Michelle du Croix Sissons and Dr. Rene Van Acker. Department of Plant Science. University of Manitoba. Winnipeg, Manitoba.

Knowing the depth of weed seedling recruitment in the field can help determine the effects of management on seed movement. Most research in the past has concentrated on the weed seed bank and seed movement in the soil, and not on actual depth of seedling recruitment. Seed bank measurements are not representative of the seedling population because of internal and external factors acting on seeds. Derksen et al. (1998) noted that the best way to predict which weeds will be a problem is to identify past years weed problems. Measuring the depth of actual weed recruitment will give a better overall picture of where weeds are coming from and will aid in the development of improved weed control.

1) To determine from what depths wild oats (Avena fatua), green foxtail (Setaria viridis), volunteer wheat (Triticum aestivum), barnyard grass (Echinochloa crusgalli) and wild buckwheat (Polygonum convolvulus) seedlings are recruited in conventional tillage compared to zero tillage fields.

2) To find a morphological marker on dicot weed species that can be used to determine depth of recruitment.

3) To determine how different farm management practices (tillage depth, speed and implement type, seeder openers and depth of seeding and fertiliser placement methods) effect recruitment depth of the above mentioned weed species.

4) To determine how different depths of planting in zero and conventional tillage soil cores affect redroot pigweed (Amaranthus retroflexus), green foxtail and wild buckwheat recruitment.

5) To document temperature changes in zero and conventional tillage fields over the growing season.

All weeds surveyed germinated shallower in zero tillage than conventional tillage, at both sampling dates (pre-seeding and pre-post emergent herbicide application). The difference in recruitment depth between the two tillage systems was greater at the pre-spray sampling period than the pre-seeding sampling period. This result was consistent over both years.

Pre-seeding weed recruitment was concentrated closer to the soil surface in both tillage systems than recruitment depth in the pre-spray sampling period.

Weeds in zero-tillage systems germinated closer to the surface than in conventional tillage. Weed recruitment was also concentrated in a narrower band close to the soil surface. In conventional tillage fields, weed recruitment tended to be distributed over a broader range of depths, and recruitment for wild oats at the pre- post emergent herbicide application could occur from as deep as 12 to 15 cm. The overall mean recruitment depth for all weeds was not greater than five centimetres in both tillage systems.

In zero tillage fields 80-90% of the weeds tended to be concentrated in the top three centimetres of the soil. In conventional tillage fields 80-90% of the weeds were recruited from up to seven centimetres deep.

Seedbank longevity did not correlate with weed seed recruitment depth. Wild oats had a shorter-lived seed bank than green foxtail, but a wider range of recruitment in the soil profile in both zero tillage and conventional tillage. The larger seed size of wild oats may allow successful recruitment from deeper in the soil than other smaller seeded weeds. Volunteer wheat has a similar seed size to wild oats, but the recruitment depth has a very narrow range. In zero-tillage, most of the recruitment for volunteer wheat is concentrated at the soil surface. Wheat has low dormancy levels and a very short-lived seed bank compared to wild oats. These factors indicate why volunteer wheat recruitment is so close to the surface. Species with longer-lived seeds such as green foxtail and wild buckwheat were expected to recruit from a wider range of depths in zero tillage than was observed. The small seed size of these two weed species restricted where in the seed bank they were successfully recruited. Even if seeds are present deeper in the soil they will not emerge if they germinate.

In Manitoba, many zero-tillage farmers are broadcasting soil-applied herbicides on the soil surface prior to snowfall to control weeds. Good weed control levels are being achieved using soil-applied herbicide in this manner. Results from this project indicate why this practice has been successful in controlling weeds. With improvements in application technology, it is possible to apply herbicides at the depth of weed germination for improved weed control.

A stable morphological marker was found for wild buckwheat. A small notch, curve or darkening of the root was identified which indicated the point of germination. This marker was found by placing seeds at known planting depths and then excavating the seedlings at different growth stages. The marker from plant in the green house proved to be consistent with plants removed from the field.

The effects of tillage on seedling recruitment are extensive and not fully understood. Past research has given some insight into weed distribution in the seedbank. Unfortunately, the weed seedbank is not representative of the seedling population. This project has focused on determining the depth of recruitment of seedlings, therefore, allowing a better understanding of the effects of tillage on weed seedling populations. A total of 88 fields were surveyed at 12 sites (Portage la Prairie, Carmen, Homewood, Mather, Crystal City, Minnedosa, Brookdale, Arden Ridge, Brandon, Souris, Hartney and Altona) in Manitoba. The expansive area and the large number of fields sampled over the past two summers ensure the results from this project are representative and inferable over a wide area of agricultural production.

Dr. Rene Van Acker

Dr. Doug Derksen

Dr. Gord Thomas

Gary Turnbull

Rufus Oree

Thank-you to all of the farmers who allowed me to sample on their fields and took time to answer my questions.

Agriculture and Agri-Food Canada

NSERC

Dow AgroSciences Canada Inc.