Evaluating cow welfare status from milk samples: effects of housing modifications on milk infrared spectra

Evaluating cow welfare status from milk samples: effects of housing modifications on milk infrared spectra | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Evaluating cow welfare status from milk samples: effects of housing modifications on milk infrared spectra M. Bahadi, D. Warner, A. A. Ismail, D. E. Santschi, D. M. Lefebvre, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4919745/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted 13 You are reading this latest preprint version Abstract The goal of this study was to isolate spectral fingerprints from milk Fourier transform infrared spectra that may reflect potential improvements in cow welfare, specifically comfort and ease of movement, resulting from modified housing configurations. Housing configuration modification treatments were tested across 3 animal trials, consisting of modified chain length (TCL) , stall width (SW) and manger wall and stall length ( MW/SL) configurations. The spectral analyses involved the use of principal components and mixed model analysis. Principal components were calculated from averages of mid-infrared spectra collected on the last weeks of treatment application in each of the animal trials. A significant effect of housing configuration was revealed. As an indication of animal comfort improvement, milk of cows assigned to longer chains revealed a trend of changes in multiple milk components (e.g., milk NPN, trans fatty acids, fat, and protein) that are consistent with changes in ruminal pH. These conclusions were inline with those drawn from the analysis of animal-based responses such as behavioral data and other outcomes. This study was able to reveal that housing modifications had a significant effect on milk spectra, with differences observed between the most and least restrictive treatments, translating into improved or reduced animal welfare status. Biological sciences/Biological techniques/Spectroscopy Biological sciences/Biological techniques/Spectroscopy/Near infrared spectroscopy Figures Figure 1 Figure 2 Figure 3 INTRODUCTION Animal welfare is increasingly becoming a concern for dairy producers and society in general; hence, more research has been conducted to improve the different aspects of animal welfare [ 1 ]. According to the World Organization of Animal Health, an animal that experiences good welfare is one that is “healthy, comfortable, well nourished, safe, is not suffering from unpleasant states such as pain, fear and distress, and is able to express behaviours that are important for its physical and mental state.” [ 2 ]. Based on this definition, the tie stall housing system of dairy cows has been criticized for its restrictions that it imposes on the cows’ ability to move and to engage with social interactions with other cows [ 3 ], and for higher prevalence of lameness and cow comfort issues, which negatively affects cow welfare, public perceptions, and producer profitability [ 4 ]. Nevertheless, the tie stall system remains a prevailing housing system for dairy cows in North America. For example, in Canada it accounts for 73.8% of dairy operations in farms [ 3 ]. The current study is part of a comprehensive study conducted over a 3-yr period aimed at investigating the individual effect of different stall design aspects on ease of movement and comfort of dairy cows housed in tie stalls. The studied elements of the stall design that were covered by this comprehensive study were the tie-rail height and forward position [ 5 ], the chain length (TCL) [ 3 ], the stall width (SW) [ 6 ] and the combined effect of stall length and manger height ( MW/SL) [ 7 ]. To assess the effects of modifying those elements on cow ability to move and comfort, multiple outcomes were measured during each trial including injury scores [ 5 , 7 ], resting behaviour [ 3 , 5 , 6 , 7 ], lying down and rising events [ 3 , 5 , 6 , 7 ], tracking cow’s movement in the stall [ 3 ] and postures and position in space of head, body and limbs during lying hours [ 6 ]. While these outcomes remain the accepted measures to evaluate animal comfort and welfare in dairy housing systems [ 4 ], we hypothesize that cows’ comfort improvement might lead to physiological changes, which might be reflected in milk chemical composition. However, affected milk components might not necessarily be those that are currently reported in dairy herd improvement (DHI) programs; hence, we decided to mine milk infrared (IR) spectra to study the changes in milk composition that can be attributed to changes in the housing environment. Milk IR spectra contain signals from all molecules present in milk that can absorb IR energy; therefore, milk IR spectra represent a comprehensive snapshot of the chemical composition of a milk sample. In a previous publication [ 8 ], we presented a novel hybrid approach for spectral analysis that combined mixed modeling and multivariate analysis of milk Fourier transform infrared ( FTIR ) spectra in which we attributed changes in milk composition to modifications to the tie-rail of a tie stall described elsewhere [ 5 ]. In this approach, we considered principal component analysis (PCA) scores as a proxy value for the concentrations of multiple milk components since they are calculated from multivariate measurements (i.e., milk FTIR spectra) that contain spectral contributions from the concentrations of multiple milk components, and we tested them for a housing treatment effect (i.e., different tie-rail configurations). The analysis pinpointed a tie-rail configuration treatment among multiple ones whose milk samples’ scores (i.e., principal component 7 scores) were significantly different from the scores of milk samples of other treatments. The loadings of that principal component revealed an inverse relationship between lactose and energy metabolism related molecules, such as b-hydroxy butyrate (BHB) , acetone and citrate, which suggested that cows in that tie-rail configuration treatment were experiencing higher level of body fat mobilization. Cows enrolled in that tie-rail configuration treatment also recorded increased injuries on two locations on the cow’s neck [ 5 ] and those injuries might have been obstructing the cows from accessing feed. We concluded that analyzing milk FTIR spectra directly and without relying on predicted values of any analyte of interest was a viable option in capturing changes in milk composition that can be attributed to modifications in housing conditions. The objective of this paper is to determine potential changes in milk components that can be attributed to changes in TCL, SW and MW/SL configurations by applying the hybrid spectral analysis approach [ 8 ] to milk spectra collected in those trials. More specifically, this approach will be evaluated for its potential to capture a milk spectral fingerprint of cows’ comfort that can later be used to predict a score for cow comfort and welfare from milk composition data or spectra, which is the long-term objective of this work. It must be noted that none of the treatments in these trials were designed to decrease cow comfort in the tie stall. On the contrary, all suggested treatments in these trials were designed to increase cows’ ease of movement; therefore, increasing their level of comfort and welfare. As a results, we expect changes to milk composition to be subtle. Routine animal welfare outcomes were used to categorize the level of animal comfort and ease of movement (i.e., improved or not) provided by each housing treatment (fully reported elsewhere [ 6 , 7 ]) and key findings will be used in this manuscript to support any trends detected by the hybrid analysis approach of spectral data. According to our knowledge, no previous work was reported in the literature in which milk FTIR spectra were used in designed trials to assess the level of animal comfort and ease of movement in housing environment. MATERIALS AND METHODS Experimental Setup A series of trials have been conducted to test different housing treatments intended to improve the level of comfort and welfare of individual cows by increasing the opportunity of movement given to the cow in her stall. The animal trials were conducted at the Macdonald Campus Dairy Complex of McGill University (Sainte-Anne-de-Bellevue, QC, Canada) using animals from the dairy herd. Details on the experimental setup and animal handling can be found elsewhere [ 3 , 6 , 7 ]. Use of animals and all procedures were approved by the Animal Care Committee of McGill University and affiliated hospitals and research institutes (protocol #2016–7794), and all experiments were performed in accordance with relevant guidelines and regulations. In this section, the experimental setups of the 3 trials are briefly described. An overview of the housing configuration treatments is shown in Table 1 . From hereafter, T1 will be referring to the control configuration of the tie stall element and T2 will be referring to the suggested modification to that element. Table 1 Housing configuration treatments across each of the 4 trials Trial Housing configuration Treatments Experimental design Week Cows Control (T1) Suggested modification (T2) TCL Chain length 1.0 m 1.4 m Randomized block design 10 24 SW Stall width Single 1 Double 2 Randomized block design 6 16 MW/SL Manger wall and stall length 20 cm 3 5 cm 3 Cross over 2 periods, 1 week adaptation, 6 weeks treatment 24 1 Single: average width of 139 cm (determined as 2 x (width of cow at hips) + 5 cm), 2 Double: average of 284 cm (determined as 2 x Single), 3 Both treatments were assigned to two stall lengths: short stall length 178 cm (L1) and long stall length 188 cm (L2) Trial 1: Chain length configuration Twenty-four lactating Holstein cows were assigned to 2 tie chain length (TCL) treatments. T1 was the control, and T2 was a suggested treatment intended to increase the cow’s movement ability in her stall [ 3 ] (Table 1 ). Cows were assigned to 12 different blocks of two cows to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation (average DIM 129) and were placed evenly in two rows facing a wall within the barn. Five cows (4 long chain, 1 control) were removed at different points (1 in week 6, 1 in week 8, 2 in week 9, and 1 in week 10) during the trial for different reasons [ 3 ]. The trial lasted for 10 weeks from February 20, 2017, to May 1, 2017. Trial 2: Stall width configuration Sixteen lactating Holstein cows were assigned to 2 stall width (SW) treatments. T1 was the control and T2 was a suggested treatment intended to increase the cow’s comfort, and specifically cow’s rest, in her stall [ 6 ] (Table 1 ). Cows were assigned to 8 blocks of 2 cows to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation (average DIM: 157) and were placed evenly in two rows within the barn. The trial lasted for 6 weeks from June 5, 2017, to July 17, 2017. Trial 3: Manger wall and stall length configuration The lengths of two rows in the barn were modified as follows: row 1 was 178 cm (i.e., short or L1), which is the length commonly found on Quebec’s farms, and row 4 was 188 cm (i.e., long or L2), which was a suggested modification. Two manger wall heights were applied randomly to stalls within each row: T1, which was the upper limit of the recommendation, and T2 (Table 1 ). Twenty-four cows were randomly divided into 4 groups. Two groups were assigned to each row and subjected to both manger wall treatments in a crossover design (1 week for adaptation, 6 weeks of data collection per treatment). Cows were assigned to 6 different blocks to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation. Increasing the stall length and reducing the manger wall increased the space available for the cow, which eased her movement and lying, thus reducing her susceptibility to injuries [ 7 ]. The trial lasted for 14 weeks from February 26, 2018, to June 4, 2018. All 24 cows enrolled in the study remained in the final data set for period 1, but 3 were removed from period 2 [ 7 ]. This trial will be referred to as MW/SL in this paper. Milk Component Analysis One composite milk sample per week was collected from each cow participating in the trials. The sample consisted of equal volumes of milk collected during the evening milking and the morning milking of the next day. All collected milk samples were analyzed for milk composition by FTIR spectroscopy at the Lactanet laboratory (Sainte-Anne-de-Bellevue, QC, Canada) using the same CombiFoss FT + analyzer (FOSS, Hillerød, Denmark). The concentrations of fat, protein (i.e., crude protein), lactose, urea nitrogen, and BHB, the fat-to-protein ratio, and somatic cell count (SCC; log transformed) were determined (Supplementary Table S 1). The milk component concentration data were analyzed to determine whether they differed among housing configurations within each trial. For the TCL trial, the treatment short-term effect was tested on milk components’ concentrations averages per cow that were calculated from samples collected in weeks 1,2 and 3. The treatment long-term effect was tested on milk components’ concentrations averages per cow that included samples collected in weeks 8, 9 and 10. For the SW and MW/SL trials, the treatment short-term and long-term effects were tested on samples that were collected from week 1 and from week 6, respectively. Data analysis was done in R (version 4.1; R Foundation for Statistical Computing, Vienna, Austria) using the add-on package lmerTest [ 9 ]. Marginal means were estimated for treatment groups using the add-on package emmeans [ 10 ]. Spectral Analysis Milk FTIR Spectra, Outliers Check and Spectral Pre-treatments For each milk sample, a full FTIR spectrum was collected, which contained 1060 spectral variables (i.e., wavenumbers) between 5008 and 925 cm − 1 . Principal component analysis ( PCA ) in JMP Pro 13.2.1 (SAS Institute, Cary, NC, USA) was applied to the collected spectra to detect outliers and none were observed in the PCA scores plot (PC1 vs. PC2). Only spectral regions containing information related to milk composition were retained for spectral analysis, namely, 1612 − 925 cm − 1 , 1797 − 1681 cm − 1 , and 3061 − 2803 cm − 1 [ 11 , 12 , 13 ]. The total number of spectral variables that were retained for analysis was 278 wavenumbers. First derivative with a derivative window of 1 and vector normalization [ 14 ] were applied, separately or combined, as pre-treatments to milk spectra using codes written in MATLAB R2018a (MathWorks, Natick, MA, USA). As a result, four sets of milk FTIR spectra were obtained for each trial: raw, vector normalized raw ( VN raw), first derivative ( FD ), and vector normalized first derivative ( VN-FD ). For the TCL trial, short-term treatment application average spectra were calculated for each cow from spectra of samples collected in weeks 1,2 and 3. Long-term treatment application average spectra were calculated for each cow from spectra of samples collected in weeks 8,9,10. These averages were calculated for the raw, VN-raw, FD and VN-FD spectral data sets. For the SW and MW/SL trials, spectra of weeks 1 and 6 were used as short-term and long-term treatment application spectra, respectively, for data analysis. Hybrid Spectral Analysis Approach The effect of housing configuration treatments on milk FTIR spectra was evaluated by combining multivariate analysis with mixed modeling detailed in a previous publication [ 8 ], 2021). In short, principal components (PC) were calculated by applying PCA in JMP Pro 13.2.1 to the four versions (i.e., raw, FD, VN-raw, VN-FD) of the short-term and long-term treatment application spectra or spectral averages. Only PC with eigenvalue equal to or greater than 1 and that explained at least 1% of the variance in the respective spectral dataset were retained for further analysis. This step retained informative PC and discarded noisy ones. Scores of the retained PC were used as an input [ 15 ] to test for the treatment effect by PROC MIXED procedure in SAS 9.4 (SAS Institute, Cary, NC, USA) (see model description below). If a significant treatment effect is revealed at P ≤ 0.05, then the least-squares means of its scores were examined to determine the treatment levels that were significantly different from the other levels using a Scheffé adjustment. If PC was calculated from raw spectral dataset, then the influential spectral features could be directly interpreted from the PC’s loading spectrum. If PC was obtained from FD spectral datasets, then the spectral integral of the PC’s loading spectrum was calculated before interpretation of the influential spectral features. The cumulative trapezoidal numerical integration function in MATLAB R2018a (MathWorks, Natick, MA, USA) was used to calculate the spectral integral of the loading spectrum in question. A peak-fitting procedure in the Peak Resolve feature in OMNIC™ 7.3 (Thermo Fisher Scientific, Waltham, MA, USA) was applied to regions in the integrated loading spectrum where no clear peaks were present. The Voigt function [ 16 ] with low or high sensitivity was used and the baseline correction was set to none. The noise and the full width at half height of the narrowest peak in the region of interest were determined by the software. The fitting process was repeated several times until an acceptable residual spectrum was obtained. Statistical Analysis PC scores calculated from milk spectra were analyzed separately to test for the treatment effect in each trial using the models outlined below using the PROC MIXED procedure in SAS 9.4 (SAS Institute, Cary, NC, USA). For the TCL trial \({Y_{ijk}}=\mu +tr{t_i}+ro{w_j}+bloc{k_k}+{e_{ijk}}\) where \(\:{Y}_{ijk}\) is the dependent variable, the PC scores isolated from spectra of milk samples from the \(\:{k}^{th}\) block (parity and lactation stage) in the \(\:{j}^{th}\) row on the \(\:{i}^{th}\) chain length, \(\:{trt}_{i}\) was the fixed effect of the \(\:{i}^{th}\) TCL treatment, \(\:{block}_{k}\) was the fixed effect of the \(\:{k}^{th}\) parity and lactation stage combination, \(\:{row}_{j}\) was the random effect of the \(\:{j}^{th}\) row in the barn and \(\:{e}_{ijk}\) was the random residual error. For the SW trial \({Y_{ijk}}=\mu +tr{t_i}+ro{w_j}+bloc{k_k}+{e_{ijk}}\) where \(\:{Y}_{ijk}\) is the dependent variable, the PC scores isolated from spectra of milk samples from the \(\:{k}^{th}\) block (parity and lactation stage) in the \(\:{j}^{th}\) row on the \(\:{i}^{th}\) stall width, \(\:{trt}_{i}\) was the fixed effect of the \(\:{i}^{th}\) SW treatment, \(\:{block}_{k}\) was the fixed effect of the \(\:{k}^{th}\) parity and lactation stage combination, \(\:{row}_{j}\) was the random effect of the \(\:{j}^{th}\) row in the barn and \(\:{e}_{ijk}\) was the random residual error. For the MW/SL trial \({Y_{ijkmnq}}=\mu +lengt{h_i}+se{q_{ij}}+bloc{k_k}+co{w_{ijkm}}+perio{d_n}+tr{t_{iq}}+{e_{ijkmnq}}\) where \(\:{Y}_{ijkmnq}\) is the dependent variable, the PC scores isolated from spectra of milk samples from the \(\:{m}^{th}\) cow of the \(\:{k}^{th}\) block in the \(\:{j}^{th}\) sequence of the \(\:{i}^{th}\) length, the \(\:{q}^{th}\) treatment of the \(\:{i}^{th}\) length, and the \(\:{n}^{th}\) period, \(\:{length}_{i}\) was the fixed effect of the \(\:{i}^{th}\) stall bed length, \(\:{seq}_{ij}\) was the fixed effect of the \(\:{j}^{th}\) sequence on the \(\:{i}^{th}\) stall bed length, \(\:{block}_{k}\) was the fixed effect of the \(\:{k}^{th}\) parity and stage of lactation combination, \(\:{cow}_{ijkm}\) was the random effect of the cow from the \(\:{k}^{th}\) block on the \(\:{j}^{th}\) sequence of the \(\:{i}^{th}\) stall bed length, \(\:{period}_{n}\) was the fixed effect of the \(\:{n}^{th}\) period, \(\:{trt}_{iq}\) was the fixed effect of the \(\:{q}^{th}\) manger wall height treatment on the \(\:{i}^{th}\) stall bed length treatment and \(\:{e}_{ijkmnq}\) was the random residual error. The significance level was defined at P ≤ 0.05 for all trials. RESULTS AND DISCUSSION A first analysis of milk component concentrations predicted through FTIR spectroscopic analysis of milk samples did not reveal any significant difference between housing configurations within each trial (Supplementary Table S 2). These findings suggest that analysis of major milk components (i.e., fat, protein, and lactose) individually may not be sufficient to reflect an impact of cow comfort and ease of movement on milk composition through changing housing configurations. In contrast, when a more comprehensive analysis of the spectra was conducted by employing PCA, significant housing treatment effects for the TCL ( P = 0.032) and SW ( P = 0.042) trials were revealed in PC6 and 5, respectively (Table 2 ). For the TCL trial, PC6 was extracted from the VN-FD long-term treatment application spectral averages datasets, and it described 1.70% of the variation. This PC did not reveal significant effects for the other studied factors that were included in the statistical model. This observation suggests that the TCL might have influenced milk composition during the trial. For the SW trial, PC5 was extracted from the VN-raw long-term treatment application spectral dataset, and it described 1.13% of the variation in that dataset. In addition to the treatment effect, this PC revealed a strong block effect ( P < 0.001); for this reason, the interpretation of the loading spectrum for this PC will take into consideration all factors accounted for by blocking, which were cow parity (i.e., number of lactations) and days in milk in the current lactation (DIM). For the MW/SL trial, the manger wall height treatment and its combined effect with the stall length did not reveal any significant treatment effect, which means that the manger wall treatment did not affect milk composition during the trial. However, PC6 extracted from the long-term treatment application FD spectral dataset revealed significant length effect ( P = 0.036), which means that the stall length might have had a significant effect on milk composition during the trial. This PC described 1.77% of the variation in its respective dataset. In the following sections, we will refer to the short and long stalls as L1 and L2, respectively, to distinguish them from the two treatments (i.e., T1 and T2), which represent the combined effect of the stall length and manger wall height. Table 2 Principal components extracted from long-term spectral datasets that revealed significant effects in the four trials. The table also lists P values obtained from the SAS PROC MIXED Procedure for tested effects in each trial Trials 1 Spectral Dataset 2 PC 3 Eigenvalue Explained Variation % P Values TCL VN FD 6 4.75 1.70 Trt 4 Block 0.032 0.088 SW VN Raw 5 3.13 1.13 Trt 4 Block 0.042 < 0.001 MW/SL FD 6 4.92 1.77 length Seq Block Trt 4 Period 0.036 0.038 0.375 0.276 0.912 1 TCL = tie chain length trial, SW = stall width trial, MW/SL = manger wall/stall length trial, 2 FD = first derivative, VN = vector normalized, 3 PC = principal component, 4 Trt = treatment Table 3 summarizes the differences in the least-squares means produced by the PROC MIXED procedure for the scores of PC6, 5 and 6, which revealed a significant treatment effect. For the TCL, SW and MW/SL trials, there were significant differences in milk samples’ scores for T1 vs. T2 ( P = 0.032), T1 vs. T2 ( P = 0.042) and L1 vs. L2 ( P = 0.036), respectively. Table 3 Differences of least squares means for the scores of the principal components that revealed a significant treatment effect for each trial Trial 1 Treatment Treatment Estimate Standard Error DF 2 t Value P Value Scheffé Adj. P Value TCL T1 T2 -1.6819 0.6650 9 -2.53 0.032 0.032 SW T1 T2 0.6021 0.2224 5.01 2.71 0.042 0.042 MW/SL L1 L2 -1.3150 0.5998 33 -2.19 0.036 0.036 1 TCL = tie chain length trial, SW = stall width trial, MW/SL = manger wall/stall length trial, 2 Degrees of freedom Interpretation of loading spectra of principal components that revealed significant housing configuration treatment effect on milk composition Trial 1: Chain length configuration Inspection of the integral of the PC6 loading spectrum (Fig. 1 ) revealed peaks at the following wavenumbers: 2919, 2851, 1715, 1576, 1541, 1461, 1419 and 968 cm − 1 . The peaks at 2919, 2851, 1715, 1541, 1419 and 968 cm − 1 can be assigned to the following vibrational modes: the asymmetrical stretching vibration ( \(\:{\nu\:}_{as}{CH}_{2}\) ) of methylene groups in fatty acids [ 17 ], the symmetrical stretching vibration ( \(\:{\nu\:}_{s}{CH}_{2}\) ) of methylene groups in fatty acids [ 17 ] and histamine [ 18 ], the \(\:C=O\) stretching vibration of the carboxyl functional group in free fatty acids [ 17 ], the deformation vibration \(\:\left(\delta\:{NH}_{2}\right)\) of the amidine functional group in creatine [ 19 ], the symmetrical \(\:C-O\) stretching vibration of the carboxylate ion in acetate, and the \(\:C-H\) bending vibration of the trans double bond in trans fatty acids [ 17 ], respectively. The peak at 1576 cm −1 may be assigned to: the ring stretching vibration of the imidazole ring in histamine [ 18 ], the \(\:N-H\) bending vibration ( \(\:\delta\:NH\) ) of hippuric acid [ 20 ] or the symmetrical \(\:C-O\) stretching vibration of the carboxylate ion in citrate and deprotonated fatty acids [ 17 ]. The peak at 1461 cm − 1 may be assigned to: the \(\:C-N\) stretching vibration of urea [ 17 ], the \(\:N-H\) bending vibration of the \(\:{NH}_{4}^{+}\) ion or the \(\:{NH}_{3}^{+}\) symmetrical bending vibration \(\:\left({\delta\:}_{s}{NH}_{3}^{+}\right)\) of histamine [ 18 ]. In addition, negative loadings were observed at 1748 cm − 1 , which can be assigned to the \(\:C=O\) stretching vibration of the ester linkages in milk fat [ 17 ]. The peak fitting process for the region 1365 − 1160 cm − 1 revealed peaks at 1347, 1299 and 1249 cm − 1 , which can be assigned to the \(\:{CH}_{3}\) deformation vibration [ 21 ], which appears in the FTIR spectrum of an aqueous solution of acetate, the bending vibration ( \(\:\delta\:{CH}_{2}\) ) of the methylene group in hippuric acid [ 20 ] and the \(\:C-O\) stretching vibration in citrate [ 17 ], respectively. These assignments are consistent with the results of the spiking experiments reported elsewhere [ 8 ]. The result of the spectral analysis suggests that the average FTIR spectra of milk samples collected from cows enrolled in both TCL treatments had significant differences in the last 3 weeks of the trial (i.e., long-term effect of chain length configuration treatment application). The loading spectrum of the principal component that revealed the significant treatment effect showed positive loadings for spectral features that might be assigned to trans fatty acids, acetate, citrate, and non-protein nitrogen (NPN) compounds, including histamine. These positive loadings suggest a direct relationship among these spectral features; hence, they might have been similarly affected by the TCL treatment. Accounting for 35–48% of milk NPN [ 22 ], milk urea averages that were determined by the FTIR milk analyzer were 12.28 mg/dL and 12.14 mg/dL for T1 and T2, respectively, during the last 3 weeks of the trial. Considering that milk urea average in T1 is greater than that of T2 and that spectral features related to urea and other minor milk components (e.g., other milk NPN and trans fatty acids) are showing loadings in the same direction, we could conclude that the concentrations of the above-mentioned minor milk components are following the same trend as urea. Increased trans fatty acids and NPN in milk, and histamine specifically, have previously been reported as markers of decreased pH in the rumen [ 23 , 24 , 25 ]. In addition, negative loadings were observed at a spectral feature that can be assigned to milk fat, which suggests an inverse relationship between milk fat and the previously mentioned minor milk components. In fact, decreased milk fat [ 24 ], decreased milk protein and increased milk NPN [ 23 ] were reported in cases of reduced ruminal pH. Indeed, we noticed a decrease in the average concentrations of protein and milk fat for cows enrolled in T1 in comparison to T2 during the last 3 weeks of the trial. The averages of protein were 3.28% and 3.34% for T1 and T2, respectively, and the averages of fat were 3.86% and 3.90% for T1 and T2, respectively, for that period. The observed trend in changes of multiple milk components (e.g., milk NPN, trans fatty acids, fat, and protein) in this trial is consistent with changes that have been reported in the literature in cases of changing ruminal pH and they suggest that cows with longer chains (T2) might have more stable ruminal pH, which was not measured for cows enrolled in this trial, during the last three weeks of the trial. Trial 2: Stall width configuration Inspection of the PC5 loading spectrum (Fig. 2 ) revealed positive loadings at 1721 cm − 1 and in the region ~ 1250 − 1140 cm − 1 centered at ~ 1205 cm − 1 . Positive loadings were also observed around 2920 and 2855 cm − 1 and within the region ~ 1565 − 1520 cm − 1 . Milk fat is predominantly composed of triglycerides, which are tri-esters derived from glycerol and three fatty acids, and the bands at 2920, 2855 and 1721 cm − 1 can be assigned to the asymmetrical stretching vibration ( \(\:{\nu\:}_{as}{CH}_{2}\) ) of the methylene groups in fatty acids of milk fat [ 17 ], the symmetrical stretching vibration ( \(\:{\nu\:}_{s}{CH}_{2}\) ) of the methylene groups in fatty acids of milk fat [ 17 ] and the \(\:C=O\) stretching vibration of the ester linkages in milk fat [ 17 ], respectively. The methylene twisting and wagging vibrations of fatty acids and esters occur in the region between ~ 1250 and ~ 1140 cm −1 [ 17 ], and the \(\:C-O\) stretching vibration of esters shows strong absorption in the 1210 − 1163 cm − 1 region [ 17 ]. In addition, bands in the spectral region between 1565 cm − 1 and 1520 cm − 1 can be assigned to the Amide II band of milk proteins [ 26 ]. PC5 scores of milk samples collected during the last week of the trial (i.e., week 6, long-term effect of stall width configuration treatment application), differed mainly in spectral features related to fat and protein, which suggests a trend that can be confirmed by the analysis results reported by the FTIR milk analyzer. The averages of fat were 3.96% and 3.76% and the averages of protein were 3.34% and 3.28% for T1 and T2, respectively. This observation suggests that cows assigned to T1 had higher milk fat and milk proteins during the last week of the trial. Trial 3: Manger wall and stall length configuration Inspection of the spectral integral of the PC6 loading spectrum (Fig. 3 ) revealed positive loadings at the following wavenumbers: 1575, 1460, 1408, 1033, 980 and 946 cm − 1 . The peak at ~ 1575 cm − 1 is observed in the FTIR spectrum of histamine in aqueous solution [ 8 ] and can be assigned to the stretching vibration of the imidazole ring [ 18 ]. The peak at ~ 1460 cm − 1 is observed in the FTIR spectra of milk samples spiked with urea and ammonium chloride [ 8 ]; hence, it can be assigned to the \(\:C-N\) stretching vibration in urea [ 17 ] or to the \(\:N-H\) bending vibration in the ammonium ion [ 17 ]. The peak at ~ 1408 cm − 1 is observed in milk samples spiked with BHB [ 8 ] and can be assigned to the \(\:C-O-H\) bending vibration or the symmetrical \(\:C-O\) stretching vibration of the carboxylate ion [ 17 ]. The peaks at ~ 1033 and 980 cm − 1 are observed in the FTIR spectra of milk samples spiked with histamine [ 8 ]. The peak at ~ 946 cm − 1 is observed in FTIR spectra of aqueous solutions of histamine and BHB [ 8 ]. The spectral analysis suggested that the stall length might have had an effect on BHB, urea and other milk NPN including histamine during the last week of the trial (i.e., long-term effect of manger wall and stall length configuration treatment application). Milk composition determined by the FTIR milk analyzer revealed that milk samples collected from cows assigned to L1 (i.e., the short length row) had higher levels of BHB and urea during the mentioned period. During week 6, the averages for period 1 of BHB content were 0.06 mmol/L and 0.05 mmol/L for L1 and L2, respectively, and 0.05 mmol/L and 0.04 mmol/L for L1 and L2, respectively, for period 2. During the same week, the averages for period 1 of urea content were 13.02 mg/dL and 11.79 mg/dL for L1 and L2, respectively, and 14.01 mg/dL and 12.43 mg/dL for L1 and L2, respectively, for period 2. Since spectral features that can be assigned to histamine are showing positive loadings, similar to features related to BHB and urea, we could assume that milk samples collected from cows assigned to L1 had higher levels of histamine too. Comparison between behavioural data and spectral analysis results Trial 1: Chain length configuration The analysis of the animal behavior and other animal-based outcome measures revealed that increasing the chain length improved the cows’ ease of movement and transitions [ 3 ]. Results indicate that the time spent outside of the stall by the withers of the cows (i.e., the ridge between the shoulder blades of an animal) assigned to the longer chain treatment (T2) was significantly greater in comparison to cows assigned to the recommended length (T1) (11 ± 1.1 vs 7 ± 1.1% of daily time; P = 0.05) [ 3 ], and it significantly increased across the trial (+ 3% of daily time; P ≤ 0.05) [ 3 ]. This measurement indicates that cows assigned to longer chains were spending more time outside the stall perimeter in the manger area. In addition, the distance outside of the stall perimeter for the withers, which represents the average distance outside of the stall in the manger area, increased significantly from the beginning and the end of the trial (+ 0.9 cm; P ≤ 0.05) for both treatments [ 3 ]; however, this measurement did not differ between treatments. From these results, we can conclude that cows assigned to the longer chain treatment (T2) were spending more time in the manger and were capable of stretching further in the manger area during the last three weeks of the trial. One possible behaviour of dairy cows at the manager is feeding; however, the combined “eating/rumination time” measurement did not reveal significant differences between the two treatments [ 27 ] and the device that was used for this measurement was not capable of reliably differentiate the two behaviours separately and therefore were combined [ 28 ]. Another possible behaviour for dairy cows at the manager is feed sorting. Cows tend to sort their ration initially in favour of the smaller particles and against longer forage particles [ 29 , 30 ]. This behaviour results in greater intake of highly-fermentable carbohydrates and lesser intake of effective fiber than intended and is associated with reduced rumen pH and altered milk composition. However, when dairy cows are given the opportunity through management and feeding environment that affect time available to manipulate feed, dairy cows will adjust their sorting pattern in favor of physically effective fiber to attenuate low rumen pH. Feed sorting behaviour also affects milk composition [ 30 ]. Milk fat can increase by 0.1 percentage points for every 10% selection in favor of long ration particles and milk protein can decrease by 0.05% for every 10% refusal of long ration particles. Indeed, average fat and protein concentrations in milk samples from cows with longer chains (T2) were higher by 0.04% and 0.06%, respectively, during the last three weeks of the trial. According to these results, we can hypothesize that cows with longer chains (T2) were given the opportunity to manipulate feed for longer time, since they spent more time in the manger, and they were at ease to access the manger a bit further, which gave them the chance to sort for effective fiber when ruminal pH was low. This hypothesis is inline with the trend that was revealed by the spectral analysis in which we observed changes of spectral features that are assigned to milk NPN, trans fatty acids and milk fat, which were consistent with changes that have been reported in the literature in cases of changing ruminal pH [ 23 , 24 , 25 ]. However, it must be mentioned that feed sorting behaviour and ruminal pH were not measured in this trial. Trial 2: Stall width configuration The spectral analysis suggested that cows assigned to the single stall width (T1) were synthesizing more milk fat and proteins during the last week of the trial. PC5, the principal component that revealed a significant treatment effect ( P = 0.042), also revealed a strong block effect ( P < 0.001). This observation suggests that the treatment effect might have been confounded by factors accounted for by blocking. In addition to nutrition and herd management practices, several factors affect milk composition. These factors are region, season, breed, individuality, age, disease, diurnal rhythm, stage of lactation and parity [ 31 ]. The trial was conducted on the same premises during one season, all cows were of the same breed (Holstein), and milk samples combined portions from morning and evening milkings to eliminate the variations in milk composition related to diurnal rhythm. Blocking was implemented in the experimental design of the trial to account for variations related to the remaining factors, which are stage of lactation and parity. However, a milk sample of cow 2057 from block 8 in the double width treatment (T2) was missing in week 6 of the trial. This cow was the only primiparous one in T2. In addition, T2 had a cow in its seventh parity (i.e., cow 7097) while the greatest parity in T1 was the fifth (i.e., cow 419). Owing to the missing milk sample of cow 2057, the median parity of the cows assigned to T2 became greater than that of the cows assigned to T1, namely, 3 and 2, respectively. Hence, cows of T2 were older and one lactation higher than those of T1. Milk fat and protein decline as the animal becomes older [ 32 ]. Milk fat falls about 0.2% each year from the first to fifth lactation because of higher production and more udder infections, while protein decreases 0.02 to 0.05% each lactation as animals age. This observation might explain the strong block effect ( P < 0.001) and the treatment effect ( P = 0.042) on PC5. In this case, the treatment effect might have originated from the imbalance of cow age (parity) in blocks of T1 and T2 that resulted from the exclusion of cow 2057 owing to the missing milk sample in week 6. The effect of a greater age (parity) in T2 might have overshadowed possible effect on milk composition related to the stall width treatment. When the spectral analysis was applied to the spectral datasets of week 5, none of the extracted PC revealed a significant treatment effect. In this trial, the spectral analysis possibly detected effects of multiple factors on the FTIR spectra of milk samples. The results of the spectral analysis were consistent with specific experimental details of the trial, which limits our capacity to corroborate improved comfort status shown by the behavioral data (i.e., larger width equals better resting postures [ 6 ]). Trial 3: Manger wall – stall length configuration Histamine was the most interesting candidate among the molecular species that could have accounted for features with positive loadings in the loading spectrum of the PC that revealed a significant stall length effect. Histamine in milk originates from the blood serum [ 22 ], and elevated concentrations of histamine can be attributed to the corium tissue breakdown (i.e., resulting in skin lesions or injuries) or stress [ 23 ]. The results of the trial show that injury severity decreased at several different locations on the cows over time, regardless of treatment. It has been found that cows had 4–8 times fewer contacts with the tie-rail while they were rising in long stalls (L2) regardless of the manger height but only in comparison to short stalls with low manger, which may have led to possible reduction of injuries on the cow’s neck [ 7 ]. However, reduction in injuries on the cow’s neck was greater for cows in short stalls with high manger as compared to low manger. The key finding on the outcome measures of comfort comparing long stalls to short stalls is the increase of 1 h per day in lying time [ 7 ]. Increased lying time of 1 h per day for cows assigned to longer stalls (L2) indicates a more comfortable environment (i.e., less contact with stall elements) and thus may explain the reduced histamine concentrations [ 23 ]. It must be noted that no physiological indicator related to stress was measured in this trial. Another source of elevated histamine in the blood is protein degradation that is associated with necrotic diseases such as mastitis and metritis [ 23 ]. In week 6 of the trial, the averages of SCC for short stall (L1) and long stall (L2), regardless of the manger wall height, were 281,435 and 133,136 cells/mL (data not shown), respectively. If these numbers are broken down by period, the averages of SCC for L1 and L2 for week 6 become 528,273 cells/mL and 203,700 cells/mL, respectively, for period 2. In Canadian Holstein cows, SCC greater than 200,000 cells/mL is considered a sign of mastitis for cows that are more than 30 DIM [ 33 ]. This observation suggests that cows assigned to shorter stalls might have been releasing more histamine into their blood than those assigned to longer stalls due to udder inflammation. However, no solid conclusion can be made regarding the relationship between the stall length, milk composition and susceptibility to mastitis in this trial. In fact, due to a strict maintenance of stall and bedding, no differences in stall cleanliness or dryness or in udder health were found between treatments and over time [ 7 ]. While the literature reports a few epidemiological studies that find that longer stalls increase udder dirtiness [ 34 , 35 ], the link between stall length, stall cleanliness and udder cleanliness is not clearly demonstrated [ 36 , 37 , 38 , 39 ]. The dynamics of infectious microorganisms’ proliferation on dairy farms is complex and the stall length and manger wall trial was not designed to consider factors that have significant effect on the proliferation of mastitis, such as management stall cleaning practices [ 34 ]. CONCLUSION In this paper, we studied changes in milk composition that can be attributed to modifications in the housing environment by altering specific elements in the tie-stall. None of the proposed modifications to the chain length, stall width, stall length and manger wall height were designed to stress the cows or inflect pain on them. On the contrary, all these modifications were proposed to increase cows’ comfort and ease of movement in their stalls. When changes in major milk components and few minor ones (i.e., urea and BHB) were considered individually, they did not reveal any differences between the control and the suggested modifications. However, principal component scores calculated from milk FTIR spectra, a multivariate measurement, revealed significant differences between treatments. These scores were capable of capturing simultaneous minor changes in multiple milk components and could relate those subtle changes to a treatment effect. The conclusions drawn from the spectral analysis were consistent with those drawn from the measured behavioral responses to treatment application and the results suggest that a proxy value for multiple milk components’ concentrations calculated from FTIR spectra can be developed to assess animal comfort, which will be the focus of the next steps in this research. Milk FTIR spectra contain more detailed information and are more likely to be successful in assessing cow comfort and welfare through regular milk samples that are routinely collected for dairy herd improvement programs. However, more research needs to be done to increase our potential to understand the welfare status of cows by adding other additional markers of welfare and health status (e.g., oxytocin) and/or combining milk FTIR spectroscopy with other analytical tools (e.g., metabolomics). We also need to extend the research to other aspects of animal welfare, such as affective states. Declarations ACKNOWLEDGEMENTS The authors would like to acknowledge the funding support provided by NSERC, Novalait, Dairy Farmers of Canada, and Valacta (now Lactanet) as a part of the NSERC Industrial Research Chair in the Sustainable Life of Dairy Cattle. The authors would also like to acknowledge the funding contribution of FRQNT, CRIBIQ and Novalait- (project No 2017-LG-202046) as part of the Programme de recherche en partenariat pour l'innovation en production et en transformation laitières - VII 2e concours, as well as Valacta (now Lactanet) for providing the milk component and infrared data used in this study. We thank and acknowledge Dr. Jacqueline Sedman (McGill University) for reviewing the manuscript, Tania Wolfe (McGill University) for providing guidance on the statistical models used in this study, and former graduate students Véronique Boyer and Sarah McPherson (McGill University) for providing understanding in their animal trial designs and results. AUTHOR CONTRIBUTION MB has contributed to the design of the work, the acquisition, analysis and interpretation of the data, and to the writing and revision of the written draft. DW, AAI and EV have contributed to the design of the work, the analysis and interpretation of the data, and to the writing and revision of the written draft. DES, DML and RD have contributed to the design of the work, the analysis and interpretation of the data, and to the writing and revision of the written draft. COMPETING INTERESTS The author(s) declare no competing interests. DATA AVAILABILITY Data can be obtained upon request to the corresponding author. References Von Keyserlingk, M., Rushen, J., de Passillé, A.M. & Weary, D.M. Invited Review: The welfare of dairy cattle – Key concepts and the role of science. J Dairy Sci. 92 (9) , 4101-4111 (2009). World Organization for Animal Health. Terrestrial Animal Health Code, Vol. 1 (31 st ed.). 333. (World Organization for Animal Health, 2023). Boyer, V., de Passillé, A.M., Adam, S. & Vasseur, E. Making tiestalls more comfortable: II. Increasing chain length to improve the ease of movement of dairy cows. J Dairy Sci. 104(3) , 3316-3326 (2021). Shepley, E. & Vasseur, E. Graduate Student Literature Review: The effect of housing systems on movement opportunity of dairy cows and the implications on cow health and comfort. J Dairy Sci . 104(6) , 7315-7322 (2021). St John, J., Rushen, J., Adam, S. & Vasseur, E. Making tiestalls more comfortable: I. Adjusting tie-rail height and forward position to improve dairy cows' ability to rise and lie down. J Dairy Sci . 104(3) , 3304-3315 (2021). Boyer, V. et al . Making tiestalls more comfortable: III. Providing additional lateral space to improve the resting capacity and comfort of dairy cows. J Dairy Sci . 104(3) , 3327-3338 (2021). McPherson, S. & Vasseur, E. Making tiestalls more comfortable: IV. Increasing stall bed length and decreasing manger wall height to heal injuries and increase lying time in dairy cows housed in deep-bedded tiestalls. J Dairy Sci . 104(3) , 3339-3352 (2021). Bahadi, M., Ismail, A. A. & Vasseur, E. Fourier Transform Infrared Spectroscopy as a Tool to Study Milk Composition Changes in Dairy Cows Attributed to Housing Modifications to Improve Animal Welfare. Foods 10(2) , 450; 10.3390/foods.10020450 (2021). Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. lmerTest package: tests in linear mixed effects models. J Stat. Softw. 82(1) ,1-26 (2017). Lenth, R. V. emmeans: Estimated Marginal Means, aka Least-Squares Means (ed. 1.6.1). R package (2021). Soyeurt, H. et al . Mid-infrared prediction of bovine milk fatty acids across multiple breeds, production systems, and countries. J Dairy Sci. 94(4) , 1657-1667 (2011). Eskildsen, C. E. et al . Quantification of individual fatty acids in bovine milk by infrared spectroscopy and chemometrics: understanding predictions of highly collinear reference variables. J Dairy Sci. 97(12) , 7940-7951 (2014). Visentin, G., Penasa, M., Gottardo, P., Cassandro, M. & De Marchi, M. Predictive ability of mid-infrared spectroscopy for major mineral composition and coagulation traits of bovine milk by using the uninformative variable selection algorithm. J Dairy Sci. 99(10) , 8137-8145 (2016). Esbensen, K. et al . Multivariate Data Analysis : An Introduction to Multivariate Analysis, Process Analytical Technology and Quality by Design (6th edition). CAMO, Oslo, Norway (2018). Lehman, A., O’Rourke, N., Hatcher, L. & Stepanski, E. JMP for Basic Univariate and Multivariate Statistics: Methods for Researchers and Social Scientists (2 nd edition). 471-511 (Sas Institute, 2013). Bradley, M. Curve fitting in Raman and IR spectroscopy: basic theory of line shapes and applications (Thermo Fisher Scientific, Application Note 50733). https://assets.thermofisher.com/TFS-Assets/CAD/Application-Notes/AN50733_E.pdf (2007). Silverstein, R. M., Webster, F. X. & Kiemle, D. J. Spectrometric Identification of Organic Compounds (8th edition). 78-107 (John Wiley & Sons, 2015). Collado, J. A. & Ramírez, F. J. Infrared and Raman spectra of histamine‐Nh4 and histamine‐Nd4 monohydrochlorides. J Raman Spectrosc. 30(5) , 391-397 (1999). Socrates, G. Infrared and Raman Characteristic Group Frequencies: Tables and Charts (John Wiley & Sons, 2004). Karabacak, M., Cinar, M., & Kurt, M. Molecular structure and vibrational assignments of hippuric acid: A detailed density functional theoretical study. Spectrochim Acta A Mol Biomol Spectrosc. 74(5) , 1197-1203 (2009). Ito, K. & Bernstein, H. J. The vibrational spectra of the formate, acetate, and oxalate ions. Can JChem. 34(2), 170-178 (1956). Alston-Mills, B. Nitrogenous Components of Milk: G. Nonprotein Nitrogen Compounds in Bovine Milk in Handbook of milk composition (ed. Jensen, R.G.) . 468-472. (Academic Press, 1995). Nocek, J. E. Bovine Acidosis: Implications on Laminitis. J Dairy Sci. 80(5) , 1005-1028 (1997). Abdela, N. Sub-acute Ruminal Acidosis (SARA) and its Consequence in Dairy Cattle: A Review of Past and Recent Research at Global Prospective. Achiev Life Sci. 10(2) , 187-196 (2016). Humer, E. et al . Invited review: Practical feeding management recommendations to mitigate the risk of subacute ruminal acidosis in dairy cattle. J Dairy Sci. 101(2) , 872-888 (2018). van de Voort, F. R. & Ismai. A. A. Proximate analysis of foods by mid-FTIR spectroscopy. Trends Food Sci Technol. 2 , 13-17 (1991). Boyer, V., de Passillé, A.M., Adam, S. & Vasseur, E. Making tiestalls more comfortable: II. Increasing chain length to improve the ease of movement of dairy cows. J Dairy Sci . 104(3) , Supplemental Table S1, https://escholarship.mcgill.ca/concern/articles/p5547x13z?locale=en. Zambelis, A., Wolfe, T., & Vasseur, E. Validation of an ear-tag accelerometer to identify feeding and activity behaviors of tiestall-housed dairy cattle. J Dairy Sci. 102(5) , 4536-4540 (2019) DeVries, T., Dohme, F. & Beauchemin, K. Repeated ruminal acidosis challenges in lactating dairy cows at high and low risk for developing acidosis: Feed sorting. J Dairy Sci. 91(10) , 3958-3967 (2008). Miller-Cushon, E. & DeVries T. Feed sorting in dairy cattle: Causes, consequences, and management. J Dairy Sci. 100(5) , 4172-4183 (2017). Jensen, R. G. Miscellaneous Factors Affecting Composition and Volume of Human and Bovine Milks. in Handbook of milk composition. 237-271 (Academic Press, 1995). Heinrichs, J., Jones, C. & Bailey, K. Milk components: Understanding the causes and importance of milk fat and protein variation in your dairy herd. Dairy Anim Sci. 5 , 1e-8e (1997). Arnould, V. M.-R., Reding, R., Bormann, J., Gengler, N. & Soyeurt, H. Review: milk composition as management tool of sustainability. Biotechnol Agron Soc Environ. 17(4) , 613-621 (2013). Plesch, G. & Knierim, U.Effects of housing and management conditions on teat cleanliness of dairy cows in cubicle systems taking into account body dimensions of the cows. Animal . 6(8) , 1360-1368 (2012). Bouffard, V. et al . Effect of following recommendations for tiestall configuration on neck and leg lesions, lameness, cleanliness, and lying time in dairy cows. J Dairy Sci. 100(4) , 2935-2943 (2017). Veissier, I., Capdeville, J. & Delval, E. Cubicle housing systems for cattle: Comfort of dairy cows depends on cubicle adjustment. J Anim Sci. 82(11) , 3321-3337 (2004). Martiskainen, P., Koistinen, T. & Mononen, J. Cubicle dimensions affect resting-related behaviour, injuries and dirtiness of loose-housed dairy cows. in Proceedings of the XIII (13th) International Congress in Animal Hygiene (ed. Aland, A.). 175 (International Society for Animal Hygiene, 2007). Lombard, J. E., Tucker, C. B., von Keyserlingk, M. A. G., Kopral, C. A. & Weary, D. M. Associations between cow hygiene, hock injuries, and free stall usage on US dairy farms. J Dairy Sci. 93(10) , 4668-4676 (2010). McPherson, S. & Vasseur, E. Graduate Student Literature Review: The effects of bedding, stall length, and manger wall height on common outcome measures of dairy cow welfare in stall-based housing systems. J Dairy Sci. 103(11) , 10940-10950 (2020). Additional Declarations No competing interests reported. Supplementary Files BahadietalSupplementalData.docx Cite Share Download PDF Status: Published Journal Publication published 17 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 16 Jun, 2025 Reviews received at journal 16 May, 2025 Reviews received at journal 07 May, 2025 Reviewers agreed at journal 05 May, 2025 Reviewers agreed at journal 05 May, 2025 Reviews received at journal 07 Oct, 2024 Reviewers agreed at journal 23 Sep, 2024 Reviewers agreed at journal 23 Sep, 2024 Reviewers invited by journal 21 Sep, 2024 Editor assigned by journal 14 Sep, 2024 Editor invited by journal 23 Aug, 2024 Submission checks completed at journal 23 Aug, 2024 First submitted to journal 15 Aug, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4919745","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":357569939,"identity":"a1c52ddf-70fa-4ec2-ae55-c2ff6eb65a49","order_by":0,"name":"M. Bahadi","email":"data:image/png;base64,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","orcid":"","institution":"McGill University","correspondingAuthor":true,"prefix":"","firstName":"M.","middleName":"","lastName":"Bahadi","suffix":""},{"id":357569941,"identity":"678a5a4c-b394-4f94-9e99-67fb4d07cb6a","order_by":1,"name":"D. Warner","email":"","orcid":"","institution":"Lactanet, Sainte-Anne-de-Bellevue","correspondingAuthor":false,"prefix":"","firstName":"D.","middleName":"","lastName":"Warner","suffix":""},{"id":357569943,"identity":"44031e31-70a5-4e83-aea7-6ad3fcd07173","order_by":2,"name":"A. A. Ismail","email":"","orcid":"","institution":"McGill University","correspondingAuthor":false,"prefix":"","firstName":"A.","middleName":"A.","lastName":"Ismail","suffix":""},{"id":357569944,"identity":"95a06b36-0950-4185-8ff8-cff3fc23bbb8","order_by":3,"name":"D. E. Santschi","email":"","orcid":"","institution":"Lactanet, Sainte-Anne-de-Bellevue","correspondingAuthor":false,"prefix":"","firstName":"D.","middleName":"E.","lastName":"Santschi","suffix":""},{"id":357569946,"identity":"68e9eb1c-e431-4ae8-94f4-9ad0e90ac3c4","order_by":4,"name":"D. M. Lefebvre","email":"","orcid":"","institution":"Lactanet, Sainte-Anne-de-Bellevue","correspondingAuthor":false,"prefix":"","firstName":"D.","middleName":"M.","lastName":"Lefebvre","suffix":""},{"id":357569948,"identity":"594f4274-71c5-4adf-96f5-f6a48862d7d9","order_by":5,"name":"R. Duggavathi","email":"","orcid":"","institution":"McGill University","correspondingAuthor":false,"prefix":"","firstName":"R.","middleName":"","lastName":"Duggavathi","suffix":""},{"id":357569950,"identity":"6148fc32-8fdf-4095-b22e-63bc8995ad1c","order_by":6,"name":"E. Vasseur","email":"","orcid":"","institution":"McGill University","correspondingAuthor":false,"prefix":"","firstName":"E.","middleName":"","lastName":"Vasseur","suffix":""}],"badges":[],"createdAt":"2024-08-15 14:01:54","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4919745/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4919745/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-28557-7","type":"published","date":"2025-12-17T15:58:04+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":65088905,"identity":"fb532726-f03b-4e06-8ef4-7479c7829846","added_by":"auto","created_at":"2024-09-23 13:26:57","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":243802,"visible":true,"origin":"","legend":"\u003cp\u003eThe spectral integral of principal component 6 loading spectrum extracted from long-term treatment application vector normalized first derivative spectral average dataset for the tie chain length trial. Shaded regions can be assigned to: 1) trans fatty acids 968 cm-1, 2) citrate, hippuric acid and acetate 1365-1160 cm-1, 3) acetate 1419 cm-1, fatty acids and NPN (i.e., urea, ammonium and histamine)1461 cm-1, 4) creatine 1541 cm-1, histamine, hippuric acid, citrate and fatty acids 1576 cm-1, 5) Carboxylic group of free fatty acids ~ 1715 cm-1, 6) C=O stretching vibration of ester linkages of triglycerides ~1748 cm-1, 7) CH stretching of fatty acids 3000-2800 cm-1.\u003c/p\u003e","description":"","filename":"BahadietalFigure1withlabel.png","url":"https://assets-eu.researchsquare.com/files/rs-4919745/v1/31c5b53770064fa83177c09d.png"},{"id":65088104,"identity":"198879eb-b1c8-419b-980e-fce8e22e5c98","added_by":"auto","created_at":"2024-09-23 13:18:57","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":254567,"visible":true,"origin":"","legend":"\u003cp\u003ePrincipal component 5 loading spectrum extracted from long-term treatment application vector normalized raw spectral dataset for the stall width trial. Shaded regions can be assigned as follows: 1) 1250-1140 cm-1 to the methylene twisting and wagging vibrations of fatty acids, esters and BHB, and to the C-O stretching vibration of esters, 2) 1565-1520 cm-1 to the Amide II band of milk proteins, 3) 1721 cm-1 to the C=O stretching vibration of ester linkages in milk fat, 4) 3000-2840 cm-1 to the asymmetrical stretching (ν_as 〖CH〗_2) and the symmetrical stretching (ν_s 〖CH〗_2) of the methylene group in fatty acids of milk fat.\u003c/p\u003e","description":"","filename":"BahadietalFigure2withlabel.png","url":"https://assets-eu.researchsquare.com/files/rs-4919745/v1/3a2b61eb3b3aec48ceca6fbc.png"},{"id":65088107,"identity":"4535ffda-f6c7-4050-b56b-aaf1d5a8b45e","added_by":"auto","created_at":"2024-09-23 13:18:57","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":242402,"visible":true,"origin":"","legend":"\u003cp\u003ePrincipal component 6 loading spectrum extracted from long-term treatment application first derivative spectral dataset for the stall length and manger wall trial. Shaded regions can be assigned as follows: 1-3) peaks at ~1033, 980 and 946 cm-1are found in FTIR spectra of milk samples spiked with histamine and BHB and their aqueous solutions, 4) peak at ~1408 cm-1 is observed in FTIR spectra of milk samples spiked with BHB, 5) peak at ~1460 cm-1 is observed in FTIR spectra of milk samples spiked with urea and ammonium, 6) peak at ~1575 cm-1is observed in FTIR spectrum of histamine in aqueous solution.\u003c/p\u003e","description":"","filename":"BahadietalFigure3withlabel.png","url":"https://assets-eu.researchsquare.com/files/rs-4919745/v1/c4c19f57a931f5bf49ea2df7.png"},{"id":98813972,"identity":"b8f03761-ea7b-4532-b8d4-6e8c9a8718ec","added_by":"auto","created_at":"2025-12-22 16:08:47","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1979211,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4919745/v1/a996425b-cb83-4713-8a05-ae7b93916f9a.pdf"},{"id":65088106,"identity":"0e47f69f-0a65-43eb-8bce-efabaf953211","added_by":"auto","created_at":"2024-09-23 13:18:57","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":26191,"visible":true,"origin":"","legend":"","description":"","filename":"BahadietalSupplementalData.docx","url":"https://assets-eu.researchsquare.com/files/rs-4919745/v1/5553cbcbe0d80d86db0d01ff.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Evaluating cow welfare status from milk samples: effects of housing modifications on milk infrared spectra","fulltext":[{"header":"INTRODUCTION ","content":"\u003cp\u003eAnimal welfare is increasingly becoming a concern for dairy producers and society in general; hence, more research has been conducted to improve the different aspects of animal welfare [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. According to the World Organization of Animal Health, an animal that experiences good welfare is one that is \u0026ldquo;healthy, comfortable, well nourished, safe, is not suffering from unpleasant states such as pain, fear and distress, and is able to express behaviours that are important for its physical and mental state.\u0026rdquo; [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Based on this definition, the tie stall housing system of dairy cows has been criticized for its restrictions that it imposes on the cows\u0026rsquo; ability to move and to engage with social interactions with other cows [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], and for higher prevalence of lameness and cow comfort issues, which negatively affects cow welfare, public perceptions, and producer profitability [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Nevertheless, the tie stall system remains a prevailing housing system for dairy cows in North America. For example, in Canada it accounts for 73.8% of dairy operations in farms [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The current study is part of a comprehensive study conducted over a 3-yr period aimed at investigating the individual effect of different stall design aspects on ease of movement and comfort of dairy cows housed in tie stalls. The studied elements of the stall design that were covered by this comprehensive study were the tie-rail height and forward position [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], the chain length \u003cb\u003e(TCL)\u003c/b\u003e [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], the stall width \u003cb\u003e(SW)\u003c/b\u003e [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] and the combined effect of stall length and manger height (\u003cb\u003eMW/SL)\u003c/b\u003e [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. To assess the effects of modifying those elements on cow ability to move and comfort, multiple outcomes were measured during each trial including injury scores [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], resting behaviour [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], lying down and rising events [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e], tracking cow\u0026rsquo;s movement in the stall [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] and postures and position in space of head, body and limbs during lying hours [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. While these outcomes remain the accepted measures to evaluate animal comfort and welfare in dairy housing systems [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], we hypothesize that cows\u0026rsquo; comfort improvement might lead to physiological changes, which might be reflected in milk chemical composition. However, affected milk components might not necessarily be those that are currently reported in dairy herd improvement \u003cb\u003e(DHI)\u003c/b\u003e programs; hence, we decided to mine milk infrared \u003cb\u003e(IR)\u003c/b\u003e spectra to study the changes in milk composition that can be attributed to changes in the housing environment.\u003c/p\u003e \u003cp\u003eMilk IR spectra contain signals from all molecules present in milk that can absorb IR energy; therefore, milk IR spectra represent a comprehensive snapshot of the chemical composition of a milk sample. In a previous publication [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], we presented a novel hybrid approach for spectral analysis that combined mixed modeling and multivariate analysis of milk Fourier transform infrared (\u003cb\u003eFTIR\u003c/b\u003e) spectra in which we attributed changes in milk composition to modifications to the tie-rail of a tie stall described elsewhere [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In this approach, we considered principal component analysis \u003cb\u003e(PCA)\u003c/b\u003e scores as a proxy value for the concentrations of multiple milk components since they are calculated from multivariate measurements (i.e., milk FTIR spectra) that contain spectral contributions from the concentrations of multiple milk components, and we tested them for a housing treatment effect (i.e., different tie-rail configurations). The analysis pinpointed a tie-rail configuration treatment among multiple ones whose milk samples\u0026rsquo; scores (i.e., principal component 7 scores) were significantly different from the scores of milk samples of other treatments. The loadings of that principal component revealed an inverse relationship between lactose and energy metabolism related molecules, such as b-hydroxy butyrate \u003cb\u003e(BHB)\u003c/b\u003e, acetone and citrate, which suggested that cows in that tie-rail configuration treatment were experiencing higher level of body fat mobilization. Cows enrolled in that tie-rail configuration treatment also recorded increased injuries on two locations on the cow\u0026rsquo;s neck [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] and those injuries might have been obstructing the cows from accessing feed. We concluded that analyzing milk FTIR spectra directly and without relying on predicted values of any analyte of interest was a viable option in capturing changes in milk composition that can be attributed to modifications in housing conditions.\u003c/p\u003e \u003cp\u003eThe objective of this paper is to determine potential changes in milk components that can be attributed to changes in TCL, SW and MW/SL configurations by applying the hybrid spectral analysis approach [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] to milk spectra collected in those trials. More specifically, this approach will be evaluated for its potential to capture a milk spectral fingerprint of cows\u0026rsquo; comfort that can later be used to predict a score for cow comfort and welfare from milk composition data or spectra, which is the long-term objective of this work. It must be noted that none of the treatments in these trials were designed to decrease cow comfort in the tie stall. On the contrary, all suggested treatments in these trials were designed to increase cows\u0026rsquo; ease of movement; therefore, increasing their level of comfort and welfare. As a results, we expect changes to milk composition to be subtle. Routine animal welfare outcomes were used to categorize the level of animal comfort and ease of movement (i.e., improved or not) provided by each housing treatment (fully reported elsewhere [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]) and key findings will be used in this manuscript to support any trends detected by the hybrid analysis approach of spectral data. According to our knowledge, no previous work was reported in the literature in which milk FTIR spectra were used in designed trials to assess the level of animal comfort and ease of movement in housing environment.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eExperimental Setup\u003c/h2\u003e \u003cp\u003eA series of trials have been conducted to test different housing treatments intended to improve the level of comfort and welfare of individual cows by increasing the opportunity of movement given to the cow in her stall. The animal trials were conducted at the Macdonald Campus Dairy Complex of McGill University (Sainte-Anne-de-Bellevue, QC, Canada) using animals from the dairy herd. Details on the experimental setup and animal handling can be found elsewhere [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Use of animals and all procedures were approved by the Animal Care Committee of McGill University and affiliated hospitals and research institutes (protocol #2016\u0026ndash;7794), and all experiments were performed in accordance with relevant guidelines and regulations.\u003c/p\u003e \u003cp\u003eIn this section, the experimental setups of the 3 trials are briefly described. An overview of the housing configuration treatments is shown in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. From hereafter, T1 will be referring to the control configuration of the tie stall element and T2 will be referring to the suggested modification to that element.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eHousing configuration treatments across each of the 4 trials\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTrial\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eHousing configuration\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eTreatments\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eExperimental design\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eWeek\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eCows\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003cp\u003e(T1)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSuggested modification (T2)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTCL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eChain length\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.0 m\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.4 m\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRandomized block design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStall width\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSingle\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDouble\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eRandomized block design\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMW/SL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eManger wall and stall length\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e20 cm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5 cm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCross over\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2 periods, 1 week adaptation, 6 weeks treatment\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003e\u003csup\u003e1\u003c/sup\u003e Single: average width of 139 cm (determined as 2 x (width of cow at hips)\u0026thinsp;+\u0026thinsp;5 cm), \u003csup\u003e2\u003c/sup\u003eDouble: average of 284 cm (determined as 2 x Single), \u003csup\u003e3\u003c/sup\u003e Both treatments were assigned to two stall lengths: short stall length 178 cm (L1) and long stall length 188 cm (L2)\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eTrial 1: Chain length configuration\u003c/h2\u003e \u003cp\u003eTwenty-four lactating Holstein cows were assigned to 2 tie chain length \u003cb\u003e(TCL)\u003c/b\u003e treatments. T1 was the control, and T2 was a suggested treatment intended to increase the cow\u0026rsquo;s movement ability in her stall [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Cows were assigned to 12 different blocks of two cows to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation (average DIM 129) and were placed evenly in two rows facing a wall within the barn. Five cows (4 long chain, 1 control) were removed at different points (1 in week 6, 1 in week 8, 2 in week 9, and 1 in week 10) during the trial for different reasons [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The trial lasted for 10 weeks from February 20, 2017, to May 1, 2017.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eTrial 2: Stall width configuration\u003c/h2\u003e \u003cp\u003eSixteen lactating Holstein cows were assigned to 2 stall width \u003cb\u003e(SW)\u003c/b\u003e treatments. T1 was the control and T2 was a suggested treatment intended to increase the cow\u0026rsquo;s comfort, and specifically cow\u0026rsquo;s rest, in her stall [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Cows were assigned to 8 blocks of 2 cows to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation (average DIM: 157) and were placed evenly in two rows within the barn. The trial lasted for 6 weeks from June 5, 2017, to July 17, 2017.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eTrial 3: Manger wall and stall length configuration\u003c/h2\u003e \u003cp\u003eThe lengths of two rows in the barn were modified as follows: row 1 was 178 cm (i.e., short or L1), which is the length commonly found on Quebec\u0026rsquo;s farms, and row 4 was 188 cm (i.e., long or L2), which was a suggested modification. Two manger wall heights were applied randomly to stalls within each row: T1, which was the upper limit of the recommendation, and T2 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Twenty-four cows were randomly divided into 4 groups. Two groups were assigned to each row and subjected to both manger wall treatments in a crossover design (1 week for adaptation, 6 weeks of data collection per treatment). Cows were assigned to 6 different blocks to account for age of the cow (i.e., parity or number of lactations) and days in milk within current lactation. Increasing the stall length and reducing the manger wall increased the space available for the cow, which eased her movement and lying, thus reducing her susceptibility to injuries [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The trial lasted for 14 weeks from February 26, 2018, to June 4, 2018. All 24 cows enrolled in the study remained in the final data set for period 1, but 3 were removed from period 2 [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. This trial will be referred to as \u003cb\u003eMW/SL\u003c/b\u003e in this paper.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eMilk Component Analysis\u003c/h2\u003e \u003cp\u003eOne composite milk sample per week was collected from each cow participating in the trials. The sample consisted of equal volumes of milk collected during the evening milking and the morning milking of the next day. All collected milk samples were analyzed for milk composition by FTIR spectroscopy at the Lactanet laboratory (Sainte-Anne-de-Bellevue, QC, Canada) using the same CombiFoss FT\u0026thinsp;+\u0026thinsp;analyzer (FOSS, Hiller\u0026oslash;d, Denmark). The concentrations of fat, protein (i.e., crude protein), lactose, urea nitrogen, and BHB, the fat-to-protein ratio, and somatic cell count (SCC; log transformed) were determined (Supplementary Table S 1).\u003c/p\u003e \u003cp\u003eThe milk component concentration data were analyzed to determine whether they differed among housing configurations within each trial. For the TCL trial, the treatment short-term effect was tested on milk components\u0026rsquo; concentrations averages per cow that were calculated from samples collected in weeks 1,2 and 3. The treatment long-term effect was tested on milk components\u0026rsquo; concentrations averages per cow that included samples collected in weeks 8, 9 and 10. For the SW and MW/SL trials, the treatment short-term and long-term effects were tested on samples that were collected from week 1 and from week 6, respectively. Data analysis was done in R (version 4.1; R Foundation for Statistical Computing, Vienna, Austria) using the add-on package lmerTest [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Marginal means were estimated for treatment groups using the add-on package emmeans [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eSpectral Analysis\u003c/h2\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003eMilk FTIR Spectra, Outliers Check and Spectral Pre-treatments\u003c/h2\u003e \u003cp\u003eFor each milk sample, a full FTIR spectrum was collected, which contained 1060 spectral variables (i.e., wavenumbers) between 5008 and 925 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Principal component analysis (\u003cb\u003ePCA\u003c/b\u003e) in JMP Pro 13.2.1 (SAS Institute, Cary, NC, USA) was applied to the collected spectra to detect outliers and none were observed in the PCA scores plot (PC1 vs. PC2). Only spectral regions containing information related to milk composition were retained for spectral analysis, namely, 1612\u0026thinsp;\u0026minus;\u0026thinsp;925 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, 1797\u0026thinsp;\u0026minus;\u0026thinsp;1681 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and 3061\u0026thinsp;\u0026minus;\u0026thinsp;2803 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The total number of spectral variables that were retained for analysis was 278 wavenumbers. First derivative with a derivative window of 1 and vector normalization [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e] were applied, separately or combined, as pre-treatments to milk spectra using codes written in MATLAB R2018a (MathWorks, Natick, MA, USA). As a result, four sets of milk FTIR spectra were obtained for each trial: raw, vector normalized raw (\u003cb\u003eVN\u003c/b\u003e raw), first derivative (\u003cb\u003eFD\u003c/b\u003e), and vector normalized first derivative (\u003cb\u003eVN-FD\u003c/b\u003e). For the TCL trial, short-term treatment application average spectra were calculated for each cow from spectra of samples collected in weeks 1,2 and 3. Long-term treatment application average spectra were calculated for each cow from spectra of samples collected in weeks 8,9,10. These averages were calculated for the raw, VN-raw, FD and VN-FD spectral data sets. For the SW and MW/SL trials, spectra of weeks 1 and 6 were used as short-term and long-term treatment application spectra, respectively, for data analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003eHybrid Spectral Analysis Approach\u003c/h2\u003e \u003cp\u003eThe effect of housing configuration treatments on milk FTIR spectra was evaluated by combining multivariate analysis with mixed modeling detailed in a previous publication [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], 2021). In short, principal components (PC) were calculated by applying PCA in JMP Pro 13.2.1 to the four versions (i.e., raw, FD, VN-raw, VN-FD) of the short-term and long-term treatment application spectra or spectral averages. Only PC with eigenvalue equal to or greater than 1 and that explained at least 1% of the variance in the respective spectral dataset were retained for further analysis. This step retained informative PC and discarded noisy ones. Scores of the retained PC were used as an input [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] to test for the treatment effect by PROC MIXED procedure in SAS 9.4 (SAS Institute, Cary, NC, USA) (see model description below). If a significant treatment effect is revealed at \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;0.05, then the least-squares means of its scores were examined to determine the treatment levels that were significantly different from the other levels using a Scheff\u0026eacute; adjustment. If PC was calculated from raw spectral dataset, then the influential spectral features could be directly interpreted from the PC\u0026rsquo;s loading spectrum. If PC was obtained from FD spectral datasets, then the spectral integral of the PC\u0026rsquo;s loading spectrum was calculated before interpretation of the influential spectral features. The cumulative trapezoidal numerical integration function in MATLAB R2018a (MathWorks, Natick, MA, USA) was used to calculate the spectral integral of the loading spectrum in question. A peak-fitting procedure in the Peak Resolve feature in OMNIC\u0026trade; 7.3 (Thermo Fisher Scientific, Waltham, MA, USA) was applied to regions in the integrated loading spectrum where no clear peaks were present. The Voigt function [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] with low or high sensitivity was used and the baseline correction was set to none. The noise and the full width at half height of the narrowest peak in the region of interest were determined by the software. The fitting process was repeated several times until an acceptable residual spectrum was obtained.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003ePC scores calculated from milk spectra were analyzed separately to test for the treatment effect in each trial using the models outlined below using the PROC MIXED procedure in SAS 9.4 (SAS Institute, Cary, NC, USA).\u003c/p\u003e \u003cp\u003eFor the TCL trial\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({Y_{ijk}}=\\mu +tr{t_i}+ro{w_j}+bloc{k_k}+{e_{ijk}}\\)\u003c/span\u003e \u003c/span\u003e \u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e is the dependent variable, the PC scores isolated from spectra of milk samples from the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e block (parity and lactation stage) in the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e row on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e chain length, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{trt}_{i}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e TCL treatment, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{block}_{k}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e parity and lactation stage combination, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{row}_{j}\\)\u003c/span\u003e\u003c/span\u003e was the random effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e row in the barn and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e was the random residual error.\u003c/p\u003e \u003cp\u003eFor the SW trial\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({Y_{ijk}}=\\mu +tr{t_i}+ro{w_j}+bloc{k_k}+{e_{ijk}}\\)\u003c/span\u003e \u003c/span\u003e \u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e is the dependent variable, the PC scores isolated from spectra of milk samples from the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e block (parity and lactation stage) in the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e row on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e stall width, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{trt}_{i}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e SW treatment, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{block}_{k}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e parity and lactation stage combination, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{row}_{j}\\)\u003c/span\u003e\u003c/span\u003e was the random effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e row in the barn and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{ijk}\\)\u003c/span\u003e\u003c/span\u003e was the random residual error.\u003c/p\u003e \u003cp\u003eFor the MW/SL trial\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\({Y_{ijkmnq}}=\\mu +lengt{h_i}+se{q_{ij}}+bloc{k_k}+co{w_{ijkm}}+perio{d_n}+tr{t_{iq}}+{e_{ijkmnq}}\\)\u003c/span\u003e \u003c/span\u003e \u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{Y}_{ijkmnq}\\)\u003c/span\u003e\u003c/span\u003e is the dependent variable, the PC scores isolated from spectra of milk samples from the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{m}^{th}\\)\u003c/span\u003e\u003c/span\u003e cow of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e block in the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e sequence of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e length, the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{q}^{th}\\)\u003c/span\u003e\u003c/span\u003e treatment of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e length, and the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{n}^{th}\\)\u003c/span\u003e\u003c/span\u003e period, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{length}_{i}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e stall bed length, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{seq}_{ij}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e sequence on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e stall bed length, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{block}_{k}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e parity and stage of lactation combination, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{cow}_{ijkm}\\)\u003c/span\u003e\u003c/span\u003e was the random effect of the cow from the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}^{th}\\)\u003c/span\u003e\u003c/span\u003e block on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{j}^{th}\\)\u003c/span\u003e\u003c/span\u003e sequence of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e stall bed length, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{period}_{n}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{n}^{th}\\)\u003c/span\u003e\u003c/span\u003e period, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{trt}_{iq}\\)\u003c/span\u003e\u003c/span\u003e was the fixed effect of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{q}^{th}\\)\u003c/span\u003e\u003c/span\u003e manger wall height treatment on the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{i}^{th}\\)\u003c/span\u003e\u003c/span\u003e stall bed length treatment and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{e}_{ijkmnq}\\)\u003c/span\u003e\u003c/span\u003e was the random residual error. The significance level was defined at \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;0.05 for all trials.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS AND DISCUSSION","content":"\u003cp\u003eA first analysis of milk component concentrations predicted through FTIR spectroscopic analysis of milk samples did not reveal any significant difference between housing configurations within each trial (Supplementary Table S 2). These findings suggest that analysis of major milk components (i.e., fat, protein, and lactose) individually may not be sufficient to reflect an impact of cow comfort and ease of movement on milk composition through changing housing configurations.\u003c/p\u003e \u003cp\u003eIn contrast, when a more comprehensive analysis of the spectra was conducted by employing PCA, significant housing treatment effects for the TCL (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032) and SW (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042) trials were revealed in PC6 and 5, respectively (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). For the TCL trial, PC6 was extracted from the VN-FD long-term treatment application spectral averages datasets, and it described 1.70% of the variation. This PC did not reveal significant effects for the other studied factors that were included in the statistical model. This observation suggests that the TCL might have influenced milk composition during the trial. For the SW trial, PC5 was extracted from the VN-raw long-term treatment application spectral dataset, and it described 1.13% of the variation in that dataset. In addition to the treatment effect, this PC revealed a strong block effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001); for this reason, the interpretation of the loading spectrum for this PC will take into consideration all factors accounted for by blocking, which were cow parity (i.e., number of lactations) and days in milk in the current lactation (DIM). For the MW/SL trial, the manger wall height treatment and its combined effect with the stall length did not reveal any significant treatment effect, which means that the manger wall treatment did not affect milk composition during the trial. However, PC6 extracted from the long-term treatment application FD spectral dataset revealed significant length effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036), which means that the stall length might have had a significant effect on milk composition during the trial. This PC described 1.77% of the variation in its respective dataset. In the following sections, we will refer to the short and long stalls as L1 and L2, respectively, to distinguish them from the two treatments (i.e., T1 and T2), which represent the combined effect of the stall length and manger wall height.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePrincipal components extracted from long-term spectral datasets that revealed significant effects in the four trials. The table also lists P values obtained from the SAS PROC MIXED Procedure for tested effects in each trial\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTrials\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSpectral Dataset\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePC\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEigenvalue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eExplained Variation %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"5\" nameend=\"c10\" namest=\"c6\"\u003e \u003cp\u003eP Values\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eTCL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVN FD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e4.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eTrt\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBlock\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.088\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVN Raw\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e3.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eTrt\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBlock\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eMW/SL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eFD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e4.92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003elength\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSeq\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eBlock\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003eTrt\u003csup\u003e4\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003ePeriod\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0.375\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e0.276\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e0.912\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"10\"\u003e\u003csup\u003e1\u003c/sup\u003e TCL\u0026thinsp;=\u0026thinsp;tie chain length trial, SW\u0026thinsp;=\u0026thinsp;stall width trial, MW/SL\u0026thinsp;=\u0026thinsp;manger wall/stall length trial, \u003csup\u003e2\u003c/sup\u003e FD\u0026thinsp;=\u0026thinsp;first derivative, VN\u0026thinsp;=\u0026thinsp;vector normalized, \u003csup\u003e3\u003c/sup\u003e PC\u0026thinsp;=\u0026thinsp;principal component, \u003csup\u003e4\u003c/sup\u003e Trt\u0026thinsp;=\u0026thinsp;treatment\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eTable\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e summarizes the differences in the least-squares means produced by the PROC MIXED procedure for the scores of PC6, 5 and 6, which revealed a significant treatment effect. For the TCL, SW and MW/SL trials, there were significant differences in milk samples\u0026rsquo; scores for T1 vs. T2 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.032), T1 vs. T2 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042) and L1 vs. L2 (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036), respectively.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDifferences of least squares means for the scores of the principal components that revealed a significant treatment effect for each trial\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTrial\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTreatment\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTreatment\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEstimate\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eStandard\u003c/p\u003e \u003cp\u003eError\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eDF\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003et\u0026nbsp;Value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e Value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eScheff\u0026eacute; Adj. \u003cem\u003eP\u003c/em\u003e Value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTCL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eT2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-1.6819\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.6650\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e-2.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eT1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eT2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.6021\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.2224\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e5.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e2.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMW/SL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eL1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eL2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e-1.3150\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.5998\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e-2.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.036\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"9\"\u003e\u003csup\u003e1\u003c/sup\u003e TCL\u0026thinsp;=\u0026thinsp;tie chain length trial, SW\u0026thinsp;=\u0026thinsp;stall width trial, MW/SL\u0026thinsp;=\u0026thinsp;manger wall/stall length trial, \u003csup\u003e2\u003c/sup\u003e Degrees of freedom\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eInterpretation of loading spectra of principal components that revealed significant housing configuration treatment effect on milk composition\u003c/b\u003e \u003c/p\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eTrial 1: Chain length configuration\u003c/h2\u003e \u003cp\u003eInspection of the integral of the PC6 loading spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) revealed peaks at the following wavenumbers: 2919, 2851, 1715, 1576, 1541, 1461, 1419 and 968 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. The peaks at 2919, 2851, 1715, 1541, 1419 and 968 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e can be assigned to the following vibrational modes: the asymmetrical stretching vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\nu\\:}_{as}{CH}_{2}\\)\u003c/span\u003e\u003c/span\u003e) of methylene groups in fatty acids [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], the symmetrical stretching vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\nu\\:}_{s}{CH}_{2}\\)\u003c/span\u003e\u003c/span\u003e) of methylene groups in fatty acids [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and histamine [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C=O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the carboxyl functional group in free fatty acids [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], the deformation vibration \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left(\\delta\\:{NH}_{2}\\right)\\)\u003c/span\u003e\u003c/span\u003e of the amidine functional group in creatine [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], the symmetrical \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the carboxylate ion in acetate, and the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-H\\)\u003c/span\u003e\u003c/span\u003e bending vibration of the \u003cem\u003etrans\u003c/em\u003e double bond in trans fatty acids [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], respectively. The peak at 1576 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e may be assigned to: the ring stretching vibration of the imidazole ring in histamine [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:N-H\\)\u003c/span\u003e\u003c/span\u003e bending vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\delta\\:NH\\)\u003c/span\u003e\u003c/span\u003e) of hippuric acid [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] or the symmetrical \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the carboxylate ion in citrate and deprotonated fatty acids [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The peak at 1461 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e may be assigned to: the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-N\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of urea [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:N-H\\)\u003c/span\u003e\u003c/span\u003e bending vibration of the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NH}_{4}^{+}\\)\u003c/span\u003e\u003c/span\u003e ion or the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{NH}_{3}^{+}\\)\u003c/span\u003e\u003c/span\u003e symmetrical bending vibration \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left({\\delta\\:}_{s}{NH}_{3}^{+}\\right)\\)\u003c/span\u003e\u003c/span\u003e of histamine [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. In addition, negative loadings were observed at 1748 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, which can be assigned to the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C=O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the ester linkages in milk fat [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe peak fitting process for the region 1365\u0026thinsp;\u0026minus;\u0026thinsp;1160 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e revealed peaks at 1347, 1299 and 1249 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, which can be assigned to the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{CH}_{3}\\)\u003c/span\u003e\u003c/span\u003e deformation vibration [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e], which appears in the FTIR spectrum of an aqueous solution of acetate, the bending vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\delta\\:{CH}_{2}\\)\u003c/span\u003e\u003c/span\u003e) of the methylene group in hippuric acid [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] and the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration in citrate [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], respectively. These assignments are consistent with the results of the spiking experiments reported elsewhere [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe result of the spectral analysis suggests that the average FTIR spectra of milk samples collected from cows enrolled in both TCL treatments had significant differences in the last 3 weeks of the trial (i.e., long-term effect of chain length configuration treatment application). The loading spectrum of the principal component that revealed the significant treatment effect showed positive loadings for spectral features that might be assigned to trans fatty acids, acetate, citrate, and non-protein nitrogen \u003cb\u003e(NPN)\u003c/b\u003e compounds, including histamine. These positive loadings suggest a direct relationship among these spectral features; hence, they might have been similarly affected by the TCL treatment. Accounting for 35\u0026ndash;48% of milk NPN [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], milk urea averages that were determined by the FTIR milk analyzer were 12.28 mg/dL and 12.14 mg/dL for T1 and T2, respectively, during the last 3 weeks of the trial. Considering that milk urea average in T1 is greater than that of T2 and that spectral features related to urea and other minor milk components (e.g., other milk NPN and trans fatty acids) are showing loadings in the same direction, we could conclude that the concentrations of the above-mentioned minor milk components are following the same trend as urea. Increased trans fatty acids and NPN in milk, and histamine specifically, have previously been reported as markers of decreased pH in the rumen [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. In addition, negative loadings were observed at a spectral feature that can be assigned to milk fat, which suggests an inverse relationship between milk fat and the previously mentioned minor milk components. In fact, decreased milk fat [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], decreased milk protein and increased milk NPN [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] were reported in cases of reduced ruminal pH. Indeed, we noticed a decrease in the average concentrations of protein and milk fat for cows enrolled in T1 in comparison to T2 during the last 3 weeks of the trial. The averages of protein were 3.28% and 3.34% for T1 and T2, respectively, and the averages of fat were 3.86% and 3.90% for T1 and T2, respectively, for that period. The observed trend in changes of multiple milk components (e.g., milk NPN, trans fatty acids, fat, and protein) in this trial is consistent with changes that have been reported in the literature in cases of changing ruminal pH and they suggest that cows with longer chains (T2) might have more stable ruminal pH, which was not measured for cows enrolled in this trial, during the last three weeks of the trial.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eTrial 2: Stall width configuration\u003c/h2\u003e \u003cp\u003eInspection of the PC5 loading spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) revealed positive loadings at 1721 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and in the region\u0026thinsp;~\u0026thinsp;1250\u0026thinsp;\u0026minus;\u0026thinsp;1140 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e centered at ~\u0026thinsp;1205 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Positive loadings were also observed around 2920 and 2855 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and within the region\u0026thinsp;~\u0026thinsp;1565\u0026thinsp;\u0026minus;\u0026thinsp;1520 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. Milk fat is predominantly composed of triglycerides, which are tri-esters derived from glycerol and three fatty acids, and the bands at 2920, 2855 and 1721 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e can be assigned to the asymmetrical stretching vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\nu\\:}_{as}{CH}_{2}\\)\u003c/span\u003e\u003c/span\u003e) of the methylene groups in fatty acids of milk fat [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], the symmetrical stretching vibration (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\nu\\:}_{s}{CH}_{2}\\)\u003c/span\u003e\u003c/span\u003e) of the methylene groups in fatty acids of milk fat [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] and the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C=O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the ester linkages in milk fat [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], respectively. The methylene twisting and wagging vibrations of fatty acids and esters occur in the region between ~\u0026thinsp;1250 and ~\u0026thinsp;1140 cm\u003csup\u003e\u0026minus;1\u003c/sup\u003e [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e], and the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of esters shows strong absorption in the 1210\u0026thinsp;\u0026minus;\u0026thinsp;1163 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e region [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. In addition, bands in the spectral region between 1565 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e and 1520 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e can be assigned to the Amide II band of milk proteins [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e \u003cp\u003ePC5 scores of milk samples collected during the last week of the trial (i.e., week 6, long-term effect of stall width configuration treatment application), differed mainly in spectral features related to fat and protein, which suggests a trend that can be confirmed by the analysis results reported by the FTIR milk analyzer. The averages of fat were 3.96% and 3.76% and the averages of protein were 3.34% and 3.28% for T1 and T2, respectively. This observation suggests that cows assigned to T1 had higher milk fat and milk proteins during the last week of the trial.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eTrial 3: Manger wall and stall length configuration\u003c/h2\u003e \u003cp\u003eInspection of the spectral integral of the PC6 loading spectrum (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e) revealed positive loadings at the following wavenumbers: 1575, 1460, 1408, 1033, 980 and 946 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. The peak at ~\u0026thinsp;1575 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is observed in the FTIR spectrum of histamine in aqueous solution [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] and can be assigned to the stretching vibration of the imidazole ring [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. The peak at ~\u0026thinsp;1460 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is observed in the FTIR spectra of milk samples spiked with urea and ammonium chloride [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]; hence, it can be assigned to the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-N\\)\u003c/span\u003e\u003c/span\u003e stretching vibration in urea [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] or to the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:N-H\\)\u003c/span\u003e\u003c/span\u003e bending vibration in the ammonium ion [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The peak at ~\u0026thinsp;1408 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is observed in milk samples spiked with BHB [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] and can be assigned to the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O-H\\)\u003c/span\u003e\u003c/span\u003e bending vibration or the symmetrical \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:C-O\\)\u003c/span\u003e\u003c/span\u003e stretching vibration of the carboxylate ion [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The peaks at ~\u0026thinsp;1033 and 980 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e are observed in the FTIR spectra of milk samples spiked with histamine [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. The peak at ~\u0026thinsp;946 cm\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e is observed in FTIR spectra of aqueous solutions of histamine and BHB [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe spectral analysis suggested that the stall length might have had an effect on BHB, urea and other milk NPN including histamine during the last week of the trial (i.e., long-term effect of manger wall and stall length configuration treatment application). Milk composition determined by the FTIR milk analyzer revealed that milk samples collected from cows assigned to L1 (i.e., the short length row) had higher levels of BHB and urea during the mentioned period. During week 6, the averages for period 1 of BHB content were 0.06 mmol/L and 0.05 mmol/L for L1 and L2, respectively, and 0.05 mmol/L and 0.04 mmol/L for L1 and L2, respectively, for period 2. During the same week, the averages for period 1 of urea content were 13.02 mg/dL and 11.79 mg/dL for L1 and L2, respectively, and 14.01 mg/dL and 12.43 mg/dL for L1 and L2, respectively, for period 2. Since spectral features that can be assigned to histamine are showing positive loadings, similar to features related to BHB and urea, we could assume that milk samples collected from cows assigned to L1 had higher levels of histamine too.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eComparison between behavioural data and spectral analysis results\u003c/h2\u003e \u003cdiv id=\"Sec17\" class=\"Section3\"\u003e \u003ch2\u003eTrial 1: Chain length configuration\u003c/h2\u003e \u003cp\u003eThe analysis of the animal behavior and other animal-based outcome measures revealed that increasing the chain length improved the cows\u0026rsquo; ease of movement and transitions [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Results indicate that the time spent outside of the stall by the withers of the cows (i.e., the ridge between the shoulder blades of an animal) assigned to the longer chain treatment (T2) was significantly greater in comparison to cows assigned to the recommended length (T1) (11\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1 vs 7\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1% of daily time; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.05) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], and it significantly increased across the trial (+\u0026thinsp;3% of daily time; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;0.05) [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. This measurement indicates that cows assigned to longer chains were spending more time outside the stall perimeter in the manger area. In addition, the distance outside of the stall perimeter for the withers, which represents the average distance outside of the stall in the manger area, increased significantly from the beginning and the end of the trial (+\u0026thinsp;0.9 cm; \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026le;\u0026thinsp;0.05) for both treatments [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]; however, this measurement did not differ between treatments. From these results, we can conclude that cows assigned to the longer chain treatment (T2) were spending more time in the manger and were capable of stretching further in the manger area during the last three weeks of the trial. One possible behaviour of dairy cows at the manager is feeding; however, the combined \u0026ldquo;eating/rumination time\u0026rdquo; measurement did not reveal significant differences between the two treatments [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] and the device that was used for this measurement was not capable of reliably differentiate the two behaviours separately and therefore were combined [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Another possible behaviour for dairy cows at the manager is feed sorting. Cows tend to sort their ration initially in favour of the smaller particles and against longer forage particles [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. This behaviour results in greater intake of highly-fermentable carbohydrates and lesser intake of effective fiber than intended and is associated with reduced rumen pH and altered milk composition. However, when dairy cows are given the opportunity through management and feeding environment that affect time available to manipulate feed, dairy cows will adjust their sorting pattern in favor of physically effective fiber to attenuate low rumen pH. Feed sorting behaviour also affects milk composition [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Milk fat can increase by 0.1 percentage points for every 10% selection in favor of long ration particles and milk protein can decrease by 0.05% for every 10% refusal of long ration particles. Indeed, average fat and protein concentrations in milk samples from cows with longer chains (T2) were higher by 0.04% and 0.06%, respectively, during the last three weeks of the trial. According to these results, we can hypothesize that cows with longer chains (T2) were given the opportunity to manipulate feed for longer time, since they spent more time in the manger, and they were at ease to access the manger a bit further, which gave them the chance to sort for effective fiber when ruminal pH was low. This hypothesis is inline with the trend that was revealed by the spectral analysis in which we observed changes of spectral features that are assigned to milk NPN, trans fatty acids and milk fat, which were consistent with changes that have been reported in the literature in cases of changing ruminal pH [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. However, it must be mentioned that feed sorting behaviour and ruminal pH were not measured in this trial.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eTrial 2: Stall width configuration\u003c/h2\u003e \u003cp\u003eThe spectral analysis suggested that cows assigned to the single stall width (T1) were synthesizing more milk fat and proteins during the last week of the trial. PC5, the principal component that revealed a significant treatment effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042), also revealed a strong block effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). This observation suggests that the treatment effect might have been confounded by factors accounted for by blocking. In addition to nutrition and herd management practices, several factors affect milk composition. These factors are region, season, breed, individuality, age, disease, diurnal rhythm, stage of lactation and parity [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The trial was conducted on the same premises during one season, all cows were of the same breed (Holstein), and milk samples combined portions from morning and evening milkings to eliminate the variations in milk composition related to diurnal rhythm. Blocking was implemented in the experimental design of the trial to account for variations related to the remaining factors, which are stage of lactation and parity. However, a milk sample of cow 2057 from block 8 in the double width treatment (T2) was missing in week 6 of the trial. This cow was the only primiparous one in T2. In addition, T2 had a cow in its seventh parity (i.e., cow 7097) while the greatest parity in T1 was the fifth (i.e., cow 419). Owing to the missing milk sample of cow 2057, the median parity of the cows assigned to T2 became greater than that of the cows assigned to T1, namely, 3 and 2, respectively. Hence, cows of T2 were older and one lactation higher than those of T1. Milk fat and protein decline as the animal becomes older [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Milk fat falls about 0.2% each year from the first to fifth lactation because of higher production and more udder infections, while protein decreases 0.02 to 0.05% each lactation as animals age. This observation might explain the strong block effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and the treatment effect (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042) on PC5. In this case, the treatment effect might have originated from the imbalance of cow age (parity) in blocks of T1 and T2 that resulted from the exclusion of cow 2057 owing to the missing milk sample in week 6. The effect of a greater age (parity) in T2 might have overshadowed possible effect on milk composition related to the stall width treatment. When the spectral analysis was applied to the spectral datasets of week 5, none of the extracted PC revealed a significant treatment effect. In this trial, the spectral analysis possibly detected effects of multiple factors on the FTIR spectra of milk samples. The results of the spectral analysis were consistent with specific experimental details of the trial, which limits our capacity to corroborate improved comfort status shown by the behavioral data (i.e., larger width equals better resting postures [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eTrial 3: Manger wall \u0026ndash; stall length configuration\u003c/h2\u003e \u003cp\u003eHistamine was the most interesting candidate among the molecular species that could have accounted for features with positive loadings in the loading spectrum of the PC that revealed a significant stall length effect. Histamine in milk originates from the blood serum [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], and elevated concentrations of histamine can be attributed to the corium tissue breakdown (i.e., resulting in skin lesions or injuries) or stress [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. The results of the trial show that injury severity decreased at several different locations on the cows over time, regardless of treatment. It has been found that cows had 4\u0026ndash;8 times fewer contacts with the tie-rail while they were rising in long stalls (L2) regardless of the manger height but only in comparison to short stalls with low manger, which may have led to possible reduction of injuries on the cow\u0026rsquo;s neck [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. However, reduction in injuries on the cow\u0026rsquo;s neck was greater for cows in short stalls with high manger as compared to low manger. The key finding on the outcome measures of comfort comparing long stalls to short stalls is the increase of 1 h per day in lying time [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Increased lying time of 1 h per day for cows assigned to longer stalls (L2) indicates a more comfortable environment (i.e., less contact with stall elements) and thus may explain the reduced histamine concentrations [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. It must be noted that no physiological indicator related to stress was measured in this trial.\u003c/p\u003e \u003cp\u003eAnother source of elevated histamine in the blood is protein degradation that is associated with necrotic diseases such as mastitis and metritis [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. In week 6 of the trial, the averages of SCC for short stall (L1) and long stall (L2), regardless of the manger wall height, were 281,435 and 133,136 cells/mL (data not shown), respectively. If these numbers are broken down by period, the averages of SCC for L1 and L2 for week 6 become 528,273 cells/mL and 203,700 cells/mL, respectively, for period 2. In Canadian Holstein cows, SCC greater than 200,000 cells/mL is considered a sign of mastitis for cows that are more than 30 DIM [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. This observation suggests that cows assigned to shorter stalls might have been releasing more histamine into their blood than those assigned to longer stalls due to udder inflammation. However, no solid conclusion can be made regarding the relationship between the stall length, milk composition and susceptibility to mastitis in this trial. In fact, due to a strict maintenance of stall and bedding, no differences in stall cleanliness or dryness or in udder health were found between treatments and over time [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. While the literature reports a few epidemiological studies that find that longer stalls increase udder dirtiness [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], the link between stall length, stall cleanliness and udder cleanliness is not clearly demonstrated [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. The dynamics of infectious microorganisms\u0026rsquo; proliferation on dairy farms is complex and the stall length and manger wall trial was not designed to consider factors that have significant effect on the proliferation of mastitis, such as management stall cleaning practices [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e"},{"header":"CONCLUSION","content":"\u003cp\u003eIn this paper, we studied changes in milk composition that can be attributed to modifications in the housing environment by altering specific elements in the tie-stall. None of the proposed modifications to the chain length, stall width, stall length and manger wall height were designed to stress the cows or inflect pain on them. On the contrary, all these modifications were proposed to increase cows\u0026rsquo; comfort and ease of movement in their stalls. When changes in major milk components and few minor ones (i.e., urea and BHB) were considered individually, they did not reveal any differences between the control and the suggested modifications. However, principal component scores calculated from milk FTIR spectra, a multivariate measurement, revealed significant differences between treatments. These scores were capable of capturing simultaneous minor changes in multiple milk components and could relate those subtle changes to a treatment effect. The conclusions drawn from the spectral analysis were consistent with those drawn from the measured behavioral responses to treatment application and the results suggest that a proxy value for multiple milk components\u0026rsquo; concentrations calculated from FTIR spectra can be developed to assess animal comfort, which will be the focus of the next steps in this research. Milk FTIR spectra contain more detailed information and are more likely to be successful in assessing cow comfort and welfare through regular milk samples that are routinely collected for dairy herd improvement programs. However, more research needs to be done to increase our potential to understand the welfare status of cows by adding other additional markers of welfare and health status (e.g., oxytocin) and/or combining milk FTIR spectroscopy with other analytical tools (e.g., metabolomics). We also need to extend the research to other aspects of animal welfare, such as affective states.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eACKNOWLEDGEMENTS\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe authors would like to acknowledge the funding support provided by NSERC, Novalait, Dairy Farmers of Canada, and Valacta (now Lactanet) as a part of the NSERC Industrial Research Chair in the Sustainable Life of Dairy Cattle. The authors would also like to acknowledge the funding contribution of FRQNT, CRIBIQ and Novalait- (project No 2017-LG-202046) as part of the Programme de recherche en partenariat pour l\u0026apos;innovation en production et en transformation laiti\u0026egrave;res - VII 2e concours, as\u0026nbsp;well as Valacta (now Lactanet) for providing the milk component and infrared data used in this study. We thank and acknowledge Dr. Jacqueline Sedman (McGill University) for reviewing the manuscript, Tania Wolfe (McGill University) for providing guidance on the statistical models used in this study, and former graduate students V\u0026eacute;ronique Boyer and Sarah McPherson (McGill University) for providing understanding in their animal trial designs and results.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAUTHOR CONTRIBUTION\u003c/p\u003e\n\u003cp\u003eMB has contributed to the design of the work, the acquisition, analysis and interpretation of the data, and to the writing and revision of the written draft. DW, AAI and EV have contributed to the design of the work, the analysis and interpretation of the data, and to the writing and revision of the written draft. DES, DML and RD have contributed to the design of the work, the analysis and interpretation of the data, and to the writing and revision of the written draft.\u003c/p\u003e\n\u003cp\u003eCOMPETING INTERESTS\u003c/p\u003e\n\u003cp\u003eThe author(s) declare no competing interests.\u003c/p\u003e\n\u003cp\u003eDATA AVAILABILITY\u003c/p\u003e\n\u003cp\u003eData can be obtained upon request to the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eVon Keyserlingk, M., Rushen, J., de Passill\u0026eacute;, A.M. \u0026amp; Weary, D.M. Invited Review: The welfare of dairy cattle \u0026ndash; Key concepts and the role of science. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e92 (9)\u003c/strong\u003e, 4101-4111 (2009).\u003c/li\u003e\n\u003cli\u003eWorld Organization for Animal Health. \u003cem\u003eTerrestrial Animal Health Code, Vol. 1 \u003c/em\u003e(31\u003csup\u003est\u003c/sup\u003e ed.). 333. (World Organization for Animal Health, 2023).\u003c/li\u003e\n\u003cli\u003eBoyer, V., de Passill\u0026eacute;, A.M., Adam, S. \u0026amp; Vasseur, E. Making tiestalls more comfortable: II. Increasing chain length to improve the ease of movement of dairy cows. \u003cem\u003eJ Dairy Sci. \u003c/em\u003e\u003cstrong\u003e104(3)\u003c/strong\u003e, 3316-3326 (2021).\u003c/li\u003e\n\u003cli\u003eShepley, E. \u0026amp; Vasseur, E. Graduate Student Literature Review: The effect of housing systems on movement opportunity of dairy cows and the implications on cow health and comfort. \u003cem\u003eJ Dairy Sci\u003c/em\u003e. \u003cstrong\u003e104(6)\u003c/strong\u003e, 7315-7322 (2021).\u003c/li\u003e\n\u003cli\u003eSt John, J., Rushen, J., Adam, S. \u0026amp; Vasseur, E. Making tiestalls more comfortable: I. Adjusting tie-rail height and forward position to improve dairy cows\u0026apos; ability to rise and lie down. \u003cem\u003eJ Dairy Sci\u003c/em\u003e. \u003cstrong\u003e104(3)\u003c/strong\u003e, 3304-3315 (2021).\u003c/li\u003e\n\u003cli\u003eBoyer, V. \u003cem\u003eet al\u003c/em\u003e. Making tiestalls more comfortable: III. Providing additional lateral space to improve the resting capacity and comfort of dairy cows. \u003cem\u003eJ Dairy Sci\u003c/em\u003e. \u003cstrong\u003e104(3)\u003c/strong\u003e, 3327-3338 (2021).\u003c/li\u003e\n\u003cli\u003eMcPherson, S. \u0026amp; Vasseur, E. Making tiestalls more comfortable: IV. Increasing stall bed length and decreasing manger wall height to heal injuries and increase lying time in dairy cows housed in deep-bedded tiestalls. \u003cem\u003eJ Dairy Sci\u003c/em\u003e. \u003cstrong\u003e104(3)\u003c/strong\u003e, 3339-3352 (2021).\u003c/li\u003e\n\u003cli\u003eBahadi, M., Ismail, A. A. \u0026amp; Vasseur, E. Fourier Transform Infrared Spectroscopy as a Tool to Study Milk Composition Changes in Dairy Cows Attributed to Housing Modifications to Improve Animal Welfare. \u003cem\u003eFoods\u003c/em\u003e \u003cstrong\u003e10(2)\u003c/strong\u003e, 450; 10.3390/foods.10020450 (2021).\u003c/li\u003e\n\u003cli\u003eKuznetsova, A., Brockhoff, P. B. \u0026amp; Christensen, R. H. lmerTest package: tests in linear mixed effects models. \u003cem\u003eJ Stat. Softw.\u003c/em\u003e \u003cstrong\u003e82(1)\u003c/strong\u003e,1-26 (2017).\u003c/li\u003e\n\u003cli\u003eLenth, R. V. emmeans: Estimated Marginal Means, aka Least-Squares Means (ed. 1.6.1). R package (2021).\u003c/li\u003e\n\u003cli\u003eSoyeurt, H. \u003cem\u003eet al\u003c/em\u003e. Mid-infrared prediction of bovine milk fatty acids across multiple breeds, production systems, and countries. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e94(4)\u003c/strong\u003e, 1657-1667 (2011).\u003c/li\u003e\n\u003cli\u003eEskildsen, C. E. \u003cem\u003eet al\u003c/em\u003e. Quantification of individual fatty acids in bovine milk by infrared spectroscopy and chemometrics: understanding predictions of highly collinear reference variables. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e97(12)\u003c/strong\u003e, 7940-7951 (2014).\u003c/li\u003e\n\u003cli\u003eVisentin, G., Penasa, M., Gottardo, P., Cassandro, M. \u0026amp; De Marchi, M. Predictive ability of mid-infrared spectroscopy for major mineral composition and coagulation traits of bovine milk by using the uninformative variable selection algorithm. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e99(10)\u003c/strong\u003e, 8137-8145 (2016).\u003c/li\u003e\n\u003cli\u003eEsbensen, K. \u003cem\u003eet al\u003c/em\u003e. \u003cem\u003eMultivariate Data Analysis : An Introduction to Multivariate Analysis, Process Analytical Technology and Quality by Design\u003c/em\u003e (6th edition). CAMO, Oslo, Norway (2018).\u003c/li\u003e\n\u003cli\u003eLehman, A., O\u0026rsquo;Rourke, N., Hatcher, L. \u0026amp; Stepanski, E. \u003cem\u003eJMP for Basic Univariate and Multivariate Statistics: Methods for Researchers and Social Scientists \u003c/em\u003e(2\u003csup\u003end\u003c/sup\u003e edition). 471-511 (Sas Institute, 2013).\u003c/li\u003e\n\u003cli\u003eBradley, M. \u003cem\u003eCurve fitting in Raman and IR spectroscopy: basic theory of line shapes and applications\u003c/em\u003e (Thermo Fisher Scientific, Application Note 50733). https://assets.thermofisher.com/TFS-Assets/CAD/Application-Notes/AN50733_E.pdf (2007).\u003c/li\u003e\n\u003cli\u003eSilverstein, R. M., Webster, F. X. \u0026amp; Kiemle, D. J. \u003cem\u003eSpectrometric Identification of Organic Compounds \u003c/em\u003e(8th edition). 78-107 (John Wiley \u0026amp; Sons, 2015).\u003c/li\u003e\n\u003cli\u003eCollado, J. A. \u0026amp; Ram\u0026iacute;rez, F. J. Infrared and Raman spectra of histamine‐Nh4 and histamine‐Nd4 monohydrochlorides. \u003cem\u003eJ Raman Spectrosc.\u003c/em\u003e \u003cstrong\u003e30(5)\u003c/strong\u003e, 391-397 (1999).\u003c/li\u003e\n\u003cli\u003eSocrates, G. \u003cem\u003eInfrared and Raman Characteristic Group Frequencies: Tables and Charts\u003c/em\u003e (John Wiley \u0026amp; Sons, 2004).\u003c/li\u003e\n\u003cli\u003eKarabacak, M., Cinar, M., \u0026amp; Kurt, M. Molecular structure and vibrational assignments of hippuric acid: A detailed density functional theoretical study. \u003cem\u003eSpectrochim Acta A Mol Biomol Spectrosc.\u003c/em\u003e \u003cstrong\u003e74(5)\u003c/strong\u003e, 1197-1203 (2009).\u003c/li\u003e\n\u003cli\u003eIto, K. \u0026amp; Bernstein, H. J. The vibrational spectra of the formate, acetate, and oxalate ions. \u003cem\u003eCan JChem.\u003c/em\u003e \u003cstrong\u003e34(2),\u003c/strong\u003e 170-178 (1956).\u003c/li\u003e\n\u003cli\u003eAlston-Mills, B. Nitrogenous Components of Milk: G. Nonprotein Nitrogen Compounds in Bovine Milk in \u003cem\u003eHandbook of milk composition \u003c/em\u003e(ed. Jensen, R.G.)\u003cem\u003e.\u003c/em\u003e 468-472. (Academic Press, 1995).\u003c/li\u003e\n\u003cli\u003eNocek, J. E. Bovine Acidosis: Implications on Laminitis. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e80(5)\u003c/strong\u003e, 1005-1028 (1997).\u003c/li\u003e\n\u003cli\u003eAbdela, N. Sub-acute Ruminal Acidosis (SARA) and its Consequence in Dairy Cattle: A Review of Past and Recent Research at Global Prospective. \u003cem\u003eAchiev Life Sci.\u003c/em\u003e \u003cstrong\u003e10(2)\u003c/strong\u003e, 187-196 (2016).\u003c/li\u003e\n\u003cli\u003eHumer, E. \u003cem\u003eet al\u003c/em\u003e. Invited review: Practical feeding management recommendations to mitigate the risk of subacute ruminal acidosis in dairy cattle. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e101(2)\u003c/strong\u003e, 872-888 (2018).\u003c/li\u003e\n\u003cli\u003evan de Voort, F. R. \u0026amp; Ismai. A. A. Proximate analysis of foods by mid-FTIR spectroscopy. \u003cem\u003eTrends Food Sci Technol.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 13-17 (1991).\u003c/li\u003e\n\u003cli\u003eBoyer, V., de Passill\u0026eacute;, A.M., Adam, S. \u0026amp; Vasseur, E. Making tiestalls more comfortable: II. Increasing chain length to improve the ease of movement of dairy cows. \u003cem\u003eJ Dairy Sci\u003c/em\u003e. \u003cstrong\u003e104(3)\u003c/strong\u003e, Supplemental Table S1, https://escholarship.mcgill.ca/concern/articles/p5547x13z?locale=en.\u003c/li\u003e\n\u003cli\u003eZambelis, A., Wolfe, T., \u0026amp; Vasseur, E. Validation of an ear-tag accelerometer to identify feeding and activity behaviors of tiestall-housed dairy cattle. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e102(5)\u003c/strong\u003e, 4536-4540 (2019)\u003c/li\u003e\n\u003cli\u003eDeVries, T., Dohme, F. \u0026amp; Beauchemin, K. Repeated ruminal acidosis challenges in lactating dairy cows at high and low risk for developing acidosis: Feed sorting. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e91(10)\u003c/strong\u003e, 3958-3967 (2008).\u003c/li\u003e\n\u003cli\u003eMiller-Cushon, E. \u0026amp; DeVries T. Feed sorting in dairy cattle: Causes, consequences, and management. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e100(5)\u003c/strong\u003e, 4172-4183 (2017).\u003c/li\u003e\n\u003cli\u003eJensen, R. G. Miscellaneous Factors Affecting Composition and Volume of Human and Bovine Milks. in \u003cem\u003eHandbook of milk composition.\u003c/em\u003e 237-271 (Academic Press, 1995).\u003c/li\u003e\n\u003cli\u003eHeinrichs, J., Jones, C. \u0026amp; Bailey, K. Milk components: Understanding the causes and importance of milk fat and protein variation in your dairy herd. \u003cem\u003eDairy Anim Sci.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1e-8e (1997).\u003c/li\u003e\n\u003cli\u003eArnould, V. M.-R., Reding, R., Bormann, J., Gengler, N. \u0026amp; Soyeurt, H. Review: milk composition as management tool of sustainability. \u003cem\u003eBiotechnol Agron Soc Environ.\u003c/em\u003e \u003cstrong\u003e17(4)\u003c/strong\u003e, 613-621 (2013).\u003c/li\u003e\n\u003cli\u003ePlesch, G. \u0026amp; Knierim, U.Effects of housing and management conditions on teat cleanliness of dairy cows in cubicle systems taking into account body dimensions of the cows. \u003cem\u003eAnimal\u003c/em\u003e. \u003cstrong\u003e6(8)\u003c/strong\u003e, 1360-1368 (2012).\u003c/li\u003e\n\u003cli\u003eBouffard, V. \u003cem\u003eet al\u003c/em\u003e. Effect of following recommendations for tiestall configuration on neck and leg lesions, lameness, cleanliness, and lying time in dairy cows. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e100(4)\u003c/strong\u003e, 2935-2943 (2017).\u003c/li\u003e\n\u003cli\u003eVeissier, I., Capdeville, J. \u0026amp; Delval, E. Cubicle housing systems for cattle: Comfort of dairy cows depends on cubicle adjustment. \u003cem\u003eJ Anim Sci.\u003c/em\u003e \u003cstrong\u003e82(11)\u003c/strong\u003e, 3321-3337 (2004).\u003c/li\u003e\n\u003cli\u003eMartiskainen, P., Koistinen, T. \u0026amp; Mononen, J. Cubicle dimensions affect resting-related behaviour, injuries and dirtiness of loose-housed dairy cows. in \u003cem\u003eProceedings of the XIII (13th) International Congress in Animal Hygiene\u003c/em\u003e (ed. Aland, A.). 175 (International Society for Animal Hygiene, 2007).\u003c/li\u003e\n\u003cli\u003eLombard, J. E., Tucker, C. B., von Keyserlingk, M. A. G., Kopral, C. A. \u0026amp; Weary, D. M. Associations between cow hygiene, hock injuries, and free stall usage on US dairy farms. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e93(10)\u003c/strong\u003e, 4668-4676 (2010).\u003c/li\u003e\n\u003cli\u003eMcPherson, S. \u0026amp; Vasseur, E. Graduate Student Literature Review: The effects of bedding, stall length, and manger wall height on common outcome measures of dairy cow welfare in stall-based housing systems. \u003cem\u003eJ Dairy Sci.\u003c/em\u003e \u003cstrong\u003e103(11)\u003c/strong\u003e, 10940-10950 (2020).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-4919745/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4919745/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe goal of this study was to isolate spectral fingerprints from milk Fourier transform infrared spectra that may reflect potential improvements in cow welfare, specifically comfort and ease of movement, resulting from modified housing configurations. Housing configuration modification treatments were tested across 3 animal trials, consisting of modified chain length \u003cb\u003e(TCL)\u003c/b\u003e, stall width \u003cb\u003e(SW)\u003c/b\u003e and manger wall and stall length (\u003cb\u003eMW/SL)\u003c/b\u003e configurations. The spectral analyses involved the use of principal components and mixed model analysis. Principal components were calculated from averages of mid-infrared spectra collected on the last weeks of treatment application in each of the animal trials. A significant effect of housing configuration was revealed. As an indication of animal comfort improvement, milk of cows assigned to longer chains revealed a trend of changes in multiple milk components (e.g., milk NPN, trans fatty acids, fat, and protein) that are consistent with changes in ruminal pH. These conclusions were inline with those drawn from the analysis of animal-based responses such as behavioral data and other outcomes. This study was able to reveal that housing modifications had a significant effect on milk spectra, with differences observed between the most and least restrictive treatments, translating into improved or reduced animal welfare status.\u003c/p\u003e","manuscriptTitle":"Evaluating cow welfare status from milk samples: effects of housing modifications on milk infrared spectra","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-09-23 13:18:52","doi":"10.21203/rs.3.rs-4919745/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-06-16T10:44:55+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-16T13:19:09+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-07T15:19:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"17731356015421050153177552041610283876","date":"2025-05-05T13:47:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"234410850418487612079362589922394821651","date":"2025-05-05T08:11:17+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-10-07T14:38:44+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"297243877356331806230146109620523007037","date":"2024-09-23T12:03:02+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"324526308680296159091367462745244254955","date":"2024-09-23T10:49:37+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-09-21T06:03:39+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-09-14T12:46:26+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-08-23T08:28:15+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-08-23T08:22:52+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-08-15T14:00:33+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"20b738f3-0fa4-49eb-abfc-e44c8cc6280d","owner":[],"postedDate":"September 23rd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":38036173,"name":"Biological sciences/Biological techniques/Spectroscopy"},{"id":38036174,"name":"Biological sciences/Biological techniques/Spectroscopy/Near infrared spectroscopy"}],"tags":[],"updatedAt":"2025-12-22T16:02:07+00:00","versionOfRecord":{"articleIdentity":"rs-4919745","link":"https://doi.org/10.1038/s41598-025-28557-7","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-12-17 15:58:04","publishedOnDateReadable":"December 17th, 2025"},"versionCreatedAt":"2024-09-23 13:18:52","video":"","vorDoi":"10.1038/s41598-025-28557-7","vorDoiUrl":"https://doi.org/10.1038/s41598-025-28557-7","workflowStages":[]},"version":"v1","identity":"rs-4919745","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4919745","identity":"rs-4919745","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}