Animals Free FullText Optimising Puppy SocialisationndashShort And LongTerm Effects Of A Training Programme During The Early So

4. Discussion
We investigated the short-term and long-term effects of “challenge exercises” (problem-solving games, novel objects, and potentially startling stimuli) during the early socialisation period on puppies’ behaviour. At the age of 6–7 weeks, the presentation of these challenges was associated with indicators of improved stress resilience. However, at the age of six months, there was no significant effect of treatment on owner-reported personality traits between the treatment and the control group. In contrast, significant effects of litter on several personality factors were found. This indicates that genetic effects and/or effects of the environment unique to each litter (independent of the training sessions) had a greater influence on behaviour at six months than the exercises between the 4th and 7th week of life. It is likely that a continuation of the treatment over a longer time period would be necessary to obtain lasting beneficial effects, since during the behaviour test at 6–7 weeks, the treatment group showed signs of higher boldness and better stress resilience than the control group in several contexts.

Puppies from the treatment group touched the novel object sooner and for a longer period of time, and they whimpered less than puppies from the control group during the novel object test. The latter spent more time in the proximity of a person during this test, which could be interpreted as seeking social support (c.f. [57]). Importantly, all of the puppies in the current study were well socialised by their breeders and had exposure to most of the enrichment items presented in the study by Vaterlaws-Whiteside and Hartmann [45]. Thus, the beneficial effect of the exercises presented to the treatment group on boldness go beyond those investigated by Vaterlaws-Whiteside and Hartmann [45].In response to the startle stimulus (loud noise), treatment puppies showed a reduced startle response and faster recovery (lower latency to start playing) compared to the control puppies. This suggests that the repeated—but carefully controlled—exposure to startling stimuli during the early socialisation period had enabled them to habituate and to generalise to novel sounds. Nonetheless, the effect was not sustained, as treatment groups did not differ in owner-reported sound sensitivity at six months of age (but see below regarding limitations of using a questionnaire to assess behaviour). If preventive training over a longer time period (such as until the puppies are adopted) can be shown to reduce the risk of noise fears in dogs, this would have major implications for canine welfare: up to half the pet dog population are affected by noise fears [58,59,60], and fearfulness displayed at a young age has been shown to further increase as dogs mature [48,60].Two previous studies have used presentations of sounds to young puppies. In Chaloupková, et al. [42], puppies from three litters received auditory stimulation (radio broadcasts including spoken-word programmes and music three times a day for 20 min) between 16 and 32 days. The control group comprised 18 puppies from 8 different litters. In a behaviour test at seven weeks of age, puppies from the treatment group showed more favourable reactions than the control group to different types of sudden loud noises, suggesting a beneficial effect of the stimulation [42]. Unfortunately, possible genetic effects were not controlled for in this study, since different litters made up the treatment and control groups.Alves, et al. [43], exposed a treatment group (N = 21) to two hours of auditory stimulation per day, starting at three weeks of age, while a control group with no such stimulation included 46 puppies. The stimulation included commercial music and radio talk shows, as well as noises such as car noise, sirens, and gunshots, played at the volume of a conversation. From week five on, noises were presented at the natural volume of gunshots and police sirens, and this was paired with play and food. Contrary to the prediction, no difference in reactions to loud noises between the treatment groups could be discerned when the puppies were tested in a puppy test at seven weeks. However, the two groups differed in other subtests including social interactions and reactions to restraint, with the control group receiving higher (i.e., for police dogs, better) scores. The authors concluded that “a rich and varied environmental stimulation may be more important and that no sole source of stimulation is essential” [43].Thus, while previous results regarding the benefits of auditory stimulation for young puppies have been inconclusive, our controlled method of presenting stimuli with increasing intensity was associated both with less intense immediate startle reactions and with faster recovery after a sudden loud noise in the puppies (even though this noise was louder than any of the noises used during treatment). Several explanations for the success of our method may apply: firstly, it is possible that real-life noises are more beneficial than recordings. For instance, it is known that in the context of treating noise fears, recordings may not be realistic enough. Acoustics are affected by the quality of the recording or the speakers and the setup of the room, and some noise-phobic dogs show no reaction to recordings [58,61,62].Secondly, exposure in the current study was carefully controlled. At the discretion of the experimenter, presentation was performed in such a way as to allow immediate startle recovery, and intensity was increased only when the puppies no longer showed a reaction at all. Although Alves, et al. [43], presented environmental noises at a reduced volume during the first two weeks, it cannot be ruled out that the presentation at full volume was too much for some puppies and might have induced a sensitisation, rather than a desensitisation.Thirdly, different litters were assigned to the treatment and the control group in the aforementioned studies, whereas we were able to control for genetic effects by assigning half of each litter to the treatment group and the other half to the control group. Thus, possible between-litter variation likely had less of an effect on the outcome than in the previous studies.

Despite the apparent benefit of the treatment when the puppies were tested at the ages of 6–7 weeks, no long-term effect of the treatment could be found. Notably, all puppies in our study were raised in a highly enriched environment during everyday life, which has been found to trump effects of specific treatments early during ontogeny also in previous studies [43,63]. In comparison, in Vaterlaws-Whiteside and Hartmann [45], the control group was raised in a relatively impoverished environment, and additionally, puppies’ keeping conditions were standardised. Lower genetic variation (all dogs in Vaterlaws-Whiteside and Hartmann [45] were purpose-bred retrievers) as well as lower environmental variation (breeding facilities as opposed to private breeders in our study) would further favour the detection of smaller effects than would be necessary in our study to reach statistical significance.Perhaps most important to consider is that our treatment took place very early in the socialisation period—the puppies only had three weeks of treatment, and so after treatment was terminated, they experienced another six to eight weeks of the socialisation period (assuming that this period ends at 12 to 14 weeks [26,27,32,39,64]). Viewed from this angle, the lack of long-term effect may not be surprising, and future studies should assess the effects of continuation of “challenge” exercises over a longer time period, such as until puppies are rehomed at eight weeks or older.The treatment and the control group differed in their responses to a friendly stranger during the behaviour test: on average, the control group approached her more quickly and spent more time interacting with her. While we did not code subtle behavioural signs to infer possible fear during this test, from personal observations, we would exclude that the lower interaction in the treatment group reflected fear-based avoidance.

We hypothesise that the differences might result from differential handling experiences between the treatment and control group, resulting in increased social attraction to an unfamiliar person in the control group. While the total time the experimenter spent with each group was the same, in the treatment group, she spent a proportion of this time preparing the exercises. Although the experimenter took care to perform picking up and feeding equally often in both groups, overall, the control group experienced more petting and playing with the experimenter than puppies from the treatment group, who were engaged in the exercises during this time. Future studies could include a control group with no interaction with the experimenter to prevent additional socialising effects, since strictly speaking, both groups in the current study received enrichment (social enrichment vs. environmental challenges).

If our hypothesis is confirmed, this would also be an important finding in relation to puppy socialisation during the COVID-19 pandemic, which made it less easy to invite a large number of visitors for puppy socialisation. The current results suggest beneficial effects of repeated visits even by the same person, if she interacts with the puppies in a pleasant way. In line with this, also in Vaterlaws-Whiteside and Hartmann [45], increased human interaction and handling by a trained researcher experienced by the treatment group was associated with better scores in responsiveness to a human assessor at six weeks. Moreover, Foyer, et al. [65], found that puppies from smaller litters scored higher on sociability as adults, with one explanation put forward being that each puppy might receive more human attention in smaller litters. In our study, there was no effect of litter size on sociability, possibly due to the large variety of breeds included.No significant effects of treatment on the puppies’ behaviour during the exploration subtest (“PC 4—Exploration”) nor on whimpering (all subtests—“PC 2—Whimpering”) were found. Thus, it may particularly be the more stressful situations of the behaviour test where the effects of “stress inoculation training” become apparent. Whimpering is commonly interpreted as signifying distress or anxiety [66,67], as well as seeking attention [68]. In our study, whimpering formed a separate component and was not associated with the puppy test components encompassing reactions to startling stimuli, problem-solving ability, or sociability, nor with treatment group. In Simpson [69], multiple functions were attributed to whimpering/whining (the two are combined in the source): greeting, frustration, pain, attention seeking, and defence. Although there were consistent individual differences in the amount of whimpering in our study, as reflected by the fact that whimpering in all subtests loaded on a single component, the relevance of this behaviour is not clear in this context since the component was independent from the others.After Fox ([64], as cited in Jensen [70]), whimpering (like crying and whining) is an infantile sound, and in line with this, this vocalisation decreased with age in the current study. This finding can be relevant regarding exact timings of future puppy tests, if whimpering is to be used as a variable of interest. It would not affect the current study results, however, since each litter included both treatment groups so that age was identical for the two treatment groups.There were significant effects of litter on “PC 1—Social-Startle” and “PC 3—Novel Object”, whereas no litter effects were found for “PC 2—Whimpering” and “PC 4—Exploration”. In line with our results, Riemer et al. [48], who used a similar puppy test, report litter effects for greeting, interaction (behaviours which in our study loaded on “PC 1—Social-Startle”) and “low boldness” (c.f. our “PC 3—Novel Object”). Additionally, unlike in the current study, litter effects also manifested in exploration/inactivity (c.f. our “PC 4—Exploration”). In any case, the existence of litter effects on at least some of the test outcomes at 6–7 weeks of age indicates the role of genetic and maternal effects and/or the shared early environment on behaviour during the test.Our long-term results also show highly significant effects of litter on the factors “Fearfulness” and “Aggression towards people”. The effect of litter on noise sensitivity became non-significant after correcting for multiple testing. No litter effects were found for “Aggression towards animals” (although there was a trend before correction), “Activity/excitability” and “Trainability”. It is possible that the latter two factors were more affected by individual training differences than the prior factors. The existence of litter effects, but not treatment effects, at six months reinforces the importance of genetic factors and/or early environmental and maternal influences for puppies’ development beyond the specific training programme. The time-limited exposure to challenge exercises over three weeks early in the socialisation period in the treatment group was not sufficient to create lasting measurable effects.

Limitations
The methodology employed has some limitations. The enrolled puppies belonged to various breeds and originated from different breeders; thus, there was significant variation in experiences during rearing (although too much stimulation provided by the breeder was an exclusion criterion for participation). Moreover, for reasons of puppy welfare, the treatment was not perfectly standardised. As puppies sleep a lot at this age, the order of the exercise types (novel object; problem solving task; startle recovery) within a day was adjusted to puppies’ activity level. For animal welfare reasons, we decided against certain interventions, e.g., waking puppies up to participate in a particular exercise at a particular time.

There was slight variation in the exercises presented to the different litters. The experimenter adapted the presentation of novel objects and problem-solving tasks for each litter a little, because some litters had previously been exposed to some objects in their home environment. Future follow-up studies would benefit from a more structured approach, in which puppies’ experiences (apart from the treatment) are more standardised, and the types of exercises and their order of presentation is identical for all litters (even if timing of presentation during the day may need to be adapted to activity level).

One puppy was excluded from the treatment group for two startle and recovery exercise sessions (both in the same day), having shown an enhanced reaction previously. After participating in two sessions with lower intensity stimuli, separate from the group, from the next day on, the puppy was successfully trained with the rest of the group again. We retained this puppy in the analysis so as not to bias the data (as removing less resilient individuals from the treatment group would exaggerate the measured effect of the treatment).

In the behaviour test, the first subtest (room exploration) was adapted for two litters and the breeder was present for these, but not for the remaining litters. Again, this decision was made for animal welfare reasons, as these litters had showed a higher tendency for fearful behaviour.

Furthermore, three litters were tested, not with the burst balloon, but with a lower intensity sound as startling stimulus, to accommodate either notable sound sensitivity in the whole litter or the setup of the testing room. We would expect treatment effects to become especially apparent under stressful conditions; thus, we could say that we were working against our own hypothesis by reducing the intensity of stimulation for some litters. Despite this, we found a significant difference between the treatment groups both in immediate reactions to the loud noise and in recovery. Therefore, we feel that the methodology was suitable to demonstrate the predicted effect of the treatment on noise sensitivity. On the other hand, no treatment effects were found for the exploration component (where 2 of 12 litters were tested with the breeder present). It cannot be ruled out that effects would have been stronger under more stressful conditions.

Finally, the behavioural assessment at six months was questionnaire-based. We used a validated personality questionnaire; however, this questionnaire targeted the traits of interest less specifically than in the behaviour test. It is a possibility that we would have seen group differences had we performed a similar behaviour test with novel objects, problem-solving tasks and startling stimuli at six months. It is a limitation of a questionnaire-based study that each dog is assessed by a different individual, and the owners may be biased to an extent. Thus, the standardised behaviour test is likely to be more sensitive to differences in reactivity to sounds than the questionnaire, and future studies should re-test dogs in a behaviour test as juveniles or (ideally) adults.