Houndstongue is a biennial or short-lived perennial forb. The following description of houndstongue is based on reviews by Upadhyaya and others [99] and de Jong and others [26]. It presents characteristics of houndstongue that may be relevant to fire ecology, and is not meant to be used for identification. Keys for identification are available [17,35,37,42,43,63,104,107]. A detailed discussion of the biology of houndstongue is provided by de Jong and others [26].
Houndstongue forms a rosette in its 1st year with softly pubescent leaves 4 to 12 inches (10-30 cm) long and 0.8 to 2 inches (2-5 cm) wide. It has a thick, black, branching taproot, extending to depths > 40 inches (100 cm). An individual plant consists of 1 or several rosettes on a single root system. Flower stems, 12 to 48 inches (30-120 cm) tall, are produced the 2nd year, or sometimes later. The upper part of the stem is branching. Stem leaves are generally broadest near the base and narrower towards the tip. Houndstongue inflorescences have up to 35 but usually no more than 10 flowers each and are axillary to leaves or terminating short branches. Each flower can produce up to 4 fruits. The fruits are nutlets, the surfaces of which are flat and densely covered with small, barbed hooks (glochidia) that facilitate dispersal by animals.
Houndstongue rosettes can withstand drought stress, enabling the plant to survive water deficits and to delay flowering until conditions are favorable. The thick, deep taproot can exploit lower soil strata for water and nutrients. The root also serves as an underground food reserve which may offer protection against winter injury [11,98,99].
Houndstongue often occurs in dense stands, but may also occur as a small but regular proportion of regional flora. Seedlings are usually strongly clustered around parent plants in densities of up to 405 seedlings per ft2 (4500/m2) [26]. Houndstongue may grow as an isolated plant or at densities up to 1 individual per ft2 (8/m2) on dunes in England. On coastal dunes in the Netherlands, its abundance ranges from rare to abundant. Vesicular-arbuscular mycorrhizae are usually present on houndstongue roots [26,39].
The center of origin of houndstongue is thought to be the mountains of western Asia and eastern Europe. Houndstongue also occurs in apparently natural communities in Great Britain. It is widely distributed through Europe, where its northern limit is approximately 60°N, though there are isolated occurrences up to 68°N in Sweden. Towards the Mediterranean region it becomes rare. Houndstongue does not grow in the southernmost regions of Europe [26].
A review by Upadhyaya and others [99] suggests houndstongue was introduced to North America as a crop seed contaminant from Europe. Herbarium specimens of houndstongue were collected in Ontario as early as 1859 and in the western provinces between 1922 and 1934. Houndstongue was noted in 1884 as "common" around Montreal, and as "a pest" in Ontario. As of 1988, houndstongue occurred in all provinces in Canada except Prince Edward Island and Newfoundland, and appeared to be most abundant in southern British Columbia and Ontario. Houndstongue occurs throughout the contiguous U.S., in all but 6 southern states. Its occurrence has not been reported in Alaska or Hawaii [96]. Houndstongue is reported as a problem plant in natural areas and parks in several states including Michigan, Missouri, Indiana [73], Colorado, and Oregon [80]. The Plants database provides a distribution map of houndstongue in the United States.
The following biogeographic classification systems are presented as a guide to demonstrate where houndstongue may be found. Precise distribution information is limited. Because it is so widespread and has broad ecological tolerances, it is difficult to exclude many ecosystems as potential hosts of houndstongue plants or populations. Therefore, these lists are speculative and not necessarily exhaustive.
Fire adaptations: Houndstongue reproduces by seed, some of which may be dispersed over moderate to long distances by animals. Fire creates conditions that are favorable for establishment of houndstongue (i.e. open canopy, reduced competition, areas of bare soil), so if houndstongue seeds are present and competition minimal, it may be favored in the postfire community. Houndstongue plants may also survive fire, since nutrient reserves in the taproot acquired during the 1st year are sufficient for normal seed production the following year, even if the plants are completely defoliated early in the spring [11,98]. More research is needed regarding adaptations of houndstongue to fire.
FIRE REGIMES: Introduced species can alter the rate of spread of fire, the probability of occurrence of fire, and the intensity of fire in an ecosystem [18]. It is unclear how the presence of houndstongue may alter the fire regime of a given site, and it is unclear how a historical fire regime might affect the presence or abundance of houndstongue at a given site. It has been suggested that the exclusion or alteration of natural processes, such as fire and flooding, can encourage the establishment and persistence of houndstongue on prairie sites in Colorado [80]. Houndstongue did not occur in any of these communities at the time in which historic fire regimes were functioning, but has established since fire exclusion began. More information is needed about ecosystems in which houndstongue is likely to be invasive in North America.
The following table provides some fire return intervals for communities or ecosystems in which houndstongue may be found. Because houndstongue is widespread, it is difficult to exclude many ecosystems as potential hosts of houndstongue plants or populations. Find further fire regime information for the plant communities in which this species may occur by entering the species name in the FEIS home page under "Find FIRE REGIMES".
Community or Ecosystem Dominant Species Fire Return Interval Range (years) silver fir-Douglas-fir Abies amabilis-Pseudotsuga menziesii var. menziesii > 200 grand fir A. grandis 35-200 [3] maple-beech-birch Acer-Fagus-Betula > 1000 silver maple-American elm A. saccharinum-Ulmus americana sugar maple A. saccharum > 1000 sugar maple-basswood A. s.-Tilia americana > 1000 [105] California chaparral Adenostoma and/or Arctostaphylos spp. 72] bluestem prairie Andropogon gerardii var. gerardii-Schizachyrium scoparium 55,72] Nebraska sandhills prairie A. g. var. paucipilus-S. s. bluestem-Sacahuista prairie A. littoralis-Spartina spartinae sagebrush steppe Artemisia tridentata/Pseudoroegneria spicata 20-70 [72] basin big sagebrush A. t. var. tridentata 12-43 [84] mountain big sagebrush A. t. var. vaseyana 15-40 [4,13,66] Wyoming big sagebrush A. t. var. wyomingensis 10-70 (40**) [103,109] coastal sagebrush A. californica desert grasslands Bouteloua eriopoda and/or Pleuraphis mutica 5-100 plains grasslands Bouteloua spp. blue grama-needle-and-thread grass-western wheatgrass B. gracilis-Hesperostipa comata-Pascopyrum smithii blue grama-buffalo grass B. g.-Buchloe dactyloides grama-galleta steppe B. g.-Pleuraphis jamesii blue grama-tobosa prairie B. g.-P. mutica cheatgrass Bromus tectorum California montane chaparral Ceanothus and/or Arctostaphylos spp. 50-100 [72] sugarberry-America elm-green ash Celtis laevigata-Ulmus americana-Fraxinus pennsylvanica 105] paloverde-cactus shrub Cercidium microphyllum/Opuntia spp. 72] curlleaf mountain-mahogany* Cercocarpus ledifolius 13-1000 [6,85] mountain-mahogany-Gambel oak scrub C. l.-Quercus gambelii blackbrush Coleogyne ramosissima Arizona cypress Cupressus arizonica northern cordgrass prairie Distichlis spicata-Spartina spp. 1-3 [72] beech-sugar maple Fagus spp.-Acer saccharum > 1000 [105] California steppe Festuca-Danthonia spp. 72] black ash Fraxinus nigra 105] juniper-oak savanna Juniperus ashei-Quercus virginiana Ashe juniper J. a. western juniper J. occidentalis 20-70 Rocky Mountain juniper J. scopulorum tamarack Larix laricina 35-200 [72] western larch L. occidentalis 25-100 [3] creosotebush Larrea tridentata Ceniza shrub L. t.-Leucophyllum frutescens-Prosopis glandulosa 72] yellow-poplar Liriodendron tulipifera 105] wheatgrass plains grasslands Pascopyrum smithii 72] Great Lakes spruce-fir Picea-Abies spp. 35 to > 200 northeastern spruce-fir Picea-Abies spp. 35-200 [30] Engelmann spruce-subalpine fir P. engelmannii-A. lasiocarpa 35 to > 200 [3] black spruce P. mariana 35-200 [30] blue spruce* P. pungens 35-200 [3] red spruce* P. rubens 35-200 [30] pine-cypress forest Pinus-Cupressus spp. 3] pinyon-juniper Pinus-Juniperus spp. 72] whitebark pine* P. albicaulis 50-200 [3] jack pine P. banksiana 30] Mexican pinyon P. cembroides 20-70 [68,93] Rocky Mountain lodgepole pine* P. contorta var. latifolia 25-300+ [2,3,83] Sierra lodgepole pine* P. c. var. murrayana 35-200 [3] shortleaf pine P. echinata 2-15 shortleaf pine-oak P. e.-Quercus spp. 105] Colorado pinyon P. edulis 10-49 [72] slash pine P. elliottii 3-8 [105] Jeffrey pine P. jeffreyi 5-30 western white pine* P. monticola 50-200 Pacific ponderosa pine* P. ponderosa var. ponderosa 1-47 [3] interior ponderosa pine* P. p. var. scopulorum 2-30 [3,8,61] Arizona pine P. p. var. arizonica 2-15 [8,16,86] Table Mountain pine P. pungens 105] red pine (Great Lakes region) P. resinosa 10-200 (10**) [30,33] red-white-jack pine* P. r.-P. strobus-P. banksiana 10-300 [30,40] pitch pine P. rigida 6-25 [12,41] pocosin P. serotina 3-8 eastern white pine P. strobus 35-200 eastern white pine-eastern hemlock P. s.-Tsuga canadensis 35-200 eastern white pine-northern red oak-red maple P. s.-Quercus rubra-Acer rubrum 35-200 loblolly pine P. taeda 3-8 Virginia pine P. virginiana 10 to sycamore-sweetgum-American elm Platanus occidentalis-Liquidambar styraciflua-Ulmus americana 105] galleta-threeawn shrubsteppe Pleuraphis jamesii-Aristida purpurea eastern cottonwood Populus deltoides 72] aspen-birch P. tremuloides-Betula papyrifera 35-200 [30,105] quaking aspen (west of the Great Plains) P. t. 7-120 [3,38,65] mesquite Prosopis glandulosa 64,72] black cherry-sugar maple Prunus serotina-Acer saccharum > 1000 [105] mountain grasslands Pseudoroegneria spicata 3-40 (10**) [2,3] Rocky Mountain Douglas-fir* Pseudotsuga menziesii var. glauca 25-100 [3,4,5] coastal Douglas-fir* P. m. var. menziesii 40-240 [3,69,82] California mixed evergreen P. m. var. m.-Lithocarpus densiflorus-Arbutus menziesii California oakwoods Quercus spp. 3] oak-hickory Q.-Carya spp. 105] oak-juniper woodland (Southwest) Q.-Juniperus spp. 72] northeastern oak-pine Q.-Pinus spp. 10 to 105] coast live oak Q. agrifolia 3] white oak-black oak-northern red oak Q. alba-Q. velutina-Q. rubra 105] canyon live oak Q. chrysolepis blue oak-foothills pine Q. douglasii-Pinus sabiniana 3] northern pin oak Q. ellipsoidalis 105] Oregon white oak Q. garryana 3] bear oak Q. ilicifolia 105] California black oak Q. kelloggii 5-30 [72] bur oak Q. macrocarpa 105] oak savanna Q. m./Andropogon gerardii-Schizachyrium scoparium 2-14 [72,105] chestnut oak Q. prinus 3-8 northern red oak Q. rubra 10 to post oak-blackjack oak Q. stellata-Q. marilandica black oak Q velutina live oak Q virginiana 10 to105] interior live oak Q. wislizenii 3] blackland prairie Schizachyrium scoparium-Nassella leucotricha little bluestem-grama prairie S. s.-Bouteloua spp. 72] redwood Sequoia sempervirens 5-200 [3,32,92] pondcypress Taxodium distichum var. nutans 70] western redcedar-western hemlock Thuja plicata-Tsuga heterophylla > 200 [3] eastern hemlock-yellow birch Tsuga canadensis-Betula alleghaniensis > 200 [105] western hemlock-Sitka spruce T. heterophylla-Picea sitchensis > 200 mountain hemlock* T. mertensiana 35 to > 200 [3] elm-ash-cottonwood Ulmus-Fraxinus-Populus spp. 30,105] *fire return interval varies widely; trends in variation are noted in the species summaryFire as a control agent: Observations by managers suggest that a lack of normal ecological processes such as fire and flood can promote establishment by invasive plants such as houndstongue [80]. In some ecosystems, re-establishing historic FIRE REGIMES can be effective at controlling invasive species by encouraging growth and vigor of native species (e.g. [19,29,44,79]). Research is needed regarding the potential of prescribed burning to control houndstongue.
Postfire colonization potential: General precautions should be followed to prevent houndstongue establishment after fire. The USDA Forest Service's "Guide to noxious weed prevention practices" [95] provides several fire management considerations for weed prevention in general that can be applied to houndstongue. Wildfire managers might consider including weed prevention education and providing weed identification aids during fire training; avoiding known weed infestations when locating firelines, monitoring camps, staging areas, and helibases to be sure they are kept weed free; taking care that equipment is weed free; incorporating the cost of weed prevention and management into fire rehabilitation plans; and acquiring restoration funding. Careful postfire vigilance to identify and record the establishment of new populations is critical. About 1 month after fire, survey for signs of new or resprouting weeds. Repeat surveys will be needed, with the frequency and intensity of the survey guided by local conditions [7].
Potential weed problems must be addressed during prefire planning of prescribed burns, and following both wild and prescribed fires. When planning a prescribed burn, preinventory the project area, evaluate cover and phenology of any houndstongue present on or adjacent to the site, and evaluate the potential for increased houndstongue populations in the area [7]. Avoid ignition and burning in areas at high risk for weed establishment or spread, and/or plan for follow-up treatments in succeeding years. Avoid creating soil conditions that promote weed germination and establishment. Discuss weed status and risks in burn rehabilitation plans [95].
To prevent new infestations, re-establish vegetation on bare ground as soon after fire as possible, using either natural recovery or artificial techniques as appropriate to site conditions and objectives. When reseeding after wildfires and prescribed burns, use only certified weed-free seed. Monitor the burn site and associated disturbed areas after the fire and the following spring for emergence of houndstongue, and treat to eradicate any emergent houndstongue or other pestiferous plants. Regulate human, pack animal, and livestock entry into burned areas at risk for weed invasion until desirable site vegetation has recovered sufficiently to resist weed invasion. Additional guidelines and specific recommendations and requirements are available [7,36,95].In North America, houndstongue often occurs in "waste" areas [63,104] and disturbed sites (especially roadsides) [17,37,42,67], including logging roads and heavily grazed areas [20,60], disturbed woods and forest clearings [104,106], and lake shores [104]. Dense infestations of houndstongue are found in clearcut areas in northern British Columbia [99], and Upadhyaya and Cranston [98] suggest that houndstongue thrives and has great potential for spread in areas disturbed by logging operations and road construction.
Houndstongue is most abundant in areas with more than 10% bare ground (review by [99]). Studies in Europe also indicate that houndstongue requires open, disturbed ground and that it is most common in dry, open habitats [26], such as the dune grasslands in England. Houndstongue can exploit bare ground around rabbit warrens, benefiting from the high nitrogen levels from rabbit excreta, and the fact that rabbits do not usually eat houndstongue. Houndstongue is often found in areas grazed by livestock and wildlife [20,74,81].
In England and in the Netherlands, houndstongue occurs on sand dunes and calcareous substrates. These calcareous, sandy soils have high nitrate levels that favor houndstongue establishment, and abundant calcium that houndstongue accumulates to a marked degree, especially in its leaves [11]. Houndstongue is absent from acid coastal dunes and from acid sandy soils and does not occur on peat or clay soils [99]. In Germany houndstongue grows on both calcium-rich and weakly-acid soil, and is classified as both a drought and nitrogen indicator [26]. In British Columbia it is found on soils ranging from well-drained, relatively coarse material to clay subsoils in the open coniferous and deciduous forests. In Eastern Canada, houndstongue is often associated with rocky pastures in limestone regions [98].
Houndstongue is found in temperate regions. In British Columbia it is found on sites that are characterized by hot, dry summers and cold winters, with annual precipitation in the range of 11 to 18 inches (268-448 mm), and mean January and July temperatures of approximately 21 and 72 degrees Fahrenheit (-6 and 22 ºC), respectively. In Ontario, houndstongue-infested regions have annual precipitation and mean January and July temperature variations in the range of 31 to 41 inches (770-1020 mm), 39 to 52 degrees Fahrenheit (3.9-10.9 ºC) and 67 to 72 degrees Fahrenheit (19.2-22.2 ºC), respectively [99]. On droughty sand dunes in the Netherlands, the amount of rainfall during the growing season greatly affects growth, survival and seed production in houndstongue. Here the species may disappear from the most exposed sites after several dry years, and recolonize these areas again after a number of wet years. Plants grow poorly and eventually die after waterlogging [26]. Houndstongue is not well adapted to dry grassland sites with less than 12 inches (300 mm) annual precipitation in British Columbia. It survives well in wetter grasslands and moist draws in drier sites [98].
The distribution pattern of houndstongue in Europe suggests that at its northern limits its temperature requirements during the growing season, rather than the occurrence of winter frost, restrict the species to warmer microsites. On sand dunes in England, it occurs mostly on south or south-east facing slopes [26]. Houndstongue is completely absent from areas above 495 feet (150 m) in England. In continental Europe, it reaches the subalpine but only rarely the alpine (4,950 to 7,920 feet (1,500-2,400 m)). Some elevational ranges at which houndstongue occurs are given below:
State Elevation range Reference UT 4,880 to 9,900 feet (1,480-3,000 m) [107] CA 2,800 to 3,300 feet (850-1,000 m) [42] NM 5,000 to 8,000 feet (1,500-2,400 m) [63]Impacts: Houndstongue can establish rapidly and form dense monocultures in disturbed habitats. Populations of houndstongue displace native plant species and hinder the re-establishment of valuable range species, thereby decreasing availability of forage to wildlife and livestock [80]. It is most detrimental on rangelands and hayfields because of its toxicity to livestock, although, in most cases, the fresh plant is considered unpalatable by livestock and is generally avoided (see Importance to Livestock and Wildlife) [50,99].
The barbed seeds of houndstongue readily attach to wool and fur. This can create marketing problems and require extra time and money for removal, thus reducing the value of livestock. The seeds can also attach to the eyelashes of animals and cause eye damage, and the foliage may cause dermatitis [50,99].
Control: Houndstongue can be controlled by killing plants and/or preventing seed production. Long-term control of houndstongue requires an integrated management approach [50].
Prevention: Prevention is the most effective method for managing invasive species, including houndstongue [50,88]. Preventing or dramatically reducing seed production and dispersal, detecting and eradicating weed introductions early, containing current infestations, minimizing soil disturbances, establishing competitive grasses, and managing grazing properly will all help decrease the spread of infestations.
Cleaning livestock when they are moved from an infested area to an uninfested area is critical to prevent seed spread [23]. Houndstongue seeds also readily adhere to shoes and clothing and need to be removed and carefully disposed of (burned or bagged). It is important to clean mowers, vehicles, and tillage equipment after operating in an infested area. When seeding is necessary, use clean, certified weed-free seed and mulch to ensure that these or other weeds are not being sown.
Place a priority on controlling small infestations so they do not expand. Conducting aggressive monitoring and treatment several times each year can help with early detection and containment of infestations when they are small. Monitoring efforts are best concentrated on the most disturbed areas in a site, particularly along roadsides, parking lots, fencelines, and waterways. When an infestation is found, the location can be recorded and the surrounding area surveyed to determine the size and extent of the infestation, so these sites can be revisited on follow-up surveys. For more on monitoring see Johnson [47].
Historic overgrazing by livestock and native ungulates encourages invasion by houndstongue [80]. In areas susceptible to invasion, proper livestock grazing should include altering timing, frequency and level of defoliation to allow a full recovery of desirable grass species. This grazing regime promotes litter accumulation to allow proper nutrient cycling and enhances vigor of desirable grasses which limits invasion by rangeland weeds [50]. For more information on grazing management for weed control see Olson [71].
Weed prevention and control can be incorporated into all types of management plans, including logging and site preparation, management of grazing allotments, recreation management, research projects, road building and maintenance, and fire management. See the "Guide to noxious weed prevention practices" [95] for specific guidelines in preventing the spread of weed seeds and propagules under different management conditions.
Integrated management: The goal of any management plan should be to not only control invasive plants, but also to improve the affected community by maximizing forage quality and quantity and/or preserving ecosystem integrity, and preventing reinvasion or invasion by other invasive species. This must be done in a way that is complementary to the ecology and economics of the site [28,45]. Effective long-term control requires that invasive plants be removed and replaced by more desirable and weed-resistant plant communities [45]. Once the desired plant community has been determined, an integrated weed management strategy can be developed to direct succession toward that plant community by identifying key mechanisms and processes directing plant community dynamics (site availability, species availability, and species performance) and predicting plant community response to control measures [87]. This requires a long-term integrated management plan.
Most often, a single method is not effective for controlling an invasive plant, but there are many possible combinations of methods that can achieve the desired objectives. Methods selected for removal or control of houndstongue on a specific site will be determined by land use objectives, desired plant community, extent and nature of the infestation(s), environmental factors (nontarget vegetation, soil types, climatic conditions, important water resources), economics, and effectiveness and limitations of available control techniques [78].
Managers are encouraged to use combinations of control techniques in a manner that is appropriate to the site objectives, desired plant community, available resources, and timing of application. For information on integrated weed management without herbicides, see the Bio-Integral Resource Center (BIRC) website.
Physical/mechanical: Tillage, hoeing, and hand-pulling may provide effective control of houndstongue, providing these operations are done before the reproductive growth stages to prevent seed production. Mechanical methods may not be practical on rangeland and natural areas, but could be useful in improved pastures or roadsides.
First-year houndstongue plants are difficult to control by aboveground cutting, as the prostrate rosette resists mowing and grazing [99], and nutrient reserves of the taproot acquired during the 1st year are sufficient for normal seed production the following year, even if the plants are completely defoliated early in the spring [11,98]. Furthermore, defoliation at the rosette stage may cause the plant to delay flowering for a year and thus result in a larger plant with a greater seed output [102]. Mowing or clipping 2nd year plants can reduce seed production in houndstongue provided that it is done before seeds are formed and that defoliation is severe enough to prevent regrowth and subsequent flowering [80]. Clipping 2nd year houndstongue plants 0 to 3 inches (0-7 cm) above ground in late June reduced but did not eliminate seed production in houndstongue (Dickerson and Fay 1982, as cited by [26,98]). Sixty percent of cut plants failed to regrow, and seed production of the plants that resumed growth declined to about 25 seeds per plant compared to 364 seeds per plant in the unclipped controls. Boorman and Fuller [11], on the other hand, found that removing the leaves from 2nd year plants had little effect on seed number or seed weight. Additionally, if the flowering stalk is cut off or if flower buds are removed, axillary buds lower on the stem may be activated and develop into cymes; or the plant may respond by forming vegetative side-rosettes from the axils of old leaf-bases [26]. Response of houndstongue after serious defoliation depends on the vigor of plants and the fertility of the site, especially nitrogen availability. Plants with low growth rates respond quite poorly to defoliation, while vigorous plants may recover and set seed [102].
Plowing is said to control houndstongue [99]. However, tillage is not usually appropriate in wildlands and rangelands since it can damage important desirable species, increase erosion, alter soil structure, and expose the soil for rapid reinfestation by houndstongue and other invasive species [62]. Cutting the root crown of either young rosettes or older houndstongue plants (before seed set) 1 to 2 inches (2.5-5 cm) below the soil surface in autumn or early spring and removing top-growth can be effective in controlling small infestations [50,99]. Hand-pulling of houndstongue on the Dunstan Homestead in northeastern Oregon reduced houndstongue populations by 85% [80]. For very large infestations, it may be difficult to get enough labor for cutting or hand-pulling. The Salmon River Restoration Council (SRRC) provides an example of watershed-scale weed control using primarily mechanical control methods and volunteer labor.
Fire: See Fire Management Considerations.
Biological: Biological control of invasive species has a long history; many important considerations need to be made before implementation of a biological control program. The reader is referred to other sources [78,108] and the Weed Control Methods Handbook [94] for background information on biological control. Additionally, Cornell University and NAPIS websites offer information on biological control.
As of 1999, 5 biological control agents were being screened for their potential use on houndstongue. These include a root weevil (Mogulones cruciger), a seed weevil (M. borreginis), a stem weevil (M. trisignatus), a root beetle (Longitarsus quadriguttatus), and a root fly (Cheilosia pasquorum) [50]. Recent research on the host specificity of Mogulones cruciger indicates that this agent can complete full development on several plant species within closely related genera in the Boraginaceae, but prefers houndstongue as a host. This is a matter of concern since at least one species in the genus Cryptantha (C. crassipes) is listed as endangered in the U.S., and 6 of the 12 Cryptantha species tested supported full development of the root weevil (C. crassipes was not tested) [22]. As of the time of this writing, no further information is available regarding the status of the other biocontrol agents. Erysiphe cynoglossi is a commonly encountered pathogen on houndstongue in western North America that is being studied for its impact on vegetative plant growth and reproduction [21].
Chemical: Herbicides are effective in gaining initial control of a new invasion or a severe infestation, but are rarely a complete or long-term solution to weed management [15]. Herbicides are more effective on large infestations when incorporated into long-term management plans that include replacement of weeds with desirable species, careful land use management, and prevention of new infestations. Control with herbicides is temporary, as it does not change those conditions that allow infestations to occur [110]. See the Weed Control Methods Handbook for considerations on the use of herbicides in natural areas and detailed information on specific chemicals.
Picloram, dicamba, chlorsulfuron, metsulfuron and 2,4-D amine can kill houndstongue plants. Repeated applications may be necessary for several years to maintain adequate control [50,59,99]. Herbicide choice and rates are influenced by growth stage, stand density, and environmental conditions (e.g. drought or cold temperatures). Check with state or county weed specialists for appropriate local use rates and timing.
Cultural: No matter what method is used to kill weeds, re-establishment of competitive, desirable plant cover is imperative for long-term control. Fertilization and reseeding with competitive, adapted species is often necessary in areas without a residual understory of desirable plants [78].Houndstongue seedlings have a comparatively low growth rate and are not strongly competitive. Interspecific competition severely reduces the dry weight of 1st and 2nd year houndstongue plants [99]. Generalist herbivores play a positive role in the population dynamics of houndstongue by reducing competition from grasses in coastal dunes in the Netherlands [20,74]. Similarly, in exclosure studies in northeastern Oregon, percent canopy cover houndstongue increased over a 30-year period under grazing pressure from both cattle and wildlife [81]. These studies suggest, therefore, that planting and maintaining competitive species can effectively control houndstongue, although more research is needed.
Houndstongue causes poisoning in horses and cattle [9,14,48,53,90]. Houndstongue contains various pyrrolizidine alkaloids (PAs), whose concentrations are highest during its rosette stage and decrease as the plant matures [9,53]. PAs are known to cause liver damage or failure in livestock [90]. Kedzie-Webb and Sheley [50] suggest that PAs are toxic to horses and cattle but not to domestic sheep. Poisoning can occur when houndstongue is cut and dried with harvested hay, or when animals are confined to a small area lacking desirable forage. Most livestock poisonings occur from ingestion of contaminated hay or feed [90]. The acute or chronic nature of poisoning depends on the PA concentration, amount eaten, and rate of ingestion [9,50]. Any level of houndstongue contamination in feed should be considered potentially lethal for all livestock [90].
Kufeld and others [57] report light use of houndstongue by Rocky Mountain mule deer in winter in Montana. Domestic sheep commonly graze houndstongue leaves [26]. There are few direct effects of herbivores on houndstongue where it occurs on coastal dunes in England and the Netherlands. It is not normally eaten by rabbits [11], although rabbits have been observed digging up taproots in winter [26]. A 3-year study by Prins and Nell [74] indicates only low levels of leaf herbivory on houndstongue in coastal dunes in the Netherlands. Rabbits caused some leaf damage in early spring, and no root consumption by rabbits was found. From June to November, larvae of the oligophagous Lepidopteran, Ethmia bipunctella, are the most important herbivores on houndstongue. Captive mice eat nutlets, but it is unknown whether this occurs in the field. Ring-necked pheasants graze on the cotyledons and on whole seedlings of houndstongue (review by [26]).
Palatability/nutritional value: Green houndstongue plants have a distinctive odor that discourages animals from eating it, but when dried it becomes more palatable [9,53].
Cover value: No information
In British Columbia, houndstongue occurs predominantly in the Interior
Douglas-fir (Pseudotsuga menziesii) and ponderosa pine (Pinus
ponderosa)-bunchgrass biogeoclimatic zones [99].
In Utah, houndstongue may be found in
sagebrush (Artemisia spp.), pinyon-juniper (Pinus spp.-Juniperus
spp.), cottonwood (Populus spp.),
mountain brush, quaking aspen (Populus tremuloides), ponderosa pine, and spruce-fir
(Picea spp.-Abies spp.) communities [107]. It is a minor component in
Gambel oak
(Quercus gambelii) communities in central and northern Utah [58]. On preserves in Colorado, houndstongue has been reported in
shortgrass prairie, narrowleaf cottonwood/red-osier dogwood (Populus
angustifolia/Cornus sericea) riparian forests, and riparian meadows.
In northeastern Oregon, houndstongue invades Idaho fescue (Festuca
idahoensis), bluebunch wheatgrass (Pseudoroegneria spicata), and
prairie Junegrass (Koeleria macrantha) grasslands, and riparian habitats
[80].
Houndstongue seeds overwinter predominantly in the top (1 cm) of soil, although some seeds may remain attached to the parent plant throughout the winter [99]. Houndstongue germination starts in late winter and early spring, when soil temperature rises above the freezing point. Houndstongue plants then develop from seedling to rosette within the 1st year of growth. When night frosts begin in autumn, rosette leaves die back and the taproot remains. Houndstongue requires a vernalization period before stem elongation can occur. Stem elongation is probably stimulated by day length [26]. Flowering is strongly size-dependent, with the probability of flowering increasing with the size of the plant. There may also be a critical size below which plants do not flower [27]. The main period of flowering is about 55 days, with seed ripening requiring an additional 70 days. Most ripened seeds fall to the ground within 4 months of ripening [99].
Typical flowering dates are reported by area as follows:
Area Flowering dates Reference Carolinas May-July [76] Great Plains May to July [37] IL May-July [67] Intermountain West May to July [17] IA May-June [54] MT mid-June [50] NM May-August [63] BC May-July [98]Although usually a biennial, houndstongue can behave as a short-lived perennial. Under fertile conditions plants typically flower in their 2nd year. Under adverse environmental conditions, flowering can be delayed until the plants reach a sufficient size for vernalization [26]. Delay of flowering is common in natural populations of houndstongue on Dutch coastal dunes. Over a 7-year period, the percentage of rosettes that flowered ranged from 2 to 25%. The proportion of rosettes flowering was positively correlated (r=0.62, p=0.05) with rainfall in the previous year and was reduced after a very cold winter [27].
While most houndstongue plants die after flowering, houndstongue is not strictly monocarpic. Of 55 houndstongue plants studied, 7 flowered again in both the 2nd and 3rd years and 2 in their 3rd and 4th years [11]. Repeated flowering has also been observed in houndstongue not only after damage to the flowering plant, but also in undamaged plants under both natural and garden conditions [26].
Houndstongue is a biennial or short-lived perennial that reproduces only by seed. An annual life cycle apparently does not occur anywhere in the entire range of the species [26]. The following discussion of the regeneration processes in houndstongue derives primarily from research in British and Dutch coastal dune areas. More research is needed on houndstongue regeneration on sites where it is invasive in North America.
Breeding system: A review by Upadhyaya and others [99] suggests that houndstongue flowers are perfect (having both male and female parts), and that seed production occurs via autogamy. However, de Jong and others [26] report that houndstongue plants produce only very few nutlets if pollinators are excluded, and that bagging inflorescences in the field generally results in reduced seed set per flower. This suggests some mechanism that prevents self-pollination within the flower. Seed set per flower is greatly increased by exchanging pollen between flowers on the same plant.
Pollination: Houndstongue is pollinated by bumblebees [75]. Flowers are also visited by other bees, thrips, and butterflies [26].
Seed production: Estimates of total seed number per plant in houndstongue range from 50 to more than 2,000 [99]. In Montana, houndstongue seed production ranges from 300-675 seeds per plant [50].
Houndstongue plants typically have 1 or 2 stems, although up to 8 stems per plant have been reported on plants growing in dune grasslands in England. In this ecosystem, a single stem could produce around 300 seeds, and seed number per stem was correlated with stem basal diameter (r=0.84, p<0.01) [11]. In a sand dune area in the Netherlands, larger plants produce more seed per flower, and smaller plants produce more flowers per plant [51]. Researchers there found that watering houndstongue plants had a positive effect on seed set [26].
Each flower has the potential to produce 4 nutlets. A review by Upadhyaya and others [99] reports an average 2.75 nutlets per flower. In Dutch sand dunes, 5.2 to 7.8% of the nutlets produced in a population formed seedlings in the following spring [26]. Boorman and Fuller [11] estimated this figure at 25% on coastal dunes in England.
Seed dispersal: Houndstongue seeds are covered in a spiny husk and possess a protruding barb that enables the seed to adhere to wild and domestic animals thus promoting long-distance dispersal [50]. European studies, however, suggest that animal dispersal is rare in houndstongue [11,26], and wind is considered to be the primary dispersal mechanism [100]. In one study, the majority of houndstongue seeds (75%) fell into an area of radius 5 inches (12 cm) around the parent plant. The greatest recorded dispersal distance was 4.6 feet (1.4 m). With such limited primary dispersal range, even rare dispersal events by animals could be important [11].
It has been proposed that the presence of houndstongue around rabbit warrens in European ecosystems is the result of rabbits cleaning seeds off their fur before entering the warrens. However, seeds do not adhere well to fine rabbit fur [11]. Evidence from a study in British Columbia indicates that cattle are important dispersers of houndstongue seed, picking up about 65% of seeds per stalk in grazed paddocks [23]. Some houndstongue seeds remain on plants well into the winter [11,23,26]. These seeds are dispersed slowly over time by attaching to animal wool and hair. Dispersal via streams and irrigation ditches is unlikely due to the high specific gravity of houndstongue seeds [99]. Although seed dispersal occurs slowly over time, colonization of disturbed sites can take place very quickly [50].
Seed banking: A review by Upadhyaya and others [99] cites several studies that indicate that houndstongue does not produce a large, persistent bank of buried seeds, and that seeds remain viable no longer than 2-3 years. Newly ripened seeds of houndstongue exhibit an innate dormancy [26,99]. Most seeds fall from the plant soon after ripening; however, some seeds remain on plants for up to 2 years, especially in sheltered habitats (e.g. scrub), thus creating an above-ground seed bank. A fraction (less than 5 to 10% in exposed habitats) of seeds that fall to the ground do not germinate in the 1st year [11,100]. A small percentage of these seeds (4-7 %) may remain viable after 1 year [26]. In the Netherlands, viable seeds were found to occur almost entirely within the top 0.4 inch (1 cm) of soil, and viable seeds did not occur deeper than 2 inches (5 cm). Experimentally, all buried seeds had germinated within 9 months at both 0.8-inch (2 cm) and 6-inch (15 cm) depths [100].
Germination: Seed burial depth, ambient temperature, moisture, soil nitrate levels, and light have been shown to affect seed germination in houndstongue (reviews by [26,99]).
The viability of freshly harvested houndstongue seeds exceeds 90%, although newly ripened seeds of houndstongue exhibit an innate dormancy [26,99]. Thus, houndstongue seeds remain ungerminated throughout the fall and winter and germinate uniformly in the spring, suggesting a vernalization requirement for dormancy release. In Dutch coastal dunes, almost all houndstongue seedlings emerge in March and April, the main germination period being March, with very little to no germination in any other month [11,100].
Germination in houndstongue generally requires some form of scarification or softening of the seed coat. Seeds are likely to remain dormant until they can take up sufficient water [11]. Nitrate also breaks seed dormancy [11,100]. In laboratory studies, germination rate was highest after 6-12 weeks of moist-chilling at 32 to 50 degrees Fahrenheit (0-10 ºC). Germination of dispersed seeds depends strongly on burial. Of the viable houndstongue seeds buried 0.4 inch (1 cm) in sand, almost 100% germinated before and during the 1st spring, while germination on the surface was less than 4% [26,100]. Darkness may either stimulate or have no effect on germination of houndstongue seeds [100].
Seedling establishment/growth: Large temporal and spatial variation in the number of surviving seedlings per reproductive individual has been observed for houndstongue [24]. It is estimated that less than 1% of dispersed seed survives to produce more seed (review by [26]). Reports of 7 to 40% of dispersed houndstongue seeds were retrieved as seedlings in spring. The remaining seeds may have been dispersed secondarily or lost through predation, although death of seeds or seedlings immediately after germination was the most plausible explanation [24,75]. First year survival of seedlings from spring to fall ranged from 22 to over 50% on Dutch coastal dunes [24].
Houndstongue seedlings have a low growth rate and are not strongly competitive. Plants do well when their competitors are kept in check (e.g. grazed by rabbits) [11]. Small-scale disturbances are important for houndstongue establishment. Houndstongue occurs only in areas of the Dutch coastal dunes where grazing by horses and cows is allowed [20]. Prins and Nell [74] compared population dynamics of houndstongue with and without protection from generalist herbivores (insects, rabbits and mice). Houndstongue mortality was significantly higher in protected populations for 3 consecutive years (p<0.05). Germination, seedling establishment and number of flowering plants were also lower inside exclosures. Similar results were observed in exclosure studies in northeastern Oregon, where percent canopy cover of houndstongue increased over a 30 year period under grazing pressure from both cattle and wildlife [81]. Generalist herbivores seem to play a positive role in the population dynamics of houndstongue by reducing competition from grasses [74]. Seedling growth and survival are also reduced by intraspecific competition at high population densities [26].
Seeds of houndstongue have large nutrient reserves that allow rapid seedling root growth so that plants can be well established before drier weather begins (reviews by [26,99]). Most seedling mortality, however, is attributed to water deficit in the early summer months, before the development of a deep and extensive root system [11,20]. Seedling survival is lowest in dry summers and on exposed sites [24]. Seedling mortality was negatively correlated with the water content in the top 4 inches (10 cm) of soil (r=-0.65, p<0.001) in Dutch coastal dunes. Early seedling mortality (April to May) was also negatively correlated with the cover of shrubs and trees (r=-0.70, p<0.001) and soil humus content, (r=-0.63, p<0.01). Variation in rainfall may cause yearly differences in houndstongue survival and growth, and also affects its distribution over shaded and exposed habitats on dune sites [25]. A large proportion of houndstongue rosettes that delayed reproduction buffered the population against low seedling recruitment in some years (a bud bank versus a seed bank) [24].
Asexual regeneration: Houndstongue reproduces only by seed.
Houndstongue is an opportunistic species that exploits conditions suitable for its establishment and growth. In the variable environment of dune habitats, houndstongue populations are usually in decline, and only in rare years are conditions favorable for growth and/or seed production. For this reason, one may see large fluctuations in population density in a single locality [24]. As is typical of biennials, large population increases are observed following reproductive success and large decreases or even local extinctions following reproductive failure. Annual disturbances as well as absence of disturbance can be fatal for biennial plant development; therefore most habitats are only temporarily suitable for biennials. Grazed range provides an environment where gaps are repeatedly created and therefore suitable sites for establishment are usually available [101]. Where it has established on disturbed sites such as roads and around old buildings, it may persist indefinitely, as is evidenced by its continued presence in abandoned mining towns in southwestern Montana, even after 45 to 77 years of recovery [52].
Houndstongue is shade tolerant [98] but grows best in full sunlight, if sufficient water and nutrients are available. In sand dunes it may occur in both open and shaded sites (5-10% of full daylight). Its establishment in open sites may depend on spring and summer rainfall. In half-open habitats, at the edges of the canopies of shrubs or trees, plants are less susceptible to drought [26]. Houndstongue was significantly (p<0.05) positively associated with closed canopies at Mammoth campground in Yellowstone National Park. Here it was more consistently found under high canopy cover than any other nonnative species. Similarly, Lacey and Lacey [59] describe occurrences of houndstongue in areas of thick litter accumulation (as might be found under a forest with high canopy cover) [1].
Cynoglossum officinale[1][2] (houndstongue, houndstooth, dog's tongue, gypsy flower, and rats and mice due to its smell) is a herbaceous plant of the family Boraginaceae.
It can be either annual or biennial. Leaves are greyish and softly haired, lanceolate to oblong. Reddish-purple funnel-shaped flowers bloom between May and September.
The plant owes its common and scientific name to the long greyish leaves that are reminiscent of a dog's tongue and were once given as a remedy for dog bites.[3]
Found in most parts of Europe, and also North America, where it was accidentally introduced[4] including in British Columbia, where it is designated a noxious weed under the British Columbia Weed Control Act.[5] It lives in wet places, sand dunes, waste land and hedges.
Houndstongue may be pollinated by bees, and may also self-pollinate.
The name houndstongue (and the latin genus Cynoglossum) comes from the roughness of the leaf.
In 1725, houndstooth was presented in the family dictionary, Dictionaire oeconomique, as part of a cure for madness.[6] In that book, madness was viewed as "a distemper, not only of the understanding, but also of the reason and memory, proceeding from a cold, which drys up everything it meets with that is humid in the brain."[6] To cure madness, Dictionaire oeconomique noted:
You must shave the head of the unhappy patient, and after that, apply to it a pidgeon, or a hen quite alive; or else bathe it with some brandy distilled with rosemary, elder, hounds tooth, and the roots of bugloss, or with the oyl of elder flowers: they rub their heads and wash their feet with a decoction of the flowers of camomile, melilot, balm gentle and laurel; they put into their noses the juice of comfrey, with either two or three spoonfuls of honey-water, broth, or white-wine, wherein wormwood and sage are infus'd ; or else they do for five and twenty days together, mix with their broth in the morning, halt a dram of the ashes of tortoise, and they put into the pot bugloss, borage, with a pinch of rosemary to season it.[6]
In the 1830s, houndstooth was known in France to be made into an emollient and diuretic for daily use in inflammatory diseases, especially of the urinary organs.[7] To prepare as a diuretic, the houndstooth leaves were mashed, and then boiled in water to extract oils, volatile organic compounds, and other chemical substances.[7] The mix could be sweetened with liquorice to create Ptisan of Dog's-grass.[7] After decoction, the herbal tea was taken internally a cupful at a time.[7] In 1834, the Hospital of Paris provided a formula of 2/3 ss—J to Oij of water for houndstooth tea.[7] By the end of the 1830s, doctors in England were using houndstooth as an antiaphrodisiac to combat venereal excesses.[8]
Herbalists use the plant for piles, lung diseases, persistent coughs, baldness, sores, and ulcers but the effectiveness of all these uses is not supported by any scientific evidence.[9]
In 1891, the U.S. state of Michigan identified houndstooth, along with flea-bane, rag weed, burdock, cockle-bur, and stickseed, as some of the worst weeds in the state.[10]
Cynoglossum officinale contains tumorigenic pyrrolizidine alkaloids.[11] It is toxic to cows and is especially dangerous to pasture owners.[4]
Flowers and leaves Cynoglossum officinale (houndstongue, houndstooth, dog's tongue, gypsy flower, and rats and mice due to its smell) is a herbaceous plant of the family Boraginaceae.
