Conceived and designed the experiments: BNG CJG EAP JRL JKD JPG. Performed the experiments: BNG. Analyzed the data: BNG JPG. Contributed reagents/materials/analysis tools: JPG. Wrote the paper: BNG.
All nesting material used in this experiment were donated by FiberCore, an environmental enrichment company. All mice were also donated by Charles River Laboratories. BNG at the time of conducting, analyzing, and writing this manuscript was a student at Purdue University. Upon graduation (Aug. 2011) she was hired as a postdoctoral researcher at Charles River. Charles River had no input on the interpretation of the results. All authors had full control of all primary data and its interpretation. These competing interests do not alter the authors’ adherence to all PLoS ONE policies on sharing data and materials.
In laboratories, mice are housed at 20–24°C, which is below their lower critical temperature (≈30°C). This increased thermal stress has the potential to alter scientific outcomes. Nesting material should allow for improved behavioral thermoregulation and thus alleviate this thermal stress. Nesting behavior should change with temperature and material, and the choice between nesting or thermotaxis (movement in response to temperature) should also depend on the balance of these factors, such that mice titrate nesting material against temperature. Naïve CD-1, BALB/c, and C57BL/6 mice (36 male and 36 female/strain in groups of 3) were housed in a set of 2 connected cages, each maintained at a different temperature using a water bath. One cage in each set was 20°C (Nesting cage; NC) while the other was one of 6 temperatures (Temperature cage; TC: 20, 23, 26, 29, 32, or 35°C). The NC contained one of 6 nesting provisions (0, 2, 4, 6, 8, or 10g), changed daily. Food intake and nest scores were measured in both cages. As the difference in temperature between paired cages increased, feed consumption in NC increased. Nesting provision altered differences in nest scores between the 2 paired temperatures. Nest scores in NC increased with increasing provision. In addition, temperature pairings altered the difference in nest scores with the smallest difference between locations at 26°C and 29°C. Mice transferred material from NC to TC but the likelihood of transfer decreased with increasing provision. Overall, mice of different strains and sexes prefer temperatures between 26–29°C and the shift from thermotaxis to nest building is seen between 6 and 10 g of material. Our results suggest that under normal laboratory temperatures, mice should be provided with no less than 6 grams of nesting material, but up to 10 grams may be needed to alleviate thermal distress under typical temperatures.
The
In the wild, mice cope with temperature extremes by building nests
Ethologists and welfare scientists are often interested in investigating what resources or aspects of the environment are important to captive animals and preference testing can be used as a first step to identifying how an animal perceives the world around it
The goal of this project was to use the behavioral titration technique to determine how much nesting material is needed to alleviate potential thermal discomfort when mice are housed over a range of ambient temperatures. We hypothesized that location preference, between a warm and cool condition, should change with temperature and amount of material. We predicted that increasing amounts of nesting material would increase nest scores and that nest scores would decrease when mice had access to a warmer ambient temperature. We predicted that mice would spend more time, overall, in temperatures near their lower critical temperature (around 30°C) but this temperature preference would vary depending on the amount of nesting material provided. Previous studies show that ambient temperatures near 30°C are especially preferred when inactive
Materials and methods were adapted in part from Gaskill et al.
Seventy-two mice from each strain (C57BL/6NCrl; BALB/cAnNCrl; Crl:CD1) arrived at Purdue University, USA from Charles River Laboratories (Wilmington, MA, USA). These three types of mice were chosen because they comprise the most commonly used inbred (C57BL/6NCrl; BALB/cAnNCrl) and outbred (Crl:CD1) research mice. This selection will allow our results to be applicable to the vast majority of the research mouse population. A large difference in body size exists between BALB/c and CD-1 mice at similar ages. Since heat loss is related to the surface area to body weight ratio
Two 5 gallon glass fish tanks (
(a) Diagram showing the configuration of water baths and cages for testing cage temperature and nesting material preferences. (b) Diagram depicting elements present in water bath and cage setup. The figures are reproduced with permission from Elsevier
Males and females were tested in alternating weeks, thus the experiment required 12 weeks to complete two replicates. We took precautions to control for position bias and the potential effect of mice in adjacent cages by using visual barriers between cages and by rotating the temperature of the cages each week.
A testing session took 6 days to complete. The day before testing began, mice were trained to use the plastic tubes to transfer back and forth between the two connected cages. On the first day of testing, mice were placed in each cage of the assigned temperature-set twice, alternating every 10 minutes to make sure that the animals experienced both environments and would use the tubes. The mice were all placed in one cage (TC or NC) to begin the day, balanced across the 6 days. The NC contained one of 6 provisions of nesting material (Enviro-dri®; FiberCore, Cleveland, OH, USA): 0, 2, 4, 6, 8, or 10g, changed daily in a balanced design. After each 24 hour period, all nesting material was removed and a new amount was added to NC, so that each group of mice had access to each of the nesting treatments over the course of the testing session. The order of treatment was randomized as a latin square design. Enviro-dri® was chosen as the nesting treatment because it closely resembles materials used in the wild and C57BL/6 mice build better nests with this material than with other options
The mice were videotaped continuously over the 6 days for behavioral data collection using infrared cameras and illuminators, digital video recorder and video surveillance software (Inter-Pacific, Wheeling, IL, USA). The location (NC, TC, or Tube) and behavior (Active, Inactive, Maintenance, Nesting, Unknown-in-nest, and Unknown;
Category | Behavior | Description |
Active | General locomotion | All locomotive behavior performed on the cage lid, climbing up the cage bars by the food hopper to reach the lid, and locomotion on the floor of the cage. |
Rearing | Seen on the floor of the cage with all an animal’s weight on its hind legs and front legs off the ground. Sniffing movements while on its hind legs were commonly accompanied with this behavior. | |
Sniffing | Sniffing was also performed against the cage floor (ground), or in between the bars of the cage lid. Slight upward jerks of the head were seen. | |
Maintenance | Grooming | All grooming behavior including licking the fur, grooming with the forepaws, and scratching with any limb. Grooming was usually performed in a sitting position with the animal’s hind quarters in contact with the floor. |
Feeding or drinking | The animal would rear up to gnaw at food pellets through the bars of the hopper. The forepaws would usually be used to hold the food pellet steady. The animal would rear up and lick the nipple drinker. | |
Inactive | Sleeping | The animal was motionless, and either lying curled up on its side, or sitting curled up, with its face tucked into its body and out of sight of the camera. Occasionally interrupted by brief single twitches of the body. |
Still and alert | The animal was sitting or curled up, but in contrast to sleep, the face was lifted. The animal either sat motionless, or would appear to be orientating its head to sounds outside of the cage. | |
Inactive in nest | The animal within the nest, due to camera angles, cannot clearly be seen but no movement within the nest can be detected. It is assumed that the animal is sleeping within the nest. This is distinguishable from other behaviors within the nest because movement within the nest or of the nest itself is not observed. | |
Nesting | Pull in | Characterized by the animal reaching out of the nest and pulling sawdust or nesting material to the edge of the nest. The animal may also grasp the material in its mouth and drag it into the nest site. Gathering is distinct from locomotion in that the hind legs do not leave the nest site, and each time the animal reaches out of the nest it pulls its forelegs back in. |
Carrying | Locomotion with material, such as large pieces of bedding or nesting material in the mouth. | |
Fraying | The animal uses sideways movement of the forepaws to draw material through the beak. Gnawing movements of the jaw and jerking movements with the head are also seen. As a result the edges of the nesting material are bitten off or large pieces of bedding are split into smaller fibers. | |
Push-Dig | The forward pushing and kicking of substrate material with fast alternating movements with the forepaws often combined with forward locomotion. | |
Sorting | The deliberate action of placing specific nesting material strips or bedding material into a particular location while sitting within the nest site. | |
Digging | Removing, or apparently trying to remove, substrate material from a certain place by series of fast alternating movement of the forepaws, as a consequence of which the material heaps up under the abdomen of the animal. | |
Scrape-dig | The series of forepaw movements are alternated by a few hindwards kicking movements of both hind legs simultaneously, through which the heap under the abdomen of the animal is transported further backwards. | |
Fluffing | An unseen nesting behavior, due to insufficient camera angles or view from inside the nest, which results in the enlargement of the nest from the inside. Walls of the nest will appear to jump and the nest as a whole will enlarge. It is assumed that the animal is hollowing out the inside of the nest by pushing the walls back and up. | |
Unknown in Nest | Unknown | An animal is inside of the nest but unsure of the behavior being occurring inside of the nest. This is different from Fluffing in that the nest does not appear to be growing or occurs out of the sequence on nest building. This is also different from Inactive in Nest in that movement is seen within the nest. |
Unknown | Unknown | An animal’s behavior cannot be determined or the view of the animal is blocked while in or outside of the nest. |
Food consumption was measured before and after each 6 day testing session from both adjoined cages.
Nest scores were recorded daily from both NC and TC at the end of the 24 hour test period, before nesting treatments were changed. Nest scores were recorded from both cages because mice will attempt to build a simple nest out of bedding material when other substrate is not provided. To compare nest scores between NC and TC, the nest score from TC was subtracted from NC to get the difference in nest score between the two cages.
Population time budgets were calculated for each group of mice by counting the total number of times each category of behavior was observed in each location (i.e. NC, TC, and the tube) for each day and dividing this count by the total number of observations for that group. Following this calculation, data from the tube and unknown behaviors were excluded from the analysis. The percent observations from NC (plus the smallest observation in order to avoid zero values) were divided by TC (again plus the smallest observation). The log of this value was taken in order to normalize the ratio of observations in NC relative to TC.
Behavior, nest score, and food consumption analyses were performed as split-plot ANOVA using GLM, in JMP 6 for Windows. The assumptions of GLM (normality of error, homogeneity of variance, and linearity) were confirmed post-hoc, and appropriate transformations were made to meet these assumptions
To avoid pseudoreplication and accommodate repeated measures, analyses were blocked by Group of mice, nested within Strain, Sex, and Temperature-Set. Group of mice cannot be treated as a random effect (there is not a meaningful wider population of groups of three mice representing unique and indivisible components of variance from which we selected our groups of three mice, and to which our results could pertain)
The difference in bodyweight before and after the experiment was documented but no statistical differences due to temperature or nesting material were seen. The average body weight by each type of mouse at the beginning of the experiment was as follows: C57BL/6 Females = 20.6g; Males = 27.1g; BALB/c Females = 19.8g; Males = 27.7g; CD-1 Females = 25.5g; Males = 32.6g. Bodyweight at the end of the experiment was: C57BL/6 Females = 20.7g; Males = 27.1g; BALB/c Females = 20.3g; Males = 27.8g; CD-1 Females = 25.6g; Males = 34.4g.
We first predicted that cage sets with warmer temperatures would result in a reduction in the amount of food consumed. The overall amount of food consumed was not significantly altered in any of the temperature-sets (GLM: F5,71 = 0.43; P = 0.82). However, there was a significant interaction between temperature-set and the location (TC or NC) where they consumed the food (GLM: F5,71 = 2.91; P = 0.019;
Strain and Temperature-Set was found to alter nest scores (GLM: F10,308 = 2.76; P = 0.003). In the 20–20°C temperature-set, C57BL/6 and CD-1 mice built better nests in NC (t α/18; P<0.05), but no significant differences in nest building between the two locations was found for BALB/c mice (t α/18; P>0.05). No differences in nest building were found in the 23, 26, or 29°C temperature-sets for any of the strains (t α/18; P>0.05). BALB/c and CD-1 mice built significantly better nests in NC at 32°C (t α/18; P<0.05), but this pattern was not shown in C57BL/6 mice (t α/18; P>0.05). In the warmest temperature-set, 35°C, all the stains built significantly better nests in NC (t α/18; P<0.05).
Nest quality was altered by interactions between Strain and Sex (GLM: F2,308 = 6.76; P = 0.001). Female BALB/c mice built significantly better nests in NC (t α/6; P<0.05), but the other two strains showed no building difference between the two locations (t α/6; P>0.05). No differences in nest building for females were found between the strains (Tukey: P>0.05). Male C57BL/6s and CD-1s built better nests in NC (t α/6; P<0.05) but the BALB/c mice showed no differences in location (t α/6; P>0.05). Male CD-1s built significantly better nests in NC than BALB/c males (Tukey: P<0.05), but BALB/c and C57BL/6 male’s building was not significantly different from one another (Tukey: P>0.05). CD-1 and C57BL/6 males built significantly better nests in NC compared to females of their respective strain (Tukey: P<0.05). However, no significant differences were found between male and female BALB/c mice (Tukey: P>0.05).
Nest quality was also affected by temperature (GLM: F5,308 = 12.6; P<0.001), but nest scores changed when the mice transferred nesting material (GLM: F5,308 = 6.6; P<0.001;
Nest scores partitioned by occurrences of nesting material carryover by (a) cage sets and (b) amount of material provided. A negative value indicates a better nest built in the temperature cage and a positive value indicates a better nest in the nesting cage. LSM and SE are plotted and significant t-tests (value different from zero; α corrected for the number of comparisons) are indicated by asterisks. A diagonal line indicates a significant linear trend and a curved line indicates a significant quadratic trend.
As predicted a significant main effect of nesting material amount on nest quality was found (GLM: F5,308 = 7.53; P<0.001). However, if mice transferred nesting material from NC to TC this significantly altered the difference in nest quality at different amounts (GLM: F5,308 = 3.9; P = 0.002;
The sexes also showed a disparity in building location when material was transferred (GLM: F2,308 = 8.07; P = 0.005;
The three strains also showed differences in nest building when material was transferred (GLM: F2,308 = 12.6; P<0.001;
The transfer of nesting material from the NC to TC was an unexpected observation in this experiment. There was a significant Sex effect: females were more likely to transfer material than males (LR χ2 = 56.4; P<0.001;
Data is plotted by (a) sex and temperature; (b) strain and temperature and; (c) amount of nesting material and strain. Quadratic peaks are indicated by solid vertical lines.
Some unpredicted main effects were found based on where the mouse strains spent their time (GLM: F2,2425 = 78.41; P<0.001). BALB/c mice spent more time in NC than the other two strains (Tukey: P<0.05). While CD-1s still spent the majority of their time in NC, this amount of time was significantly less than the BALB/c mice (Tukey: P<0.05). C57BL/6 mice were the only strain to spend the majority of their time in TC (Tukey: P<0.05). The sexes also showed differences in their location preferences (GLM: F1,2425 = 120.4; P<0.001). Overall males spent more time in NC while females spent more time in the TC.
We predicted that the temperature a cool cage was paired with would affect the preference for nesting material. As predicted Temperature-Set affected preference but depended on Sex (GLM: F5,2425 = 25.6; P<0.001;
Fold difference in percent of location observations between the nesting cage relative to the temperature cage. Effects of temperature-set are plotted by interactions with (a) sex; (b) strain and; (c) behavior. LSM and SE are plotted and significant t-tests (value different from zero-α corrected for the number of comparisons) are indicated by asterisks.
Preference differences were also seen between the different strains (GLM: F10,2425 = 13.1; P<0.001;
Behavior was also altered based on temperature-set (GLM: F20,2425 = 18.2; P<0.001;
A significant interaction between Temperature-set and Amount of nesting material was found (GLM: F25,2425 = 2.54; P<0.001;
Mean difference in percent of observations between the nesting cage and the temperature cage for the temperature-set by amount of nesting material interaction. The green area indicates equal preference for NC and TC. Blue and purple shading indicate a 2 and 4 fold preference for NC. Orange and red shading indicate 2 and 4 fold preferences for TC.
As predicted, a significant interaction between the amount of nesting material provided and sex was found (GLM: F5,2425 = 5.95; P = 0.019;
Fold difference in percent of observations between the nesting cage relative to the temperature cage. Effects of the amount of nesting material provided are plotted by interactions with (a) sex; (b) strain and; (c) behavior. LSM and SE are plotted and significant t-tests (value different from zero-α corrected for the number of comparisons) are indicated by asterisks.
Where mice preferred to spend their time was also significantly affected by amount and strain (GLM: F10,2425 = 2.77; P = 0.002;
Differences in behavior were also seen depending on the amount of material provided (GLM: F20,2425 = 8.2; P<0.001;
Other behavioral observation differences were affected by the main effect of sex (GLM: F4,2425 = 36.5; P<0.001;
Strain was also found to affect the location of behavior (GLM: F8,2425 = 16.7; P<0.001;
This experiment shows for the first time the preference tradeoff between temperature and nesting material, based on the combination of the two factors. The knowledge of this tradeoff is extremely important because not all laboratory temperatures are identical, and therefore it is unknown how much material is needed to eliminate mouse thermal discomfort under various conditions. Furthermore, nesting material is increasingly being implemented in the United States and is considered a standard husbandry item in Europe
The effect of warmer or cooler temperatures on food consumption has been documented in humans
Our mice showed the expected nest building responses to both temperature and amount of nesting material, when material was found solely in NC. However, when the mice transferred the material, better nests were no longer consistently built in NC. This decision to carry over nesting material from one cage to another was a surprising result, as a similar experiment by Gaskill et al
CD-1 mice on the other hand, employ a different strategy than the other two strains. They appear to tradeoff between nest building and thermotaxis based on the amount of nesting material provided or temperature. However, the provision of nesting material seems to be the main factor they are basing this decision on. When material provision is low, they show high motivation to transfer material and combine it with temperature, or use simple thermotaxis. They do not begin to spend significantly more time in NC until 6 grams or more of nesting material was provided. Most likely this is the smallest amount of material that can be used to build a suitably insulating nest. Consequently, the motivation to transfer material also declines as the provision of material increases. The probability of material transfer for the CD-1 mice ranges from a level similar to a strain utilizing thermotaxis (C57BL/6) when there is a small amount of material, to a level similar to the nest building strain (BALB/c) as nesting provision increases. While CD-1 mice may switch strategies, they seem to favor nest building over thermotaxis.
It is possible that the reason CD-1 mice employ a strategy different from the other two strains is due to their young age or exposure to enrichment at an earlier age. Thermal preference studies have found differences in temperature selection due to age
If the probability of material transfer can be used to indicate the primary mode of behavioral thermoregulation, then it is interesting that this behavior was seen in all combinations of mice and environmental variables. It appears that mice retain some underlying motivation for nesting material, even if their thermal needs are met. It is likely the drive to build a nest may serve a purpose other than thermoregulation
Previous preference work points to slight differences in thermal preference between the sexes as well as differences in nest shape as temperatures increase
Although we have shown thermal disparity between the sexes, it appears that 20°C is a universally cool temperature for both sexes and all strains
The behavioral ethogram covered 6 categories of behavior but only a few showed any significant differences due to our treatments. Previous studies suggest that temperature or nesting material may be more essential to mice while inactive
A practical question, especially from an economic standpoint, is how much material is needed to alleviate any thermal distress and how much does that amount change under standard laboratory temperatures. Other experiments have investigated the amount of material collected by mice
Our location preference results, which do not incorporate specific behaviors, may have been slightly lowered based on the way that the data was processed. Location means were averaged over every behavioral category in our ethogram. Since mice prefer different temperatures at different parts of the day as well as for different behaviors, this analysis controls for the particular behaviors that drive these preferences, such as inactive behavior. Therefore more frequent behaviors are weighted equally with other less frequently observed behaviors, such as nest building. Thus, location preference values are the mean location preference for all behaviors.
(TIF)
(TIF)
(TIF)
A special thank you to Charles River Laboratories for donating all of the mice used in this study as well as Fiber Core for donating the nesting material. We would also like to thank Kat Rodda for assisting with running the experiment and video observations and Jason Fields for animal care.