Automating Dairy Cattle Welfare Monitoring: A New Era for Healthier Herds

The well-being of dairy cattle has always been a priority for farmers and veterinarians, but keeping a close eye on every cow in a herd is a challenging task. Thankfully, technological advancements in automatic welfare monitoring systems are now transforming cattle welfare management in exciting ways. By blending science with data-driven monitoring, these systems help farmers create healthier environments and respond to cattle needs quickly. This article explores how these automated systems work, focusing on methods, benefits, and practical tools that can improve both herd health and farm efficiency.

The Role of Automated Monitoring in Dairy Welfare

Automated welfare monitoring systems provide a window into a cow’s daily life, capturing various welfare indicators such as environmental conditions, behavior, and health parameters. These systems fall under precision livestock farming (PLF), a field designed to increase efficiency and reduce stress on animals by minimizing human interaction and maximizing data insights. This shift is crucial as herds grow larger, allowing for comprehensive monitoring while maintaining high welfare standards.

Essential Welfare Monitoring Parameters

Automated systems track both environment-based and animal-based parameters. Here’s how they work:

1. Housing Environment Parameters: Sensors measure air quality, temperature, and humidity, among other things. These indicators ensure that cows are kept in conditions that align with optimal welfare standards. Good air quality and proper ventilation are critical, and automated systems can immediately alert farmers if adjustments are needed.
2. Animal-Based Parameters: Automated monitoring systems also track animal-specific parameters like movement, milk composition, body temperature, and heart rate. These insights can highlight potential health issues (such as mastitis, detected through milk composition), reproductive cycles, and behavioral trends that may indicate stress or discomfort.

Core Components of Automated Cattle Welfare Monitoring

Effective monitoring requires a combination of several key components:

• Identification Systems: Automatically identifying each cow is essential for accurate data tracking. Modern systems can tag animals individually, enabling real-time monitoring without human intervention.
• Behavioral and Health Sensors: Sensors placed in milking robots or feeding stations measure everything from milk yield to body temperature and heart rate. Advanced pedometers can track steps, helping farmers identify lameness early, while sensors near feeding areas can alert farmers to changes in feeding habits—often the first sign of illness.
• Data Management and Analysis: Data from sensors is stored in a centralized database, often linked to a farm management information system (MIS). The database helps create welfare models that interpret collected data, flagging any indicators of distress or health concerns.

Innovative Modules for Advanced Health Monitoring

Many automated systems include advanced modules for specific welfare concerns:

• Lameness Detection: Lameness, a common issue among dairy cows, can be assessed through technologies like weight distribution and gait analysis using walk-over mats, pressure-sensitive pads, or video analysis. For example, walk-through scales can monitor vertical force distribution on each leg to identify subtle gait abnormalities.
• Milk Composition Analysis: Parameters such as milk temperature, electrical conductivity, and somatic cell count provide insights into a cow’s health. High somatic cell counts, for instance, often indicate mastitis, while changes in milk conductivity can signal udder health issues.

Automated Monitoring Systems: Examples and Applications

Many companies now offer solutions for automated welfare monitoring, including DeLaval Herd Navigator, which provides real-time alerts for health and productivity, and Lely milking robots, known for their data-driven animal care. These systems not only improve the quality of care for each cow but also streamline operations for farmers, making data-driven decisions faster and easier.

Actionable Tips for Implementing Automatic Welfare Monitoring

1. Start with Key Parameters: If new to automation, focus on milk composition and body temperature monitoring, which provide early warnings of potential health issues.
2. Use Existing Data: Many farms already collect some welfare-related data; integrating these into a centralized MIS can be an efficient start to automation.
3. Prioritize Lameness Monitoring: Lameness impacts productivity significantly, and early detection through weight or gait analysis is an effective preventive measure.
4. Collaborate with Veterinarians: Automated systems are excellent tools, but they complement, not replace, expert care. Work with veterinarians to understand system reports and implement best practices.

Summary Points for Social Media and Visual Content

• Automation in Dairy: Precision livestock farming is transforming dairy welfare by using sensors and data to keep cows healthy.
• Key Monitors: Essential parameters include housing quality, milk composition, body temperature, and activity levels.
• Advanced Tools: Technologies like walk-over mats, feeding sensors, and milking robots simplify welfare monitoring.
• Lameness Matters: Automated lameness detection tools like weight scales and gait analysis identify issues early, improving overall herd health.
• Data Management: Centralized databases create actionable insights, making data-driven care possible on a large scale.

Use these points to create a visual summary for Instagram or Canva infographics, highlighting the tools and methods used in automated dairy welfare monitoring to engage and educate your audience.

The document outlines the implementation of an Ethernet-based Animal Welfare Analysis Local Area Network (AWALAN) in livestock housing, specifically for dairy cattle welfare monitoring. Key aspects include integrating data from various sources, such as leg pressure mats, video gait analysis, and milk conductivity sensors, to monitor cow health and behavior in real-time.

The network leverages TCP/IP protocol for device communication, supporting both wired and wireless connections. AWALAN’s architecture connects multiple controllers and servers, facilitating data transfer from systems monitoring factors like cow gait, milk yield, and body temperature. The gathered data is then analyzed through probabilistic models (e.g., Bayesian networks, neural networks) designed to assess individual animal welfare and issue alerts.

The system is modular and supports research by enabling access to both raw and processed data. This framework aims to improve farm management, disease prevention, and welfare transparency across dairy production units, enhancing welfare analysis at various levels from local to potentially national networks. The document also emphasizes the need for model calibration with veterinary expertise to ensure accuracy in welfare assessment.

Continuing from the previous summary, the document further emphasizes the critical role of high-speed data exchange to manage the increasing volume of information as new measurement and analysis tools are incorporated. For example, Ethernet’s high data rates (100 to 1000 Mbps) are advantageous over traditional industrial field buses, supporting the substantial data loads from systems such as video gait analysis and sensor mats used for leg pressure monitoring. The choice of Ethernet also promotes flexibility and openness, making it easier for third-party producers and researchers to add devices and access raw data.

The AWALAN and LAN systems in this setup are closely integrated, with the AWALAN dedicated to animal welfare analysis while leveraging data from the farm’s Management Information System (MIS) database. The analysis server acts as a central hub for managing data flow, functioning as a VPN server-router and database server while also providing remote access for researchers and farm administrators. Through this server, AWALAN can retrieve essential welfare data, combine it with MIS and remote databases, and apply welfare assessment models to predict issues and generate alerts for early intervention.

The welfare assessment models rely on multiple indicators, such as locomotion, milk yield, body temperature, and leg pressure. More complex welfare assessments, like mastitis detection, utilize additional specific indicators like somatic cell count (SCC) and milk conductivity. These inputs are derived from a combination of measurement devices, MIS outputs, and external databases, which are interconnected via the farm’s LAN.

To ensure accuracy, the models undergo calibration with experienced veterinarians who help determine sensitivity and specificity, minimizing false positives and negatives. The welfare model is built using an evaluation schema based on the Welfare Quality® system, a structure that uses over 30 on-farm measures to assess 12 welfare criteria across four dimensions. This hierarchical model arrangement allows for detailed welfare analysis at multiple levels, enabling both localized monitoring of individual animals and broader data collection for overall welfare assessment.

The document concludes by highlighting the broader benefits of this automated welfare monitoring approach. It aims to enhance welfare management in milk production by reducing disease-related losses, optimizing cow productivity, and supporting innovations in traceability and labeling within the food production chain. Furthermore, this framework supports the development of regional, national, and international networks for welfare control, promoting consistent welfare standards across dairy production operations.

Finally, the document acknowledges the importance of interdisciplinary research and cooperation with veterinarians to refine the assessment models and enhance the system’s overall reliability. These automated welfare assessment systems represent a forward-thinking approach to livestock management, aiming to balance animal welfare with production efficiency and contribute to sustainable dairy farming practices.

and specifically, coughs due to non-infectious causes, like poor air quality, showed a peak frequency of around 1,574 Hz, while infectious coughs exhibited a significantly lower peak frequency, typically around 500 Hz. This difference in frequency, along with the duration of each cough (with infectious coughs lasting approximately 0.67 seconds compared to 0.43 seconds for non-infectious coughs), supports the potential for acoustic data to distinguish between infection-related and environment-related respiratory issues.

Research utilized over 500 GB of cough recordings, both in lab and field settings, to investigate and classify these acoustic features, which in turn helped refine automated detection algorithms. The algorithms use characteristics like peak frequency, RMS, and sound duration to identify the type of cough. These results suggest that such automated sound analysis systems can serve as an early warning tool for farmers, alerting them to health issues within pig herds based on real-time cough detection.

An additional aspect of the study was the development of a localization method to identify where cough sounds originated within a pig housing facility. This method calculates the time difference of arrival of sound waves to locate the source of coughs, which can inform farmers about specific areas in need of intervention. This automated, real-time monitoring has the potential to help farmers reduce antibiotic use by distinguishing between infections and environmental issues, allowing for targeted management of air quality when needed, which in turn could decrease the incidence of disease and associated treatment costs.

In conclusion, the integration of sound analysis into pig farming health monitoring systems shows promise for improving respiratory disease management. By providing a non-invasive and efficient diagnostic tool, these systems could enhance animal welfare, reduce the spread of infections, and promote a more judicious use of antibiotics, contributing to sustainable livestock practices.

with non-infectious coughs exhibiting a peak frequency of around 1,574 Hz, in contrast to the lower peak frequencies of infectious coughs, around 500 Hz. This frequency differentiation allows for effective categorization of cough types, critical for monitoring pig health conditions in real-time. Specifically, the higher peak frequency in non-infectious coughs suggests an environmental origin, like air quality issues, while lower frequencies are associated with infectious agents such as Pasteurella and Actinobacillus. This distinction is vital as it aids in reducing unnecessary antibiotic use, thus addressing both economic and health concerns.

The study further explores how these findings can be implemented in automated systems using sound detection algorithms that classify cough sounds based on features such as peak frequency, duration, and energy envelope. By deploying such a system, farmers can receive real-time alerts about the health status of pigs, allowing them to take early action, potentially adjusting ventilation to improve air quality or administering targeted treatment only when necessary.

Additionally, a localization technique was developed based on time-difference-of-arrival calculations, enabling precise tracking of where coughs occur within a facility. This feature enhances the monitoring system by pinpointing specific areas of concern, enabling farmers to take localized action, further optimizing herd health management. Through the continuous monitoring of cough sounds, this acoustic system can help detect respiratory issues early, reduce disease spread, and lower healthcare costs, while minimizing antibiotic resistance risks.

In summary, sound analysis technology provides a promising, non-invasive method for health monitoring in pig farming. By using sound features as indicators of respiratory conditions, this technology offers a proactive approach to managing animal welfare, ensuring timely interventions, and ultimately promoting sustainable farming practices.

This passage explores the development and implementation of a sound-monitoring system in pig farms to identify respiratory health issues, particularly focusing on cough detection, classification, and localization. The system works by recording and analyzing cough acoustics to detect infectious versus non-infectious coughs based on characteristics such as peak frequency, duration, energy envelope, and the use of autoregressive (AR) models. Here’s a breakdown of the key points:

1. Cough Identification and Classification:
   • Coughs are distinguishable from other pig sounds by their unique acoustic features, lacking a clear fundamental frequency and displaying broad energy distribution.
   • Infectious coughs tend to have a lower peak frequency (around 500 Hz) and a longer duration (0.67 seconds), compared to non-infectious coughs with a higher frequency (approximately 1,574 Hz) and shorter duration (0.43 seconds).
   • The energy envelope of sounds helps in distinguishing coughs from background noise, enabling the system to capture and label cough sounds.
2. Automated Cough Recognition:
   • A tool uses the Hilbert Transform to calculate the energy envelope of recorded sounds, detecting and isolating potential cough sounds based on amplitude thresholds.
   • Coughs are further processed to form a classifier using AR analysis and clustering algorithms, achieving a high accuracy rate of 92% for distinguishing sick coughs from other sounds.
3. Sound Localization:
   • Using multiple microphones, the system can estimate the position of coughing pigs by analyzing time delays in sound reception at different microphone locations. This helps to localize coughs within specific zones, or “hazard areas,” of the pig housing.
   • Localization is essential to target treatments to specific areas rather than administering antibiotics to the entire population, which helps in reducing antibiotic usage.
4. Time Constant Analysis:
   • Differences in the decay of the energy envelope in cough sounds are analyzed using time constants, with infectious coughs showing longer time constants. This provides insight into the respiratory health of pigs and helps in monitoring disease progression.
5. Application in Farm Management:
   • This system enables continuous monitoring of pig respiratory health, allowing farmers and veterinarians to detect and respond to health issues early on. It offers a way to manage respiratory diseases effectively by treating only the necessary pens, potentially reducing the reliance on antibiotics.

This automated system thus supports early disease detection and selective treatment, promising to enhance animal welfare while minimizing medication costs and antibiotic use in farming.

The study outlines the advancements in automated systems for early lameness detection in dairy cattle, essential for improving animal welfare and minimizing economic loss. Two key systems are examined: the GAITWISE system and image-based analysis.

1. GAITWISE System: This system uses a pressure-sensitive mat to monitor hoof placement over time, recording spatial, temporal, and force-related data. Through this, 20 parameters, such as step length and stance time, are calculated, identifying subtle gait asymmetries that can signal early lameness. Real-time data collection and high success rates at experimental farms demonstrate the GAITWISE system’s practicality, although commercial application may benefit from environmental controls to reduce measurement errors.
2. Image-Based Analysis: Vision technology captures cow gait via video to automatically assess “step overlap” and “back arch,” both indicators linked to lameness. The step overlap, measured through image processing algorithms, correlates with manual gait scoring but shows variability, meaning it might be more effective when combined with other features. Additionally, back arch analysis via curvature calculation accurately classified lameness in experimental settings, achieving a 96% success rate across different datasets.

Both systems show promise for commercial on-farm application, offering tools that could reduce the need for frequent manual inspections and early interventions for improving cattle welfare.

These results highlight the potential of back arch analysis as a robust indicator for lameness detection in dairy cows, achieving high classification accuracy with minimal misclassification. This success rate could be due to the consistency of back arch changes with levels of pain and discomfort in lame cows, which experts also use for manual scoring.

Overall, although both step overlap and back arch analysis methods for detecting lameness in dairy cows show promise, further validation and calibration for practical on-farm use are required. Variability among individual cows remains a challenge, especially with step overlap, which is influenced by differences in natural gait and conformation. Consequently, integrating multiple parameters or developing individual cow-specific baselines may enhance detection accuracy, allowing early intervention and promoting cow welfare.

22.4 Future Potential of Automated Lameness Detection Technologies

While advances in pressure mat and video-based systems show high potential for early lameness detection, questions remain regarding on-farm applicability. In practical settings, considerations such as environmental conditions, farm layout, ease of system use, and cost must be addressed. For example, pressure mat systems like GAITWISE, though effective, may face issues with measurement success rates if environmental factors distract cows. Furthermore, installation and maintenance of such systems require an initial investment that may be prohibitive for some farms.

On the other hand, image analysis, although non-contact and potentially less invasive, requires a controlled setup to ensure consistent results. Variations in lighting, camera angle, and cow positioning can all impact the accuracy of gait feature extraction. Innovations in camera technology, such as high-speed, low-light capable cameras, along with artificial intelligence (AI)-driven algorithms for analyzing complex gait patterns, could improve robustness. Moreover, cloud-based processing and data storage may facilitate real-time monitoring across farms.

Future research should aim to refine these automated systems by focusing on:

1. Individual cow gait patterns: Establishing individual baselines to better account for cow-to-cow variation, which would allow systems to detect subtle deviations in gait.
2. Environmental resilience: Improving system design to operate reliably under varied farm conditions.
3. Cost reduction and scaling: Developing scalable solutions that are cost-effective for broader adoption across dairy operations, potentially through modular designs or shared infrastructure.

In conclusion, as technology evolves, automated gait analysis systems hold significant promise for improving the welfare and productivity of dairy herds by providing early, objective, and continuous lameness detection. By addressing current limitations and integrating multi-parameter approaches, these tools may soon become valuable assets in precision livestock farming.

The Potential of Automatic Lameness Detection Systems

Despite over two decades of research into lameness and early detection methods, a widely adopted, commercially successful solution has yet to emerge in the dairy industry. Currently, StepMetrix™ (BouMatic) is one of the few commercial products available for automatic lameness detection. The StepMetrix™ system measures ground reaction forces as cows walk through it, analyzing the force and duration of each step to assign a specific SMX score to each hind limb. The system is technically mature and can reliably operate on farms, scoring cows individually even when multiple cows are present on the mat.

However, a study by Bicalho et al. (2007) highlighted a significant challenge with StepMetrix™, noting its high specificity rate (85.4% to 94.5%) but low sensitivity rate (20.4% to 35.2%). This imbalance means the system accurately identifies non-lame cows but often fails to detect actual cases of lameness. Bicalho and colleagues further concluded that visual locomotion scoring conducted by trained veterinarians outperformed the StepMetrix™ in terms of detection accuracy, leading some stockpersons to question the cost-benefit ratio of investing in such a system.

Similar to the GAITWISE and image analysis tools discussed in this chapter, several technical innovations hold potential for automatic lameness detection. Despite their promise, none of these systems have yet become commercially viable products. From a research perspective, these tools could serve as effective lameness detection systems, as scientific studies have shown that the biological data they measure can be correlated with lameness. However, the path to commercial application faces several obstacles:

1. Alignment with Stockperson Expectations: For practical adoption, these systems must provide results that meet the expectations of dairy workers, including clear, actionable insights.
2. Cost-Benefit Ratio: Stockpersons must feel that the investment in these systems will yield economic benefits, such as improved herd health and productivity.

An automatic detection system that stockpersons would find valuable must be efficient, accurate, reliable, and capable of functioning under any farm and climate conditions. This calls for product development beyond the proof-of-concept phase, necessitating further investment in funding, human resources, and research.

Achieving a commercially viable solution will likely require a collaborative approach, with universities, research institutes, and companies working together to bridge the gap between research and practical implementation.

Lighting for Laying Hens: The Effect of Environmental Factors on Bird Behaviour

Abstract

The quality and rhythm of light are crucial for the early development of layer chickens. Light can significantly influence behavioural issues, such as cannibalism, which tend to decrease with proper rearing practices. Early exposure to perches and light types may also impact feeding and perching behaviour, as well as preferences for light sources later in life. This study aimed to explore individual differences in perching behaviour and whether environmental enrichment enhances early perching, thereby reducing behavioural issues. The effects of natural and artificial light on perching and feeding behaviour were examined, alongside the light preferences of birds at 14 weeks. Additionally, a separate on-farm study assessed the lighting conditions in common Swedish henhouses and tested the HATO® lighting system in relation to bird health and welfare requirements.

Key Findings

• Early perching behaviour was positively correlated with time spent under perches but negatively correlated with social interactions.
• Environmental enrichment did not significantly affect the latency to start perching, although birds with floor enrichment tended to roost earlier.
• Day length influenced feeding behaviour, and access to daylight appeared to promote earlier perching.
• Night-time roosting initiation was associated with daytime perch use.
• Chicks raised under incandescent light preferred incandescent lighting, whereas those raised in natural light preferred natural lighting.
• No severe health or behavioural issues were noted in farms utilizing various lighting systems, except for instances of feather pecking.

Introduction

Domestic laying hens, descended from the red jungle fowl, are naturally diurnal and gregarious. Their natural environment consists of a 12-hour light and 12-hour dark cycle, highlighting the importance of the lighting environment in egg production.

The light source’s characteristics, such as intensity (illuminance) and wavelength (spectrum), differ significantly. Natural light offers an even distribution of wavelengths between 400 and 700 nm and contains ultraviolet A (UVA) light, while incandescent light has a different spectral distribution with more red and less blue wavelengths. The bird’s visual system, which includes a fourth type of cone allowing for UVA perception, provides them with greater spectral sensitivity than humans. As a result, traditional illuminance measures (lux) may not adequately represent how birds perceive light; a proposed alternative unit, Gallilux, has been suggested for poultry environments.

Light quality and rhythm are essential for the early development of chickens. Natural behaviour includes roosting at night, a behaviour not always exhibited in commercial settings due to suboptimal lighting conditions. The European Union’s directive mandates the phasing out of battery cages, necessitating non-cage systems that enhance welfare but may increase risks for feather pecking and cannibalism. Early rearing conditions are known to influence these behavioural issues, with previous research indicating that early perch use reduces floor laying and cloacal cannibalism.

In Sweden, all farm animals must have access to natural daylight according to animal welfare legislation, requiring poultry houses to include windows. However, inadequate lighting management may lead to behavioural problems, raising concerns among farmers regarding the redesign of old henhouses to comply with these regulations.

Incandescent lighting has traditionally dominated commercial henhouses in Sweden, but its suitability is debated. Studies suggest that hens prefer fluorescent tubes due to the blue wavelengths, which might enhance physical activity levels compared to incandescent light.

Recent Developments

With the European Union’s ban on opaque incandescent light bulbs in 2009, there is a pressing need for alternative lighting solutions in poultry houses. The HATO® Agricultural Lighting system has been introduced in Sweden, promising a light spectrum closer to natural light, with an even distribution of wavelengths and increased UVA light.

Swedish animal welfare regulations require that all new technical equipment receive approval before use, necessitating scientific investigations into animal health and welfare impacts. The HATO® system was identified by the Swedish Agricultural Board as a new technique needing thorough evaluation before unrestricted marketing.

Objectives of the Studies

The primary objectives were:

1. To investigate individual differences in the onset of perching behaviour and assess whether environmental enrichment influences this latency.
2. To examine how different light rhythms and sources (natural vs. artificial) affect perching, feeding, and light preferences in birds.
3. To document the lighting environments on Swedish farms and assess the HATO® system’s compliance with legal welfare requirements.

Methods

Experiment 1: Ninety Lohmann white day-old chicks were assigned to groups of five across 18 pens, each equipped with two wooden perches. The study included three treatments:

• Control (C): no enrichment
• Floor enrichment (F): four wooden blocks (40×10×5 cm) on the floor
• Hanging enrichment (H): two CDs and two plastic 500 ml bottles hanging from the ceiling at chick eye level.

Chick behaviour was observed through direct scan sampling from day 5 to day 40.

Experiment 2: One hundred twenty-six day-old LSL-chicks were grouped into seven individuals across 18 littered pens with two wooden perches. Three lighting treatments were tested:

• 8 hours of incandescent light (A8)
• 16 hours of incandescent light (A16)
• 8 hours of natural light (N8).

Natural light came from windows with clear double glass. Each group was video recorded every third day from 42 to 76 days of age. At 14 weeks, light preferences were assessed in a test pen with compartments illuminated by natural and incandescent light.

A preliminary study was conducted on six commercial farms using HATO®, fluorescent, or incandescent lighting, focusing on production data, light intensity, bird health, and behavioural assessments. Clinical inspections and behavioural studies involved a random sample of 100 birds across various age groups and were scored based on established methodologies.

---

This detailed summary captures the main points from the provided text regarding the influence of lighting on laying hens and outlines the associated studies and methodologies. If you have any specific sections or topics you’d like to delve into further, let me know!

The provided text outlines research findings on the effects of lighting conditions on the behavior and welfare of laying hens, particularly focusing on their perching behavior and feeding patterns. Here’s a summary and analysis of the key points:

23.2.1 Statistics

1. Experiment 1:
   • Individual behavior and the impact of treatment on the start of perching were analyzed using ANOVA.
   • The effect of treatment on learning through social facilitation was analyzed using the Prentice-William-Peterson (PWP) model of survival analysis.
2. Experiment 2:
   • Daytime and nighttime feeding behaviors were modeled using one-way ANOVA.
   • Onset of night perching was analyzed via Cox proportional hazards modeling.
   • Preference for natural versus incandescent light was assessed using linear mixed modeling, accounting for repeated measures.

23.3 Results

• Experiment 1 Findings:
  • Positive correlation between latency to perch and time spent under perches during the first 2 weeks (P=0.01).
  • Negative correlation between latency to perch and time spent under a heating lamp and interactions with other chicks (P<0.01).
  • The treatments did not significantly affect perching latency (P=0.21) or social facilitation (P=0.15), although a slight positive effect of enrichment was noted.
• Experiment 2 Findings:
  • A8 birds had a significantly lower feeding proportion compared to A16 birds (P<0.001).
  • N8 birds had a borderline earlier onset of night perching compared to A8 birds (P=0.056).
  • A8 birds had a strong preference for incandescent light, while N8 birds had a significantly higher probability of choosing natural light (P=0.04).
  • The predicted proportions of birds choosing natural light were 0.36 for N8 and 0.13 for A8.
• On-farm Study:
  • Median flock size was 15,840, with a median mortality of 0.9%.
  • Median laying rate was 93%, and daily feed consumption averaged 120 g per bird.

23.4 Discussion

• Impact of Rearing Conditions:
  • Early behaviors significantly influence later perching behavior and welfare.
  • Lighting conditions affect feeding behavior; longer nighttime periods lead to more daytime feeding.
  • Birds reared in natural light showed earlier perching behavior than those in incandescent light.
• Health Observations:
  • Mild keel bone deviations were observed, increasing with age.
  • Feather loss was prevalent, particularly in aviary birds compared to those in modified cages.
• Feather Pecking:
  • Although feather pecking occurred, its causes were not clearly linked to housing or lighting systems.

23.5 Conclusion

• Early perching behavior was linked to night-time roosting, with no significant effect from environmental enrichment.
• Natural light exposure positively affected the onset of night perching.
• Incandescent light preference was noted in birds reared under such conditions.
• Despite variations in lighting environments across farms, no significant health issues were found, except for feather pecking, which had unclear correlations with housing or lighting systems.

Implications

The findings suggest that rearing conditions, particularly light exposure, significantly influence the behavior and welfare of laying hens. The preference for natural light indicates a potential area for improving the welfare of poultry in production settings. The study also highlights the complexities of feather pecking, suggesting that further research is needed to identify contributing factors beyond lighting and housing systems.

These insights could be valuable for poultry producers aiming to enhance animal welfare and productivity through informed management practices regarding lighting and housing.

Introduction: Lighting Standards to Meet Pig Welfare Guidelines

Light plays a crucial role in regulating various physiological and behavioral processes in animals (Robbins et al., 1984). It influences motivation for behaviors such as foraging and social interaction through the visual system’s stimulation. According to the European Directive (Commission Directive 2001/93/EC), pigs must be kept under a light intensity of at least 40 lx for a minimum of eight hours daily. This regulation seeks to address the common practice of keeping pigs in dim light to reduce fighting and competition. However, it lacks protocols for measuring light intensity and does not clarify why 40 lx is deemed necessary for swine welfare.

Literature presents mixed findings regarding light’s effects on pig behavior and welfare. Some studies suggest that light intensity is not a significant factor influencing pig welfare. For instance, Van Putten (1980) found no evidence that light presence or absence affected pig behavior. However, low light levels have been linked to reduced aggression and tail biting (van Putten, 1968). In contrast, studies show that low light environments can lower aggression among recently grouped pigs (Barnett et al., 1994).

While some studies indicate that supplemental light can improve productivity in breeding sows (Stevenson et al., 1983), the spatial distribution of light intensity in livestock houses has not been sufficiently studied. The present study investigates the light intensity within two compartments for fattening pigs, focusing on varying floor types and ventilation systems under different illumination sources.

---

2. Materials and Methods

2.1 Static Measurements

The study was conducted from May to September 2005 in a piggery in Northern Italy housing fattening pigs for Parma ham production (90-160 kg). The pig house dimensions were 21 m long and 12.5 m wide, featuring concrete block walls 2.05 m high, with ventilation systems and varying light sources.

Each compartment contained eight pens (5.63 m long and 2.63 m wide) with vertical windows for natural light. Compartment 1 had a concrete slatted floor and mechanical ventilation, while Compartment 2 had a solid concrete floor with an external dunging area and natural ventilation.

Light was artificially illuminated from 9:00 am to 5:00 pm in designated areas, while other areas relied solely on natural light. Each pen had four lamps positioned 3 m above the floor.

2.2 Light Intensity Measurements

Static Measurements

Light intensity was continuously recorded using MW8501.7 LSI light sensors, positioned 1.5 m above the floor in the central corridor to monitor daily light intensity. Data collection frequency was set to one minute.

3D Measurements

On two sunny days in September, light intensity was measured at five heights (0, 20, 75, 100, and 150 cm) in six positions within the pens. Measurements were taken in both compartments over two days. Each measurement lasted eight minutes per height for a total of 48 minutes per pen, with animals removed beforehand.

The recorded data were processed using Matlab for visualizing the 3D light distribution. The interpolation method was applied to estimate light intensity across the entire pen area based on the six measured points.

---

Results and Discussion

The findings reveal that while the required 40 lx standard was achieved in artificially illuminated areas, it was frequently unmet in pens relying solely on natural light. The study highlights the importance of adequate light distribution for animal welfare and challenges the effectiveness of current EU regulations in ensuring appropriate lighting conditions in swine housing. The low light levels recorded within the pens may negatively impact the welfare and behavior of the pigs, indicating a need for further research and potential revision of lighting standards in swine husbandry practices.

The data was analyzed using SAS software (2008) to evaluate mean values regarding light exposure in pig housing. A one-way ANOVA was conducted to examine the effects of room type and lighting (artificial vs. natural) on light intensity. Additionally, a frequency analysis (Proc FREQ) was performed to identify periods of light intensity lower or higher than 40 lux from 9:00 am to 5:00 pm. Light distribution was assessed using MATLAB’s graphical ‘patch’ function.

24.3 Results and Discussion

The mean daily light intensities recorded from 9:00 am to 5:00 pm across four building areas are depicted in Figure 24.2. On average, the light levels were significantly higher (P<0.05) in artificially illuminated areas (A and C) at 57±6 lux compared to naturally illuminated areas (B and D) at 40±5.56 lux. The maximum recorded light intensity was also higher in artificially illuminated areas, showing an average of 63±8 lux compared to 42±6 lux in naturally illuminated areas. Notably, area B, solely illuminated by natural light, had the lowest average intensity of 34±2.44 lux, while area D, also naturally lit, averaged 45±4 lux, exceeding area B by 10 lux.

Frequency analysis indicated that light intensity was above or equal to 40 lux for 89% of the time in area A (artificially illuminated), 0% in area B, 97% in area C (artificially illuminated), and 83% in area D.

3D light distribution analyses, presented in Figure 24.3, were performed at three heights: 0 m, 0.75 m, and 1.5 m. These heights corresponded to the resting level of pigs (0 m), their eye level (0.75 m), and human eye level (1.5 m). At 0 m (Figure 24.3a), light intensity approached zero, likely due to the distance of inlets from the floor, which may positively influence pig welfare by providing darkness during rest.

At 0.75 m (Figure 24.3b), the maximum light intensity reached only 25 lux (red color) in areas A and B, reflecting the higher illumination from windows on the west side. Conversely, in areas C and D, with dunging areas, natural light was absent due to structural features. At 1.5 m (Figure 24.3c), human eye level showed higher light intensity, primarily from windows and openings in areas A and B, but values remained low (<27 lux), insufficient for effective inspection of animals.

The analysis revealed that pigs need a minimum of 40 lux for at least 8 hours daily, yet area B failed to meet this standard, as did artificially illuminated areas, though the discrepancy was minimal. Even at human height, only area B did not reach the required light levels (Figure 24.2). In-depth 3D measurements indicated that light intensity was below 40 lux across all areas, with a maximum of only 27 lux.

In a naturally illuminated pen (pen 2 in area B), light penetration was limited despite a window, resulting in high variability in light intensity from 27 to 4.5 lux over a distance of 3 m. Most positions recorded around 9 lux, highlighting that improved window designs could enhance lighting.

Area D had higher natural light levels due to dunging areas and southern exposure, yet artificially illuminated areas (A and C) consistently outperformed naturally lit areas (B and D) in light levels.

24.4 Conclusions

This study measured light intensity in a pig house through static measurements at human height and 3D assessments at ground, animal, and human eye levels. Results showed that the 40 lux standard for pig welfare was met 89% and 97% of the time in artificially lit pens without and with dunging areas, respectively. However, naturally lit areas did not meet the standard, particularly in pens without dunging areas, while pens with dunging areas met it 83% of the time.

Variability in light intensity was linked to building structure, pen positioning, and orientation. Overall, artificially illuminated areas had significantly higher average light intensity than naturally illuminated ones. The study suggests that enhancing window size and placement, alongside effective artificial lighting, could improve overall illumination, benefiting both animal welfare and farmer inspections. Moreover, implementing sensors could optimize artificial lighting, potentially reducing energy costs.

The 3D light distribution analysis under clear skies demonstrated low light intensity across all levels, particularly at floor level (4.5 lux), complicating inspections by farmers or veterinarians. The measured gradients in light intensity indicated substantial non-homogeneity within the building, emphasizing the need for further research to better understand light distribution in pig housing. Additionally, clearer official guidelines on light intensity measurement procedures are necessary for establishing welfare standards.