Host-specificity testing of Puccinia xanthii var. parthenii-hysterophorae, a potential biocontrol agent for Parthenium hysterophorus in South Africa

Abstract

Following its introduction into Australia, the rust fungus Puccinia xanthii var. parthenii-hysterophorae has played an important role in the integrated approach to manage Parthenium hysterophorus. The rust was imported from Australia and established within quarantine facilities in South Africa. Supplementary host-specificity testing of the rust on indigenous members from the tribe Heliantheae and 13 commercial sunflower cultivars was conducted. The rust was shown to be highly specific to parthenium, with only a single incident of symptom development on leaves of the native Spilanthes mauritiana. The symptoms on S. mauritiana were considered to be a laboratory artifact as no further symptoms were observed on this species during subsequent inoculations. Microscopic evaluations revealed that although basidiospores germinate on the surface of S. mauritiana leaves, they do not penetrate the epidermis. Spilanthes mauritiana is host to Puccinia africana, but cross-inoculation greenhouse studies showed that although P. africana is assumed to belong to the P. xanthii complex, it can be separated as a morphospecies from P. xanthii var. parthenii-hysterophorae as it is highly specific to its host. Based on these results it was concluded that P. xanthii var. parthenii-hysterophorae was suitable for release as a biological control agent on Pa. hysterophorus in South Africa.

Keywords:

• biological control
• host specificity
• Parthenium hysterophorus
• parthenium weed
• Puccinia xanthii var. parthenii-hysterophorae

Introduction

Parthenium hysterophorus L. (Asteraceae) originates from tropical America around the Gulf of Mexico (Haseler 1976) and occurs from the southern USA to central South America (Dale 1981). This highly invasive, noxious weed has become widespread in Africa, many South Pacific Islands, Asia and Australia (Navie et al. 1996, Evans 1997, McConnachie et al. 2010). In South Africa this weed was first observed in KwaZulu-Natal in the 1880s (Wood 1897), became prevalent in the 1980s in the aftermath of cyclone Demoina in 1984 (van der Laan 2006), and has since spread to the Mpumalanga and North West provinces (McConnachie et al. 2010).

Parthenium is an annual herb that can reach a height of 2 m, producing thousands of seeds that remain viable for several years (Navie et al. 1996). It invades disturbed areas and has a high incidence in areas that are regularly disturbed by floods (McFadyen 1992). Its allelopathic nature is well documented (Kanchan and Jayachandra 1980, 1981, van der Laan 2006), as well as its ability to reduce crop production and the carrying capacity of natural rangelands (McFadyen 1992, Navie et al. 1996, Evans 1997). Parthenium causes severe allergic responses in humans and livestock (Towers and Subba Rao 1992).

Existing management practices for parthenium weed in South Africa consist of mechanical and chemical approaches, which are often uneconomic and carry associated health implications from exposure to the plant. It is widely recognised that they should be supplemented with biological control, which is sustainable, cost effective and environmentally friendly (Evans 1997). Surveys conducted in the native range of parthenium weed revealed that there were numerous organisms that suppressed growth of the plant, suggesting that implementing a biological control programme was a good management option (Haseler 1976). Eleven agents were released in Australia against parthenium weed (McFadyen 1992), and evaluation studies revealed that biological control (using insect and fungal agents) significantly improved grass production in Australian rangelands (Dhileepan 2007).

The two most important pathogens associated with Pa. hysterophorus in its native range are two rust fungi, Puccinia abrupta Dietel & Holw. var. partheniicola (H.S.Jacks.) Parmalee and Puccinia xanthii Schwein. var. parthenii-hysterophorae Seier, H.C.Evans & Á.Romero (previously referred to as Puccinia melampodii Dietel & Holw.) (Evans 1987a, 1997, Parker 1989). Puccinia abrupta var. partheniicola was found to be already present in South Africa (Wood and Scholler 2002), presumably having been introduced along with its host plant. In Mexico, this rust was damaging to parthenium weed in semi-arid, high-altitude areas, but in humid, lowland situations infection was very light and scattered (Evans 1987b). The optimal temperature for germination of urediniospores and infection of plants is 15 °C (Fauzi et al. 1999). Therefore, this rust fungus will have little impact in the low-altitude, high-temperature areas in which the weed is currently spreading in South Africa. However, it may impact the weed in high-altitude, low-temperature areas. For this reason, the research focus was mainly on P. xanthii var. parthenii-hysterophorae, which was collected from the humid lowland areas in Mexico (Parker 1989) and since 2000 has been successfully used as a biological control agent against Pa. hysterophorus in Australia (Tomley 2000).

The life cycle, biology and host-specificity of P. xanthii var. parthenii-hysterophorae were extensively investigated in Mexico (Romero and Duran 2000) and by CABI-UK (Seier et al. 1997, Tomley 2000) before release of the rust fungus in Australia. Puccinia xanthii var. partheniihysterophorae is an autoecious, microcyclic rust fungus, producing both telia and basidiospores on one host (Parmelee 1967). The teliospores germinate over a wide temperature range (optimum being 25 °C) and produce basidiospores (optimum at 22 °C), which directly penetrate the host epidermis (Romero and Duran 2000, Tomley 2000, Kelaniyangoda and Ekanayake 2010). Symptoms of P. xanthii var. parthenii-hysterophorae first appear approximately 9 d after infection and can be observed as chlorotic spots on the leaves (Seier et al. 1997). Ten to 12 d after infection, telia erupt from the centre of these chlorotic spots, which gradually extend outwards (Tomley 2000). Under severe infection these chlorotic spots can coalesce, leading to leaf necrosis and die-back. Stem infections have also been observed (Seier et al. 1997, Tomley 2000).

Prior to the release of P. xanthii var. parthenii-hysterophorae in Australia, host-specificity testing was conducted on over 60 plant species within the family Asteraceae, as well as on 13 sunflower cultivars (M Seier, CABI, UK, pers. comm.). Limited infection was observed on a few plants, but in most cases the infection consisted of abnormal or abortive sporulation and the level of sporulation was much less than those on parthenium weed (Seier et al. 1997, Tomley 2000). A field experiment was conducted in Australia after the release of P. xanthii var. partheniihysterophorae, to confirm the host range under natural conditions (Tomley 2000). Although all plants on which symptoms developed under quarantine conditions were included, limited sporulation occurred on Calendula officinalis only (Tomley 2000). Since its release in Australia in 2000 there have been no reports of P. xanthii var. parthenii-hysterophorae infecting any plant other than Parthenium hysterophorus.

Puccinia xanthii is associated with various species within the tribe Heliantheae, namely Ambrosia spp. and Xanthium spp. (Bremer 1994). However, molecular and crossinoculation studies concluded that P. xanthii consists of a complex of morphospecies that can be separated by their hosts (Seier et al. 2009). In addition, Spilanthes mauritiana (Pers.) DC. is host to Puccinia africana Cooke in South Africa (Doidge 1950). This rust is morphologically indistinguishable from Puccinia xanthii (Henderson 1970) and is likely part of the P. xanthii complex.

The objective of this study was to evaluate the risk associated with introducing the Puccinia xanthii var. parthenii-hysterophorae strain originally collected in Mexico then released into Australia, into South Africa. This paper reports on the host-specificity testing conducted in South Africa on native species supplementary to those tested in Australia. In addition, cross-inoculation studies were performed using P. africana and P. xanthii var. partheniihysterophorae to determine whether they are specific to their known hosts.

Materials and methods

Plant material

Seeds of Pa. hysterophorus were collected from mature plants in the field and sown in seedling trays in a greenhouse maintained at 20–25 °C (day) and 18–20 °C (night). Seedlings were transplanted into 18 cm diameter plastic pots. Seeds were collected regularly for propagation. These plants were used to culture the rust fungus, P. xanthii var. parthenii-hysterophorae and for tests. Test plants were obtained either as seeds, seedlings or mature plants from the field or from local nurseries, and were maintained under the same conditions as parthenium plants.

Maintenance of rust culture and inoculation of plants

Parthenium leaves with air-dried teliospores of P. xanthii var. parthenii-hysterophorae strain W1496 were obtained from the Alan Fletcher Research Station, Brisbane, Australia, and imported into the plant pathogen quarantine facility of the ARC–PPRI Stellenbosch campus in 2004. Potted parthenium plants were inoculated using the inverted leaf-disk method of Morin et al. (1993), in which leaf discs with rust pustules were placed on water agar (1.5%, w:v) in a Petri dish. The Petri dish was attached with an adhesive to the bottom of a 25 cm diameter pot. The surface of the plant and the inoculum were misted with water before the pot was inverted over a pot of similar size containing the test plant. The pots were sealed together to form an incubation chamber. The plants were incubated at 22 °C for 24 h, after which the inoculation chambers were dismantled and the plants were transferred to a glasshouse with 20-25 °C (day) and 18-20°C (night) temperatures.

Host-specificity testing

The tribe Heliantheae, to which both parthenium weed and sunflower belong, has its major diversity in the Americas. There are few members indigenous to South Africa, with six genera and nine species in total. None of the native Heliantheae required for testing were available commercially, but attempts were made to field-collect as many species as possible. Representative species from four indigenous genera of the Heliantheae were tested (Table 1), as well as 13 locally produced commercial sunflower cultivars.

Table 1 : Host-specificity testing of Puccinia xanthii var. partheniihysterophorae conducted in South Africa

Test plants of tribe Heliantheae (including sunflower cultivars)	Presence or absence of rust fungus symptoms
Aspilia natalensis ^a	No sporulation
Bidens schimperi ^a	No sporulation
Blainvillea gayana ^a	No sporulation
Spilanthes mauritiana ^a	Three pustules on two leaves
Melanthera scandens	No sporulation
Thelechitonia trilobata	No sporulation
Zinnia peruviana	No sporulation
Helianthus annuus PAN 7010	No sporulation
H. annuus PAN 7033	No sporulation
H. annuus PAN 7034	No sporulation
H. annuus PAN 7044	No sporulation
H. annuus PAN 7371	No sporulation
H. annuus PAN 7031	No sporulation
H. annuus PAN 7050	No sporulation
H. annuus PAN 7043	No sporulation
H. annuus Hysun 338	No sporulation
H. annuus Hysun 350	No sporulation
H. annuus PNR 6488	No sporulation
H. annuus Orit	No sporulation
H. annuus Hysun 333	No sporulation
^a Heliantheae indigenous to South Africa

Parthenium and test plants were inoculated as previously described. Inoculations were conducted using batches of plants (six pots per batch), each pot in its own incubation chamber. Parthenium plants were inoculated at the rosette stage and sunflower plants were inoculated at the six-leaf stage. Each pot contained four sunflower plants or a single plant of the other test species. The experiment was repeated three times, using six pots per species for each experiment. In each batch one parthenium plant was included as a control. The plants were monitored for presence or absence of symptoms, i.e. pustule development, 26 d after inoculation when the control plants were heavily infected with well-developed telia.

A subsample of inoculated leaves of the indigenous species were harvested and, after clearing in Carnoy's solution (75% alcohol, 25% acetic acid) for a period of at least 48 h, the leaves were removed and stained with 0.1% aniline blue in lactophenol for a period of 6–24 h, depending on the thickness of the leaf. They were mounted in lactic acid (50% aqueous solution) and examined microscopically.

Cross-inoculation

The rust pathogen Puccinia africana was collected from Spilanthes mauritiana plants in Umhlanga, KwaZulu-Natal (29°42′44.9″ S, 31°05′37.5″ E). This rust was established on S. mauritiana plants in the quarantine facilities at ARC-PPRI, Stellenbosch, and subsequently multiplied for inoculation studies. Cross-inoculation studies were conducted using nine Pa. hysterophorus and S. mauritiana plants each. In each experiment three S. mauritiana plants and three Pa. hysterophorus plants were inoculated with each respective rust fungus, using the incubation chamber technique, as previously described and the experiment was repeated three times. A subsample of the inoculated leaves was harvested from each plant and examined microscopically, using the technique described above.

Results and discussion

No symptoms or pustules developed on any of the 13 sunflower cultivars tested. During host-specificity testing in Australia, two Australian sunflower cultivars (not cultivated in South Africa) were infected under glasshouse conditions (Seier et al. 1997). However, teliospores from these infections were shown to be infertile and field testing undertaken in Australia demonstrated that P. xanthii var. parthenii-hysterophorae poses no risk to sunflower production. Another P. xanthii form that occurs on Xanthium strumarium has been shown to be able to infect a few sunflower cultivars under artificial conditions (Alcorn 1976, Morin et al. 1993, Pretorius et al. 2000) and field conditions (Alcorn and Kochman 1976, Gulya 2002). However, in Mexico, sunflower is cultivated extensively where P. xanthii var. parthenii-hysterophorae occurs naturally, and to date there has been no record of this rust affecting any sunflower cultivar (Farr and Rossman 2012). It was concluded therefore that P. xanthii var. partheniihysterophorae does not pose a risk to sunflower production in South Africa.

No sporulation was recorded on any of the other plant species tested, except for a very limited number of pustules on Spilanthes mauritiana. On one occasion only, two leaves, each on a different plant, bore a total of three pustules. Microscopic examination of S. mauritiana leaves revealed that although basidiospores of P. xanthii var. parthenii-hysterophorae germinated, they failed to penetrate the leaf epidermis (Figure 1). It was concluded that the once-off infection of S. mauritiana by P. xanthii var. parthenii-hysterophorae was a false negative result because of artificial high inoculum loads characteristic of testing under glasshouse conditions (Evans et al. 2001), as found for some species tested in Australia (Seier et al. 1997).

Figure 1:  Light micrographs (1000×) of the infection process of Puccinia xanthii var. parthenii-hysterophorae on Parthenium hysterophorus and Spilanthes mauritiana. (a) Puccinia xanthii var. parthenii-hysterophorae germinated basidiospore (B) with a short germ-tube (GT) that has formed a simple appressorium (A) on the host epidermis at 24 h after inoculation on Parthenium hysterophorus. (b) Puccinia xanthii var. parthenii-hysterophorae intra-epidermal vesicle (IV) with a primary hypha (H) 24 h after inoculation on Parthenium hysterophorus. (c) Puccinia xanthii var. parthenii-hysterophorae germinated basidiospore (B) producing a long germ-tube (GT), without penetration, 24 h after inoculation on Spilanthes mauritiana. Scale bars = 10 μm

Spilanthes mauritiana was highly susceptible to Puccinia africana, but in contrast Pa. hysterophorus was not affected by this rust (Figure 2). This indicates that although P. africana might form part of the Puccinia xanthii complex, it differs from the race occurring on parthenium and is considered to be a different form.

Figure 2:  Light micrographs of the infection process of Puccinia africana on Spilanthes mauritiana and Parthenium hysterophorus. (a) Puccinia africana intra-epidermal vesicle (IV) with a branching primary hypha (H) 48 h after inoculation on Spilanthes mauritiana. (b) Puccinia africana basidiospore (B) not germinating on Parthenium hysterophorus 48 h after inoculation. Scale bars = μ10 m

Given the high level of host specificity demonstrated before release of P. xanthii var. parthenii-hysterophorae in Australia (M Seier, pers. comm.), together with the additional testing reported here, it is considered that P.xanthii var. parthenii-hysterophorae is safe for introduction into South Africa for the biological control of Pa. hysterophorus. Furthermore, this rust fungus has been demonstrated to be an effective agent under favourable climatic conditions. In Mexico P. xanthii var. parthenii-hysterophorae appeared to be well adapted to dry conditions as it was found to persist in areas where rain was absent for four months, but was most aggressive during the wet season (Tomley 2000). In Australia, P. xanthii var. parthenii-hysterophorae established rapidly and over a wide range in northern Queensland (Dhileepan et al. 2006). The performance of the rust, however, varied between years depending on rainfall, with the highest impact during a higher-rainfall year (Dhileepan 2007). It is expected that P. xanthii var. parthenii-hysterophorae will contribute greatly to the management of parthenium weed infestations in the warm, lower-altitude regions of South Africa, where there are currently no biological control agents implemented against this weed.

Acknowledgements

We thank Gwen Samuels, Lea Orien and Christopher Xhegwana of ARC-PPRI for assistance with plant and fungal culture maintenance. We also gratefully acknowledge financial support from the Department of Water and Environmental Affairs ‘Working for Water Programme’.