Subjects
The experiments of this study were carried out with wild cavies of the species Cavia aperea ERXLEBEN, 1777, derived from breeding stocks established at the Department of Behavioural Biology, University of Muenster. The animals were descendants from feral cavies trapped in the province of Buenos Aires, Argentina, in 1995 and from lineages belonging to the Universities of Bayreuth and Bielefeld, Germany. Animals of our breeding population were regularly replaced by animals stemming from different breeding stocks in Germany to ensure a reduction of potential inbreeding effects in our population. Since wild cavies do not show natural colour markings, the individuals were marked by bleaching the fur with 32% hydrogen peroxide once per month. The fur of newborn animals was bleached directly after birth for individual identification.
Housing conditions of pregnant and lactating females
Sixteen groups were composed, consisting of one male and two female wild cavies along with their preweaning offspring. Within these groups, eight unstable as well as eight stable social environments were established (for more detail see
Establishment of unstable and stable social environments
). Each group was kept in a 1.5 m2 enclosure. Pups that were not used as experimental animals were removed from the enclosures at the age of 20 ± 1 days.
All animals were kept under the following standardised conditions: temperature 22 ± 2 °C, relative humidity about 55 ± 10%, light/dark cycle 12:12 h with the light phase starting at 7:00 h in the morning. Commercial guinea pig diet (Höveler “Spezialfutter” 1070 for guinea pigs, Höveler Spezialfutterwerke, Germany; altromin 3023, Altromin Spezialfutter, Germany), oat flakes (Fortin Mühlenwerke, Germany), hay and water were available ad libitum. Vitamin C was added to the water twice per week as it is a daily requirement in the diet of guinea pigs and wild cavies since they lack one of the enzymes that are necessary to convert glucose to ascorbic acid.
All animals were housed in wooden enclosures in our animal facility. The floors were covered with wood shavings for bedding (Allspan, Germany) and cleaned every four weeks. The enclosures were enriched with a brick, two wooden branches and two cardboards as houses for shelter.
Establishment of unstable and stable social environments
Unstable social environment (UE): In eight groups, one of the two females was transferred to the clockwise neighbouring enclosure every second week. With a one-week shift, the other female was rotated counter-clockwise. This regularly exchange of the pregnant and lactating females between different groups led to a change in group compositions once a week. Lactating females were transferred with their offspring.
Stable social environment (SE): In contrast to unstable groups, the group composition of stable groups remained constant throughout the study. To prevent a handling bias, all females and their pups were handled in the same way at corresponding times.
The handling process of each animal of both social environmental conditions lasted approximately two minutes. It comprises catching the animal and transferring it into a new (UE) or the same (SE) social group.
Housing conditions of the sons
The experiment was conducted with 22 sons (SE-sons: 11, UE-sons: 11) originating from the females’ second litters after the establishment of SE- and UE-groups. Second litters were used in order to ensure that the whole development of the offspring during the prenatal phase took place in established stable and unstable social environmental conditions. Since all females gave birth to offspring already before we started the manipulation of social conditions and the males were experienced breeding males, we did not expect differences in parental care between the first and second litters after the establishment of stable and unstable social conditions. All sons stayed in their mothers’ groups (stable or unstable) until day 40 ± 2 days of age. The standardised housing conditions were the same as mentioned in chapter
Housing conditions of pregnant and lactating females
.
At day 40 ± 2 days of age, one SE- and one UE-son were placed simultaneously into an unknown enclosure (Fig. 1). In each enclosure, an unfamiliar social group consisting of one adult male and two adult females along with their preweaned offspring were already living at least for one week before the introduction of the two males took place. The offspring was removed from the enclosures at the age of 20 ± 1 days.
All enclosures were located outside in order to simulate semi-natural conditions. Temperature, relative humidity and the light/dark cycle followed prevailing natural conditions including spring, summer, autumn and winter conditions. Each enclosure was about 13 m2 and consisted of a less sheltered outdoor part and a more sheltered indoor part (less sheltered outdoor part 10 m2, more sheltered indoor part 3 m2). They were connected via a ramp and a plastic tube and enriched with hiding places (2 cardboards and 4 wooden houses) and wooden branches. In addition, the less sheltered outdoor part was subdivided into different areas by plastic walls. This additional fragmentation was important for a successful introduction of SE- and UE-sons in this new social groups since it is known for wild cavies that adult males and females can be highly aggressive against non-related animals [23, 24]. During wintertime, a heat radiator was placed into the more sheltered indoor parts of each enclosure to ensure that the animals had the chance to warm-up in a heated area.
All animals were fed in the same way as described above (see
Housing conditions of pregnant and lactating females
). Floors of the more sheltered indoor parts were covered with wood shavings, floors of the less sheltered outdoor parts were covered with sand. The whole enclosure was cleaned every four weeks.
Measurement of activity of the animals via radio-frequency identification
The radio-frequency identification (RFID) method used in this study allowed a computer-based analysis of activity of the animals of each social group of the outdoor enclosures. For this purpose, all animals carried an RFID-transponder (Type ID 100, TROVAN, Ltd.) under the skin of their neck. The transponder were injected subcutaneously under isoflurane anesthesia by using a specially designed injector (IID-100). The injection took place for UE- and SE-sons after weaning. In addition, two coiled antennae were placed around a plastic tube in front of each of the four wooden houses of the less sheltered outdoor parts as well as at the ramp which connected the less sheltered outdoor and the more sheltered indoor part. By passing the antennae, an electromagnetic field was induced which, in turn, leads to a transmission of the ID of the passing animal towards two reading devices (Typ LIDD665N, TROVAN, Ltd.). IDs of passing animals and time stamps were automatically registered by a SQL database over the whole observation period. For analyses, the observation period (in total 60 days) was summarised in intervals of 10 days: day 1–10, day 11–20, day 21–30, day 31–40, day 41–50 and day 51–60.
Blood sampling and body weights
Blood samples were collected at day 40 ± 2 of age directly before and 2 h after the introduction of SE- and UE-sons into the new social groups. The latter blood sample was collected in order to investigate the short-term reaction of the animals to this social challenge since it is known that maximum values of cortisol in wild cavies are reached on average two hours after the onset of a stressor [34]. The long-term influence on cortisol and testosterone concentrations due to this social challenge was investigated by sampling blood on day 5, 20, 30 and 40 after the introduction had taken place.
All blood samples were taken at a fixed point in time set at 13.00 h ± 15 min in order to avoid possible influences of circadian rhythm in hormone concentrations (domestic guinea pigs show diurnal variations in plasma cortisol titres with a peak around 13:00 h [35]). An increase in cortisol levels at this time point can be assigned to the challenging situation and is not related to circadian influences anymore since maximum values are already reached. The blood used for cortisol determination was collected within the first three minutes, blood samples for testosterone analysis within six minutes after entering the enclosure. Both, cortisol and testosterone concentrations are known to be unaffected by the occurrence of stress during this short time interval [35]. Blood samples were collected from the blood vessels of the ears. During blood sampling, one experimenter held the animal in his/her lap, while a second person collected the sample. In a first step, a heat rub (Finalgon, Boehringer Ingelheim Pharma, Germany) was applied to the ear of the animal to stimulate the blood circulation. After removing the salve, the vessels were illuminated with a hand-held LED light and pricked with a sterile injection needle. About 0.02 - 0.05 ml blood was collected in heparinized capillary tubes (Brand, Germany). Afterwards, the plasma was separated by centrifugation (11.700 ×g for 7 min) and subsequently deep-frozen at −20 °C until assayed. After each blood sampling, the body weights were measured.
Endocrine analyses
Plasma cortisol concentrations were analysed in duplicate using a luminescence immunoassay (Cortisol luminescence immunoassay RE62011, IBL International, Germany). The antibodies used cross-reacted with relevant steroids as follows: cortisol 100%, prednisolone 30%, 11-deoxycortisol 8.5%, cortisone 4.5%, prednisone 2.1%, corticosterone 2.0%, and 6ß-hydroxycortisol 1.0%. The intra- and interassay coefficients of variation were 4% and 7.2%, respectively.
Plasma testosterone concentrations were determined in duplicate using a solid phase enzyme-linked immunosorbent assay (ELISA; Testosterone ELISA Kit, Demeditec Diagnostics, Germany). The antibody cross-reacted with relevant steroids as follows: testosterone 100%, 11β-hydroxytestosterone 3.3%, 19-nortestosterone 3.3%, androstenedione 0.9%, 5α-dihydrotestosterone 0.8%, 17α-methyltestosterone 0.1%, epitestosterone, oestradiol, progesterone, cortisol, oestrone and danazol < 0.1% each. The intra- and interassay coefficients of variation were 5.74% and 7.23%, respectively.
Paternity tests
SE- as well as UE-sons stayed in the new social groups until both females of each group gave birth to their offspring after SE- and UE-sons had reached sexual maturity (i.e. around day 75 of age [25]). Since female wild cavies show a postpartum oestrus [36], all three males of one group (one adult male, one SE- and one UE-son) had the chance to mate once with each female. Paternities of in total 51 pups were determined by directly comparing alleles between potential fathers and offspring born in the groups as the genotypes of the potential fathers and the mothers were known. For analyses, tissue samples of the ear were collected via an ear punch after weaning. Genomic DNA was purified by first digesting the tissue samples using protein kinase, followed by phenol/chloroform extraction and DNA precipitation with ethanol. DNA pellets were washed with 70% ethanol several times and re-suspended in TE-buffer. Eleven microsatellites were amplified by PCR [30, 37] and sequenced. Alleles were analysed using GeneMarker®.
Statistical analysis
The data of plasma cortisol and testosterone concentrations as well as body weights were analysed using a generalized linear mixed model analysis with hierarchical random effects (GLMM, package “lme4”, routine “lmer”, R version 3.1.1 (2014-07-10)) where treatment (stable or unstable), experimental day and litter size (covariate) were fixed factors and pair ID a random factor. The model was fitted to the data by maximum likelihood techniques and statistically tested for significance by Wald test. The statistical method was used for cortisol and testosterone since this kind of analysis is able to deal with rare missing values at different time points, which occurred within this study due to technical reasons during our blood sampling procedure and the following analyses. No significant influence of pair ID could be revealed.
Activity data were analysed using a two-way analysis of covariance with repeated measures based on the effects of treatment and time (repeated measures factor) and the covariate litter size (Statistical program package BMDP, [38]). A log transformation of activity data and cortisol and testosterone data was performed since their statistical distributions were skewed to the right and deviated from a normal distribution shape. This was inspected by residual analyses by means of Q-Q-plots and histograms (R routine “3.1.1 (2014-07-10)”). Three animals per group (three UE- and three SE-sons) had to be excluded from the analysis due to missing values.
SigmaPlot 10.0 for Windows software (SPSS Inc., Chicago, USA) was used to create graphs. These show the means and standard errors of activity data and plasma cortisol and testosterone concentrations. In general, for each observed parameter in each phase of the experiment, a statistical significance level of α = 0.05 was used to test the two-sided hypotheses (first kind error rate related to each comparison).