Background

Wildfires are a common phenomenon in Mediterranean regions, partly due to poor land-use planning and management practices. In an average year in Portugal, circa 100.000 ha of woodlands are consumed by wildfire. This figure is not expected to decrease noticeably in the foreseeable future, also because climate change scenarios predict an increase in meteorological conditions propitious for fires. These factors underscore the importance of assessing the direct and indirect consequences of wildfires. Amongst the latter, fireinduced runoff generation and associated soil erosion is widely recognized to endanger on-site ecosystem functioning and sustainability as well as to increase off-site risks such as silting-up and eutrophication of downstream dams. At the same time, however, there still exist important knowledge gaps on the hydrological and erosion effects of wildfires. The gaps with respect to soil losses, catchment-scale processes and their modeling are being addressed by the EROSFIRE-I/II projects (http://www.erosfire.org) but the - even greater - gaps regarding organic matter, carbon and nutrient exports could only be touched upon in these projects, especially due to the implied human and laboratory analysis resources.

Aims

The main goal here is to further the knowledge of direct as well as indirect wildfire effects on topsoil organic matter, carbon and key nutrient (N, P, K) contents, and, thereby, to lay the foundations for the respective transport modules in the model-based tool that the EROSFIRE projects are developing for erosion (then, "land degradation") risk assessment following wildfires.

Approach

The proposed work is structured according to the following research questions and corresponding Tasks:

A. direct, immediate wildfire effects

1) how does fire-induced combustion and heating change organic matter, carbon and key nutrient contents and composition at and below the soil surface (Task2)?

B. indirect wildfire effects

2) how does fire occurrence and severity modify seasonal patterns in topsoil organic matter, carbon and key nutrient contents and composition (Task3)?

3) how does fire occurrence and severity alter possible losses of topsoil organic matter, carbon and key nutrients (Task5) through runoff and associated soil particle transport (Task4)?

4) how well can existing runoff-erosion models and, in particular, those being tested in the EROSFIRE-projects (but with additional, flux-specific transport modules), represent the runoff and associated exports observed in Tasks 4 and 5 (Task6)?

Since it is difficult (if not impossible) to foresee where exactly wildfires will occur during the next summer(s) and, thus, to study the pre-fire situation, this proposal envisages an experimental set-up that is based on concurrent paired observations rather than in-situ pre- and post-fire measurements. In other words, possible fire-induced changes in Tasks2 to 5 will be inferred from comparison of neighboring, presumed identical sites within and close to the burnt area. Since direct and indirect effects may well depend markedly on fire severity, here the paired observations will involve neighboring triplets of high- and medium/low-severity, burnt and (long-)unburned sites. At the catchment-scale, however, this set-up will not be implemented (for budgetary reasons) but longer-term wildfire effects at this scale will still be addressed, albeit less directly. Background runoff loads of organic matter, etc. will be estimated at existing stations based on continuous readings (sensor be purchased) and incidental samples (existing sampler).

Further important methodological aspects are that:

i) sampling is limited to a single land-use/cover, especially to adequately address temporal, spatial and soil-depth variability patterns;

ii) soil and runoff sampling occurs at 1- to 2-monthly/-weekly intervals, respectively;

iii) soil and overland flow sampling concerns the same sites, permitting a grey-box assessment of monthly to annual carbon and nutrient balances;

iv) standard laboratory methods are used for sediment, organic matter, carbon and N, P, K contents and less commonly-used methods for soil organic matter composition;

v) two contrasting but complementary models are applied for event-wise peak exports (MEFIDIS) and longer-term on-site losses (MMF).

CESAM members on this project

Ana Isabel Oliveira Machado
project grant holder

Dalila do Rosário Encarnação Serpa
project grant holder

Researcher

Maria Isabel da Silva Nunes
Researcher

Mário M. A. Cerqueira
Researcher

Martinho António Santos Martins
project grant holder

Valdemar Esteves
Researcher