Effects of replacing soybean meal with alternative sources of protein on nutrient digestibility and energy value of sheep diets

The study determined the potential of eight locally produced protein sources to replace soybean meal (SBM) in the diet of sheep. Three in vivo digestibility trials were conducted using a Latin square (3x3, 4x4 and 4x4) experimental design with castrated Chios rams. The authors estimated nutrient digestibility coefficients and energy value of diets with soybean meal (control), lupin seeds (LS), rapeseed meal (RSM), sunflower meal (SFM), fababean seeds (FBS), vetch seeds (VS), pea seeds (PS), flaxseeds (FS), and chickpea seeds (CS). The results showed that the SBM, RSM, and PS diets had similar nutrient digestibility and energy value. The SFM diet had lower organic matter (OM) digestibility than the SBM diet. The FS and SFM diets had similar crude protein (CP) digestibility to the SBM diet. Additionally, FS, VS, and SBM diets had similar energy value. Furthermore, SBM, SFM, VS and FS diets had similar dry matter (DM), neutral detergent fibre (NDF), hemicelluloses, and cellulose digestibility coefficients. The SBM, LS, FBS, and CS diets had similar DM, OM, NDF, hemicelluloses, and cellulose digestibility coefficients. Additionally, SBM, LS, and CS diets had similar energy value. This study reveals that diets with RSM, PS, FS, and LS, compared with diets with SBM, did not have adverse effects on nutrient digestibility and energy value. These results tend to support the idea that some locally available protein sources seem to have the potential to replace SBM in sheep diets.


Introduction
The rapid growth in human population has increased the demand for proteins from animals.Soybean meal (SBM) has been established as the main protein source for animal nutrition.However, high prices, because of the increased costs of production and transportation, and fluctuation in production, have necessitated the search for alternative locally produced protein sources for feeding livestock.Moreover, customers are aware of the usage of genetically modified SBM, although there is no clear scientific evidence for its association with health problems (Domingo & Bordonada, 2011).These challenges were recognized by international bodies such as the European Union, which encourage self-reliance via the support of cultivation and production of protein supplements from locally produced ingredients.
The interest in legume seeds as a source of protein in ruminant feeding has increased, mainly because of the positive ecological functions in crop rotation of these plants.Moreover, SBM could be replaced in animal diets by many other feeds and by-products, such as oilseeds and oilseed residues.However, these two categories of feeds have different CP and ether extract (EE) content.According to Fiorentini et al. (2013), lipid metabolism is highly variable in the rumen and total tract owing to factors such as the nature and concentration of lipids in the diet, the types of chemical and physical treatments applied to the feeds, and their characteristics, and the forage to concentrate ratio in the diet.Additionally, genetic selection for better crops normally focuses on better crop yield, with limited interest in their nutritional value.It should not be forgotten that local climate may affect the nutritional value of crops.Other factors include processing methods.For these reasons, continuous evaluation of the nutritional value of animal feeds should be carried out for each production area.With this in mind, the eight most promising protein sources were selected for comparison with SBM in terms of digestibility of diet nutrients.The protein sources included two meals (SFM and RSM) and six untreated ground seeds (LS, FBS, VS, PS, linseed or FS, and CS).Most of these feeds used to be cultivated extensively in past decades in Europe and the wider Mediterranean area.
There are many types of SFM, depending on the variety of sunflower seeds and the method of processing.Molina et al. (2003) reported that SFM could be a good supplement to low degradable protein feedstuffs.Additionally, the increased interest in biodiesel has renewed the demand for oils from rapeseed.The RSM is one of the main protein sources used in animal feeding, and has been used successfully as a protein feed in dairy cow diets (Mulrooney et al., 2009).
The PS and FBS are interesting crops to grow in cool temperate climate conditions, as they can partly replace imported protein sources owing to their relatively high CP content (Larsen et al., 2009).Additionally, LS can successfully replace soybean meal in dairy cow diets (Froidmont & Bartiaux-Thill, 2004).Compared with PS, LS contains more nitrogen (N) and EE, and therefore qualifies as a high-quality feedstuff for ruminants (Froidmont & Bartiaux-Thill, 2004).
The chickpea seed is cultivated principally as a legume crop.According to the literature, it could replace SBM in lamb fattening diets (Hadjipanayiotou, 2002).Furthermore, increasing FS inclusion in cow diets did not affect digestibility or milk yield (Martin et al., 2016), while, according to Schroeder et al. (2014), feeding steers with FS did not affect CP total tract digestibility.The VS is a multipurpose, cool season, annual legume grown for livestock feed and soil fertility improvement in Mediterranean environments.The hay and grain could be used as sources of protein in ruminant diets (Gul et al., 2008).
The objectives of this study were to determine the effects of eight alternatives to SBM protein sources on apparent total tract nutrient digestibility and energy value of sheep diets.Since no single study has ever compared these N supplements in terms of energy value and nutrient digestibility of diets, the main goal of this study was to evaluate the capability of these protein sources to totally replace SBM in sheep diets and to provide crucial information that could be used to make this replacement feasible.

Materials and methods
The management and care of animals were approved by the protocol of Aristotle University of Thessaloniki, Greece.Three experiments were carried out.An in vivo digestibility trial (Trial 1) was conducted with three castrated rams, aged 8 to 9 months and weighing 42±1 kg, using three diets in a 3x3 Latin square experimental design.The three diets were isocaloric and were formulated to meet the maintenance energy requirements of rams (Table 1).The tested feeds were SBM, RSM and PS.All the seeds were ground into meal.Experimental diets were fed in two equal amounts at 08:00 and 17:00 each day (Tables 1 and 2).Each of the three periods consisted of 8 days' adaptation and 7 days of sample collection.Water was freely accessible through individual drinkers.Faeces and urine were collected using the total collection for seven days for each treatment, weighed at approximately 08:00 each day, and composted by treatment and ram.Samples were stored at a temperature of 2-3 °C until all the samples for that collection period had been taken.Diet samples were taken for laboratory analysis by grab sampling as the feed allowances were being weighted.Diets were made up of alfalfa hay (450 g/day), ground corn grain (390 g/day) and 145 g/day SBM (SBM diet), 170 g/day RSM (RSM diet) and 145 g/day PS (PS diet).
The same procedure was used for the second digestibility trial (Trial 2).This experiment used rams aged 15 to 16 months, with a live bodyweight of 58±1kg.The experimental design was a 4x4 design.Diets corresponded to maintenance requirements for energy with protein sources being SBM, SFM, VS or FS (Tables 1 and 2).Diets were made of alfalfa hay (580 g/day), ground corn grain (470 g/day), and 170 g/day FS (FS diet), 245 g/day SFM (SFM diet), 200 g/day VS (VS diet) and 200 g/day SBM (SBM diet).
The same procedure was used for the third digestibility trial (Trial 3).This time the experiment used rams aged 20 to 21 months with a live bodyweight of 61±1kg.The tested feeds were SBM, LS, FBS, and CS (Tables 1 and 2).Diets were made up of alfalfa hay (580 g/day), ground corn grain (490 g/day) and 205 g/day SBM (SBM diet), 190 g/day LS (LS diet), 215 g/day FBS (FBS diet) and 215 g/day CS (CS diet).
Feed and composite faecal samples were ground to pass through a 1-mm screen (AOAC, 2006; method 950.02B).Dry matter (DM) was determined by drying in an oven (AOAC, 2006;method 934.01).Ash was determined by ignition in a muffle furnace (AOAC, 2006;method 942.05).The CP was measured as Kjeldahl N x 6.25 (AOAC, 2006;method 984.13).Ether extract (EE) was determined according to AOAC (2006;method 920.39).Acid detergent fibre (ADF) and acid detergent lignin (ADL) were determined using AOAC (2006;method 973.18) and NDF using AOAC (2006;official 2002.04).The NDF was assayed without a heat stable amylase and expressed inclusive of residual ash.Additionally, ADF was expressed inclusive of residual ash.The hemicellulose value was measured as the difference between NDF and ADF, and the analyses were sequential on the same sample.Gross energy (GE) was measured with an adiabatic bomb calorimeter (Parr Instrument Co., 1970)., 1990).Urinary energy (UE) was calculated using the equation UE (kcal/g) = 0.027 + 0.119 (UN%) (Street et al., 1964).According to Blaxter & Clapperton (1965), gaseous energy (G) is broadly proportional to the apparent digestibility of the diet and at a maintenance level of nutrition can be calculated by the following equation: 100G/I = 3.67 + 6.22 (I-F)/I Where: G is the energy lost as methane I is the intake of energy F is the faecal energy Additionally, digestible energy (DE) was calculated as DE = GE-F and ME by the difference ME= DE-UE-G.For all the methods, measurements were made in triplicate and standards were included in each run of each method.Digestibility of nutrient coefficients and energy value were analysed statistically using S-Plus (2001).Significance was declared at P <0.05.

Results
The results from Trial 1 are given in Table 3. Statistical analysis revealed that the SBM, RSM and PS diets had no differences (P >0.05) in terms of DM, OM, CP, EE, NDF, ADF, hemicelluloses, cellulose DE/GE, ME/GE and ME/DE parameters.The results from Trial 2 are given in Table 4. Statistical analysis showed that the SFM diet is not similar to the SBM diet in terms of OM and DE/GE parameters.Also, the SFM diet is not similar to the FS diet in terms of DE/GE parameters.Additionally, the VS diet is not similar to the SBM diet in terms of CP parameter.Furthermore, the FS diet is not similar to the SFM, VS, and SBM diets in terms of EE parameter.Additionally, the FS diet is not similar to the SFM and VS diets in terms of ADF parameter.
The results from Trial 3 are given in Table 5. Statistical analysis revealed that the SBM diet is not similar to the FBS diet in terms of CP, DE/GE, and ME/GE parameters.Also, the SBM diet is not similar to the CS diet in terms of the CP parameter.Additionally, the CS diet is not similar to the FBS diet in terms of the EE parameter.Furthermore, the LS diet is not similar to the FBS diet in terms of the ADF parameter.

Discussion
The results of Trial 1 suggest that the nutritional value of sheep diets would not be negatively affected by replacing 100% SBM with RSM and PS.Similarly, Maxin et al. (2013) reported that RSM could replace SBM in dairy cow diets, while Petit et al. (1997) reported that PS and SBM have similar DE.Additionally, Gilbery et al. (2007) reported that replacing RSM with PS or CS in cow diets did not affect CP digestibility.It seems that RSM and PS could totally replace SBM in sheep diets, despite the significant differences in their chemical composition (CP, NDF and ADF).Additionally, the present study revealed that OM digestibility of the RSM diet was 71.8%, a value similar to that of 73.2% reported by Eghbali et al. (2011) for the same diet.Furthermore, Górka et al. (2015) reported that replacing RSM and barley grain with high-lipid (9%) byproduct pellets in heifer diets affected total tract nutrient digestibility negatively.This is not inconsistent with the current results because in the present study, replacing SBM with RSM or FS had no effect on diet digestibility coefficients, probably because lipid metabolism is highly variable owing to many factors, such as the nature and concentration of lipids in the diet (Fiorentini et al., 2013).Furthermore, in line with the current results, Hentz et al. (2012) reported that feeding wethers with increasing levels of RSM (up to15 g/kg of bodyweight) had no effect on diet NDF digestibility.On the other hand, Gilbery et al. (2007) reported that replacing RSM with PS positively affected the diet's ADF digestibility, which is in contrast with the current results, possibly because the animals in the current study were fed at maintenance level, and the digestibility of nutrients did not compromise nutrient demand at this level.As a result, differences in nutrient requirements (maintenance vs growth vs reproduction vs wool growth) might influence the digestibility coefficients of the diets in different ways.Eweedah et al. (1996) reported similar CP and ADF, and lower OM digestibility values for lamb diets with SFM or SBM.These findings are in agreement with the current results.Furthermore, Irshaid et al. (2003) reported that replacing SBM with increasing levels of SFM (up to100%) in lamb diets had no effect on the DM, OM, CP, NDF, and ADF digestibility of the diets.Additionally, it was reported in a previous study (Richardson et al., 1981), that replacing SBM with SFM did not affect DM, CP, and ADF digestibility of the diets.Data of the present study indicate the same effects.Furthermore, according to Molina et al. (2003), the addition of increasing amounts of SFM improved the in vitro DM and OM digestibility of olive cake and olive leaves.It seems that SFM diets could replace SBM in ruminant diets, with limited negative effects on the digestibility coefficients of the most nutritious components (excluding OM digestibility) and on their energy value (excluding digestibility of energy).The chemical composition of commercial flours varies widely, which also affects their nutritional value.
Similar energy values of the VS diet compared with the SFM, SBM and FS diets, but lower CP and similar DM, OM, EE, NDF and ADF digestibility in comparison with the SBM diet in Trial 2, are consistent with results reported by Seifdavati et al. (2012).Replacing SBM with VS affected the CP digestibility of the diets negatively.Although these two diets were isocaloric, the VS diet had a higher energy to protein ratiothe VS diet had lower CP contentwhich negatively affected the nitrogen sufficiency in the rumen, resulting in a possible increase in rumen fermentable protein to meet the needs of microorganisms, thereby reducing the efficiency of CP utilization.Furthermore, it was confirmed in the present study that the digestibility of NDF is not affected by the protein source of the diet, which is consistent with the findings of Seifdavati et al. (2012).Additionally, this study confirmed the results of Ohsihita et al. (1997) that the addition of SBM in ruminant diets has no effect on digestibility of hemicelluloses.
According to the literature, the addition of FS to steer and cow diets had limited effects on digestibility and milk yield (Schroeder et al., 2014;Martin et al., 2016).Furthermore, Gonthier et al. (2004) reported that the addition of FS to cow diets tended to increase their DM, OM, CP, NDF, and GE digestibility.These reports are in line with the current findings.Additionally, the FS diet showed higher EE digestibility than the SFM, VS and SBM diets, probably because of the superior EE content of FS.In line with the results of the present study, Bernard (1990) reported that diets with high EE content had high EE digestibility coefficients.The high EE digestibility of the FS diet was observed concurrently with a higher percentage of endogenous EE in the faeces of the animals fed with diets of low EE content (SFM, VS, and SBM diets).This might result in recording lower apparent total tract digestibility than it is, because of the high proportion of endogenous EE in the faeces, in comparison with EE derived from the feed.The results of the present study indicate that FS did not negatively affect the energy value and nutrient digestibility of sheep diets.These findings are significant and support the central idea of this study that sheep diets with different chemical composition exhibit similar nutritional value.Furthermore, the authors found that replacing FS with SFM or VS affected the ADF digestibility of diets negatively, suggesting that FS is a better ingredient than SFM or VS.This effect may be related to the high ether extract content of FS, which overlaps the feed, and prevents microorganisms from digesting ADF.Seifdavati et al. (2012) reported that replacing SBM with VS in lamb diets had a negative effect on ADF digestibility.Furthermore, Eweedah et al. (1996) reported that digestibility of ADF remained unaffected when SFM replaced SBM in diets.These results are partially in agreement with the current results, probably owing to differences in the chemical composition of SFM (Irshaid et al., 2003).
Results of Trial 3 tend to support the notion that CS and FBS seem to have the potential to replace SBM.Similarly, Tufarelli et al. (2012) reported that FBS could replace SBM in dairy cow diets.Furthermore, Mustafa et al. (2000) reported that replacing SBM with CS did not affect the digestibility coefficients of nutritious components of diets.Additionally, Abreu & Bruno-Soares (1998) reported that CS, PS, and VS showed similar OM digestibility, but higher OM digestibility, when compared with LS.This variation in results could be attributed to differences in chemical composition in these feeds.Furthermore, consistent with the current result, Gebru et al. (2015) reported that the inclusion of white LS in sheep diets improved total DM and nutrient intake, nutrient digestibility, average daily gain, and feed conversion efficiency, and could be recommended not only for maintenance, but also for optimum performance of ruminants.Earlier, Moss et al. (1997) reported that feeding cattle with increasing levels of LS improved the DE of the diets.Therefore, it would seem that LS is a good protein and energy source (equal to SBM) for sheep diets, even without further processing except for grinding, with similar energy value and nutrient digestibility coefficients to SBM.
Finally, taking into account that no single study has ever compared all these N supplements in terms of diets' nutrient digestibility and energy value, the authors can say that RSM, PS, FS, and LS could be used as alternatives to SBM protein sources to form sheep diets with different chemical composition, but similar nutritional value.Additionally, SFM, CS, FBS, and VS seem to have the potential to replace SBM in sheep diets.These feeds (SFM, CS, FBS, and VS) could be used in diets of moderately productive sheep.Further investigation is needed to examine the effects of these protein sources and their treated forms on the characteristics of digestion, energy value, growth performance, and milk production of sheep in different feeding conditions.

Conclusions
This study revealed that diets with RSM, PS, FS, and LS compared with diets with SBM have no adverse effect on energy value and nutrient digestibility coefficients.These results support the idea that some locally available protein sources could replace SBM in sheep diets.

Table 3
Mean values of nutrient digestibility and energy value (expressed as DE/GE, ME/GE and ME/DE) SBM: soybean meal; RSM: rapeseed meal; PS: ground pea seed NS: not significant; 1 DE/GE: Digestible Energy/Gross Energy; ME/GE: Metabolizable Energy/Gross Energy; ME/DE: Metabolizable Energy/Digestible Energy

Table 4
Mean values of nutrient digestibility and energy value (expressed as DE/GE, ME/GE and ME/DE) Mean values of nutrient digestibility and energy value (expressed as DE/GE, ME/GE and ME/DE) a, b Means in the same row with different superscript are different at P <0.05; NS: not significant; * Treatment effect; FS: ground linseed; SFM: sunflower meal; VS: ground vetch seed; SBM: soybean meal 1 DE/GE: Digestible Energy/Gross Energy; ME/GE: Metabolizable Energy/Gross Energy; ME/DE: Metabolizable Energy/Digestible Energy Table 5 a, b Means in the same row with different superscript are different at P <0.05; NS: not significant; *Treatment effect; SBM: soybean meal; LS: ground lupin seed; FBS: ground fababean seed; CS: ground chickpea seed 1 DE/GE: Digestible Energy/Gross Energy; ME/GE: Metabolizable Energy/Gross Energy; ME/DE: Metabolizable Energy/Digestible Energy