The Asian brush-clawed shore crab Hemigrapsus takanoi has high tolerance for environmental changes facilitated the establishment of non-native populations along the Atlantic European coast. The self-maintenance and potential spread of this invasive crab will partially depend on its ability to disperse during the larval period. Larvae are not equipped with efficient osmoregulatory mechanisms to tolerate low salinity conditions; therefore, they evolved specific swimming behavior that facilitates exportation offshore for development in more stable and higher salinity conditions. To study the salinity tolerance, we quantified the survival of newly hatched larvae subjected to salinities ranging from 2 to 35 over a 24 h period. We observed that more than 50% of larvae could survive 24 h only at salinities higher than 20, and that shorter incubation periods of 2-6 h could produce high mortality at salinities lower than 10. We used video-tracking techniques to quantify swimming in newly hatched larvae at different levels of salinity, and under starvation or food availability conditions. The results showed that apparent swimming speed increases as salinity increases, and that upward trajectories are faster than downward ones. When food was available, the larvae reduced the frequency of helical swimming trajectories, turned out to be faster, straighter and more vertical. At salinities lower than 20, the swimming trajectories became more random, and the described patterns tended to disappear. Our results indicate that lower survival and reduced swimming performance may constrain the dispersal capacity of the non-native populations located at low salinity habitats.

No abstract

No abstract

Marine plankton classification is important for monitoring marine biological populations and understanding marine ecosystems. However, it is difficult to obtain abundant annotated image data of marine plankton to train classification models of high quality. To classify marine plankton with a small number of image samples, a cross‐domain few‐shot learning model (CDFM) is proposed. First, CDFM learns knowledge from existing labelled images and then transfers the knowledge to new images. In this process, there is an issue of domain differences between the new images and the existing datasets due to differences in data acquisition time and locations. To address this issue, the authors pre‐train a model in the source domain and then use fine‐tuning to adapt it to the target domain. Second, the graph neural network is used as a meta‐learning module to learn a feature distance metric for marine plankton classification with limited samples. Extensive experiments on four marine plankton image datasets are conducted, including Kaggle Plankton, miniPPlankton, ZooScan and WHOI, and CDFM outperforms existing methods for marine plankton classification.