Monitoring mosquito eco-systems and vector-control strategies using a stand-off optical sensor.

利用远距离光学系统监测蚊子生态系统和病媒控制策略 传感器。 PI：B.Thomas-NIH R21 项目总结： 媒介生物控制策略仍然是保护人类种群免受 大量蚊子传播的疾病，如疟疾、登革热、寨卡病毒或西尼罗河病毒。 通常使用物理诱捕器来监测蚊子的数量，然而这种方法受到许多 劣势。它需要由合格的人员进行漫长而昂贵的实验室分析，这极大地 减少观察到的昆虫数量以及陷阱部署的时间。陷阱也提供了一个很差的估计。 实际人口规模或人口密度，因为诱捕器的吸引范围通常是未知的 并且可能会随着天气条件的变化而变化。这些限制是我们评估 各种病媒控制战略(化学品、生物、环境改良)的有效性 等)。劣质方法不一定被识别，而这些方法最终会导致传染病的传播 疾病。在这种情况下，我们认为监测昆虫种群动态的新方法是 改进控制程序性能的必要努力。 新泽西理工学院的一个团队与哈德逊蚊子团队合作 该计划寻求支持，以使用一种新的光学传感器进行一系列现场实验 实时识别其视野内的蚊子的科、种和性别。基于激光的 该仪器是一种双波长偏振敏感测距传感器。对于每一只过境的飞虫 通过红外激光，传感器可以提取机翼和机身的光学特性 昆虫以及它的翅膀拍打频率。来自实验室原型和数值模拟的初步数据 仿真表明，使用有监督的机器学习分类器，该仪器可以识别出 300米以外蚊子的种类、性别和孕次。该仪器将部署在一个较高的 新泽西州蚊子密度区全季持续监测蚊虫种群 2021年4月至10月。连续的测量将允许识别一些昆虫，这是目 比物理陷阱高出一个数量级。由于探测到的空气量已知，数据分析将提供 每一类昆虫的种群密度，将从这些密度中推导出种群动态。此外, 每一次昆虫凌日的时间和日期允许研究昼夜节律、高峰活动和行为作为 气象站测量的大气条件的作用。在2022年，一个类似的实验将是 在同一地点进行，而哈德逊蚊子计划将进行媒介控制运动 目标是库蚊和伊蚊，这两种蚊子都是各种传染病传播的罪魁祸首。这个 多次使用空气中拟除虫菊酯杀虫剂对靶标和非靶标昆虫的影响将是 通过研究每个物种的死亡率和种群动态进行评估。这两年的数据都将是 对比现场实物圈闭，采用现行黄金标准方法，进行进一步分析验证。

Monitoring mosquito ecosystems and vector-control strategies using a stand-off optical sensor. PI: B. Thomas – NIH R21 Project Summary: Vector control strategies remain one of the most effective ways to protect human populations from the large number of mosquito borne diseases such as malaria, dengue fever, zika virus, or West Nile virus. Mosquito populations are generally monitored using physical traps, however this method suffers from many disadvantages. It requires long and expensive laboratory analysis by qualified personnel which drastically reduces the number of observed insects as well as time of trap deployment. Traps also provide a poor estimate of the actual population size or population density because the attractive range of traps is generally unknown and may change with weather conditions. These limitations are strong drawbacks in our ability to evaluate the effectiveness of various types of vector-control strategies (chemicals, biological, environmental modifications etc.). Inferior methods are not necessarily identified which ultimately contributes to the spread of infectious diseases. In this context, we argue that new methodologies to monitor insect population dynamics is key in the necessary effort to improve control program performance. A team from the New Jersey Institute of Technology in collaboration with the Hudson Mosquito Program seeks support to carry out a series of field experiments using a new optical sensor capable of identifying in real-time the family, species, and gender of mosquitoes in its field of view. The laser-based instrument is a dual-wavelength polarization-sensitive stand-off sensor. For each flying insect transiting through the infrared laser beams, the sensor can retrieve the optical properties of the wings and body of the insect as well as its wing beat frequency. Preliminary data from a laboratory prototype and numerical simulations indicate that the instrument, using a supervised machine learning classifier, can identify the species, gender, and gravidity of mosquitoes up to 300 m away. The instrument will be deployed in a high mosquito density area in New Jersey to continuously monitor the mosquito population over the whole season from April to October 2021. Continuous measurements will allow to identify a number of insects that is orders a magnitude higher than physical traps. As the probed volume of air is known, data analysis will provide the population density for each class of insects from which the population dynamics will be derived. In addition, the time and date of each insect transit allow to study the circadian rhythm, peak activities, and behavior as a function of atmospheric conditions measured by a weather station. In 2022, a similar experiment will be conducted at the same location while the Hudson Mosquito Program will conduct a vector control campaign targeting Culex and Aedes mosquitoes, both responsible for the spread of various infectious diseases. The impact of multiple applications of airborne pyrethroid insecticide on targeted and non-targeted insects will be evaluated by studying the mortality rates and population dynamics for each species. Both years, the data will be compared to physical traps on site, the current gold standard method, for further analysis and validation.