Genista monspessulana
(Montpellier broom)

In its native range in the Mediterranean region, G. monspessulana is widespread but only locally abundant. It tends to form small populations. In its introduced range, however, it is an invasive environmental weed of national, state and urban parks and fallow land, often forming dense monospecific stands. It is listed as a controlled weed in Australia and California, USA.

Broadleaved
Perennial
Seed propagated
Shrub
Woody

In its native range, G. monspessulana is most abundant in the coastal and sub-montane slopes of the Atlas Mountains in Morocco and Algeria, southwest Corsica, France (where it co-occurs with Cytisus scoparius) and in northwest Catalonia (Spain and France). Other significant native areas for this species are France (Massif des Maures and les Cevennes), Italy (Catena Costiera and Aspromonte; Sicily in Monti Nebrodi and Madonie; Sardinia in Monte Ortobene, Nuoro), Albania, Greece (Ilía - western Peloponissos), certain eastern Aegean Islands (e.g. Rhodes), Portugal (Estremadura, western Lisboa), Spain (southwestern Andalucia) and Turkey (southern coastal Mediterranean slopes) (Gibbs and Dingwall, 1971; Sheppard, 2003; Sheppard and Thomann, 2004). Populations in the east part of this distribution have very few specific natural enemies suggesting a western centre of evolution of this species, while populations in Portugal were also species poor for natural enemies suggesting a recent expansion there (Sheppard and Thomann, 2004).

Native populations in the Mediterranean tend to be small (90% of populations observed had fewer than 100 individuals) and scattered, either as a low density and spindly understorey component of disturbed cork oak/pine forest or as dense but transient populations in post-fire regenerating maquis communities in zones of rich acidic soil and relatively high rainfall (>600 mm). This association with higher rainfall areas in the Mediterranean becomes immediately apparent in Greece where the plant is only found on one 100-km² area in the western Peloponissos. This is the only place in Greece where acid soils occur at sufficient altitude for the necessary rainfall, in an area that does not experience hard continental winters (Sheppard, 2003). Apart from G. monspessulana, the other species within the Teline group are restricted in native distribution to either the western Mediterranean or various Atlantic islands (twelve species are endemic to the Atlantic islands; Percy and Cronk, 2002). G. monspessulana's habitat restriction in its native range and its scattered occurrence around the Mediterranean give it a very patchy distribution and localized abundance across the Mediterranean region.

It its exotic range, G. monspessulana is present in a number of countries surrounding the native distribution (Syria, the Caucasus, the Azores and Armenia; Gibbs and Dingwall, 1971; ILIDIS, 2004). It is also present in North and South America, South Africa and Australasia. In Australia, it is a widespread environmental weed of national, state and urban parks and fallow land mainly in South Australia and Victoria, but also in Tasmania and southern New South Wales (Sheppard, 2000). It has been estimated to invade at least 600,000 ha in Australia. In South Australia and Victoria it is classed as a noxious weed, while in Tasmania it is a 'secondary weed'. In South Australia it is widespread but with large infestations in the Mount Lofty Ranges, Belair National Park and the Clare Valley (Crossman and Kochergen, 2002). It also poses direct costs to the Forestry Industry (D McGuire, Forestry South Australia, Meadows, Australia, personal communication, 2005). In Victoria, Montpellier broom is in the top 30 most widespread weeds with 1000 ha of dense infestation, 50,000 ha of medium infestation and 550,000 ha of scattered infestation (Lane et al., 1980). Montpellier broom is a regionally controlled weed in the Wimmera, Glenelg, Corangamite, Port Phillip East, Goulburn, North East, West Gippsland and East Gippsland Catchment and Land Protection Regions. It is widespread with the largest infestations in the Dandenong ranges National Park (Gillespie, 1991), Australian Alps National Park, Central Highlands and the Wonangatta Valley. In Tasmania, Montpellier broom is widely scattered in the east, occurring commonly in the north-east in the Great Western Tiers and on Mount Wellington. In New South Wales it occurs on the coast, the tablelands, the western slopes and southwestern plains and there are scattered infestations totalling 1200 ha in 26 local government area shires and seven reserves in five districts of the New South Wales National Parks (Leys, 1998). Significant areas infested are in the Illawarra district and New England Tablelands. A few plants have been recorded in the Margaret River area of Western Australia (Parsons and Cuthbertson, 1992).

In the USA, G. monspessulana is a widespread environmental weed of national, state and urban parks and fallow land, mainly in central California and southern Oregon (Jepson, 1979). In California it has been estimated to invade at least 40,000 ha and was first introduced as an ornamental into the San Francisco Bay region. It is classed as an A-1 weed by CalEPPC (California Exotic Pest Plant Council) and a class C weed by the Californian Department of Food and Agriculture. It is particularly prevalent along the coast from Monterey county north to Mendicino county and inland in Lake Solano and Contra costa counties and also in the Sierra Nevada foothill counties to 800 m (Bossard, 2000). It is the most important weed in Jackson State Experimental Forest. Invaded habitats include coastal plains, mountain slopes, riverbanks, road cuts, forest clear-cuts, grassland and open canopy forest on a wide range of soil types.

As G. monspessulana is an important horticultural species, risks remain for significant further spread both through fresh introductions into countries associated with uninformed horticultural practices and illicit passage of seeds through the post. Only a few horticultural varieties of broom show a capacity for naturalization and spread. The difficulty of clear identification of the different varieties by the inexperienced will remain a significant risk, however many official horticultural bodies are aware of the risks and discourage sales of varieties known to be invasive (Atkinson and Sheppard, 2000). Similarly horticultural use requires assessment of the risks biological control agents pose for the profitability of sales of harmless varieties. As with all weeds that spread by seed, phytosanitary measures are required to prevent the movement of seeds locally via the movement of livestock or of mechanical equipment through infestations during seed pod dehiscence or when seeds cover the soil. In key areas it co-occurs with other brooms (e.g., Cystisus scoparius) posing a risk of weed replacement following control efforts.

G. monspessulana is native to the Mediterranean basin where it is a widespread, but only locally common, component of high rainfall (>600 mm) Mediterranean maquis plant communities on acid granite, schist or old river-based soils (Gibbs, 1968; Sheppard, 2000). As part of this community, it is present in the understorey of Mediterranean evergreen or semi-deciduous Quercus or Pinus woodland or in more open situations of Quercus coccifera scrub. It co-occurs typically with other shrubs including Arbutus unedo, Erica arborea, Cistus salviifolius or C. monspeliensis and other Genisteae and is not considered invasive, appearing as flushes only after disturbance, particularly fire, and along roadsides. Amongst other Genisteae, is it usually subdominant to Cytisus villosus, Calicotome villosa or Calicotome spinosa. Cytisus arboreus, Adenocarpus telonensis and Spartium juncaeum and, occasionally, C. scoparius can also co-occur with it.

In exotic communities, G. monspessulana appears less restricted by habitat type and soil pH. It has become a major widespread weed in native and commercial forest ecosystems, in disturbed habitats, and along water courses. Invaded habitats include sclerophyllous recreational and commercial forests, open woodland, along roadsides, railways and river systems from 0-1000 m altitude, coastal plains, mountain slopes, riverbanks, road cuts, forest clear-cuts, grassland and open canopy forest.

G. monspessulana is a significant weed of forestry, particularly in pine and eucalypt plantations, where it either smothers planted saplings and reduces their growth or prevents natural regeneration. In native woodland situations (e.g. Californian redwood forests, USA), it can reduce natural regeneration by shading. Natural regeneration in woodland can be reduced following the hotter natural fires which result from the higher fuel load provided by the infestation.

Genetics

Significant morphological variation observed between infestations in different locations (variation in bushiness and leaf size) can largely be associated with habitat (i.e., in the open or understorey). While there is little outward morphological variation associated with geographic distribution (e.g. flower colour), natural populations scattered around the Mediterranean will have had significant long-term reproductive isolation, which is likely to be associated with degrees of genetic distance. In exotic populations, the historic horticultural importation and associated hybridization activities with other species in the Teline group suggest that exotic populations may nonetheless exhibit genetic variation and as such may not be identical to native G. monspessulana material (D Cooke, Department of Water, Land and Biodiversity Conservation, Adelaide, Australia, personal communication, 2005). A study of such variation is required to clarify such suppositions. Several commercially available broom varieties are of hybrid stock, e.g. Cytisus 'Porlock' is considered to be a hybrid between G. monspessulana and Cytisus x spachianus sensu hort. Cytisus x spachianus has a range of varietal names (Cytisus racemosus Hort, Cytisus 'Racemosus Nana', Cytisus praecox 'Nana', Cytisus 'Racemosus Scoparius Nanus', Genista racemosa Hort, Genista hispanica sensu Macoboy, and Genista x spachiana) and is considered to be a hybrid between G. canariensis and G. stenopetala (Rowell, 1991; Atkinson and Sheppard, 2000) and has also naturalized in some areas (Parsons and Cuthbertson, 1992).

Physiology and Phenology

G. monspessulana can live up to 13 years, but this is extremely rare with most mature plants dying between the ages of 3 and 7 years (Lloyd, 2000). Old plants are also rarer in the native range. Annual growth is such that axillary meristems, produced on the main axis, grow up into lateral vegetative shoots at the same time as extension of the main axis. The top of the vegetative shoots produces racemes of lateral axillary flower clusters in the subsequent year while the main axis of the shoot continues to grow and produce leaves above the flowering zone as the season progresses. Secondary flowers may be produced as part of a second flowering higher up the shoot a few months later in the same season on young actively growing plants. Photosynthesis occurs in the green stems and leaves. Young stems remain green for at least 1 year and the persistent leaves allow photosynthesis to occur throughout the growing season. Leaves are shed one year later as new leaves are produced in spring higher up the shoot. Shaded plants adapt by having lower branches and larger leaves.

Adult Montpellier broom is fairly drought tolerant, although the ca. 600 mm minimum rainfall required suggests some life stages, probably the seedlings, are relatively drought intolerant. Germination occurs following seasonal rains mainly in the spring and autumn (mainly in autumn in the native range). Big flushes of germination follow large-scale disturbance, droughts and fires after rain and can probably occur at any time of the year.

G. monspessulana is a typical Mediterranean species and is likely to have relatively high optimal temperatures for photosynthesis. The period of most rapid growth is in spring/summer, but plants can grow all year round if humidity and temperature conditions are suitable as there is no obligate winter dormancy typical of more temperate species such as Cytisus scoparius. Flowering is early season; February until early June (or July in Coastal California, USA; Bossard, 2000) with seed pod dehiscence in July in the northern hemisphere and August until November with seed pod dehiscence in January in the southern hemisphere.

Reproductive Biology

G. monspessulana reproduces only by seed from the second year and twice a year when young. Age structures in the native range are mostly bell shaped with age classes of 2-4 years dominant, few new recruits and few older plants (Lloyd, 2000). This is consistent with populations being generated by pulsed opportunities for recruitment associated with fire and other disturbance. Native populations appear to be mainly early successional, being replaced by later successional species. Age structures in the exotic range fit a 'reversed J' distribution better, with highest numbers in the youngest age class and numbers declining with increasing age. This indicates that invading alien populations, in contrast to the native range, have continuous recruitment with seedlings regaining spaces left by dying older plants in what are clearly more persistent populations. This agrees with observations that Montpellier broom in invading infestations tends to form dense monospecific stands.

Flowers can self pollinate, but the fertilization rate is 50% lower with autogamous pollen and successful fertilization requires that the flowers are tripped by a pollinator (Parker and Haubensak, 2002). Flower abortion rate in exotic stands varies from 70 to 95% and this could be only partially explained by pollen limitation, suggesting resources were the main factor limiting seed production (Parker and Haubensak, 2002). Seed production increases asymptotically with plant age for all populations, peaking at ca. 100,000 seeds per plant for 5- to 6-year-old individuals growing without competition (Lloyd, 2000). In the exotic range in California, USA, seedbanks, up to 9000 seeds/m² have been recorded (Parker and Kersner, 1989; Alexander and D'Antonio, 2003) and seed banks were found not to increase significantly with stand age suggesting annual seedbank losses quickly equalled annual seed rain (Alexander and D'Antonio, 2003). In Australia, annual seed rain can range from 2500 to 13,500 seeds/m², setting up seedbanks of between 30,000 and 100,000 m² with an annual seedbank decay rate of 23-50% largely resulting from failed germination.

The annual probability that a seed in the seedbank will become a seedling is about 0.01, while the probability of this seedling surviving the first year is about 0.07 following either anthropogenic disturbance or fire (Pareja, 1999). In the native Mediterranean range, the plant generates smaller seedbanks 500-900/m². In the native range, G. monspessulana suffers high levels of herbivory from goats, stem miners, insects in seedpods and post dispersal seed predation by rodents (J Lloyd, Weeds CRC, University of Adelaide, Australia, unpublished data).

Environmental Requirements

G. monspessulana can be found at up to ca. 1000 m altitude. Altitudinal and latitudinal limits are set by persistent low winter temperatures and rainfall on seedling survival (González-Andrés and Ortiz, 1996a). It is also poorly adapted to consistently cold temperate winters, showing little capacity to shut down growth over winter. In a Mediterranean climate, however, it will quickly overtop more winter dormant temperate species, like C. scoparius, because it can grow as soon as conditions are suitable such as during warm weeks in winter. In Spain it was found to grow 140 cm in two growing seasons and faster than other brooms (González-Andrés and Ortiz, 1996a). The plant appears able to tolerate windy conditions and will occur on steep, sparsely vegetated slopes. G. monspessulana seedlings require persistent moisture to survive and seedlings are less frost tolerant than those of Cytisus scoparius (González-Andrés and Ortiz, 1996a). G. monspessulana infestations do not collectively senesce in the way C. scoparius populations do (Peterson and Prasad, 1998), instead individuals die and drop out of the stand over a broad age range and are quickly replaced by juveniles (exotic range) or other species (native range). In drier climates broom can remain a highly invasive species associated with the banks or braided riverbeds of watercourses or along drainage lines. In its native range, broom is a calcifuge (Polunin and Smythies, 1973). In the exotic range, however, this broom occurs on a broader range of soils derived from a wide variety of substrates, particularly river sand, schist, granite, basalt or siliceous soils, and can even grow reasonably well on alkaline soils up to pH 8 (González-Andrés and Ortiz , 1996b). It does not flourish on calcareous soils. Plants establish best after soil or vegetation disturbance, caused by animals, fire or herbicide treatments, for example. Montpellier broom is quite tolerant of heavy shade as it persists as an understorey plant at least in forest gaps and trails even with quite dense canopy cover. Seedlings can tolerate up to 80% shade (C Bossard, St Mary's College, Moraga, California, USA, unpublished data) allowing broom to persist in woodland and forests. Seedling survival tends to be significantly lower beneath parental or other relatively dense canopy cover.

Associations

G. monspessulana is associated with specific nitrogen-fixing bacteria of the genus Bradyrhizobium (González-Andrés and Ortiz, 1999). Bacterial associations in root nodules allow this plant to grow faster survive better and have higher nitrogen content assisting competition and invasion on nitrogen-poor soils.

The arthropod natural enemies of G. monspessulana have been the subject of recent studies (Sheppard, 2000, 2003; Sheppard and Thomann, 2004). At least 90 arthropod species are associated with Montpellier broom in Europe, and several other generalist species are now found on broom in its exotic range (Sheppard, 2003; Sheppard and Thomann, 2004). Of these, 52 species are considered specific at least to the level of the tribe. Currently only four species seem likely to be specific to G. monspessulana; the psyllid Arytinnis hakani, the apionids (Lepidapion sp. nov. and Oryxolaemus sp. nov.) and the nepticuliid (Trifurcula serotinella). The eriophyid mite Aceria genistae has also been recorded from G. monspessulana in the USA (Chan and Turner, 1998). Natural enemies that have caused noticeable damage in the native range sufficient to affect Montpellier broom growth and spread are listed in the table. In the native range, this broom is regularly defoliated by the moth Uresiphita polygonalis and a combination of stem boring insects regularly kills immature plants in eastern France. The pathogen Uromyces genistae attacks the older leaves in spring/summer in Europe and the USA (Guynot and Massenot, 1958; Sheppard, 2000).

In the native range, goats are probably another significant factor limiting the abundance of this plant, given their ubiquitous occurrence, and that most Mediterranean species in the Genisteae are spiny. No published data are available on the impact of natural enemies of Montpellier broom in the native range. In the USA, Uresiphita reversalis can defoliate G. monspessulana, but plants just regrow leaves after larval stages finish development (Montllor et al., 1990, 1995).

Environmental

• Agroforestry

General

• Ornamental

Many other exotic weed species occur in the same tribe as G. monspessulana. These include five other Genista species and members of the genera Cytisus (four species), Retama (two species), Calicotome (one species), Spartium (one species) and Ulex (one species) (Holm et al., 1979; Jepson, 1979). All Genista species recorded as invasive weeds outside their native range are within the Teline group and occur predominantly in Mediterranean climates on acid soils with relatively high rainfall (>600 mm). Of these G. stenopetala, G. canarinesis and G. linifolia are the most widespread and problematic, but other species (e.g., Genista maderensis and G. tinctoria and horticultural hybrids) have also naturalized (Groves and Hosking, 1997). G. canariensis grows to 2 m with leaves borne on short petioles (1-6 mm) and intense yellow flowers (to 12.5 mm long) with a pubescent standard in compact terminal racemes. G. stenopetala is a taller shrub (to 6 m) but with terminal (and lateral) 5- to 26-flowered racemes, 2-9 cm long; petioles 4-12 mm long; leaflets obovate to oblanceolate, 18-50 mm; and flowers 12.5-15.5 mm (Gibbs and Dingwall, 1971). G. stenopetala is from the Canary Islands and may hybridize with G. monspessulana in some exotic infestations (native distributions are allopatric). G. linifolia is quite different from G. monspessulana with long narrow sessile leaves and leaflets (9-61 x 2-10 mm) with flowers in terminal racemes (4-20 flowers) with the standard uniformly sericeous (silky). Although many other species of Genista have been moved around the world for botanical gardens etc., the prevalence of weeds within the Teline group may be explained by their horticultural attractiveness and hence more widespread importation and planting.

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.

Cultural Control

Unpublished work in Australia, New Zealand and the USA on broom management with goats, suggests this is a potential option. It can be effective on disturbed/agricultural land where native regeneration is not being encouraged. Meat or dairy goats should be used because these are the easiest to handle and cannot jump the fences. Goats can be trained to be quite selective at least within the vegetation structure, for example they can effectively strip flowers. Goat management may require the provision of access trails and follow-up herbicide treatments on resprouts once the goats have been moved on.

Burning uncut broom in late spring/early summer can be used successfully (Boyd, 1994), although appropriate conditions for an effective burn are hard to obtain safely. Burnt sites showed >80% loss in seedbank density (Alexander and D'Antonio, 2003). Reburning the same site 2-4 years later may also help if there is a sufficient fuel load and the frequent fires do not prevent native bush regeneration. However, reburning does not further reduce the seedbank (Alexander and D'Antonio, 2003). Fire is best used as part of an integrated strategy and post-fire monitoring is a key requirement.

Mechanical Control

Mechanical removal of G. monspessulana is labour intensive and so best applied to small infestations amongst desirable vegetation. Mower or bush hogs are sometimes a viable option for monospecific stands but there are often problems of accessibility. Slashing, whacking or brush cutting are other alternatives (Archbald, 1996). Bulldozing is not recommended due to the soil disturbance generated and the effect this has on burying and prolonging the seedbank. Cut individuals can resprout if not cut at ground level. For individual large plants, weed wrenches assist targeted removal, but can disturb the soil prolonging the need for monitoring for regeneration (Bossard, 2000). Hand removal of individuals prior to seed fall in sensitive areas, particularly for isolated individuals at flowering is an effective strategy in National Parks where the walking public can be trained to recognize and remove broom. Mechanical control does not deal with the long-lived seedbank and so integrated strategies associated with reseeding desirable species, or at least allowing those present to seed effectively, are often required (Bossard, 2000). Over-sowing with a perennial grass layer has been suggested for other brooms and may also be effective following fire. All forms of mechanical removal require monitoring of treated areas for regrowth from stumps and the seedbank. Sites should be visited once every 1-2 years after treatment in late spring for 5-10 years and every 2 years thereafter to treat any new patches and flowering individuals prior to seed set.

Chemical Control

The main chemicals used to control brooms are picloram, triclopyr, glyphosate, fluroxypyr and metsulfuron (Parsons and Cuthbertson, 1992). Specific chemicals are appropriate for specific situations, such as proximity to water courses. The addition of some surfactants to glyphosate and metsulfuron increase the level of control achieved by these chemicals. Triclopyr ester in Hasten or Penevator oil in low volume basal bark application with a wick has proved very effective at killing mature plants (Bossard et al., 1995). Herbicide injection (drilling, filling and frilling) has also been trialed (Gillespie, 1991). Chemical control of G. monspessulana is most usually applied as a high-volume foliar spray. Applications need to be made in periods of active growth after flower formation, but before seed dehiscence. Seedlings are least resistant to hormonal herbicides at the 10- to 15-cm size (Bossard, 2000). Regular follow-up spraying is essential for effective control, but standing dead biomass also presents a major fire hazard.

Biological Control

Work on the biological control of G. monspessulana started in 1998. So far only one agent, Arytinnis hakani, has received a release permit and, as yet, only in Australia (Sheppard, 2003). Several other species are being considered. No releases have been made. Bruchidius villosus has recently been introduced into a number of countries to control Cytisus scoparius (Syrett et al., 1999) and is also likely to feed on G. monspessulana.

Integrated Control

Only two studies have been found that look in depth at possible management strategies for G. monspessulana and both these are unpublished. The first relates to an experiment carried out in Jackson State Experimental Forest in California, USA (Bossard, 1995), while the second was a similar study carried out in the Adelaide hills, South Australia (Lloyd, 2000). The results were similar. Management is most effective using fire-based integrated weed management in areas that cannot wait for the long-term biocontrol solution. Before burning, curing the broom using herbicides (triclopyr basal applications or glyphosate to protect natives) increases fire intensity, thereby stimulating or killing a maximum amount of the resident weed seed bank. Fire is most effective on pre-sprayed or slashed broom because this also keeps fire low to the ground. Lloyd (2000) also found that smoke stimulated seed germination in plots not directly affected by the fire. Targeting the seed bank in this way with fire achieved 85-95% losses of G. monspessulana in both trials even with quite low intensity burns. Follow-up treatments must be carried regularly to kill regenerating plants just before seed set, i.e. about 2 years after the fire. Such follow-up treatments kept broom cover to <5% after 3 years in California. Many land managers perceive management with fire as too risky and also little is known about the effects of such control burns on the native species that should be encouraged to replace the weed.

Best practice in broom management will require a specific strategy for each situation. Areas should be designated for containment versus treatment where resources are limiting. Fire or goats can be the basis of integrated broom management strategies depending on the situation. Without these options (e.g., in containment areas) biological control remains the only solution. All successful management requires a communication strategy to explain and encourage adoption and adaptive management (trying new ideas in parallel to traditional wisdom) should be encouraged at all times.