There is anecdotal evidence from farmers that cyclic animals grouped together tend to synchronise their oestrous cycles and the study of Izard and Vandenbergh [16
] supports this. However, no study has ever provided indisputable proof of inter-female primer pheromones causing synchronisation of oestrous cycles in cattle. This could be because there is no such pheromone or because the complexity of large mammals makes it difficult to design and carry out the necessary experiments. In the present study, unique in that it was performed on animals in isolation, we found no effect of oestrous urine and vaginal mucus on the cycle length, neither when measured as the interval between LH surges nor as the interval between ovulations.
There are several possible explanations why there was no significant effect of the type of exposure on oestrous cycle length, which was the underlying hypothesis of this study.
Firstly, there is of course the possibility that no oestrous synchrony pheromone exists. We did, however, see differences in expression of oestrus and also a tendency for an effect on LH concentration between the two types of exposure, which may be caused by the exposure to oestrous substances, so the hypothesis of a bovine inter-female pheromone cannot be rejected. However, the two treatment cycles investigated in the present study may not be enough time for the effect to fully manifest. In the present study we focused on the follicular phase. Investigating the transition between luteal and follicular phases, in particular the release of prostaglandin, might have provided additional information and is of great interest for future studies.
Secondly, there may be a risk that bioactivity of the samples was lost during the handling and storage of the samples, either through evaporation or through degradation. Studies on bovine vaginal mucus [18
] and female elephant urine [29
], however, indicate that this risk is slight under present conditions.
Thirdly, our assumption was that the cycle length within an unexposed animal is constant. It is possible that the inter-cycle variability is greater than presumed, which would prevent the detection of any influence of treatment on the cycle length. We did, however, find a significant effect of cycle on peak interval, which was due to the difference in length of the control cycles. These different patterns are difficult to explain, but they show that an effect of the treatment cannot be rejected.
Fourthly, the theoretical coupled-oscillator model for oestrous synchrony that has been described previously in humans and in rodents [15
], includes two different pheromones with adverse actions, one that shortens the cycle length and a second that has a lengthening effect. In humans, the first pheromone is thought to be released during the follicular phase (2-4 days before the preovulatory LH surge) and the second during the ovulatory phase (the day of the LH surge to two days after) [15
]. In the present study samples were collected during the 48 h preceding ovulation, i.e. both before and after the preovulatory LH surge. If the coupled-oscillator theory is true for the cow, this may have caused a counteraction mechanism, masking any effect on the cycle length. It is imperative to state, however, that this theoretical model [32
] and oestrous synchrony as such has been called into question [33
Finally, there may also be a synergistic effect between pheromones and other biostimuli, i.e. that it's not only the substance that is needed for the effect to manifest, but also the oestrous behaviour of the other animal. Since the animals in the present study were kept isolated, they were not exposed to such visual or tactile stimuli from other animals. Another possibility is that the effect is mediated by a signature mixture, rather than a pheromone. A signature mixture is an individual-specific mix of chemical signals that need to be learnt by the receiver [34
]. Nishimura et al. [21
] demonstrated that heifers, smeared with their own oestrual discharge during dioestrus, were nearly always mounted by their herd mates, while dioestrual heifers, that were smeared with the oestrual discharge of another animal, were not. These results support the theory of a signature mixture, rather than a pheromone, acting between females, which could explain the lack of an effect on the cycle length in the present study, where the animals were exposed only to the substances. It may also be that different animals release different amounts of pheromone, so that a threshold concentration is not surpassed, or that the receptivity of the receivers may vary. The limited numbers of experimental animals and, to some extent, donor animals would make the present study vulnerable to such variability.
We did see a tendency for an effect of type of exposure on the preovulatory LH surge, as suggested by Aron [11
]. However, this tendency concerned the peak LH concentration and sum of LH concentrations during the surge, with higher values for the control cycles, and not the timing of the surge. In a previous study we investigated the effect of oestrous urine and vaginal discharge on the LH pulsatility pattern preceding the preovulatory surge and found that the pattern differed significantly between the two types of exposure, with increased nadir concentrations and decreased amplitude of peaks during treatment cycles [35
]. It is possible that there really is an inhibiting effect of treatment on the two LH parameters, but that the number of animals in this study is too limited for this difference to manifest significantly. Moreover, the large variation in several of the hormonal and reproductive variables might, to some extent, also explain the lack of differences between the two types of exposure.
The effect of cycle within type of exposure for maximum oestradiol concentration is interesting, indicating that control cycles are more homogenous and that the variability of this parameter is greater during the treatment cycle.
Even though neither the duration nor the strength of expression of oestrus (measured as the total score) differed between the two types of exposure, we found a significant interaction between time period and type of exposure for the variable restlessness. During the control cycles increased restlessness occurred mainly during the period of standing oestrus (14-26 h before ovulation), while it occurred earlier (26-38 h before ovulation) during the treatment cycles. There were also indications that the same may apply to the variable position. The score for genitalia showed the opposite pattern with similarly high scores during the first two periods for the control cycles and the maximum score for the treatment cycles occurring during the period of standing oestrus. These effects might be caused by endocrinological changes due to the exposure, even though we could not detect such differences in our data.
The score for genitalia and discharge increased between control cycles 1 and 2, while it decreased between the treatment cycles. This cycle effect might be caused by the long time in isolation. Such an effect could have been avoided by randomising the order in which the two types of exposure occurred. However, due to the unexplored nature of bovine primer pheromones the present schedule was chosen to avoid a possible spillover effect from periods of treatment to periods of control and also to avoid contamination when handling the substances simultaneously. It may also be that the treatment has an additional or delayed effect, so that the effect can't be seen until the last cycle of the treatment.