Individual stability in vocalization rates of preweaning piglets

Abstract

Piglet vocalization rates are used as welfare indicators. The emission rates of the two gross categories of piglet calls, namely low frequency calls ("grunts") and high frequency calls ("screams"), may contain different information about the piglet's internal state due to differing communicative functions of the two call types. More knowledge is needed about the sources of variation in calling rates within and between piglets. We examined to what extent the emission rates of the two call types are codetermined by individual and litter identity, i.e., whether the rates are repeatable within individuals and similar between littermates. We recorded frequency of grunts and screams in one mildly negative (short-term Isolation) and one moderately negative (manual Restraint) situation during the first week (week 1) and the 4th week (week 4) of life and asked the following questions: 1) Are within-individual vocalization rates stable across the suckling period? 2) Are within-individual vocalization rates stable across the two situations? 3) Is there within-litter similarity in vocalization rates? 4) Does this within-litter similarity increase during the suckling period? Within-individual vocalization rates were stable between week 1 and week 4 (grunts in Restraint P < 0.05; grunts in Isolation P < 0.001; screams in Restraint P < 0.001; screams in Isolation P < 0.001). Across the two situations at the same age, the vocalization rates were not stable for grunts but were stable for screams at week 1 and week 4 (P < 0.05). Vocalization rates were more similar between littermates than between piglets belonging to different litters (grunts in Restraint P < 0.001; grunts in Isolation P < 0.01; screams in Restraint P < 0.001; screams in Isolation P < 0.001). This litter effect did not grow stronger from week 1 to week 4 as the within-litter coefficient of variance did not decrease between the two ages. Sex of the piglet had no influence on vocalization rates while greater body weight was associated with lower screaming rates in the Restraint situation (P < 0.05). In conclusion, our study demonstrates that both individuality of the piglet and litter identity affect the vocalization rates of piglets in negatively valenced situations. For screams, the repeatability of individual vocalization rates holds even across situations, while for grunts, the rates are repeatable during ontogeny within the situations, but not across situations.

Figure 1.

Relationships between individual vocalization rates at week 1 and week 4 for grunts in Restraint (a), grunts in Isolation (b), screams in Restraint (c), and screams in Isolation (d). Linear regression lines were added to illustrate trends (a–c). For screams in Isolation (d), the heights of the bars represent the number of piglets that screamed in week 1 (Yes—left bar, No—right bar), while colors represent how many piglets screamed in week 4 (Yes—gray, No—white). For instance, there were five piglets that did not scream in week 1 but screamed in week 4.

Figure 2.

Relationships between vocalization rates in Isolation and in Restraint for grunts at week 1 (a), grunts at week 4 (b), screams at week 1 (c), and screams at week 4 (d). Linear regression lines were added to illustrate trends (a, b). Means and 95% confidence intervals are presented (c, d).

Figure 3.

Effect of age on vocalization rates (numbers of calls, proportion of piglets emitting screams) of piglets during Restraint (a, c) and Isolation (b, d). Means and 95% confidence intervals (a–c) and proportion of piglets emitting scream and 95% confidence intervals (d) are presented.

Figure 4.

Illustration of the effect of weight on screaming rate during Restraint situation. Weight residuals are residuals from linear regression of common logarithm of weight on age. Data for week 1 and week 4 are plotted together. A simple linear regression line is added to illustrate the overall trend (y = −28.7x + 33.6).