Research Project: Biological Control of Invasive Weeds from Eurasia and Africa

Location: European Biological Control Laboratory

2019 Annual Report

Objectives
Objective 1: Explore for natural enemies, including arthropods and microorganisms, of invasive weeds identified as high priority targets by the ARS Office of National Programs, performing collections, importations and exportations in compliance with local and international regulations. High priority pests include hoary cress, swallow-worts, thistle species, giant reed, tree of heaven, French broom, and medusahead grass. Objective 2: Perform taxonomic, population genetic, and phylogeographic studies as necessary to support the successful development of biological control agents for invasive weeds. Objective 3: Identify the biological and physical parameters that affect the efficacy and safety of potential agents, including climatic conditions, host specificity, effective rearing conditions, and how interactions that occur between a target weed (e.g. hoary cress) and multiple natural enemies (e.g., a root gall weevil and a root pathogen) enhance or inhibit biocontrol.

Approach
The goal of the proposed research is to improve management of several important, invasive alien weeds of the contiguous United States. The Bureau of Land Management estimates that over 4,600 acres of western rangelands are lost to the spread of invasive weeds each day. Successful invaders usually lack natural enemies, which control them in their native ranges. Classical biological control is a highly cost-effective, sustainable, environmentally-sound method of controlling exotic weeds over large areas. This method requires research aimed at understanding the basic biology and ecology of target weeds and their interactions with the environment and natural enemies. Genetic characterization of target weeds in their native and adventive ranges can identify regions of origin of invasive populations and help to understand the invasion process. Foreign exploration in the region of origin provides candidate biological control agents such as insects and mites for evaluation. Host specificity and potential effectiveness of candidates will be evaluated in field and laboratory studies. Experiments will also examine the evolutionary, ecological, and physiological aspects of plant-insect interactions relevant to weed biological control. Research on yellow starthistle, hoary cress, giant reed, medusahead, and French broom will continue, and additional weeds will be targeted in response to stakeholder demand and available resources. EBCL plays a key role providing research and prospective agents to federal and state cooperators necessary for the successful control of target weeds. The research proposed here is critical to achieving ecologically rational, sustainable management of some of the most important invasive weeds in the United States.

Progress Report
Objective 1, the phytophagous wasp TETRAMESA sp. was collected from galls observed on Medusahead (TAENIATHERUM CAPUT-MEDUSAE) in Eastern Greece. One to two galls per plant were found, impacting seed production. The wasp larva develops inside the spikelets. The eriophyid mite ACULOPS MOSONIENSIS was collected from tree of heaven (AILANTHUS ALTISSIMA) in Italy, and a colony was established at EBCL. Adults of a seed feeder weevil (LEPIDAPION ARGENTATUM), a prospective agent of French broom (GENISTA MONSPESSULANA), were collected throughout the season in southern France. Field studies were conducted to understand its life cycle, and its impact. Adults of a leaf beetle (CHRYSOCHUS ASCLEPIADEUS), a prospective agent of swallow-worts (VINCETOXICUM SPP.), were collected in the French Jura for setting up an open field test. Objective 2a, a genetic analysis of the invasive grass, VENTENATA DUBIA was conducted by our cooperator at Boise State University, ID. Out of the 26 invasive U.S. populations, which were analyzed with allozymes markers, 8 multilocus genotypes (MLG) were identified. The most common genotype revealed was widely distributed throughout the Pacific Northwest and was found in ca. 60% of the populations analyzed. Out of the 15 native European populations analyzed, 10 MLGs were identified. More populations contain admixtures in the native range (46%), compared to invasive populations (26%). Only one population in Slovakia was a match for the most common MLG found in the pacific northwest (WA, OR, ID, MT, WY, UT, NV). Another match was found (Slovakia, France) for a more rare genotype (western OR). We may hypothesize, from these data that the origin of VENTENATA DUBIA is from central Europe. Collections of spurges (EUPHORBIA ESULA, E. CYPARISSIAS) were made in France, Spain, Italy, Greece, Belgium for a total of 13 populations. Genetic analysis, to be performed by cooperating ARS scientists in Sidney, MT, is aimed to pinpoint the origin of spurges invasive in the US. Collections of medusahead (TAENIATHERUM CAPUT-MEDUSAE), cheatgrass (ANISANTHA TECTORUM), and red brome (ANISANTHA RUBENS) were made by a visiting ARS scientist from Reno, NV and BBCA cooperators in three Eurasian countries (Italy, Spain, Bulgaria) and in one African country (Kenya). Panicles containing seeds from 17 populations were brought back to EBCL. Seeds were germinated under strict confined conditions. The DNA of the seedlings was extracted and shipped ARS Reno for genetic analysis using high sequencing technology in order to direct where to look for prospective biological control agents. The stem-feeding shootfly (CRYPTONEVRA sp.), a prospective biocontrol agent of the giant reed (ARUNDO DONAX) in USA, has several genetic lineages, suggesting the presence of several species, one of which can develop on native US common reed (PHRAGMITES AUSTRALIS). DNA was extracted non-invasively from CRYPTONEVRA sp. specimens sampled in ARUNDO DONAX and PHRAGMITES from several populations in France and Spain, and corresponding voucher specimens are being analyzed by a taxonomist. The preliminary analysis of the DNA Barcode region would rather suggest that there is one single species associated with both ARUNDO DONAX and PHRAGMITES. In addition, a general taxonomic examination of the dipteran specimens collected emerging from ARUNDO DONAX shoots over the period 2001-18 was completed with 500+ specimens examined, coupled with genetic analyses of representative samples of each species identified. It revealed that CRYPTONEVRA NIGRITARSIS is the only species infesting ARUNDO DONAX shoots in the south of France. However, it was also shown that CRYPTONEVRA NIGRITARSIS is also able to infest galls of LIPARA LUCENS in PHRAGMITES AUSTRALIS, casting doubts on the specificity of this prospective agent. However, a total of 13 dipteran species were formally identified in the collections, and little is known on the biology, ecology, and specificity of these species. The weevil LEPIDAPION ARGENTATUM, a prospective agent of French broom (GENISTA MONSPESSULANA), has two oviposition behaviors: forming galls in stems, and boring seedheads, suggesting the presence of two species. DNA analyses of the weevil sampled in France and corresponding to the two behavioral observations were conducted, and showed the presence of a single species. In addition, the French origin of the weevil belongs to a different genetic lineage than the lineage present in Sicily and in Spain, though all three lineages belong to the same species. Objective 2b, The analysis of the genetic diversity of the root gall weevil (CEUTORHYNCHUS ASSIMILIS) associated with hoary cress (LEPIDIUM DRABA) was finalized, hence confirming that the specialist lineage of the weevil which was considered as a prospective biocontrol agent of hoary cress, and the generalist lineage of the same species which develops on several Brassicaceae can interbreed, which might represent an issue for a release permit. CABI in collaboration with EBCL has evidenced that the specialist lineage was in fact not specific to the target hoary cress (LEPIDIUM DRABA) as it was able to develop to adult on three federally listed threatened and endangered species. Under Objective 2c, EBCL finalized population genetic analyses using microsatellites and sequence data from two regions in the chloroplast genome of yellow starthistle (CENTAUREA SOLSTITIALIS) sampled throughout 7 countries in the native range in support of research being conducted in California. It revealed a clear population genetic structure of potential source populations in Eurasia including a deep differentiation of one lineage found in Sicily. Counting of chromosomes evidenced 18 chromosomes in this lineage versus 16 chromosomes in the others. Specimens of this lineage were shipped to a botanist in Spain who considered this lineage to be a species new to Science. We found that western European populations were derived from eastern European populations and that western populations were genetically more closely related to US populations than eastern populations. Objective 3a, a field experiment with four conditions (hoary cress [LEPIDIUM DRABA] alone, hoary cress with the gall weevil (CEUTORHYNCHUS ASSIMILIS), hoary cress with the soilborne pathogenic fungus RHIZOCTONIA SOLANI, and all three organisms) was conducted at EBCL to measure possible interactions among the three organisms. Plant mortality rate tended to be higher in the weevil plus pathogen condition (21.8%) than in the weevil alone condition (13.8%), probably because the insect damage provided entry points for RHIZOCTONIA to attack the plant. In order to study the change in microbial community in response to the introduction of the biocontrol agent, bacterial and fungal communities from the different experimental conditions were analyzed by sequencing the 16S rRNA gene and the ITS region. ITS metagenomics revealed a low abundance of RHIZOCTONIA SOLANI in all the experimental conditions with no correlation with the introduction of the 06f12 isolate RHIZOCTONIA SOLANI in two conditions. Cultivable bacteria and fungi have been isolated and screened against the O6f12 isolate RHIZOCTONIA SOLANI used in the experiment. Three bacterial isolates and seven fungi were able to strongly inhibit the growth of RHIZOCTONIA SOLANI. Whereas RHIZOCTONIA SOLANI represents less than 1% of the fungal community in all the conditions, the seven fungi inhibiting RHIZOCTONIA constituted approximately 15% of the fungal community in all the conditions. The abundance of the fungal inhibitors could thus explain the low proportions of RHIZOCTONIA (native and introduced). In the presence of the weevil, a significantly higher abundance of the three genera SCLERODERMA, TUBER and LEUCOAGARICUS and a lower abundance of the COPRINELLUS genus was observed. Our field experiment suggested a non-significant modification of the structure and the diversity of bacterial and fungal populations in presence of the weevil and/or RHIZOCTONIA at the community level. Objective 3b, an open field experiment to measure the herbivory and the oviposition behavior of CHRYSOCHUS ASCLEPIADEUS, a prospective agent of swallow-worts (VINCETOXICUM SPP.), was conducted in Lyon (France). Five circular gardens with potted plants of VINCETOXICUM HIRUNDINARIA, V. ROSSICUM, and ASCLEPIAS SYRIACA were tested. A total of 100 couples were released. No feeding damage was observed outside the genus VINCETOXICUM (30 plants attacked/30 plants exposed). A second, open field test with potted VINCETOXICUM HIRUNDINARIA and ASCLEPIAS SYRIACA was set up in the Jura (France). Plants were randomly displayed in a pasture where CHRYSOCHUS ASCLEPIADEUS naturally lives. Foliage damage was only reported from VINCETOXICUM HIRUNDINARIA (5 plants attacked/5 plants exposed). Based on our data, CHRYSOCHUS ASCLEPIADEUS is sought to be petitioned for release. Sentinel GENISTA MONSPESSULANA plants were exposed from April to October in Southern France. We determined the seasonal activity and natural density of the French broom weevil (LEPIDAPION ARGENTATUM), inducing both gall formation, and seedhead damage on the target weed. A peak population from galls was observed on mid June, and a second peak occurred in late July from seedheads. The life cycle of CRYPTONEVRA NIGRITARSIS, a prospective agent of ARUNDO DONAX was elucidated. CRYPTONEVRA NIGRITARSIS is a multivoltine species with a summer population peak. It mainly overwinters as pupae, but may also overwinter as larvae. A new rearing method was designed. The larvae need water-stressed plants to develop but also require high levels of humidity to complete their development. The new rearing method consisted of cutting live A. donax shoots and placing them in a thick layer of water agar. As the shoot slowly died, it retained most of its moisture. This rearing method is effective and inexpensive and can replace the traditional method in potted plants.

Accomplishments
1. A beetle impacting swallow-worts, but safe for milkweed?. Invasive swallow-worts are composed of two species of VINCETOXICUM that invade the north-eastern USA. We aimed to release a host-specific agent, safe for the environment that did not attack the Monarch butterfly host-plant, the common milkweed. We then exposed our beetle candidate, CHRYSOCHUS ASCLEPIADEUS, to both swallow-worts and milkweeds in an open field test in France to measure host preference and herbivory impact. Our data showed zero attack on milkweeds (0/20 plants exposed), meaning no foliage feeding by the beetle, and substantial defoliating on the weed target (30/30 plants exposed). If preliminary feeding data are confirmed by oviposition behavior data (to be obtained by 2020), then our data will be used to implement previously obtained data for a petition to release the beetle to be submitted to TAG. Controlling swallow-worts in the U.S. and Canada will increase native biodiversity understory in natural parks and reserves, and recreational areas.

2. Taxonomic inventory of the Chloropidae attacking giant reed in its native range. The giant reed ARUNDO DONAX is a highly invasive weed in the southern U.S. where it invades riparian habitats, displaces native flora and fauna, and creates environmental conditions favorable to the spread of invasive ticks. A diverse range of natural enemies have been described for ARUNDO DONAX in its native range, but the assemblage of fly species from the Chloropidae family (Diptera) has never been definitely established. A taxonomic examination of all specimens collected by EBCL over the 2001-2018 period (> 500 specimens examined) was done at EBCL in 2019 and revealed, after complementary genetic analyses, that a total of 13 species from the Chloropidae family can be found in giant reed shoots, including two species new to science. This important accomplishment brings new perspectives for the biological control of ARUNDO DONAX, as the ecology, efficiency, and specificity of most of these species is currently unknown.

Review Publications
Bon, M., Guermache, F., De Simone, D., Cristofaro, M., Vacek, A., Goolsby, J. 2018. PCR based screening of nematodes and fungi associated with pupae of Lasioptera donacis Coutin (Diptera: Cecidomyiidae), a biological control agent of the giant reed. Florida Entomologist. 101(3):505-507. https://doi.org/10.1653/024.101.0309.
Lesieur, V., Martin, J., Hinz, H., Fumanal, B., Sobhian, R., Bon, M. 2018. Implications of a phylogeographic approach for the selection of Ceutorhynchus assimilis as a potential biological control agent for Lepidium draba. Biological Control. 123(43-52). /doi.org/10.1016/j.biocontrol.2018.05.001.
Bitume, E.V., Moran, P.J., Sforza, R. 2019. Impact in quarantine of the galling weevil Lepidapion argentatum on shoot growth of French broom (Genista monspessulana), an invasive weed in the western U.S. Biocontrol Science and Technology. 29(7):615-625. https://doi.org/10.1080/09583157.2019.1573417.
Kerdellant, E., Thomann, T., Vitou, J., Sheppard, A., Guisto, C., Simonot, O., Sforza, R. 2019. The double life cycle of Lepidapion argentatum, a potential candidate for the biological control of French Broom. Biocontrol Science and Technology. 10.1080/09583157.2019.1597334.