Our results reveal that the new ancient ?-designed strength contour are a fair qualitative malfunction of trip speed–stamina relationships within the, at the very least, specific bug taxaparing this on numerous J-designed energy contours which have before been discovered various other pests, it’s interesting one to a variety that’s very determined by hanging for its foraging (the brand new specialistboscis does not expand if the wings is located at other individuals) is the very first to show indications away from lowest speed requiring large energy than simply intermediate increase.
It ought to be recalled that both patterns made use of here, like the C
sexta displayed a preference for flight speeds of approximately 3 ms ?1 when flying freely in a room . Our predictions place this speed between ump and umr. The moths in our study were not able to fly at higher speeds than 3.8 ms ?1 without crashing, and similar behaviour has been shown in previous studies (e.g. ). However, we do not exclude the possibility that individuals in the wild could fly faster, as this is typically the case for wind tunnel studies with free-flying animals. Even though the characteristic flight speeds that were derived from the data correspond rather well to these two observations, the speeds varied somewhat between the two individuals. This could represent true differences in the flight and morphology of the moths, or be an artefact due to the limited number of data points for, especially, M2 at 3.8 ms ?1 . The flight speed predictions based on the blade-element model (equations (2.5)–(2.7)) were similar to those calculated by the experimental data. However, the range of possible speeds that could be predicted by varying these two parameters showed that the model is quite sensitive to parameter values. Pennycuick’s model (equations (2.5), (2.6), (2.8)), which was only used with the default parameter value of k = 1.2 (CD,pro is not explicitly in the model), predicted significantly higher flight speeds than both the experimental data and the blade-element model. The difference in curve shape and predicted flight speeds between the two models is due to the different scaling of profile power with air speed (blade-element: P?u 2.5?3 , Pennycuick: P?1), which in turn is caused by different assumptions on how CD,pro is affected by flight speed. In our blade-element model, we use the range between a constant CD,expert and one that scales with u ?1/2 (table 4), while Pennycuick’s model assumes that CD,pro scales with 1/u 3 and thus fully compensates for the u 3 factor in the Ppro equation. As Pennycuick’s approximation of a constant profile power originates from studies on birds, for which this component appears to be relatively constant over cruising speeds , it is unsurprising that 
the blade-element model predicts flight power more accurately for hawkmoths. In addition, Pennycuick’s model is only valid between ump and umr, a range which starts at the maximum air speed at which our moths were able to fly stably.
Hence, it design might not be ideal for modeling insect flight
D,pro estimations, are based on quasi-steady-state aerodynamics. How would the presence of a LEV affect our conclusion that M. sexta has a ?-shaped power curve? It is likely that an LEV would cause CD,expert to increase, as the vortex creates additional drag on the wing . This would make the ? shape more pronounced. As LEV is a lift-enhancing effect most useful at low speeds [13,42], one would imagine that CD,specialist is higher at low speeds, while close to steady-state values when the flight is faster, resulting in a flatter curve. However, the LEV in M. sexta has previously been found to be similar in size at all speeds , and even increasing in size with flight speed . Therefore, using unsteady values of CD,pro would probably not change the prediction of a ?-shaped power curve with a rather small span of predicted characteristic flight speeds.