Bison bull calves (n=78) were allotted to four treatments in a Latin square study to compare effect of feed delivery systems on animal performance. Feeding systems evaluated were 1) a totally mixed ration, 2) concentrate fed in a fenceline bunk with free choice hay, 3) concentrate fed in an automatic feeder with free choice hay, and 4) concentrate offered free choice in a self feeder with free choice hay. Intake of hay and concentrate varied (P<.05) by feeding system but dry matter intake per day was the same for all treatments. Concentrate intake was higher (P<.05) for the totally mixed ration and hay intake lower (P<.05) with the inverse response from the feeding systems offering free choice hay. Dry matter intake in grams/kg body weight was lower (P<.05) for the totally mixed ration than the other three treatments. Feeding period affected gains (P<.05) with reduced gains occurring during the winter feeding period from Nov 14 to Jan 3. It appears that feeding system has little impact on total dry matter intake and average daily gain but plays a role in feed choice by bison.

Key Words: Bison, Feedlot, Feeding System

Introduction

Bison bulls fed for meet are offered high energy diets to improve rate of gain, feed efficiency, and cost of gain. In addition, animal will reach market weight sooner for a younger, higher quality meat according to the bison carcass grading criteria. Bison feeders are using a variety of feeding systems to deliver the roughage and concentrate components of the ration. Animals confined in a small paddock or pen will establish a definite pecking order which may affect when individual animals eat, how often they eat, how much they eat, and what they eat. Feeding high energy diets may become competitive if feeder space limits access.

Bison feeders who finish their own bulls lack the economy of scale to provide a fenceline totally mixed ration on a daily basis. The most frequently used small scale feeding systems is a self feeders filled with a pelleted or ground concentrate and free choice hay in a ring feeder. (Anderson and Sexhus, 1996). This paper summarizes an on farm study to evaluate animal performance using four different feeding systems for feedlot bison.

Materials and Methods

Seventy eight yearling bison feeder bulls (avg. wt 653 +/- 17.3 lb.) were allotted randomly by order through the chute to one of four identical pens with one treatment administered in each pen. Treatments were the methods of delivering feed to the animals in each pen. They were 1) a totally mixed ration (TMR) of 75% pelleted concentrate and 25% chopped grass hay fed once daily to appetite in a fenceline bunk; 2) pelleted concentrate fed daily to appetite in a fenceline bunk with long grass hay offered separately in a ring feeder (SEP); 3) pelleted concentrate fed in an automatic feeder (AUTO) with long grass hay offered free choice in a ring feeder, and 4) pelleted concentrate offered free choice in a self feeder (SF) with grass hay offered free choice in a ring feeder for large round bales.

The Latin Square design study was conducted at the Double MM Bison Ranch, south of Carrington ND. The trial started August 6, 1996 and animals were rotated to a new feeding system in a pre-planned random order ever 49 days. Animals were weighed individually, identified with two ear tags, and de-wormed at the start of the trial. The pelleted concentrate was formulated using 75% wheat screenings, 15% corn grain, 5% dried molasses and 5% salt and mineral supplement. The commercially manufacture pelleted bison ration averaged 90.32% dry matter, 14.96% crude protein, 15.24% ADF and 28.67% NDF, and 8.17% ash. Grass hay averaged 89.16% dry matter, 7.55% protein, 43.02% ADF, 72.63% NDF, and 9.58% ash. The same low quality grass hay was used for all treatments throughout the trial. The TMR diet was mixed approximately weekly in a three auger "Little Augie" mixer wagon equipped with a scale and stored inside a paved pole barn until fed. The mostly native grass hay was procession in a tub grinder with a 4 inch screen and stored inside until mixed with pellets. The TMR and SEP animals were fed to appetite based on daily bunk readings. The setting for feed delivery in the AUTO treatment was adjusted based on bunk readings. The AUTO feeder is a commercially manufactured feeding system developed and marketed by Sheyenne Advanced Feeding Systems, Cooperstown, ND. The feeder was set to deliver feed 5 times throughout the day by battery powered auger system at approximately 8 AM, 10 AM, 1 PM, 4 PM and 6 PM for a pre-selected amount of time. The tapered feeder with an auger at the bottom delivered feed to a bunk placed under the five discharge spouts. Auger run time for feed delivery was adjusted as required to provide adequate feed to animals in the pen. SF animals were fed using a conventional creep feeder placed in the middle of the pen. Feed was weighed into the AUTO and SF feeders. The feed delivered by the AUTO feeder was collected in pails at each discharge spout periodically and weighed to insure equal dispensing.

Pelleted feed and hay were totaled by pen for each period. Data were analyzed by period and by treatment using ANOVA procedures. Pen was the experimental unit and period was the replicate.

Results and Discussion

The results of this study suggest feeding system may have some impact on the feed intake but limited effect on the performance of bison. Table 1 contains performance data for feedlot bison bulls. Average dry matter intake was less (P<.05) for TMR at 24.26 compared to the other three treatments averaging 26.61 pounds per day. This may be due to differences in hay utilization as the TMR diet was consumed with virtually no waste while the ring feeders had substantial spoiled hay that was not consumed. The TMR hay level was established at 25% of the ration as fed. This appears at first observation to be a lower level of hay intake than the other three treatments with free choice hay. Past experience suggests substantial waste occurs in feeding large round bales to bison. No estimate of hay consumption vs waste was made in this study but hay placed in the feeders averaged 13.27 pounds per head per day vs 6.79 pounds consumed in the TMR (P<.05). Animals compensated for the apparent increased hay intake by reducing (P<.05) pelleted concentrate intake in the three diets with free choice hay (20.37 lb. pellets for TMR vs 16.82 for diets with free choice hay) but not to the extent that dry matter intake was equalized.

Total feed disappearance was 2.98 pounds less as fed and 2.65 pounds less in dry matter terms for the TMR vs the other three diets (P<.05). A greater amount of hay should possibly have been used in the TMR for equivalent ingredient intake. The higher concentrate level in the TMR (P<.05) would be expected to produce improved daily gains, which it did not.

Gains were less than desired overall with TMR and SEP treatments producing 1.40 and 1.36 pounds of gain per day compared to 1.50 and 1.49 for AUTO and SF respectively during the 195 day trial. Individual treatments means for average daily gain were not different (P>.10), nor were pooled means for bunk feeding (TMR and SEP) vs self feeding systems (AUTO and SF).

Observation of the TMR diet consumption indicates animals were reluctant to clean up the chopped hay, especially the more finely chopped particles. Animals in the SEP treatment were also reluctant to consume the fines from pellets remaining the bunk. This agrees with observations by commercial bison feeders that once started on one form of feed, bison are reluctant to consume a different form, especially from pellets to fines, or long hay to chopped hay. Bison exposed to the AUTO feeder became conditioned to the sound of the auger motor running, and readily came to eat. However, the natural tendency of bison is to consume smaller amounts of feed and eat more often. The AUTO feed delivery system worked well throughout the feeding trial and proved to be a convenient method of delivering a concentrate. The self feeder was a conventional rectangular calf creep feeder accessible from both sides.

Differences in gain were observed during the third feeding period compared to periods 1, 2 and 4 when the most severe winter weather occurred and the days were shortest. Daily gains average .66 pounds from Nov. 14 to Jan 3. During the two previous feeding periods, gains averaged 1.91 pounds. This finding agrees with other observations on feedlot performance of bison during the winter (Anderson and Miller, 1997, and Christopherson et. al, 1979). It is apparent that management strategies need to be developed to counter this reduction or alter the diet to reduce cost of feed during periods of naturally slower growth for bison bulls fed for meat.

Gains observed in period 1, (Aug 6 - Sep 25), period 2 (Sep 26 - Nov 14), period 3 (Nov 15 - Jan 3) and period 4 (Jan 4 - Feb 22) were 1.97, 1.85, .66 and 1.27 respectively. Dry matter intake by period averaged 25.01, 28.43, 27.56, and 23.96 respectively. Reduced intake in period 4 is difficult to explain. Animals averaged 931 +/- 11.6 pounds at the end of the trial. Some animals were ready for market and may have reduced intake based on excellent body condition.

Implications and Future Trials

Feeding systems appear to have some impact on feed intake patterns, but it is difficult to separate actual intake from spoiled feed using large bale ring feeders. Intake of concentrates appears to be inversely related to hay intake. The relatively modest energy commercial bison pellet used in this study is used by many commercial bison growers. In this study it produced gains that were less than most producers would consider economical. Higher energy diets are recommended as determined by Anderson and Miller, (1996). Additional studies are needed to determine the effect of forage quality on animal performance and feeding methods for reducing waste in free choice hay feeders. Other feeding systems may offer some potential advantages as well, such as round self feeders, or increased bunk space. Bulls in this trial were horned. Dehorned bulls may prove to be more adapted to conventional fenceline feeding systems.

Conducting Latin square studies with such long feeding periods, especially when confounded with season, is not ideal. Considering the amount of information gained and the logistical difficulties of doing nutrition research with bison, the data developed in this study is most useful. Bison research is needed to develop optimum nutritional strategies. On-farm studies as well as university trials, when adequate resources are available, can provide data to commercial bison growers. On-farm studies can be replicated as a Latin square design used in this study or with multiple producers comparing the same two (or more) treatments to gain appropriate statistical confidence in the response.


Literature Cited

Anderson, Vern and Bryan Miller. 1997. Influence of season and diet on feedlot performance of bison. Prof. An. Sci 13:14-17.

Anderson, V. L. and Dennis Sexhus. 1996. Current practices and priorities of bison feeders in the Northern Plains. Carrington Research Extension Center-NDSU Beef and Bison Field Day Proceedings Vol 19:225-29.

Christopherson, R. J., R. J. Hudson, and M K. Christopherson. 1979. Seasonal energy expenditures and thermoregulatory responses of bison and cattle. Can J. Anim. Sci. 59:611.

Miller, Bryan, and Vern Anderson. 1996. Comparison of energy level, corn, type, and corn processing for bison bulls in the feedlot during four seasons of the year. Carrington Research Extension Center-NDSU Beef and Bison Field Day Proceedings Vol 19:21:24.

Table 1. Effect of feeding system on feedlot performance of bison bulls during four feeding periods.

	TMR -1	Separate Grain/Hay	Automatic Feeder	Self Feeder	St. Err.
Initial wt, lb.	649	648	657	653	9.88
Pellet Intake, lb.	20.37-a	17.64-b	16.99-b,c	15.83-c	1.09
Hay Intake, lb.	6.79-a	11.82-b	12.88-b	15.10-c	.73
DM Intake, lb./day.	24.26-a	26.34-b	26.70-b	27.65-b	1.06
DM Intake, g/kg body wt.	30.09-a	32.81-b	33.24-b	33.79-b	1.48
Average Daily Gain, lb.	1.40	1.36	1.50	1.49	.30
Feed/unit gain	21.92	21.62	22.24	27.44	3.92

¹ Totally mixed ration

^{a, b, c} Values with different superscripts are significantly different (P<.05)

Appreciation is expressed to the North Dakota Agricultural Products Utilization Commission

and the North Dakota Buffalo Association for support for this study.