Detection of small moving objects is an important research area with applications including monitoring of flying insects, studying their foraging behavior, using insect pollinators to monitor flowering and pollination of crops, surveillance of honeybee colonies, and tracking movement of honeybees. However, due to the lack of distinctive shape and textural details on small objects, direct application of modern object detection methods based on convolutional neural networks (CNNs) shows considerably lower performance. In this paper we propose a method for the detection of small moving objects in videos recorded using unmanned aerial vehicles equipped with standard video cameras. The main steps of the proposed method are video stabilization, background estimation and subtraction, frame segmentation using a CNN, and thresholding the segmented frame. However, for training a CNN it is required that a large labeled dataset is available. Manual labelling of small moving objects in videos is very difficult and time consuming, and such labeled datasets do not exist at the moment. To circumvent this problem, we propose training a CNN using synthetic videos generated by adding small blob-like objects to video sequences with real-world backgrounds. The experimental results on detection of flying honeybees show that by using a combination of classical computer vision techniques and CNNs, as well as synthetic training sets, the proposed approach overcomes the problems associated with direct application of CNNs to the given problem and achieves an average F1-score of 0.86 in tests on real-world videos.

A diverse supply of pollen is an important factor for honey bee health, but information about the pollen diversity available to colonies at the landscape scale is largely missing. In this COLOSS study, beekeeper citizen scientists sampled and analyzed the diversity of pollen collected by honey bee colonies. As a simple measure of diversity, beekeepers determined the number of colors found in pollen samples that were collected in a coordinated and standardized way. Altogether, 750 beekeepers from 28 different regions from 24 countries participated in the two-year study and collected and analyzed almost 18,000 pollen samples. Pollen samples contained approximately six different colors in total throughout the sampling period, of which four colors were abundant. We ran generalized linear mixed models to test for possible effects of diverse factors such as collection, i.e., whether a minimum amount of pollen was collected or not, and habitat type on the number of colors found in pollen samples. To identify habitat effects on pollen diversity, beekeepers’ descriptions of the surrounding landscape and CORINE land cover classes were investigated in two different models, which both showed that both the total number and the rare number of colors in pollen samples were positively affected by ‘urban’ habitats or ‘artificial surfaces’, respectively. This citizen science study underlines the importance of the habitat for pollen diversity for bees and suggests higher diversity in urban areas.

Monitoring virus infections can be an important selection tool in honey bee breeding. A recent study pointed towards an association between the virus-free status of eggs and an increased virus resistance to deformed wing virus (DWV) at the colony level. In this study, eggs from both naturally surviving and traditionally managed colonies from across Europe were screened for the prevalence of different viruses. Screenings were performed using the phenotyping protocol of the ‘suppressed in ovo virus infection’ trait but with qPCR instead of end-point PCR and a primer set that covers all DWV genotypes. Of the 213 screened samples, 109 were infected with DWV, 54 were infected with black queen cell virus (BQCV), 3 were infected with the sacbrood virus, and 2 were infected with the acute bee paralyses virus. It was demonstrated that incidences of the vertical transmission of DWV were more frequent in naturally surviving than in traditionally managed colonies, although the virus loads in the eggs remained the same. When comparing virus infections with queen age, older queens showed significantly lower infection loads of DWV in both traditionally managed and naturally surviving colonies, as well as reduced DWV infection frequencies in traditionally managed colonies. We determined that the detection frequencies of DWV and BQCV in honey bee eggs were lower in samples obtained in the spring than in those collected in the summer, indicating that vertical transmission may be lower in spring. Together, these patterns in vertical transmission show that honey bee queens have the potential to reduce the degree of vertical transmission over time.

The varroa mite (Varroa destructor), a complex ectoparasite with a very distinctive biology (Rosenkranz et al., 2010) is considered a key pest affecting beekeeping and honey bee health worldwide (T...