For all trials, the ration consisted of a free choice grain mixture as well as approximately 160 grams/head/day of alfalfa pellets which were hand fed. The grain mix formula was 820 kg of whole barley mixed with 120 kg of a custom lamb grower supplement. The resulting mixture was 14.7% crude protein and 72.5% TDN. The supplement was fortified with minerals and vitamins and was medicated with lasalocid sodium at a rate of 126 mg/kg as an aid in the prevention of coccidiosis. Clean water was also available at all times. All feed was weighed into the feeders and any wasted or refused feed was weighed back.

Lambs were weighed at the start of each trial: December 22/99 (Trial 1), October 06/00 (Trial 2) and November 16/00 (Trial 3) and every 14 days thereafter. Weighing conditions were the same for all groups and over all weighing periods. Lambs were shipped to the University of Guelph Meat Laboratory when they reached a live weight of 45 kg or higher. Lambs were shipped every 2 weeks. At the Meat lab, further data such as carcass weight, GR measurement, fat scores, and rib eye area were measured. This paper deals only with the feeding data.

In trial 1, three pens of each genotype (SUx, COx, and OLIBS) were used. The assignment of genotype to pens was done randomly. In trials 2 and 3, only a small number of COx lambs were available, and no OLIBS lambs were available. Thus, only one pen of COx lambs was used in trials 2 and 3. Three pens of SUx lambs were used in trial 2 and two pens in trial 3.

Results of Trial 1 were analysed using analysis of variance. Where the analysis indicated that significant differences existed, the least significant difference method was used to compare means. Trials 2 and 3 could not be statistically analyzed due to a lack of replication on the COx pens.

Results:
Trial 1:
Starting weights were significantly higher for the SUx lambs as compared to the OLIBS lambs (Table 1). Lambs were somewhat lighter than anticipated at the start of the trial but were in good health.

During the first 56 days of the feeding period, the SUx lambs gained significantly faster than the COx and OLIBS lambs (Table 1). Absolute gains for the SUx lambs were 365 grams/head/day (0.8 lb/hd/d) which is acceptable but not outstanding for this feeding program and type of lambs. Average daily gain calculated until each lamb reached its target weight was again highest for the SUx lambs, averaging 372 grams/head/day. In both cases, the average daily gain was not significantly different between the CO cross lambs and the OLIBS lambs.
The SUx lambs reached their target weight in an average of 73 days (range 58 to 98 days) as compared to 80 days (range 70 to 98 days) for COx lambs and 83 days (range 70 - 98) for OLIBS lambs. However, there was not a statistically significant difference between the 3 genotypes for days to finish.

Feed to gain ratio is defined as the kilograms of grain mixture required to produce a kilogram of lamb bodyweight gain. In this trial, feed:gain ratios were quite similar across genotypes, averaging 3.4:1 for the SUx lambs and 3.6:1 for the COx lambs. Feed:gain ratios were slightly higher when calculated over the entire feeding period. There were no significant differences in feed:gain ratio in either case.

Top of Page
Top of Report

Trial 2:
Trial 2 consisted of 1 pen of 7 COx lambs, and 3 pens with a total of 51 SUx lambs. The SUx lambs were sorted into pens by starting weight (Table 2). Over the first 42 days, the average daily gain ranged from 349 to 373 g/hd/d. The feed:gain ratio over the same period varied widely, ranging from 2.8:1 for the lightest group of SUx lambs to 4.9 for the COx lambs. Over the entire feeding period until the lambs reached the target weight, the average daily gain ranged from 330 to 348 g/hd/d. For the entire feeding period, the feed:gain ratio ranged from 4.2:1 to 4.9:1 for the SUx lambs, and was 6.0 for the single pen of COx lambs. While the difference in feed:gain ratio between SUx lambs and COx lambs is relatively wide, statistical analysis cannot be applied since there was no replication of the COx pen.

Trial 3.
In trial 3 a single pen of 8 COx lambs and 2 pens of SUx lambs with a total of 23 lambs were used. Starting weights were higher than for the previous two trials. Average daily gain for the first 28 day period was widely variable, ranging from 238 to 340 grams/hd/d. The feed:gain ratio for the same period was 4.3 to 4.8 for the SUx and 5.7 for the COx. Over the total feeding period (up to day 69), average daily gains were 264 g/hd/d for the COx lambs and from 313 to 332 for the SUx lambs. Feed:gain ratio for the total period averaged 6.0 for the COx lambs and 4.7 for the SUx lambs. As with trial 2, lack of replication in the COx pens did not allow statistical analysis to be done.

Top of Page
Top of Report

Cost of Gain
In the spring of 2001, the supplement pellet cost approximately $400/mt (fob mill) and local barley was about $120/mt. At a mixing rate of 18% supplement and 82% barley, the mixture cost per tonne was $170.40. This equates to a cost of 7.7 cents per pound.

Using the performance data from Table 1 above, the calculated cost for grain supplement per pound of liveweight gain ranged from 26.3 cents to 28.1 cents (Table 4). Feed cost per day was between 20 and 21 cents and total grain cost to reach the target weight ranged from a low of $15.33 for the SUx to a high of $16.60 for the OLIBS cross. Note that these costs do not include the alfalfa pellets, which cost about 3.5 cents per head per day. On most farms this would be replaced with alfalfa hay at a much lower cost.

Summary:
Overall average daily gains on these trials were somewhat below expectations but still reasonable. Cold weather in trials 2 and 3 may have affected daily gains and feed efficiency as compared to Trial 1. Temperatures inside the feeder barn are only slightly above ambient. Feed to gain ratios in Trial 1 were very good as compared to industry standards. This series of trials indicates that while daily gains did vary between genotypes, differences in feed efficiency were less clear. Trials 2 and 3 suggested that the SUx lambs had superior feed efficiency, but this cannot be stated with certainly due to the lack of replication discussed earlier. The cost figures presented above can serve to provide guidelines for feedlot lamb cost of production. Carcass quality data will be presented in a following article.

Top of Page
Top of Report

Establishment Methods for Forages
Jim Johnston and Matt Bowman
New Liskeard Agricultural Research Station

Background:
On improved land, most forages mixtures include alfalfa. Alfalfa is usually established by seeding the forage mixture under a cereal companion crop. In the year of seeding, a crop of grain is harvested along with the straw. In the following year (first production year), the forage mixture is harvested for hay or silage. This establishment method can be problematic since the small grain crop competes with the underseeded forages for light, water, and nutrients. It has become more common in recent years to remove the small grain crop in the vegetative stage as whole plant silage. This allows the underseeded forage most of the summer (from July onwards) to grow without competition from the small grain crop.

Another option for establishing alfalfa-based mixtures is direct seeding. Direct seeding means that no companion crop is used. Some concerns with direct seeding include increased risk of erosion in spring since there is less ground cover during forage establishment and weed control can be more critical since there is no companion crop to compete with the weeds.

In cash-crop rotations, it is desirable to include a legume ploughdown crop to return nitrogen and organic matter to the soil. Red clover is most commonly considered as a ploughdown crop in Ontario. Ploughdown crops such as red clover are often sown with spring cereals, but some producers in have reported success establishing red clover under a canola crop.

If this method was reliable, it would provide another option in the crop rotation for short-season areas where cropping choices are limited to small grains, canola, and pulse crops like field peas.

This trial was designed to evaluate the effectiveness of establishing either alfalfa-timothy or red clover-timothy mixtures via direct seeding or by a range of companion crops including spring grains and canola.

Top of Page
Top of Report

Methods:
The forage mixtures to be established were alfalfa-timothy and red clover-timothy (Table 1). The establishment options were: direct seeding, barley silage, Polish canola in 7" rows, Polish canola in 14" rows, Argentine canola in 7" rows, Argentine canola in 14" rows, and barley grain (Table 1). These treatments were established at New Liskeard in 1999 and in 2000.

In the seeding year, the grain crops were all direct combined (no swathing). Forage harvests were taken from the 1999 establishment in 2000, and from the 2000 establishment in 2001.

The trial utilized a randomized complete block design laid out as a split plot. Main plots were establishment method and subplots were legume mixture. Data collected included forage yields and mixture composition. Data was analyzed using analysis of variance. When significance was indicated, mean separations were done using the protected LSD method.

Results:
Tests were established in 1999 and 2000. Each test had forage harvested in the first production year only (ie: 2000 forage harvest from the 1999 seeding, and 2001 forage harvest from the 2000 seeding). Each test will be referred to by the year in which it was harvested (2000 and 2001). Results for the 2000 and 2001 harvests are presented separately.

In the first cut of the 2000 harvest, forage mixtures that were established in 1999 by direct seeded or with a barley silage companion crop were the highest yielding (Table 2). Forage yield following establishment with Polish canola were intermediate in yield, while those established with Argentine canola or a barley grain crop were lowest in yield. In the second cut, no differences occurred in response to establishment method. Total seasonal yield showed similar results to the first cut, where the highest forage yields were from direct seeded forage and barley companion crops, Polish canola companion crops gave intermediate forage yields, and Argentine canola and barley grain companion crops gave the lowest forage yields.

Top of Page
Top of Report

In all of the 2000 harvests (1^st, 2^nd, and total yield), red clover-timothy mixtures outyielded alfalfa-timothy mixtures (Table 2). Results from the second cut and from total seasonal yield showed a significant interaction between establishment method and forage mixture (Table 2). In both cases, the cause of this interaction was that the barley silage companion crop had either a higher yield from the alfalfa mixture as compared to red clover mixture or else had no difference between the mixtures. In all other cases, the red clover mixture outyielded the alfalfa mixture by a wide margin. The grass content of the alfalfa mixtures was significantly higher than the red clover mixtures. This indicates that conditions for establishment of red clover were more favourable than for alfalfa. Field notes also indicated more weeds in the alfalfa plots, but sufficient records for statistical analysis of this point are not available.

Only one cut was taken from the 2001 harvest (2000 seeding). Forage yields from the 2001 harvest were surprising; the top forage yields resulted from a Polish canola companion crop seeded in 35 cm rows (Table 3). Other establishment methods with statistically equal forage yields were Polish canola in 17 cm rows, direct seeded forage, and Argentine canola in 35 cm rows. Intermediate forage yields resulted from the barley silage companion crop, while the lowest forage yields resulted from the Argentine canola in 18 cm rows and the barley grain companion crops. Legume content was highest in the barley silage, barley grain and direct seeded treatments. Grass content was highest in the canola treatments as well as the direct seeded forage. Weed content was highly variable and not significantly different among treatments.

While there was a tendency for red clover mixtures to outyield alfalfa-mixtures in the 2001 harvest, the difference was not significant (Table 3). Red clover mixtures had higher legume content and lower grass and weed content than alfalfa mixtures in the 2001 harvest (Table 3). Interactions between establishment method and legume type were all non-significant in the 2001 harvest.

Top of Page
Top of Report

Discussion:
Results of the 2000 harvest were as expected: namely that the highest forage yields resulted from establishment methods that maximize the amount of sunlight getting to the forage plants and minimize the competition for water and nutrients. Direct seeded forage fits this description best since no companion crop is present to compete with the forage plants. Barley silage also fits this description since the crop is removed in July, leaving half of the growing season for the forage plants to establish. Canola and barley grain fall at the other end of the spectrum, since they are longer season crops that leave little time after maturity for the forage crop to grow unimpeded.

In the 2001 harvest, direct seeded forage was still among the best treatments and barley grain was the poorest; however, 3 of the 4 canola companion crop treatments were much better than expected. Given the 4 canola treatments tested, we might expect 35 cm rows to show an advantage (to the underseeded forage) since more light would penetrate the canopy. Similarly, we would expect Polish canola to show an advantage since it matures much earlier, leaving more time for the forages grow without competition. Given these considerations, it is reasonable that we found Argentine canola in 17.5 cm rows to be poorer in terms of forage establishment than the other three canola treatments combinations. What is surprising is that the other canola treatments were equivalent in subsequent forage yield to the direct seeded forages. Also surprising is that the barley silage treatment had forage yields that were slightly lower than the Polish-35 cm treatment. The barley silage treatment did however, have subsequent forage yields statistically equal to the direct seeded forage and the other two canola treatments. These results indicate that canola may be more successful as a companion crop than previously thought. It should be noted that two factors could affect these results when tested on a field scale. The first is that the small plots were kept weed free or with low weed pressure without herbicides. In the field, herbicides used on canola will damage forage seedlings. It is likely that with early spraying onto a thick canopy of canola and weeds, little damage to the forage legumes would occur, but this could not be assessed from the present trial. The other factor is that in the field, canola often lies in the swath curing for several weeks. This could lead to smothering of the underseeded forages. In our plot trials, canola was direct combined, thus we have no estimate of the damage that may occur to underseeded forages from canola swaths.

In the 2000 harvest, red clover mixtures outyielded alfalfa mixtures, and this trend was evident in 2001 as well. In addition, red clover mixtures had less grass in both years, indicating a stronger stand of the legume component for red clover. This was expected, since red clover is known to be quite vigorous in the seeding year. Given the good stands of red clover obtained from canola companion crops, it seems probable that a red clover ploughdown could be established in the canola year of a cash crop rotation.

Conclusions:

1. Based on subsequent forage yields, direct seeded forages are generally the most reliable way to establish legumes, provided that seedbed preparation and weed control are good.

2. Barley for silage is also a reliable method of establishing forage legumes.

3. Canola gave a variable response as a companion crop, with better results from Polish types and from wider (35 cm) row spacings. Canola underseeded to red clover appears to be viable as a ploughdown option, assuming the legume is not severely damaged by the canola herbicide.

4. Barley for grain was consistently the poorest option for establishing forages. Under normal barley management (full seeding rate and fertility), this should not be considered as a means of establishing a vigorous, high yielding forage crop.In the spring of 2001, the supplement pellet cost approximately $400/mt (fob mill) and local barley was about $120/mt. At a mixing rate of 18% supplement and 82% barley, the mixture cost per tonne was $170.40. This equates to a cost of 7.7 cents per pound.

Top of Page
Top of Report

Alfalfa Harvest Management Strategies for Northern Ontario:
1. Yield and Persistence
Jim Johnston and Matt Bowman
New Liskeard Agricultural Research Station

Alfalfa is the most common forage legume used in Ontario for hay and silage production. In longer season areas, three cuts per year is recommended, with a critical fall rest period being set aside to allow root reserves to build prior to winter. Research conducted in Ontario during the 1970's on alfalfa harvest management and the critical fall rest period included sites as far north as Dufferin county, but did not include sites in northern Ontario. A general recommendation in the north is to take only two cuts per year and allow a rest period from approximately Aug 15 to September 25 or until two killing frosts have occurred. However, in some northern areas, such as the Temiskaming district, there is frequently sufficient growth to take three cuts per year. The effect of this management on stand survival and productivity has not been formally assessed. This trial was conducted over a 6 year period to evaluate several harvest management options for forage yield and quality, winter survival, and subsequent stand performance. This report discusses forage yield under various cutting systems and the subsequent forage yields in response to those treatments.

Methods:
All trials were direct seeded to pure alfalfa, variety Centurion, at 13 kg/ha. Trials were seeded at New Liskeard in 1994 and 1995. Each alfalfa plot had a plot of timothy on each side to act as a guard. Six harvest management treatments were used (Table 1). Harvest management treatments were based on a combination of number of cuts, plant development, and calendar date. Plant development was determined using the system of Mueller and Fick (1989) in which 100 plants were selected randomly and hand-separated by stage of development using a numbering system where 3 = early bud, 4 = late bud; 5 = early bloom, and 6 = late bloom. The cutting systems used were intended to mimic systems that would have practical applications on farm. For example, a standard and standard plus systems would be typical for a cow-calf or cash-crop hay operator, while the intense, short rest, and medium rest systems would more likely be used on a dairy operation.

Top of Page
Top of Report

To evaluate the effect of the harvest management treatments on future stand productivity, the treatments were applied in the first production year, then 2 cuts were taken in the second production year to measure the response to the treatments the previous year. The harvest management treatments were then applied again in the third production year, followed by only two cuts in the fourth production year, again to measure the response to the treatments imposed in the 3^rd year. By using this harvesting sequence, we were able to judge the effect of the three cuts systems on crop persistence over several winters (Table 2).

Results:
Alfalfa established by direct seeding in 1994 was subjected to the six harvest management treatments in 1995. Forage yields were highest under the long rest and standard + treatments (Table 3). The standard 2 cut, intense (3 cut), and short rest (3 cut) systems were lowest in yield.

The 1996 harvest measured the response to the 1995 treatments. Results showed that the standard 2 cut system and the long rest system produced the highest yields, while the intense and short rest systems produced the lowest yields. The 6 harvest management treatments were again applied in 1997. Results were generally similar to 1995, with highest yields resulting from the long rest and the standard + systems, while the intense system was lowest yielding. In response to these treatments, the 1998 yields showed no significant difference among treatments, indicating that the harvest management treatments applied in 1997 did not affect winter survival and subsequent yield in 1998.

Top of Page
Top of Report

Alfalfa was established by direct seeding in 1995 and was subjected to the six harvest management treatments in 1996. Forage yields were similar among the standard +, short rest, medium rest, and long rest treatments. (Table 4) Yields were lowest for the intense treatment. In response to these treatments, the 1997 yields showed no significant difference among the six treatments. In 1998, the six harvest management treatments were to be imposed, but only the standard 2 cut and the intense system were completed. Poor fall growth and very wet soil conditions prevented the final harvest of the other treatments. However, there was no difference in yield between the standard two cut and the intense treatment in 1998, nor was there any yield difference in response to these treatments in 1999 (only a single cut was taken in 1999).

Discussion:
When considering the advantages of any forage harvest management system, forage yield, forage quality, and stand persistence must be considered. The relative importance of each of these factors will vary with the type of farm operation, but in general, yield and stand persistence are very important. High yields are desirable to offset land-based costs such as rent or mortgage costs. In some operations, persistence may be needed for only 3 or 4 years before the land is rotated to other crops. In other operations, a forage stand may be kept in production for 10 or more years if persistence is adequate.

The results of this study indicate that 3 cut systems often, but not always, outyield 2 cut systems. The total yield from two cuts taken at the early bloom stage is usually similar in yield to a 3 cut system where all cuts are taken at the mid to late bud stage, or where all 3 cuts are completed prior to mid-August. Higher yields occur where two cuts are taken at early bloom plus a third cut in late September, or where the first two cuts are taken at the mid to late bud stage and the third cut is taken at full bloom stage. In practice it is difficult to use stage of development (bud or bloom) to time fall harvests, since alfalfa flowering is reduced greatly by short days in the fall. Thus, if both cut 1 and cut 2 are taken at the bud stage, a long rest period of 40 to 50 days would be required to simulate a bloom-stage harvest in the fall.

Top of Page
Top of Report

In this study, stand persistence was based on the yield of forage in the year following the 3 cut treatments. Our expectation was that 3 cut systems that did not allow the plants to flower at least once during the summer (ie: intense system) or that ignored the critical harvest period (short rest system) would have lower yields the following year. However, this result occurred only in the 1996 harvest of the 1994 seeding. In all other response years (1998 - Table 3; 1997 and 1999 - Table 4) there was no difference in yield among the harvest management treatments. This suggests that at New Liskeard, a range of 3 cut systems can be utilized without a high risk of stand damage, but the risk of damage is greater when sufficient rest periods are not given. Finally, it should be noted that several factors other than harvest management will affect alfalfa survival. These factors include stand age, disease pressure, soil drainage, and soil fertility, especially potassium levels. Management to improve soil drainage and fertility in particular, will increase the chances of alfalfa survival under three cut management.

Conclusions:

1. Three cut systems that have all cuts taken at the bud stage will yield equal to or less than two cuts taken at the early bloom stage.

2. Three cut systems that allow one cut to reach the bloom stage will usually outyield a 2 cut system.

3. Three cut systems that have all cuts taken at the bud stage or that do not allow a long rest period prior to the final cut, are at a higher risk of winterkill, but lower yields as a result of such management occurred only once in four years in our tests. Note that our site is tile drained and has high potassium soil test values.

4. Three cut systems that allowed the plants to reach the bloom stage or allowed a long rest (40 days or more) prior to the final cut consistently yielded equal to the 2 cut system in our tests. While one would expect to eventually see a yield reduction under these three cut systems, it did not occur up to the 4^th production year of the stands we tested.

Top of Page
Top of Report

Harvest Management Strategies for Northern Ontario:
2. Yield and Persistence
Jim Johnston and Matt Bowman
New Liskeard Agricultural Research Station

Alfalfa is the most common forage legume used in Ontario for hay and silage production. In longer season areas, three cuts per year is recommended, with a critical fall rest period being set aside to allow root reserves to build prior to winter. Research conducted in Ontario during the 1970's on alfalfa harvest management and the critical fall rest period included sites as far north as Dufferin county, but did not include sites in northern Ontario. A general recommendation in the north is to take only two cuts per year and allow a rest period from approximately Aug 15 to September 25 or until two killing frosts have occurred. However, in some northern areas, such as the Temiskaming district, there is frequently sufficient growth to take three cuts per year. The effect of this management on stand survival and productivity has not been formally assessed. This trial was conducted over a 6 year period to evaluate several harvest management options for forage yield and quality, winter survival, and subsequent stand performance. This report discusses forage quality in each of the harvest management systems.

Methods:
All trials were direct seeded to pure alfalfa, variety Centurion, at 13 kg/ha. Trials were seeded at New Liskeard in 1994 and 1995. Each alfalfa plot had a plot of timothy on each side to act as a guard. Six harvest management treatments were used (Table 1). Harvest management treatments were based on a combination of number of cuts, plant development, and calendar date. Plant development was determined using the system of Mueller and Fick (1989) in which 100 plants were selected randomly and hand-separated by stage of development using a numbering system where 3 = early bud, 4 = late bud; 5 = early bloom, and 6 = late bloom. The cutting systems used were intended to mimic systems that would have practical applications on farm. For example, the standard and standard plus systems would be typical for a cow-calf or cash-crop hay operator, while the intense, short rest, and medium rest systems would more likely be used on a dairy operation.

To evaluate forage quality, samples were taken from each harvest management treatment at each harvest date. Samples were analyzed for crude protein (CP), acid detergent fibre (ADF), and neutral detergent fibre (NDF). Crude protein is estimated from the total nitrogen in a sample, with a target value of 20% in dairy quality hay. ADF provides an estimate of forage digestibility, with a target value of 30% in dairy quality hay. NDF provides an estimate of forage intake potential, with a target value of 40% in dairy quality hay. All quality parameters are expressed on a 100% dry matter basis.

Top of Page
Top of Report

Results:
Cutting dates were determined by visual observation of the plots and estimating when the growth was at the appropriate stage of development for each treatment. When the plots were cut, samples were taken to determine mean stage by weight (MSW). Cutting dates for all harvests in 1995 (94 seeding) and 1996 (95 seeding) are shown in Table 2 and Table 3. In addition, the MSW calculated for each treatment in cuts 1 and 2 are shown in association with the corresponding cutting date. MSW values are not shown for cut 3 due to the lack of bud and flower development on late summer and fall growth.

In 1995, the actual MSW values for treatments 1 and 2 were lower than the targets (Table 2), indicating that visual observation tended to rate the stands as being more mature than they actually were based on the MSW. In 1996, we left these two treatments longer before taking the first cut and the MSW values were much closer to the target values (Table 3). The MSW values for treatments 3 to 6 were reasonably close to the target values in both years for cut 1. For cut 2, MSW values were close to target in 1995 but lower than target in 1996.

Forage quality for cut 1 and cut 2 in each year is reported as the mean of Treatments 1 and 2 and the mean of Treatments 3 to 6, since these groups had the same cutting date (Table 4 and Table 5). For cut 3, each individual treatment mean is presented, except for treatment 1, which is the 2 cut system.

As expected, first cut forage harvested at a lower MSW (earlier calendar date) had much better quality values (higher CP, lower ADF and NDF) (Table 4 and Table 5). Absolute quality in 1995 was close to expected values, with Treatments 1 and 2 giving average quality hay, and treatments 3 to 6 being close to dairy quality values. In 1996, absolute quality was much poorer, only CP levels were comparable to 1995. The fact that the MSW values for treatments 3 to 6 in 1996 were similar to 1995, yet the forage fibre levels were much higher in 1996 indicates that MSW may not be a reliable indicator of forage quality at this location.

At cut 2, forage harvested at a lower MSW (Treatments 3 to 6) had better quality. Absolute values were somewhat poorer than cut 1 in 1995 and better than cut 1 in 1996.

hird cut quality varied widely but was related to rest period and cutting dates among treatments (Table 4 and Table 5). In both years, the best quality in cut 3 was from treatments 2 and 3. The result for treatment 2 in 1995 is somewhat surprising since there was a relatively long rest period prior to cut 3 in that treatment (Table 2). In 1995, all cut 3 quality results were of dairy quality for fibre levels, although protein levels dropped below the 20% target for treatments 4 to 6. In 1996, treatments 4 to 6 had much higher fibre levels. This is related to the much later cutting dates for cut 3 in 1996.

Top of Page
Top of Report

Discussion:
The MSW system was developed at Cornell University in New York state. In New York, the MSW system was shown to be a reliable indicator of alfalfa quality. Within each year of the test, lower MSW values clearly were related to higher forage quality. However, between the two harvest years, similar MSW values corresponded to quite different absolute quality values. In fact, for cut 2, MSW values of 3.2 to 3.3 in 1995 corresponded to better quality results than MSW values of 2.8 to 3.0 for the same cut in 1996. The expected result would have been that the lower MSW value should result in higher quality. While we have not completed our analysis of the MSW system, observations indicate that it may not be suitable for predicting alfalfa quality in our location, except in a very general way. Variation in daylength and temperature may play a role. We also found that visual observations often estimated alfalfa maturity to be more advanced than the MSW system, particularly at the bloom stage. Further evaluation of the relationship between MSW and alfalfa quality needs to be done at our location.

For cuts 1 and 2 in both years, there were very clear differences between quality at the bloom stage (Treatments 1 and 2) and the bud stage (Treatments 3 to 6). For cut 1, these quality differences resulted from a 9 day later cutting date in 1995 and a 13 day later cutting date in 1996. The absolute values indicate that for animals with high nutritional requirements, first cut needs to be taken prior to the bloom stage. However, for stock with lower requirements, the bloom stage quality was acceptable and this cutting stage is associated with higher dry matter yields. On dairy operations, where the bulk of the forage must be of high quality, it is clear that both of the first 2 cuts should be taken prior to the bloom stage. This means that management of the third cut is quite important for assuring adequate persistence (see article 1 in this series).

Quality at cut 3 was quite good for all treatments in 1995 and for treatments 2 and 3 in 1996. For certain farm operations were only a small quantity of the total forage required needs to be of the highest quality, a system such as treatment 2 could be used to obtain some high quality forage and still retain maximum yields. It is important to consider, however, that cut 3 generally provides only 20% to 35% of total seasonal yields and that third cuts taken after mid-September often have to be made into silage due to poor drying weather. Late fall harvests after a very long rest period (Treatment 6 in 1996) are bound to result in higher fibre levels since the plant continues to form lignin even though the signs of maturity such as flowering may be absent.

The balance between forage yield, quality, and persistence varies between farm operations. On dairy farms, forage quality is the overriding concern and yield and persistence follow from that management. Other operations can time their cuttings to increase yield and persistence since quality requirements are lower. Given the lower proportion of total yield from cut 3, dairy operators are advised to obtain the desired forage quality from cuts 1 and 2, then managed cut 3 to increase stand persistence. Other operators can obtain high yields and ensure persistence from cuts 1 and 2, then use cut 3 for some higher quality forage.

Conclusions:

1. For dairy quality forage, producers must harvest cuts 1 and 2 at the pre-bloom stage, then focus on managing cut 3 for persistence (medium or long rest system).

2. For average quality forage, cutting at the early bloom stage is sufficient. This will provide higher forage yields and should increase persistence following severe winters.

3. Two cuts taken at the early bloom stage can be followed by a late September harvest to obtain a portion of the total forage at high quality levels. This system may be useful for cow herds calving in March and April.

4. The Mean Stage by Weight (MSW) system is useful as a gross indicator of forage quality, but is inconsistent over years at this location. More investigation of the MSW system is needed at New Liskeard to assess its usefulness on farm.

Top of Page
Top of Report

Management of Alfalfa in the Seeding Year
Jim Johnston and Karen Davies
New Liskeard and Thunder Bay Agricultural Research Stations

Alfalfa can be established by direct seeding or by using a companion crop. If forage is required in the year of seeding, then direct seeding or a cereal silage companion crop are used. In short season areas (less then 2600 corn heat units), it is generally recommended that only 1 harvest of forage be taken in the seeding year. Work in northern Ontario has shown that in some years, the second growth following a cereal silage harvest is sufficient to justify harvesting it for stored feed. It is unclear whether taking a second cut in the establishment year is damaging to the subsequent alfalfa crop. This study was undertaken to assess forage yield in the seeding year and in the first production year from direct seeded alfalfa and from alfalfa seeded with an oat companion crop, when the seeding year growth was harvested either 1 or 2 times.

Methods:
Alfalfa stands were established at the Thunder Bay Research Station in 1998 and 1999. Subsequent forage yields were measured in the year following establishment. Treatments were: establishment method (direct seeding vs. oat companion crop) and number of cuts in the establishment year (1 vs. 2). Details of the treatments are shown in Table 1.

Top of Page
Top of Report

Results:
In both establishment years, oat companion crops yielded significantly more than direct seeded alfalfa (Table 2). The response to number of cuts in the seeding year varied. In 1998, direct seeded alfalfa yielded significantly more from 2 cuts than 1 cut, but there was no significant difference between 1 and 2 cuts for the oat companion crop. In 1999, yield was higher from 2 cuts than 1 cut for both the direct seeded alfalfa and for the oat companion crop.

Yield distribution in 1998 was weighted heavily in favour of the first cut, while in 1999 the yield was more evenly distributed.

Top of Page
Top of Report

Forage yields in the year following establishment were generally higher when the establishment method was direct seeding as compared to an oat companion crop (Table 3). The advantage in favour of direct seeding was significant in all harvests in 1999. The same trend occurred in the 2000 harvest for first cut and total yield, but the differences were not significant.

In the 1999 harvest, forage yield was not affected by the number of cuts taken in the previous (seeding) year. In the 2000 harvest, forage yields were significantly greater when only one cut had been taken in the seeding year (Table 3).

Discussion: Our results show that oat companion crops will yield more forage in the seeding year than direct seeded alfalfa. Therefore, if forage yield in the seeding year is very important, a cereal silage companion crop would be recommended. Taking a second cut in the seeding year also increased forage yield in 3 of 4 situations. Under direct seeding, the second harvest in the seeding year is usually pure alfalfa. With an oat companion crop, the second cut often contains oat regrowth along with newly established alfalfa.

Top of Page
Top of Report

This is especially true when the oat companion crop is harvested at the late boot or early heading stage. Oats cut at this stage can have very heavy regrowth which can continue to compete with the underseeded alfalfa. Delaying the first cut until the oat heads are fully emerged will reduce the thickness and vigour of the cereal regrowth, providing better conditions for the underseeded alfalfa. Under dry soil conditions, oat regrowth will be minimal and the new seeding of alfalfa may also be slow to establish. Under this scenario, a second harvest in the seeding year will not be economical.

In the first production year following the 1998 seeding, direct seeded alfalfa outperformed alfalfa established under an oat companion crop. This finding has since been supported by other trials at New Liskeard. If forage yield in the seeding year is not critical, direct seeding of alfalfa should be considered due to increased hay yields in the first production year. The effect of taking 1 vs. 2 cuts in the seeding year varied. No effect on subsequent alfalfa yields occurred in 1999, but in 2000, taking 2 cuts in the seeding year resulted in a significant reduction in subsequent alfalfa yields. This is not unexpected, since winter survival varies from year to year depending on weather conditions. It is apparent that taking a second cut in the seeding year does impose more stress on the alfalfa, leaving it more susceptible to damage during a harsh winter. In this study, the maximum yield obtained from a second cut in the seeding year was about 1200 kg/ha. In the first production year, the yield reduction between 1 and 2 treatments was over 3000 kg/ha. It is quite possible that this yield reduction would continue into the second production year, although we did not measure those yields. Given the relatively low yields to be gained by taking a second cut in the seeding year relative to the potential loss of yield in the following year(s), it seems wise to refrain from taking a 2^nd cut in the seeding year unless a serious feed shortage dictates that the crop be harvested a second time.

In the event that taking a second cut in the seeding year was necessary, one could delay harvest until after the critical period to ensure maximum root reserves in the new alfalfa plants. This would still be risky in windswept areas were little snow accumulates.

Conclusions:

1. In the seeding year, cereal companion crops usually yield more forage than direct seeded alfalfa.

2. In the first production year, alfalfa established by direct seeding will usually outyield alfalfa established via an oat companion crop.

3. Taking 2 cuts of forage in the seeding year will, in certain years, increase the risk of winter damage and result in lower forage yields in the first production year. Taking the 2^nd cut in the seeding year is not recommended for short season areas unless exceptional circumstances dictate otherwise.

Top of Page
Top of Report

Yield of Barley and Canola When Grown as a Companion Crop
Jim Johnston and Matt Bowman
New Liskeard Agricultural Research Station

Perennial forage crops such as alfalfa are normally established using a cereal companion crop such as oats or barley. The cereal companion crop can be harvested for grain and straw or cut green and ensiled. Red clover is also underseeded with crops such as wheat for use as a ploughdown to improve soil structure. Canola has not traditionally been used as a companion crop for forage establishment, but there has been interest in having a red clover ploughdown crop grown in association with canola.

To ensure good forage establishment, a companion crop should allow good light penetration of the canopy and be mature by mid-summer so that water and nutrients are available to the forage plants. Using these criteria, barley cut green for silage is considered to be a better companion crop than barley harvested as grain. Canola is not considered a good companion crop since it tends to be very competitive, allows little light to penetrate to the ground, and with Argentine varieties, is a long-season crop. To make it more likely that forage would establish under a canola crop, one could use a shorter season type (ie: Polish) and/or use wider row spacing to allow more light penetration through the crop canopy. This article discusses the effect of different companion crop options on companion crop yield. Subsequent forage yields are discussed in a separate article.

Methods:
The trial was conducted at New Liskeard in 1999 and 2000. Barley was sown for silage and for grain (Table 1). Both Argentine and Polish canola were grown and each was sown in 17.5 and 35 cm row spacings. Each of the companion crop treatments was underseeded to alfalfa-timothy and red clover-timothy mixtures. The forage mixtures were also direct seeded (no companion crop). The barley silage plots were cut about 10 days after head emergence. The grain crops were all direct combined at grain maturity.

Top of Page
Top of Report

Results:
Direct seeded forages were not harvested in 1999 due to slow establishment and low yield (Table 2). The alfalfa mixture suffered from potato leafhopper feeding and was quite stunted. In 2000, both legume mixtures produced excellent yields. Barley silage produced the highest overall yield in 1999 and 2000. For producers who can use cereal silage, this appears to be a good option to establish forages without sacrificing yield in the seeding year.

Among the grain crops, barley outyielded canola in both years. Within the canola treatments, very different results occurred in 1999 as compared to 2000. In 1999, canola in 17.5 cm rows outyielded canola in 35 cm rows, but within each row spacing, there was no yield difference between Argentine and Polish types. This was surprising since it is generally accepted that Polish canola will yield only ½ to 3/4 that of Argentine. In 2000, Argentine outyielded Polish canola, but there was no significant difference between 17.5 and 35 cm row spacings.

The type of legume underseeded (alfalfa vs. red clover) had no effect on seeding year yields of forage or grain in either test year. No interactions between companion crop and underseeded legume occurred in either year.

Top of Page
Top of Report

Discussion:
These results support previous work that New Liskeard that showed a high yield potential from cereal silage. This is a good option for producers who want successful forage establishment but also need forage produced in the seeding year. Direct seeded forage was very good in 2000, but did not produce a harvestable crop in 1999. The 1999 crop could have been grazed lightly, but this is often not an option due to lack of fencing or water supply. It should be noted from observations at New Liskeard that under good management, direct seeded legumes should produce a yield of 3 to 5 t/ha at least 3 years in 5.

Barley grain yields were normal for this area. Canola yields in general were average to above average. The difference in yield across canola treatments in 1999 and 2000 is difficult to explain. The 2000 result would be expected: i.e.: lower yields from Polish types and little effect of row spacing. Canola has the ability to branch readily if the stand is thin (as with wide rows) and therefore the effect of row spacing should be minimal. The 1999 results may be related to particular conditions that reduced branching in the wide-row treatments. Additionally, the Argentine canola yields in 1999 may have been reduced by seed pod shattering, which can be a concern with direct harvested canola.

Conclusions:

1. If forage production in the seeding year is required, cereal silage is the best option, although direct seeded legumes can also produce high yields under good conditions.

2. Growing Polish canola as a companion crop is likely to result in reduced canola yields as compared to Argentine types.

3. Growing canola in wide rows as a companion crop may result in lower canola yields, but this effect is variable.

4. Regardless of what companion crop is grown, grain yield is not affected by the legume that is underseeded (alfalfa or red clover).

Top of Page
Top of Report

Return to the Main Page