Pilosella caespitosa
(yellow hawkweed)

P. caespitosa is a stoloniferous rosette plant which has spread rapidly to exotic locations outside its native range, namely to North America after its introduction as a garden ornamental or contaminant of agricultural seed. It continues to be available as an ornamental and can be easily transported by machinery and as such is likely to spread further. P. caespitosa is an undesirable invader on account of its vigorous growth due to stolon, rhizome and adventitious root bud production and wind-dispersed seeds which are produced in high numbers. P. caespitosa displaces desirable pasture plants leading to loss of forage and biodiversity. Early infestations of P. caespitosa which are still very small may be difficult to spot since this weed occurs on upland pastures and forest meadows in parts of the USA, and it is a declared weed in the states of Idaho and Washington (USDA-NRCS, 2016).

The natural distribution of P. caespitosa is Euro-Sibiric (Zahn, 1987) and the cohesive area of distribution extends from the Rhine to the Altey (Gottschlich, 1996). Besides populations in France near Paris, none of the other populations west of the Rhine (e.g. in Brittany, France, UK and Norway) are native (Gottschlich, 1996). P. caespitosa can be found from Siberia to subarctic Europe and from Scandinavia and northern Russia to northern Greece and Bulgaria and occasionally Turkey (Gottschlich, 1996). P. caespitosa has been introduced into North America and New Zealand.

Further spread is highly probable, owing to the risks of both accidental movement as a contaminant of seed or soil, or attached to machinery, and deliberate introduction as an ornamental. This is encouraged by the availability from commercial nurseries via mail-order catalogues and websites.

In its native range, P. caespitosa occurs in patches on nutrient-poor, moist pastures, loam or clay soils, and also peat. Its typical growth habitats are swamp meadows, swamps and swamp margins, but P. caespitosa is also often found on oligotrophic sites disturbed by man such as a pioneer plant at moist roadsides, waysides, gravel pits and along railway lines (Gottschlich, 1996). For most Pilosella species the best growing conditions in Europe were available in the middle of the nineteenth century due to the huge diversity of the countryside, i.e. many pastures and forest margins and the permanent lack of nutrients on most agricultural land (Gottschlich, 1996). In North America, P. caespitosa is primarily a weed of moist pastures, forest meadows, abandoned fields, clearings and roadsides (Fernald, 1950; Wilson and Callihan, 1999). However, they have shown a tendency to invade mid- to high-elevation meadows and abandoned farmland (Wilson and Callihan, 1999). In New Zealand, P. caespitosa grows in extreme wet and dry conditions and is a notable weed associate in shrublands, tall-tussock and red-tussock grassland. It is often abundant up to 1500 m (Hunter, 1992). 

P. caespitosa is not known to be a weed in crops. It does not persist in cultivation because crops can out-compete other species of Pilosella, especially where herbicides are used in the cropping system (Wilson and Callihan, 1999). It is however, a serious weed of both natural and managed pastures where it out-competes with forage species.

Genetics

The chromosome number of P. caespitosa is 2n=18, 27, 36, 45 (Sell and West, 1976). P. caespitosa populations from southern Poland are tetraploid (2n=36) and apomictic whereas populations from eastern Poland are diploid (2n=18) and sexual (Gottschlich, 1987). Hybridization is possible with numerous other species of Pilosella including P. aurantiaca, H. cymosum, P. floribunda, P. lactucella and P. officinarum, with the most closely related species being P. aurantiaca (Meusel and Jäger, 1992). Two species assumed to have resulted from interspecific hybridization including P. caespitosa are H. flagellare and P. glomerata.

Physiology and Phenology

P. caespitosa seeds germinate well at high temperatures after only a short period of after-ripening, though are probably not viable in soil for a long period of time (Panebianco and Willemsen, 1976). However, even though seed and seedling mortality appear to be very high, the establishment of only a small number of rosettes is needed to ensure the success of P. caespitosa since it is a perennial and multiplies vegetatively. P. caespitosa plants flower between May and July in Europe, but later flowering can occur (Gottschlich, 1996); between May and August in North America (Fernald, 1950) and from November to January in New Zealand, but individuals flower until March and fruits mature between December and February (Webb et al., 1988). No references were found regarding allelopathic substances in P. caespitosa (e.g. umbelliferone, caffeic acid and chlorogenic acid) as recorded for P. officinarum (Makepeace et al., 1985) which does not mean, however, that they are absent in P. caespitosa.

Reproductive Biology

Non-flowering P. caespitosa rosettes originate either from seedlings or establish at the end of leafy stolons from adventitious root buds located on the fibrous roots (Wilson and Callihan, 1999) or from rhizomes. Since each flowering stem produces (5-) 10-25 (-50) capitula, considerable amounts of seeds can be produced. The flowering stems are (20-) 30-50 (-80) cm tall and seeds are wind-dispersed. However, even though seed and seedling mortality appear to be very high, the establishment of only a small number of rosettes is needed to ensure the success of P. caespitosa, since it is a perennial and multiplies vegetatively (Panebianco and Willemsen, 1976). When the plant initiates flowering, stolons and rhizomes initiated from axillary buds at the base of rosette leaves begin to grow (Wilson and Callihan, 1999). There is no information comparing numbers of newly established rosettes originating from vegetative production and numbers arising from seeds, although it is likely that vegetative reproduction within a population is more important than rosettes derived from seeds, as recorded for P. officinarum and the closely related P. floribunda. A large difference in turnover of P. officinarum plants between field sites in New Zealand was recorded, with field populations of 5-173 new rosettes per 100 existing rosettes and that spread occurs mainly by vegetative means (99%) (Makepeace, 1985). Similar results were obtained in North America for the closely related P. floribunda where only 1% of new plants in a population were derived from seedlings (Thomas and Dale, 1975). Nonetheless, seeds remain important for long-distance spread. A mixed strategy of clonal growth and reproduction by seeds may be necessary to maintain populations of this species in the presence of high interspecific competition and a shortage of open space (Winkler and Stöcklin, 2002). By spreading vegetatively, P. caespitosa rosettes can form dense mats with 3500 plants per m² (Wilson and Callihan, 1999).

Environmental Requirements

Throughout their geographic range in New Zealand, species of Hieracium/Pilosella are particularly abundant in sub-humid to humid montane to lower sub-alpine bioclimates. The optimal rainfall range for vigorous Hieracium/Pilosella-dominated communities appears to be 600-1200 mm. However, P. caespitosa also occurs locally where annual rainfall exceeds 3000 mm (Hunter, 1992). Possible reasons for the past 50 year decline of P. caespitosa populations throughout Europe include nitrogen fertilizer application, nitrogen contamination through rain and intensive agricultural use of the landscape (Gottschlich, 1996). However, locally, P. caespitosa is becoming more abundant due to disturbance (e.g. in gravel pits and at roadsides) (Gottschlich, 1996). In North America, P. caespitosa poses the greatest threat to cooler, sub-humid to humid sites in the northern regions of the United States. Habitats most susceptible to invasion range from the lowlands of the northern Pacific Coast to elevations of 1500 m or more in the mountain states (Wilson and Callihan, 1999). P. caespitosa plants tolerate very low winter temperatures, with -32.5°C recorded in Poland.

P. caespitosa is not attacked to any noticeable degree by phytophagous insects in New Zealand (Syrett and Smith, 1998) and may therefore have a competitive advantage over native rangeland species. In its native range, P. caespitosa is associated with a range of specialized phytophagous insects (Grosskopf et al., 2001; 2003), the main insect groups being Aulacidea spp., Oxyptilus spp., Cheilosia spp., gall midges and tephritids developing in the flower heads. In contrast to other weeds (e.g. Alliaria petiolata, Cynoglossum officinale or Euphorbia spp.), so far, no host-specific Coleoptera have been found.

P. glomerata, H. cymosum and P. dubia are all very similar to P. caespitosa. P. glomerata has yellowish-green to bluish-green leaves whereas P. caespitosa always has yellowish- or grass-green leaves. P. glomerata leaves have a dense cover of short hairs which are 1 (-2) mm in length (which are responsible for the rough texture of the leaves) and the upper and lower leaf surfaces are sparsely covered with stellate hairs. In contrast, leaves of P. caespitosa are covered with numerous long hairs (2-6 mm long) but the upper leaf surface has no stellate hairs; only the margin and the lower leaf surface are sparsely covered with stellate hairs. H. cymosum only rarely has stolons and when present, are below-ground, pale and thin. The number of flower heads per stem is 20-50 (up to 100) and the stylus is yellow. Flowering P. caespitosa plants produce stolons which are 5-15 cm long and green spatulous leaves. The flowering stems have 10-25 (up to 50) flower heads and the stylus is dark. P. dubia usually has no stolons, but when present, stolons are thin and often below-ground. The rosette leaves of P. dubia are bluish green to leek-green.

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.

Control

Cultural Control

In its native range, P. caespitosa occurs mainly on nutrient-poor soils. Thus, fertilizers can control these species by increasing the competitive ability of more desirable pasture species (Wilson and Callihan, 1999). The main method of control at present is to improve land by the application of fertilizer and over sowing with pasture species. However, these species remain a problem on areas with low economic growth potential, abandoned land, reserves and national parks (Grundy, 1989).

Mechanical Control

Mechanical control of P. caespitosa has had limited success. Digging the plants or otherwise disturbing the stolons, rhizomes, or roots only serves to spread and multiply the weed, since plants can grow from buds on small roots, stolons and rhizome fragments (Wilson and Callihan, 1999). Due to its mat-forming growth, mowing does not kill P. caespitosa.

Chemical Control

Chemical control is especially promising on small infestations to prevent further spread. P. caespitosa is effectively controlled by clopyralid, picloram, and picloram + 2,4-D, (Whitson et al., 2000). However, total kill is unlikely and follow up with fertilizer and top dressing is considered essential (Grundy, 1989). A problem with chemical control is that the chemicals have as much, or more, impact on many desirable pasture species as they do onP. caespitosa. (Grundy, 1989). In many areas in New Zealand, chemical control is uneconomic and thus not considered on a large scale (Grundy, 1989).

Biological Control

Since chemical and mechanical control of P. caespitosa are ineffective and/or not economical in New Zealand, a programme to develop biological control with insects and a pathogen was initiated in 1992 (Syrett and Smith, 1998). P. caespitosa is not attacked to any noticeable degree by specialized phytophagous insects in New Zealand (Syrett and Smith, 1998), and may therefore have a competitive advantage over native rangeland species. Five insect species of European origin were studied and tested, and are approved for release in New Zealand (Syrett et al., 1999; Grosskopf et al., 2002; Klöppel et al., 2003) and four of them have the potential to develop on P. caespitosa. These are the plume moth Oxyptilus pilosellae (Lepidoptera, Pterophoridae), the larvae of which feed on the above-ground plant parts; the gall midge Macrolabis pilosellae (Diptera, Cecidomyiidae) which galls stolon tips and rosettes; the syrphid Cheilosia urbana (Diptera, Syrphidae), the larvae of which feed externally on the roots; and the closely related species Cheilosia psilophthalma (Diptera, Syrphidae), the larvae of which feed in rosette centres, leaf axils and on stolons. P. caespitosa is not a host plant of the fifth species, the cynipid Aulacidea subterminalis Niblett (Hymenoptera, Cynipidae) which galls the stolon tips of P. officinarum and P. aurantiaca.

Since P. caespitosa is a noxious weed in North America and is often not effectively controlled by other means, especially in remote areas, a consortium was formed to initiate a biological control programme in North America. There are, however, major differences between the programmes in New Zealand and North America. These are firstly the presence of closely related and native Hieracium/Pilosella species in North America, The second difference is a different range of weedy Hieracium species present in New Zealand (P. caespitosa, H. lepidulum, P. officinarum and P. piloselloides subsp. praealta) as compared to North America (P. aurantiaca, P. caespitosa, P. glomerata, P. piloselloides, locally also P. officinarum). Insect species approved for release in New Zealand are being tested by CABI-CH for their suitability as potential biological control agents of P. caespitosa in North America, and additional surveys to explore other insect species are being made (Grosskopf et al., 2001; 2003). Research on the potential of O. pilosellae,M. pilosellae,A. hieracii and C. psilophthalma as biocontrol agents of P. caespitosa have been discontinued as a result of agents not being host-specific. Aulacidea pilosellae causes small galls on the midrib of leaves, stolons and flower stalks of several Hieracium/Pilosella target species, and appears to have a restricted host range. The wasp is being studied in collaboration with Agriculture and Agri-Food Canada (AAFC) and Montana State University (MSU).

A joint petition for the field release of the root-feeding hoverfly Cheilosia urbana in North America was submitted to the United States Department of Agriculture – Animal and Plant Health Inspection Service (USDA-APHIS) Technical Advisory Group (TAG) and the Canadian Biological Control Review Committee in December 2014. The agent was approved for release in Canada by the Canadian Food Inspection Agency (CFIA) in April 2016 and recommended for release by TAG for the USA in May 2016. Surveys in Switzerland to collect the hoverfly in view of future releases in Nord America commenced in 2016.

Integrated Control

Strategies for sustainable, long-term management include prevention, early detection, minimizing disturbance (e.g. avoidance of heavy grazing and burning), maintaining soil health, controlling grazing, judicious use of herbicides, periodic applications of fertilizer, and cooperation between private and public land managers (Wilson and Callihan, 1999). Classical biological control could complement these management practices by further stressing or even killing P. caespitosa plants (Grundy, 1989; Wilson and Callihan, 1999).

Since P. caespitosa infests permanent upland pastures and forest meadows in remote areas in parts of the USA, conventional surveys to determine infestations of P. caespitosa are not practical (Carson et al., 1995). Therefore, Carson et al. (1995) and Lass and Callihan (1997) developed a method to identify P. caespitosa infestations with remote multispectral digital imagery taken from aircraft. However, small and moderate P. caespitosa infestations with less than 20% P. caespitosa cover are not detectable using this technique (Carson et al., 1995).

19/12/2016 Updated by:

Ghislaine Cortat, CABI-CH, Switzerland

29/09/2003 Original text by:

Gitta Grosskopf, CABI-CH, Switzerland