|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL NUTRITION |
Department of Animal and Veterinary Science, University of Idaho, Moscow 83844-2330
| Abstract |
|---|
|
|
|---|
Key Words: dairy cow duodenal digesta microbial protein synthesis reticular digesta
| INTRODUCTION |
|---|
|
|
|---|
Microbial protein flow estimated from duodenal sampling is also quite variable (Titgemeyer, 1997
). Therefore, researchers have sought alternative approaches to estimate MPS. Hristov and Broderick (1996)
proposed a ruminal sampling method and Huhtanen et al. (1997)
proposed an omasal sampling technique to estimate digesta flow and MPS. The latter method was refined and repeatedly used in consequent experiments (Ahvenjärvi et al., 2000
; Reynal et al., 2003
). The reticulum, as the organ propelling digesta through the reticulo-omasal orifice, regulates the flow of nutrients to the lower digestive tract (Sissons et al., 1984
; Sutherland, 1988
). The composition of reticular digesta is rather consistent (Dardillat and Baumont, 1992
), and particles found in the reticulum are likely to leave the reticulorumen (McBride et al., 1984
). Thus, digesta located in the reticulum appears to closely represent digesta leaving the reticulorumen.
The objective of this experiment was to investigate the possibility of estimating the outflow of nutrients and microbial protein from the rumen based on sampling reticular contents as an alternative to the conventional duodenal sampling technique.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Markers, Sampling, and Related Analyses
Ruminal ammonia and, consequently, microbial N were labeled through a continuous, 6-d intraruminal infusion of 10 atom percent excess of (15NH4)2SO4 (Cambridge Isotope Laboratories Inc., Andover, MA) dissolved in McDougalls buffer (McDougall, 1948
). The infusion rate was 0.25 g of 15N/d. The amount of solution infused was 2,016 mL/d. The infusion began at 0600 on d 15 of each experimental period (4 d before the first reticular and duodenal sampling) and continued through d 21 of each period. Flow of nutrients was estimated from duodenal or reticular samples by using the triple-marker method of France and Siddons (1986)
, with indigestible NDF (iNDF), Co, and Yb as markers (Ahvenjärvi et al., 2000
). Flow of OM was not corrected for VFA flow (Ahvenjärvi et al., 2002
). Lithium-Co-EDTA (Udén et al., 1980
) and Yb acetate (Rhodia Inc., Shelton, CT) were infused continuously from d 15 through 21 of each period at rates of 15 and 4 g/d, respectively. Microbial protein outflow from the rumen was also estimated by using the ruminal sampling technique (Hristov and Broderick, 1996
).
Duodenal Sampling.
Eight duodenal samples (300 mL per sampling) were collected from d 19 through 21 of each experimental period. Sample collection began after discarding digesta accumulated in the cannula neck. The sampling schedule was as in Hristov and Ropp (2003)
. Individual samples were immediately frozen at –80°C. After thawing, digesta samples were composited on a weight basis (per cow and period). The composited samples were separated into fluid and large and small particulate phases (F, LP, and SP, respectively). The large particle phase (LPDuo) was separated by filtering the whole digesta through a 100-µm fabric (Sefar America Inc., Depew, NY). The filtrate was centrifuged at 10,000 x g for 30 min to separate the fluid (supernatant, FDuo) and the small particle (pellets, SPDuo) phases (Ahvenjärvi et al., 2000
).
Samples from all phases were freeze-dried, ground through a 1-mm sieve, and analyzed for ash-OM (AOAC, 1999
; method 942.05), Co and Yb (Iris ICP atomic emission spectrophotometer, Thermo Jarrell Ash Corp., Franklin, MA), iNDF (LPDuo and SPDuo only; Rinne et al., 2002
), N and 15N enrichment of non-ammonia N (NAN; Ahvenjärvi et al., 2002
), purines (Hristov et al., 2005
), NDF (Ankom 200 Fiber Analyzer, Ankom Technology), and starch (starch analysis kit, Megazyme International Ireland Ltd., Wicklow, Ireland; McCleary et al., 1994
). A heat-stable amylase (
-amylase, EC 232.560.9, Sigma Chemical Co.) was used for the NDF analysis; sodium sulfite was not used for the analysis (Van Soest et al., 1991
). Nitrogen and 15N enrichment of samples were analyzed on a Costech ECS 4010 C/N/S elemental analyzer (Costech Analytical Technol. Inc., Valencia, CA) interfaced to a Deltaplus isotope ratio mass spectrometer (Thermo Finnigan MAT GmbH, Bremen, Germany). Samples were pulverized (Retsch MM200 micro mill; F. Kurt Retsch GmbH & Co. K. G., Haan, Germany) prior to the analysis.
Reticular Sampling.
Samples from the reticulum (400 mL per sample) were collected at the same time as the duodenal samples. A wide-mouth, 250-mL Nalgene bottle (Nalge Nunc Int., Rochester, NY) was suspended twice at approximately 5 to 10 cm from the bottom of the reticulum with the lid on. While on the bottom of the reticulum, the bottle was opened and allowed to fill with digesta. A 100-mL aliquot was used to isolate ruminal bacteria. Bacterial pellets were isolated as described by Hristov and Ropp (2003)
. This procedure isolates bacteria from fluid- and loosely associated particulate matter and does not account for the different composition of bacteria firmly associated with digesta particles. The remainder of the sample was immediately frozen at –80°C. After thawing, the reticular samples were composited on a weight basis and fluid and large and small particulate phases (FRet, LPRet, and SPRet, respectively) were separated as for the duodenal samples.
The proportion of microbial N in duodenal and reticular digesta phases was estimated based on 15N enrichment of the bacterial standard isolated from the reticular samples: 15N enrichment of NAN in the duodenal or reticular phase (F, LP, SP) divided by 15N enrichment of bacteria recovered from the composite reticular sample.
Ruminal Sampling.
Outflow of microbial NAN from the rumen was also estimated by using the rumen sampling approach (Hristov and Broderick, 1996
; Hristov et al., 2000
). Samples of whole ruminal contents were segregated into fluid (FRum) and particulate (PRum) phases by filtering through a 100-µm fabric. Proportions of microbial NAN in FRum and PRum were estimated based on 15N enrichments of FRum, PRum, and reticular bacterial NAN during the continuous infusion of (15NH4)2SO4, as described in the previous paragraph. Ruminal pools of FRum and PRum DM, OM, NAN, and NDF (PRum only) were found through evacuation of the ruminal contents on d 21 of each period 6 h after the morning feeding and by respective analyses of aliquot samples. Fractional outflow rates of FRum and PRum were calculated as ln-transformed Co or Yb concentrations plotted vs. time after the cessation of Li-Co-EDTA and Yb acetate infusions. Whole ruminal content samples were collected at 0, 2, 4, 6, 8, 10, 14, 18, 24, and 30 h after marker infusion ceased (at 0600 on d 21 of each period). Ruminal samples were collected from 4 locations in the rumen (approximately 250 g each) and segregated into FRum and PRum by filtering through a 100-µm fabric. The 2 phases were analyzed for Co or Yb, respectively (Iris ICP atomic emission spectrophotometer).
Statistical Analysis
All data were analyzed as a Latin square repeated measures by using PROC MIXED (SAS Inst. Inc., Cary, NC). The model used was:
![]() |
where µ was the overall mean; C, P, T, and S were cow, period, treatment, and sampling site (analogous to method of calculating nutrient flows), respectively; and e was an error term under the usual assumptions for ANOVA. Sampling site represented repeated measures on each animal. Cow, period, and cow x period x treatment were included as random effects. All other effects were assumed to be fixed. Reported values are least squares means. Statistical difference was declared at P
0.05. When the main effect was P
0.10, means were separated by pairwise t-tests.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
|
Nonammonia N concentration was greater (P < 0.001) in LPRet than in LPDuo, but the difference was much larger in the SP phase (P < 0.001). Concentration of NAN was 49% greater (P < 0.001) in FDuo than in FRet. The substantially greater NAN concentration in FDuo is probably a reflection of the large amount of enzymes secreted in the abomasum and perhaps lysis of microbial cells leaving the rumen. In agreement with this hypothesis, FRet NAN was entirely of microbial origin, whereas 40% of NAN in FDuo was nonmicrobial N. Similar to the 15N marker, purine concentration in F was greater (P = 0.03) for FRet than for FDuo. The largest difference in NAN content between the 2 sampling sites was observed for the SP phase. It is unlikely that this difference was due to selective retention or passage of particles and microbial mass through the reticulo-omasal orifice. The SP phase was isolated through filtration and consequent centrifugation at 10,000 x g for 30 min. Most of the bacterial markers from ruminal contents (purines and 15N) can be recovered in the pellets at a centrifugal force of 15,000 x g (Hristov and Zaman, 2006
). Thus, the N in SPRet (and perhaps SPDuo) would represent small particulate matter colonized with microbial cells, fluid bacteria, and protozoa. Both sites (SPRet and SPDuo) had similar concentrations of microbial markers (purines, P = 0.87; 15N, P = 0.14) and estimated proportions of microbial N in NAN (15N-based; P = 0.11). Endoscopic observations by McBride et al. (1984)
indicated that the reticulo-omasal orifice does not prevent large particles (10 mm) from entering the omasum. This individual observation, however, was not supported by the greater proportion of large particles in reticular compared with duodenal digesta observed in the current study and by Ahvenjärvi et al. (2001)
. Harmeyer and Michalowski (1991)
concluded that the reticulo-omasal orifice does not have a significant discriminating function for the selective passage of particles and protozoa into the omasum. These authors reported similar pH, protozoal concentrations, and particle size of samples collected from the reticulum and the effluent leaving the reticulorumen. Based on these data and results from the present experiment, selective retention of protein-rich protozoal mass in the reticulum, for example, as a cause for the observed large difference in NAN concentration between SPRet and SPDuo, may be ruled out. The likely explanation for this difference is the greater concentration of ammonia N in SPDuo compared with SPRet. Our analyses showed that ammonia N represented on average 45% of the total N in SPDuo and was negligible in SPRet (3.4% of the total N).
Starch concentration was similar between the 2 sampling sites (P = 0.89 for LP; P = 0.79 for F), except for the greater (P = 0.04) starch content observed with SPRet compared with SPDuo. Ytterbium concentration was greater in LP (P = 0.003) and SP (P < 0.001) recovered from the reticulum than in the respective particles recovered from the duodenum. An opposite effect was observed for F; Yb concentration was greater (P = 0.002) in FDuo than in FRet. Most likely, these differences represent marker migration from the particulate to the fluid phase at the low pH in the abomasum and the proximal duodenum (Combs et al., 1992
). In the present experiment, Co concentration in the 3 phases was not affected (P = 0.34 to 0.89) by sampling site.
As expected from the composition of the individual phases, whole duodenal digesta had greater (P < 0.05) OM and lower (P < 0.05) NAN and NDF concentrations than reconstituted reticular digesta (Table 2
). Whole ruminal contents had the greatest (P < 0.05) concentrations of OM and NDF but the lowest (P < 0.05) concentration of NAN compared with the duodenal and reticular samples. Duodenal and reticular digesta had similar (P = 0.90) concentrations of starch; ruminal samples were not analyzed for starch. Concentration of microbial NAN was greater (P < 0.05) in reticular than in duodenal digesta or ruminal contents. Cobalt and Yb concentrations were not different (P > 0.56) between the duodenal and reticular digesta phases during marker infusion, but reticular digesta had a greater (P = 0.004) iNDF concentration.
|
Estimated true ruminal digestibility of DM and OM was greater (by 12%, P = 0.05 and by 13%, P = 0.03, respectively) and that of NDF was lower by 17% (P = 0.004) for the reticular compared with the duodenal sampling. Ruminal digestibility of NAN and starch were similar (P = 0.38) between sampling sites.
The duodenal, reticular, and ruminal methods gave similar estimates for microbial NAN outflow (MPS) from the reticulorumen. The greater microbial NAN concentration in reticular compared with duodenal digesta was counteracted by the greater estimated DM flow with the duodenal method, which resulted in similar MPS. Because the proportions of microbial NAN in LP and SP were similar between the 2 sampling sites and the phases represented similar proportions in the sample DM (71 and 72%, duodenal and reticular samples, respectively), the difference in MPS was caused primarily by the difference in the proportions of microbial NAN in F (60 vs. 107%, duodenal and reticular samples, respectively). Using similar digesta markers to estimate flow (purines were used as a microbial marker), Ahvenjärvi et al. (2000)
reported similar microbial NAN flows estimated through omasal or duodenal sampling, irrespective of the sampling site for the bacterial standard.
Microbial NAN flow estimated through the ruminal method was only numerically greater (P = 0.32) than the duodenal and reticular estimates (by 14 and 7%, respectively). This numerical difference was caused mostly by the greater estimate for NAN flow by the ruminal, compared with the duodenal or the reticular, methods. Flow of nutrients other than NAN, however, was grossly overestimated by the ruminal method. Similar unrealistic flows can be calculated from Hristov and Broderick (1996)
. Apparently, this approach cannot be used to estimate outflow of nutrients accurately, other than NAN and microbial N from the rumen. Ruminal particulate matter is not a uniform entity and the existence of escapable and nonescapable pools has been suggested (Pond et al., 1988
; Huhtanen et al., 1995
; Lund et al., 2006
). Estimating outflow of nutrients, particularly fiber, based on the dilution rate of a single particulate marker (Yb in this case) ignores the different kinetic patterns of digesta particles varying in digestibility, size, and flow characteristics (specific gravity, for example; Hristov et al., 2003
). The ruminal approach uses the dilution rate and the pool sizes of the fluid and particulate phases to estimate flow of nutrients (Hristov and Broderick, 1996
). Thus, it assumes that all nutrients retained in the solids phase would obey the kinetics of the particulate marker used, ignoring the multicompartmental nature of solid digesta in the rumen. As a result, outflow of DM, and consequently of other nutrients, from the rumen is overestimated.
However, the ruminal method seems to provide reasonable estimates for microbial NAN flow (Hristov and Broderick, 1996
; Hristov et al., 2000
; present data); more recently, an identical (Hristov, 2003
) approach was proposed by Ondarza and Sniffen (2002)
. Microbial mass leaving the rumen with the particulate matter would most likely be associated with the small digesta particles, for which the kinetic pattern would more closely resemble that of Yb infused intraruminally. In addition, microorganisms are known to have a high affinity for rare-earth elements (Johnson and Kyker, 1961
), and high concentrations of Yb (and Ce) in bacterial pellets isolated from ruminal contents have been reported (Combs et al., 1992
; Hristov et al., 2003
). Thus, the true outflow rate of the microbial mass in the rumen may not differ significantly from the dilution rate of an extrinsic marker, such as Yb, infused intraruminally. In the case of fiber, passage and digestion rates and exchange between digestible and indigestible and escapable and nonescapable pools control the retention time of fiber in the rumen, and a single-pool model would not accurately predict ruminal turnover of fiber.
Compared with the duodenal estimates, the reticular method underestimated the flow of DM and OM, but overestimated NDF flow. However, as discussed earlier, the greater flow of NDF estimated from reticular samples is most likely a result of the digestion of fiber occurring in the omasum. Technically, the difference in outflow of DM between the 2 methods was mostly due to the greater concentration of iNDF in reticular digesta, which in the triple-marker system would result in lower estimated flow of DM compared with the duodenal estimates. Organic matter flow estimates were additionally influenced by the lower OM concentration in reticular compared with duodenal samples (particularly SP and F), which was probably a result of the endogenous OM secreted into the abomasum and the absorption of minerals in the omasum (Engelhardt and Hauffe, 1975
; Punia et al., 1988
). In addition, contamination of the reticular sample with inorganic material (sand, for example) cannot be ruled out. The reticulum does accumulate heavy inorganic matter from ingested feed, and care must be taken not to contaminate digesta during sampling. Ahvenjärvi et al. (2000)
also reported 7% lower OM flow when estimated based on omasal vs. duodenal sampling. To avoid interference from the variable ash content of digesta, it appears that nutrient flows should be calculated on an ash-free (OM) basis. Apparently, a similar problem was encountered by Ahvenjärvi et al. (2001)
, because these authors expressed digesta composition and nutrient flows on an OM basis. The numerically greater estimated flow of microbial NAN with the reticular digesta would have a much smaller impact on DM and OM flow estimates. The greater concentration of NDF in reticular particles, and consequently in whole reticular digesta, was the main reason for the greater estimated flow of NDF with the reticular compared with the duodenal method. Similar to our results, Ahvenjärvi et al. (2000)
reported greater flow of OM measured at the duodenum compared with estimates based on omasal sampling. As a result, apparent digestibility of OM was greater when measured at the omasum than at the duodenum. The authors explained this phenomenon with the large amount of OM secreted into the abomasum. In the study by Ahvenjärvi et al. (2000)
, NDF flow was lower (and NDF digestibility greater) when measured at the duodenum than at the omasum, which was explained by the role of the omasum in fiber digestion in ruminants. Results from the present experiment confirm these observations, because NDF digestibility was 17% greater when estimated based on duodenal, compared with reticular, flow of NDF. The greater iNDF:NDF ratio in LPDuo vs. LPRet (0.74 vs. 0.63, respectively; estimated from data in Table 1
) is also indicative of significant NDF digestion in the omasum. Similarly, Okine and Mathison (1991)
reported from 0.14 to 0.25 (depending on level of intake) of the total tract NDF digestion taking place between the reticulum and the duodenum.
This study supports the concept that microbial protein outflow from the rumen can be measured based on sampling of ruminal or reticular digesta. The reticular sampling technique can also provide reliable estimates for ruminal degradability of DM and its constituents. These findings need to be confirmed in experiments with basal diets varying in structure and forage-to-concentrate ratios.
| Footnotes |
|---|
2 Corresponding author: ahristov{at}uidaho.edu
Received for publication December 31, 2006. Accepted for publication June 13, 2007.
| LITERATURE CITED |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |