Biological control of silverleaf nightshade, Solanum … · 2015. 7. 6. · elaeagnifolium Cav....

.4griculture. Ecosystems and Environment, 37 ( 1991 ) 137-155 Elsevier Science Publishers B.V., Amsterdam

137

Biological control of silverleaf nightshade, Solanum elaeagnifolium, and bugweed, Solanum

mauritianum, (Solanaceae) in South Africa

T. Pickers ~ and H.G. Zimmermann b "~Plant Protection Research Institute, P.O. Box 330, Uitenhage 6230, South Africa

"Private Bag X134, Pretoria 000 i, South ,,Ifrica

(Accepted 25 March 1991 )

ABSTRACT

Pickers, T. and Zimmermann, H.G., 1991. Biological control of silverleaf nightshade, Solarium elaeagn(folium, and bugweed, Solarium mauritianum, ( Solanaceae ) in South Africa. Agric. Ecosys- tems Era'iron., 37:137-155.

The weed status and biological control of two introduced solanaceous plant species in South Africa are reviewed. Solanum elaeagn~folium Cav. (silverleaf nightshade, satansbos), indigenous to the southern USA, Mexico and Argentina, is a weed of arable and pastoral land. Mechanical and herbi- cidal attempts at control have been largely unsuccessful. Faunistic surveys in the countries of origin of S, elaeagn~folium revealed at least 15 potential bioconlrol agents, of which only a fruit-boring ge- lechiid moth, Frumenta nephelomicta (Meyrick), has so far been released, but failed to establish.

Solamcm mauritiamml Scop. (bugweed, bugtree), indigenous to Argentina, Brazil and Uruguay, is a weed mainly of commercial forests in the high-rainfall areas of South Africa. The search for potential biocontrol agents has been hampered because the taxonomic status orS. mauritianum is unclear and only three insect species have been introduced into quarantine.

Biocontrol of both Solanum weeds has been hindered by the oligophagous habits of the herbivorous species tested so far. The problem is compounded by the large number of agronomic species in the Solanaceae. In particular, Solanum melongena L. (eggplant) is acceptable to most solanaceous folivorous agents, at least under cage conditions, and only the more specialized, endophagous herbivore species may be suitable for introduction into South Africa. However, evidence is presented that the threat of folivorous agents to eggplant cultivations may have been overrated and prospects for the eventual release of these oligophagous agents are discussed.

INTRODUCTION

The genus Solanum, which comprises about 75% of the Family Solanaceae (Symon, 1981 ), is of major agricultural importance worldwide. Solanum tuberosum L. (potatoes) and Solanum melongena L. (eggplant, brinjals) are the most important cultivated species, with Solanum macrocarpon L. (a type of brinjal) a lesser known crop in some parts of Africa (Jaeger and Hepper,

01 ~7-8809/91/$03.50 © 1991 Elsevier Science Publishers B.V. All rights reserved,

138 T. OLCKERS AND H.G. ZIMMERMANN

1986; Bukenya and Hall, 1987). Potatoes are cultivated extensively in South Africa, while eggplant is of lesser importance (Annecke and Moran, 1982 ).

In South Africa, the genus Solarium comprises 51 indigeneous and 13 exotic species (Gibbs Russell et al., 1987 ). The Solanum species in South Africa are ecologically similar to those in Australia (Symon, 1981 ), occupying mainly disturbed sites, and are rarely components of climax vegetation. Conse- quently, about 40% of South Africa's Solanum flora ( 15 indigenous and 10 exotic species) are regarded as problem plants (Wells et al., 1986 ).

The two most invasive of the exotic species in South Africa, Solanum elaeagnifolium Cav. (silverleaf nightshade, satansbos) and Solanum mauritianum Scop. (bugweed, bugtree), are weeds of agriculture and commercial forests, respectively. Solanum mauritianum is already widespread in South Africa and continues to spread (Neser et al., 1990), while S. elaeagnifolium infestations are localized, but with the potential for further expansion (Was- sermann et al., 1988). Neither mechanical nor chemical control procedures have contained the spread of either of these weeds.

In 1973, South Africa became the first country to import and test natural enemies against Solanum weeds. Previously, no Solanaceae had been sub- jected to biological control attempts (Julien, 1987). In Australia, biological control of S. elaeagnifolium was considered, but rejected as unlikely to suc- ceed (Wapshere, 1988).

The aim ofthis review is to document the biological control ofS. elaeagnijblium and S. mauritianum in South Africa, and to discuss the future of this work. In this paper, the origin and taxonomic status of the weeds, their his- tory of introduction and their subsequent spread in South Africa are summarized. Attempts at chemical and mechanical methods of control are de- scribed, and the search for suitable natural enemies and the biology of the introduced agents are considered. The unusual problems facing biological control of Solanum weeds may dictate the adoption of unconventional ap- proaches that may be of international interest.

ORIGIN AND TAXONOMIC STATUS OF S. EL,4E.IGNIbDLIUM

Solanum elaeagn(folium (Fig. 1 ) is indigenous to northeastern Mexico, the southwestern USA and possibly Argentina (Goeden, 1971; Zimmermann, 1974; Boyd et al., 1983; Wapshere, 1988). It has now spread to India, Aus- tralia, Chile, North Africa, Mediterranean countries and southern Africa (Wassermann et al., 1988 ), and is also troublesome in some American states such as Arizona, California and Texas (Robinson et al., 1978). Solanum elaeagn(folium displays considerable morphological variation in the Ameri- cas, which has confused its taxonomic status. Although numerous subspecies were proposed by Morton (1976), Symon (1981) regards these as natural variants. The problem of the exact identity and origin of the form in South

BIOI,()(II('AL CONTR()L OF SOL..I.~'I 'M EL..II-21GNII~OLIUM AND S. M..IL'RI1L.INLIM 139

1;

8

W

Fig. I. Silverleaf nightshade, S. elaeagn(h~lium. (a ) Mature slem: I b) young stem with flowers: (c) fruits. (Drawn by A. Waiters, National Botanical Institute, Pretoria. )

Africa is exacerbated by secondary weedy distributions of S. elaeagnifof~:~m in the USA. Goeden ( 1971 ) and Wapshere ( 1988 ) used the diversity of herbivorous insects on the plants as an indicator of the centre of origin, which is probably in the southwestern Texas or northeastern Mexico region, but may be further south in Mexico.

INTRODUCTION AND SPREAD OF S. E L A E A G N i ~ O L I U M IN SOUTH AFRICA

Solanum elaeagnifolium was initially thought to have been imported as a contaminant of hay during the 1940s or 1950s (De Beer, 1985). There are, however, indications of an earlier accidental importation with pig fodder at Kendrew (32°31'S 24°31'E) in the Eastern Cape in about 1905 (Wasser-


mann et al., 1988). An infestation observed at Wolmaransstad (27 ° 12'S, 25°31'E) in the western Transvaal, in about 1919 (De Beer, 1985), may have originated from Kendrew. The plant was officially recorded as a problem in 1952 and subsequently declared a weed in 1966 (Henderson and An- derson, 1966).

The largest infestations of S. elaeagnifolium occur at Kendrew (De Beer, 1985 ). There are also large infestations in the Western Transvaal and on the Springbok Flats in the Northern Transvaal, while smaller populations occur in Natal, the Southern Cape interior and the Orange Free State (De Beer, 1985).

Solanum elaeagniJblium requires an annual rainfall in ¢;:ccss of 30~ mm, but is apparently not restricted by soil type (De Boer, ~ 0',~,5 ). I1 occurs mainly on disturbed soil, e.g. neglected and grazed land, along roads and fences, and in water furrows, in arable lands, it forms dense stands which compete with cultivated crops. The weed may replace natural vegetation in areas of over- grazed veld and in heavily trampled areas around watering holes.

Propagation occurs via seeds and vegetatively through the spreading root system which sprouts to form new plants. Sprouting is enhanced by the re- moval of the aerial parts of the parent plants or by cultivation which stimu- lates coppice growth from cut root sections. Seeds are dispersed by water, agricultural implements and vehicles, in bales of hay or lucerne and in the dung of cattle, sheep and guinea fowl which ingest the fruit (De Beer, 1985 ) Although the plants die back in winter, ripe dry fruit are retained on dead branches and may be dispersed by wind.

USEFUL ATTRIBUTES OF X EL.,IF..IGNll.'OLIU,II

Solanum species are a rich source of steroidal alkaloids (Schreiber, 1979) used in the synthesis of contraceptive and corticosteroid drugs. Solasodine, the most important of these precursors, has been commercially extracted from S. elaeagnijblium in india (Maiti and Mathew, 1967). Exploitation of this resource is under investigation in South Africa.

In spite of confirmed toxicity in some situations (De Beer, 1985), S. elaeagniJblium has apparent value as a fodder with a high protein content. Wassermann et al. (1988) recorded that ~attle and wild antelope browse on the weed during spring and early summer, when natural grasses are unpalat- able, without showing symptoms of poisoning. The plant is thus viewed by some as a useful, drought-resistant fodder. This contrasts with the situation in Texas (Doilahite and Allen, 1960) and in Australia (Molnar and Mcken- zie, 1976 ) where stock losses through poisoning have been reported.

BIOLOGICAL CONTROL OF SOLANUM EL..IE..IGNIFOLIUAI AND 5: M..IURIIT.4NUM ] 41

MECHANICAL AND CHEMICAL CONTROL OF S. E L A E A G N I F O L I U M

Trials in the USA showed that appropriate management practices in arable land can bring S. elaeagnifolium under control within 3 years (Davis et al., 1945; De Beer, 1987). In winter crop situations, regular tillage weakens the S. elac.:;gn(fohum plants because no winter growth occurs. With summer crops, where there is competition for moisture and light, dense stands of a crop are able to suppress the weed. Regular tillage or clearing in summer also prevents seed set and depletes plant reserves, while lack of irrigation in cleared areas discourages regrowth. Mechanical control is more difficult outside arable land, although intensive browsing by stock can reduce fruit set (Wassermann et al., 1988).

A wide range of herbicides, including 2-methyl-4-chlorophenoxyacetic acid (MCPA), Picloram and 2,4-dichlorophenoxyacetic acid (2,4-D), have been used in chemical trials against S. elaeagn~folium in South Africa since 1952. However, inconsistent efficacy of the herbicides has precluded them from for- mal registration and recommendations for chemical control are presently restricted to the use of 2,4-D to inhibit grewth and to prevent seed set (Wasser- mann et al., 1988).

BIOLOGICAL CONTROL OF S. E L 4 E A G N I F O L I U M

The herbivorous insects associated with S. elaeagn(folium were first surveyed in 1966-67 in the southern states of the USA (Goeden, 1971 ). Later, in 1971, searches for potential control agents were conducted in Argentina (Zimmermann, 197.~). The insect herbivore faunas in Mexico, Arizona and New Mexico were mo~e diverse and included more apparently monophagous species than those in Argentina, suggesting that S. elaeagniJblium is a relatively recent introduction into Argentina. Consequently, later surveys have been largely confined to the USA an,'~ Mexico (Neser, 1984).

A nematode and seven insect species have been imported into quarantine in South Africa and evaluated. Four species were rejected because of insuffi- cient host specificity (Table I ), one was released, and the remainder are still under consideration (Table 2 ). The biology, host specificity and status of the most promising agents is discussed.

Gratiana lutescens (Boh.) and Gratiana pallidula (Boh.) ( .'hr?,somelidae: Cassidinae)

The biology and host preferences of Gratiana lutescens, from Argentina, and Gratiana pallidula, from the southern USA, are very similar. In fact, under insectary conditions, successful cross-breeding produced fertile offspring which, together with their morphological similarities, suggests that they are geographically isolated subspecies rather than true species (Siebert, 1975).


TABLE I

Phytophagous insects from Argentina (Arg) and Texas (Tex) introduced, but subsequently rejected, as potential biocontrol agents ofS. elaeagnifolium in South Africa

Agent From Year Damage Reason for rejection

Gratiana Arg 1 9 7 3 Defoliation by Narrow host range in genus lutescens adults and Solanum, but attacks (Chrysomelidae) larvae eggplant and some

indigenous Solatmm species (Siebert, 1975 )

Gratiana pallidula (Chrysomelidae)

Tex 1 9 7 3 Asabove Asabove

..lrvelius Arg 1974 Adults and albopunctatus nymphs ( De Geer) destroy seeds (Pentatomidae)

('onotrachelus Arg 1974 Larvae t,isignatus ( Boh. ) destroy seeds (Curculionidae)

Wide host range within Solanaceae includes crops, e.g. eggplant, tomatoes and peppers (Siebert, 1977). Also listed as a pest of tomatoes, cotton and soybeans in South America (Zimmermann, 1974 )

Wide host range includes crops, e.g. tomatoes and eggplant (Zimmermann, 1974; Neser and Siebert, 1977)

The eggs are deposited in small packets against the leaf veins. The first instars prefer young leaves, but later instars also feed on older leaves, petioles and even green stems. The larvae, particularly late instars, are voracious feed- ers and may skeletonize or devour entire leaves and young growth. Pupation occurs on the leaves after a short prepupal stage. The adults feed on the leaves and, to a lesser degree, the young shoots, but are less damaging than the larvae. A single generation (egg to egg) takes 30-35 days in mid-summer, allow- ing several generations annually in the field. Rates of adult feeding decrease towards the end of summer, prior to overwintering in leaf litter, but increase again in spring when mating and oviposition commences.

Tests in Argentina and in quarantine in South Africa showed that adults oviposited and larvae developed to maturity on S. melongena (eggplant) and on an indigenous weed, Solanum linnaeanum Hepper and Jaeger (=So- lanum hermannii Dun.), but not on 12 other test plants, including tomatoes (Lycopersicon spp. ), peppers (Capsicum spp. ) and potatoes (Solanum tuberosum) (Zimmermann, 1974; Siebert, 1975). Survival "nd rates of development on eggplant and S. linnaeanum were similar to those on S. elaeagni-

BIOLOGICAL CONTROL OF SOL,4NUM EL4L:4GNIFOLIUM AND S. MA URIIZ4N UM 143

TABLE 2

Phytophagous agents from Mexico (Mex) and Texas (Tex) released, or under consideration for i c- lease, for the biocontrol ofS. elaeagn(folium in South Africa

Agent From Year Damage Status

Frumenta Mex 1976 Destroys fruit nephelomicta Tex 1989 contents and (Gelechiidae) seeds

Dit.t,lenchus Tex 1984 Destroys ph.vllobius leaves by (Nematoda) galling

Leptinotarsa Tex 1985- Defoliation by texana 1989 adults and (Chrysomelidae) larvae

Leptinotarsa Tex 1985- defecta 1987 (Chrysomelidae)

Host specific. Released on three occasions, but failed to establish (Neser et al., 1990)

Narrow range in Solarium, but induces slight galling on eggplant and some native species (Scott, 1985 ). Release under consideration

Narrow range in Solanum, but develops on eggplant and some native species (Zimmermann, 1987). Release under consideration

Defoliation As above

folium, and adults showed no preference for either plant species. Neither G. lutescens nor G. pallidula was thus considered suitable for release.

Frumenta nephelomicta Meyrick (Gelechiidae)

Larvae of Frumenta nephelomicta destroy a high proportion of S. elaeagnifolium fruit in Mexico (Zimmermann, 1974). The females apparently scat- ter their minute eggs on the soil around the plants. These hatch intermittently, some immediately, while others diapause for long periods. The tiny neonates enter the ovaries of flowerbuds singly and develop within the maturing fruits, preventing seed formation (Neser and Siebert, 1977). Infested fruits have a spongy texture and are larger and slightly more elongated than healthy fruit. When eggs hatch in the absence of flowers, the larvae become stem gallers. Before pupating in the fruit or stem gall, each larva chews an emergence chan- nel, leaving a thin epidermal window which is breached when the moth emerges.

Frumenta nephelomicta proved host specific and did not develop on indigenous Solanum species or solanaceous crops (Neser and Siebert, 1977; Neser


et al., 1990). Three releases of eggs were made, starting in November 1979 and ending in November 1985, but the insect failed to establish. A follow-up survey of the final release of about 50 000 eggs at Kendrew revealed six infested fruits. Only about 60% of the eggs had hatched, although most of the remainder were still viable and have probably hatched since the survey. Ad- verse climatic conditions at the time of release caused a diminished fruit set and may have contributed to the poor survival of neonates (Zimmermann, 1986). Frumenta nephelomicta was reintroduced from Texas in 1989 and more suitable methods of release are now being investigated.

Ditylenchus phyilobius (Thorne, 1934) Filip'ev. 1936 (Nematoda)

This nematode is common in the southern USA (Orr, 1980), but also occurs in Argentina and India (Wassermann et al., 1988). Research in the USA suggested that Ditylenchus phyllobius ( = Orrina phyllobia was a promising biocontrol agent (Robinson et al., 1978, 1979).

Infective larvae migrate from the soil to the apical meristems of plants where they enter a?~tively growing tissue, causing large leaf galls. Populations increase quick!y an6 disperse by wind, by water and by movement in the soil. The nematcfdts kill plants and in the USA a 90% reduction in plant density was recorded in infestations of S. elaeagn(lblium within 4 years of release (Orr, 1980). Ditylench,ls phyllobius is largely confined to S. elaeagnifolium, although slight galling occasionally occurs on S. melongena and Solanum carolinense L. in the USA.

Between 1984 and 1986, Scott ( 1985 ) tested the host specificity ofD. phyl- h~bius on 14 plant species, including six crop plants, in a greenhouse and under confined field conditions in South Africa. Although a high percentage of S. mehmgena plants were colonized, damage was minimal and fruit development was unaffected. Three indigenous Solanum weeds, Solanum panduri- ./brine E. Mey, Solanum coccineum Jacq. and Solanum burchellii Dun., were also slightly galled, but only S. elaeagnifolium was affected to any degree. Per- mission for release has been withheld pending further investigations.

Leptinotarsa texana (Schaeffer) and Leptinotarsa defecta (StttO (Chrysomelidae)

These beetles inflict mainly foliar damage on S. elaeagnifolium in the Southern USA (Goeden, 1971 ) and Mexico (Wapshere, 1988). They attack leaves, flowers, young fruits and stems during their larval and adult stages, causing considerable damage and defoliation. The biology of both species is similar (Hoffmann, 1985 ). Females lay batches of 20-40 eggs on the lower leaf surfaces. The neonates consume the egg shells before feeding on the plant

BIOLOGICAL CONTROL OF SOLANUM EL..IE..IGNIFOLIUM AND S. M.4 UR117,,INUM 145

TABLE 3

Phytophagous insects found on S. elaeagn(folium in Argentina (Arg), Mexico (Mex) and Texas (Tcx), considered but as yet unevaluated, as biocontrol agents in South Africa

Agent From Damage Status

Trici~obaris Tex Stem boring by larvae texana Mex slows and stunts (Curculionidae) vegetative growth

Carpophilus sp. Arg Feeding on stamens ( Nitidulidae ) causes abscission of

flower buds

A:rmmetrischema Arg Larvae damage pistils ardeola and stamens, and (Gclechiidae) prevent fruit

formation

Unidentified Mex Stem galling by gall midge larvae stunts (Cecidomyiidae) vegetative growth

and prevents flowering

.,lnthonomus sp. Mex (Curculionidae) Tex

Destroy flower buds

Gargaphia Mex Cell sucking causes ari:onica defoliation (Tingidae)

Zonosemata Mex Larvae destroy fruit vittigera Tex contents (Tephritidae)

Has narrow range in Solanum (Cuda, 1983 ). Introduced to South Africa, but did not survive in quarantine

Tests in Argentina showed no attack of crops, e.g. potatoes, tomatoes or peppers (Zimmermann, 1974 ]. Introduced to South Africa, but did not survive in quarantine

Tests in Argentina showed no attack of crops, e.g. potatoes, tomatoes or peppers (Zimmermann, 1974 ). Not introduced

Little biological data. May be host specific (Zimmermann, 1974: Wapshere, 1988 )

Little biological data. Host specificity unknown (Zimmermann, 1974 )

Little biological data. Host specificity unknown (Goeden, 197 I; Wapshcre, 1988}

Apparently host specific, but causes little seed damage (Goeden and Ricker, 1971 )

and the larvae undergo seven moults before pupating in the soil. The adults diapause for the winter, moving into the soil as the plants senesce in autumn.

Solanum elaeagnifolium is the primary host of both species, although Lep- tinotarsa defecta has also been occasionally collected on Solanum rostratum Dun. and Solanum tridynamum Dun. (Jacques, 1985; Hsiao, 1986). Neither species has been recorded feeding on any solanaceous crops, including S. melongena (Goeden, 197 l; Neck, 1983; Hsiao, 1986; L. Knutson, personal communication, 1987; R.W. Neck, personal communication, 1987), although Hsiao ( 1974, 1981 ) recorded moderate to slight feeding on S. carolinense, Solanum dulcamara L. and S. melongena when adults were confined on these hosts in cages.


In South Africa, five native Solanum species and S. rnelongena supported full development of both Leptinotarsa species under caged conditions. How- ever, larval mortality was considerably higher and fecundity of adults reared on these hosts lower, suggesting that the plants are unlikely to maintain viable field populations after release (Zimmermann, 1987). In walk-in cages, adult Leptinotarsa texana and L. defecta showed clear ovipositional preference for S. elaeagnijblium, although levels of feeding damage on two of the plant species tested, S. melongena and S. coccineum, were the same as the levels on S. elaeagn([blium (Zirnmermann, 1987 ). Releases of both L. te,cana and L. de- .li'cta have not been approved pending further enquiry into their host specificity and their ability to establish on cultivated eggplants.

Other instz'ts

The surveys of Goeden (1971), Zimmermann (1974) and Wapshere (1988) in the Americas revealed numerous other herbivores injurious to S. elaeagntilblium. These include Carpophilus sp. (Nitidulidae), Symmetris- chema ardeola (Meyrick) (Gelechiidae), Anthonomus sp. (Curculionidae), Gargaphia ari-onica Drake and Carvallo (Tingidae) and an unidentified gall midge (Cecidomyiidae). Studies were conducted on two other potential agents in the USA, namely the curculionid Trichobaris texana Le Conte (Cuda, 1983 ) and the tephritid Zonosemata vittigera (Coquillet) (Goeden and Ricker, 1971 ). Information on these species, none of which were evaluated in South Africa, is summarized in Table 3.

ORIGIN AND TAXONOMIC STATUS OF S. ,iLl I_.'RITI.,INU,~!

Solanum mauritianum (Fig. 2), indigenous to Argentina, Brazil and Uru- guay, has been introduced to Africa, Australia, India and islands of the Atlan- tic, Indian and Pacific oceans (Roe, 1972). Solanum mauritianum appears to be part of a complex of species which includes Solanum granuloso-lepro- sum Dun., Solanum riparium Pers., Solanurn erianthum D. Don. and So- lanum verbascifolium L., which all show considerable morphological overlap (Neser, 1984). The taxonomic status and exact origin of the form in South Africa is unclear. Faunistic surveys on the S. mauritianum complex revealed a greater diversity of phytophagous insect species in Southern Brazil than in Argentina, suggesting a more northern centre of origin (Neser et el., 1990).

INTRODUCTION AND SPREAD OF 5". ,~L.ILIRITI,.INUM IN SOUTH AFRICA

Solanum mauritianum was possibly introduced to Africa, Madagascar, Mauritius and India via the Portuguese trade routes in the early 16th century (Roe, 1972, 1979). It was first noted in Natal around 1862 (Wright, 1904)

BIOLOGICAL CONTROL OF SOLANL!M EL.4E..t(iNIFOLIL'M AND S. M.4 ~'R177.~lNUM 147

; i ~ "..

' \ x "

Fig. 2. Bugweed. S. maurittanum. (a) Flowers and immature fruits: (b) branch with cluster of fruits. (Drawn by G. Condy. National Botanical Institute. Pretoria. )

and had become more widespread by 1881 (Harding, i 938; Anon., 1984). It was proclaimed a noxious weed in 1937 (Harding, 1938).

Solanum mauritianum has spread to most parts of South Africa and is established particularly in the higher rainfall regions of the Transvaal, Transkei, Eastern and Western Cape, and Natal (Hinze, 1985 ). The plant is highly invasive where natural vegetation has been destroyed or plantations have been cut down (Byford-Jones, 1981; Anon., 1984 ), but it also invades undisturbed riverine habitats. The fast-growing plants compete with young pine trees, in- hibiting growth and causing stem deformation (Le Roux, 1982; Hinze, 1985 ). The plants also produce large numbers of fruit which are utilized by the Natal fruit fly, Ceratitis rosa Karsch. The availability of this food source enables populations of C. rosa, which is an economic pest (Annecke and Moran, 1982), to survive over the winter in large numbers (Ripley and Hepburn,


1930, 1935 ). The fruit are favoured by a number of frugivorous birds (Olck- ers and Hulley, 1989). Rameron pigeons, Columba arquatrix Temminck, feed extensively on the fruits in Natal and are probably the main dispersal agents in plantation areas (Oatley, 1984). Bugweed also propagates vegetatively, particularly when cut or uprooted, and will coppice from root sections.

Solanum mauritianum has no fodder value and the plants are generally avoided by grazing animals. In Australia, some regard the plant as a useful nursery shrub which forms a canopy under which indigenous forest can regen- erate (Van Dyck, 1979 ). This contrasts with the South African situation where the plant is considered harmful to young commercial forests and indigenous vegetation (Le Roux, 1982; Hinze, 1985 ).

M E C H A N I C A L A N D C H E M I C A L C O N T R O L O F S. MAURITL4NUM

Although mechanical control of S. mauritianum may limit fruiting, it is generally ineffective because severed roots and stems initiate regrowth (Hinze, 1985 ) and needs to be supplemented with chemical treatments. In situations where chemical treatments are not advisable, such as along water courses or among cultivated crops, repeated mechanical treatments conducted during adverse environmental conditions (e.g. winter frost) may control the weed (Hinze, 1983). Herbicides registered for S. mauritianum include Giyphos- ate, 2,4,5-trichlorophenoxyacetic acid (2,4,5-T) and Triclopyr ( Hinze, 1983, 1985).

B I O L O G I C A L C O N T R O L O F S. M..I URITL4NUM

Faunistic surveys on S. mauritianum were initiated in 1984 in Argentina and southeastern Brazil (Neser, 1984; Neser et al., 1990). Although some of the identifications and host records are doubtful, most of the insects surveyed are apparently oligophagous or polyphagous species that attack a range of plants within the Solanaceae.

Two promising species, a stem-borer, Nealcidion bicristatum (Bates) (Cer- ambycidae) and a leaf-feeder, Corythaica cyathicollis (Costa) (Tingidae), were imported for specificity testing. Although both were restricted to Solan- aceae, they were rejected because they utilized a broad range of host plants in quarantine and because the published list of host records is vague. Both species are reported as minor pests of solanaceous crops in South America.

A leafminer, Acrolepia xylophragma (Meyrick) (Acrolepiidae), was imported during 1984 and found to be restricted to Solanaceae. Although it failed to develop on plants in the genera Nicotiana, Withania, Lycopersicon, Datura and Physalis under cage conditions, the larvae completed development on potato, eggplant and nine of 15 other indigenous Solanum species tested. However, in choice tests under cage conditions, the adult females deposited 51 times more eggs on S. mauritianum than on potato (H. De Beer, unpub-

BIOLOGICAL CONTROL OF SOL.INUM ELAIfAGNll.OLICM AND 5: .t£4URIIZ.INU.~I ! 49

lished data, 1990). Furthermore, 65 and 70 time~ more adults emerged from S. rnauritianum than from eggs oviposited on potatoes and eggplant, respectively. Acrolepia xylophragma has never been associated width or recorded from any solanaceous crop in Argentina, where all the collected specimens originated from S. bonariense L. This insect is still in quarantine.

D I S C U S S I O N

Prospects for the biological control of Solanum weeds in South Africa have been lessened because the host ranges of all, but one, of the agents tested under cage conditions have been unacceptably wide. Although more specialized herbivores are likely to be found among the endophagous gall-formers, stem- and fruit-borers, and florivores, such agents would probably be ineffective on S. elaeagnifolium where reproduction is mostly vegetative. However, flower- and fruit-feeding agents could be effective on S. mauritianum where reproduction is mostly sexual and long-range seed dispersal is a major concern.

Problems arising from the general lack of host specificity among the potential agents are compounded by the large number of agronomic species in the Solanaceae. In particular, eggplant (S. melongena) is acceptable to all the folivores tested under cage conditions so far. At face value, this implies that eggplant is a 'neutral' host and that it will not be immune to any of the solanaceous folivores. A point has thus been reached where the biological control of Solanum weeds in South Africa needs to be critically analysed and future objectives defined.

A fundamental requirement is to determine if the results of specificity tests conducted under cage conditions reflect the real host ranges of the agents and thus whether or not the folivores will attack eggplant and indigenous Solanurn species in the field in South Africa. Most evidence suggests that the specificity tests have been too conservative and that the threat posed by the potential agents may have been overrated for the tbllowing reasons.

(i) Although eggplant is cultivated extensively in the Americas, none of the potential agents (Tables l, 2 and 3) are listed as pests of this or any other solanaceous crops in the New World (Hayward, 1958; Costa Lima, 1968; Goeden, 1971; Neser et al., 1990). Leptinotarsa texana and L. defecta are the most promising candidates for the biocontrol of S. elaeagnifolium. Although S. rnelongena is extensively cultivated in the Rio Grande valley of Texas, where there are large populations of L. texana on S. elaeagnifofium within range, the beetles have never been collected or observed feeding on S. melongena (R.W. Neck, personal communication, 1987 ) and there appears to be no reason why they should do so if released in South Africa.

(ii) In South Africa, none of the many phytophages of indigenous Solanum species (Olckers and Hulley, 1989) are recorded pests of Solanum crops, ex- cept for a coccinellid, Epilachna hirta (Thunberg) (Ballard, 1914; Annecke


and Moran, 1982 ). However, the majority of indigenous Solanum insects will attack S. melongena in neglected cultivations that are no longer subject to insecticide treatments. This contrasts with the situation in the Americas where even unsprayed plants in neglected cultivations are not attacked by any of the insects listed in Tables 1, 2 and 3.

Because South African solanaceous insects are effectively controlled by in- secticides in eggplant cultivations, in the unlikely event of any imported agents moving onto eggplant, they will almost certainly also be controlled by the in- secticides and thus should be no more of a potential problem than the indigenous insects.

(iii) The reliability of starvation, and even choice tests, under cage conditions has often been questioned (Harris and Zw/51fer, 1968; Wapshere, 1974; Andres et al., 1976; Dunn, 1978). Many cases have been cited where candidates under cage conditions have displayed uncharacteristically broad host ranges which also included economic plants. Because of their expanded host ranges under cage conditions, some of these candidates have been rejected, although they have never been recorded as pests (e.g. Rizza and Pecora, 1980), with others released but which never became pests (Miyazaki and Naito, 1974; Ward et al., 1974; Harris, 1984; Peschken, 1984).

The arrangement of test plants in cages can induce females to oviposit on non-hosts (Marquis and Braker, 1987). Thiery and Visser (1986) showed that the long-range olfactory orientation of Leptinotarsa decemiineata (Say) towards potato hosts could be disrupted by the presence of other non-host Solanum spp. This phenomenon of'host plant masking' strongly suggests that the behaviour of candidates like Gratiana spp. and Leptinotarsa spp., which do not attack eggplant in the field, but did so in cages, may be artefacts of the cage conditions.

(iv) There is a minimal risk of Leptinotarsa spp. switching to eggplant in South Africa because ovipositing females show a strong preference for their natural hosts (Zimmermann, 1987 ) and because the infestations ofS. elaeag- n(folium are distant from eggplant cultivations. In addition, populations of Leptinotarsa spp. are unlikely to persist in eggplant cultivations because the beetles suffer higher larval mortality rates, extended developmental durations and reduced fecundity on eggplants.

(v) Cultivation practices would inhibit establishment of the agents on eggplant because this crop represents an unpredictable food source in both space and time. Eggplant is grown for relatively short periods during the year, often during winter, and is generally rotated with non-solanaceous crops.

(vi) The sparse and discontinuous distribution of indigenous Solanum spp. would make them less vulnerable to attack by introduced folivorous agents because oligophages tend to attack the most abundant hosts available (Har- ris, 1990). Also, the most common indigenous Solanum spp. that would be most at risk are themselves listed as weeds in South Africa (Wells et al., 1986 ).

BIOLOGICAL CONI'ROL OF SOL.4NI M EL.gEAGNIFOLll_'M AND S. M..IURIIZ4NUM 151

In conclusion, it follows that the established precedents and protocols in the biological control of weeds may not be appropriate in the case of the biological control of Solanum species. The dilemma is well illustrated by the ex- ample ofL. decemlineata (Jermy et al., 1988) which developed on lettuce for several generations without loss of fitness, relative to their normal potato hosts (Hsiao and Fraenkel, 1968). Leptinotarsa decemlineata, however, is not a recorded pest on lettuce because the chemical stimuli required for host plant recognition are lacking. Therefore, conventional host preference tests used in biological control may be obstructively conservative in this case.

The authors propose that restrictions should be relaxed for herbivorous agents ofSolanum spp. that have extend, ~ host ranges under cage conditions. The evidence discussed here is compelling enough to justify the release of L. t~:x'ana and L. defecta, and perhaps some of the other agents under consideration, against Solarium weeds in South Africa.

ACKNOWLEDGEMENTS

Thanks are expressed to S.G. Compton, J.H. Hoffmann, P.E. Hulley, V.C. Moran, J.K. Scott and colleagues of the Plant Protection Research Institute for useful discussion and co,nments on the manuscript. The authors are grate- ful for unpublished data obtained from H. de Beer.

REFERENCES

Andres, L.A., Davis, C.J., Harris, P. and Wapshere, A.J., 1976. Biological control of weeds. In: C.B. Huffaker and P.S. Messenger (Editors), Theory and Practice of Biological Control. Academic Press, New York, pp. 481-499.

Annecke. D.P. and Moran V.C., 19~2. Insects and Mites of Cultivated Plants in South Africa. Butterworths, Durban, 3d3 pp.

Anonymous, 1984. Pests and Diseases: N 3xious plant must be rooted out. Farm. Weekly Small- holder, September 28: 6-7.

Ballard, E., 1914. A list of the more important insect pests of crops in the Nyasaland Protecto- rate. Bull. Entomol., Res., 4: .'47-351.

Boyd, J.W., Murray, D.S. and Tyrl, R.J., 1983. Silverleaf nightshade, Solanum elaeagn~folium: Origin, distribution and relation to man. Econ. Bot., 38:210-217.

Bukenya. L.R. and Hall, J.B., 1987. Six cultivars of Solanum macrocarpon (Solanaceae) in Ghana. Bothalia, 17: 191-195.

Byford-Jones, C., 1981. Bugged by bugtrees? Farm. Weekly, May 20, pp. 22-23. Costa Lima, A.M., 1968. Quarto Catalogo dos lnsetos que Vivem nas Plantas do Brasil seus

parasitos e predatores. Parte 2-2 tomo, lndice de insetos e indice de plantas. Ministerio da Agricultura, Departamento de Defesa e lnspecao Agropecuaria, Rio de Janeiro, Brasil, 265 pp.

Cuda, J.P., 1983. The biology of three species of Trichobaris Leconte (Coleoptera: Curculioni- dae) in central Texas. PhD Thesis, Texas A & M University, pp. 66-92.


Davis, C.H., Smith, T.J. and Hawkins, R.S., 1945. Eradication of white he" ~e nettle in Southern Arizona. Agricultural Experiment Station, University of Arizona, Tucson, Bull. 195, 14 pp.

De Beer, H., 1985. Silverleafbitter apple causes concern. Farm. S. Air., Weeds A.6/1985, 4 pp. De Beer, H., 1987. Control of silverleaf bitter apple. Farm. S. Afr., Weeds A.6.1 / 1987, ! p. ~ollahite, J.W. and Allen, T.J., 1960. Silverleaf nightshade poisoning in livestock. Tex. Agric.

Exp. Stn. Prog. Rep. No. 2146. Dunn, P., 1978. Shortcomings in the classical test of candidate insects for the biocontrol of

weeds. In: T.F. Freeman (Editor), Proceedings of the Fourth International Symposium on Biological Control of Weeds, 1976, at Gainesviile, FL, pp. 51-56.

Gibbs Russell, G.E., Welman, W.G., Retief, E., Immelman, K.L., Germishuizen, G., Pienaar, B.J., Van Wyk, M., Nicholas, A., De Wet, C., Mogford, J.C. and Mulvenna, J., 1987. List of Species of Southern African Plants, Edition 2, Part 2. Memoirs of the Botanical Survey of South Africa No. 56, Botanical Research Institute, Department of Agriculture and Water Supply, South Africa, 144 pp.

Goeden, R.D., 197 I. Insect ecology of silverleaf nightshade. Weed Sci., 19:45-5 I. Goeden, R.D. and Ricker, D.W., 1971. Biology of Zonosemata vittigera relative to silverleaf

nightshade. J. Econ. Entomol., 64: 417-421. Harding, J.J., 1938. Die luisboom ('bugtree"). Onkruidplante Suid-Afrika, I: 11 - 1 2 .

Harris, P., 198a. Linaria vulgaris (Miller), yellow toadflax and L. dahnatica (L.) Mill., broad- leaved toadflax (Scrophulariaeeae). In: J.S. Kelleher and M.A. Hulme (Editors), Biological Control Programmes against Insects and Weeds in Canada 1969-1980. C.A.B. International, Slough, UK, pp. 179-182.

Harris, P., 1990. Environmental impact of introduced biological control agents. In: M. Mack- auer, L.E. Ehler and J. Roland (Editors), Critical Issues in Biological Control. Intercept, Andover, pp. 289-300.

Harris, P. and Zw61fer, H., 1968. Screening of phytophagous insects for biological control of weeds. Can. Entomol., 100: 295-303.

Hayward, K.J., 1958. lnsectos Tucumanos Perjudiciales. Revista Industrial y ~gricola de Tu- cuman, Tome 42, Publicacion de la Estacion Experimental Agricola de la Provincia de Tu- cuman, 144 pp.

Henderson, M. and Anderson, J.G., 1966. Common Weeds in South Africa. Mem. Bet. Surv. S. Air. No. 37, Botanical Research Institute, Department of Agricultural Technical Services, South Africa, 440 pp.

Hinze. W.H.F., 1983. Solamon mauritiamon as a forest invader plant. In: S. Neser and A.L.P. Cairns (Editors), Proceedings of the Fifth National Weeds Conference of Southern Afi'ica, 1983, at Stellenbosch, South Africa, pp. 327-331.

Hinze, W.H.F., 1985. Solamon mauritianum Scop. Bugweed, luisboom. Plant Invaders, Pam- phlet 365/I, Directorate Forest Management, South Africa, 7 pp.

Hoffmann, J.H,, 1985. Motivation for the release of two chrysomelid beetles (Leptmotarsa spp. ) in South Africa for biological control of Solarium elaeagn(folium. Report of the Plant Protec- tion Research Institute, Department of Agriculture and Water Supply, Pretoria, 17 pp. (unpublished).

Hsiao, T.H., 1974. Chemical influence on feeding behaviour of Leptmotarsa beetles. In: L.B. Browne (Editor), Experimental Analysis of Insect Behaviour. Springer-Verlag, New York, pp. 237-247.

Hsiao, T, H., 1981. Ecophysiological adaptation among geographic populations of the Colorado potato beetle in North America. In: J.H, Lashomb and R. Casagrande (Editors), Advances in Potato Pest Management. Stroudsberg, Hutchinson Ross, pp. 69-85.

Hsiao, T.H., 1986. Specificity of certain chrysomelid beetles for Solanaceae. In: W.G. D'Arcy (Editor), Solanaceae: Biology and Systematics. Columbia University Press, New York, pp. 345-363.

BIOLOGICAL CONTROL OF SOLANUM EL.4EAGNIFOLIU/tl AND X M..IURITi.,INUM 153

Hsiao, T.H. and Fraenkel, G., 1968. Selection and specificity of the Colorado potato beetle for solanaceous and non solanaceous plants. Ann. Entomol. Soc. Am., 61: 493-503,

Jacques, R.L., 1985. Overview on the taxonomy of the genus Leptinotarsa. Presented at the 1985 meeting of the Entomological Society of America, 9 pp. (unpubl.).

Jaeger P.~M.L. and Hepper, F.N., !986. A review of the genus Solanu:n in Africa. In: W.G. D'Arcy (Editor), Solanaceae: Biology and Systematics. Columbia University Press, New York, pp. 41-55.

Jermy, T., Szentesi, A. and Horvfith, J., 1988. Host plant finding in phytophagous insects: the case of the Colorado potato beetle. Entomol. Exp. Appl., 49: 83-98.

Julien, M.H., 1987. Biological control of weeds: a world catalogue of agents and their target weeds. C.A.B. International, Wallingford, UK, 144 pp.

Le Roux, P.J., 1982. Plant invader species in pine plantations in South Africa. In: H.A. Van de Venter and M. Mason (Editors), Proceedings of the Fourth National Weeds Conference of Southern Africa, 1982, at Stellenbosch, South Africa, pp. 155-168.

Maiti, P.C. and Mathew, R., 1967. Rich sources ofsolasodine. Curt. Sci., 36: 26. Marquis, R.J. and Braker, H.E., 1987. Influence of method of presentation on results of plant-

host preference tests with two species of grasshoppers. Entomol. Exp. Appl,, 44: 59-63. Miyazaki, M. and Naito, A., 1974. Studies on the host specificity of Gastrophysa atrocyanea

Mots. (Col.: Chrysomelidae), a potential biological control agent against Rumex obtus(folius L. (Polygonaceae) in Japan. Proceedings of the Third International Symposium on Biolog- ical Control of Weeds, 1973, at Montpellier, France, pp. 97-107.

Molnar, V,M. and McKenzie, D.N., 1976. Progress report on siiverleaf nightshade research. Pamphlet Vermin and Noxious Weeds Destruction Board, Department of Crown Lands and Surveys, No. 61, Keith Turnbull Research Institute, Victoria, Australia, 12 pp.

Morton, C.V., 1976. A revision of the Argentine species of Solanunl. Academia Nacional de Ciencas, Cordoba, Argentina, 260 pp.

Neck, R.W., 1983, Foodplant ecology and geographical range of the Colorado potato beetle and a related species (Leptinotarsa spp. ) (Coleoptera: Chrysomelidae ). Coleopt. Bull., 37:177- 182.

Neser, S., 1984. Report on a visit to Australia, Texas and Canada and countries en route with observations on biological control of weeds. Report of the Plant Protection Research Insli- lute, Department of Agriculture and Water Supply, Pretoria, pp. 18-27 (unpublished).

Neser, S. and Siebrn, M.W., 1977. Insects tested for release against satansbos (Solamtm elaeag- nilbliton) in South Africa. Proceedings of the Second Entomological Congress, 1977, at Pre- toria, South Africa, p. 50.

Neser, S., Zimmermann, H.G., Erb, H.E. and Hoffmann, J.H., 1990. Progress and prospects for the biological control of two Solanum weeds in South Africa. In: E.S. Deifosse (Editor), Proceedings of the Seve~:th International Congress on the Biological Control of Weeds, 1988, at Rome, Italy, pp, 37!-381.

Oatley, T.V., ~ '~ 84. E~ ~ltmation of a new niche by the Ra,me~on Pigeon Columba arquatrix in Natal. Proo:edings of the Fifth Pan-African O~'nithological Congress. 1984, Lilongwe, Ma- lawi, pp. 323-330.

Olckers, T. and Hulley, P.E., ! 989. Insect herbivore diversity of the exotic weed Solanum maur- itidnum Scop. and three other Solanum species in the eastern (;ape. J. Entomol. Soc. South. Afr., 52: ~!-q3.

Orr, C.C.. 1 ~60. Nematode for ~-~,r trol ofsilverleafnightshade, Biocontrol News Inf., I: I. Peschken, D.P., 1984 Host range of Lema t3'anella (Coleoptera: Chrysomelidae), a candidate

for biocomrol of Canada Thistle, and of four stenophagous, foreign thistle insects in North America. Car:. Entomol., 116: 1377-1384.

Ripley, L.B. and Hepburn, G,A., 1930. The wintering of the Natal fruit-fly. Farming in South Africa, Reprint No. 89. Department of Agriculture, Union of South Africa, 5 pp.


Ripley, L.B. and Hepburn, G.A., 1935. Wild host plants for the fruit-fly. Farming in South Africa. Reprint No. 28, Departmeat of Agriculture, Union of South Africa, 1 p.

Rizza, A. and Pecora, P., 1980. Biology and host specificity of ChtTsomela rossia, a candidate for the biological control ofdalmation toadflax, Linaria dahnatica. Ann. Entomol. Soc. Am., 73: 95-99.

Robinson, A.F., Orr, C.C. and Abernathy, J.R., 1978. Distribution of Nothanguina phyllobia and its potential as a biological control agent ofsilverleaf nightshade. J. Nematol., 10:361- 366.

Robinson, A.F., Orr, C.C. and Abernathy, J.R., 1979. Behavioural response of NothanguSna phyllobia to selected plant species. J. Nematol., 1 I: 73-77.

Roe, K.E., 1972. A revision of Solanum sect. Brevantherum (Solanaceae). Brittonia, 24: 239- 278.

Roe, K.E., 1979. Dispersal and speciation in Solanum section Brevantherum. In: J.G. Hawkes, R.N. Lester and A.D. Skelding (Editors), The Biology and Taxonomy of the Solanaceac Academic Press, London, pp. 563-567.

Schreiber, K., 1979. The steroid alkaloids of Solanum. In: J.G. Hawkes, R.N. Lester and A.D. Skelding (Editors), The Biology and Taxonomy of the Solanaceae. Academic Press, London, pp. 193-202.

Scott, M.B., 1985. Potential for the use of a nematode, Orrina phyllobia (Thorne, 1934), as a biocontrol agent of the weed, satansbos, Solanum elaeagn~lblium Cav. Report of the Univer- sity of Fort Hare, 7 pp. (unpublished).

Sicbert, M.W., 1975. Candidates for the biological control ofSolanum elaeagn(lblimn Cav. (So- lanaceae) in South Africa. I. Laboratory studies on the biology ofGratiana lutescens (Boh.) and Gratiana pallidula (Boh.) (Coleoptera: Cassidinae). J. Entomol, Soc. South. Afr., 38: 297-304.

Siebert, M.W., 1977. Candidates for the biological control ofSolanum elaeagn(lblium Cav. (So- lanaceae) in South Africa. 2. Laboratory studies on the biology of Arvelius albopunctattls ( De Geer) ( Hemiptera: Pentatomidae ). J. Entomol. Soc. South. Afr., 40:165-170.

Symon, D.E., 1981. A revision of the genus Solammt in Australia. J. Adelaide Bow. Gard., 4: 1- 367.

Thiery, D. and Visser, J.H., 1986. Masking of host plant odour in the olfactory orientation of the Colorado potato beetle. Entomol. Exp. Appl., 41: 165-172.

Van Dyck, S,, 1979. Destruction of wild tobacco trees (Solmmm mauritianum Scopoli) by mountain possums ( Trichosm'lts canimts Ogilby ). Mem. Quecnsl. Mus., 19:367-371.

Wapshere, A.J., 1974. A strategy for evaluating the safety of organisms for biological weed control. Ann, Appl. Biol., 77:201-21 I.

Wapshere, A.J., 1988. Prospects for the biological control of silver-leaf nightshade, Solanum elaeagn~lblium, in Australia. Aust. J. Agric. Res., 39: 187-197.

Ward, R,H., Pienkowski, R.L. and Kok, L.T., 1974. Host specificity of the first-instar of Ceu- thorhynchidius horridus, a weevil tbr biological control of thistles. J. Econ. Entomol., 67: 735-737.

Wassermann, V.D., Zimmermann, H.G. and Neser, S., 1988. The weed silverleaf bitter apple ("satansbos") (Solanum elaeagn~lblium Cav. ) with special reference to its status in South Africa. Teeh. Commun, No. 214, Plant Protection Research Institute, Department of Agri- culture and Water Supply, South Africa, I 9 pp.

Wells, M.J., Balsinhas, A.A., Joffe, H., Engelbrecht, V.M., Harding, G. and Stirton, C.H., 1986. A Catalogue of Problem Plants in South Africa. Mem. Bot. Surv. S. Aft'., No. 53. Botanical Research Institute, Department of Agriculture and Water Supply, South Africa, 658 pp.

Wright, C.H., 1904. Solanat:eae. In: W.T. Thisclton-Dyer (Editor), Flora Capensis Vol. 4. Lov- ell Reeve, London, pp. 87-121,

Zimmermann, H.G., 1974. Preliminary work on the biological control of satansbos, Solanum

BIOLOGICAL CONTROL OF SOL..INUM EL.IEAGNIFOLIUM AND S.M.4URiTZ4NUM 155

elaeagn(lblium Cav. Proceedings of the First National Weeds Conference, at Pretoria, pp. 151-168 (unpublished).

Zirnrnermann, H.G., 1986. The biological control of the weed Solanum elaeagnifohum (satansbos). Report of the Plant Protection Research Institute, Department of Agriculture and Water Supply, Pretoria, 2 pp. (unpublished).

Zimmerrnann, H.G., 1987. Host specificity ofLeptinotarsa texana. Report of the Plant Protec- tion Research Institute, Department of Agriculture and Water supply, Pretoria, 22 pp. ( unpublished ).

Biological control of silverleaf nightshade, Solanum … · 2015. 7. 6. · elaeagnifolium Cav....

Documents

Transcript of Biological control of silverleaf nightshade, Solanum … · 2015. 7. 6. · elaeagnifolium Cav....