Basic HTML Version
Methodology – Limitations, Constraints and Biases
17
Limitations, Constraints and Biases
The survey protocol has been designed to reduce bias in sampling, and to obtain objective information on the weed
flora that remains after control practices have been used by the farmer. In the development of the protocol, several
constraints were placed on the eligibility of areas and fields for inclusion in the survey. The survey covers the main
irrigated area of the province. Private irrigation and reserve land was omitted. Only the major irrigated crops were
considered. Other crops may have different weed floras. Fields were limited to those with an area greater than 16
ha. Fields were only surveyed if producers had used the field for a minimum of five years, if they were accessible
by road, and if the producers were willing to cooperate. These constraints will have altered the spectrum of fields
slightly from a completely random sample.
Occasionally, when the surveyor went to the field the crop did not match the qualified crop for the site. If the farmer
was not available to resolve the discrepancy with the surveyor, the field was generally surveyed. Fields that were
not planted to the selected crop types were excluded from the summaries. The exclusion of fields may have slightly
altered the weed spectrum, as areas with excluded fields are underrepresented (Table 3). In four cases, the field was
planted to crop types being surveyed. These included two cereals crops (triticale and oats) and two annual broad-
leaved crops (field peas and flax). These fields were included in overall summaries.
Only a small portion of each field was surveyed. This portion intentionally did not include sloughs, field edges,
shelterbelts, etc. It also did not extend into the less accessible areas of the field. This may limit the representation of
some species, such as foxtail barley, that are found more commonly near saline sloughs, or brome grass that is often
found near field margins. The small area in the field was sufficient to illustrate the distribution of major species and
minor species that might be important to agriculture. This level of sampling is not sufficient to give an exhaustive
list of the flora, or to illustrate the distribution of rare species.
The identification of some weed species is difficult in the field. Surveyors were trained, and asked to send in
unknowns for identification. However, mistakes may have been made. In some cases, the common names of
species differ in different areas. For instance, this might have caused some confusion between lamb
=
s-quarters and
redroot pigweed. The distinction between spiny annual sow-thistle, annual sow-thistle and perennial sow-thistle
may have been difficult for some surveyors as the annual sow-thistles are often not distinguished from each other.
Generally, the identification and counts by the field surveyors were used without alteration.
The weed survey recorded the numbers of individuals of each species. It does not show the vigour, the biomass or
the competitive ability of the plants found in the field. The survey shows what was there, not its effect on the crop.
In the survey analysis, all species are considered separately. The complex interrelationships among species have not
been considered. These will be the subject of later analyses. Relationships between species and farm management
practices will be considered in a further publication in this series. The relationships between species and field
variables (for instance, distribution patterns in the field) are not considered.
The results in this report provide a snapshot of the size and extent of weed populations in agricultural ecoregions of
Alberta in 2009. In general, the precipitation was low in the winter of 2008/2009
growing season was very dry
until July and the temperatures tended to be average to low until September
. Differences in snow cover,
precipitation and temperature may favour some weeds over others, or may result in higher or lower weed numbers in
some areas than usual for the period as a whole. Consequently, only dramatic or consistent trends in the weed data
are considered as important. Minor fluctuations may result from simple year-to-year variation.
14
National Agroclimate Information Service (NAIS).
2009. Precipitation Compared to Historical Distribution
(Prairie Region) - Static Scale - November 1, 2008 to March 31, 2009 [Online] Available:
http://www.agr.gc.ca/pfra/drought/prpwi09_e.html [11 January 2010]
15
National Agroclimate Information Service (NAIS).
2009. Monthly Mean Temperature Difference from
Normal (Prairie Region) [Online] Available: http://www.agr.gc.ca/pfra/drought/ArcPrTemp2009_e.html[11 January
2010]