It was found that both plant species produced a very small amount of oil, but L. camara produced nearly three times more (0.09 ± 0.005% w/w) than L. radula (0.03 ± 0.006% w/w). The moisture content for the leaves of L. camara was 79 ± 0.51% and for L. radula 67 ± 0.01%. Previous studies of L. camara reported oil content ranging from 0.06% w/w to 0.22% w/w [6,21]. The components found for the oils of L. camara (19 compounds) and L. radula (nine compounds) are presented in Table 1.

The empirical observation made previously of a clear smell difference between the two species of Lantana was confirmed by gas chromatography and mass spectrometry analyses that revealed different compositions for their volatile oils. The common constituents for the two species are: tetradecane, E-caryophyllene, α-humulene, β-E-farnesene, germacrene-D, bicyclogermacrene and phytol. The major components in the oil of L. camara are germacrene D (19.8%), E-caryophyllene (19.7%), bicyclogermacrene (11.7%) and α-humulene (9.3%). These compounds represent approximately 60.5% of the oil. The four major constituents of the oil of L. radula are E-caryophyllene (25.3%), germacrene-D (17.6%), E-nerolidol (19.0%) and phytol (29.2%), representing 91.1% of the oil.

The greater chemical diversity of the volatile oils of L. camara as compared to L. radula could be connected with the presence of secretor idioblasts in L. camara [19] which are absent in L. radula but further studies are required to confirm this hypothesis.

As observed on Figure 1, oils obtained from each of these species caused a significant inhibition on fungal growth, as compared to controls.

In the first two days of evaluation the oil of L. camara at a concentration of 1,000 mg L⁻¹, caused a reduction in the growth of the fungal colonies of 25.0% and 19.6%, respectively. At higher concentrations the fungus growth was completely inhibited during the first day and for 10,000 mg L⁻¹ a full inhibition was still observed during the second day (Figure 1A and Figure 2).

At the eighth day of evaluation the reduction of the development of the colonies was dependent on the oil concentration tested and the following results were found: 12% inhibition at 1,000 mg L⁻¹; 17% at 3,000 mg L⁻¹; 27% at 5,000 mg L⁻¹ and 49% at 10,000 mg L⁻¹. These results are similar to published results where inhibitory effect of the oil from L. camara on the growth of Aspergillus niger van Tiegh and Fusarium solani Mart. (Sacc) are reported [15].

For the oil of L. radula at the concentrations of 1,000 mg L⁻¹ and 3,000 mg L⁻¹, 17.2% and 40.6% inhibition were observed, respectively, on the first day of evaluation (Figure 1B and Figure 3). At higher concentrations (5,000 and 10,000 mg L⁻¹) the oils caused complete inhibition of the fungus growth. Even after eight days of experiment the volatile oils of L. radula caused a significant reduction on fungus development (4.7% inhibition at 1,000 mg L⁻¹; 23.2% at 3,000 mg L⁻¹; 42.3% at 5,000 mg L⁻¹ and 60.8% at 10,000 mg L⁻¹).

The results demonstrated that the volatile oils from L. radula have a greater effect on the growth of C. cassiicola compared with the oils obtained from L. camara. Although bioassays with each individual component of oils from each plant species were not performed, it is likely that the greater activity of L. radula’s oil could be associated with its higher concentration of E-nerolidol, since this compound has already been shown highly active against Escherichia coli [37], to have larvicidal activity on Aedes aegypti larvae and anti-parasitic activity against a strain of Leishmania amazonensis [38]. Another significant difference between the oils is the higher concentration of phytol in L. radula oil, although there is no published report of any anti-fungal activity for this compound. Although caryophyllene is present in high concentration in both oils, previous investigation have demonstrated that it has no effect, at least against the basidiomycete species Laetiporus sulphureus, Lenzites betulina and Trametes versicolor [39]. The higher growth-inhibiting activity observed for L. radula oil on C. cassiicola may in part explain the observed resistance to attack by this fungus [19].

Aerial parts of L. camara and L. radula, from plants grown in a greenhouse, were used. Voucher specimen of each plant species were deposited in the VIC Herbarium [registration numbers 30159 (L. camara) and 30160 (L. radula)] for reference.

Leaves of L. camara and L. radula were collected and each sample was subdivided into three portions of 40 g each and then subjected to 3 h of hydrodistillation in a Clevenger apparatus. The resulting oils were separated from the aqueous phase, weighed and the reported yields were calculated with respect to the mass of dry material. All distillations were repeated three times and the oils obtained were separated from the aqueous phase and stored under a nitrogen atmosphere, maintained at 0 °C, until they were chromatographically analyzed. Dry leaf weight was calculated by drying each sample (2 g, held at 103 ± 2 °C for 24 h) [40]. Each analysis was carried out in triplicate.

Qualitative analyses were conducted on a GC-MS-QP5050A system with a mass-selective detector (Shimadzu, Kyoto, Japan), equipped with a DB-5 (J & W Scientific, Albany, NY, US) fused silica column (30 m × 0.25 mm i.d., film thickness 0.25 µm). Column temperature was 40 °C (2 min), increased at 3 °C/min to 240 °C, and kept at this temperature for 10 min. Injector temperature was 220 °C. Helium was the carrier gas at a flow rate of 1.8 mL/min. An amount of 1 μL (1% w/v solution of the oil in dichloromethane) was injected and the split ratio was 1:10. The column pressure corresponded to 100 kPa. Mass detector conditions were as follows: temperature source 240 °C; electron impact (EI) mode at 70 eV; scan rate 1 scan/s; mass acquisition range 29–450 u.

The identification of the components was performed by comparison of their retention indexes (RI), relative to a standard alkane series (C₉-C₂₄), and comparison of the mass spectra with those on record in the Wiley library data base (Wiley 330000) or from the literature [41].

GC analyses were carried out in triplicate, and accomplished with a GC-17A series instrument (Shimadzu) equipped with a flame ionization detector (FID). Chromatographic conditions were as follows: Fused silica capillary column (30 m × 0.22 mm) with a DB-5 bonded phase (0.25 μm film thickness); carrier gas, N₂ at a flow rate of 1.8 mL/min; injector temperature 220 °C, detector temperature 240 °C; column temperature was programmed to start at 55 °C (isothermal for 2 min), with an increase of 3 °C/min, to 240 °C, isothermal at 240 °C for 15 minutes; injection of 1.0 μL (1% w/v in dichloromethane); split ratio 1:10; column pressure of 115 kPa.

The oils obtained from each plant were dissolved in Tween 20 and tested in vitro for antifungal activity against C. cassiicola, using the “Poison Food Technique” [42]. Each oil was incorporated in the vegetable broth agar VBA [13] at 1,000, 3,000, 5,000 and 10,000 mg L⁻¹ (Tween 0.1%), and were vigorously agitated and poured sterilized Petri dishes (60 mm in diameter). The plates were centrally seeded with culture disks (5 mm in diameter) obtained from the margin of actively growing cultures of C. casiicola f. ssp. lantanae (isolate JMP 17) from the culture collection of the Department of Plant Pathology at UFV and incubated at 25 ± 2 °C in the dark. Plates containing the fungus but without Lantana oil served as controls and were incubated under the same conditions. The incubation lasted 8 days in the dark. The effect of the oil on the mycelial growth (mm) was determined by measuring the radial growth of C. cassiicola in intervals from the 1st to the 8th day after the inoculation. The percentage of fungus growth inhibition was calculated in relation to the radial growth in control plates. Treatments were carried out in a completely randomized design with four replications and the data was analyzed by the Tukey’s test at 0.05 probability level.