Effect of maize processing on diet selection by bovine *

Effects of maize processing on diet selection was studied in 12 fi stulated dry cows. Subexperimental period was repeated twice (8 animals/treatment) and lasted 20 days. Treatments allowed selection between roughage and processed maize: (CG) coarsely-ground, (FG) fi nely-ground and (SF) steam-fl aked. Urea was used with sugar cane (10% of crude protein) to avoid N infl uence over selection. Intake of maize in SF was 45.1 and 42.1% lower than in CG and FG, decreasing total DM and TDN intake, and increasing NDF intake. However, degradable starch intake did not differ. Animals were able to recognize differences among maize physical characteristics.


INTRODUCTION
The concept of processing grains involves increase of starch availability for enzymatic attack by ruminal microbiota and by the host.Methods as steamfl aking, which combines heat and humidity, result in increase of grain superfi cial area with increase of starch digestibility (Theurer et al., 1995).Forbes (1995) summarized many studies involving the hypothesis that animals would be able to distinguish differences among feedstuffs and that they would be capable of choosing the most compatible diet with their physiological status (growth, gestation, lactation).Such choices would be based not only on sensorial perceptions, but also on nutritional composition of feed.

660
The aim of the present study was to relate nutritional differences among maizes processed in different ways to the hypothesis that animals could perceive such differences, being able to balance their energy intake.

MATERIAL AND METHODS
Twelve non-pregnant and non-lactating Holstein cows (650 kg of body weight) with ruminal cannula were housed in individual stalls with subdivided feed bunks, which allowed evaluating feed intake separately.
A randomised block design formed in function of BW and experimental period was used.Sub-experimental period was repeated twice, performing 8 animals per treatment.Treatments corresponded to different methods of maize grain processing, available for animals to manifest their preferences: (CG) sugar cane plus urea and coarsely ground maize (1.6 mm of average particle size-APS), (FG) sugar cane plus urea and fi nely ground maize (0.8 mm of APS) and (SF) sugar cane plus urea and steam-fl aked maize (8.25 mm of APS and density of 270 g/L).
Sugar cane plus urea (10% CP) and maize were supplied ad libitum.Twice daily, at 08.00 h and 16.00 h, 40 g of mineral supplement and 40 g of a mixture of urea and ammonium sulphate were administered directly inside the rumen of each animal (80 g of each mixture/day/animal), in order to avoid that preference would be infl uenced by minerals needs and to avoid nitrogen defi ciency, respectively.
Identifi cation if processing method infl uenced choice process was accomplished through comparison of chemical composition of diets selected by animals.
Each sub-experimental period lasted 20 days.From day -5 to -1, animals received only roughage diet.From day 0 up to day 14, maize was added to animals' diet.During the whole period (-5 day until 14 day), individual intake of all available ingredients was measured.Final intake was obtained from average intake among days 11 and 14.
On day 14, just prior to morning meal (0 h) and at 3, 6, 9 and 12 h postfeeding, samples of ruminal fl uid were taken to proceed determination of pH, ammonia nitrogen by colorimetry and VFA by gas chromatography.
Analysis of variance separated as variation sources the effects of treatments, blocks and periods.Separation of means was done using LSD test, chosen due to its high sensibility, recommended when the objective is to demonstrate equality among treatments (Lentner and Bishop, 1993).

RESULTS AND DISCUSSION
Animals that received treatment SF consumed 45.1 and 42.1% less concentrate in kg than animals from treatments CG and FG (Table 1), respectively.This resulted in decreases of 24.5 and 22.6% in total DM intake, 29.8 and 27.6% in MAIZE PROCESSING AND DIET SELECTION TDN intake in kg, 9.3 and 8.6% in TDN intake in percentage and increases of 45.7 and 47.2% in NDF intake in percentage, respectively.In this trial, a fi xed value was used to estimate maize TDN.
The decrease of concentrate intake and, consequently, of DM may be explained by starch physical characteristics, resulting higher amount of available energy, in steam-fl aked maize than in the other dry processed maizes used in this trial.
With the increase of energy availability, the decrease of concentrate intake by animals from treatment SF may be explained by the theory of metabolic limitation (Conrad, 1966) or related to the process of physiopathological control in the rumen, where increase in acids production would be one of the factors responsible for intake control in those animals (Van Soest, 1994).
Another possibility to explain changes in selected diet is that animals refused the physical appearance of steam-fl aked maize.However, it is possible to suspect that animals were able to chose their diet in function of available energy instead of maize physical appearance once degradable starch intake of the three groups did not differ signifi cantly.

CONCLUSIONS
Offering to bovines the opportunity to choose ingredients differing in physical characteristics, it was observed that choices were non-random and degradable starch intake was kept constant.

Table 1 .
Daily mean intake of feed and nutrients by selection from different treatments Additionally, besides the large difference between quantity and composition of selected diet, ruminal fermentation did not differ too much between treatments, indicating that animals were able to keep rumen environment constant.RODRIGUES P.H.M. ET AL.

Table 2 .
Ruminal fermentation pattern obtained with treatments Treat.-probability for treatment effect; Time×Treat -probability for time×treatment interaction effect, C.V. -coeffi cients of variation; NS -non-signifi cant