Hakea sericea
(silky hakea)

The factors contributing to the success of H. sericea as an invader in South Africa are its prolific production of canopy borne seeds in the absence of natural enemies, the high seed longevity in the canopy and efficient seed dispersal. In dense stands of H. sericea, seed densities of up to 7500 seeds per m² have been reported and the winged seeds of H. sericea facilitate dispersal over several kilometres in some cases. An additional factor that has contributes to the invasiveness of H. sericea is the regular wildfires that occur in the areas invaded by H. sericea. While H. sericea is a weed of pasture it is principally an environmental weed seriously threatening the biodiversity of the 'fynbos' Cape Floral Kingdom.

Perennial
Seed propagated
Shrub
Tree
Woody

H. sericea is native to south-eastern Australia and is found in Queensland (Mt Barney, Mt Maroon and Mt Mee) and New South Wales, with naturalized occurrences overseas, e.g. in South Africa, New Zealand and south-west Europe (Barker, 1996). H. sericea has been planted for reclamation of arid lands in Spain and Portugal and has become locally naturalized (Royal Botanic Garden Edinburgh, 2003). In South Africa, dense stands of H. sericea occur in the Western and Eastern Cape Provinces. Isolated plants have been recorded in Kwa-Zulu Natal. In New Zealand, H. sericea occurs north of Auckland and is considered an important invader because of its ability to invade Leptospermum and gumland communities (Beever, 1988).

Further spread is possible if accidentally introduced as an ornamental. H. sericea has been declared a noxious weed in South Africa in terms of 'The Conservation of Agricultural Resources Act' (Act No. 43 of 1983).

H. sericea in its native range grows naturally in heaths and as the understorey of dry sclerophyllous forests on the coastal regions of south-eastern Queensland to south-eastern New South Wales, Australia. It has become naturalized in South Africa and New Zealand. In South Africa H. sericea occurs in sclerophyllous vegetation known as mountain fynbos (or 'macchia') which is virtually devoid of tree species.

H. sericea is not a weed of crops but has been known to invade valuable pasture land in South Africa (Phillips, 1938). It is today, however, principally regarded as an environmental weed.

Physiology and Phenology

H. sericea flowers between late autumn and late spring. The newly set fruit is green, soft and juicy but as fruits mature they become woodier. The heat-resistant fruits, produced annually, are stored in the canopy and the seeds are only released once the plant has died. There is no seed bank in the soil. The nutrient-rich, thin-coated seeds can germinate throughout the year provided the soil is moist from an initial good precipitation. After germination, a taproot develops even before the first true leaves are formed which may contribute to the seedling's resistance to drought (Fugler, 1979). The juvenile stage is two years and mature H. sericea trees can live for up to thirty years.

Reproductive Biology

Propagation is by seeds, which accumulate on the plant throughout its lifetime. Large numbers of seeds are released following the death of the plant, usually by fire. The prolific seed production of H. sericea can result in estimated seed densities of up to 7500 seeds per m² in the ash bed following fires (Neser and Kluge, 1986). Beadle (1940) found that the seed of H. sericea was capable of withstanding temperatures of 110°C for four hours without a significant reduction in the germination percentage. No cases of vegetative reproduction have been found in South Africa although the plant can regrow vegetatively if damaged. The seed production of H. sericea is far more prolific in South Africa than in its native range. This has been attributed to the absence of natural predators in South Africa (Neser, 1968).

Environmental Requirements

H. sericea occurs mainly on well drained soils derived from sandstone and quartzite with low nutrient levels. These soils are low in available nitrogen and phosphorus and often lack copper, zinc and molybdenum (Specht and Rayson, 1957). The spread of H. sericea in South Africa has been restricted to some extent by barriers of unsuitable nutrient-rich substrata (Fugler, 1979). H. sericea is native to areas in Australia where the rainfall is relatively evenly distributed throughout the year. In South Africa however, it occurs in rainfall regimes including a Mediterranean-type climate with a summer drought, a summer rainfall region and an area with rainfall throughout the year.

Surveys by South African researchers have discovered more than 40 insects in over 20 families that attack H. sericea in Australia (Kluge and Neser, 1991). Four host-specific herbivorous insects have been introduced from Australia for the biological control of H. sericea in South Africa. In addition, an indigenous fungus has also been utilized in the H. sericea biological control programme. Recent progress in determining the host range of a flowerbud-feeding weevil, Dicomada rufa, may soon allow the release of this agent. Thirteen indigenous herbivorous insects in nine families have been recorded on H. sericea in South Africa (Swain and Prinsloo, 1986), although none of these do any significant damage to the plant.

H. sericea is very similar to H. decurrens and has also been confused with H. gibbosa. The most obvious characteristic distinguishing H. decurrens from H. sericea are its thin fruits. H. gibbosa has longer, distinctly hairy leaves and larger fruits than H. sericea and H. decurrens.

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.

Mechanical Control

The most successful method to control H. sericea is the fell and burn technique where adult plants are cut down and left for 12 to 18 months before they are burnt. Shortly after the plants are cut down, the fruits, which have accumulated over the plant's lifetime, split open and fall to the ground. The released seeds germinate the following winter. The area is then burnt before the seedlings become reproductively mature. One or two follow-up operations are necessary after the burn to eradicate any regenerating or coppicing plants. This is an extremely important facet of the operation as it ensures that no plants are left to produce viable seeds. Although this is a very effective control method, the increased fire intensities using this technique can have a negative effect on sensitive ecosystems (Breytenbach, 1989). It is not advisable just to cut and then leave the cut plants unburnt, as fire is needed to kill any seedlings that may germinate. Burning standing plants can be effective in some cases but may result in dense stands of seedlings and widespread dispersal. The manual eradication of seedlings is both time consuming and expensive.

A massive campaign was initiated in South Africa in 1996 called the 'Working for Water' programme. The objective of this programme is to increase and sustain water supplies by eradicating alien invasive plants from all rivers, wetlands, catchments and other water resources countrywide. Alien invasive plants, especially Hakea spp., Pinus spp. and Australian Acacia spp. are known to reduce water run-off and restrict stream flow and are being cleared as part of this campaign.

Chemical Control

Chemical control has not played a large role in the control of H. sericea as it can have a negative effect on the indigenous vegetation. The costs of chemical control are also high as H. sericea occurs in dense thickets and inaccessible areas. Tebuthiuron is recommended for the control of H. sericea in South Africa.

Biological Control

A biological control programme against H. sericea was initiated in South Africa in 1962. Priority was given to seed-attacking insects and the first insect releases were made in 1970 (Kluge and Neser, 1991). The first agent released was the hakea seed weevil, Erytenna consputa (Curculionidae: Erirhininae), the larvae of which destroy the green developing fruits. The adults of E. consputa live for 2-3 years and feed on shoots, buds, flowers and young fruit as these become available. The eggs are laid in or near young developing fruits from late winter to mid-summer. The larvae develop singly in the young fruit and can consume more than one fruit before they complete their development. Following its establishment, E. consputa has spread throughout the South African range of the plant. The weevil has drastically reduced the annual seed production of H. sericea at some sites (Neser and Kluge, 1985; Kluge and Neser, 1991; Gordon, 1999).

The hakea seed moth, Carposina autologa (Lepidoptera: Carposinidae), the larvae of which destroy the seeds in mature fruits of H. sericea, was introduced in 1972 to augment E. consputa. The moth is univoltine with no diapause or quiescent phase. The eggs are laid singly on the surface of mature fruits in autumn. The larvae enter the fruits along the suture on the axial surface of the fruit. Only one larva develops per fruit. The mature larva emerges from the fruit in summer and pupates in the soil. Despite difficulties rearing and releasing this agent, it has now become established at a number of sites in South Africa (Dennill, 1987; Dennill et al., 1987; Gordon, 1993). Studies and observations at release sites in the field showed that the moth has reduced the mean number of accumulated seeds on H. sericea by up to 80% (Gordon, 1999). Despite these promising results several factors are limiting its effectiveness. Firstly, an indigenous fungus, Colletotrichum gloeosporioides [Glomerella cingulata] causes death and die-back of H. sericea in some areas and the fruits on infected trees split open and seeds fall to the ground resulting in larval mortality as the larvae are unable to move to new fruits. Secondly, the moths are unable to distinguish between healthy and previously attacked fruits for oviposition resulting in larval mortality as few neonate larvae successfully find and enter pristine fruits. Thirdly, regular wild fires in the Western Cape cause local extinction of C. autologa and they take a long time to recolonize regenerating plants.

A second weevil, Cydmaea binotata (Curculionidae: Erirhininae) was introduced into quarantine as a contaminant in one of the shipments of E. consputa and was successfully cultured in quarantine. They were found to be so damaging to H. sericea seedlings that it was felt that their feeding would enhance the infection rate of C. gloeosporioides [Glomerella cingulata]. Their biology and host-specificity was studied and when the weevils were found to be host-specific permission was granted for their release in 1979. The larvae are most damaging as they tunnel down the leaves or distal sections of soft stems. Releases were made at 36 sites throughout the range of the weed but weevils have only since been recovered at four sites (Kluge and Neser, 1991). The impact of the weevil on H. sericea has not been investigated in South Africa because their effect on the density of seedlings has been negligible.

The stem-boring beetle, Aphanasium australe (Coleoptera: Cerambycidae) was recently introduced not only to add impetus to the biocontrol of H. sericea project but also because it attacks the related weed Hakea gibbosa (Gordon, 2003). The larvae tunnel gregariously at the base of stems and in the sub-surface roots of the plant leading to stem bases developing a characteristic thickening due to the formation of scar tissue. Although A. australe does not kill mature plants growing under natural conditions, it is envisaged that trees subjected to additional stress i.e. from drought, may be killed by larval damage. The first releases of this agent were made during January 2000.

A flowerbud-feeding weevil, Dicomada rufa, is a promising agent that is being considered for release to negate perceived weaknesses in the programme for biological control of H. sericea. The effectiveness of E. consputa and C. autologa is being hampered by periodic wildfires. Regenerating H. sericea plants only set seed 2-3 years after a burn, causing local extinction of E. consputa and C. autologa as both agents require fruits for development. As D. rufa feeds on buds, flowers and succulent growth it is believed D. rufa could make a significant contribution by limiting fruit production at this critical stage (Gordon, 1999). The adults of D. rufa are 2-3 mm in length and dull grey-brown in colour. The larva is a cream coloured, legless grub that feeds on the flowers and succulent growth. Due to problems culturing this insect in quarantine, data on its host-specificity was obtained by developing a type of open-field testing method, the fixed- plot survey method, to show that D. rufa is host specific to H. sericea (Kluge and Gordon, 2004).

An indigenous fungus, Glomerella cingulata f.sp. aeschynomene, has been exploited for the biological control of H. sericea in South Africa. The fungus causes stem and branch lesions exuding quantities of colourless gum. Plants may be killed if lesions occurring on the lower regions of the main stem girdle the stem (Morris, 1981; 1991). Individual branches that are girdled can also die. The growing points of young seedlings are most susceptible resulting in necrosis extending down the stem killing the plant (Morris, 1981). Various methods of culturing and inoculating plants have been successfully developed. These include wound inoculations, knapsack and aerial applications of spore suspension of the fungus and the application of a dried preparation of fungal-colonized wheat bran onto young seedlings (Morris, 1989; Morris et al., 1999).

Integrated Control

The 'Working for Water' programme in South Africa is primarily involved with the mechanical control of H. sericea but has identified biological control as the only long-term solution to prevent further spread of the weed and the re-invasion of cleared areas. Although mechanical control is an extremely efficient method to control H. sericea, biological control needs to be in place to prevent re-invasion of the weed and limit the need for follow-up operations. However, large-scale eradication of H. sericea can lead to the local extinction of established biocontrol agents. The seed-feeding agents are particularly at risk because seedlings recolonizing burnt areas take a number of years before they set fruit. It is therefore essential that insect refuges or reserves be established in areas to be cleared. These insect refuges can then act as foci from which recolonization of re-invading H. sericea populations can occur and collections of agents for redistribution can be made. These reserves should be 1-5 ha in size, 10 km apart and consist of reproductively mature plants (Gordon, 1999).