboratory (MLML) to estimate the magnitude of drag associated with sea turtle biotelemetry devices. Together, the group is seeking ways to minimize drag through redesign of the biotelemetry devices and their attachments. Minimizing drag will help ensure the applicability of biotelemetry data to tag-free turtles in the wild and lessen the adverse impacts on tagged turtles.

The most recent experiments have focused on attachment devices used with leatherback turtles, including a backpack harness, a tag which attaches directly to the longitudinal ridge of the turtle, and a submersible camera. In addition to the leatherback studies, trials were conducted to determine the effects of double tagging hard-shelled turtles (e.g., loggerhead, olive ridley, and green turtles) for post-hooking mortality studies. For all turtle tagging studies, increases in drag force created by the tag cause a direct and proportional increase in power output and energy consumption of swimming turtles. Therefore, the increased drag caused by biotelemetry tags has many implications for the migratory energetics and welfare of the outfitted turtles.

Studies of drag in leatherback attachments

Measurements were conducted using a life-size fiberglass cast of an adult leatherback turtle placed in the Large Boundary Layer Wind Tunnel of the Department of Mechanical Engineering, UBC. The drag force and coefficient were determined for the simulated turtle without any attachments and with various biotelemetry devices added.

A fiberglass cast of an adult leatherback turtle with satellite harness attachment. Yarn was placed on the cast and biotelemetry devices to help measure flow patterns and turbulence. Drag was measured for various experimental attachments such as this submersible camera.
Drag measurements were made in the UBC wind tunnel. Left: A fiberglass cast of an adult leatherback turtle with satellite harness attachment. Middle: Yarn was placed on the cast and biotelemetry devices to help measure flow patterns and turbulence. Right: Drag was measured for various experimental attachments such as this submersible camera.

The harness attachment caused a 93% to 167% increase in drag, depending on wind velocity. The ridge-mounted tag (Wildlife Computers) increased drag by 17% to 65% at lower simulated swim speeds (wind velocity) but increased drag by only 0.1% to 8% at speeds typical of leatherback movements. Greater drag increases measured at the slower swim speeds are probably due to differences in how the attachment changes flow patterns (and turbulent flow) around the turtle as speed is increased.

Studies of drag in double tagging configurations

There is substantial interest in understanding the probability of post-hooking mortality of sea turtles that have been returned to sea after interacting with longline fishing gear. One method to estimate post-release mortality has been to release the turtle with a satellite tag and determine the turtle’s movements and likelihood of mortality through analysis of tag transmission data. However, such methods are flawed; a tag can cease to transmit data due to tag failure and other reasons, not simply because of the turtle’s death. One way to help resolve these issues is to increase the number of tags placed on each released turtle, enabling direct estimation of the rate of tag loss. NOAA has received requests for permits allowing such studies with double-tagged turtles. However, it is not known how a second tag will affect drag, whether it will simply have an additive effect or cause an interaction effect of the two tags (known as interference drag).

Drag forces on a green turtle body form with no tag, a single tag, and double tags in various 
            arrangements. The tags had equal areas and were rectangular in shape. From left to right: the turtle cast 
            alone; cast with one tag; two tags in parallel separated by 1 tag width; two tags in parallel separated by 
            2 tag widths; two tags in parallel with no separation; two tags in series with no separation; two tags in 
            series separated by 1 tag length; two tags in series separated by 2 tag lengths.
Drag forces on a green turtle body form with no tag, a single tag, and double tags in various arrangements. The tags had equal areas and were rectangular in shape. From left to right: the turtle cast alone; cast with one tag; two tags in parallel separated by 1 tag width; two tags in parallel separated by 2 tag widths; two tags in parallel with no separation; two tags in series with no separation; two tags in series separated by 1 tag length; two tags in series separated by 2 tag lengths.

To study this, scientists measured the drag associated with a green turtle cast with and without tags, in various configurations. Wooden mock tags were placed first singly and then as a pair (both parallel and in series) on the green turtle cast. The graph above shows that there was an interaction or interference drag. Two tags side by side (parallel), separated by 1 or 2 tag widths, caused more than twice the drag of a single tag alone. Placing the tags parallel but touching each other caused only a doubling of the drag. Placing the two tags one behind the other (in series) did not increase the drag over a single tag alone. However, as the separation between the tags grew to 1 tag length or more there was a slight interaction effect.

In general, the results showed significant energetic demands placed on sea turtles carrying tags. Several recommendations resulted: The frontal area of the tags should be reduced; the tags should have a teardrop shape and low profile; the antenna length and diameter should be minimized; and the tags should not be placed at the peak height of the carapace. Studies incorporating multiple tags should reduce the interference drag by bunching transmitters or separating them by several tag widths. With respect to the very large drag of the leatherback harness, it is recommended that the alternative of using ridgemount tags directly attached to the longitudinal ridge be pursued, with careful scrutiny of turtle health issues associated with drilling into the ridge. These questions will have to be answered through field experiments.