Inﬂuence of Cooling and Heating Systems on Pen Fouling, Lying Behavior, and Performance of Rearing Piglets

: The rearing of piglets is a demanding phase of pig production partly because of the changing temperature requirements of the piglets during rearing. Piglets need high temperatures in the resting area, especially at the beginning, while the optimal temperature is lower toward the end of rearing. To meet the changing temperature demands of the piglets and also to optimize the pen structure, one ﬂoor cooling and two heating systems were examined in this study. In two rearing compartments, four pens with 48 piglets each were equipped with a heated cover above a heated/cooled lying area. The lying behavior and performance of 1152 piglets, in addition to the pen fouling, were recorded over six rearing periods. There was no difference between the two heating systems in the lying behavior. However, the acceptance of the lying area was very high in all pens and periods with heating. The cooling had a signiﬁcant inﬂuence on the lying behavior depending on the rearing week. Signiﬁcantly more piglets lay on the cooled lying area compared with the control pen during the last weeks. The fouling of the pens was not affected by the cooling or heating systems; however, the fouling in all pens was very low. The tested pen structure in combination with a heating and cooling system is a well-functioning way of integrating a solid lying area. data curation, S.O. and B.K.; writing—original draft preparation, S.O.; writing—review and editing, S.O., H.-P.P., and E.G.; visualization, S.O.; supervision, E.G.; project administration, E.G. and H.S.; funding acquisition, E.G. and H.S. All authors have read and agreed to the version of the manuscript.


Introduction
The rearing of piglets is a demanding phase of pig production because of their high stress levels and changing needs during these weeks. The piglets are exposed to massive stress directly after weaning by being separated from the sow, moving into another housing system, and changing from milk to a solid diet [1]. The housing environment should be optimal to reduce this stress and the resulting negative effects on their performance and health. Many different factors influence the piglets, but one of the most important is the ambient temperature. The optimal temperature changes for the piglets during the rearing because of the increasing age and liveweight [2]. In the beginning, the weaned piglets need higher temperatures, while nearing the end, the optimum temperature is lower and it could become too warm, especially in the summer. The optimal air temperature for rearing piglets between 10 and 30 kg liveweight is 20-30 • C [3]. The optimum temperature also depends on the type of floor (solid or slatted) and whether there is straw in the lying area. Piglets need higher temperatures, especially for resting and sleeping areas, to avoid poor performance or poor health. On the other hand, cool temperatures in the feeding and activity areas can improve the performance of the piglets [4][5][6]. It is assumed that reaching the thermal comfort zone of the piglets, especially in the resting area, increases animal welfare.
The heating or cooling of a whole building or room is high in cost and not quickly adaptable to outside temperatures or other factors. Therefore, a good alternative is the use of zone heating or cooling to create an optimal microclimate within the pen. These systems have the advantages of reducing the energy cost for heating and guiding the piglets in the division of the functional areas [1,7]. Different systems could be used to create such a microenvironment, such as heated covers, floors, or walls [7].
In the past, the focus in piglet rearing was on optimal heating, but due to climate change, cooling has become more important. The animals in Germany are more frequently exposed to heat stress because of extremely hot temperatures over 40 • C and rising annual temperatures during the summer months [8]. Heat stress in fattening pigs and sows can cause inconsistent and reduced growth, decreased carcass quality, and increased morbidity, mortality, and processing issues [9,10]. In addition, high ambient temperatures can lead to increased fouling of the pens [11,12]. This can lead to an increased workload since the surfaces are usually cleaned by hand, and furthermore, fouling also negatively affects the air quality and thus the animal welfare. Pigs lie on cool surfaces, wet their skin, or wallow to lower or regulate their body temperature due to their inability to sweat [13][14][15]. Cooling of the surface can reduce pen fouling, heat stress, and wallowing in fattening pigs and sows, while increasing feed intake and the acceptance of a solid lying area for resting [16][17][18][19][20]. Therefore, cooling the surface in piglet rearing systems could reduce heat stress and its negative effects.
Two heating systems (heated cover and heated floor) and a floor cooling system were tested in four rearing pens in the present study. All pens were equipped with a solid lying area that could be heated or cooled by water, and a cover over this lying area that could be heated. Whether the two heating systems caused differences in the behavior or the performance of the rearing piglets was investigated during the cold periods (mid-October to March). In addition, the influence of a cooled solid lying area on the performance, lying behavior, and pen fouling was investigated during the warm periods (April to mid-October). One hypothesis was that a cooled lying area leads to more piglets lying in the lying area especially at the end of the rearing. Another hypothesis was, that a lying area with a heated floor leads to a better acceptance and lower fouling of the solid lying area compared to a lying area heated by a cover.

Materials and Methods
From January 2017 to July 2020 the project "Label-Fit" approached the implementation of the animal welfare label "The German Animal Welfare Association-For more Animal Welfare" (Deutscher Tierschutzbund-Für mehr Tierwohl) in conventional pig husbandry systems. Consequently, a solid lying area, different temperature zones, and automatic distribution of exploration and manipulation materials were installed in four rearing pens in the federal research and training center "Bildung-und Wissenszentrum Boxberg." Data over six rearing periods were collected between November 2018 and December 2019. All data of this study were recorded at a licensed farm that produces, rears, fattens, and markets pigs (VVVO-Number: 08 128 0140 538). The piglets from this study were fattened and marketed within the frame of normal farm practice at the end of each rearing period. Housing, management, and data acquisition were conducted under farm conditions and in accordance with German legislation (TierSchNutztV) [21] for farm animals. The animals were visually inspected for health issues daily. There were no signs and outbreaks of infectious diseases during the experiment.

Animals
A total of 1152 undocked rearing piglets (German Hybrid × German Piétrain) were used. The piglets were weaned and came into the rearing compartments at an age of four weeks. The mean weight at the time of weaning was 7.7 ± 1.5 kg and 29.5 ± 5.2 kg at the end of rearing. The piglets were kept in two compartments (D5 and D6) with two pens each (D501/D502 and D601/602) (Figure 1). A group of 48 randomly selected piglets was Agriculture 2021, 11, 324 3 of 14 allocated to each pen. The piglets were distributed according to an equal distribution of sex and weight in all pens.

Animals
A total of 1152 undocked rearing piglets (German Hybrid × German Piétrain) were used. The piglets were weaned and came into the rearing compartments at an age of four weeks. The mean weight at the time of weaning was 7.7 ± 1.5 kg and 29.5 ± 5.2 kg at the end of rearing. The piglets were kept in two compartments (D5 and D6) with two pens each (D501/D502 and D601/602) ( Figure 1). A group of 48 randomly selected piglets was allocated to each pen. The piglets were distributed according to an equal distribution of sex and weight in all pens.

Housing
The rearing compartments were temperature-controlled and had forced ventilation (25% minimum ventilation rate) with an air supply from outside through the attic and a porous ceiling. The set ventilation temperature during all rearing periods decreased from

Housing
The rearing compartments were temperature-controlled and had forced ventilation (25% minimum ventilation rate) with an air supply from outside through the attic and a porous ceiling. The set ventilation temperature during all rearing periods decreased from 25 • C (day 0) to 22.5 • C (day 15) to 18.5 • C (day 35) and to 17 • C until the end of the rearing period. From 7 a.m. to 6 p.m., the light in the compartments was switched on.
Each pen had a dimension of 7.95 m × 2.85 m (0.45 m 2 per piglet), including a lying, activity, and elimination area ( Figure 1). The solid concrete lying areas had 3% perforation for drainage, the concrete slatted floor had a slat width of 13 mm and 57 mm between the slats (<15% perforation), and the triangular grid had a slat width of 12 mm and 12 mm between the slats (50% perforation). The lying area in pen 501 and pen 602 was 1.5 m × 6 m (0.19 m 2 per piglet) and 1.4 m × 6 m (0.17 m 2 per piglet) in the other two pens due to the structural conditions. The activity area in each pen had also been adapted to ensure that all pens were the same size. The activity area in pen 501 and 602 was 1.35 m × 6 m (0.16 m 2 per piglet) and 1.45 m × 6 m (0.17 m 2 per piglet) in pen 502 and 601. The elimination area was split into two parts of a 0.95/1 m × 2.85 m triangular grid (0.11 m 2 per piglet).
The lying area under the covers in all pens could be cooled or heated by water. Therefore, solid concrete floor elements with built-in stainless steel water pipes were installed (DELA GmbH, Töging am Inn, Germany). The cover above the lying area in all pens was 1.1 m × 6 m in total, with two 0.9 m × 3 m infrared heating aluminum covers (REXLAN EUROPE ApS, Sorø, Denmark). These covers could be heated by water, and the heating function was based on the black radiation principle. Lamellas were attached to the cover; hence, the heat generated could not escape, and a warm microclimate was created under the cover.
The piglets were fed by plateau combination feeders with a 4:1 animal:feeding place ratio (L. Verbakel BV, Sint-Oedenrode, The Netherlands). These feeders were located in the middle of the partition of the two pens ( Figure 1). All piglets were fed ad libitum in a two-phase feeding regime ( Table 1). All pigs had free access to water at four drinking bowls, "Suevia 95S" (Suevia Haiges GmbH, Kirchheim/Neckar, Germany). The piglets had free access to two hamper rows, two wooden beams, and hay/straw pellets in a material dispenser, for exploration and manipulation. Twice per day (11 a.m. and 3 p.m.) 15 g chopped straw per piglet was spread automatically on the lying area under the cover by the automatic straw distribution system "Spotmix Welfare" (Schauer Agrotonic GmbH, Prambachkirchen, Austria). The lying area was cleaned three times a week if necessary during the whole rearing period in order to prevent a bad indoor climate.

Heating and Cooling
Two different heating systems were compared during three rearing periods with lower outside temperatures (mid-October to March). Accordingly, the lying area in each compartment was heated by the floor elements (heated floor) in one pen and heated by the infrared heating cover (heated cover) in the other pen. Since the provision of heating and covers is understood as good practice and standard in piglet rearing, in this research, no explicit control without equipment (and heating) was part of the research. Therefore, it was not investigated whether heating itself influences the behavior but how different types of heating influence the behavior of rearing piglets.
The set temperature curve of the floor heating was 35 • C (day 0), 33 • C (day 15) and 22 • C (day 50) and was controlled by the return temperature. Only hot water was added through the flow in the system and no cooling was practiced during the periods with heating. The lying area of both compartments in the other pen was heated by the infrared covers. The heating of the cover was controlled depending on the air temperature under the cover. The cover was heated in such a way that the air temperature underneath the Agriculture 2021, 11, 324 5 of 14 cover was the same in both pens to ensure comparability. The temperatures of the flow and the return were measured continuously to control both heating systems.
The floor cooling system was tested during the warmer season (April to mid-October). The lying area in one pen of each compartment was cooled by water. For this purpose, water was cooled down to about 12 • C with a heat exchanger (Buderus, BoschThermotechnik GMBH, Wetzlar, Germany) and then mixed with warm water until the target temperature was reached. To control the floor cooling system, the temperatures of the flow and the return were measured continuously. The set temperature curve of the floor cooling system ranged from 30 • C (day 0), 28 • C (day 15) to 22 • C (day 50) and was controlled by the return temperature. The other pen of each compartment remained unchanged (control).

Data Collection and Aggregation
The relative humidity and indoor and surface temperature were measured in the two compartments by 32 sensors ( Table 2). The sensors were compared and their functionality checked before each run.  The lying behavior of the piglets was recorded by video cameras. Two video cameras (HIKVision DS-2CD2125FWD-1, Neu-Isenburg, Germany) were installed in each compartment, one at the front and one at the back of each compartment. The cameras recorded 24 h per day during all rearing periods. The lying behavior was analyzed three days a week, at three time slots each day, i.e., 5-6 a.m. (morning), 10.30-11.30 a.m. (noon), and 8-9 p.m. (evening), using scan sampling. The lying behavior was usually analyzed 62 times per pen and period; however, due to defects or contamination of the camera, failures in data were possible. In every time slot, a screenshot of the first moment when all or most of the piglets were lying was used. These screenshots of the pen were segmented into five parts: triangular grid front and back, active area solid and slatted, and lying area, and the piglets lying in these areas were counted ( Figure 2). Piglets lying on the line were assigned to the part in which more than half of the body or the head was located. Piglets lying on the side or belly with legs stretched out or folded underneath were defined as lying. For each pen and time of sampling, the percentage of piglets lying on the solid floor was calculated for the statistical analysis. Based on direct observations in previous periods, it was found that the animals under the cover lay almost exclusively. Therefore, it was assumed that the animals under the cover that could not be seen on the screenshot were lying. The number of piglets lying in all areas except the lying area was counted, from which the number of piglets lying on the lying area under the cover was then deduced. One person exclusively executed the evaluation of the lying behavior by video analysis.
Agriculture 2021, 11, 324 6 of 14 each pen and time of sampling, the percentage of piglets lying on the solid floor was calculated for the statistical analysis. Based on direct observations in previous periods, it was found that the animals under the cover lay almost exclusively. Therefore, it was assumed that the animals under the cover that could not be seen on the screenshot were lying. The number of piglets lying in all areas except the lying area was counted, from which the number of piglets lying on the lying area under the cover was then deduced. One person exclusively executed the evaluation of the lying behavior by video analysis. The fouling of the pen with urine and feces was recorded in order to assess the acceptance of the functional areas by the piglets even more precisely. Weekly scores were used to monitor the fouling of the pen (usually seven observations per rearing period and pen). The same five segments that were used for the lying behavior were also used for recording the fouling. The fouling was recorded by following scores: Grade 0 (0-10% fouling), 1 (>10-25% fouling), 2 (>25-50% fouling), 3 (>50-75% fouling), and 4 (>75-00% fouling). The weekly scores were carried out mainly by the same person.
For statistical analysis, the measurements of the lying behavior were summarized for each rearing week. For each rearing period, means were calculated for each time of day (morning, noon, and evening). Thus, there were 21 means for each pen and rearing period used in the data analysis. For the analysis of the pen fouling, only a weekly average score was used (grade between 0-4).

Experimental Design and Data Analysis
The rearing periods represented the temporal and the two compartments the spatial repetitions of the experimental treatment in the experimental design. Thus, the measurements in the pens were considered statistically independent. With regard to determining which treatment to test in which pen (heated floor or above, cooling, or control), types of treatment were randomly distributed.
The effect on the daily weight gain was analyzed with a linear mixed model (procedure GLIMMIX, SAS 9.4 ® ). The fattening compartments (D5 and D6), the heating (heated floor and heated cover), or cooling (cooling and control), and their interaction were used as fixed effects. The rearing period (1-3) was set as a fixed design effect. The data of daily feed intake and feed conversion ratio served as informative accompanying parameters and were only analyzed descriptively. The fouling of the pen with urine and feces was recorded in order to assess the acceptance of the functional areas by the piglets even more precisely. Weekly scores were used to monitor the fouling of the pen (usually seven observations per rearing period and pen). The same five segments that were used for the lying behavior were also used for recording the fouling. The fouling was recorded by following scores: Grade 0 (0-10% fouling), 1 (>10-25% fouling), 2 (>25-50% fouling), 3 (>50-75% fouling), and 4 (>75-00% fouling). The weekly scores were carried out mainly by the same person.
For statistical analysis, the measurements of the lying behavior were summarized for each rearing week. For each rearing period, means were calculated for each time of day (morning, noon, and evening). Thus, there were 21 means for each pen and rearing period used in the data analysis. For the analysis of the pen fouling, only a weekly average score was used (grade between 0-4).

Experimental Design and Data Analysis
The rearing periods represented the temporal and the two compartments the spatial repetitions of the experimental treatment in the experimental design. Thus, the measurements in the pens were considered statistically independent. With regard to determining which treatment to test in which pen (heated floor or above, cooling, or control), types of treatment were randomly distributed.
The effect on the daily weight gain was analyzed with a linear mixed model (procedure GLIMMIX, SAS 9.4 ® ). The fattening compartments (D5 and D6), the heating (heated floor and heated cover), or cooling (cooling and control), and their interaction were used as fixed effects. The rearing period (1-3) was set as a fixed design effect. The data of daily feed intake and feed conversion ratio served as informative accompanying parameters and were only analyzed descriptively.
The effect on the lying behavior was analyzed with a linear mixed model (GLIMMIX procedure, SAS 9.4 ® ). The factors rearing week (1-7), time of day (morning, noon, evening), and heating (heated floor and heated cover) or cooling (cooling and control), and their interactions were taken as fixed effects. The design effect for the period × rearing week × time of day was also set as a fixed effect. The random effects for period × compartment and period × compartment × pen were assumed to be serially correlated across levels of rearing week and time of day, thus accounting for the repeated measures nature of the data. The anisotropic power covariance structure was assumed for both random effects, allowing for separate autocorrelations for rearing week and time of day.
The data of pen fouling were not normally distributed and were transformed by taking the square root. The fouling of pens was also analyzed with a linear mixed model (GLIMMIX procedure, SAS 9.4 ® ). In these analyses, the factors of week (1-7), heating (heated floor and heated cover) or cooling (cooling and control), and their interaction were taken as fixed effects. The rearing period x rearing week was set as a fixed design effect. The effect of the rearing week (1-7) was also considered as a random effect for the interactions period × compartment and period × compartment × pen. The anisotropic power covariance structure was used to estimate the random effects.
The approximate normal distribution and homogeneity of variance of the studentized residuals were confirmed graphically using QQ-plots and plots of residuals versus predicted values, respectively, for all models. The method of Kenward and Roger [22] was used for computing the denominator degrees of freedom for the tests of fixed effects for both models. No fixed or design effects were removed from the models. Comparisons of means were conducted using the Edwards-Berry procedure for controlling the familywise type I error rate at a level of α = 5% [23]. Furthermore, estimated effects with adjusted p < 0.1 were considered to be suggestive of real effects.

Temperature and Relative Humidity
The mean outdoor and indoor temperature and the relative humidity of all sensors for each compartment and period are shown in Table 3. There were no huge differences between the temperature or relative humidity in the two compartments (D5 and D6) during the rearing periods. Therefore, the two compartments could be used as a spatial repetition for the statistical analysis.  (1) Data from the weather station in Boxberg (www.wetter-bw.de, accessed on 14 October 2020); (2) Minute data recorded by two sensors for temperature and relative humidity in each compartment.
The air temperature in the compartments was higher during the warmer rearing periods when the cooling was tested (period 3 to 6) than during the rearing periods with heating (period 1, 2, and 6). The mean relative humidity during all rearing periods was between 44% and 62%. These values correspond with the requirements of the DIN 18910 [3]. Accordingly, the relative humidity in housing systems without heating should be between 60% and 80%, and between 40% and 70% with heating.

Performance
The data of 1112 rearing piglets were used for the analysis of the daily weight gain, feed intake, and feed conversion ratio. Data from 40 pigs were not used as they were removed from the pens due to active tail biting (biters), injury, or disease (Table 4). Due to the lack of individual animal-related data on feed consumption, the daily feed consumption and thus feed conversion ratio could only be calculated based on the total feed consumption in a pen over one rearing period. The daily weight gains of the rearing piglets corresponded, with weight gains of 435 ± 88 g/d, to the German average from 2019 with 435 g/d [24]. Therefore, the lower room temperature over all periods (25-17 • C) had no negative effect on the performance of the piglets. Johnston et al. [25] showed that a reduction of the room temperature by about 8 • C during the nighttime is effective in reducing energy costs without compromising pig performance. Table 4. Average daily weight gain (± standard deviation), feed intake, and feed conversion ratio for each rearing period and treatment. (n) = number of piglets, respectively, pens.
The feed intake and conversion ratio in this study were comparable to the values of other farms in this region [26]. Collin et al. [27] found that piglets fed ad libitum and exposed to 23 • C ambient temperature had a higher daily feed intake and daily weight gain (intake: 1117 g/d, gain: 914 g/d) than piglets exposed to 33 • C ambient temperature (intake: 837 g/d, gain: 751 g/d). Additionally, the piglets in this study showed better feed intake and daily weight gain in cool periods than during warm periods. There were no huge differences between the two heating systems in feed intake or conversion ratio, but the cooling seems to lead to a higher daily feed intake and conversion ratio compared with the control pens. Additionally, in the study of Collin et al. [27], the feed conversion ratios were higher in the group exposed to lower ambient temperatures.
Neither the heating systems nor the floor cooling had a significant effect on the daily weight gain, but the rearing compartment and the fattening period had a significant influence. The daily weight gain in the rearing periods with cooling was lower in D5 than in D6 (D5 393 g/d vs. D6 439 g/d, standard error of a difference = SED = 7.1, p < 0.0001). Furthermore, the rearing periods had a significant effect on the daily weight gain in all treatments. The daily weight gain in period 1 with 438 g/d for periods with cooling was significantly higher than in period 2 with 407 g/d (SED = 8.6, p = 0.0003) or period 3 with 404 g/d (SED = 8.7, p < 0.0001). For heating periods, period 1 with 468 g/d was tendentiously higher than period 2 with 451 g/d (SED = 8.7, p = 0.0519) and significantly higher than in period 3 with 441 g/d (SED = 8.7, p = 0.0024). Unfortunately, even a closer look at all animal and sensor data could not find any explanation for the effect of the compartment or the period. However, a descriptive examination of the average weight gains from the heating and the cooling periods reveals that the daily weight gain during the heating periods was higher (Table 4). This could be due to the lower room temperatures during the heating period than during the cooling periods ( Table 3). Results of further studies showed that lower ambient temperatures increase the performance of piglets during rearing and, furthermore, during fattening [4,5].

Function of Heating and Cooling Systems
The temperature data were examined in more detail to check the function of the heating and cooling systems. Accordingly, the data taken every minute were combined to mean values for each rearing week over all heating ( Figure 3) and cooling periods (Figure 4), respectively. Data from all air temperature sensors (six per compartment) and surface temperature sensors (six per compartment) in the lying area were used for this analysis. A microclimate could be created under the covers in all pens, which was clearly different from the room temperature (mean difference during heating periods 5.8 ± 1.2 • C; during cooling periods 3.8 ± 1.4 • C). The functional reliability of the different functional areas could be increased by creating a clear temperature difference between the lying area and the active/elimination area [1,7]. The room temperature of 24-19 • C and the temperature under the cover of 28-24 • C (Figures 3 and 4) corresponded to the temperature optimum for piglets from 10-30 kg of 26-18 • C [3]. from the room temperature (mean difference during heating periods 5.8 ± 1.2 °C; during cooling periods 3.8 ± 1.4 °C). The functional reliability of the different functional areas could be increased by creating a clear temperature difference between the lying area and the active/elimination area [1,7]. The room temperature of 24-19 °C and the temperature under the cover of 28-24 °C (Figures 3 and 4) corresponded to the temperature optimum for piglets from 10-30 kg of 26-18 °C [3].   An average profile of all surface temperatures (ST) of the lying areas (12 sensors) and the air temperatures under the covers (TC) (eight sensors) and in the compartments (four sensors), in addition to the set air temperature in the compartments over all three heating periods. The mean temperatures for each rearing week, which were formed from the records taken every minute over the three rearing periods, are shown. The bars drawn at each data point indicate the standard deviation.  The temperature under both covers was kept as similar as possible to ensure that the two heating systems could be compared. The mean temperatures under the covers over all time was 25.7 ± 3.3 • C for pens with heated covers and 25.9 ± 1.5 • C for heated floors. Thus, the comparability of the two heating systems was provided. There was a difference in the surface temperature of the lying area in the pens during the heating periods, especially in the first four weeks. The surface temperate of the lying area in pens with a heated floor was higher (31.7 ± 3.5 • C) than in pens with heated covers (29.2 ± 4.2 • C) (Figure 3).
In addition, the temperature under the covers of all pens during the cooling periods was nearly similar (Figure 4). The surface temperature at the beginning was warmer than in the control pens due to the set temperature of the floor cooling given. As intended, the surface temperature of the lying area in pens with cooling could be decreased by the floor cooling system during the last weeks. The set air temperature in the compartments was exceeded from week three during the cooling periods ( Figure 4). This could have meant that it was too warm for the rearing piglets. However, this higher temperature was also an advantage for testing the floor cooling system.

Influence of Heating on Lying Behavior
The average proportion of piglets lying at the three time slots chosen were calculated to verify the choice. The average proportions for the three times of day were morning 99.1%, noon 91.1%, and evening 97.3%. The data also show that the lying area was well accepted because over 90% of the piglets were lying on the lying area during heating and over 80% during cooling periods. The high acceptance could be due to the solid surface, the higher temperatures, and/or the lower light intensity under the covers. Pigs avoid slatted areas for resting under optimal ambient temperatures and prefer solid areas [28]. Furthermore, the lower light intensity under the cover may have improved the acceptance of this area for lying [29][30][31]. However, the most important factor influencing rearing piglets is temperature. As has already been mentioned, piglets need warmer temperatures directly after weaning, especially for resting [1,2]. Due to the higher temperatures under the cover in this study and their temperature demands, the piglets were influenced in their choice of lying area.
The lying behavior was not affected by the heating by cover or floor ( Figure 5). However, the acceptance of the lying area as such by the piglets was very high in both systems (94% with heated floor and 93% with heated cover). This suggests that both heating systems are suitable for rearing piglets. Heating with a cover has the advantage that it can be easily integrated into existing housing systems and is usually less expensive than floor heating. In this study, the heated cover for one pen cost about EUR 1400, while the heatable/coolable floor cost about EUR 1700. The advantage of a heated floor, on the other hand, is that the set temperatures have an unadulterated effect on the animal through direct body contact and can also be adjusted relatively quickly. However, due to this direct contact with the animal, the control of floor heating must be more precise because temperatures that are too warm or too cold affect the animals directly.

Influence of Cooling on Lying Behavior
The cooling alone had a tendentiously significant effect on the lying behavior of the piglets: Fewer animals lay on the lying area in pens with cooling than in pens without cooling (cooling 83%, heating 85%; SED = 1.33, p = 0.0933) ( Figure 6). These results contrast with previous studies with fattening pigs in which floor cooling increased the number of piglets on the lying area [18][19][20]. However, the piglets showed a significant influence of the cooling on the lying behavior depending on the rearing week (p = 0.0004). In rearing weeks three and four, significantly fewer piglets lay on the cooled lying area (cooling 70%; control 77%), compared to the control pens. In week seven, there was an opposite effect of the cooling since 88% of the piglets lay on the cooled lying area, while only 83% used the lying area for lying in the control pens. This could suggest that the surface temperature was set too low too early. If the temperature was too low, smaller piglets especially would avoid the lying area.

Influence of Cooling on Lying Behavior
The cooling alone had a tendentiously significant effect on the lying behavior of the piglets: Fewer animals lay on the lying area in pens with cooling than in pens without cooling (cooling 83%, heating 85%; SED = 1.33, p = 0.0933) ( Figure 6). These results contrast with previous studies with fattening pigs in which floor cooling increased the number of piglets on the lying area [18][19][20]. However, the piglets showed a significant influence of the cooling on the lying behavior depending on the rearing week (p = 0.0004). In rearing weeks three and four, significantly fewer piglets lay on the cooled lying area (cooling 70%; control 77%), compared to the control pens. In week seven, there was an opposite effect of the cooling since 88% of the piglets lay on the cooled lying area, while only 83% used the lying area for lying in the control pens. This could suggest that the surface temperature was set too low too early. If the temperature was too low, smaller piglets especially would avoid the lying area.
Regarding the surface or air temperature data, it is not clear whether it was too cold or too warm. The surface temperature of the lying area in week three was higher in the pens with cooling than in the control pens (cooling 28.1 ± 2.0 • C; control 27.3 ± 4.2 • C), while it was the opposite in week four (cooling 26.8 ± 3.0 • C; control 28.1 ± 4.0 • C). The air temperature under the cover was lower in pens with cooling in both weeks where fewer piglets lay on the lying area. Accordingly, the combination of the air and the surface temperature did not meet the piglets' requirements. Significantly more piglets lay in the cooled lying area, compared with the control pen during the last week. In addition, the number of pigs in the cooled lying area increased during the last three weeks of rearing ( Figure 6) compared to week four and were higher than in the control pens. These results could be confirmed by other studies showing that cooling has the potential to increase the acceptance of a lying area by fattening pigs during warmer seasons [18][19][20]. Regarding the surface or air temperature data, it is not clear whether it was too cold or too warm. The surface temperature of the lying area in week three was higher in the pens with cooling than in the control pens (cooling 28.1 ± 2.0 °C; control 27.3 ± 4.2 °C), while it was the opposite in week four (cooling 26.8 ± 3.0 °C; control 28.1 ± 4.0 °C). The air temperature under the cover was lower in pens with cooling in both weeks where fewer piglets lay on the lying area. Accordingly, the combination of the air and the surface temperature did not meet the piglets' requirements. Significantly more piglets lay in the cooled lying area, compared with the control pen during the last week. In addition, the number of pigs in the cooled lying area increased during the last three weeks of rearing ( Figure 6) compared to week four and were higher than in the control pens. These results could be confirmed by other studies showing that cooling has the potential to increase the acceptance of a lying area by fattening pigs during warmer seasons [18][19][20].

Fouling of the Pens
The fouling of the entire pen and the different areas was very low over all six rearing periods (Figure 7).

Fouling of the Pens
The fouling of the entire pen and the different areas was very low over all six rearing periods (Figure 7). The area with the highest fouling was the triangular grid, but the fouling even in this area with grade <1 (>10-25% fouling) was very low. The solid lying area was only slightly fouled at the sides that bordered the triangular grid. The activity area was not fouled, The area with the highest fouling was the triangular grid, but the fouling even in this area with grade <1 (>10-25% fouling) was very low. The solid lying area was only slightly fouled at the sides that bordered the triangular grid. The activity area was not fouled, except for the area around the exploration material dispenser with the pellets. Neither the cooling nor different heating systems had any influence on the fouling of the solid lying area. The low level of fouling by feces and urine suggests that the division of the pen into functional areas is well accepted by the rearing piglets. Previous studies showed that it is difficult to keep a solid area clean, especially in summer when the temperatures increase [11,32]. The pen fouling during this study remained low even during warmer summer conditions.

Conclusions
The results show that both heating systems were suitable for the rearing of piglets. There were no differences in the lying behavior, performance, or pen fouling between the heated cover or floor. The cover is a good option for existing rearing farms to set up a microclimate and, thus, reduce heating costs. However, floor heating offers a great advantage in terms of high temperatures due to the possibility of cooling. When using a cooling system, it is very important to adapt the set temperature of the cooling to the air temperature and use it mainly toward the end of the rearing period. The structuring of the pen with different surfaces and two climate zones turned out to be practicable since all functional areas were well accepted by the piglets. An intelligent control system that automatically adjusts the cooling and heating, depending on the indoor temperature or the behavior of the piglets, would be an option for optimization. By utilizing a heat exchange concept for cooling and heating, the system can also be economically optimized.