Occurrence and Epidemic Adaptation of New Strains of Ralstonia Solanacearum Associated
Causes, Prevention and Cure of Invasive Weed Problems in New Zealand: Cytisus Scoparius, a Case Study
Paynter Quentin1, Fowler, Simon V.2 and Hayes, Lynley M2
1 Landcare Research New Zealand Ltd, Private Bag 92170, Auckland, New Zealand ()
2Landcare Research New Zealand Ltd, PO Box 69, Lincoln, New Zealand
Summary
The number of alien plant species established in the wild in New Zealand already outnumbers the native flora and more species naturalise every year from a pool of introduced species that amounts to approximately 10% of the world’s vascular plants.
Strategies for management of invasive weeds include risk assessment of novel plants and border controls to prevent new weed incursions. Work on recognising emerging nascent weed problems from newly naturalising species is still in its infancy. Attempts have been made to eradicate some species from some discrete regions, such as offshore islands, and methods to predict the impacts and spread of naturalising species are being developed so that management strategies can be enhanced.
Widely established species are controlled by a variety of measures, including herbicides, mechanical means and biological control. An example of a widely established invasive weed in New Zealand is Scotch broom Cytisus scoparius. Recent research to determine why it is invasive and how it might be controlled is discussed.
- Introduction – the problem
Approximately 2,100 alien plant species have naturalised in New Zealand, outnumbering the 1,900 native seed-producing plants (Wilton & Breitwieser 2000). Of the 240 serious environmental weeds listed on the New Zealand Department of Conservation database, approximately one-third (35%) originate from Eurasia with the rest coming from the Americas (mainly South America, 24%), Asia (16%), Africa (15%), and Australia (8%) (Owen 1998). Of these, about three-quarters (74%) of terrestrial weeds were deliberately introduced as ornamental plants, with a further 14% deliberately introduced for agriculture, horticulture or forestry (Owen 1998). Only 10% were introduced accidentally as contaminants with soil, animals or other plants. Similarly over half (54%) of aquatic plants were introduced as ornamental plants (Owen 1998). The number of known invasive weeds has been growing steadily indicating that the problems posed by invasive weeds are likely to significantly worsen. New Zealand is home to approximately 10% of the world’s c. 272,000 known plant species due to the importation of alien species for cultivation from around the globe (Duncan & Williams 2002). On average four new species naturalise each year (Esler 1988) in the Auckland region alone, where over 615 species were known to have naturalised by the 1980s (Esler & Astridge 1987). Furthermore, lag times are common between the introduction, establishment and spread of environmental weeds (Binggeli 2000). Many of today’s most seriously invasive weeds in New Zealand naturalised long ago. For example, gorse Ulex europaeus and Scotch broom Cytisus scoparius were first recorded as naturalised in 1867 and 1872, respectively. This indicates more recently established species are likely to become serious weeds in the future.
The costs of invasive weeds include loss of production (mainly agricultural and forestry products) and indirect costs including monitoring the border, implementing regional pest management strategies and weed control. The total cost of pastoral weeds alone (production losses plus weed control expenditure) was estimated in 1982 to be $340 million (losses) + $53 million (expenditure) = $393 million p.a. (Anon. 1982), values which equate to $968 million + $151 million = $1.1 billion present value (using Reserve Bank CPI index). The importance of individual species in this total cost is highlighted by more recent analyses: (1) current infestations of giant buttercup Ranunculus acris cause annual losses in milk solid production revenue of $156 million (2000-2001 prices; Bourdôt et. al. 2003); (2) current expenditure on pastoral farms in Otago and Southland on Californian thistle Cirsium arvense control and vaccine for associated scabby mouth disease in sheep is $27 million (Labes 2000) with a probable national expenditure of $135 million pa; (3) annual expenditure on management of nassella tussock Nassella trichotoma in Canterbury is $1.3 million (Brown Copeland 1995) while the projected loss in the absence of control is $25 million p.a. (Graeme Bourdôt pers comm.).
Damaging environmental losses are also incurred due to invasive alien weeds. In New Zealand initial research suggests that succession trajectories and nutrient flows are altered by weeds which are estimated to cause a loss in native biodiversity valued at $1.8 billion (1994 dollars; Williams & Timmins 2002) or $2.22 billion at present value.
2. Strategies for control
Prevention, containment and eradication
Strategies being developed for the management of invasive weeds include use of weed risk assessment models (WRA’s) and border controls to prevent new weed incursions (Rahman et al 2003). Work on recognising and prioritising emerging weed problems from newly naturalising species is still in its infancy but local government is already using WRA to prepare their pest plant strategies (e.g. Champion 1995) and attempts have been made to eradicate some species from some discrete regions, such as offshore islands (Coulston 2003). However, many introduced weeds are too widely established to attempt eradication. For these species attempts are being made to reduce their impacts.
Biological Control
Many weeds infest rangelands of relatively low economic value. This hinders chemical and mechanical control because of the relative costs involved. Biological control, which can be time consuming and initially costly to implement and has no guarantee of success, often represents the only economically attractive long-term management option (Paynter et al. 2003a). This is because the benefits of a successful biological control program are permanent, and the benefit-to-cost ratios are often highly favourable (e.g. Coombs et al. 1996). Although there have been recent concerns regarding the safety of biological control (e.g. Louda et al. 1997; Pemberton 1995), in New Zealand the safety record of weed biological control programmes has been generally good (Paynter et al. 2004), as has the success rate (Fowler et al. 2000a). Nevertheless, ongoing research is striving to further enhance the safety and effectiveness of biological control.
A major challenges for weed control scientists is to understand the causes of a weed invasion and use this information to improve the efficiency and probability of success of a biological control programme through the strategic selection of biological control agents and integration of biological control with other management techniques, where appropriate (e.g. Paynter et al. 2003b). Studying the population ecology of a weed is an important step towards these goals.
3. Scotch broom Cytisus scoparius as a case study
Throughout the world, biological control programmes are being conducted against introduced populations of the European shrub Scotch broom, Cytisus scoparius (L.) Link (Fabaceae), henceforth broom (Syrett et al., 1999), which is occasionally a minor weed in Europe (e.g. Rousseau & Loiseau 1982). Like many introduced shrubs, it has formed near-monocultures over large tracts of land in New Zealand, Australia (Parsons & Cuthbertson 1992), the USA and elsewhere (Syrett et al. 1999 et ante). In New Zealand its annual cost as a pasture weed and a weed of forestry, was estimated to exceed $13m in 1999 (Jarvis et al. 2003). The apparent vigour of broom in exotic habitats, compared to its native range was assumed to be due to an absence of specialist herbivores. A large-scale, but unreplicated insecticide exclusion study performed in England (Waloff & Richards 1977) indicated insect herbivory reduced plant growth by 50%, fecundity by 75% and increased the mortality rate of broom plants by 50% (Fig 1). Surveys indicated that, in contrast to the native 
range, broom plants in Australia and New Zealand are largely devoid of specialist insect 
herbivores (Hosking et al. 1998; Syrett et al. 1999; Memmott et al. 2000).
Figure 1. Effect of insecticide treatment on (a) height, (b) fecundity and (c) mortality of broom plants (Redrawn from Waloff & Richards 1977).
Models of broom population dynamics indicated that differences in three parameters, longevity, disturbance and the probability of stand regeneration following plant death, should explain the disparity in abundance between exotic and native populations (Fig 2; Rees & Paynter 1997). However, spatially explicit models are difficult to parameterise and seldom validated (Higgins et al. 2001) and the relative importance of these parameters was not known. Quantitative studies performed in both Europe, where broom is native, and in New Zealand and Australia, where it is an introduced weed were used to test predictions made by Rees & Paynter 
(1997).
Figure 2. Predicted relationships between (a) disturbance, (b) maximum plant longevity and (c) the probability that broom can regenerate following stand senescence on the amount of space occupied by broom (redrawn from Rees & Paynter 1997).
Longevity
Paynter et al. (2003b) studied the age structure of native and exotic broom populations, by counting the growth rings of broom plants and found that maximum plant ages did not vary significantly between countries. This seems counterintuitive, based on the findings of Waloff & Richards (1977). However, few of the populations that Paynter et al. (2003b) studied contained senescent individuals, so the maximum ages they recorded do not necessarily indicate the maximum potential longevity of broom. The probability of major disturbance that kills mature plants may be similar across countries so that few plants survive to their maximum potential age in both native and exotic habitats.
Disturbance
The importance of large-scale disturbance is more equivocal. Broom is clearly adapted to exploit disturbances that eliminate competing vegetation; its seeds are stimulated to germinate by high temperatures caused by fire (Tarrega et al. 1992) and weedy outbreaks have been linked to fire (Paynter et al. 1998; Rousseau & Loiseau 1982), soil cultivation or instability (e.g. Paynter et al. 1998; 2000; Rousseau & Loiseau 1982; Williams 1983) and overgrazing by cattle (Williams 1983). Therefore, it was no surprise that major disturbance (cutting down existing stands and cultivating the soil) enhanced seedling establishment and survival in both Europe (Paynter et al. 1998; Paynter et al. 2000) and Australia (Downey & Smith 2000; Sheppard et al. 2002). However, Sheppard et al. (2002) reported that broom seedlings also recruited beneath relatively undisturbed mature broom stands at two localities in Australia, whereas seedlings did not recruit beneath mature stands in two French populations (Paynter et al. 1998). Similarly, simulating the early death of a broom stand (cutting down broom plants, but leaving the soil undisturbed) resulted in extensive regeneration of broom in Australia (Sheppard et al. 2002), but in Europe, competing herbs and grasses smothered many broom seedlings (Paynter et al. 1998). It is also suggested that the ‘rain’ of specialist insect herbivores and pathogens from parent plants is responsible for high seedling mortality beneath European stands (Fowler et al. 1996).This result is unlikely to be due to the Australian flora being intrinsically less competitive than in Europe because broom seedling establishment was actually lower in native Australian grasslands than in improved pastures, where European pasture grasses had been introduced (Sheppard et al. 2002).
Although these results support Waloff & Richards (1977) notion that broom should be more vigorous and therefore more competitive in exotic habitats due to a lack of natural enemies, Paynter et al. (2003b) found no evidence that broom grows faster or taller in exotic habitats. This unexpected result could be explained by differences in broom population densities, which were found to be significantly higher in exotic habitats than in Europe (Paynter et al. 2003b; Fig 3):
Figure 3. Mean population densities of all broom plants and mature broom plants only (excluding seedlings) in Australia (solid), New Zealand (stippled) and Europe (diagonal bars). Columns within the same category with the same letter are not significantly different (LSD). (Figure from Paynter et al. (2003b)).
Intraspecific competition reduced the growth rate of broom seedlings in both native (Paynter et al. 1998) and exotic (Sheppard et al. 2002) habitats. More intense intraspecific competition could, therefore, have concealed any potential release from the impact of specialist insect herbivores on these aspects of plant performance in broom’s introduced range. However, Paynter et al. (1998) noted that phytophagous arthropods were slow to colonise broom seedlings in France. Similarly, Waloff & Richards (1977) noted that in England, while aphids and psyllids rapidly invaded young broom plants, populations of Phytodecta (Gonioctena) olivacea (Forster) and Sitona regensteinensis (Herbst.) did not become abundant until plants were in their sixth year. Therefore, it seems that the impact of insect herbivory on broom is greatest on mature, established plants rather than younger, rapidly growing plants.
Probability of stand regeneration following plant death
Paynter et al. (2003b) found that the frequency of broom populations with relatively stable age distributions (Fig. 4) – i.e. described by a significant ‘reversed-J’ shaped relationship between Loge (number of individuals) and Loge (plant age) (see Ågren & Zackrisson 1990) significantly varied between countries and was highest in Australia (73% of populations) and lowest in Europe (18% of populations). This supports the hypothesis that exotic populations are more commonly self-replacing in the absence of large-scale disturbance than native populations, where it is more usual to find stands of relatively uniform age that have been created by disturbance. Following senescence and death, native stands are more likely to persist only as a soil seed bank until the next disturbance event creates conditions for re-establishment of another uniformly aged stand (Rees & Paynter 1997). By contrast, in exotic habitats stands may persist in the absence of disturbance (Sheppard et al. 2002).
Figure 4. Examples of age structures of Cytisus scoparius populations (a-f, exotic populations; g-i, Native populations) ordered as follows: (a) Kakapo Brook, New Zealand, recently replanted forestry; (b) Hanmer, New Zealand, braided river levee; (c) Glenroy, New Zealand, wasteland; (d) Havelock, New Zealand, Forestry; (e) Tomalla, Australia, undisturbed native bushland; (f) Krawaree, Australia, invaded pasture; (g) L’Esperou, France, periodically burnt heathland; (h) St André, France, native heathland; (i) Tapis, Spain, roadside verge. The data are presented in 1-year age classes. (Figure reprinted from Paynter et al. (2003b)).
Implications for biological control of broom
Together, these studies indicate how broom dominates exotic habitats: Increased seed production, due to an absence of specialist insect herbivores (Waloff & Richards 1977) results in increased fecundity and, therefore, increased population densities because broom seedling densities are directly related to seed bank size. At densities commonly encountered in the field, intraspecific competition reduced seedling growth, but not survival to flowering age or probability of flowering, in both native (Paynter et al. 1998) and exotic (Sheppard et al. 2002) habitats.
Furthermore, in their native range mature broom plants rarely die in one season: branches die and plants degenerate over time in a process accelerated by insect herbivory (Waloff 1968; Waloff & Richards 1977). Therefore, it seems that increased fecundity and reduced morbidity of exotic broom plants results in populations that are so dense and healthy that competing species are more likely to be excluded. This would explain why competition with regenerating grasses and herbs curtailed broom seedling regeneration after cutting down broom stands in Europe (Paynter et al. 1998), but not in Australia (Sheppard et al. 2002). Therefore, disturbance that diminishes competing vegetation is less frequently necessary for broom re-establishment in exotic populations.
It follows that exotic populations of broom should be regulated by biological control if seed-feeders are used in conjunction with agents that defoliate mature plants, enabling competing vegetation to establish beneath stands, further inhibiting regeneration from the seed bank. Indeed, a similar woody legume shrub Sesbania punicea (Cav.) Benth. has already been successfully controlled by such a combination of biological control agents in South Africa (Hoffmann & Moran 1998). Nonetheless, the broom biological control programme has made slow progress in New Zealand (Fowler et al. 2000b). Reasons for this and potential solutions are discussed below.
Barriers to the successful biological control of broom
Scotch broom is only distantly related to native members of the native New Zealand Fabaceae, so specialised insect herbivores that feed on broom should pose little threat to New Zealand plant species (e.g. Briese & Walker 2002). However, there is significant interest by AgResearch Ltd (New Zealand) in promoting the use of another exotic plant, the closely related tree lucerne (tagasaste) Cytisus proliferus L.F. for forage production in dry areas, and by some regional councils for erosion control on marginal hill country.
A submission made to the Ministry of Agriculture and Forestry (MAF) by Landcare Research to import the European Chrysomelid beetle Gonioctena olivacea, a potential biological control agent of broom into New Zealand (Syrett et al. 1997) was rejected in 1998. MAF required better cost-benefit data on broom as a weed in New Zealand, and a more detailed assessment of the risk to tree lucerne from potential G. olivacea damage.
A report was prepared in response to these requirements, which considered the costs and benefits of broom control, assuming non-target attack to tree lucerne were to occur (Jarvis et al. 2003). The report considered three scenarios, a 25%, 50% or a 95% reduction in broom. The benefits of tree lucerne were shown to be minor, compared to the costs of Scotch broom and both 50% and 95% control had significant net benefits to the New Zealand economy. However, farmers and foresters indicated there would be only minor benefits at the 25% control level, whereas beekeepers, who consider Scotch broom a valuable pollen source, contend that major costs would still be accrued at this level of reduction. Therefore the net benefit to the New Zealand economy was predicted to be rather minor at the 25% control level (c. NZ$500k p.a.).
However, there are reasons to suppose beekeepers have overestimated the impact of a 25% reduction in broom on the availability of pollen to honeybees. Broom flowers are fused shut and require forced ‘tripping’ by a pollinator to gain access to pollen (Parker 1997). It is, therefore, possible to detect flowers that have been successfully visited by a pollinator, and those that have not. In the Pacific northwest of the USA, where broom is an invasive weed, only 3-28% of broom flowers were tripped by a combination of honeybees and bumblebees (primarily bumblebees, indicating honeybees harvested pollen from <30% of flowers; Parker 1997). Similarly, only 40% of flowers were visited by effective pollinators in Japan (Suzuki 2000). Fecundity of both the Japanese and US broom populations was pollinator-limited. Studies in Europe and Australia, also found evidence for low reproductive success, with only c. 25% of labelled flowers in both countries, resulting in pods (Groves & Paynter 1998).