Our research focuses on the population dynamics of plants and how they are influenced by impacts of natural disturbances and global environmental change. We are particularly interested in the interactive effects of fire, grazing and drought in grasslands and woodlands in southern Australia, and how climate change, fragmentation and shrub encroachment affect ecosystems.

Monday, 27 June 2011

Why ecologists should know a little about geomorphology

I like rocks almost as much as I like plants!! There, I've said it!

And I particularly like the processes that give rise to the landforms that we see - geomorphology. Being able to read a landscape - such as the rock types and the way that they affect topography and drainage, as well as their impact on base level nutrient availability - is one of the most important skills that a plant ecologist can acquire. Indeed, an understanding of geomorphology broadly helps explain the distribution of native vegetation types in southern Australia.

Native grasslands, for example, occur predominantly on plains of low elevation, both in northern and southern Victoria. One might logically think that there is a similar underlying reason for their distribution and the absence of trees. Nothing could be further from the truth! These land surfaces have very different geological and geomorphological histories that have shaped these systems in different ways.

In western Victoria, volcanoes (more than 350 of them) have spewed out lava over the last 20,000 yrs to 5 M yrs, producing the third largest larva plain in the world, exceeded only by the Deccan in western India, and the Snake River Plateau in the United States! The volcanic activity was probably similar to that now active in Hawaii, with the dominant volcanic product being fluid basalt lava with only a small component of pyroclastic material (mainly scoria). Lavas of this type can spread rapidly across the landscape, and in places extend over 50 km from the volcano.

Lava flows must have produced an initially barren surface that required extensive denudation (i.e. modification by weathering) to be a suitable plant habitat, with primary succession proceeding from species derived in the surrounding landscape - this possibly explains why the western plains flora consists of many generalist species, and few endemics have evolved in the relatively short timeframes since volcanism. The soils that developed in situ are fine-textured cracking clays and are very nutrient-rich (indeed they are amongst the most productive in Australia; it also probably explains why weed invasions are so pronounced here too). As a consequence, trees are restricted to stony rises and cinder cones where drainage is best and soil cracking least.

Mt Elephant, as seen from Dundonnell, is one of the larger of the volcanoes found
 on the western plains. It is an example of a steep-sided scoria volcano - true 'fire mountains'
when they erupted. They formed when magma interacted explosively with groundwater,
blasting molten rock high into the air. The ejected material cooled before it hit the ground,
forming fragments of frothy red or black rock called scoria. These fragments quickly
settled around the vent, building cones with deep central craters. (Photo: John Morgan)

The Volcanic Plains of western Victoria - this geological map is a
good approximation of the distribution of the native grasslands
(source: http://home.iprimus.com.au/foo7/volcmap.html)
By contrast, the vast native grasslands of northern Victoria are the product of an entirely different land forming process. Here, the landscapes were formed by river flooding spreading coarse alluvium. The alluvial plains are built of sediment derived from the erodible sandstones, mudrocks and igneous rocks of the Victorian Highlands and spread by rivers down the mountain flanks. These sedimentary surfaces are quite unlike the volcanic surfaces of western Victoria and formed under a completely different geological regime. These sediments, deposited and redistributed by rivers and wind, buried the older bedrock surfaces and produced a complex landscape of low relief and gentle slope. Like the volcanic plain, it is a mosaic of materials, ages and forms. And like the volcanic plains, the soils are fine-textured clays that easily waterlog in winter, preventing the growth of trees.

Riverine plains grassland - formed by alluvial processes.
(Photo: Eris O'Brien)

Grasslands also occur up in the highlands on mountain summits, plateaus and high plains, but these are not generally not due to the underlying rocks. Rather, they occur where low temperature or cold air drainage - the so called 'frost hollows' - suppresses tree growth. See Wearne & Morgan for a description of these interesting grasslands in the Mt Hotham region. At the moment, that low temperature envelope is very narrow and occurs generally above 1600 m, but in colder Pleistocene times the tree line may have been as low as 1000 m.
So, I've provided a simple example of how land forming processes are responsible for the landscapes we see today. Unfortunately, there's no textbook that adequately introduces geomorphology of Victoria that I can recommend. Rather, you'll need to observe the landscape and ask: what are the rocks that underlay this area? when did this occur? how have aeolian and fluvial processes shaped the landform? Are there obvious associations of vegetation when the above change?

Friday, 17 June 2011

Experimenting with Fire

My PhD primarily revolved around the effects of fire frequency on regeneration dynamics and species coexistence in the endangered temperate grasslands of western Victoria. Here, fire plays an indirect role - frequent fire prevents competitive exclusion of the intertussock forbs from the dominant C4 tussock grasses. In this case, it was the frequency of fire, not the type of fire that seemed most important to the conservation of plant diversity.

In 2003, landscape-scale fires burnt the alpine
vegetation of Victoria. But it was clearly very patchy.
(Photo: John Morgan)
But fires can come in many guises - fires ain't fires - you only need to look at a wildfire to see that it can burn thoroughly or patchily. It's clear to me, however, that we don't understand very much about plant community responses to difference in fire "type". Rather, much of our knowledge (and research) is from the standpoint of the time-since-last fire and, perhaps, the fire frequency (and these assume that fires are much the same). Yet, it is the type of fire that might ultimately affect mortality of established plants, germination cues, and resource levels. Not to mention how much C is returned to the atmosphere.

To learn more about fire and how to measure it, I've just spent a week burning tropical savanna in the Northern Territory Wildlife Park with Dick Williams from CSIRO. This was excellent fun, but also highly informative. The Burning for Biodiversity experiment is an amazing field study examining the effect of fire frequency and timing on a variety of taxa and the dynamics of carbon. Importantly, it relates these outcomes to aspects of fire behaviour. So, what better way to learn about fires than to visit one of the few experiments in Australia that is quantifying fire!

A common measure of fire behaviour is fire intensity – the amount of energy released per unit length of fire front (kW m-1). It is defined as the product of rate of spread (ROS), fuel load, and the heat released from the fuel during combustion. The higher the fuel load and the ROS, the higher the fire intensity.

Fuel loads are easy to calculate - the amount of fine (<6 mm diameter) fuel is sampled in quadrats pre-fire and weighed. It is the fine fuel that will rapidly combust (flamming combustion) and affect properties of the fire front. Larger diameter fuels burn more slowly (in a process called smouldering combustion), typically after the fire front has passed. These fuels are important to quantify, as they will release much more C into the atmosphere.

Harvesting fuels prior to ignition
(Photo: James Camac)
Savanna in the Northern Territory - awaiting burning
(Photo: John Morgan)

ROS is a little harder to quantify because it is much more dynamic, but plays a critical role on fire intensity. In Darwin, we used two techniques to estimate ROS from our contolled burns. First, we measured the time the fire front takes to reach pre-defined points in the landscape - using points marked with numbered poles (we used six) and a stopwatch, the average rate of spread of the fire between the points can be calculated. As a  backup to the estimates by eye (it can get quite hairy when the fire front is moving quickly), we also used specially designed automatic timers buried in the soil with a small thermocouple left exposed above-ground - these timers record the time at which the thermocouple heats to >200 deg C and, somewhat ingeniously, the residence time (the time that the temperature stayed above 200 deg C).

Lighting the fireline with a drip torch
(Photo: James Camac)

Timers, attached to thermocouples, are buried in the soil.
These record the time at which fire passes, and how long
flaming continues at the point. (Photo: James Camac)

Flamming combustion
(Photo: John Morgan)

Smouldering combustion
(Photo: John Morgan)

These are very simple measures that can address fundamental research questions - they should be in the toolbox of all fire ecologists because, if measured, they allow quantification of how fire intensity might affect biodiversity.

We certainly saw large differences in fire intensity in hectare-scale plots burnt on the same day! I look forward to coming back to Darwin in 2012 to observe just why these differences might matter. At this point, I'm not sure when, where, and why fire intensity affects biodiversity - there are simply too few examples in the literature to provide a coherent review. But I know that in future, I'll be quantifying fire intensity in my research (both within and across different fires) to get a better understanding of this primary aspect of fire behaviour.