Tree topkill, not mortality, governs the dynamics of savanna-forest boundaries under frequent fire in central Brazil

Tropical savanna and forest are recognized to represent alternate stable states, primarily determined by feedbacks with fire. Vegetation-fire dynamics in each of these vegetation types are largely determined by the influence of the vegetation on fire behavior, as well as the effects of fire behavior...

Full description

Saved in:

Bibliographic Details
Published in:	Ecology (Durham) Vol. 90; no. 5; pp. 1326 - 1337
Main Authors:	Hoffmann, William A, Adasme, Ryan, Haridasan, M, Carvalho, Marina T. de, Geiger, Erika L, Pereira, Mireia A.B, Gotsch, Sybil G, Franco, Augusto C
Format:	Journal Article
Language:	English
Published:	Washington, DC Ecological Society of America 01-05-2009
Subjects:	alternate stable states Animal and plant ecology Animal, plant and microbial ecology bark Biological and medical sciences Brazil calcium Cerrado Ecosystem evergreen forest fire ecology fire intensity fire regime Fires Forest & brush fires Forest and land fires Forest ecology forest ecosystems Forest fires Forest soils Forest trees Forests Fundamental and applied biological sciences. Psychology gallery forest General aspects Grasses Grasslands hysteresis magnesium Mortality Phytopathology. Animal pests. Plant and forest protection Plant Bark Plant Stems Population Dynamics positive feedback riparian forest Riparian forests savanna Savanna soils Savannas soil nutrients topkill tree growth tree mortality Trees Trees - physiology Tropical forests Vegetation Weather damages. Fires wildfires Brazil South America America Savannah Mortality Tropical zone topkill riparian forest Evergreen plant positive feedback Fire ecology Biome Feedback Dynamics Riparian Gallery forest Fires alternate stable states savanna Evergreen forest Brazil Tree Cerrado Hysteresis
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Tropical savanna and forest are recognized to represent alternate stable states, primarily determined by feedbacks with fire. Vegetation-fire dynamics in each of these vegetation types are largely determined by the influence of the vegetation on fire behavior, as well as the effects of fire behavior on tree mortality, topkill (defined here as complete death of the aerial biomass, regardless of whether the plant recovers by resprouting), and rate of growth of resprouts. We studied the effect of fire on three savanna-forest boundaries in central Brazil. Fire intensity was greater in savanna than forest, as inferred by a twofold greater height of stem charring. Despite lower fire intensity, forest tree species exhibited higher rates of topkill, which was best explained by their thinner bark, relative to savanna species. Following topkill, there was no tendency for sprouts of savanna trees to grow faster than those of forest species, contrary to expectations, nor was whole-plant mortality higher in forest than in savanna. This contrasts with observations of high rates of postburn mortality in many other tropical forests. The low tree mortality in these transitional forests suggests that the dynamic of these natural savanna-forest boundaries is fundamentally different from that of forest boundaries originating from deforestation in the humid tropics. The forests studied here appear to be much more resilient to occasional incursion of fire from the savanna, despite being unable to invade frequently burned savanna. The thin bark of forest species makes them particularly susceptible to the "fire trap," whereby repeated topkill of small trees prevents recruitment into adult size classes. Rapid growth will be particularly important for forest species to escape the fire trap, so we predict that, where fire is frequent, forests should be restricted to high-resource sites. Here, Mg²⁺ and Ca²⁺ concentrations had particularly strong effects on postburn growth rates, suggesting that these elements may most strongly limit the distribution of forest in these fire-prone savannas.
Bibliography:	http://dx.doi.org/10.1890/08-0741.1 Corresponding Editor: T. P. Young. ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23
ISSN:	0012-9658 1939-9170
DOI:	10.1890/08-0741.1