Abstract: For centuries, scientists have attempted to identify and document analytical laws that underlie physical phenomena in nature. Despite the prevalence of computing power, the process of finding natural laws and their corresponding equations has resisted automation. A key challenge to finding analytic relations automatically is defining algorithmically what makes a correlation in observed data important and insightful. We propose a principle for the identification of nontriviality. We demonstrated this approach by automatically searching motion-tracking data captured from various physical systems, ranging from simple harmonic oscillators to chaotic double-pendula. Without any prior knowledge about physics, kinematics, or geometry, the algorithm discovered Hamiltonians, Lagrangians, and other laws of geometric and momentum conservation. The discovery rate accelerated as laws found for simpler systems were used to bootstrap explanations for more complex systems, gradually uncovering the “alphabet” used to describe those systems.

Abstract: Social behaviour was observed in individually marked goats in two herds. The goats from one herd (n = 98) were horned, those of the other herd (n = 83) were polled. By recording agonistic interactions within the herds, a dominance index was determined for each animal. In both herds, intervention took place. Intervention is defined as one animal pushing in between two fighters, and thus ending the fight. More cases of intervention took place per individual animal amongst the horned goats than amongst the polled ones. Goats which intervened in fights on several occasions usually had a high dominance index. Members of the herd which were observed intervening only once had an average dominance index in both herds of almost 0.5. In some cases, goats very low in the rank order intervened a fight. Only rarely did the intervenors have a lower dominance index than the two fighters. In 103 cases, the direct dominance relationship between a fighting animal and the intervenor was known. In 95 cases (92.2%), the intervenor was dominant to the herd member in this fight and in just eight cases (7.8%), it was subordinate. It could not be determined what advantage the intervenor gained from its activity. It is possible that, at least in certain cases, a particularly relationship existed between the intervenor and one of the fighters.

Abstract: Abstract Models of swarming (based on avoidance, alignment and attraction) produce patterns of behaviour also seen in schools of fish. However, the significance of such similarities has been questioned, because some model assumptions are unrealistic [e.g. speed in most models is constant with random error, the perception is global and the size of the schools that have been studied is small (up to 128 individuals)]. This criticism also applies to our former model, in which we demonstrated the emergence of two patterns of spatial organization, i.e. oblong school form and high frontal density, which are supposed to function as protection against predators. In our new model we respond to this criticism by making the following improvements: individuals have a preferred ‘cruise speed’ from which they can deviate in order to avoid others or to catch up with them. Their range of perception is inversely related to density, with which we take into account that high density limits the perception of others that are further away. Swarm sizes range from 10 to 2000 individuals. The model is three-dimensional. Further, we show that the two spatial patterns (oblong shape and high frontal density) emerge by self-organization as a side-effect of coordination at two speeds (of two or four body lengths per second) for schools of sizes above 20. Our analysis of the model leads to the development of a new set of hypotheses. If empirical data confirm these hypotheses, then in a school of real fish these patterns may arise as a side-effect of their coordination in the same way as in the model.

Abstract: Summary We compared the genetic diversity and distance among six German draught horse breeds to wild (Przewalski's Horse), primitive (Icelandic Horse, Sorraia Horse, Exmoor Pony) or riding horse breeds (Hanoverian Warmblood, Arabian) by means of genotypic information from 30 microsatellite loci. The draught horse breeds included the South German Coldblood, Rhenish German Draught Horse, Mecklenburg Coldblood, Saxon Thuringa Coldblood, Black Forest Horse and Schleswig Draught Horse. Despite large differences in population sizes, the average observed heterozygosity (Ho) differed little among the heavy horse breeds (0.64�0.71), but was considerably lower than in the Hanoverian Warmblood or Icelandic Horse population. The mean number of alleles (NA) decreased more markedly with declining population sizes of German draught horse breeds (5.2�6.3) but did not reach the values of Hanoverian Warmblood (NA = 6.7). The coefficient of differentiation among the heavy horse breeds showed 11.6% of the diversity between the heavy horse breeds, as opposed to 21.2% between the other horse populations. The differentiation test revealed highly significant genetic differences among all draught horse breeds except the Mecklenburg and Saxon Thuringa Coldbloods. The Schleswig Draught Horse was the most distinct draught horse breed. In conclusion, the study demonstrated a clear distinction among the German draught horse breeds and even among breeds with a very short history of divergence like Rhenish German Draught Horse and its East German subpopulations Mecklenburg and Saxon Thuringa Coldblood.

Abstract: Personality in animal behaviour describes the observation that behavioural differences between individuals are consistent over time and context. Studies of group-living animals show that movement order among individuals is also consistent over time and context, suggesting that some individuals lead and others follow. However, the relationship between leadership and personality traits is poorly studied. We measured several personality traits and leadership of individual barnacle geese, Branta leucopsis. We measured body size and scored the dominance of individuals living in a stable group situation before subjecting them to an open-field test, an activity test, a novel-object test, and a leadership test in which the order of the movement of individuals in pairs towards a feeding patch was scored. We found high repeatability for activity and novel-object scores over time. Leadership was strongly correlated with novel-object score but not with dominance rank, activity or exploration in an open field. These results provide evidence that leadership is closely related to some aspects of personality. Interestingly, an individual's arrival at the food patch was affected not only by the novel-object score of the focal individual, but also by the novel-object score of the companion individual, indicating that movement patterns of individuals living in groups are affected by the personality traits of other group members and suggesting that movement patterns of a group may be shaped by the mix of personality types present in the group.