Rubus argutus
(sawtooth blackberry)

R. argutus, commonly known as the sawtooth blackberry, is native to central and eastern USA and has been introduced to the Hawaiian Islands, Chile (Markin et al., 1992), New Zealand (USDA-ARS, 2012) and Japan (Mito and Uesugi, 2004).

It has been declared a noxious terrestrial weed in Hawaii, Smith (1985) listed it among the 12 most significant environmental weeds and Meyer (2000) listed it among the 19 dominant invading plant species. It is capable of spreading both via seed and vegetative growth and competes with native vegetation (Tunison, 1991). It spreads most rapidly with disturbance; seedling emergence is stimulated by soil disturbance and vegetative growth is much more rapid if the canopy is removed. Even when above-ground shoots are killed by cutting or spraying, it may re-sprout from well-developed roots (Smith, 1985). R. argutus may work synergistically with other invasive species such as wild pigs to reduce environmental quality (Stone, 1985), but not in all settings (Aplet et al., 1991). While it is effective in recolonizing areas with a moderate or opened canopy is disrupted, growth of R. argutus is inhibited by deep shade.

R. argutus is native to central and eastern USA.

Although known to be established on Kauai, Oahu, Maui and Hawaii throughout most of the 20^th century, R. argutus was first reported on Molokai in 1992, naturalized and forming scattered patches from Hanalilolilo to Kawela between 700 and 1120 m above sea level (Hughes, 1995).

In New Zealand it occurs in North Auckland, specifically Pipiwai, Whangarei district and Kamo (Webb et al., 1988).

Although apparently only introduced to four major areas outside its home range (Hawaii, Chile, New Zealand and Japan), R. argutus is capable of surviving in many other similar temperate-subtropical regions such as many of the South Pacific Islands. R. argutus is not yet present in Australia, but blackberry (R. anglocandicans) was declared a weed of national significance there in 1999 (Evans and Weber, 2003). A weed risk assessment of R. argutus completed for Hawaii scored a relatively high value of 21.5 (PIER, 2012).

In Hawaii, R. argutus is naturalized in disturbed habitats within mesic to wet forest and subalpine grasslands, ranging between 200 and 2300 m in elevation (Wagner et al., 1999). In its native range, R. argutus is prevalent in grassland habitats, riparian areas, open forest habitats and forest edges (Hilty, 2012).

R. argutus competes with pine seedlings or saplings during natural regeneration of pine stands such as loblolly, Pinus taeda and shortleaf Pinus echinata pines in southeastern USA (Cain and Shelton, 2003; Schabenberger and Zedaker, 1999). The plant may also invade areas such as roadsides planted with grasses such as bahiagrass (Paspalum notatum) in the coastal plain region of southern USA (McCarty et al., 1996). When invading natural habitats, it may compete with rare species, as in Hawaii where it threatens the IUCN red-listed Alsinidendron obovatum and Alsinidendron lychnoides (Bruegmann and Caraway, 2003a; 2003b).

Genetics

R. argutus is tetraploid (Naess et al., 1998). Chromosome numbers have been reported as either 2n = 14 or 21 (Wagner et al., 1999).

The invasive R.phoenicolasius exhibits less genetic variation than R. argutus in the latter’s native habitat. It is likely that R. argutus exhibits more outcrossing through a higher proportion of sexual reproduction than R. phoenicolasius (Innis et al., 2011).

Reproductive Biology

R. argutus has been reported as self-incompatible (Keep, 1968), and by comparison to R. phoenicolasius was seen to rely more on pollination for fruit maturation (Innis, 2005). R. argutus is regularly visited by pollinators in May-June (Goltz, 1987).

R. argutus propagates by means of seed and by vegetative growth. Spreading clonal growth occurs via root sprouting; aerial shoots also occasionally root when contacting soil (Smith, 1985; Tunison, 1991).

Seeds of R. argutus were not resistant to fire when placed directly on litter prior to burning; however, seeds some seeds retained in fruits during burning were still able to germinate and had germination percentages equivalent to the control seeds (Cain and Shelton, 2003).

In Arkansas, an average of 63 seeds per fruit was calculated by Shelton and Cain (2002). A rough estimate of fruit production was calculated at 50 fruits per square m (PIER, 2012).

R. argutus produces a persistent seed bank; after a third year of storage in soil on a forest floor, 7-19% of R. argutus seeds were still viable (Shelton and Cain, 2002). The germination of R. argutus improves with exposure to sulfuric acid for up to 1 hour, demonstrating the adaptation of the seeds to animal ingestion (Brinkman, 1974).

Physiology and Phenology

In its native range, R. argutus flowers in April-May and its fruits mature in May-June (Johnson and Hoagland, 1999). Normally first year canes (primocanes) do not flower, whereas second year canes flower (floricanes) and subsequently senesce (Tunison, 1991; Hilty, 2012).

R. argutus is more or less deciduous in Hawaii, particularly in more exposed sites, upper elevations or when exposed to low winter temperatures (Tunison, 1991). In New Zealand flowering occurs from February to May and fruiting during May (Webb et al., 1988).

Although aerial portions may be destroyed in fire, R. argutus can regrow from its roots (University of Hawaii Botany Department 2012).

In the Hawaiian environment, R. argutus exhibited a specific leaf area of 187.4 square cm per g, a foliar nitrogen level of 1.76 % and a foliar phosphorus level of 0.08% (Baruch and Goldstein, 1999). The specific leaf area for R. argutus was higher than the average for invasive species than for native Hawaiian plants (167.6 square cm per g) and the foliar nitrogen level higher than the average native value (1.36%), whereas the foliar phosphorus level was the same as for native species (0.08%) (Baruch and Goldstein, 1999). Inositol has been isolated from tissues of R. argutus (Sando, 1926).

It was demonstrated that exposure to high levels of tropospheric ozone reduced the nutritional quality of R. argutus, thus potentially impacting wildlife (Ditchkoff et al., 2009).

Longevity

The roots are perennial but the canes biennial. Second year canes senesce, but new canes arise from the roots, and thus clonal patches of R. argutus may persist for many years.

Activity patterns

Spread of R. argutus is favoured by disturbance; soil disturbance encourages seedling recruitment whereas canopy opening encourages clonal recruitment (Tunison, 1991). In some areas, R. argutus reproduces only by vegetative means (Harrington et al., 1997).

Population Size and Structure

In its native habitat in eastern USA, R. argutus is an earlier colonizer, forming nascent populations more rapidly than many other woody species (Rossell et al., 2008). It may occasionally reach high populations in its native range, forming dense thickets (White and Judd, 1985). Unlike other invasive plants in Hawaii such as Clidemia hirta or Psidium cattleianum, R. argutus does not always form high densities after colonizing an area, but tends to form widely distributed interspersed patches and only forms monotypic stands under certain conditions (Tunison, 1991). In open areas, stems tend to be more erect or somewhat arching by comparison to shaded areas, where the stems are more trailing or decumbent (Tunison, 1991).

Nutrition

At a phosphorus-limited site in Hawaii, densities of R. argutus increased with the addition of phosphorus or nitrogen and phosphorus (Ostertag and Verville, 2002).

Associations

Arbuscular mycorrhizal fungi were associated with R. argutus in its native habitat (Innis, 2005).

Environmental Requirements

R. argutus is adapted to a broad range of soil textures (USDA-ARS, 2012). The plant grows best in full or partial sun but under drier conditions can become desiccated (Hilty, 2012). Habitats ranging from wet to mesic are suitable (Smith, 1985), including boggy sites (Tunison, 1991). Its drought tolerance is only moderate and it does not tolerate salinity (USDA-NRCS, 2012). R. argutus exhibits resistance to frost, fire and waterlogging.

A number of natural enemies of R. argutus have been evaluated as biological control agents. Five insects, Schreckensteinia festaliella, Croesia zimmermani, Priophorus morio, Bembecia marginata, and Chlamisis gibbosa, were released as biocontrol agents for R. argutus in Hawaii (Nagata and Markin, 1986). Only the lepidopterans S. festaliella and C. zimmermani have successfully established and provided some control, although they also impact native Hawaiian Rubus species (Nagata and Markin, 1986). Three fungal rust pathagens have also been evaluated for biological control: Gymnoconia nitens, Kuehneola uredinis and Phragmidium violaceum. None of these have been recommended for Hawaiian release because of a lack of specificity to R. argutus (Gardner and  Hodges, 1983; Markin et al., 1992; Gardner et al., 1997). It does not seem that P. violaceum attacks R. argutus, although it is utilized in Chile for biological control of other Rubus species (Markin et al., 1992; 1993).

Environmental

• Amenity
• Landscape improvement
• Wildlife habitat
• Windbreak

General

• Sociocultural value

Human food and beverage

• Beverage base
• Fruits

Due to frequent hybridization and introgression events, identification of Rubus species can be challenging (Davis, 1990; Johnson and Hoagland, 1999). Thimbleberry (Rubusrosifolius Sm.) is also naturalized on Hawaii and bears white flowers like R. argutus; however, the mature fruits are red. The fruit of the endemic Rubusmacraei A. Gray may be dark purple but range from dark red to dark purple and its leaves are larger and broadly ovate in comparison to those of R. argutus. The Himalayan blackberry, R. armeniacus [R. discolor] has recently invaded  Hawaii; it also bears white flowers and black fruit, although the flowers may also be pinkish. The leaves of R. armeniacus are light greenish/gray below and somewhat less toothed, larger and more oval than those of R. argutus; new growth in R. armeniacus tends to bear leaves with five leaflets (older canes having three) and it tends to form, thick, shrubby patches (Boersma et al., 2006). R. argutus is closely related to R. ostryifolius (Johnson and Hoagland, 1999), although R. ostryifolius is also considered a synonym (Hilty, 2012). Both R. argutus and R. allegheniensis have racemes with relatively small bracts, but the racemes of R. allegheniensis have stalks with sticky-glandular hairs whereas those of R. argutus have woolly non-glandular hairs (Hilty, 2012).

In comparing the ecology of R. argutus in its native range with the invasive R. phoenicolasius, Innis (2005) determined that R. phoenicolasius populations persisted better under drought conditions and competed more vigorously with other plant species.

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.

Prevention

In Australia, the risk of invasion by R. argutus has been evaluated and the species is included in efforts to prevent new invasive plants from entering the country and becoming established (DPI, Victoria, 2011).

Prevention of re-establishment of R. argutus is important in large-scale restoration projects, such as a restoration effort in Hawaii Volcanoes Park where thousands of invasive plants were removed (Loh et al., 2007).

Cultural Control and Sanitary Measures

In many areas where R. argutus is invasive in Hawaii nutrients are limited. Increasing nutrient levels increase the abundance of R. argutus and at the same time decrease populations of native species such as Metrosideros polymorpha that are adapted to lower nutrient levels (Ostertag and Verville, 2002).

Physical/Mechanical Control

It is possible to reduce vegetative growth in R. argutus by various physical means of control such as mowing, roller chopping or fire, but other weed growth may be stimulated by these actions (Mislevy et al., 1987). Populations of the closely related thornless blackberry, R. canadensis (Davis, 1990), may increase with burning, mowing or deep cultivation (Coladonato, 1994).

Biological Control

There has been a priority placed on developing biological control methods for R. argutus in Hawaii because large-scale applications of herbicides or manual control in the environments where R. argutus is found is impractical (Gardner et al., 1997).

Five insects were released to target R. argutus for biological control between 1963 and 1969, three of which became established by the mid-1960s and continued to impact R. argutus populations subsequently (Nagata and Markin, 1986). The two most successful insect agents were Schreckensteinia festaliella, a leaf skelentonizer, and Croesia zimmermani, a leaf roller (Nagata and Markin, 1986). The third biocontrol insect persisting in Hawaii is the defolilating sawfly Priophorus morio, but its incidence is much lower than that of the other two introduced insects that are still prevalent (Nagata and Markin, 1986). The chrysomelid beetle Chlamisus gibbosa was also introduced to Hawaii in 1969 as a biological control agent for R. argutus (Davis, 1970).

When these insects were released in the 1960s, impacts on native Hawaiian Rubus species were not considered, but these insects do have negative impacts on native Rubus species (Tunison, 1991). The emphasis has since shifted to finding fungal agents rather than additional insects that might interfere with the three insects already present in Hawaii (Markin et al., 1992). Monitoring and research on these insect agents continues. C. zimmermani and S. festaliella were found to be incapable of acting as hosts to parasitoids introduced to Hawaii for controlling other pests (Henneman and Memmott, 2001). Pheromone-based monitoring methods for C. zimmermani and S. festaliella have been developed (Suckling et al., 2006).

In Chile, a European rust fungus (Phragmidium violaceum) was tested against R. argutus after being utilized successfully against other Rubus species in Australia (Oehrens and Gonzales, 1974; Bruzzese and Field, 1984; Markin et al.,1992). Although it was introduced to Chile to control Rubus species, it apparently does not attack R. argutus (Markin et al., 1992; 1993).

The rust fungus Gymnoconia nitens attacks R. argutus throughout its native range (Gardner et al., 1997). Gardner et al. (1997) tested G. nitens on a variety of Rubus species, either native or introduced to Hawaii, for possible utilization as a biological control agent in Hawaii. Although G. nitens showed some promise for controlling invasive Rubus species, including R. argutus, because it also caused disease on native the Rubus species R. hawaiensis and R. macraei, its use for biological control on islands where these species occur was not advised (Gardner et al., 1997). Similarly, the leaf rust Kuehneola uredinis, although effective as a potential agent against R. Argutus, had some impact (although less severe) on both native Hawaiian species (Gardner and Hodges, 1983).

Chemical Control

Cut stem application of glyphosate can be effective for small, scattered populations of R. argutus, with stems cut as close to the soil as possible (Smith, 1985; Tunison, 1991). The cut-stem method can be quite labour-intensive when high populations of R. argutus are present (Tunison 1991). Glyphosate or metsulfuron may also be used as a foliar application, but care must be taken to avoid damage to non-target plants (Tunison, 1991). Both triclopyr and picloram had high efficacy against R. argutus whereas dicamba did not (Motooka, 1986; Motooka et al., 2003). For clearing trails of R. argutus in Kauai, drizzle applications of triclopyr and glyphosate have proven effective (Motooka et al., 2003). Various formulations of triclopyr utilizing surfactants have been also tested in Hawaii (Motooka et al. 2003).

In a 3-year investigation of R. argutus control in bahiagrass in Florida, McCarty et al. (1996) found triclopyr was most effective. Combining triclopyr with 2,4-D, dicamba, hexazinone or sulfometuron added some measure of control, but also increased damage to bahiagrass. Fluroxypr provided 60-80% control of R. argutus. Although glyphosate and glufosinate also provided some control of R. argutus, the damage to bahiagrass turf was unacceptable and glyphosate is not always effective during the growing season (Yonce and Skroch, 1989; McCarty et al., 1996). Santos et al. (1992) specified that although glyphosate may not be 100% effective, a measure of control is useful in some situations. McCarty et al. (1996) found that emergence of blackberry seedlings was not controlled by application of oxidiazon.

10/10/12 Original text by:

David R. Clements, Consultant, Canada