Matthias M. Boer

Differential responses of hillslope and channel elements to rainfall events in a semi-arid area

The degree of hydrological connectivity of hillslope elements in a semi-arid climate was studied at the season and event timescales. Field data were obtained in Rambla Honda, a Medalus project field site situated in SE Spain, on micaschist bedrock and with 300 mm annual rainfall. The season timescale was assessed using correlation analysis between soil moisture and topographic indices. The event timescale was studied by a quasi-continuous monitoring of rainfall, soil moisture, runoff and piezometric levels. Results show that widespread transfers of water along the hillslope are unusual because potential conditions for producing overland flow or throughflow are spatially discontinuous and extremely short-lived. During extreme events, runoff coefficients may be locally high (ca. 40% on slope lengths of 10 m), but decrease dramatically at the hillslope scale (<10% on slope lengths of 50 m). Two mechanisms of overland flow generation have been identified: infiltration excess, and local subsurface saturation from upper layers. The former occurs during the initial stages of the event while the latter, which is quantitatively more important, takes place later and requires a certain time structure of rainfall intensities that allow saturation of the topsoil and the subsequent production of runoff. Hillslopes and alluvial fans function as runoff sources and sinks respectively. Permanent aquifers are lacking in Rambla Honda. Variable proportions of hillslope areas may contribute to flash floods in the main channel, but their contribution to the formation of saturated layers within the sediment fill is very limited.

Mapping soil depth classes in dry Mediterranean areas using terrain attributes derived from a digital elevation model

Abstract Modern land management increasingly demands quantitative information on spatially variable soil properties. Traditional soil survey maps do not provide this information. In this paper we report on the application of terrain attributes to mapping soil depth classes at high spatial resolution over large areas under dry Mediterranean conditions. Soil data were collected in 111 georeferenced field plots of 30 m × 30 m, more or less equally distributed over three lithological units, phyllites, shales and limestones. Topographic attributes were computed from a digital elevation model at 30 m resolution. A principal components analysis was carried out on the map overlays of the terrain attributes in order to obtain uncorrelated topographic factors that enabled us to apply a probability approach. Using the maximum likelihood classifier with the field plots as a dispersed training area, predictions were made of mean soil depth class and the probability of occurrence of shallow or deep soils. A cross-validation revealed a 65%, 81% and 61% accuracy for the three maps of the shale area, a 50%, 55% and 40% for the maps of the phyllite area, and a 78%, 72% and 75% accuracy for the maps of the limestone area. Explanations for both the good results in the shale and limestone areas, and the poor results in the phyllite area, focus on the effects of the spatial scale of topographic variation, sediment transport mechanisms, and the impact of land use.

Evaluating the long-term water balance of arid zone stream bed vegetation using evapotranspiration modelling and hillslope runoff measurements

The difference between long-term actual evapotranspiration (AET) and precipitation (P) provides a useful indication of the extent to which a site retains or loses water resources, and therefore of the likely occurrence of specific land degradation processes. Sink areas OAET q PU receive lateral water inputs from other parts of the catchment. In arid and semi-arid environments these areas are frequently found in, or next to, the stream beds of ephemeral rivers and are often characterised by intensive land use or high conservation values. For both types of land use it is important to know if, and how much, AET exceeds P, and where the lateral water inputs come from. Thick sedimentary fills in the stream bed, variable climate conditions and ephemeral flow conditions pose specific difficulties to the evaluation of the water balance of these sites. The objective of this study was to develop an approach to explore the relative importance of lateral water inputs to shrub stands growing in thick sedimentary fills of semi-arid ephemeral rivers. The approach is based on (i) estimating long-term AET2 P balances in the channel sediments and (ii) assessing whether these inflows originate mainly from surrounding hillslopes or from the upstream part of the catchment. A physically based evapotranspiration model for sparse vegetation was used to estimate the long-term AET rates. The relative importance of hillslope runoff and channel flow was evaluated in a semi-quantitative fashion from a combination of surface area estimates and mostly published values of soil hydrological parameters. The approach was developed and tested in a selected stand of Retama sphaerocarpashrubs in a stream bed at the Rambla Honda field site (Tabernas,

Spatial scale invariance of southern Australian forest fires mirrors the scaling behaviour of fire-driving weather events

Power law frequency-size distributions of forest fires have been observed in a range of environments. The scaling behaviour of fires, and more generally of landscape patterns related to recurring disturbance and recovery, have previously been explained in the frameworks of self-organized criticality (SOC) and highly optimized tolerance (HOT). In these frameworks the scaling behaviour of the fires is the global structure that either emerges spontaneously from locally operating processes (SOC) or is the product of a tuning process aimed at optimizing the trade-offs between system yield and tolerance to risks (HOT). Here, we argue that the dominant role of self-organized or optimised fuel patterns in constraining unplanned-fire sizes, implicit in the SOC and HOT frameworks, fails to recognise the strong exogenous controls of fire spread (i.e. by weather, terrain, and suppression) observed in many fire-prone landscapes. Using data from southern Australia we demonstrate that forest fire areas and the magnitudes of corresponding weather events have distributions with closely matching scaling exponents. We conclude that the spatial scale invariance of forest fires may also be a mapping of the meteorological forcing pattern.

Divergent responses of fire to recent warming and drying across south‐eastern Australia

The response of fire to climate change may vary across fuel types characteristic of differing vegetation types (i.e. litter vs. grass). Models of fire under climatic change capture these differing potential responses to varying degrees. Across south-eastern Australia, an elevation in the severity of weather conditions conducive to fire has been measured in recent decades. We examined trends in area burned (1975-2009) to determine if a corresponding increase in fire had occurred across the diverse range of ecosystems found in this part of the continent. We predicted that an increase in fire, due to climatic warming and drying, was more likely to have occurred in moist, temperate forests near the coast than in arid and semiarid woodlands of the interior, due to inherent contrasts in the respective dominant fuel types (woody litter vs. herbaceous fuels). Significant warming (i.e. increased temperature and number of hot days) and drying (i.e. negative precipitation anomaly, number of days with low humidity) occurred across most of the 32 Bioregions examined. The results were mostly consistent with predictions, with an increase in area burned in seven of eight forest Bioregions, whereas area burned either declined (two) or did not change significantly (nine) in drier woodland Bioregions. In 12 woodland Bioregions, data were insufficient for analysis of temporal trends in fire. Increases in fire attributable mostly to warming or drying were confined to three Bioregions. In the remainder, such increases were mostly unrelated to warming or drying trends and therefore may be due to other climate effects not explored (e.g. lightning ignitions) or possible anthropogenic influences. Projections of future fire must therefore not only account for responses of different fuel systems to climatic change but also the wider range of ecological and human effects on interactions between fire and vegetation.

Modelling the potential for prescribed burning to mitigate carbon emissions from wildfires in fire-prone forests of Australia

Prescribed fire can potentially reduce carbon emissions from unplanned fires. This potential will differ among ecosystems owing to inherent differences in the efficacy of prescribed burning in reducing unplanned fire activity (or ‘leverage’, i.e. the reduction in area of unplanned fire per unit area of prescribed fire). In temperate eucalypt forests, prescribed burning leverage is relatively low and potential for mitigation of carbon emissions from unplanned fires via prescribed fire is potentially limited. Simulations of fire regimes accounting for non-linear patterns of fuel dynamics for three fuel types characteristic of eucalypt forests in south-eastern Australia supported this prediction. Estimated mean annual fuel consumption increased with diminishing leverage and increasing rate of prescribed burning, even though average fire intensity (prescribed and unplanned fires combined) decreased. The results indicated that use of prescribed burning in these temperate forests is unlikely to yield a net reduction in carbon emissions. Future increases in burning rates under climate change may increase emissions and reduce carbon sequestration. A more detailed understanding of the efficacy of prescribed burning and dynamics of combustible biomass pools is required to clarify the potential for mitigation of carbon emissions in temperate eucalypt forests and other ecosystems.

Canopy leaf area of a mature evergreen Eucalyptus woodland does not respond to elevated atmospheric [CO2] but tracks water availability.

Canopy leaf area, quantified by the leaf area index (L), is a crucial driver of forest productivity, water use and energy balance. Because L responds to environmental drivers, it can represent an important feedback to climate change, but its responses to rising atmospheric [CO2] and water availability of forests have been poorly quantified. We studied canopy leaf area dynamics for 28 months in a native evergreen Eucalyptus woodland exposed to free-air CO2 enrichment (the EucFACE experiment), in a subtropical climate where water limitation is common. We hypothesized that, because of expected stimulation of productivity and water-use efficiency, L should increase with elevated [CO2]. We estimated L from diffuse canopy transmittance, and measured monthly leaf litter production. Contrary to expectation, L did not respond to elevated [CO2]. We found that L varied between 1.10 and 2.20 across the study period. The dynamics of L showed a quick increase after heavy rainfall and a steady decrease during periods of low rainfall. Leaf litter production was correlated to changes in L, both during periods of decreasing L (when no leaf growth occurred) and during periods of increasing L (active shedding of old foliage when new leaf growth occurred). Leaf lifespan, estimated from mean L and total annual litter production, was up to 2 months longer under elevated [CO2] (1.18 vs. 1.01 years; P = 0.05). Our main finding that L was not responsive to elevated CO2 is consistent with other forest FACE studies, but contrasts with the positive response of L commonly predicted by many ecosystem models.

Woody clockworks: circadian regulation of night‐time water use in Eucalyptus globulus

The role of the circadian clock in controlling the metabolism of entire trees has seldom been considered. We tested whether the clock influences nocturnal whole-tree water use. Whole-tree chambers allowed the control of environmental variables (temperature, relative humidity). Night-time stomatal conductance (gs ) and sap flow (Q) were monitored in 6- to 8-m-tall Eucalyptus globulus trees during nights when environmental variables were kept constant, and also when conditions varied with time. Artificial neural networks were used to quantify the relative importance of circadian regulation of gs and Q. Under a constant environment, gs and Q declined from 0 to 6 h after dusk, but increased from 6 to 12 h after dusk. While the initial decline could be attributed to multiple processes, the subsequent increase is most consistent with circadian regulation of gs and Q. We conclude that endogenous regulation of gs is an important driver of night-time Q under natural environmental variability. The proportion of nocturnal Q variation associated with circadian regulation (23-56%) was comparable to that attributed to vapor pressure deficit variation (25-58%). This study contributes to our understanding of the linkages between molecular and cellular processes related to circadian regulation, and whole-tree processes related to ecosystem gas exchange in the field.

Climatic anomalies drive wildfire occurrence and extent in semi-arid shrublands and woodlands of southwest Australia

Variation in the frequency, extent and intensity of wildfires can drive changes in the composition, structure, diversity and functioning of ecosystems in fire-prone regions worldwide. However, relationships between climatic variation and wildfire occurrence remain poorly understood in many fire- prone regions. We investigated fire occurrence and extent across 15,500 km 2 of semi-arid southwest Australia in relation to inter-annual and/or seasonal variation in regional climate and broad-scale circulation patterns. Superposed epoch analysis (SEA) was used to determine whether wildfire occurrence was related to anomalously high or low regional rainfall or temperature. In particular, we tested if years of minor fire extent (i.e., ,250 km 2 burnt) and major fire extent (i.e., .1,000 km 2 burnt) occurred under different climatic conditions. We also used SEA to determine if wildfires occurred during or following periods of extremes of drivers of regional climate, including the El Nio southern oscillation (ENSO), the Indian Ocean dipole, atmospheric blocking in the adjacent Southern Ocean, and the southern annular mode (SAM). Fire occurrence was linked to dry and hot conditions typically associated with the El Ni˜ no phase of ENSO, with few or no fires in years with cool and wet summers. However, major fire years tended to occur during drought conditions that followed wet and cool conditions in spring and summer of the preceding year. These wet and cool periods were typically associated with the presence of blocking highs in the Southern Ocean to the south of Western Australia. We hypothesise that high rainfall in spring and summer favours the growth of ephemeral plants while subsequent drought conditions promote fuel drying, resulting in more continuous and highly flammable fuel beds capable of sustaining larger fires. Regional climatic patterns are likely driven by interactions among the SAM, atmospheric blocking, and decaying tropical cyclones. As climatic extremes are expected to increase in intensity and frequency in the future, it is likely that the occurrence of extensive wildfires in semi-arid southwest Australia will also increase, potentially driving changes in the distribution and composition of fire-sensitive plant communities.

Assessment of dryland condition using spatial anomalies of vegetation index values

A landscapes long‐term capacity to retain, utilize and recycle local resources is an objective basis for assessing its ecological functionality or condition. In subtropical and tropical drylands, where plant growth is moisture‐limited for much of the time, land condition is reflected in the local water balance. The ratio of long‐term actual evapotranspiration and precipitation (E a/P) is proposed as an objective indicator of dryland condition. A spatial modelling framework is developed for the quantification of E a/P over large areas using remotely sensed vegetation density patterns. Model parameters are defined by two particular situations: (i) non‐vegetated sites, where E a/P depends on the long‐term runoff coefficient of bare soil surfaces (RCbs), and (ii) non‐degraded sites with a vegetation density close to the potential value for which E a/P≅1.0. Spechts evaporative coefficient is used as an independent variable for the prediction of the potential vegetation density, whereas RCbs is estimated with the curve number method. The performance of the method was evaluated in a 900 km2 area in south‐east Spain, where predicted land condition was found to be in good agreement with qualitative field observations on the nature and intensity of land degradation processes.