Effect of formaldehyde-treated urea on rumen fermentation , ration digestibility and nitrogen utilization

The study comprises two experiments in which young Finn-sheep were used as test animals. The experimental rations consisted of equal pans of NaOH-treated wheat (Exp. 1) or barley (Exp. 2) straw and a concentrate mixture of barley-molassed beet pulp (Exp. 1)or barley-oats-molassed beet pulp (Exp. 2 ). Feeding was performed twice a day. In addition 20 grams of urea/animal/day was mixed into the concentrates just before feeding. The urea was treated with the following percentages of formaldehyde, on a weight basis: 0 (F 0), 1.0 (F 10), 3.0 (F 3,0) and 5.0 (F5O) in Exp. 1 and 0, 1.0 and 1.5 (F,.5 ) in Exp. 2. The digestibility of the total ration decreased, when F 3 0 and F 5 0 urea was used, but the decrease was significant (P< 0.05) only when the apparent digestibility of crude protein was compared between theF 0 and F 3 ,o diets. The amount of rumen bacteria was decreased (P< 0.05) and the amount of protozoa increased (P< 0.01) by formaldehyde treatment levels above F lO and F3O , respectively. The concentration of the totalVFA in the rumen tended to decrease with treatment levels higher than F3O . No significant differences were found in the composition of the VFA. When treated urea was used, the excretion of nitrogen in the faeces increased but its excretion in the urine decreased. The percentage retention of the nitrogen ingested by the animals on diets F 0) F, 0, F 3 and F5 ,o in Exp. 1 was 15.0, 10.8, 13.2 and 12.2 and on diets F O,F 10 and Fl 5 in Exp. 2it was 20.5, 20.2 and 21.2, respectively.


Introduction
In earlier papers of SETÄLÄ and SYRJÄLÄ-QVIST(I9B2 a,b) it was reported that degradation of urea to ammonia in vitro could be decreased by treatment with formaldehyde.We also suggested that it would be possible to find the optimum treatment level for microbial protein synthesis.The aim of these experiments was to study the effects of formaldehyde-treated urea on rumen fermentation, ration digestibility and nitrogen utilization in vivo.
Maataloustieteellinen Aikakauskirja JOURNAL OF THE SCIENTIFIC AGRICULTURAL SOCIETY OF FINLAND

Materials and methods
The rumen physiological and digestibility experiments were made in two trials arranged according to 4 X 4 (Experiment 1) and 3x3 (Experiment 2) Latin square designs with young Finn-sheep.In each treatment the transfer period lasted nine days and the collection period five days.In both trials the experimental diets were based on equal parts of NaOH-treated wheat (Exp. 1) or barley (Exp. 2)straw and a concentrate mixture (Table 1).In addition, urea was mixed into the concentrate just before feeding.The urea tested in experiment 1 was untreated urea (F 0), and urea treated with 1.0 (F 1 0), 3.0 (F 3.0) and 5.0 (F 5 0 ) % formaldehyde on a weight basis.In experiment 2 the urea was untreated urea, and urea treated with 1.0 and 1.5 (F l-5 )% formaldehyde.The treatment of the urea was described by SETÄLÄ and SYRJÄLÄ- QVIST (1982 a).
The animals also received water and a mineral mixture (24.0 % Ca, 6.0 % P, 3.0 % Mg and 14.5 % NaCl) ad libitum.The amounts consumed were calculated and recorded.
The feeds were sampled every day during the collection period and on the two first and last days of the transfer period.Thirty per cent of the total daily amount of the faeces was taken for analysis after thorough mixing.In Exp. 2 the urine was treated and sampled as described by SETÄLÄ et ai. (1980).In Exp. 1 urine was sampled 1, 3 and 4 hours after feeding and preserved without H 2 S0 4 .After this period the treatment and sampling were done as in Exp. 2.
Blood samples from the jugular vein (Vena jugularis) were taken in Exp. 1 on the first and fifth day of the collection period, five hours after feeding.This sampling time was chosen in acccordance with the results and sugges- tions of LEWIS (1957).
The chemical analyses of the feeds and faeces were made according to standard methods.Before analysis the samples were vacuum-dried (+6O°C) for two days and milled through a 1-mm screen.The dry matter content of the samples was determined by keeping them in +lO5°C for 24 hours.The nitrogen content of the urea and urine was determined by the Kjeldahl method.In Exp. 1 the urea content of the urine sampled 1-4 hours after feeding was determined colorimetrically (ANON. 1973), and the urea in the faeces, excluding ammonia, was determined by the method of AO AC (ANON 1970).In Exp. 2 the non-dietary faecal nitrogen (NDFN) was determined according to MASON and FREDERIKSEN (1979), by using neutral detergent solutions.In Exp. 1 the pH, ammonia and VFA in the rumen samples were determined as explained by SETÄLÄ et ai.(1980).The numbers of protozoa were calculated according to WESTERLING (1970) and those of bacteria in a Helbers counting chamber, with bottom area Ixl mm, depth 0.02 mm and volume 0.02 mm 3 (SYRJÄLÄ et ai. 1976).
The blood samples were treated for the analyses as described by NÄSI (1979).Plasma urea-N and total proteins were determined according to CHANEY and MARBACH (1962) and REINHOLD (1953), respectively.
The statistical treatment of the results was made according to LUCAS (1975) and the differences between treatments were analyzed by the Tukey test (STEELE and TORRIE 1960).

Rumen fermentation
The total concentration of VFA in the rumen tended to be lower, when F 5 o urea was used (Fig. 1).The differences between diets were not, however, significant.This kind of change was also noticed by KULASEK et al. (1975,1976), who used similar treatment levels for urea, and by SYRJÄLÄ et ai. (1978),when formaldehyde-treated skimmed-milk powder was studied.
Formaldehyde treatment tended to increase the proportion of acetid acid and decrease the proportion of isobutyric and butyric acids in the total VFA.Similar changes in acetic acid were found by FAICHNEY (1974) when using feed with a high content of rumen-degradable nitrogen treated with formal- dehyde.However, this was not observed by SYRJÄLÄ et ai.(1978) or KEMPTON and LENG (1979).
The proportions of isovaleric and valeric acids in the total VFA were relatively small.MÄKINEN (1972) showed that on urea-rich diets the amounts of branched-chain fatty acids in the rumen are low.In our experiment.VFA and the molar propor- tions of VFA according to mole-% in the rumen fluid of sheep before and I-4 hours after feeding.OF 0 , AF10.
however, the proportions of isovaleric and valeric acids varied considerably, although not significantly, within the period studied.
Formaldehyde treatment caused a significant decrease (P< 0.05) in the peak of ammonia formation in the rumen (Fig. 2).Although differences were found between the urea treatments in vitro (SETÄLÄ and SYRJÄLÄ-QVIST a), they were not found in vivo.One explanation for this could be the change in the relationship between bacteria and protozoa in the rumen (Table 2).The amount of bacteria decreased (P< 0.05) and the amount of protozoa increased (P< 0.01) with a treatment level higher than 1 % and 3 % formaldehyde, respectively.If the decrease in the amounts of bacteria is caused by degradation of bacterial cells digested by protozoa, which is indicated by the changes in the concentrations of isovaleric and valeric acids in the rumen, then this degradation releases more ammonia into the rumen (COLEMAN 1975).
It appears also that the contribution of recycled urea nitrogen to the rumen ammonia levels was small (THORNTON 1970, KENNEDY andMILLI- GAN 1978), and the main pathway was recycling in saliva (NOLAN and LENG 1972).Besides damage of bacteria cells by protozoa, decreased utilization of ammonia due to decreased fermentation and lack of fermentable energy for microbial protein synthesis (HAGEMEISTER et al. 1980) was another reason why the ammonia concentrations in the rumen did not differ between treatment levels in Exp. 1.
The changes in the amounts of different protozoa species were not very clear, although some significant differences were found.In treatments corres-Table 2. Composition of rumen microbiota on different urea-containing diets (Experiment 1).F(j-F 5 ,0, see  ponding to the present levels of 3 to 5 % formaldehyde, the proportion of bacteria has generally increased and the proportion of protozoa decreased (HEMPEL-ZAWITKOWSKA and KULASEK 1974, KULASEK et al. 1976, SYRJÄLÄ et al. 1978), but compared with those results, the amounts of protozoa were rather low.
On the other hand, bacteria are removed from the rumen mainly in the liquid phase (BERGEN and YOKOYAMA 1977), while less than 30 %of protozoa use this phase (WELLER and PILGRIM 1974).Since the water consumption increased with the formaldehyde treatment level (r = +0.93xx ), this may have had an effect on the liquid flow from the reticulo-rumen (OYAERT and BOUCKAERT 1961, ROGERS et al. 1979) and hence on the proportions of bacteria and protozoa.

Digestibility of the ration
Formaldehyde treatment decreased the digestibility of the total ration most clearly at treatment levels higher than 1.5 % (Table 3).When the apparent digestibility of crude protein is considered, significant (P< 0.05) differences are found only between the F 0 and F 5O diets.When the true digestibility of crude protein is calculated according to MASON (1979) using the values of non-dietary faecal nitrogen (Table 4), the results with F O , F t-0 and F l 5 feeding in Exp. 2 were 89.7, 88.0 and 89.1 % respectively.These values are similar to the results given by KAUFMANN (1977) and MASON and FREDERIKSEN (1979).
The main reason for the decreased digest i bility at the highest treatment Table 3.The digestibility %of the total ration in different urea-containing diets.F 0 -F 5 .o,see Table 1   Table 4. Excretion of nitrogen in the faeces and urine, plasma urea-N and total protein, and nitrogen balance of the animals on different urea-containing diets.
F 0 -F5 0, see   levels may be the change in the rumen microbiota.The ability of protozoa to digest crude fibre is poor compared with that of bacteria (HUNGATE 1966, HARMEYER 1973), although at least some species can digest and utilize hemicellulose and cellulose (HAYER et al. 1976, COLEMAN 1978).

Utilization of nitrogen
There were no significant differences in nitrogen retention of the animals on different diets (Table 4).
According to ANDERSON et al. (1959), increased excretion of nitrogen in faeces is caused by overprotected and hence undigested urea nitrogen.In Exp. 1 the method used for urea determination revealed no significant differences between treatments in the urea content of the faeces.It is possible that part of the treated urea not degraded in the rumen was degraded to ammonia in the caecum and was used for synthesis of microbial protein (TELLER et al. 1979).This could be an important factor for the apparent digestibility of crude protein.In Exp. 2 however, no significant differences were found in the excretion of non-dietary faecal nitrogen below the treatment level of 1.5 % formaldehyde.
Since the digestibility of organic matter and the fermentation in the rumen were higher on the F l 0 diet than on the other diets containing treated urea in Exp. 1, less fermentable energy was available in the caecum on this diet.This decreases utilization of ammonia in the caecum and, together with possible degradation of F] 0 urea in the digestive tract before the caecum, affects the blood urea-N levels (CHALMERS et al. 1976).

Figure 1 .
Figure 1.Rumen pH, total P < 0.05 a -c, means with different letters differ significantly.P < 0.01 d -f, means with different letters differ significantly.

Table
1 . Chemical composition of the feeds. -