RT Dissertation/Thesis
T1 Comparison of plant cell wall degrading community in the rumen of N?Dama and N?Dama x Jersey crossbred cattle in relation to in vivo and in vitro cell wall degradation
T3 n.z.
A1 Nouala-Fonkou,Simplice
WP 2004/07/12
AB This thesis presents a unique combination of an in vivo feeding trial, the analysis of the microbial community structure in the rumen, and in vitro fermentation studies, in order to assess the impact of breeds and diets on animal performance in a West African production setting. Pure N?Dama and N?Dama x Jersey crossbred cattle were fed two basal diets, baby corn and groundnut hay, supplemented with graded levels of either conventional concentrate or moringa leaf meal, to compare animal responses in productivity. In this context, Moringa oleifera leaf meal constitutes a locally available, potential alternative to commercial concentrate for cattle production. The cell wall digesting community of N?Dama and its crosses was analysed using phylogenetically based hybridisation probes to account for the contribution of rumen microbes to differences in fermentation patterns and animal response. In vitro fermentation studies were carried out using the same diets and supplementation levels as fed in vivo, to test the accuracy of the in vitro gas production technique in predicting the optimum level of supplementation. 
The in vivo feeding experiment focussed on the comparison of breed performance with diets relevant for local production conditions. Six N?Dama and six N?Dama x Jersey (crossbred) animals were used in a cross over design. They were fed consecutively three combinations of roughage and supplement, baby corn stover and concentrate (BCS:Co), groundnut hay and concentrate (GNH:Co) and groundnut hay and moringa meal (GNH:Mo), each at 5 levels of supplementation (0, 10, 20, 30 and 40%). Results from this study showed that there was a clear difference in animal response to different feeding regimes between the two breeds. When averaged over all diets organic matter intake (OMI) was higher in crossbred compared to N?Dama (94 and 87.6 g/kg 0.75 d-1, respectively). When analyzed for the diets and averaged over the breeds OMI was higher when animals were fed the baby corn based diet compared to groundnut based diets (95 against 88 g/kg 0.75 d-1). Only when the diet consisted of BCS:Co, and at low levels of supplementation, N?Dama ingested more than crossbred, but the difference was not significant. With GNH:Co crossbred ingested significantly more at levels of supplementation less or equal to 20%. With GNH:Mo crossbred ingested more, whatever the level of supplementation. The optimum level of supplementation in vivo, estimated with the single slope broken line model, was 10% and 20% for both breeds when they were fed BCS:Co and GNH:Co respectively, but 30% for N?Dama and 10% for the crossbreds when animals were fed GNH:Mo.
Organic matter digestibility (OMD) was higher in N?Dama (64.6% against 60.7% in crossbreds) when animals were fed BCS:Co and supplementation had no effect on OMD of BCS:Co whatever the breed. When animals were fed groundnut hay as basal diet, OMD was also significantly higher in N?Dama at low levels of supplementation, but the differences became insignificant beyond 10% and 20% of concentrate or moringa, respectively. With GNH:Co OMD showed a quadratic response (p<0.001) with increasing level of supplementation when it was fed to crossbreds and was not affected when it was fed to N?Dama. Increasing levels of moringa meal supplementation increased OMD in both breeds up to a peak at 20 and 30% for N?Dama and crossbred, respectively. 
Average daily weight gain (ADG) was not affected by the breed, however it was higher on BCS:Co compared to other diets. On BCS:Co ADG increased with the level of supplementation, reaching a peak at 30%, whereas supplementation had no effect on ADG when animals were fed groundnut hay based diets. 
As N?Dama could take in and digest more of the low quality BCS:Co diet, they were less efficient in feed conversion under this feeding regime (FCE: 14 vs. 9 for the crossbreds). On GNH based diets, however, N?Dama surpassed the crossbreds in feed conversion efficiency with ratios of 11 vs. 13 for GNH:Co and 9 vs. 27 on GNH:Mo.
Rumen microbes play the key role for the digestibility of a given feed and thus also for feed intake and finally animal performance. Obviously, the community composition and activity is highly dependent on the diet. With the present set-up, however, with identical external conditions and three different, well defined diets fed to both, N?Dama and crossbred cattle, a comparison of the microbial community structure between breeds could be attempted. The in vivo and in vitro data taken in the other parts of the study allow a sensible interpretation of potential changes in microbial composition. Rumen fluid was collected from three fistulated N?Dama and three crossbred animals adapted to the experimental diets at medium supplementation level. The cell wall degrading community was analyzed using the phylogenetically based 16S rRNA hybridisation probes. The results showed that on BCS:Co diet the Fibrobacter and R. flavefaciens RNA concentrations were higher in rumen fluid of N?Dama compared to crossbred. These concentrations were also significantly affected by the diet, such that they were higher on baby corn stover compared to groundnut hay based diets. The results of the microbial community analysis suggested that the differences between breeds observed in digestibility could be partially explained by the composition of the cell wall degrading community. 
Parallel to the in vivo experiment, in vitro fermentation studies were undertaken to evaluate the predictability of the in vivo response to supplementation by the in vitro data. Rumen fluid from 3 N?Dama and 3 crossbred donor animals was used for 24 hour in vitro fermentations. The donor animals were fed consecutively the same three diets used in vivo (BCS:Co, GNH:Co and GNH:Mo) at 20% level of supplementation. Each of these inocula was incubated with in vitro substrates consisting of all the combinations tested in vivo (i.e. 3 diets, 5 levels of supplementation) plus supplement alone. This design should allow to analyze for both, the impact of donor breed as well as that of the donor diet and to conclude which factors may be varied while maintaining predictability. The breed of the donor animals did not significantly affect 24 hour gas production, but short chain fatty acid concentration was higher with rumen fluid from crossbreds when donors were fed BCS:Co and GNH:Mo. Moringa meal as supplement to donor animals changed the fermentation pattern of all the substrates, such that gas production and SCFA increased substantially in groundnut hay based substrates, whereas gas production of BCS:Co substrates decreased and SCFA did not substantially change. In vitro digestibility was higher with rumen fluid from N?Dama whatever the diet of donor animals and the substrate incubated. GNH:Mo as donor diet also increased IVTD of all the substrates. Even though there was no clear response in vivo, this indicates a general stimulation of microbial activity in the rumen and renders moringa leaf meal a promising supplement.
Averaged over all data there was a positive correlation (r2=0.53 p<0.001) between IVTD and in vivo OMD. This correlation was much stronger when calculated for a specific diet (e.g. r2=0.90 p<0.001 for BCS:Co, averaged over the breeds). Analyzing the data for the individual breeds affected correlations only to a minor degree. Thus, when testing a supplementation strategy in vitro, it should be important that donor animals are fed the same components (roughage and supplement) that will be combined at different levels in vitro, whilst the breed of donor animals may be of second importance.
This work provides conclusive evidence that in vitro incubations may be used to design supplementation strategies, thus reducing the need for in vivo experiments. Moringa leaf meal is a promising local resource to substitute for conventional concentrate. Differences in productivity between breeds could be correlated to (and may be partially manifested through) a divergent community structure of rumen microbes. That, in turn, indicates that animals of different breeds might have a ?genetic background? that favours the establishment of a certain community, even if the animals are kept under identical conditions. This relationship should be investigated by more advanced molecular techniques.

K1 Pansen
PP Hohenheim
PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
UL http://opus.uni-hohenheim.de/volltexte/2004/59