Climate change represents one of the most relevant and discussed topics in the agricultural sector, with global warming beginning to show serious consequences also in the livestock sector. To analyze the impact of exposure to extreme climatic events, such as heatwaves, on the productivity of dairy cows, this Ph.D. project utilized data collected directly from farms located in the Arborea area (Oristano, Italy). The research was divided into three main areas of investigation: i) analysis at the farm level; ii) individual phenotypic and genotypic study; iii) modeling of the response of cows to heatwaves. Productive and environmental data were collected using sensors installed in barns and through milking robot software. The first study examined the effects of heatwaves on milk production and concentrate intake, classifying farms based on their seasonality in milk delivery. Data from 11 intensively managed farms (121 ± 55 cows), equipped with automatic milking systems (AMS), were analyzed, focusing on heat events occurring between August 8 and 23, 2021. The temperature-humidity index (THI), daily milk yield (MY), and caloric intake (CI) were monitored for each cow. During the heatwave, a significant increase in THI was recorded in the initial days, followed by a gradual reduction. This caused a decrease in bovine performance, which proved to be significantly lower compared to the baseline levels before the heatwave (P<0.05; data not reported). The minimum MY and CI were recorded 3 days after the THI peak and 6 days after the start of the HW. Seasonal farms (summer-to-winter production ratio (S:W) < 0.97) showed inferior performance. During the entire HW, MY decreased by 2.7% and 4.6% compared to the baseline for non-seasonal and seasonal farms, respectively (P<0.001). The second study investigated the individual response to heat events both phenotypically, identifying and quantifying parameters associated with heat stress, and genotypically, estimating the heritability of the observed traits and analyzing associated genes. Using a linear regression model with farm and parity order as predictive variables, the percentage of milk lost (MYl), milk recovered (MYr), and the duration of heat stress (DHS) were estimated, in relation to previous performance. Statistically significant differences (p < 0.001) were observed for the metrics MYl, DHS, and MYr between farms. MYl ranged between -29.95±1.57% and -6.96±1.77%, DHS between 12.4±0.39 and 14.0±0.30 days, and MYr between +17.63±1.51% and +3.68±1.53%, respectively. Parity significantly affected DHS (p < 0.001) and MYr (p < 0.05). Two SNPs were associated with MYl, one with MYr, and 29 with DHS. The heritability values for MYl, MYr, and DHS were 0.08±0.05, 0.07±0.04, and 0.04±0.03, respectively. Finally, the third study developed a dynamic model to represent the response to heat stress using a system thinking approach. The model was developed following the steps of system dynamics and subsequently evaluated on 20 cows from the studied farm. Relevant parameters of their response to HS were quantified through calibration. The proposed model was able to capture the effect of HS with high accuracy for 11 cows with MAPE < 5%, CCC > 0.6, and R2 > 0.6. The other nine cows did not show heat-sensitive behavior.
Il cambiamento climatico rappresenta una delle tematiche più rilevanti e discusse nel settore agricolo, con il riscaldamento globale che inizia a manifestare serie conseguenze anche nel comparto zootecnico. Per analizzare l'impatto dell'esposizione ad eventi climatici estremi, come le ondate di calore, sulla produttività delle bovine da latte, il presente progetto di dottorato ha utilizzato dati raccolti direttamente dagli allevamenti situati nell'area di Arborea (Oristano, Italia). La ricerca si è articolata in tre principali ambiti di indagine: i) analisi a livello di allevamento; ii) studio fenotipico e genotipico individuale; iii) modellizzazione della risposta delle bovine alle ondate di calore. I dati produttivi e ambientali sono stati raccolti mediante sensori installati in stalla e tramite i software dei robot di mungitura. Il primo studio ha esaminato gli effetti delle ondate di calore sulla produzione di latte e sull'assunzione di concentrati, classificando gli allevamenti in base alla loro stagionalità nel conferimento del latte. Sono stati analizzati dati provenienti da 11 aziende con gestione intensiva (121 ± 55 vacche), dotate di sistemi di mungitura automatica (AMS), focalizzandosi sugli eventi di calore verificatisi tra l'8 e il 23 agosto 2021. L'indice di temperatura-umidità (THI), la produzione giornaliera di latte (MY) e l'assunzione calorica (CI) sono stati monitorati per ciascuna vacca. Durante l'ondata di calore, si è registrato un aumento significativo del THI nei primi giorni, seguito da una riduzione graduale. Questo ha causato una diminuzione delle performance dei bovini, che si sono rivelate significativamente inferiori rispetto ai livelli precedenti l'ondata di calore (P<0,05; dati non riportati). Il minimo di MY e CI è stato registrato 3 giorni dopo il picco di THI e 6 giorni dall'inizio della HW. Gli allevamenti stagionali (rapporto produzione estiva e produzione invernale (S:W) < 0.97) hanno mostrato performance inferiori. Durante tutta la HW, MY è diminuito del 2,7% e del 4,6% rispetto alla baseline rispettivamente per gli allevamenti non stagionali e stagionali (P<0,001). Il secondo studio ha approfondito la risposta individuale agli eventi di calore sia dal punto di vista fenotipico, identificando e quantificando parametri associati allo stress da caldo, sia dal punto di vista genotipico, stimando l'ereditabilità dei caratteri osservati e analizzando i geni associati. Utilizzando un modello di regressione lineare con azienda e ordine di parto come variabili predittive, si è stimata la percentuale di latte perso (MYl), il latte recuperato (MYr) e la durata dello stress da caldo (DHS), in relazione alle performance precedenti. Differenze statisticamente significative (p < 0,001) sono state osservate per le metriche MYl, DHS e MYr tra gli allevamenti. Il MYl variava tra -29,95±1,57% e -6,96±1,77%, il DHS tra 12,4±0,39 e 14,0±0,30 giorni, il MYr tra +17,63±1,51% e +3,68±1,53%, rispettivamente. La parità ha influito significativamente sul DHS (p < 0,001) e sul MYr (p < 0,05). Due SNP sono stati associati a MYl, uno a MYr e 29 a DHS. I valori di ereditabilità per MYl, MYr e DHS sono stati rispettivamente 0,08±0,05, 0,07±0,04 e 0,04±0,03. Infine, il terzo studio ha sviluppato un modello dinamico per rappresentare la risposta allo stress da caldo, utilizzando l’approccio system thinking. Il modello è stato sviluppato seguendo i passaggi della system dynamics e successivamente valutato su 20 vacche dell'azienda oggetto dello studio. I parametri rilevanti della loro risposta all'HS sono stati quantificati con la calibrazione. Il modello proposto è stato in grado di catturare l'effetto dell'HS con elevata accuratezza per 11 vacche con MAPE < 5%, CCC > 0,6 e R2 > 0,6. Le altre nove vacche non hanno mostrato un comportamento sensibile al caldo.
L'adattamento delle vacche da latte allo stress da calore: modellizzazione e gestione
CRESCI, ROBERTA
2025
Abstract
Climate change represents one of the most relevant and discussed topics in the agricultural sector, with global warming beginning to show serious consequences also in the livestock sector. To analyze the impact of exposure to extreme climatic events, such as heatwaves, on the productivity of dairy cows, this Ph.D. project utilized data collected directly from farms located in the Arborea area (Oristano, Italy). The research was divided into three main areas of investigation: i) analysis at the farm level; ii) individual phenotypic and genotypic study; iii) modeling of the response of cows to heatwaves. Productive and environmental data were collected using sensors installed in barns and through milking robot software. The first study examined the effects of heatwaves on milk production and concentrate intake, classifying farms based on their seasonality in milk delivery. Data from 11 intensively managed farms (121 ± 55 cows), equipped with automatic milking systems (AMS), were analyzed, focusing on heat events occurring between August 8 and 23, 2021. The temperature-humidity index (THI), daily milk yield (MY), and caloric intake (CI) were monitored for each cow. During the heatwave, a significant increase in THI was recorded in the initial days, followed by a gradual reduction. This caused a decrease in bovine performance, which proved to be significantly lower compared to the baseline levels before the heatwave (P<0.05; data not reported). The minimum MY and CI were recorded 3 days after the THI peak and 6 days after the start of the HW. Seasonal farms (summer-to-winter production ratio (S:W) < 0.97) showed inferior performance. During the entire HW, MY decreased by 2.7% and 4.6% compared to the baseline for non-seasonal and seasonal farms, respectively (P<0.001). The second study investigated the individual response to heat events both phenotypically, identifying and quantifying parameters associated with heat stress, and genotypically, estimating the heritability of the observed traits and analyzing associated genes. Using a linear regression model with farm and parity order as predictive variables, the percentage of milk lost (MYl), milk recovered (MYr), and the duration of heat stress (DHS) were estimated, in relation to previous performance. Statistically significant differences (p < 0.001) were observed for the metrics MYl, DHS, and MYr between farms. MYl ranged between -29.95±1.57% and -6.96±1.77%, DHS between 12.4±0.39 and 14.0±0.30 days, and MYr between +17.63±1.51% and +3.68±1.53%, respectively. Parity significantly affected DHS (p < 0.001) and MYr (p < 0.05). Two SNPs were associated with MYl, one with MYr, and 29 with DHS. The heritability values for MYl, MYr, and DHS were 0.08±0.05, 0.07±0.04, and 0.04±0.03, respectively. Finally, the third study developed a dynamic model to represent the response to heat stress using a system thinking approach. The model was developed following the steps of system dynamics and subsequently evaluated on 20 cows from the studied farm. Relevant parameters of their response to HS were quantified through calibration. The proposed model was able to capture the effect of HS with high accuracy for 11 cows with MAPE < 5%, CCC > 0.6, and R2 > 0.6. The other nine cows did not show heat-sensitive behavior.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD SDC Thesis_Cresci Roberta.pdf
accesso aperto
Licenza:
Tutti i diritti riservati
Dimensione
2.09 MB
Formato
Adobe PDF
|
2.09 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/310097
URN:NBN:IT:IUSSPAVIA-310097