Monitoring animals in their natural habitat is essential for advancement of animal behavioural studies, especially in pollination studies. Non-invasive techniques are preferred for these purposes as they reduce opportunities for research apparatus to interfere with behaviour. One potentially valuable approach is image-based tracking. However, the complexity of tracking unmarked wild animals using video is challenging in uncontrolled outdoor environments. Out-of-the-box algorithms currently present several problems in this context that can compromise accuracy, especially in cases of occlusion in a 3D environment. To address the issue, we present a novel hybrid detection and tracking algorithm to monitor unmarked insects outdoors. Our software can detect an insect, identify when a tracked insect becomes occluded from view and when it re-emerges, determine when an insect exits the camera field of view, and our software assembles a series of insect locations into a coherent trajectory. The insect detecting component of the software uses background subtraction and deep learning-based detection together to accurately and efficiently locate the insect among a cluster of wildflowers. We applied our method to track honeybees foraging outdoors using a new dataset that includes complex background detail, wind-blown foliage, and insects moving into and out of occlusion beneath leaves and among three-dimensional plant structures. We evaluated our software against human observations and previous techniques. It tracked honeybees at a rate of 86.6% on our dataset, 43% higher than the computationally more expensive, standalone deep learning model YOLOv2. We illustrate the value of our approach to quantify fine-scale foraging of honeybees. The ability to track unmarked insect pollinators in this way will help researchers better understand pollination ecology. The increased efficiency of our hybrid approach paves the way for the application of deep learning-based techniques to animal tracking in real-time using low-powered devices suitable for continuous monitoring.

Climate change has the potential to enhance or disrupt biological systems, but currently, little is known about how organism plasticity may facilitate adaptation to localised climate variation. The bee-flower relationship is an exemplar signal-receiver system that may provide important insights into the complexity of ecological interactions in situations like this. For example, several studies on bee temperature preferences show that bees prefer to collect warm nectar from flowers at low ambient temperatures, but switch their preferences to cooler flowers at ambient temperatures above about 30° C. We used temperature sensor thermal probes to measure the temperature of outdoor flowers of 30 plant species in the Southern regions of the Australian mainland, to understand how different species could modulate petal temperature in response to changes in ambient temperature and, potentially, influence the decision-making of bees in the flowering plant’s favour. We found that flower petal temperatures respond in different ways to changing ambient temperature: linearly increasing or decreasing relative to the ambient temperature, dynamically changing in a non-linear manner, or varying their temperature along with the ambient conditions. For example, our investigation of the difference between ambient temperature and petal temperature (ΔT), and ambient temperature, revealed a non-linear relationship for Erysimum linifolium and Polygala grandiflora that seems suited to bee temperature preferences. The temperature profiles of species like Hibertia vestita and H. obtusifolia appear to indicate that they do not have a cooling mechanism. These species may therefore be less attractive to bee pollinators in changing climatic conditions with ambient temperatures increasingly above 30° C. This may be to the species’ detriment when insect-pollinator mediated selection is considered. However, we found no evidence that flower visual characteristics used by bees to identify flowers at close range, such as colour or shape, were straightforward modulators of floral temperature. We could not identify any clear link to phylogenetic history and temperature modulation either. Mapping our test flower distribution on the Australian continent however, indicates a potential clustering that suggests different flower responses may constitute adaptations to local conditions. Our study proposes a framework for modelling the potential effects of climate change and floral temperature on flower pollination dynamics at local and global scales.

In this paper the authors pose 10 questions they consider the most important for understanding generative computer art. For each question, the authors briefly discuss its implications and suggest how it might form the basis for further discussion.