Abstract: Horses are a highly social species living in complex social systems which should require them to memorise and generalise social experiences and distinguish between familiar and unfamiliar conspecifics. In the main part of my thesis I concentrated on the specific conflict situation of a horse being introduced into a new social group, and investigated its behaviour and stress level. Horses were either introduced (1) immediately, (2) after an observation period, or (3) together with an integration horse after an observation period. Additionally, in the second part of my thesis I arranged several experiments to elaborate additional aspects which could affect the behaviour of horses during introductions. In this study I could describe a simplified method for measuring stress through the analysis of faecal GCMs in horses. An enzyme immunoassay (EIA) for 11-oxoaetiocholanolone using 11-oxoaetiocholanolone-17-CMO: BSA (3?,11-oxo-A EIA) as antigen showed high amounts of immunoreactive substances. The new assay increases the accuracy of the test and lowers the expenses per sample; also storing of samples at room temperature after collection is less critical. This is a big advantage both in the field of wildlife management of equids and in the field of equestrian sports (chapter 1). Comparing the different introduction techniques, the introduction with an integration horse led to significantly less total interactions and lower levels of aggression than the introduction of single horses, both immediately and after several days of observing the new group. Additionally, by observing the behaviour of the horses during everyday sociality I could develop a formula describing the interrelationship between expected aggression level and enclosure size per horse. The curve takes an exponential shape. Starting from a space allowance of 300 m2 and more per horse, the amount of aggressions per hour approaches zero. For the reduction of aggression levels and injury risks in socially kept horses I recommend an enclosure size of at least 300 m2 per horse (chapter 2). I further investigated the stress level of the introduced animals. Horses which were immediately introduced did not show elevated faecal GCMs. In contrast, horses which were introduced after an observation period had slightly elevated values 2 and 3 days after the introduction. For horses introduced together with an integration horse faecal GCMs were significantly above the baseline value on the day of introduction and 1 day after it. These differences between introduction techniques indicate that the introduction event itself is not as stressful as previously assumed. Rather standing together with an integration horse and not being able to integrate immediately into the complete group elicits stress in horses (chapter 3). In the commentary of chapter 4 several studies are discussed which failed to demonstrate social learning in horses. It is argued that they did not consider important aspects which could have an influence, such as the dominance status or the social background of the horses (chapter 4). In chapter 5 a social feeding situation was investigated. The social rank as well as the position of conspecifics affected the feeding strategy of horses. Domestic horses used social cognition and strategic decision making in order to decide where to feed. When possible they tended to return to the same, continuously supplied feeding site and switched to an ?avoidance tendency? in the presence of dominant horses or when another horse was already feeding there (chapter 5). One possibility to recognize group members is through olfactory recognition. In chapter 6 it is shown that horses are able to distinguish their own from their conspecifics? faeces. In addition, they paid most attention to the faeces of those group members from which they received the highest amount of aggressive behaviour (chapter 6). Horses show cognitive abilities because they are able to use humans as local enhancement cues when searching for food, independently of their body posture or gaze consistency when the persons face them. Moreover, they seem to orientate on the attention of familiar persons more than of unfamiliar persons (chapter 7). Altogether, the results of this thesis provide further support for the view that horses show good conflict resolution strategies. They are perfectly able to deal with the conflict situation of being introduced to new group members, and the introduction event itself is not as stressful as previously assumed. It is rather suggested that standing together with an integration horse and not being able to integrate immediately into the complete group elicits stress in horses. All additional experimental set-ups could demonstrate that horses are well capable of social cognition.

Abstract: Durch die vorliegende Studie wurde der Einfluss von Strukturelementen auf das Liege- und Aggressionsverhalten von Pferden in Gruppenhaltung untersucht. Die Strukturelemente sollten Rückzugsmöglichkeiten bieten, sowie in ihrem Bereich zur Aufhebung der Individualdistanz führen und dadurch die Fläche relativ vergrößern. Für die Versuche stellte das Haupt- und Landesgestüt Marbach drei unabhängige Pferdegruppen mit verschiedener Herdengröße zur Verfügung, die jeweils in Einraum- Innenlaufställen gehalten wurden. Die Datenerfassung fand zwischen 23 und sieben Uhr statt. Insgesamt wurden 366 Stunden Videomaterial ausgewertet. Beim Ruheverhalten wurden mit Hilfe des event-sampling- Verfahrens die Parameter Gesamtliegedauer, Dauer in Seitenlage, Dauer der Einzelphasen in
Seitenlage und Abliegehäufigkeit pro Nacht erfasst. Es wurden Versuchsphasen ohne
Strukturelemente und mit hängenden Planen als Strukturierung durchgeführt. Im ersten Stall
kamen außerdem über einander gestapelte Strohballen zum Einsatz. Diese Art der
Strukturierung stellte sich jedoch als nicht praktikabel heraus und führte im Vergleich zu den Planen zu einer signifikanten Verschlechterung der Gesamtliegedauer. In Stall 1 konnte man eine tendenzielle Verbesserung der Parameter Gesamtliegedauer und Gesamtdauer in Seitenlage durch das Anbringen von Planen sehen. In Stall 3 dagegen verschlechterte sich das Ruheverhalten in den Versuchsphasen mit Planen gegenüber den Versuchsphasen ohne Struktur. Die Werte der Gesamtdauer in Seitenlage nahmen signifikant ab. In Stall 2 zeigten sich keine Unterschiede in den verschiedenen Versuchsphasen. Möglicherweise ist die Wirkung der Strukturelemente auf das Liegeverhalten abhängig von der Flächengröße. Stall 1 hatte bezogen auf die Leitlinien des BMELV die größte und Stall 3 die kleinste Fläche. Bei den anderen Parametern des Liegeverhaltens gab es keine signifikanten Unterschiede in den verschiedenen Versuchsphasen. Mit Hilfe des time-sampling-Verfahrens wurde die Anzahl gleichzeitig liegender Pferde und gleichzeitig liegender Pferde in Seitenlage bestimmt. Auch hier zeigten sich keine Unterschiede in den verschiedenen Versuchsphasen. Bei der Gegenüberstellung der Werte der Gesamtliegedauer und der Gesamtdauer in Seitenlage mit dem Alter der Pferde (Stall 1 und Stall 3) und mit dem Integrationszeitpunkt (Stall 1) konnte kein Zusammenhang festgestellt werden. Beim Aggressionsverhalten wurden mit Hilfe des
event-sampling-Verfahrens in den Ställen 2 und 3 verschiedene Arten von Aggressionen
erfasst, die dann in die drei Intensitätsgrade Low-Level-, Mid-Level- und High-Level-
Aggressionen unterteilt wurden. Neben der Anzahl wurde die Dauer der verschiedenen
Aggressionen bewertet. Insgesamt konnte eine positive Wirkung der Planen auf das
Aggressionsverhalten beobachtet werden. Die Gesamtanzahl an Aggressionen nahm in beiden
Ställen tendenziell in den Versuchsphasen mit Planen ab. In Stall 3 konnte, wenn man die
Aggressionen stundenweise betrachtet, ein signifikanter Unterschied festgestellt werden.
Auch der Hinterhandschlag und die Aggressionen, welche das Ruheverhalten stören,
verringerten sich tendenziell nach dem Anbringen von Strukturelementen. In beiden Ställen
nahm die relative Häufigkeit von Mid-Level-Aggressionen nach dem Anbringen von Planen
zu. Dagegen konnte bei den High-Level-Aggressionen und in Stall 3 bei den Low-Level-
Aggressionen eine relative Abnahme beobachtet werden. Sowohl die Anzahl als auch die
Dauer der Mid-Level-Aggressionen verringerten sich in Stall 3 stundenweise betrachtet
signifikant in den Versuchsphasen mit Planen. In Stall 2 war bei den High-Level-
Aggressionen sowohl bei der Dauer als auch bei der Anzahl eine signifikante Abnahme zu
sehen. Wenn man das Aggressionsverhalten in Bezug zu der Fläche in den zwei Ställen
betrachtet, schien diese vor allem einen Einfluss auf die High-Level-Aggressionen zu
nehmen. Mit Hilfe des Rangindex der Pferde im Stall 3 wurde eine Rangordnung aufgestellt.
Zwischen dem Platz der Pferde in der Rangordnung und den Parametern des Ruheverhaltens
(Gesamtliegedauer, Gesamtdauer in Seitenlage) sowie dem Alter der Pferde konnte kein
Zusammenhang festgestellt werden. Zwischen dem Rangindex und der Gesamtanzahl an
Aggressionen bestand dagegen ein hoch signifikanter Zusammenhang. Im Rahmen dieser
Studie ist das Anbringen von Strukturelementen in Bezug auf das Aggressionsverhalten in
Einraum-Innenlaufställen von Pferden zu empfehlen. Die Wirkung auf das Liegeverhalten der
Pferde sollte in weiteren Studien untersucht werden.
[In this study the influence of structural elements on lying and aggression behavior in horses
kept in groups was exanimated. Structural elements should offer the possibility of retreat.
Furthermore, within the respective area, the individual distance should be limited, thus
increasing the overall space relatively. For the studies the “Haupt- und Landesgestüt
Marbach” provided three non-related groups of different sized horses, each kept in separate
“Einraum-Innenlaufställen”. Data collection took place between 11 pm and 7 am. Altogether,
366 hours of video material were analyzed. Using the event-sampling-method, the following
parameters within the horses´ resting behavior were determined: The total lying period, the
total lying period in lateral position, the periods of the single lying in lateral position and the
frequency of lying down. Test phases were carried out without structural elements and with
hanging canvases as structural elements. In the first stable there were also three bales of straw
piled up on top of each other. This kind of structuring emerged as impractical and in
comparison to the canvas, led to a significant deterioration to the total period of lying. After
the canvases were fixed in the first stable, a slight improvement with the parameters of total
lying period and total lying period in the lateral position could be seen. In contrast, the resting
behavior in stable 3 worsened in the test phases with the canvases in comparison to the test phases without structural elements. The values for the total lying period in the lateral position
decreased significantly. In stable 2, there were no differences in the different test phases.
Possibly the effect of the structural elements is dependent of the space. Stable 1 had the
largest space and stable 3 the least space referred to the guideline of the BMELV. The other
parameters of the lying behavior showed no significant differences in the different test phases.
The number of horses lying at the same time, and the number of horses lying at the same time
in the lateral position were detected with the aid of the time-sampling-method. There were
also no differences between the different test phases. If the values of the total lying period and
the total lying period in lateral position were compared to the age of the horses (stable 1 and
stable 3) and to the point of integration (stable 1), there was no correlation.
Different kinds of aggression were detected by the means of the event-sampling-method in
stable 2 and 3 and were subdivided into the three levels of intensity: Low-level-, mid-level
and high-level-aggressions. Both the number and the duration of the different aggressions
were evaluated. All in all a positive influence of the canvas on the aggression behavior could
be observed. In both stables, the total number of aggressions decreased in the test phases with
canvases. In stable 3 a significant difference could be detected, if the aggressions were
observed per hour. Also hind limb kicks and aggressions disturbing the resting behavior,
decreased by trend, after fixing the canvases in the stable. In both stables the relative
frequency of mid-level-aggressions increased after the canvases were mounted. In contrast
there was a relative decrease at the high-level-aggressions and in stable 3 at the low-levelaggressions.
The number and the duration of the mid-level-aggressions in stable 3 decreased,
when considered by the hour in the test phases with canvases. In stable 2 a significant
decrease in duration and number of high-level-aggressions was observed. When examining
aggression behavior in reference to space in the two stables, high-level-aggressions seemed to
be mostly influenced. In stable 3 a rank order was established with the aid of the rank index of
the horses. There was no correlation between the horses´ rank and the parameters of the
resting behavior (the total lying period, the total lying period in lateral position) and between
horses´ rank and age. In contrast, there was a high significant correlation between the rank
index and the total number of aggressions. In context to this study, structural elements in
“Einraum-Innenlaufställen” (BMELV 2009) seem to have a positive influence on the
aggression behavior and thus seem to be recommendable. The influence on the lying behavior
should be investigated in further studies.]

Abstract: Hawk-dove games have been extensively used to predict the conditions under which group-living animals should defend their resources against potential usurpers. Typically, game-theoretic models on aggression consider that resource defense may entail energetic and injury costs. However, intruders may also take advantage of owners who are busy fighting to sneak access to unguarded resources, imposing thereby an additional cost on the use of the escalated hawk strategy. In this article we modify the two-strategy hawk-dove game into a three-strategy hawk-dove-sneaker game that incorporates a distraction-sneaking tactic, allowing us to explore its consequences on the expected level of aggression within groups. Our model predicts a lower proportion of hawks and hence lower frequencies of aggressive interactions within groups than do previous two-strategy hawk-dove games. The extent to which distraction sneakers decrease the frequency of aggression within groups, however, depends on whether they search only for opportunities to join resources uncovered by other group members or for both unchallenged resources and opportunities to usurp.

Abstract: This is the first study to investigate the short-term effects of high population density on captive chimpanzees (Pan troglodytes). Subjects of the study were 45 chimpanzees living in five different groups at the Yerkes Regional Primate Research Center. The groups were observed under two conditions: 1) when they had access to both the indoor and outdoor sections of their enclosures; 2) during cold days when they were locked into the indoor runs, which reduced the available space by more than half. Under the high-density condition, allogrooming and submissive greetings decreased, but juvenile play increased. Remarkably, the rate of various forms of agonistic behavior, such as aggression, bluff charge, bluff display, and hooting, occurred less frequently under the high-density condition. This general decrease in adult social activity, including agonistic behavior, can be interpreted as an inhibition strategy to reduce opportunities for conflict when interindividual distances are reduced. This strategy is probably effective only in the short run, however. Behavioral indicators of anxiety, such as rough scratching and yawning, showed elevated rates, suggesting increased social tension under the high-density condition.

Abstract: A consistent conclusion in reconciliation research is that animals that reconcile are likely to have strong social bonds. This has led to the hypothesis that reconciliation occurs most often between valuable social partners. We tested this hypothesis in a group of Assamese macaques (Macaca assamensis) living near a temple in Assam, India. Using focal sample and ad libitum data collection, we recorded the occurrence of reconciliation, grooming, and agonistic aiding, and the outcomes of approach. We used matrix association methods (TauKr correlation) to correlate reconciliation with grooming, aiding, and approach outcome. Females reconciled more often with females with which they had stronger grooming and aiding relationships. The correlation between reconciliation and aiding was significant for support to the aggressor and the victim. In contrast, no such correlations with reconciliation were found for males. This study provides evidence that females reconcile most often with valuable and compatible social partners. The results do not support the relationship-quality hypothesis for males, and we suggest that future studies give more consideration to the possibility that males reconcile for reasons other than to repair relationships with valuable partners.