Cotticelli, Alessio (2024) Precision livestock farming technologies addressing steroid biomarkers associated with animal welfare in modern livestock productive systems. [Tesi di dottorato]

Anteprima

Testo Cotticelli_Alessio_36.pdf Download (2MB) | Anteprima

Abstract

Animal welfare is a debated topic in modern society. Its meaning is a very extensive concept since several definitions have been proposed. A theoretical approach to animal welfare takes into consideration not only the physical and health conditions of animals, but also the psychological well-being and the ability to express species-specific behaviours. Therefore, the welfare status can be assessed and measured through indicators that must consider both the environmental conditions of the farm (management, facilities and climatic parameters) and the adaptive efforts of the animal itself. In recent years, animal care professionals working in managed settings have focused on identifying effective measures for systematically monitoring and assessing welfare. Different indicators can be considered: direct (or animal based), such as behavioural, physiological, pathological and productive parameters, and indirect (or environmental factors), as farming structures and systems, management and human-animal relationships. Among the animal-based measures, several studies have been carried out on non-invasive physiological biomarkers to gain insight into an animal’s physical condition, psychological health and overall welfare status. For this reason, glucocorticoids (especially cortisol), are frequently used to evaluate the physiological response to stress, since they provide information about the activity of the hypothalamic–pituitary–adrenal (HPA) axis. Recent research has incorporated other biomarkers of HPA activity, namely dehydroepiandrosterone (DHEA) and its sulphate ester (DHEA-S). Despite the extensive literature concerning steroid biomarkers, conflicting results have been reported about the HPA axis functioning at different phases of productive career of farm animals, depending on factors such as inflammatory and reproductive status, chronicity of exposure to stressors, and so on. However, as stated above, a correct evaluation of welfare cannot leave an accurate monitoring of the animal environment out of consideration. In this view, recent studies focused on precision livestock farming (PLF), that can be defined as “the management of livestock by continuous automated real-time monitoring of production/reproduction, health and welfare of livestock, and environmental impact”. PLF technologies consist of sensors at animal level (accelerometers, Radio-Frequency Identification, (RFID), rumen boluses, temperature and pH sensors, etc.) and in the environment (cameras, temperature loggers, gas sensors, microphones, etc.). The specific applications of these tools can help the decision-making processes by providing early detection of health or welfare problems in individual animals and the application of targeted corrective practices. Therefore, the potential of PLF technologies to address livestock welfare is promising. The aim of this thesis was to study cortisol and DHEA(S) as biomarkers of the HPA functioning and provide essential evidence for the practical purpose of targeting PLF technologies towards an animal welfare improvement within the modern livestock productive systems. The integrative approach between PLF and endocrinological measurements has been used in different species and across some sensitive phases of their productive career. In the first experiment, a radioimmunoassay (RIA) method for cortisol in buffalo milk was validated. Three formulations of milk and three solvents were tested: whey cortisol concentrations showed a significant correlation with whole extracted (methanol) milk and were not affected by fat content variation during the milking session. It was concluded that the RIA suited the cortisol measurement in buffalo milk and the ranges could be employed in the calibration of a biosensing method for non-invasive assessment of cortisol directly integrated in milking parlour systems. The second experiment aimed at studying the relationship of cortisol in blood, milk, whey and hair with parity, lactation stage and productive level in dairy buffaloes and to study their predictive potential. Multiparous (n = 30) and primiparous (n = 38) buffaloes were assigned to 4 productive classes and 3 lactation stages and cortisol concentrations were measured using an in-house RIA method. Parity did not show a significant effect on cortisol concentrations of the four media, conversely the stage of lactation largely influenced cortisol concentrations in all matrices. Moreover, hair cortisol concentrations were negatively correlated to mature equivalent milk yield, mature equivalent protein content and mature equivalent corrected milk. Finally, milk had the predictive potential to estimate cortisol levels in other matrices. Through the third experiment the hair concentrations of cortisol, DHEA, DHEA-S and their ratios in dairy calves during postnatal and postweaning periods were investigated. Hair sampling was conducted on healthy dairy calves at the ages of 64.8±0.65 days (postnatal) and at 155.3±0.85 (postweaning) days. Hair cortisol concentrations were higher during the postnatal compared to the postweaning period. Similarly, the cortisol:DHEA and cortisol:DHEA-S ratios were higher in the first period, showing a higher animal allostatic load at birth. A reduction of the allostatic load of the calves was demonstrated by the reduction in hair cortisol concentrations at 5 months compared to those measured at 60 days, as well as by the significant reduction in the cortisol:DHEA and cortisol:DHEA-S ratios. The fourth experiment investigated testicular ultrasonography and steroid concentrations (cortisol, DHEA-S, cortisol/DHEA-S ratio, testosterone) in hair for their utility in the bull breeding soundness evaluation (BBSE). Beef and dairy bulls underwent routine semen collection twice weekly for 12 weeks. Ultrasonography through a B-mode ultrasound scanner equipped with a linear array probe and hair sampling for steroid RIA were performed at the last semen collection. Semen was analysed immediately after collection and post-thawing. Bulls with homogeneous parenchyma had a higher (p < 0.05) percentage of motile sperm post-thawing and hair DHEA-S was positively related to motile sperm, progressively motile sperm and motility yield. It was concluded that the inclusion of testicular ultrasonography and hair DHEA-S in the standard BBSE could provide a more integrated and comprehensive assessment of fertility in bulls. Cortisol, DHEA and their ratio in hair were also evaluated as biomarkers of allostatic load and resilience in 296 pregnant sows around farrowing in the fifth experiment. Four different models of farrowing crates and four different batches were considered. Each sow was sampled for the first time (ST1) 2.6 ± 1.6 (mean ± SE) days before parturition and for a second time (re-growth hair, ST2) at 88.9 ± 3.3 days after parturition. Thus, the sows were allocated to the February, March, April, May, or June batch. No differences in terms of hair steroids concentration were found between the four models of farrowing crates and between the sampling times (P > 0.05). A significant interaction emerged between the batch and the sampling time for all the biomarkers considered. In particular, ST1 showed always higher HC/HDHEA ratios than ST2. Thus, the pregnancy period in collective pens was more challenging than the early postpartum and lactation in individual crates. The sixth experiment aimed at assessing the possible association between the maternal concentrations of plasma progesterone (P4) and cortisol and the number of fetuses in Teramana goats. Twenty- four pregnant does were enrolled in the study. Two to one week before the expected date of parturition, each doe was submitted to blood sampling. Plasma cortisol and P4 concentrations were determined by RIA. At birth, the number of kids for each doe was recorded and does were retrospectively grouped based on the number of fetuses (single, twin or triplet). Three (13%) does delivered single kids, 16 (69.6%) twins, and 4 (17.4%) triplets. Does bearing triple fetuses showed significant higher concentration of cortisol in comparison to does with single pregnancy. On the contrary, P4 concentrations did not differ between does bearing different number of fetuses. Significant positive correlations emerged between plasma P4 and cortisol concentrations and the number of fetuses, and between the two hormones. It was concluded that the single measurement of cortisol one week before the expected parturition might be useful to distinguish between does bearing singleton and triplet pregnancies. The seventh experiment focused on the management of weaning at individual level by investigating how PLF technologies can be used to monitor individual dry matter intake. A method was proposed using a 3D depth camera and a proper algorithm to measure the volume and weight of eaten feed. To preliminarily assess the feasibility of the proposed method, a suitable measurement setup was implemented in laboratory conditions, using a 3D Depth camera (RealsenseTM D455, Intel©). A dedicated MATLABTM code was implemented to control the camera and retrieve the distances measured in the framed scene. The MATLAB function ginput() was exploited to graphically determine the extent of the considered regions according to shown distance image. The 3D camera was positioned at 65 cm from the reference plane whose distance could be modified and controlled thanks to a hand crank. The volume of feed was associated with MATLAB reconstructions, showing differences lower than 2%. The eighth experiment studied the use of an RFID-linked walk-over-weighing (WoW) system in a pastoral sheep production system to predict future liveweight (LW) of sheep with different lead times. The experiment lasted 94 days, the flock consisted of 144 Merino and crossbred lambs (White Suffolk × Merino) of both genders. Each animal was tagged with an electronic RFID ear tag and the remote walk-over-weighing station was installed at the entrance of a yard enclosing the single water source to record the liveweight of animals each time they came across. The station was configured to allow free flow of animals. Growth rate was calculated as the first derivative throughout the predicted LW curve. The future predicted LW (PW) was calculated daily on the 20, 30, 40, 50, and 60 days ahead of any actual day throughout the trial by multiplying each animal’s actual growth rate by the target days and adding the actual observed LW (OW). The accuracy of the weight predictions was assessed using a linear mixed-effects and Lin’s concordance correlation coefficient (CCC). As expected, the accuracy and the precision of the PWs calculated showed a decreasing trend. The concordance correlations coefficients showed an overall agreement between the PWs and the OWs, with CCCs ranging from 0.692 and 0.967. The same trend emerged by studying the PWs and the OWs through multiple linear regressions. It was concluded that the WoW system allowed to record LW of individual sheep daily.

Downloads

Actions (login required)

Modifica documento