Nutritive evaluation of olive cake and tomato pomace for small ruminant feeding

Resumen

The use of agroindustrial by-products in animal feeding requires precise information on their nutritional value. However, agroindustrial by-products are heterogeneous and as consequence, values for chemical composition and nutritive parameters are highly variable, and the information related is scarce. For that reason, the objective of this Doctoral Thesis is to evaluate the nutritive value of two by-products that are produced in huge amounts in Spain: olive cake (OC) and tomato pomace (TP). In order to achieve this objective, six Experiments (five in vitro and one in vivo) were conducted. The objective of Experiment 1 (Paper 1) was to analyse the influence of storage time and further processing: crude OC (COC), exhausted OC (subjected to a second oil extraction; EOC) and cyclone OC (obtained from a cyclone separator; CYOC) on nutritive value of OC samples. Twelve samples (six COC and six EOC) were obtained monthly from the same pond from 1 to 6 storage months, and nine samples (three COC, three EOC and three CYOC) were obtained monthly from a different pond from 6 to 9 months storage. Chemical composition was analysed, and OC samples were fermented in vitro with sheep rumen fluid. Increasing storage time up to 6 months decreased sugars and total soluble polyphenols (TSP) content but increased fibre content in OC. Dry matter effective degradability (DMED) decreased linearly (P < 0.001) by 35.9 and 45.5% as storage time augmented from 1 to 6 months for COC and EOC, respectively. COC had lower DMED values than EOC (averaged values 0.255 and 0.294 g/g, respectively). Both potential gas production (PGP) and average gas production rate (AGPR) were lower (P ≤ 0.018) in COC compared with EOC, which was attributed to the high ether extract (EE) content of COC (≥ 86 g/kg dry matter (DM)). For samples stored longer than 6 months, CYOC had greater (P < 0.05) DMED than COC and EOC (averaged values 0.207, 0.164 and 0.164 g/g, respectively). The results indicate that ruminal degradability of OC is reduced with advancing storage time, but only subtle changes were observed during the first two months. CYOC showed greater degradability than COC and EOC, but differences between COC and EOC became negligible after five storage months. The objective of Experiment 2 (Paper 2) was to determine the variability in the chemical composition and in vitro ruminal fermentation of OC by-products. Forty-two OC samples with different storage times (1–14 months) and processing (25 COC, 9 EOC and 9 CYOC) were fermented in vitro with sheep ruminal fluid. EOC samples had a lower EE content than COC and CYOC (15.9, 110 and 157 g/kg DM, respectively), but greater neutral detergent fibre (NDF) (645, 570 and 441 g/kg DM) and acid detergent insoluble nitrogen (ADIN) (9.76, 8.10 and 8.05 g/kg DM) content. EOC had the greatest (P < 0.05) AGPR, whereas the greatest fermented organic matter (FOM) was obtained for EOC and CYOC. The best single predictor of the AGPR was total sugars content (R2 = 0.898), whereas NDF was the best one for FOM (R2 = 0.767; P < 0.001). Statistical models using storage time as a predictor variable had lower accuracy and R2 values than those from the chemical composition. In summary, the nutritive value of OC was highly dependent on its processing, but its EE content did not negatively affect ruminal fermentation parameters, which could be estimated from either carbohydrate composition or storage time. The aim of this Experiment 3 (Paper 5; Appendix) was to estimate in vitro the nutritive value of four samples of COC with different particle size obtained during the drying process from the same trommel at different stages of the screening procedure. Particles size was greater than 3 mm (COC3), 2 mm (COC2) and 1 mm (COC1) or smaller than 1 mm (COC0). Alfalfa hay and barley straw were included in the study for comparative purposes. Chemical composition was analysed in all samples. Gas production kinetics was determined in 120 h in vitro incubations with sheep rumen fluid as inoculum, and fermentation parameters and in vitro dry matter digestibility (DMD) were analysed after 24 h of incubation. COC3 and COC0 samples had lower contents of NDF (415 and 391 g/kg DM), but greater crude protein (CP) (96.6 and 114 g/kg DM) and EE (220 and 242 g/kg DM) contents, than COC2 and COC1 (627 and 624; 42.4 and 48.7; 102 and 124 g/kg DM, respectively). DMD of COC samples ranged from 39.4 to 58.7%, and was lower than that for alfalfa hay (67.5%) for all samples, but greater than that for barley straw (42.4%) for COC0 and COC3 samples. COC3 and COC0 samples had greater (P < 0.05) PGP (103 and 66.2 ml/g DM, respectively) and total volatile fatty acids (VFA) production (608 and 600 μmol, respectively) than COC2 and COC1, but values were lower than those for alfalfa hay (202 ml/g DM and 1023 μmol, respectively) and barley straw (198 ml/g DM and 733 μmol). The results indicate that decreasing particle size to <1 mm increased the fermentation potential of COC, but its nutritive value was still slightly lower than that of barley straw. The similar quality observed for COC3 and COC0 samples was attributed to the high sugar content of COC3 samples. The objective of Experiment 4 (Paper 3) was to determine the variability in nutritive value for ruminants of TP samples and analyze its effect on in vitro fermentation when it was included in a high-concentrate diet. Twelve TP samples were obtained from two processing plants at weekly intervals and analysed for chemical composition, in vitro rumen fermentation, and intestinal digestibility. The chemical composition of TP did not differ between processing plants and only slight variations were observed among sampling times. Tomato pomace had a low DM content (< 300 g/kg), a high content of NDF, CP and EE (572, 160, and 82.7 g/kg DM on average, respectively), and was rapidly fermented in the rumen. Protein degradability at 16 h in situ incubation was 510 g/kg and in vitro intestinal digestibility of CP (IDCP) was low (430-475 g/kg). Replacing soybean meal and barley straw by dried TP increased the in vitro fermentation rate and the production of VFA and reduced NH3-N concentrations without affecting methane. In summary, TP samples showed little variability in nutritive value over sampling time and TP of up to 180 g/kg could be included in highconcentrate diets without negatively affecting rumen fermentation. The objective of Experiment 5 (Paper 6; Appendix) was to determine the effect of including agroindustrial by-products, such as dried distillers grains with solubles (DDGS), dried citrus pulp (DCP) and EOC in diets for dairy goats on nutrient digestibility, ruminal fermentation, methane production, urinary excretion of purine derivatives (PD) and milk yield and composition. Twelve Murciano-Granadina goats were used in a cross-over design trial with two periods. In each period, six goats received daily a control diet composed of 1 kg of alfalfa hay and 1 kg of high-cereal concentrate and other six goats received a diet (BYP) composed of 1 kg of alfalfa hay and 1 kg of a concentrate including corn DDGS, DCP and EOC in proportions of 180, 180 and 80 g/kg concentrate (as fed basis), respectively. Diet had no effect on total DM intake, but intake of alfalfa hay, CP and EE was greater (P ≤ 0.042) for the BYP than for control group. There were no differences between diets in nutrients apparent digestibility, with the exception of fat which was greater (P = 0.002) for BYP compared with control diet. Although fecal nitrogen (N) tended to be greater (P = 0.063) for BYP-diet, there were no differences in N utilization. Compared to control, milk yield tended to be greater (P = 0.056) and daily production of milk CP, fat, whey protein and total solids (TS), as well as milk gross energy, were greater (P ≤ 0.031) for BYP diet. Diet BYP promoted a healthier fatty acids (FA) profile in milk, as the concentration of C12:0, C14:0 and C16:0 FA was or tended to be lower (P ≤ 0.080) and that of polyunsaturated FA (PUFA) was greater (P = 0.001) in the milk of BYPfed goats compared with those fed the control diet. Diet had no effect on ruminal parameters (pH, VFA and NH3-N concentrations) and methane emissions, but urinary excretion of total PD tended to be lower (P = 0.070) in BYP-fed goats than in those fed the control diet. A mixture of corn DDGS (180 g), DCP (180 g) and EOC (80 g) could replace 44% of cereal grains and protein feeds in the concentrate for dairy goats without compromising nutrient utilization, ruminal fermentation or milk yield, and led to a healthier FA profile in milk. The objective of Experiment 6 (Paper 4) was to verify lower 15N abundances of ADIN than of total N and neutral detergent insoluble N (NDIN) in different feeds, and to assess the magnitude of ADIN contamination. Samples from one corn crop (aerial biomass, grain and root) and two sunflower (SF) crops (aerial biomass and seeds) were analysed for 15N abundance and content of total N, NDIN and ADIN. Samples had been 15N-labelled by fertilizing with either 15NH4NO3 or NH4 15NO3, and there were three replicates per type of sample. The 15N-abundance of total N and NDIN was similar (P > 0.05) for all samples, whereas that of ADIN was lower (P < 0.001 to 0.037) than values for total N and NDIN for all samples, excepting for SF seeds from one crop. Estimates of N-contamination of ADIN from cetyltrimethylammonium bromide (CTAB) were not affected by the fertilizer used to label N (P ≥ 0.207), and values (g/g) ranged from 0.129 for corn grains to 0.447 for corn aerial biomass. N-contamination of ADIN was positively correlated with NDF, acid detergent fibre (ADF), hemicelluloses and cellulose content (P < 0.001; r = 0.770 to 0.905; n = 18 and 22 for corn and SF, respectively), and negatively (P < 0.001 to 0.048) with ADIN content (expressed as proportion of ADF). The negative correlations (P < 0.001) between N-contamination of ADIN and acid detergent lignin (ADL)/ADF ratio observed for all samples would indicate that N-contamination of ADIN was reduced by cell wall lignification. In conclusion, the results showed that the use of CTAB in the ADF analysis resulted in increased ADIN values, and therefore overestimated the amount of N associated to ADF, questioning the use of ADIN as a parameter indicative of N availability.