Post-fire management: effects and new tools application
- Autores: Javier González Romero
- Directores de la Tesis: Jorge de las Heras Ibáñez (dir. tes.), Manuel Esteban Lucas Borja (codir. tes.)
- Lectura: En la Universidad de Castilla-La Mancha ( España ) en 2021
- Idioma: inglés
- Tribunal Calificador de la Tesis: Joan Josep Estrany Bertos (presid.), Pablo Ferrandis Gotor (secret.), Demetrio Antonio Zema (voc.)
- Programa de doctorado: Programa de Doctorado en Ciencias Agrarias y Ambientales por la Universidad de Castilla-La Mancha
- Materias:
- Enlaces
- Tesis en acceso abierto en: RUIdeRA
- Resumen
Wildfires have historically been a natural alteration in Mediterranean ecosystems. Despite Mediterranean ecosystems' high resilience, the expected climate change scenarios, may lead into more recurrent and severe wildfires, and erosion and degradation processes may be enhanced. Under these circumstances an effective and well-planned post-fire management seems crucial to mitigate wildfire damages and preserve ecosystems’ functions in the future. Post-fire management strategies must be founded in a thorough preliminary damage assessment to clearly identify restoration priority areas. To properly understand the multiple interactions of the different measures with the ecosystem and to establish solid criterions for future mitigation plans, these measures effects must be thoroughly evaluated. To do that initial evaluation, and to evaluate the effects of the measures, the application of new management tools based on the emergent concept of sediment connectivity (SC) may be of great interest.
The objectives of this PhD thesis are two: To evaluate the use of SC indices to optimize the location of restoration measures and evaluate their effects in Mediterranean catchments; To assess the local effect of the construction of post-fire mitigation measures (check-dams) on Mediterranean ephemeral streams. To achieve these goals, five studies have been conducted and are part of this dissertation.
The study I evaluates the ability of four SC indices (IC-Borselli, IC-Cavalli, IC-Persichillo and AIC) to assess SC changes in three Mediterranean headwater sub-catchments affected by a wildfire in 2012. To do so, three temporal scenarios were considered, pre-fire (before fire), Fire (immediately after fire) and post-fire (two years after the fire). In addition, two computation targets were chosen, the stream network (Hillslope-channel SC; SCOUTLET) and the outlet of each sub-catchment (Hillslope-outlet SC; SCSTREAM). Burn severity classes were classified based on dNBR index values. The obtained results were analyzed according to geomorphic (landforms), mathematic (significance, percentile, and frequency distribution), ecological (burn severity) and sedimentological (measured specific sediment yield – SSY) criteria.
The final output of the computation consisted in a total of 66 SC maps with a spatial resolution of 2x2 m. The IC-Borselli and AIC were the most responsive approaches to the effects of wildfire on SC at catchment scale, whereas the IC-Persichillo was the most sensitive index to the increasing burn severities. The overlap between the fire severities and the geomorphic features (landforms) was a key factor to understand the hydrological response at both the stream-network and outlet targets. We found a good and positive agreement between the measured SSY in the three check-dams and the changes in the estimated SCOUTLET due to the fire, especially with the IC-Borselli and AIC. For a better implementation of post-fire programs, the use of the AIC index targeting the outlet to assess sediment transport in streams, which is dominated by the deposition process. The use of the same index but targeting the stream may be recommended to place sediment control measures at hillslopes for intense rainfall events when effective sediment transport happens.
Considering the results of the study I, the “aggregated index of connectivity” (AIC) was selected to assess the effect of post-fire management on SC in five Mediterranean catchments located in the SE of Spain, affected by the same wildfire in 2012. The studies II and III used the AIC index and a similar methodology but in the study II, outlets were selected as computation target, with the aim of evaluating the medium- and long-term sediment connectivity processes where sediment accumulation and transport take place also in the channels. On the other hand, in study III the stream system of each catchment was selected as computation target, to obtain the spatial patterns and values of hillslope–channel SC (SCSTREAM) that is the predominant process of sediment delivery during short and intense episodes defined as those within which all eroded soils are transported out the sub-basins without deposition during a hydrological event. For this reason, despite following similar methodologies with regard experimental design and index computation inputs, the results must be interpreted in separated ways.
For the two studies, two temporal scenarios (pre-man: immediately after the fire and before the implementation of the post-fire practices; and post-man; two years after the fire including all post-fire management measures). The index was computed under two different approaches, a structural one (without considering the effect of the rainfall on sediment dynamics) and a functional (considering the rainfall of each temporal scenario). For each study, a total of 20 SC maps were generated and were normalized to make the results comparable.
In the study II the five catchment outlets, were established as the computation target output analysis included the comparison of the different SC distribution along the catchments, the specific effect of the hillslope and channel measures on SC, and a sedimentological analysis using observed specific sediment yield (SSY). The five catchment outlets, which drain into the rivers Segura and Mundo, were established as the computation target, and two sediment dynamics were calculated: Structural and functional SC.
The results of this study reveal that, at the catchment scale, despite the normalization the outlet position of the catchments as an important factor as well as catchment morphology and slope gradients. The structural SC maps allowed evaluating the actual role played by the post-fire practices removing the rainfall influence, while functional SC was linked to the actual sedimentary response. Hillslope treatments resulted in significant structural changes on SC at local scale. However, the DEM resolution was not enough to reflect the effect of check-dams on SC paths through the channel network. A good and positive correlation was found between the SSY and the changes of SC due to the post-fire practices and vegetation recovery.
In the study III, to evaluate the effect of intense rainfall-runoff events, the stream system of the five catchments was set as the computation target (predominance of effective sediment delivery outside the catchments). The results of this study reveal that the stream density and slope variables were the most influential factors on SC at the sub-catchment scale. Hillslope treatments, (barriers (when built in high densities) and afforestation) significantly reduced SC in comparison with untreated areas in both structural and functional approaches. Under more erosive rainfall conditions at the post-man scenario (functional approach) hillslope treatments were only able to reduce the increase of SC between scenarios. The manual inclusion of check-dam disconnectivity allowed to overcome the lack of resolution of the utilized DEM and reflect their effect in the simulation. A positive correlation was found between the measured SSY and the AICN-STR changes due to the post-fire practices and vegetation recovery.
The studies IV and V were focused on the assessment of the local effect of post-fire check-dam construction on ephemeral streams' soil functionality (defined as a combination of physicochemical and biological indicators) over time, and vegetation floristic composition, diversity, and richness. Three zones were stablished to study the effects of check-dams in ephemeral channels: the wedge zone (immediately upstream the dam; W), the downstream zone (immediately downstream the dam, DS) and control zone (out of the influence area of the check-dam; C).
In the study IV, soils from eight check dams located in two semi-arid adjacent areas at SE Spain (affected by wildfire in 2012 and 2016), were sampled. 48 composite soil samples were collected along two transects in the W, DS and C zones during July 2017, five years after fire in the first area, and 1 year after fire in the second. Physicochemical (OM, EC, texture, bulk density, pH and infiltration) and biological (enzymatic activity, soil microbial biomass carbon and glomalin content) parameters were measured.
This soil data was analyzed with a two-way ANOVA, where the number of years after fire and the survey zones were settled as fixed factors and a post-hoc Tukey HSD test was applied. Additionally, a Pearson’s correlation analysis and a principal component analysis (PCA) were performed. The study findings reveal that by altering sediments cycle and transport, check dams influence soil's main physicochemical and biochemical characteristics. Significant differences were found between check dam-affected zones and the control ones for many indicators such as organic matter content, electrical conductivity, or enzymatic activity. According to the ANOVA results, interaction between check dams influence and time after fire, was a crucial factor.
The study V was focused on how ephemeral streams' vegetation, is affected locally by check-dam construction in the years following a wildfire. Vegetation and soil samplings were carried out in 17 check dams throughout a semiarid area in SE Spain, which was affected by a wildfire in 2012. Soil and vegetation surveys were carried out in W, DS and C zones for each check-dam. Transects of at least 50 m2 were placed in each zone and using a grid by number method the vegetation was surveyed, placing ten additional 1m2 at each transect to evaluate coverture. To characterize soils, at each transect several physicochemical parameters were measured in each transect. Similarity, diversity, and richness analysis were conducted to characterize plant communities’ composition in each zone. Additionally, coverture plots’ data was classified in life forms to evaluate their distribution in terms of coverture. To analyze the results, ANOSIM and SIMPER analysis was used to evaluate the between- and within-zone differences and similarities. In Addition, a non-metric multi-dimensional scaling (MDS) analysis was performed to confirm the similarities within the previously established zones. Diversity, richness, and coverture data were analyzed using a one-way ANOVA and a Tukey HSD post-hoc test, setting the surveyed zone (W, DS and C) as a fixed factor. Finally, A canonical correspondence analysis (CCA) was used to analyze the influence of the physicochemical soil parameters on the different floristic compositions.
The results of the study revealed that check-dams had a significant influence given the accretion of fine sediments in depositional wedges. This accumulation of fine sediments, organic matter and nutrients resulted in the formation of a suitable environment for ruderal plant development by creating a differentiated community upstream of dams. Check-dam construction also affected species diversity, with slightly higher values for the Shannon and Simpson indices at those transects directly influenced by the structure.
In general, the results of the present PhD dissertation, show that the AIC index can be a suitable and useful tool for post-fire management. The versatility of this index allows it not only to assess fire affection and its influence on SC (to identify priority restoration areas), but also to reflect the effect of post-fire management measures on SC. Using this tool, a better zoning of expensive channel measures like check-dams, which affect to ephemeral channels’ soil and vegetation characteristics can be achieved.
