Deliberate (prescribed) fire management had not previously been used in nature reserve management in New Zealand. An accidental fire burned 70 ha of the 100 ha Flagstaff Scenic Reserve on the outskirts of Dunedin City (southeast South Island), in autumn (March) 1976. Subsequently another 20 ha was deliberately burnt in spring (September) 1976 to assess the potential of prescribed burning to perpetuate narrow-leaved snow tussock Chionochloa rigida grassland.
Four areas were sampled: unburnt tussock grassland/scrub (altitude 500-540 m); spring-burnt grassland (510-600 m); lower autumn-burnt grassland (500-590 m); and upper autumn-burnt grassland (580-668 m; reserve summit).
In 1978, 1980, 1984 and 1988, within each area 10 (8 m-long) transects were established at random. Four 0.5 × 0.5 m quadrats were placed along each transect, within quadrats species shoot frequency was determined in 25, 10 x 10 cm squares.
Average species frequencies were calculated for the four quadrats in each transect. This data was used to develop a computer model to predict vegetation change.
Average maximum leaf lengths of three important perennial, tussock grassland species after 1976 changed as follows:
New Zealand flax Phormium spp. - in 1978 leaves were shorter (significantly so) in autumn-burnt (51 cm) than in spring-burnt areas (60 cm). A significant difference remained in 1980 and 1984, but by 1988 they had converged to around 70 cm.
Chionochloa rigida- in 1978 leaves in autumn-burnt (around 40 cm) were shorter than in spring-burnt areas (55 cm). By 1988 the two treatments had converged (leaves around 75 cm). Shoot frequency was low after fire, but generally increased.
Astelia nervosa - leaves were shorter initially in 1978 in autumn-burnt (around 35 cm) compared to spring-burnt areas (45 cm). By 1988, leaf length had increased, but was still greater in the spring burnt area (60 cm vs. 51cm)
Of numerous non-native species, after both fires common bent Agrostis capillaris frequency increased dramatically (>60% shoot frequency) compared to unburnt controls (20%), but declined more quickly in spring-burnt areas. By 1988, in the unburnt area, a native shrub, manuka Leptospermum scoparium, was replacing Chionochloa. The model predicted medium-term replacement of non-native grassland by Chionochloa tussock grassland, with succession towards Leptospermum scrub after several decades.
Based on these results, the authors recommended that Chionochloa grassland at this altitude should be burnt every 15-40 years to maintain the tussock grassland community. Such a low frequency burning regime would prevent loss of tussock grassland to any high intensity fires that might otherwise occur (i.e. if a high fuel load was allowed to accumulate as a result of no burning management taking place), and retard scrub invasion.
If using or referring to this published study, please read and quote the original paper, this can be viewed at: http://www.sciencedirect.com