Aspects of management of poplar rust in South Africa.
An assessment of infection on poplar clones grown at the Lion Match Company Redclyff Nursery in Seven Oaks, KwaZulu-Natal was made to ascertain the nature of infection of the three common species of Melampsora infecting poplars in South Africa. These three species are M. larici-populina, M. medusae and the hybrid, M. medusa-populina. Their contrasting wall echinulations were used to differentiate these species using a scanning electron microscope. A visual rating scale measuring percentage leaf area infected (LAI) was used to determine disease severity. Rust development was slow in new material obtained from New Zealand, indicating rust resistance. This new material was not infected with M. larici-populina. Plant material from Europe showed severe susceptibility to M. larici-populina. Melampsora medusae-populina was the most prevalent species found at Lion Match Company plantations, Seven Oaks, KwaZulu-Natal, South Africa in the survey conducted from January to April, 1998. A once-off survey of rust infected poplar leaves from ten different locations (over 1 500 km apart) was conducted to ascertain the effect of geographical and meteorological conditions on the presence and severity of rust on poplars. The most popularly grown clones in South Africa are Clone 65/29, Clone 65/31, Clone 1488, Clone 129 and the old clone Populus deltoides var. missouriensis. Clone 65/31 had the greatest severity of disease (10.4%) closely followed by Clone 1488 (9.5%). Clone 129 had the least amount of disease (1.8%). The most common rust species occurring in South Africa was M. larici-populina. The hybrid, M. medusae-populina, was the least prevalent, although race studies found this species to be the most virulent. Four trials were conducted to determine the potential of fungicides to control rust infections on poplars. Sixteen fungicides were tested. Naturally infected poplar trees of the clone 65/31, grown in pots, were used as test material. The first trial had 16 fungicide treatments and an untreated control. Four treatments were significantly more effective than others: Alto (cyproconazole) (at 0.3 ml/L) with and without the adjuvant Armoblen 600 (at 0.75 ml/L), Anvil (hexaconazole) (at 0.2 ml/L) and Early Impact (flutriafol + carbendazim) (at 0.6 ml/L), respectively. Oxycarboxin appeared to have enhanced disease progression. Two experimental strobilurins, Stroby WG and Quadris FL (kresoxim-methyl, BASF and azoxystrobin, Zeneca) (at 0.12 ml/L, and O.4 ml/L, respectively) and a new class of fungicide, Astor WG40 (experimental, Novartis) (at 2g/L) controlled rust poorly. Four treatments were used in the second trial: Quadris as a foliar spray (O.4 ml/L), Impact applied on superphosphate granules «1 ml + 5g)/tree), and two controls; superphosphate alone (5g/tree) and untreated. The key finding of the second trial was that Impact gave complete control as a granular application over a 56 day period. Superphosphate alone enhanced rust development slightly. A third trial was conducted which corroborated results obtained in the first two trials: Alto plus Armoblen 600 was the best treatment, Early Impact the next best, then Alto, Anvil, superphosphate coated with Impact, Impact alone, the untreated control, Duett and lastly, superphosphate alone. The superphosphate treatment again slightly increased the disease percentages. A fourth trial was conducted with different rates of Alto (0.1, 0.2, 0.3, 0.5, 0.7 ml/L, and an untreated control), applied with the standard rate of Armoblen 600. All rates of Alto gave control of the disease, the highest rate being the most effective. A 3 x 3 x 3 factorial design was used to determine the effect of nitrogen (N), phosphorus (P) and potassium (K) on the growth of poplar trees and development of rust infection. Nitrogen (limestone ammonium nitrate (LAN) at 28% N) was applied at 0, 15.5 and 31 kg/ha, K (KCI at 50% K) was applied at 0, 16.7 and 33.3 kg/ha and P (single superphosphate at 10% P) at 0, 5.3 and 10.6 kg/ha. Over one year the single best tree grew 4.1 m, having received a treatment of 31 kg N/ha, 10.6 kg P/ha and 16.7 kg K/ha. This same treatment gave the best mean growth of 3.1 m. The poorest treatment was 15.5 kg N/ha, 5.3 kg P/ha and 33.3 kg K/ha, with a mean of 1.7m in growth. The treatment of 15.5 kg N/ha, O kg P/ha and 33.3 kg K/ha resulted in the lowest disease level with a mean of 23.5% leaf area infected (LAI). Treatment with 31 kg N/ha, O kg P/ha and 33.3 kg K/ha resulted in the highest disease level with 39.2% LAI. The results suggested that higher N applications increased disease susceptibility, although this trend was not significant.