Provenance variation in Podocarpus totara (D. Don): Growth, tree form and wood density on a coastal site in the north of the natural range, New Zealand

Podocarpus totara (D. Don) (totara), a highly valued, naturally-durable softwood tree species, is widely distributed in indigenous lowland forest throughout New Zealand. As harvesting from natural forests has declined there has been increasing interest in planting and managing P. totara for timber production. Seed from 36 provenances was collected from throughout the species’ natural range (latitude 35°21′–45°52′) in 1985, and a resulting field trial planted on an open, frost-free coastal site near Auckland at latitude 36° was assessed for growth and tree form in 1994 and 1999. Survival of all but one provenance exceeded 80%, with an overall average height of 5.5 m and diameter at breast height (DBH) of 8.6 cm.Provenance mean height and DBH varied significantly from 4.8–6.5 m and 6.9–11.2 cm, respectively, 11 years after planting. Height was correlated with latitude and mean summer temperature of site of origin (r = −0.58, 0.54), with a similar trend apparent at the end of the nursery phase three years after sowing of seed. Stem form, assessed on height to first double leader, was also negatively correlated (r = −0.49) with latitude, although less closely than height. The seedlots from northern latitudes tended to grow faster and have better stem form than seedlots from southern latitudes, suggesting that genetic factors related to seed source are influencing growth of P. totara 11 years after planting. In contrast, there was no correlation of provenance mean heights with altitude which may reflect the fact that seed dispersal of P. totara is by birds which can spread seed over wide altitudinal ranges within relatively short distances. However, the wide latitudinal and thus climatic range has resulted in some genetic differentiation of local populations.The study indicates that there is scope for selection for both growth rate and stem form of P. totara for production of improved seed for planting. There were no significant correlations found between wood density sampled in 1999, with geographical and growth variables.