NOVEL AND CONVENTIONAL CONTROL OF LINARIA GENISTIFOLIA SSP. DALMATICA IN
A SEMI-ARID GRASSLAND OF BRITISH COLUMBIA’S SOUTHERN INTERIOR

By

JACOB LAWRENCE BRADSHAW

A THESIS SUBMITTED IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
BACHELOR OF NATURAL RESOURCE SCIENCE (HONS.)
IN THE DEPARTMENT
OF
NATURAL RESOURCE SCIENCES
AT
THOMPSON RIVERS UNIVERSITY

THESIS EXAMINING COMMITTEE
Wendy Gardner (Ph.D.), Thesis Supervisor and Associate Professor,
Dept. of Natural Resource Sciences
Peggy Jo Broad (B.Sc.F.), Lecturer, Dept. of Natural Resource Sciences
Catherine Tarasoff (Ph.D.), Adjunct Faculty, Dept. of Natural Resource Sciences

Dated the 19th day of April 2017, in Kamloops, British Columbia, Canada

Abstract
Thesis Supervisor: Associate Professor Wendy Gardner (Ph.D.)
Dalmatian toadflax (Linaria genistifolia ssp. Dalmatica) is a major concern in disturbed
areas of British Columbia’s interior due to its pronounced ability to displace native
vegetation. To investigate the most appropriate control method, including efficiency, a
study was established in Kenna Cartwright Nature Park, Kamloops, BC in 2016. Manual
removal and three herbicide treatments were compared by collecting live and dead stem
count in 6m by 6m replicates (n=30) during baseline, post-treatment and end-of-growing
season assessments. Tordon 22K was applied at 213 g a.e. ha-1 picloram mixed with Agral
90 surfactant (0.025% by volume) using broadcast spraying, spot spraying and hand
wicking (August 23rd). Treatment had a significant effect; broadcast and spot spraying
provided the highest overall success resulting in mean stem counts of 1.7 ± 1.07 and 6.33 ±
2.67 at the end of the growing season, respectively. Wicking offered the same overall
success as spot spraying in reducing total live stems. However, wicking contained 61% of
the live stems at the end of the growing season, relative to baseline, whereas spot spraying
only contained 31%. Broadcast spraying offered the greatest success in which survivorship
was 13%. Only broadcast and spot spraying provide statistically significant reduction in
vegetative stems from baseline to end-of-season; broadcast spraying contained zero stems
in this life stage. Despite eliminating all stems at time of treatment, manual removal
experienced high regrowth which resulted in statistical similarity to the control and
appeared to encourage vegetative regrowth. Nearly all of the regrowth experienced in the
control, manual removal, spot spraying and wicking at the end of the growing season were
rosettes. Seedlings did not greatly contribute to reproduction in any treatments and were
absent from broadcast spraying. For high density sites (>31 stems/m2), broadcast spraying
offers the greatest chance of success based on both success and efficiency. Low to medium
density sites can be controlled by spot spraying, however the lower residual soil activity
must be considered. Ultimately, the effect on the native plant community must
be evaluated when selecting one of the recommended treatment methods and future
research should focus on tracking the communities through time.

ii

Table of Contents
Abstract.............................................................................................................................................................. ii
Table of Contents .......................................................................................................................................... iii
List of Tables ................................................................................................................................................... iv
Table of Figures .............................................................................................................................................. v
Acknowledgements ...................................................................................................................................... vi
Dedication .......................................................................................................................................................vii
Introduction ..................................................................................................................................................... 1
Economic and Environmental Impacts of Invasive Species ...................................................... 1
Dalmatian Toadflax Life History and Response to Disturbance .............................................. 3
Knowledge Gaps ......................................................................................................................................... 6
Objectives ..................................................................................................................................................... 7
Methods ............................................................................................................................................................. 7
Site Description .......................................................................................................................................... 7
Baseline Sampling .................................................................................................................................. 12
Experimental Design ............................................................................................................................. 13
Treatments ........................................................................................................................................... 13
Plot Description and Design........................................................................................................... 15
Field Data Collection.............................................................................................................................. 16
Statistical Analysis ................................................................................................................................. 18
Results ............................................................................................................................................................. 19
Treatment Success ................................................................................................................................. 19
Treatment Efficiency ............................................................................................................................. 26
Discussion ...................................................................................................................................................... 28
Evaluating Treatment Success and Efficiency ............................................................................. 28
Management Implications ................................................................................................................... 32
Future Research ...................................................................................................................................... 33
Literature Cited............................................................................................................................................ 34
Appendix A: Plant Community Summary .......................................................................................... 39
Appendix B: Plot Data ............................................................................................................................... 40

iii

List of Tables
Table 1. Historic Kenna Cartwright Restoration Plan Dalmatian toadflax inventory and
management recommendations. From Tarasoff (2002). ............................................................ 11
Table 2. Dalmatian toadflax stem density categories per m2, based on Dodge et al.
(2008). ............................................................................................................................................................. 12
Table 3. Treatment methods and type, chemical application rate and active ingredient
concentration. .............................................................................................................................................. 14
Table 4. Classification of Dalmatian toadflax life stage according to field assessment
category. ......................................................................................................................................................... 17
Table 5. Analysis of variance within life stage between treatments for end-of-season
Dalmatian toadflax mean live stem count. Asterisk identifies where Welch robust tests
of equality of means were performed. ................................................................................................ 20
Table 6. One-sample t-test summary used for post hoc comparison between broadcast
spraying all other treatments. Test statistic set to zero (equal to broadcast standard
deviation). ...................................................................................................................................................... 20
Table 7. Analysis of variance within life stage between treatments on the difference
between baseline and end-of-season Dalmatian toadflax mean live stem count. Asterisk
identifies where Welch robust test of equality of means was performed instead of oneway ANOVA. .................................................................................................................................................. 21
Table 8. Average total solution volume, herbicide content, active ingredient
concentration and treatment duration applied to replicates. ................................................... 26
Table 9. Herbicide application method efficiency summary highlighting herbicide
solution applied to each replicate and the baseline, post-treatment and end of growing
season mean total-live stem count (m-2). .......................................................................................... 27

iv

Table of Figures
Figure 1. Study location in Kenna Cartwright Nature Park, Kamloops, British Columbia. 8
Figure 2. Representative photograph of Dalmatian toadflax invasion within study area
in Kenna Cartwright Nature Park, Kamloops, B.C. ......................................................................... 13
Figure 3. Herbicide application methods of broadcast and spot spraying via backpack
sprayer (a) and hand wick (b). .............................................................................................................. 14
Figure 4. Study plot locations with treatment method and gross study area. .................... 15
Figure 5. Plot design with sampling stratification method. Six replicates of each
treatment (N = 30)...................................................................................................................................... 16
Figure 6. Mean stem count (m-2) of total-live, vegetative, mature and dead life stages at
baseline, post-treatment and end of growing season. Letters denote similarity within
each life stage within each sampling period. Asterisks identify where the end-of-season
live stem count is statistically different from its baseline. ......................................................... 23
Figure 7. Seedling and rosette mean stem count (m-2) at baseline, post-treatment and
end of season assessments. Letters denote statistical similarity at the end of the
growing season. Error bars are 95% C.I. ........................................................................................... 25

v

Acknowledgements
I would like to sincerely thank my supervisor Dr. Wendy Gardner for the countless
discussions, ability to help me pull together my thoughts in frustrating times and steadfast supervision. I especially appreciated your genuine excitement to see my thesis
come to fruition in the final weeks. To my committee members Peggy Jo Broad and Dr.
Catherine Tarasoff, thank for your valuable guidance and support. The diverse range of
approaches to both the experiment and what avenues to pursue provided me with a
broad range of possibilities to explore. I would also like to thank Dr. Shane Rollans and
Jacqueline Sorensen for their advice on statistical analysis and encouragement to keep
searching for a better method. Without you I would still be stuck in the statistics vortex.
A great deal of gratitude goes to the City of Kamloops for funding this research and
trusting this investigation with me in such a valuable nature park. A special thanks to
Kirsten Wourms for facilitating this opportunity and having such encouraging
excitement for the project. Thanks to Adam Fehr for helping me implement the
treatments and providing valuable logistical advice. I would also like to thank Andrea
Seager and Chantelle Gervan for helping me establish the experiment and collect data.

vi

Dedication
This thesis is in memory of my late brother and dedicated to my mother, father and
brother for supporting my move across the country to pursue this degree and for
encouraging me to chase all opportunities, no matter where they take me.

vii

Introduction
Invasive alien species play a significant role in the fate of native ecosystems, specifically
whether our environment will continue to prosper and retain the same resilience that
has encouraged diverse landscapes across Canada (Vitousek et al. 1997; EC 2004;
Winstral and Marks 2014). Globally, they are considered one of the most significant
threats to biodiversity by the World Conservation Union, second only to habitat loss,
and are a critical priority for the government of Canada (EC 2004; GOC 2016). Invasive
species are those that have become established outside of their natural past or present
distribution, largely through intentional or accidental human introduction, and threaten
biological diversity (Vitousek et al. 1997; CBD 2008; Pyšek and Richardson 2010;
IUCN). Although typically only 5% to 20% of introduced species spread and impact the
destination ecosystem in this manner, the negative outcomes can be extensive due to
their pronounced ability to out-compete native organisms (CBD 2008; Pyšek and
Richardson 2010; IUCN 2017). Additionally, the sheer number of introductions has a
cumulative impact on native biodiversity, as has been observed during the 76%
increase in invasive species in Europe since 1970 (Butchart et al. 2010). This large
increase is common across the globe; in British Columbia there are currently 175
invasive plants (ERBC 2015). The influence of these species transcends biological
kingdoms. For example, about 400 of 958 threatened or endangered species are at risk
due to nonindigenous species in the United States (Wilcove et al. 1998 as cited in
Pimentel et al. 2005)
Economic and Environmental Impacts of Invasive Species
The total annual economic burden of invasive species in Canada is nearly $30 billion,
$2.2 billion of which is directly related to invasive plants (CFIA 2014). The high cost is
due to a combination of chemical and mechanical control, loss of crop productivity and
increased need for specialized equipment and personnel (EC 2004; GOC 2016). In 2008,
total economic burden for British Columbia was $65 million and without continued
intervention, invasive species are projected to cost B.C. $139 million a year by 2020
(IPCBC 2008). Today these species continue to place a significant financial burden on

1

landowners and tax payers throughout the province, costing the agricultural industry
alone $50 million each year (ISCBC 2017). However, more concerning is the threat to
society far beyond the direct revenue and mitigation costs. Alien species have the
potential to reduce water quality, eliminate traditional food and medicinal plants,
degrade human health and reduce recreational opportunities (Vitousek et al. 1997;
Pejchar and Mooney 2009; Pyšek and Richardson 2010; ISCBC 2017).
With the potential to influence energy and nutrient cycling, invasive species interrupt
the traditional synergy of wildlands and place a burden on community structure that
has been consistently compounded by human development and environmental
manipulation (Pyšek and Richardson 2010). As a result of these negative impacts,
invasive species have been identified as a threat by local organizations such as the
Thompson Nicola Regional District and the City of Kamloops (City of Kamloops no date;
TNRD no date). Local efforts for controlling invasive plants are in-line with the four
stages through which the Government of Canada intends to respond to invasive alien
species, prevention of new invasions, early detection of new invaders, rapid response to
new invaders, and management of established and spreading invaders (EC 2004; Pyšek
and Richardson 2010). Although many institutions have played a role in these four
responses, historically the Canadian Food and Drug Agency has spear-headed
prevention and provincial bodies such as the B.C. Early Detection Rapid Response
coordinate efforts on the second and third stages (EC 2004; Pyšek and Richardson
2010; BCIMISWG 2014). On-the-ground invasive plant management has typically been
implemented by municipal government and private individuals (Pyšek and Richardson
2010). It is in the final response, management, that this study intends to provide a
greater understanding of the potential control methods for the invasive alien,
Dalmatian toadflax (Linaria genistifolia ssp. dalmatica; henceforth toadflax).

2

Dalmatian Toadflax Life History and Response to Disturbance
One of the most formidable species threatening our region’s natural areas is Dalmatian
toadflax, an invasive plant that is considered provincially noxious in British Columbia
(Scott 1999; Ralph et al. 2014; TNRD no date; City of Kamloops no date). This species is
native to the Mediterranean region of Dalmatian and was brought to North America as
an ornamental species in 1874 (Alex 1962 as cited in Ogden and Renz 2005). When in
dense stands, this plan reduces wildlife habitat quality and competitively excludes
native flora (Lajeunesse 1999 as cited in Kyser and DiTomaso 2013). Compounding this
issue is the self-incompatible nature of toadflax which is speculated to have lead to the
high level of genetic variability within the species (Docherty 1982 as cited in Kyser and
DiTomaso 2013). This plant is characterized by pale green waxy leaves clasping the
stem and conspicuous “snapdragon-like” yellow flowers (2.5 cm to 4 cm long) (Ralph et
al. 2014). Toadflax stands 1.2 m tall and produces a creeping perennial root system
(Scott 1999; Ralph et al. 2014). The species is now an aggressive invader, occurring
from sea level to 2800 m in B.C. (Scott 1999; Jacobs 2006; Ralph et al. 2014).
Understanding the biology of toadflax allows for the appropriate control method to be
utilized at the appropriate stage in the plant’s life cycle. This species has several
morphological adaptations that allow it to prosper in stressful environments. These are
expressed in the dynamic relationship between toadflax’s phenology, competing
vegetation and climate. Reproductive potential is perhaps the greatest of such
advantages. Dalmatian toadflax is a prolific seed producer with rapid spring
germination, allowing it to gain an advantage on native species (Scott 1999). This is, in
part, due to it’s ability to sprout from adventitious root buds as early as 14-21 days
after germination and to produce up to 500,000 seeds per plant on low competition
sites, which may remain viable for ten years (Robocker 1968; Robocker 1970; Scott
1999; Jacobs 2006). Dalmatian toadflax is able to out-compete neighbouring vegetation
under diverse disturbance regimes due to these life history attributes.
Toadflax is an effective pioneer following fire due to its extensive perennial sprouting
root system (Zouhar 2003; Jacobs 2006). Fire also has the potential to increase seed
3

production and is therefore not a recommended treatment option (Zouhar 2003; Jacobs
2006). There is a high potential for post fire colonization in semi-arid grasslands due to
toadflax’s proliferation on dry areas with low competition (Zouhar 2003). This has lead
to several cases where the presence of toadflax was not known before fire but there
was rapid population growth post-fire (Zouhar 2003). Due to these traits, early
detection is key in preventing toadflax from establishing in post-burn areas. This is the
least costly and most effective control measure, especially when toadflax cover is
greater than twenty percent (Goodwin and Sheley 2001; Clark 2003).
Biological control of Dalmatian toadflax has been underway in Canada and the United
States since the 1960s (Kyser and DiTomaso 2013). However, considering the wide
range of environmental conditions this plant tolerates, the seven biocontrol species that
have been released have each experienced a wide range of establishment and control
success (Kyser and DiTomaso 2013). The high genetic variability of the plant also
negatively impacts biological control success and has lead to a lack of long-term stand
reductions (Jamieson and Bowers 2010 as cited in Kyser and DiTomaso 2013; Kyser
and DiTomaso 2013). One of the most successful biological agents for Dalmatian
toadflax is the stem dwelling weevil, Mecinus janthinus, which can exert top-down
effects on the plant during high densities and attack rates when stem diameter is large
enough for it oviposit and pupate (Jamieson et al. 2012). Other common agents include
beetles, moths and weevils (Kyser and DiTomaso 2013).
Manual and mechanical removal of toadflax is most successful when the population is
small and easy to isolate (Scott 1999). Hand-pulling is a viable option in moist sandy
soils. This method is recommended in wilderness areas due to it’s low disturbance,
relative to cultivation, but could require ten years of repeated treatment to eradicate a
population (USDA 2012). To ensure maximum effectiveness, the entire root must be
removed as toadflax has the ability to reproduce from root fragments as small as 1 cm (
Bakshi and Coupland 1960 as cited in Ogden and Renz 2005; USDA 2012; Lajeunesse et
al. 1993, Jacobs and Sing 2006 as cited in Kyser and DiTomaso 2013; Kyser and
DiTomaso 2013; Wilson et al. 2005). Another alternative is cutting, which can be
4

effective in specific situations (Scott 1999; USDA Forest Service 2012). However, since
above-ground removal of biomass can stimulate adventitious root buds, cutting must be
timed appropriately considering the plant’s extensive carbohydrate stores (Scott 1999;
Sheley and Clark 1999; Zouhar 2003; Bakshi and Coupland 1960 as cited in Ogden and
Renz 2005; USDA 2012). The remaining mechanical control methods, mowing and
tillage, are not recommended for toadflax because of its ability to vigorous re-sprout
from its roots (Wilson et al. 2005).
Herbicide treatment is an important component of toadflax control, but requires high
rates of application and does not result in long-term control of the species when used
independently of other control measures (Scott 1999; USDA 2012; Kyser and DiTomaso
2013). An important consideration in chemical treatment is whether seeding will be
required; this is not necessarily required in areas of high native grass abundance and
cover (USDA 2012). In wilderness or natural areas, the use of backpack or hand-held
herbicide applicators is preferred, both due to lower potential for non-target mortality
and regulatory framework (USDA 2012). Picloram is commonly cited as one of the most
successful herbicides for Dalmatian toadflax control because it s translocated through
the plant and causes growth abnormalities (Robocker 1968; Duncan 1999 as cited in
Kyser and DiTomaso 2013; Kyser and DiTomaso 2013; KCE 2014; USDA 2012). As a
commercial product Tordon 22K from DOW AgroSciences offers picloram at 0.24%
concentration and is a recommended off-the-shelf management tool because it provides
the necessary soil-residual activity of approximately three years (Ogden and Renz
2005; DowAS 2009; USDA 2012; Kyser and DiTomaso 2013). Tordon 22K is selective at
the broadleaf level. There is contradicting literature concerning which time of
application of picloram is most appropriate, all of which are referenced by management
agencies throughout North America. Jacobs and Sheley (2005) and Jacobs (2006)
suggest spring application before toadflax develops its waxy cuticle, where as Robocker
(1968) suggests fall application in granular form to create soil residual herbicide.
Additionally, Kadrmas and Johnson (2002), Ogden and Renz (2005) and DiTomaso et al.
(2013) recommend application at any stage of growth is appropriate when considering
that picloram can has soil residual properties, greater success following first frosts, and
5

is applied to target rapidly growing stages. The active ingredient concentration (a.e.)
when using picloram to control toadflax varies depending on the application method,
soil pH, soil texture, interspecific competition and timing of application (Robocker et al.
1972; Jacobs and Sheley 2005). The range is wide, from 0.14 kg a.e. ha-1 to 2.52 kg a.e.
ha-1 (Robocker 1968; Robocker et al. 1972; Ferrell and Whitson 1989; Denny 2003;
Jacobs and Sheley 2005; Ogden and Renz 2005; Jacobs 2006; DiTomaso et al. 2013). To
increase the amount of picloram delivered to Dalmatian toadflax, surfactants are
recommended at 0.25% to 0.5% by volume to counteract the waxy cuticle on the leaf
which easily sheds liquid and may interfere with herbicide uptake (USDA 2012;
Lajeunesse et al. 1993, Sing 2006 as cited in Kyser and DiTomaso 2013; DiTomaso et al.
2013). Considering the variability in treatment methodology and constituents, toadflax
is most effectively controlled by combining treatments that act upon multiple life stages
of the plant (Scott 1999). With all treatment methods, especially herbicide, the effect on
non-target species must be considered.
Knowledge Gaps
Society is demanding less use of chemicals in our natural environment under the
perception that it will have an immediate positive effect on biodiversity. However,
invasive species will continue to have the opportunity to proliferate until integrated
management strategies are perfected. To reduce the spread of invasive species in the
interim, it is important to identify effective low-impact control methods. This research
will contribute to the conservation of natural ecosystems through an increased
understanding of the herbicide and manual eradication possibilities for Dalmatian
toadflax. Reduced use of herbicide is a priority for many jurisdictions throughout
Canada, including the City of Kamloops, and is critically important in protected areas,
such as Kenna Cartwright Nature Park. Therefore, it is also important to consider the
efficiency of treatment options while maintaining high eradication rates on an
operational level.

6

Objectives
The purpose of this research was to determine the most effective control method for
Dalmatian toadflax in a semi-arid grassland ecosystem that experiences frequent
anthropogenic use. Considering the limitations of previous research, the eradication
success of manual removal, broadcast spraying, spot spraying, and wicking were
investigated. To develop an evidence-based system for analysing this, the following
measurable objectives were evaluated:
1. To compare survivorship of Dalmatian toadflax between treatment methods;
2. To contrast treatment success and efficiency of treatment method;
3. To investigate the change within each life stage from baseline to post-treatment.

Methods
Site Description
This study was established in Kenna Cartwright Nature Park, Kamloops, British
Columbia in 2016. The dominant vegetation type in healthy grassland communities of
the park is characterized by bunchgrasses and forbs. This region experiences daily
mean growing season temperatures between 5.2°C in March and late October to 21.5°C
in July (ECCC 2016). Over this interval, monthly precipitation averages between 12.8
mm and 37.4mm (ECCC 2016). At approximately 800 hectares, Kenna Cartwright
Nature Park is one of the largest urban parks in North America and experienced over
230,000 visits in 2016, of which 121,968 were through the nearest entrance relative to
the study area (pers. comm. K Wourms) (Figure 1). Residents participate in hiking,
mountain biking, snowshoeing and there are a significant number of off-leash dog
walkers. The park is an important education tool through which approximately 300
students participate in annual tree planting, weed pulling and general nature
programming. Bear, deer, coyote, small mammals, raptors and owls utilize habitat
within the park (pers. comm. K Wourms). Due, in part, to the diversity within the
protected area, there are multiple on-going and historic studies (pers. comm. K
Wourms).

7

Figure 1. Study location in Kenna Cartwright Nature Park, Kamloops, British Columbia.

8

The primary recent ecosystem management tool in the Park has been prescribed fire and
this study is situated in the most recent burn area. In 2016, 14.3 hectares were burned
between March 14 to 23. Approximately 65-75% of this area experienced low intensity fire
(Morrow 2016a). The study is situated in the western half of the burn area (pers. comm. K
Wourms). Prior to the prescribed fire there was a fuel management project, during which
pine was bucked, piled and burned from January 18 to 22 (Morrow 2016b). This 8.7 ha
area did not include any of the study site. Other historic management activities include
helicopter spraying with Bacillus thuiringiensis var. kurstaki for Douglas fir tussock moth
management, goat grazing for noxious weed management, biocontrol release for invasive
plant control and danger tree falling following the 2006 – 2008 mountain and western pine
beetle epidemics (pers. comm. K Wourms).
Goats were used as a biological control between 2012 and 2015 for 10 days per year,
typically in June, and the target for seed head consumption was set at 90%. During 2014
and 2015, the goat grazing was concentrated on the area where this study is now
established. This control method was terminated in 2016 due to lack of monitoring and
understanding of true success. Anecdotal evidence suggested that the goats were not
decreasing the amount of target noxious weeds, such as Dalmatian toadflax, for which they
were intended. Biological releases are continually made in the park targeting knapweed
and Dalmatian toadflax using a variety of agents such as Larinus spp., Cyphocleonus achates
and Mecinus janthinus. On-going trail maintenance and facility improvement projects occur
using foot crews, vehicles and heavy machinery.
In the 2002 Park Restoration Plan, herbicide treatment was not recommended for toadflax
due to both the high chemical concentration levels required as well as the fact that M.
janthinus biological control was effective at that time in eliminating large infestations of
toadflax (Tarasoff 2002). Detailed information on the number of actions taken in response
to these recommendations is not known. In 2002, there were 3.75 ha, 5.57 ha and 13.84 of
light, moderate and heavy toadflax densities, respectively (Table 1). Additionally, there was
a total area of 164.29 ha of Dalmatian toadflax mixed with knapweed, in varying
compositions (Tarasoff 2002). The size of toadflax infestations throughout the park ranged
9

drastically from 0.02 ha to 12.02 ha and when found with knapweed, the infestations
ranged from 0.02 ha to 68 ha. Biocontrol release, manual removal of sparse weeds and
revegetation with native grasses was recommended for all toadflax composition and
densities. Mechanical treatment (weed wack) of trails was advised in specific situations.

10

Table 1. Historic Kenna Cartwright Restoration Plan Dalmatian toadflax inventory and management recommendations. From Tarasoff (2002).
Dalmatian Toadflax Inventory

Management Recommendations

Infestations/

Average Size and Range

Weed Wack

Revegetate with

Total Area

(ha)

Trails

native grasses

Light Toadflax

18/3.75 ha

Yes

No

Yes

Moderate Toadflax

12 released needed

Yes

Yes

Yes

3.36 (0.24-12.02)

5 releases needed

Yes

Yes

Yes

3/36.19 ha

12.06 (0.94-31.36)

6 toadflax releases needed

Yes

No

Yes

Light toadflax w/ Mod knapweed

3/2.29 ha

0.76 (0.04-1.34)

2 toadflax releases needed

Yes

Yes

yes

Light toadflax w/ Heavy knapweed

5/37.63 ha

7.53 (0.51-25.4)

8 toadflax releases needed

Yes

Yes

Yes

Mod toadflax w/ Mod knapweed

7/20.18 ha

3.01 (0.06-8.9)

8 toadflax releases needed

Yes

Yes

Yes

Heavy toadflax w/ Light knapweed

1/68.0 ha

68.0

8 toadflax releases needed

Yes

Yes

Yes

Description

Biocontrol sites >0.1 ha

Hand pull

0.21 (0.02-1.4)

10 releases needed

21/5.57 ha .

0.27 (0.02-0.97)

Heavy Toadflax

4/13.84 ha

Light toadflax w/ Light knapweed

11

Baseline Sampling
Mean Dalmatian toadflax density was consistent across the study area (p = 0.056). On
average, toadflax density is in the upper range of a medium density classification (𝑥̅ = 26
m2) according to the limited assessments of this kind available in the literature (Dodge et
al. 2008). For the purpose of this study, stem count density followed the same categories as
Dodge et al. (2008) and are used to explain site-specific management implications (Table
2).
Table 2. Dalmatian toadflax stem density categories per m2, based on Dodge et al. (2008).
Category

Toadflax Stem Density (m2)

Low

1-10

Medium

11-29

High

≥ 30

Plant community data was collected at baseline sampling but is not considered in this
research. An import graminoid species within the area is Pseudoroegneria spicata which
provides critical forage and wildlife habitat opportunities (Johnston 2008). Other grasses
found onsite include Hesperostipa comata, Koeleria macrantha and Poa secunda. The forb
community consists of primarily Achillea millefolium, Goodyera oblongifolia, Antennaria spp.
and Calochortus macrocarpus. In addition to Dalmatian toadflax, alien species within the
area include Centaurea maculosa, Aruncus dioicus and Poa pratensis. Although Bromus
tectorum is found within 350 m of the study area, none was detected in field surveys.
Artemisia tridentata and Ericameria nauseosa are scattered throughout the study area;
however, plot stratification reduced the presence of shrubs to approximately zero to
maintain consistency and to reduce the likelihood of mortality to these species via
herbicide. Pinus contorta poles and veteran Pinus ponderosa are scattered throughout the
hillside and were avoided in plot layout (Figure 2). The study area has a mean slope of 20%
and south-western aspect. Gross study area is approximately 2.81 ha and total net
treatment area is 0.108 ha. The site is boarded by a two-track road to the south and a single
trail intersects the study area in the north western portion.

12

Figure 2. Representative photograph of Dalmatian toadflax invasion within study area in Kenna Cartwright
Nature Park, Kamloops, B.C.

Experimental Design
Treatments
Manual and chemical treatments were implemented in an effort to identify the most
successful and efficient control method for Dalmatian toadflax in this region. Manual
removal was performed by hand pulling the plant and chemical treatments included
broadcast and spot spraying using a backpack sprayer, as well as herbicide application
through wicking (Figure 3).

13

a
b
Figure 3. Herbicide application methods of broadcast and spot spraying via backpack sprayer (a) and hand
wick (b).

Three chemical treatments were applied on August 23, 2017 during the early senesces life
stage. The label recommends application when plants are actively growing through full
bloom at a rate of 90 mL in 18 L solution per m2 (DowAS 2009). Considering the seminal
research and these regulatory limits, application rate was set at 213 g a.e. ha-1 picloram
(Table 3). Non-ionic surfactant Agral 90 was added to all chemical solutions at a rate of
0.25% v/v to increase adhesion (SCI 2010). The manual removal concentrated efforts on
pulling the greatest amount of root biomass from the soil as possible when soils were
slightly moist and warm. All treatments were replicated six times and these plots were 6 m
by 6 m.
Table 3. Treatment methods and type, chemical application rate and active ingredient concentration.
Commercial
Application
Tordon 22K A.E.
Actual A.E.
Treatment
Product
Rate
Concentration
Application Rate
Control

-

-

-

-

Manual Removal

-

-

-

-

Tordon 22K

8.8 L ha-1

Picloram 240 g a.e. L -1

213 g a.e. ha -1

Agral 90

0.25% v/v

Broadcast Spraying
Spot spraying
Wicking

Broadcast application was calibrated by determining the rate of movement required to
spray each plot with 762 mL of solution (approximately three minutes). The precision of
14

this method was ± 20 ml remaining at the end of all replicates. The total solution used and
time taken for all herbicide treatments was recorded, as well as the duration for manual
removal. Volume of solution used in the wick treatment could not be precisely recorded
due to the design of the hand applicator. It was, however, approximated using a graduated
cylinder.
Plot Description and Design
Treatments were randomly assigned to pre-established plots (Figure 4). Two of the
treatments (broadcast spraying and wicking) appear clumped but were not reassigned due
to the uniformity of vegetation community, aspect and slope across the study area, as
described previously in Site Description. Permeant sample points (PSP) are located in the
bottom left corner of each plot (when facing north-east) and marked with a metal stake.

Figure 4. Study plot locations with treatment method and gross study area.

Each PSP was recorded in Universal Transverse Mercator coordinate system using the
World Geodetic System (1984) projection. Azimuths of the “bottom” (to looker’s right)

15

edge and left edge of the plot square were recorded from the PSP to allow for reestablishment (Appendix B: Plot Data).
There is a one meter sampling “buffer” around the margins of the plot resulting in an inner
4 m by 4 m square available for sampling (Figure 5). This inner area represents 16 m2 of
the total 36 m2 area. Toadflax density sampling was conducted in five 1 m2 subplots per
replicate at three sampling periods, one-week pre-treatment (August 17, 2016), two weeks
post treatment (September 12, 2016) and end of growing season (October 22, 2017). The
five subplots were randomly selected, with removal, from the available 16 and established
using a 1 m by 1 m plot square as a measuring tool. This was completed at each sampling
period and so, individual plots were not sampled through time.

Figure 5. Plot design with sampling stratification method. Six replicates of each treatment (N = 30).

Field Data Collection
Count of toadflax stems in one of seven growth stage classifications was collected in the
field and subsequently reclassified into vegetative, mature or dead after further literature
review and statistical consideration (Table 4). Critical vegetation and site observations
were recorded.

16

Table 4. Classification of Dalmatian toadflax life stage according to field assessment category.
Classification

Field Assessment Category
Seedling

Vegetative

Rosette
New stem growth (creeping)
Single stem flowerless

Mature

Single stem flowers present
Branched stem flowers present

Dead

Dead

Herbicide effectiveness was determined by surveying the toadflax basal region for living
buds in addition to the above-ground evidence of herbicide uptake such as epinasty, callus
tissue formation, leaf malformations and flag leaves (Gunsolus et al. 1999). The end-ofgrowing-season assessment was completed in the same manner to investigate residual
mortality of toadflax, as well as the amount of regrowth within each treatment. Rosettes
were an important metric at this sampling stage considering the vigorous root sprouting
that toadflax is capable of. Each subplot was inspected for evidence of Mecinus janthinus
during all sampling periods by looking for ovipositor and emergence holes in the stem as
well as the weevil larvae, pupae and adults. There was no evidence of Mecinus janthinus
found in any of the subplots or in surveyed populations of Dalmatian toadflax nearby.
To determine the current plant community, sampling of cover-by-layer was performed
(Appendix A: Plant Community Summary). The canopy-by-species sampling method was
modified from Daubenmire (1958) such that three Daubenmire rectangles were randomly
assigned to one of the sampling corners and the actual cover (percent) was used instead of
the cover classes outlined by Daubenmire. Bare exposed mineral soil and litter cover were
recorded as separate layers. The cover sampling method was completed using a top-down
analysis of the absolute cover of vegetative species, woody material, bare ground, feces,
rock, and cryptogrammic crust.
Environmental data recorded for each site included weather, bearing, slope, aspect,
elevation, and site descriptors. One photo was taken at each permanent sample point and
17

followed the methodology described in Chapter 6 of the Grassland Monitoring Manual for
B.C. to capture local conditions (Delesalle et al. 2009). The photographs were taken from
two meters behind the permanent sample point and captured all four flagged corners,
when possible.
Statistical Analysis
Mean living toadflax by treatment within each life stage (vegetative, mature and total-live)
was used to investigate the treatment success. The analysis of variance between each
treatment method and the control within the total live category dictates overall success.
Investigation into the difference in means between vegetative and mature life stages offers
further insight into appropriate timing of application.
Two sample groups were marginally non-normal (Shapiro-Wilk’s test). Further
investigation analysing normality quantile-quantile plots demonstrated no need to correct
the normality. More important, though, was the failure to pass Levene’s test of
homogeneity of variances in five of nine groups. The residual scatter plots did not
demonstrate a clear suitability for transformation and attempts to implement natural
logarithm and square root transformations were unsuccessful.
Considering these limitations, a one-way analysis of variance was used when appropriate
(mature life stage) and a Welch robust test of equality of means was used for nonhomogeneous variances (total-live and vegetative life stages). The Welch ANOVA is
effective for reducing the probability of Type I error when analysing data with great
heteroscedasticity (McDonald no date). I performed a post-hoc analysis of the one-way
ANOVA using Tukey-HSD. For the Welch ANOVA, I used a Games-Howell post hoc for a
portion of vegetative and all of total-live life stages. Since the broadcast spray treatment
was constant at zero stems per m2 in the vegetative life stage, this variable was not suitable
for inclusion in the analysis of variance since the standard deviation is zero. Instead, the
remaining treatments were compared to zero using a one-sample t test (i.e. test value set to
zero). This allowed for post-hoc comparison between broadcast spraying and all other

18

treatments without jeopardizing the reliability of the analysis of variance between the
remaining treatments.
To investigate whether there were significant differences between the baseline sampling
and end-of-season count data within each life stage by treatment, I used paired t-tests for
mature and total-live data. If data indicated non-homogenous variance I used relatedsamples Wilcoxon signed rank tests.
All test significance levels were set to α = 0.05, two-tailed. All Welch ANOVA’s were
asymptotically F distributed. Visual investigation of normality assumption was completed
in R version 1.22 using DAAG package. All statistical tests were performed in IBM SPSS
Statistics for Windows, Version 24.0.

Results
Treatment Success
Treatment had a significant effect on end-of-season mean total-live stem count (Table 5).
Further investigation into the aggregated data (by splitting vegetative and mature life
stages) offers increased insight into the effect of treatments on specific life stages. At this
level, there was a statistically significant difference in the vegetative life stage between
treatments and not the mature life stage (Table 5).

19

Table 5. Analysis of variance within life stage between treatments for end-of-season Dalmatian toadflax mean
live stem count. Asterisk identifies where Welch robust tests of equality of means were performed.
Life Stage
Treatment
N
D.F. Between; Within
F
p-value

Total Live*

Vegetative*

Control

6

Manual Removal

6

Broadcast

6

Spot Spray

6

Wick

6

Control

6

Manual Removal

6

Spot Spray

6

Wick

6

4; 11.031

24.134

<0.001

3; 10.189

14.947

<0.001

See

Broadcast1

Mature

Table 6

Control

6

Manual Removal

6

Broadcast

6

Spot Spray

6

Wick

6

4; 25

1.015

0.418

1 Welch robust test of equality of means used for control, manual removal, spot spray and wick

treatments. One sample t test used to compare broadcast spraying to all other treatments.

Table 6 summarizes the post hoc analysis method used for determining statistical
difference between broadcast and all other treatments within the vegetative life stage.
Table 6. One-sample t-test summary used for post hoc comparison between broadcast spraying all other
treatments. Test statistic set to zero (equal to broadcast standard deviation).
Treatment
T Statistic
D.F.
p-value (2-tailed)
Control

8.844

5

<0.001

Manual removal

4.635

5

0.006

Spot Spray

2.699

5

0.043

Wick

6.781

5

0.001

The amount of increase or decrease in density between baseline and end-of-season stem
count data showed a significant treatment effect for the total-live and vegetative categories
and not within the mature life stage (Table 7).
20

Table 7. Analysis of variance within life stage between treatments on the difference between baseline and
end-of-season Dalmatian toadflax mean live stem count. Asterisk identifies where Welch robust test of
equality of means was performed instead of one-way ANOVA.
Life Stage
Treatment
N
D.F. Between; Within
F
p-value

Total Live

Vegetative*

Mature

Control

6

Manual Removal

6

Broadcast

6

Spot Spray

6

Wick

6

Control

6

Manual Removal

6

Broadcast

6

Spot Spray

6

Wick

6

Control

6

Manual Removal

6

Broadcast

6

Spot Spray

6

Wick

6

4; 25

12.050

<0.001

4; 12.022

23.783

<0.001

4; 25

3.098

0.034

Total-live stem count post-hoc analysis identifies statistical similarity between broadcast
and spot spraying (p=0.309), both of which had the highest overall success (Figure 6).
Broadcast spraying is different from all other treatments except spot spraying, whereas
spot spraying is also similar to manual removal (p = 0.1) and wicking (p = 0.588)
treatments. The wicking treatment is statistically different from the broadcast treatment (p
= 0.004) and control (p = 0.007), and similar to manual removal (p = 0.201). The control is
only statistically similar to the manual removal treatment (p = 0.843). Most of these
changes are evident by the post-treatment sampling period indicating a quick treatment
effect and no real change in regrowth between this time and the end of the growing season.
The vegetative life stage mirrors the total-live except that broadcast spraying is different
from all treatments in this life stage, including spot spraying (p = 0.043) and spot spraying
is different from manual removal (p = 0.048).

21

Overall, only the control increased in total-live stem count by the end of the growing season
and this increase was not statistically different from its pre-treatment density. The
broadcast and spot spraying treatments offer the greatest reduction in density, relative to
baseline, and these are statistically different from all other treatments (p < 0.001). The
broadcast and spot spraying were the only treatments to reduce vegetative stem count
from the pre-treatment to end-of-season assessments. However, the broadcast treatment
alone is statistically different from its pre-treatment density (p=0.027) (Figure 6; see
asterisks). All treatments experienced significant reductions in mean stem count within the
mature life stage. The hand pulling treatment removed all stems and it is clear that the only
living stems at the end of the season are new growth since treatment. The number of dead
stems in the post treatment sampling is greater than in the end of season assessment and
can likely be attributed to variability within the sampling with removal methodology – see
Future Research for details.

22

a

a

a
a

a
a

a

a
b
*

a

b

b

b
c
*

b
*

c

b
c
*

a
*

a
*

a
*

a

a

a
b

c
*

c

a
*

a
b
*

a

Mean stem count (m2)

a

a

a

a

a

a
a

a

a

a

b

b

b

b

b
c

a

a

d
*

c

a

a
c

a
b

b

a
*

a
*

a
b

c

Figure 6. Mean stem count (m-2) of total-live, vegetative, mature and dead life stages at baseline, posttreatment and end of growing season. Letters denote similarity within each life stage within each sampling
period. Asterisks identify where the end-of-season live stem count is statistically different from its baseline.

23

The vegetative category encompassed seedlings, rosettes and new growth (creeping)
(Table 3). Of interest is the change in seedling versus rosette density which is displayed in
Figure 7. Seedling density remained relatively constant. However, the rosette density
increased greatly in the control and manual removal treatments and marginally in the spot
spray and wicking treatments by the end of the growing season. In the broadcast, there
were no seedlings or rosettes detected during the final assessment. As a result of this,
broadcast is only statistically similar to spot spraying in both of these life stages. Similar to
the trend observed in total-live, manual removal is statistically similar to the control and
different from all other treatments. In the rosette life stage, all chemical treatments appear
to be different from the control and manual removal.

24

a
b

b

a b

a

a

Mean stem count (m2)

a
b

a

a

a

a
a
b

a

a
b

b
b

b

b

c

a
a

a

a

a
b b

a

c

b
c b

Figure 7. Seedling and rosette mean stem count (m-2) at baseline, post-treatment and end of season
assessments. Letters denote statistical similarity at the end of the growing season. Error bars are 95% C.I.

25

Treatment Efficiency
The duration of treatments varied greatly. Manual removal plots took 28 minutes longer on
average than the longest herbicide treatment. Contrarily, the differences in duration of the
three herbicide application methods were not as drastic (Table 8). Wicking was the
slowest, followed by spot spraying and finally broadcast spraying.
Table 8. Average total solution volume, herbicide content, active ingredient concentration and treatment
duration applied to replicates.
Method
Duration (min.) Solution (mL) Tordon 22K (mL) Picloram (g a.e. replicate-1)
Control
Manual Removal
Broadcast
Spot Spray
Wick

32 ± 7
3
3.5 ± 1.5
4.3 ± 1

762
567 ± 316
371 ± 170

32
26 ± 16
16 ± 6

7.8
6.4 ± 3.9
3.8 ± 1.4

Broadcast spraying used the greatest volume of herbicide solution, on average. The wick
method used half the amount of herbicide solution as broadcast spraying. Spot spraying
had a large amount of variability, which was related to toadflax density (Table 9). Above 31
stems per m2 (high density sites), spot spraying used more herbicide than broadcast
spraying; this was the case in 50% of the replicates. Due to this variability, the actual
picloram concentration in broadcast and spot spraying replicates was relatively close,
compared to the wick application method.
The average reduction in stem count was 21 in the broadcast application and 20 in the spot
spraying (Table 9). However, broadcast spraying had only 13% of the live stem density at
the end of the growing season relative to baseline and spot spraying had 31%. Wicking
replicates contained an average of 61% of the baseline density and had a mean reduction of
7 stems per m2.

26

Table 9. Herbicide application method efficiency summary highlighting herbicide solution applied to each
replicate and the baseline, post-treatment and end of growing season mean total-live stem count (m-2).
Herbicide Solution
PostEnd of
Treatment
(mL)
Baseline
Treatment
Season
Difference*

Broadcast
Spraying

Average

Spot Spraying

Average

Wick
Application

762

38

3

1

-37

762

22

3

5

-17

762

26

3

4

-22

762

18

3

5

-13

762

12

2

1

-11

762

25

2

3

-22

762

24

3

3

-21

250

23

8

3

-20

780

37

6

8

-29

880

32

7

18

-14

700

30

9

6

-24

800

30

6

15

-15

400

23

4

5

-18

635

29

7

9

-20

250

16

8

10

-5

450

28

6

17

-11

375

29

7

11

-18

450

7

9

9

2

200

10

8

10

0

500

19

9

11

-8

Average
371
18
8
11
-7
* Difference = (end of season – baseline) to represent survivorship rather than mortality.

27

Discussion
Evaluating Treatment Success and Efficiency
The objective of this research was to identify the most successful treatment method for
removal of Dalmatian toadflax and contrast these results with application efficiency. There
are several key trends in the response of Dalmatian toadflax to treatment methods that
offer new understanding of the eradication possibilities for this invasive alien species.
When selecting an appropriate control method, the broad-scale differences between
manual removal and chemical treatment present the first insight into toadflax growth
suppression.
Considering the similarity to the control in all analyses and long treatment duration,
manual removal of Dalmatian toadflax offers little chance of successful eradication. When
compared to the success of the chemical options, especially broadcast spraying, the manual
removal is not a feasible or reliable method, which is consistent with previous studies
(Scott 1999; Sheley and Clark 1999; Zouhar 2003; Kyser and DiTomaso 2013). Since hand
pulling removed all stems, there was a decrease in the total-live stem count at the end of
the growing season, relative to baseline. However, this must be contrasted with
reproduction experienced by October of the same year. The increase in rosettes under this
treatment method at both the post-treatment and end of season assessments suggests that
it did in fact promote adventitious buds and has the potential to greatly influence the
amount of growth in proceeding years. Further to this point we must consider the
morphology of Dalmatian toadflax and how this influences the environmental impact of
hand pulling; the extensive root system results is a great deal of soil disturbance.
Therefore, while manual removal may remove current toadflax biomass, the exposed
mineral soil following this treatment is highly susceptible to seedling establishment
(Robocker 1970; Jacobs 2006). In an area occupied by various invasive and agronomic
plants, this type of disturbance has the potential to promote further invasion and have a
negative impact on the native ecosystem into the future (Clark 2003; Lajeunesse 1999 as
cited in Kyser and DiTomaso 2013).

28

Despite the variety of chemical timing and application rates identified in the literature, very
few studies have been able to truly evaluate the interaction between these. Adding to this
lack of clarity is the fact that seminal research by Robocker (1968) applied incredibly high
rates of picloram in (1.68 kg a.e. ha-1), where more recent studies are within the 180-260 g
a.e. ha-1 range (Ogden and Renz 2006; Jacobs 2006; DiTomaso et al. 2013). Considering
these limitations, this study intended to identify the most appropriate control method at a
moderate application rate of 213 g a.e. ha-1 with a focus on application method.
Broadcast and spot spraying provided the highest overall success resulting in mean stem
counts of 1.7 ± 1.17 and 6.3 ± 2.7 per m2, respectively, in the total-live category at the end
of the growing season. The application of herbicide via wicking produced similar results as
the manual removal in both the total-live and vegetative life stages. Contrarily, spot
spraying only had this similarity in the total-live category. Given the small population size
required to select wicking over spot spraying, the overall impact on the plant communities
through time will offer more insight into the appropriate use of these two methods.
Furthermore, when considering the survivorship experiences in each treatment, there are
more pertinent details exposed. Wicking offered the same overall success as spot spraying in
reducing total live stems based on mean stem count. However, wicking contained 61% of the
live stems at the end of the growing season, relative to baseline, whereas spot spraying only
contained 31%. Furthermore, the lack of soil residual herbicide activity, an important

component of chemical toadflax eradication, is not offered by the wick treatment (Jacobs
2006).
Spot spraying within the total-live category is similar to the broadcast spraying and may
offer a potential control method on low to medium density sites, since it is only on high
density sites that this method uses more herbicide than broadcast spraying. Additionally,
only spot and broadcast spraying provide statistically significant reduction in vegetative
stems from pre-treatment to end-of-season. The density and distribution of toadflax must
be considered when selecting between spot and broadcast spraying. Since in three of six
replicates spot spraying used more herbicide than broadcast, the true efficiency of spot
treatment is difficult to justify when density surpasses 31 stems per m2. The duration of
29

this treatment can vary greatly and may negatively impact operational success and cause
increased financial costs that counter balance the savings in herbicide. However, on low
density sites with patchy distribution of toadflax, spot spraying may allow land managers
to reduce the environmental impact of chemical treatment on native broadleaf species,
relative to broadcast spraying (Clark 2003; Jacobs 2006). Considering the intentions of the
majority of land managers and society at large, this transferred and possibly increased
financial cost is worth the environmental benefits.
Broadcast spraying was the only treatment to eliminate vegetative growth and
reproduction at the end of the growing season and nearly did so two weeks post-treatment.
It is for this reason that on high density sites broadcast spraying is the most appropriate
treatment method. Broadcast application also offers reliable delivery rates and a
standardized duration which will increase operational efficiency. This treatment method
was quickest for high density sites and had the lowest average duration when considering
all sites. Additionally, broadcast spraying produced the greatest success by resulting in a
survivorship rate of only 13%. Broadcast application of picloram offers one of the most
important components of toadflax control, soil residual activity. When applied during
stages of active growth, before waxy cuticle development, this method has the greatest
chance of maximizing eradication success via herbicide (Jacobs and Sheley 2005; Sheley
2006; Robocker 1968; Ogden and Renz 2005; DiTomaso et al. 2013). This is important to
consider when rosettes were absent from the broadcast spraying at the end of the growing
season and nearly all of the regrowth experienced by the four other treatments were
rosettes.
The statistical analysis of treatment success was subject to a large degree of variability in
the manual removal, spot spraying, wick and control treatments. Only the broadcast spray
had low standard deviation throughout the experiment. Despite how this potentially
influences test results, clear (and often substantial) statistical differences were detected
between treatments. Additionally, there was a clear trend throughout the life stages that
chemical treatments were a further distance from manual removal and control than they

30

were from each other. Future research should standardize sampling in a way that
minimizes variability and then focus on tracking the plant communities through time.

31

Management Implications
This research highlights the importance of planning invasive plant management on small
scales and provides land managers with multiple integrated management components to
consider when removing Dalmatian toadflax from semi-arid grasslands:
1. On high density sites, broadcast spraying offers the greatest success as well as
highest efficiency and predictability. It also removed all vegetative reproduction at
the end of the growing season and offers soil residual activity;
2. On low to medium density sites spot spraying is an efficient and effective method.
However, it does not eliminate vegetative reproduction by the end of the growing
season which may impact future biomass accumulation;
3. Manual removal is not recommended as it is not statistically different from the
control. It also has the potential to increase soil disturbance and appears to promote
adventitious root buds;
4. Wicking is time consuming relative to other herbicide application methods and has
the lowest chemical eradication success. Further research through time may lend
more insight into the applicability of this method on low density sites where soil
residual activity is not desired.

32

Future Research
The high variability within stem count data is a result of the sampling method and should
be addressed before proceeding with future research. In herbicide efficacy studies,
mortality is evaluated on a stem-by-stem basis and each individual plant is marked and
studied. Since this was a field-based operational study, this strict level of stem count data at
an entire plot level was not feasible and thus introduced variability into the data that may
have been reduced using other sampling methods. It is recommended that a permeant area
be marked for sampling and all stems within this area should be counted. Timing of
application for this study was within the range of recommendations cited in the literature.
However, an earlier application (pre-bolting) has recently shown to be successful and could
be targeted for future management of Dalmatian toadflax (Kyser and DiTomaso 2013).
Considering the high statistical evidence, timing of application is less important than the
community-level impacts of treatments in future research.
The primary question going forward is how the plant community as a whole will respond
to the treatment methods, which were solely researched for and targeted at Dalmatian
toadflax control. Future research should take the baseline plant community data and track
the response of individual species through at least the next two years. Given the selectivity
of Tordon 22K, native forbs and other invasive species could respond in different ways to
both the herbicide residual activity and the absence of Dalmatian toadflax. Additionally, soil
information should be collected (specifically pH and texture) to evaluate the success of
Tordon 22K relative to other studies when considering the potential residual activity.

33

Literature Cited
[BCIMISWG]. BC Inter-Ministry Invasive Species Working Group. 2014. Invasive Plant
Early Detection and Rapid Response Plan for BC. [accessed 2017 Mar 26]. Available
from: https://www.for.gov.bc.ca/hra/invasivespecies/Publications/Prov_EDRR
_IS_Plan.pdf
Butchart SHM, Walpole M, Collen B, van Strien A, Scharlemann JPW, Almond REA, Baillie
JEM, Bomhard B, Brown C, Bruno J, et al. 2010. Global Biodiversity: Indicators of
Recent Declines. Science (80). 328:1164–1168. [accessed 2017 Mar 25]. Available
from: https://pubs.er.usgs.gov/publication/70004540
[CFIA]. Canadian Food Inspection Agency. 2014. 2013-2014 Departmental Performance
Report – Table C: Invasive Alien Species. [accessed 2016 Feb 26]. Available from:
http://www.inspection.gc.ca/about-the-cfia/accountability/reports-to-parlia
ment/2013-2014-dpr/eng/1409769354767/1409769355486?chap=0#c32s3c
City of Kamloops. 2017. Noxious Weeds & Invasive Plants. [accessed 2016 Feb 29].
Available from:http://www.kamloops.ca/ipm/weeds-overview.shtml#.VtPs-8fCrzI
Clark J. 2003. Invasive Plant Prevention Guidelines. Bozeman, MT; Centre for Invasive Plant
Management. [accessed 2016 Nov 1]. Available from: http://www.weedcenter.org/
store/docs/CIPM_prevention.pdf
[CBD]. Convention on Biological Diversity. 2008. Alien Species that Threaten Ecosystems,
Habitats, or Species (Article 8 (h)).
Delesalle B, BJ C, SJ W. 2009. Grassland monitoring manual: A tool for ranchers. Coupe B,
editor. Grasslands Conservation Council of British Columbia. Available from:
[accessed 2016 Nov 1]. http://www.bcgrasslands.org/images/stories/whatwedo
/PDFs/grassland_monitoring_manual_for_bc_high_res_compressed.pdf
Denny MK. 2003. Maintaining and establishing culturally important plants after landscape
scale disturbance. Montana State. [accessed 2016 Nov 1]. Available from: http://
scholarworks.montana.edu/xmlui/bitstream/handle/1/8314/31762103820229.pd
f?sequence=1
[TNRD]. Thompson Nicola Regional District. no date. Invasive plants affect everyone.
[accessed 2016 Feb 29]. Available from: https://www.tnrd.ca/content/noxiousweed-control
DiTomaso J., Kyser G., Al E. 2013. Weed control in Natural Areas in the Western United
States. :1–3. [accessed 2017 Mar 26]. Available from: http://wric.ucdavis.edu/
information /natural areas/wr_L/Linaria_dalmatica.pdf

34

Dodge RS, Fulé PZ, Hull Sieg C. 2008. Dalmatian toadflax (Linaria dalmatica) response to
wildfire in a southwestern USA forest. Ecoscience 15:213–222. Available from:
[accessed 2017 Mar 26].https://www.fs.fed.us/rm/pubs_other/rmrs_
2008_dodge_r001.pdf
[DowAS]. Dow AgroSciences LLC. 2009. Specimen Label Tordon 22K. Indianapolis, IN.
[accessed 2017 Feb 18]. https://s3-us-west-2.amazonaws.com/greenbookassets/L11608.pdf
[ECCC]. Environment and Climate Change Canada. 2016. Canadian Climate Normals 19812010. Station Data - Kamloops A.
[EC]. Environment Canada. 2004. An Invasive Alien Species Strategy for Canada.
[accessed 2017 Mar 25]. Available from: http://publications.gc.ca/collections
/collection_2014/ec/CW66-394-2004-eng.pdf
[ERBC]. Environmental Reporting B.C. 2015. Status of Invasive Species in B.C. [accessed
2017 March 29]. Available from: http://www.env.gov.bc.ca/soe/indicators/plantsand-animals/print_ver/EnvReportBC_Invasive_Species_Sept2015.pdf
Ogden JA, Renz MJ. 2005. Dalmatian Toadflax (Linaria genistifolia ssp. Dalmatica).
[accessed 2017 Mar 26]. Available from: http://weeds.nmsu.edu/pdfs
/dalmatian_toadflax_factsheet_11-06-05.pdf
Ferrell M, Whitson T. 1989. Picloram/fluroxypyr combinations for Dalmatian toadflax
control. Res. Prog. Rep. - West. Soc. Weed Sci.
Goodwin K, Sheley R. 2001. What To Do When Fires Fuel Weeds. Rangelands 23:15–21.
[accessed 2016 Nov 1]. Available from: https://journals.uair.arizona.edu/index.ph
p/rangelands/article/viewFile/11530/10803https://www.mendeley.com/library/
[GOC]. Government of Canada. 2016. Environment and Climate Change Canada - Nature –
Why Are Invasive Alien Species a Problem? [accessed 2017 Mar 25]. Available from:
https://www.ec.gc.ca/eee-ias/default.asp?lang=En&n=4612AC81-1
Gunsolus JL, Curran WS, Fennelly WK. 1999. Herbicide Mode of Action and Injury
Symptoms. Univ. Minnesota 87:263–271.
Morrow B. 2016a. Kenna Cartwright Park 2016 Prescribed Fire Activities.
Morrow B. 2016b. Kenna Cartwright Park Pine Removal Project Summary.
IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM
Corporation

35

[IUCN]. International Union for the Conservation of Nature. 2017. Invasive species.
[accessed 2017 Mar 25]. Available from: https://www.iucn.org/theme/species
/our-work/invasive-species
[IPCBC]. Invasive Plant Council of BC. 2008. 2006 Economic Impacts Baseline Report #6.
Econ. Impacts Baseline Rep.:46. [accessed 2016 Feb 29]. Available from:
http://bcinvasives.ca/documents/Baseline_Report6_WEB.pdf
[ISCBC]. Invasive Species Council of BC. 2017. What are their impacts for BC? [accessed
2016 Feb 26]. Available from: http://bcinvasives.ca/invasive-species/about/whatare-their-impacts-for-bc/
Jacobs J. 2006. Invasive Species Ecology and Management of Dalmatian toadflax (Linaria
dalmatica (L.) Mill.).
Jacobs JS, Sheley RL. 2005. The Effect of Season of Picloram and Chlorsulfuron Application
on Dalmatian Toadflax (Linaria genistifolia) on Prescribed Burns. Weed Technol.
19:319–324. [accessed 2016 May 29]. Available from: http://www.jstor.org.
ezproxy.tru.ca/stable/3989713?seq=1#page_scan_tab_contents
Jamieson MA, Knochel D, Manrique A, Seastedt TR. 2012. Top-down and bottom-up
controls on Dalmatian toadflax (Linaria dalmatica) performance along the Colorado
Front Range, USA. Plant Ecol. 213:185–195. [accessed 2016 May 29]. Available from:
http://www.jstor.org/stable/41412371?&seq=1#page_scan_tab_contents
Johnston K. 2008. Reducing the Competitive Advantage of Spotted Knapweed (Centaurea
maculosa) over Bluebunch Wheatgrass (Pseudoroegneria spicata) through Seed Bed
Preparation and Soil Nitrogen Reduction. Thompson Rivers University.
Kadrmas T, Johnson WS. 2002. Managing Yellow and Dalmatian Toadflax. [accessed 2017
Mar 26]. Available from: https://www.unce.unr.edu/publications/files
/nr/2002/FS0296.pdf
[KCE]. Kitsap County Extension. 2014. Dalmatian Toadflax Control. [accessed 2017 Mar 26].
Available from: http://extension.wsu.edu/kitsap/wp-content/uploads/
sites/31/2014/01/DalmationToadflax.pdf
Kyser GB, DiTomaso JM. 2013. Effect of Timing on Chemical Control of Dalmatian Toadflax
(Linaria dalmatica) in California. Invasive Plant Sci. Manag. 6:362–370. Available
from: [accessed 2016 May 29]. http://dx.doi.org/10.1614/IPSM-D-12-00071.1
Maindonald JH and Braun JW. 2015. DAAG: Data Analysis and Graphics Data and Functions.
R package version 1.22. Available from: https://CRAN.R-project.org/package=DAAG
McDonald JH. No date. One-way anova - Handbook of Biological Statistics. [accessed 2017
Mar 13]. Available from: http://www.biostathandbook.com/onewayanova.html
36

Pejchar L, Mooney HA. 2009. Invasive species, ecosystem services and human well-being.
Trends Ecol. Evol. 24:497–504. [accessed 2017 Mar 26]. Available from:
http://ac.els-cdn.com/S0169534709001761/1-s2.0-S0169534709001761main.pdf?_tid=3632daa4-1286-11e7-ae0d-00000aab0f6b&acdnat=
1490575510_2cc47be90bd963773fb0d6d09d4725c4
Pimentel D, Zuniga R, Morrison D. 2005. Update on the environmental and economic costs
associated with alien-invasive species in the United States. Ecol. Econ. 52:273–288.
[accessed 2017 Mar 25]. Available from: http://linkinghub.elsevier.com/
retrieve/pii/S0921800904003027
Pyšek P, Richardson DM. 2010. Invasive species, environmental change, and health. Annu.
Rev. Environ. Resour. 35:25–55. [accessed 2017 Mar 26]. Available from:
http://faculty.washington.edu/wgold/BES301/invasive species impacts review
2010.pdf
Ralph D, Miller V, Hougen C, Leekie J, Wikeem B, Cranston R. 2014. Field Guide to Noxious
Weeds and other selected Invasive Plants of British Columbia. 8th ed. Inter-Ministry
Invasive Species Working Group. [accessed 2016 Oct 31]. Available from:
http://bcinvasives.ca/documents/Field_Guide_to_Noxious_Weeds_Final_WEB_0925-2014.pdf
Robocker W, Schirman R, Zamora B. 1972. Carbohydrate Reserves in Roots of Dalmatian
Toadflax. Weed Sci. 20:212–214. [accessed 2016 May 29]. Available from:
http://www.jstor.org/stable/4041936?&seq=1#page_scan_tab_contents
Robocker WC. 1968. Control of Dalmatian Toadflax. J. Range Manag. 21:94. [accessed 2016
Nov 1]. Available from: http://www.jstor.org/stable/3896363?origin=crossref
Robocker WC. 1970. Seed Characteristics and Seedling Emergence of Dalmatian Toadflax.
Weed Sci. 18:720–725. [accessed 2017 Mar 27]. Available from: https://www.
jstor.org/stable/pdf/4041641.pdf
Scott L. 1999. WEEDS - Dalmatian toadflax. [accessed 2016 Nov 1]. Available from:
https://www.for.gov.bc.ca/dos/programs/range/docs/dt.pdf
Sheley RL, Clark J. 1999. Biology and management of noxious rangeland weeds. Oregon
State University Press.
[SCI]. Syngenta Canada Inc. 2010. Agral 90 Non-ionic wetting and spreading agent. Guelph,
ON. [accessed 2017 Feb 18]. Available from: https://www.syngenta.ca/
pdf/Labels/Agral_90_11809_en_label.pdf
Tarasoff C. 2002. Restoration Plan Kenna Cartwright Municipal Park. Kamloops, BC.
[accessed 2017 Feb 17]. Available from: https://www.for.gov.bc.ca/hfd
/library/FIA/2002/FIA2002MR001.pdf
37

[USDA]. USDA Forest Service. 2012. Field Guide for Managing Dalmatian and Yellow
Toadflaxes in the Southwest. USDA For. Serv. F. Guid. [accessed 2016 May 29].
Available from: http://www.fs.usda.gov/detailfull/r3/forestgrasslandhealth/invasivespecies/?cid=stelprdb5410251&width=full
Vitousek PM, D’Antonio CM, Loope LL, Rejmanek M, Westbrooks R. 1997. Introduced
Species: A Significant Component of Human-Caused Global Change. N. Z. J. Ecol.
21:1–16. [accessed 2017 Mar 26]. Available from: http://newzealandecology.
org/system/files/articles/NZJEcol21_1_1.pdf
Wilson LM, Sing SE, Piper GL, Hansen RW, Clerck-Floate R De, MacKinnon DK, Randall CB.
2005. Biology and Biological Control of Dalmation and Yellow Toadflax. [accessed
2017 Mar 27]. Available from: https://www.fs.fed.us/rm/pubs_other/
rmrs_2005_wilson_l001.pdf
Wourms K. 2017. Personal Communication. Nature Parks Crew Leader, City of Kamloops.
Zouhar K. 2003. Linaria spp. In: Fire Sciences Laboratory, editor. Fire Effects Information
System. U.S. Department of Agriculture

38

Appendix A: Plant Community Summary
Table 1. Study area plant community summary based on selectively placed plot locations.
Functional Group
Species

Grasses

Forbs

Shrubs
Trees

Invasive

Ground Cover (∑100%)

Cover (%)

Pseudoroegneria spicata

21

Hesperostipa comata

7

Koeleria macrantha

2

Poa secunda

6

Achillea millefolium

2

Goodyera oblongifolia

<1

Antennaria spp.

2

Calochortus macrocarpus

<1

Agoseris glauca var. dasycephala

1

Geum triflorum var. ciliatum

<1

Artemisia tridentata

2

Ericameria nauseosa

1

Pinus contorta

<1

Pinus ponderosa

<1

Linaria genistifolia ssp. dalmatica

34

Centaurea maculosa

4

Aruncus dioicus

2

Poa pratensis

2

Grindelia squarrosa var. serrulata

1

Bare Ground

13

Rock

4

Cryptogrammic Crust

63

Litter

20

39

Appendix B: Plot Data

Figure 1. Plot re-establishment guide. Permanent sample point corresponds with UTM coordinates outlined in
Table 1.

40

Table 1. Plot and treatment information summary including permanent sample point coordinates, block azimuths, treatment duration, herbicide application amount and
volume of chemical applied.
Azimuth
Left

Time (min.)

135

45

3

Solution
Volume
(mL)
762

50

140

4

250

5616328

48

138

3

5616320

132

42

3

682685.3

5616313

114

24

2

10 U

682724.2

5616271

142

52

-

-

-

-

-

10 U

682693.2

5616308

130

40

3

762

711.00

32.00

19.00

Wick

10 U

682712.2

5616270

112

22

4.25

450

420.56

18.93

10.51

Wick

10 U

682745.5

5616268

118

28

4.5

375

350.47

15.77

8.76

10

Broadcast Spray

10 U

682748.9

5616251

120

30

3

762

711.00

32.00

19.00

11

Wick

10 U

682751.4

5616240

138

48

3.25

450

420.56

18.93

10.51

12

Wick

10 U

682764

5616225

178

88

5.25

200

186.92

8.41

4.67

13

Wick

10 U

682778.9

5616221

134

44

4

500

467.29

21.03

11.68

14

Control

10 U

682783.5

5616253

118

28

-

-

-

-

-

15

Spot Spray

10 U

682802.3

5616263

138

48

3.5

780

728.97

32.80

18.22

16

Hand Pull

10 U

682801.1

5616276

116

26

33

-

-

-

-

17

Control

10 U

682821.3

5616286

116

26

-

-

-

-

-

18

Hand Pull

10 U

682823.1

5616295

134

44

38

-

-

-

-

19

Spot Spray

10 U

682829.1

5616305

128

38

5

400

373.83

16.82

9.35

20

Spot Spray

10 U

682835.8

5616297

118

28

4.5

250

233.64

10.51

5.84

21

Hand Pull

10 U

682831

5616291

116

26

24

-

-

-

-

22

Broadcast Spray

10 U

682854.1

5616238

104

14

3

762

711.00

32.00

19.00

23

Control

10 U

682856.1

5616200

126

36

-

-

-

-

-

24

Hand Pull

10 U

682853.3

5616181

81

351

35

-

-

-

-

25

Control

10 U

682889.4

5616220

100

10

-

-

-

-

-

26

Hand Pull

10 U

682908.3

5616226

122

32

34

-

-

-

-

27

Control

10 U

682967.6

5616208

118

28

-

-

-

-

-

28

Spot Spray

10 U

682982.3

5616174

91

1

3

800

747.66

33.64

18.69

29

Hand Pull

10 U

682944.4

5616223

90

0

27

-

-

-

-

30

Spot Spray

10 U

682812.3

5616261

110

20

3

700

654.21

29.44

16.36

Plot

Treatment

UTM
Zone

Easting

Northing

Azimuth
Bottom

1

Broadcast Spray

10 U

682613.1

5616353

2

Wick

10 U

682660.3

5616339

3

Broadcast Spray

10 U

682659.9

4

Broadcast Spray

10 U

682676.5

5

Spot Spray

10 U

6

Control

7

Broadcast Spray

8
9

41

Water (mL)

Tordon
(mL)

Agral 90
(mL)

711.00

32.00

19.00

233.64

10.51

5.84

762

711.00

32.00

19.00

762

711.00

32.00

19.00

800

747.66

33.64

18.69