Difference of nitrogen contents determined by the combustion and Kjeldahl method in response to nitrate nitrogen in some feedstuffs

Two experiments were conducted to examine the difference of nitrogen contents determined by the combustion (Cm) and Kjeldahl (Km) method and its response to nitrate nitrogen in ruminant feedstuffs. In Experiment 1, 14 ruminant feedstuffs were determined for combustion nitrogen (Cn), Kjeldahl nitrogen (Kn) and nitrate nitrogen (NO3-N). Although NO3-N resulted in a difference between the Cm and Km, the correlation coeffi cient of the NO3-N content and the value of “Cn-Kn” was low (R 2 = 0.6341), suggesting other factors infl uencing the difference of N determination between the Cm and Km. In Experiment 2, recoveries of net NO3-N were determined using Chinese wild rye-grass (CWG), maize grain (MG) and soyabean meal (SBM) supplemented with sodium nitrate at the level of 5, 10, 15, 20, 25, 30, 35, 40 and 45% on DM basis, respectively. The results showed that different recoveries of NO3-N by Km rather than Cm would account for the difference of Cn and Kn in some ruminant feedstuffs.


INTRODUCTION
Many researchers have compared N contents of feedstuffs determined by the combustion (Cm) and Kjeldahl (Km) methods (Koenig, 1991;Jakob et al., 1995).The results revealed a high linear correlation (0.992~0.999) of N contents for the majority of feedstuffs between the two methods.However, when the feedstuff is rich of NO 3 -N, the higher N content was obtained from Cm than Km (Watson and Galliher, 2001).Simonne et al. (1997) stated that Cm and Km may recover different forms of N in plant tissues respectively, e.g., nucleic acids N and NO 3 -N.Afterward, Simonne et al. (1998) further demonstrated that under a wide range of NO 3 -N in 130 leaf samples, NO 3 -N alone did not account for the difference between Cm and Km.
When we routinely determined the N content of ruminant feedstuffs including forages and vegetables, Cn contents are always higher than Kn, regardless of their NO 3 contents.This observation suggests that there may exist other factors responsible for the difference of N contents determined between Cm and Km.Therefore, the present study was conducted to: 1. prove if the NO 3 -N alone can account for the difference between Cm and Km in ruminant feedstuffs, and 2. explore the factors infl uencing the difference of N determination between the two methods, when incremental amounts of NO 3 -N were included in the feedstuffs.

MATERIAL AND METHODS
Two experiments were conducted in this study.In Experiment 1, 14 feedstuff samples were selected for determination of N contents: rye-grass, Chinese wild rye-grass (CWG), perfoliate rosin-weed, awnless bromegrass, lucerne, India lettuce, crested wheat-grass, dahuria lyme-grass, sorghum hybrid Sudan-grass, maize stalks, wheat straw, rice straw, grass and cabbage.All samples were dried in a forced-air oven at 70 o C for 72 h and ground in a vortex mill (0.5 mm sieve, Perten Laboratory Mill 3100).The Kjeldahl-N (Kn), combustion-N (Cn) and nitrogen nitrate (NO 3 -N) content were then determined for each sample.In Experiment 2, ten incremental levels of sodium nitrate (S5506, Sigma-Altrich) were mixed with Chinese wild rye-grass (CWG), maize grains (MG) or soyabean meal (SBM) to form the sodium nitrate concentration of 5,10,15,20,25,30,35,40 and 45 g/100 g dry matter.Nitrogen contents of the samples were determined by the procedure of AOAC (2000) for Kn with the N Analyzer (Foss Model 2300, Sweden) and by the procedure of AOAC (990.03) for Cn with the N Analyzer (Rapid N III, Elementar, Germany).All results were expressed in g N/100 g dry matter (%).Recovery of NO 3 -N in three serial mixed samples in Experiment 2 was calculated based on the assumption that all ammonia-N was totally captured into Kn.
In order to determine the NO 3 -N, 1 g sample was weighed into a beaker with 50 ml deionized water, then mixed for 1 h and fi ltered.NO 3 -N in solution was then measured by the procedure of Jones and Case (1990) using a spectrophotometer (UV-VIS 8500, Shanghai Tianmei Scientifi c Instrument Co., Ltd., China).
The N content of feedstuffs was analysed as a single factor design using the GLM procedure of SAS (2003).Simple linear correlation analysis and signifi cance test of the coeffi cient were performed on N according to the procedure described by Steel and Torrie (1960).Analysis of correlation between NO 3 -N and Cn -Kn was evaluated using regression analysis procedure of SAS (2003).The recovery of NO 3 -N in samples was evaluated with an analysis of variance using the GLM procedure of SAS (2003).

RESULTS AND DISCUSSION
Nitrogen (N) contents of 14 nitrate-contained feedstuffs were determined by both Cm and Km (Table 1; Experiment 1).Higher N content was obtained for Cm compared with Km for all feedstuffs (P<0.05;CV<5%; Cn: Kn=1.04~1.21).These results were in agreement with the observations of Simonne et al. (1997), who found that sulphuric acid could convert all protein N and only part of NO 3 -N into ammonium N. In contrast, when the sample contained substantial NO 3 -N, Cm provided a higher N measurement (Watson and Galliber, 2001).In the present study, although a linear correlation (r=0.9960) of N contents of 14 feedstuffs between the two determination methods as shown in Figure 1, the slope of the regression equation was signifi cantly (P<0.01)different from that of Y=X, suggesting other forms of N (e.g., nitrate N) not only ammonium N included in the feedstuffs.In order to prove the nitrate N responsible for the difference of N measurements between the two methods, we made a correlation between nitrate nitrogen (NO 3 -N) content and Cn-Kn for 14 feedstuffs (Figure 2).No signifi cantly linear correlation was obtained between the NO 3 -N contents and Cn-Kn (R 2 =0.6341; Figure 2), suggesting that NO 3 -N alone does not account for the difference between the two methods.This observation coincides with the result of Simonne et al. (1998), who found not only nitrate N but other factors, such as the nucleic acid, the matrix effect or a combination of both, may explain these differences in vegetable leaves.Further studies are required to prove the validity of this hypothesis.As shown in Tables 2 and 3 (Experiment 2), the recovery of net NO 3 -N of 3 nitrate contained feedstuffs determined by Cm ranged from 89.2 to 100.5%, whereas the range of Km was from 34.2 to 54.9%.The Cm can well capture 48.0 a 43.9 b 0.33 0.000 1 means within the same line with different superscript letters differ (P<0.05) 2 no differences were obtained between NO 3 -N addition levels (P>0.05) 3 CWG -Chinese rye-grass; MG -maize grain; SBM -soyabean meal nitrate N from the feedstuffs, with an acceptable N recovery range from 89.2 to 100.5%.However, Km not only partially captures the nitrate N refl ected by the lower nitrate recoveries, but also the captured amount of nitrate N was varied with the feedstuff source because of signifi cantly different (P<0.05)recoveries obtained among the different feedstuffs (Table 2).Simonne et al. (1998) stated that besides nitrate N, nucleic acid N may also be accountable for the difference between Cn and Kn of vegetable leaves.However, when we determine the N content of yeast products rich of nucleic acids using Cm and Km, there was no difference obtained.The different recoveries of net NO 3 -N for Km was likely attributed to certain intrinsic substances (e.g., high concentration of lipids, lipoproteins, etc.) within the biological samples that resulted in their incomplete mineralization for mixtures of sulphuric acid.Based on these observations, different captures of nitrate N in Km rather than Cm would be responsible for the difference of Cn and Kn in some ruminant feedstuffs.

CONCLUSIONS
The combustion method is superior to the Kjeldahl method when the feedstuff contains signifi cant quantities of nitrate nitrogen.Nitrate nitrogen alone does not account for the difference between the two methods.Low and different nitrate captures of feedstuff sources determined by Kjeldahl method rather than combustion method would be accountable for the difference between combustion and Kjeldahl nitrogen in some ruminant feedstuffs.

Figure 1 .
Figure 1.Relation of N contents of 14 feedstuffs as determined by the Cm and Km (Experiment 1) Figure 2. Relation between NO 3 -N contents and Cn -Kn in 14 feedstuffs (Experiment 1)

Table 2 .
The recovery of nitrate nitrogen in 3 feedstuffs determined by Kjeldahl method 1 (Experiment 2)

Table 3 .
The recovery of nitrate nitrogen in 3 feedstuffs determined by combustion method 1 (Experiment 2)