Wilding pines are an invasive species in New Zealand which were reported to be spreading at a rate of 5% per year in 2011 and were estimated to cover 20% of New Zealand by 2030 if no additional control effort was made (Ministry of Primary Industries, 2022a). Vive la Résistance (led by SCION) is the most recent research programme which received MBIE funding in 2021 to create resistance strategies against re-invasion of wilding pines on treated land.

The ability to predict the spread of wilding pines is of particular interest. Wilding pines have winged seeds which autorotate during their descent (thus are termed samaras). This evolutionary adaptation allows wilding pines to spread across very large distances via wind dispersal. Previous studies have identified the terminal velocity of these samaras and investigated their mechanism of achieving high lift. Vive la Résistance has developed models of wilding pine dispersal that use the measured terminal velocity as the only aerodynamic input parameter for samaras. This assumes that samaras are always in a steady state of autorotation during their flight which may not always be true. In turbulent conditions samaras can become perturbed (complete cessation of autorotation) which rapidly increases their fall velocity until they regain autorotation.

A low-speed (<4.2 ms⁻¹) vertical wind tunnel was constructed to observe the flight of samaras for extended durations. A flow profile with a broad minimum in velocity was obtained using a flow conditioning device. This allowed samaras to achieve stable flight at constant slip speed localized in the centre of the tunnel. Samaras were perturbed using a jet of pressurized air. The flight behaviour was captured with high-speed cameras at 250 FPS from the side and top of the wind tunnel. Computer vision was used in Python with the OpenCV library to measure the FSS (fall slip speed), RPM and coning angle of the samara throughout each recording.

The perturbation response of Contorta Pine, Douglas Fir and Radiata Pine samaras were measured. When the perturbation was strong enough to stop autorotation, the FSS was seen to rapidly increase immediately after perturbation before relaxing back to steady state. A mass-spring-damper model was fitted to the relaxation period. The mean relaxation time was 0.60 s for Contorta Pine, 1.14 s for Douglas Fir and 1.35 s for Radiata Pine. The threshold for perturbation from a jet at an angle of 0° to horizontal was measured to be 1.19 ms⁻¹ for Contorta Pine, 1.55 ms⁻¹ for Douglas Fir and 3.56 ms⁻¹ for Radiata Pine. Further investigation was performed on the effect of changing the jet angle which showed that the velocity required to stop autorotation for a jet angle of -90° and 90° was only 65% of the velocity of the 0° jet angle.

The knowledge of how perturbation occurs and its effect on the FSS of wilding pine samaras can be used to increase the accuracy of current dispersal models. It is proposed that time-averaging FSS depending on the frequency of perturbation is a practical method of integrating with CFD and other modelling approaches.