MANAGEMENT EFFECTS ON LONG-TERM WEED POPULATIONS

A. Gordon Thomas

Agriculture and Agri-Food Canada, Saskatoon Research Centre

107 Science Place, Saskatoon, SK S7N 0X2

Introduction

In 1995 Agriculture and Agri-Food Canada and Saskatchewan Agriculture and Food initiated a collaborative project to update the weed survey information for Saskatchewan and collect additional information on crop and weed management practices. The project produced two sets of data. The first data set contained the results from a field survey of weed occurrence and density in major cereal, oilseed and pulse crops in the province. The second data set contained the detailed management history of the surveyed fields from a questionnaire survey. Information in these two data sets can be used to examine the effects of management on weed populations.

Field Survey for Weeds

The objective of the 1995 provincial weed survey was to provide a "snapshot" of the size and extent of weed populations in the province. The survey was conducted on 1178 randomly selected fields in Saskatchewan. The type of tillage operation used on the field was not considered in the random site selection process. The survey included 10 crops: spring wheat (407 fields), durum wheat (164), barley (161), oat (47), canary seed (11), canola (216), mustard (25), flax (53), field pea (60), lentil (34).

Fields were surveyed from mid-July to mid-August. All weeds were counted in 20 quadrats in a field. Each quadrat was a square area 50 cm by 50 cm in size. Quadrats were arranged in a "W" pattern 100 paces of the edge of the field. By surveying fields in late summer, the weeds that had not been controlled were counted. Weeds found in a field in late summer were, in part, a result of the agronomic management decisions made by the producer at various times during the crop year. The impact of these management decisions on weed populations was recorded in the late summer survey. Weed counts at this time of the crop year indicated the size and extent of troublesome weed populations.

A relative abundance index is used to summarize the results of the weed counts in the surveyed fields. The index considers the frequency of occurrence of weeds in fields, the frequency of occurrence of weeds in the 20 quadrats per field and the density of the weeds in the field. A high value of the index shows that a weed occurred frequently and at high densities.

Questionnaire Survey of Management Practices

A detailed questionnaire that requested information on the management of the surveyed field was mailed to each of the 1178 producers. The questionnaire was designed to help in the evaluation of the field survey of weeds carried out in late summer. It included sections on cropping, tillage and fertilizer practices, weed control, and insect pests. The questionnaire consisted of Yes/No, multiple choice and short answer questions.

Completed questionnaires were returned by 750 producers. The assistance of the producers who took the time to complete the questionnaire is greatly appreciated. Their answers provided detailed information for 315 management variables. Not all producers answered every question. Therefore, each of the summary tables includes a line showing the number of questionnaire responses. This report will include only a few of the variables that relate to the type of tillage operation used on surveyed fields. Complete information from the questionnaire will be available soon in a separate report.

Major Ecoregions in Saskatchewan

The arable portion of the province is divided into four large areas called ecoregions. These areas have similar climate, natural vegetation and soils. The Boreal Transition ecoregion is the most northern area of arable agriculture. This ecoregion has the coldest temperatures and highest annual precipitation of the four major ecoregions. Agriculture is found mainly on the Dark Gray Chernozemic soils. South of the Boreal Transition ecoregion is the Aspen Parkland. It is warmer and drier than the Boreal Transition ecoregion and the soils are typically Black Chernozems. South of the Aspen Parkland is the Moist Mixed Grassland ecoregion. The climate is considered semi-arid with warmer temperatures and less annual precipitation than the two previous ecoregions. Soils are Dark Brown Chernozems. The Mixed Grassland ecoregion is found in the southwestern portion of the province. It is the driest and warmest part of the province with Brown Chernozemic soils.

The distribution of individual weed species is correlated with these four major ecoregions. Therefore, some detailed information from the field survey and the management questionnaire is summarized by ecoregion.

Long-term Trends in Weed Populations

The 1995 Saskatchewan field survey for weeds documented the changes in the distribution and abundance of weeds that have occurred since the first surveys in 1976-79 and more recently in 1986. All surveys have used the same methodology. The relative abundance index is used to summarize the changes that have occurred in weed groups during the past 19 years.

Changes have occurred in the relative abundance index for some weed groups during the past 19 years (Table 1). The index for annual grassy weeds has remained almost constant. Green foxtail, wild oats, Persian darnel and barnyard grass are the common grassy weeds. The frequency of occurrence of these grasses in the province has not changed in any significant way. Overall the index for annual broad-leaved weeds has decreased but not all weeds have decreased. The frequency of occurrence for Russian thistle has decreased from 42% of the fields to 23% and cow cockle from 32% to 9%. On the other hand, the frequency of occurrence of shepherd's-purse has increased from 8% to 15% and cleavers from 2% to 19%. The index for perennial weeds had more than doubled since the 1986 survey. The frequency of occurrence of Canada thistle has increased from 19% to 46% of the fields and dandelion from 7% to 18%. The index for volunteer crops has increased from 1976-9 until 1995. For example, the frequency of occurrence for volunteer canola has increased from 5 to 15% of the fields.

Table 1. Changes in the relative abundance index of weed groups

Are these changes in the relative abundance index of weed groups related to changes in farm management practices? This is not an easy question to answer. It is difficult because of the multiplicity of management factors that affect weed populations and because the effects of individual factors are not easily isolated. A reduction in the frequency of fallow, an increase in the diversity of crops grown in rotations, a reduction in tillage, and an increase in direct seeding have occurred during the past 19 years. Accompanying these changes there have been changes in equipment, fertilizer placement and herbicide usage. Also, management practices are not uniform throughout the province because of different climatic constraints in various areas of the province. This geographic variability in management practices must also be considered in the analysis of weed populations. The information presented in the remainder of this report is an attempt to examine specifically the effect of changes in the type of tillage operations on weed populations.

Tillage Practices

Type of Tillage Operation

Producers were asked to indicate the type of tillage operation used on the field for the surveyed crop. They were given three choices: conventional tillage, direct seeding - low disturbance, direct seeding - high disturbance. Agronomic inputs and weed populations are summarized based on the type of tillage operation used on the surveyed field.

Direct seeding (both low and high disturbance) was used on 29% of all the fields reported in the questionnaire; however, the percentage varied from a low of 17% in the Boreal Transition ecoregion to a high of 37% in the Moist Mixed Grassland ecoregion (Table 2). Only 6 of the 86 fields in the Boreal Transition were managed using a low disturbance direct seeding operation. In contrast, 33 of the 190 fields in the Aspen Parkland ecoregion used low disturbance direct seeding.

Table 2. Types of tillage operations used on surveyed fields based on number of responses in management questionnaire.

The proportion of fields with low and high disturbance direct seeding changes from the south to the north (Table 2). In the Mixed Grassland ecoregion in the south, low and high disturbance fields are equal in number and in the Boreal Transition ecoregion low disturbance fields are less than half the number of high disturbance fields. The unequal number of fields with the three types of tillage operations makes it difficult to investigate associations between tillage and weed populations and between tillage and other management variables.

Amount of tillage

The amount of tillage was estimated from the number of implement passes used on the surveyed field in the fall of the preceding year and in the spring before seeding the surveyed crop. For example, if a tillage operation included both a cultivator and harrow, then two implement passes were recorded.

The number of implement passes used on surveyed field managed with conventional tillage operations varied from 0 to 13 with the median number of passes equal to 3 (Table 3). The median number of passes on fields with high and low disturbance direct seeding operations was 1 and 0 respectively. Forty-nine fields under conventional tillage management did not receive any tillage before seeding. Most of these fields were seeded with a discer. In other cases, seeding was a one pass operation with a cultivator pulled ahead of a seeding implement. Twenty fields under low disturbance direct seeding management received one or more tillage passes.

Table 3. Number of implement passes used to prepare field for seeding including tillage the previous fall.

Cropping Practices

Surveyed crop

Surveyed crops were grouped into cereal, oilseed and pulse and the percentage of each crop and crop type reported in the management questionnaire was compared among types of tillage operations.

Cereals were grown on an equal proportion of the fields managed with the three types of tillage operations (Table 4). Producers with a low disturbance direct seeding operation grew more pulse and less oilseed crops than producers with conventional tillage operations. Weed populations will be affected by the choice of crops because different herbicides are used.

Table 4. Percentage of the fields reported in the management questionnaire for crop types and specific crops.

Previous crop type or fallow

Crops grown on the surveyed fields the previous year were grouped into fallow, cereal, oilseed and pulse categories. The percentage of responses in each category was compared among the three types of tillage operations.

About half the fields with conventional tillage management were fallowed the previous year and 30% of the surveyed crops were grown on cereal stubble (Table 5). In contrast, about 15% of the fields with direct seeding management were fallowed the previous year. Most of the fields with direct seeding management were grown on cereal stubble but oilseed and pulse stubble was also used.

Table 5. Crop types grown previous to surveyed crop, expressed as a percentage of fields reported in the management questionnaire.

Crop rotation

Producers were asked to indicate the specific sequence of crops grown or occurrence of fallow on the surveyed field for the previous five years. The frequency of fallow in the crop rotation was determined using the data from the current surveyed crop and the previous five years. The rotation could vary from continuous cropping for six years to use of fallow every other year.

Table 6. Frequency of fallow in the crop rotation used on the surveyed field, expressed as a percentage of fields reported in the management questionnaire.

Nearly half the producers using a low disturbance direct seeding operation indicated that fields had been continuously cropped for 6 years and only 10% of the producers had used fallow in 3 of the 6 years (Table 6). The frequency of continuous cropping with high disturbance direct seeding operations was less than with low disturbance. Fallowing 3 of 6 years was the most common rotation used by producers with conventional tillage operations.

Method used to maintain fallow

If the field had been fallowed in the year before the survey, then producers were asked to indicate the method used to maintain the fallow. They were given three choices: tillage and herbicide, tillage only and herbicide only.

Conventional tillage producers relied mainly on tillage to maintain the summer fallow the year before the field survey (Table 7). Very few of them used herbicides as the only method. A combination of tillage and herbicide may be the main method used by producers with direct seeding operations but the number of responses is probably too small to be reliable.

Table 7. Method used to maintain summerfallow based on the number of responses in the management questionnaire.

Seeding implements

Information was requested on the implement used to seed the surveyed field. Besides five common seeding implements, producers could specify a particular implement. The "other" category included broadcast seeders and generic descriptions of seeding implements such as "air drill."

A good correlation was found between the seeding implement used to seed the surveyed field and the type of tillage operation used to manage the field (Table 8). Most direct seeding producers used a high or low disturbance air seeder; however, 33% of low disturbance direct seeding producers and 20% of high disturbance direct seeding producers used an implement other than an air seeder. Insufficient information was given by the producer to categorize the large number of "other" seeding implements used on fields managed with low disturbance direct seeding management. Only a few of the fields with conventional tillage management were seeded with a low disturbance air seeder. Other seeding implements were used with equal frequency by conventional tillage producers

Table 8. Seeding implement used based on number of responses in management questionnaire.

Fertilizer Practices

Use of fertilizer in fall and spring

Producers were asked to indicate if they applied any fertilizer in the fall, in the spring before seeding, or at the time of seeding the surveyed field. Responses were not separated by nutrient.

Table 9. Use and time of fertilizer applications based on number of responses in management questionnaire.

The use of fertilizer differed among the types of tillage operations (Table 9). Fertilizer was not used on 31%, 14% and 8% of the fields managed with conventional, high disturbance and low disturbance tillage operations. A single spring application of fertilizer was the most common time for all three types of tillage management.

Method of fertilizer application

A summary of methods of application was restricted to those questionnaires where a single application of fertilizer in spring was used since most the producers applied fertilizer at this time.

Most producers applied their fertilizer with the seed to the surveyed field (Table 10). Banding was the second most frequently used method for all three types of tillage management. Banding used more frequently by direct seeding producers than by conventional tillage producers.

Table 10. Methods used to apply single applications of fertilizer in the spring based on the number of responses in management questionnaire.

Herbicide Practices

Time of herbicide application

Information was requested on herbicides applied during the fall of the previous year (pre-harvest and post-harvest), during the spring of the survey year (before seeding and crop emergence) and during the summer on the surveyed crop (in-crop).

The use of herbicides differed among the types of tillage operations (Table 11). Eleven percent of the fields managed with a conventional tillage operation did not receive a herbicide treatment whereas less than 5% of the direct seeded fields were not treated. Herbicides were applied more frequently to the low disturbance direct seeding fields. Herbicides were applied at three times of the year to 8% of the fields and twice to 57% of the fields. In contrast herbicides were applied as an in-crop treatment to 55% of the conventional tillage fields.

Table 11. Time of herbicide applications based on number of responses in management questionnaire.

Herbicide groups

Producers indicated on the questionnaire the specific products they had used during the fall, spring and summer. Individual products were classified into herbicide groups. Herbicide groups are based on modes of action as described in "Pesticide Resistance Management Labeling", a regulatory proposal document (Pro96-03) published by the Submission and Information Division, Pest Management Regulatory Agency, Health Canada, Ottawa. Eight groups of herbicides are used on annual crops in Saskatchewan.

Direct seeding operations have resulted in changes in herbicide use patterns. Group 3, 4 and 9 usage differed among the types of tillage operations (Table 12). Group 3 usage (trifluralin and similar soil incorporated products) was least on low disturbance direct seeded fields. Fields with direct seeding tillage operations had 30% greater usage of group 4 herbicides, such as 2,4-D, MCPA and dicamba, than did conventional tillage operations. Only 5% of the conventional tillage field area was treated with group 9 (glyphosate) compared with 24% of the high disturbance operations and 66% of the low disturbance operations. A comparison of the total area treated with all the herbicide groups indicates that direct seeding operations relied more heavily on herbicides for weed control than conventional tillage operations. Fields with low disturbance management were treated most frequently with herbicides.

Table 12. Herbicide group use based on percentage of area reported in management questionnaire.

Weed Populations

Common weeds

The relative abundance index is used to compare weed communities among the three types of tillage operations.

Similar weed species are found with the three types of tillage operations and species are ranked in a very similar order based on relative abundance; however, some trends with tillage operations are evident (Table 13). For example, the index for green foxtail, Russian thistle, volunteer canola, dandelion, thyme-leaved spurge and narrow-leaved hawk's-beard increases when tillage is reduced with direct seeding. Wild oats and cow cockle have the largest index in conventional tillage operations. The mean weed density is lowest in low disturbance direct seeding.

Weed types

Weeds were grouped into four weed types: annual grasses, annual broad-leaved, perennial and volunteer crops. The relative abundance index of individual weeds was summed for each group and the summed values compared among the three types of tillage in each of the four ecoregions.

Table 13. Relative abundance index for 30 commonly occurring weeds found in surveyed fields with completed management questionnaires.

Annual grassy and annual broad-leaved weeds account for most of the weed abundance in the Mixed Grassland and Moist Mixed Grassland ecoregions (Table 14). In the Aspen Parkland ecoregion the relative abundance index for grassy weeds is less and perennial weeds are more abundant than in the two ecoregions to the south. Perennial weeds have a higher relative abundance index than annual grassy weeds in the Boreal Transition ecoregion.

The total relative abundance index for the annual grassy weeds is highest for low disturbance direct seeding in the Mixed Grassland ecoregion but lowest in the other three ecoregions. Perennial and volunteer crop weed groups have higher index values in direct seeding tillage operations than in conventional tillage.

Table 14. Relative abundance index for four weed types in the surveyed ecoregions.

Mean weed density

Density is the mean for all weeds occurring in the fields and is given for the three types of tillage operations in each of the four ecoregions.

Weed densities increase from a low in the Mixed Grassland ecoregion in the south to a high in the Boreal Transition ecoregion in the north (Table 15). Within an ecoregion, weed densities are usually lowest in fields with low disturbance direct seeding management. The high mean density in high disturbance direct seeding fields in the Moist Mixed Grassland ecosystem is due to high densities of barnyard grass and thyme-leaved spurge in a few fields.

Table 15. Mean density of all weeds expressed as number of plants per m².

Association of weeds with direct seeding

Associations of weeds with types of tillage operations were investigated using a multivariate statistical method called redundancy analysis. Weed density data were used in the analysis. The significance of the association between the weed community and type of tillage was tested with a Monte Carlo resampling procedure. Separate analyses were done for each of the four ecoregions. The association between weeds and tillage in the Boreal Transition ecoregion was not significant. This ecoregion had fewer direct seeded fields than the three ecoregions to the south.

Four species were associated with conventional tillage. Wild oats were associated with conventional tillage in the Moist Mixed Grassland and Aspen Parkland ecoregions, wild buckwheat in the Mixed Grassland ecoregion, redroot pigweed in the Moist Mixed Grassland ecoregion and cow cockle in the Mixed and Moist Mixed Grassland ecoregions.

Table 16. Associations of weeds with high or low disturbance direct seeding management in ecoregions.

Twenty-one species representing each of the four weed types were associated with direct seeding operations (Table 16). Volunteer wheat was the only species to be consistently associated with low disturbance seeding in all three ecoregions. Flixweed and barnyard grass were associated with low disturbance in one ecoregion and with high disturbance in another ecoregion. More weeds were associated with low disturbance than with high disturbance direct seeding.

The 21 species in Table 15 have few characteristics in common. Species included three annual grasses, eight annual broad-leaved weeds, six perennials, and four volunteer crops. Two of the perennials, Canada thistle and quack grass, reproduce vegetatively from underground parts. Four other perennials, dandelion, foxtail barley, biennial wormwood and vetch are biennial or short-lived perennials that reproduce from seed. Three of the annual broad-leaved weeds, flixweed, narrow-leaved hawk's-beard and Canada fleabane, are winter annuals and two species, kochia and Russian thistle are tumbleweeds.

These species varied in rank from 1st to 57th in the survey based on relative abundance. Seven species were ranked less than 30th and do not appear in the list of commonly occurring weeds (Table 13).

Associations of some species with specific tillage operations in this survey agree with the results from experimental plot work and weed surveys in other areas.

Troublesome weeds

Producers were asked to list their five most troublesome weeds on the surveyed field. Only weeds listed by 5% or more of the producers are included in the summary for each type of tillage operation.

Table 17. Weeds listed by growers as troublesome in the management questionnaire.

Wild oats was considered the most troublesome weed by producers in all three types of tillage (Table 17). Several weeds, such as wild buckwheat, green foxtail, stinkweed and others, are considered equally troublesome by all producers. Flixweed, dandelion and narrow-leaved hawk's-beard were troublesome to low disturbance producers. In contrast, cow cockle, cleavers, shepherd's-purse and hemp-nettle were more troublesome to conventional and high disturbance producers than low disturbance producers. High disturbance producers listed Canada thistle and quack grass more frequently than producers in the other two tillage types. Volunteer canola was a problem for direct seeding operations.

Less than 5% of the low disturbance producers listed Persian darnel, foxtail barley, Canada fleabane, round-leaved mallow, thyme-leaved spurge and volunteer wheat as troublesome although these weeds were associated with a low disturbance tillage operation (Table 16). None of the producers in any tillage system recognized biennial wormwood as troublesome.

Summary

Questionnaire responses from producers have shown the significant changes that have occurred in management inputs with a shift from conventional to direct seeding and, in particular, to low disturbance operations. These changes included a reduction in tillage, extended rotations with more crop diversity, lower frequency of fallow, greater usage of air seeders, more frequent usage of fertilizer and herbicides, greater usage of group 4 and 9 herbicides and less usage of group 3. Currently, changes in management inputs associated with direct seeding have not had a major impact on weed populations. Similar weed communities are found in the fields of all producers. Changes in the relative abundance of weed groups were not necessarily associated with the use of direct seeding management. A few species are associated with direct seeding operations but many of them occur infrequently in the weed survey. The abundance of these species may be indicators of changes that are occurring and their abundance should be closely monitored

Acknowledgments

This research was funded in part by the Canada-Saskatchewan Agriculture Green Plan Agreement and by Herbicide Manufacturers and Distributors in Saskatchewan. The enthusiastic support of this project by Len Juras, the weed control specialist with the Sustainable Production Branch of Saskatchewan and Food, is greatly appreciated. It is with special pleasure that I acknowledge the contribution made by the 1178 producers who agreed to cooperate in the project and the 750 producers who took the time to fill out and return the questionnaire.