A deeper and unique dive into water usage data

Using room level data to characterizes heat stress occurrences between natural, tunnel-ventilated facilities

By Felipe Rodrigues Picchi and Brett C. RamirezHeat stress continues to be a major environmental stressor for all stages of swine production, with severe productivity and economic losses. For growing pigs, heat stress can be evaluated using different physiological and behavioral responses (Mayorga et al., 2019). Core-body temperature is the gold standard and is approximated by rectal temperature; however, continuous monitoring of rectal temperature in a commercial setting is challenging due to physical and technological constraints. Feed intake and daily gain can also be quantified but requires labor and special facilities. Behavioral responses can be captured using cameras, but this requires computationally expensive algorithms to classify heat stress behavior, such as wallowing, laying patterns and feeder/drinker visits. There is a need to explore inexpensive and rapid approaches to identify heat stress patterns in growing pigs to compare different environments and management approaches.Room-level water monitoring can provide insight into the responses of growing pigs to changes in environmental conditions, health challenges, and diet composition. However, at the room level, total water usage includes water intake by the pigs and wastage (Patience, 2012). Water intake is defined as the water ingested by the pigs, which is influenced by physiological responses and behavioral mechanisms. Water wastage is affected by the water delivery device, water flow rates, environmental conditions, and pig behavior (Patience, 2012). If health status, diet composition and other management practices remain constant, deviations in water usage patterns may be attributed to environmental conditions. This preliminary study aimed to characterize heat stress occurrences between natural and tunnel-ventilated facilities using room-level water usage data and to highlight the importance of timing for management during hot weather. Two wean-to-finish barns located in west-central Iowa were used as examples in this study. One double-wide barn featured a power-tunnel ventilation system with a nominal capacity of 2,400 pigs (1,200 per room). The second barn featured a double-long, power-natural ventilation system with a nominal capacity of 1,200 pigs per room. Both barns were equipped with sensors to measure temperature, relative humidity, and room-level water usage (BarnTalk, BarnTools, Urbandale, Iowa). Additionally, an outdoor temperature sensor monitored ambient conditions.During heat stress conditions, it is well-documented that pigs decrease activity in response to warming environmental conditions, as pigs seek to regulate their body temperature by reducing movement and feed intake and, consequently, water intake. When analyzing the data, two distinct patterns emerged: when the room temperature was above thermoneutral conditions, water usage had two peaks during the day, versus when the room temperature was below thermoneutral conditions, water usage typically peaked once during midday. We deployed several algorithms to automatically detect these peaks in water usage to classify them into heat stress and thermoneutral conditions. An example is shown in Figure 1.

Figure 1: Relationship between water usage and room/outdoor temperatures for tunnel-ventilated facility. (a) An example of a temperature above thermoneutral conditions shows two peaks in water usage, aligning with fluctuations in the room temperature before and after a significant rise in ambient temperature. (b) An example of the same variables on a day with lower temperatures where water usage peaks only once during the day.

We investigated what times of the day may experience increases in water usage. Based on the histograms presented in Figure 2, there are distinct peak frequency patterns for the tunnel-ventilated and natural-ventilated systems across a 24-hour period. For the tunnel-ventilated barn (Figure 2a), the peak frequency is noticeably higher during the evening hours, particularly between 16:00 and 19:00, where it reaches a maximum of 30 occurrences – suggesting a significant increase in peak formation during these hours. In contrast, the naturally ventilated system (Figure 2b) exhibits a distribution pattern of peak frequencies, with notable increases around 8:00, 10:00, and again between 17:00 and 20:00. The maximum peak frequency for this system is lower, around 25 occurrences, indicating a less predictable peak formation throughout the day compared to the tunnel ventilated system.

Figure 2: Peak frequency histograms for (a) tunnel ventilation and (b) naturally ventilated systems on a 24-hour day. The x-axis represents the hour of the day, and the y-axis represents the peak frequency throughout the two months of recorded data.

Taking a deeper dive in water usage data collected at frequent intervals can be potentially used to inform environmental control strategies and when caretakers need to be active in the barn. Under heat stress conditions, if caretakers/managers go through the barn during the middle of the day, when the pigs are least active (based on this data), this may result in improper feeder adjustments and/or trying to get pigs active when they typically are not. Getting pigs up and moving to feeders on the ends of the peaks in water usage may result in better distribution of pigs at the feeder, resulting in increased productivity. Acknowledgments
Special thanks to AMVC for access to the barns and BarnTools for their support and help with accessing and processing data.

Picchi is a graduate research assistant and Ramirez is the interim director of the Egg Industry Center and an associate professor in the Department of Agricultural and Biosystems Engineering, Iowa State University.