Herding of sheep by dogs is a powerful example of one individual causing many unwilling individuals to move in the same direction. Similar phenomena are central to crowd control, cleaning the environment and other engineering problems. Despite single dogs solving this ‘shepherding problem’ every day, it remains unknown which algorithm they employ or whether a general algorithm exists for shepherding. Here, we demonstrate such an algorithm, based on adaptive switching between collecting the agents when they are too dispersed and driving them once they are aggregated. Our algorithm reproduces key features of empirical data collected from sheep–dog interactions and suggests new ways in which robots can be designed to influence movements of living and artificial agents.

Flocking is a striking example of collective behaviour that is found in insect swarms, fish schools and mammal herds. A major factor in the evolution of flocking behaviour is thought to be predation, whereby larger and/or more cohesive groups are better at detecting predators (as, for example, in the 'many eyes theory'), and diluting the effects of predators (as in the 'selfish-herd theory') than are individuals in smaller and/or dispersed groups. The former theory assumes that information (passively or actively transferred) can be disseminated more effectively in larger/cohesive groups, while the latter assumes that there are spatial benefits to individuals in a large group, since individuals can alter their spatial position relative to their group-mates and any potential predator, thus reducing their predation risk. We used global positioning system (GPS) data to characterise the response of a group of 'prey' animals (a flock of sheep) to an approaching 'predator' (a herding dog). Analyses of relative sheep movement trajectories showed that sheep exhibit a strong attraction towards the centre of the flock under threat, a pattern that we could re-create using a simple model. These results support the long-standing assertion that individuals can respond to potential danger by moving towards the centre of a fleeing group.

The pronounced cachexia (unexplained wasting) seen in Huntington’s disease (HD) patients suggests that metabolic dysregulation plays a role in HD pathogenesis, although evidence of metabolic abnormalities in HD patients is inconsistent. We performed metabolic profiling of plasma from presymptomatic HD transgenic and control sheep. Metabolites were quantified in sequential plasma samples taken over a 25 h period using a targeted LC/MS metabolomics approach. Significant changes with respect to genotype were observed in 89/130 identified metabolites, including sphingolipids, biogenic amines, amino acids and urea. Citrulline and arginine increased significantly in HD compared to control sheep. Ten other amino acids decreased in presymptomatic HD sheep, including branched chain amino acids (isoleucine, leucine and valine) that have been identified previously as potential biomarkers of HD. Significant increases in urea, arginine, citrulline, asymmetric and symmetric dimethylarginine, alongside decreases in sphingolipids, indicate that both the urea cycle and nitric oxide pathways are dysregulated at early stages in HD. Logistic prediction modelling identified a set of 8 biomarkers that can identify 80% of the presymptomatic HD sheep as transgenic, with 90% confidence. This level of sensitivity, using minimally invasive methods, offers novel opportunities for monitoring disease progression in HD patients.

With their large brains and long lives, sheep offer advantages over rodents for studying progressive neurodegenerative disease in timeframes relevant to humans. Perentos et al. demonstrate the validity of an ovine model of Batten disease by revealing in vivo EEG abnormalities that resemble those in affected children.

Huntington’s disease (HD) patients show reduced flexibility in inhibiting an already-started response. This can be quantified by the stop-signal task. The aim of this study was to develop and validate a sheep version of the stop-signal task that would be suitable for monitoring the progression of cognitive decline in a transgenic sheep model of HD. Using a semi-automated operant system, sheep were trained to perform in a two-choice discrimination task. In 22% of the trials, a stop-signal was presented. Upon the stop-signal presentation, the sheep had to inhibit their already-started response. The stopping behaviour was captured using an accelerometer mounted on the back of the sheep. This set-up provided a direct read-out of the individual stop-signal reaction time (SSRT). We also estimated the SSRT using the conventional approach of subtracting the stop-signal delay (i.e., time after which the stop-signal is presented) from the ranked reaction time during a trial without a stop-signal. We found that all sheep could inhibit an already-started response in 91% of the stop-trials. The directly measured SSRT (0.974 ± 0.04 s) was not significantly different from the estimated SSRT (0.938 ± 0.04 s). The sheep version of the stop-signal task adds to the repertoire of tests suitable for investigating both cognitive dysfunction and efficacy of therapeutic agents in sheep models of neurodegenerative disease such as HD, as well as neurological conditions such as attention deficit hyperactivity disorder.