Rural R&D for Profit round 2 project (2016-2020)

Achievements of Rural R&D for Profit round 2 project: Biocontrol solutions for sustainable management of weed impacts to agricultural profitability

Undertook a literature review on taxonomy and distribution of African boxthorn and known natural enemies of the weed in the introduced and native ranges

Taxonomy

Conyza is mostly a New World genus in the tribe Conyzinae of the family Asteraceae (or Compositae), the largest of all plant families (c. 25,000 species worldwide). There are three main species of Conyza in Australia – Conyza bonariensis (flaxleaf fleabane), Conyza canadensis (Canadian fleabane) and Conyza sumatrensis (tall fleabane). Conyza bonariensis is the most widespread, occurring in all states and territories, followed by C. sumatrensis and C. canadensis. There are four other Conyza species present in Australia, but their distribution is limited ‒ C. bilbaoana, C. parva, and C. primulifolia (PlantNET 2020). It is noteworthy that the curator of the Australian National Herbarium has doubts as to the circumscription and recognition of Conyza bilbaoana in Australia.

Conyza bonariensis has the narrowest leaves at the rosette stage when compared to other Conyza species (Thébaud and Abbott 1995). It has a more compact stature, with many short branches and bearing large capitula, while C. canadensis is essentially a single-stemmed taxon with few long branches and with small and elongated capitula. Conyza bonariensis is a genetic allopolyloid (arose through hybridisation; hexaploid (2n=54)) and strictly semelparous (has a single reproductive episode before death) (Thébaud and Abbott 1995). It is self-compatible and seems not to be pollinated by insects (Thébaud et al. 1996).

Evidence of genetic variation in several morphological traits of C. bonariensis was found in a common garden experiment conducted by Thébaud and Abbott (1995) in Europe. Hybrids of unknown fecundity, originating from crosses between C. bonariensis and C. canadensis and between C. bonariensis and C. sumatrensis, have been reported in Europe (McClintock and Marshall 1988). However, a subsequent isozyme survey of five Conyza species in Europe failed to find intermediate individuals (Thébaud and Abbott 1995). Zelaya et al. (2007) highlighted that loss of vigour in Conyza hybrids is apparently a common trait. They speculated that ploidy differences may be a significant barrier determining successful hybridisation between Conyza spp., e.g. more compatible hybrids would be expected from crosses between allopolyploids such as C. sumatrensis and C. bonariensis compared to crosses with the diploid (2n = 18) C. canadensis. The extent of genetic diversity in C. bonariensis and existence of hybrids in Australia are unknown.

Distribution

Conyza bonariensis is present in all states of Australia, occurring predominantly in temperate and Mediterranean coastal regions, and with restricted distributions in semi-arid to arid central regions (GBIF.org 2nd November 2018). It is native to warm temperate South America, (Michael 1977). It is considered widespread in Argentina, Uruguay, Paraguay and Brazil, and has been recorded in coffee plantations in Colombia and Venezuela (Mangolin et al. 2012).

Defined goals for management of Flaxleaf fleabane

Between May and July 2017, the online platform SurveyGizmo was used to survey key stakeholders in the grains industry affected by C. bonariensis and Sonchus oleraceus, about the impacts and desired management goals for these weeds. Both weeds were included in the same survey because they are problems in the similar areas and land uses. for the A total of 60 responses were received; 51 complete (85% answered all questions), and 9 partial responses. Respondents identified as either agricultural landholders (55.2%) or agricultural land manager, agronomist or extension officer (48.3%).

Respondents indicated that most of the agricultural impacts listed for these weeds were relevant as four of the six impact statements received > 50% response (respondents were able to select more than one impact statement). The leading impact statement was that these weeds are “difficult to control” (79.2%) followed by “reduces stored water supplies in fallow” (66%).

Eighty-seven percent of respondents indicated that they felt that a new tool such as biocontrol would be useful in the management of fleabane and sowthistle (11.1% neutral, 1.9% disagreed). Fallow, roadsides and fence lines were considered by most respondents to be areas in which biocontrol could contribute to management as they received >50% responses.

The respondents selected the following top four management objectives to which biocontrol needs to contribute to be considered successful were:

• Decrease weed management cost and effort,
• Reduce herbicide inputs required,
• Decrease the need for follow-up control, and
• Reduce the occurrence of new infestations.

A review on the costs of weed to the grain industry (Llewellyn et al. 2016), stated that C. bonariensis and S. oleraceus are high impact summer fallow weeds responsible for revenue losses of $43.2 and $4.9 million per year respectively. These two cropping weeds are also responsible for $3.6 (C. bonariensis) and $1.3 (S. oleraceus) million annually in additional herbicide costs due to the development of herbicide resistance in populations these species. The economic and chemical inputs required to control these weeds are set to increase if herbicide resistance increases in frequency and distribution. Based on the outputs of the management objectives’ survey, these fiscal impacts serve as a baseline against which the economic and management success of any introduced biocontrol agents can be judged in the future.

Nominated Flaxleaf fleabane as a biocontrol candidate

The project prepared the documentation to support the nomination of C. bonariensis as a target for biocontrol. The documentation was submitted to the IPAC (now EIC), by the Queensland Department of Agriculture and Fisheries in May 2017 and endorsed by the Committee in November 2017 (Rafter and Morin 2017).

Conducted genetic analysis on samples of Flaxleaf fleabane from different regions in Australia and the native range

A total of 375 putative individuals of C. bonariensis: 239 individuals from 18 sites in Australia, 60 individuals from 8 sites in Brazil, 9 individuals from 1 site in Argentina and 67 herbarium specimens from the Americas were analysed with Diversity Arrays Technology (DArT)seq. DArTseq Single Nucleotide Polymorphism (SNP) data for these samples consisted of 100,629 loci which was reduced to 18,110 loci following filtering of DArT parameters.

Due to challenges with collecting fresh specimens in the native range for this study, the dataset comprised samples from herbarium specimens. This is a newly emerging approach to supplement sampling in cases where it is difficult to collect fresh material from the field. It is noteworthy that 50% of the 137 samples from putative C. bonariensis herbarium specimens processed were successfully extracted and sequenced.

Phylogenetic trees using the whole data set were generated using both Geneious (https://www.geneious.‌com/‌‌geneious/) and SplitsTree (http://splitstree.org/) softwares. These trees indicated the presence of two groups of samples, with no evidence of hybridisation between them. A further phylogenetic tree was generated in Geneious using a reduced dataset comprising three randomly selected plants from each Australian sites where fresh material was collected and all other samples in the dataset. We also excluded from this analysis the six West Mackay samples previously found to cluster with Group 1, as this indicated that sampling at this site was performed on more than one Conyza species.

 

Group 1 comprised samples primarily from South and Central America, with just one sample from the United States (close-up figure of Group 1 available on request). Group 2 comprised all Australian samples and the remaining 18 United States samples as well as samples from Brazil and a small number of other samples from South and Central America. Group 2 thus corresponded to C. bonariensis, as it is known in Australia. In contrast, Group 1 indicated that several of the herbarium samples included in the analysis had been misidentified as C. bonariensis. Furthermore, the analysis revealed that not all fresh samples collected from putative C. bonariensis plants at the same sites in Brazil were in the same group, indicating that plants from more than one Conyza species were sampled by our collaborators. This analysis indicated that C. bonariensis collected in Australia were more closely related to herbarium samples from Chile, Costa Rica and Guatemala, followed by a sample from Bahamas and a subset of samples from the United States.

The Bayesian clustering program STRUCTURE and the Poppr R package were subsequently used to assess the extent of population genetic structure among samples in Group 2. Since this group was dominated by Australian samples, we undertook a different approach that firstly determined how many genetic clusters (K) occurred within the invaded Australian range without any prior knowledge of population affinities. The analysis showed that two genetic clusters were most likely present within Australia and that two populations from Western Australia (Mocardy; H and Oakabella; O) were distinct to the other Australian populations. Assignment of native range samples revealed that they were all admixed and that some samples in the United States had higher levels of admixture with the genetic cluster that characterised the two different Western Australian populations. The levels of admixture in native range samples makes it challenging to determine where Australian samples have originated from, although there is a suggestion that the two Western Australian populations have come from the United States.

Undertook bioclimatic modelling to identify optimal locations and conduct native range surveys and host-specificity tests for potential biocontrol agent(s) and imported at least one potential agent in quarantine

Bioclimatic modelling

The approaches taken for the bioclimatic modelling of C. bonariensis were like those used for Lycium ferocissimum; Match Climates and Compare Locations models developed using the CLIMEX package (Kriticos et al. in preparation). An experiment was also conducted to measure the growth rate of C. bonariensis at different temperatures.

 

 

Native range surveys

Pathogens

Surveys for pathogens on Conyza species were performed in different departments of Colombia between November 2017 and May 2019. Surveys concentrated on Colombia because several rust fungi had only been recorded from Conyza sp. in this country in South America. A total of 136 Conyza sp. samples with disease symptoms were collected across all surveys during that period. High morphological diversity in Conyza species was observed in the field and it was difficult to categorically identify plants to species level. Consequently, plant tissue was collected from representative specimens and DNA extracted and sequenced to obtain a more reliable identification. Although some collections were made on C. bonariensis, most were made on C. sumatrensis. Collaborators in Colombia had the opportunity to present a few of their Conyza sp. herbarium specimens to Dr John Pruski was visiting the plant herbarium in Medellín during winter 2019. Dr Pruski is an expert taxonomist on the Asteraceae family based at the Missouri Botanical Garden in St. Louis, MO, USA. He identified each of the five herbarium specimens presented to him as either C. sumatrensis var. leiotheca or C. sumatrensis var. sumatrensis.

Disease symptoms observed during the field surveys ranged from chlorotic or necrotic lesions on leaves or stems, mildew and rust. Several fungi were recovered from the symptoms, including Basidiophora entospora, Oidium sp., Cercosporella sp., Septoria sp., Wentiomyces sp., Diaporthe sp. (confirmed with sequencing) and Alternaria sp. Two rust fungi were also found: a species with aecia on Conyza sp. and the microcyclic species identified as Puccinia cnici-oleracei. The former rust fungus was later confirmed in a cross-inoculation study to be an heteroecious, macrocyclic rust species, with Cyperus sp. as its main host and Conyza sp., thus, unsuitable for biocontrol.

Insects

Surveys for insects have been undertaken on Conyza species in Argentina, Paraguay, Brazil, Colombia and the southern USA (Louisiana, Texas, Alabama) between November 2017 and February 2020. These surveys have identified herbivorous insects from some 14 families (Agromyzidae; Cecidomyiidae; Cerambycidae; Curculionidae; Lixidae; Membracidae; Miridae; Mordellidae; Pseudococcidae; Pterophoridae; Tephritidae; Tingidae; Tortricidae; Coccidae) associated with Conyza spp. in the native range. Some 35 species/morphospecies have been identified to date, and among these the most promising are two species of gall flies (Trupanea bonariensis that forms stem galls; an unindentified fly that forms leaf-blister galls), a weevil (Lixus sp., a stem- and root-feeder) and two species of moths (both unidentified, both leaf rollers). These species have all been recorded in Argentina and Brazil, in areas of strong bioclimatic similarities to where C. bonariensis occurs as a weed in Australia. Several of the insect species in the native range are new to science and are in the process of being identified by taxonomic experts.

Importation of candidate agents into quarantine 

Pathogens

Puccinia cnici-oleracei (ex. Conyza) was deemed the most promising candidate biocontrol agent to investigate. The necessary export permit was obtained from the relevant Colombian authorities. Concurrently, a permit to import the fungus in the CSIRO quarantine facility in Canberra was obtained from DAWE. The fungus was imported on 20 November 2018. A single-telium isolate was generated from the material imported and a culture established in quarantine.

Insects

Trupanea bonariensis, the stem gall forming tephritid fly, was imported in a quarantine facility in Australia in November 2019 and February 2020. Colonies of this fly were in the process of being established as a precursor to detailed biological studies and host-specificity testing. This work was undertaken as part of the following Rural R&D for Profit project (Agrifutures Australia Project number: PRJ-12377).

Host-specificity tests for candidate biocontrol agents

The proposed list of non-target species for host-specificity testing of candidate biocontrol agents for C. bonariensis was submitted to DAWE in December 2018 for posting on their website for feedback (Hunter et al. 2018).

Pathogens

A cross-inoculation experiment was performed in Colombia to obtain an initial indication of the specificity of P. cnici-oleracei. The experiment included accessions of P. cnici-oleracei recovered from Conyza sp. and Emilia sonchifolia, which were growing in proximity at the same site. Both Conyza (=Erigeron) and Emilia species are recorded as hosts of P. cnici-oleracei, although the fungus on Emilia is also referred to as Puccinia emiliae by some authors (Farr and Rossman 2020). Results demonstrated that the rust accessions were capable of only infecting the host species they originated from. Plants of Conyza sp., but not of E. sonchifolia, developed disease symptoms when exposed to rust-infect Conyza sp., and vice versa. Based on these results, we concluded that the fungus from Conyza sp. was probably highly specific and thus decided to refer to it as P. cnici-oleracei (ex. Conyza).

Comprehensive host-specificity testing in quarantine began in February 2019 and only one experiment remained to be performed. The test list comprises a total of 50 closely related, non-target species in the subfamily Asteroideae of the family Asteraceae that occur in Australia (ornamental, weed and native). Most species have been tested in at least two separate experiments using different accessions of each plant species, with C. bonariensis plants used as positive controls in all experiments.

Our results thus far showed that P. cnici-oleracei (ex. Conyza) is highly host specific to C. bonariensis. The fungus successfully developed and produced telia only on the nine Australian accessions of C. bonariensis tested. While Conyza sumatrensis and Bidens pilosa developed necrotic flecks and in some instances a few large necrotic blotches, the fungus never produced any telia on these species. Chlorotic flecks developed on one accession of Calendula officinalis, and a few, rare pin-sized telia were observed on one replicate of Eschenbachia leucantha. Inoculation of C. bonariensis using these pin-sized telia did not result in any infection. All other non-target plant species tested did not develop any visible symptoms and were rated as either immune or highly resistant based on microscopic examinations of the development of the fungus on these species.

Insects

Field observations of host-specificity have been made on insects recorded to date. As part of these field surveys, assessments were made on co‐occurring Asteraceae species (including species in the genera Eupatorium, Chromolaena, Ageratrum, Senecio, Bidens and Baccharis) to see if the insects being recorded on Conyza species are also found on these species. Trupanea bonariensis that forms stem galls, an unindentified fly that forms leaf-blister galls, a stem- and root-feeder weevil Lixus sp. and two species of moths (both unidentified, both leaf rollers) show promise for biocontrol. They cause significant damage and are seldom seen on co-occurring species in the Asteraceae in the field. Colonies of these species have been established in Brazil to elucidate their biology prior to importation into a quarantine facility in Australia to undergo detailed host-specificity testing (as part of a new Rural R&D for Profit project (Agrifutures Australia Project number: PRJ-12377; 2019-2022)).

Pending risks to non-target plants are acceptable, submitted application to the Commonwealth regulators seeking approval to release at least one potential agent. Upon receiving approval, released biocontrol agent(s)

Host-specificity testing with the rust fungus P. cnici-oleracei (ex. Conyza) was completed at the end of April 2020. A draft of the release application was prepared and submitted (Morin et al. 2020) to DAWE.

Explored options for integration of biocontrol with other management techniques

Conyza spp. (incl. all three exotic Conyza spp.) and S. oleraceus are similar in terms of their impacts in grain production systems. So, the options for integration of biocontrol with other management techniques are somewhat analogous.

Both weeds are principally managed through chemical and cultural control techniques in agricultural/cropping systems during the growing season (Wu 2007; Widderick 2014; Widderick & van der Meulen 2016; Widderick et al. 2004). This involves a combination of chemical control using postemergence herbicides of two herbicides (double-knock; either as a mix or applied sequentially) for in-crop control, and the use of a residual herbicide in fallows. The use of this single tactic has resulted in herbicide resistance in both weeds. Cultural tactics include decreased row spacing for in-crop weed management (Wu & Walker 2004), and the strategic use of tillage for burial of weed seed to below 2 cm depth in fallows (Werth & Walker 2007); the latter disrupts the benefits of minimum tillage farming.

Chemical approaches can manage both weeds effectively at a cost, but there is an opportunity for integration with biocontrol to possibly reduce costs and to preserve the utility of the effective herbicides by delaying resistance. Biocontrol agents (e.g. like Puccinia cnici-oleracei (ex. Conyza) being developed as part of this project) could serve as chronic stressors to the weed in fallows and outside cropped areas and limit reproductive output of the weed, thereby reducing the risk of seedbank build-up in fallows and also the risk of spreading into fields from surrounded non-cropped areas. The utility of biocontrol suppressing weed performance in unmanaged contexts (i.e. beyond crop fields and fallows) could further limit the rate of in-crop incursion of weed seeds within the growing season.

Integration of biocontrol with chemical and cultural control tactics will require coordination among land managers and consultants recommending/deploying management tactics for Conyza spp. and S. oleraceus.

References

Farr DF, Rossman AY (2020) Fungal databases. U.S. National Fungus Collections, ARS, USDA, Available at: https://nt.ars-grin.gov/fungaldatabases/ Accessed 15 March 2020

GBIF.org (24th July 2018) GBIF Occurrence Download https://doi.org/10.15468/dl.pbd67p.

*Hunter GC, Rafter MA, Raghu S and Morin L (2018) Proposed plant host test list for assessing risk of candidate biological control agents for Conyza bonariensis. Prepared by CSIRO.

Llewellyn R, Ronning D, Clarke M, Mayfield A, Walker S, Ouzman J (2016) Impact of weeds on Australian grain production – The cost of weeds to Australian grain is in the adoption of weed management and tillage practices. CSIRO, Australia.

Mangolin CA, Silvério de Oliveira Junior R, de Fátima P.S. Machado M (2012) Genetic Diversity in Weeds. In: Alvarez-Fernandez R (ed) Herbicides – Environmental Impact Studies and Management Approaches. IntechOpen, Available from: https://www.intechopen.com/books/herbicides-environmental-impact-studies-and-management-approaches/genetic-diversity-and-structure-of-weed-plant-populations. doi:DOI: 10.5772/32733.

McClintock D, Marshall JB (1988) On Conyza sumatrensis (Retz) E. Walker and certain hybrids in the genus. Watsonia 17: 172–173.

Michael PW (1977) Some weedy species of Amaranthus (amaranths) and Conyza/Erigeron (fleabanes) naturalized in the Asian-Pacific region Proceedings of the 6th Asian-Pacific Weed Science Society Conference, Indonesia, 1977:87–95

PlantNET 2020, NSW Flora Online (http://plantnet.rbgsyd.nsw.gov.au/floraonline.htm)

*Rafter M, Morin L (2017) Nomination of a target weed for biological control: Conyza bonariensis L. (Asteraceae). Prepared by CSIRO.

Scott JK, Yeoh PB, Michael PJ (2016) Methods to select areas to survey for biological control agents: An example based on growth in relation to temperature and distribution of the weed Conyza bonariensis Biological Control 97: 21–30 doi:10.1016/j.biocontrol.2016.02.014

Thébaud C, Abbott RJ (1995) Characterization of invasive Conyza species (Asteraceae)in Europe: Quantitative trait and isozyme analysis. American Journal of Botany 82: 360–368.

Thébaud C, Finzi AC, Affre L, Debussche M, Escarre J (1996) Assessing why two introduced Conyza differ in their ability to invade Mediterranean old fields. Ecology 77: 791–804.

Werth J, Walker S (2007) Tillage effects on fleabane emergence Proceedings of a fleabane workshop held at Queensland Department of Primary Industries and Fisheries in Toowoomba on 7th February 2007, pp. 22–23.

Widderick M and van der Meulen A. (2016) Ecology and management of common sowthistle. Available at: https://www.youtube.com/watch?v=I_5gm-ZhGE0  Accessed 17 March 2017: Grains Research and Development Corporation.

Widderick M (2014) Weed 15: Common sowthistle (Sonchus oleraceus), in Storrie, A. M. (ed), Integrated weed management in Australian cropping systems. Grains Research and Development Corporation.

Widderick M, Walker S and Sindel B (2004) Better management of Sonchus oleraceus L. (common sowthistle) based on the weed’s ecology. In Proceedings of the 14^th Australian Weeds Conference. 6–9 September 2004, Wagga Wagga, New South Wales, Australia. Sindel, B.M. and Johnson, S.B. (eds), pp. 535-537, Weed Society of New South Wales, Sydney, New South Wales.

Wu H (2007) The biology of Australian weeds 49. Conyza bonariensis (L) Cronquist. Plant Protection Quarterly, 22: 122–131.

Wu H, Walker S (2004) Flaxleaf fleabane a difficult-to-control weed in dryland cropping systems associated with zero-tillage. Australian Grain, northern focus iii-v, November-December 2004.

Zelaya IA, Owen MDK, VanGessel MJ (2007) Transfer of glyphosate resistance: evidence of hybridization in Conyza (Asteraceae). American Journal of Botany 94: 660–673.