Hawkweeds in the Northwestern United States

Home , Cichorieae, Cichorium, Hieracium, Lactuca, Pilosella, Pilosella officinarum

18 RANGELANDS 19(4), August 1997 Hawkweeds in the Northwestern United States

Linda M. Wilson, Joseph P. McCaftrey, Paul C. Quimby, Jr., and Jennifer L. Birdsall

everal species of hawkweed (Hieracium spp.) are re- being genetic forms of a "related" species (Hitchcock et al. cent additions to the weedy flora of the Inland and 1955) S Pacific Northwest, the region encompassing The genus is divided into 4 subgenera, 3 of which are Washington, northern Idaho and northwestern Montana. found in north-temperateflora and are distinguished by the Over 60,000 acres are currently estimated to be infested. following key (adaptedfrom LePage 1971): Private landownerconcern about the rapid rate of spread of 1a. South Americanflora Mandonia hawkweeds is shared land wildlife by public managers, lb. North temperateflora managers, timbercompanies, livestock farmers, and recre- 2a. Plants normally stoloniferous (except. ationists. Lack of control of these weeds has become a se- H. Leaves rious land piloselloides Vill.): always managementconcern in the last decade. Various basal; involucreformed of two or three chemical and cultural methods have successfully controlled series of imbricatedbracts small Piosella infestations, but larger, less accessible populations, 2b. Plants not stoloniferous: comprising the bulk of the infestation, are more difficult to 3a. Involucreformed of more than two control. series Biological control is being investigated as a tool of bracts Hieracium with which to unequal length combat the increasing hawkweed problem. 3b. Involucre formed of two series of The of this is to summarize purpose paper whatis currently unequal length bracts Stenotheca known about the introduced hawkweeds in the United There are considered to be about 36 of States in order to facilitate future research and manage- generally species ment of hawkweeds. Hieracium belonging to 3 subgenera in America north of Mexico. About 25 species are native to the United States and Canada and fall into 2 Taxonomy subgenera, Hieracium (3 species) and Stenotheca (22 species) (Table 1). Subgenus The hawkweeds (Hieracium spp.) are members of the Hieracium is circumboreal in distribution, and is represented in North tribe Lactuceae (= Cichorieae) in the family Asteraceae. America by H. umbellatum L., H. robinsonii The name Hieracium is derived from the Greek hierax, Zahn and H. canadense Michx. Species in this subgenus meaning hawk; allegedly keen-sighted hawks of yore ate reproduce sexually and apomictically. Subgenus the sap of the brightly colored plants to sharpen their Stenotheca is restricted to the New World and contains eye- most of our native sight (Fernald 1950). Hieracium is an extremely diverse taxa. Species in this subgenus are strict- genus morphologically.It includes a huge number of taxa in ly sexual. Conflicting characters used to distinguish species which apomixis, hybridization, and are all com- have generated considerable debate regarding classifica- polyploidy tion. For mon (Stebbins 1950). Apomixis, or asexual reproduction, example, Guppy (1978) suggestedthat H. scouleri includes both agamospermy (production of seeds without Hook., H. albertinum Farr., and H. cynoglossoides L. are too pollen) and vegetative reproduction. These reproductive closely correlated in characters to be distinguished into and strategies generate enormous morphological and cytologi- separate species, instead form a complex of species. cal variation within and among species, making Hieracium Similarly, Guppy (1978) suggested that H. umbellatum L., one of the most taxonomically difficult genera. Problems and its close relative H. canadense Michx., may be 2 forms of the same with taxonomy are worsened by the lack of clearly defined, species. Deardorff (1977) added H. longiberbe stable characters that are important in classifying species Howell and H. nudicale Heller to the H. scouleri complex. The hawkweed within the genus. For example, species are broadly distin- remaining species found in the United guished by the character and amount of leaf States and Canada are introduced from Europe and belong pubescence, to 2 venation, flower color, stolon morphology, and the size, subgenera (Table 2). Four species belong to the sub- Hieracium: H. murorum H. shape, and color of the inflorescence (Bailey and Bailey genus L., argillaceum Jordan 1976, Voss and Bählke 1978). The result has been the de- group, H. sabaudum L., and H. groenlandicum Arv.-Touv. Plants in this scription of a huge number of species, subspecies, vari- subgenus have not become weedy in North America. The eties, and forms that differ slightly (Sell 1974). Zahn (1921- remaining 7 species belong to the subgenus 22) conducted the most comprehensive treatment of the Pilosella.As in other subgenera,these plants are difficult to genus in which he described 756 species world-wide. It is distinguish morphologically. From a practical standpoint, in estimated that while over 10,000 Hieracium species, sub- however, plants subgenus Pilosella fall into 3 groups. species, varieties and forms have been described world- The first group is represented by H. pilosella, mouse-ear wide, only about 700 species are valid, with the remainder hawkweed, easily distinguished by its single, yellow flower. RANGELANDS 19(4), August1997 19

Table 1. List of Hieracium native to America north of species Species Subgenus Mexico,by subgenus (H=Hleraclum,S= Stenothoca). = H. idahoense Gand. H. nudicaleHeller Species Subgenus = H. paddoense Gand. H. canadense Michx. (Canada h.) H = H. washingtonense Gand. = H. columbianum Rydb. = H. cascadorum Zahn. = H. umbellatum var. scathisculum Farw. = H. chelannense Zahn. = H. laevigatum ssp. canadense Zahn. = H. chapacanum Zahn. H. robinsonji (Zahn) Fern (Robinson h.) H = H. scoulen var. scouleri = H. scoulen var. nudicale H. umbellatum L. (narrow-leaved h.) H (Gray) = H. scabrisculum Schwein. H. traillii(Green's h.) S = H. canadense var. scabrum = H. greenei Porter & Britton (? Gray) = H. macranthum Nutt. = H. howellhiGray. = H. canadense var. angustifolium 1. & G. = H. bathigerum Greene = H. kalmii = H. oregonicum Zahn. = H. suksdorfiiGand. = H. cinereum Howell. H. albiflorum Hook. (whitef lowered h.) S H. triste Wilid. (alpine h.) S = H. vancouvenanum Arv.-Touv. H. venosum L. (rattlesnake weed) S = H. hefleriGand. = H. leptopodanthum Gand. = H. candelabrum Gand. While predominantly an eastern weed, H. piosella has a = H. albiflorum subvar. rosendahlii Zahn. limited western distribution in coastal Washingtonand = H. albiflonimf. lyalliiZahn. Oregon (Roche 1992). The second group includes H. au- H. albertinum Farr (western h.) s rantiacum L., orange hawkweed, the only introduced = H. absonumMacbr. species with orange flowers. The remaining group consists = H. scoulenvar. albertinum of species with multiple, yellow flowers and includes H. H. argutum Nutt.(S. Californiah.) s pratense Tausch. (= H. caespitosumDumort), H. praealtum H. bolanden Gray.(Bolander's h.) Vill., H. piloseioides (= H. florentinum All.), H. floribundum = H. siskiyouense Peck. Wimm. & Grab., and H. flagellare Wilid. (= H. auricula L.). H. cameum Greene s Known variously as yellow hawkweed,meadow hawkweed, field devil and devil H. cynoglossoides L. (houndstongue h.) hawkweed, king hawkweed, yellow = H. griseumRydb. hawkweed, this group represents most of the taxonomic = H. cusickilGand. complexities. = H. flettliSt. John & Warren Until recently, it was thought that only H. pratense (Figure = H. panyiZahn. 1 occurred in the Inland Northwest, however, H. flonbun- = ) H. iydbeipiiZahn. dum and H. pioselloides have also been identified in the = H. scoulen var. gnseum (Rydberg) past 6 years. Other populationscontain plants that are mor- H. fendleriSchultz (Fendler h.) S phologically similar to H. pratense and H. pilose/loides. H. gracile Hook. (slender h.) S Lepage (1967) suggested that H. pioseioides is a hybrid = H. hookenSteud. between H. praealtum and H. flagel/are. Similarly, he treat- = H. tnstevar. detonsum ed H. flage/lareas a hybrid between H. pratense and H. p1- H. gronoviiL. (hairy h.) S lose/Ia (Lepage 1967, 1971). Guppy (1976) suggestedthat H. homdum Fries. (shaggy h.) 5 H. floribundumand H. pratense, distinguishableonly by the = H. brewenGray degree of pubescence, may be different phases of a single H. lemmoniiGray species that exhibits a broad morphologicalrange. Lepage H. longiberbe Howell (long-beaked h.) S (1967) recognized the largest number of hybrids among the = H. piperiSt. John &Warren introduced hawkweeds, particularly between H. auranti- = H. scoulen var. longiberbe acum, H. pilose/la, H. floribundum and H. pratense. It is H. longipilum Tori. (long-bearded h.) 5 likely that the combined influences of morphologicalvaria- H. mananum Willd. (Maryland h.) tion, polyploidy, frequent apomixis, and widespread hy- dridization the introduced has led to a H. paniculatum L. (panicled h.) among species huge of intermediate and that distinctions H. array types, species pringlelGray. (woolly h.) can no longer be made in North America based solely on H. rusbyiGreene 5 descriptionsfrom Europeanflora. H. scabrum Michx. (rough h.) S H. scouleri Hook. (Scouler's h.) s Biogeography = Pilosella scoulen Schu. & Schu.-Bip. = H. amplumGreene H. was introduced into the United States in = pratense likely H. absonum Macbride & Payson 1828 (Britton and Brown 1970), and is now commonly = H. cinereum Howell found from Québec and Ontario (Frankton and Mulligan 20 RANGELANDS 19(4), August1997

Table 2. List of Hieracium species Introduced into America north of Mexico,by subgenus (P=PIIoseIIa, H=Hleraclum).

Species Subgenus H. aurantiacum L. (orange h.) P = H. brunneocroceum Pugsley. H. flonbundum Wimm. & Grab. P (yellow devil h., glaucous hawkweed) H. ilagellare Willd. (whiplash hawkweed) P = H. auncula L. = H. factucella Wailr. H. pilosellaL. (mouse-ear h.) P H. pitoselloides Viii.(King devil h.) P = H. florentinum All. = H. praealtumVill. H. praealtumVill. (King devil h.) P H. pratense Tausch (meadow h. field h., King devil) P = H. caespitosum Dumort (YellowKing devil) H. argilfaceum Jordangroup (European h.) H = H. Iachenalii Gmelin (common h.) = H. vulgatum Fries (common h.) = H. mollePursh. = H. stnimosum L. H. muronimL. (wall h., golden lungwort) H H. groen!andicum Arv.-Touv. (Greenland h.) H H. sabaudum L.(hawkweed) H

1970) southward to Georgia and Tennessee (Rickett 1987). The first record of H. pratense in the western United States was in Pend Oreille County, Washington, in 1969 (Marion Ownbey Herbarium, Wash. State Univ.). H. aurantiacum was introduced into Vermont in 1875 as an ornamentaland within 25 years spread throughout much of northeastern United States and southeastern Canada (Voss and BOhIke FIg. 1. Meadowhawkweed (Hieracium pratense) 1978). First recorded in Spokane, Washington, in 1945 (Marion Herbarium, WSU), H. aurantiacum has Ownbey In H. and H. aurantiacumare now been collected from gardens in Nevada County, California Idaho, pratense widely distributed the 6 northern and is re- (Munz and Keck 1959), on the east slope of the Rocky throughout counties, Mountains in Colorado coastal and portedly spreading south. In Washington, H. pratense and (Weber 1990), Oregon H. floribundum Washington (Rickett 1987), and in southwestern British are found primarily west of the Cascade Mountains as well as in the 4 northeastern counties. Columbia (Guppy 1976). H. aurantiacum is more wide- Montana H. H. and H. spread than other hawkweed species primarily because it is reports aurantiacum, pratense, p1- as an ornamental and often cultivation. lose/bides variously distributed in 5 northwestern counties planted escapes in addition Both H. pioselloides and H. floribundumwere first to both Glacier and Yellowstone National Parks reported Glacier National in northern New York 1879 (Voss and BOhIke 1978) and (Robin Cox, Vegetation Manager, Park, 1900 respectively (Kennedy 1902). pers.comm.) Introduced hawkweeds infest similar habitats in and Washington, Idaho, Montana, and British Columbia. Found Morphology Biology predominantly in mountain meadows and clearings in forest zones, hawkweeds also infest and Hawkweeds are perennial, stoloniferous plants that con- permanent pastures tain a the introduced H. hayfields, cleared timber units, and abandoned farmland milky sap. Among species, only pilose/bides is not stoloniferous. Each consists where the soil is well drained, coarse-textured,and moder- young plant low in matter. are found at elevations of a rosette of 5 to 12 hairy leaves arising from a short, ately organic They thick with a fibrous root from 1,500 feet to over 5,000 feet, with the rhizome, shallow, system. Rosette ranging largest leaves are to entire to dentate. Plants infestations occurring around 3,300 feet. None of the intro- oblong elliptical, pro- duced are found in the natural or shrub- duce 1 to several flowering stalks with 5 to 20 ligulate species grasslands heads in Cauline if steppe of the northern lntermountain West, and are not ex- arranged compact corymbs. leaves, pre- sent, are much smaller and are pected to become problem weeds in any dry habitat usually arranged alternately along the lower half of the stalk. The flowers are yellow or associated with western rangelands. orange (H. aurantiacum).The involucre is cylindrical, with unequal RANGELANDS 19(4), August1997 21 bracts arranged in 2 or 3 imbricated rows. Achenes are (Bishop and Davy 1994) and H. aurantiacum (Stergios spindle-shaped,slender, and have a pappus of a single cir- 1976). cle of slender, tawny bristles. Hawkweeds have long been used in the study of apomix- Aspects of the biology and ecology of introduced hawk- is and cytology. Having a base chromosome number of 9, weeds have been studied in the United States, Canada and polyploids range from triploid (2n = 27) to decaploid (2n = New Zealand. Seeds that germinate in the spring have a 90). The introduced species are typically triploid (2n=27), higher survival rate than fall germinated seeds (Panebianco tetraploid (2n = 36) or pentaploid (2n = 45) (Skalinska and Willemsen 1976). Aftera rapid period of rosette growth 1967). Skalinska and Kubien (1972) reported that while all in the spring and early summer, and following sufficient ex- of the polyploid H. pratense they examined from Poland posure to long days (Veung and Peterson 1972), plants were apomictic, the diploid biotypes (2n = 18) were strictly flower in late June and form viable seed shortly after anthe- sexual. Deardorff (1977) stated that H. pratense is sexual, sis. Under favorable conditions, a single generationcan be but he did not reportthe geographic origin of the specimens completed in about 4 months. The plant will overwinteras a he collected. Diploids have not been reported for the other rhizome and regrow the following spring. In greenhouseex- introduced hawkweeds. periments conducted in 1996 by the senior author, plants The tribe Lactuceaeis well known for its diverse flavonoid flowered an average of 61 daysfollowing germination. chemistry. In the genus Hieracium, umbelliferone and a Long-distance transport of seeds is likely through the ac- number of other phenolic compounds have been identified tivities of birds and animals (including man). In field experi- and are known to have phytotoxic properties (Dawes and ments with H. aurantiacum, Stergios (1976) observed that Maravolo 1973). Some researchers have even suggested about 95% of seeds fell within 3.5 feet of the patch, and that a chemotaxonomic treatment of the genus is valid thus were not carried far by the wind. While seeds are based upon the high degree of correlation between chemi- physiologically ready to germinate in the summer following cally derived characters (Bate-Smith et al. 1968). In New a short period of afterripening, dry soil conditions in the Zealand, Makepeace (1976) studied the chemical ecology summer and cool temperatures in the fall cause most of mouse-ear hawkweedand attributed much of its compet- seeds to delay germination until the following spring itive ability to allelopathy. (Panebiancoand Willemsen 1976). Stolons develop only after flowering is initiated (Yeung Control and Peterson 1972). Each of the 1-6 leafy stolons arises from axillary buds at the base of rosette leaves. Depending A number of studies in both North America and New on plant health, time of year, and environmentalconditions, Zealand have addressed problems with hawkweed control. stolons will either terminate with rosettes of leaves or devel- The use of herbicides,while effective, is limited to relatively op into a secondary flowering stalk (Thomas and Dale accessible sites. Herbicides such as 2,4-D, clopyralid, and 1976). If a new rosette is formed, it will establish roots, and picloram have resulted in the greatest degree of control as the stolon degeneratesthe rosette becomes an indepen- when applied at the normally recommended rate for peren- Lass and Callihan dent plant. nial pasture weeds (Noel et al. 1979, The stoloniferous habit of hawkweeds results in the for- 1992). Studies conductedat the University of Idaho showed mation of dense mats of weeds that can persist for 30 that over 50% control was achieved for 6 years following years and reach densities of 3,500 plants m2 (Thomas and treatment with a one pint rate of clopyralid (Lass and Dale 1975). In H. pioselloides, axillary buds develop into Callihan 1992). Other herbicides either failed to control secondary flowering shoots rather than stolons. Another hawkweed or suppression was achieved for fewer than 3 means of vegetative reproduction by introduced hawk- years (Lass and Callihan 1992). weeds is by adventitious root buds. Plants derived from Marked reductions in hawkweed density and vigor have root buds have been reported for H. pratense (L. M. Wilson been obtained by fertilizer treatments in the United States and J. P. McCaffrey, unpubl. data) and H. florentinum (Reader and Watt 1981), Canada (Hay and Ouellette (Petersonand Thomas 1971). 1959), and New Zealand (Scott et al. 1990 a, 1990b). The proportion of flowering plants in the population, the However, Reader and Watt (1981) found that repeated fer- number of flower heads, seed production,the number and tilizer application had no effect on dense patches of hawk- length of stolons, and leaf size are all highly density-depen- weed that contained few grasses or otherforbs. It appears dent (Thomas and Dale 1975, Stergios 1976) Thomas and that the best results of hawkweed control are achieved Dale (1975) showed that only 10% of the rosettes in the where fertilizer and herbicidesare applied together. center of a patch flowered in a given year, and that more Soil analyses from hawkweed infestations in British plants on the periphery of the patch flowered than at the Columbia indicated that soils there are low to deficient in center. Moreover,they reportedthat only 1% of the rosettes sulfur. Similar analysesfrom Benewah County, Idaho, while in an established population are derived from seedlings, preliminary, showed that upland meadow soils are univer- and concluded that vegetative reproduction is more impor- sally deficient in sulfur, whether inside or outside a hawk- tant than seeds to the maintenance of older patches. weed patch. However, the Idaho studies showed that soils Similar phenomena have been reported for H. pilosella under hawkweed are low in organic matter (L. Wilson and RANGELANDS 19(4), August 1997

P. J. McCaffrey, unpubi. data). More detailed analyses are combination with interspecific competition and variable fer- needed to characterizethese soil-hawkweedrelationships. tility. Additional research is planned to conduct hybridiza- Most of the hawkweed infestations in the Inland tion studies and to examine polyploidy in from select- Northwestoccur in areas plants with limited access where the use ed populationsaround the region. of herbicides and fertilizers is impractical. These areas may be better suited to classical biological control. Interest in Summary such a program has spread quickly in the Inland Northwest and has resulted in the formation, in 1992, of the Moist, humid regions in the northwestern United States Hawkweed Action Committee, Inc. in St. Manes, Benewah are likely to be highly susceptibleto invasion by the aggres- County, Idaho. Explorationfor insects attacking hawkweeds sive, rapid-spreading hawkweeds. The spread of intro- in their native ranges is being conducted by USDA-ARS duced hawkweeds in the Inland Northwest during the past scientists in and France, the International Institute of decade has alerted land managers to the need to develop Biological Control (IIBC) in Delémont, Switzerland. long-termmanagement strategies to limit further hawkweed One of the major constraints often limitingthe introduction spread and to control the larger areas of infestation. The of one or more natural enemies is host specificity. It must tools available to accomplish this include biological, chemi- be shown that insects selected for biocontrol feed only on cal, and cultural control. Cultural practices include fertiliz- the target weed. Host specificity problems may arise with ing, good grazing practices, early detection and control of hawkweed since the native flora is replete with Hieracium small hawkweed infestations. Good land stewardship prac- species, some of which are threatened and endangered. tices will be imperativeto ultimately manage the introduced Furthermore, closely related genera such as Crepis, hawkweedsat an acceptable level. Taraxacum, Tragopogon, Cichorium, and Lactuca are well Biological control of hawkweeds may be more easily represented in the native flora and may further complicate achieved with natural enemies that attack stolons and/or host specificity testing. However, since most of the intro- are specific to the subgenus Piosella. If a biological control duced hawkweeds are stoloniferous and belong to the sub- program is established, it will be a valuable component of genus Pilosella, potential biological control agents need an integrated program of hawkweed managementbecause only be specific to the subgenus level. Insects that attack it can impact remote and inaccessibleinfestations. stolons would be particularly beneficial as these structures are specific to most of the species in subgenus Pilosella. ReferencesCited Flower head insects may be less important in a program for biological control of hawkweeds because sexual Bailey, L.H. and E.Z. Bailey(eds.). 1976. Hortus Third. McMiIlan reproduc- New York. tion has been shown to contribute little to the Publishing, populationdy- Bate-Smith, E.C., P.D. Sell, and C. West. 1968. and namics of these weeds, and the risk of native Chemistry attacking taxonomy of Hieracium L. and Pilosella Hill. Phytochemistry hawkweedsmay be greater. 7:1165-1169. Plant taxonomy and biology studies will be to Bishop, G.F. and A.J. Davy. 1994. Hieracium pilosella L. required officinarum properly address host specificity testing to ensure the accu- (Pilosella F. Schultz & Schultz-Bip.). J. Ecol. rate evaluation of insect-host associations. 82(1):195-210. Pre-release sur- Britton, N.L. and A. Brown. 1970. An illustrated flora of the north- veys were conducted by University of Idaho scientists dur- ern United States and Canada. Vol. III: Gentianaceaeto ing 1 994—i 995 to identify insects associated with the native Compositae.Dover Pubs., N.Y. and introduced hawkweeds in northern Idaho Dawes, D.S. and N.C. Maravolo. 1973. Isolation and characteris- (Wilson, tics of a Johnson and in These studies revealed possibleallelopathic factor supporting the dominant role McCaffrey, prep.). of Hieracium aurantiacumin the a suite of insects that feed on a number of bracken-grasslandsof northern native hawk- Wisconsin.Trans. Wis. Acad. Sci. Arts. Lett. 61: 235-251. weed species. Two species of tephritid flies and 1 species Deardorff, D.C. 1977. Biosystematic studies in Hieracium of moth were found in the flower heads, and a gall wasp (Asteraceae), section Thrysoidea. Ph.D. Thesis, University of was found in the axils of leaf in the leaf Washington,Seattle., Wash. petioles, mid-ribs, ML. 1950. and inside the flower stalk. Insects were also found in simi- Fernald, Grays Manual of Botany. 8th edition. AmericanBook Company,N.Y. lar structures of H. pratense, but only where it occurred with Frankton, C. and G.A. the Mulligan. 1970. Weeds of Canada. native species. Future surveys and continued testing Canada Departmentof Agriculture.Public. 948. will determine the range of feeding by these insects. These Guppy, G.A. 1976. The hawkweeds (Hieracium) of British studies coincide with research Columbia. Davidsonia 7(2):13—17. presently underway by the G.A. 1978. USDA-ARS in Montpellier, France, and IIBC in Guppy, Species relationships of Hieracium the (Asteraceae) in British Columbia. Can. J. Bot. DelOmont, Switzerland. A successful biological control pro- 9. for 56(23):3008—301 gram hawkweeds in the United States will require the Hay, J.R. and G. J Ouellette. 1959. The role of fertilizer and 2,4- coordinated inputs of scientists and land managers in the D in the control of pasture weeds. Can. J. Plant Sci. United States, Canada, Europe, and New Zealand. 39:278—283. research Hitchcock, L.C., A. Cronquist, M. Ownbey, and J.W. Currently, by the authors at the University of 1955. Vascular Idaho includes Thompson. plants of the Pacific Northwest.Part field and greenhouse studies to determine 5: Compositae.Univ. Wash. Press, Seattle, Wash. how hawkweeds compensate for damage, especially in Kennedy, G.G. 1902. The Maine coast at Culter. Rhodora 4:23—26. RANGELANDS 19(4), August 1997 23

Lass, L.W. and R.H. Callihan. 1992. Response of yellow hawk- Skalinska,M. and E. Kubien. 1972. Cytologicaland embryologi- weed to range herbicides in a non-crop site. Reg. Prog. Rep. cal studies in Hieracium pratense Tausch. Acta. Biol. Cracov. West. Soc. Weed Sd. Ser. Bot. 15(1):39—49. Lepage, E. 1971. Les épervières du Québec (The hawkweedsof Stebbins, G.L. 1950. Variation and evolution in plants. Columbia Quebec (Hieracium)).Nat. Can. 98(4):657—674. Univ. Press, N.Y. Lepage, E. 1967. Etude de quelques hybrides chez nos Stergios, B.G.1976. Achene production,dispersal, seed germina- Epérvieres(Hieracium) adventices. Nat. Can. 94:609—619. tion and seedling establishment of 1-lieracium aurantiacumin an Makepeace, W. 1976. Allelopathy of mouse-ear hawkweed abandonedfield community.Can. J. Bot. 54:1189—1197. (Hieracium pilosella). Proc. New Zealand Weed Pest Control Thomas, A.G. and H.M. Dale. 1976. Cohabitationof three Conference.29th:1 06—109. Hieracium species in relation to the spatial heterogeneity in an Munz, P.A. and D.D. Keck. 1959. A California flora, with supple- old pasture. Can. J. Bot. 54:2517—2529. ment. Univ. Calif. Press, Los Angeles, Calif. Thomas, A.G. and H.M. Dale. 1975. The role of seed production Noel, W.O., D.W. Wattenburger,W.S. Belles, and G.A. Lee. in the dynamicsof establishedpopulations of Hieraciumfloribun- 1979. Chemical control of orange hawkweed on rangeland. dum and a comparison with that of vegetative reproduction. Proc.West. Soc. Weed Sci. 32:77. Can. J. Bot. 53:3022—3031. Panebianco, R. and R.W. Willemsen.1976. Seed germinationof Voss, E.G. and M.W. Böhlke. 1978. The status of certain hawk- Hieracium pratense, a successional perennial. Bot. Gaz. weeds (Hieracium subgenus Pilosella) in Michigan. Mich. Bot. 1 37(3):255—261. 17(2):35—47. Peterson, R.L. and A.G. Thomas 1971. Buds on the roots of Weber, WA. 1990. Colorado flora: Eastern slope. Univ. Press of Hieracium florentinum(hawkweed). Can. J. Bot. 49:53—54. Colorado,Niwot, Cob. Reader, R.J. and W.H. Watt. 1981. Response of hawkweed Yeung, E.C. and R.L. Peterson. 1972. Studies on the rosette (Hieracium floribundum) patches to NPK fertilizer in an aban- plant of Hieraciumfloribundum. I. Observationsrelated to flower- doned pasture. Can. J. Bot. 59:1944—1949. ing and axillary bud development.Can. J. Bot. 50(1):73—78. Rickett, H.W. 1987. Wildflowers of the United States. Vols. 1-6. Zahn, K.H. 1921-1922. Compositae-Hieracium.Part II, Sect. XVI NewYork BotanicalGarden. McGraw-Hill, N.Y. Tridentata (Fortsetzung und Schluss). In: A. Engler (ed.), Das Roche, C. 1992. Mouse-ear hawkweed (Hieracium pilosella L.). PflanzenreichIV. 280 (Heft 79) 865—1146. PNW Coop. Ext. Pub. PNW4O9. Washington State University, Pullman,Wash. Scott, D., J.S. Robertson, and W.J. Archie. 1990a. Plant dynam- Authors are Research Support Scientist and Professor of tussock infested with Hieracium ics of New Zealand grassland Entomology, Department of Plant, Soil, and Entomological Sciences, pilosella. I. Effects of seasonal grazing, fertilizer, and over- University of Idaho, Moscow, Ida. 83844-2339; Plant Physiologist, drilling. J. AppI. Ecol.27:224—234. USDA-ARS, 1500 N Central Ave., Sidney, Mont. 59270 and Botanist, Scott, D., J.S. Robertson, and W.J. Archie. 1990b. Plant dy- USDA Forest Service, 1648 S 7th, Bozeman,Mont. 59717. namics of New Zealand tussock grassland infested with Wewould like to acknowledgethe financial support ofthe Hawkweed Hieracium pilosella. II. Transition matrices of vegetation Action Committee, Inc., and especially HACI member, Mr. Harry Eder, J. Ecol. 27:235—241. forhis leadership and dedicationto hawkweedcontrol in northern Idaho. changes. AppI. We also thank J. J. B. Johnsonand C. Talbott Rochefor their P.D. 1974. Hieracium. In: Tutin et al. Flora Campbell, Sell, (eds.). Europaea review of the manuscript. This article is published as research paper Vol. 4. CambridgeUniv. Press, London. pp. 71 3—754. 96739of the Idaho AgnculturalExperiment Station, Moscow. Skalinska, M. 1967. Cytological analysisof some Hieracium species, subg. Pilosella, from mountains of southern Poland. Acta. Biol. Cracov. Ser. Bot. 10:128—141.

I Ptruax1 INC L COMPANY. J 3717 Vera Cruz Ave MinneapolIs, MN 55422 Phcne 612 537-6639 Native Grass Drill ACCURATELY PLANTS ALL TYPES OF SEED • Fluffly native grasses • Tiny legumes S MeOtum sed wheat grasses