Insect wing design: evolution and biomechanics

昆虫是地球上最多样化的动物，飞行可能是它们成功的关键。然而，尽管经过了数亿年的进化，昆虫的翅膀并没有汇聚成一个单一的最佳形状。相反，今天世界上可以看到的翅膀形态非常多样（甚至更多的化石），但从根本上说，它们都执行同样的任务——使飞行成为可能。这让我问：“为什么没有一种翅膀形状最适合扑翼飞行？”答案很可能存在于各种各样的局部最优解决方案中，这些解决方案专门适用于每种昆虫在其一生中所承担的任务。每一种昆虫的飞行任务都是独一无二的，因此对翅膀形态和运动的选择压力也是物种特有的。蜻蜓用翅膀捕捉猎物，并与对手进行空战，必须速度快，机动性强。与此形成对比的是，在欧洲各地迁徙的死亡头天蛾，会袭击蜜蜂的巢穴。它们必须非常高效，因为在迁徙过程中能量是非常宝贵的，但也要足够强壮，以抵御蜜蜂在蜜库中的攻击。了解超过一百万种已被描述的飞行昆虫的形态是不可能的，但它们的趋势贯穿其中，这对空气动力学工程来说是令人兴奋的，因为它们为已经尝试、测试和证明成功的特定要求提供了解决方案。我的研究旨在了解昆虫翅膀形状如何以及为什么会有这样的变化，尽管空气动力学性能的强烈选择压力，以及为什么在生态位之间过渡时形态会发生变化。检验这一点的最好方法是看看趋同进化的例子，这些物种具有相似的生态和形态，但起源于不同的分类分支。选择彼此完全不相关的物种可以区分翅膀形状的各个方面，这是设计优化的一部分，而不是那些简单地由于其历史起点。因此，我的实验采用了比较的方法来评估整个班级的代表性物种。在我的研究计划的第一步，一个博士后将直接测量飞行昆虫的空气动力学输出，因为它对了解空气在翅膀周围和尾流中的运动速度和方向是至关重要的。通过在空气中播种轻雾，并用脉冲激光照射粒子，将昆虫拴在风洞中，计算昆虫周围的气流速度。这种技术被称为数字粒子图像测速，是工程师研究复杂流动的首选技术。最近，我成功地将这项技术应用到飞虫身上，尽管它们体型小，振翅频率高。昆虫的翅膀上没有肌肉组织。所有复杂的扑动变形都是由机翼铰链处的肌肉主动控制，或者被动地由机翼结构上的惯性和空气动力控制。气动输出是机翼运动的结果，因此了解机翼形状在扑动过程中的变化是至关重要的。在我的研究的轨道2中，一个博士生将记录同一代表性昆虫个体的运动学。这名学生将通过人为地选择不同形态的果蝇（如更细的翅膀），并描述新形态的飞行性能，来测试关于翅膀形状在生态学中作用的预测。同时，学生将验证他们的结果，根据飞行性能选择应变，并测量机翼形态的变化。这两个轨道的输出将是：1)从生物力学适应的角度解释昆虫翅膀形状的多样性；2)计算流体动力学研究的详细运动学数据；3)为工程师建造昆虫大小的飞行器提供明确的设计准则。

Insects are the most diverse order of animals on earth and flight may be the key to this success. However, despite hundreds of millions of years of evolution, insect wings have not converged on a single optimal shape. Instead, there is an extraordinary range of wing morphologies visible in the world today (and even more fossilized), yet fundamentally, they all perform the same task - to enable flight. This led me to ask 'why is there no single wing shape that is best-suited to flapping flight?'The answer may well lie in assorted locally optimal solutions, specifically adapted to the tasks each insect undertakes during its life. The mission-profile of flight is unique for each insect species and so the selection pressures on wing morphology and kinematics is also species specific. A dragonfly that catches its prey on the wing and engages in aerial combat against rivals must be fast and manoeuvrable. Contrast this with the death's-head hawkmoth, migrating across Europe raiding bees' nests. They must be highly efficient since energy is at a premium during migration, but also robust enough to withstand attacks from bees when in their honey-stores. Understanding the morphologies of over a million described flying insect species is unfeasible, yet trends run through them which are exciting for aerodynamic engineering because they show solutions to specific requirements that have been tried, tested, and proven to succeed.My research seeks to understand how and why insect wing shapes have such variation despite intense selective pressure for aerodynamic performance, and why morphologies change when transitioning between ecological niches. The best way to examine this is to look at examples of convergent evolution, species which have similar ecology and morphology, yet originate from disparate taxonomic branches. Selecting species which are quite unrelated from one another allows discrimination of the aspects of wing shape which are part of design optimisation as opposed to those which are simply due to their historical starting point. My experiment therefore utilizes a comparative approach to evaluate representative species from across the class.In Track 1 of my research programme, a Postdoc will measure the aerodynamic output of flying insects directly, because it is essential to know how fast and in which direction the air is moving around the wings and in the wake. Flow velocities will be calculated around insects tethered in a wind tunnel by seeding the air with a light fog, and illuminating the particles with pulsing laser light. This technique is called Digital Particle Image Velocimetry and is the technique of choice for engineers studying complex flows. Recently, I successfully applied the technique to flying insects despite their small size and high wingbeat frequencies.Insects have no musculature in their wings. All the deforming complexities of the flapping cycle are controlled either actively by muscles at the wing hinge, or passively by inertial and aerodynamic forces on the wing architecture. The aerodynamic output is a result of wing motion so it is vital to know how the wing shape changes during flapping. In Track 2 of my research, a PhD student will record the kinematics of individuals from the same representative insects. The student will test predictions about the role of wing shape in ecology, by artificially selecting strains of fruit fly for alternate morphologies (e.g. more slender wings) and characterising the new morphs' flight performance. Simultaneously, the student will validate their results, by selecting strains based upon flight performance, and measuring the resulting modification in wing morphology.The output from these two tracks will be: 1) an explanation for the diversity of insect wing shapes from the perspective of biomechanical adaptation; 2) detailed kinematic data for Computational Fluid Dynamics studies; 3) clear design guidelines for engineers constructing insect-sized vehicles.