Introduction
By-products obtained from the olive oil industry (essentially, olive cake and olive leaves) offer an alternative source of nutrients for ruminant livestock in the Mediterranean areas, where 90% of the world olive oil production takes place. Olive cake is obtained in high quantities (Alburquerque et al., Reference Alburquerque, Gonzalez, García and Cegarra2004) and its use, in practical ruminant feeding, has been widely reported, either fresh, dry or ensiled. However, a new centrifugation system for olive oil extraction (two-stage system) has been adopted over the last 10 years in Spain, Italy and Greece. In this system, in comparison with the traditional three-stage procedure, two main products are obtained after centrifugation of the olives: olive oil and a wet by-product, two-stage olive cake (TSOC). The TSOC includes water extract and, as a consequence, polyphenols with potential antinutritional activity, such as condensed tannins (CT). A better understanding of the comparative digestive capacity between sheep and goats in using TSOC is essential in order accurately to formulate diets including this by-product. Comparative research on sheep and goat digestive capacity is scarce, especially on the use of tannin-containing foods (Núñez-Hernandez et al., Reference Núñez-Hernandez, Wallace, Holechek, Galyean and Cardenas1991). Landau et al. (Reference Landau, Perevolotsky, Bonfil, Barkai and Silanikove2000) reported that goats utilise CT-rich foods better than do sheep. In a previous work (Yáñez Ruiz et al., Reference Yáñez Ruiz, Moumen, Martín García and Molina Alcaide2004) we also pointed out that microbial protein synthesis was only significantly increased by polyethylene glycol (PEG) treatment, a tannins-binding compound, in sheep fed diets containing TSOC, in comparison with goats. However, the comparative interspecies effect of PEG on total tract digestion, nitrogen (N) losses and, the possible systemic toxic effect of CT have not been assessed yet. This information would help to understand the different response observed in sheep and goats when fed CT-rich foods.
On the other hand, the toxic effect of CT, reflected by damage in the liver, kidneys and the epithelium of the digestive tract (Kumar and Singh, Reference Kumar and Singh1984; Reed, Reference Reed1995; Hervás et al., Reference Hervás, Pérez, Giradlez, Mantecón, Almar and Frutos2003) has been considered as one of the most important factors in explaining the deleterious influence of tannins on nutrient digestibility (Bryant et al., Reference Bryant, Provenza, Pastor, Reichardt, Clausen and Du Toit1991) although this effect varies depending on type and concentration of CT.
The objective of this work was to study comparatively, in goats and sheep, the nutritive utilisation of practical diets including two-stage olive cake in the whole digestive tract. The effects of CT, present in the by-product, on nutritive utilisation of diets and on liver and kidney functions were also studied.
Material and methods
Animals and experimental diets
Six adult dry non-pregnant Granadina goats (45 ± 2.8 kg live weight) and six Segureña wethers (71 ± 3.4 kg live weight) were used. Three experimental diets were formulated: alfalfa hay (AH), AH and pellets (CO) containing rolled barley and TSOC without (AHCO) or with (AHCOP) PEG. This paper reports further results from an experiment previously published (Yáñez Ruiz et al., Reference Yáñez Ruiz, Moumen, Martín García and Molina Alcaide2004) on rumen fermentation, in sacco degradability and microbial protein synthesis. Diets were provided in two equal meals at 0900 and 1600 h. Ingredients and chemical composition of experimental diets are shown in Table 1.
Ingredients and chemical composition of the experimental diets†
† AH = alfalfa hay, AHCO = alfalfa hay and pellets (TSOC plus barley plus mineral-vitamin mixture), and AHCOP = AHCO plus polyethylene glycol (PEG). Mineral-vitamin mixture was formulated with 300 g sodium chloride and 700 g commercial vitamin-mineral (Ovifort, NANTA, SA) with the composition (per kg): Ca 0.155 mg, P 0.097 mg, Mg 0.097 mg, Se 0.006 μg, Co 0.014 μg, Fe 0.0027 mg, Zn 0.0039 mg, Mn 0.003 mg, S 0.028 mg, retinal 120 mg, cholecalciferol 2 mg and α-tocopherol 150 mg.
‡ Chemical composition (g/kg dry matter): dry matter 876, organic matter 889, nitrogen 15.8, ether extract 2.5, neutral-detergent fibre 632, acid-detergent fibre 430, acid-detergent lignin 236, 473 total CT.
Animals were housed in individual metabolism crates and had free access to fresh water. The experimental design consisted of a Latin Square with two animals of each species allocated randomly at the start of the experiment to one of the three experimental diets. Each period consisted of 20-day adaptation, 7 days of total faecal and urine collection and 3 days of blood sampling. The experiment was carried out by following the established guidelines of the Spanish Research Council (approval no. 123/03) concerning the use of animals in research, which are in compliance with European Directive 86/609.
Liver and kidney function
Blood samples (10 ml) were collected from the jugular vein of each animal for serum profile studies before the morning feeding, on three consecutive days. Blood was kept at 4°C for 4 h, and then serum collected, transferred to a 5 ml tube and centrifuged at 500 g for 10 min. Supernatant was transferred to eppendorf tubes and kept at -20°C for analysis. Samples were assayed for alkaline phosphatase activity (ALP, IU/l) and creatinine content (mg/dl). The activity of ALP and creatinine concentration in blood have been used to examine the potential toxic effect of tannin-containing foods in ruminants and have been shown to reflect damage in the liver and kidneys, respectively (Silanikove et al., Reference Silanikove, Gilboa, Nir, Perevolotsky and Nitsan1996).
In vivo digestibility and N losses
Total faeces and urine production and feed refusals were collected for 7 days from each animal, then weighed and aliquots representing 10, 20 and 30% of daily collection, respectively, were stored at -20°C for chemical analyses. Urine was collected in buckets containing 10% (vol/vol) sulphuric acid to keep final pH below 3 and avoid N losses. Samples of food offered were also collected daily, pooled and stored at -20°C.
Laboratory analyses
Samples of offered and refused food and of faeces were dried at 60°C for 48 h, then mill-ground (1-mm screen) and analysed for dry matter (DM), organic matter (OM), ether extract (EE) and N, according to the Association of Official Analytical Chemists (1984) methods. The gross energy (GE) content was determined in an adiabatic calorimeter. The neutral- and acid-detergent fibre (NDF and ADF) and acid-detergent lignin (ADL) analyses were performed by the sequential procedure of Van Soest and Masson (Reference Van Soest and Masson1991) by using the Ankom 200/220 fibre analyser (Ankom, 2000). The NDF was assayed with sodium sulphite and without α-amylase, and both NDF and ADF were expressed without residual ash.
Free, protein and fibre bound CTs were determined in feed samples using the procedure proposed by Pérez Maldonado and Norton (Reference Pérez Maldonado and Norton1996). CTs from quebracho powder (Roy Wilson Dickson Ltd, Mold, UK) were used as standard.
The ALP activity was determined by measuring the increase in the absorbance (405 nm) due to the formation of p-nitrophenol from p-nitrophenolphiosphate esther (Inquebor, SA, Granada, Spain). The creatinine content in serum was measured as the absorbance (492 nm) after incubation of samples with picric acid and using external standard (Inquebor, SA Granada, Spain).
Statistical analysis
Experimental data were subjected to analysis of variance using the mixed linear model procedure of Statistical Analysis Systems Institute (SAS, 1985) with a model that included the random effect of the animal and fixed effects of animal species, diet and the interaction. If a value of P < 0.05 appeared, differences among means and variable interactions were tested with Bonferroni t test.
Results
Feed composition and daily intakes
Chemical composition of diets formulated with AH plus TSOC was not very different from that of AH, although a reduction of N and an increase of total CT were observed (Table 1).
The average DM offered and consumed (g) by goats were 918 and 829 ± 42.5 (AH diet), 835 and 810 ± 51.4 (AHCO diet) and 861 and 831 ± 45.6 (AHCOP diet). Corresponding amounts for wethers were 1377 and 1319 ± 39.1 (AH diet), 1250 and 1250 ± 35.1 (AHCO diet) and 1305 and 1276 ± 34.2 (AHCOP diet).
Nutrient apparent digestibility
The inclusion of TSOC (AHCO and AHCOP diets) reduced nutrient digestibility compared with AH diet in goats and wethers (Table 2), especially for NDF and ADF. A comparison between AHCO and AHCOP diets revealed only differences in NDF and ADF digestibility: values corresponding to animals fed diet AHCOP were higher (P < 0.001) than those in animals fed AHCO diet, independently of the animal species. A significant interaction between animal species and diet was observed for DM, OM and CP digestibility, which evidences the different response of wethers and goats to PEG treatment.
Effect of the experimental diets and animal species the apparent nutrient digestibility coefficients in wethers and goats
a,b,c Within a row and animal species, means without a common superscript letter differ (P < 0.05).
† AH = alfalfa hay, AHCO = alfalfa hay and pellets (TSOC plus barley plus mineral-vitamin mixture), and AHCOP = AHCO plus PEG.
‡ AS = animal species effect, D = diet effect.
N losses
In goats, N intake was not different (P>0.05) across diets (Table 3). However, wethers consumed significantly less N when TSOC was included in the diet. In both animal species faecal N excretion was increased when they were fed AHCO diet in comparison with AH and, decreased (P < 0.05) as a consequence of PEG treatment. Urinary N excretion was also higher (P < 0.05) in goats and wethers fed diet AHCO compared with AH and, unaffected by PEG treatment
Daily nitrogen intake, faecal and urinary losses, nitrogen (N) retention, total absorbable amino acids and percentage of requirements met in goats and wethers fed the experimental diets
a,b,c Within a row and animal species, means without a common superscript letter differ (P < 0.05).
† AH = alfalfa hay; AHCO = alfalfa hay and pellets (TSOC plus barley plus mineral-vitamin mixture), and AHCOP = AHCO plus PEG.
‡ AS = animal species effect, D = diet effect.
Liver and kidney function
The serum creatinine concentration was similar in goats and wethers fed the different experimental diets (Table 4). The PEG supply did not (P>0.01) modify the serum creatinine concentration in any of the animal species. The activity of ALP was increased (P < 0.05) in wethers fed diets AHCO and AHCOP in comparison with AH diet.
Creatinine concentration and alkaline phosphatase (ALP) activity in serum of goats and wethers fed the experimental diets
a,b,c Within a row and animal species, means without a common superscript letter differ (P < 0.05).
† AH = alfalfa hay, AHCO = alfalfa hay and pellets (TSOC plus barley plus mineral-vitamin mixture), and AHCOP = AHCO plus PEG.
‡ AS = animal species effect, D = diet effect.
§ Physiological ranges for creatinine (mg/dl): 0.7–1.5 (goats) and 0.7–2.0 (sheep) and for ALP (IU/l): 27–156 (goats) and 61–283 (sheep), (Merck Veterinary Manual, 2006).
Discussion
Information about TSOC chemical composition is scarce although, in general, it is similar to that of olive cake obtained from the previously used three-stage extraction technology (Molina Alcaide et al., Reference Molina Alcaide, Yáñez Ruiz, Moumen and Martín García2003). However, the two-stage olive cake includes water extract, which were obtained separately in the three-stage technology, being rich in polyphenols (Ramos Cormenzana et al., Reference Ramos Cormenzana, Juárez Jiménez and García Pareja1996). As a consequence, higher tannins content could be expected in TSOC in comparison with three-stage olive cake. In fact, we found 43.4 g/kg DM of CTs in TSOC and Nefzaoui (Reference Nefzaoui1985) found 13.6 g/kg DM in the three-stage olive cake.
Nutrient apparent digestibility
The beneficial effect of PEG supply on NDF and ADF digestibility may reflect a deleterious effect of CT contained in TSOC on the digestibility of structural carbohydrates, in agreement with other observations (Ben Salem et al., Reference Ben Salem, Nefazaoui, Ben Salem and Tisserand1999). The accuracy of the conventional detergent extraction techniques for fibrous components analysis has been criticised in the case of tannin rich forages (Reed, Reference Reed1986) by a confounding in the analysis, so the results reported here should be interpreted with caution. Nevertheless, others have used this technique and found deleterious effects of CT on NDF (Priolo et al., Reference Priolo, Waghorn, Lanza, Biondi and Pennisi2000) and ADF (Ben Salem et al., Reference Ben Salem, Nefzaoui, Ben Salem and Tisserand2000) digestibility in sheep fed, respectively, carob pulp and Acacia cyanophylla.
Treatment with PEG increased OMD of AHCO diet. Silanikove et al. (Reference Silanikove, Nitsan and Perevolotsky1994 and Reference Silanikove, Gilboa, Nir, Perevolotsky and Nitsan1996) reported improvement in energy utilisation by goats and sheep fed tannin containing shrubs (5 to 20% DM) when supplying PEG to the animals. The different degree of improvement found in both cases may be due to, among other reasons, the different quality of the experimental diets, concentration and type of CT and method of PEG application (Getachew et al., Reference Getachew, Makkar and Becker2001).
The PEG supply had a significant effect (P < 0.001) on CP digestibility only in wethers, which shows the superiority of goats to digest protein in the presence of CT. This could be explained either by a different rumen microbial population in goats and sheep as some works suggest (Brooker et al., Reference Brooker, O'Donovan, Skene, Sellick and Brooker2000; Yáñez Ruiz et al., Reference Yáñez Ruiz, Moumen, Martín García and Molina Alcaide2004) or by a more efficient overall N recycling in goats (Narjisse et al., Reference Narjisse, Elhonsali and Olsen1995). Núñez-Hernández et al. (Reference Núñez-Hernandez, Wallace, Holechek, Galyean and Cardenas1991) reported a higher N digestibility in goats than in lambs when fed mountain mahogany (Cercocarpus montanus) leaves treated or untreated with PEG. Several nutritional studies comparing sheep and goats reported no differences in nutrient digestibility when moderate- or good-quality forages were fed (Isac et al., Reference Isac, García, Aguilera and Molina Alcaide1994; Molina Alcaide et al., Reference Molina Alcaide, Martín García and Aguilera2000) but fibre digestibility of low-quality forages was greater in goats than in sheep (Doyle et al., Reference Doyle, Egan and Thalen1984). In our work only ADF digestibility of AHCO and AHCOP diets differed between goats and wethers, being higher for goats, while values for AH diet remained similar for both animal species. In other studies involving low dietary N availability, goats also showed greater N digestibility than sheep (Bohra, Reference Bohra1980; Doyle et al., Reference Doyle, Egan and Thalen1984).
N losses
The different faecal N excretion response across the experimental diets between goats and wethers should be reflected in the amount of N retained. There is a lack of consistency regarding differences in N losses between these two species. Bohra (Reference Bohra1980) found that goats had lower N losses than sheep when fed the same diet. Doyle et al. (Reference Doyle, Egan and Thalen1984) reported similar N losses when goats and sheep were fed subterranean clover hay or grass clover hay. However, with a ryegrass diet, goats lost less N than sheep did. Higher (P < 0.05) faecal N losses observed in wethers fed AHCO than AHCOP diet might indicate that some tannin-protein complexes are not fully hydrolysed in the gastro-intestinal tract, hence reducing net protein absorption in the gut. Ben Salem et al. (Reference Ben Salem, Nefazaoui, Ben Salem and Tisserand1999) also reported a significant decrease in N loss in sheep after treating a diet based on Acacia cyanophylla with increasing amounts of PEG. Our data, regarding overall N losses, confirm the superiority of goats over sheep in challenging nutritional situations such as CT-containing diets or low N intakes as it has been previously pointed out by Silanikove (Reference Silanikove2000).
The increased urinary losses in animals fed AHCO compared with those fed AH might be explained by the decrease in the protein quality of the diets, especially in amino acid content, hence increasing the inefficiency of its use (Rotz, Reference Rotz2004). Indeed the amino acid nitrogen content and profile of the by-product has been reported to be very poor (Martín García et al., Reference Martín García, Moumen, Yáñez Ruiz and Molina Alcaide2003). The lack of significant effect of PEG on urinary N losses suggests the absence of post absorptive effect of CT in both animal species.
Liver and kidney function
Previous works carried out with quebracho tannins reported the lack of toxicity (Frutos et al., Reference Frutos, Hervás, Giraldez, Fernández and Mantecón2000) at concentrations equivalent to those used in the present work. However, different results might be expected for tannins depending upon its chemical structure (Min et al., Reference Min, Barry, Attwood and McNabb2003) and, there is no information on TSOC tannins. The absorption of unmodified CT molecules is believed to be unlikely, due to their large size (Mueller-Harvey and McAllan, Reference Mueller-Harvey, McAllan and Morrison1992). However, a possible intestinal damage could lead to absorption of CT and consequent systemic toxic effects (McLeod, Reference McLeod1974).
Damage to the kidneys would most likely lead to renal failure and to changes in serum creatinine (Zhu et al., Reference Zhu and Flippish1995). The lack of effect on the creatinine serum concentrations above the normal physiological values after including TSOC in the diet and treating with PEG suggests that damage to the kidney did not occur to the experimental animals. This agrees with previous data (Yáñez Ruiz et al., Reference Yáñez Ruiz, Moumen, Martín García and Molina Alcaide2004) that showed similar urinary creatinine excretion in goats and wethers fed the same diets.
Some authors have also considered a potential hepatic damage as a consequence of tannins absorption (Silanikove et al., Reference Silanikove, Gilboa, Nir, Perevolotsky and Nitsan1996). The ALP levels are used to detect bile obstruction, i.e. a mild and progressive damage to the liver (Silanikove and Tiomkin, Reference Silanikove and Tiomkin1992). Garg et al. (Reference Garg, Agrawal and Nath1984) used a similar approach to demonstrate that feeding growing cattle with Shorea robusta did not induce systemic toxicity. However, changes in these parameters were found when tannin-related hepatoxicity occurred (Zhu et al., Reference Zhu and Flippish1995). The ALP activity levels in the serum of animals were within the physiological range although affected by PEG treatment, suggesting that no damage to the liver occurred (Merck Veterinary Manual, 2006). A longer-term study of the effect of feeding similar diets on serum metabolites might be worthy.
Conclusion
Condensed tannins from two-stage olive cake had different effects on nitrogen losses and nutrient digestibility in wethers and goats. In wethers, N digestion was affected as a result of complexation between tannins and proteins, while no effect was found in goats. Our results report different digestive response in sheep and goats fed diets containing condensed tannins and show that at the concentration provided in the experimental diets, condensed tannins from TSOC are not toxic for both animal species. However, the impact of the by-product on the digestibility, even in goats, make questionable the suitability of its use in practical conditions.
Acknowledgements
This work was funded by Junta de Andalucía (Project CAO01-003).
