A convolutional neural network to identify mosquito species (Diptera: Culicidae) of the genus Aedes by wing images.

Accurate species identification is crucial to assess the medical relevance of a mosquito specimen, but requires intensive experience of the observers and well-equipped laboratories. In this proof-of-concept study, we developed a convolutional neural network (CNN) to identify seven Aedes species by wing images, only. While previous studies used images of the whole mosquito body, the nearly two-dimensional wings may facilitate standardized image capture and reduce the complexity of the CNN implementation. Mosquitoes were sampled from different sites in Germany. Their wings were mounted and photographed with a professional stereomicroscope. The data set consisted of 1155 wing images from seven Aedes species as well as 554 wings from different non-Aedes mosquitoes. A CNN was trained to differentiate between Aedes and non-Aedes mosquitoes and to classify the seven Aedes species based on grayscale and RGB images. Image processing, data augmentation, training, validation and testing were conducted in python using deep-learning framework PyTorch. Our best-performing CNN configuration achieved a macro F1 score of 99% to discriminate Aedes from non-Aedes mosquito species. The mean macro F1 score to predict the Aedes species was 90% for grayscale images and 91% for RGB images. In conclusion, wing images are sufficient to identify mosquito species by CNNs.

Friend or foe? A parasitic wasp shifts the cost/benefit ratio in a nursery pollination system impacting plant fitness.

Nursery pollination systems are species interactions where pollinators also act as fruit/seed herbivores of the plant partner. While the plants depend on associated insects for pollination, the insects depend on the plants' reproductive structures for larval development. The outcome of these interactions is thus placed on a gradient between mutualism and antagonism. Less specialized interactions may fluctuate along this gradient with the ecological context, where natural enemies can play an important role. We studied whether a natural enemy may impact the level of seed consumption of a nursery pollinator and how this in turn may influence individual plant fitness. We used the plant Silene latifolia, its herbivore Hadena bicruris, and its ectoparasitoid Bracon variator as a model plant-herbivore-natural enemy system. We investigated seed output, germination, survival, and flower production as proxies for individual plant fitness. We show that B. variator decreases the level of seed consumption by H. bicruris larvae which in turn increased seed output in S. latifolia plants, suggesting that parasitism by B. variator may act as a regulator in the system. However, our results also show that plant survival and flower production decrease with higher seed densities, and therefore, an increase in seed output may be less beneficial for plant fitness than estimated from seed output alone. Our study should add another layer to the complex discussion of whether parasitoids contribute to plant fitness, as we show that taking simple proxies such as seed output is insufficient to determine the net effect of multitrophic interactions.