Williams, P.R. (2002) The effect of fire regime on tropical savannas of north-eastern Australia: interpreting floristic patterns through critical life events. PhD thesis, James Cook University.

Preview

PDF Download (4MB)

Abstract

The degree to which fire influences plant species composition and abundance (referred to in this thesis as floristic patterns) in Australian tropical savannas has been the subject of debate. Available information emphasises floristic patterns rather than the mechanisms that determine patterns. The assessment of critical events in a plant’s life cycle (referred to in this thesis as critical life events) is necessary to understand population dynamics and therefore the processes that determine floristic patterns. Fire may affect population size through its influence on one or more critical life events including seed dormancy release, seed bank dynamics, seedling emergence, seedling survival and the survival and growth of juvenile and mature plants.

This thesis presents data on floristic patterns in response to fire regimes in savannas of north-eastern Australia, a region with limited quantitative information on fire response at a community level; and contributes to the broader understanding of savanna function, through an examination of life cycle processes. The aims of the research were to examine the effect of fire regime on:

a) floristic patterns in eucalypt savannas in the vicinity of Townsville, north-eastern Australia; and b) critical life events, to investigate mechanisms determining floristic patterns.

To achieve these objectives, laboratory and shadehouse experiments were carried out and surveys of seedling emergence and vegetative post-fire regeneration were undertaken at four sites in the Townsville region. At one of these sites, Cape Cleveland, floristic patterns and critical life events were studied in detail using a manipulative fire experiment which examined three fire regimes between 1997 and 2001:

1. Control: savanna burnt in the mid dry season (July) 1997 only. 2. Early burnt: savanna burnt in the mid dry season 1997 and early dry season (May) 1999. 3. Late burnt: savanna burnt in the mid dry season 1997 and late dry season (Oct) 1999. Five annual surveys of permanent plots at Cape Cleveland demonstrated a high degree of stability in species composition, however the abundance of several species was affected by fire regime. A significant increase in the abundance of various subshrubs, ephemeral and twining perennial forbs, and grasses was detected in the first year after fire, particularly following late dry season burning. The abundance of these species declined towards pre-fire levels in the second year after fire. The dominant grass Heteropogon triticeus significantly declined in abundance, and tree density increased, with fire intervals of four years.

Investigations of critical life events detected mechanisms that determine floristic patterns. A pulse of seedling emergence of subshrubs and herbaceous species occurred in the post-fire wet season, with greatest seedling density following late dry season fires. This pulse of seedling emergence explained the increase in plant abundance documented in the year following fire. It resulted from seed dormancy release, triggered by multiple fire-related agents, including the removal of herbaceous cover, which increased germination of a range of species; and the exposure to: heat-shock, which increased the germination of a range of forbs; smoke, which promoted the germination of upright perennial forbs and grasses; and nitrate, which interacted with heat shock to increase germination of at least two forbs.

The germinable soil seed bank was dominated by grasses and forbs, and seed banks of trees and shrubs were scarce and short-lived. Seasonal fluctuations were evident in the seed bank, which increased through the dry season to a maximum density in the late dry season, and declined to a minimum by the mid wet season. The high intensity late dry season fires reduced seed production by the dominant tree Corymbia clarksoniana in the year following burning, however the effect of fire on soil seed bank dynamics was limited due to the annual wet season decline in seed reserves that occurred irrespective of fire. These data indicate that late dry season fires promoted the emergence of a greater density of herb seedlings, compared with early dry season fires, by stimulating seed dormancy release in a higher proportion of seeds, and because more soil seed reserves were available in the late dry season to be affected by fire.

Seedling survival of herbaceous species was higher for those that emerged from recently burnt rather than unburnt savanna. High rates of seedling mortality of perennial species, and the short lifespan of ephemeral forbs, explained the decline in plant abundance documented in the second year following fire. A moderate proportion of seedlings of common sprouter species developed the capacity to survive fire within one year of germination. Only five percent of C. clarksoniana seedlings survived in unburnt, and fewer in burnt savanna, three years after germination.

In savanna that remained unburnt for four years, growth of juvenile trees above 2 m in height accounted for the observed increase in tree density. The capacity of the majority of species to sprout after fire, and rapid seed production of non-sprouter species, explained the stability in species composition documented in the annual floristic surveys at Cape Cleveland. Fire-promoted stem reproduction via root-suckering explained the increase in abundance detected in the subshrub Breynia oblongifolia following burning.

This thesis provides important information on the processes that drive tropical savanna function. It indicates that seedling recruitment is a process that limits population growth. Seedling recruitment of trees and shrubs is scarce and episodic, apparently linked to high rainfall years. The seedling recruitment of subshrubs and herbaceous species is driven by opportunities for seed dormancy release, which are primarily restricted to fire events, and limited by the low to moderate soil seed bank density and poor levels of seedling survival. Growth of juvenile trees beyond 2 m in height, and hence to reproductive maturity, requires fire-free intervals of at least four years and is therefore an event that is restricted by fire and limits population growth. The high proportion of sprouting species enables plant persistence, so that juvenile and mature plant survival are life events that provide population resilience. For ephemeral species, the soil seed bank provides a storage mechanism that provides population resilience.

This thesis provides practical information to improve the management of this widespread and important ecosystem. Conservation management of these savannas will need to balance the role of regular fires in maintaining the diversity of herbaceous species with the requirement of fire intervals of at least four years for allowing recruitment, and therefore population maintenance, of trees. While late dry season fires may cause some tree mortality, the use of occasional late fires will promote dense seedling recruitment of herbaceous species, which may be necessary to maintain sustainable populations of many grasses and forbs.

Item ID:	205
Item Type:	Thesis (PhD)
Keywords:	Fire regime, Eucalypts, Legumes, Germination, Northern Australia, Savanna, Seed bank
Date Deposited:	20 Sep 2006
FoR Codes:	06 BIOLOGICAL SCIENCES > 0602 Ecology > 060208 Terrestrial Ecology @ 50% 06 BIOLOGICAL SCIENCES > 0607 Plant Biology > 060703 Plant Developmental and Reproductive Biology @ 50%
SEO Codes:	96 ENVIRONMENT > 9610 Natural Hazards > 961004 Natural Hazards in Forest and Woodlands Environments @ 100%
Downloads:	Total: 1540 Last 12 Months: 15
	More Statistics