Mothers of large experimental litters abandoned their suckling pups more often than mothers of small experimental litters when separation calls of another pup were broadcast. Moreover, mothers of small litters decreased their responsiveness to pup calls from day 8 to day 20, whereas those of large litters remained strongly responsive. Mothers of small experimental litters also weaned the pups earlier than did mothers of large experimental litters. Thus, guinea pig mothers adjusted their level and time period of responsiveness to experimental litter size.
Earlier experiments had shown that guinea pig females responded, apart from a general decrease of nursing activity over time , little if at all, to changes in pup demand by increasing milk yield  or by adjusting nursing performance [14, 16]. These experiments suggested that females pay surprisingly little attention to pup demand or pup state. However, these studies described nursing behaviour and milk yield and did not observe maternal behavioural responsiveness depending on litter size as our playback experiments did. Thus the playback experiments complement previous studies by showing that females indeed adjust behavioural responsiveness to litter size, even if the litter size they are rearing is not the one they had produced. These results also fit with earlier findings that larger litters are weaned later [15, 16] and suggest that maternal motivation is increased through stimuli provided by larger litters. Previous studies on other rodents also had shown that maternal responsiveness depends on litter size. Maternal nest attendance decreased with increasing litter size in golden hamsters (Mesocricetus auratus)  and rats (Rattus norvegicus)  which may be due to the increased temperature in the nest or disturbances caused by the activity of many pups. Mongolian gerbil mothers (Meriones unguiculatus) with larger litters spent less time in the nest, but licked and sniffed more than mothers with small litters . In contrast to these measures of maternal care, our playback experiments provide evidence for changes in maternal responsiveness to pup separation calls in relation to litter size.
The pup separation calls we tested, differ functionally from begging calls tested in other species. They are emitted only when pups are out of contact with their mother and are not given before and during suckling interactions. Playback experiments on pigs (Sus scrofa domestica) demonstrated a stronger response of mothers to needy piglets . In these experiments, the smallest and slowest growing young in the litter which had just missed a nursing and were isolated in a relatively cool enclosure, called most intensely. Similarly, in birds a positive relationship between begging intensity and parental feeding rate could be found [4, 28–31]. Great tit (Parus major) nestlings showed an increase in mean begging rates from experimentally reduced to enlarged broods. Their parents adjusted the feeding rates that were similar per nestling over three brood size .
Also, it has been suggested that piglet calls function mainly as a signal to the sow by piglets that are excluded from the current nursing episode . Even though our experiments could be interpreted similarly, this function does not seem to apply to the separation calls as used in guinea pigs. Fey and Trillmich  never observed pups to utter separation calls when they were near their mother but had temporarily no access to the teats in litters of four. These pups would rather dig in under the mother, presumably to wait for an opportunity to access a teat. For female guinea pigs, finding lost pups may play a major role in protection of those pups and may also be important in thermoregulation, particularly for young pups that have only limited energy reserves to maintain thermoregulation. As energy input via milk plays only a minor role late in lactation , the benefits of lengthening the lactation period may rather be of social function or to support pups' thermoregulation.
Causally, differences in maternal responsiveness might have been linked to differences in hormonal state, as suggested by correlative studies in humans. In humans, maternal approach behaviour was directly associated with levels of cortisol and multiple regressions revealed that the infants' vocal behavior significantly predicted maternal level of cortisol . Mothers showing the highest levels of maternal approach responses were those with a high cortisol concentration and either a positive maternal attitude, or a vocally more active infant. Based on these data, they speculate that for the new mother to exhibit a high level of responsiveness to her infant, she must attain a certain level of arousal , which can be produced by elevated cortisol . Moreover, Fey & Trillmich  showed that maternal cortisol levels in guinea pigs decreased as pups grew older, and mothers rearing a litter of four pups maintained, although not significantly so, higher cortisol levels than those with litters of two pups. Thus, auditory stimuli may affect maternal responsiveness via the general adaptive functions of arousal and evocation of maternal behavior.
As predicted, guinea pig mothers with large litters actively interrupted nursing and responded to pup separation calls by approaching the loudspeaker whereas mothers of small litters most often did not. This indicates that mothers pay attention to litter size and do not respond when calling pups cannot be their own. We previously showed that females can recognize the calls of their own offspring  but nevertheless are responsive to calls of unfamiliar pups. This suggests that the costs of such false alarms are lower than the costs of missed detections . Such a pattern of response is not unusual as also in ungulate hider species a similar unspecific response of mothers to separation calls has been reported . In these species, females do not know the exact hiding location of their offspring and use vocalizations to reunite with the fawns [38, 39]. Mule deer (Odocoileus hemionus) and white-tailed deer (Odocoileus virginianus) mothers defend their young vigorously against predators. The fawns' calls activated the mothers and in mule deer, females even responded to separation calls of white tailed deer calves [17, 37]. These authors argue that it may pay more to mistakenly defend a foreign young than to lose the own fawn. However, as females also approached the loudspeaker in an aggressive manner when they were together with their own fawn, factors other than the separation from their fawn must also affect responsiveness to playback by female deer. In contrast to larger mammals, in guinea pigs and their wild ancestors, females have little possibility to actively defend their offspring, as mammalian and bird predators are able to kill the mother as well as the offspring. Therefore, the response by guinea pig mothers is likely to function to reunite with young gone astray rather than in directly defending them. Mothers with larger offspring numbers then presumably respond stronger than those with fewer offspring as it is more likely to lose young in larger litters. Reuniting with these young quickly and thereby preventing them from calling is likely to reduce the danger of attracting the attention of predators.