Female moths call in vain: Streetlights diminish the promise of mating

Artificial light at night has increased strongly in recent decades and is now affecting moths, a key contributor to pollination networks. The global shift to light-emitting diode (LED) streetlights changes the nocturnal light environment even further, especially because of their high variability in spectrum and intensity. To date, the impact on mating success of moths is only little  known, making it essential to investigate their behavioural responses.

We recorded the flight behaviour of male moths (Sphinx ligustri) using a symmetrical flight tunnel. Two different light environments with a female positioned on one side of the tunnel were used to test the effect of different LEDs (1800K Amber, 2200K Warm white, 3900K Neutral white) and intensities (0.05 lux, 150 lux, 370 lux, 590 lux) on arrival location, flight duration, and direction changes of males.

We have created two different light habitats within the tunnel. A homogeneous light environment, with equal light on both sides of the tunnel, and a heterogeneous light environment, with light only on one side. In both habitats, the flight behavior of male moths was tested in the presence of females. To test whether mating behavior was impaired, a female was placed on only one side of the tunnel. In the heterogeneous habitat, the opposite side of the light was used to create a conflict.

Methods
We used a flight tunnel to observe and analyse the response of male moths (Sphinx ligustri L., Lepidoptera, Sphingidae) to different light stimuli. This flight tunnel consisted of plexiglass and had a total length of 3.50 m with an inner edge length of 30 x 30 cm. A sliding door enabled us to position the moth in the middle of the tunnel. On top of each end, opaque boxes were positioned that contained the different light sources as well as a hook to add a cage with a female. These boxes had an opening of 20 x 20 cm to connect them with the flight tunnel. To prevent males from entering these boxes, the opening was covered with Gauze (Nobamull, Danz GmbH u. Co KG). To exclude any impact of the room, all non-white or reflective surfaces were covered with white cloths, creating a uniform environment around the flight tunnel.

We tested three different light-emitting diodes (LEDs) provided by Osram (TRILUX GmbH & Co. KG): 1. Amber 1800 K (λ max = 596 nm) with almost no blue light components, 2. Warm white 2200 K (λ max = 615 nm) with a small proportion of blue in the light spectrum, specially developed to be particularly insect-friendly streetlight and still achieve a good colour effect, and 3. Neutral white 3900 K (λ max = 444 nm) with a high proportion of blue. The spectra of all LEDs were measured with a high-sensitivity spectrometer (Ocean Optics Maya2000 PRO; Figure 1b). Light intensity was measured in a distance of 50 cm directly below the light source using a luxmeter (RS Pro ILM Light Meter 1337).

For the experiments, we use a species from the Sphingidae family, which are primarily nocturnal or crepuscular and are among the most important pollinators among moths (Winfree et al., 2011). Sphinx ligustri are robust, good fliers that have been reliably attracted to females in other experiments (Storms et al., 2022). The pupae were stored during winter in the refrigerator at a temperature of 5°C without any light and removed between four and six weeks before the experiment. For hatching, the pupae were exposed to indirect daylight in a natural day-night rhythm. To prevent mating, the light pollution naïve males and females were kept separately in a uniquely labelled plastic bottle. During the day, the plastic bottles were not covered lightproof to maintain with daylight the natural rhythm. All experiments were carried out at night. Before experiments, all moths were fed under red light (Storm 400 Headlamp, Black Diamond Equipment Europe GmbH) with sugar solution (2M) to ensure they were not hungry. Afterwards, males were placed beneath separate metal tins standing on small wooden boards to enable easy transportation to the release platform within the flight tunnel. Furthermore, the tins ensured that the moths could rest for at least one hour before experiments in complete darkness and the experimenter was unaware of their body alignment in relation to the flight tunnel at the beginning of a trial. A virgin female older than two but younger than seven days was placed in a small cage covered with gauze. This cage was positioned in one of the two opaque boxes that could be easily switched from one side to the other. During all experiments, the position of the female was swapped after half of the males were tested. To ensure that all pheromones were removed before the next trials started, the entire tunnel was cleaned with ethyl alcohol and both, the tunnel as well as the room were ventilated. We performed a series of experiments to investigate the behaviour of male moths responding to various stimuli in two different light habitats.

A total of 38 male moths were used for all experiments. Each moth had only one flight attempt per night. If the trial was cancelled, the experiment was repeated the following night. As long as the individuals were able to fly the aim for each was to successfully complete all experiments once if possible (3 to 12 were achieved). Missing flights were replaced by individuals that hatched later. The order of the experiments per animal was randomized to avoid a possible effect of the age. To exclude a room bias, the sides were swapped for each experiment after half of the flights. The male moth beneath the metal tin was placed in the middle of the tunnel on a little starting platform (9 x 9 cm, height 5 cm). Once placed, the metal tin was removed to release the male in a randomised orientation. At this moment, observation of behaviour started and was recorded using an ethogram in BORIS (Friard & Gamba, 2016). Parameters recorded were flight duration (in seconds from flight start to end of flight), change in flight direction (we considered a change of >180° to be a change in direction, indicating that an individual flew the opposite direction compared to the previous direction), intermediate stops and destination. One trial was completed when the male moth reached a pre-defined line, namely the transition from the tunnel to the last segment of the tunnel with the opaque box on top. If a male moth did not arrive on one side of the tunnel within 20 minutes or stopped its flight somewhere else in the tunnel, the trial was cancelled. All experiments were performed at approximately 25°C.

近几十年来，夜间人工光照强度显著上升，如今已对作为传粉网络关键类群的蛾类造成影响。全球范围内向发光二极管（light-emitting diode, LED）路灯的转型进一步改变了夜间光环境，尤其是这类路灯的光谱与光照强度存在极高变异性。截至目前，学界对夜间人工光对蛾类交配成功率的影响仍知之甚少，因此探究蛾类的行为响应至关重要。

本研究采用对称式飞行隧道（symmetrical flight tunnel）记录了霜天蛾（Sphinx ligustri）雄蛾的飞行行为。我们设置了两种光照环境，隧道的一侧放置雌蛾，以此测试不同LED光源（1800K琥珀色、2200K暖白色、3900K中性白色）及其光照强度（0.05勒克斯、150勒克斯、370勒克斯、590勒克斯）对雄蛾的抵达位置、飞行时长与转向行为的影响。

我们在飞行隧道内构建了两种不同的光照生境：一种为均匀光照环境，隧道两侧光照强度一致；另一种为非均匀光照环境，仅隧道单侧存在光照。两种生境下均设置雌蛾以测试雄蛾的飞行行为。为验证交配行为是否受光照干扰，我们仅在隧道单侧放置雌蛾；在非均匀光照生境中，光照缺失的一侧会形成行为冲突。

## 研究方法

本研究采用飞行隧道观察并分析鳞翅目天蛾科霜天蛾（Sphinx ligustri L.）雄蛾对不同光刺激的响应。该飞行隧道由有机玻璃（plexiglass）制成，总长度为3.50米，内部尺寸为30×30厘米。隧道内设有滑动门，可将试验蛾放置于隧道中部。隧道两端顶部均安装不透光盒子，内置不同光源，同时配备挂钩用于悬挂携带雌蛾的饲养笼。盒子与隧道相连的开口尺寸为20×20厘米，为防止雄蛾进入盒子，开口处覆盖有纱布（Gauze，Nobamull，Danz GmbH u. Co KG）。为排除实验室环境的干扰，所有非白色或具有反光性的表面均用白色布料覆盖，以在飞行隧道周围营造均匀的实验环境。

本研究测试了欧司朗（Osram，TRILUX两合公司）提供的三款不同LED光源：1. 1800K琥珀色LED（峰值波长λ_max=596nm），几乎不含蓝光成分；2. 2200K暖白色LED（峰值波长λ_max=615nm），光谱中仅含少量蓝光，为专门开发的昆虫友好型路灯，同时具备良好的显色效果；3. 3900K中性白色LED（峰值波长λ_max=444nm），蓝光占比极高。所有LED的光谱均通过高灵敏度光谱仪（high-sensitivity spectrometer，Ocean Optics Maya2000 PRO；图1b）进行测量。光照强度采用照度计（luxmeter，RS Pro ILM Light Meter 1337）在光源正下方50厘米处进行测定。

本实验选用天蛾科（Sphingidae）物种，该类群主要为夜行性或晨昏性，是蛾类中最重要的传粉类群之一（Winfree等，2011）。霜天蛾体型健壮、飞行能力优异，过往研究已证实其可被雌蛾有效吸引（Storms等，2022）。实验用蛹在冬季于5℃无光环境的冷藏柜中储存，于实验前4至6周取出。孵化阶段，蛹置于散射日光下，维持自然昼夜节律。为避免交配发生，未接触过光污染的雄蛾与雌蛾需单独饲养于带有唯一标识的塑料瓶中。日间塑料瓶不做遮光处理，以自然光维持其昼夜节律。所有实验均于夜间开展。实验开始前，所有蛾类在红光（Storm 400头灯，Black Diamond Equipment Europe GmbH）照射下喂食2M糖水溶液，确保其处于饱腹状态。随后，雄蛾被放置于独立金属罐内，金属罐置于小木板上，便于转运至飞行隧道内的释放平台。金属罐可确保蛾类在实验前于完全黑暗环境中至少休息1小时，同时避免实验者知晓试次开始时蛾类相对于隧道的身体朝向。选取日龄2至7天的处女雌蛾，放置于覆盖纱布的小型饲养笼中，该笼被安置于两个不透光盒子中的一个，盒子可便捷地在隧道两侧切换。所有实验中，每测试半数雄蛾后，便调换雌蛾的放置位置。为确保下一试次开始前所有信息素均被清除，整个飞行隧道需用乙醇清洗，同时隧道与实验室均需通风换气。本研究开展了一系列实验，以探究雄蛾在两种不同光照生境中对各类光刺激的行为响应。

本研究共使用38只雄蛾开展所有实验。每只蛾类每晚仅可进行一次飞行试次。若试次被取消，则于次日夜间重复实验。只要试验蛾能够飞行，每只个体均尽可能完成全部实验（实际完成3至12次实验不等）。未能完成飞行的个体由后续孵化的蛾类替代。每只蛾的实验顺序均经过随机化处理，以排除日龄可能带来的影响。为排除实验室环境偏差，每完成半数飞行试次后，便调换隧道两侧的实验位置。将置于金属罐内的雄蛾放置于隧道中部的小型起始平台（9×9 cm，高5 cm）上，随后移除金属罐，以随机朝向释放雄蛾。此时即刻开始行为观察，并采用基于行为谱的分析软件BORIS（Friard & Gamba, 2016）记录行为数据。记录的参数包括：飞行时长（从飞行开始至飞行结束的秒数）、飞行转向（将转向角度>180°记为一次转向，即个体飞行方向与前一阶段完全相反）、中途停留次数与最终抵达位置。当雄蛾抵达预设界线，即隧道与末端带有不透光盒子的隧道段之间的过渡区域时，试次宣告完成。若雄蛾在20分钟内未抵达隧道任意一侧，或在隧道内其他区域停止飞行，则该试次被取消。所有实验均在约25℃的环境下开展。