Breeding systems of some cold tolerant eucalyptus species.

Abstract

Seasonal flowering times for Eucalyptus nitens, E. dunnii, E. smithii, E. macarthurii and E. grandis were evaluated in clonal grafted orchards located at the Shaw Research Centre (SRC) in KwaZulu-Natal, South Africa. The orchards are situated at 29° 29 'South, 30° 11 'East at 1100 m above sea level. The climate is cool (MAT 16.7° C) with a January mean monthly maximum of 25.8° C and July minimum of 4.4° C. An estimated mean annual rainfall of 998 mm and median annual rainfall of 899 mm has been reported (PALLETT and MITCHELL 1993). It is evident that the different species flower consistently from one year to the next during the same period with similar mean flowering peaks. Long reproductive sequences where identified for all species relative to E. grandis, particularly E. smithii and E. dunnii. Paclobutrazol was used to initiate flowering to facilitate the study of the breeding systems of the different species. When applied as a soil drench during early summer an increase in the flower bud production in E. nitens, E. smithii and E. grandis was achieved. The use of various cytochemical methods to test pollen viability, were shown to be mere indicators of potential viability and lack the reliability for adequate testing of stored pollen. From the range of in vitro, pollen viability studies the most successful media for all species tested was 30 % sucrose with 150 mg r¯¹ boric acid. Without boric acid in the media, the response after 24 h was significantly poorer (p<0.001). Significant differences (p<0.05) in the area of pollen grains were found between and within species. There was no significant difference between E. dunnii and E. macarthurii at the species level. Pollen of E. smithii, E. grandis and E. nitens were significantly smaller than that of both E. dunnii and E. macarthurii. From isolation experiments which limited potential pollinators it is apparent that a reduction of pollinators not only leads to poorer capsule survival but also poorer seed set. Following an initial survey of pollinators of E. grandis, very few insects were recorded relative to surveys conducted in the natural habitats with indications that an association does exist between the presence of active pollinators and temperature. The potential of flowers to set seed is clearly demonstrated by the difference between open pollinated flowers and controlled pollinated flowers following intraspecific crosses where differences in seed yield per capsule are very often more than double for species such as E. nitens and E. macarthurii. Similarly with interspecific crosses, higher seed yields are extracted from crosses between closely related species. An extensive survey of orchards clearly demonstrates that E. nitens has the lowest clean seed recovery (13.8 %) significantly less than that of E. smithii (18.0 %) and both E. macarthurii and E. dunnii at 26.1 % and 26.0 % respectively.