Linaria dalmatica
(dalmatian toadflax)

L. dalmatica is a herbaceous plant native to western Asia and south east Europe. It is of particular concern in North America where it was introduced in the late nineteenth century. It has since spread across most of the western areas of the USA and Canada (De Clerck-Floate and Turner, 2013). There it has invaded rangelands, rights-of-way and natural habitats. It is classified as noxious in many of the western states of both countries. It is also classed as invasive in South Africa. Its success may be attributed to its high specific leaf area, aggressive root system, prolific seed production and the presence of alkaloids which discourages grazing by livestock, thus allowing it to become dominant (De Clerck-Floate and Turner, 2013).

The native distribution of L. dalmatica is restricted to the Balkan Peninsula, Turkey, Iran, Iraq, Syria and Romania. The Royal Botanic Garden Edinburgh (2015) notes that it is locally naturalized in Central Europe therefore, the records from Germany and Hungary presumably refer to introductions. It also occurs as an occasional escape from gardens in Switzerland, UK and elsewhere in Europe. It is now widely naturalized and invasive across Canada and western USA (Leo’s Dictionaries, 2015). L. dalmatica has more recently been spreading in South Africa (Invasive Species South Africa, 2015) and has also naturalized in Australia (Parsons and Cuthbertson, 1992) and in Argentina (Puntieri and Brion, 2003). As it is widely cultivated in gardens, it is likely to occur as an occasional escape in many other countries.

The risk of introduction of L. dalmatica continues to be high, though there must be few temperate countries to which it has not already been introduced as an ornamental. Thus the risk of spread is mainly accidental, from careless disposal of material from gardens.

In its native range L. dalmatica is a plant of uncultivated fields, vineyards, roadsides and waste places, mainly on coarse textured soils. It may occur on limestone hills and generally favours alkaline soils. In North America, its introduced range, it occurs in similar habitats but has also invaded forest plantations, wood and herbaceous species in the genus Artemisia, pastures, rangelands and to some extent crops such as lucerne/alfalfa (Medicago sativa) (Vujnovic and Wein, 2005).

L. dalmatica has a wide temperature tolerance but on a mountain area in Canada, establishment on disturbed roadsides was limited by the low temperatures experienced at higher elevation (Polinac and Rew, 2014).

The crops affected by L. dalmatica are mainly pasture and rangeland species, but alfalfa (Medicago sativa) and forest species such as ponderosa pine (Pinus ponderosa) may also be affected.

Genetics

The chromosome number 2n = 12 has been reported for L. dalmatica (Missouri Botanical Garden, 2015).

Reproductive Biology

L. dalmatica spreads both by seed and by its spreading root system. The shoots arising from the roots are described as ‘sterile’, but presumably will develop fully if separated from the parent plants.

Under laboratory conditions some seeds will germinate the same season in which produced. Germination from dry storage increases for two to three years, with a periodic low in November and a high in April. Over 90% germination is obtained with two to three year old seeds in the laboratory. In the field, germination occurs in both spring and autumn and seedlings can emerge from a soil depth of 2-2.5 cm. Seeds may be dormant for a few weeks after maturation. Thereafter they have no particular requirement (Robocker, 1970).

Seedlings which emerge in the autumn generally do not survive the following year. Most emergences occur in the spring, following stratification over winter. Cyclic alterations of 10 days stratification at 2°C and 10 days drying at room temperature for two cycles induced higher rates of germination than did any other factor. The highest germination of seeds planted at different depths in five soil textures was found to be from a depth of 1.27 cm in loamy sand (Alex, 1962). Seeds germinate from 2.5 cm in clay and loamy soils and from 5 cm in sandy soils. Seedling recruitment can be seriously reduced by plant competition (Grieshop and Nowierski, 2002).

Shoots growing from the crown flower from May and fruit by August, producing large numbers of seeds, up to 250 per capsule and up to 500,000 per plant. The flowers are self-sterile and are insect pollinated (Vujnovic and Wein, 2005). Seventeen species of bumble bee (Bombus spp.), honey bee (Apis mellifera) and the leaf cutting bee Megachile periherta were identified as pollinators, with bumble bees being most common (97% of observations) (De Clerck-Floate et al., 1997).

Physiology and Phenology

A study by Alex (1962) found that roots of seedlings penetrated downward ca. 1 cm/day during the first 31 days in fine sandy loam. Thereafter the rate was slower but roots reached a depth of 56 cm after 105 days. The total length of all roots of one 105 day old plant grown in fine sandy loam exceeded 27 m. Adventitious stem buds were present on the primary axis after 22 days of age and on the second and third order branch roots at 105 days of age. Stem buds on branch roots were mostly more than 16 cm distant from the primary axis and at depths of 2 cm to more than 10 cm below the ground surface (Alex, 1962). Once established, seedlings can flower and set seed within the first year (Vujnovic and Wein, 2005).

L. dalmatica has been shown to benefit greatly from enhanced carbon dioxide levels. In one study, at 600 ppm CO₂, the biomass of L. dalmatica increased by 13 fold, seed production by 32 fold and clonal expansion by seven fold, much larger than the effects observed on the native C₃ grass Pascopyrum smithii [Elymus smithii] (Blumenthal et al., 2013). Increased temperatures of 1.5/3.0°C (day/night) had little effect (Blumenthal et al., 2013). However a study by Dijkstra et al. (2010) found much less pronounced increases and emphasized the potential interactions with nitrogen and water levels.

Longevity

The average life of a plant is about three years. The life of a stand varies and depends on the interactions of the plant's growth habits with environmental factors. A stand may disappear within three years under severe competition. However the development of floral stems from secondary crown points on lateral roots may prolong or sustain the stand beyond the life of the original plants (Robocker, 1974).

Seeds persist in the field for at least 10 years (Robocker, 1974). A study by Bruns (1965) found that seeds could still germinate after 12 months of submergence in canal water.

Nutrition

Nitrogen addition has a strong positive effect on plant performance, with a two fold increase in both biomass and seed production (Jamieson et al., 2010).

Associations

In North America, L. dalmatica is associated with pastures and rangelands dominated by Bromus tectorum, B. japonicus, Stipa comata, Koeleria cristata [Koeleria pyramidata], Festuca idahoensis and Artemisia tridentata (Vujnovic and Wein, 2005).

Environmental Requirements

L. dalmatica is favoured by high water and nitrogen availability, but is relatively unpalatable to cattle. Cover of L. dalmatica was high on steeper slopes, particularly those with southern aspects (Jamieson et al., 2010). The beneficial effects of nitrogen on L. dalmatica can be neutralized by addition of labile carbon to the soil (Blumenthal, 2009).

Wilson et al. (2005) indicate that L. dalmatica is particularly favoured in areas with dry summers. Nevertheless, snow can be highly beneficial to L. dalmatica, apparently in terms of providing high moisture during its establishment phase (Blumenthal et al., 2008).

Post wildfire conditions in high severity burned areas favour increased density, cover, reproduction and spread of L. dalmatica. Whilst native species richness may be reduced, suggesting that the invasive species would persist, at least in the short term, at the expense of natives (Dodge et al., 2008). L. dalmatica was also found to benefit from prescribed fire in a rangeland situation (Jacobs and Sheley, 2003).

The natural enemies of L. dalmatica have been well studied as a part of efforts to achieve biological control. Potential biocontrol organisms have included Calophasia lunula, Mecinus janthiniformis, Eteobalea intermediaella, Rhinusa neta, R. linariae and R. antirrhini. Recent studies include the stem boring weevils M. laeviceps and M. peterharrisi and the stem-galling weevil R. rara.

Environmental

• Amenity

General

• Sociocultural value

Ornamental

• Propagation material
• Seed trade

L. dalmatica is quite distinct from the co-occurring L. vulgaris, which has narrow strap like leaves. L. genistifolia, which also occurs in the USA, also differs from L. dalmatica in its much narrower leaves and relatively small corolla (2-5 cm), while the seeds are without a winged margin.

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Prevention

SPS Measures

L. dalmatica is listed as noxious in Arizona, California, Colorado, Idaho, Minnesota, Montana, Nevada, North Dakota, New Mexico, Oregon, south Dakota, Utah, Washington and Wyoming, USA (USDA-NRCS, 2015). In British Colombia, Canada, it is listed as noxious in Alberta and Manitoba (De Clerck-Floate and Turner, 2013).

Control

Cultural Control and Sanitary Measures

It is well recognized that L. dalmatica is quick to exploit gaps and weaknesses in established grass stands. It can be kept out of areas by maintenance of a dense healthy stand. Grasses may vary in their effectiveness and in one study of grasses to suppress L. dalmatica, the greatest reduction (95%) was found with Elymus lanceolatus (Rose et al., 2001). However, once established it may not be readily suppressed by newly sown grasses. Early detection is vital for the control of new infestations.

Physical/Mechanical Control

Mowing is relatively ineffective for controlling L. dalmatica as the weed can readily recover from the root systems and the removal of competition from other vegetation may allow it to increase. Similarly burning may encourage growth rather than suppress it. In crop land intensive cultivation for at least two years with at 8-10 cultivations in the first year may be required. Hand pulling can prevent seeding but only annual grubbing and pulling over a 10-15 year period can control it fully (Vujnovic and Wein, 2005).

Biological Control

L. dalmatica has been the subject of a number of biological control programmes. According to De Clerck-Floate and Turner (2013), five insect biocontrol agents have been tested and intentionally introduced into Canada for control of L. dalmatica. The first field release took place in 1963.

Of the five insects released, three have established on L. dalmatica. These include the seed capsule weevil Rhinusa antirrhini, the defoliating moth Calophasia lunula and the stem boring weevil Mecinus janthiniformis. The latter had previously been referred to as ‘M. janthinus’ but differing results from different populations of this ‘species’ led to the realisation that two species were involved. Only M. janthiniformis is effective on this L. dalmatica while L. janthinus is effective on L. vulgaris (Toševski et al., 2011). Toševski et al. (2013) have now described a fast and accurate way of distinguishing the two Mecinus species haplotypes using PCR-RFLP diagnostic assay of the mitochondrial cytochrome oxidase subunit II (COII) gene.

R. antirrhini has established well in British Colombia, Canada and populations are increasing. It is however uncertain whether this seed feeding agent can have a significant impact on the weed (De Clerck-Floate and Turner, 2013).

In Canada, Calophasia lunula has only become established in southern British Colombia and not at all in Alberta, apparently due to low temperatures (De Clerck-Floate and Turner, 2013). Jamieson et al. (2010) found that C. lunula has a mechanism to de-activate the iridoid glycosides, defence mechanisms, in L. dalmatica but ‘M. janthinus’ does not.

M. janthiniformis is by far the most successful biocontrol agent released in Canada. Damage is caused by both adults and larvae and conspicuous reductions in the weed have been recorded. Results are less successful in Alberta almost certainly due to severe winter temperatures. Some disappointing results may also be due to the wrong Mecinus species being released. Weed and Schwarzländer (2014) suggest that the impact of M. janthiniformis may vary considerably due to site specific variation in rainfall and density dependent processes. As a result alternative control methods should be prioritized in areas where herbivore impact is expected to be low. Studies have recently started on two additional Mecinus species, M. laeviceps and M. peterharrisi (Toševski et al., 2016). A petition for the introduction of the stem-galling weevil R. rara n North America is pending.

Two species introduced but not proving successful are the root galling weevil R. linariae which proved to have a preference for L. vulgaris and is no longer being exploited and the root moth Eteobalea intermediella which established in 1998 but had died out by 2002 and is no longer being used (De Clerck-Floate and Turner., 2013). The related E. serratella was thought by Quarles (2007) to be having an impact on L. dalmatica, but Shelton (2015) notes that L. dalmatica is a relatively poor host for this species and it appears that it is no longer of interest.

Grubb et al. (2002) suggested that the naturally occurring flower feeding beetle Brachypterolus pulicarius might prove to be an important component in the integrated weed management of L. dalmatica but that the adults show a distinct preference for L. vulgaris and its further distribution in North America was not recommended (MacKinnon et al., 2007). The seed capsule weevil R. neta also occurs widely in British Colombia, Canada but is not being actively distributed (De Clerck-Floate and Turner 2013).

Wilson et al. (2005) provide extremely detailed and valuable information on the biology of the biocontrol organisms and the planning, implementation and monitoring of biological control programmes against both L. dalmatica and L. vulgaris.

Chemical Control

Phenoxy-propionic herbicides are superior to phenoxyacetic herbicides in controlling L. dalmatica. Satisfactory control can be obtained with fenoprop. Rates of fenoprop required to check or to control the plant did not injure perennial grasses. Dicamba has also been effective however it is resistant to triclopyr and fluroxypy (Vujnovic and Wein, 2005).

Picloram can be effective when applied at various stages of growth, but granules applied to the soil in the autumn can be more effective than a foliar application at the same rate in spring (Vujnovic and Wein, 2005). A combination of fenoprop and picloram has also provided control (Robocker, 1968).

Aminocyclopyrachlor (140 and 280 g ae ha^-1) and aminocyclopyrachlor+chlorsulfuron (140 g/ha+53 g/ha) gave second year control; chlorsulfuron at the dormant stage (105 and 158 g/ha) and aminopyralid at the rosette stage (245 g/ha) also gave 2 years of control. The treatments had only minor effects on grass species (Kyser and DiTomaso, 2013).

Herbicides recommended by USDA Forest Service (2012) for use in non-crop areas include imazapic, glyphosate, rimsulfuron, sulfometuron and chlorsulfuron.

IPM

Quarles (2007) suggests that integrated control of L. dalmatica could involve combinations of clean cultivation, pulling, mowing, burning, grazing, competitive planting, using biological control agents and herbicides. Suggestions for integrated control strategies are also made by the USDA Forest Service (2012). Suggestions include checking hay and straw for the presence of seed, feeding certified weed-free hay pellets to horses, combining control methods and ensuring that follow up treatment is carried out to treat plants that are missed.

Ecosystem Restoration

There are indications that the reduction of L. dalmatica in mixed natural plant communities has tended to encourage the invasion of other equally undesirable weedy species such as Bromus tectorum and Centaurea maculosa [Centaurea podospermifolia].

13/12/2016 Updated by:

André Gassmann, CABI-CH, Switzerland

12/06/2015 Original text by:

Chris Parker, Consultant, UK