|dc.description.abstract||Studies across different species have demonstrated the presence of an interplay between the
brain, the gut and the microbiome, most commonly referred to as the gut-brain axis.
Furthermore, it has been well documented that stress can affect neuroendocrinological and
immunological systems, resulting in altered behaviours, as well as physiological
dysregulations. In the past few decades, the effects of stress on the microbiome and the
implications of the microbial community structure for the host have been the focus of many
studies, aiming to shed light on this intricate and multi-pathway relationship.
Despite this, most studies have been conducted on humans and rodents, with very few on farm
animals, particularly ruminants. In ruminants, as in monogastrics, the gut accommodates high
microbial concentrations and facilitates host-microbial interactions. What differentiates
ruminants is the presence of the rumen, which hosts an equally important microbial
community. This organ acts as a primary location for fermentation of feed and plays a pivotal
role in animal metabolism, immunity and overall homeostasis. The effects of stress
susceptibility, and more specifically psychological or behavioural stress, have been poorly
explored in ruminants, despite their well-recognised and important effects on other aspects of
animal health and welfare.
This project therefore aimed to explore: 1) the effect of genetic predisposition to stress; 2) the
long-term effect of prenatal and early life stressful events; and 3) effects of repeated and
unpredictable management stress on the ovine gastrointestinal microbiome in conjunction with
various physiology and behavioural aspects.
The first experiment investigated genetic differences in gut and rumen microbial community
structure and blood cortisol concentrations in 58 adult Romane ewes, previously selected on
the basis of divergent reactivity to stress (30 ewes with high reactivity; 28 ewes with low
reactivity). The two groups differed in their behavioural reactivity towards a temporary
separation from congeners based on bouts of high bleats. Despite extensive analyses of the
microbiota at the phylum, order and genus level, there were only small significant differences
in the rumen and faecal microbiota, even when including cortisol levels in the analyses. For
example, higher levels of cortisol were positively correlated with Ruminococcus abundance in
faecal samples and Lactobacillus in the rumen, while Rikenellaceae abundance was positively
correlated with reactive EBV scores in faecal samples.
The second experiment investigated long-term effects of three prenatal stress treatments
(Control, Negative and Alternative) and the effect of two early-life treatments (Isolation or
Ewe Recognition tests) on the rumen microbial community structure of 35 8-month-old
Scottish Mule lambs, at a stage when the rumen microbiome had assumed a relative stable and
mature form. Sex and diet effects were confounded, while Prenatal Treatment did not appear
to have an effect. Neonatal Treatment had an effect on relative abundances at the phylum level.
The abundance of several bacterial species was correlated with higher or lower cortisol levels,
such as Lactobacillus in Isolated females and Proteobacteria in Isolated males; these
significant negative correlations suggested long-lasting effects of early life events.
Finally, in the third experiment, we explored the effect of a 6-week mild unpredictable Chronic
Stress paradigm on various behavioural (i.e., time budgets, reactions to a suddenness test) and
physiology aspects (i.e. hormonal levels, heart rate and VFAs), as well as the rumen and gut
microbiota structure. Forty-eight female Romane lambs were separated into two treatments:
Non-Treated (NT, n = 24) and Mild Chronically Stressed Animals (MCS, n = 24). Amongst
the most interesting results, indicating a treatment effect, were differences in synchronisation
of animals resting and sleeping, duration of resting time, and reactions to novelty, as expressed
by latency to approach the ball and contact time with the ball in the suddenness test. Microbiota
diversity indices, particularly for the non-treated group, indicated a different development of
the microbial community. In MCS animals, cortisol and serotonin levels indicated that several
bacteria proliferate in the presence/absence of these hormones, but correlation scores were
In conclusion, it appears that the microbiota community structure in the rumen is not
significantly affected by management stress or stress susceptibility, although the
communication pathways between rumen bacteria and host behaviour warrant further