Research Progress on Biological Control of Saltcedar Research on saltcedar has been underway at Temple, TX, since 1987 with partial support from the USDI-Bureau of Reclamation (DeLoach et al 1996). A review of literature and analysis of the harmful and beneficial effects of saltcedar and its potential for biological control was sent to TAG in May 1989 and received a favorable opinion to proceed in December 1989. Overseas research began in the summer of 1991 with a second grant from Bureau of Reclamation. Cooperative projects were established at the ARS European Biological Control Laboratory at Rome, Italy (later moved to Montpellier in southern France) and with Prof. Dan Gerling at the University of Tel Aviv, Israel. In Israel, testing began on two mealybugs, Trabutina mannipara and Naiacoccus (now Trabutina) serpentina. Foreign explorations also began in China, where no information was available on natural enemies, through the Sino-American Biological Control Laboratory at Beijing; testing began there in 1993, supported by ARS International Research Programs. In 1993, cooperative projects were established with Dr. Svetlana Myartseva, Academy of Sciences of Turkmenistan at Ashgabat, and in 1995 with Prof. Ivan Mityaev and Dr. Roman Jashenko at the Academy of Sciences of Kazakhstan at Almaty. At present, we are testing 19 insect species as candidates for introduction. Two have preliminary approval for release, 7 others are being tested in quarantine, overseas testing has been completed on 5, and 5 others are still being tested overseas (Table 7). Research progress was reviewed by DeLoach et al (1996).
Host range determination is the most important part of the testing program. The most reliable data are those obtained under the most natural conditions and which are best obtained overseas before introduction into quarantine. For each of the candidate insects being tested, extensive searches were first conducted of the literature and of museum records to determine which plants the insects are known to attack in nature. In addition, extensive field surveys were done by our overseas cooperators and myself at various times during the growing season, over several years, and in several locations. Not only T. ramosissima was sampled but also other species of Tamarix, species in other genera of the Tamaricaceae, and Frankeniaceae, and other more distantly related plant families as well as habitat associates growing in riparian areas with saltcedar. Insect species known to feed on plants outside the family Tamaricaceae in nature were not considered as candidates for introduction.
Two types of outdoor, uncaged, replicated, controlled tests were conducted. In the first, potted plants of the various test species were placed underneath trees infested naturally with the candidate insects - natural infestation of the test plants over the season was recorded. In the second, various test plants growing outside under natural conditions were inoculated with the candidate insects and survival, feeding and development were recorded. These types of tests and observations necessarily must be done overseas.
Also, controlled tests (either no-choice or multiple-choice) were conducted in cages in the laboratory or greenhouse. These tests measured adult and larval (or nymphal) feeding, survival and development of the candidate insect and ovipositional preference for the different test plants. A plant was not considered to be a true host unless the insect could complete its entire life cycle on it and could reproduce sufficiently to maintain its population on the plant. We have shipped cuttings of saltcedar (including athel) from several major infested areas in the United States to our overseas cooperators so they can compare host preference on the same biotypes of saltcedar that the insects will encounter in the United States.
We selected test plant species based on the phylogenetic system of Wapshire (1974) and now universally accepted by biocontrol of weeds scientists worldwide. Under this system, the first round of testing includes plants most closely related to the target weed (same genus). If the insects feed on these, then more distantly related plants are tested (different genera of the weed's family), and then even more distantly related plants (genera of different families of the plant's order), until the true host range of the insect is determined. In the case of saltcedar, the insect would be rejected if its host range includes plants beyond the family Tamaricaceae.
In addition, we include critical test plants and habitat associates. The critical test plants (in the case of saltcedar testing, these are athel and Frankenia spp.) are included at an early stage of the testing. If the candidate insect can complete its life cycle and reproduce on any critical plant, the insect is rejected, without wasting time on further testing. Habitat associates - cottonwoods, willows, mesquites, etc. - are included and even some of the major crops grown in the area of planned releases. This gives a greater degree of assurance to concerned individuals, even though the probability that these might be damaged is exceedingly low, given the previous surveys already done overseas.
The overseas studies also include observation and testing of the life cycle, seasonal cycle, degree of parasitism, behavior, and type and degree of damage to the plant, as discussed above under Research Protocol. The testing done in quarantine follows the same protocol as the above described laboratory testing.
Trabutina mannipara (mealybug) - The genus Trabutina contains 5 species including species formerly assigned to Naiacoccus, all restricted to the genus Tamarix (Danzig & Miller, in press). T. mannipara (Hemprich and Ehrenberg) was the first insect tested by our cooperator, D. Gerling at Tel Aviv. The biotype tested is known only from deciduous Tamarix from the Dead Sea and Sinai areas in Israel but other biotypes occur in central Asia. It is not known to attack T. aphylla. The adult female is wingless and excretes a tough waxy egg sac that gradually encloses her and in which she lays the eggs (Fig. 1). After hatching, the nymphs remain in the egg sac for a few days, then disperse onto the surrounding branches and plants. This species has two or three generations per year.
In quarantine at Temple, nymphs survived only on Tamarix among 14 genera of Violales and 2 habitat associates (willow and cottonwood) tested. Populations from the 1st to the 2nd generations increased 4 to 20 times on various Tamarix accessions but decreased on T. aphylla. The waxy nymphs covered young twigs and eventually killed several of the test plants. Several parasitoids and a predaceous drosophylid fly attack T. mannipara in Israel but do not attack 1st-instar nymphs, so they can easily be separated from cultures before field release. The adult females are strongly tended by ants in the field. In Israel, T. mannipara is known only from areas warm enough for citrus to grow. In the United States this zone extends north to Phoenix and Tucson, Arizona and the lower Rio Grande Valley of Texas. Thus, this mealybug is especially safe as one of the first planned releases, being restricted in host range, climatic tolerance and dispersal ability. The TAG recommended T. mannipara for field release on December 1994 but final approval awaits completion of an Environmental Assessment by APHIS-PPQ.
Diorhabda elongata (leaf beetle) - Seven species of Diorhabda are recorded from Eurasia. D. elongata Brullé (Fig. 2) occurs from France to Mongolia and China, where it attacks only Tamarix spp., mostly along the river banks. It is common everywhere and causes appreciable damage (Mityaev 1958, Lopatin 1977). In China, where Tamarix is planted to stabilize desert sandy soils, the beetle must be treated with insecticides (DDVP or Rogar) to prevent it from defoliating and killing the plantations (Bao 1989, Sha 1991).
In quarantine tests at Temple, larvae fed only on the Tamaricaceae and almost entirely on Tamarix. In tests using potted plants, only 2 adults developed from 106 neonate larvae on Frankenia; both were deformed and unable to reproduce. However, neonate larvae produced adults as well on T. aphylla as on T. ramosissima. Ovipositional host-selection tests at Temple were inconclusive because adults brought from China in 1992-1994 laid only a few eggs. However, G.Q. Lu, J.F. Wang and S.L. Jiang (unpubl. data) found that in multiple-choice tests at Beijing and Hohhot using potted plants, females laid only 25% as many eggs per plant on the 2 U.S. T. aphylla as on the 4 T. ramosissima accessions. In a smaller no-choice test at Ashgabat, also using potted plants, females laid 12.7 eggs each per day on T. ramosissima and none on T. aphylla (S.N. Myartseva, unpubl. dat). If released, D. elongata probably will damage T. aphylla only slightly, if at all, because it will not lay very many eggs on this species. We have reared no parasites from larvae or adults collected in China, but Prof. Myartseva reared Tetrastichus n. sp. from last-instar larvae in Turkmenistan. In China, Sha (1991) reported 3 generations, with adults overwintering, but in Turkmenistan, it overwinters as full-grown larvae. This indicates possible biotype differences. The TAG recommended approval of this species in December 1994, but final approval awaits completion of an Environmental Assessment by APHIS-PPQ.
Psectrosema spp. (gall midges) - At Montpellier, Drs. R. Sobhian and L. Fornasari found three new species of Psectrosema causing small stem and leaf tip galls on Tamarix in France. Psectrosema nigrum Gagné was most abundant April to early June. Two additional species, P. album Gagné and P. acuticorne Gagné, were more abundant in July and August (Gagné et al, in press). These midges were heavily parasitized and were preyed upon by a nabid. In France, outdoor host-range tests, conducted under infested trees, showed that these species attacked the local T. gallica and in the U.S. T. ramosissima, but not T. aphylla. Final testing is underway by P.E. Boldt in quarantine at Temple, TX. At Almaty, Drs. I. Mityaev and R. Jashenko are studying the biology of another species, Psectrosema noxium Marikovski, which is very damaging in Kazakhstan (Mityaev 1961) and in western China. Gerling et al (1976) studied several other gall-making insects of saltcedar from Israel and the Middle East.
Coniatus tamarisci (weevil) - Adults and larvae of at least 14 species of Coniatus feed on buds and shoots of Tamarix from France to China. They feed only on Tamarix, including T. ramosissima, or rarely on Myricaria or Reaumaria. They sometimes occur in large numbers and cause heavy damage (Sinadskii 1968). At Montpellier, Dr. Fornasari found that C. tamarisci F. produced most larvae on T. ramosissima accessions from Texas and Wyoming, next on the natural host T. gallica, a few on Myricaria, and none on T. aphylla, Frankenia, or other plant genera. Final testing has begun in quarantine at Temple, TX.
Ornativalva spp. (gelechiid moths) - Ornativalva is a genus of 14 species whose host range is entirely restricted to the genus Tamarix. The genus occurs from Spain and Algeria across central Asia to India and Mongolia. In most species, the larvae have at least a weak leaf-tying behavior, though some live in galls of other insects. Most have 4-5 generations per year. We found Ornativalva sp. commonly throughout northern China. In the laboratory at Hohhot, H. Zheng and J.F. Wang, and A.P. Liu found that its larvae developed as well on several U.S. accessions of T. ramosissima as on the local T. austromongolica, and about half as well on T. aphylla and Myricaria. In quarantine at Temple, 38% of the neonate larvae of O. grisea Sattler collected from T. chinensis south of Tienjin in eastern China developed to adults, 30% on T. chinensis from Cangzhou, 11% on T. ramosissima from Seymour, TX, and none on 3 other Tamarix species, T. aphylla, or on 2 Frankenia species tested.
Crastina, Colposcenia and Diaphorina spp. (psyllids) - These genera apparently are specific entirely to the genus Tamarix. They form "brushy-tip" galls ca. 2 cm long x 1 cm diam. from the enlarged foliage bracts, densely arrayed on the dwarfed twigs. They are common throughout China, Kazakhstan, and Turkmenistan to Israel, and often at very high populations on groups of plants. Kovalev (1995) listed 20 species of Colposcenia from Tamarix, many of them on T. ramosissima, and none are known to attack T. aphylla. At Hohhot, H. Zheng and J. Wang found that field collected galls from T. austromongolica Nakai contain up to 20 or more nymphs, some of which transferred when placed on potted laboratory plants. Female adults oviposit between the leaf bracts on growing terminals. The eggs hatched in 12 days and nymphs lived 28 days. In quarantine at Temple, nymphs on galls brought from China in 1994 transferred and produced over 300 adults on T. ramosissima but these did not reproduce and host-range testing could not be done. Rearing methods need refinement. At Almaty, Mityaev and Jashenko found that adults of Crastina tamaricina oviposit on new shoots in the spring and the developing nymphs produce galls. Adults from them emerge in July and produce a second generation which does not produce galls; the third instars overwinter.
Corimalia tamarisci (weevil) - Kovalev (1995) reported 34 species of the tribe Corimaliini (Apionidae) on Tamarix from the Mediterranean to Mongolia, 2 species of Allomalia, 27 of Corimalia, 3 of Hypophyes, and 4 of Titanomalia; he proposed that the latter genus should contain excellent biocontrol agents because they form galls in fruits or on stems that heavily damage Tamarix. At Montpellier, R. Sobhian is testing the host range and biology of Corimalia tamarisci (Gyrll.). It produces ovary galls which sterilize the flowers; this would slow the spread of Tamarix but would not prevent nectar and pollen production for honeybees or damage the shade tree values of either saltcedar or athel, even if it attacks both.
Agdistis tamaricis (plume moth) - Agdistis tamaricis (Zell.) occurs from the Mediterranean to Pakistan. It is oligophagous on Tamarix and Myricaria in Israel. At Tel Aviv, D. Gerling and V. Chikatunov tested its host range and biology, where it has three generations. In 7 collections during the summer of 1995, where both saltcedar and athel occurred side by side in nature, they collected 22 larvae from the native saltcedars (T. nilotica (Ehrenb.) Bge., T. palaestina Bertol, and T. tetragyna Ehrenb.) but none from planted athel trees, though a few were found on athel at other locations. In multiple-choice tests in 1 m2 cages, moths laid 3 times as many eggs on the Israeli saltcedar and 5 times as many on U.S. accessions of T. ramosissima as on athel. When the larvae hatched, they completely defoliated T. ramosissima from Kansas but feeding on athel was not noticeable.
Cryptocephalus (leaf beetle) - Cryptocephalus is a large genus that contains several species that attack Tamarix from Israel and the Caucasuses to Mongolia and China (Lopatin 1977). V. Chikatunov reviewed the species in Israel, and conducted biological and host range tests on C. sinaita moricei (Pic.) which occurs along the Dead Sea. In multiple-choice tests in 1 m2 outdoor cages, 38% of the eggs produced adults on T. ramosissima from Kansas, and 27% on the natural host T. nilotica; only 17% of larvae (started as 2nd instars) produced adults on T. aphylla. Also, duration of the larval stages was twice as long on T. aphylla, a sign of poor host suitability. In nature, C. sinaita moricei probably completes its development only rarely, if at all, on T. aphylla.
Other Natural Enemies - Seven other species of insects are being tested overseas, all of which appear suitable for introduction. Larvae of 2 gelechiid moths form stem galls ca. 9 mm diam x 20 mm long on young stems of Tamarix, which often kill the distal part of the stem. They occur from France to Israel and China. At Montpellier, L. Fornasari is studying the biology and host range of the univoltine Parapodia sinaica Frauenfeld. Amblypalpis olivierella Rag. causes similar damage (Loope et al 1988) and sometimes is a serious pest of Tamarix from Israel to Kazakhstan. At Almaty, I. Mityaev and R. Jashenko are testing the biology and host range of Amblypalpis tamaricella.
Trabutina serpentina (Green) occurs from Israel to western China. In laboratory tests at Tel Aviv, D. Gerling found that nymphs developed well on the local Tamarix and on U.S. accessions of T. ramosissima, but only 20% as many on T. aphylla. It sometimes reaches high populations by the Dead Sea, but is rare on T. aphylla in nature. It is also locally abundant near Almaty and is being tested there by I. Mityaev and R. Jashenko. Trabutina crassispinosa Borchs. is similar in appearance to T. mannipara but is probably more cold tolerant; S. Myartseva is testing it at Ashgabat. The scale Adiscodiaspis tamaricicola (Mal.) is being tested at both at Ashgabat and at Almaty.
In China, Q.G. Lu and S.L. Jiang collected the cerambycid, Asias halodendri (Pallas) from T. chinensis south of Tienjin. This species is abundant in stems, often killing them to the ground. It is reported in literature from several fruit trees but not from Tamarix. They are conducting tests to determine if it may be a sibling species specific to Tamarix. Few stem borers are known from Tamarix that do not attack T. aphylla and this could be very valuable, if host specific.
Several other insect species listed by Mityaev (1961), Sinadskii (1968), Gerling and Kugler (1973), and Kovalev (1995) probably could be candidates for introduction. Most of the genera discussed above have other species that may be just as promising. Stem damaging insects include several species of buprestids (though some damage T. aphylla). Few cerambycids attack Tamarix but the oligophagous Hesperophanes heydeni Back. (Kovalev 1995) is promising. Among the most damaging species may be the root-galling weevil Liocleonus clathratus Oliv. Other foliage feeders include the leaf beetles Stylosomus, the weevils Geranorhinus, and moths of several families. Many sucking insects are damaging and abundant in Asia and probably are host specific, especially the scale Circodiaspis sinensis Tang, the twig-galling mealybug Acanthococcus orbiculus Mat., leafhoppers of the genera Opsius, Tamaricades, and Tamaricella, cixiids of the genus Duilius (= Hemitropis and Bitropis), and mirids of the genus Tuponia.
Plant pathogens are little known from Tamarix. At Montpellier, A Kirk isolated and is testing Fusarium sp. from seeds in the field and Trichothecium sp. from dying branches. Seedling establishment in southern France is a rarity compared with the high rate seen in the United States.
Return to Workshop Home Page, return to the table of contents for the biological control paper, or continue on to the Plans for Release
For information on the outcome of this workshop or integrated weed management in the Pacific Region (Region 1), U.S. Fish and Wildlife Service, Portland, OR, contact: Scott_Stenquist@fws.gov