Intake and production responses of pregnant and lactating Sarda ewes fed hays of different composition and quality

This thesis made of 5 chapters, is focused on the importance of forage quality on intake and productive performances of ewes during late pregnancy and lactation. Chapter 1 reviewed the literature, focusing on the role of the forage quality, and especially on fiber, in small ruminant nutrition and production. It has now been well established, especially in dairy cattle, that fiber plays a fundamental role on animal health, feeding behavior, voluntary intake, and productive performances of animals during all different physiological stages, and particularly during late pregnancy and lactation. Unfortunately, at present, there is far less information regarding the effects of fiber quality in small ruminants, and especially dairy sheep, compared to cattle, even though dairy sheep farms use proportionally larger quantities of harvested and fresh forages compared to cattle, and it is well known that forage quality is affected by many factors. Indeed, forage quality and digestibility depend on several specific plant characteristics and species, phenological cycles and maturation processes, technological harvesting aspects, such as haymaking, and the interaction of fiber with the whole farm system. Chapter 2 aimed to clarify the role that forage quality plays in late pregnancy and early lactation of Sarda dairy ewes. An experiment was designed to understand to which extent forage quality affected voluntary intake, BW, BCS, and the energy balance of ewes during pregnancy and the subsequent early lactation. For that reason, 28 pregnant Sarda dairy ewes were divided into two groups, balanced for body weight, days in pregnancy, and expected litter number (mean±SD; BW, 54±8.7 kg; DIP 115±2; 1.24 litter number). The experiment started 30 days before the expected pregnancy and ended at 30 days of lactation. Chapter 3 had the objective to evaluate the effects of three different forages with different quality and composition on the productive performances of 21 Sarda dairy ewes in their mid lactation (150 days in milk). Three groups balanced for DMI, milk yield (MY), body weight (BW) and BCS, mean±SD; DMI 2.22±0.05 kg/d MY 2.00±0.03 kg/d; BW 53.9±2.04 kg; BCS 2.98±0.08, were created. One of the three diets were randomly assigned to each group. Chapter 4 aimed to evaluate existing DMI prediction models by using the DMI data collected in the Chapter 2. Three DMI models (Pulina et al., 1998, Model 1; Gallo and Tedeschi, 2021, Model 2; and Serra, (1998), Model 3) for gestating and lactating ewes were compared to observed DMI and evaluated using the software Model Evaluation System version 2.3.4. In addition, we integrated the model of Serra, (1998) with the continuous adjusted factor (fDMI) proposed by Gallo and Tedeschi (2021), developing Model 4. The results showed that Model 4 (Serra, 1998, adjusted) was the most precise and accurate in fitting the observed DMI in early lactation in both the AAM group (r2=0.82, CCC=79, Cb=0.86 and RMSEP=0.20, Pvalue<0.001), and the OH group (r2=0.91, CCC=79, Cb=0.83 and RMSEP=0.20, Pvalue<0.001). These findings suggested that more observed data are necessary to improve these models, including in them predictors associated to diet quality. Chapter 5 reports the general conclusions and practical implications of this thesis.