Huddling can be defined as "an active and close aggregation of animals". It is a cooperative group behaviour, permitting individuals involved in social thermoregulation to minimize heat loss and thereby lower their energy expenditure, and possibly allowing them to reallocate the saved energy to other functions such as growth or reproduction. Huddling is especially important in the case of animals faced with high heat loss due to a high surface-to-volume ratio, poor insulation, or living in cold environments. Although numerous experimental studies have focused on the huddling behaviour of a wide range of species, to our knowledge, this is the first attempt to review the various implications of this widely used behavioural strategy. Huddling allows individuals to maximise energy savings by (1) decreasing their cold-exposed body surface area, (2) reducing their heat loss through warming of ambient temperatures surrounding the group, and (3) eventually lowering their body temperature through physiological processes. Huddling provides substantial energy savings and is estimated to reduce energy expenditure by between 6 and 53%. Broad variations in the energetic benefits of huddling depend on the number of individuals and species involved in huddles, the ambient temperatures to which individuals are exposed and the density of the aggregations. It has been shown that huddling individuals have increased survival, a lower food intake, a decreased body mass loss, increased growth rate, reduced water loss, and/or a more constant body temperature together with a significant reduction in metabolic rate. Though huddling has been studied widely, this review reveals the intricacies of this adaptive behaviour.

Acute stress triggers peripheral vasoconstriction, causing a rapid, short-term drop in skin temperature in homeotherms. We tested, for the first time, whether this response has the potential to quantify stress, by exhibiting proportionality with stressor intensity. We used established behavioural and hormonal markers: activity level and corticosterone level, to validate a mild and more severe form of an acute restraint stressor in hens (Gallus gallus domesticus). We then used infrared thermography (IRT) to non-invasively collect continuous temperature measurements following exposure to these two intensities of acute handling stress. In the comb and wattle, two skin regions with a known thermoregulatory role, stressor intensity predicted the extent of initial skin cooling, and also the occurrence of a more delayed skin warming, providing two opportunities to quantify stress. With the present, cost-effective availability of IRT technology, this non-invasive and continuous method of stress assessment in unrestrained animals has the potential to become common practice in pure and applied research.

1. Infrared thermography (IRT) involves the precise measurement of infrared radiation which allows surface temperature to be determined according to simple physical laws. This review describes previous applications of IRT in studies of thermal physiology, veterinary diagnosis of disease or injury and population surveys on domestic and wild mammals. 2. IRT is a useful technique because it is non-invasive and measurements can be made at distances of <1 m to examine specific sites of heat loss to >1000 m to count large mammals. Detailed measurements of surface temperature variation can be made where large numbers of temperature sensors would otherwise be required and where conventional solid sensors can give false readings on mammal coats. Studies need to take into account sources of error due to variation in emissivity, evaporative cooling and radiative heating of the coat. 3. Recent advances in thermal imaging technology have produced lightweight, portable systems that store digital images with high temperature and spatial resolution. For these reasons, there are many further opportunities for IRT in studies of captive and wild mammals.

This study examined the foraging locations of adult male and female Antarctic fur seals (Arctocephalus gazella) in the Scotia Sea during the postbreeding period. Satellite transmitters were used to track adult males and females and to obtain information about dive depths. Male fur seals migrated away from the breeding area during the postbreeding period whereas females remained close to the breeding grounds and foraged in the same area during two consecutive years. The most intensive foraging by females was associated with the edge of the continental shelf of South Georgia. Males dived deeper than females. Counts of males at South Georgia and at the South Orkney Islands support the result from satellite tracking data showing that males move from South Georgia to the South Orkney Islands at the end of the breeding season. Unlike males, females were limited in their foraging range by the necessity to return to feed dependent young, so breeding sites are likely to be located close to foraging areas that are optimal for females. Locations used for feeding by females were avoided by males, either because they were suboptimal for males or because foraging by females at South Georgia causes local depletion of food, and males, which have the option to forage further afield, can forage more successfully in regions where there are no females. Comparison with fisheries data also suggests that these fur seals are targeting the most abundant exploitable prey.

Stress, a central concept in biology, describes a suite of emergency responses to challenges. Among other responses, stress leads to a change in blood flow that results in a net influx of blood to key organs and an increase in core temperature. This stress-induced hyperthermia is used to assess stress. However, measuring core temperature is invasive. As blood flow is redirected to the core, the periphery of the body can cool. This paper describes a protocol where peripheral body temperature is measured non-invasively in wild blue tits (Cyanistes caeruleus) using infrared thermography. In the field we created a set-up bringing the birds to an ideal position in front of the camera by using a baited box. The camera takes a short thermal video recording of the undisturbed bird before applying a mild stressor (closing the box and therefore capturing the bird), and the bird's response to being trapped is recorded. The bare skin of the eye-region is the warmest area in the image. This allows an automated extraction of the maximum eye-region temperature from each image frame, followed by further steps of manual data filtering removing the most common sources of errors (motion blur, blinking). This protocol provides a time series of eye-region temperature with a fine temporal resolution that allows us to study the dynamics of the stress response non-invasively. Further work needs to demonstrate the usefulness of the method to assess stress, for instance to investigate whether eye-region temperature response is proportional to the strength of the stressor. If this can be confirmed, it will provide a valuable alternative method of stress assessment in animals and will be useful to a wide range of researchers from ecologists, conservation biologists, physiologists to animal welfare researchers.