School of Agriculture, Food and Ecosystem Sciences - Theses
Permanent URI for this collection
Search Results
1 - 3 of 3
-
ItemEffects of wildfire on forest structure and plant functioning in resprouting forests: implications for catchment water balance( 2013)Globally, forests release large amounts of water that are critically important for urban and industrial water supply. Forests also account for the majority of terrestrial carbon sequestration. In southern Australia, much of the water supply comes from protected catchments vegetated by eucalypt forests. The primary natural disturbance in these forests is wildfire, with close to 3 million hectares burnt over the last decade in the state of Victoria (DSE 2013). Given that forest water-use (the volume of water released to the atmosphere) can change following disturbance, thereby changing the volume of streamflow discharge, understanding the effects of wildfire on forest water-use is vital for water resource planning. Concern around the effects of wildfire on water supply were heightened in 2009 when, following a prolonged drought that necessitated water restrictions, the Black Saturday wildfires burnt through 430 000 ha, including 30% of the water supply catchments for Melbourne, Australia’s second largest city. There are two different ecological responses of eucalypt forests to wildfire. The mixed eucalypt species forests are facultative resprouters, meaning they primarily regenerate vegetatively, with some seedling recruitment. In contrast, ash-type forests are obligate seeders, meaning they are largely killed by fire and regenerate from seed. Despite mixed eucalypt species forests dominating southern Australia’s water catchments, the effect of wildfire on evapotranspiration and streamflow from this forest type is unknown. Research to date has instead been focused on ash-type forests where post-fire evapotranspiration can be up to double that of long unburnt forest, leading to a corresponding reduction in streamflow. The central aim of this thesis is thus to elucidate how post-fire changes in forest structure and plant functioning in resprouting forests affects catchment water balance. Focusing research on the resprouting mixed eucalypt species forests not only addresses this knowledge gap in ecohydrology in Australian catchments, but also provides an opportunity to investigate the effects of altering foliage distribution, while largely maintaining other attributes of forest structure. The overall approach of the thesis is to evaluate the influence of post-fire changes in forest structure and functioning on evapotranspiration and streamflow. This is a process based approach which identifies the mechanisms underlying any observed changes in catchment water balance. This approach is critical for modelling the impacts of topographic driven variability in forest type and variability in fire severity on catchment water balance; and for predicting changes in evapotranspiration under circumstances other than those measured, such as at other locations or under a different climate. Such an approach also provides insights into the functioning of other resprouting vegetation types, which are found across the globe. The research presented in this thesis found that observed post-fire evapotranspiration was a function of both fire severity and landscape position. In forest subject to high intensity wildfire (100% canopy scorch), evapotranspiration was substantially less than in unburnt forest, over 1-3 years post-fire. The magnitude of change in evapotranspiration in forest burnt at moderate severity (<30% canopy scorch), was much less than for high intensity fire. Evapotranspiration was consistently lower in damp forest (located on southern slopes and gullies) than dry forest, although the rate of recovery was similar. These reductions in evapotranspiration in burnt forest were driven by lowered stand-scale transpiration. This was a function of partial tree mortality, 100% shrub mortality, and reduced transpiration within surviving trees. Reductions in stand-scale transpiration were partially offset by increased interception and evapotranspiration close to the forest floor, this in turn was driven by regenerating seedlings which increased the total leaf area of burnt forest. Despite lower transpiration per unit sapwood area in surviving trees, transpiration per unit leaf area was higher compared to unburnt trees. This was related to the post-fire canopy architecture of surviving trees, with more foliage located at lower heights than in unburnt trees. This lower foliage had higher rates of gas exchange, consistent with the hydraulic limitation hypothesis which predicts higher gas exchange in lower foliage due to a shorter hydraulic path length and lesser gravitational force. Recovery of evapotranspiration is predicted to occur within 10-15 years post-fire, over which time net evapotranspiration, when compared with unburnt forest, is predicted to be lower in forest burnt at high severity, but higher in forest burnt at moderate severity. These changes in forest water-use are expected to result in net streamflow increases following severe wildfire, but net decreases following moderate severity fire, subject to soil water storage dynamics. Streamflow observations from a mixed eucalypt species forested catchment burnt at light-moderate severity supports these predictions, with streamflow lower than expected due to fire induced changes in evapotranspiration dynamics over the initial five years post-fire. The research findings presented in this thesis demonstrate the importance of both forest structure and plant functioning in governing catchment water balance. These findings also demonstrate that the ecological response of species to disturbance is a critical factor in determining rates of recovery of carbon and water fluxes.
-
ItemPost-fire debris flows in southeast Australia: initiation, magnitude and landscape controls( 2013)Surface runoff and sediment availability can increase after wildfire, potentially resulting in extreme erosion, flash floods and debris flows. These hydro-geomorphic events supply large amounts of sediment to streams and can represent a hazard to water supply systems, infrastructure and communities. This thesis combines observations, measurements and models to quantify and represent the post-fire processes that result in hazardous catchment responses. The processes that constitute risk to water quality and infrastructure were identified through field surveys of burnt catchments in the eastern upland of Victoria (southeast Australia) where impacts had occurred. The survey established that the majority of high-impact events after wildfire were linked to runoff-generated debris flows, a process previously undocumented in the region. The debris flows were initiated through progressive sediment bulking, and occurred in response to short duration and high intensity rainfall events, within one year after wildfire. Debris flows were confined to dry sclerophyll forests that had been subject to crown fire. Wet forest types displayed comparatively subdued responses, a pattern attributed to the relatively high infiltration capacity in these systems. Infiltration and sediment availability were isolated as the key hillslope components that were sensitive to burning and which strongly influenced catchment processes and debris flow susceptibility. The aims of subsequent work were therefore to develop models of infiltration and sediment availability as controls on hillslope response and use these to quantify changes in key parameters during recovery from wildfire. Infiltration was modelled as function of surface storage (H), matrix flow (Kmat) and macropore flow (Kmac). Macropore flow was found to be the main parameter driving the temporal trends in infiltration capacity during recovery from wildfire. Water repellency was ubiquitous in headwater recovering from wildfire, although the strength diminished during prolonged wet weather conditions, a dependency which could be modelled as a function of monthly weather patterns. Sediment availability was highly variable with soil depth, a feature which contrasts with assumptions underlying commonly used erosion models, typically developed in agricultural systems. The majority of erosion following wildfire was found to occur in a shallow layer of highly erodible material which could be represented through dnc, a parameter describing the depth of non-cohesive soil. This depth of available soil decreased exponentially during recovery. The models of sediment availability and infiltration were effective at capturing both spatial variability and recovery processes and form a basis on which to model debris flow initiation and magnitude in variable landscapes during recovery from wildfire.
-
ItemThe impact of fire disturbance and simulated climate change conditions on soil methane exchange in eucalypt forests of south-eastern Australia( 2013)Soils in temperate forest ecosystems globally act as sources of the greenhouse gas carbon dioxide, and both sinks and sources of the greenhouse gases nitrous oxide and methane (CH4), with well-drained aerated soils being one of the most important sinks for atmospheric CH4. Soil CH4 uptake is driven by aerobic CH4 oxidation through methanotrophic bacteria that oxidize CH4 at atmospheric to sub atmospheric concentrations with soil gas diffusivity being one of the key regulators of soil CH4 uptake in these systems. Climate change predictions for south-eastern Australia indicate a high probability of increasing temperatures, lower average rainfall and an increase in the frequency and severity of droughts and extreme weather events. As a further consequence of climate change in south-eastern Australia, there is a predicted increase in days with high fire risk weather and an increased probability of severe wildfires. In response to these predictions, the use of planned burning as a management strategy within Australian temperate forests and woodlands has increased significantly in an attempt to mitigate this risk of uncontrolled wildfire. Changes in soil moisture regimes, temperature regimes and soil disturbance have the potential to alter soil CH4 uptake, however this has generally been studied in the deciduous and coniferous forests of the northern hemisphere. Currently there is a lack of knowledge regarding temporal and spatial regulators of soil CH4 uptake in temperate Australian forest systems and results from northern hemisphere studies cannot be confidently applied to the eucalyptus dominated Australian forests. Consequently, it is difficult to assess how climate change might affect this important soil based CH4 sink, resulting in significant uncertainty around the magnitude and future trends of the CH4 sink strength of forest soils in south-eastern Australia. To help address this uncertainty, this study investigated both the seasonal drivers of soil CH4 uptake and the sensitivity of soil CH4 uptake to altered soil conditions caused by wildfire, planned burning or simulated climate change scenarios in south-eastern Australian temperate eucalypt forests. This thesis encompasses four field studies: (i) To investigate the possible impacts of the predicted decrease in average rainfall and increase in temperature on soil CH4 uptake we measured soil CH4 flux for 18 months (October 2010 – April 2012) after installing a passive rainfall reduction system to intercept approximately 40% of canopy throughfall (as compared to control plots) in a temperate dry-sclerophyll eucalypt forest in south-eastern Australia. Throughfall reduction caused an average reduction of 15.1 ± 6.4 (SE) % in soil volumetric water content, a reduction of 19.8 ± 6.9 (SE) % in water soil filled pore space (WFPS) and a 20.1 ± 6.8 (SE) % increase in soil air filled porosity (φair ). In response to these changes, soil CH4 uptake increased by 54.7 ± 19.8 (SE) %. Increased temperatures using open top chambers had a negligible effect on CH4 uptake. Relative changes in CH4 uptake related more to relative changes in φair than to relative changes in WFPS indicating a close relationship between φair and soil gas diffusivity. Our data indicated that soil moisture was the dominant regulating factor of seasonality in soil CH4 uptake explaining up to 80% of the seasonal variability and accounting for the observed throughfall reduction treatment effect. This was confirmed by additional soil diffusivity measurements and passive soil warming treatments. We further investigated non-linear functions to describe the relationship between soil moisture and soil CH4 uptake and a log-normal function provided best curve fit. Accordingly, soil CH4 uptake was predicted to be highest at a WFPS of 15%. This is lower than in many other ecosystems, which might reflect a drought tolerant local methanotrophic community. However, the applicability of the log-normal function to model CH4 uptake should be evaluated on global datasets. Soil moisture during our study period rarely fell below 15% WFPS and the observed mean was approximately 40% WFPS. It is therefore likely that soil CH4 uptake will increase if rainfall reduces in the dry-sclerophyll forest zone of Australia as a consequence of climate change. (ii) Planned burning is a management strategy applied in south-eastern Australia that aims to reduce fuel loads and therefore mitigate the risk of large, uncontrolled wildfires. Recent government policy changes have led to a significant increase in the total area of public land subject to planned burning activities within the region. To investigate the impact of fire frequency (as a result of planned burning) on soil CH4 uptake, soil methanotrophic activity and soil CO2 fluxes we measured these three variables in six campaigns across all seasons (March 2009 – February 2011) in a dry sclerophyll eucalypt forest in the Wombat State Forest, Victoria. Three different fire frequency treatments had been applied since 1985: planned burning in autumn i) every 3 years, ii) every 10 years, and iii) not burned since before 1985. Mean soil CO2 emissions were significantly higher in the planned burn treatments compared to the unburnt treatments. In contrast, soil CH4 oxidation did not show the same response to planned burning. Our data indicate that differences in soil CO2 fluxes in response to planned burning might be driven by increased autotrophic root respiration most likely related to decreased nutrient and water availability to overstorey plants. This theory contrasts with alternative explanations that focus on post fire changes in soil nitrogen dynamics, increased heterotrophic respiration and increase soils surface temperatures. Given the long-term nature of the applied burning treatments (implemented for over 25 years) it is therefore unlikely that increases in planned burning will have an impact on the CH4 uptake capacity of these fire resistant eucalypt forests. (iii) Wildfire is the most important disturbance event that alters composition and stand age distribution in forest ecosystems in south-eastern Australia. Wildfire impacts often alter environmental conditions that influence CH4 uptake of forest soils. The impact of wildfire on the CH4 uptake capacity of forest soils is currently unknown. In 2010/2011 we measured soil atmosphere CH4 exchange along a chronosequence in a Tasmanian wet sclerophyll eucalypt forest where the time since the last stand-replacing disturbance ranged between 11 years and approximately 200 years and was due to either wildfire or wildfire emulating harvest operations. Our results indicate an initial increase in soil atmosphere CH4 uptake from the most recently disturbed sites (11 years post-disturbance) to ‘mature’ sites (46 and 78 years post-disturbance). This initial increase was followed by a time-since-last-disturbance (TSLD) related decrease in soil atmosphere CH4 uptake. Our data indicate the initial increase in CH4 uptake is related to a decrease in soil bulk density and an associated increase in soil gas diffusivity. However, the subsequent decline in CH4 uptake with increasing TSLD (from 78 to 200 years) was more likely driven by an increase in soil moisture status and a decrease in soil gas diffusivity. We hypothesize that the observed increase in soil moisture status for the stands aged 78 years and older was driven by forest succession related changes in soil organic matter quality/quantity, an increase in throughfall and an overall decrease in stand water use as demonstrated for tall mixed wet sclerophyll eucalyptus forests elsewhere. (iv) In order to gain a better understanding of seasonal and inter-annual variation in soil CH4 exchange for temperate eucalypt forests in south-eastern Australia, we measured soil CH4 exchange in high temporal resolution (every 4 hours or less) over two consecutive years (March 2010 – March 2012) in the Wombat State Forest, Victoria and over one year (October 2010 – February 2012) at the Warra, Tasmania. These two sites are both temperate Eucalyptus obliqua (L. Her) dominated forest systems however they have contrasting annual precipitations (Victoria Site= 870 mm yr-1, Tasmania Site = 1700 mm yr-1). Both systems were continuous CH4 sinks with the Victorian site having a sink strength of -1.79 kg CH4 ha-1 yr-1 and the Tasmanian site having a sink strength of -3.83 kg CH4 ha-1 yr-1 in 2011. Our results show that CH4 uptake was strongly regulated by soil moisture with uptake rates increasing when soil moisture decreased, which explained up to 90% of the temporal variability in CH4 uptake at both sites. Furthermore, when soil moisture was expressed as soil air filled porosity (φair) we were able to predict the CH4 uptake of one site by the linear regression between φair and CH4 uptake from the other site, indicating a generic relationship. Soil temperature only had an apparent control over seasonal variation in CH4 uptake during periods when soil moisture and soil temperature were closely correlated. The natural fluctuation in generally low soil nitrogen levels did not influence soil CH4 uptake at either site. Comparing our measured site data to modelled data utilising a process based methane uptake model (Curry 2007), our two sites showed reasonable agreement providing scaling factors used to account for soil temperature (rT) response and moisture response (rSM) of methane oxidation rate (k) were forced to unity. Under these conditions CH4 uptake was primarily regulated by diffusivity in the model, indicating that observed seasonal variability in soil CH4 uptake at both sites was primarily regulated by soil moisture related changes in soil gas diffusivity. This study filled some important knowledge gaps with regards to information about magnitude and controls of temporal variability but also with regards to climate changes sensitivity of soil CH4 uptake in temperate eucalypt forests in south-eastern Australia and provides important datasets that will enable better predictive modelling of changes in soil CH4 uptake across the temperate forest landscape in south-eastern Australia. The results indicate it is likely that soil CH4 uptake will increase if rainfall reduces in the dry-sclerophyll forests of Australia as a consequence of climate change. Our findings on the impact of wildfire on soil CH4 exchange highlight the potentially large spatial variability in CH4 uptake across the landscape within the same forest and soil type, a factor that would need to be accounted for in global CH4 uptake models. This issue could be partially addressed for tall wet temperate eucalypt forests in case the here theorized relationship between forest succession and CH4 uptake can be verified in further studies.The finding that low intensity planned burning does not have an effect on soil CH4 uptake suggests that fire may need to be of a particular severity before changes in soil properties and the associated changes in soil CH4 uptake can be observed. Our long term monitoring results further highlight the importance of long-term field measurements in establishing relationships between soil environmental drivers and soil CH4 uptake and are therefore useful for the calibration of models that calculate the global CH4 sink distribution and magnitude.