Feeding Date-Palm Leaves Ensiled with Fibrolytic Enzymes or Multi-Species Probiotics to Farafra Ewes: Intake, Digestibility, Ruminal Fermentation, Blood Chemistry, Milk Production and Milk Fatty Acid Profile

Simple Summary Under semi-arid and arid conditions, trees and shrubs such as date palm can be used as an adequate source of feed for goats and sheep to reduce feed cost. However, the low nutritive value of such materials determines its nutritive value. Ensiling with fibrolytic enzymes or lactic-acid bacteria can be used to enhance the nutritive value of date palm leaves and other agricultural byproducts before feeding to animals. Exogenous enzymes can alter the structure of the tissue while lactic-acid bacteria improve ensiling in enhancing nutrient digestibility, resulting in improved performance (daily gain or milk production). This may enhance farmers’ gain and animal health. This is the first experiment to utilize ensiling with fibrolytic enzymes or lactic-acid bacteria to enhance the nutritive value of date palm leaves as an unconventional feed. Abstract The present experiment evaluated the feeding of date palm leaves (DPL) ensiled with fibrolytic enzymes (ENZ) or multi-species probiotics (MSP) on nutrient utilization and lactational performance of ewes. Fifty multiparous lactating Farafra ewes were used in a completely randomized design for 90 d. The treatments consisted of the control diet with a concentrate feed mixture and date palm leaves (at 60:40, DM basis) ensiled without additive (control) or DPL ensiled with ENZ or MSP replacing control DPL at 50 or 100%. Both ENZ and MSP increased (p < 0.01) DPL and total intakes, digestibility of all nutrients, concentrations of ammonia, total volatile fatty acids, acetate and propionate in the rumen. Increased milk production, concentrations of fat, lactose and energy in milk, and feed efficiency were observed with MSP and ENZ compared to the control treatment. Moreover, ENZ and MSP increased (p < 0.05) the concentrations of total n3, n6 fatty acids, polyunsaturated fatty acids and conjugated linoleic acids and decreased (p < 0.001) the atherogenicity. The differences between ENZ and MSP and between the low and high replacement levels were minor for all measured parameters. Ensiling of DPL with MSP or fibrolytic enzymes is recommended to improve feed efficiency and improve lactational performance of ewes.


Introduction
In semi-arid and arid regions, improvement of utilization of available feed resources and search of alternative feeds for ruminants are required due to the shortage of green fodders. Egypt and many other countries lack adequate availability of animal-feed ingredients, for animals. Moreover, we hypothesized that inoculating silage with MSP containing LAB, along with other probiotic bacteria such as Bacillus subtilis and Bacillus lichenifomis would facilitate maintenance of the ensiling conditions and improve the nutritive value of ensiled DPL before feeding. The present study aimed to evaluate the effects of inoculating DPL with MSP or fibrolytic enzymes during ensiling on its nutritive value as a feed for lactating Farafra ewes under the arid conditions in the New Valley area in Egypt.

Study Location
This experiment was carried out at the experimental farm of the Department of Animal Production, Faculty of Agriculture of New Valley, Al Kharga, Egypt (25 •
For assessment of the ensiling process, 200 g (fresh weight of silage was mixed with 800 mL distilled water, homogenized for 3 min with a blender and filtered through 4-layer cheesecloth. The filtrate was collected for measurement of pH using a digital pH meter, ammonia-N (NH 3 -N) according to AOAC [14], and volatile fatty acids (VFA) according to AOAC [14]. Aflatoxin F 1 concentration was measured in silage with the use of a Fluorometer, Series-4 (Vicam, Milford, MA, USA) based on the methods described by AOAC [14].

Ewes and Management
Two weeks before expected parturition, fifty lactating Farafra ewes (mean ± standard deviation: 2 ± 1.2 parity; 33.3 ± 3.04 kg body weight; 24 ± 3.3 months of age; 550 ± 10/4 g/d of previous milk production) were assigned randomly to five dietary treatments (n = 10 ewes/treatment). Ewes were randomly stratified to treatments in a completely randomized design. Ewes were individually kept in semi-opened concrete floor pens (1.5 m 2 /sheep) with free choice fresh water. Sheep were fed a diet comprising (per kg DM) 600 g of a concentrate feed mixture and 400 g of DPL ensiled without additives in the control treatment. In the other four diet treatments, ensiled (without fibrolytic enzymes of MSP) DPL of the control treatment diet was replaced with DPL ensiled with fibrolytic enzymes (ENZ50 and ENZ100) or MSP (MSP50 and MSP100) at 50 or 100% level, respectively. Ewes were first offered the allotted amounts of concentrate feed mixture in the feeder, followed by DPL after the consumption of concentrate feed. Diets were prepared to meet nutrient requirements for milk production of ewes according to NRC [15] recommendations. To ensure orts collection, feeds were offered 1.10 times above the NRC recommendations. The experiment lasted for 90 d. Individual animals were weighed at monthly intervals. Table 1 shows the chemical compositions of ingredient and experimental diets. The daily samples of diets were composited weekly and dried at 60 • C in a forced-air oven for 48 h [14] (method 930.15) before storing for chemical analyses.

Feed Intake and Nutrient Apparent Digestibility
Three digestibility trials were conducted during the last 10 d of each month using acid-insoluble ash as an internal indigestibility marker. The equations of Ferret et al. [16] were used to calculate the coefficients of apparent digestion. Feed intake was calculated as the difference between feed offered and orts from the previous day's feeding. Individual fecal grab samples were collected twice daily during the collection period at 07:00 and 15:00 h, dried at 60 • C in a forced-air oven for 48 h, and pooled per ewe.
Composited samples of dried feeds, orts and feces were ground to pass through a 1mm screen using a mill and analyzed for DM, ash, nitrogen, and ether extract (EE) according to AOAC [14] official methods. Neutral detergent fiber and lignin contents were determined according to Van Soest et al. [17]. Acid detergent fiber (ADF) content was analyzed according to AOAC [14] and expressed exclusive of residual ash. Nonstructural carbohydrates, cellulose, hemicellulose, and organic matter (OM) concentrations were calculated.

Sampling and Analysis of Rumen Fluid
On d 30, d 60 and d 90 of the experiment, ruminal fluid samples were collected from all animals in the morning at 3 h post feeding to analyze fermentation variables (VFA and NH 3 ). About 100 mL of ruminal fluid was collected from each ewe and strained through 4 layers of cheesecloth for NH 3 -N analysis [14] and VFA [18] determination. The collected samples were preserved at −20 • C pending analyses. Concentration of VFA and its individual molar proportions were determined using a gas chromatograph (Thermo Fisher Scientific, Inc., TRACE1300, Rodano, Milan, Italy) fitted with an AS3800 autosampler and equipped with a capillary column HP-FFAP (19091F-112; 0.320 mm o.d., 0.50 µm i.d., and 25 m length; J & W Agilent Technologies Inc., Palo Alto, CA, USA). A mixture of known concentrations of individual short-chain fatty acids was used as an external standard (Sigma Chemie GmbH, Steinheim, Germany) to calibrate the integrator.

Sampling and Analysis of Blood Serum
On d 30, d 60 and d 90 of the experiment, blood samples (10 mL) were collected at 4 h post feeding from the jugular vein of each ewe into clean dry tubes without anticoagulants. Collected samples were centrifuged at 4000× g for 20 min at 4 • C, and serum was decanted into 2-mL Eppendorf tubes and frozen at −20 • C pending analysis using specific kits (Stanbio Laboratory, Boerne, TX, USA) according to manufacturer instructions. Globulin concentration was calculated (total protein-albumin).

Milk Sampling and Composition
Ewes were hand-milked during the last 10 d of each experimental period at 09:00 and 21:00 h, and 10% of recorded milk yield samples was taken at each milking and composited daily for the analysis of milk components (fat, lactose, total solids, and protein) using infrared spectrophotometry (Lactostar Dairy Analyzer, Funke Gerber, Berlin, Germany).
Gross energy content in milk, fat-corrected milk (4% FCM, kg/day) and energycorrected milk (ECM, kg/day) were calculated according to Tyrrell and Reid [20]. Feed efficiency was calculated and expressed as milk yield, FCM, and ECM per unit of DM intake.

Statistical Analyses
The Shapiro-Wilk test was used to test the normal distribution of data. For the small number of variables that showed significance for the Shapiro-Wilk test, data transformation (e.g., natural log, inverse of the natural log, square root, or inverse of the square root) was applied before statistical analysis. Data were analyzed using a completely randomized design with repeated measurements in time, in which each ewe was an experimental unit using PROC MIXED of SAS (Online Version, SAS ® On Demand for Academics, SAS Inst., Inc., Cary, NC, USA). The following model was used as: where Y ijkl expressed each observation of the jth ewe in the kth sampling time given ith diet, T i expressed the diet's effect, A(T) ji expressed the ewe within each diet, P k expressed the sampling week effect, (T × P) ik expressed the interaction between the diets and sampling period, and E ijkl expressed the experimental error. Polynomial (linear and quadratic) contrasts were used to examine level responses to increasing the level of DPL separately for enzyme or MSP effect. Additionally, contrast between enzyme vs. MSP treatments was applied. The period and diet × period interactions were non-significant (i.e., p > 0.05) for most of the measurements; thus, only the main effects of diets were reported. Significance was declared at a level of p < 0.05.

Feed Intake and Apparent Nutrient Digestibility
Without differences between enzyme and MSP treatments, ensiling of DPL with enzymes or MSP increased (linear and quadratic effects, p < 0.01) both DPL and total intakes compared to the control ewes (Table 2). Feeding DPL ensiled with MSP or enzymes in-creased (p < 0.01) the digestibility of nutrients linearly (all nutrients) and quadratically (all nutrients except for NDF for MSP treatment and EE for both enzyme and MSP treatments). The digestibility of DM was greater for MSP vs. enzyme (p = 0.001), whereas NSC digestibility (p = 0.005) and cellulose digestibility (p = 0.029) were greater for enzyme compared with the MSP treatment.

Ruminal Fermentation
No differences were observed between MSP and enzyme treatments for all measured ruminal parameters except for an NH 3 -N concentration that was higher (p = 0.029) for MSP treatment than enzyme treatment (Table 3). Without affecting ruminal pH, the butyrate concentration or acetate:propionate ratio, enzyme and MSP treatments increased (linear p < 0.001 and quadratic effects p < 0.05) the concentrations of ruminalNH 3 -N, total VFA, acetate and propionate.

Blood Chemistry
Treatments did not affect serum globulin, urea-N, aspartate aminotransferase (AST), alanine aminotransferase (ALT), low density lipoprotein (LDL), high density lipoprotein (HDL), and beta-hydroxy butyric acid (BHBA) concentrations (Table 4). Both enzyme and MSP treatments increased (linear and quadratic effects, p < 0.01) the concentrations of serum total protein, albumin, glucose and antioxidant capacity. Concentration of non-esterified free fatty acid (NEFA) increased linearly with increasing PDL level treated with MSP. No differences between MSP and enzyme treatments were noted for most of the variables except for urea-N and NEFA that were greater for MSP treatment than for enzyme treatment.

Milk Yield, Composition and Fatty Acids
There were no differences between enzymes and MSP treatments for milk production and yields of milk components. However, DPL ensiled with enzymes or MSP linearly increased (p < 0.01) the production of milk, ECM, FCM, and yields of milk components (Table 5). Without affecting milk protein and ash contents, enzymes and MSP treatments increased (linear p < 0.001, and quadratic effects, p < 0.01) the concentrations of milk total solids, solids not fat, fats, lactose, and milk energy. MSP treatments showed higher (p < 0.001) contents of solids not fat and lactose compared to enzyme treatments. Moreover, treatments linearly improved (p < 0.05) feed efficiency calculated as milk:intake, ECM:intake or FCM:intake ratios.

Feed Intake and Nutrient Apparent Digestibility
Fibrolytic enzymes and MSP increased DPL intake by 7.3, 8.6, 7.1 and 8.7%, for ENZ50, ENZ100, MSP50 and MSP100 treatments, respectively compared to the control treatment indicating improved palatability with the fibrolytic enzymes and MSP addition. The improved nutrient digestibility and ruminal fermentation with treatments are another probable reason for greater feed intake. Increasing the rate of fermentation of the insoluble fraction following the use of fibrolytic enzymes, and MSP may reduce the rumen fill and consequently increase the feed intake of low-energy feed [10]. Moreover, the administration of enzymes during ensiling allows enzymes to be quickly attached to fiber and reduces the lag time resulting in increasing feed intake and digestibility [11].
It is well-documented that DPL contain high concentrations of various plant secondary metabolites such as tannins, and flavonoids, waxes, isoflavones, and lignans, which can adversely affect nutrient digestibility [21]. In the present experiment, fibrolytic enzymes and MSP improved all nutrient digestibility, with different modes of action. Improved ruminal fermentation with MSP and fibrolytic enzymes may result from greater nutrient digestibility, particularly the fiber fractions. Hamdon et al. [7] reported that DPL have low nutrient digestibility when wheat straw was replaced by DPL due to the high fiber concentration in DPL. In the present experiment, the improved CP and cellulose digestibility indicates that ensiling with the additives (enzymes and MSP) may loosen the association among the fiber bundles and between protein and fiber fractions in DPL [7]. The highest digestibility of DM and cellulose were observed with MSP50 and MSP100 treatments compared to other treatments, which indicate better rumen fermentation rates and patterns with MSP treatment [8]. Although activity of ruminal cellulolytic microbial populations was not measured in the present experiment, MSP supplementation could produce a tonic level of lactate, which would then boost a basal abundance of lactate utilizing bacteria in the rumen, thus stabilization and fluctuation of pH [22,23]. The improved ruminal environment may increase the fiber-degrading microbial communities in the rumen, resulting in improved nutrient digestion and synthesis of microbial proteins [24]. Moreover, the increased digestibility with MSP treatment may improve the interaction of MSP with the ruminal microbial flora [9,25]. Generally, probiotics may increase enzyme activity in the gastrointestinal tract and improve digestibility [8,26].
The higher digestibilities of nutrients with fibrolytic enzymes' treatment are some of the advantages of adding exogenous enzymes in ruminant diets. Exogenous fibrolytic enzymes can break the cross-linkages between cell wall substances and lignin, and solubilize cell-wall components (mainly hemicellulose) [3]. Moreover, fibrolytic enzymes might cause some changes in the rate of potentially degradable NDF in the rumen [27] and the activity and number of ruminal non-fibrolytic and fibrolytic microbiota [12]. Fibrolytic enzymes increase the ruminal degradability rate of the potentially digestible NDF [27], changing the nutrient digestibility site, enhancing ruminal microbiota attachment and plant cells colonization and the synergy between exogenous enzymes and ruminal endogenous enzymes and microflora [12,28]. Ensiling of DPL with fibrolytic enzymes may cause hydrolysis of complex carbohydrates into simple sugars that may be utilized by ruminal microbiota for growth and stimulating more microbial growth, resulting in changes in the overall rumen microbial population and enzymatic activity in the rumen [11].

Ruminal Fermentation
Both MSP and fibrolytic enzymes have almost the same effects on ruminal fermentation. Treatments did not influence ruminal pH values that were greater than the optimum level (5.6) for ruminal fiber degrading microbial activities and growth [29]. Preventing a decline in ruminal pH is important to avoid a change in ruminal microbiota from predominantly fibrolytic to amylolytic microbial communities [30]. Fibrolytic enzymes and MSP treatments increased the concentrations of ruminalNH 3 -N, which may be a result of the increased CP digestibility. The observed NH 3 -N concentrations ranged between 30.1 to 32.9 mg/dL, which was greater than the level (8.5 to over 30 mg NH 3 -N/dL) for optimum rumen microbial proliferation and activity [31].
The increased concentrations of ruminal total VFA, with fibrolytic enzymes and MSP treatments, may be attributed to improved nutrient digestibility (e.g., organic matter and NSC digestibility). Increased total and individual VFA concentrations of the enzymesand MSP-treated diets likely resulted from the increased feed intake, N and fermentable carbohydrate availability, ruminal microbial activity and fermentation rate. Similar results were observed by Abid et al. [10]. Additionally, the treatments increased the concentrations of ruminal acetate with fibrolytic enzymes and MSP, which could be the result of improved apparent fiber degradation [5]. Fibrolytic enzymes and MSP treatments increased the ruminal propionate concentrations, which may result from improved apparent nutrient degradation, especially NSC digestibility by ruminal enzymes. Treatment of DPL with MSP generally improved ruminal fermentation in vitro when berseem hay was replaced with MSP-treated DPL at a 25% level [5].

Blood Chemistry Measurements
In the present experiment, all the measured serum biochemical variables were within the standard physiological ranges for healthy ewes, indicating good health with normal nutritional and physiological status of the ewes. Treatments did not influence the concentrations of serum globulin or urea-N, indicating minimal effects of treatments on ewes' nutritional status, muscle protein catabolism and unaltered kidney function [32]. Moreover, treatments did not affect the concentrations of serum ALT or AST, suggesting minimal effects of treatments on liver health [33]. Hamdon et al. [7] observed that feeding DPL to ewes increased serum ALT levels. The unaffected concentrations of serum HDL or LDL with treatments indicate the uninfluenced treated DPL on fat metabolism, liver dysfunction, and fat malabsorption [34]. Additionally, the unchanged or minor changes in concentrations of serum NEFA and BHBA indicate that body-fat breakdown was not changed and the ewes were not in a negative energy balance in the DPL treatments [35].
Both the MSP and enzyme treatments increased the concentrations of serum total protein and albumin, which are important indicators for improved nutritional and physiological status of the ewes due to increased nutrient intake and digestibility. Increased serum total protein and albumin can be related to higher feed intake and nutrient supply in the ewes fed treated DPL. Additionally, MSP and enzyme treatments increased the concentrations of serum glucose, which may be associated with the observed enhanced apparent organic matter and NSC digestibilities. Serum glucose concentration has a strong relationship with the concentration of ruminal propionate that increased in the present study due to enzymes and MSP treatment of DPL because blood glucose is synthesized from ruminal propionate in the liver [36]. This result, additionally, corroborates with the unchanged NEFA and BHBA concentrations in serum that ewes in DPL treatments were not in deficient energy balance despite greater milk production in these groups. Although ruminal NH 3 -Nconcentration elevated with the additive use, this was not reflected in blood urea-N concentration. Blood urea-N in ruminants is a function of several factors, including absorption of ruminal NH 3 -N, efficiency of utilization of absorbed amino acids, catabolism of protein, and transfer of blood urea to milk and its excretion rate.
Increasing the antioxidant capacity in ewes fed diets treated with MSP and enzymes is paralleled with the results of Sharifi et al. [21], who observed that feeding low-quality date palm to lactating goats improved total antioxidant capacity in milk and blood. Ensiling of phenolic-rich leaves increased the concentrations of phenolic acids and flavonoid compounds that have antioxidant properties. The antioxidants and phenolic compounds in DPL, which may be more available due to treatments with MSP and enzymes than the untreated DPL, may increase antioxidant status in blood.

Milk Yield, Composition and Fatty Acids
Fibrolytic enzymes and MSP treatments increased the daily production of milk (9, 10.7, 8.1 and 9.5%), ECM (15.3, 17.5, 14.7 and 16.6%), and FCM (15.6, 18.1, 14.2 and 15.6%) for ENZ50, ENZ100, MSP50 and MSP100 treatment, respectively. Increasing milk production in comparison with the feed intake reflected enhanced feed (milk) efficiency. Many experiments [8,13] observed a positive relationship between MSP supplementation and enzymatic treatments and milk production. The use of probiotics has been observed to improve microbial ecology, feed conversion ratio, and nutrient intake, resulting in better performance [9]. The supplementation with MSP causes some changes in ruminal bacterial community composition, including bacteria in the family of Lactobacillales [9], which plays a vital role in stabilizing ruminal pH [24]. The cumulative effect of greater nutrient intakes and digestibility and improved ruminal fermentation (i.e., propionate concentration) may be considered the main reasons for greater daily milk. Additionally, higher ruminal propionate concentration, which is a precursor for glucose and lactose synthesis, has favorable effects on milk yield as propionate appears to augment energy availability [37]. As earlier mentioned, greater blood glucose suggests a good energy status of animals, and can be another reason for increases in milk production in the DPL-fed ewes [37].
Multi-species probiotics treatments showed higher concentration of milk lactose compared to enzyme treatment, but the reason is unclear in this study. It is well-known that lactose concentration depends on nutrient digestibility (especially OM and NSC) and ruminal propionate concentration, and all of them followed the same trends in both fibrolytic enzymes-and MSP-treated diets. Nutritional factors contribute about 50% of the variations in milk composition and yields [38]. Fibrolytic enzymes and MSP treatments increased the concentrations of milk fats and milk energy, which may be attributed to the increased ruminal acetate concentration because of enhanced fiber digestion with DPL treatments. Ruminal acetate is the major precursor for mammary gland fatty-acid synthesis [38].
Treatments had minimal effects on individual fatty acids; however, fibrolytic enzymes and MSP treatments increased the concentrations of C18:1n9 trans, C18:2 trans-10, cis-12, C18:3n3, C18.3n6, C20:5n3, PUFA, total CLA and UFA:SFA ratio, but decreased the atherogenicity index. More than half of milk fatty acids arise from plasma uptake and the rest are synthesized in mammary glands [39]. As previously noted, enzymes and MSP treatments improved fiber digestion, which might be associated with altered milk fatty-acid profiles as a result of changes in the acetate-to-propionate ratio in the rumen.
The increased PUFA concentration and UFA:SFA ratio suggest that DPL treatments affected the ruminal bacterial activities responsible for biohydrogenation of dietary PUFA [38]. Minor improvements in a few n-3 FA contents in milk were also noted due to probiotic or enzyme treatments. It is well-documented that PUFA concentrations in milk depend mainly on the amount absorbed in the small intestine [38] as a result of escaping of ruminal biohydrogenation, which makes them available for incorporation into milk fat. Milk CLA are produced in the rumen when linoleic acid is partially biohydrogenated by ruminal bacteria. In addition, CLA is synthesized in the mammary glands by desaturation of rumenic acid (a partial hydrogenation product of linoleic acid) contributing about 70% of total milk CLA [40]. The extracts of DPL were found to contain many phytochemicals, including polyphenols, flavonoids, tannins, saponins, and quinines with antimicrobial activities [41]. Ensiling with MSP or enzyme may release more phenolic and saponin compounds from DPL, and these compounds can reduce ruminal microbial biohydrogenation of UFA that may be absorbed from intestine to blood and subsequently to milk increasing PUFA and CLA contents in meat and milk. Greater n-3 FA concentrations in milk have also been reported due to lactobacillus probiotic feeding to goats [42]. Increased proportion of PUFA and CLA in milk caused by treated DPL would be beneficial for human health.

Conclusions
Ensiling of date-palm leaves with fibrolytic enzymes or MSP before feeding to lactating Farafra ewes increased feed intake, improved nutrient digestibility, positively altered ruminal fermentation, and improved lactational performance and milk nutritive value (milk fatty acid profile) compared to the date-palm leaves ensiled without additives. Fibrolytic enzymes-treated DPL at 20% of diet improved fiber digestibility, while DPL at 40% increased milk production and feed efficiency compared to other treatments. Date-palm leaves ensiled with MSP or fibrolytic enzymes with minor differences between treatments may be used to improve milk production performance and milk quality in ewes under arid conditions. Fibrolytic enzyme treatment is recommended over MSP treatment. Data Availability Statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conflicts of Interest:
The authors declare no conflict of interest.